Brain derived neurotrophic factor.

DESCRIPTIVE MEMORY

1. INTRODUCTION The present invention relates to nucleic acid sequences encoding brain-derived neurotrophic factor, substantially pure proteins, peptide fragments or derivatives thereof, and antibodies directed against protein, peptide fragments or derivatives thereof. of the BDNF. In addition, the invention relates to genes that are members of a newly defined BDNF / NGF gene family and to the products of their genes. 0 The invention also describes the preparation of pharmaceutical compositions comprising effective amounts of BDNF gene products or, alternatively, antibodies directed against BDNF gene products, as well as methods for diagnosing and treating various neurological diseases and disorders, including Alzheimer's disease and Parkinson's disease. In particular, the BDNF gene products of the invention have value in the diagnosis and therapy of dopaminergic neuron disorders, such as Parkinson's disease, as well as sensory neuron disorders and degenerative retinal diseases.

2. BACKGROUND OF THE INVENTION

2.1. NORTH NEURONAL CELLS AND THE ROLE OF FACTORS

NEURQTROPHICS IN NERVOUS SYSTEM DEVELOPMENT

In many parts of the vertebrate nervous system many more neurons are present in early development than in the adult animal. Early developmental periods are characterized by natural death cell waves of neuronal cells (Carr and Simpson, 1978, J. Comp. Neurol., 182 = 727-740; Cowan et al., 1984, Science 225 = 1258-1265) . Survival, differentiation and maturation of developing neurons can be regulated by environmental or epigenetic factors rather than by a rigorous intrinsic genetic program. For example, experimental manipulations of the chick embryo have shown that transplantation or extirpation of peripheral target fields, such as the limb bud or eye in the early stages of chick development, may result in an increase or decrease.

493

<img file="PT95152B_D0001.tif" />

6526-019-118 corresponding, respectively, to the number of sensitive, sympathetic, parasympathetic or motor neurons adjacent to the enlarged or reduced target field (Hamburger, 1934, J. Exp. Zool. 68 = 449; Hollyday and Hamburger, 1976 J. Comp. Neurol., 170: 311-521; Landmesser and Pilar, 1976, J. Cell. Biol., 68 = 557-374). A target field can only support a limited number of neurons, and a normal phase of the developmental process may be the suppression of too many neurons to simulate the neurotrophic capacity of the target tissue. The discovery and isolation of the protein now called nerve growth factor (N6F = Nerve Grouth Factor) led to a molecular hypothesis of how a target might be able to regulate the number of neurons that survive and innervate this tissue (Levi-Montalcini et al. (1968, Physiol Rev., 48 = 524-569; Thoenen and Barde, 1980, Physiol Rev., 60 = 1284-1335).

It is now well established that, at least in the peripheral nervous system, such neuronal target tissues synthesize and release limited amounts of various neurotrophic molecules that are critical for the survival of specific types of neurons (Korsching and Thoenen, 1983, Proc. Natl. Acad USA 80 = 3513-3516; Heumann et al., 1984, EMBO J. 3 = 3183-3189; Shelton and Reichardt, 1984, Proc. Natl. Acad, Sci USA, 81 = 7051-7955; Korsching and Thoenen , 1985, Neurosci. Lett, 54 = 201-205).

2.2. NERVO GROWTH FACTOR Nerve Growth Factor (NGF) is by far the most fully characterized of these neurotrophic molecules and has been shown, both in vitro and in vivo, to be essential for the survival of sympathetic and sensitive neurons derived from neural Christian during the first stage of development of both chick and rat (Levi-Montalcini and Angeletti, 1963, Develop. Biol. 7 = 653-659; Levi-Montalcini et al., 1968, Physiol, Rev, 48 = 524-569), Injections of purified NGF into the developing chick embryo have been found to cause massive hyperplasia and hypertrophy of spinal sensory and sympathetic neurons (Levi-Montalcini et al., 1968, Physiol, Rev, 48 = 524-569). Montalcini and Booker, 1960, Proc Natl, Acad, Sci, USA, 46 = 575-584; Hamburger et al.

493

6526-019-118

-41981, J. Neurosci. 1 = 60-71). Conversely, removal or sequestration of endogenous NGF by daily injection of anti-NGF antibodies in neonatal rats has been associated with virtual destruction of the sympathetic nervous system (Levi-Montalcini and Booker, 1960, Proc. Natl. Acad. Sci. USA 46 = 384-391; Levi-Montalcini and Agneletti, 1966, Pharmacol. Rev. 18 = 619-628). Exposure to NGF antibodies, even earlier in development by injections of in utero antibodies or passive transplacental transfer of maternal antibodies, has resulted in a substantial loss of neural Christian-derived sensory neurons, such as spinal sensory neurons and trigeminal middle dorsum. (Goedert et al., 1984, Proc. Natl. Acad. Sci. USA 81 = 1580-1584; Gorin and Johnson, 1979, Proc. Natl. Acad. Sci. USA 76 = 5382-5386). Until recently, almost all studies of NGF have focused on its role in the peripheral nervous system, but it now appears that NGF also influences the development and maintenance of specific central nervous system neuron populations (Thoenen et al., 1987, Rev Physiol Biochem Pharmacol 109 = 145-178 (Whittemore and Seiger 1987 Brain Res Rev 12 = 439-464).

The accidental discovery of large amounts of NGF protein in the submandibular gland of the adult male mouse (Cohen, 1960, Proc. Natl. Acad. Sci. USA 46 = 302-511) and an earlier finding of NGF levels in venom. (Cohen and Levi-Montalcini, 1956, Proc. Natl. Acad. Sci. USA, 42 = 571-574), randomly provided sufficient amounts of NGF to permit studies of the physiology, protein chemistry and, more recently, molecular biology of NGF (ie, molecular cloning of NGF and its receptor). The function of large amounts of NGF in the salivary gland of the adult male mouse remains unknown; However, it seems that this abundant source of NGF may play no role in the development or maintenance of the central or peripheral nervous system. In target tissues innervated by NGF-sensitive neurons (neurons that have been shown to depend on NGF for survival, which have high affinity to NGF receptors and which, with high

<img file="PT95152B_D0002.tif" />

Specificity, internalize and transport NGF backwards) it was found that in the stabilized state measurable levels of NGF were extremely low, on the order of picograms or nanograms per gram of tissue, when compared with the 1,000-fold higher levels in NGF tissue. salivary gland of adult male mouse. NGF has not been found at any appreciable level in serum and therefore does not appear to be a circulating hormone or growth factor (Suda et al., 1978, Proc. Natl. Acad. Know. USA 75: 4042-4046).

In addition to the important discovery of a large source of NGF protein in the mouse submandibular gland, the progress of sensitive, reliable and efficient biological assays has played an important role in elucidating the biology and biochemistry of NGF. The dorsal root ganglia (DRG) of the developing chick embryo is one of the first neuronal types to be sensitive in vitro to NGF. Explants of E8-E12 chick DRG cultures in a plasma clot and, more recently, dissociated neuron-enriched chick DRG cultures have proven to be very useful in bioassays of NGF activity (eg during purification) and in studies in vitro biology of NGF (Levi-Montalcini et al., 1954, Cancer Res. 4: 49-57; Levi-Montalcini and Angeletti, 1963, Develop. Biol. 1:653-659; Greene, 1977, Develop. Biol 58:96, ibid 58: 106). 0 The fact that more than forty DRGs can be dissected from a single chick embryo has led to widespread use of NGF bioassays in many laboratories.

In addition to the availability of large amounts of NGF protein and efficient NGF screening systems, a third important contributor to knowledge of NGF biology is the relative fecundity with which NGF antibodies can be obtained in guinea pigs, rabbits, sheep, etc .; it appears that mouse NGF is highly immunogenic. Cohen (1960, Proc. Natl. Acad. Sci. USA 46: 502-511) raised antibodies to the NGF that had been purified from the mouse submandibular gland and showed, with Levi-Montalcini and Booker (1960, Proc. Natl. Acad. Sci. USA 46: 584-591) that these antibodies destroyed the sympathetic ganglia

493

6526-019-118

<img file="PT95152B_D0003.tif" />

-6or caused immunosympathotomy when administered daily to newborn rats (Levi-Montalcini and Angeletti, 1966, Pharmacol. Rev. 18: 619-628).

The abundance of protein from the NGF allowed the main sequence to be determined by relatively conventional protein chemistry (Angeletti and Bradshaw, 1971, Proc. Natl. Acad. Sci. 68: 2417-2420). The NGF gene has now been cloned into several species including the mouse (Scott et al., 1983, Nature 302: 558-540). the man (Ullrich et al., 1983, Nature 505: 821-825) the cow and the chick (Meier et al., 1986, EMBO J. 5 = 1489-1493) and the rat (Whittemore et al., 1988, J. Neurosci. Res., 20 = 402-410) using essentially conventional molecular biology based on the availability of the mouse NGF protein sequence to prepare suitable oligonucleotide probes. The availability of abundant NGF has also greatly facilitated studies of the NGF receptor, which ultimately led to the molecular cloning of the human and rat NGF receptor (Johnson et al., 1986, Cell, 47: 545-554; Radeke et al. al., 1987, Nature 525 = 595-597).

It is now well established that NGF is not a ubiquitous neurotrophic factor. Within the peripheral nervous system, NGF does not appear to be a survival factor for parasympathetic neurons, sensory neurons derived from neural placodes, or enteric neurons, as determined in in vitro and in vivo studies. Moreover, NGF does not appear to be a survival factor for developing motor neurons (Oppenheim, 1982, J. Comp. Neurol. 210 = 174-189), although these neurons appear to express at least a reduced form of NGF receptor affinity during development (Raivich et al., 1985, EMBO J. 4 = 637-644). The lack of effect of NGF on these neuronal types stimulated the search for other neurotrophic factors, especially factors that would support the survival of spinal cord motor neurons and / or parasympathetic ciliary ganglion neurons.

2.3. OTHER NEURQROPHIC FACTORS

In the past decade there have been numerous reports on the

493 Ó52Ó-019-118

<img file="PT95152B_D0004.tif" />

neurotrophic activity in extracts from a wide variety of tissues and in conditioned culture media of very different cell types. In almost all cases, however, progress in purification and characterization of these activities has stumbled on the fact that these activities are present in extremely small amounts, from the order of picograms to nanograms per gram of tissue.

In addition, although appropriate biological assays for peripheral neurons have been established, the proposal for reliable, reproducible, and specific tests for the central nervous system has been shown to be problematic. While individual types of peripheral neurons appear as discrete ganglia, dissectable facilments, central nervous system (CNS) neurons are invariably highly heterogeneous in their distribution. Specific markers are therefore required both for identification and improvement of certain classes of CNS neurons. Progress in the production of these markers has been very limited, e.g. cell surface antibodies or cytoskeletal components or specific histological strains. Thus, characterization of neurotrophic factors that are (i) not as abundant as NGF, (ii) difficult to assess, and (iii) not available in sufficient quantities to elicit antibody production, proved to be an extremely difficult process.

2.3.1. COMPARISON OF THE BRAIN DERIVED GROWTH FACTOR WITH 0 NERVE GROWTH FACTOR

Neurotrophic activity capable of maintaining the survival of chick embryo dorsal root ganglion neurons in vitro has been identified in the conditioned medium where rat C-6 glioma cells had been cultured (Barde et al., 1978, Nature 274 -818). Activity was not neutralized by rat NGF antibodies, suggesting the presence of another neurotrophic factor in the conditioned medium. Similar activities that cannot be blocked by NGF antibodies were then revealed in mouse brain astroglial cell cultures.

<img file="PT95152B_D0005.tif" />

493

6526-019-118

-8 normal adult (Lindsay, 1979, Nature, 282 = 80-82; Lindsay et al., 1982, Srain Res. 243 = 529-345) and extracts of the adult rat and growing rat (Barde et al. , 1980, Proc. Natl. Acad. Sci. USA, 77 = 1199-1203) and mature or developing chick spinal cord (Lindsay and Peters, 1984, Neurosci., 12 = 45-51). However, in no case was any active factor identified or identified and it remains questionable to what extent the observed activities were due to the same or different factors.

lymph nodes have been found

Using the pig brain as starting material. Barde et al. (1982, EMBO J. 1 = 549-553) mentioned a factor, now called brain-derived neurotrophic factor (BDNF) that appears to improve survival of dorsal root neurons in E10 / E11 chick embryos. that neurotrophic activity resided in a highly basic protein (isoelectric point, pi&gt; 10.1) that migrated during sodium dodecyl sulfate (SDS) gel electrophoresis as a single 12.3 kD molecular weight band. 0 Purification factor was estimated at 1.4 x 10 6 but the yield was very low, with only approximately 1 pg of BDNF purified from 1.5 kg of pig brain. In addition, as the final step of the purification process was preparative gel electrophoresis, BDNF activity could not be completely restored due to the presence of residual SDS (Barde and Thoenen, 1985, in Hormones and Cell Regulation, Vol. 9, Dumont et al., Eds Elsevier Science Publishers, pp. 385-390). It was noted that the highly basic nature and molecular size of BDNF were very similar to those of the NGF monomer. 0 BDNF however appeared to have properties that differed from the known properties of NGF since (a) in the evaluation of dorsal root ganglia of the chick, NGF antibodies had no apparent effect on the biological activity of BDNF; (b) the effects of BDNF and NGF appeared to be additive; and (c) unlike NGF, BDNF was found to have no effect on the survival of E12 chick sympathetic neurons. In addition, during the first studies with extracts of

493

6526-019-118

In the brain, neurotrophic activity in these sources was observed to appear to act on sensory neurons at later stages of development than when associated with NGF. Using separate cultures of chick embryo neurons grown on a polycationic substrate such as polylysine or poliornitine, BDNF was found to support the survival of more than 30% of E10-11 chick embryo dorsal root ganglion neurons (embryonic day). ten or eleven) but appeared to have little effect on the survival of the same neurons in E6 (Barde et al., 1980. Proc, Natl. Acad, USA 77: 1199-1203 supra). Under similar conditions, NGF supported the survival of 30-40¾ dorsal root ganglion (DRG) neurons in E6. It is interesting to note that it was later found that when grown on a substrate coated with the extracellular glycoprotein laminin matrix, both NGF such as BDNF supported the survival of about 50% of DRG neurons from E6-E12 chick embryos (Lindsay et al., 1985, Develop. Biol. 112: 519-528). Subsequent studies found that the effects of NGF and BDNF were additive when both were present at saturation concentrations.

Previous studies by Levi-Montalcini (1966, The Harvey Lectures 60: 217-259) on the neuronal specificity of NGF suggested that NGF was not an ubiquitous neurotrophic factor even for sensory neurons, as NGF seems to have no effect on NGF. neurons of certain sensory ganglia of the chick skull, especially the knotted ganglion of the tenth cranial nerve. Further in vivo studies (Johnson et al., 1980, Science 210: 916-918; Pearson et al., 1983. Develop Biol. 96: 32-3) showed that exclusion of NGF during embryogenesis had no effect on neuron survival of most rat cranial sensory ganglia, while similar treatment lowered neuronal count in sensory ganglia derived from neural Christian More detailed in vitro studies (Lindsay and Rohrer, 1985, Develop. Biol. 112: 50-48; Davies and Lindsay, 1985, Develop. Biol, 111: 62-72; Lindsay et al., 1985, J. Cell. Know Suppl 3: 115-129) showed

493

6526-019-118

- Clearly, NGF supported the survival of most sensory neurons derived from the neural Christian but had no apparent effect on the survival of cranial sensory neurons derived from neural placodes,

The first demonstration of a distinct BDNF neuronal specificity from that of NGF was the in vivo demonstration that purified BDNF supports the survival of 40-50¾ of the separate sensory neurons derived from the E6, E9, or chick embryo-derived nodular ganglion. E12 (Lindsay et al., 1985, J, Cell Sci-Supp. 3: 115-129). NGF had no apparent effect on these neurons either by itself or in conjunction with BDNF. It was later shown in explant culture studies that BDNF appeared to support the survival and development of neuritis from other sensory ganglia derived from neural placodes, including the geniculate, ventrolateral trigeminal (petrous) ganglia (Davies et al. , 1986, J. Neurosci. 6: 1897-1904), none of which had been sensitive to NGF. In all of the above studies, NGF antibodies had no effect on observed BDNF activity. In addition to its effect on cultured neurons from peripheral lymph nodes, BDNF has been found to stimulate survival and neuronal differentiation of cultured quail Christian neural cells (Kalcheim and Gendreau, 1988, Develop. Brain. Res. 41:79 -86).

Prior to the present invention, the inability to produce sufficient amounts of BDNF for immunization prevented the production of anti-BDNF antibodies for comparison with anti-NGF antibodies for their effects on neuronal populations and precluded BDMF / MGF cross-neutralization experiments. Two studies BDNF (Kalcheim et al., 1987, EMSO J. 6: 2871-2873; Hofer and Barde, 1988, Nature 351: 261-262) have, however, indicated a physiological role of BDNF in the development of SNP in birds. If an in ovo mechanical barrier is placed between the DRG sm E3 / E4 (dorsal root ganglia of 3 or 4 days) and its CNS target in the neural tube, it is observed that many DRG neurons die (Kalcheim and Le Douarin, 1986, Develop, Biol. 116: 451-466). It has been postulated that this neuronal death could be due to

493

6526-019-118

Loss of a CNS-derived neurotrophic factor (neural tube). It was then observed that BDNF, bound to a laminin-coated sialastic membrane, could prevent this cell death (Kalcheim et al., 1987, EMBO J. 6 = 2871-2873). Injections of BDNF in developing quail eggs have been found to reduce naturally occurring cell death in the knotted ganglia, an effect not yet observed with NGF (Hofer and Barde, 1988, Nature, 331 = 261-262). In addition to its effect on peripheral sensory neurons of both Christian neural and neural placode origin, BDNF has been shown to support the survival of developing CNS neurons. Johnson et al. (1986, J. Neurosci., 6 = 3031-3938) reported data indicating that BDNF supports survival of cultured retinlan ganglion cells from E17 rat embryos. This confirmed previous studies showing that conditioned media and brain extracts prepared from retinal ganglion cells in certain regions seemed to support the survival of these neurons (McCaffery et al., 1982, Ex. Brain Res. 48 = 37 -386; Sarthy et al., 1983, J. Neurosci. 3 = 2532-2544; Turner et al., 1983, Dev. Brain Res. 6 = 77-83).

In addition to the effects on survival of cultured developing neurons, BDNF has been shown to have effects on cultured neurons in the adult peripheral and central nervous system. BDNF as well as NGF have been found to stimulate axonal regeneration of cultured rat adult neuronal neurons (Lindsay, 1988, J. Neurosci. 8 = 2394-2405) even though adult sensory neurons do not appear to require neurotrophic factors for in vitro maintenance for 3 or 4 weeks. In addition, in adult rat retinal cultures, BDNF was found to improve both survival and axonal elongation of retinal ganglion cells (Thanos et al., 1989, Eur. J. Neurosci. 19-26). Table I shows the comparison of the biological effects of NGF and BDNF.

493

6526-019-118

TABLE I

COMPARISON OF BDNF AND NGF BIOLOGICAL ACTIVITIES *

SURVIVAL**

<td>SYSTEM</td><td>Nervous Peripheral</td><td></td><td>BDNF</td><td>NGF</td>

<td>(i)</td><td>E6 chick DRG</td><td></td><td> -</td><td> ++</td>

<td></td><td>E10 chick DRG</td><td></td><td> -</td><td> ++</td>

<td></td><td>Friendly neurons of the</td><td>chick E12</td><td> -</td><td> ++</td>

<td>(ii)</td><td>E6-E12 chick DRG</td><td></td><td> ++</td><td> ++</td>

<td></td><td>Chick Nose E6-E12</td><td></td><td> ++</td><td> -</td>

<td></td><td>Friendly neurons of the</td><td>chick E12</td><td> -</td><td> ++</td>

<td></td><td>E12 chick ciliary</td><td></td><td> -</td><td> -</td>

<td></td><td>(Lindsay et al., 1985,</td><td>above)</td><td></td><td></td>

<td>(iii)</td><td>Chick E3-E14:</td><td></td><td></td><td></td>

<td></td><td>Jugular</td><td></td><td> +/++</td><td> 4-+</td>

<td></td><td>DM-Trigeminal</td><td></td><td> +/++</td><td> ++</td>

<td></td><td>Rock</td><td></td><td> +/++</td><td> -</td>

<td></td><td>Geniculate</td><td></td><td> +/++</td><td> -</td>

<td></td><td>VL-Trigeminal</td><td></td><td> ++</td><td> -</td>

<td></td><td>Entrance exam</td><td></td><td> -</td><td> -</td>

<td></td><td>Mesencephalic</td><td></td><td> ++</td><td> -</td>

<td></td><td>(Davies et al., 1986</td><td>. above)</td><td></td><td></td>

<td></td><td>(Barde et al., 1987,</td><td>Prog.</td><td></td><td></td>

Brain Res., 71 = 185-189)

CENTRAL NERVOUS SYSTEM (i) E17 ++ rat retinal ganglion cells (Johnson et al., 1986,

J. Neurosci. 63031-3038) in chronological order of publication dates; effects on in vitro tests without survival- (-); moderate survival (+); good survival (++)

<img file="PT95152B_D0006.tif" />

2.3.2. NEURAL TARGETS DQ NEUROTROFICQ FACTOR

493

6526-019-118

-13 DERIVATIVE OF BRAIN

Sensory neurons of peripheral nerve ganglia have been found to arise from either of two distinct transient embryological structures, namely the Christian neural and the neural placods. The neural Christian seems to give rise to both the neurons and the satellite cells of autonomous ganglia and sensory spinal nerve ganglia, ie DRG. The contribution of the neural Christian and the neural placodes to the cranial nerve sensory ganglia has been studied using the quail / chick chick transplantation system proposed by Le Douarin (Le Douarin, 1973, Develop, Biol. 20 = 217- 222; Noden, 1978, Develop. Biol. 67: 513-329; Narayanan and Narayanan, 1980, Anat. Rec. 196: 71-82; Ayer-Le Lievre and Le Douarin, 1982, Develop. Biol. 94: 291- 310; D⁵Amico-Maratel and Noden, 1983, Am. J. Anat., 166: 445-468). According to the review by Lindsay et al. (1985, J. Cell. Sci. Supp. 3: 115-129) it is now believed that, at least for birds, the neurons of the distal ganglia of the 72, 92 and 102 nerves. cranial ganglia (geniculate, rock and nodular ganglia, respectively) and neurons of the cranial nerve vestibular-acoustic complex are exclusively of neural placode origin. 0 The cranial nerve trigeminal ganglion contains neurons of both Christian and placode origin (predominating, in the ventrolateral pole of the maxillomandibular lobe, placode-derived neurons) while satellite cells of all cranial ganglia have been shown to be Christian in origin. neural.

In in vitro experiments using separate, neuron-enriched explant cultures of sensory neurons from cranial and spinal nerves, it was found that sensory neurons of Christian neural origin respond to NGF; In contrast, neurons derived from neural placodes (including neurons from the ventrolateral portion of the trigeminal ganglion and the entire neuronal population of the vestibular, geniculate, cliff and nodular ganglia) have been shown to be largely insensitive to NGF throughout embryonic development. Yours

<img file="PT95152B_D0007.tif" />

<img file="PT95152B_D0008.tif" />

493

6526-019-118

Differences in need and response to NGF, both neural Christian-derived and placode-derived sensory neurons (Table I) have been shown to be sensitive to BDNF neurite survival and activity (Lindsay et al., 1985, J. Cell. Sci. Supp. 3 = 115-129; Lindsay et al., 1985, Develop Biol. 112 = 519-528 (Kalcheim and Gendreau, 1988, Develop, Brain Res. 41 = 79-86). Tebar and Barde (1988, J. Neurosci. 8 = 3337-3342) studied the binding parameters of radiolabelled BDNF to dorsal root ganglion neurons in the chick embryo; Their results are consistent with the existence of 2 BDNF receptor classes, one with high affinity for BDNF, the other with low affinity. No high affinity receptors were observed in sympathetic neurons.

The known neuronal targets of BDNF were further reviewed by Barde et al. (1987, Prog, Brain Res. 71 = 185-189). Prior to the present invention, it was not possible to identify cells that synthesize BDNF due to the lack of BDNF-specific nucleic acid or antibody probes. There have been unsuccessful attempts to prepare both polyclonal and monoclonal antibodies to BDNF. This failure to produce antibodies has prevented the molecular cloning of BDNF, the determination of the physiological effect of depriving developing BDNF neurons in vivo, the quantification of BDNF in tissues using immunoassays, and the localization of BDNF using immunocytochemistry.

<img file="PT95152B_D0009.tif" />

-150UADRO II

ESDNF SENSITIVE AND OUR SENSITIVE NEURONS ~

A. Sensitive Neurons

I. Sensory neurons of the chick of origin in the neural Christian:

(a) in the dorsal root ganglion (b) in the jugular ganglion (c) in the dorsomedial trigeminal ganglion (d) in the mesencephalic trigeminal nucleus ^ *

II. Chick sensory neurons of origin in the ectodermal placode:

(a) nodular ganglion (b) vestibular ganglion (c) rock ganglion (d) geniculate ganglion (e) ventrolateral trigeminal ganglion

III. Rat retinal ganglion cells

IV, Chick retinal ganglion cells

B. Non-Sensitive Neurons

I. Sympathetic chick and mouse neurons

II. Sympathetic ciliary chick neurons

De Barde et al., 1987, Prog. Brain Res. 71: 185-189 v_er Davies et al., 1986, Nature 519: 497-499

3 SUMMARY OF THE INVENTION The present invention describes the preparation of nucleic acid sequences encoding brain-derived neurotrophic factor (BDNF), its protein, substantially pure and in quantity, peptide or derivative fragments and antibodies directed against protein, peptide fragments or derivatives of BDNF. 0 The present invention provides for the first time sufficient amounts of BDNF to prepare anti-8DNF antibodies and permits diagnostic and therapeutic applications of BDNF.

<img file="PT95152B_D0010.tif" />

<img file="PT95152B_D0011.tif" />

493

6526-019-118 —16—

In various embodiments of the invention the BDNF nucleic acids, proteins, peptides, derivatives or antibodies of the invention may be used in methods of diagnosis and treatment of various neurological disorders and disorders, and in particular in the diagnosis and treatment of disorders. sensory neurons and retinal degeneration. In addition, in specific embodiments of the invention, BDNF nucleic acids or BDNF gene products may be used in the diagnosis and treatment of neuroblastoma tumor, Parkinson's disease and Alzheimer's disease. BDNF gene products may also be used to facilitate incorporation of implants into nerve tissues or, alternatively, to promote nerve regeneration following trauma, stroke, infection or post-surgery damage in other specific embodiments of the invention.

The invention also describes the preparation of pharmaceutical compositions comprising effective amounts of BDNF gene products or, alternatively, antibodies directed against BDNF gene products, which may be used in the diagnosis or treatment of various neurological diseases or disorders.

Furthermore, by providing the complete nucleotide sequence of BDNF, the present invention allows the comparison of BDNF and NGF genes, thereby identifying homologous regions and defining a BDNF / NGF gene family. Thus, the present invention describes a process for identifying other members of the

<td colspan="2">BDNF / NGF gene family. In a specific arrangement, the process</td><td>of</td>

<td>invention</td><td>is used to identify a new non-NGF member,</td><td>no</td>

<td>BDNF,</td><td colspan="2">of the BDNF / NGF gene family. The invention further provides</td>

<td>new</td><td>members of the identified BDNF / NGF gene family</td><td>in</td>

<td>wake up</td><td>with the process described and with their gene products 3.1. ABBREVIATIONS AND DEFINITIONS</td><td></td>

<td>BDNF</td><td>brain derived neurotrophic factor</td><td></td>

<td>hBDNF</td><td>BDNF of man</td><td></td>

<td>CAT</td><td>choline acetyltransferase</td><td></td>

<td>CNS</td><td>central nervous system</td><td></td>

<td>DRG</td><td>dorsal root ganglia (ganglion)</td><td></td>

493

6526-019-118

<img file="PT95152B_D0012.tif" />

-17EDTA

NGF

PBS

PCR

SNP

SDS

Tris Ethylenediamine tetraacetic acid Nerve growth factor (Phosfate Buffered Saline) Phosphate buffered saline polymerase chain reaction peripheral nervous system sodium dodecyl sulfate tris (hydroxymethyl) aminomethane

4 DESCRIPTION OF THE FIGURES

Figure 1. Nucleotide sequence and deduced amino acid sequence of porcine prepro-BDNF cDNA. The complete sequence of two juxtaposed cDNA clones is shown in this figure. The sequences of peptides obtained by microsequencing (Table III) are underlined. The only consensus sequence for N-glycosylation is double underlined. The beginning of the sequence of mature BDNF is double-carat marked.

Figure 2. Sequence comparison between NGF and BDNF. Areas where more than two amino acids are in identical positions are indicated. Sequences begin with the first amino acids of mature proteins and end with the last before the stop codons, 51 amino acids are common to BDNF and the various NGFs (Schwarz et al., 1989, J. Neurochem. 52 = 1203-1209), including the six cysteine ​​residues.

Figure 3. Autoradiography of Southern EcoRI cut Southern blots of genomic DNA from man, monkey, rat, mouse, dog, cow, rabbit, chicken and yeast hybridized to 32p-labeled NGF and BDNF probes_no

Figure 4. Northern blot analysis. From each tissue, 20 pg total RNA was applied per lane, and hybridized with a 32 p-labeled cRNA probe. Mouse BDNF, Note a strong signal at about 1.45 kB with brain tissue, but none with the other 7 tissues analyzed.

Figure 5, Human BDNF cDNA sequence and deduced amino acid sequence, and comparison of rat, pig and chicken DNA sequences.

<img file="PT95152B_D0013.tif" />

493

6526-019-118

-18—

Figure 6. BDNF expression PCMV1-pBDNF plasmid.

Figure 7. (a) Results of ELISA determination of B5 peptide antisera binding, using serial dilutions of antisera.

(b) Quantification of binding of various antisera, diluted 1 = 500, with 50 ng BDNF.

Figure 8. Immunohistochemical staining results of tyrosine hydroxylase in BDNF-treated ventral midbrain cultures (dashed bars) and control (solid bars).

Figure 9. Dopamine uptake by ventral midbrain cultures. BDNF supplemented cultures are represented by dashed bars; control cultures are represented by solid bars.

Figure 10 (a) The effect of BDNF on the number of CAT positive cells in forebrain cholinergic neuron cultures.

(b) cultures of forebrain cholinergic neurons, at densities of 260,000 cells per well (black bar) or 150,000 cells per well (dashed bar), were treated with 150 ng / ml NGF. The number of CAT immunopositive cells in non-NGF treated versus treated cells is compared.

Figure 11. Changes in intensity of CAT enzyme activity in minute catalysed substrate picomoles as a function of BDNF concentration in forebrain cholinergic neuron cultures.

Figure 12. Astroglial cell cultures, at about 60% confluence, were treated with (a) epidermal growth factor (EGF) or (b) BDNF for 42 h, then incubated with [α H] methylthymidine. The amount of incorporated was measured in relation to the concentration of EGF and BDNF.

Figure 13. Chicken, mouse and rat restricted EcoRI Southern blot hybridized to BDNF / NGF R1B / 2C probe. The positions of the NGF and BDNF genomic EcoRI fragments are

J

<img file="PT95152B_D0014.tif" />

493

6526-019-118

-19 indicated by N and 8, respectively.

Figure 14 Comparison of NGF and BDNF sequences with a novel PCR-identified member of the BDNF / NGF family of mouse DNA with Box 3 / Box 4 primers: of the new gene [referred to herein as M3 / 4], also known as Neurotropin-3 (NT-3) shows only the coding chain sequence and deduced amino acid sequence aligned with mature mouse NGF and mature pig BDNF , rat, mouse, or male Q dashes indicate a position where deletion of a codon in NGF is used to optimize alignment relative to BDNF and M3 / 4. □ italics indicate coincidence in amino acid sequences and / or conservative amino acid substitutions.

Figure 15. Northern blots of RNA from various human tumor derived cell lines hybridized with human BDNF probe.

Figure 16. Depolarization effect of BDNF mRNA levels in hippocampal neurons.

(a) The course of BDNF mRNA expression in primary cultures of hippocampal neurons in the presence of 50 mM KCl. Total cellular RNA was extracted from 0.5x10 4 cells, glyoxylated and analyzed by 1.3¾ agarose gel electrophoresis (Biziere,

K. and Coyle, T., Neurosci., 1978, Lett. 8: 303; McGeer et al., 1978, Brain. Res. 139: 581). Hybond N -transferred RNA was hybridized to a ^ 2p-labeled cRNA probe (specific activity 10 ^ cpm / µg), specific for mouse BDNF, and produced by in vitro transcription by flow. The upper two bands (4 and 1.5 kb) correspond to BDNF mRNA; the two bands of BDNF (4 kb and 1.5 kb) are RNase A resistant and represent two different transcripts that appear to be similarly regulated though) and the lower one (700 bp) corresponds to 10 pg of a short pattern. of BDNF mRNA recovery that was added to the samples prior to RNA extraction.

(b) Calcium dependence. Neurons were incubated for 3 h in normal (CM) culture medium, modified calcium-free (-Ca) medium, or in 10 µM nifedipine culture medium

493

6526-019-118

-20 (TIN). Where indicated (+ K) 50 mM KCl was added.

Figure 17. Increases in NGF mRNA levels in hippocampal neurons by potassium and kainic acid. Neurons were incubated for 3 h in control medium (C) or in the presence of 50 mM KCl (K +) or 25 µM kainic acid (KA). RNA was extracted and the NGF mRNA levels determined by quantitative PCR (11). Values ​​are mean ± SEM (n = 6).

Figure 18. Dose response curve on the effect of kainic acid. Hippocampal neurons were incubated for 3 h in the presence of various concentrations of kainic acid. RNA was extracted and analyzed as described in Fig. 16. Values ​​represent means ± SEM of three experiments.

Figure 19. Course of increases in BDNF and NGF mRNA levels due to kainic acid treatments. Total cellular RNA was extracted, and analyzed as shown in Fig. 16 from the hippocampus (a) or cortex (b) at the indicated times (in hours) after intraperitoneal injection of kainic acid (12 mg / kg) 'A single dose of diazepam (10 mg / kg) was injected. 90 min after kainic acid (the two bands of BDNF (4 kb and 1.5 kb) are RNase A resistant and represent two different transcripts which appear to be similarly regulated however). The values ​​corresponding to 1.5 kb BDNF (o) mRNA and 1.3 kb NGF (o) mRNA are indicated. Similar increases were observed for the 4 kb BDNF mRNA. The values ​​given represent means ± SEM of 3 to 4 experiments.

Figure 20. Effect of anticonvulsants on kainic acid-induced BDNF mRNA expression. Rats were injected intraperitoneally with diazepam (10 mg / kg) (DZ); MK-801 (1 mg / kg) (MK) or ketamine (20 mg / kg) (KET) 15 min before injection of cainic acid (12 mg / kg) (+ KA) or saline (controls). Three hours later hippocampal tissue was used to prepare total RNA as shown in Fig. 16. Values ​​represent means ± SEM of three experiments.

Figure 21, Septal Cell Cultures.

AF. Phase contrast micrograph of the course of the

J

493

6526-019-118

<img file="PT95152B_D0015.tif" />

growing cells in culture. Cells were plated at a density of 1.3x10 4 cells / cm²_and kept at 5 HS / NS as described in the text. Registration times were 1 (A, B), 2 (C, D) and 4 (E, F) days. Cell type-specific markers were used to identify cell populations, AChE histochemically labeled neurons (G, K) and NGF receptor immunopositive neurons (I). Scale = 25 µm.

Figure 22. Comparison of Dissociated Septal Cell Response to BDNF or NGF Treatment, Based on AChE Number of Cells.

(THE). Comparison of response to BDNF (25 ng / ml), NGF (50 ng / ml) or the combination of the two ligands at different plating densities. (B) Comparison of cholinergic neuron response to a concentration zone (0-50 ng / ml) of BDNF or NGF. For these results the cells were grown over a period of 11-12 days in serum containing medium. Points represent means ± SEM of 6-9 determinations.

Figure 23. The bar graph depicts the effect of delaying the addition of BDNF or NGF to dissociated septal cultures. BDNF (50 ng / ml) or NGF (50 ng / ml) were added to cells from dissociated cultures with delays of 5-6 h (+12), 5 days (-5 / + 7) or 7 days (-7 / + 5). Cultures were maintained for 12 days. Cells were scored based on AChE positive histochemical stain. Results are expressed as mean ± SEM of 4 to 5 determinations.

Figure 24. Bar graph showing the ability of BDNF or NGF to regulate the number of NGF receptor immunopositive cells.

NGF receptor immunoregulation was conducted using monoclonal antibody 192-IgG at a dilution of 1 = 1000. Dissociated septal cells were plated at a density of 1.3x10⁵ cells / cm² and treated continuously over a period of 12 days. A comparison is made between the saturating dose of NGF and a dose zone of BDNF (O-100ng / ml). The

493

6526-019-118

-22results are the mean ± SEM of n = 4.

Figure 25, Dose Response Curve of Septal Cholinergic Neurons to BDNF (A) and NGF (B) CAT-Inducing Activities „

The cultures were plated in 6 mm wells covered with polyornitine and laminin and maintained for 12 days at 5 HS / N3. Cells were exposed to BDNF or NGF 5 to 6 h after plating. Factors were replaced every 3 days when the medium was changed. The indicated results represent the mean ± SEM of an n of 5 to 6 determinations.

Figure 26. Dose Response Effect of BDNF and NGF on AChE Enzyme Activity.

Rat septal cholinergic neurons were developed for 12 days in serum-containing medium with hormonal supplements. Cells were treated with BDNF (light squares) and NGF (solid squares) as described in Figure 25. Results are the mean ± SEM of a n of 6 to 9. AChE activity in untreated cultures was 16.4 ± 2 nmol / h / well.

Figure 27, Course of Increased BDNF-Induced CAT Enzyme Activity compared to that of NGF.

The time required to stimulate CAT activity was determined on plaque cells at a density of 2.3x10 4 cells / cm 2 and grown over various time periods in medium containing serum with hormonal supplements and exogenous factors. The cells were initially exposed. BDNF (light squares) or NGF (solid squares) 5-6 hours after plating. Results represent the percentage of activity determined in treated versus untreated cultures (n = 6, 12 day BDNF n = 3).

Figure 28. Comparison of the Effect of Glial Cells on BDNF Ability to Induce CAT Enzyme Activity. Septal cells were plated at a density of 2.3 x 10⁵cells / cm 2. After a period of 5-6 h in the plate, the medium was changed to either serum free medium or medium containing

J

<img file="PT95152B_D0016.tif" />

493

6526-019-118

-23 serum, both supplemented with 1¾ N3 as indicated in the text. Arabinoside cytosine (1 μΜ) was added for 24 h to further reduce the number of astrocytes. The cultures were then maintained for 12 days. Results represent the mean ± SEM of 4 determinations.

Figure 29. 0 Bar Graph Shows the Effect of BDNF or NGF Treatment on High-Affinity Hill Taking Level,

Choline uptake was conducted on cells maintained for 11 days after plating at a density of 1.3x10 4 cells / cm 2. BDNF (50 ng / ml) or NGF (50 ng / ml) were present throughout the culture period. The points represent the mean ± SEM of 5 BDNF and 10 NGF determinations.

Figure 30, SDS PAGE Resolution of 35g-labeled reaction products_? which enzymatic cleavage of neurotrophic factor derived from the brain of man by trypsin and endoproteinase Arg-C

Figure 31. Graphical representation of DRG bioassay comparing untreated CHO-hBDNF with endoproteinase-cleaved CHO-hBDNF. Neuritis growth is scored from 0 to 5, with 5 representing the maximum bioactivity.

Figure 32. Phase (A) and brightfield (8, C) contrast photomicrographs of dissociated cultures of mouse E14 rat ventral mesencephalus cells stained after 9 days in culture with tyrosine hydroxylase (TH) monoclonal antibodies .

A and B. Phase and brightfield contrast photographs of the same field show the scarce abundance of TH + neurons in these cultures. Only a single TH + neuron (arrows) is seen in a field containing many bright phase neurons with long neurites. Some non-neuronal cells are evident by morphology or GFAP staining (not indicated).

C. Bright field photograph showing two TH + neurons (small arrows) in a field containing about 98 bright phase cells of neuronal morphology. Scale = 100 1M.

J

<img file="PT95152B_D0017.tif" />

493

6526-019-118

-24Figure 33. High-magnification photomicrographs of 6-day cultures of E14 rat ventral midbrain cells showing various morphologies of TH + neurons, some with extensive neuritic growth and extremely elaborate growth cones (A, B). Cultures were established and stained as described in the legend of Fig. 32. Scale = 25 µM

Figure 34. BDNF promotes the survival of dose-dependent, mesencephalic dopaminergic TH + neurons, evident in longer culture periods than 3 days.

A. Comparison of the number of TH + neurons in E14 rat ventral midbrain cultures, maintained with (solid bars) or without BDNF (light bars). In treated cultures, BDNF (50 ng / ml) was added once over day 2 culture. No differences were observed between control and day 3 treated cultures, but the number of TH + neurons in BDNF treated cultures was 1.8 times higher than in Day 8 controls. Values ​​are the average of duplicate samples.

B. The effect of BDNF on increasing survival of TH + neurons is dose dependent.

The number of TH + neurons was determined in duplicate cultures after 8 days in culture in the absence of BDNF or in medium containing increasing concentrations of BDNF added on day 2.

C. 0 NGF has no effect on neuron survival

TH +.

To assess the specificity of BDNF, cultures were also grown in the presence or absence of NGF (50 ng / ml) for 8 days and analyzed for TH + cells. NGF, added on day 2, did not increase the survival of TH + neurons in cultures examined on either day 6 or day 8. Results are the average of duplicate plaques.

Figure 35, Repeated doses of BDNF further enhance the survival of TH + neurons as compared to the single addition of BDNF on day 2.

Cultures were prepared as described in the text. Purify

<img file="PT95152B_D0018.tif" />

6526-019-118

Human recombinant BDNF was grown from COS M5 cells. Cultures were treated with a single addition (SA; dotted bars) of BDNF (50 ng / ml) on day 2, or with repeated doses of BDNF (50 ng / ml) on days 1, 2, 4, 6 and 8 (MA - multiple addition; solid bars) or developed without BDNF (control; light bars). At the indicated times, cultures were fixed and stained for determination of TH immunoreactivity. 0 BDNF replenishment by multiple additions resulted in a 2.7 X greater number of TH + neurons on day 11 as compared to the control, while the maximum increase observed with a single BDNF addition was only 2 X after 11 days. Typical results from several experiments are the average of duplicate samples.

Figure 36. Delayed addition of BDNF does not increase the number of TH + neurons to the same extent as when BDNF is added on day 2.

Cultures were prepared as described in the text, except for the addition time of exogenous BDNF. Every. cultures were switched to serum free medium on day 2 and BDNF (50 ng / ml, pig brain BDNF) was added at one time either on day 2, day 5 or day 7 (CD2, CD5, CD7). On culture days 6, 8 and 10, the number of TH + neurons was determined in duplicate cultures. In this experiment, a single addition of BDNF on day 2 gave a 3 X larger number of TH + neurons on day 10. When BDNF joined late on day 5, an increase in TH + neurons over controls but less than 2 X was observed on day 10, and this difference in number of TH + cells between treated and control cultures did not increase. more with longer culture time. Controls, light bars; addition of BDNF on day 2, dotted bars; BDNF added on day 5, solid bars; BDNF added on day 7, dashed diagonal bars.

Figure 37. Dose-dependent response of BDNF in high affinity GABA uptake in hippocampal cultures.

Partially purified nBDNF (rotophor purified from COS M5 supernatants) was added to

J /?

493

6526-019-118

<img file="PT95152B_D0019.tif" />

various dilutions. BDNF at Ix10 dilution was found^-2had maximal neurite promoting activity in the dorsal root ganglion of the E8 chick. High affinity uptake of H-6ABA was measured after 8 days of in vitro BDNF treatment.

Figure 38. BDNF protects cultured nigral dopaminergic neurons from the neurotoxic effects of MPP +. Cultures were established from E14 rat embryos as described in Figure 1. After 24 h in vitro the culture medium was changed to serum free formulation. After 3 days in culture, the cultures were divided into 4 groups of 6 35 mm plates. One group was maintained as a control group and the others received (i) basic fibroblast growth factor, 10 ng / ml (bovine brain bFGF);

(ii) mouse NGF 50 ng / ml or (iii) BDNF 50 ng / ml. Cultures were maintained for an additional 24 h before exposing 3 plates of each group to MPP + · 1 μ, Μ (Research Biochemicals, Inc, Natick, MA) for 48 h. At the end of the experiment, 6 days total, all plates were evaluated for TH immunoreactivity and the number of TH + cells in each group was determined. The results presented represent the number of TH + neurons after treatment with MPP +, calculated as a percentage of the number of TH + neurons in similar cultures not exposed to MPP +. All values ​​are means ± no of triplicate cultures,

5 DETAILED DESCRIPTION OF THE INVENTION The present invention describes the preparation of nucleic acid sequences encoding brain derived neurotrophic factor (BDNF) as well as protein, peptide fragments and BDNF derivatives prepared in bulk using these sequences. Further, the invention describes the preparation of pharmacological compositions and describes the therapeutic uses of BDNF, providing for the first time the means for obtaining sufficient amounts of substantially pure BDNF for clinical use. The invention also describes the preparation of antibodies directed against BDNF or fragments thereof, providing a process for obtaining sufficient immunogen. In addition, by allowing the comparison of BDNF and NGF nucleic acid sequences, the present invention allows the identification of homologous regions of the sequences.

<img file="PT95152B_D0020.tif" />

493

6526-019-118

Nucleic acid sequences between BDNF and NGF, thus defining a family of BDNF / NGF genes; The invention provides a process for identifying and isolating new members of this gene family.

For the sake of clarity of description, and not by way of limitation, the invention will be described in the following parts:

(i) BDNF purification (ii) BDNF bioassay (iii) BDNF protein microsequencing (iv) BDNF encoding DNA cloning (v) BDNF expression (vi) BDNF genes and proteins (vii) obtaining anti-antibodies -BDNF (viii) identification of new gene family members

BDNF / NGF (ix) (x) Utility of the Invention pharmaceutical compositions

5.1. PURIFICATION OF BRAIN DERIVED NEUROTROPHIC FACTOR

In order to identify BDNF-encoding nucleic acids, a few micrograms of BDNF protein can be obtained from tissues to determine amino acid sequences that can then be used to obtain oligonucleotide probes. The extreme rarity of BDNF protein practically limits the amount of amino acid sequences that can be determined. BDNF can be prepared from pig brain by the methods described in Barde et al. (1982, EMBO J. 14: 549-553), or Hofer and Barde (1988, Nature 331 = 261-262).

Preferably BDNF may be prepared according to the following detailed protocol, given by way of example and not by way of limitation; Various modifications may be considered for use by the skilled artisan. Due to the rarity of BDNF, about 6 kilograms of brain tissue may preferably be used in a purification process. Brain tissue can be homogenized in sodium phosphate buffer at a concentration of about 0.2 M sodium phosphate and at a pH of approximately

<img file="PT95152B_D0021.tif" />

6526-019-118

-286, containing 1mM EDTA and newly added 1mM phenylmethanesulfonyl fluoride, so that the brain tissue / fluid ratio is approximately 1 kg brain to 21 µl buffer. Hydrochloric acid may then be used to bring the pH of the mixture to about 4, and the mixture may then be stirred for about 2 h at 4 ° C. The mixture can then be centrifuged for 25 min at 20,000 g. After centrifugation the supernatant is collected, adjusted to a pH of about 6.0 using sodium hydroxide and then stirred with about 1 l of pre-swollen carboxymethylcellulose (per 6 kg brain tissue) which had solid was equilibrated with 0.1 M sodium phosphate at a pH of 6. After several washes with a total of about 20 l of 0.1 M sodium phosphate, pH 6, the suspension may be poured onto a column and washed with same buffer containing 0.13 M NaCl, preferably overnight. Active fractions, identified by BDNF-sensitive bioassay (see Section 5.2, below), can then be eluted with phosphate buffer containing 0.5 M NaCl and subsequently dialyzed against various changes of about 5 1 of 5 mM potassium phosphate at pH from 6.8. The dialyzed fractions may then be applied to a hydroxyapatite column with a bed volume of about 20 ml per kg of brain tissue processed; the hydroxyapatite column should be pre-equilibrated with 5 mM potassium phosphate at a pH of 6.8 prior to sample application. The column can then be eluted with a linear gradient composed of 500 ml of 5 mM potassium phosphate and 500 ml of 700 mM potassium phosphate, both at a pH of about 6.8. BDNF activity should be optimally eluted at approximately 500 mM potassium phosphate. The mixed active fractions can then be adjusted to a final molarity of about 700 mM potassium phosphate and applied to a phenyl sepharose column (with a volume of about 5 ml per 6 kg of brain tissue processed), equilibrated with a 700 mM potassium phosphate solution with a pH of 6.8. After washing with approximately 40 ml of the same buffer, BDNF activity may be eluted with 0.1 M potassium phosphate, pH 6.8, and then BDNF activity may be dialyzed against distilled water, and then lyophilized. 0 material

493

6526-019-118

<img file="PT95152B_D0022.tif" />

The lyophilized gel may be taken into an SDS gel electrophoresis sample buffer containing 0.1% SDS but preferably not containing mercaptoethanol, and applied to an SDS gel comprising a linear gradient of 10 to 25% acrylamide. After electrophoresis separation is complete, the gel can be stained for about 10 min with Coomassie blue and bleached for about 20 min. A migrating band at the level of a cytochrome C marker may be removed and eluted from the gel by electrophoresis. SDS can be largely removed as described by Weber and Kuter (1971, J. Biol. Chem. 246: 4504-4509). Note that SDS removal is not generally complete. A process for BDNF microsequencing has been modified by Hofer and Barde (1988, Nature 331: 261-262). and does not use gel electrophoresis as a last step of purification.

5.2. BIOAVALIATION DQ BRAIN DERIVED NEURQTROPIC FACTOR

Any method which qualitatively or quantitatively indicates the activity of BDNF according to the present invention may be used. These bioassays may be useful in identifying and / or measuring the activity of natural or recombinant BDNF.

Any BDNF bioassay known in the art may be used. For example, chick embryo dorsal root ganglion (DRG) neurons may be used in a BDNF assessment, as described in Barde et al. (1980, Proc. Natl. Acad. Sci. USA 77: 1199-1203).

By way of example, dorsal root ganglia of chick embryos from day 6 to 14, and preferably chick embryos of 10 to 12 days, can be collected using the dissection techniques already known in the art and immediately placed in a small F14 medium volume (made of GIBCO F-12 powder, supplemented as in Vogel et al., 1972, Proc. Natl. Acad. Sci. USA 69: 3180-3184), which should be replaced at the end of saline dissection Phosphate Buffered Buffer (PBS)²⁺ in G²⁺containing about 0.1% trypsin. After about 20 min incubation at 37 ° C, the ganglia can be centrifuged and washed.

493

6526-019-118

<img file="PT95152B_D0023.tif" />

2X with F14 medium containing about 10¾ v / v heat-inactivated horse serum. The ganglia can then be dissociated by gentle trituration (about 10-15 aspirations) using a small diameter (about 1 mm) siliconized pasteur pipette. The remaining tissue agglomerates may then preferably be removed by passing the cell suspension through a nylon mesh with about 40 µm pores. The cell suspension may then be pre-plated for about 210 min on a plastic tissue culture plate; During this time, most non-neuronal cells should adhere to the plastic surface, leaving a neuron-enriched cell population in suspension. Cells may be diluted to a concentration of about 5x10 4 cells per ml of plate medium (preferably F-14 medium supplemented with 10¾v / v heat-inactivated horse serum along with antibiotics) and placed in polynitine. or preferably on polynitine / laminin coated tissue culture plates (see below).

Alternatively, and not as a limitation, BDNF bioassessions that are relatively insensitive to NGF may in some circumstances be preferable to systems such as the DRG system described above which may under certain conditions respond to both BDNF and NGF These relatively specific BDNF systems include retinal ganglion cultures as well as neuronal cultures derived from neural placodes.

Peri natal retinal cells can be cultured according to the methods described in Johnson et al. (1986, J. Neurosci. 6: 3031-3038). For example, retinas may be removed from peri-natal animals (in the rat the term prenatal refers to embryos from embryonic day 17 to 48 h postnatal pups) washed in phosphate-buffered saline ( Calcium and magnesium free PBS), then incubated in PBS containing about 0.05 - 0.1 0.1 trypsin for approximately 15 min at about 37 ° C. After proteolytic digestion, retinas may be washed in F14 culture medium containing about 10 d v / v horse serum (HS), inactivated by

493

6526-019-113

<img file="PT95152B_D0024.tif" />

heat, (F14-HS). Retinas are then dissociated by gentle pipetting into a small volume (about 1-10 ml) of fresh F14-HS. The undissociated tissue is allowed to settle and the remaining cells and medium are pipetted out of the culture.

Alternatively, BDNF bioassays comprising cells derived from neural placodes may be used in accordance with the invention. Explants of the embryonic cranial nerve, p. ex. The ventrolateral portion of the trigeminal ganglion or the vestibular, geniculate, rohedo, or nodose ganglia may be cultured on culture media-covered collagen gel, as described in Davies et al. (1986, J. Neurosci. 6: 1897-1904) or dissociated according to Barde et al. (1980, Proc. Natl. Acad. Know 77: 1199-1203).

As it has been observed that BDNF responsiveness can be increased by 10X by changing the growth substrate from poliornvine to laminin-poliornitine (Barde et al., 1987, Prog. Brain Res. 71 = 185-189) bioassays BDNF are preferably made on laminin substrates. Crop surfaces can be prepared, e.g. (i) coating the culture surface for about 8-10 h with a sterile polyornitine solution; (ii) washing several times with sterile water; and (iii) coating the culture surface for about 2 h with laminin at a concentration of approximately 25 µg / ml in PBS (Johnson et al., 1986, J. Neurosci. 6 = 3031-3038).

In any bioassay system according to the invention, a BDNF dose response curve can be established using methods known in the art. Using an evaluation of the dissociated sensory ganglion of the chick, half of the maximum survival at a concentration of about 5 ng / ml of purified BDNF was observed and a maximum survival at a concentration of about 10 to 20 ng / ml was observed. of purified BDNF (Barde et al., 1987, Prog. Brain Res. 71 = 185-189).

5.3. HYDROCHNOSIS OF PROTEIN D0 BRAIN DERIVATED NEURRICH FACTOR

BDNF protein prepared from the brain can be

<img file="PT95152B_D0025.tif" />

493

6526-019-118

Sequenced; It should be noted, however, that the extreme rarity of the protein makes it unlikely that a substantial portion of the BDNF protein sequence can be reliably obtained. The protein may be directly sequenced or initially cleaved by any protease or other compound known in the art, including (but not limited to) Staphylococcus aureus V8, trypsin and cyanogen bromide. The peptides may be sequenced by automated Edman degradation in a gas phase microsequencer according to the methods of Hewickk et al. (1981, J. Biol. Chem. 256-7990-7997) and Hunkapillar et al. (1983, Methods Enzymol, 91 = 227-236). Amino acid detection by phenylthiohydantoin can then be conducted according to Lottspeich (1985, Chromatography 326 = 321-327). Overlapping fragments of the amino acid sequence can be determined and used to deduce longer lengths of contiguous sequences.

5.4. CLONATION OF DNA NEUROTROPHIC FACTOR ENCODER

BRAIN DERIVATIVE

The rarity of BDNF effectively precludes the use of classical strategies for cloning the BDNF gene. For example, if the available protein sequence is used to create a complementary labeled oligonucleotide probe and this probe was used to screen the cDNA library obtained from the tissue presumed to synthesize BDNF, the number of positive clones will be so small that it is evanescent. 0 The present invention provides a combination of methods for cloning the BDNF gene, comprising purifying adequate amounts of BDNF protein, sequencing BDNF protein, obtaining an oligonucleotide probe, constructing a cDNA library, amplification based on the amino acid sequence derived from BDNF; and finally, selection of the BDNF gene. In this method

<td>of the invention,</td><td>the preferred process of</td><td colspan="2">amplification</td><td colspan="2">uses the</td>

<td>amplification</td><td>of acid sequences</td><td>nucleic</td><td>of</td><td>fabric</td><td>per</td>

<td>reaction on</td><td>polymerase chain (PCR)</td><td>(Saiki</td><td>et</td><td>al.,</td><td> 1985,</td>

<td>Science 250</td><td>= 1350-1354) with a view to</td><td>expand</td><td>O</td><td>number</td><td>in</td>

BDNF sequences available for cloning. A detailed description of the preferred process follows.

493

6526-019-118

<img file="PT95152B_D0026.tif" />

First, the amino acid sequence derived from the purified BDNF protein can be used to deduce the oligonucleotide primers for use in the polymerase chain reaction. Due to the degeneracy of the genetic code in which multiple nucleic acid triplets may specify the same amino acid, multiple oligonucleotides should be synthesized in a given amino acid sequence in order to obtain potential multiple combinations of amino acid sequences; the resulting oligonucleotides are referred to as degenerate primers.

Polymerase Chain Reaction (PCR) requires chain primers, sense as well as antisense. Thus a degenerate oligonucleotide primer corresponding to the BDNF amino acid sequence can be used as primer for a DNA strand and a second primer, homologous to a commonly occurring DNA sequence, e.g. ex. A portion of thymidine residues resulting from reverse transcription of the mRNA polyadenosine tail can be used as the second strand primer of DNA. These primers may then be used in the polymerase chain reaction with a nucleic acid standard that is presumed to contain the BDNF coding sequences, e.g. ex. Genomic DNA or preferably cDNA prepared from mRNA collected from tissues presumed to synthesize BDNF. DNA reaction products can then be analyzed by electrophoresis to determine if the DNA reaction product has a molecular size similar to the expected size. BDNF gene and preferably by nucleotide sequence analysis.

However, as the use of two degenerate primers in PCR increases the possibility of amplifying non-BDNF-encoding nucleic acid sequences, a preferred method utilizes only one degenerate primer, with the other primer corresponding to the exact sequence of BDNF. In order to identify the exact sequence of BDNF, the amino acid sequence determined using purified BDNF protein can be used to obtain both sense and antisense oligonucleotide primers. 0 DNA reaction product resulting from the use of these primers in PCR using? as we shape BDNF-encoding nucleic acids,

493

6526-019-118

<img file="PT95152B_D0027.tif" />

it should be a nucleic acid fragment encoding the amino acid stretch used to obtain the primers and can be of a predictable size (eg at least the number of amino acid residues multiplied by 3 base pairs in length). ). Sequential analysis of the DNA reaction product can now be compared with that of the investigated amino acid sequence to confirm that the amplified amino acid sequence can indeed encode the BDNF peptide fragment. While any nucleic acid sequencing process known in the art may be used, it is preferable to use the dideoxynucleotide chain termination process (Sanger et al., 1979, Proc. Natl. Acad. Sci. USA 72 = 5918-5921 ). Sequencing may be accomplished using purified gel or preferably cloned DNA reaction product. The DNA reaction product can then be used to obtain an oligonucleotide primer corresponding to the exact BDNF coding sequence. This Primer may then be used in conjunction with a second, degenerate primer to increase the length of the BDNF sequence beyond that represented by the exact sequence fragment initially determined. For example, and not by way of limitation, the loop sense primer may correspond to the exact nucleotide sequence of the BDNF, while the antisense primer may be a degenerate primer homologous to a sequence region likely to be downstream. '. · Of the sequenced fragment, e.g. ex. the polyadenosine tail of the mRNA inversely transcribed into the cDNA. It may be necessary to use a similar method to find the upstream sequence of the sequenced fragment; for example, the antisense strand primer may correspond to the exact BDNF nucleotide sequence and the sense strand primer may be a degenerate primer, homologous to a region upstream of the BDNF sequence above the sequenced fragment, e.g. . ex. a 5 'tail of polyadenosine added to the 5' end of cDNA using a terminal deoxynucleotide transferase. The sequence of the complete gene or BDNF mRNA can thus be assembled.

493 6526-019-1IS

<img file="PT95152B_D0028.tif" />

DNA reaction products may be cloned using any method known in the art. A large number of host vector systems known in the art may be used. Possible vectors include, but are not limited to, modified cosmids, plasmids or viruses, with the restriction of the vector system being compatible with the host cell used. These vectors include (without limitation) bacteriophages such as lambda derivatives or plasmids such as pBR322, pUC or the plasmid derivatives of Bluescript (R) (Stratagene). Recombinant molecules can be introduced into host cells via transformation, transfection, infection, electroporation, etc. »BDNF gene is inserted into a cloning vector that can be used to transform, transfect or infect appropriate host cells to create many copies of gene sequences. This can be accomplished by ligating the DNA fragment into a cloning vector that has complementary cohesive limits. However, if complementary restriction sites used to fragment DNA are not present in the cloning vector, the ends of the DNA molecules may be enzymatically modified. It may be advantageous to incorporate restriction endonuclease cleavage sites into the oligonucleotide primers used in the polymerase chain reaction to facilitate insertion of the vectors. Alternatively, any desired site may be produced by linking nucleotide sequences (linkers) to the DNA boundaries; these linked linkers may include chemically synthesized specific oligonucleotides, restriction endonuclease recognition sequence coders. In an alternative process, the cleaved vector and the BDNF gene may be modified by homopolymeric tails.

In certain specific arrangements, transformation of host cells with recombinant DNA molecules incorporating an isolated BDNF gene, or a synthesized DNA sequence, allows multiple copies of the gene to be obtained. Thus the gene can be obtained in large quantities by the growth of transformants, isolating recombinant DNA molecules from the transformants and, when necessary, separating the inserted gene from the transformants.

493

6526-019-118

Isolated recombinant DNA.

According to a preferred embodiment of the invention, once a BDNF-encoding cDNA-derived clone is created, a BDNF-encoding genomic clone can be isolated using classical techniques known in the art. For example, a labeled nucleic acid probe can be obtained from the BDNF clone and used to screen genomic DNA libraries by nucleic acid hybridization using, e.g. eg, the method established by Benton and Davis (1977, Science 196 = 180) for bacteriophage libraries and Srunstein and Hogness (1975, Proc. Natl. Acad, Sci. USA 72 = 3961-3965) for plasmid libraries. The recovered clones can then be analyzed by restriction fragment maps and sequencing techniques according to methods well known in the art.

In addition, other cDNA clones can be identified from a cDNA library using sequences obtained according to the invention.

5.5. BRAIN DERIVED NEUROTROPHIC FACTOR GENE EXPRESSION

The nucleotide sequence encoding the do-BDNF protein or portion thereof may be inserted into an appropriate expression vector, that is, a vector containing the elements necessary for transcription and translation of the inserted sequence encoding the protein. The necessary transcription and translation signals may also be provided by the native BDNF gene and / or its flank regions. Various host vector systems may be used to express the protein coding sequence. These include, but are not limited to, virus-infected mammalian cell systems (e.g., vaccinia virus, adenovirus, etc.); virus-infected insect cell systems (e.g. baculovirus); microorganisms such as yeast containing yeast vectors, or bacteria transformed by bacteriophage DNA, plasmid DNA or cosmid DNA. The expression elements of these vectors vary in strength and specificity. Depending on the host-vector system

<img file="PT95152B_D0029.tif" />

493

6526-019-118

-37 used, any of several suitable transcription and translation elements may be used,

Any of the above described methods for inserting DNA fragments into a vector can be used to construct expression vectors containing a chimeric gene consisting of transcription / translation control signals and protein coding sequences. These processes may include in vitro recombinant DNA, synthesis techniques and in vivo recombination (genetic recombination). Expression of the nucleic acid sequence encoding a BDNF protein or peptide fragment may be regulated by a second nucleic acid sequence such that the BDNF or peptide protein is expressed in a host transformed by the DNA molecule. recombinant. For example, BDNF expression may be controlled by any promoter / enhancer element known in the art. Promoters that can be used to control BDNF expression include, but are not limited to, the SV40 early promoter region (Bernoist and Chambon, 1981, Nature 290: 304-310), the promoter contained in the long terminal repeat motif. Rous sarcoma virus 3 '(Yamamoto et al., 1980, Cell 22: 787-797), the herpes thymidine kinase promoter (Wagner et al., 1981, Proc. Natl. Acad. Sci. USA 78: 144-1445). the regulatory sequences of the metallothionine gene (Brinster et al., 1982, Nature 296: 39-42); prokaryotic expression vectors such as the? 6-lactamase promoter (Villa-Kamaroff et al., 1978, Proc. Natl. Acad, Sci. USA 75: 3427-3731) or the tac promoter (DeBoer et al., 1983, Proc Natl. Acad. Sci. USA 80: 21-25). see also Useful proteins from recombinant bacteria in Scientific American, 1980, 242: 74-94; plant expression vectors comprising the nopaline synthetase promoter region (Herrera-Estrella et al., Nature 303: 209-213) or the cauliflower S35 nosal virus RNA promoter (Gardner et al., 1981, Nucl , Acids Res. 9: 2871) and the photosynthetic enzyme promoter ribulose bisphosphate carboxylase (Herrera-Estrella et al., 1984, Nature 310: 115-120); promoters of yeast or other fungi such as the Gal 4 promoter, the ADC

493

6526-019-118

<img file="PT95152B_D0030.tif" />

-hydrogenase), PGK promoter (phosphoglycerol kinase), alkaline phosphatase promoter and the following animal transcriptional control regions showing tissue specificity and which have been used in transgenic animals: elastase I gene control region that is active in cells pancreatic acinar (Swift et al., 1984, Cell 58 = 659-646; Ornitz et al., 1986, Cold Spring Harbor Symp. Quant. Biol. 50 = 399-409; MacDonald, 1987, Hepatology 7 = 425-515); control region of the insulin gene that is active in pancreatic beta cells (Hanahan, 1985, Nature 315 = 115-122). control region of the immunoglobulin gene that is active on lymphoid cells (Grosschedl et al., 1984, Cell 58 = 647-658; Adames et al., 1985, Nature 318 = 535-558; Alexander et al., 1987, Mol Cell Biol. 7 = 1436-1444), mouse mammary tumor virus control region that is active in testicular, cervical, lymphoid, and mast cell cells (Cleder et al., 1986, Cell 45 = 485-4-95), albumin gene control region that is active in the liver (Pinkert et al., 1987, Genes and Devi. 1 = 268-276), alpha-fetoprotein gene control region that is active in the liver (Krumlauf et al ., 1985, Mol. Cell, Biol. 5 = 1639-1648; Hammer et al., 1987, Science 255 = 55-58); alpha-1-antitrypsin gene control region that is active in the liver (Kelsey et al., 1987, Genes and Devi. 1 = 161-171), beta-globin gene control region that is active in myeloid cells (Mogram et al. , 1985, Nature 515 = 558-540; Kollias et al., 1986, Cell 46 = 89-94; myelin basic protein gene control region that is active in oligonucleotide cells: in the brain (Beadhead et al., 1987, Cell 48 = 705-712); myosin light chain 2 gene control region that is active in the skeletal muscle (Sani, 1985, Nature 514 = 285-286) and the gonadotropic hormone releasing gene control region that is active in the hypothalamus (Mason et al ., 1986, Science 254 = 1572-1578).

Expression vector insertions containing the BDNF gene can be identified by 3 general procedures = (a) DNA-DNA hybridization, (b) presence or absence of gene marker functions and (c) expression of inserted sequences. In the first process, the presence of a foreign gene inserted into a vector

<img file="PT95152B_D0031.tif" />

493

6526-019-118

Expression can be detected by DNA-DNA hybridization using probes that include sequences homologous to those of the inserted BDNF gene. In the second process, the recombinant host / vector system can be identified and selected based on the presence or absence of certain gene marker functions. (e.g. thymidine kinase activity, resistance to antibiotics, transformation phenotypes, occlusion body formation in baculovirus, etc.) caused by insertion of foreign genes into the vector. For example, if the BDNF gene is inserted into the vector marker gene sequence, recombinants containing the BDNF insert may be identified by the absence of marker gene function. In the third process, recombinant expression vectors may be identified. evaluating the foreign gene product expressed by the recombinant. Such assessments may be based, for example, on the physical or functional properties of the BDNF gene product in bioassessment systems described above in Section 5.2.

Once a certain recombinant DNA molecule has been identified and isolated, various methods already known in the art may be used to propagate it. Once growth conditions and a suitable host system have been established, recombinant expression vectors can be propagated and prepared in appreciable amounts. As noted above, expression vectors which may be used include (but are not limited to) the following vectors or derivatives: human or animal viruses such as vaccinia virus or adenovirus; insect viruses such as baculovirus; yeast vectors; bacteriophage vectors (e.g. lambda) and plasmid and cosmid DNA vectors, to name but a few.

In addition, a host cell strain that modulates expression of the inserted sequences or that modifies and processes the gene product in the specific manner desired may be chosen. Expression of certain promoters may be increased in the presence of certain inducers; thus, genetically engineered expression of BDNF protein can be controlled. It also happens that the different host cells have

493

6526-019-118

<img file="PT95152B_D0032.tif" />

characteristic and specific mechanisms for the processing of translation, post-translation and modification (eg glycosation, cleavage) of proteins. Appropriate cell lines or host systems may be chosen to ensure the desired modification and processing of the expressed foreign protein. For example, expression in a bacterial system may be used to produce a non-glycated core protein product. Expression in yeast will produce a glycoside product. Mammalian cell expression can be used to ensure native glycosation of the heterologous BDNF protein. In addition, different vector / host expression systems can effect processing reactions such as proteolytic cleavages of different intensities.

In a specific arrangement of the invention, prepro BDNF encoding DNA may be cloned into pCMV plasmid, amplified and then used to transfect COS cells by the calcium phosphate method (Chen and Okayama, 1987, Mol. Cell. Biol. 7: 2745- 2752); BDNF activity can then be collected from a cell culture medium (See Sections of Example 10, infra).

5.5.1. PRODUCT IDENTIFICATION AND PURIFICATION

GENEEXPRESSO

Once the recombinant expressing the BDNF gene has been identified, the gene product can be analyzed. This is achieved by testing based on the physical or functional properties of the product.

Once BDNF protein is identified, it can be isolated and purified by classical methods such as chromatography (e.g. ion exchange, affinity and column classification), centrifugation, differential solubility or any other classical technique for protein purification. Functional properties may be assessed by any suitable assay including (but not limited to) dorsal root of the embryo, perinatal rat retinal cells or neurons derived from the neural placode.

493

6526-019-118

It is important to note that processes used to prepare BDNF from brain tissues, as they involve, as a final step, preparative gel electrophoresis, produce BDNF that is not fully active due to the presence of residual SDS (Barde and Thoenen, 1985). , in Hormones and Cell Regulation, Vol. 9, Dumont et al., eds., Elsevier Science Publishers, pp. 385-390). In contrast the present invention allows the isolation of BDNF which is produced from recombinant nucleic acid molecules and which, being free of SDS, therefore has complete activity. For example (and without limitation) the anti-BDNF antibodies of the invention ( eg antibody directed against porcine BDNF amino acid fragment B5-33 (described in Section 11, below) can be used to collect recombinant BDNF by immunoprecipitation or affinity chromatography, thus producing fully active, detergent free BDNF »

In another embodiment of the invention, prepro BDNF can be enzymatically converted to active mature BDNF using ρ, e.g., Arg-C endoproteinase (see Example Section 17, below).

5.6. GENES AND PROTEINS OF BRAIN DERIVED NEUROTROPHIC FACTOR

Using the methods referred to above and those of Sections of Example 6 and 9, below, the following nucleic acid sequences were determined and their corresponding amino acid sequences deduced. The genomic sequence of the pig BDNF cDNA depicted in Figure 1. The genomic sequence of the human BDNF cDNA depicted in Figure 5 which also shows the sequences of the pig, rat and chicken DNA was determined. Each of these sequences or their functional equivalents may be used in accordance with the invention. In addition, the invention relates to BDNF genes and proteins isolated from porcine, bovine, feline, poultry, equine or canine animals, as well as primates or any other species where BDNF activity exists. The invention is also directed to BDNF nucleic acid subsequences comprising at least ten nucleotides, these subsequences comprising hybridizable portions of the BDNF sequence having application, e.g. for example, in hybridization assays of

493

6526-019-118

-42 nucleic acids, Southern and Northern blot analyzes, etc. The invention also provides BDNF protein and fragments and derivatives thereof, according to the amino acid sequences set forth in Figures 1 and 5 or their functional equivalents. The invention also prepares BDNF protein fragments or derivatives comprising antigenic determinants or functionally active. Functionally active terms herein mean that they have positive activity in assays for known BDNF functions, e.g. ex. in chick embryo GRD trials,

For example, the nucleic acid sequences depicted in Figures 1 and 5 may be altered by substitutions, additions or deletions that provide functionally equivalent molecules. Due to the degeneracy of the nucleotide coding sequences other DNA sequences which substantially encode the same amino acid sequences represented in Figures 1 and 5 may be used in the practice of the present invention. They include (but are not limited to) nucleotide sequences comprising all or part of the BDNF genes depicted in Figures 1 and 5 which are altered by substituting the different codons encoding a functionally equivalent amino acid residue in the sequence, thereby producing a change without symptoms (silent change). Similarly, the BDNF proteins or fragments or derivatives thereof of the invention include (without limitation) those which contain, as a major amino acid sequence, all or only part of the amino acid sequence, substantially as in Figures 1 and 5, including altered sequences where functionally equivalent amino acid residues replace residues within the sequence, resulting in a symptomless change. For example, one or more amino acid residues within the sequence may be replaced by other amino acids of similar polarity that act as functional equivalents, resulting in a symptom-free change. Substitutes for an amino acid within the sequence may be chosen from among other members of the class to which the amino acid belongs.

J

<img file="PT95152B_D0033.tif" />

493

6526-019-118

Non-polar (hydrophobic) amino acids include alanine, leucine, isoleucine, valine, proline, phenylalanine, tryptophan and methionine. Polar neutral amino acids include glycine, serine, threonine, cysteine, tyrosine, asparagine and glutamine. Positively charged (basic) amino acids include arginine, lysine and histidine. Negatively charged (acidic) amino acids include aspartic acid and glutamic acid. Also included within the scope of the invention are BDNF proteins or fragments or derivatives thereof which are differentially modified during or after translation, e.g. ex. by glycosylation, proteolytic cleavage, binding to an antibody molecule or another cell ligand, etc. '

In addition, a given BDNF may, by in vitro or in vivo mutation, create and / or destroy translation, initiation and / or termination sequences, or create variations in coding regions and / or form new restriction or endonuclease sites. destroy pre-existing ones to facilitate in vitro modifications later. Any mutagenesis technique already known in the art, including (without limitation) site-directed in vitro mutagenesis (Hutchinson et al., 1978, J. Biol. Chem. 253 = 6551), the use of TAS (^) links (Pharmacia), etc.

5.7. CREATION OF NEUROTROPHIC ANTI-FACTOR ANTIBODIES

BRAIN DERIVATIVE

According to the invention, the BDNF protein or fragments or derivatives thereof may be used as an immunogen to raise anti-BDNF antibodies. Previous attempts to produce an anti-BDNF immune response were unsuccessful possibly due to the small amounts of purified BDNF available, producing relatively abundant amounts of BDNF protein using recombinant protein synthesis techniques (based on nucleic acid sequences). BDNF of the invention), the problem of insufficient amounts of BDNF has been overcome.

To increase the likelihood of producing an anti-BDNF immune response, the amino acid sequence of BDNF can be analyzed at

493

6526-019-118

It is difficult to identify those parts of the molecule that may be associated with increased immunogenicity. The amino acid sequence may, e.g. be subjected to a computer analysis to identify surface epitopes according to the method of Hopp and Woods (1981, Proc. Natl. Acad. Sci. USA 78: 3824-3828) which has been successfully applied to to identify hepatitis B surface antigen antigenic peptides Alternatively, the deduced BDNF amino acid sequences for the different species can be compared and the relatively non-homologous regions identified; These non-homologous regions are more likely to be immunogenic for the various species.

To prepare monoclonal antibodies directed to BDNF any technique that produces antibody molecules by continuous cell line culture can be used. For example, the hybridoma technique, originally developed by Kohler and Hilstein (1975, Nature, 256: 495-497), as well as the trioma technique, the human 8-cell hybridoma technique (Kozbor et al., 1983). , Immunology Today 4:72) and the EBV-hybridoma technique for producing monoclonal antibodies in man (Cole et al., 19S5, in Nonoclonal Antibodies and Cancer Therapy, Alan R. Liss, Inc, pp. 77-96) and others. similar are within the scope of the present invention.

Monoclonal antibodies for therapeutic use may be human monoclonal antibodies or chimeric mouse (or other species) monoclonal antibodies. Human monoclonal antibodies can be prepared by any of the numerous techniques known in the art (eg, Teng et al., 1983, Proc. Natl. Acad. Sci. USA 80: 7308-7312; Kozbor et al., 1983, Immunology Today 4: 72-79 (Olsson et al., 1982, Meth. Enzymol. 92: 3-16). Chimeric antibody molecules may be prepared containing a mouse antigenic binding domain with human constant regions (Morrison et al., 1984, Proc. Natl. Acad. Sci, US, A, 81: 6851, Takeda et al., 1985 , Nature 314: 452).

Various processes already known in the art may be used to

J

<img file="PT95152B_D0034.tif" />

493

6526-019-118

Prepare polyclonal antibodies against BDNF epitopes. For antibody production, various host animals may be immunized by injection with BDNF protein or fragments or derivatives thereof, including (without limitation) rabbits, mice, rats, etc. Various adjuvants may be used to enhance the immune response, depending on the host species, including (without limitation) Freund's reagent (complete and incomplete), mineral gels such as aluminum hydroxide, surfactants such as lisolecithin, pluronic polyols, polyanions, peptides, oily emulsions, keyhole limpet hemocyanins dinitrophenol and potentially useful adjuvants for man such as BCG (Bacillus Calmette-Guerin) and Corynebacterium parvum.

A molecular clone of an antibody against a BDNF epitope can be prepared by known techniques. Recombinant DNA methodology (see, e.g., Maniatis et al., 1982, Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York) can be used to construct nucleic acid sequences encoding a monoclonal antibody molecule or antigen binding region thereof.

Antibody molecules may be purified by known techniques, e.g. by immunoabsorption or immunoaffinity chromatography, by chromatographic methods such as HPLC (high performance liquid chromatography) or a combination of these methods.

Antibody fragments containing the idiotype of the molecule may be prepared by known techniques. For example, such fragments include (but are not limited to): the F (ab ') 2 fragment that can be produced by pepsin digestion of the antibody molecule; Fab 'fragments which can be obtained by reducing the disulfide bridges of fragment F (ab') p, and Fab or Fab 2 fragments which can be obtained by treating the antibody molecule with papain and a reducing agent.^ç

Example Section 11 describes the preparation of polyclonal antisera directed against the B5-33 peptide fragment of

J

493

6526-019-118

-46 BDNF protein.

5.8. IDENTIFICATION OF NEW BDNF / NGF GENE FAMILY MEMBERS

Sequential analysis of the gene encoding BDNF and deduction of its amino acid sequence revealed that this protein bears many structural similarities to NGF (Figure 2). The main sequence of mature BDNF, as well as the general structure and likely mode of processing of a protein precursor, strongly suggest that the NGF and BDNF genes may have evolved from a common ancestral gene. In the mature polypeptide region, if only 3 intervals are inserted into the NGF sequences to optimize comparison, a total of 51 amino acid identities are found to be common to those of the NGF already known from many pig and human BDNF © species. These identities include six cysteine ​​residues, suggesting that NGF and BDNF share a very similar secondary structure. In addition, four segments of six or more amino acids can be seen where the NGF of all the above species and the pig BDNF are identical or differ by no more than one conservative amino acid substitution. It is therefore reasonable to conclude that NGF and BDNF appear to be closely related members of a gene family.

A rational search for other members of the BDNF / NGF gene family can be performed using an approach that takes advantage of the unanticipated existence of conserved, strongly homology NGF to BDNF segments. For example, other members of the selected BDNF gene family can be identified from various nucleic acid sequences, those that are homologous to BDNF and NGF, and then identify, among selected sequences, those that also contain nucleic acid sequences that are not. homologous to NGF and BDNF. By non-homologous terms is meant a region containing at least about 6 contiguous nucleotides where at least about 2 nucleotides differ in the NGF and BDNF sequences.

One of the preferred specific arrangements of the invention features

J

<img file="PT95152B_D0035.tif" />

493

6526-019-118

-47the following process. Corresponding to each of the four conserved segments (boxes) set out in Table III below, degenerate sets of degenerate oligonucleotide probes of at least 18 nucleotides, representing all possible amino acid coding sequences found in either NGF or NGF, can be synthesized. BDNF, for about 6 contiguous codons. By numbering with respect to the amino terminus mature polypeptides (so that His prepro BDNF His 134 is treated with His 1 on mature protein), the four boxes can be characterized as follows (numbered relative to mature human proteins; DNA sequence , represented in Figure 1),

TABLE III

DNA sequence

<td>Box 1 =</td><td>NGF</td><td>Gly10</td><td>- Serl9</td><td></td>

<td></td><td>BDNF</td><td>Gly8</td><td>- Ser17</td><td> 587-616</td>

<td>Box 2 =</td><td>NGF</td><td>LysSO</td><td>- Cys58</td><td></td>

<td></td><td>BDNF</td><td>LysSO</td><td>- Cys58</td><td> 713-739</td>

<td>Box 3:</td><td>NGF</td><td>Gly67</td><td>- Asp72</td><td></td>

<td></td><td>BDNF</td><td>Gly67</td><td>- Asp72</td><td> 764-781</td>

<td>Box 4 =</td><td>NGF</td><td>Trp99</td><td>- CysllO</td><td></td>

<td></td><td>BDNF</td><td>TrplOQ</td><td>- Cyslll</td><td> 863-898</td>

Synthetic oligonucleotides, derived from the sequence pairs of boxes indicated in Table III, can be used as primers to PCR amplify sequences from a source (RNA or DNA) of potential interest. This may include mRNA or cDNA or genomic DNA of any eukaryotic species which may express a polypeptide closely linked to BDNF or NGF. Performing only six PCR reactions (namely = a Box 1 primer with a Box 2 primer; Box 1 with Box 3; Box 1 with Box 4; Box 2 with Box 3; Box 2 with Box 4; Box 3 with Box 4) it may be possible to detect a gene or gene product by sharing any 2 of the above 4 segments of conserved sequence between NGF and BDNF. If someone decides to synthesize several different degenerate initiators from each box, it may still be possible to perform a

J

<img file="PT95152B_D0036.tif" />

493

6526-019-118

Complete screening with a reasonably small number of PCR reactions. It is also possible to vary the stringency of the hybridization conditions used at the beginning of the PCR reactions to allow more or less similarity between the unknown gene nucleotide sequences and those of NGF or BDNF. If a segment of a previously unknown member of the BDNF / NGF gene family is successfully amplified, that segment can be molecularly cloned and sequenced, and used as a probe to isolate a complete cDNA or genomic clone. This, in turn, will allow determination of the complete nucleotide sequence of the unknown gene, analysis of its expression and production of its protein for functional analysis.

The approach described above has been used to identify a novel gene related to both BDNF and NGF and is described in Example Section 13, infra.

In addition, the present invention discloses the use of sequential BDNF / NGF homologies in the preparation of novel recombinant molecules that are members of the BDNF / NGF gene family but cannot occur in nature. For example (and not as a limitation), a recombinant molecule may be constructed according to the invention comprising portions of NGF and BDNF genes. Such a molecule may show properties associated with both NGF and BDNF, and present a new profile of biological activities, including not only agonists, but also primary sequence of BDNG and NGF. It may also predict the tertiary structure of the molecules using computer simulation (Hopper and Woods, 1981, Proc. Natl. Acad. Sci., USA 78 = 3824-3828); BDNF / NGF chimeric recombinant genes can be prepared in light of the correlations between tertiary structure and biological function. Chimeric genes may also be constructed comprising portions of either or. more members of the BDNF / NGF gene family, including the new member described in Section 13.

antagonists. To be used for

5.9. UTILITY OF THE INVENTION The present invention describes the preparation of the sequence of

J

<img file="PT95152B_D0037.tif" />

493

6526-019-118

-49 nucleic acids of BDNF and its protein, substantially pure peptide fragments or derivatives. BDNF may be produced for the first time in sufficient quantities for diagnosis and therapeutic applications. Also, anti-BDNF antibodies, also available for the first time as a result of the invention, and the nucleic acid probes of the invention may be used in diagnosis and therapeutic applications. For most purposes it is preferable to use BDNF genes or gene products of the same species for diagnostic and therapeutic purposes, although the use of cross-species BDNF may be useful in specific embodiments of the invention.

5.9.1. DIAGNOSTIC APPLICATIONS The present invention which relates to BDNF-encoding nucleic acids and proteins, peptide fragments or derivatives as well as antibodies to BDNF protein, peptides or derivatives may be used to diagnose diseases and disorders of the nervous system that may be associated with BDNF expression pattern changes.

In various embodiments of the invention, BDNF genes and their nucleic acid sequences and subsequences including complementary sequences may be used in diagnostic hybridization assays. BDNF nucleic acid sequences, or their subsequences comprising about 15 nucleotides, may be used as hybridization probes. Hybridization assays may be used to detect, predict, diagnose or monitor conditions, disorders or diseases associated with changes in BDNF expression, including in particular those resulting from damage to sensory neurons. These diseases and conditions include (but are not limited to) CNS trauma, infarction, infection, degenerative nerve diseases, malignant conditions, or postoperative changes including (but not limited to) Alzheimer's disease, Parkinson's disease, Huntington's chorea and degenerative retinal diseases . For example, total RNA can be evaluated in a patient's tissue sample by the presence of BDNF mRNA, where a change in the amount of BDNF mRNA is indicated.

<img file="PT95152B_D0038.tif" />

493

6526-019-118

-50-minute neuronal degeneration.

Relatively high levels of BDNF mRNA have been identified in neuroblastoma tumor cells (see Section 14, infra, for more details); thus, in a specific arrangement of the invention, detection of increased mRNA levels using BDNF nucleic acid probes can be used for the diagnosis of neuroblastoma tumor. The extremely low levels of BDNF synthesized by most tissues make BDNF mRNA a particularly useful tumor identifier.

In other embodiments of the invention, protein-directed antibodies, peptide fragments or BDNE- derivatives may be used to diagnose disorders and disorders of the nervous system, including in particular sensory disorders and degenerative retinal disorders, as well as disorders of the retina. and above diseases. Antibodies of the invention may be used e.g. ex. in in situ hybridization techniques using tissue samples obtained from a patient in need of such evaluation. In another example, the antibodies of the invention may be used in ELISA procedures to detect and / or measure amounts of BDNF present in tissue or fluid samples; Similarly, antibodies of the invention may be used in Western blot analysis to detect and / or measure the -BDNF present in tissue or fluid samples. An antibody of the invention that binds to BDNF in ELISA and Western blots is described in Section 11, infra.

In other embodiments of the invention, protein, peptide fragments or BDNF derivatives may be used to diagnose diseases and disorders of the nervous system. In a particular arrangement (and without limitation), protein or fragment of labeled peptides may be used to identify BDNF receptor expressing tissues or cells with a view to identifying aberrations of BDNF receptor expression and, consequently, potential tissue or tissue abnormalities. in the cellular response to BDNF.

5.9,2. THERAPEUTIC APPLICATIONS The present invention relates to nucleic acids which

493

6526-019-118

<img file="PT95152B_D0039.tif" />

encoding BDNF and its protein, peptide fragments or derivatives thereof, as well as antibodies directed against BDNF protein, peptides or derivatives, may be used to treat diseases and disorders of the nervous system that may be associated with changes in the pattern of BDNF expression or who may benefit from exposure to BDNF or anti-BDNF antibodies.

In various embodiments of the invention, protein, peptide fragments or BDNF derivatives may be administered to patients whose nervous system has been affected by trauma, surgery, ischemia, infection, metabolic disease, nutritional deficiency, malignant states or toxic agents. 0 The invention may in particular be used to treat conditions in which damage to sensory neurons or retinal ganglion cells has occurred by administering effective therapeutic amounts of BDNF protein, peptide fragments or derivatives. In various specific embodiments of the invention, BDNF may be administered locally to sensory neurons that have been impaired by including (without limitation) neurons in the dorsal root ganglia or in any of the following tissues: geniculate, rock and nodular ganglia; the acoustic vestibule complex of the cranial nerve; the ventrolateral pole of the maxillo-mandibular lobe of the trigeminal ganglion, the mesencephalic trigeminal nucleus ganglia and the sympathetic ganglia. It may be desired to administer BDNF-related peptides or BDNF protein by membrane adsorption, e.g. eg, a silastic membrane that may be implanted near the affected nerve. The present invention may also be used e.g. ex. to accelerate the recovery of patients suffering from diabetic neuropathies, e.g. e.g., multiplex mononeuropathies.

In other embodiments of the invention, the BDNF protein or its peptide fragments or derivatives thereof may be used to treat congenital conditions or neurodegenerative disorders, including (but not limited to) Alzheimer's disease, Parkinson's disease, Parkinson's syndrome. Plus (where parkinsonian symptoms result from degeneration of dopaminergic neurons), such as Progressive Supranuclear Palsy (Steele-Richardson-Olszewski Syndrome), Olivopontocerebellar phytrophy

<img file="PT95152B_D0040.tif" />

-52 (OPCA), Shy-Drager Syndrome (multiple system atrophy) and the Guamanian Parkinsonism dementia complex, and Huntington's chorea; In particular, the invention may be used to treat congenital or neurodegenerative disorders associated with sensory nerve dysfunction and degenerative retinal diseases. For example, protein or peptide fragments or BDNF derivatives may be used in the treatment of hereditary spinal paraplegia with retinal degeneration (Kjellin and Barnard-Scholz syndrome), retinitis pigmentosa, Stargardt's disease, Usher's syndrome (retinitis pigmentosum pigmentosa). congenital hearing loss) and Refsum syndrome (retinitis pigmentosa, hereditary hearing loss and polyneuropathy), to name but a few. It is possible that a defect in BDNF synthesis or responsiveness may be the underlying etiology of syndromes characterized by a combination of retinal degeneration with other sensory dysfunction.

In a specific arrangement of the invention, administration of the BDNF protein, or its peptide fragments or derivatives thereof, may be used in conjunction with surgical tissue implantation in the treatment of Alzheimer's disease and / or Parkinson's disease. Sections 12 and 18, infra, BDNF can be used to improve the survival of dopaminergic substantia nigra neurons (Figure 34), This confirms the use of BDNF in the treatment of CNS dopaminergic neuron disorders, including (but not limited to) Parkinson's disease [particularly in light of the results presented in Section 21, below, which shows that BDNF can be used to prevent neurotoxicity caused by MPP, a toxin associated with the induction of a Parkinson's disease-like syndrome]. In addition, it has been observed that BDNF supports the survival of CNS cholinergic neurons (Sections 12 and 16, infra) and in particular basal forebrain cholinergic neurons, indicating that BDNF may be useful in the treatment of disorders involving neurons. cholinergic agents including (but not limited to) Alzheimer's disease. Approximately 35% of Parkinson's disease patients were found to be

493

6526-019-118

<img file="PT95152B_D0041.tif" />

Colds of Alzheimer's dementia; BDNF produced according to the invention may prove to be a unique therapy agent useful for this disease complex. Similarly, BDNF prepared according to the invention may be used to therapeutically treat Alzheimer's disease in conjunction with Down's syndrome. The BDNF produced according to the invention may be used to treat various dementias as well as congenital learning disorders. BDNF has also been found to appear to suppress the proliferation of astroglial cells, which recommends the use of BD-NF to decrease scarring. In the CNS (eg after surgery, trauma or infarction) as well as the use of BDNF in the treatment of CNS tumors derived from astroglials. In yet another embodiment of the invention, BDNF may be used to improve regulation of NGF receptor expression and may therefore be advantageously used, prior to or concurrently with NGF, to a patient in need of such treatment.

As indicated in Section 15, below, administration of kainic acid may be used to induce increased expression of BDNF (and, to a lesser extent, expression of NGF) in neurons, including hippocampal as well as cortical neurons. It has been observed (Section 15, below) that inhibition of non-NMDA receptors blocks this kainic acid-mediated increase in BDNF expression. Thus, expression of the BDNF and BDNF-related peptides of the invention may be induced in vitro or in vivo by administration of kainic acid or related compounds or carbacol, histamine or bradykinin, or related compounds thereof, which have been found to have similar effects in vitro or in vivo or by non-NMDA glutamate receptor agonists or other drugs that increase or mimic the action of acetylcholine, histamine or bradykinin. Induction of NGF expression may also be achieved by administration of kainic acid or related molecules or non-NMDA receptor agonists.

In other embodiments of the invention, BDNF protein, fragments or derivatives may be used in conjunction with other cytokines to achieve a desired neurotrophic effect. Per

<img file="PT95152B_D0042.tif" />

493

6526-019-118

-54example (and not as a limitation¹), the BDNF prepared by the invention may be used together with NGF or with skeletal muscle extract to achieve a stimulatory effect '. synergistic .. in the growth of neurons where the term synergistic means that the effect of combining protein, peptide fragments or BDNF derivatives with a second agent achieves a greater effect than the substance used alone. It is intended that BDNF may function synergistically with other CNS-derived peptide factors not yet fully characterized in the growth, development, and survival of a large order of central nervous system neuronal subpopulations.

It is also envisaged that, based on the full characterization of the BDNF molecule, new fragments of BDNF peptides, derivatives or mutants capable of acting as antagonists of some or all of the biological functions of BDNF may be developed. These BDNF antagonists may be useful in the selective elimination of sensory neurons, e.g. eg in the treatment of chronic pain syndromes.

In still other embodiments of the invention, antibodies directed to the BDNF protein or its peptide fragments or derivatives thereof may be administered to patients suffering from various neurological disorders and diseases and in need of such treatment. For example, patients suffering from excessive BDNF production may require such treatment. Anti-BDNF antibodies can be used to prevent abnormal regeneration of sensory neurons (e.g. after surgery) or, as discussed above, in the treatment of chronic pain syndromes. In light of the high levels of BDNF mRNA found in neuroblastoma tissues, BDNF may serve as an autocrine tumor growth factor for neuroblastoma; thus, anti-BDNF antibodies can be therapeutically administered to achieve tumor regression in a specific arrangement of the invention.

<img file="PT95152B_D0043.tif" />

493

6526-019-118

-555.10. PHARMACEUTICAL COMPOSITIONS

Active compositions of the invention, which may include part or all of the BDNF gene product, including protein, peptide fragments or derivatives thereof or antibodies (or antibody fragments) directed against the protein, BDNF peptide or derivative fragments or a combination of BDNF and - a second agent such as NGF or skeletal muscle extract may be administered in any pharmaceutically biocompatible sterile vehicle, including (but not limited to) saline, buffered saline, dextrose and water.

The protein, peptide fragments or BDNF derivatives may comprise an amino acid sequence or a sequence thereof, substantially as shown in Figure 1 or Figure 5; It may be preferable to use BDNF protein which in particular comprises part or all of the amino acid sequence from about amino acid 134 to about amino acid 252, as shown in Figure 1, or a functionally equivalent sequence, as it is believed that this subsequence comprises the functional part of the do-BDNF molecule. BDNF may be derived from sequences that correspond to the BDNF genes of any suitable species including (but not limited to) the homer. ', The' pig, the rat, the chicken, the cow, the dog, the sheep, the goat, the cat, the rabbit, etc ..

The amount of protein, peptide fragments or BDNF derivatives, or antibodies that are effective in treating a particular disorder or condition depends on the nature of the disorder or condition and can be determined by classical clinical techniques. Wherever possible, it is desirable to determine the dose response curve of the pharmaceutical compositions of the invention first in vitro, e.g. ex. BDNF bioassay systems described above, and then to useful animal model systems, prior to testing them in man. Based on the in vitro results, in one specific embodiment of the invention, a pharmaceutical composition effective in promoting the survival of sensory neurons may provide a local BDNF protein concentration of between about 5 and 25 ng / ml and preferably between 10 and 10 ng. and 20 ng / ml BDNF.

<img file="PT95152B_D0044.tif" />

493

6526-019-118

In another specific embodiment of the invention, a pharmaceutical composition effective in promoting growth and survival of dopaminergic or cholinergic neurons may provide a local BDNF protein concentration of from about 10 ng / ml to 100 ng / ml.

Methods of introduction include (without limitation) intradermal, intramuscular, intraperitoneal, intravenous, subcutaneous, oral and intranasal. Further it may be desirable to introduce the pharmaceutical compositions of the invention into the central nervous system by any suitable route, including intraventricular and intrathecal injection; intraventricular injection may be facilitated by an intraventricular catheter, e.g. eg connected to a reservoir, e.g. ex. an Ommaya reservoir.

In addition it may be desirable to administer the compositions of the invention locally in the area in need of treatment; this can be achieved, e.g. ex. and not as a limitation, by local infusion during surgical operation, by injection, through a catheter or by an implant, said implant being a porous, non-porous or gelatinous material including membranes, e.g. ex. sialastic membranes or fibers.

The invention also provides pharmaceutical compositions comprising proteins, peptide fragments or BDNF derivatives administered via liposomes, microparticles or microcapsules. In various embodiments of the invention it may be useful to use such compositions to achieve delayed release of BDNF or BDNF related products.

It is intended to be possible to introduce cells actively producing BDNF, BDNF-related substances, BDNF antagonists or anti-BDNF antibodies into areas in need of increased or reduced BDNF concentrations.

6 EXAMPLE: MOLECULAR CLQNATION Ε CHARACTERIZATION

Brain-Derived NEUPOTROPHIC FACTOR ADNC

PIGS

The extreme rarity of BDNF protein in tissues has prevented the use of classical BDNF gene cloning strategies. Instead

493

6526-019-118

<img file="PT95152B_D0045.tif" />

From this, using DNA amplification technology, some protein sequence results were obtained as follows:

(i) micrograms of BDNF was purified from kilograms of porcine brain.

(ii) Purified BDNF was analyzed by protein microsequencing techniques and the amino acid sequence of a stretch of 36 amino acid residues was determined.

(iii) based on the amino acid sequence determined in (ii), oligonucleotides were synthesized which were used as polymerase chain reaction primers using the pig's upper colliculus cDNA as a standard to amplify the DNA fragment encoding DNA. defined amino acid.

(iv) the DNA reaction product from (iii) was sequenced and (v) the primer oligonucleotides were synthesized and used in the polymerase chain reaction with the pig's upper collector cDNA to create juxtaposed DNA fragments representing upstream of the BDNF mRNA as well as downstream of sequences encoding the original 36 amino acid fragment. Then the complete pig BDNF coding region was molecularly cloned into two juxtaposed fragments,

Details of each of these steps, as well as further characterization of the BDNF gene, are given below.

6.1. MATERIALS AND METHODS

6.1.1. BDNF PURIFICATION FROM BRAIN PIG

Purification of BDNF from porcine brain was performed essentially by the procedure set forth in Hofer and Barde (1988, Nature 551: 261-262).

Six kg of pig brain was homogenized with an Ultra-Turrax homogenizer in 0.2 M sodium phosphate buffer, pH 6.0, containing 1 mM EDTA, and while fresh, phenyl methanesulfonyl fluoride (1 mM) was added. ) in a proportion

<img file="PT95152B_D0046.tif" />

J

493

6526-019-118

-58 kg of brain to 2 l of buffer. The pH of the supernatant was washed to pH 4.0 with 1 N HCl and stirred for 2 h at 4 ° C. After centrifugation for 25 min at 20,000 xg, the combined supernatants (adjusted to pH 6.0 with 1 N NaOH) corresponding to 3 kg brain were shaken for 2 h with 1 l of pre-intumescent carboxymethylcellulose which had been equilibrated with 0.1 M sodium phosphate, pH 6.0. After 2 washes with a total of 20 l of 0.1 M sodium phosphate, pH 6.0, the suspension was poured into a column and washed overnight in the same buffer containing 0.13 M NaCl, The active fractions were eluted with phosphate buffer containing 0.5 M NaCl and then dialyzed against 2 x 5 1 5 mM potassium phosphate, pH 6.8. The dialyzed fractions resulting from processing 2 x 3 kg of starting material were applied to a 130 ml bed volume hydroxyapatite column which had been previously equilibrated with 5 mM potassium phosphate, pH 6.8, A The column was then eluted with a linear potassium phosphate gradient from 500 ml at 5 mM to 500 ml at 700 mM, both at pH 6.8, and BDNF activity eluted at approximately 500 mM potassium phosphate. (see Barde et al., 1982, EMBO J. 1 = 549-553). By the addition of 1.0 M potassium phosphate, the active fractions together were adjusted to a final molarity of 700 mM potassium phosphate and applied to a 5 ml column of potassium phosphate-equilibrated phenyl sepharose. mM, pH 6.8. After washing with 40 ml of the same buffer, BDNF activity was eluted with 0.1 M potassium phosphate, pH 6.8, dialyzed against lyophilized distilled water. 0 Lyophilized material was taken into SDS gel electrophoresis sample buffer containing 0.1% SDS and without mercaptoethanol as described in Barde et al. (1982, EMBO J, 1 = 549-553) before applied to an SDS + gel consisting of a linear (rather than exponential) gradient of 10 to 25¾ acrylamide. After electrophoresis separation was complete, the gel was rapidly stained (10 min) with Coomassie blue, discolored for 20 min and a band that migrated to cytochrome C level was removed and eluted from the gel by electrophoresis. The SDS was withdrawn as described by Barde et al. (1982, EMBO

J., 1. = 549-553). The purification process that eventually

493

6526-019-118

<img file="PT95152B_D0047.tif" />

would lead to the sequencing of BDNF? has been modified (according to Hofer and Barde? 1988? Nature 331: 261-262) and does not use sm gel electrophoresis? as the last step of purification.

6.1.2. PROTEIN SEQUENCING BDNF was either sequenced directly (55 pmols determined by amino acid analysis - with an initial yield of 40 pmole for amino terminal histidine) or split as follows: 5 to 10 µg EDNF (from 5 different preparations) were split according to the following processes. Joined V8: l ug S. V8 aureus (Miles) at 5 ng BDNF in 0.2 M NH4CO3? pH 8.0? containing 10% acetonitrile (total volume 50 µl) and incubated overnight at room temperature. Trypsin: 1 µg TPCK-treated trypsin (Sigma XIII-like bovine pancreas) was added to 3 ng BDNF in Tris-HCl (0.1 ii, pH 8.0) containing 10 mM CaCl2 (total volume: 40 µl) ε incubated overnight at 37 ° C. Cyanogen bromide (CNBr): 10 æg of BDNF were incubated for 3 h at room temperature (total volume: 60 æl) with 10¾ (ρ / v) CNBr in 70 ¾ (v / v) formic acid (final concentrations) '. After addition of 500 µl H2O at the end of the reaction? The sample was concentrated to 50 µl in rapid vacuum. 50 µl Tris-HCl (1.0 M, pH 8 µG) was added together with 5 µl β-mercaptoethanol and the sample was incubated overnight at 37 ° C. After addition of 5 µl of iodomethane the sample was evaporated to dryness under rapid vacuum. Reduction and alkylation of BDNF after CNBr cleavage was found: no fragments without reduction? which was revealed by HPLC and Swank Munkres gel-SDS gel electrophoresis (Swank and Munkres, 1971 final. Biochem. 59: 462-477). Does this indicate that disulfide bridges are present in BDNF? so arranged that none of the splits can give rise to free peptides. After all splits, the dried samples were resuspended in 0.1¾ trifluoroacetic acid (TFA) and applied to an HPLC? C8 microdimeter? reverse phase (Applied Biosystems) and the peptides were eluted at 0.1 ml / min using a linear gradient? 60 min? from 0 to 60¾ acetonitrile in 0.1¾ TFA. Detection was at 214 nm with a Waters 441 UV detector. The peptides were sequenced by automated degradation.

J

<img file="PT95152B_D0048.tif" />

493

6526-019-118

Edman on a gas phase microsequencer (model 470A from

Applied Biosystems) according to Hewick (1981, J. Biol. Chem.

256 = 7990-7997) and Hunkapillar (1983, Methods Enzymol. 91 = 227-236).

Detection of PTH amino acids was performed according to Lottspeich (1985,

J. Chromatogr. 326 = 321-327).

-60—

6.1.3. DNA MOLD PREPARATION

Pig genomic DNA was isolated according to the method of Hsrraann and Frischauf (1987, Methods Enzimol. 152 = 180-183).

For the preparation of cDNA, total RNA was obtained from a 6 g sample of superior pig brain colliculus. The tissue specimen was obtained, dissected and frozen in liquid nitrogen in a local slaughterhouse. Classic methods (Okayama et al., 1987, Methods Enzymol. 154 = 3-28) were used to extract RNA. 80 j4.g of total RNA were transcribed with Moloney murine leukemia virus reverse transcriptase (SRL according to manufacturer's instructions except addition of 1 ul RNasin and omission of actinomycin D) using the CGCATCCGAATTCTGCAGTTTTTTTTTTTTT primer mixture. A, C or G as terminal 3 nucleotide⁷ (oligo 3, designed to simulate a 3 'poly-A stretch containing SamHI, EcoRil and PstI recognition sites).

6.1.4. POLYMERASE CHAIN ​​REACTION

The polymerase chain reaction (PCR) was performed according to the method described in Saiki et al (1985, Science 230 = 1550-1554).

6.2. RESULTS AND DISCUSSION

6.2.1. PROTEIN MICROSQUENCHMENT RESULTS

The protein sequencing results are presented in Table IV.

<img file="PT95152B_D0049.tif" />

493

6526-019-118

-61 TABLE IV

BDNF PEPTITE SEQUENCE EXPERIMENTALLY DETERMINED

N YS Terminal

V8

CNBr

CNBr

Trypsin

Trypsin

HSDPARRGELSV

XVTAADKKTAVD

KVPVSKGQLKQYFYE

XGGTVTVLEKVP (V) (S)

GYTKEGXRGIXRGI (T) AVDMSGGTVTVLEX

ALTMDSK

The amino acids of the combined sequence VTAADKKTAVDHSGGTYTVLEKVPYSKGQLXQYFYE represent sequences encoded by the primer oligonucleotides used in PCR; see Section 6.1.2.

Four contiguous segments of amino acid sequences were determined using Edman degradation which in turn could be used to deduce the sequence of 36 amino acids representing approximately one third of the entire amino acid sequence.

6.2.2. OLIGONUCLEOTE SYNTHESIS AND USE OF PCR TO OBTAIN CODING DNA FROM AN AMINO ACID FRAGMENT

The two fully degenerate 17-mer primer oligonucleotides were chemically synthesized based on the coding sequences of the 6 amino acid segments near the amino terminus and carboxy terminus, respectively, of the 36 amino acid residue fragment described in Section

6.2.1., Supra. In particular, two separate mixtures of 17-mer oligonucleotides (corresponding to all possible codons and designated oligo 1 and oligo 2) based on the sequence AADKKT (sense) and 'KQYFYE (antisense) (underlined in Table IV) were synthesized. 150 µmol of each primer was added to 1 pg of the pig genomic DNA standard. 35 rounds of amplification were performed using the polymerase chain reaction (PCR). per manufacturer's instructions (GeneAmp ™, Perkin-Elmer Cetus). Denaturation was at 94 ° C

J

493

6526-019-118

<img file="PT95152B_D0050.tif" />

for 1 min, primer quench was at 45 ° C for 2 min and primer extension was at 72 ° C for 2 min. The 3% agarose gel (ethidium bromide stained) was cut from the predicted size DNA band (101 bp) and amplified asymmetrically as described by Innis et al. (1988, Proc. Natl. Acad. Sci. USA 85: 9456-9440) with a 100-fold excess of either oligo 1 or oligo 2.

6.2.3. CDNA FRAGMENT NUCLEOTE SEQUENCE The resulting sense and antisense cDNA fragments were sequenced by the dideoxynucleotide chain termination method (Sanger et al., 1979, Proc. Natl, Acad, Sci, USA, 72: 3918-3921) using, as primers, oligonucleotides 1 and 2 with ends labeled by²P, The observed nucleotide sequence contained only an open reading frame uninterrupted by a chain stop codon. The deduced amino acid sequence for this reading frame was completely consistent with the actual amino acid sequence determined for this porcine BDNF region.

6.2.4. CLONATION OF ALL PORCINE BDNF D0 cDNA The complete coding region of the pig BDNF was molecularly cloned into two juxtaposed segments. To obtain the 3 '(sense chain relative) portion of the BDNF cDNA, a 30-mer primer containing 21 bases of the exact BDNF chain sequence from within the region described above was synthesized to serve as a initiator of meaning. The nucleotide sequence of this primer was oligo 4, (5 ') AAACTAGTCGACQGCAGTGGACATGTCGGG (5^?) [underlining indicates the sequence corresponding to the BDNF sense chain, from position 643 to 663 in Figure 1, encoding the amino acid sequence Thr-AXa-Val-Asp-iet-Ser-Gly (the first two bases of the codon Gly)]. The nine additional nucleotides at the 5 'end of this primer were included to provide convenient splice sites for restriction endonuclease, Spel and Sal, for use in molecular cloning. A three-degenerate 31-mer primer oligonucleotide was created to be complementary to a poly-A stretch preceded by any nucleotide (T, G or C) in the

J

493

6526-019-118

<img file="PT95152B_D0051.tif" />

cDNA sense strand and containing recognition sites for the BamHI, EcoRI and PstI restriction endonuclear constructs. The sequence of this antisense (oligo 3) primer was (5 ') CGGATCCGAATTCTGCAGTTTTTTTTTTTTX (3'). where X = A, C or 'G. Synthetic primer oligonucleotides were used to amplify sequences from: superior coliiculus DNA; '. pork prepared above by PCR. Specifically, the 3 'amplified DNA was obtained using 10 µl reverse transcription reaction, 150 pmole of sense primer (oligo 4) and 150 pmole of oligo 3 as antisense primer in a PCR. Southern blot analysis was performed on the amplified DNA products and the band giving a signal of hydration with the AAGGATCCTGCA6TTGGCCTTTCGA8ACGG end-labeled oligonucleotide (oligo 5, used as the antisense primer in the 5 'reaction described below, containing recognition of BamHI and PstI) was cut-off, extracted by the glassmilk method (.impo gene

EcoRI and SalI digested TM cloned into a plasmid

Bluescript SK⁺ (Stratagene) and sequenced),

In order to obtain the remaining BDNF coding sequence (the 5 'portion), cDNA was prepared as described above and poly-A tails were added to the 5' ends using terminal deoxynucleotide transferase. . The same mixture of three 31-mer oligonucleotides each containing a fragment of 12 consecutive T residues described above was used to obtain a complementary primer for the added poly-A tails. A single 30-mer oligonucleotide primer containing 17 bases corresponding to the complementary strand of the BDNF coding sequence and BamHI and PstI restriction sndonuclease recognition sites was synthesized. The sequence of this primer was (5⁷) AAGGATCCTGCAGTTGGCCTTTCGAGACGG (3⁷) (oligo 5, supra; underlining indicates the complementary region of the sense chain of the BDNF coding sequence, from position 709 to 693 in Figure 1); this corresponds to the coding segment of the Pro (last two bases of codon) amino acid sequence -Val-Ser-Lys-Gly-GIn. Primers were added to the poly-A tail cDNA and PCR amplified as above. The

<img file="PT95152B_D0052.tif" />

493

6526-019-118

<img file="PT95152B_D0053.tif" />

Reaction products were cut with PstI and cloned into the Bluescript vector and the nucleotide sequence was determined by the dideoxynucleotide chain termination method.

6.2.5. PORCINE BDNF PO cDNA NUCLEOTE SEQUENCE The combined nucleotide sequence determined from juxtaposed pig BDNF cDNA clones is shown in Figure 1, together with the deduced amino acid sequence. The sequence comprises an open reading frame of a 252 amino acid polypeptide. Identification of the start codon Met (ATG) is based on the presence of two adjacent chain end codons (TAG-TGA) in the same open structure starting 36 base pairs upstream. The amino terminus of porcine BDNF, determined by direct sequential analysis of purified protein, corresponds to the His 134 residue of this polypeptide. Thus, sequencing results show that mature BDNF is derived from the processing of a precursor polypeptide. 0 His 134 residue is immediately preceded by the sequence Arg-Val-Arg-firg. These sequences of an amino acid residue followed by a neutral residue and then two more basic residues have been considered as target sites for proteolytic processing of precursor polypeptides. The deduced amino acid sequence of mature BDNF predicts a 119 amino acid protein (molecular weight 13 511 daltons) with a basic charge (pl - 9.99), properties that are consistent with the previous characterization of BDNF by biologically active factor evaluation, after fractionation by two-dimensional gel electrophoresis. The amino acid sequence of parts of BDNF determined by protein sequencing (a total of 64 amino acid residues) is in complete agreement with the amino acid sequence deduced from the nucleotide sequence of cDNA clones (underlined in Fig. 1). . The precursor polypeptide sequence is consistent with BDNF sm processing at least two steps: first, a signal peptide of perhaps 18 residues would be cleaved from the amino terminus, followed by cleavage between Arg 133 and His 134 to release the mature polypeptide. If this model is correct, then the

J

<img file="PT95152B_D0054.tif" />

493

6526-019-118

The precursor could be designated prepro-BDNF.

7, EXAMPLE: BDNF GENE IS DISTINCT PQ NGF GENE IN

VARIOUS SPECIES OF VERTEBRATES

7.1. MATERIALS AND METHODS

7.1.1. PREPARATION OF MGF AND BDNF DNA PROBES In British Biotechnologies Limited a plasmid containing a synthetic mature human NGF coding gene was incorporated, incorporating the normal human NGF coding sequence with some conservative codon substitutions to introduce convenient cleavage sites. restriction endonucleases. A pair of 18-mer oligonucleotide primers were synthesized to allow amplification of a 270 base pair segment of this gene, corresponding to the coding region for amino acid residues 9-11, by PCR. labeled DNA probe, a 10-cycle PCR reaction with 32p_or Similar procedures yielded a BDNF probe, except for the amplification that was performed using porcine genomic DNA as the original standard, and the amplified segment corresponded to the coding region of amino acids 28 to 111 of the mature BDNF. A complementary strand (antisense) primer corresponding to amino acid region 106 to 111 was then synthesized and had the sequence (5 ') ACATACACAGGAAGTGTC (3'). A sense strand oligonucleotide primer was prepared for amino acid residue region 28-33 [(5¹) GCAGTGGACATGTCGGGT (3 ')]. Under spot conditions in 2 x SSC at 68 ° C, Southern blot hydration (Southern, 1975, J. Mol. Biol. 98: 503-517) was performed with these two probes.

7.1.2. SEQUENCING OF DIFFERENT SPECIES BDNF The same two 18-mer oligonucleotides surrounding the coding region for amino acids 28 to 111 of the mature pig BDNF described above were used as primers to amplify, under classical PCR conditions, 252 base pair segments of the pig, rat, chicken and man genomic DNA, and the resulting DNA reaction products were sequenced by the dideoxynucleotide chain termination method (Sanger et al., 1979. Proc. Natl. Acad. Sci. USA 72: 5918J

<img file="PT95152B_D0055.tif" />

493

6526-019-118 —66—

-3921). In some cases the amplified DNA band was cut off and then extracted by agarose gel electrophoresis and reamplified prior to sequencing. In other cases reamplification was not essential.

7.2. RESULTS AND DISCUSSION

Prior to the present invention, only pig BDNF protein had been purified. It was of the utmost importance to unequivocally demonstrate that BDNF was not simply the beta subunit of the nerve nerve growth factor (beta-NGF or simply referred to herein as NGF) of the pig, which to date was not known for purification and cloning. molecular. This was especially critical since the previously described physical properties of pig BDNF are virtually identical to those of the β-NGF monomer of various species. The lack of a neutralizing antibody against pig BDNF and the previous lack of knowledge of the amino acid or nucleotide sequence of pig BDNF made it impossible to determine the exact relationship between BDNF and NGF. It was conceivable that the observed differences in biological activity between BDNF and NGF could, for example, simply reflect differences in NGF between pig and certain other species (eg mouse) or could result from differential modification of the protein protein molecule. NGF in different tissues (eg pig brain versus mouse salivary gland) or from inadvertently introduced protein modifications in some steps of purification from pig brain.

If BDNF were found to be distinctly distinct from NGF, it would be of great interest to determine whether the BDNF gene is present in other species, notably man. Prior to the present invention, no information was available about this point as BDNF was purified from only one species. The presence of apparently distinct neurotrophic activity from NGF in various crude extracts and conditioned media did not imply the existence of an identical or substantially equivalent substance to pig BDNF.

Comparison of the predicted amino acid sequence of BDNF of

J

<img file="PT95152B_D0056.tif" />

493

6526-019-118

-67 pig with known NGF sequences from certain species (human, bovine, guinea pig, mouse, chicken and snake) indicated that BDNF is significantly less related to NGF of any vertebrate than NGF to each other (Figure 2). A striking feature of the primary structure of mature BDNF is its similarity to that of NGF; With only 3 intervals introduced into the NGF sequences to optimize the comparison, 51 identities are common to the various NGFs (from snake to man) and BDNF (Figure 2). It is important to note that such identical parts include all 6 cysteine ​​residues. Although the exact arrangement of the BDNF disulfide bridges is not yet known, it is evident that such bridges are present (Table III, legend). The 3 tryptophan and 2 phenylalanine residues observed in BDNF are in identical positions in NGF. We also note that 6 aspartic acid residues (out of 7 in BDNF) and 7 inaine (out of 9) are present at identical positions in mammalian NGFs and BDNF. These 5 amino acids constitute about half of the amino acid identities between the 2 proteins. Conversely, there are some notable differences between BDNF and all NGFs: In addition to the 3 ranges mentioned, there are still 21 positions where amino acids are identical in all NGFs but different in BDNF.

Most of the BDNF precursor sequence is unrelated to that of NGF, with 2 exceptions: the supposed BDNF secretory signal sequence shows 5 amino acid identities (of the 18 amino acids) and a shocking overall relationship to the BDNF signal sequence. Mouse NGF, where splitting has been shown to occur after an alanine residue found at position 18 after translation initiation methionine (Edwards et al., 1988, Mol. Cell Biol. 8 = 2456-2464). It seems likely that alanine, which is also at position 18 in the BDNF, represents a potential cleavage site for removal of the BDNF signal sequence. The other similarity to NGF begins in the single N-glycosation consensus sequence (double underlined in Fig. 1) corresponding to asparagine 126. This asparagine is located 8 amino acids before the splice site that gives rise to mature BDNF. The same arrangement is found in several NGFs as well as the Arg-XJ

<img file="PT95152B_D0057.tif" />

493

6526-019-118

Basic-Arg of the last 4 amino acids of the precursor (Schwarz et al., 1989, J. Neurochem. 52: 1203-1209)

Evidence that NGF and BDNF are encoded by distinct genes in various vertebrate species was obtained by preparing DNA probes from molecularly cloned human NGF and porcine BDNF and by Southern blot hybridization with DNA. Genomic DNAs were digested with EcoRI restriction endonuclease. Genomic DNAs were analyzed from the following sources: man, monkey, rafo, mouse, dog, cattle, rabbit, chicken and yeast. Q DNA was digested with EcoRI and analyzed by Southern blot on duplicate filters with a β2p-labeled human NGF probe _{It's an OS}Porcine BDNF assay »Single bands were observed with each probe in all analyzed organisms except yeast» In most cases, the bands that hybridize to the NGF and BDNF probes in any organism were of different electrophoretic mobility, yet that in some cases (e.g., mouse DNA), the EcoRI fragments that hydrate with NGF and BDNF probes were approximately the same size and could not be separated under the electrophoretic conditions used (Figure 3).

A portion of the mature BDNF coding sequences were amplified by PCR from pig, rat, chicken and human genomic DNA, and the nucleotide sequences were determined. By analyzing the amplified region DNA sequence of the pig genomic DNA, the results obtained for the sequence with molecular clones from the pig brain cDNA were accurately confirmed. The rat, man and chicken BDNF genomic sequences for this 252 bp segment were also determined (Figure 5). In the rat and man the deduced amino acid sequence for the region (at least) of amino acids 28 to 111 was identical to that of pig BDNF, although some nucleotide differences (eg conservative changes in the third position of a codon) were observed between the various species. In chicken there was a single amino acid substitution in this region; residue 61 of the mature BDNF protein in chicken is lysine whereas it is methionine in the BDNFs of the

493

6526-019-118

-69 mammals. The sequence results, together with the Southern blot hybridization experiment described above, provided unequivocal proof that BDNF is encoded by a highly conserved gene, distinct from that encoding NGF.

8 EXAMPLE: BDNF DQ RNA EXPRESSION IN FABRICS

Neuronal Versus Non-Neuronal

8.1. MATERIALS AND METHODS

8.1.1, RNA PREPARATION

Total RNA was extracted from adult female mouse tissues according to (Okayama et al. (1987, Methods Enzymol. 154: 3-28). Briefly, frozen tissue was homogenized in 5.5 M guanidine thiocyanate , the lysate was centrifuged to remove impurities and the supernatant was placed on a layer of adjusted cesium trifluoroacetate to a density of 1.51 g / ml. Following a 24 hr 125 ° C xg centrifugation on a SN 27 rocker rotor (Beckmann), the RNA was resuspended and precipitated with ethanol and 8 M ammonium acetate and stored at -70 ° C. Electrophoresis was performed according to Lshrach et al (1977, Biochemistry 16: 4743-4751) on 1,3-agar-formaldehyde gels. 0 RNA was transferred to nylon membranes (Hybond-N, Amersham) and hybridized overnight at 62 ° C in 1 ml of 2 x SSC containing 50¾ formamide with a ^ 2p-labeled cDNA mouse BNF probe ( iq7 _ÇP_{m? to see} below). Washed for 60 min at_? 65°0_; in 0.1 x SSC. After washing, the spot was incubated for 60 min at room temperature with 0.1 pg / ml KNase. A (Pharmacia) and the film was exposed at -70 ° C (with intensifying filter) for 48 h.

8.1.2. CRNA Probe Preparation

A mouse brain cDNA library with 2 independent BDNF oligonucleotides was screened. Two double clones were isolated which were subcloned at the EcoRI site of an SK plasmid.⁺ (Stratagene). The nucleotide sequence corresponding to nucleotides 350 to 829 of the pig sequence was determined (see Fig. 1). In this sequence a total of only 4 amino acids in difference between mouse and pig BDNF was found, indicating a remarkable degree of conservation of this domain.

J

493

6526-019-118

-70 between pig and mouse BDNF. A single loop ARM probe was prepared using this standard and T3 polymerase (Prorrtega). The specific activity of this probe was 10θ cpm / µg.

8.2. RESULTS AND DISCUSSION

Northern blot analysis was used to assess BDNF mRNA expression in neuronal versus non-neuronal tissues. Northern blot analyzes were performed using mouse tissues allowing for faster RNA extraction processing than in pig tissues. A probe detected a signal at about 1.45 kb in the brain (Fig. 4) and spinal cord (results not shown) - Significantly, no signal was detected in any other tissue including kidneys, intestines, lungs, liver, spleen, heart and muscles (Fig. 4). Assuming that the size of the pig mRNA is similar to that of the mouse, the cDNA sequence shown in Figure 2 represents more than 80¾ of the complete mRNA sequence.

A significant observation in establishing the physiology of BDNF is that mRNA coding for this protein was found only in the central nervous system and not in 7 non-CNS tissues. BDNF mRNA was found not only in the whole brain (Fig. 4) but also in the spinal cord and upper colliculus (the sequence shown in Figure 1 is derived entirely from a superior cDNA / colliculus pattern). These results support the idea that BDNF is a target-derived neurotrophic factor and that BDNF-responsive neurons are intrinsic to the CNS or directly related to CNS structures. In fact, all neurons known so far to respond to BDNF or project into the CNS, such as the dorsal root or cranial sensory ganglia (Lindsay et al., 1985, Dev. Biol. 112 = 519-328; Davies et al ., 1986, J. Neurosci.6 = 1897-1904), or are CNS neurons like retinal ganglion cells (Johnson et al., 1986, J. Neurosci. 6 = 3031-3038). Clearly, detailed studies are needed to examine the exact distribution of BDNF synthesis sites in the CNS, but it is already clear that the distribution of BDNF mRNA is very different from that of NGF mRNA found in many non-belonging tissues.

J

493

6526-019-118

The CNS (Hseumann et al., 1984, EMBO J. 3: 3183-3189; Shelton et al., 1984, Proc. Natl. Acad, Sci. USA 81: 7951-7955).

9, EXAMPLE: MOLECULAR CLONATION AND CHARACTERIZATION OF

BDNF GENES OF MAN AND MOUSE

9.1. MATERIALS AND METHODS

9.1.1. DNA AND cDNA GENOMIC LIBRARIES

From Stratagene, an adult male retinal cDNA library was obtained from lambda-ZAPII. A human placental genomic DNA library was obtained from Clontech from EMBL3 / SP6 / T7. From Clontech, a human fetal brain cDNA library was obtained from gt11. A mouse genomic DNA library from EMBL3 / SP6 / T7 was obtained from Clontech. Both genomic libraries were prepared by partial digestion of Sau 3A restriction endonuclease genomic DNA and ligation at the Bam HI site of the vector. From Stratagene, a lambda-ZAPII rat brain cDNA library was obtained.

9,1.2, PREPARATION OF DQ BDNF DNA PROBES The 2p-labeled BDNF DNA probes were prepared using the same primer oligonucleotides described in Section

7,1.1., Supra, in PCR with human genomic DNA to amplify the coding region of human BDNF residues 28-111. In parallel, a mouse BDNF-specific probe, labeled with, was obtained using mouse genomic DNA as the standard for PCR.

9.1.3, LIBRARY SEARCH

The libraries of ijjjjjdahda were screened according to classical methods (Benton and Davis, 1977, Science 196: 180-182; Maniatis et al., 1978, Cell. 15: 687-701), hybridizing to 50% formaldehyde with sulfate. dextran and Denhardt's solution at 42 ° C. The filters were prehybridized at 42 ° C in 50¾ formamide, 5XSSCPE, 10¾ Denhardt's solution, 0.5 mg / ml salmon sperm DNA, 0.1¾ SDS and 10¾ dextran-sufate. Hybridization was performed in the same buffer except for Denhardt's solution which was 2% in salmon sperm DNA which was 0.1 mg / ml, and in the SDS and hexane sulfate which were omitted. Then the water filters were washed.

493

6526-019-118

<img file="PT95152B_D0058.tif" />

hybridization at 68 ° C. The human BDNF probe was used to screen the human genomic DNA and retinal cDNA libraries; The mouse BDNF probe was used to screen genomic DNA and mouse brain cDNA libraries. Libraries were also screened with human and rat NGF probes prepared as described in Section 7.1.1. supra.

9.2. RESULTS AND DISCUSSION

From each library at least 670,000 plates were searched. Human positive clones were considered to be those that hybridized to the human BDNF probe, but not to the human NGF probe described above. BDNF genomic clones were obtained from both human and rat libraries at a frequency consistent with representation of the BDNF gene at one copy per haploid genome. From both the mouse and human genomic libraries, approximately one million plaques were screened. Three positive from the mouse genomic library and one from the human genomic library were obtained. Positive clones from the mouse brain and retinal cDNA libraries were obtained at a frequency consistent with a gene expressed at very low levels; in mouse brain cDNA 2 positives were identified out of 670,000; in the human retinal cDNA library, one in 670,000 clones was identified. No positive clones were detected among 670,000 from a commercial cDNA library prepared from human fetus brain. Nucleotide sequencing was performed on genomic clones and human BDNF cDNAs using synthetic primer oligonucleotides representing exact sequences from within the human and rat BDNF coding region as determined in Section 7.1.2. supra, the longest human cDNA clone obtained had an insert of approximately 1.6 to 1.8 kbp and, as expected, contained the exact sequence of the human BDNF portion determined after direct amplification of the genomic DNA. as described in Section 7.2, supra. Detailed sequence analysis of this cDNA clone (Figure 5) revealed an open reading frame encoding a similar but not identical 247 amino acid polypeptide. the full length of the precursor

<img file="PT95152B_D0059.tif" />

493

6526-019-118

-73 of pig BDNF. Within the region corresponding to the mature BDNF polypeptide (eg between the His 134 codon and the chain termination codon) no differences were found in the deduced amino acid sequence; all nucleotide substitutions between pig and man were conservative with respect to coding specificity. The remainder of the so-called BDNF precursor polypeptide showed some differences in amino acid sequence between man and pig, and, remarkably, the absence in the human BDNF gene of 5 out of 6 consecutive codons. a slightly shorter polypeptide results (247 versus 252 amino acids).

Libraries prepared in the EMBL3 vector have inserts between 10 and 23 kbp of foreign DNA. The exact insert size in the human genomic BDNF clone analyzed in detail has not been determined. However, the clone contained a single approximately 4 kbp EcoRI restriction endonuclease fragment that hybridized to the labeled BDNF probe used to screen the library. This fragment is the expected size, based on Southern blot hybridization results of human genomic DNA with the porcine BDNF probe described above. Sequential analysis of the human clone was performed using synthetic oligonucleotides representing the cDNA sequence to prepare DNA synthesis from the bacteriophage DNA template. 0 Sequencing the coding sequence of the alleged human BDNF precursor was considered identical to that of the human cDNA clone, except for a single nucleotide substitution in the prepro region, corresponding to a methionine amino acid substitution (ATG) instead of Valine (GTG). ) at nucleotide 785 in Figure 5. This change may reflect a polymorphism in the human genome. As with the human NGF gene, no intermediate sequences were detected within the coding sequence of the alleged human prepro BDNF. Rat cDNA sequence information is also shown in Figure 5.

493

6526-019-118

-7410. EXAMPLE: RECOMBINANT BDNF EXPRESSION

10.1. MATERIALS AND METHODS

10,1.1. PREPARING A BDNF EXPRESSION VECTOR

The sequence corresponding to the pig prepro BDNF was obtained using the oligonucleotide primers (150 pmole each) ATAATCTAGATGACCATCCTTTTCCTT (sense)

ATAATCTAGACTATCTTCCCCTCTTAAT (ANTISENTIATED) in a PCR using 1 æg of pig mold genome (each primer has an added XbaI site). The amplification reaction was as described except at a quench temperature of 50 ° C. Following XbaI digestion, amplified DNA was ligated to the XbaI site of plasmid pCMV⁴ to give pCMV ^ -pBDNF^-(-1 indicates the direction of orientation in Figure 6 and corresponds to CQS + in Table V) and the plasmid was introduced into the XL-1 bacterium by electroporation.

10.1.2. DQ BDNF EXPRESSION IN CQS CELLS

Plasmid pCMV1-pBDNF DNA from positive clones (verified by oligo 5 hybridization, Fig. 2) was cut with both XbaI and PstI. The size of the resulting products allowed us to determine the orientation of the inserts and both plasmids were used to transfect COS cells by the calcium phosphate method (Chen et al., 1987, Mol. Cell Biol, 7 = 2745-2752), and growth medium was harvested after 24 h. BDNF activity was assessed by bioassay in the dorsal root ganglion of the chick embryo, as discussed above in detail. 10.2. RESULTS AND DISCUSSION

Sensitive neurons from the E8 chick spine were plated (6,000 per well), incubated for 24 h and counted after 24 h (Lindsay et al., 1985, Develop. Biol. 112 = 319-328). Values ​​are the means of triplicate determinations ± standard deviation. BDNF and NGF were used at 1 ng / ml, concentration at which, with either factor, maximum survival was observed. CQS + refers to cells transfected with plasmids containing the sense orientated BDNF insert; COS refers to cells transfected with plasmids containing the BDNF inserts in the reverse orientation; CQS refers to untransfected cells. At dilutions above 1 = 20, no

J

<img file="PT95152B_D0060.tif" />

493

6526-019-118

Survival beyond control with either COS conditioned or COS conditioned medium. In all experiments without NGF, a 1 ng / ml monoclonal anti-NGF antibody was used (Korsching et al., 1988, Proc, Natl. Acad. Sci. USA 80: 5513-3516).

As shown in Guideline V, only the medium obtained from COS cell cultures bearing the sense-orientated plasmid pCMV1-pBDNF was associated with survival above control values ​​of chick sensory neurons. The recombinant BDNF is therefore biologically active. Furthermore, the addition of isolated BDNF isolated from the pig brain did not significantly increase the survival of sensory neurons above the survival levels resulting from recombinant BDNF used alone, indicating that recombinant BDNF is able to saturate the receptors of sensory neurons. BDNF. Recombinant BDNF has also been found to have an additive or synergistic effect in promoting survival of chick sensory neurons when used in conjunction with NGF.

GUADRO V

SURVIVAL OF DQ PINTO SENSITIVE NEURON CULTURES

<td>COS medium (final dilution)</td><td> 1:20</td><td> 1:50</td><td> 1+200</td>

<td>COS +</td><td></td><td> 2.510+263</td><td> 2,833±171</td>

<td>WAISTBAND-</td><td> 211+16</td><td></td><td></td>

<td>CQS</td><td> 250+87</td><td></td><td></td>

<td>BDNF + COS +</td><td></td><td> 2.516+209</td><td></td>

<td>NGF + COS +</td><td></td><td> 5.770+72</td><td></td>

BDNF alone 2,718 + 424

11 EXAMPLE: OBTAINING ANTIBODIES AGAINST 0 BDNF

11.1 »MATERIALS AND METHODS

A BDNF-specific polyclonal antiserum, designated sera 4, was prepared on a NZ white rabbit by immunization with a synthetic peptide.

<img file="PT95152B_D0061.tif" />

493

6526-019-118

-7611-1-1- SUMMARY OF PEPTITES AND CONNECTION TO A SUPPORT

A peptide consisting of 34 amino acid residues, designated 85, has been synthesized by conventional procedures. It has the following amino acid sequence corresponding to 33 mature BDNF residues (residues 153 to 185 of the porcine complete prepro DNF sequence indicated Figure 1), with an additional amino-terminal cysteine ​​residue (indicated in italics) to allow binding to a support protein using m-maleimidobenzoic acid-N-hydroxy succinimide (MBS):

Cys-Val-Thr-Ala-Ala-Asp-Lys-Lys-Thr-Ala-Val-Asp-Met-Ser-Gly-Gly-Thr-Val-Thr-Val-Leu-Glu-Lys-Val-Pro Val-Ser-Lys-Gly-Gln-Leu-Lys-Gln-Tyr peptide B5 was bound to bovine serum albumin (BSA) using bis-diazobenzidine (BDB). Fresh BDB was prepared by dissolving 46 mg of benzidine-HCl (p-diaminodiphenyl-HCl, obtained from Sigma) in 9.0 mL of 0.2 N HCl. 35 mg of NaNO2 was dissolved in 1 mL of H2O and added. to the benzidine solution while stirring for 1 h at 4 ° C. 21 mg BSA was dissolved in 3.0 ml 0.16 M borate, 0.13 M NaCl, pH 9.0. Approximately 15 mg of peptide B5 was dissolved in 1.5 ml borate-NaCl buffer, pH 9.0. The peptide solution was added to the BSA solution and placed on ice. 1.0 ml of. BDB to the BSA-peptide solution and the reaction mixture was incubated with shaking for 2 h at 4 ° C; pH was controlled and maintained at 9.0 by the addition of small amounts of 0.5 M NaOH as needed. The reaction was terminated by the addition of 0.2 ml of a 1% phenol buffered solution. Excess reagents were removed by dialysis against phosphate buffered saline (PBS).

11-1-2. IMMUNIZATION

A total of six rabbits were immunized according to the following protocol:

rabbits 1 and 4 - B5 peptide linked at its C-terminus to BSA using BDB;

rabbits 2 and 3 - B5 peptide linked at its N-terminus to BSA using MBS;

493

6526-019-118

Rabbits 5 and 6 - peptide B5 mixed with nitrocellulose powder.

In all cases, the first immunization used 1 mg of immunogen (100 ng B5 / 500 µg nitrocellulose for rabbits 5 and 6) in 0.5 ml PBS plus 0.5 ml Freund's complete adjuvant. The mixture was injected subcutaneously at multiple locations in the back. The second immunization was performed three weeks later and was identical to the first except that incomplete Freund's adjuvant was used instead of Freund's complete adjuvant. Subsequent booster doses occurred at 4 to 6 week intervals. Rabbits were bled 1 week after immunization and antisera were routinely checked for binding to pure B5 peptide by enzyme linked immunoadsorbent (ELISA) evaluation.

11.1.3. D0 ANTIBODY A0 BDNF CONNECTION DETECTION

100 µg of antigen (B5 peptide) in H2O was added to the wells of a microtiter plate and allowed to dry overnight, then washed briefly with H2O and blocked with 100 µg gelatin at 1 ° for 30 min. at room temperature. The wells were washed three times with distilled water and then 100 µg of antiserum was added and allowed to incubate at 4 ° C overnight. The wells were then washed three times in PBS / 0.05% triton X-100, after which 100 µg of labeled anti-rabbit immunoassay was added. peroxidase (1/1000 dilution) and incubated at room temperature for 3 h. The wells were washed twice, 100 µg ABTS solution (10 mg ABTS (Sigma) dissolved in 10 ml was added. In 0.1 M citrate, pH 4.0 plus 10 µg H2O2) and incubated for about 2 5 minutes, until development. color. The reaction was quenched by the addition of 10 µg of 1N NaN3. Samples were diluted 1: 5 with H2O and optical density at 415 nm was measured.

11.2. RESULTS AND DISCUSSION Rabbit antiserum 4 (sera 4) showed the highest titer (Figure 7a) and was used in the following experiments. Sera 4 antibodies were partially purified by precipitate.

6526-019-118

<img file="PT95152B_D0062.tif" />

Ammonium sulfate tincture; To a portion of antiserum, an equal volume of saturated ammonium sulfate was slowly added under stirring, and the solution was stirred for a further 15 minutes, then centrifuged at 2,000 x g. The pellet was washed twice in saturated 50% ammonium sulfate, then redissolved in PBS in a volume corresponding to the original serum volume. Ammonium sulfate was removed by dialysis against various PBS changes. The dialyzed solution was divided into 1.0 ml aliquots and lyophilized using rapid vacuum. A sera 4 antibody sample was resuspended in 0.5 ml H 2 O, reacted with peptide B5 by ELISA and found to react at a dilution of 1 = 4000.

Polyclonal antibodies against a synthetic peptide (B5) corresponding to a 33 amino acid fragment of pig BDNF were prepared by immunization of rabbits as described above. Sera 4 which was shown to have the highest titer against the synthetic peptide showed reactivity with purified porcine brain BDNF by ELISA (Figure 7b). Poor immunoblot reactivity was also noted (Results not indicated). However, the antiserum was unable to block BDNF activity in a biological evaluation on sensitive neurons of the dorsal root ganglion of the chick embryo.

12 EXAMPLE = NQVQS BIOLOGICAL EFFECTS DQ BDNF

The following observations indicate that BDNF is able to (i) aid survival and induce the completely differentiated state of CNS dopaminergic neurons; (ii) assist the survival of CNS cholinergic neurons; and (iii) suppress the proliferation of astroglial cells. These biological effects of BDNF have never been described before. Because dopaminergic neurons, cholinergic neurons, and astroglial cells may each be associated with neurological diseases or disorders, BDNF may prove to be useful in treating neuropathologies involving these cell populations, w

w

493

6526-019-118

<img file="PT95152B_D0063.tif" />

12.1. MATERIALS AND METHODS

12.1.1 / METHODS OF CULTURE OF BLACK SUBSTANCE DOPAMINGER NEURONS Ventral midbrain was dissected from the brains of rat embryos, ranging in age from embryonic day 13 to embryonic day 15, Normally 2 liters were used each experience. The dissection solution had the following composition: NaCl, 136.8 mM; KCl, 2.7 mM; At₂HPQ4.7H₂Q_J 8.0 mM; KH2PO4; 1.5 mM; glucose 6 mg / ml and BSA, 0.1 mg / ml, pH 7.4 This solution was prepared and then sterilized by filtration through a 0.2 µ pore filter. »Dissection was performed under non-sterile conditions. Once the tissue was dissected from all brains, the remaining process was performed under sterile conditions. The tissue fragments were placed on a 35 mm culture dish and cut with sharp scissors. . 2 ml of nutrient medium F-12 containing 0.125% trypsin was then added and incubated at 37 ° C. At the end of this incubation period, Diiasel was added to the suspension such that the final concentration was 80 ng / ml. Another identical incubation was performed and to the tissue suspension was then added 8.0 ml growth medium consisting of Minimum Essential Medium (MEM) supplemented with 2 mM glutamine, 6 mg / ml glucose, 5 units / ml penicillin. , 5 mg / ml streptomycin and 7.5¾ fetal calf serum (FCS). The sample was centrifuged in a benchtop centrifuge at room temperature at 500 rpm for a period of 5 min. The medium was aspirated and 2 ml growth medium was added to the cell pellet. A 1 mm aperture fire-optimized pipette was used to shred the cells 8 times. The remaining tissue fragments were allowed to settle for gravity and a small aliquot of the supernatant was taken to assess cell number by counting on a hemocytometer. After cell density was determined, cells were placed in tissue culture dishes at a density of 50,000 / cm 2.

Culture plates were prepared the day before dissection. Tissue plates (24 wells, 2 cm 2 / well) were pre-coated with polyornitine (molecular weight 30,000 - 70,000

493

6526-019-118

<img file="PT95152B_D0064.tif" />

-80g / mol) at 0.5 mg / ml at room temperature for 3 h. The plates were thoroughly washed with water and then treated with mouse larninin, 5 µg / ml at room temperature for 3 minutes.

H. The plates were then washed with water as above and incubated overnight at 37 ° C in a humidified atmosphere constituted in the presence of growth medium. The following day the medium was removed from the plates and replaced with fresh growth medium.

Once the cells were placed in the culture plates, they were placed in a 37 ° C incubator with 5 ° C / 2 ° C for a period of 24 h. The culture medium was changed to a serum free formulation (SFM = serum free medium) with the following composition: a 1: 1 (v / v Basal Eagle Medium (BEM) mixture and F-12 nutrient mixture with glucose (33 mM), glutamine (2 mM), NaHCQ3 (15 mM), HEPES (10 mM), supplemented with insulin (25 µg / ml), transferrin (100 µg / ml), putrescine (60 µM), progesterone (20 nM), sodium selenite (30 nM), penicillin (5 U / ml), streptomycin (5 mg / ml) and T3 (30 nM). In some experiments purified BDNF was added to the cultures after the medium was changed to SFM on the 2nd day of culture.

Solutions for culturing neurons were prepared. dopaminogens using water taken from a Milli-Q reagent water system. Tissue culture medium formulations were obtained from Gibco Laboratories (Santa Clara, California), as well as fetal calf serum (lot NO, 43N1086) and mouse larninine. All other media components were purchased from Sigma Chemical (St, Louis, M0) and were cell culture verified. ”Poliornitine and a / DNasel were also obtained from Sigma. Trypsin was obtained from Wothington (Freehold, NJ), lot # 2. 3667. Commercial chemicals were analytical grade purchased from Baker Chemical (Phillipsburg, NJ). The BDNF used in these experiments was purified from pig brain by Dr, Y.-A. Barde, according to the process of Barde et al., 1982 (supra).

493

6526-019-118

<img file="PT95152B_D0065.tif" />

12.1.2. METHODS FOR IMMUNOCYCHEMICAL COLORING OF VENTRAL MESENCÉFALQ CROPS

Fresh fixative solutions were prepared for each experiment. For the tyrosine hydroxylase (TH) staining the fixative was 4.0% formaldehyde in Sorenson phosphate buffer. Sorenson buffer was prepared by adding a 0.2 M solution of KH2PO4 to 0.2 M Na2HP04 until the pH reached 7.3. Paraformaldehyde was then added to the solution and briefly heated to allow it to dissolve, and cooled to room temperature before use.

To initiate the process, the culture medium was removed from the culture plates by gentle suction and the appropriate fixing solution was carefully added to the plates. Incubate at room temperature for 20 min. This was followed by three washes in Sorenson phosphate buffer of 5 minutes each with gentle rotation. The cells were then incubated in a quenching solution for 30 min at room temperature with gentle rotation. The quench solution of the cultures to be stained by TH consisted of Sorenson phosphate buffer containing 2% of normal horse serum. The cultures were then incubated in permeabilization buffer at room temperature for 30 min under gentle rotation. For the cultures to be stained by TH, the Sorenson Buffer contains 0.2¾ saponin and 1.5¾ normal horse serum. After the permeabilization step, cultures were incubated in the presence of primary antibody overnight at 4 ° C. The mouse TH antibody was a mouse monoclonal antibody of the isotype IgG2a. It was used at 40.4 g / ml in a solution of 10 mM NaP04, 50 mM NaCl, 0.2¾ saponin, pH

7.5. After incubation with the primary antibody, the cultures were washed 5 times, 15 min each, in the appropriate permeabilization buffer. The cultures were then incubated with biotin conjugated secondary antibody, ie, biotinylated horse anti-mouse IgG. This incubation was performed at room temperature for 2 h under gentle rotation. Washings identical to those described above followed and cultures were incubated in the presence of a preformed peroxidase complex of the

<img file="PT95152B_D0066.tif" />

Avidinbiotinylated horseradish (ABC reagent, Vector Laboratories, Burlingame, CA) prepared according to the manufacturer's protocol. After 30 min incubation at room temperature under gentle rotation, the cultures were washed as described above. The cultures were then incubated with 55 mM Tris-Cl, pH 7.3, containing 0.5 mg / ml diaminobenzidine and 0.01% hydrogen peroxide. The reaction product was allowed to develop over

2-5 min, after which the solution was removed and the cultures were washed several times in ice cold PBS. The number of positive cells / cm 3 was then determined.².

paraformaldehyde and glutaraldehyde were obtained from Fluka Chemical. Vectastain kits contained normal serum (used as a blocking agent), biotinylated affinity purified anti-immunoglobulin, avidin DH, and biotinylated HRP-H were purchased from Vector Laboratories. Diaminobenzidine was obtained from BRL (Gaithersberg, MD),

12.1.3. METHODS USED FOR MEASURING THE³H-DOPAMINE EH VENTRAL MESENCIFALO CROPS

The making of ³H-dopamine (³H-DA) was performed according to Dal Toso et al. (1988, J. Neurosci. 8 = 733-745) with minor modifications. The plug buffer had the following composition: NaCl, 136.8 mM, KCl, 2.7 mM, Ν32ΗΡ04.7Η₂8.0 mM, 1.5 mM KH 2 PO 4, 5.0 mM glucose, 1.0 mM CaCl 2, 1.0 mM MgSO 4, 0.1 mM ascorbic acid, 0.1 mM pargillin, pH 7.4. When necessary, 5.0 µM benzotropin mesylate (BST) was added to the plug buffer.

Cells were washed once with preheated (37 ° C) plug buffer and replenished with 0.4 ml plug buffer / 2 cm² Well The cultures were then preincubated for 5 min at 37 ° C. At the end of this pre-incubation 0.1 ml³H-DA in plug buffer (250 nM, 40 Ci / nMol) such that the final concentration of ³H-DA in buffer was 50 mM. Cultures were incubated at 37 ° C for 15 min, followed by four washes at 4 ° C of 0.5 ml each with plug buffer. Two additional washes with ice-cold PBC (10 mM NaP04, 150 mM NaCl,

493

6526-019-118

<img file="PT95152B_D0067.tif" />

-83pH 7.6). After the last wash was completed, 0.2 ml NaOH, 0.5 N / 2 cm 2 well was added to the cells, and allowed to stand at room temperature for two hours. The NaOH extract was then collected and taken to the scintillation counter (Packard, LS 500TD) with 10 ml of last image scintillation fluid. The specific socket was defined as the socket that was abolished in the presence of 5 µM BZT. Typically this represented 70-90% of the total observed take.

³H-DA was obtained from NEN (Boston, MA), Ascorbate, Pargillin, BZT, and glucose were obtained from Sigma (St. Louis, M0). Ultimate scintillation fluid was purchased from Packard (Sterling, VA).

12,1.4. METHODS FOR PRODUCING CULINERGIC NEOUNCULAR D0 CULTURES

Primary basal forebrain cholinergic cell cultures were produced from rats on the 17th embryonic day. Specifically, the cholinergic neurons used in this study came from the middle septal nucleus and the Broca diagonal band nucleus. This neuronal population projects mainly in the hippocampus. Dissociated mixed cultures (neurons and glia) were produced by the following process. The septal region was dissected out of the surrounding tissue and removed from the fetal brain. Tissue pieces were then mixed, cut with scissors and treated for 20 min at 37 ° C with 12.5% ​​trypsin. Trypsin was inactivated by dilution in the plate culture medium (Dulbecco's Modified Eagle Medium (DMEM) containing 1% penicillin and streptomycin, 5% horse serum and 1% N3, a hormonal supplement). An isolated cell suspension was obtained by grinding the digested tissue fragments with a fire-enhanced pasteur pipette. Dissociated cells were counted on a hemocytometer and plated at an appropriate density in the plate culture medium. Assay ligands were pooled 5 to 6 hours after plating, and cells were grown in vitro for 10 days, with medium changes every three days.

<img file="PT95152B_D0068.tif" />

-8412.1.5. EVALUATION WITH ACETILCOLIN TRANSFERASE

After the treatment period the cells were either used in acetylcholine transferase enzyme (CAT) assays or processed by CAT immunohistochemical staining using the following protocol. The monoclonal antibody against CAT was purchased from Boehringer Mannheim Biochemical Co. and characterized elsewhere. At the end of the experimental period the cells were washed twice with DMEM. Cultures were fixed by a two step process; 50 μ.1 of 4¾ paraformaldside were added at 50 μΐ ds DMEM over a 10 min incubation period. This solution was removed and replaced with 100 μΐ of 4 paraformaldehyde and incubation continued for 30 min at room temperature. After fixation, cells were washed 3 times with phosphate buffered saline (PBS) and permeabilized by a 30 min incubation with saponin (0.5 mg / ml). Detergent was removed with 3 PBS washes and a normal 5% rabbit serum blocking solution was added for 30 min. The primary antibody was added at a dilution of 1 = 3 in normal rabbit serum after removal of the blocking solution, and the cultures were incubated overnight at 4 ° C. With a PBS wash the solution containing the primary antibody was removed. Bound immunoglobulin was detected by the Vectastain ABC method. Diaminobenzidene tetrahydrochloride (DAB) was used as the substrate for the peroxidase reaction which usually takes from 1 to 5 minutes. The reaction was stopped by washing the cultures 2 times with 0.1 M tris-HCl pH 7.2. The cultures were stored in 50 mM tris, pH 7.6, containing 0.15 M NaCl, 42¾ glycerol and 0.15¾ Zephiran (Pierce Chemical Co., Rockville, IL) at 4 ° C.

12.1.6. METHOD FOR PREPARATION OF PURIFIED ASTROGLIAL CELL CROPS

Purified glial cultures were prepared by the McCarthy and DeVellis method (McCarthy, KD, and DeVellis, J., 1980, J. Cell Biol. 85 = 890-902) from the 1-2 day postnatal rat hippocampus. The hippocampi were removed from 5 newborns and cut with scissors. The tissue pieces were then digested in 2 ml of 0.125 trypsin for 20 min at 37 ° C. The protease was

493

6526-019-118

<img file="PT95152B_D0069.tif" />

inactivated by dilution with plate medium (10¾ Gibco fetal bovine serum, DMEM, 0.5¾ penicillin (5000 mcg / ml) and 0.5¾ glutamine). An isolated cell suspension was obtained by passing the digested tissue fragments through a tapered pasteur pipette. The cells were pelleted by centrifugation at

900 rpm for 5 min, resuspended in the plate medium and then counted in a hemocytometer. The single cell suspension was divided into three flasks. 75 cm² tissue culture and the cells were allowed to grow to about 80 ° C confluence. The cells were then subcultured by a trypsinization protocol similar to that described above. Glial cells were counted and plated. plate at a density of 10,000 cells / 0.9 cm².

12.2. RESULTS

12.2.1. BDNF D0 EFFECT ON TYRQINSIN HYDRQXYLASE

PRESENT IN VENTRAL MESENCEPHALE CROPS

Immunocytochemical staining described in section 12.1.2 above was used to measure the effect of BDNF on the number of tyrosine hydroxylase (TH) positive cells (Fig. 8). On day 8, a maximum increase of more than 200% of the control value was observed in ventral midbrain cells from BDNF-stimulated cultures. A slight increase was observed very early on day 3 of culture in BDNF stimulated cultures.

12.2.2. DQ BDNF EFFECT ON DOPAMINE TAKE BY VENTRAL MESENPHALE CROPS

The making of ³H-dopamine (³H-DA) was measured according to the method of Dal Taso et al (1988, J. Neurosci. 8: 733-745) with minor modifications as described in section 12.1.3., Supra. A slight increase in dopamine uptake on day 8 of culture was observed in BDNF-stimulated ventral midbrain cultures (Fig. 9).

12.2.3. DQ BDNF EFFECT ON EXPRESSION OF ACETILCOLINE TRANSFERASE BY CHOLINERGIC PROSENCÉPHAL NEURONS

Figure 10 (a) shows the effect of BDNF on CAT positive cell number after a growth period of 12

493

6526-019-118

-86 days in vitro. A 5.9X increase in CAT cell number was observed with the addition of 100 ng / ml BDNF, with the EC50 calculated at 10 ng / ml. As a positive control, cultures were treated at 260,000 (black bars) or 150,000 (dashed bar) cells per well by the same procedure as NGF (Figure 10 (b)). The density of 260,000 cells per well corresponds to that used in the BDNF study. This increase in the number of CAT-positive cells is similar to that previously reported. 0 The potential of BDNF to act on cholinergic neurons was also examined by measuring the activity of the CAT enzyme (F. Fonnum; J. Neurochemistry, 1975, 24 = 407-409). Figure 11 shows the variations of CAT produced by BDNF treatment. In this case, an increase of 1.8X with 100 ng / ml BDNF was achieved and the EC50 was calculated at 61 ng / ml.

12.2.4. BDNF OR EGF EFFECT ON ASTROGLIAL CELL CROPS

Type II astroeites have been shown to have high affinity receptors for various neurotransmitters and neuropeptides. Thus astrocytes are capable of responding to signals of neuronal origin. For this reason and because type II astrocytes represent a cellular component in primary cultures, a possible direct effect of BDNF on glial cells has been investigated. Cells were maintained in vitro for 4 days prior to the addition of growth factors to allow them to reach about 60¾ confluence and were then treated for 42 h with EGF or BDNF. During the last 18 h incubation was present in the medium [ H] methylthymidine. The effect of 6F is indicated in Figure 12 (a). As noted above, EGF was found to be mitogenic to astrocytes. Maximum response was observed with 10 ng / ml EGF which produced a 5.2X increase in [1 H] methylthymidine uptake. Figure 12 (b) shows the effect of BDNF on [1 H] methylthymidine uptake. The BDNF response appears to be biphasic: very low doses (0.1 ng / ml) produced a slight increase in thymidine incorporation, while doses greater than 1 ng / ml BDNF inhibited [4 H] methylthymidine incorporation. 0 BDNF at a dosage of 5 ng / ml

<img file="PT95152B_D0070.tif" />

-87 produced a 24% inhibition, suggesting a decrease in glial cell proliferation rate during the treatment period.

12.3. DISCUSSION by staining in cultures of these

These in vitro experiments clearly show that BDNF maintains survival or induces the completely differentiated state in dopaminergic neurons of rat substantia nigra development, as shown to be tyrosine hydroxylase and dopamine uptake neurons from rat embryonic midbrain. these are the neurons that cause Parkinson's disease, it is highly likely that BDNF has therapeutic potential over Parkinson's disease, either by reducing neuron loss or by increasing levels of tyrosine hydroxylase (a limiting enzyme in dopamine synthesis) or possibly both.

Also, like nerve growth factor (NGF), BDNF appears to have an effect on survival of cholinergic neurons from the rat basal forebrain, as shown by increased acetylcholine transferase (CAT) staining, increased CAT activity and increased acetylcholine transferase staining in cultures of the middle septal nucleus of the rat embryo and Broca diagonal band nucleus. Thus, BDNF, alone or in conjunction with NGF, may be useful in the treatment of diseases or disorders affecting basal forebrain cholinergic neurons, including, for example, Alzheimer's disease.

13 EXAMPLE: IDENTIFICATION OF A NEW GENE OF THE FAMILY OF

BDNF / NGF GENES

The first attempt to identify new members of the NGF / BDNF gene family using degraded oligonucleotide PCR, based on conserving amino acid sequence segments between NGF and BDNF (Boxes 1-4, see section 5.8), was to determine whether pairs of these primers could be used to amplify both NGF and BDNF gene sequences from genomic DNA of various species.

493

6526-019-118

<img file="PT95152B_D0071.tif" />

This method was then used to identify a novel gene that shared homologies with NGF and BDNF in all 4 boxes of said homology.

13.1. MATERIALS AND METHODS

13.1.1. POLYMERASE CHAIN ​​REACTION

PCR was essentially performed as described in Section 6, supra.

13.2, RESULTS

13.2.1. SEQUENCE AMPLIFICATION BOTH DQ NGF CQMQ DQ BDNF FROM GENomic DNA

Degraded synthetic oligonucleotide primers corresponding to the conservation portions of the Box 1 and Box 2 amino acid sequences were synthesized between NGF and BDNF (see Section 5.8, above), and used in a PCR reaction with rat genomic DNA, by default. The exact sequences of the primers were as follows (degradation positions, mixed with two or more bases included in the oligonucleotide synthesis step, shown in parentheses; underlined italics indicate tails with multiple cleavage sites by restriction endonuclease for facilitate vector binding in the next cloning step: A = adenine, G - guanine, C = cytosine, T = thymine, N = mixture of A, G, C, T) =

Box 1 (sense), initiator 1B =

5'-GACTCGAGTCGACTCGGTGTG (C, T) GACAG (C, T) (A, G) T (C, T_sA) A6-3 '

Box 2 (antisense), initiator 2C =

LÇCAAGCTTCTAGAATTCCA (C, T) TT (N) GT (C, T) TC (A, G) (A, T) A (A, G) AA (A, G), TA (C, T) TG-3 '

300 ng of each degraded primer mixture was added to 500 ng of rat genomic DNA in 100 w.1 of the classic PCR reaction mix. 35 cycles were performed, each consisting of a 1 min incubation at 94 ° C, 2 min at 43 ° C and. 2 min at 72 ° C. The predicted size of the PCR amplification product of both the BDNF gene and the NGF gene using these primers would be 175 base pairs, including the 2 tails.

493

6526-019-118

<img file="PT95152B_D0072.tif" />

17 — mere (underlined) included for convenience in subsequent cloning steps. Electrophoresis of the reaction mixture on an 8% polyacrylamide / 5% glycerol gel gave a predicted size main strand of amplified DNA, 175 bp.

13-. 2·. 2. DETECTION OF ADDITIONAL SEQUENCES OF

BDNF / NGF PROBE IN GENERIC DNAS OF VARIOUS SPECIES

The 175 bp band was removed from the acrylamide gel by electro-elution and then amplified in a second 7-cycle PCR reaction under reaction conditions identical to the first except at dGTP, dATP, and TTP concentrations which dropped to 50 µM each and at the marker. ^ 2_The^ f_The_p ~ cJCTP used instead of unlabeled dCTP. Radiolabeled DNA product was separated from the reaction components by chromatography on a calibration column. This probe, called R1B / 2C (for rat DNA amplified from primers IB and 2C), was then used to detect complementary sequences in DNAs. genomics of several vertebrate species (results for rats, mice and chickens are shown in Figure 13) after digestion with EcoRI restriction endonuclease and nitrocellulose staining by the Southern hybridization method on EcoRI digested genomic DNAs as shown in Figure 13

The size of the EcoRI fragments of genomic DNA containing NGF and BDNF sequences was determined in parallel spot controls using radiolabeled human NGF and BDNF probes prepared by PCR from cloned genes. The positions of the genomic fragments by the EcoRI of NGF and BDNF are shown in Figure 13 in N and B, respectively. Figure 3 presents results from a similar analysis; Equivalent results are shown in Figure 3 using the human BDNF probe and the porcine BDNF probe. As stated above, each of the NGF and BDNF probes hybridized to isolated EcoRI fragments in various vertebrate genomic DNAs. For example, in mouse DNA the BDNF probe detected a band of approximately 8.8 kb, while the NGF probe detected a band of approximately 10.4 kb.

493

6526-019-118

<img file="PT95152B_D0073.tif" />

As can be seen in Figure 13, in each species tested (results shown for chicken, mouse and rat) a. R1B / 2C probe hybridized to a DNA band indistinguishable from that identified by the NGF probe as well as a band indistinguishable from that identified by the BDNF probe (in mouse, the NGF and BDNF genomic EcoRI fragments have the same electrophoretic mobility, corresponding to approximately 11 , 5-12.0 kb). This demonstrates that degraded primer oligonucleotides 1B and 2C can be used to amplify both BDNF and NGF gene sequences. It is worth noting that in some cases additional bands were observed in the Southern genomic spot hybridized to the R1B / 2C probe. For example, in the EcoRI digested mouse genomic DNA, at least two additional bands (labeled XI and X2 of approximately 19.0 kb and 1.5 kb, respectively) were observed which corresponded neither to NGF nor to BDNF. . Similarly, at least two additional bands were observed in rat BDNF hybridization (XI, X2, approximately 7.3 and 1.2 kb, respectively) and at least one in chicken DNA (X, approximately 2.6 kb). ). In some cases additional bands were also observed that are not explicitly marked in this figure. The presence of bands not attributable to NGF or BDNF suggests the possible existence of additional gene family members. Similarly, bands clearly distinct from those known to contain BDNF and NGF gene sequences were found using other sets of primer pairs and genomic DNA templates (data not shown).

13.2.3. IDENTIFICATION OF A NEW CONNECTED GENE

AO BDNF AND NGF

A specific verification of the hypothesis that new NGF and BDNF-related genes could be identified by PCR using degraded primer oligonucleotides was performed using Box 3 and Box 4 primers (see Section 5.8, above) and mouse genomic DNA, as a mold. Degraded primers were synthesized from the sequences:

493

6526-019-118

<img file="PT95152B_D0074.tif" />

Box 3 (Direction):

5³-6GGGATCCGCGGITG (T, C) (C, A) GIGGIAT (T, C, A) GA-3 '

Box 4 (antisense):

5³-TC6AATTCTAGATIC (T, G) IAT (AG) AAIC (T, 6) LCCA-3³ (G = guanine, A = adenine, C = cytosine, T = thymidine, I = inosine; mixtures of more than one base in the oligonucleotide are indicated in parentheses). Note that inosine was used at some positions corresponding to the third base (wobble = oscillation) of a codon to allow degradation of the genetic code. It is also possible to use a mixture of the conventional 4 bases of DNA instead of inosine and the results obtained have been essentially identical to those obtained with these primers.

With the Box 3 / Box 4 degraded primer pair, PCR from NGF and BDNF genomic sequences in mouse DNA would be expected to amplify a segment of approximately 90 bp. Using the above primers, PCR was performed for 4 cycles with a quench temperature of 45 ° C, followed by 31 cycles with a quench temperature of 49 ° C. The products were analyzed by gel electrophoresis and a major band of the expected size was observed. In the mouse, the NGF gene contains a HindII restriction endonuclease cleavage site in the region between Box 3 and Box 4, while the BDNF gene contains a cleavage site for the Apal enzyme in this region. Therefore, digestion of the PCR amplification product with HindII and Apal would be expected to eliminate the NGF and BDNF sequences from the main product band. However, when the PCR product was apparently completely digested with these two restriction enzymes, a digestion resistant band persisted in the amplified DNA. This suggests that in addition to the NGF and BDNF genes, at least one new gene has been amplified.

13.2.4. CHARACTERIZATION OF A MOVEMENT MEMBER OF THE BDNF / NGF GENES FAMILY

The digestion resistant PCR product was eluted in a. gel are used as a mold '. in asymmetric PCR reactions where a

<img file="PT95152B_D0075.tif" />

493

6526-019-118

<img file="PT95152B_D0076.tif" />

Any of the original degraded primers was present in 100 x molar excess over the other primer. This asymmetric amplification allowed the preparation of single-loop DNA templates suitable for sequentially by the chain termination method. Sequential analysis of the new gene (referred to herein as M3 / 4 but also referred to as Neurotrophin-3) was further continued by PCR amplification of the sequences between exactly the primer located between Boxes 3 and 4, and the end-3 poly-A sequence.⁵ of the gene transcript using the cDNA extremes rapid amplification (RACE) strategy described by MA Frohman, Μ. K. Dush, and 6. R. Martin, Proc. Natl. Acad. Know. USA, 85: 8998-9002 (1988). DNA sequencing results revealed that a new gene that comprised an open reading frame capable of encoding a distinct polypeptide from both NGF and BDNF had been amplified, but with an amino acid sequence closely related to both (Figure 14).

Preliminary evidence was obtained that this new gene encoded a neurotrophic factor by determining its expression pattern in rat tissues by Northern blot hybridization. This analysis indicated that the new gene expresses itself much more strongly in brain tissues than in any other tissue examined.

13.3. DISCUSSION

As shown above (Figure 13), the DNA probe obtained by PCR amplification using Box 1 and 2 primers with mouse genomic DNA such as mRNA (R1B / 2C) hybridized to new bands beyond those known to contain sequences. of NGF and BDNF genes in EcoRI digested DNA of all species tested. Similar analysis was conducted using a radiolabeled probe for the new gene amplified using Box 3 and 4 primers on mouse genomic DNA (M3 / 4). In each case the M3 / 4 probe hybridized to a single major genomic DNA band. digested by EcoRI, which was distinct from the bands known to contain BDNF and NGF sequences. It should be noted that the EcoRI fragments of mouse, rat and chicken genomic DNA,

493

6526-019-118

<img file="PT95152B_D0077.tif" />

observed by hybridization with the M3 / 4 probe are in all cases coincident with one of the new bands obtained by hybridization with the R1B / 2C probe. These are the 19.0 kb EcoRI fragment in mouse DNA (XI in Figure 14), the 7.3 kb fragment in mouse DNA (XI in Figure 14) and the 2.6 kb band in chicken DNA. (X in Figure 14). This suggests that portions of the same gene were amplified from mouse DNA using primer pair 1B / 2C and mouse DNA using primer pair 3/4. Thus, at least one new gene apparently shares homology with the NGF and. BDNF in all 4 homology boxes above.

The concept of homology boxes between NGF and BDNF and other members of the gene family (eg M3 / 4 also known as NT-3) has been expressed primarily in terms of the main amino acid sequence and methods for identifying new genes. of the family. It is however important to note that there are probably additional implications for the secondary and tertiary structure of these neurotrophic factors, their interactions with specific receptors, and rational design of new molecules with potential therapeutic value.

For example, there are 6 Cys residues in NGF and it has been found that they are all involved in disulfide bonds. Numbering them from the most terminal N to the most terminal C, such as Cys 1-Cys 6 disulfide bridges are known to involve Cys 1-Cys 4, Cys 2-Cys 5, Cys 3-Cys 6. The positions of all 6 Cys residues are conserved between NGF and BDNF, and the positions of 3 Cys residues in the now sequenced M3 / 4 portion align exactly with Cys 4, Cys 5, NGF Cys 6 and BDNF. This suggests that the secondary structure of all members of the gene family may be closely related and largely determined by conserved Cys residues. It should be noted that the indicated homology boxes for BDNF and NGF cover 5 of the 6 Cys residues (Cys 1 in Box 1, Cys 2 in Box 2, Cys 3 in Box 3, Cys 5 and Cys 6 in Box 4). This supports the idea that these Cys residues and their immediate neighbors play an important role in determining the overall structure of these neurotrophic factors. The structural determinants of receptor-specific interaction

<img file="PT95152B_D0078.tif" />

493 6526-019-118

-94 high affinity for each of the rieurotrophic factors presumably resides in unique portions of each molecule,

Thus, the new chimeric genes can be produced by combining between family members (e.g., in vitro recombination or direct gene synthesis) in any of the 4 homology boxes already described or somewhere in the molecule. These chimeric proteins have probably a similar secondary structure, due to the conservation of Cys residues, and other amino acid residues, but may have new biological properties. For example, a chimeric BDNF / NGF protein may be bifunctional in terms of interaction with both BDNF and NGF receptors. Chimeric proteins may also differ from related molecules in terms of dimerization and other physicochemical properties. Chimeric proteins may also be able to function as antagonists of any of the related molecules.

Active BDNF / NGF fragments can be used to prepare other family members based on knowledge of the core regions critical for appropriate involvement, as well as information on regions that are required for type-specific interaction with the receptors.

Comparison of new family members (eg M3 / 4) with known family members can be used to uncover new homology boxes that may help guide the search for other members of the BDNF / NGF gene family, For example, comparing M3 / 4 with BDNF reveals some useful homology boxes. Of particular interest is the conserved fourth Cys residue, the only one not found in Boxes 1-4 above. There is actually a rather long conserved amino acid substitution or identity segment shared by BDNF and M3 / 4 that includes this Cys residue, namely: His-Trp-Asn-Ser-Gln-Cys- (Arg or Lys) -Thr (Thr or Ser) —Gin— (Ser or Th?) —Tyr-Val — Arg-Ala-Leu — Thr. Within this region at least two homology boxes could be chosen for the synthesis of useful degraded primer oligonucleotides (e.g.

<img file="PT95152B_D0079.tif" />

493

6526-019-118

-95 ex. His-Trp-Asn-Ser-Qln-Cys requires only 96X degradation for an 18-mer primer or 48X degradation for a 17-mer; Tyr-Val-Arg-Wing-Leu-Thr would also be a useful box).

14- EXAMPLE: INCREASE OF BDNF EXPRESSION IN CELLS

NEUROBLASTOMAS

14.1, MATERIALS AND METHODS 14.1,1 »CELL LINES

CHP100, CHP126, CHP134, CHP234, LAN1, LAN5, NB9, SY5Y, Y79, FOI, BU2, H01, HL60 and C0L320 are cell lines maintained in Dr. Fred Alt's laboratory that provide RNA for the Northern blot used in Figure 15. All cell lines were human tumor cell lines. CHP100 is a neuroepithelioma cell line; CHP126, CHP134, CHP234, LAN1, LAN5, NB9, and SY5Y are neuroblastoma cell lines; Y79 is a retinoblastoma cell line; FOI, BU2, H01 are melanoma cell lines, HL60 is a promyelocytic leukemia cell line, C0L320 is a colon neuroendocrine carcinoma cell line.

14.1,2. RNA PREPARATION

RNA was prepared and Northern blots were performed essentially as described in Section 8.1.1., Supra, using a full length probe of human DNA. 10 µg total RNA was placed in each gel lane used to produce the Northern blot of Figure 15, except for the less loaded LAN1 RNA and SY5Y where 1 µg poly (A) was used.⁺ from RNA »

14.2. RESULTS

Figure 15 shows the results of Northern blot analysis of the BDNF 'probe hybridized to a panel of RNA samples obtained from various human cell lines. Abundant BDNF probe-hybridized RNA was detected in CHP234 and LAN5 cell lines, and smaller amounts in CHP126 and CHP134. All positive lines came from human neuroblastoma tumors.

<img file="PT95152B_D0080.tif" />

493

6526-019-118

-9615. EXAMPLE: REGULATIONS DEPENDENT ON THE ACTIVITY OF

PO BDNF and NGF RNAs in MOUSE HIPOCAMP is MEDIATED

BY NON-NMDA GLUTAMATE RECEPTORS

15.1. MATERIALS AND METHODS

15.1.1. TREATMENT OF CAINIC ACID RATS

Rats usually received diazepam (Valium) 90 min after intraperitoneal injection of kainic acid at 12 mg / kg to suppress intense activity due to capture. As shown in Fig. 19, Valium given after kainic acid did not interfere with further increases in BDNF and NGF mRNA levels. Rats not receiving Valium showed similar increases in hippocampal BDNF and NGF mRNA levels 3 hours after cainic acid compared to animals receiving kainic acid followed by Valium.

15.1,2. HIPOCAMP CROP PREPARATION

The hippocampi were prepared from embryos of E17 rat (days old), dissected and incubated for 20 min at 37 ° C in phosphate buffered saline (PBS) without calcium or magnesium ions but containing 10 mM glucose, 1 mg / ml albumin, 6 µg / ml DNase and 1 mg / ml papain. After washing with the papain-free solution, the hippocampal cells were carefully dissociated by a fire-enhanced pasteur pipette. Cells were harvested by low speed centrifugation, resuspended in DMEM supplemented with 10¾ fetal calf serum and plated in culture dishes (0.5 x 10 6 cells per 35 mm) previously coated with poly-DL-ornithine ( 0.5 mg / ml) and laminin (5 µg / ml). 3 h after plating, the medium was changed to a serum-free medium containing the supplements described by Brewer and Cotman (Brain Res. 497-65. 1989), but without glutamate. The neurons remained viable until three weeks of culture and were usually used in the 7th day experiments after plating.

15.1.3. D0 ARN AMPLIFICATION

Total cellular RNA was extracted according to Chomcynski and Sacci (Anal. Biochem., 1987, 162 = 156-159), from 0.5 x 1G &amp; cells after addition of a one-strand cRNA recovery standard.

<img file="PT95152B_D0081.tif" />

493

6526-019-118

-97NGF shortened - '. . (30 fg). 0 NGF mRNA and cRNA were co-amplified in a polymerase / reverse transcription (PCR / RT) chain reaction together in a tube containing 1/5 of the extracted RNA, IxRT / PCR buffer (10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl 2, 0.1 mg / ml gelatin, 0.1% Triton X-100), 0.25 mM dNTP, 0.1 μΜ of each 5 'and 3' primer, 5U RNasin (Promega ), 3.2 U AMV reverse transcriptase (Life Science) and 2 U Taq Polymerase (Genofit) in a total volume of 25 μ.1_ The mixture was covered with mineral oil, incubated at 41 ° C for 30 min, heated at 92 ° C for 60 seconds, with primer quench at 55 ° C for 60 sec and primer extension at 72 ° C for 60 sec. Amplification products (203 bp NGF mRNA at 153 bp recovery standard) were separated over 3% NuSieve / 3: 1 Agarose gel (FMC Bioproducts), Hybond N plus membrane alkali stained (Amersham) and hybridized. second (Heumann, R. and Thoenen, H., 1986, J. Biol. Chem. 261: 9246; Lindhold et al., 1988, J. Biol. Chem. 263 = 16348). For absolute quantification, in parallel reactions, known amounts of in vitro transcribed NGF mRNA and the recovery pattern were co-amplified. A similar method was recently described by Hang et al. (PNAS 86 = 9717-9721, 1989).

15.2. RESULTS AND DISCUSSION

BDNF and NGF are members of a gene family of about 50% identical amino acids (Leibrock et al., 1989, Nature 541: 149). The molecules reveal strictly conserved domains. Contained within these domains are the 6 cysteine ​​residues most likely implicated in stabilizing the three-dimensional structure of these molecules necessary for their biological activity. There are, however, also variable domains in BDNF and NGF that determine their different neuronal specificity (Lindsay et al., 1985, Dev. Biol 112 = 319; Johnson et al., 1986, J. Neurosci. 6 = 3031; Hofer, M. and Barde, Y. 1988, Nature 551: 261; Rodriguez-Tebar et al., 1989, Dev. Biol. 156 = 296). In addition the difference between these two neurotrophic molecules is also indicated by their synthesis sites. NGF is expressed both on the periphery (Korsching, S. and Thoenen, J., 1983, Proc. Natl.

J

<img file="PT95152B_D0082.tif" />

493

6526-019-118

-98EMBO

Natl. Natl,

1986, Proc, Natl. Acad, Science 254 = 552). In ftcad. Know. USA -80 = 5515; Ebendal et al., 1983, Exp. Cell Res. 148 = 311; Heumann et al., 1984, EMBO J. 3 = 3183; Dhelton, DL and Reichardt, LF, 1984, Proc. Natl. Acad. Know. USA 81 = 7951) as in the central nervous system (CNS) (Korsching et al., 1985,

J 4 = 1389; Shelton, DL and Reichardt, LF, 1986, Proc.

Acad. Know. USA 83 = 2714; Whittemote et al., 1986, Proc. ftcad Know. USA 85 = 817; Large et al., 1986, Science

254 = 352). The innervation density of NGF-responsive neurons reflects the levels of NGF in the corresponding target tissues (Korsching, S. and Thoenen, 1983., 1983, Proc. Natl. Acad. Sci. USA 80 = 3513; Ebendal et al. 1983, Exp. Cell Res, 148 = 511; Heumann et al., 1984, EMBO J. 3 = 3183; Dhelton, DL and Reichardt,

LF, 1984, Proc. Natl. Acad. Sci, USA 81 = 7951; Korsching et al., 1985, EMBO J. 4 = 1389; Shelton, DL and Reichardt, LF, 1986, Proc, Natl. Acad. Know. USA 85 = 2714; Whittemore et al., I know. USA, 83 = 817; Large et al., 1986, in contrast to NGF, BDNF is predominantly expressed in the CNS in neurons, and BDNF mRNA levels are considerably higher than NGF mRNAs, e.g. eg, about 50x in the hippocampus. In the peripheral nervous system, NGF is synthesized by various non-neuronal cell types (Bandtlow et al., 1987, EMBO J. 6 = 891), whereas in the brain it is mainly located in neurons, as demonstrated by in-situ hybridization. (Rennert, PD and Heirich, G., 1986, Biochem. Biophys. Res. Commun. 158 = 815; Ayer-LeLievre et al., 1988, Science 240 = 1559; Whittemore et al., 1988, J. Neurosci. Res .20 = 405). However, it has also been observed that type 1 culture astrocytes produce substantial amounts of NGF (Lindsay, RM,

282 = 80; Furukawa et al., 1987, Biochem. Biophys.

/

142 = 595). The relative contribution of neurons and astrocytes to synthesized NGF in the brain is not yet known.

1979, Nature Res. Commun.

In view of the predominant expression of both BDNF and NGF mRNAs in central nervous system neurons, we investigated whether levels of these two mRNAs are influenced by neuronal activity and, if so, which transmitters

J

<img file="PT95152B_D0083.tif" />

493

6526-019-118

-99 could be involved in regulation. In a first series of experiments, we prepared mouse embryonic hippocampal (E17) neuronal cultures. As shown in Figure 16a, depolarization of very potassium (50 mM) hippocampal neurons causes an increase in BDNF mRNA. Maximum levels are obtained between 3 and 6 h after increasing potassium concentration. 0 Potassium-mediated increase in BDNF mRNA could be inhibited by omission of calcium ions in the medium and was also reduced by inhibition of calcium influx by calcium channel blocker nifedipine (Fig. 16b).

Since the hippocampal cultures used consisted of a mixed population of neurons showing different transmitter and receptor expression patterns, we studied the effect of various physiological and synthetic receptor agonists on BDNF and NGF mRNA expression. The results presented in Table VI show that of all the tested substances eainic acid is a glutamate receptor agonist (Monaghan et al., 1989, Annu. Rev. Pharmacol. Toxicol. 29: 365), produced by far the largest increase in BDNF mRNA in hippocampal neurons. In contrast, other molecules such as carbacol (a muscarinic receptor agonist) and, to a lesser extent, histamine and bradykinin, slightly but significantly elevate BDNF mRNA (Table VI). Because NGF mRNA levels in hippocampal neurons are very low, we have established a suitable polymerase chain reaction (PCR) method for determining NGF mRNA variations. Using this method we observed that, as with BDNF mRNA, NGF mRNA levels were also increased by potassium and kainic acid in hippocampal neurons (Figure 17).

As shown in Figure 18, the maximum increase in BDNF mRNA in hippocampal neurons was obtained with about 25 µl kainic acid. A further increase in kainic acid concentration resulted in a decrease in BDNF mRNA levels, which most likely reflects toxic effects of high kainic acid concentrations on hippocampal neurons (Figure 18). Glutamate receptor mediated neurotoxicity

J

<img file="PT95152B_D0084.tif" />

493

6526-019-118

-100 has already been reported (Choi et al., 1987, J. Neurosci. 7 = 357; Rothman, S., M. and Olney, JW, 1987, Trends Neurosci. 10 = 299; Choi, DW, 1988, Neuron 1 = 623 ) to various central neurons after application of excitatory amino acid glutamate analogs. However, the kainic acid concentrations needed to improve BDNF and NGF mRNA expression in hippocampal neurons can clearly be. separated from concentrations that result in neurotoxicity.

To investigate in more detail the effect of kainic acid we went to see if the increase in BDNF mRNA could be blocked by any of the known glutamate receptor antagonists (Monaghan et al., 1989, Annu. Rev. Pbarmacol. Toxicol. 29 = 565). Quinurenic acid, a broad spectrum glutamate receptor antagonist, as well as CNQX, a competitive inhibitor of non-NMDA receptors (Monaghan et al., 1989, Annu. Rev. Pharmacol. Toxicol, 29 = 365). completely blocked the increase in kinic acid-mediated BDNF mRNA in hippocampal neurons (Table VII). On the other hand, ο MK801, which specifically blocks NMDA receptors, was ineffective in blocking the increase in BDNF mRNA in these neurons, In addition, NMDA itself did not alter BDNF mRNA levels (Table VI). This shows that kainic acid acts directly through its receptors and that the effect is not due to the release of endogenous glutamate (which also acts on NMDA receptors). It can thus be concluded that in vitro kainic acid elevates BDNF mRNA levels through non-NMDA glutamate receptors.

493

6526-019-118

-101-

<img file="PT95152B_D0085.tif" />

QUADR0 VI

EFFECT OF VARIOUS RECEPTOR AGONISTS ON EXPRESSION

Of BDNF mRNA in CULTURAL NEURONS

<td>Addition</td><td>% of control</td>

<td>none</td><td> 100 ± 5</td>

<td>carbacol (50 μΜ)</td><td> 220 ± 25</td>

<td>carbacol (50 μΜ) + atropine (10 μΜ)</td><td> 85 ± 15</td>

<td>nicotine (100 μΜ)</td><td> 110 ± 7</td>

<td>histamine (50 μΜ)</td><td> 150 ± 12</td>

<td>serotonin (100 μΜ)</td><td> 95 ± 6</td>

<td>dopam i na (100 μΜ)</td><td> 115 ± 7</td>

<td>norepinephrine (25 μΜ)</td><td> 75 ± 10</td>

<td>substance P (1 μΜ)</td><td> 85 ± 9</td>

<td>Somatostatin (1 μΜ)</td><td> 110 ± 8</td>

<td>bradykinin (1 μΜ)</td><td> 155 ± 15</td>

<td>cainic acid (25 μΜ)</td><td> 1365 ± 70</td>

<td>NMDA (25 μΜ)</td><td> 105 ± 10</td>

TABLE VII

EFFECT OF ANTAGONISTS FROM DIFFERENT RECEIVERS OF

GLUTAMATE ABOUT INDUCED BDNF MRI EXPRESSION

BY CAINIC ACID

Addition no kainic acid (25 μ quin) quinurenic acid (1 mM) quinurenic acid (1 mM) + kainic acid (25 CNQX (10 μΜ)

CNQX (10 μΜ) + kainic acid (25 μΜ)

MK-801 (5 μΜ)

MK-801 (5 μΜ) + kainic acid (25 μΜ)% of control

WM)

<td> 100</td><td> ±</td><td> 60</td>

<td> 1250</td><td> ±</td><td> 60</td>

<td> 70</td><td></td><td> 6</td>

<td> 84</td><td> ±</td><td> 7</td>

<td> 109</td><td> ±</td><td> 6</td>

<td> 105</td><td> ±</td><td> 7</td>

<td> 96</td><td>go</td><td> 5</td>

<td> 1130</td><td></td><td> 80</td>

493

6526-019-118

To assess the degree of physiological significance of our observations, we examined whether similar mechanisms also operated in vivo. Adult Wistar rats of both sexes, weighing from ISO to 200 g, were treated with 12 mg / kg of kainic acid. After several time periods, we determined the variations of BDNF and NGF mRNAs in the hippocampus and cortex by Northern blot analysis. Fig. 19 shows that kainic acid elicited - an increase in BDNF and NGF mRNA levels at both levels. brain regions. The increase in BDNF mRNA was substantially greater than that in NGF mRNA. In addition, even after 24 h after administration of kainic acid, mRNA levels remained elevated. 0 The course of changes in BDNF and NGF mRNAs showed that the maximum increase in hippocampus was reached approximately 3 h after administration of kainic acid (Fig. 19). Interestingly, the increase in BDNF and NGF mRNAs in the hippocampus was preceded by an increase in c-fos-mRNA; the two phenomena may be causally related, as has recently been shown in the case of the sciatic nerve following injury (Hengerer et al., 1990, Proc. Natl. Acad. Sci. USA, 87: 3899). In addition, there is a delay in the increase of both BDNF and NGF mRNAs in the cerebral cortex compared with the increase in the hippocampus, indicating that the signal leading to increased BDNF and NGF expression spreads from the hippocampus to other regions. of the brain. This repair agrees with earlier reports (Morgan et al., 1987, Science 257: 192) showing that the increase in c-fos elicitia by convulsive metrazole occurred first in the hippocampus and then in the cortex.

After previous studies have shown that NMDA receptors are implicated in the regulation of hippocampal mRNAs such as e.g. N6FI-A (Cole et al., 1989, Nature 540: 474), we then examined whether or not NMDA receptor antagonists, MK801 and ketamine, would block the increase of BDNF and NGF mRNAs in vivo. However, confirming our in vitro results, ο MKS01 did not inhibit the kainic acid-mediated increase of hippocampal BDNF mRNA, although it effectively suppressed seizures resulting from the activation of the BDNF receptor.

493 6526-019-118

-103·

<img file="PT95152B_D0086.tif" />

NMDA (Fig. 20), Most likely this NMDA receptor activation results from endogenous glutamate released by kainic acid (Biziere, K, and Coyle, T., Neurosci., 1978, Lett. 8: 303; McGeer et al. , 1978, Brain Res. 139: 381) .. Similarly, the increase in hippocampal NGF mRNA following treatment with kainic acid was not prevented by MK801 or ketamine. In an earlier study (Gall, CM and Isackson, PJ, 1989, Science 245: 758), it has been shown that electrolytic lesions causing limbic seizures elevate NGF mRNA in the rat hippocampus. The present results, obtained with ο MK801 and kainic acid, however, showed a clear dissociation between seizure induction and increased expression in BDNF and NGF. 0 Treatment of rats with diazepam (Valium) prior to kainic acid injections completely blocked the increase of BDNF and NGF mRNAs in the rat hippocampus (Figure 20). The blocking effect of diazepam on the increase of BDNF and NGF mRNA most likely results from suppression of neuronal activity via the GABAergic inhibitor system. However, 90 minutes after administration of kainic acid (ie approximately 30 min after the onset of seizures) the increase in BDNF and NGF mRNA was no longer blocked by diazepam, indicating that a relatively short period of activity - The increased rate may be sufficient to initiate cascading events leading to increased expression of the mRNAs of the two neurotrophic factors.

The present investigation has shown that non-NMDA receptors may be implicated in the regulation of BDNF and NGF mRNA in the hippocampus, thus distinguishing this regulatory mechanism from that previously described by Cole et al. (Cole et al., 1989, Nature 340 = 474) who concluded that hippocampal mRNAs encoding early genes appear to be regulated through the NMDA subtype glutamate receptors. It has been reported that some of the effects of kainic acid on neurons may also be mediated by quisqualate glutamate receptors (Ionaghan et al., 1989, Annu. Rev, Pharmacol., Toxicol. 29: 365).

In conclusion, the present study showed that neuronal activity regulates the levels of factor-encoding mRNAs.

<img file="PT95152B_D0087.tif" />

6526-019-118

493

-104neurotrophic BDNF and NGF in rat hippocampus and cortex. Of all the substances tested, kainic acid, acting via non-NMDA glutamate receptors, elevated BDNF and NGF mRNAs both in vivo and in hippocampal neuronal cultures. In contrast, in the periphery there is no evidence that NGF synthesis in non-neuronal cells is regulated by conventional transmitting substances or neuropeptides released by innervation (NGF-sensitive) neurons (Barth et al., 1984, J. Cell Biol. 99: 839; Hellweg et al., 1988, Exp. Cell Res, 179: 18). 0 BDNF is predominantly expressed in the CNS and there is at least partial overlap between central circuits, using glutamate as a neurotransmitter, and regions that are shown to be relatively rich in BDNF and NGF mRNAs (Hofer et al., EMBO J, in press; Monaghan et al., 1989, Annu. Rev. Pharmacol, Toxicol. 29: 365). In particular, there are striking similarities between the distribution of NGF mRNA, as demonstrated by in situ hybridization (Hofer et al., EMBO J., in press) and that of kainic acid receptors, visualized by hippocampal autoradiography, cerebral cortex and cerebellum (Monaghan et al., 1989, Annu, Rev, Pharmacol, Toxicol. 29: 365). The observations made are consistent with the interpretation that glutamate represents the physiological transmitter that regulates BDNF and NGF mRNAs in the CNS,

16 EXAMPLE: Q Brain Derived Neutrophic Factor

INCREASES SURVIVAL AND DIFFERENTIATED FUNCTIONS OF

Septic Cholinergic Neurons of the Rat, C Culture

16.1. MATERIALS AND METHODS

16.1.1, PREPARATION OF DISSOCIED CELLS AND CELL CULTURE CONDITIONS

The rat septal region (Sprague-Dawley) after 17 days gestation was dissected to be free of surrounding tissues. Tissue fragments were mixed, washed 3 times with Ham's F-OO and then transferred. to a 35 mm tissue culture plate where they were cut. Single cells were isolated by incubating the tissue with 0.25¾ trypsin for 20 min at 37 ° C. After trypsin inactivation by incubation at room temperature for 5

<img file="PT95152B_D0088.tif" />

493

6526-019-118

In -105min growth medium (infra) containing 50 µg / ml deoxyribonuclase type 1 (Sigma), cells were dissociated by repeatedly passing the fragments through a tapered-tip Pasteur pipette. Dissociated cells were then centrifuged at 500 xg for 45 s. The supernatant was removed and recentrifuged. The loose cell pellets were resuspended together in the growth medium and cell production was determined by the use of a hemocytometer. Finally the cells were plated in 6 mm wells that had been coated with polyornitine (10 µg / ml) and laminin (10 µg / ml). Cell viability after 24 h in culture was verified by the ability of cells to reject trypan blue.

Normal growth medium, 5HS / N3, for cultures composed of neurons and glia, contained: 5¾ (v / v) horse serum (Gibco), 1¾ N3 (v / v) additives (Romijn et al., 1982, Dev. Brain Res. 2 = 583-589), 0.5% (v / v) glutamine (200 mM, Gibco) and 0.25% (v / v) penicillin-streptomycin (10,000 units / ml, 10,000 mcg / / ml, respectively, Gibco) in Dulbecco's Modified Eagle's Medium (DMEM). Neuron-enriched cultures were prepared by replacing growth medium 5 to 6 h after plating with DMEM containing: 1 N3 additives, 0.5% glutamine and 0.25% penicillin and streptomycin. Under both conditions cytosine arabinoside treatment (1 rM for 24 h) was used to limit glial cell proliferation.

16.1.2, ACETILCQLINE TRANSFERASE ACTIVITY ASSESSMENT

The culture medium was removed by washing the cells 2 times with 100 µl PBS. Cells were lysed by a freeze-thaw cycle and a 15 min incubation at 37 ° C in 50 mM KH2PO4, pH 6.7, containing 200 mM NaCl and 0.25% (v / v) Triton X-100 . From the cell lysate 2 µs were removed and their CAT activity assessed according to the micro-Fonnum process (Fonnum, F, 1975, J. Neurochem. 24: 407-409), The final substrate composition consisted of 0.2 mM [? C] Acetyl-CoA (NEN, 54.4, mCi / mmol, 300 mM NaCl, 8 mM choline bromide, 20 mM ethylenediaminetetraacetic acid and 0.1 mM neostigmine in 50 mM NaH 2 PO 4 buffer (pH 7.4) at these concentrations 71 493 6526-019-118

-106--

<img file="PT95152B_D0089.tif" />

enzyme and substrate reactions, the enzymatic reaction was linear for 90-120 min. The specificity of acetylcholinatransferase induction was tested by the addition of a specific CAT activity inhibitor, N-hydroxyethyl-4- (1-naphthylvinyl) pyridium (HNP), during the evaluation (White, HL and Cavallito, CJ, 1970, J. Neurochem 17 = 1579-1589).

16.1.3. ACETILCOLINE ESTERASE (AChE) ACTIVITY ASSESSMENT The level of AChE present in the lysates was measured by the Potter methods (Potter, LT, 1967, J. Pharmacology ^ nd

Experimental Therapeutics 156 = 500-506) and Johnson and Russell (Johnson, CD, and Russell, RL, 1975, Anal. Biochem. 64 = 229-258)^-) with some modifications. Lysates prepared in 50 mM KH2PO4 buffer, pH 6.7, containing 200 mH NaCl and 0.25¾ (v / v) Triton X-100, were added to [α] Acetylcholine Hiodide (NEN, NET-113,

73.3 mCi / mmol), etopropazine (0.1 mM) in 50 mM KH2PO4, ρΗ 7.0. After a 15 min incubation at 37 ° C, the reaction was terminated by the addition of 100 µl of a stop buffer solution containing: 1 M chloroacetic acid, 0.5 M NaOH and 2.0 M NaCl. AChE specific activity was determined in the presence of 1.0 (10^_^) M neostigmine.

16.1.4. High Affinity Hill Taking Measurement High affinity hill turning by a high affinity Na-s-dependent mechanism was assessed essentially by the process of Vaca and Pilar, 1979 (Vaca, K. and Pilar, G., 1979, J (Gen. Physiol. 75 = 605-628). Cells were washed once with 100 μ.1 Ham F-10, followed by a wash with 100 æl Tyrodes solution. Following a 10 min depolarization induced by the 25 mM Tyrodes solution. potassium, cells were incubated for 20 min at room temperature with [1 H] choline (NEN, NET-109, 0.11 µM, 86.7 Ci / mmol) in normal Na-containing Tyrodes solution.⁺ or Li⁺. The uptake reaction was stopped by washing the cells 2 times with PBS. The accumulated choline was extracted by incubating the cultures for at least 20 min with 100 µl of cold acetone containing 1 N formic acid. The soluble fraction was then removed and counted. Choline specific uptake is the difference between the

493

6526-019-118

The total outlet level is the Na + ~ independent hill outlet.

16.1.5. HYTOCHEMICAL MARKING BY ACETILCOLINAESTERASE

Cholinergic cells were identified by. histochemical labeling by acetylcholine esterase by a modification of the Geneser-Jensen and Blackstadt labeling method (1971, Z. Zellforsch. 114 = 460-481). After fixation of paraformaldehyde ersi 4¾ cultures, cells were incubated 5-6 days at 4 ° C in the presence of the AChE substrate solution composed of the following: 4 mM acetylthiocholine iodine, 2 mM copper sulfate, 10 mM glycine and 10 wg / ml gelatin in 50 mM acetate buffer, pH 5.0. Visualization of the reaction product was obtained as previously described (Hartikka, J. and Hefti, F., 1988, J. Neurosci. 8 = 2967-2985).

16.1.6. IMMUNO-HISTOCHEMICAL MARKING BY NGF RECEIVER

The cultures were washed 2 times with DMEM and then fixed with 4¾ paraformaldehyde. The cells were then treated for 3-4 h with a pH 7.4 sodium phosphate buffer containing 5¾ bovine serum albumin, 0.02¾ triton X-100 and 5¾ sucrose (Hartikka, J. and Hefti, F., 1988). , J. Neurosci. 8 = 2967-2985). NGF receptor (NGF-R) was detected using monoclonal antibody 192-IgG (Taniuchi, M. and Johnson, E., 1985, J. Cell Biol. 101 = 1100-1106; Chandler et al., 1984, J Biol Chem.

259 = 6882-6889). at a dilution 1 = 1000 in buffer containing 5¾ horse serum. Cultures were incubated with diluted antibody for 15-18 h at 4 ° C. Bound mouse immunoglobulin was detected using biotinylated horse anti-mouse IgG (1 = 200 Vector). Immunoreactive cells were identified using 3 ', 3'-diaminobenzidine as substrate for the bound peroxidase enzyme.

16.1.7. P0 BDNF AND NGF PURIFICATION

Purification of BDNF from pig brain and NGF from male mouse submaxillary gland were conducted according to previously published protocols (Barde et al., 1982, EMBO J, 33 = 549-553; Lindsay et al., 1985, Dev. Biol. 112 = 319-328). Recombinant BDNF used for a part of these

493

6526-019-118

-108-

<img file="PT95152B_D0090.tif" />

studies, has been previously characterized (Maisonpierre et al., 1990, Science. 247: 1446-141).

16.2. RESULTS

Septic cell cultures produced the growth of elaborated neurites in polyornithine-laminin-coated wells with many cells showing bright phase one: characteristic neuronal soma at 24 h (Figure 21A and B). Continued growth of neurites resulted in a progressive increase in the extent of cell-cell contact on days 2 and 4 in vitro (compare in Figure 21 C, D with E, F). There were very few non-neuronal cells in the cultures even after 4 days, which was noted by the absence of dark phase flattened cells. To assess the effects of BDNF on the survival in culture of septal cholinergic neurons, an AChE histochemical tagging and an NGF receptor immunohistochemical tagging were used. Typical morphologies of AChE positive neurons are shown in Figure 21 6, H. Several NGF receptor positive neurons are indicated in Figure 21 I.

BDNF caused a 2.4x increase in the number of AChE positive cells in low density cultures (defined as cell densities between 66 700 and 1333 300 cells / cm2) of embryonic rat septal cells, as shown in Figure 22A . At the same cell densities the effect of NGF was slightly lower (1.9x). The discovery that both BDNF and NGF can produce an increase in the number of AChE positive neurons in septal cultures has suggested whether these two factors act on similar or distinct neuronal populations. As shown in the third panel of Figure 22A, the simultaneous addition of BDNF and NGF saturating levels does not lead to a greater increase in the number of AChE positive neurons than observed with either factor alone. In terms of neuronal survival, these results suggest that cholinergic neurons in rat septal cultures that respond to BDNF or NGF should be partially juxtaposed populations, the effect of BDNF on the survival of β-positive neurons.

493

6526-019-118

-109-

<img file="PT95152B_D0091.tif" />

AChE was dependent on the density at which cells developed (Figure 22A). At high plaque densities (200,000 - 266,000 cells / cm2) neither BDNF alone nor in combination with NGF produced - an increase in the number of neurons expressing AChE. This may be due to the high endogenous levels of neurotrophic activity. or, alternatively, the enhancing effects of survival of increased cell-cell contact occurring at higher cell densities.

At low cell density, the number of AChE positive cells increased as a function of BDNF concentration, with a maximum observed response at a dose of 10 ng / ml, which increased the number of AChE positive cells / well from control values ​​from 169.2 ± 12.9 to 388.0 ± 26.2 (an increase of 2.5x) in treated cultures. In comparison, a maximum NGF response at 50 ng / ml was observed and an increase in AChE / well positive cells from 121.0 ± 7.6 to 231.7 ± 12.9 (an increase 1.9x). The stationary phase on the response curve appeared stable for both BDNF and NGF, as there was no reduction in the number of AChE positive neurons at doses taken as 100 ng / ml.

Apparently, in low density cultures, both BDNF and NGF are able to increase the number of cells detected by AChE histochemistry. This increase may be due to the recovery of cholinergic cells that are dying in the culture due to absence of growth factor, the recruitment of precursor cells or the induction of the cholinergic marker, AChE. To try to distinguish these possibilities, delayed addition experiments (Figure 23) were performed on series of cultures maintained for a total of 12 days. One set of cultures, treated 5 to 6 h after plating, was maintained in the presence of BDNF or NGF for the entire 12 day period. A second group of cultures without growth factors were created for 5 days and then treated with BDNF or NGF in the last seven days (-5 / + 7). When comparing the two sets of cultures the response to NGF or BDNF added after 5 days was very similar to that observed when the cells grew.

493

6526-019-118

-110continuously in the presence of factors (compare the solid and dotted bars in Figure 23). Results were very different, however, in a third set of cultures where BDNF or NGF were only present in the cultures of the last 5 days (-7 / + 5). Under these conditions, neither NGF nor BDNF was able to increase number of AChE positive cells. In other experiments, exposure of septal neurons to both NGF and BDNF for a short time, 3-4 days, was found to be sufficient to produce a large increase in AChE enzyme activity and thus to detect cells by labeling. by AChE. It is thus seen that the absence of a detectable increase in the number of AChE positive neurons in the '-7 / + 5' cultures¹ it was not due to the possibility that the growth factor exposure time was insufficient to allow an induction of AChE positive cells.

(Montero, C. and Hefti, F., 1988, J. Neurosci. 8 = 2986-2999; Higgins et al., 1989, Neuron. 3: 247-256) have been shown to NGF both in vivo and in vitro. It is capable of regulating its own receptor levels measured at both mRNA and protein levels. Since BDNF shows similar effects to NGF on cholinergic phenotype induction in cell survival in septal cultures, we tested the potentiality of BDNF to regulate NGF receptor levels, quantified by the number of IgG192-positive cells. At 10 ng / ml, BDNF produced a 3x increase in the number of NGF receptor immunopositive cells, as shown in Figure 24. It is interesting to note that at higher concentrations (25-100 ng / ml) the effect of BDNF is progressively less marked. Thus, the dose response in this case was markedly different from that observed in the effect of BDNF on other parameters in these cultures, e.g. eg on the number of AChE positive cells. As a positive control, cells were treated with NGF at a concentration of 50 ng / ml which also resulted in a 3x increase in the number of positive cells. Therefore, based on the labeling of XgG-192, BDNF has an effect approximately equal to that. of NGF in improving regulation of NGF receptor expression.

493

6526-019-118

-111-

<img file="PT95152B_D0092.tif" />

In addition to the effects on cell survival, the possibility of BDNF promoting cholinergic phenotypic characteristics has been investigated. A linear increase in CAT activity was observed in septal cultures grown in the presence of increasing BDNF levels up to 50 ng / ml, where the response saturated the control values ​​at a level of 1.8x, as shown in Figure 25. Induction of CAT activity in parallel NGF-treated cultures reached a platform at 25 ng / ml, which represents a 3.4x increase in control values. Responses to either BDNF or NGF showed no significant decline at 3x concentration sufficient to produce a saturation response. Induction of CAT activity with either BDNF or NGF did not affect the level of HNP independent transferase, [N-hydroxyethyl-4- (1-naphthylvinyl) pyridium].

In light of the fact that both BDNF and NGF produce increased CAT activity in septal cultures, the response to simultaneous treatment by both factors was studied (Table VIII). For these experiments two concentrations of BDNF, 5 and 25 ng / ml, were used, which increased CAT activity by 1.4x and 2.6x respectively. When NGF (50 ng / ml), which alone increased enzyme activity by 2, Sx, was combined with BDNF, the CAT activity level reflected at least one additive effect (Table VIII). At the concentrations of BDNF used, a purely additive effect to that of NGF would have led to increases of 4.2 and 5.4x, respectively, relative to the control values. The observed values ​​were actually slightly more than additives,

<img file="PT95152B_D0093.tif" />

493

6526-019-118

<img file="PT95152B_D0094.tif" />

-112CHAPTER VIII

BDNF AND NGF CO-APITION EFFECTS ON CAT ACTIVITY

<td></td><td>CAT activity</td><td></td>

<td>Concentration (ng / ml)</td><td>(pmol / h / well)</td><td>% control</td>

<td>control</td><td> 561,3 ± 34,8</td><td></td>

<td>NGF (50)</td><td> 1546,3 ± 89,7</td><td> 280</td>

<td>BDNF (5)</td><td> 768,0 ± 25,4</td><td> 140</td>

<td>BDNF (25)</td><td> 1470,1 ± 66,2</td><td> 260</td>

<td>NGF (50) + BDNF (5)</td><td> 2767,6 ± 177,2</td><td> 490</td>

<td>NGF (50) + BDNF (25)</td><td> 3742,0 ± 669,3</td><td> 670</td>

Septic cells were grown over a 12 day period at 5HÕ / N3 in the presence of the indicated concentrations of the trophic factors. CAT activity was determined as described. Values ​​represent means ± SEM of 4-5 determinations.

The possibility of the biological response observed with BDNF treatment was due to a BDNF-dependent release of endogenous NGF. To examine this issue, a monoclonal NGF? Antibody (antibody 27/21, Korsching and Thoenen, 1983) was used to block any NGF-related response. As shown in Table IX, the effect of NGF on CAT activity was greatly reduced by anti-NGF, while the BDNF-induced response was not essentially affected. Note, however, that the basic level of CAT activity was reduced by about 20% in anti-NGF-treated cultures alone as compared to untreated controls. Nevertheless, the observed increase in CAT activity in septal cultures produced by BDNF appears to have a direct effect and not mediated by an increase in endogenous NGF.

<img file="PT95152B_D0095.tif" />

<td colspan="4">TABLE IX</td>

<td colspan="3">EFFECT OF ANTI-NGF ON NGF CAPACITY</td><td>And BDNF</td>

<td>Induce</td><td>D0 ENZYME ACTIVITY</td><td>CAT</td><td></td>

<td>CAT activity</td><td></td><td></td><td></td>

<td>Concentration (ng / ml)</td><td>(pmol / h / well)</td><td> ¾</td><td>control</td>

<td>Control</td><td> 517,53+ 3,2</td><td></td><td></td>

<td>BDNF (25)</td><td> 1021,6 ±106,1</td><td></td><td> 197</td>

<td>NGF (50)</td><td> 1139,0 ± 99,1</td><td></td><td> 220</td>

<td>Control + Anti-NGF</td><td> 418,7 ± 27,0</td><td></td><td> 81</td>

<td>BDNF + Anti-NGF</td><td> 819,2 + 68,8</td><td></td><td> 158</td>

<td>NGF + Anti-NGF</td><td> 551,0 ± 27,0</td><td></td><td> 107</td>

After a 12-day growth period in 5HS / N3, in the presence of the indicated trophic factor concentrations, the septal cells (plated at a density of 2.3 (10⁵) cells / cm 2) were collected. CAT activity was determined as described in the Experimental Process section. The values ​​represent the mean ± SEM of 6 determinations,

The possibility of AChE activity was also co-regulated with CAT activity by BDNF or NGF (Figure 26). In contrast to CAT activity, the dose of AChE response to BDNF or NGF- was quite similar as enzyme activity increased linearly to a concentration of 50 ng / ml. At this concentration NGF and BDNF increased AChE 274¾ and 234¾ activity respectively, when compared to untreated control values.

The course of the BDNF or NGF-induced increase in CAT activity was investigated (Figure 27). CAT activity in cultures treated with 50 ng / ml BDNF increased to 170¾ from control values ​​over 3 days. This increase continued until day 6, when it stabilized the control values ​​at 2.5x over the remainder of the test period. In contrast, CAT's response to NGF administration (50 ng / ml) was faster, reaching

493

6526-019-118

-114-

<img file="PT95152B_D0096.tif" />

the induced level 2.5x that of the control on day 3. The increase in CAT activity continued to rise linearly, reaching 3.2x those of controls after 12 days.

Astrocytes, grown in culture for some time, have been shown to synthesize various neurotrophic activities, including NGF (Lindsay, RM, 1979, Nature 282 = 80-82; Lindsay et al., 1982, Brain Res. 243 = 329-543 Alderson et al., 1989, Dev. Brain, Res. 48 = 229-241). Although we have excluded the possibility that our observations on BDNF were mediated through an increase in NGF levels, it is conceivable that BDNF may act indirectly by modulating the expression of other neurotrophic factors, especially in glial cells. the possibility of non-neuronal cells influencing the effect of BDNF on CAT activity in septal cultures, We compared the response of cholinergic neurons to BDNF in glial-neuronal and mixed cultures. in neuron-enriched cultures (Figure 28). In neuron-enriched cultures, the response of CAT activity to BDNF showed a bell-shaped curve; A maximum increase at a dose of 5 ng / ml BDNF was achieved, and at a concentration of 25 ng / ml there was a significant decline in enzyme activity (p&gt; 0.001 compared with 15 to 25 ng / ml BDNF). Similar to Figure 25, the CAT response to BDNF in the presence of a confluent glial layer was linear, in this case up to 25 ng / ml BDNF, the maximum dose used. In the presence of non-neuronal cells higher levels of BDNF were required to produce an equivalent increase in CAT activity compared to those obtained in similarly treated neuron enriched cultures. Still, the ability of BDNF to stimulate to the fullest. CAT activity in septal cholinergic neurons was essentially the same in the presence or absence of glial cells.

Based on CAT and AChE activity, it can be deduced that BDNF can act similarly to NGF in improving cholinergic phenotypic markers. To further examine this issue, the effect of BDNF on choline uptake level was compared. in

493

6526-019-118

-115-

<img file="PT95152B_D0097.tif" />

high affinity, Na⁺-dependent, with that of NGF (Figure 29). Cells grown for 12 days in the presence of 50 ng / ml BDNF accumulated choline to a level 3.8x that of untreated controls. In parallel cultures, NGF (25 ng / ml) induced a 2.3x increase in choline accumulation.

16.3. DUSCUSSION

Compared to their already demonstrated effects on peripheral neurons (Barde et al., 1982, EMBO J. 16549-553; Lindsay et al., 1985, Dev. Biol. 112: 319-328; Davies et al., 1986 J. Neurosci 6: 1897-1904; Hofer, M, and Barde, Y., 1988, Nature 331: 262-262), the neuronal specificity of BDNF within the central nervous system is less well characterized. To date, the only known sensitive central nervous system neurons constitute a small subpopulation of Thy-1 positive retinal ganglion cells first identified in E17 rat retinal cultures (Johnson et al., 1986, J. 6: 3031-3038), Other studies have also demonstrated the effects of BDNF on adult retinal ganglion cells in explant cultures (Thanos et al., 1989, Eur. J. Neuro. 1: 19-26) . In this study we demonstrated that BDNF enhances the survival of cultured cholinergic neurons from the embryonic rat septum, as shown by histochemical labeling by AChE, and enhances expression of various cholinergic phenotypic markers, ie the enzymatic activity of AChE and CAT and high affinity choline uptake. In addition, BDNF has been found to increase NGF receptor expression in septal cultures.

In the early experiments, we found that BDNF increased the number of AChE-positive neurons in E17 septal cultures by approximately 2x. As in the case of NGF (Hartikka, J. and Hefti, F., 1988, J. Neurosci 8: 2967-2985) this effect was most noticeable at low cell densities. It is of interest to note that the addition of NGF and BDNF was ineffective in further increasing the number of AChE positive cells at levels observed in the presence of either growth factor. This observation strongly suggests that BDNF and NGF act largely on the same population of cholinergic septal neurons.

493 6526-019-118

-116-

<img file="PT95152B_D0098.tif" />

As was initially said for NGF (Hefti et al., 1985, Neuroscience 1: 55-68), BDNF did not improve survival of cholinergic neurons in established cultures at relatively high cell densities. In proportion to the cells plated, survival of chrysergic neurons was found to be higher at high cell density than at low cell density even in the absence of. any neurotrophic factor. There are several possible explanations for this case including increased cell-cell contact, beneficial effects of increased number of non-neuronal cells or a more than proportional increase in endogenous neurotrophic activity. With regard to the latter possibility, we found no indication of substantially increased levels of endogenous NGF since the addition of an excess of anti-NGF to untreated high density cultures only reduced CAT base activity by 20%. In these cultures, exogenous NGF was able to produce a large 3.4x increase in CAT activity.

There are now many reports, both on protein and mRNA levels, indicating that NGF can improve the regulation of its own receptor expression in both cultured and in vivo neurons. In this study, we used the IgG-192 monoclonal antibody to detect the NGF receptor. This monoclonal has already been characterized in both rat sensory neurons and rat brain (Taniuchi, M. and Johnson, E., 1985, J. Cell Biol. 101: 1100-1106), In addition to confirming Hartikka and Hefti's findings that NGF can improve the regulation of its receptor level in septal cultures, as noted by the increase in the number of IgG 192 immunopositive neurons, we also found BDNF increases the number of IgG-192 immunopositive cells to about 3x. The response to BDNF was biphasic because at high concentrations, in the order of 25-100 ng / ml, it failed to produce as large an increase in the number of positive cells as 10 ng / ml. Interestingly, this type of dose response was quite different from that observed for the BDNF-induced increase in the number of positive neurons.

493

6526-019-118

-117 AChE. In the latter case no reduction in the maximum effect of BDNF at high concentrations was observed. It should be noted that it has recently been shown that the low affinity NGF receptor also binds to similar affinity BDNF, suggesting that the low affinity NGF and BDNF receptors may be identical (Rodriguez-Tebar et al., 1990, Binding of Brain-Derived Neurotrophic Factor to the Nerve Growth Factor Receptor Neuron. 4, 487-492),

Possible similarities in the mechanism of action of BDNF and NGF were suggested by the discovery that BDNF was equipotent with NGF in inducing AChE activity. Although BDNF was also found to promote the survival of cholinergic neurons of similar intensity to NGF, the effect of BDNF on the enzymatic activity of CAT was not as great as that of NGF. Over several experiments, the maximum induction of CAT by BDNF ranged from 1.8 to 2.6x whereas values ​​greater than 3x were routinely observed with NGF. Thus, while there are some similarities in the mechanisms of action of BDNF and NGF, it is quite likely that there are subtle differences in the expression and regulation of their respective receptors and in the binding of these receptors to the secondary messenger pathways necessary to induce CAT activity.

The assumption that differences in the induction of CAT activity produced by BDNF or NGF could represent different regulatory pathways was further supported by the results of studies of the phenomena over time. Culture septal neurons responded to BDNF with a rise in CAT activity until day 6, at which time the response was shown to stabilize. In contrast, the induction of CAT activity in response to NGF showed a prolonged linear increase from the 3rd to the 12th day, and stopping the rise in the induction of CAT activity on the 6th day in BDNF-treated cultures may be indicative of a altering the development of expression of either the BDNF receptor or a given component in the response path. Given that the number of cholinergic neurons (AChE positive neurons) obtained in high density cultures after 12 days is

<img file="PT95152B_D0099.tif" />

493

6526-019-118

Regardless of exogenous BDNF or NGF, it seems unlikely that different types of neuronal cell death could explain the differences observed over time in the induction of CAT activity by these two growth factors.

In early experiments involving BDNF treatment of cholinergic cells of the basal forebrain, no hypothesis was made as to the type of primary, neuronal or glial cell that responded to this neurotrophic factor. Although studies on highly enriched peripheral neuron cultures strongly suggest that BDNF has its effects directly on neurons (Lindsay et al., 1985, Dev. Biol. 112 = 319-328; Lindsay, 1988, J. Neurosci. 7 = 2394-2405), it is conceivable that the neuronal effects of BDNF observed in the present study may have been mediated through a first BDNF action on astroglia or other non-neuronal cells. However, BDNF was found to be equally effective in inducing CAT activity in the presence of a confluent monolayer of predominantly astroglial cells or in highly depleted glial cell cultures by treatment with the mitotic inhibitor cytosine arabinoside. An intriguing observation in these experiments was that the shape of the BDNF dose response curves was distinctly different in both cases. In the presence of a glial monolayer, BDNF response was linear as a function of increasing BDNF concentration. In neuron enriched cultures, the BDNF dose response was shifted to the left. One possible explanation for these results is that the glia themselves can express the BDNF receptor, thereby lowering the effective concentration of ligand available to the receptor located on the neurons. . Alternatively, the suppressive effect of astrocytes could be related to an effect on concentration on receptor expression, and mediated, e.g. by the greatly increased degradation of BDNF. Similar observations were made by Hartikka and Hefti (Hartikka, J and Hefti, F., 1988, J. Neurosci, 8 = 2967-2985) and Honegger and Lenoir (Honegger, P. and Lenoir, D., 1982, Dev. Brain Res, 3 = 229-238). Indirect cells, not ligand or

<img file="PT95152B_D0100.tif" />

493

6526-019-118

-119glials of different brain regions can greatly influence the morphology of neurons associated with them (Prochiantz et al., 1979, Proc. Natl. Acad. Sci. USft. 76: 5387-5391; Prochiantz et al., 1981, Nature 295 : 570-572), Thus, septal glia may exhibit intrinsic properties of either membrane or secretory nature which may act to suppress cholinergic phenotype expression. The regulatory properties of hippocampal astronites under similar test conditions are now being investigated.

Based on the results presented, it can be deduced that both BDNF and NGF are capable of increasing septal cholinergic neuronal function. It is interesting to speculate as to the possible physiological relevance of the two highly homologous neurotrophic factors apparently acting on a single neuronal population. There have been some suggestions that the actions of BDNF and NGF may initially overlap during early development of peripheral neurons, especially over subpopulations of dorsal root ganglion neurons or their precursors, and then later separate to act. independently over distinct populations (Lindsay et al., 1985, Dev. Biol. 112: 319-328; Ersberger, U. and Rohrer, Η., 1988, Dev. Biol. 126: 420-432); Barde, YA 1989, Neuron. 2: 1525-1534) However, this has not been definitively proved by the use of appropriate markers to define subpopulations of sensory neurons. Since the present study focused on cultured embryonic septal neurons at a single point in development time, E17, it is possible that a similar species of segregative phenomena could occur at later stages of septal development. temporal and spatial effects on the responsiveness of septal cholinergic neurons to BDNF or NGF.

It would now be interesting to examine the effects of BDNF on in vivo cholinergic neurons of the basal forebrain since there is now evidence that hippocampal NGF may be implicated in the normal development and maintenance of in vivo cholinergic neurons of the basal forebrain. It would be interesting

<img file="PT95152B_D0101.tif" />

493

6526-019-118

-120demonstrate whether or not BDNF plays a similar role in vivo. The recent finding that BDNF mRNA is particularly abundant in the hippocampus supports the existence of this role.

17, EXAMPLE 'ENZYMATIC CONVERSION OF PREPRO-BDNF IN

MATURE AND ACTIVE BDNF

17.1. MATERIALS AND METHODS

Commercially available Arg-C endoproteinase (purified from mouse submaxillary glands) was used to enzymatically convert the large precursor form of hBDNF (prepro BDNF) to biologically active mature protein. This enzymatic cleavage was achieved in an in vitro reaction. The enzyme reaction substrate was preproBDNF synthesized in CH0-DG44 cells, which was secreted into the cell culture medium and collected. The culture medium consisted of Ham (nucleoside-free) F12 medium with 1¾ fetal bovine serum (FBS) and 1¾ each penicillin and streptomycin. The reaction was carried out at 37 ° C for 5 min using 5 enzyme units and 50 U.1 CH0-DG44 cell supernatant containing preprohBDNF. The split of preprohBDNF into mature hBDNF was complete under these conditions.

17.2, RESULTS AND DISCUSSION

The prepro form of recombinant hBDNF (approximately 31,000 dalton) was completely processed into the mature bioactive form of hBDNF (approximately 12,000 dalton) in vitro using Arg-C endoproteinase. Recombinant human brain derived neurotrophic factor has already been successfully produced in mammalian cells (CH0-DG44 cells). 0 The hBDNF gene was stably integrated into the CHO cell genome and the hBDNF gene amplified using the methotrexate amplification strategy. Recombinant hBDNF was secreted into the medium and found to be biologically active. Metabolic labeling followed by SDS polyacrylamide gel electrophoresis (15¾ gel) demonstrated that most of the synthesized, [Î ± S S] -marked product, secreted into the culture medium, migrated as prepro form with apparent molecular weight. of 31,000 (Figure 30, lane 2).

<img file="PT95152B_D0102.tif" />

493

6526-019-118

CH0-DG44 cell breakdown are shown in Figure 30, lane 1. In order to predominantly create a mature form of hBDNF (12,000 apparent molecular weight), we have devised an in vitro strategy to enzymatically cleave the prepro form of hBDNF.

from da

Trypsin digestion products are shown in Figure 30, lanes 3-6. Bovine pancreatic trypsin (purchased from Worthington; chymotrypsin-free) was dissolved in phosphate-buffered saline and added to 50 µl aliquots of [^ ¾] labeled CHO cell hBDNF. The trypsin concentrations used were 25, 50, 75 and 100 µg / ml. The reaction products labeled with [^⁵S] are indicated in lanes 3-6, respectively. The enzymatic fission reaction was performed at 37 ° C for 5 min. A gradual decrease in labeling intensity of prepro-hBDNF 31,000 was observed with a corresponding increase of a product of 18,500 molecular weight. The mature form of hBDNF (molecular weight 12,000) was not created under these conditions. Figure 30, lane 7, shows reaction products labeled with after enzymatic digestion of cell hBDNF

CHO with Arg-C endoproteinase isolated from mouse submaxillary gland, from Boehringer Manheim Biochemicals, Inc ,. 100 units of this enzyme were reconstituted from a lyophilized preparation with 100 μ.1 Milli-Q water and stored at -20 ° C. For enzymatic digestion of CHO cell hBDNF we used 5 units (5 w.1) of enzyme and 50 μ.1 of protein, labeled with preprohBDNF-containing CHO cell supernatant (Figure 30, lane 2). As shown in Figure 30, lane 7, 5 Arg-C endoproteinase units were able to convert the prepro form of hBDNF (molecular weight 31,000) to the mature form of hBDNF (12,000 molecular weight) in 5 min at 37 ° C. ° C.

Enzymatic treatment with Arg-C endoproteinase of CHO cell supernatants expressing hBDNF increased the level of BDNF biological activity relative to unprocessed supernatants. CHB-DG44 cell supernatants expressing hBDNF were treated for 5 min with Arg-C endoproteinase at 37 ° C. 200 µl of supernatant was treated with 20 enzyme units and both treated and untreated hBDNF

493

6526-019-118

-122-

<img file="PT95152B_D0103.tif" />

were evaluated for their biological activity using E8 chick dorsal root ganglia explants. 24 h after the addition of hBDNF, neurite growth was qualitatively evaluated. We found that endoproteinase-treated BDNF samples were significantly more bioactive than control (untreated) hBDNF samples. In Figure 31 a concentration curve of these results was plotted.

In conclusion, purified Arg-C endoproteinase can be used in an in vitro enzymatic reaction to create the mature bioactive form of recombinant human brain neurotrophic factor from incompletely processed (i.e. proBDNF) precursor molecules. will enable efficient large-scale production of bioactive human BDNF from mammalian cell culture systems. it may be possible to immobilize the Arg-C endoproteinase in a solid matrix and establish an efficient column chromatography step.

18- EXAMPLE: BRAIN-DERIVED NEUROTROPHIC FACTOR PROMOTES SURVIVAL AND MATURATION OF DOPAMINERGIC NEURONS

OF THE VENTRAL MOUSE MESENCÉPHAL

18.1. MATERIALS AND METHODS

18.1.1. CELL CROPS

Dissociated cultures by enzymatic and mechanical dissociation of dissected ventral midbrain tissue from mouse embryos E14-E15 were established. Mixed tissues from 2 or 3 L of rat embryos from Sprague-Dawley mice living together were trypsinized (0.125%; Worthington) in F12 medium (Gibco) for 20 min._;at 37 ° C. After washing in growth medium (MEM containing the following supplements: 2 mM glutamine, 6 mg / ml glucose, 0.5 U / ml penicillin 6, 5 wg / ml streptomycin, 7.5% fetal calf serum) tissue It was briefly centrifuged at low speed for 5 min and the pellet was dissociated by grinding. After allowing 1-2 min to settle for the undispersed cell pellets, the single cell suspension was seeded in 35 mm plates (pre-coated with polyornitine and laminin; Lindsay et al., 1985, Dev. Biol. 112 :

<img file="PT95152B_D0104.tif" />

493

6526-019-118

-123: 319) containing growth medium to give a density of 5 x 10 10 cells per cm 2. After an overnight incubation in growth medium to allow cell attachment, cells were cultured in the presence or absence of BDNF in a defined serum-free medium as described by Bottenstein and Sato (Bottenstein and Sato, 1979). , Proc, Natl, Acad, Sci, USA 76 = 514), except for insulin which was included at 20 ng / ml. To visualize the dopaminergic cells the cultures were fixed with 4¾ paraformaldehyde, washed thoroughly, were permeabilized with 0.02% Saponin in Sorensen buffer with 1.5% horse serum and labeled with a mouse monoclonal antibody against TH (Soehringer-Manheim) rat, antibody binding was visualized using a ABC Vectastain kit (Vector Labs).

18,1.2, MEASUREMENT OF PE-DOPAMINE TAKE Cultures were prepared as described in 18.1.1 above and grown for the number of days indicated, both in the presence and absence of porcine BDNF at 50 ng / ml. Dopamine uptake was determined in triplicate at the indicated intervals. Cells were pre-washed in uptake buffer of the following composition: 136.8 mM NaCl, 2.7 mM KCl, 8 mM Na2HP04-7H20, 1_;5th mM KH2 PO4, 5 mM glucose, 1 mM CaCl3, 1 mM MgSO4, 0.1 mM ascorbate s 0.1 mM pargillin, pH 7.4, Samples to be measured for non-neuronal H2 H-dopamine uptake had benzotropin (BZT) in the take-up buffer at a concentration of 5 µM. After washing, the cells were preheated to 37 ° C for 5 min in fresh take-up buffer, at which time H-dopamine (NEN) was added. 40 Ci / mmol) to a final concentration of 50 nM. The cultures were incubated at 37 ° C for 15 min, the withdrawal solution was removed, the cells were placed on ice and washed 4 times with ice cold outlet buffer. The cells were harvested by addition of NaOH. 0.5 N to culture plates and NaOH extract was counted in a Beckman scintillation counter with 15 ml Ultima Gold scintillation fluid (Packard Instruments). Specific neuronal dopamine uptake is defined as the uptake (cpm) observed in the absence from BZT, observed in the presence of BZT. In the routine, it was found that from 70 to

<img file="PT95152B_D0105.tif" />

<img file="PT95152B_D0106.tif" />

493

6526-019-118

-12490¾ of the total plug was inhibited by BZT. In replicate cultures at each time interval, the number of TH + neurons was determined by immunocytochemical labeling and the results reported represent normalized uptake on a TH + basis.

18.1.3. BDNF TRANSITORY EXPRESSION

COS M5 cells were transfected with the CDM8 vector containing the human BDNF coding sequence. At 72 hours after transfection, 50 ml of culture supernatant were collected and dialyzed against 6 M urea. 0 dialyzed supernatant was then combined with ampholins (pH 3.5-1.0, BioRad) for 4% h. Fractions were collected, dialyzed against 25 mM NaPO 4 buffer, pH 7.6 using standard pre-washed BSA dialysis (0.5 mg / ml) to avoid unspecified absorption. Fractions were evaluated for their neurite growth promoting activity in E8 chick dorsal root ganglion explant cultures. The active fractions were mixed. Analysis of the active mixture by SDS PAGE and subsequent silver staining (Wray et al., 1981, Anal. Biochem. 118: 197) on 8-18% gel, revealed a 68 kD faded band corresponding to BSA and a single band. at about 12 kD, corresponding to that previously observed for porcine BDNF (Barde et al., 1982, EMBO J. 1: 549) (results not indicated). In comparative bioassays performed on chick dorsal, knotted, and sympathetic root ganglion explants, the specific activity and neuronal specificity of purified recombinant human BDNF were similar to those of purified porcine BDNF.

18.1.4. HBDNF PRODUCTION FROM TRANSFORMED CHO CELLS CHQ-DG44 cells were a lovely gift from Dr. L. Chasin (Columbia University, NY). These cells are deficient in dihydrofolate reductase (dhfr-) (Urlaub and Chasin, 1980, Proc. Natl. Acad. Sci. USA 77: 4216-4220). CH0-DG44 cells were maintained in Ham's F12 medium with 10¾ fetal bovine serum, 1¾ penicillin and streptomycin and 2 mM glutamine. The cells were subcultured twice a week and routinely examined for mycoplasmic contamination. All plasmid DNA was purified by double association with cesium chloride prior to transfection.

6526-019-118

-125-

<img file="PT95152B_D0107.tif" />

dog. The human brain-derived neurotrophic factor gene was subcloned into the mammalian expression vector pCDMS to obtain pCShB as previously described (Maisonpierre et al., 1990, Science 247 = 1446-1451). A weakened dihydrofolate gene -reductase, p410, was kindly offered by Dr. L. Chasin. These two constructed genes were co-transfected into CH0-DG44 cells by electroporation (reviewed by Shigekawa and Dower, 1988, BioTechniques 6: 742-751). Approximately 1 x 10 4 were transfected.^is cells with 20 µg pCShB and 0.2 µg p410. 48 hours after transfection, CH0-DG44 cells were divided into a nucleoside-free Ham F12 selection medium (without thymidine and hypoxanthine; -HT) plus 10¾ dialysed fetal bovine serum and 1¾ penicillin and streptomycin. Individual colonies were isolated approximately 10 days after cell placement in nucleoside-free selection medium. Colonies individually chosen to be resistant to growth in nucleoside-free Ham F12 (-HT) medium were treated with 0.02 or 0.05 or 0.1 µM methotrexate (MTX) to induce gene amplification ( Alt et al., 1978, J. Biol. Chem. 253: 1357-1370). MTX resistant colonies were grown in mixtures and then amplified with 1.5, or 2.0 '- or 2.5 µM MTX. A single 2.5 µM MTX resistant colony was isolated and chosen for a culture for progressive growth and hBDNF production. Southern blot analysis of this clone (DGZ1000-B-3-2.5) indicated approximately 50-fold amplification of the BDNF gene against wild-type CH0-DG44 cells. Bioassays with embryonic dorsal root ganglia (DRG) explants (E8) were used to estimate that this clone produced approximately 9.5 æg hBDNF / ml (results not indicated). Large-scale recombinant hBDNF was produced by cultivating D6Z1000-B-3-2.5 cells in 480 cm 2 rolling bottles. The medium was harvested approximately 4 days after these cells reached confluence. Purification of hBDNF from culture supernatant was performed as above for COS supernatant.

<img file="PT95152B_D0108.tif" />

6526-019-118

493

-12618.2. RESULTS AND DISCUSSION

The characterization of molecular signals that control neuronal survival and specificity of target innervation is of fundamental interest, not only because it helps to elucidate how the nervous system is shaped, but also because it is likely to lead to a better understanding of the mechanisms involved. in maintaining and repairing mature neurons in appropriate synaptic configurations. Although it is widely recognized that survival and maintenance of neurons depends on specific neurotrophic factors derived from target cells (Levi-Montalcini and Angeletti, 1968, Physiol. Rev. 48: 534; Thoenen and Barde, 1980, Physiol. Rev. , 60: 1284; Purves, D., 1988, in Body and Brain, Harvard University Press, Cambridge, MA; Barde, YA, 1989, Neuron, 2: 1525; Lindsay, RM 1988, The Making of the Nervous System, Oxford University Press, Oxford, p. 148) very few of these molecules have been fully characterized. So far, only two neurotrophic factors, nerve growth factor (NGF) and brain derived neurotrophic factor (BDNF), have been shown to selectively support the survival of distinct neuronal populations in vivo (Levi-Montalcini and Angeletti, 1968, Physiol, Rev, 48: 534; Barde, YA, 1989, Neuron, 2 = 1525; Leibrock et al., 1989, Nature 341 = 149).

Neuronal cell culture has been widely used to establish bioassays to identify neurotrophic activity in cell and tissue extracts (Levi-Monfâlcini and Angeletti, 1968, Physiol, Rev. 48 = 534; Thoenen and Barde, 1980, Physiol, Rev , 60 = 1284; Lindsay, RM, 1988, The Making of the Nervous System, Oxford University Press, Oxford, 148, Varon and Adler, 1981, Adv. Cell, Neurobiol., 2 = 118) and to determine the neuronal specificity of purified neurotrophic factors (Ebendal, T., 1989, in Nerve Growth Factors, John Wiley, London, p. 81; Barbin et al., 1984, J, Neurochem 43: 1468; Barde et al., 1982, EMBO J. 1: 549; Davies et al., 1986, J, Neurosci, 6 = 1897; Lindsay et al., 1985, Dev. Biol. 112 = 319) . To date, most studies have focused on the neurotrophic requirements of different classes of neurons.

<img file="PT95152B_D0109.tif" />

<img file="PT95152B_D0110.tif" />

493

6526-019-118

Of the peripheral nervous system (PNS). The high availability of NGF has made it - by far the best-characterized neurotrophic factor: NGF has been shown to be essential, both in vitro and in vivo, for the survival and maintenance of rustonal subpopulations of either SNP or CNS (Levi-Montalcini. and Angeletti, 1968, Phisiol, Rev. 48: 534; Thoenen and Barde, 1980, Physiol. Rev., 60 = 1284; Nhittemore and Seiger, 1987, Brain Res. Rev. 12 = 439; Snider and Johnson, 1989, Ann Neurol 26 = 489; Hefti et al., 1989, Neurobiology of Aging 10 = 515; Martinez et al., 1987, Brain Res. 412 = 295; Takei et al., 1988, J. Neurochera. 51 = 1118; Hartikka and Hefti, 1988, J. Neurosci. 8 = 2967), The greatest difficulty in identifying CNS specific neurotrophic growth factors for neurons has been the reduced availability of purified preparations of neurotrophic factors other than NGF, and the difficulty in obtaining homogeneous preparations of specific neuronal populations. . Two recent observations led us to study the effects of BDNF on the survival of CNS dopaminergic (nigral) neurons. Firstly, it is now evident that the BDNF gene is expressed in the CNS and at levels comparable to or higher than the NGF gene (Leibrock et al., 1989, Nature 341 = 149). Secondly, it has been reported that the partially purified protein obtained from bovine striatum (a target of nigral dopaminergic neurons), with characteristics apparently similar to those of BDNF, may increase in vitro survival of dopaminergic neurons prepared from the midbrain (Dal Toso et al., 1988, J. Neurosci 8 = 733).

Figure 32A illustrates the normal appearance of dissociated ventral midbrain cells after nine days in culture.- Virtually all cells have bright phase pericaria and long outgrowth. Some cells with fibroblast morphology were evident and no astroglial cells were detected when cultures were labeled with astroglial marker antibodies, the glial fibrillar acidic protein (GFAP; Bignami et al., 1972 Brain Res. 43 = 429), To visualize dopaminergic neurons in this culture, immunocytochemical

J

<img file="PT95152B_D0111.tif" />

493

6526-019-118

-128 monoclonal against TH. As expected, the number of TH + neurons was small (arrows, Fig. 328, 32C) and varied, depending on the culture time, between 0.1 and 0.5¾ of the cells in the plate. Several neuronal morphologies were noted between TH + cells, as shown in Figure 33.

In all cultures, there was invariably a gradual decline in the number of TH + cells (Figs. 34A, 35) after the first 3-4 days in vitro. yarns 8 days of culture, e.g. For example, the number of TH + cells in control cultures was only 25% of that found in similar cultures on day 2 (Fig. 35). However, in BDNF-treated cultures, the loss of TH + cells was very small compared to of controls. Every day after 8 °, in vitro, the number of TH-s- cells in BDNF-treated cultures was higher than untreated controls, e.g. 1.8x higher after a single addition of BDNF (Fig. 34A) and 5x higher after multiple additions (Fig. 35). Even when added only once to cultures maintained for 8 days, the effect of BDNF was observed to be dose dependent (Fig. 34B). In accordance with reports (Dal Toso et al., 1988, J. Neurosci 8: 733; Knusel et al. 1990, J. Neuroscience 10: 558), NGF (50 ng / ml) had no effect on cell number (Fig. 34C),

As a new method of examining the effect of BDNF on dopaminergic neurons in these cultures, the ability of TH + cells to take H-dopamine was studied. As shown in Table X, dopamine uptake capacity (normalized by TH + neurons) increased markedly in the first 8 days in culture but then declined. However, it was observed that BDNF did not alter the ability of TH + cells to take dopamine, as similar values ​​were obtained over time in the presence or absence of BDNF (Table X). From this result, we hypothesized that BDNF would probably act directly in promoting the survival of dopaminergic neurons in midbrain cultures, as opposed to inducing expression of a dopaminergic phenotype in neurons that do not necessarily require BDNF to survive. For the best

493

6526-019-118

-129-

<img file="PT95152B_D0112.tif" />

To examine this subject, we performed experiments in which the number of TH + neurons was determined in cultures where the addition of BDNF was initially delayed by several days. We found that when the addition of BDNF was delayed until day 5 and the number of TH + cells was subsequently determined on days 6, 8 or 10, fewer dopaminergic cells were seen in these cultures than in parallel cultures where BDNF had been present. day 2 onwards (Fig. 36). Although delayed addition of BDNF markedly increased the number of TH + neurons compared to controls, the effect of delayed addition, when examined even at equivalent times (in time), was never such as to save as many TH + neurons as the addition of BDNF on day 2.

Taken together, these results testify to the possibility that BDNF acts only to increase TH gene expression in a fixed number of cells, and suggest that BDNF exerts its action by increasing the survival of dopaminergic neurons that would otherwise be lost if neurotrophic factor had not been added in the early stages of the process,

Several reports have documented the stimulation of maturation and improved in vitro survival of mesencephalic dopaminergic neurons by either target-derived extracts or various growth factors (Dal Toso et al., 1988, J, Neurosci, 8: 733 = Knusel et Al, 1990, J. Neurosci, 10 = 558; Prochiantz et al., 1979, Proc. Natl, Acad. Sci, USA 76 = 5387 = DiPorzio et al., 1980, Nature 288 = 370; Prochiantz et al., 1981 , Nature 293 = 570; Denis-Donini et al., 1983, J. Neurosci. 3 = 2292). However, there has been no report on a fully purified molecule that directly and selectively supports the survival of TH + neurons in the complete absence of glial cells or division-enhanced cells (as noted, e.g., with Ig1 or FGF, Dal Toso et al., 1988, J. In accordance with previous reports using early embryonic stage mouse tissue (Dal Toso et al., 1988, J, Neurosci 8 = 733), it was found that E14 rat ventral midbrain cells in without serum were essentially free of astrocytes or fibroblasts.

J

<img file="PT95152B_D0113.tif" />

493

6526-019-118

-130 relatively low cells plated, 50,000 cells / cm², used in this study diminished the effects of any possible endogenous neurotrophic activity and allowed a more accurate cell count. In view of the results presented herein, it is shown that neurotrophic factor, partially purified from bovine striatum by Del Toso et al. (Dal Toso et al., 1988, J. Neurosci 8 = 733), is probably BDNF. This suggests that BDNF is produced in the striatum and carried by the terminals of dopaminergic neurons that protrude from the substantia nigra. So far, however, BDNF mRNA (or even mRNA from other members of the neurotrophin family) has not yet been observed to exist in abundance in striatum tissue by Northern blot analysis.

It is evident that the addition of BDNF, even in the case of multiple additions, does not result in the complete survival of nigral dopaminergic neurons during the analyzed culture periods. This phenomenon had already been observed, under experimental conditions very similar to ours, by Dal Toso et al. (Dal Toso et al., 1988, J, Neurosci 8 = 733) which showed that this loss could be related to the cell density used; at higher cell densities (4 times as used herein) a greater number of catecholamine fluorescent cells could be seen and the spontaneous disappearance of these cells would be decreased, cell-cell contact is possibly necessary for long term survival, of these cells.

Further studies, particularly those examining the effects of BDNF on the development and maintenance of ventral midbrain dopaminergic neurons in vivo, will be needed to determine the physiological relevance of the effects of BDNF described herein. In view of the specific loss of nigral dopaminergic neurons in Parkinson's disease, it is particularly desirable to obtain a neurotrophic factor that selectively affects these cells. Thus, the present finding that BDNF supports the survival of these neurons, cells that are refractory to NGF, provides rationale for animal experiments that will determine whether BDNF can protect dopaminergic neurons.

J

<img file="PT95152B_D0114.tif" />

493

6526-019-118

Neurotoxic effects of either 6-hydroxydopamine or MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine). These studies may lead to a therapeutic approach to Parkinsonism.

TABLE X

DQPAMINE TAKING IN VENTRAL MESENCIFALO CULTURES

NOT DIRECTLY INCREASED BY BDNF

Outlet ³H-dopamine

<td></td><td>(cpm / TH / (t)</td><td>neuron / 15 minutes)</td>

<td>Culture Days</td><td>Control</td><td>BDNF Treated</td>

<td> 3</td><td> 0,51 ± 0,1</td><td> 0,95 ± 0,2</td>

<td> 6</td><td> 4,5 ± 0,3</td><td> 3,1 ± 0,6</td>

<td> 8</td><td> 18,5 ± 5,3</td><td> 27,3 ±1,2</td>

<td> 10</td><td> 3,37 ± 0,24</td><td> 2,21 ± 0,18</td>

<td>19. EXAMPLE:</td><td>0 BDNF PRODUCES · ONE</td><td>INHIBIT, DEPENDENT</td>

DOSE GAMMA-AMINQUTYRIC ACID OUTLET

19.1. MATERIALS AND METHODS

19.1.1. HIPOCAMP CELL CROPS

The hippocampes of Sprague-Dawley rat E18-19 embryos were dissected and collected on FIO medium. Tissues were cut, washed 2x with FIO medium (Gibco) and trypsinized with 0.25% trypsin (Gibco) for 20 min. At 37 ° C. Trypsin was inactivated by the addition of serum-containing medium composed of minimal essential medium (MEM) supplemented with fetal calf serum (FCS, 10%), glutamine (2 mM), penicillin (25 U / ml) and streptomycin. (25 wg / ml) Dissociated cells obtained by gentle trituration were harvested and centrifuged at low speed (500 rpm) for 30 s. Centrifugation was repeated 2x and the cell pellets were then resuspended in serum containing medium. The cells were then plated, 6 mm wells or 35 mm plates that had been coated with polyornitine (10 wg / ml) and laminin (10 wg / ml). In most experiments, cells were plated at a low density of approximately 71,000 cells / cm². Five to six hours after plating cells, the medium was changed to a

493

6526-019-118

-132-

<img file="PT95152B_D0115.tif" />

serum-free medium containing 1¾N3 and penicillin-streptomycin (2.5 units / ml and 25 ng / ml, respectively), at which time BDNF was added. The medium was changed every 3 to 4 days with factor addition.

19.1.2. HIGH AFFINITY MEASUREMENT OF GAMMA-AMUTURIC ACID (GABA)

High affinity GABA uptake was measured using a modified process by Tomozawa and Appel (1986, Brain Res. 399 = 111-124). Cells were washed in GABA plug buffer containing 140 mM NaCl, 2.6 mM KCl, 1 mM &amp; 1 mM H2PO4

Ν &amp; 2ΗΡ04.6 mg / ml glucose, 1 mM MgCl2, 1 mM CaCl2, 0.1¾ BSA.

After washing, cells were incubated with GABA uptake buffer for 5 min at 37 ° C. (NEN, NET-191X, 111.4 Ci / mmol) was then added to a final concentration of 12 nM and incubated at 37 ° C for 10 min. The cells were then kept on ice and washed 3x with plug buffer. The cells were incubated with 0.14 N NaOH for 2 h at room temperature and the extract of H2 H-GABA was counted. It has been found that³H-GABA was linear up to at least 30 min. GABA uptake by non-neuronal cells was inhibited by the addition of 2 mM β-alanine, while neuron-specific uptake was inhibited by 1 mM nipecopic acid.

19.2. RESULTS AND DISCUSSION

We have recently detected abundant levels of BDNF message in neuron-enriched hippocampal cultures but not in hippocampal astrocytes. These results strongly suggest that BDNF is located in hippocampal neurons. To examine the effect of BDNF on cultured hippocampal neurons, cells with various concentrations of BDNF (rotorphor purified from COS supernatants) were treated. At the end of an 8-day treatment period, high affinity uptake of GABA by the neurons was measured. As shown in Figure 37, BDNF produced a dose dependent inhibition of GABA uptake. Thus, BDNF may affect survival and / or phenotypic expression of GABAergic neurons. BDNF had no effect on glutamate containing neurons (verified by medi71 493

6526-019-118

-133-

<img file="PT95152B_D0116.tif" />

glutamate uptake) nor on neurofilament protein levels. Thus, the effect of BDNF on hippocampal cultures shows ssr specific to GABAergic neurons.

20 »EXAMPLE: BDNF GIVES A PROTECTIVE EFFECT AGAINST

The Toxic Effects of MPP +. 20.1. MATERIALS AND METHODS

20.1.1. MEASUREMENT OF EFFECTS OF NEURQTROPHIC FACTORS ON SH-SY5Y CELLS HANDLED WITH MPP +

SH-SY5Y cells were seeded in 24-well plates at a density of 1 x 10 4 cells per well. Twenty-four hours after sowing, cells were treated with neurotrophic factors. Twenty-four hours after treatment with neurotrophic factors, cells were treated with MPP + (10 μΜ). Forty-eight hours after the addition of MPP +, the number of viable cells was quantified by the trypan blue exclusion assay. Neurotrophic factors used were mNGF-COS (dilution 1 = 5), hBDNF-COS (1: 5 dilution), nCNTF purified from E. coli (25 ng / ml), purified bFGF from bovine brain (25 ng / ml). ml), rNT3-CQS (1: 5 dilution) and purified hEGF (25 ng / ml).

20.1.2. Measurement of the effects of neurochemical factors

ABOUT VENTRAL MESENCIFALO CROPS TREATED WITH MPP +

Dissociated ventral midbrain neurons cultures were established from rat E14 embryos according to the above protocols (See Section 18, supra). 48 h after establishment of these cultures, they were treated with appropriate neurotrophic agents as if indicated. Neurotrophic factors included hBDNF (rotophor purified from CHO cells, &lt;50 ng / ml) purified mNGF (50 ng / ml)) rNT-3 (from COS supernatants, 1:50 dilution), bFGF, purified bovine brain (10 ng / ml) and purified rCNTF (25 ng / ml) (from E. coli). Twenty-four hours after the addition of neurotrophic factor, the cultures were treated with MPP + (1 μΜ). Forty-eight hours after MPP + treatment, immunohistochemistry was identified and TH positive neurons were quantified. Results represent the average of three independent experiments performed with triplicate samples per experiment. 0

<img file="PT95152B_D0117.tif" />

<img file="PT95152B_D0118.tif" />

493

6526-019-118

The number of positive neurons following MPP + treatment is represented by dashed bars. Light bars represent the number of TH positive neurons without MPP + treatment,

20.2. RESULTS AND DISCUSSION

When BDNF, NGF, CNTF, NT-3, bFGF and EGF were tested separately for their ability to protect SH-SY5Y cells from 1-methyl-4-phenylpyridinium (MPP +) toxicity by the By trypan blue exclusion, only BDNF and NGF showed significantly protective activity against MPP + toxicity (Tables XI and XII).

TABLE XI

PROTECTION BY BDNF AND NGF. SH-SY5Y CELL AGAINST

MPP + TOXICITY

Results: NQ. of viable cells

Treatment: (¾ viable)

<td>COS-transfected by simulation (1 = 5)</td><td> 0,4</td><td>X</td><td> 10⁴</td><td> (5¾)</td>

<td>SDNF-COS (1: 5)</td><td> 1,2</td><td>X</td><td> 10⁵</td><td> (67¾)</td>

<td>NGF-C03 (1 = 5)</td><td> 1,7</td><td>X</td><td>1G^s</td><td> (84¾)</td>

<td>CNTF</td><td> 0,6</td><td>X</td><td>IO⁴</td><td> (14¾)</td>

<td>NT-3</td><td> 0,7</td><td>X</td><td> 10⁴</td><td> (20¾)</td>

<td>bFGF</td><td> 0,4</td><td>X</td><td> 10⁴</td><td> (11¾)</td>

<td>EGF</td><td> 0,1</td><td>X</td><td> 10⁴</td><td> (3¾)</td>

493

6526-019-118

-135-

<img file="PT95152B_D0119.tif" />

TABLE XII

BDNF AND NGF CONCENTRATIONS CURVE IN EXPERIENCES

COS-MOCK Treatment (1 = 2) (1 = 5) (1 = 10) (1 = 20) (1 = 50) (1 = 100)

TOXICITY BY MPP-8¾ viable cells

3

COS-BDNF (1 = 2) 73 (1 = 5) 67 (1 = 10) 65 (1 = 20) 42 (1 = 50) 30 (1 = 100) 25

C0S-N6F (1 = 2) 38 (1 = 5) 84 (1 = 10) 63 (1 = 20) 50 (1 = 50) 47 (1 = 100) 34

In vsntral midbrain cultures, the results indicate that BDNF shows a significant protective effect against MPP + toxicity, while bFGF has a minor effect (Figure 38). MPP + treatment has been found to reduce the number of tyrosine hydroxylase positive neurons by 85¾ in control cultures but only by 40¾ in BDNF pretreated cultures prior to exposure to MPP +. 0 MPP + is the alleged active metabolite of MPTP toxin that has been observed to induce a Parkinson's disease-like syndrome, which is an accepted model system of Parkinson's disease. Thus, the protective effect of BDNF and NT-3 indicate that these compounds may prove to be useful.

<img file="PT95152B_D0120.tif" />

<img file="PT95152B_D0121.tif" />

493

6526-019-118

-136 in the treatment of Parkinson's disease or to prevent neurological damage following toxic exposure.

21 Deposit of Microorganisms

The following DNA-. recombinant plasmid and recombinant bacteriophage were deposited on August 30, 1989 in the American Type Culture Collection, 12301 Parklawn Drive, Rockville, Maryland 20852:

No. phBDMF-C-1 40648 Access Point

XbBBNF-Gl 40649 The present invention is not limited in scope by the specific arrangements described herein. Indeed, various modifications of the invention in addition to those described herein will be apparent to those skilled in the art from the foregoing and accompanying drawings. These modifications are intended to be within the scope of the appended claims.

Various publications are cited herein, their disclosures being incorporated herein by reference only for their content.