WO1993000922A1 - Peptides aiding nerve regeneration - Google Patents

Abstract

Use of a first peptide of the formula D-Ala-Gln-Tyr-Phe-Arg-Trp-Gly-NH2 and a neurotrophic analog or fragment of ACTH in the preparation of a medicament for the treatment of neuromuscular degeneration in a mammal.

Description

FEPTIDES AIDING NERVE REGENERATION

Background of the Invention

This invention relates to therapeutic peptides. ACTH-like neuropeptides can facilitate nerve regeneration in the central and peripheral nervous systems. Bijlsma et al. Eur. J. Pharm.. 76 (1981) 73-79 and Bijlsma et al. Eur. J. Pharm.. 92 (1983) 231-236 describe using ACTH-like neuropeptides, including H-Met(0₂)-Glu-His-Phe-D-Lys-Phe-OH and α-MSH, to restore sensorimotor function in rats having crushed sciatic nerves (where no D- or L-isomeric designation is given herein, the naturally occurring L-isomer is intended) .

Summary of the Invention In general the invention features a method for treating muscle wasting which results from motor nerve trauma (e.g. , severed or compressed motor nerves) or a disease that causes neuromuscular degeneration, by administering to a mammal, particularly a human patient, a combination of the analog of the invention having the formula D-Ala-Gln-Tyr-Phe-Arg-Trp-Gly-NH₂ (designated BIM 22015) , with a second neurotrphic analog or fragment of ACTH, e.g., ACTH (4-10).

Administration of the two agents can be sequential, separate and simultaneous, or by use of a composition including the two agents provided together in admixture, preferably with a pharmaceutically acceptable carrier substance such as saline. Administration of the combination composition can be by any method as described herein for administration of the single analog alone. We have discovered that the analog, BIM 22015, affects muscle strength and speed, enhances muscle fiber diameter, and inhibits muscle atrophy; myotrophic properties which are distinctly different than those observed in other neurotrphic ACTH analogs. Thus, neuromuscular degenerative diseases such as muscular dystrophy, multiple sclerosis, progressive infantile spinal atrophy, hypotonia, and amyotrophic lateral sclerosis can be treated by a combination of this compound and other neurotrophic ACTH analogs (e.g.., ACTH 4-10) . Such therapy will treat both the neurodegenerative and the myodegenerative aspects of the disease.

The neurotrophic ACTH analog or fragment can have the formula:

A¹-A²-A³-A⁴-Phe-Arg-Trp-A₅, wherein A¹ is H or acetyl; A² is Ala, D-Ala, Nle, or Met; A³ is Glu or Gin; A⁴ is His or Tyr, A⁵ is NH₂, Gly-NH₂, or D-Ala-NH₂, provided that when A³ is Glu and A⁴ is His, A² cannot be Met or Nle; or a pharmaceutically acceptable salt thereof.

In preferred embodiments of the heptapeptide, A¹ is H, A² is Nle, A³ is Gin, A⁴ is His, and A⁵ is Gly-NH₂; A¹ is H, A² is Nle, A³ is Glu, A⁴ is Tyr, and A⁵ is Gly-NH₂; A¹ is H, A² is Met, A³ is Gin, A⁴ is Tyr, and A⁵ is Gly-NH₂; A¹ is H, A² is Ala, A³ is Gin, A⁴ is Tyr, and A⁵ is Gly-NH₂; and A¹ is H, A² is D-Ala, A³ is Gin, A⁴ is Tyr, and A⁵ is Gly-NH₂.

The therapeutic peptide can be combined with a pharmaceutically acceptable carrier substance, e.g., magnesium carbonate or lactose, together form a therapeutic composition for aiding in regenerating nerves of the peripheral nervous system and the brain. The composition can be in the form of a pill, tablet, capsule, liquid, or sustained release tablet for oral administration; a liquid for nasal administration; or a liquid for intravenous, subcutaneous, a topical, transder al, or sustained release tablet, surgical suture, glue, or chamber for parenteral, or intraperitoneal administration.

Other features and advantages of the invention will be apparent from the following description of the preferred embodiments thereof, and from the claims.

Description of the Preferred Embodiments We first briefly describe the drawings.

Fig. 1 shows a schematic illustration of gait differences for different injuries, and is taken from Bain et al., 1989 Plastic and Reconstructive Surgery.83:129. Fig. 2 is a line graph of the functional recovery after treatment with saline or different dosages of BIM 22015 (100% = 100% recovery; data in mean and its SEM) .

Fig. 3 is a bar graph representation of the walking test (SEM is adjusted by the 95% confidence multiplier) .

Fig. 4 is an endplate area bar graph.

Fig. 5 is an endplate perimeter graph (p < 0.025) . Fig. 6 is an endplate graph of internal branching.

Fig. 7 is an endplate muscle fiber diameter (p < 0.05) .

Fig. 8 is a bar graph representation of muscle contraction in response to stimulation with increasing numbers of pulse pairs (n=5) .

Fig. 9 is a set of representative traces which show peak amplitude and decline from peak amplitude of intact EDL muscle, denervated EDL muscle treated with saline, denervated EDL muscle treated with ACTH 4-10, or denervated EDL muscle treated with BIM 22015, during peroneal nerve stimulation with 300 pulse-pair units of supramaximal strength (peak from end function; UNKELSCOPE) . Fig. 10 is a set of bar graph representations of the frequency distribution of EDL motor unit size recruited from threshold to 1.8 treshold stimulus strength in intact EDL muscle treated with saline, denervated EDL muscle treated with saline, denervated EDL muscle treated with ACTH 4-10, or denervated EDL muscle treated with BIM 22015.

Fig. 11 is a line graph representation of the return of toespread as indicated by the ratio of the denervated left foot digit 1-5 distance to that of the intact right foot over a period of 5 days following peroneal nerve crush and treatment with varying levels of BIM 22015.

We now describe the structure, synthesis, testing, and use of preferred embodiments of the invention. Structure

The heptapeptides of the invention have the general formula recited in the Summary of the Invention above. They all have an NH₂ at the carboxy terminal end, in addition to Phe at position 4, Arg at position 5, and Trp at position 6.

The heptapeptides can be provided in the form of pharmaceutically acceptable salts. Examples of preferred salts are those of therapeutically acceptable organic acids, e.g., acetic, lactic, maleic, citric, malic, ascorbic, succinic, benzoic, salicylic, methanesulfonic, toluenesulfonic, or pamoic acid, as well as polymeric acids such as tannic acid or carboxymethyl cellulose, and salts with inorganic acids such as hydrohalic acids, e.g., hydrochloric acid, sulfuric acid, or phosphoric acid.

Synthesis

The synthesis of H-Nle-Gln-His-Phe-Arg-Trp-Gly- NH₂ follows.

Other heptapeptides of the invention can be prepared by making appropriate modifications of the following synthetic method.

The first step is the preparation of Nle-Gln- His-Phe-Arg-Trp-Gly-benzyhydrylamine-resin, as follows.

Benzyhydrylamine-polystyrene resin (Bachem, Inc.) (1.00 g, 0.5 mmole) in the chloride ion form is placed in the reaction vessel of a Beck an 990B peptide synthesizer programmed to perform the following reaction cycle: (a) CH₂C1₂; (b) 33% trifluoroacetic acid CH₂C1₂ (2 times for 1 and 25 min. each) ; (c) CH₂C1₂; (d) ethanol;

(e) CH₂C1₂;

(f) 10% triethylamine in CHC1₃; and (g) CH₂C1₂.

The neutralized resin is stirred with Boc-glycine and diisopropylcarbodiimide (1.5 mmole) in CH₂C1₂ for 1 hour and the resulting amino acid resin is then cycled through steps (a) to (g) in the above wash program. The following amino acids (1.5 mmole) are then coupled successively by the same procedure:Boc-Trp, Boc-tosyl-Arg, Boc-Phe, Boc-carbenzoxy-His, Boc-Gln, and Boc-Nle.

After washing and drying, the completed resin weighs 1.60 g.

From the above resin is prepared H-Nle-Gln-His-Phe-Arg-Trp-Gly-NH₂, as follows.

A mixture of the above heptapeptide resin (1.85 g, 0.5 mmole) and a solution of 4 ml anisole, and 36 ml hydrogen fluoride is stirred at 0°c for 45 minutes. Excess hydrogen fluoride is evaporated rapidly under a stream of dry nitrogen, after which the free peptide is precipitated and washed with ether.

The peptide is then dissolved in a minimum volume of 50% acetic acid and eluted on a column (2.5 x 100 mm) of Sephadex G-25. Fractions containing a major component, as determined by u.v. absorption and thin layer chromatography (tic) are pooled and evaporated to a small volume in vacuo. This solution is applied to a column (2.5 x 50 cm) of octadecylsilane-silica (Whatman LRP-1, 15-20 m mesh size) which is eluted with a linear gradient of 0-40% acetonitrile in 20% acetic acid in water. Fractions are examined by tic and analytical high performance liquid chromatography (hplc) and pooled to give maximum purity. Repeated lyophilization of the solution from water gives 81 mg of the product as a white, fluffy powder.

This material is found to be homogeneous by hplc and tic. Amino acid analysis of an acid hydrolysate confirms the composition of the heptapeptide. H-Nle-Glu-Tyr-Phe-Arg-Trp-Gly-NH₂, H-Met-Gln-

Tyr-Phe-Arg-Trp-Gly-NH₂, H-Ala-Gln-Tyr-Phe-Arg-Trp-Gly- NH₂, and H-D-Ala-Gln-Tyr-Phe-Arg-Trp-Gly-NH₂ are prepared in similar yields in an analogous fashion by appropriately modifying the above procedure. Testing of Heptapeptides

Heptapeptides of the invention may be testing for their effect on nerve and muscle degeneration as described below.

One heptapeptide of the invention, BIM-22015, which has the amino acid formula

D-Ala-Gln-Tyr-Phe-Arg-Trp-Gly-NH₂, was tested on two different types of peripheral nerve lesions: crushed and sectioned nerves. In rats, the effect of different dosages and modes of administration of BIM-22015 on regeneration of the crushed peroneal nerve or the sectioned sciatic nerve was investigated. In rabbits, morphological studies were performed in which the sciatic nerve was sectioned and sutured prior to treatment with

BIM-22015. In addition, electrophysiological and morphological parameters, as well as motor function, of denervated muscle were studied following treatment with

BIM-22015 or ACTH 4-10. The results of these combined studies, presented below, suggest that BIM-22015 may be more effective on the neuromuscular junction (i.e., endplate) than on the axon, and that both muscle and nerve response to this peptide differ when compared to other ACTH analogs.

Example 1. Effect of BIM-22015 on Crushed Peroneal Nerve Regeneration in Rats. The effect of BIM-22015 regeneration of the crushed peroneal nerve in the rat was assessed using both morphological and physiological studies, as follows. Both locomotor pattern and endplate morphology during the early stages of crushed nerve regeneration were assessed after treatment with BIM-22015. Adult Sprague-Dawley male rats (Taconic Farms) were maintained at 22° C on a 12 hour light and a 12 hour dark cycle and were supplied with standard rat chow and water ad lib. All surgery was performed under anesthesia with 8% chlorohydrate (0.6 ml/lOOg I.P.). Denervation was accomplished using a #5 Dumont forceps with a uniformly filed tip to crush the peroneal nerve. This crush produced a 1 mm wide lesion just proximal to the EDL muscle. The wound was closed with surgical staples. Following surgery, the animal was immediately administered saline or BIM-22015 in one of the following dosages: (1) 0.9% saline, (2) 10.0 ug BIM-22015, (3) 5.0 ug BIM-22015, (3) 1.0 ug BIM-22015, or (4) 0.1 ug BIM-22015 (per animal). Administration continued once every 48 hours for 5 days; on day 5, the rats were killed and tissues taken for morphological analysis. All injections and surgeries occurred between 10:00 and 12:00 noon.

Locomotor pattern was tested by measuring the walking pattern of the animal. The animal was allowed to walk up a 20° inclined platform. The track was well-lit by fluorescent tubes and a dark box was placed at the end of the track. This dark and light differences was used to entice the animal to walk up the track. The track was lined with 13.5 inch wide continuous computer paper (201b.). The rat's hind feet were dipped in non-toxic ink, and the animal allowed to walk naturally. The walking pattern test was performed the day prior to surgery and on the 3rd, 4th, and 5th days after surgery. All walking tests were performed between 10:00 and 12:00 noon. Six undisturbed and consistent walking patterns were chosen for analysis of toespread differences. Fig. 1 is a schematic illustration of the possible patterns. The distance between the first and fifth digit was determined using a Manostat mechanical caliper (type 6911) with accuracy of + 0.005 mm. The contralateral foot was used as a control to negate the size differences in individual animals. The Student T test was used to determine the statistical differences, and data were expressed as means of the six sharpest toespreads and the standard error of the mean (using 95% confidence formula) , standard deviation, and variances.

Although the results show no statistical significance in any of the parameters measured, probably due to the small animal number per group and the small magnitude of treatment-related improvement, the O.lmg dosage of BIM-22015 resulted in a very consistent upward recovery rate. These results, presented in Figs. 2 and 3, show that the optimal dosage of the BIM-22015 analog is 0.1 g. The other dosages outperformed the saline control animal but only by a slight margin. The O.lmg dosage slightly outperformed the 5.0mg. These results illustrate the beneficial effect of BIM-22015 on the recovery rate of locomotion at the lowest dose tested (O.lug) . Morphology

Morphological studies of crushed nerve regeneration were performed as follows. On day 5, the rats were anesthetized with 8% chlorohydrate (0.6ml/100g IP) before surgery. The animal was placed ventral side down and a skin incision was made on the hindlimb from the lateral epicondyle of the femur to the lateral malleolus in the ankle. The extensor digitorum longus muscle (EDL) was then carefully exposed from its surrounding connective tissue and extreme care taken to preserve the blood supply. The EDL is a heterogenous muscle consisting of both fast white muscle fibers and slow red muscle fibers. It is classified as a fast muscle based upon its isometric twitch parameters. The origin of the EDL is the epicondyle of the femur to the distal tendon which divides into four parts with insertions into the third phalange base of the second to the fifth digits. It is innervated by the peroneal nerve and its blood is supplied by the anterior tibial artery. The muscle was prepared for staining as follows. The silver-cholinesterase method of staining of

Pecot-Dechavssine et al. (1979 Stain. Technol. 54:25) was used to stain the endplates, or neuromuscular junctions, of the EDL. Briefly, the muscle was fixed in situ by 1% neutral formalin in normal Ringer's solution. The muscle was then excised and placed in fresh 1% neutral formalin solution for extra 3-4 minutes. The endplates were localized with the stain described by Karnovsky et al. (1964, J. Hist. Chem. Cytochem. 12:219) for 7-15 minutes under visual control. After localization, the EDL was soaked in 5% neutral formalin solution for 20 minutes, then washed in Triton X-100 detergent for 1 hour. After the detergent step, the EDL was fixed in 80% ethanol solution for 1 hour. Finally, the endplates were impregnated with silver using 0.5% protargol solution overnight (12-30 hrs. , at 37°C) . The following day, the excess silver was reduced using a Bodian developer for 18 minutes. The muscle was rinsed in distilled water, finely teased, and mounted using a glycoglyerol mounting gel. The endplates were observed using Olympus BH series binocular microscope at a magnification of lOOx using an oil immersion objective lens. A low light sensitive video camera mounted on top of the Olympus projected the slide onto a black and white monitor. A Bioquant image digitizer with a manual optic mouse was used to determine the following parameters: (1) endplate parameter, (2) endplate area, (3) endplate internal branching, and (4) muscle fiber diameter. The Bioquant statistical package was used to determine statistical differences, and the data are expressed as means and standard error of the mean. The Student's t-test and the ANOVA statistical tests were performed. The Bioquant image digitizer was attached to a Hipad Digitizing pad (Houston Instruments, Austin, Texas) and was linked to the Apple lie personal computer using a morpho etry software, Bioguant JE (R & M Biometrics, Nashville, Tennessee) . Results

Treatment of the crushed nerve with BIM-22015 results in a compensatory effect at the neuromuscular junction which was most evident at the lowest dose administered (0.1 g) . The results showed a significant increase in endplate perimeter at O.lmg (p < 0.025) (Figs. 4 and 5) and 5.0mg (p < 0.05) of BIM-22015, and internal branching (Fig. 6 and Table 1) at O.lmg BIM-22015 (p < 0.025). Measurement of the endplate muscle fiber diameter revealed a significant (p < 0.005) high muscle hypertrophy at two dosages of BIM-22015, l.Omg and lO.Omg. At the 0.1 mg and 10 mg doses, there was also a dose-effect relationship, which may be due to enhancement of muscle-sparing effects similar to the one observed with some peptides, such as MSH and 3-endorphin. Hugh et al. (1989, Neuroscience Letts. 103:169) have shown endplate over-expression of a-MSH and 3-endorphin in animals suffering from motoneurone disease, e.g., in wobbler or dystrophic mice.

Example 2. Effect of BIM-22015 on Muscle Weight After Sciatic Nerve Section in Rats.

BIM-22015 was tested for its effect on muscle weight after sciatic nerve section, and rejoining as follows. Four groups of between eight and ten Wistar rats, eight weeks of age and approximately 200 g each, were tested. The sciatic nerve of each rat was dissected and a clean section was performed with surgical scissors. The proximal and distal stumps were then glued together with tissucol, and BIM-22015 was administered topically in surgical glue (i.e., tissucol) at 100 ug/0.5ml (group 2), at l.l ug/0.5ml (group 3), or by s.c. administration at 5ug.kg.day for 18 days. The control group (group 1) was treated with tissucol alone. The sectioned nerve was allowed to regenerate for 35 days before evaluation. On day 35, each rat was anesthetized with phenobarbital prior to surgery. The exterior leg was surgically opened in order to free the common extensor muscle of the finger. The distal tendon was cut at the level of the heel. Following electrophysiological measurements of muscle tension in response to acute stimulus, muscle tentanos at 100Hz, muscle exhaustion following a 5 second contraction at 500 Hz, the muscle was sectioned at the proximal level, the tendons removed, and the muscle was dried and weighed. Table 2 shows results of muscle measurements. The results show a dependence on the route of administration of BIM-22015, as topical administration (groups 2 and 3) was largely ineffective, but the s.c. route of administration (group 4) was associated with significant improvement of muscle strength. For example, when muscle weight is normalized to mean animal weight (Table 2, column 2), group 4 (s.c. administration) showed an improvement of 15% over the control group 1, but groups 2 and 3 showed little improvement. The increase of muscle weight by s.c. administration of BIM-22015 may be due to an increase in the number of fibers which are innervated or in the volume of a few reinnervated fibers. It was also determined that the increase in muscle weight was not associated with edema or lipid infiltration. The inter-animal variability in this experiment was too large to achieve statistical significance.

The effect of BIM-22015 on muscle strength is shown in Tables 3 and 4. Group 4 (Table 3) shows a significant (p = 0.01) improvement of 30% over control group 1 (column 2, Table 3). This improvement is also evident when the muscle strength (i.e., contractile response) is normalized to the muscle weight (column 3, Table 3) . Group 4 (Table 4) shows an improvement of 28% (p = 0.05) over control group 1 (column 1, Table 4). However, when muscle strength is normalized to muscle weight (column 2) , the improvement is no longer significant compared to the control group. Table 4 also shows that the speed of muscle contraction (column 3) is dramatically improved (300%) in group 4 compared to group 1; this difference is statistically significant (p = 0.01). The speed of relaxation (column 4, Table 4) is also dramatically improved in group 5 compared to group 1. However, no effect was found on fatigability of the muscle at both low (100Hz) and high (500 Hz) frequency. The morphological results show an increase in muscle weight after s.c. administration of BIM-22015, which can only be accounted for by an increase in the contractile protein from the muscle fiber. The results of the functional tests suggest that muscle reinnervation was accelerated by treatment with BIM-22015. These results confirm the conclusion that BIM-22015 prevents muscle wasting, as was found above for experiments involving crushed peroneal nerve regeneration. Example 3.Morphological Studies of the Effect of

BIM-22015 After Section and Suture of Sciatic Nerve in Rabbits.

The sciatic nerve of the rabbit after transection followed by end-to-end suture was studied, using light and transmission electron microscopy. The effects of BIM-22015 on the number of newly formed myelinated nerve fibers, the axonal diameter, and the thickness of the myelin sheath were examined, as follows.

Preparation of Nerve Samples 52 young, adult male rabbits of the White New

Zealand strain were anesthetized under sterile conditions with sodium pentobarbital injected into the marginal ear vein in a dose of 30 mg/kg body weight. A neurosurgeon performed a transection of the left sciatic nerve followed by a neurosuture under the operating microscope. A lateral skin incision was made on the thigh between the trochanter major and the knee joint. The thigh musculature was exposed to reveal the nerve which was then transected at a point about 1.5 cm distal to where it emerged from below the gluteal muscle. The point of transection was located distal to its division into the muscular branch of the sciatic nerve. By preserving this branch of the nerve, the animals showed virtually no signs of paralysis after the operation. The two ends of the nerve could be fixed by two to three tension-free sutures around the peri- and epineurium. The wound was closed by a continuous skin suture and no antibiotics were used. Daily administration of BIM-22015 to 25 animals began one hour after the start of the operation. The animals received the substance subcutaneously in a dose of lmg/kg bodyweight per day in 1 ml sodium chloride solution. After the operation, the rabbits were divided into 4 groups each of 12 animals (except Group 4 had 16 animals) . One-half of each group (6 animals, except for 7 animals in Group 4) was treated with BIM-22015 (Test animals) and the remaining animals were given a sodium chloride solution (Control animals) .

After sacrifice of the animals under Nembutal anesthesia the left sciatic nerve was removed from each of 6 test animals and 6 control animals after 3, 8 or 32 days, and from 7 test and 9 control rabbits after

96 days. In each case, an approximately 4 cm piece of the nerve could be obtained. Immediately after its removal, the nerve was stretched and stuck to a Styropor frame and placed in a 5% cooled (4°C) cacodylate buffered glutaraldehyde solution (pH 7.2) for 1-2 hours.

After a two to three hour pre-fixation, tissue blocks were removed from the following regions: the area of the transection, 4 mm proximal and 3, 6 and 10 mm distal to this point. The main areas of interest for investigating changes were to be the region of the transection and 1 cm distal to this. The samples obtained were fixed for another 24 hours in special glass vessels with cacodylate buffered, cooled (4°C) glutaraldehyde solution. The samples were then rinsed for 1-3 hours in cacodylate buffer (pH 7.2) and then postfixed in 1% osmium tetroxide for 2 hours. After repeated rinsing with cacodylate buffer (4 x 30 minutes) and dehydration in graded alcohol solutions up to an alcohol concentration of 100% (in each case for 30 minutes) , the samples were placed for 60 minutes in propylene oxide and then in a propylene oxide-Epon 812 mixture (proportion 1:1). The samples were then transferred to pure Epon 812 (Serva, Heidelberg) for embedding. The polymerization of the plastic medium took place in a special polymerization cupboard for one day each at 35°C and 45°C and for three days at 60°C.

With an ultramicrotome 0.5 mm thick semithin sections were cut. The sections were then stained with 0.25% toluidine blue at 80°C and used for evaluation of the tissue changes by light microscopy and the subsequent preparation of ultrathin sections. The latter were obtained using an ultramicrotome (Reichert OM U3) and placed on coppermesh microscope slides. The contrast process was carried out in two steps in the automatic contrast machine (2168 Ultrastainer Carlsberg System, LKB Sweden) . In the first step, a 30 minute treatment with Ultrastain 1 (uranyl acetate solution) was carried out at 40°C, followed by a washing process and, finally, a counter-contrast was undertaken for 80 seconds with Ultrastain 2 (lead citrate solution) . The ultrastructural evaluation was undertaken with the electron microscope EM (Zeiss, Oberkochen) at 60kV. The results were documented on dimensionally-stable film material. Image Analysis

In order to analyze the influence of BIM 22015 on the morphology of myelinated nerve fibers during regeneration, a computer-controlled semi-automatic interactive image analysis was carried out. With this method, geometric data were collected and evaluated. The system used consists of a coordinates measuring table with a processor for measuring geometric parameters and an electronic pencil (MOP AM 02 KONTRON, Eching) with a computer (HP 9845 B, Hewlett-Packard) to take data from MOP AM 02 and transform it from geometric sizes into the desired parameters using a special program.

The measurements were carried out on electron micrographs with a magnification of 6,400. At this magnification, the structures were large and clear enough to be able to trace them exactly with the electronic pencil. In addition, at this magnification, a sufficiently large number of nerve fibers were portrayed, so that the amount of material in the micrograph remained within a certain range.

The morphometric measurements were taken on image material from 6 test and control animals with a survival time of 32 days and 7 test and 9 control animals with a survival time of 96 days. 10 micrographs each from the region of the operation site (loc 2) and from a region 1 cm distal from this site (loc 5) were chosen at random.

The outer and the inner perimeters of the myelin sheath and the external perimeter of the axon were traced with the electronic pencil. In the majority of cases this is identical with the inner perimeter of the myelin sheath. The MOP AM 02 measured the following geometric data: the total area of the nerve fibre, i.e., the area of the axon and the myelin sheath, the area of the axon, the external perimeter and the outer shape of the nerve fibre, and the inner perimeter of the myelin sheath and the shape of the axon in cross-section. From these values, a computer program calculated the thickness of the myelin sheath and the diameter of the axon as diameter of a circle of equal area. As the third variable, the program determined a form factor, that gave the deviation from circularity.

In the next step, the individual values of all test and control animals were collated separately for the variables; myelin sheath thickness and axon diameter per localization and time point, and the distribution of the individual values within the treated and untreated animals determined. A second statistics program provided the mean value, the standard deviation and minimal and maximal values for each variable within a group of animals according to localization and time point. Finally, histograms could be were prepared with these data that represented the data in graphical form. Analysis of Variance All the individual values of the two variables, myelin sheath thickness and axon diameter per animal, localization and time point were first converted separately to logarithms. The mean value was calculated from these new values, and then converted back to ordinary numbers. The two-way analysis of variance could be carried out with the resulting representative values using the statistics program ANOVA (NVA2U) . The aim of this analysis of variance was to show what effect the factor time and the factor treatment have on the growth of the myelin sheath and the axon. In addition, the time-dependency of the effect was investigated. Results

The light microscopic evaluation of the semi-thin sections of 0.5 mm following the surgical severance and microsurgical end-to-end suture of the sciatic nerve did not disclose definite differences between the control and experimental groups. The light microscopically demonstrable effects of the treatment with BIM-22015 cannot be confirmed due to the great variation of findings among the animals. Since the sciatic nerve is not very long in the rabbit and therefore only permits the evaluation of a localization 10 mm distal from the surgical site, it is very difficult to estimate the speed of the growing nerve. The nerve grows 1.2 to 1.4 mm per day so that numerous nerve fibers have already reached the localization 10 mm distal from the surgical site by day 8 p.o. In addition, differences between control and experimental animals were not observed by electron microscopy. The morphometric findings in the region of transection and in the region 1 cm distal to the transection are summarized in Table 5 and Table 6, respectively.

The myelinated fibers could only be seen after 8 days with electron microscopy, because the myelin sheaths were still very thin. However, they were so few in number that no statistical evaluation could be made at this time. This finding confirms the observation of Gutman et al. (1943, J.Physiol. 101:489), that the regeneration of peripheral nerve fibers after crushing occurs faster than after a nerve transection. Since in the early regeneration phase (3 and 8 days after surgery) , remyelinisation is not yet very far advanced, the number of nerve fibers, the axon diameter and the myelin sheath thickness were analyzed statistically at a time when the myelinated fibers seen on light microscopy were present in larger numbers (32 and 96 days after the operation) .

In the region of the transection, the differences between treated and untreated animals were not confirmed statistically, although the myelin sheath is, on average, thicker in the treated animals both after 32 days and after 96 days. In the region 1 cm distal of the transection, the myelin sheath is also only slightly thicker in the treated animals after 32 days than in the untreated controls. In the period from the 32nd to the 96th day, the myelin sheath grows almost twice as fast in the treated animals. After 96 days, the myelin sheath in the treated animals is significantly thicker than in the controls. Thus the effect of BIM-22015 on the increase in thickness of the myelin sheath over the period from the 32nd to the 96th day is statistically confirmed. A statistically significant effect of BIM-22015 on the axon diameter could not be demonstrated either on the 32nd or on the 96th day p.op. Although also not statistically confirmed, on average, a higher number of myelinated nerve fibers could be counted in the treated animals. After 32 days, the number of medullated fibers was considerably higher in the region of the transection, and in the area 1 cm distal of the lesion, both after 32 and 96 days. The fact that these differences were not statistically significant is probably due to the large scatter of the individual values in the groups of both treated and untreated animals. These results should not be ignored in evaluating the effect of BIM 22015 on the regeneration of peripheral nerves.

The macroscopically observed thickening in the region of the transection due to the formation of granulation tissue in two control animals is to be regarded in both cases as a random result and independent of the effect of BIM 22015 treatment. Overall, the formation of granulation tissue after the operation was so small that it did not affect the nerve regeneration. Therefore, BIM-22015 exerts an effect not only in the early regeneration phase, but also at later times after nerve trauma, with both the number of axonal sprouts and also remyelinisation and thickness of the myelin sheath being affected.

Example 4. Comparison of BIM-22015 and ATCH (4-10) on Crushed Nerve Regeneration in Rats.

Electrophysiological and morphological parameters, as well as motor function, of the rat extensor digitoru longus (EDL) muscle were measured at varying times following treatment of a crush lesion of the peroneal nerve with BIM-22015 or ACTH 4-10. Male Sprague Dawley rats, weighing 100-150g, were obtained from Blue Spruce Farms (through Wards Scientific) one week prior to the experiment, housed in groups of 4 or 5 in a 12h light/12h dark cycle, and fed lab chow and tap water ad lib. All surgeries were performed between 10 a.m. and 12 noon and all injections were given during these times. Injections were administered commencing immediately after surgery and given subsequently every 48 h. Under 8% chloral hydrate anesthesia, the left peroneal nerve of each rat was exposed at the point of entrance into the peroneus longus muscle, then crushed with a #5 Dumont forceps with a uniformly filed tip, producing a 1 mm wide lesion. The wound was closed with surgical staples. This technique gives well defined axonal interruption without interrupting the connective tissue sheaths in the region. Sham-denervation consisted of exposing the peroneal nerve and then suturing the incision without damage to the nerve. Intact rats were not exposed to any surgery. Electrophysiology

Treatment groups . The rats were divided into 4 groups, each consisting of 5 animals: (1) intact rats; (2) denervated rats administered physiological saline i.p.; (3) denervated rats administered 40μ*g/kg ACTH 4-10 i.p.; (4) denervated rats administered 40μ.g/kg BIM 222015 i.p. The volume of each injection was 0.2 ml and all injections were administered according to the pattern described above. Preparation for electrophysiological recording.

Under sodium pentobarbital anesthesia (50 mg/kg body weight i.p.) the EDL muscle was freed from the surrounding tissue. The tendon-to-tendon length (resting length) was recorded before the distal tendons were attached to an isometric transducer (Thornton type 420) calibrated before the experiment to convert voltage to grams pull. The foot and femur were pinned to prevent movement. The area was bathed throughout the experiment with paraffin oil maintained at 37°C ± 1. The rat was rigidly attached in the prone position to a temperature- controlled surgical table. The peroneal nerve was exposed and a miniature stimulating electrode (SNEX 200; 1 mm diameter hook; Rhode Medical Instrument) placed under it and held with a rubber sleeve for insulation. Response to Pulse Pairs: Characteristics of

Twitch and Tetanus . The peroneal nerve was stimulated, using a constant voltage isolated physiological stimulator (Coulborn Instruments model E13-51) , with single pulses (duration 200 /sec) increasing in strength until maximum contraction amplitude of the EDL muscle was obtained. Three EDL responses were recorded at 2X maximal stimulus strength and used to determine the characteristics of the twitch. The stimulator was then programmed to generate pairs of pulses. The duration of each pulse was set of 200 μsec. The interpulse was set at 2,000 μsec. The number of pulse pairs was increased from 1 to 5 (in steps of 1) , then to 10, 20 and 40 and the data analyzed for parameters of the tetanus.

Characteristics of the muscle twitch measured were contraction amplitude, duration and rate of rise, and half-relaxation duration. For the tetanus, contraction amplitude, tetanic tension, and maximum rate of tension development were determined, and the twitch/tetanus ratio calculated. Recruitment of Motor Units . Motor units (MUs) were recruited by selective activation by a method similar to the one introduced by McComas et al. (J. Neurol. Phychiatry, 3_4:121-131, 1971), and modified by Harris and Wilson (J. Neurol. Neurosurg. Psychiatry, 3_4:512-20, 1971). EDL muscles of intact rats were used for biological calibration according to the guidelines suggested by Swett and Bourassa (Electrical Stimulation Research Techniques, Patterson et al., eds., Academic Press, NY pp. 244-295) . The stimulus intensity was normalized with respect to minimal activation of motor units. Small graded pulses were delivered to the peroneal nerve to detect the smallest possible contraction of the EDL muscle. The stimulus intensity that evoked this response was considered the threshold stimulus (T) . Stimulus intensity was increased gradually from T to 1.2T, 1.4T, 1.6T, 1.8T and 2.0T. Each resulting increment in EDL contraction amplitude was interpreted as the recruitment of a putative MU. The mean increment in the first 5 putative MUs between T and 2T was used to estimate the number of functional MUs in the EDL. During reinnervation, MUs activated within the range of the stimulus strength were easily separated, whereas those that were activated at greater stimulus strength overlapped considerably.

Motor Unit Estimate (MUE) was calculated using the formula:

MUE = Maximum contraction amplitude (twitch)

Mean increment of first 5 putative MUs

Motor Unit Performance . The stimulator was programmed to deliver pairs of monophasic pulses. The duration of each pulse was set at 200 sec. The interpulse was set at 2,000 μsec. The ability of the indirectly stimulated EDL muscle to reach and maintain peak contraction amplitude was tested with 300 pulse-pair units. The pattern of stimulation was synthesized from 4 sequential intervals. A pulse-pair unit is defined as one "pass" through one sequence of the 4 intervals. A pulse-pair unit is defined as one "pass" through one sequence of the 4 intervals. The EDL muscle responses were displayed on the screen of the microcomputer and peak contraction amplitudes measured with 2 cursers. The decline in contraction amplitude was evaluated using a "peak from end" function: this function set the signal at a time "t" equal to the absolute maximum peak and the percent decline from this peak to that at the last sampled time was calculated. The following parameters were evaluated: estimated number, size, combined contraction amplitude of the first 5 MUs recruited at 1.8T, maximum twitch amplitude of the EDL muscle, contribution of the first 5 MUs to maximum EDL twitch amplitude and the frequency distribution of MU population.

EDL muscle responses to indirect stimulation were displayed on a Tektronix oscilloscope (model 2220) and transferred to a microcomputer (Epson; Apex 100/20) by means of an analog-digital converter (Thornton SPI- 501) . Motor unit responses were measured by means of UNKEL-SCOPE software (MIT) . Contraction amplitude (at resting muscle length) of each recruited MU was displayed on the screen of the microcomputer and measured with 2 cursers. Numerical data were transferred to Lotus 123. Statistical analysis was performed with Statistix II software. Analysis of variance (ANOVA) was used to establish significance of differences between pairs at the 95% confidence level. Results While denervation decreased twitch contraction amplitude and increased twitch duration and half- relaxation time, there were no significant differences in any of the twitch parameters, nor in the twitch/tetanus ratio, in denervated rats treated with saline, BIM 22015 or ACTH 4-10. However, when the muscle was stimulated at a high rate for a short period of time, BIM 22015 treatment resulted in an improved muscle response as the number of pulses was increased. Figure 8 shows that increasing the number of pulse pairs significantly augmented the contraction amplitude of the denervated, peptide treated muscles as compared to the enervated saline treated controls. This improvement in amplitude appeared at a lower pulse number in the rats treated with BIM 22015 than those treated with ACTH 4-10. The rate of tension development in the indirectly stimulated EDL muscle was significantly increased by treatment with BIM 22015 when tested 11 days after crush denervation. A significant improvement in rate of tension development was not seen in the ACTH 4-10 treated rats until the number of pulse pairs was increased to 20 (Table 7) . Peptide treatment was not able to restore any of the electrophysiological parameters tested in denervated EDL muscles to those of the sham-denervated muscles.

Peak amplitude following repetitive indirect stimulation was reduced by denervation to approximately 27% of precrush values (Table 8) . Both peptides were able to improve peak amplitude but the effect of BIM 22015 differed from that of ACTH 4-10 on this characteristic of muscle function. Following BIM 22015 treatment, the initial response was better: the peak was higher and the muscle was able to maintain this peak for the first 0.4 sec of stimulation and its superiority over the saline controls for up to 1.0 sec after the initial peak response. However, while the ACTH 4-10 muscle did not perform as well for the first 0.4 sec of stimulation, it fatigued less rapidly (Fig. 9).

Crush denervation markedly reduced total EDL motor unit number and increased the mean size of the first 5 MUs. Table 9 shows that peptide treatment, while not restoring these parameters to precrush values, increased total MU number and reduced the mean size of the first 5 MUs. In addition, BIM 22015 altered the size distribution of the reinnervated MUs, producing more intermediate sized units than in the saline treated rats (Figure 10) .

As shown in Table 9, denervation caused a severe drop in maximum twitch amplitude and a marked increase in the summed contraction amplitude of the first 5 MUs; their percent contribution to maximum twitch amplitude was increased more than 7x. While neither peptide restored these values to precrush figures, the effect of ACTH 4-10 was significantly better than the saline treatment. BIM 22015 was ineffective. Morphology

Treatment groups . The animals were divided into 5 groups, a sham denervated group and 4 denervated groups each treated according to one of the following regimes: (1) administered saline; (2) administered 0.1 /j,g/kg/48 h BIM 22015; (3) administered 10 g/kg/48 h BIM 22015 and (4) administered ACTH 4-10 10 g/kg/48 h for 7 days after nerve crush. In the second study, a dose- response experiment, the denervated rats were divided into 5 groups, each group receiving one of the following dosages of BIM 22015: 0.1; 1.0; 5.0; 10.0 μg/kg/48 h, or saline, for 5 days. Each group consisted of 5 rats.

Morphological techniques . The animals were anesthetized with 8% chloral hydrate and the EDL fixed in situ and prepared for light microscopy using a modified silver-cholinesterase stain. Whole mounts of the finely teased muscle were observed under an Olympus BH lOOx oil immersion objective lens (total magnification lOOOx) . Thirty-six endplates from each group were measured. The endplate area, perimeter, internal nerve terminal branching and muscle fiber diameter were obtained with a BioQuant image analysis system (R & M Morphometrics) . Student's t-test was used to determine statistical significance. Results Peroneal nerve crush caused a significant decrease in muscle fiber diameter measured 5 days after the lesion. As shown in Table 10, treatment with varying dosages of BIM 22015 increased muscle fiber diameter as compared to saline treated controls, although only the results with 0.1 and 10.0 μg/kg were statistically significant. Using the dosage of 10 μ*g/kg, the potency of ACTH 4-10 and BIM 22015 on muscle fiber diameter was compared at both 5 and 7 days after nerve crush. Table 11 shows that 5 days after nerve crush, the saline treated muscles were reduced to 76% of precrush diameter, as evaluated by the diameter of the sham-crushed, saline treated muscles. Those muscles treated with ACTH 4-10 atrophied to 71%, whereas the BIM 22105 treated EDL muscles retained 92% of their fiber diameter. There was essentially no further change in any of these groups at 7 days.

All endplate parameters were markedly reduced by nerve crush 7 days after the lesion. At this time, BIM 22015 dosages of 0.1 and 10.0 μ-g/kg were tested (Table 12) . Only the lower dosage was effective, significantly improving both nerve terminal branching and area. Perimeter was not affected. In an extended dose- response study at 5 days after nerve crush, we found once again that it was the lowest dosage (0.1 g/kg) that improved endplate parameters, although at this early stage in reinnervation, endplate perimeter and nerve terminal branching were increased, while area was unchanged (Table 13) .

Return of motor function: Peroneal Function Index and Toespread Treatment groups . Following peroneal nerve crush, the rats were divided into 5 groups, each receiving one of the following injection regimes: saline, 0.1, 1.0, 5.0 or 10.0 μ-g/kg/48 h BIM 22015. There were 5 rats per group.

Peroneal Function Index (PFI) . The PFI was determined by a modification of the sciatic function test. This index is determined empirically using several gait parameters, including print length and toespread. The rat's hindfeet were dipped in non-toxic india ink and the animal induced to walk up an inclined plane (76 cm long) lined with clean white paper, with a dark box at the end of the runway. Gaits and toespreads were collected from the ink-prints on the paper track. Toespread. The distance between the first and

5th digits is the toespread and this value, together with the print-length (the distance between the tip of the third toe and the base of the pad) was determined for the injured foot using the BioQuant image analysis program. The contralateral footprint was used as a control to negate the effect of growth differences between animals. Walking tests were performed pre-crush and on the 3rd, 4th and 5th days after crush.

For the statistical evaluation of motor function, Student's t-test was used to determine the statistical significance between 2 groups: ANOVA was used for differences between several groups. Results

In this evaluation of the return of motor function, zero is the normal precrush value and this decreases to negative values following nerve crush. Complete recovery is taken as a return to the zero precrush value. Three days after nerve crush, this index decreased to -46% for the saline controls, and to -53% for the BIM 22015 treated rats. Control and peptide treated groups were approximately -36% 5 days after nerve crush; at 7 days after crush saline treated rats were - 40% whereas the BIM 22015 treated animals appeared slightly better at -33%. However, none of the differences between control and peptide treated animals were statistically significant.

When the ratio of the toespread of the injured foot to that of the contralateral foot was calculated, and the precrush ratio taken as 100%, rats treated with 0.1, 1.0, 5.0 or 10.0 μg BIM 22015 appeared, in general, to have a better ratio than saline treated controls. However, none of these values differed significantly from the control values (Fig. 11) .

BIM 22015 in many respects has neurotrophic functions similar to those of ACTH 4-10 and another ACTH analog, ORG 2766, in that it affects the contractile properties of indirectly stimulated, reinnervated muscle as well as the morphology of the endplate. However, BIM 22015 also demonstrates a number of myotrophic properties which were not observed after treatment with ACTH 4-10 or other ACTH analogs.

The electrophysiological studies indicate that BIM 22015 differs from other ACTH peptides in that treatment with ACTH 4-10 (Fig. 2) or other ACTH-like peptides improves muscle endurance, whereas BIM 22015 has less effect on endurance but improves muscle strength and speed. The increased muscle speed is demonstrated by the increased rate at which tetanic tension is generated; increased muscle strength is shown by the greater amplitude of muscle contraction when tested with increased numbers of pulse pairs. BIM 22015 treated muscle responds especially well to this type of increased demand. In ACTH 4-10 treated rats, an improvement in EDL tetanic parameters under these conditions is not evident. The analysis of changes in MU number, size and function indicate that the marked increase in the size of the regenerating MUs following nerve crush, is a compensatory mechanism for the reduction in number of functional MUs. During reinnervation, large MUs may contribute effectively to the maximum twitch tension however, large MUs do not efficiently sustain contraction amplitude during prolonged stimulation. The experiments presented here indicate that the pattern of MU reinnervation is affected differently by BIM 22015 and

ACTH 4-10. BIM 22015 influences the recovery not only of the fast (Type II) muscle fibers, but also the rare, intermediate-slow (Type I) fibers of the EDL.

The lack of statistical significance in the tests for the return of motor function is probably due to the small number of animals (5) in these groups. While these numbers were adequate for other experiments, tests of motor function have a great deal of individual variability due to the incorporation of a great many complex components such as the compensatory involvement of other leg muscles.

The effectiveness of BIM 22015 in preventing muscle atrophy and in increasing muscle fiber diameter contrasts markedly to the denervation atrophy observed in ACTH 4-10 treated muscle fibers during development and regeneration (Table 11) . The fact that these analogs have opposite effects may be due to the difference in peptide structure, with the modifications of BIM 22015 at the N-terminus (D-Ala-Gln-Tyr-Phe) responsible for positive yotrophic effects, the C-terminal fragment (Phe-Arg-Tyr-Gly) responsible for the attenuating effects. Use

When administered to a mammal (e.g. , orally, topically, transdermally, intravenously, parenterally, nasally, or by suppository) , the heptapeptides are effective in aiding in regenerating nerves of the peripheral and central nervous systems following nerve damage, and in preventing muscle degeneration or increasing muscle mass. The heptapeptides are administered beginning directly following the injury, for a period of ten days or more. Administration is daily or every other day.

The heptapeptides of the invention can be used to treat nerve crush lesions and neuropathies of alcoholic, diabetic, or toxic substance exposure origins. The heptapeptides can also be used to aid in suturing severed nerves. The heptapeptides can promote the growth of new nerve processes, enhance the connection of nerves to muscles, and prevents muscle wasting. Where the heptapeptides are used to treat nerve trauma, an initial dose may be administered immediately after nerve injury, and a second dose at a later time, e.g., two weeks, at a higher dose to preserve the muscle mass. Where muscle degeneration has already occurred, such as in degenerative neuromuscular disease, the heptapeptides may be administered at intervals, e.g., daily or weekly, in dosages sufficient to cause muscle hypertrophy.

In addition, combining the potent myotrophic effects of BIM 22015 with the neurotrophic attributes of other ACTH analogs provides a dual benefit which other neuropeptides, either alone or in combination, do not possess. This combination accords a clinical potential for the treatment of neuromuscular degenerative conditions, whether pure or "mixed", such as muscular dystrophy, progressive infantile spinal atrophy, and hypotonia.

The heptapeptides can be administered to a patient in a dosage of 0.1 μg/kg/day to 250 μg/kg/day, preferably 5-100 μg/kg/day.

Claims

Claims

1. Use of a first peptide of the formula D- Ala-Gln-Tyr-Phe-Arg-Trp-Gly-NH₂ and a neurotrophic analog or fragment of ACTH in the preparation of a medicament for the treatment of neuromuscular degeneration in a mammal.

2. The use of claim 1 wherein said ACTH fragment or analog has the formula: A¹-A²-A³-A⁴-Phe-Arg-Trp-A⁵, wherein

A¹ is H or acetyl; A² is Ala, D-Ala, Nle, or Met; A³ is Glu or Gin; A⁴ is His or Tyr; and

A⁵ is NH₂, Gly-NH₂, or D-Ala-NH₂, or a pharmaceutically acceptable salt thereof.

3. The use of claim 1 wherein said second peptide is of the formula Met-Glu-His-Phe-Arg-Trp-Gly, or a pharmaceutically acceptable salt thereof.

4. The use of claim 1 wherein said muscle degeneration is a result of motor nerve trauma.

5. The use of claim 1 wherein said muscle degeneration is a result of a neuromuscular degenerative disease.

6. The use of claim 5 wherein said disease is chosen from the group including muscular dystrophy, multiple sclerosis, progressive infantile spinal atrophy, hypotonia, and amyotrophic lateral sclerosis.

7. A therapeutic composition for treating neuromuscular degeneration comprising a myotrophic amount of a peptide of the formula D-Ala-Gln-Tyr-Phe-Arg-Trp- Gly-NH₂ and a neurotrophic analog or fragment of ACTH.

8. The composition of claim 7 wherein said ACTH fragment or analog has the formula:

A¹-A²-A³-A⁴-Phe-Arg-Trp-A⁵, wherein A¹ is H or acetyl;

A² is Ala, D-Ala, Nle, or Met;

A³ is Glu or Gin;

A⁴ is His or Tyr; and

A⁵ is NH₂, Gly-NH₂, or D-Ala-NH₂, in a pharmaceutically acceptable carrier substance.

9. The composition of claim 7 wherein said second peptide is of the the formula Met-Glu-His-Phe-Arg- Trp-Gly.

10. The composition of claim 7 wherein said composition is in the form of a pill, tablet, capsule, or sustained release tablet for oral administration to said mammal.

11. The composition of claim 7 wherein said composition is in the form of a liquid capable of being administered nasally or parenterally to said mammal.

12. The composition of claim 7 wherein said composition is in a topical or transdermal formula.

13. The composition of claim 7 wherein said composition is in the form of a sustained release tablet, surgical suture, glue, or chamber capable of being administered parenterally to said mammal.