EP0557290A4

EP0557290A4 - Treatment of aids dementia, myelopathy, peripheral neuropathy, and vision loss.

Info

Publication number: EP0557290A4
Application number: EP19910916598
Authority: EP
Inventors: Stuart A Lipton
Original assignee: Childrens Medical Center Corp
Current assignee: Childrens Medical Center Corp
Priority date: 1990-08-23
Filing date: 1991-08-23
Publication date: 1993-07-14
Also published as: JPH06500554A; EP0557290A1; WO1992003137A1

Abstract

A method of reducing damage to neurons in a human patient infected with a human immunodeficiency virus, involving administering to the patient levemopamil, or a physiologically acceptable salt thereof, in a concentration effective to cause reduction in the gp120-responsive rise in free Ca++ ion concentration in, and subsequent injury of, the neurons of the patient.

Description

TREATMENT OP AIDS DEMENTIA, MYELOPATHY, PERIPHERAL NEUROPATHY, AND VISION LOSS Background of the Invention This invention relates.to the treatment of nervous system disorders caused by infection with human immunodeficiency virus (HIV) .

HIV infection in humans causes general immunosuppression and involves other disorders, such as myelopathy, vision loss, peripheral neuropathy, and other neurological manifestations including a dementing neurological disorder, i.e., the AIDS dementia complex, the latter of which is a common and important cause of morbidity in HIV-infected patients. HIV infection has been documented in various areas of the CNS, including the cerebral cortex, spinal cord, and retina. Price et al. (1988, Science 239:586) and Ho et al. (1989, Annals in Internal Medicine 111:400) review the clinical, epidemiological, and pathological aspects of the AIDS dementia complex, and suggest that the mechanism underlying the neurological dysfunction may be indirect tissue damage by either viral- or cellular-derived toxic substances released by infected cells, such as macrophages and microglia. In addition to the previously known white matter lesions present in the brains of patients with AIDS related dementia, there is evidence that approximately 20% of the neurons are damaged or die even in the absence of opportunistic superinfections (Ketzler et al., 1990, Acta Neuropatholoqica 80:92). Pomerantz et al. (1987, New Enq. J. Med. 317:1643) document the presence of HIV type I infection of the retina in two patients with AIDS. Tenhula et al. (1990, Invest. Opthamol. Vis. Sci. 3_1:365) report the loss of axons in the optic nerve in HIV-infected patients, suggesting loss of retinal ganglion cell neurons. Brenneman et al. (1988, Nature 335:639) found that gpl20, the coat protein of HIV, killed hippocampal neurons in vitro. Summary of the Invention

The invention features a method of reducing damage to neurons in a human patient infected with a human immunodeficiency virus, by administering to the patient levemopamil, or a physiologically-acceptable salt thereof, in a concentration effective to cause a reduction in the gpl20-responsive rise in intracellular free Ca⁺⁺ ion concentration in, and subsequent death of, the neurons of that patient. By "human immunodeficiency virus" is meant all types and variants of human immunodeficiency virus (HIV) including HIV-1, HIV-2, LAV, and others. In preferred embodiments, the neurons may derive from the central or peripheral nervous system. Preferably, the blood of the infected patient contains antibodies to HIV; most preferably, the patient manifests symptoms of AIDS related complex or of acquired immune deficiency syndrome. The levemopamil most preferably is administered orally or intravenously. The method may also include administration to the patient of a second compound that is capable of reducing the gpl20-responsive rise in intracellular free Ca⁺⁺ ion concentration, in a concentration effective to cause such reduction. Most preferred second compounds are the calcium channel antagonists listed in Tables 1, 2, and 3, including nifedipine, verapamil, nitrendipine, diltiazem, Smith Kline drug no. 9512, nimodipine, a lodipine, nicardipine, flunarizine, and diproteverine. Other preferred second compounds include antagonists of the N- methyl-D-aspartate (NMDA) receptor-channel complex, including channel blockers, receptor antagonists, or agents acting at either the glycine co-agonist site or any of several modulation sites such as the Zinc site, the Magnesium site, the polyamine site, the pH-sensitive site, or the redox modulatory site. Among the preferred NMDA receptor-channel antagonists are those listed in Table 4, for example, MK-801 and D-2-amino-5- phosphonovalerate (APV) .

The invention is useful for the reduction or prevention (including prophylactic treatment) of dementia, myelopathy, peripheral neuropathy, vision loss or other neurological manifestations associated with infection by a human immunodeficiency virus. Brain uptake index (BUI) analysis indicates that levemopamil passes readily from the blood to the brain, facilitating a therapy which is both extremely rapid and unusually potent.

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 The drawings will first briefly be described.

Drawings

Fig. 1 is a dose-response curve showing retinal ganglion cell death at different gpl20 concentrations;

Fig. 2 is a graph of retinal cell survival in the presence of recombinant gpl20 and antiserum;

Fig. 3(a) is a graph of kinetics of intracellular free Ca²⁺ concentration ([Ca²⁺]i) in a retinal ganglion cell in response to various doses of gpl20, and (b) is a steady state dose-response graph of gpl20 concentration versus [Ca²⁺]i; Fig. 4 is a bar graph of gpl20-promoted rise in [Ca²⁺]i in the presence of gpl20 alone or gpl20 that has been immunoprecipitated with gpl20 antiserum. Therapy The drug levemopamil (also known as (S)-emopamil;

U.S. Patent No. 4,596,820) is a calcium channel blocker which readily passes through the blood-brain barrier (i.e., is CNS permeable). This drug may be used to treat central nervous system disorders, such as myelopathy, vision loss, and dementia, associated with HIV-I infection.

Levemopamil may be administered by any one of a number of routes. For example, because levemopamil is CNS-permeable, administration may be oral or intravenous, at a suggested divided dosage of 0.1 to 50 mg/kg body weight, two or three times daily. Less preferably, administration may be intrathecally or intravitreally, at a suggested dosage of 0.1 to 50 mg/kg body weight, two or three times daily. If desired, a second Ca++ channel antagonist may be administered to a patient in conjunction with levemopamil. Dosage of preferred second compounds is as follows. Nimodipine or Smith Kline drug no. 9512, nicardipine, flunarizine, or diproteverine may be administered orally or intravenously, at a suggested divided dosage of 60 to 1200 mg/day. MK-801 may be used at a daily dose of 0.01 to 5.0 mg/kg. Nifedipine is administered at a dosage of 20 to 800 mg/day; verapamil at a dosage of 80 to 720 mg/day; and diltiazem at a dosage of 60 to 960 mg/day. Each of these compounds is preferably administered intrathecally or intravitreally. The effective dose of other suggested second compounds (e.g., those in Tables 1-4) will range from 0.01 to 1000 mg/kg. Following is a detailed description indicating that V-I infection is detrimental to cells associated with the nervous system; such nerve cell damage is the result of a gpl20 protein-mediated increase in intracellular Ca⁺⁺. The detailed description also outlines assays which provide one skilled in the art with the necessary guidance to determine levemopamil's maximum efficacy level. In Vitro Neuronal Cell Death Assay Neuronal cell death may be assayed by incubating retinal ganglion cells in vitro with purified native or recombinant gpl20 and scoring live cells. The ability of the levemopamil to reduce neuronal cell death is determined by scoring live cells which have been incubated overnight with both gpl20 and the drug.

Retinal ganglion cells from postnatal rats are identified and their viability ascertained as follows. Under general anesthesia, the fluorescent dye granular blue (Mackromolekulare Chemic, Umstadt, FRG) is injected as approximately a 2% (w/v) suspension in saline into the superior colliculus of 4- to 7-day-old Long Evans rats (Charles River Laboratory, Wilmington, MA) . Two to 7 days later, the animals are killed by decapitation and enucleated, and the retinas quickly removed. The retinas are then dissociated and cultured in Eagle's minimum essential medium (MEM, catalog #1090, Gi .no Grand Island, NY), supplemented with 0.7% (w/v) methylcellulose, 2 mM glutamine, 1 μg/ml gentamicin, 16mM dextrose, and 5%(v/v) rat serum, as described in Lipton et al., J. Physiol. 385:361. 1987. The cells are plated onto 75 mm² glass coverslips coated with poly-L-lysine in 35 mm tissue culture disL.es; gpl20 is then added. Sibling cultures receive various doses of levemopamil, with and without gpl20. Cell survival is assayed after one day in culture. Incubations last 20-24 h at 37°C in an atmosphere of 5% C0₂/95% air. Ganglion cells can be unequivocally identified by the continued presence of the fluorescent blue dye. The ability of retinal ganglion cells to take up and cleave fluorescein diacetate to fluorescein is used as an index of their viability as described in detail in Hahn et al., 1988, Proc. Natl. Acad. Sci. USA J5:6556). Dye uptake and cleavage correlates well with normal electrophysiological properties assayed with patch electrodes.

To perform the viability test, the cell-culture medium is exchanged for physiological saline containing 0.0005% fluorescein diacetate for 15-45 s, and then cells are rinsed in saline. Retinal ganglion cells that do not contain the fluorescein dye (and thus were not living) often remain visible under both phase-contrast and UV fluorescence optics, the latter because of the continued presence of the marker dye granular blue; other dead retinal ganglion cells disintegrate and only debris remains. In contrast, the viable retinal ganglion cells display not only a blue color in the UV light but also a yellow-green fluorescence with filters appropriate for fluorescein. Thus, the use of two exchangeable fluorescence filter sets permits the rapid determination of viable ganglion cells in the cultures, which are found as solitary neurons or lying among other cells in small clusters (usually in the ratio of approximately 1:10 solitary to clustered) . Gpl20 is produced and purified as follows. Two native isolates of gpl20, RF2 and 3B, (Brenneman et al., 1988, supra; Nature 335:445; National Cancer Institute, Frederick, MD) are purified by immunoaffinity chromatography, as described by Robey et al. (1986, Proc. Nat. Acad. Sci. USA 83:7023). Pyle et al. (1988, AIDS Re^ Hum. Retrovir. 3_:387), and Kornfeld et al. (1988, Nature 335:4-5) In addition, reco binant gpl20 derived from the gene encoding gpl20-3B, is obtained as follows.

Recombinant gpl20 is produced by transfection of a Chinese Hamster Ovary (CHO). cell line (ATCC) with a plasmid containing isolate 3B envelope coding sequences encoding amino acids 61-531 (Lasky et al., 1986, Science 233:209) . The gene is truncated to remove the native amino-terminal signal sequence and the carboxy-terminal transmembrane domain, and then ligated in-frame to the herpes simplex virus glycoprotein-D signal sequence (Ber an et al., 1985, Science 227:1490) to allow the envelope protein to be constitutively secreted by the CHO cell line. Production in a mammalian cell ensures that the envelope protein is glycosylated. This envelope glycoprotein, rgpl20-3B, is purified by immunoaffinity chromatography to 5 parts in a million (99.995%) pure based on estimates from polyacrylamide gel electrophoresis and Western blotting. The preparations of gpl20 (at low concentrations) are highly labile in that they have to be freshly thawed (with refreezing avoided) in order to display activity. qpl20 Increases Neuronal Cell Death In Vitro

Neuronal cell death is assayed by incubating retinal ganglion cells in vitro with purified native or recombinant gpl20 and scoring live cells.

Fig. 1 is a dose-response curve for concentrations of gpl20 ranging from 10^"9 M to less than 10~¹³ M, and shows that incubation of native purified gpl20 (RF2) or recombinant gpl20 preparations with cultured retinal cells resulted in the death of a significant number of ganglion cells within 24 h. A significant increase (P<0.01) in cell death was observed at gpl20 concentrations above 2 x 10~¹²M. Living cells were scored without knowledge of treatment in each culture. Control cultures had counts ranging from 95 to 250 retinal ganglion cells per 75 mm coverslip, which corresponded to approximately 1% of the total retinal cell population. Each point was determined in at least triplicate in each experiment, and the data represented 3 experiments. Values shown are means +/- their standard errors. Statistical analysis consisted of a one-way analysis of variance followed by a Scheffe multiple comparison of means. These results indicate that neuronal cell death in this experiment is due to the addition of gpl20. To confirm this result, a control experiment is done in which neuronal cell death is prevented by the addition of antiserum specific for the gpl20 envelope protein. qpl20 Antiserum Prevents qpl20-Associated Neuronal Cell Death

Goat anti-gpl20 immune sera is prepared by primary and secondary intramuscular immunizations of recombinant fragments of the 3B isolate of gpl20 plus Freund's adjuvant. Prior to these injections, preimmune serum is collected from each goat. The presence of gpl20 antibodies in the postimmune serum is verified by polyacrylamide gel electrophoresis of immunopurified gpl20 and Western blotting. Fig. 2 shows that anti-gpl20 antisera prevented neuronal cell killing of rat retinal ganglion cells by recombinant envelope protein (rgpl20-3B) . Retinal cultures from 8 day-old rats were plated in the presence or absence of 20 pM gpl20 plus either postimmune serum containing polyclonal gpl20 antibodies from goat G1084, or preimmune serum from the same goat. Sera were used at a concentration of 1:500 diluted in control growth medium. Fig. 2 represents data from 5 separate assays; error bars represent standard error of the mean. Statistical analysis was performed as described above. These results show that anti-gpl20 neutralizes the effect of purified preparation of gpl20; that is, cell death in cultures treated with 20 pM gpl20 can be prevented by the addition of goat anti-gpl20 serum to the culture medium. Neither the antibody preparation nor its preimmune serum alone significantly affected survival compared to that observed in control culture medium. In contrast, the addition of gpl20 or gpl20 plus preimmune serum resulted in significant killing of retinal ganglion cells (P<0.01). Furthermore, compared to cultures treated with gpl20 alone or gpl20 plus preimmune serum, the anti-gpl20 serum saved a significant proportion of retinal ganglion cells exposed to gpl20 (P<0.01). Treatment with antiserum G1084 in the presence of gpl20 resulted in neuronal cell counts that were not statistically different from cultures incubated in control medium alone. These figures present results for both solitary and clustered cells in culture; however, the most germane results are those for clustered cells, since neurons are interconnected in the brain.

Measurement of Intracellular Ca²⁺

The concentration of intracellular free Ca ([Ca²⁺]i) is measured in postnatal rat retinal ganglion cells and hippocampal neurons by digital imaging microscopy with the Ca²⁺ sensitive fluorescent dye fura 2, as follows. Retinal ganglion cells are cultured from 1-to 2-week-old Long Evans rats as described (Leifer, et al. , Science 224: 303, 1984; Lipton et al., J. Physiol. 385:361. 1987). The ganglion cells are identified by the presence of retrogradely transported red fluorescent dyes (Dil or rhodamine-labeled microspheres) or by morphological and immunofluorescence criteria (e.g., the large or α-like ganglion cells which stain with neurofilament antibodies; Drager et al., Nature 309:624,. 1984) . During Ca²⁺ measurements, unless otherwise stated the fluid bathing the neurons consists of Hanks' balanced salts: 136.7 M NaCl, 1 mM NaHC0₃, 0.34 mM Na₂HP0₄, 0.44mM KH₂P0₄, 5.36 mM KC1, 2.5 mM CaCl₂, 0.5 mM MgS0₄, 0.5 mM MgCl₂, 5 mM Hepes NaOH, 22.2 mM glucose, and phenol red indicator (0.001% v/v) ; pH 7.2. Coat protein gpl20 and other substances are usually applied to the neurons by pressure ejection after dilution in this bath solution. High K⁺ solutions are prepared by substituting KC1 for NaCl. Neuronal [Ca²⁺]i is analyzed with fura 2-acetoxy-methyl ester (AM) as described [Grynkiewicz, et al., J. Biol. Chem. 260:3440 (1985); Williams et al., Nature 318:558 (1985); Connor et al., J. Neurosci. :1384 (1987); Connor et al. , Science 240:649 (1988); Cohan et al., J. Neurosci. 7:3588 (1987); Mattson, et al., ibid, :3728 (1989)]. After adding Eagle's minimum essential medium containing 10 μM fura 2-AM to retinal or hippocampal cell neurons, the cultures are incubated at 37°C in a 5% C0₂/95% air humidified chamber and then rinsed. The dye is loaded, trapped, and deesterified within 1 hour, as determined by stable fluorescence ratios and the effect of the Ca ionophore ionomycin on [Ca²⁺]i is measured. During Ca²⁺ imaging, the cells are incubated in a solution of Hepes-buffered saline with Hanks' balanced salts. The [Ca²⁺]i is calculated from ratio images that are obtained by measuring the fluorescence at 500 n that is excited by 350 and 380 nm With a DAGE MTI 66 SIT or QUANTEX QX-100 Intensified CCD camera mounted on a Zeiss Axiovert 35 microscope. Exposure time for each picture is 500 ms. Analysis is performed with a Quantex (Sunnyvale, CA) QX7-210 image-processing system. Since cells are exposed to ultraviolet light only during data collection (generally less than a total of 20 s per cell) , bleaching of fura 2 is minimal. qpl20 Affects the Intracellular Concentrati •on of Ca2+

Gpl20-promoted neurotoxicity was shown to involve an increase in intracellular Ca concentration in the following experiment. Intracellular Ca²⁺ was measured as described above. Application of 200 pM of highly purified gpl20 from a recombinant source, as described above, produced a striking increase in [Ca²⁺]i (Fig. 3). Compared to control levels (Ca²⁺ = 63 + 4 nM, mean + SEM, n = 42) obtained before the addition of coat protein, levels increased 33-fold within 7 min of gpl20 application (2100 + 330 nM, n = 10; range of values 934 to 3943 nM) . Other preparations of gpl20 purified from natural isolates (RF2 and 3B) yielded similar results. All experiments shown here used the highly purified recombinant gpl20. Similar effects are seen when gpl20 was applied to hippocampal neurons. Hippocampal cortices of embryonic day 18 CD rats was dissociated with trypsin (0.027% w/v) and plated at a density of 600,000 cells per 35-mm culture dish, each dish containing five poly-L-lysine-coated glass coverslips. Growth medium (Rosenberg et al, Neurosci. Lett.. 103:162. 1989) was changed three times per week. In these experiments, Ca²⁺ measurements were made after 14 to 21 days in culture. Overall, 200 pM gpl20 produced an increase in [Ca ]i in 76% of the neurons tested (n = 75) .

Several experiments indicate that gpl20 was responsible for this rise in [Ca²⁺]i. Application of normal bathing medium did not produce a change in [Ca 2+]i, although subsequent addition of gpl20 to the same retinal ganglion cell neurons increased [Ca²⁺]i to -2μM (n = 10). Treatment of gpl20 with trypsin followed by neutralization with soybean trypsin inhibitor (Sigma Chemical Co. , St. Louis, MO.) resulted in a preparation that was no longer active in increasing [Ca²⁺]i. As the recombinant gpl20 was produced from a construct with the herpes simplex virus glycoprotein D signal sequence, a glycoprotein D control, made in exactly the same medium as recombinant gpl20, was applied to retinal ganglion cells. Glycoprotein D exerted either no effect or resulted in a modest increase in [Ca ⁺]i (<200 nM, n = 6), but never to the micromolar level typically observed after the addition of equimolar gpl20.

Pressure ejection of 50 mM KC1 on these neurons yielded an increase in [Ca ⁺]i to the range of 600 nM (600 + 99 nM, n = 5) . Brief exposures to KC1 for 30 s to 3 min produced these levels of Ca , which peaked within 1.5 min of the beginning of the addition and recovered to levels of -250 nM over the next few minutes (240 + 21 nM, n = 5) . In contrast, at least a 1 min addition of 200 pM gpl20 was necessary to produce any rise in [Ca²⁺]i, and the effect was persistent and irreversible during the course of an experiment on a single cell (10 to 30 min of [Ca ]i monitoring, with measurements every 30 s) . The time course of the change in [Ca ]i evoked by a 3 min application of various doses of gpl20 was shown in Fig. 3A. The peak level was reached -7 min after the beginning of the addition. Thus, there are both qualitative and quantitative differences in the observed increase in [Ca ]i in response to K⁺ as opposed to gpl20. Extremely low doses of gpl20, in the picomolar range, were effective in increasing [Ca ]i in a graded, dose-dependent fashion (Fig. 3B) . As little as 20 pM gpl20 produced increases in [Ca²⁺]i. Very high levels of free Ca (>2 μM) were obtained with concentrations of gpl20 at or above 200 pM.

The increases in [Ca²⁺]i observed with gpl20 may have been caused by a contaminant in the purified preparation of the viral envelope protein, although this seems unlikely with the highly purified recombinant gpl20. As additional confirmation, immunoprecipitation experiments with goat antibody to gpl20 (anti-gpl20) coupled to protein A-coated Sepharose beads was performed as follows.

Immunoprecipitation of gpl20 was performed as described above with some modifications. A 1:100 or

1:500 dilution of anti-gpl20 or preimmune serum from the same goat, was bound to protein-A-coated Sepharose beads, washed, and incubated with a solution containing 7 nM gpl20 for 18 hours at 4°C; this was followed by centrifugation. The supernatant of the material treated with preimmune serum had gpl20 activity, as evidenced by immunoblotting and by increased [Ca²⁺]i and cell death (after a dilution of 1:350 to ~20 pM) ; t e material exposed to anti-gpl20 had little or no activity. One of three such experiments is shown in Fig. 4.

Treatment with preimmune serum did not significantly alter the ability of gpl20 to increase [Ca ⁺]i (compare striped and grey bars; although the mean [Ca ]i was greater after application of gpl20 treated with preimmune serum than after gpl20 alone, this difference did not reach statistical significance) . Both gpl20 and preimmune serum-treated gpl20 produced a significant increase in [Ca²⁺]i, compared to the control (P <0.01, analysis of variance (ANOVA) followed by Scheffe multiple comparison of means; significance indicated by an asterisk) . In contrast, immunoprecipitation with postimmune serum containing anti-gpl20 completely abrogated the gpl20 effect (last column on right of Fig. 4). Levemopamil's Effect on qpl20-Mediated Neuronal Cell Death

The effects of levemopamil on gp120-mediated neuronal cell death and its ability to antagonize calcium channels may be tested as follows. Cultures of retinal ganglion cells receive gpl20 (20 pM) and/or levemopamil (10 μM) at the time of plating (as described above) , and neuronal cell survival is assayed one day later (also as described above) . Each experiment is reproduced in 4 replicate tissue culture dishes and a mean value calculated. Survival of clustered retinal ganglion cells in the control group (i.e., those treated with gpl20 only) is compared to those treated with gpl20 plus levemopamil. Preferably this experiment is repeated a number of times. An increase in retinal ganglion cell survival with gpl20 plus levemopamil treatment in comparison to gpl20 alone (preferably, including the 3B and RF2 natural isolates and the recombinant 3B form) indicates that levemopamil is effective in reducing gpl20-mediated neuronal cell damage. Evidence from experiments involving other calcium channel-blocking drugs (e.g., nifedipine) indicates that high concentrations of such drugs induce significant neuronal cell killing on their own (see WO90/11761, hereby incorporated by reference) . One possible explanation for these phenomena invokes the hypothesis that there is an opti .mal level of i .ntracellular Ca 2+ necessary for neuronal health and welfare. Too little [Ca²⁺]i may inhibit survival while too much [Ca²⁺]i may also lead to cell death. In between these two extremes, survival may be enhanced. According to another theory, the drug alone may be toxic to neurons for some other, unrelated, reason although in that case it would be difficult to explain the finding that survival in the case of nifedipine was slightly better with the combination of gpl20 plus nifedipine compared to nifedipine alone.

The effect of using lower doses of levemopamil on neuronal survival following gpl20 treatment may be tested as follows. Treated cultures receive gpl20-3B (20 pM) and/or levemopamil at a range of concentrations (e.g., between 10 nM and 1 μM) at the time of plating, and retinal ganglion cell survival is assayed one day later. By fine tuning the dose-response curve, it is possible to find an optimal level of levemopamil that produces minimal death on its own and yet substantially blocks the toxicity mediated by gpl20. Evidence from equivalent experiments involving other calcium channel antagonists, such as nifedipine and nimodipine, indicate an optimal drug concentration of 0.1 μM; this concentration is suggested as an optimal levemopamil concentration as well.

Levemopamil's Effect on the qpl20-Mediated Increase in Intracellular f^Ca2+1

To test levemopamil's effect on the intracellular concentration of free Ca²⁺ ions in retinal ganglion cells incubated with gpl20, retinal ganglion cells are loaded with fura 2, as described above. The viral envelope protein ^*pl20 (200 pM) is then applied by puffer pipette to neurons previously bathed in normal medium or in medium containing levemopamil (e.g., 100 nM to 10 μM) for several min. Intracellular Ca concentration is measured as described above. gpl20 produces an increase in [Ca ]i (see above) . One measure of levemopamil efficacy would be evidenced by a partial or total block of such an increase in the intracellular Ca concentration; however, the best measure of efficacy is prevention of neuronal cell injury in the face of HIV-related insult. Levemopamil's Effect on Current Flow Through Calcium Channels in the Presence of qpl20 The following assay of neuronal cell function tests the effect of a calcium channel antagonist such as levemopamil on Ca²⁺ ion flow through Ca²⁺ channels. Without being bound to any theory as to the mechanism whereby gpl20 increases cell death, it is possible that gpl20 increases current flow across Ca channels. As a precautionary measure in screening for a compound capable of reducing the gpl20-associated rise in intracellular Ca²⁺ concentration, the following assay of Ca²⁺ current should be performed in the presence of gpl20. Ca²⁺ currents are measured in the presence of gpl20 in retinal ganglion cells in the presence or absence of levemopamil (e.g., 100 nM - 10 μM) . Alternatively, current carried by Ba through calcium channels is measured during the application of 20 pM gpl20-RF2 with or without levemopamil (e.g., 100 nM - 10 μM) . All traces are obtained in the presence of gpl20. A particularly sensitive assay is accomplished using a voltage step e.g., V_c = -10 mV) initiated from a depolarized holding potential (V_H = -40 V) (Karschin and Lipton, 1989, J. Phvsiol. 418:379). Concentration ranges for antagonism of Ca²⁺ current with levemopamil may be experimentally determined in this way, starting with an effective concentration range of 100 nM -10 μM. Other Neurons are Sensitive to gpl20 Another important consideration is whether or not Ca ⁺ levels in mammalian central neurons other than retinal ganglion cells will be sensitive to gpl20 and, therefore, amenable to treatment with a calcium channel blocker like levemopamil. Miller (Science 235:46. 1987) suggests that dihydropyridines affect Ca influx, at least to some degree, in over 90% of neurons from various brain areas; there were, however, regional differences in the effectiveness of these drugs. Other classes of calcium channel antagonists, such as phenylalkylamines, may prove effective in different areas of the brain. For example, the Ca ⁺ current in hippocampal neurons has been shown to be partially suppressed by verapamil (100 μM) (Yaari et al., 1987, Science 235:680) ; in addition, novel calcium channel blockers or G proteins and intracellular messengers that affect their efficacy, may prove useful in this regard (Olivera et al., 1985, Science 230:338, Dolphin et al., 1987 J. Phvsiol. 386:1: and Yaari et al., 1987, Science 238:1288) . Miller (supra) showed that hippocampal neurons were more sensitive to calcium channel antagonists than were striatal neurons. This v™iability can probably be attributed to the fact that c _y a prolonged component of Ca ⁺ current (similar to L-type current) is sensitive to dihydropyridines and probably levemopamil.

Administration ,<f a Second Ca++ Channel or NMDA-Receptor Channel Antagonists

If desired, a second compound capable of either antagonizing Ca²⁺ channels or antagonizing NMDA-receptor channels (or their effects) may be administered to a patient in conjunction with levemopamil. Preferred calcium channel antagonists are listed in Tables l, 2 and 3. TABLE 1

Antagonists of the Voltage Dependent Calcium Channel (General Classes) dihydropyridines (e.g., nimodipine) phenylalkylamines (e.g., verapamil, D-600, D-888) benzothiazepines (e.g., diltiazem and others) bepridil and related drugs diphenylbutylpiperdines diphenylpiperazines (e.g., flunarizine/cinnarizine series)

HOE 166 and related drugs fluspirilene and related drugs toxins and natural compounds (e.g., snail toxins - ωconotoxin GVIA and GVIIA, maitotoxin, taicatoxin, tetrandine, hololena toxin, plectreurys toxin, funnel-web spider venom and its toxin fraction

TABLE 2

DIHYDROPYRIDINE CALCIUM CHANNEL ANTAGONISTS nifedipine KW3049 niludipine oxodipine

PY108-068 (darodipine) CD349 mesudipine TC81

GX 1048 YM-09730-5 or (4S)DHP floridine MDL72567 nitrendipine Rθlδ-3981 nisoldipine DHP-218 ni odipine nilvadipine nicardipine a lodipine felodipine 8363-S

PN200-110 (Isradipine) iodipine

CV4093 azidopine

TABLE 3

OTHER CALCIUM CHANNEL ANTAGONISTS

diclofurime D-600 pimozide D-888 prenylamine Smith Kline 9512 fendiline ranolzine perhexiline lidoflazine mioflazine CERM-11956 flunarizine/cinnarizine R-58735 series R-56865 verapamil amiloride dilfiazine phenytoin dipropervine thioridazine tricyclic antidepressents

Preferred second compounds include the following drugs, of which the most preferred are those that are capable of crossing the blood-brain barrier, for example, nimodipine (Miles Pharmaceuticals, West Haven, CT) , Smith Kline drug no. 9512 (Smith Kline, French-Beecham, Philadelphia, PA) , diproteverine (Smith, Kline, French-Beecham) , and flunarizine. Less preferred antagonists are those that are less CNS permeable, for example, verapamil (Calan, G.D. Searle & Co., Chicago, 111.; Isoptin, Knoll, Whippany, NJ) , nitrendipine, diltiazem (Cardize , Marion, Kansas City, MO) , and nifedipine, U.S. Patent 3,485,847, hereby incorporated by reference (Procardia, Pfizer, NY, NY; Adalat, Miles) . Other Ca²⁺ channel antagonists which may be useful are mioflazine, flunarizine, bepridil, lidoflazine, CERM-196, R-58735, R-56865, Ranolazine, Nisoldipine, Nicardipine, PN200-110, Felodipine, Amlodipine, R-(-)-202-791, and R-(+)-Bay-K-8644 (Miles, Bayer), whose chemical formulae are described in Boddeke et al., Trends in Pharmacologic Sciences, 1989, .10:397 (hereby incorporated by reference) and Triggle et al. , Trends in Pharmacologic Sciences (1989) 10:370. Various calcium channel antagonists have been identified (Hosey, J. Membrane Bio. 104:81. 1988; Ohtsuka et al., General Pharmacology 2_0:539, 1989; Greenberg, Annals Neurol. 21:317, 1987; and Lin et al. , Proc. Natl. Acad. Sci. USA jT7:4538, 1990; hereby incorporated by reference) .

For any second calcium channel antagonist compound (e.g., those described above), effectiveness in preventing neurological disorders associated with HIV-1 (or other HIV) infection is determined by screening the drug using one or more of the following assays of neuronal cell function: neuronal cell death, detection of intracellular free Ca ion concentration in neurons, and detection of current flow through Ca²⁺ channels (see above) .

Any suitable antagonist of the N-methyl-D- aspartate (NMDA) subtype of glutamate receptor-channel complex may also be used as a second compound in conjunction with levemopamil. Preferred antagonists are listed in Table 4. The antagonist can be a channel blocker, a receptor antagonist, or act at the glycine co- agonist site or at any of several modulation sites such as the Zinc site, the Magnesium site, the polyamine site, the pH sensitive site, or the redox modulatory site (see, e.g., Table 4). Many antagonists of the NMDA receptor have been identified (Watkins et al. , Trends in Pharmacological Sci. 11:25, 1990, hereby incorporated by reference) . Other substances that interact with the NMDA receptor-channel complex in a manner that attenuates the gpl20-responsive rise in intracellular Ca ion concentration may also be used in the method of the invention. Such modulatory substances include those known to cause oxidation of the redox site of the NMDA receptor (U.S.S.N. 391,778, filed August 9, 1989, and incorporated herein by reference) , and oxidized or reduced glutathione.

TABLE 4

NMDA Receptor-Channel Complex Antagonists

MK-801 (dizocilpine) and newer derivatives of this

Dibenzocycloheptene (non-competitive, open-channel blocker) (Merck)

Dextrorphan, dextromethorphan and derivatives or morphinans (noncompetitive NMDA receptor antagonist) (Hoffman La-Roche)

Phencyclidine (PCP) and derivatives and pyrazine compounds.

Hoechst 831917189 (non-competitive NMDA antagonist) Kynurenate (antagonist at the glycine co-agonist site and elsewhere) . 7-chloro-kynurenate, 5,7-chloro-kynurenate, and derivatives

(antagonist at the glycine co-agonist site) (Merck) k-opioid receptor agonist U50488H (Upjohn) Adenosine and derivatives (e.g., cyclohexyladenosine - to reduce presynaptic glutamate release) .

MDL 27,266 (Merrell Dow) and triazole-one derivatives. Monosialogangliosides (e.g., GM1 of Fidia Corp.) (to inhibit Protein Kinase C translocation involved in NMDA receptor-mediated neurotoxicity) . CGS-19755 and other piperidine derivatives (CIBA-GEIGY) D-2-amino-5-phosphonovalerate (NMDA recertor antagonist) D-2-amino-7-phosphonoheptanoate (AP7, selective NMDA receptor antagonist) . CPP [3-(2-Caroxypiperazin-4-y)propyl-l-phosphonic acid], a selective NMDA receptor antagonist that is a rigid analog of AP7. Indole-2-carboxylic acid (competitive antagonist of potentiation at glycine co-agonist site of NMDA receptor) . DNQX (6,7-dichloroquinoxaline-2-3-dione and other

Quinoxaline or Oxadiazole derivatives including CNQX) (DNQX is a specific glycine co-agonist site antagonist of the NMDA receptor) . Ketamine (non-competitive open channel blocker of NMDA receptor-operated channels) .

O-phosphohomoserine (an NMDA antagonist) Tiletamine and other cyclohexane derivatives (non¬ competitive NMDA antagonists) Arcaine or related biguanidines and biogenic polyamines (antagonist of the polyamine modulatory site)

Ifenprodil and related drugs (antagonist of polyamine site)

Diethylenetriamine (antagonist of the polyamine site) 1,10-Diaminodecane (inverse agonist of the polyamine site) 21-aminosteroid (lazaroids) such as U74500A, U75412E and U74006F (Upjohn) (interfere with events linking the NMDA receptor to overstimulation by calcium)

Oxidized and reduced glutathione (to block NMDA-receptor mediated neurotoxicity) Memantine, Amantadine, Rimantadine, and Derivatives (non- competitive use-dependent NMDA antagonists)

Generally an antagonist may be tested for utility in the method of the invention using various types of neuronal cells from the central nervous system and the assays described herein, as long as the cell can be isolated intact using conventional techniques. Retinal cultures were used in the assays described herein (but hippocampal cortex neurons have also been used, e.g., in assays of neuronal death and intracellular calcium) , because retinal cells can be produced from postnatal mammals, are well-characterized, and contain a central neuron, the retinal ganglion cell, that can be unequivocally identified with fluorescent labels. A substantial portion of retinal ganglion cells in culture display both functional synaptic activity and bear many, if not all, of the neurotransmitter receptors found in the intact retina and brain. An effective antagonist will cause a decrease in HIV-1-associated neuronal cell damage or death, and will prevent the rise in intracellular Ca²⁺ ion concentration that occurs in the presence of gpl20. In addition, an effective antagonist will decrease Ca⁺⁺ ion influx through neuronal calcium channels to a degree sufficient to reduce neuronal cell death, while not completely blocking Ca⁺⁺ ion influx, an event which itself might kill neuronal cells. The antagonist may be compounded into a pharmaceutical preparation, using pharmaceutical compounds well-known in the art; the exact formulation of the antagonist compound depends upon the route of administration. Other Embodiments Other embodiments are within the following claims. For example, the method of the invention may be used for treatment of dementia, myelopathy, peripheral neuropathy, or vision loss associated with infection by a human immunodeficiency virus. The method can be used whether or not the patient manifests symptoms of the AIDS related complex or AIDS itself, and thus the method of the invention may be used as a prophylactic treatment for damage to CNS neurons, after HIV infection. In the method of the invention, levemopamil may be administered by any means that allows the compound access to the central nervous system. Preferably, levemopamil is administered orally or intravenously; alternatively, it may be administered intrathecally to the brain and/or spinal cord, or intravitreally to the retina. The method of the invention also includes administering to a single patient levemopamil and one or more of the useful compounds from multiple categories of modulators of intracellular Ca ⁺ concentration (e.g., those listed in Tables 1-4) .

Claims

Claims 1. A method of reducing damage to neurons in a human patient infected with a human immunodeficiency virus, comprising ad inistering levemopamil, or a physiologically acceptable salt thereof, to said patient in a concentration effective to cause a reduction in the gpl20-responsive rise in free intracellular Ca++ ion concentration in, and subsequent injury of, said neurons of said patient.

2. The method of claim 1, wherein said neurons derive from the central nervous system.

3. The method of claim 1, wherein said neurons derive from the peripheral nervous system.

4. The method of claim 1 wherein the blood of said human patient contains antibodies to HIV.

5. The method of claim 1 wherein said human patient manifests symptoms of the AIDS related complex or acquired immunodeficiency syndrome.

6. The method of claim 1, said compound being administered to said patient orally or intravenously.

7. The method of claim 1, further comprising administering to said patient a second compound capable of reducing the gpl20-responsive rise in free Ca++ ion concentration in, or injury of, said neurons of said patient, in a concentration effective to cause such reduction.

8. The method of claim 7, wherein said second compound is one of those listed in Tables 1, 2 or 3. 9. The method of claim 7, said compound being one or more of nifedipine, verapamil, netrendipine, diltiazem, Smith Kline drug no. 9512, nimodipine, amlodipine, nicardipine, flunarizine, or diproteverine.

10. The method of claim 1, further comprising administering to said patient a second compound which is an antagonist of the NMDA receptor-channel complex, in a concentration effective to cause such antagonism.

11. The method of claim 10, wherein said second compound is one of those listed in Table 4.

12. The method of claim 10, wherein said second compound is 2-amino-5-phosphonovalerate or MK-801.