Search

The three-dimensional structure of the immunosuppressive cyclic peptide cyclosporine (Cs), determined in crystal by X-ray analysis and in solution in aprotic solvents by n.m.r., differs mainly by the orientation of the 7 carbon side chain of residue 1. Because of its poor solubility in water, the conformation of Cs in aqueous medium cannot be studied by n.m.r. methods, which require concentrations of the substance of the order of milligram/mL but can be analyzed by immunochemical methods in which concentrations in the nanogram/mL range are detected. In the present study, the ability of a series of monoclonal antibodies (McAbs) raised against Cs to recognize different parts of residue 1 of Cs was determined from the cross-reactivity of different Cs-analogues modified in residue 1. The results show that when Cs is dissolved in aqueous buffer, the terminal atoms of residue 1 side chain are not available for binding to antibodies recognizing the face of the molecule defined by residues 1, 2, 3, 10, 11, suggesting that the chain is probably folded back under the molecule, as observed in the crystal structure. Binding of McAbs to Cs was also affected by conformational modifications of the peptide ring that occur in some Cs-analogues. The results illustrate the potential of McAbs for probing the conformation of Cs-derivatives for which no structural data are available.

The introduction of highly active antiretroviral therapy has led to significant changes in disease progression and mortality of human immunodeficiency virus (HIV) type 1 (HIV-1) infection and its sequela, AIDS. HIV-1 infection has been turned into a chronic condition that can be treated and that may be manageable over many years (90). However, a significant proportion of patients still fail to have a complete response to treatment over a prolonged time and are at risk of virological rebound, which may lead to the emergence of drug-resistant virus variants (20). The 21 anti-HIV-1 drugs used at present target only three steps in the viral replication cycle, namely, virus fusion, reverse transcription, and the proteolytic processing of viral proteins (22). An inhibitor directed against a fourth step, envelope glycoprotein binding to the CCR5 coreceptor (24, 99), and an inhibitor of a fifth step, integration of viral DNA in the cellular genome (34), were recently approved by FDA. Further inhibitors of the last two steps are in clinical evaluation (23, 86). Thus, the four virus-coded proteins, the envelope glycoprotein, reverse transcriptase (RT), protease, and integrase, are the molecular targets of anti-HIV-1 chemotherapy in the clinic at present. This limits the number of possible combinations that may be used, because cross-resistance of virus strains against drugs targeted to the same viral molecule is common (61). Therefore, novel anti-HIV drugs directed against other steps in the viral replication cycle are highly needed. Cyclophilin A (CypA) was established more than a decade ago to be a valid target in anti-HIV-1 chemotherapy (69, 70). The cellular protein CypA fulfills an essential function early in the HIV-1 replication cycle. It was found to bind specifically to the HIV-1 Gag polyprotein (48). A defined amino acid sequence around G89 and P90 of capsid protein p24 (CA) was identified as the binding site for CypA (17, 31). The affinity of CypA for CA promotes the incorporation of CypA into the virion particles during assembly (10, 30, 87). Experimental evidence indicates that the CypA-CA interaction is essential for HIV-1 replication; inhibition of this interaction impairs HIV-1 replication in human cells (10, 37, 81, 82). The step in the viral replication cycle where CypA is involved was demonstrated to be an event after penetration of the virus particle and before integration of the double-stranded viral DNA into the cellular genome (13, 51, 82). CypA is a member of the immunophilin class of proteins. These ubiquitous cellular proteins possess cis-trans prolyl isomerase (PPIase) activities (27) and are assumed to be involved in protein folding and to function as chaperones in intracellular transport (78). Cyclophilins are also known to be the intracellular receptor molecules for cyclosporines (35), a class of cyclic undecapeptides produced by Trichoderma polysporum (25, 71). Binding of cyclosporines to cyclophilins leads to the blockade of the isomerase activity. The most prominent representative of this class of compounds is cyclosporine (Cs), a potent inhibitor of T-cell activation widely used in the clinic as an immunosuppressant in organ transplantation (11). When Cs is bound to CypA it forms a ternary complex with calcineurin. This binding inhibits the phosphatase activity of calcineurin, which is crucial for signal transduction in the activation cascade of T cells. Inhibition of calcineurin function is thus the molecular basis of the immunosuppressive action of Cs (8). The structure of Cs bound to its ligands has been elucidated: two separate domains in the undecapeptide Cs that are involved in binding to CypA and calcineurin, respectively, can be distinguished (40, 43, 66, 94, 98). The immunosuppressive capacity of Cs can therefore be separated from its affinity to CypA by chemical modification. The anti-HIV-1 activity of Cs was first reported in 1988 (91). Evaluation of this drug and many derivatives for inhibition of HIV-1 replication revealed that nonimmunosuppressive Cs analogs had anti-HIV-1 activities equal to or even superior to those of immunosuppressive analogs (7, 10). A clear correlation was evident, however, between their antiviral activities and the ability of cyclosporines to bind to their cellular receptor protein, CypA. Modification of amino acids in the calcineurin binding domain of Cs, in particular, substitution of the undecapeptide at position 4, was shown to abolish the formation of the ternary complex with calcineurin and, thus, led to the loss of its immunosuppressive capacity (62, 96, 105). Binding to CypA and anti-HIV-1 activity were not impaired by such modifications. (Methyl Ile4)cyclosporine (NIM811), a nonimmunosuppressive cyclosporine with a high affinity for binding to CypA and also potent and selective anti-HIV-1 activity, was characterized in more detail (68). Mode-of-action studies revealed that NIM811 inhibited the CypA-CA interaction in a dose-dependent manner (10, 87) and that an early step in the replication cycle of HIV-1 was impaired in the presence of the compound (10, 51, 82). NIM811 was produced by fermentation of the fungus Tolypocladium niveum, followed by extraction and purification (89). Chemical derivatization procedures for the production of a large variety of Cs derivatives are well established. A large amount of knowledge exists about the structure-activity relationships regarding immunosuppressive capacity, CypA binding, and anti-HIV-1 activity (7, 10, 41, 96, 105). We set out to synthesize a series of novel cyclosporines using Cs as a starting material. The goal was to produce by chemical derivatization nonimmunosuppressive cyclosporines with increased affinities for cyclophilin, which was expected to yield derivatives with improved anti-HIV-1 activities. The most promising compound in this series of derivatives turned out to be (d-MeAla3-EtVal4)cyclosporine (Debio-025; where Me and Et are methyl and ethyl, respectively) (97). The in vitro pharmacological profile of this novel CypA-blocking compound and, in particular, its anti-HIV-1 potential are described in this report.

Cyclosporin A (Sandimmun) achieves immunosuppressive activity by complex formation with cyclophilin and subsequent binding of the binary complex to and inhibiting protein phosphatase 2B (calcineurin). Complexes of nonimmunosuppressive cyclophilin binding cyclosporin analogues do not inhibit protein phosphatase 2B, suggesting a crucial role for this enzyme in T cell activation. Binding of cyclosporin A to cyclophilins A, B, and C, respectively, results in complexes of significantly different inhibitory potency. The cyclosporin molecule thus has two functional domains, one mediating cyclophilin binding and a second one endowing affinity of the complex to calcineurin, thereby inhibiting its enzyme activity. Structure-activity studies and x-ray crystallography of cyclosporin-cyclophilin complexes indicate a crucial role of leucine side chains in positions 4 and 6 of the cyclosporin macrocycle for the calcineurin interaction.

Association of cyclophilin A with renal brush border membranes: Redistribution by cyclosporine A. Background Administration of the immunosuppressive agent cyclosporine A (CsA) is associated with nephrotoxicity. The main target for CsA, cyclophilin A (CypA), was found in high levels in epithelial cells of renal proximal tubules. In the present study, CypA was immunodetected and characterized following CsA treatment in subcellular fractions of renal cortex. Method The renal content and distribution of CypA was evaluated in untreated rats and in rats treated with a subcutaneous injection of CsA (10 mg · kg -1 · day -1 ) for 10 days. Results In untreated rats, membrane-bound CypA represents 0.25% of total brush border membrane (BBM) proteins, similar to the proportion found in the soluble fraction. High ionic strength treatment was unable to extract CypA from BBMs, whereas alkaline treatment (Na 2 CO 2 , pH 11) and detergent 3 – [(3 – cholamidopropyl) – dimethyl – ammonio] – 1 – propanesulfate (CHAPS) released it from BBMs. These results indicate that CypA is associated with renal BBMs, and that hydrophobic interactions are involved in this association. The CypA distribution was strongly modified in both BBMs and the soluble fraction after CsA treatment, but its affinity for CsA estimated by photoaffinity labeling was unaffected. The CypA expression level decreased by 45% in BBMs, while it increased by 33% in the soluble fraction, compared with control rats. CypA remained associated with the membranes following in vitro incubation of renal BBMs with CsA. However, incubation of CypA with one of its substrates released CypA from renal BBMs. Conclusions These experiments suggest that renal BBMs contain a significant amount of CypA and chronic exposure to CsA, and acute exposure to one of CypA substrates may modify its subcellular distribution.

Cyclosporine (CsA) is an immunosuppressive and antimicrobial drug which, in complex with cyclophilin A, inhibits the protein phosphatase calcineurin. We recently found that Cryptococcus neoformans growth is resistant to CsA at 24°C but sensitive at 37°C and that calcineurin is required for growth at 37°C and pathogenicity. Here CsA analogs were screened for toxicity against C. neoformans in vitro. In most cases, antifungal activity was correlated with cyclophilin A binding in vitro and inhibition of the mixed-lymphocyte reaction and interleukin 2 production in cell culture. Two unusual nonimmunosuppressive CsA derivatives, (γ-OH) MeLeu 4 -Cs (211-810) and D-Sar (α-SMe) 3 Val 2 -DH-Cs (209-825), which are also toxic to C. neoformans were identified. These CsA analogs inhibit C. neoformans via fungal cyclophilin A and calcineurin homologs. Our findings identify calcineurin as a novel antifungal drug target and suggest nonimmunosuppressive CsA analogs warrant investigation as antifungal agents.

The interaction between P-glycoprotein (P-gp) from membranes isolated from multidrug-resistant Chinese hamster ovary cells and cyclosporin A (CsA) analogues and its metabolites was characterized. Screening of these latter as chemosensitizers was performed using three different assays: (i) vinblastine uptake, (ii) photoaffinity labeling by [125I]iodoaryl azidoprazosin, and (iii) P-gp ATPase activity. Oxidation of the hydroxyl group at position 1 of CsA (200-096), CsG (215-834), or CsD (PSC-833) increased their inhibition of P-gp. CsA analogues (208-032, 208-183) modified at position 11 retained their ability to inhibit P-gp while analogues modified at position 2 (CsC and CsD) lost their efficiency. The inhibitions induced by metabolites of CsA were also compared to those obtained with CsG metabolites. From all the molecules tested, PSC-833 and 280-446 peptolide were the strongest inhibitors. Our results indicate that modifications of CsA analogues at position 1 and 2 are critical for their interaction with P-gp and that CsA metabolites retain a portion of the inhibitory activity of the parent drug.Key words: P-glycoprotein, cyclosporin A, vinblastine uptake, photolabeling, ATPase activity.

Syntheses of cyclosporine analogues are reported wherein the peptide couplings were achieved in solid phase. The Wang resin was used as the solid support, and the peptide couplings commenced with the residue 11 of the cyclosporine skeleton. The couplings proceeded in a stepwide manner up to the residue MeBmt1, using symmetric anhydrides. The peptides were then cleaved off the resin, and the cyclization was achieved in solution using Castro's reagent. The solid-phase synthesis described herein offers a very efficient method for the rapid synthesis of structurally diverse cyclosporine analogues in small quantities. The biological activities of the synthetic cyclosporine analogues were evaluated in two in vitro assays, including the IL-2 reporter gene assay and the cyclophilin binding assay. The structure-activity relationship is discussed.

The interaction between P-glycoprotein (140-180 kDa) from the multidrug-resistant Chinese hamster ovary cell line CHRC5 and cyclosporin A was characterized using three different photoactivable cyclosporin A analogs. Two monoclonal antibodies, which are able to discriminate between two major domains of cyclosporin A (the cyclophilin and calcineurin binding domains), were used to detect the photolabeled proteins. A protein of 155 kDa corresponding to P-glycoprotein was much more strongly photolabeled in membranes of CHRC5 cells than in membranes of their drug-sensitive parent cell line AuxB1. The antitumor drug vinblastine and the reversal agents verapamil and cyclosporin A inhibited the photolabeling, and the nonimmunosuppressive derivative PSC-833 caused a stronger inhibition than cyclosporin A. P-glycoprotein photolabeled with cyclosporin A analogs was only detected with the monoclonal antibody that recognizes cyclosporin A and its metabolites, indicating that the calcineurin binding domain recognized specifically by the other antibody is not exposed. These results suggest that the portion of cyclosporin A that binds to calcineurin plays a role in the interaction of cyclosporin A with P-glycoprotein.

Cyclosporins, in particular the nonimmunosuppressive derivative SDZ NIM 811, exhibit potent anti-human immunodeficiency virus type 1 (HIV-1) activity in vitro. SDZ NIM 811 interferes at two stages of the viral replication cycle: (i) translocation of the preintegration complex to the nucleus and (ii) production of infectious virus particles. Immunosuppressive activity is not correlated with anti-HIV-1 activity of cyclosporins. However, binding to cyclophilin A, the major cellular receptor protein of cyclosporins, is a prerequisite for HIV inhibition: all structural changes of the cyclosporin A molecule leading to loss of affinity to cyclophilin abolished the antiviral effect. Cyclosporin derivatives did not interact directly with HIV-1 proteins; cyclophilin was the only detectable receptor protein for antivirally active cyclosporins. There is no evidence that inhibition of HIV occurs via a gain of function of cyclophilin in the presence of cyclosporins: the complex of cyclophilin A with SDZ NIM 811 does not bind to calcineurin or to any other viral or cellular proteins under conditions in which calcineurin binding to the cyclophilin A-cyclosporin A complex is easily detectable. Thus, the loss of function caused by binding of cyclosporins to cyclophilin seems to be sufficient for the anti-HIV effect. Cyclophilin A was demonstrated to bind to HIV-1 p24gag, and the formation of complexes was blocked by cyclosporins with 50% inhibitory concentrations of about 0.7 microM. HIV-2 and simian immunodeficiency virus are only weakly or not at all inhibited by cyclosporins. For gag-encoded proteins derived from HIV-1, HIV-2, or simian immunodeficiency virus particles, cyclophilin-binding capacity correlated with sensitivity of the viruses to inhibition by cyclosporins. Cyclophilin A also binds to HIV-1 proteins other than gag-encoded proteins, namely, p17gag, Nef, Vif, and gp120env; the biological significance of these interactions is questionable. We conclude that HIV-1 Gag-cyclophilin A interaction may be essential in HIV-1 replication, and interference with this interaction may be the molecular basis for the antiviral activity of cyclosporins.

Strong bases (lithium diisopropylamide (LDA) or BuLi) convert cyclosporin A (CS) to hexalithio derivative containing a Li alkoxide, four Li azaenolate, and one Li enolate units. The Li6 compound is solubilized in tetrahydrofuran (THF) by addition of excess LDA or LiCl. Reactions with electrophiles (alkyl halides, aldehydes, ClCO2R, CO2, (RS)2, D2O) at low temperatures give products containing new side chains in amino-acid residue 3 of the cyclic undecapeptide (see 1–13, Schemes 1, and 2, and Figs. 1 and 2) in moderate to high yields and, with Re- or Si-selectivities, depending upon the conditions of lithiation of up to 7:1, Pure CS derivatives (Scheme 2, Table 1 in the Exper. Part) can be isolated by column chromatography. N-Alkylations or cleavage of the peptide backbone by carbonyl addition occur only at higher temperatures and/or with prolonged reaction times (see 14 and 15 in Scheme 4). Very little or no epimerization of stereogenic centers occurs under the conditions employed. Possible reasons for the feasibility of these surprizing conversions of CS are discussed (Schemes 4 and 5 and Fig. 3). For comparision, [MeAla3]CS (2b) and [D-MeAla3]CS (2a) were also prepared by conventional peptide synthesis in solution (Schemes 6 and 7). Their 1H- and 13C-NMR spectra are compared with those of CS (Table 2 in the Exper. Part).

The immunosuppressant cyclosporine A (CSA) has been shown to bind to the ubiquitous cellular protein, cyclophilin, and to inhibit its rotamase activity. In the present study, 3H-cyclosporine diazirine analogue was used to photolabel viable human cells of lymphoid and fibroblast origin in order to identify the intracellular targets for the drug. While cyclophilin was strongly labeled in situ, additional minor cyclosporine-protein complexes of 25, 40, 46 and 60 kDa were identified in the T cell leukemia cell line Jurkat. These proteins bound specifically, since only active CSA but not inactive CSH or FK506 competed for binding. Photolabeling of MRC5 cells, a CSA resistant human fibroblast cell line, revealed a 25 kDa complex as the major product, while the 46 and 60 kDa bands were not detectable and cyclophilin labeling was only faint, even though both MRC5 and Jurkat cells contain similar cyclophilin concentrations. Thus, our data suggest that the intracellular targets of CSA and/or the accessibility to cyclophilin varies considerably in drug sensitive and resistant cell types, which may contribute to explaining the lymphocyte selectivity of the drug.

The tripeptide and hexapeptide derivatives Boc-Gly-Sar-MeLeu-OH (5b), Boc-Ala-Sar-Sar-OH (6b), Boc-Ala-Sar-MeLeu-OH (7b), and Boc-Abu-Sar-MeLeu-Val-MeLeu-Ala-OH (12b) can be poly-deprotonated (tri- and pentalithio derivatives K and P, respectively), and thus C-alkylated on sarcosine (Sar) moieties with MeI and allyl or PhCH2Br. The polylithiated species are solubilized in THF, and their reactivity modified by excess base (lithium diisopropylamide (LDA)), by added LiCl, and/or the cosolvent N,N′-dimethylpropyleneurea (DMPU). Optimization of the reaction conditions for methylation in the cases of 7b (Table 3) and 12b (Scheme 8) gave products in which the Sar residue of the educt has been transformed into a Me-D-Ala unit in yields of 80 (9c/8c) and 67% (14c/13c), respectively, and with a diastereoselectivity of ca. 4:1. Less selective methylations and benzylations were observed with the tripeptides 5b and 6b containing only one stereogenic center; also, excess base and alkyl halide may lead to double alkylations in those latter two cases (Tables 1 and 2). No epimerization of stereogenic centers was detected under the strong-base conditions. The analysis of the products was accomplished by a combination of NMR and FAB-MS spectroscopy, as well as by hydrolysis to the parent amino acids, subsequent formation of derivatives with isopropyl isocyanate, and GC analysis on the chiral column Chirasil-Val®.

Since the discovery of the immunosuppresive activity of Cyclosporin A (CsA), considerable work has been devoted to the chemical synthesis of analogues. More recently, the finding of potential anti-HIV I activity of CsA evoked interest for the design of more selective cyclosporins active against HIV I but devoid of immunosuppresive activity. Based on previous observations that a N-methyl group at residue 4 is involved in one of the main metabolic degradation pathways [1], the synthesis of CsA analogues disposing various N-ethyl substituted residues at position 4 appeared particularly appealing for developing potential anti-HIV drugs [2]. Here we present the synthesis of new NEtXaa4CsA derivatives and their biological activities.

The immunosuppressant drug cyclosporin A exists as various conformers in water. Up to 1 h is needed to reach maximum complex formation after mixing the drug with its receptor, cyclophilin, or an antibody, indicating that only a fraction of the conformers in aqueous solution adopts a conformation suitable for binding. In the present study we compare the binding behavior of cyclosporin to that of two analogs, using a biosensor instrument (BIAcore, Pharmacia). The amount of complex formation was measured as a function of time after adding the peptides to cyclophilin. The equilibrium affinity constants of cyclophilin for these analogs have been measured. The slow binding of cyclosporin to cyclophilin compared to the instant binding of the cyclosporin analogs supports the hypothesis that cyclophilin recognizes a well defined conformation of cyclosporin that exists in water prior to binding.

Vincamine (13) was prepared from the aldehyde 8 in 5 steps using methyl dimethyl-phosphono-methoxy-acetate as a synthon for the pyruvic ester residue. A method to convert apovincamine (12) into vincamine (13) has hen worked out. Optically active 8 was synthesized from ethyl-malonaldehydic acid ethyl ester in 9 steps including optical revolution of the acid intermediate 3 by means of salt formation with L-(+)-pseudoephedrine. The two diastereomeric salts were distinguishable by NMR. in carbontetrachloride.

Boc-D-Ala-MeLeu-MeLeu-MeVal-OH (DLLL) and its isomers DLDL, DLDD and DLLD were synthesized using several different strategic approaches and a modification of the mixed pivalic anhydride method for carboxyl activation. Alternatively, the tert-butoxy-carbonyl (Boc) or benzyloxycarbonyl (Z) amino-protecting groups and the benzyloxy (OBzl) or tert-butoxy (OtBu) carboxyl-protecting groups were used to protect the reacting peptides. By monitoring the reaction temperature, it was possible to synthesize, starting from the tripeptide DLL as peptide model, either the tetrapeptide DLLL (−20°) or the tetrapeptide DLDL (+20°), selectively. Using 1H-NMR spectroscopy to follow the mixed pivalic anhydride formation of the DLL- and DLD-tripeptides, it could be shown that anhydride formation is strongly dependent on the temperature. It is slow at −20° (several hours) and fast at +20° (20 to 40min). The isomerization of the DLL-anhydride to the more stable DLD-anhydride can be reduced to a minimum by working at −20°, while this isomerization proceeds to near completion at +20°.

The heptapeptide H-MeBmt-Abu-Sar-MeLeu-Val-MeLeu-Ala-OBzl (20) was synthesized for coupling with the previously described cyclosporine tetrapeptide sequence Boc-D-Ala-MeLeu-MeVal-OH (21). The product of the coupling, the undecapeptide Boc-D-Ala-MeLeu-MeLeu-MeVal-MeBmt-abu-Sar-MeLeu-Val-MeLeu-Ala-OBzl (22), was then deprotected and cyclized to cyclosporine (1). The tetrapeptide diastereoisomer Boc-D-ala-MeLeu-MeLeu-D-MeVAl-OH (23) could also be used as a starting material to produce selectively the desired undecapeptide 22. In this case, the N-methyl-D-valine unit, was selectively isomerized to the L-from by using the appropriate condensing agent. The diastereoisomeric undecapeptide Boc-D-ala-MeLeu-MeLeu-D-MeVal-MeBmt-Abu-Sar-MeLeu-Val-MeLeuAla-OBzl (24) was also synthesized starting from 21 by using the mixed pivalic anhydride method to selectively invert the configuration of the N-methyl-L-valine. The structure of the undecapeptide 24 was confirmed by deprotection and cyclization to ‘cyclosporin H’, a natural product known to have the structure [D-MeVal11]cyclosporine (2).

Natural Vincamine (1) has been synthesized in an enantioselective manner starting from the ethylpentenal 7. In the key step a mixture of the diastereoisomeric racemates, 14 and 15, was directly obtained from the silyl enol ether 11 and the dihydro-β-carboline 12 by the way of an intramolecular Mannich reaction of the intermediate 13(Scheme 4). The undesired stereoisomers, 14 and 15b, were recycled to 15a using the related reversible Mannich reaction 18 ⇄ 14 + 15, followed by crystallization of the salt from 15a and (+)malic acid. 15a was converted to natural vincamine (1) in several steps including the known transformation 20→1.

The ability of cyclophilin to react with derivatives of cyclosporine (CsA) was studied. Cyclophilin was found to interact preferentially with CsA-residues 1, 2, 10 and 11, which, together with residue 3, are the residues known to contribute to the immunosuppressive activity of CsA. The recognition of different CsA-derivatives by cyclophilin was correlated with their immunosuppressive activity in vitro. All CsA-derivatives showing a significant activity did bind to cyclophilin, although some of the CsA-derivatives able to bind cyclophilin exhibited only low activities. The results suggest that binding to cyclophilin might be one requirement for immunosuppressive activity of CsA derivatives. When tested with CsA-derivatives showing various conformational changes, the binding of cyclophilin was strongly specific for the peptide-ring conformation of CsA. No binding of calmodulin to CsA could be detected in several formats of solid-phase enzyme- or radioimmunoassay, suggesting that, in contrast to cyclophilin, calmodulin does not possess sufficient affinity for CsA to bind to it when immobilized on the solid phase.

Starting from R,R-(+)-tartaric acid, the synthesis of (2S,3R,4R6E)-3-hydroxy -4-methyl-2-methylamino-6-octenoic acid in 24 steps is reported. This novel amino acid is found in the cyclic undecapeptide cyclosporin A, isolated from the fungal strain Tolypocladium inflatumGAMS. Its stereospecific synthesis allowed, for the first time, the isolation and characterization of the new amino acid previously reported as the ‘C-9-amino acid’ [1].

Cyclosporine mediates its immunosuppressive effect by preventing the synthesis of lymphokine mRNA during the process of T lymphocyte activation. Although the detailed molecular mechanism by which CsA achieves this effect is unknown, two proteins have been identified as putative intracellular CsA-receptor proteins. One of these, calmodulin, is an important Ca++-binding protein and enzyme cofactor and the other, cyclophilin, is a novel protein that is reported to have protein kinase activity. In this study the CsA-binding capacity of both these proteins has been assessed using CsA-coated ELISA plates and CsA-affinity gel matrices. CsA binding was shown by cyclophilin whereas no CsA-calmodulin binding could be detected under identical conditions. However, it was not possible to demonstrate any cyclophilin-associated protein kinase activity. Jurkat cells were probed for the presence of CsA-binding proteins using the CsA-affinity gel matrix; a 17 KD protein, most probably cyclophilin, was identified as the major CsA-binding protein. In addition, a previously unidentified CsA-binding 45 KD phosphoprotein was precipitated from 32P-labeled Jurkat cells. These results would support cyclophilin as the major, if not only, intracellular receptor protein for CsA. However, the relationship between binding of CsA to cyclophilin and/or the 45 KD phosphoprotein and the immunosuppressive effects of CsA is still unknown.

Cyclosporine is a cyclic peptide with a selective immunosuppressive action. One of its 11 amino acids is a novel amino acid, (4R)-4-((E)-2-butenyl-4,N-dimethyl-L-threonine, the preparation of which allowed the synthesis of cyclosporine and specifically modified analogues. Structure-activity relationships emerging from the study of modified cyclosporins suggest that a large portion of the cyclosporine molecule is involved in interactions with its lymphocyte receptor and that this includes amino acids 1, 2, 3 and 11.

Although several cytosolic proteins including calmodulin and cyclophilin have been shown to bind cyclosporine, the direct involvement of these proteins in the immunosuppressive activity of cyclosporine remains to be established. In the present study, a quantitative immunoassay for cyclophilin was developed which made it possible to compare its relative affinity for cyclosporine and any of its analogues. The binding of cyclophilin to cyclosporine coated on a solid phase was revealed by anti-cyclophilin rabbit antiserum followed by antiglobulin-enzyme conjugate. This reaction could be inhibited by addition of free cyclosporine or certain cyclosporine analogues. By studying the binding of cyclophilin to more than fifty cyclosporine derivatives modified singly on each of the eleven amino acid residues, it could be shown that cyclophilin binds to the residues of cyclosporine known to be critical for its immunosuppressive activity. These data identify cyclophilin as a highly discriminating stereospecific binding protein for cyclosporine.

The heptapeptide H-MeBmt-Abu-Sar-MeLeu-Val-MeLeu-Ala-OBzl (20) was synthesized for coupling with the previously described cyclosporine tetrapeptide sequence Boc-D-Ala-MeLeu-MeVal-OH (21). The product of the coupling, the undecapeptide Boc-D-Ala-MeLeu-MeLeu-MeVal-MeBmt-abu-Sar-MeLeu-Val-MeLeu-Ala-OBzl (22), was then deprotected and cyclized to cyclosporine (1). The tetrapeptide diastereoisomer Boc-D-ala-MeLeu-MeLeu-D-MeVAl-OH (23) could also be used as a starting material to produce selectively the desired undecapeptide 22. In this case, the N-methyl-D-valine unit, was selectively isomerized to the L-from by using the appropriate condensing agent. The diastereoisomeric undecapeptide Boc-D-ala-MeLeu-MeLeu-D-MeVal-MeBmt-Abu-Sar-MeLeu-Val-MeLeuAla-OBzl (24) was also synthesized starting from 21 by using the mixed pivalic anhydride method to selectively invert the configuration of the N-methyl-L-valine. The structure of the undecapeptide 24 was confirmed by deprotection and cyclization to ‘cyclosporin H’, a natural product known to have the structure [D-MeVal11]cyclosporine (2).

Starting from R,R-(+)-tartaric acid, the synthesis of (2S,3R,4R6E)-3-hydroxy -4-methyl-2-methylamino-6-octenoic acid in 24 steps is reported. This novel amino acid is found in the cyclic undecapeptide cyclosporin A, isolated from the fungal strain Tolypocladium inflatumGAMS. Its stereospecific synthesis allowed, for the first time, the isolation and characterization of the new amino acid previously reported as the ‘C-9-amino acid’ [1].

More than 180 monoclonal antibodies (McAbs) to the cyclic undecapeptide cyclosporine (Cs) have been prepared. Several immunization protocols and antibody screening processes were compared. Two main groups of McAbs recognizing different “sides” of the Cs molecule could be differentiated. The antibodies belonged to the IgG and IgA classes and showed high affinity for Cs (up to 10 −10 –10 −11 mol/l). Based on their ability to discriminate Cs-derivatives modified singly at each of the 11 residues of the Cs molecule, the antigenic recognition pattern of different McAbs was studied at the level of individual residues. Closely related recognition patterns were found in each of the two main McAb groups. The apparent size of the Cs antigenic sites recognized by different McAbs varied from four to ten residues and did not correlate with antibody affinity.

The fine specificity of monoclonal antibodies for different residues of cyclosporine (Cs) has been studied. The results indicate that the conformation of Cs in aqueous medium corresponds to that observed in crystals rather than that found in aprotic solvents.

An efficient synthesis in 10 steps and overall yields up to 27% of NEtXaa(4)-cyclosporin A derivatives (Xaa = Leu, Val, Ile, Thr) starting from cyclosporin A is described. Biological activities of the new analogues show promising results for the design of cyclosporin derivatives exhibiting non-immunosuppressive and anti-HIV activity. (C) 2000 Elsevier Science Ltd. All rights reserved.