Antiretroviral therapy: strategies beyond single-agent reverse transcriptase inhibition

Characterization of HIV Reverse Transcriptases with Tyr181→Cys and Leu100→lle Mutations

Two mutants of human immunodeficiency virus (HIV) reverse transcriptase (RT), Tyr181 to Cys and Leu100 to He, have been prepared and characterized by use of various inhibitors. As compared to wild type RT the mutant RT's had lower Kcat/Km values. The Km values were lower with heteropolymeric than with homopolymeric template-primers. Inhibition by phosphonoformate was of mixed type with both wild-type and mutant RT's and the mutants were less sensitive to phosphonoformate than the wild type RT. The non-nucleoside RT inhibitors 9-CI-TIBO and L-697,661 gave a non-competitive inhibition with respect to substrate of the wild type RT. The mutant RT's were inhibited at higher concentrations, showing a mixed type of inhibition with respect to substrate. ddGTP caused a competitive inhibition of wild type and mutant RT's with respect to substrate. RT preparations with different mutations are useful in rapidly characterizing the interaction between various inhibitors and HIV RT and thus facilitate the development of new inhibitors.

Inhibition of the Cellular Rev Response and HIV-1 Replication by 8-Alkyl-2-(4-pyridyl)pyrido[2,3-d pyrimidin-5(8H)-Ones

A high-capacity, 96-well plate assay in COS-1 cells was developed to screen for inhibitors of the essential HIV-1 Rev response. The assay used Rev-induced expression and cell excretion of the p24 protein from the HIV-1 gagpol gene as a readout. Co-expression of β-galactosidase was used as a specificity control. Using this assay as a drug discovery screen, the authors discovered a series of 8-alkyl-2-(4-pyridyl)pyrido[2,3- d]pyrimidin-5(8 H)-ones that inhibited the primary Rev response in COS-1 cells with IC50s in the range 2-20 μm. These compounds also inhibited HIV-1 strain IIIB replication in human H9 cells (T-cell lymphoma) with IC50s in the same concentration range. Limited structural information suggests that alkyl substituent on N(8) influences potency of this series. These compounds might be the first reported small-molecule inhibitors of HIV-1 replication which act by inhibiting the essential Rev response; further studies in T-cells are in progress to confirm this hypothesis.

Nucleoside reverse transcriptase inhibitor toxicity and mitochondrial DNA

IMPORTANCE OF THE FIELD

HIV/AIDS is a worldwide epidemic. While there remains no cure for the HIV-1 infection, nucleoside reverse transcriptase inhibitors (NRTIs) have helped transform the HIV-1 infection from a lethal disease into a chronic illness. Though NRTIs inhibit HIV-1 replication, they exhibit side effects in human tissues that appear to result from NRTI inhibition of human mitochondrial polymerase γ (pol γ).

AREAS COVERED IN THIS REVIEW

this review discusses the current knowledge of NRTI-induced toxicity, specifically the inhibition of pol γ and the mitochondrial toxicity from incorporation of NRTIs into mitochondrial DNA. Details are discussed about general mechanisms of NRTI toxicity and how specific tissue toxicities in mitochondria relate to clinical manifestation.

WHAT THE READER WILL GAIN

a detailed knowledge of the mitochondrial toxicity resulting from NRTI-inclusive therapies and related tissue toxicities are provided. This review presents both the molecular effects of NRTI usage on mitochondrial genetic homeostasis and energy metabolism as well as the clinical manifestations associated with NRTI toxicities.

TAKE HOME MESSAGE

NRTIs remain a critical component of current HIV-1 treatment regimens. Future NRTIs should provide higher specificity for HIV-RT and lower incorporation by pol γ to minimize mitochondrial toxicity. Alternatively, therapeutic interventions to prevent or alleviate mitochondrial toxicity should be addressed.

A new class of anti-HIV-1 agents targeted toward the nucleocapsid protein NCp7: the 2,2'-dithiobisbenzamides

As part of the National Cancer Institute's Drug Screening Program, a new class of antiretrovirals active against the human immunodeficiency virus HIV-1 has been identified, and the HIV-1 nucleocapsid protein NCp7 was proposed as the target of antiviral action. The 2,2'-dithiobis-[4'-(sulfamoyl)benzanilide] (3x) and the 2,2'-dithiobis(5-acetylamino)benzamide (10) represented the prototypic lead structures. A wide variety of 2,2'-dithiobisbenzamides were prepared and tested for anti-HIV-1 activity, cytotoxicity, and their ability to extrude zinc from the zinc fingers for NCp7. The structure-activity relationships demonstrated that the ability to extrude zinc from NCp7 resided in the 2,2'-dithiobisbenzamide core structure. The 3,3' and the 4,4' isomers were inactive. While many analogs based upon the core structure retained the zinc extrusion activity, the best overall anti-HIV-1 activity was only found in a narrow set of derivatives possessing carboxylic acid, carboxamide, or phenylsulfonamide functional groups. These functional groups were more important for reducing cytotoxicity than improving antiviral potency or activity vs NCp7. All of the compounds with antiviral activity also extruded zinc from NCp7. From this study several classes of low microM anti-HIV agents with simple chemical structures were identified as possible chemotherapeutic agents for the treatment of AIDS.

(s4dU)35: a novel, highly potent oligonucleotide inhibitor of the human immunodeficiency virus type 1 reverse transcriptase

Oligodeoxycytidylates were converted to s4dUMP-containing oligomers by treatment with liquid H2S. The inhibitory potency of the modified oligonucleotides on human immunodeficiency virus type 1 reverse transcriptase depended on the chain length and on the percentage of modification. The most potent reverse transcriptase inhibitor was (s4dU)35. The inhibitory pattern was competitive, when either poly(A) x (dT)16 or poly(C) x (dG)l6 was used as template-primer (variable substrate), suggesting that the free enzyme interacts with (s4dU)35. The Ki values were 3.0 and 2.2 nM in the presence of poly(A) x (dT)16 and poly(C) x (dG)16, respectively.

Single-dose and steady-state pharmacokinetics of hypericin and pseudohypericin

Single-dose and steady-state pharmacokinetics of antivirally acting hypericin (H) and pseudohypericin (PH) were studied in 13 healthy volunteers by administration of St. John's Wort extract LI 160, a plantal antidepressant. Oral administration of 250, 750, and 1,500 micrograms of H and 526, 1,578, and 3,156 micrograms of PH resulted in median peak levels in plasma (Cmax) of 1.3, 7.2, and 16.6 micrograms/liter for H and 3.4, 12.1, and 29.7 micrograms/liter for PH, respectively. The Cmax and the area under the curve values for the lowest dose were disproportionally lower than those for the higher doses. A lag time of 1.9 h for H was remarkably longer than the 0.4-h lag time for PH. Median half-lives for absorption, distribution, and elimination were 0.6, 6.0, and 43.1 h after 750 micrograms of H and 1.3, 1.4, and 24.8 h after 1,578 micrograms of PH, respectively. Fourteen-day treatment with 250 micrograms of H and 526 micrograms of PH three times a day resulted in median steady-state trough levels of 7.9 micrograms/liter for H and 4.8 micrograms/liter for PH after 7 and 4 days, respectively; the corresponding Cssmax levels were 8.8 and 8.5 micrograms/liter, respectively. Kinetic parameters after intravenous administration of Hypericum extract (115 and 38 micrograms for H and PH, respectively) in two subjects corresponded to those estimated after an oral dosage. Both H and PH were initially distributed into a central volume of 4.2 and 5.0 liter, respectively. The mean distribution volumes at steady state were 19.7 liters for H and 39.3 liters for PH, and the mean total clearance rates were 9.2 ml/min for H and 43.3 ml/min for PH. The systemic availability of H and PH from LI 160 was roughly estimated to be 14 and 21%, respectively. Treatment with Hypericum extract, even in high doses, was well tolerated.

Synergistic anti-human immunodeficiency virus type 1 effect of hydroxamate compounds with 2',3'-dideoxyinosine in infected resting human lymphocytes

The cellular models generally used in the in vitro evaluation of anti-human immunodeficiency virus compounds are dividing cells. A model constituted by resting lymphocytes may more accurately reflect a drug's future efficacy in humans, since viral DNA synthesis is known to take place in quiescent cells, creating a reservoir of infected cells awaiting activation to complete their viral replication cycle and to produce infectious virions. We report here the activity of 3'-azido-3'-deoxythymidine, 2',3'-dideoxyinosine, 2',3'-dideoxycytidine, and two hydroxamates, D-aspartic acid beta-hydroxamate and hydroxycarbamate (hydroxyurea), alone and in various combinations, in an in vitro model based on resting lymphocytes. In our model, resting peripheral blood lymphocytes were infected with human immunodeficiency virus type 1 and treated with drugs for 7 days, at which time drugs were removed and the cells were activated by phytohemagglutinin. We show that under these conditions 3'-azido-3'-deoxythymidine, 2',3'-dideoxyinosine, and 2',3'-dideoxycytidine, alone or in combination, neither fully inhibit viral production nor protect lymphocytes from the cytopathic effect of viral replication, at concentrations corresponding to the peak plasma levels observed in a typical treatment schedule in humans. In contrast, we report the synergistic effect of treatment by each hydroxamate with 2',3'-dideoxyinosine of infected resting lymphocytes, resulting in the total suppression of viral production, total protection against the cytopathic effect induced by viral replication, and no effect on the ability of the cells to replicate in this cell culture system.

New lipophilic alkyl/acyl dinucleoside phosphates as derivatives of 3'-azido-3'-deoxythymidine: inhibition of HIV-1 replication in vitro and antiviral activity against Rauscher leukemia virus infected mice with delayed treatment regimens

The antiretroviral activity of two new lipophilic derivatives of azidothymidine (AZT), N4-hexadecyl-2'-deoxyribocytidylyl-(3',5')-3'-azido-2',3'-deoxythy midine (N4-hexadecyldC-AZT) and N4-palmitoyl-2'-deoxyribocytidylyl-(3',5')-3'-azido-2',3'-deoxythy midine (N4-palmitoyldC-AZT) was evaluated in comparison to AZT. In vitro the drugs were tested in human immunodeficiency virus 1 (HIV-1) infected CD4+ HeLa and H9 cells. The in vivo antiviral effect of these derivatives was analysed in Rauscher leukemia virus (RLV) infected mice. The derivatives were incorporated into small liposomes. In vitro both derivatives inhibited virus proliferation in both HIV-1 infected cell lines in a similar dose-responsive manner as AZT. In a plaque reduction assay, using HeLa cells, the IC50 values were 0.035 microM for AZT, 0.5 microM for N4-hexadecyldC-AZT and 4.5 microM for N4-palmitoyldC-AZT, whereas p24 antigen analysis on H9 cells gave IC50 values of 0.005 microM, 0.05 microM and 0.05 microM, respectively. RLV infected mice were treated with intermittent schedules i.p. or i.v. on days 1, 6, 11, and days 16 or 0, 3, 7, and 11 after infection. Regimens with further delayed drug application were on days 3, 7, and 11 and 7 and 11 only. While i.p. treatment with total doses of 380-1140 mg/kg free AZT resulted in 10-30% inhibition of RLV induced splenomegaly, the derivatives gave inhibitions of 37-94%. Late onset of treatment with the derivatives was significantly more effective as compared to free AZT. Intravenous treatment with N4-hexadecyldC-AZT was effective, but with AZT was inactive. The discrepancy in antiviral activity of the AZT derivatives found between the in vitro and in vivo test systems emphasizes the importance of investigating the activity of drug derivatives in vivo.

Triterpene derivatives that block entry of human immunodeficiency virus type 1 into cells

A series of triterpene compounds characterized by a stringent structure-activity relationship were identified as potent and selective inhibitors of human immunodeficiency virus type 1 (HIV-1) replication. Currently studied botulinic derivatives have 50% inhibitory concentrations (IC50) against HIV-1 strain IIIB/LAI in the 10 nM range in several cellular infection assays but are inactive against HIV-2. These compounds did not significantly inhibit the in vitro activities of several purified HIV-1 enzymes. Rather, they appeared to block virus infection at a postbinding, envelope-dependent step involved in the fusion of the virus to the cell membrane.

Lack of synergy in the inhibition of HIV-1 reverse transcriptase by combinations of the 5'-triphosphates of various anti-HIV nucleoside analogs

3'-Deoxy-3'-azidothymidine (AZT) has been shown to synergistically inhibit the replication of human immunodeficiency virus type 1 (HIV-1) in cell culture when combined with several other 2',3'-dideoxynucleoside analogs. In an effort to understand the biochemical mechanism of this synergy, we have examined the effect of combinations of the 5'-triphosphate of AZT (AZT-TP) with either ddCTP, ddATP, or the 5'-triphosphate of the carbocyclic analog of 2',3'-didehydro-2',3'-dideoxyguanosine (carbovir) on both the RNA-directed and DNA-directed DNA polymerase activity of HIV-1 reverse transcriptase. Kinetic studies, which evaluated the ability of these combinations to competitively inhibit the enzyme, showed that AZT-TP could not bind to the enzyme with either the RNA or DNA template at the same time as either of the other three inhibitors. None of these analogs could affect the incorporation of another analog into the DNA chain by the HIV-1 reverse transcriptase. These results indicated that synergistic inhibition of the HIV-1 reverse transcriptase is not responsible for the synergistic antiviral activity seen in cell culture with combinations of these nucleoside analogs.

Empirical and rational approaches for development of inhibitors of the human immunodeficiency virus--HIV-1

The human immunodeficiency virus, HIV-1, is generally accepted to be responsible for AIDS. It is imperative that all approaches, empirical and rational, be taken for development of a drug for therapy of this disease. These approaches are discussed, with emphasis on the direction being pursued in our laboratory. Empirically, we found 3'-deoxy-2',3'-didehydrothymidine, a compound first synthesized for potential anticancer activity by J. Horwitz in the 1960s, to be a potent inhibitor of HIV-1. It is now in Phase II/III clinical trials. We have also synthesized several 2,5'-anhydro pyrimidine nucleoside analogs, which have interesting chemical and biological properties. We have evaluated a natural product, gossypol and synthesized various derivatives for anti-HIV-1 activity, but none were appreciably more inhibitory than the parent compound. More recently, we have taken the rational approach and synthesized a boron-modified tetrapeptide, Ac-Thr-Leu-Asn-boro-Phe, which corresponds to the COOH-terminal of the Phe-Pro scissle bond of the gag/pol gene polyprotein product. Potent inhibition of the HIV-1 encoded protease was observed. These approaches and findings will be discussed.

Antiviral therapy in human immunodeficiency virus infections. Current status (Part II)

Part I of this article reviewed the targets against which anti-HIV drugs can be directed, problems in assessing active compounds (e.g. resistance development and use of surrogate end-points). and nucleoside analogues effective against HIV reverse transcriptase. Intensive research is currently being undertaken in laboratories and hospitals to design and evaluate new inhibitors of HIV. In this work, combining different drugs is one important approach, both to decrease toxicity and to offset the rate of resistance development, which seems to be a major problem associated with therapy directed against the ever-changing HIV. Therapeutic vaccines and immunomodulators are other modalities being actively evaluated against HIV and AIDS, although this effort has not yet yielded any licensed treatment. It appears likely that new antiviral drugs and immunotherapies will be forthcoming during the next 5 years, that they will be used in a variety of combinations, and that the treatment options available for opportunistic infections will increase. These developments should improve the survival and the quality of life of patients with HIV infection.

Characterization of two structurally novel HIV-1 protease inhibitors identified by rational selection

The human immunodeficiency virus (HIV-1), associated with the AIDS (acquired immunodeficiency syndrome) epidemic, encodes an aspartyl protease that is essential for polyprotein processing in the virus (Navia et al., 1989). It has been demonstrated that inactivation of the protease either catalytically or by an inhibitor prevents infectious virion formation (Kohl et al., 1988; Darke et al., 1989). The acquired knowledge of key molecular interactions occurring between inhibitors and aspartyl proteases, as well as the structural similarities between HIV-1 protease and human renin was used to rationally select candidates for HIV-1 screening from the pool of analogs designed as renin inhibitors. A minimal number of chosen compounds were tested in an HIV-1 protease assay system. Two structurally novel peptides emerged as potent enzymatic protease inhibitors. This study highlights the selection process and characterizes the antiviral properties of the two novel analogs.

Mechanism of action of N-butyl deoxynojirimycin in inhibiting HIV-1 infection and activity in combination with nucleoside analogs

The effects on HIV-1 infection of a glucosidase inhibitor, N-butyl deoxynojirimycin (N-buDNJ), were examined. The combinations of N-buDNJ and nucleoside analogs dideoxyinosine (DDI), dideoxycytidine (DDC), or azidothymidine (AZT) were examined in an acute infection assay. The combination of N-buDNJ and nucleoside analog reduced the yield of reverse transcriptase activity more than did either agent alone, and the effects on the number of infectious virus particles were additive or synergistic. In studies of the mechanism whereby N-buDNJ alters HIV-1 envelope fusion activity, no effects on CD4 binding were detected. However, cleavage within the V3 loop of gp120 was reduced by N-buDNJ treatment, possibly reflecting an altered conformation of this region of the envelope protein.