Acyclovir Could Help Control HIV, Study Finds
"In conclusion, our data on HIV-infected tissues coinfected with various HHVs suggest that ACV may be used to decrease the HIV load in both the peripheral blood and the genital compartments of patients infected with one or several HHVs, including the highly prevalent HHV-6 (Campadelli-Fiume et al., 1999). Although the magnitude of HIV-1 suppression by ACV as well as by the currently approved NRTIs is too low to be used in monotherapy, it is sufficient to be an important part of drug cocktails.
In general, the combination of ACV with an endogenous HHV infection to suppress HIV may constitute a new principle of anti-HIV therapy-a "binary weapon" in which one inert component is converted by another endogenous component into an active therapeutic compound. In the case of ACV, its exceptionally low toxicity and the low cost of ACV and related drugs that have been safely used in humans for more than 30 years, as well as their existing formulations as pills and creams, make them potentially applicable for HIV treatment, possibly in combination with other drugs. New targeted clinical trails will test whether ACV and its derivatives can be used for this new purpose in line with a popular trend to identify new uses for old drugs (Chong and Sullivan, 2007).....Our results provide new insights into the effect of ACV in HIV-1-infected patients. In all previous trials, the effect of ACV on HIV-1 was considered to be indirect and due to the suppression of HSV-2-mediated inflammation.....the results obtained in tissues ex vivo should be extrapolated to the in vivo situation with caution. However, the reliance on lymphotropic HHVs to create the active HIV suppressor at the site of HIV replication may, in principle, solve the critical pharmacological problem of drug delivery."
Herpes Drug Could Help Control HIV, Study Finds
Article Date: 15 Sep 2008 - 1:00 PDT
Acyclovir -- a low-cost, generic drug used to treat herpes -- also might help control HIV in people living with both viruses, according to a study published Wednesday in the journal ell Host and Microbe, Reuters reports. According to the study, acyclovir can help control HIV in infected individuals but only in tissues that also are infected with herpes. This finding might explain why some studies have shown that people taking acyclovir have lower HIV viral loads, while other studies have shown that taking the drug does not prevent HIV transmission.
Lead researcher Leonid Margolis of NIH's National Institute of Child Health and Human Development said that the herpes virus alters acyclovir into a form that can work against HIV. "If you suppress herpes, HIV also goes down," Margolis said. Acyclovir does not become active until it encounters the herpes virus, and the virus then completes a chemical reaction called phosphorylation, which turns the drug into an active compound.
According to Reuters, Margolis said several high-profile studies aimed at preventing HIV infection by treating patients for genital herpes failed because testing "acyclover against HIV in pure cell lines, it doesn't work." In addition, researchers in the prevention studies were trying to suppress herpes completely. "If you suppress herpes virus completely, there is nothing to phosphorylate," Margolis added. He also said that lower or more infrequent doses of the drug could be more effective but that this possibility will have to be tested.
Margolis said that he hopes the outcome of his team's study will help researchers find better ways to use acyclovir against HIV/AIDS and perhaps develop new products such as a microbicide. In addition, adding acyclovir to antiretroviral combination therapies could bolster the therapies, Margolis said. Duane Alexander, director of NICHD, said, "The findings open up promising new avenues of investigation in the fight against the AIDS virus." According to the study, people do not have to be living with the genital herpes virus, or herpes simplex-2, to benefit. Any herpes virus -- including the strains HHV6 and HHV7, which cause an almost universal childhood infection called roseola -- could produce the same effects, according to Reuters (Fox, Reuters, 9/10).
Acyclovir Is Activated into a HIV-1 Reverse Transcriptase Inhibitor in Herpesvirus-Infected Human Tissues
Cell Host & Microbe Sept 2008
Andrea Lisco1, 9, Christophe Vanpouille1, 9, Egor P. Tchesnokov2, Jean-Charles Grivel1, Angelique Biancotto1, Beda Brichacek1, Julie Elliott3, Emilie Fromentin4, Robin Shattock5, Peter Anton3, Robert Gorelick6, Jan Balzarini7, Christopher McGuigan8, Marco Derudas8, Matthias Gotte2, Raymond F. Schinazi4 and Leonid Margolis1, Corresponding Author Contact Information, E-mail The Corresponding Author
1Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD 20892, USA 2Department of Microbiology and Immunology, McGill University, Montreal, Quebec H3A 2B4, Canada 3Center for Prevention Research, UCLA AIDS Institute, David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA 90095-1678, USA 4Veterans Affairs Medical Center, Emory University School of Medicine, Decatur, GA 30033-4501, USA 5St. George's University of London, London, SW17 0RE, UK 6AIDS Vaccine Program SAIC-Frederick, Inc., NCI-Frederick, Frederick, MD 21702, USA 7Rega Institute for Medical Research, Katholieke Universiteit, B-3000 Leuven, Belgium 8Welsh School of Pharmacy, Cardiff University, Cardiff, CF10 3NB, UK
For most viruses, there is a need for antimicrobials that target unique viral molecular properties. Acyclovir (ACV) is one such drug. It is activated into a human herpesvirus (HHV) DNA polymerase inhibitor exclusively by HHV kinases and, thus, does not suppress other viruses. Here, we show that ACV suppresses HIV-1 in HHV-coinfected human tissues, but not in HHV-free tissue or cell cultures. However, addition of HHV-6-infected cells renders these cultures sensitive to anti-HIV ACV activity. We hypothesized that such HIV suppression requires ACV phosphorylation by HHV kinases. Indeed, an ACV monophosphorylated prodrug bypasses the HHV requirement for HIV suppression. Furthermore, phosphorylated ACV directly inhibits HIV-1 reverse transcriptase (RT), terminating DNA chain elongation, and can trap RT at the termination site. These data suggest that ACV anti-HIV-1 activity may contribute to the response of HIV/HHV-coinfected patients to ACV treatment and could guide strategies for the development of new HIV-1 RT inhibitors.
Growing epidemics of HIV-1 infection, especially in countries with limited resources, and the emergence of drug-resistant viruses make it necessary to find ever new safe, efficient, and inexpensive strategies against this virus. Strategies as effective and safe as those developed against several other human pathogens such as HSV, which can be efficiently treated with ACV, have yet to be developed against HIV. The antiherpetic specificity of ACV is primarily based on the unique ability of HHV-encoded kinases to phosphorylate ACV to its monophosphate derivative, which is subsequently converted into the antivirally active ACV-TP (Reardon and Spector, 1989). Consistent with its highly restricted antiherpetic activity, ACV is not currently used as a direct HIV-1 inhibitor.
Our findings demonstrate (1) the direct inhibitory effect of ACV-TP, but not of ACV itself, on isolated HIV-1 RT in a cell-free system, (2) the suppression of HIV-1 replication by ACV in human tissues and in T cell lines if and only if they carry HHV that phosphorylate ACV, and (3) the ability of a phosphorylated ACV prodrug to bypass the requirement of HHV to suppress HIV. Thus, we provide definitive experimental evidence of inhibition of HIV-1 RT activity by phosphorylated ACV and demonstrate that ACV phosphorylation occurring in HHV-infected human tissues transforms this widely used antiherpetic drug into an HIV-1 inhibitor.
The direct suppression of HIV-1 RT activity by ACV-TP was demonstrated in two different cell-free assays, one using recombinant HIV-1 RT and the other using RT extracted from lysed HIV-1 virions. We showed that ACV-TP is incorporated into the nascent viral DNA chain with a level of efficiency similar to that of its natural equivalent dGTP. Incorporation of ACV-TP results in the termination of reverse transcription, while the excision of the incorporated ACV-MP from the DNA chain is partially inhibited because of the fragility of the complex between HIV-1 RT and the terminated DNA chain or because of the formation of a dead-end complex. These properties of ACV resemble those of certain approved anti-HIV NRTIs ([Marchand and Gotte, 2003] and [Meyer et al., 1999]), and we showed here that ACV-TP, a nonphosphonate acyclic nucleoside analog, inhibits HIV-1 RT.
To further confirm that, in HIV-1-infected cells, ACV-triggered HIV-1 suppression is mediated by RT, we identified a pair of isogenic HIV-1 strains differing only by one amino acid (M184V) in RT but having different sensitivities to ACV in ex vivo human tissues. Future studies on selection of ACV-resistant HIV-1 isolates and testing of various multidrug-resistant HIV-1 isolates for ACV sensitivity will reveal the exact set of RT mutations associated with reduced sensitivity to ACV and the rate of their evolution. Identification of ACV-resistant mutants does not, in general, exclude the use of ACV against HIV-1 variants resistant to other NRTIs; for instance, ACV efficiently suppresses the replication of AZT-resistant HIV-1.
As demonstrated in a cell-free system, to inhibit HIV-1 RT, ACV has to be converted into ACV-TP. In human tissues, such conversion requires HHV infection. Accordingly, we found that, in the presence of HHVs that are capable of phosphorylating ACV, HIV-1 was inhibited in ACV-treated tonsils, lymph nodes, and cervico-vaginal and colorectal tissues, where the critical events of HIV-1 pathogenesis and transmission occur in vivo.
Various HHVs, including the ubiquitous HHV-6 detected in all but one tissue ex vivo, may mediate HIV-1 suppression by ACV. The level of HHV-6 replication may be essential. In tonsillar tissues, in which ACV inhibited HIV replication by more than 50%, the median HHV-6 load was significantly higher than in the tissues in which ACV suppressed HIV replication by less than 50%. Moreover, in tissues where the HIV-1 inhibition was suboptimal, there was a correlation between the ACV-mediated suppression of HIV-1 replication and the level of HHV-6. These results demonstrate again the critical role of HHVs in ACV-mediated suppression of HIV-1 and indicate the importance of HHV-6 in mediating HIV-1 suppression by ACV in our ex vivo tissue system.
HHV-6 and HHV-7 are ubiquitous viruses, and, therefore, the probability of finding an HHV-free tissue is very low. Nevertheless, by testing tissues from multiple donors, we identified one tonsillar tissue that was not infected by any HHV. ACV did not inhibit HIV-1 replication in this tissue.
However, to further demonstrate that HHVs are necessary and sufficient for the anti-HIV effect of ACV, we used the HHV-uninfected MT4 cell line. Consistent with the proposed mechanism based on HHV-mediated activation, ACV did not inhibit HIV replication in this HHV-free cell line. However, when HHV-6-infected cells were added to the HIV-1-infected cultures, ACV became an HIV-1 suppressor. This effect was dependent on both the concentration of ACV and the fraction of HHV-6-infected cells. Apparently, in HHV reconstituted MT4 cultures, ACV that is phosphorylated in HHV-6-infected cells is transferred to HIV-1-infected cells, since the majority of these cells were not coinfected with HHV-6. These results are in full agreement with the published data on the transfer of phosphorylated ACV between cells ([Burrows et al., 2002] and [Degreve et al., 1999]). In tissues, transfer of phosphorylated ACV between cells is facilitated by specialized contacts (Nicholas et al., 2003). These and probably other factors (e.g., rapid proliferation and large endogenous dNTP pools, as well as a high level of HIV-1 replication) may contribute to a lower sensitivity of HIV-infected cell lines to ACV compared with integral tissues.
To further demonstrate the necessity of ACV activation for HIV suppression, we synthesized an already monophosphorylated (activated) ACV in which the phosphate is masked by liphophilic groups. In contrast to nonphosphorylated ACV, this prodrug suppressed HIV-1 in HHV-free cultures of MT4 or CEM cells. Although enzymatic reactions mediating ACV prodrug conversion into its active form have been described (Congiatu et al., 2007), the entire process of this conversion remains to be elucidated. Nevertheless, our results provide strong evidence that, upon phosphorylation, ACV suppresses HIV-1 replication in cells. Importantly, these experiments also demonstrate the feasibility of designing a new class of anti-HIV compounds. However, unlike ACV (with a proven safety record and exhaustively studied pharmacokinetics), ACV prodrug effects in various systems have to be evaluated in order to form conclusions on their potential clinical use.
In summary, the following mechanism seems to be responsible for ACV suppression of HIV-1 in human tissues ex vivo, the majority of which carry one or several HHVs, including HHV-6: ACV is monophosphorylated by herpesviral enzymes in HHV-infected cells and then further converted to ACV-TP, which suppresses HIV by inhibiting HIV-1 RT, similarly to other NRTIs.
Our results suggest that ACV may be therapeutically beneficial for various HIV-1-infected patients, since the majority of humans are already infected with HHV-6, often together with other HHVs that activate ACV at least during reactivation. In particular, in immunocompromised patients for whom HHV replication is frequent, those HHVs that are not completely suppressed by ACV (e.g., HHV-6) can continuously generate phosphorylated ACV derivatives. The incomplete inhibition of HHV-6 by ACV confirmed in our experiments is consistent with the much higher Ki of ACV-TP for HHV-6 DNA polymerase (UL69) compared with that of HSV-2 or HSV-1 DNA polymerase (Bapat et al., 1989).
However, clinical trials are needed to test whether replication of HHVs-in particular of HHV-6, which is typically maintained in various organs, including the intestines and the vagina, even in immunocompetent individuals (De Bolle et al., 2005)-would be sufficient to suppress HIV-1 in ACV-treated individuals. Also, clinical trials should reveal whether the range of ACV concentrations used in our study to suppress HIV-1 is clinically relevant. Although ACV penetration efficiency and drug clearance were unknown for ex vivo tissues, the calculated EC50 of 3.1 μM was in the range of what was reported in vivo: a dose of 1 g of oral valacyclovir per day results in a plasma peak concentration of 29.5 μM, a minimum concentration in serum of 3 μM, and a plasma concentration time curve of 89 μM/h ACV ([Lycke et al., 2003] and [Soul-Lawton et al., 1995]). Moreover, the therapeutic dose of the orally administered ACV prodrug valacyclovir, depending on clinical indications, can be increased to as much as 3 g per day.
Recent clinical trials performed so far are in agreement with our ex vivo results and demonstrated that ACV is efficient in suppressing HIV in HSV-2-coinfected individuals ([Baeten et al., 2007], [Nagot et al., 2007], [Zuckerman et al., 2007], [Delany et al., 2007] and [Dunne et al., 2008]). Valacyclovir treatment, at the dose of 1 g per day, reduced the HIV-1 plasma load in these individuals by 50%-70%, an effect comparable to that reported here for human tissues infected with other HHVs. This HIV-1 viral load reduction was clinically beneficial (Corey, 2007) and similar to that reported for AZT or stavudine monotherapy (not, vert, similar70%) ([Delta Coordinating Committee and Delta Virology Committee, 1999], [Katzenstein et al., 2000] and [Rey et al., 1998]). In contrast to the established ACV activity in reducing HIV load in HSV-2 coinfected patients, recent trials failed to demonstrate that HSV-2 suppressive therapy prevents acquisition of HIV-1 ([Celum et al., 2008], [Cohen, 2007], [Lisco and Vanpouille, 2008] and [Watson-Jones et al., 2008]). However, none of the approved NTRIs widely used for therapy was yet developed into an efficient preventive drug when used alone.
Our results provide new insights into the effect of ACV in HIV-1-infected patients. In all previous trials, the effect of ACV on HIV-1 was considered to be indirect and due to the suppression of HSV-2-mediated inflammation. Here, we demonstrate that ACV directly suppresses HIV-1 RT in HHV-coinfected tissue. This effect depends on the levels of HHVs, which were not evaluated in these clinical trials or in previous in vitro studies (Resnick et al., 1986). Obviously, the results obtained in tissues ex vivo should be extrapolated to the in vivo situation with caution. However, the reliance on lymphotropic HHVs to create the active HIV suppressor at the site of HIV replication may, in principle, solve the critical pharmacological problem of drug delivery.
In conclusion, our data on HIV-infected tissues coinfected with various HHVs suggest that ACV may be used to decrease the HIV load in both the peripheral blood and the genital compartments of patients infected with one or several HHVs, including the highly prevalent HHV-6 (Campadelli-Fiume et al., 1999). Although the magnitude of HIV-1 suppression by ACV as well as by the currently approved NRTIs is too low to be used in monotherapy, it is sufficient to be an important part of drug cocktails.
In general, the combination of ACV with an endogenous HHV infection to suppress HIV may constitute a new principle of anti-HIV therapy-a "binary weapon" in which one inert component is converted by another endogenous component into an active therapeutic compound. In the case of ACV, its exceptionally low toxicity and the low cost of ACV and related drugs that have been safely used in humans for more than 30 years, as well as their existing formulations as pills and creams, make them potentially applicable for HIV treatment, possibly in combination with other drugs. New targeted clinical trails will test whether ACV and its derivatives can be used for this new purpose in line with a popular trend to identify new uses for old drugs (Chong and Sullivan, 2007 C.R. Chong and D.J. Sullivan Jr., New uses for old drugs, Nature 448 (2007), pp. 645-646. View Record in Scopus | Cited By in Scopus (18)Chong and Sullivan, 2007).
HIV-1 infection is usually accompanied by infection with other microbes, HIV-1 copathogens, which either pre-exist in the human body or invade it de novo. The replication of HIV-1 copathogens is frequently promoted in HIV-1-infected hosts, and their suppression is often beneficial for the clinical course of HIV disease ([Corey, 2007] and [Jacobson and Mills, 1988]). Human herpesvirus (HHV) infections are commonly associated with HIV-1. In particular, herpes simplex virus 2 (HSV-2) infection is associated with an increase in HIV-1 transmission (Cohen, 2004 M.S. Cohen, HIV and sexually transmitted diseases: lethal synergy, Top. HIV Med. 12 (2004), pp. 104-107. View Record in Scopus | Cited By in Scopus (33)[Cohen, 2004] and [Freeman et al., 2006]) and worsens the clinical course of HIV disease. HSV-2 reactivation may lead to increased plasma HIV-1 levels, thereby adversely affecting survival rates ([Corey et al., 2004] and [Schacker et al., 2002]).
Acyclovir (ACV) is a guanosine nucleoside analog particularly active against HSV-2 as well as against the other β-HHVs (HSV-1 and VZV) (Elion, 1983). It also inhibits, although with lower potency, the replication of the β-HHVs (CMV, HHV-6, and HHV-7) and of the β-HHVs (EBV and HHV-8) (De Clercq et al., 2001).
The mechanism of HHV suppression by ACV is well understood. ACV is phosphorylated in HHV-infected cells by viral-encoded kinases. The resulting monophosphate derivative (ACV-MP) is then converted into ACV triphosphate (ACV-TP) by cellular enzymes and is subsequently incorporated by the viral DNA polymerase into the nascent viral DNA chain, causing its obligate termination (Reardon and Spector, 1989). Generally, the sensitivity of different HHVs to ACV is determined by the extent of its phosphorylation by HHV kinase and by the rate of incorporation of ACV-TP into the viral DNA chain (De Clercq et al., 2001). The mechanism described above explains why ACV does not directly affect viruses other than HHVs, including HIV-1. Nevertheless, early clinical trials indicated that ACV treatment may result in survival benefits in cohorts of HIV-1-infected patients with a high incidence of clinically identifiable HHV (Ioannidis et al., 1998). In agreement with these data, recent trials showed that, in HSV-2-coinfected individuals, the suppression of HSV-2 by ACV is accompanied by reduction of the plasmatic, genital, and rectal HIV-1 load ([Baeten et al., 2007], [Nagot et al., 2007], [Zuckerman et al., 2007], [Delany et al., 2007] and [Dunne et al., 2008]). This decrease was attributed to the suppression of HSV-2-mediated inflammation indirectly reducing HIV-1 load.
ACV Suppresses HIV Replication in Human Lymphoid Tissue Coinfected Ex Vivo with HSV-2
In this study, we used an ex vivo system that preserves the cytoarchitecture of human tissues and supports replication of various viruses without exogenous stimulation (Glushakova et al., 1995).
Initially, we inoculated tissues with HIV-1LAI.04 and HSV-2 (strain G). Both viruses readily replicated in tissues. As expected, ACV at a concentration of 30 μM suppressed HSV-2 replication in both HSV-2 singly infected and HIV-1 coinfected tissues (by 97 ± 1.8% and 93 ± 5.8%, respectively, p = 3 X 10-5, p = 5 X 10-4, n = 4). However, in these HSV-2-coinoculated tissues, ACV also suppressed HIV-1 replication by 62 ± 4.2% (p = 2 X 10-3, n = 4) (Figure 1A).
(A) Blocks of human tonsillar tissue were coinoculated ex vivo with HSV-2 strain G and X4LAI.04 and treated or not with ACV (30 μM). HIV-1 replication was monitored by measuring p24gag accumulated in culture media over 3 day periods. Presented are means ± SEM of the results with tissues from 4 to 17 donors. For each donor, each data point represents pooled viral release from 27 tissue blocks. Note that ACV suppresses HIV-1 in HSV-2-coinfected tissues.
(B) HIV-1 replication was measured in tonsillar blocks infected with X4LAI.04 as in (A). ACV was added at the concentrations of 0.3, 3, 10, 30, and 100 μM, and its anti-HIV activity was evaluated from the suppression of viral replication compared with donor-matched HIV-infected tissues not treated with ACV. The 50% effective concentration (EC50) was estimated by fitting the data to four-parameter logistic regression and was estimated to be 3.1 μM (95% confidence interval: 1.85-5.24). Presented are means ± SEM of the results with tissues from three to seven donors. Note that ACV suppresses HIV-1 replication in the tonsillar tissues in a dose-dependent manner.
(C) HIV-1 proviral DNA load in tissue blocks at day 12 postinfection was measured by real-time PCR. Presented are medians and interquartile ranges of the results (n = 27). Note that ACV efficiently reduces HIV-1 proviral DNA load in blocks of tonsillar tissue.
(D) HIV-1 replication was monitored as in (A). Presented are means ± SEM of the results (n = 38). Note that ACV efficiently suppresses HIV-1 replication in the tonsillar tissues tested.
ACV Suppresses HIV Replication in Singly Infected Human Lymphoid Tissue Ex Vivo
Surprisingly, ACV was also found to suppress HIV-1 replication in tonsillar tissues from 38 donors that were not coinoculated with HSV-2 (Figures 1B-1D).
In HIV-1 singly infected tissues, the suppression of HIV-1 replication, measured as p24gag accumulation in culture medium, was dose dependent with a 50% effective concentration (EC50) of 3.1 μM (95% confidence interval: 1.85-5.24) (Figure 1B). HIV-1 replication, evaluated by the measurement of HIV-1 proviral DNA in tissue blocks at day 12 postinfection, was reduced by 94.8% (interquartile range [IQR] 65.9%-99.8%) in blocks of tonsillar tissues treated with ACV at the concentration of 30 μM compared with donor-matched untreated blocks. The median viral load was reduced by not, vert, similar1.3 log10, from 3108 gag DNA copies per 104 cells in untreated tissues (3.49 log10; IQR 2.99-4.01) to 162 gag DNA copies per 104 cells in ACV-treated tissues (2.27 log10; IQR 0.09-3.07, p < 10_4, n = 27) (Figure 1C). Consistently, ACV significantly decreased the amount of p24gag released into the culture medium on average by 80 ± 4% (p < 7 _ 10_7, n = 38) in human tonsillar tissues (Figure 1D). The average cumulative production of p24gag in the culture medium of tonsillar tissues was 46.4 ± 8 ng/ml and 16.6 ± 4.6 ng/ml, respectively, in tissues untreated and treated with 30 μM ACV.
To avoid the possible confounding effects of contaminating products in commercial ACV, we used ACV from three sources and found that all of them similarly suppressed HIV (p ≥ 0.45, n = 4). Suppression of HIV-1 replication by ACV was not associated with T cell depletion, as evidenced from the similar numbers of total T cells (CD4+CD3+ and CD8+CD3+, p ≥ 0.5, n = 3) and their subsets of activated (CD25+, HLA-DR+, or CD38), naive (CD45RA+CD62L+), or nonnaive (CD45RA_CD62L+/-) (p ≥ 0.16, n = 3) cells. Moreover, ACV antiviral activity was restricted to HSV-2 and HIV-1 since, as expected, ACV did not reduce vaccinia virus replication in tonsillar tissue ex vivo (data not shown).
ACV Suppresses Replication of CCR5- and CXCR4-Tropic HIV-1 Variants in Various Human Tissues
To test whether ACV suppressed HIV-1 replication in tissues other than tonsils, we used explants of tissues that also play an important role in HIV transmission and pathogenesis in vivo: lymph nodes and cervicovaginal and colorectal tissues. ACV suppressed HIV-1 replication, measured by p24gag accumulation in culture medium, in lymph nodes by 70 ± 13.9% (p = 2 X 10-3, n = 7), in colorectal tissues by 72.6 ± 9% (p = 2 X 10-2, n = 3), and in cervical tissues by 60.1 ± 3.5% (p = 7 X 10-3, n = 3). ACV treatment of ex vivo human lymph nodes reduced the average cumulative production of p24gag in culture medium from 67.3 ± 28.1 ng/ml to 15.4 ± 6.6 ng/ml, in ex vivo human colorectal tissues from 7.2 ± 6.8 ng/ml to 1.3 ± 1.2 ng/ml (n = 3), and in ex vivo human cervicovaginal tissues from 3.8 ± 1.7 ng/ml to 1.4 ± 0.5 ng/ml (n = 3) (Figure 2A).
(A) HIV-1 replication in blocks of human lymph nodes and colorectal and cervicovaginal tissues was monitored as in Figure 1A. For each type of tissue, the graph represents a typical result of three to seven experiments performed with tissues from different donors. Note that ACV efficiently suppresses replication of HIV-1 in lymph nodes and colorectal and cervicovaginal tissues.
(B) Examples of ACV suppression of different HIV-1 variants (X4LAI.04, R5BaL, R5SF162, and R5AD8). Each data point represents pooled viral release from 27 tissue tonsillar blocks. Note that ACV efficiently suppresses replication of all four HIV-1 variants.
(C) Presence of HHV-1, -2, -3, -4, -5, -6, -7, and -8 by real-time PCR in blocks of tonsillar tissue. All tissues were negative for HHV-1, -2, -3, and -8. Presented are means ± SEM of the results with tonsils from n donors. Note that there are no significant differences in the level of ACV suppression of HIV-1 replication in tissues infected with various HHVs.
Coinfection with Different Endogenous Human Herpesviruses Is Associated with the Anti-HIV Effect of ACV
To test whether suppression of HIV by ACV is related to the antiherpetic activity of this drug, we measured the presence of various HHVs in the tonsillar tissues used in the present work (Figure 2C). Real-time PCR analysis revealed that all 27 tonsillar tissues that were tested for the presence of HHVs were negative for HSV-1, HSV-2, HHV-3 (VZV), and HHV-8 but were infected with HHV-4 (EBV), -5 (CMV), -6, and -7 in various combinations (Figure 2C). CMV was present in 15% of tissues; EBV, in 52%; HHV-7, in 89%; and HHV-6, in all but one tissue (96%).
Similar to what was observed in immunocompromised patients (Lusso and Gallo, 1994), HHV load was increased in tissues ex vivo. There was a 5-fold increase in the median HHV-6 load at day 12 in culture, from 22.4 DNA copies per 104 cells at the time of surgery (IQR 1.8-57.7) to 116.6 DNA copies per 104 cells (IQR 33.7-808, p = 3 _ 10_4, n = 26). After ACV treatment, the median HHV-6 load at day 12 was reduced to 55.1 DNA copies per 104 cells (IQR 21.3-334.6, p = 10-2, n = 26), demonstrating that ACV significantly suppressed HHV-6 replication. The sensitivity of HHV-6 to ACV was further evaluated in separate experiments in which ex vivo tonsillar tissues were inoculated with HHV-6B (PL-1 strain). In these experiments, ACV suppressed HHV-6B replication in a dose-dependent manner with an EC50 of not, vert, similar27 μM. We found that ACV reduced the total production of HHV-6B in culture medium on average by 56.2 ± 11.4% (52.2 ± 12.4 _ 108 DNA copies per ml versus 22.5 ± 9.4 _ 108 DNA copies per ml in control and ACV-treated tissues, respectively; p = 3 _ 10-3, n = 3).
HHV-7, HHV-4, and HHV-5 median loads in tonsillar tissues were, respectively, 9.23 (IQR 0.51-56.81), 4.71 (IQR 0.06-448.3), and 0.1 (IQR 0.06-0.1) DNA copies per 104 cells at the time of surgery and 21.87 (IQR 0.9-219.7, n = 24), 31.16 (IQR 3.63-260.9, n = 14), and 388 (IQR 175.9-2416, n = 4) copies per 104 cells after 12 days of culture.
In all tissues infected with different combinations of HHVs, ACV suppressed HIV-1 replication with a similar efficiency (Figure 2C). HHV-6 was the only HHV that was present in all combinations. Moreover, it seems that the suppression of HIV-1 by ACV is related to the amount of HHV-6. In tissue blocks in which ACV inhibited HIV-1 replication by more than 50%, the median HHV-6 load on day 12 in culture was significantly higher than in the tissues in which ACV suppressed HIV-1 replication by less than 50%: 131.1 DNA copies per 104 cells (IQR 36.1-1154, n = 23) versus 21.6 DNA copies per 104 cells (IQR 0.6-77.4, n = 4; p = 4 _ 10-2), respectively. Furthermore, we documented a correlation between the ACV-mediated suppression of HIV-1 replication, as measured by p24gag release, and the level of HHV-6 in tissues where the HIV-1 inhibition was suboptimal (between 0 and 99%; r = 0.43, p = 0.03, n = 25).
Although several HHVs, such as HHV-6, are ubiquitous and transmitted early in childhood, we identified a tissue in which there was no detectable HHV. In agreement with our hypothesis, in this tissue, ACV suppression of HIV-1 (at the concentration of 30 μM) was negligible (16%) (Figure 2C).
Furthermore, we evaluated the efficiency of ACV in suppressing different HIV-1 variants. We inoculated ex vivo lymphoid tissues with CCR5- and CXCR4-tropic HIV-1 variants X4LAI.04, R5SF162, R5BaL, and R5AD8. As shown in Figure 2B, ACV suppressed replication of all of these isolates in tonsillar tissues with similar efficiency (by 88%, 95%, 89%, and 94%, respectively, as assessed from p24gag production).
Coinfection with HHVs Is Necessary and Sufficient for the Anti-HIV Effect of ACV
To prove this claim, we performed experiments with MT4 cells, an HHV-uninfected T cell line that efficiently supports the replication of HIV-1LAI.04. We found that ACV did not suppress HIV-1LAI.04 (EC50 > 250 μM) in these cultures, in agreement with earlier observations (Barral et al., 2003 K. Barral, R.C. Hider, J. Balzarini, J. Neyts, E. De Clercq and M. Camplo, Synthesis and antiviral evaluation of 3-hydroxy-2-methylpyridin-4-one dideoxynucleoside derivatives, Bioorg. Med. Chem. Lett. 13 (2003), pp. 4371-4374. Abstract | Article | PDF (157 K) | View Record in Scopus | Cited By in Scopus (2)Barral et al., 2003). To test whether HHV infection is sufficient for the anti HIV-1 effect of ACV, we added various amounts of HHV-6B-infected MT4 cells to HHV-free MT4 cultures infected with HIV-1LAI.04. As expected, ACV suppressed HHV-6. After 3 days of culture, the fraction of HHV-6B-infected cells was reduced in a dose-dependent manner, as measured by flow cytometry (Figure 3A). Importantly, in these cultures, ACV suppressed HIV replication as well, as evaluated by the number of p24gag+ T cells and p24gag release into the culture medium (EC50 of not, vert, similar50 μM) (Figures 3A and 3B). In these experiments, 9% of HHV-6B-infected cells were sufficient to suppress HIV-1LAI.04 replication by not, vert, similar70% (Figure 3A). We found that, in agreement with earlier publications (De Clercq et al., 2001), the concentration that corresponds to the EC50 in these experiments did not affect cell viability (50% cytotoxic concentration > 250 μM).
(A) HIV-1LAI.04-infected MT4 cells were cocultured with HHV-6B-infected MT4 cells in a ratio of 10:1. Presented are the distributions of HHV-6B (upper panel) and HIV-1LAI.04-infected cells (lower panel), as measured by flow cytometry at day 3 post-HIV-1 infection, in cultures untreated or treated with various concentrations of ACV. Note that the fraction of both HHV-6B- and HIV-1-infected cells is reduced in a dose-dependent manner by ACV treatment.
(B) HIV-1LAI.04-infected MT4 cells were cocultured with HHV-6B-infected MT4 cells. HIV-1 replication was monitored as in Figure 1A. Note that the replication of HIV-1 in HHV-6B-/HIV-1-infected cocultures is suppressed by ACV treatment in a dose-dependent manner.
(C) The compound acyclovir-(1-naphthyl [methoxy-L-alaninyl]) phosphoramidate (Cf2649) has been synthesized as described in the Supplemental Data.
(D) HIV-1LAI.04-infected HHV-free MT4 cells were treated with various concentrations of the ACV monophosphorylated prodrug Cf2649. Presented are the kinetics of HIV-1LAI.04 replication, monitored by measuring p24gag accumulated in culture media at days 2, 3, 4, and 6 postinfection. Presented data are representative of two experiments. Note that compound Cf2649 suppresses the replication of HIV-1LAI.04 in a dose-dependent manner.
(E) HIV-1LAI.04-infected MT4 cells not infected with any HHVs were treated with various concentrations of the compound Cf2649. The cumulative release of p24gag into culture media over 6 days of culture treated with various concentrations of the compound Cf2649 is presented as a fraction of the p24gag production in the untreated cultures. Presented data are representative of two independent experiments. Note that compound Cf2649 suppresses the replication of HIV-1LAI.04 in a dose-dependent manner.
To further prove that ACV phosphorylation is required for HIV-1 inhibition, we synthesized the monophosphorylated ACV prodrug acyclovir-(1-naphthyl [methoxy-L-alaninyl]) phosphoramidate (Cf2649) (Figure 3C). This compound bypasses the requirement of HHVs for the activation of ACV since it is already monophosphorylated. Indeed, when applied to (HHV-free) MT-4 cells, Cf2649, in contrast to nonphosphorylated ACV (EC50 > 250 μM), suppressed HIV-1 replication with an EC50 of not, vert, similar3 μM (Figures 3D and 3E). Similar results were obtained with another HHV-free T cell line (CEM) (data not shown).
In conclusion, we demonstrated that neither in HHV-free tissue nor in HHV-free cell lines does ACV suppress HIV-1 infection. Reconstitution of the cell line system with HHV-6-infected cells or bypassing the HHV-kinases requirement by applying an ACV monophosphate derivative makes HHV-free systems susceptible to HIV suppression by ACV.
ACV-TP Inhibits HIV-1 RT
Since we demonstrated that the anti-HIV-1 activity of ACV requires HHV-mediated activation and that phosphorylated derivatives of ACV are produced in HHV-infected tissues (see Supplemental Data and Figure S1 available online), we hypothesized that ACV-TP, into which ACV is ultimately converted (Elion, 1983), interferes with the activity of the HIV-1 RT and directly suppresses HIV-1 replication.
We tested whether ACV-TP suppresses RT by measuring the polymerizing activity of RT from lysed HIV-1 using an exogenously added HIV-1 template. We observed a dose-dependent inhibition of HIV-1 RT activity by ACV-TP, while we noted no suppression of HIV-1 RT activity by ACV itself, even at high concentrations (Figure 4A).
(A) Exogenous template reverse transcriptase assays were performed in the presence of various concentrations of ACV or ACV-TP as described in the Experimental Procedures. Presented are means ± SEM of the results of two experiments performed in duplicates. Note that ACV-TP inhibits HIV-1 RT in a dose-dependent manner.
(B) The dependence of the RT inhibition by ACV-TP on the concentration of dGTP was evaluated using an exogenous template reverse transcriptase assay. dGTP was used at the indicated concentrations. The reactions were performed in the presence of the indicated concentrations of ACV-TP. Presented are means ± SEM of the results of two experiments performed in duplicate. Note that inhibition of HIV-1 RT activity by ACV-TP is inversely dependent on the concentration of dGTP.
Since ACV-TP is a guanosine-5_-triphosphate analog, we investigated whether it acts as a nucleotide RT inhibitor by testing whether dGTP prevents ACV-TP inhibition of RT. We found that, at ACV-TP concentrations of 3.38 μM or 33.8 μM, HIV-1 RT inhibition was inversely dependent on the dGTP concentrations. At the highest concentration of ACV-TP tested, no competition with dGTP was observed (Figure 4B). Since ACV-TP lacks an additional hydroxyl group (present in dGTP and essential for DNA chain polymerization), we further hypothesized that the above-described suppression of HIV-1 RT by ACV-TP is similar to its suppression of HSV DNA polymerase, namely by incorporation into the nascent HIV DNA resulting in its chain termination (Reardon and Spector, 1989).
To prove this hypothesis, we used a gel-based assay ([Marchand and Gotte, 2003] and [Marchand et al., 2007]) and found that HIV-1 RT incorporated ACV-TP into the DNA primer as efficiently as the natural substrate dGTP. The efficiency of single-nucleotide incorporation events (kcat/Km) under steady-state conditions for dGTP and ACV-TP was 14.7 μM-1min-1 and 14.0 μM-1min-1, respectively (Table S1). Furthermore, ACV-TP incorporation resulted in complete DNA chain termination, as shown in Figures 5A and 5B. A primer that contained the natural dGMP at its 3' end was successfully extended in the presence of dTTP, which is the substrate for the following three consecutive template positions (Figure 5B, left panel). In contrast, DNA synthesis was effectively blocked when the primer was terminated with ACV-MP. Even the relatively high concentration of 50 μM of dTTP did not permit its incorporation (Figure 5B, right panel). These data suggest that the mechanism of action of ACV is similar to the currently approved anti-HIV nucleoside reverse transcriptase inhibitors (NRTIs).
(A) The reaction scheme shows the relevant region of the primer and template and two possible outcomes of the primer elongation. Elongation of the primer in the presence of dGTP or ACV-TP is indicated by the "Z" in red, which refers to the incorporated monophosphate dGMP or ACV-MP, respectively. Incorporation events following position Z with dTTP are in blue and underlined.
(B) The primer (S, lane C) was initially elongated to incorporate dGMP or ACV-MP, respectively, at the 3' end referred to as Z. Incubation with increasing concentrations of the next nucleotide dTTP resulted in three nucleotide incorporation events with the dGMP-terminated primer, which are labeled as products (P). Note that the ACV-MP-terminated primer is not extended, which shows that the inhibitor acts as a chain terminator.
DNA chain termination by NRTIs is not irreversible but can be subjected to phosphorolytic excision of the incorporated drug by HIV RT (Meyer et al., 1998). Binding of the next complementary nucleotide following the DNA chain terminator can lead to the formation of a dead-end complex (DEC). In this complex, HIV RT is trapped in a conformation that blocks the excision reaction. Thus, to test whether ACV-MP-terminated DNA is accompanied with the formation of a dead-end complex, we evaluated the inhibition of ACV-MP phosphorolytic excision from the 3' end of the primer by HIV-1 RT. The excision of a DNA chain terminator requires the presence of pyrophosphate (PPi) or pyrophosphate donor molecules (such as ATP). We found that HIV-1 RT is capable of excising the incorporated ACV-MP in the presence of physiologically relevant concentrations of ATP (data not shown) as demonstrated for certain NRTIs. However, to analyze whether the ACV-terminated primer permits formation of a dead-end complex, the ATP-dependent excision reaction was assayed in the presence of increasing concentrations of the next complementary nucleotide. We found that the next complementary nucleotide caused 50% inhibition of ACV-MP excision (IC50) at a concentration of not, vert, similar26 μM (Figures 6A and 6B). These data strongly suggest the formation of a dead-end complex in ACV-MP-incorporated templates. The higher IC50 (26 μM versus 3.1 μM) in comparison to reactions conducted with the control DNA chain terminator ddGMP was probably due to the acyclic nature of ACV-MP. In general, the excision reaction was inefficient; only 10%-15% of the terminated primer strands were rescued for continuation of DNA synthesis at low concentrations of the next complementary nucleotide.
ATP-dependent excision was monitored in a combined excision/rescue assay as previously described in the Supplemental Experimental Procedures.
(A) The reaction scheme shows the relevant region of the primer and template. Extension of the primer in the presence of ddGTP or ACV-TP is indicated by the "Z" in red. The excision of a DNA chain terminator requires the presence of ATP as pyrophosphate donor. The 3'-ultimate nucleotide (Z) was excised with ATP, and the simultaneous presence of dGTP and ddTTP (in blue) allowed the rescue of DNA synthesis.
(B) The combined excision/rescue reaction was studied with ddGMP and ACV-MP-terminated primers. Z refers to the terminated primer and P to the rescued product. The asterisk indicates dGMP misincorporation in the absence of the correct ddTTP substrate. Quantification of excision/rescue reactions for ddGMP and ACV-MP are plotted. The concentration of the next complementary nucleotide required to inhibit 50% of the reaction is calculated on the basis of these two curves. Note that ACV-MP-terminated primer excision/rescue reaction is inhibited due to dead-end complex formation.
(C) Incorporation of ACV-TP produces a dead-end complex with HIV-1 RT. Presented is site-specific footprinting of HIV-1 RT with ddGMP- and ACV-MP-terminated primers in the presence of increasing concentrations of the next complementary nucleotide. Lanes -Fe and +Fe show control reactions in the absence and presence of divalent Fe2+ ions that cause site-specific cleavage on the labeled template. The arrows and the sequence underneath the gel show the position of the oxidative cleavage on the template strand at positions -17 and -18, which are indicative for post- and pretranslocated complexes, respectively. Note that, for ACV-MP-terminated primer, no cleavage occurs at low next complementary nucleotide concentrations, indicating that the complex between RT and terminated primer is fragile, whereas, at higher concentrations of the next complementary nucleotide, RT is blocked in a dead-end complex.
To provide additional direct evidence of dead-end complex formation, we employed site-specific footprinting that allowed us to determine the position of HIV-1 RT on its primer/template at single-nucleotide resolution ([Marchand and Gotte, 2003] and [Marchand et al., 2007]). Nucleotide binding and formation of a dead-end complex can only occur in the posttranslocated conformation in which the nucleotide-binding site of HIV-1 RT has been cleared. In this conformation, the RT has moved a single nucleotide further downstream relative to the pretranslocated state. Our footprints are based on oxidative cleavage on the template strand at positions -17 and -18, which are indicative of post and pretranslocated complexes, respectively. We compared footprints of complexes with ddGMP- and ACV-MP-terminated primers in the presence of increasing concentrations of the next complementary nucleotide (Figure 6C). Low concentrations of the next complementary nucleotide are sufficient to stabilize the posttranslocated complex when the primer was terminated with ddGMP, proving the formation of a dead-end complex. A similar pattern is seen with ACV-MP, although the concentration of the next nucleotide required to stabilize the complex is higher. Oxidative cleavage in case of ACV-MP did not occur in the absence of or at low concentrations of the next complementary nucleotide. These findings suggest that, at the low next complementary nucleotide concentration, the complex between RT and terminated primer formed at the site of the ACV-MP termination is relatively fragile-perhaps prone to dissociation. In contrast, at high concentration of the next complementary nucleotide, RT can be trapped in a dead-end complex. The dissociation constant for ACV-terminated primers in the presence of a high concentration of the next nucleotide (111 μM) is increased not, vert, similar3-fold (Figure S2).
Thus, these data explain why the excision of ACV-MP is generally inefficient at both low and high concentrations of the next complementary nucleotide. At low concentrations, the complex between RT and ACV-MP is relatively unstable, while, at higher concentrations, a DEC is formed.
To prove that ACV targets HIV-1 RT in human tissues as in cell-free assays, we tested two pairs of HIV isogenic strains that are resistant to some of the currently approved NRTIs. The isolate AZT.4x containing the mutations at D67N, K70R, T215Y, and K219Q express complete resistance to AZT and resulted to be as sensitive to ACV in infected tissues as its parental HIV-1 variant (Figure S3A). However, an HIV-1 isolate carrying the M184V mutation (that confers resistance to Lamivudine) was less sensitive to ACV than its parental strain (Figure S3B). The EC50 for the M184V isolate was approximately four times higher than that for the parental HIV-1 isolate. Thus, the evidence that a specific isolated mutation of one amino acid in RT can reduce the sensitivity of an HIV-1 isolate to ACV further proves that HIV-1 RT is targeted and suppressed by ACV in human tissues, in agreement with the data on ACV suppression of HIV-1 RT in cell-free systems.