Abstract

TEXT

Combination antiretroviral therapies provide durable viral suppression and constitute the standard of care for HIV infection (http://www.aidsinfo.nih.gov/guidelines/) (1). For example, the combination of tenofovir disoproxil fumarate (TDF) and emtricitabine (FTC) (Truvada) is one of the preferred regimens for treatment of HIV-1 infection (2, 3). In vitro studies have shown that tenofovir (TFV) and FTC have synergistic antiretroviral activity (4, 5).

The investigational nucleoside reverse transcriptase inhibitor (NRTI), 4′-ethynyl-2-fluoro-2′-deoxyadenosine (EFdA) is presently under preclinical evaluation. Unlike other NRTIs used in the treatment of HIV infection, EFdA retains the 3′-hydroxyl moiety. It also contains a 2-fluoro group on the adenine base and a 4′-ethynyl group on the deoxyribose ring. Although EFdA is an adenosine analog, its activation is initiated by phosphorylation by 2′-deoxycytidine kinase (dCK), and the drug is highly resistant to degradation by adenosine deaminase (ADA) (6). EFdA shows exceptional antiretroviral activity in vitro (6–8) and in vivo (9, 10) as well as a favorable cross-resistance profile with current reverse transcriptase inhibitors (RTIs) used in the clinic.

Studies on potential interactions between EFdA and other antiretroviral drugs can provide information that could be useful in the development of combinatorial therapeutic strategies. The present study evaluates anti-HIV efficacy in combinations of EFdA with representative FDA-approved RTIs in vitro.

We first determined the antiviral potencies of five NRTIs (zidovudine [AZT], lamivudine [3TC], FTC, TDF, and EFdA) and three nonnucleoside RTIs (NNRTIs; efavirenz [EFV], etravirine [ETR], and rilpivirine [RPV]) against HIV-1_NL4-3 in order to obtain an optimal range of drug concentrations for use in combination assay analyses. As previously demonstrated (6, 7), EFdA inhibited HIV-1 replication several orders of magnitude more efficiently than other currently approved NRTIs (Table 1). In the same cell-based assays, we evaluated antiretroviral activity of EFdA in combination with the FDA-approved RTIs. To obtain more comprehensive evaluations of drug combinations and to reduce analysis bias, we use two algorithms provided by the software packages MacSynergy II (version 1.0; Ann Arbor, MI) and CalcuSyn (Biosoft, Ferguson, MO), which are based on the Bliss independence model (11, 12) and the median effect principle (13), respectively. Quantitative differences in data analyses by the two algorithms used by the MacSynergy II and CalcuSyn programs are not uncommon (14). In the present work, drug interactions were considered significant only if detected by both computational approaches.

Table 1

Antiviral activity of HIV-1 inhibitors

Compound class and name	EC50 (nM) for anti-HIV-1 activitya
NRTI
AZT	180 ± 60
3TC	1,210 ± 240
FTC	370 ± 70
TDF	14 ± 2
EFdA	3 ± 1
NNRTI
EFV	1.6 ± 0.4
ETR	1 ± 0.1
RPV	0.4 ± 0.1

Open in a separate window

^aEC₅₀, 50% effective concentration. Values were determined using the TZM-bl cell line obtained from the NIH AIDS Research and Reference Reagent Program. All assays were conducted in triplicates. The data shown are mean values ± standard deviations obtained from the results of at least three independent experiments.

Most drugs tested, including the adenosine analog TDF, showed little or no drug interactions in combination with EFdA (Fig. 1 and Table 2). Data analysis with CalcuSyn suggested that the combinations of both EFdA and 3TC (EFdA-3TC) and EFdA-FTC are moderately antagonistic (Fig. 1 and Table 2). Using MacSynergy, the EFdA-3TC combination was assessed as minor antagonism, whereas the EFdA-FTC combination was considered additive; however, its value (−23.4 μM²%) was very close to minor antagonism (−25 μM²%). The observed borderline antagonism may arise from competition at the first and rate-limiting phosphorylation step as EFdA, 3TC, and FTC are all initially activated by 2′-deoxycytidine kinase (6, 15, 16). Small differences in the effect of 3TC versus FTC may arise from the longer half-life of FTC. In contrast, the combination of EFdA-RPV demonstrated apparently significant synergy, as assessed by the two different methods (41 μM²% in MacSynergy and 0.64 combination index [CI] in CalcuSyn). To confirm the synergy of HIV-1 inhibition by EFdA-RPV, we further evaluated this combination in the enzymatic assay for reverse transcriptase. Primer extension assays (7, 17) were performed with Quant-iT PicoGreen reagent (Invitrogen, Carlsbad, CA) (18). As shown in Table 2, the combination of EFdA with RPV provided synergistic effects on inhibition of reverse transcription.

Open in a separate window

Fig 1

Effects of EFdA in combination with other anti-HIV-1 agents. The calculated additive surface, which represents the predicted additive interactions, is then subtracted from the experimental surface to reveal regions of greater-than-expected interactions (synergy). A resulting surface appearing as a horizontal plane at 0% inhibition above the calculated additive surface suggests that the interactions are merely additive. Peaks of statistically significant synergy (positive value) or antagonism (negative value) that deviate significantly from the expected additive drug interactions derived from 95% confidence interval data are shown in the different plots of the interaction between EFdA and other anti-HIV-1 agents in the cell-based assay, as follows: AZT (A), 3TC (B), FTC (C), TDF (D), EFV (E), ETR (F), and RPV (G). Units of μM²% are analogous to the units for area under a dose-response curve in the two-dimensional situation.

Table 2

Interactions of drug combinations for inhibition of HIV-1 virus or RT enzyme

Drug class and EFdA combination	Target	MacSynergy analysis		CalcuSyn analysis		Proposed interactionc
		Synergy/antagonism (μM2%)a	Predicted interaction	CIb	Predicted interaction
NRTI
AZT	Virus	0/0	Additive	1.18	Additive	Neutral
TDF	Virus	0/−12.8	Additive	1.36	Moderate antagonism	Neutral
3TC	Virus	0/−39.5	Minor antagonism	1.23	Moderate antagonism	Possible antagonism
FTC	Virus	0/−23.4	Additive	1.25	Moderate antagonism	Neutral
NNRTI
EFV	Virus	1.9/0	Additive	0.9	Additive	Neutral
ETR	Virus	7.8/0	Additive	1.05	Additive	Neutral
RPV	Virus	41.0/−0.04	Minor synergy	0.64	Synergy	Synergy
RPV	Enzyme	34.1/0	Minor synergy	0.56	Synergy	Synergy

Open in a separate window

^aThe volume of the peaks in the difference plots at the 95% confidence level, which corresponds to the area under a dose-response curve and is considered to provide a quantitative measure of possible drug interactions. MacSynergy II defines μM²% values as follows: 25 to 50, minor synergy; 50 to 100, moderate synergy; >100, strong synergy. The corresponding negative value ranges reflect minor, moderate, and strong antagonism, respectively. Values between 25 and −25 μM²% are considered insignificant.

^bCI, combination Index. CalcuSyn defines CI values as follows: 0.1 to 0.3, strong synergy; 0.3 to 0.7, synergy; 0.7 to 0.85, moderate synergy; 0.85 to 1.2, additive; 1.2 to 1.45, moderate antagonism; 1.45 to 3.3, antagonism; 3.3 to 10, strong antagonism.

^cProposed interaction is assessed on the basis of the predictions by the two computational methods. Neutral, a lack of drug interaction in the combination (additivity).

We further compared the antiretroviral activity of various concentrations of EFdA alone, RPV alone, and a 1:1 molar combination of EFdA and RPV against wild-type virus and two HIV-1 mutants with reverse transcriptase (RT) mutations M184V and L100I/K103N. To evaluate the ability of the drugs to establish a barrier to subsequent infection in the absence of exogenous drug, cells were pretreated with various concentrations of each drug alone or in combination, followed by removal of exogenous drug and inoculation with HIV-1. These conditions assess intracellular persistence of drug following exogenous drug clearance, which is dependent on the intracellular half-life of the test drugs. The EFdA-RPV combination provided additive to synergistic inhibition of wild-type HIV-1 and both mutant strains (Table 3). The protective effect established by EFdA pretreatment is likely the result of EFdA resistance to degradation by adenosine deaminase (ADA) (6), consistent with a longer intracellular half-life (19). Hence, our data suggest that EFdA could be a strong candidate for use in preexposure prophylaxis, an approach in which TDF has shown efficacy in clinical studies (20, 21).

Table 3

Antiviral and prophylactic activity of EFdA, RPV, and EFdA-RPV combinations

Activity type (pretreatment) and virus	Drug treatmenta	EC50 (nM)b	CI50c
Antiviral activity
Wild type	EFdA	0.9 ± 0.2
	RPV	0.7 ± 0.1
	EFdA-RPV	0.4 ± 0.2	0.55
M184V	EFdA	15 ± 3
	RPV	0.6 ± 0.2
	EFdA-RPV	0.7 ± 0.1	0.5
L100I/K103N	EFdA	1 ± 0.5
	RPV	10 ± 2
	EFdA-RPV	9 ± 2	0.45
Prophylaxis (2-h preincubation)
Wild type	EFdA	11 ± 5
	RPV	18 ± 4
	EFdA-RPV	5 ± 2	0.75
M184V	EFdA	97 ± 22
	RPV	8 ± 2
	EFdA-RPV	10 ± 3	1
L100I/K103N	EFdA	12 ± 3
	RPV	175 ± 42
	EFdA-RPV	9 ± 2	0.6
Prophylaxis (18-h preincubation)
Wild type	EFdA	3 ± 1
	RPV	21 ± 4
	EFdA-RPV	4 ± 1	0.9
M184V	EFdA	53 ± 13
	RPV	14 ± 4
	EFdA-RPV	13 ± 40	0.6
L100I/K103N	EFdA	6 ± 1
	RPV	250 ± 25
	EFdA-RPV	5 ± 2	0.75

Open in a separate window

^aEFdA and RPV were combined at a 1:1 ratio.

^bEC₅₀, 50% effective concentration. The data shown are mean values ± standard deviations obtained from the results of at least three independent experiments using P4R5-MAGI cells (32).

^cCI₅₀ is the calculated combination index at a 50% inhibitory concentration.

Finally, we evaluated the ability of EFdA-RPV pretreatment to protect MT-2 lymphoblastoid cells from infection by a mixed virus population consisting of six HIV-1 strains: the wild type; the mutants K65R, Y181C, M184V, and D67N/K70R/T215F/K219Q (thymidine analogue mutations [TAMs]); and L100I/K103N. Cells were exposed to appropriate drugs for 16 h, and then exogenous drug was removed (Fig. 2). The breakthrough virus population harvested from cells not exposed to drug contained all six input viruses, plus some recombinant strains (Fig. 2B). Rapid virus breakthrough was evident in cells pretreated with RPV alone, and the only virus detected in the RPV-breakthrough population was the NNRTI-resistant L100I/K103N (100%). In contrast, breakthrough in cells pretreated with EFdA alone was significantly delayed. The EFdA-breakthrough population contained a mixture of the NRTI-resistant M184V mutant (54%) and TAMs (32%), plus some recombinant strains consisting of M184V plus one or more TAMs. No breakthrough was noted in cells pretreated with the EFdA-RPV combination, suggesting an enhanced protective effect of the drug combination compared to either drug alone.

Open in a separate window

Fig 2

HIV-1 breakthrough in cells pretreated with EFdA, RPV, or a combination of EFdA and RPV. (A) MT-2 cells were incubated with the indicated concentrations of drug for 18 h and then extensively washed to remove exogenous drug. The washed cells were infected with a mixture of six virus strains (the wild type and five drug-resistant variants). Cells were examined daily for HIV-induced cytopathic effects. (B) Genotypic composition of breakthrough virus is shown. At least 20 clones from each breakthrough virus were sequenced. EFdA (0.1 μM) and RPV (0.1 μM) were used in combination. No virus breakthrough was noted in cells pretreated with the EFdA-RPV combination. TAMs (thymidine analog resistance mutations) included the D67N, K70R, T210F, and T219Q RT mutations. Recombinant virus possessed mutations derived from at least two of the input virus strains.

Synergistic interactions between NRTIs and NNRTIs have been previously reported in both viral (5, 22–24) and enzymatic assays (5, 25–31). NNRTIs may act synergistically with NRTIs by suppressing the phosphorolytic unblocking of NRTI-terminated primers, possibly by stabilizing the primer terminus at a posttranslocation position, where it cannot undergo phosphorolytic removal (5, 26, 27, 29). The mechanism for the apparent synergistic activity of the EFdA-RPV combination is under investigation.

In conclusion, EFdA in combination with RPV may provide a beneficial interaction against replication of drug-sensitive and certain RTI-resistant HIV-1 strains. The results of the present study indicate that EFdA may act as promising component of future antiretroviral therapies.

ACKNOWLEDGMENTS

Footnotes

REFERENCES