Development of a New Cartridge Radioimmunoassay for Determination of Intracellular Levels of Lamivudine Triphosphate in the Peripheral Blood Mononuclear Cells of Human Immunodeficiency Virus-Infected Patients

Abstract

A new sensitive method for the measurement of lamivudine triphosphate (3TC-TP), the active intracellular metabolite of lamivudine in human cells in vivo, has been established. The procedure involves rapid separation of 3TC-TP by using Sep-Pak cartridges, dephosphorylation to 3TC by using acid phosphatase, and measurement by radioimmunoassay using a newly developed anti-3TC serum. The radioimmunoassay had errors of less than 21% and a cross-reactivity of less than 0.016% with a wide variety of other nucleoside analogs. The limit of quantitation of the assay for intracellular 3TC-TP was 0.195 ng/ml (0.212 pmol/10⁶ cells), and a cell sample of only 4 million cells was ample for the assay. This procedure, combined with our previously developed method for measuring zidovudine (ZDV) metabolite levels, proved capable of measuring 3TC-TP, ZDV monophosphate (ZDV-MP) and ZDV triphosphate (ZDV-TP) in human immunodeficiency virus (HIV)-infected subjects treated with combination 3TC and ZDV therapy. In seven subjects, intracellular 3TC-TP levels ranged from 2.21 to 7.29 pmol/10⁶ cells, while intracellular ZDV-MP and ZDV-TP levels ranged from <0.01 to 1.76 and 0.01 to 0.07 pmol/10⁶ cells, respectively. Concentrations of 3TC in plasma determined in these subjects ranged from 0.34 to 9.40 μM, which was about fivefold higher than ZDV levels in plasma of 0.04 to 1.4 μM. This is the first study to determine the intracellular levels of the active metabolites in HIV-infected subjects treated with this combination. These methods should prove very useful for in vivo pharmacodynamic studies of combination therapy.

Lamivudine [2R-cis-(−)4-amino-1-[2-(hydroxymethyl)-1,3-oxathiolan-5-yl]-2(1H-pyrimidinone] (3TC) is an important 2′,3′-dideoxynucleoside that inhibits DNA synthesis by terminating the nascent proviral DNA chain; it interferes with the reverse transcriptase activity of the human immunodeficiency virus type 1 (HIV-1), HIV-2, and hepatitis B virus (HBV). This analog and zidovudine (ZDV) are synergistic in vitro against HIV-1 replication (10). A mutation in the HIV polymerase gene at codon 184 that is selected when 3TC is present confers resistance to that drug (4, 9, 15, 18, 19). 3TC has high oral bioavailability in human subjects and is quite effective in reducing HIV RNA in HIV-infected subjects, particularly when used in combination with ZDV and a protease inhibitor such as indinavir (6). 3TC is currently also being evaluated in clinical trials for the treatment of HBV infections (1). Cellular enzymes phosphorylate all known dideoxynucleosides, including 3TC, to their respective triphosphate derivatives, which are the active inhibitory metabolites of these drugs. The relationship between systemic concentration of the dideoxynucleosides and their subsequent antiretroviral effect is still poorly understood. The enzymes required for the activation (phosphorylation) of these drugs are regulated by the cell cycle such that their activities increase with the activation state of the cells (5). This varies greatly for different enzymes. For example, thymidine kinase, the enzyme involved in ZDV phosphorylation, shows extremely low activity in quiescent lymphocytes and monocyte/macrophages but high activity in stimulated lymphocytes (5, 12, 13). Conversely, deoxycytidine kinase, the enzyme responsible for the initial phosphorylation of 3TC, shows relatively small changes during the different stages of the cell cycle (5, 8, 11). Consequently, the extent to which each drug is converted to its active triphosphate varies, and the concentration of the latter, rather than the parent drug, may need to be determined to establish a relationship between the dosage and therapeutic efficacy.

An analytical method employing a coupled cartridge-radioimmunoassay (RIA) has been developed in our laboratory and applied successfully to measure the intracellular metabolites of ZDV in peripheral blood mononuclear cells (PMBCs) from HIV-infected subjects (16, 17). After their isolation from cell extracts, ZDV metabolite concentrations are determined following their complete hydrolysis by acid phosphatase to the parent nucleoside and the resulting nucleoside is quantitated by RIA. In the present study, we report on the development of a new RIA for 3TC and its application for the determination of 3TC triphosphate (3TC-TP) in extracts of PBMC from HIV-infected subjects. We further demonstrate that the two RIAs are sufficiently specific for the measurements of both ZDV and 3TC metabolites in individuals administered these drugs in combination. The immunoassay described in the present paper may also be applicable for the determination of the pharmacokinetics of 3TC in plasma and other biological fluids.

MATERIALS AND METHODS

Materials.

[³H]3TC (12 to 16 Ci/mmol) was purchased from Moravek Biochemical (Brea, Calif.). Lamivudine was provided by Glaxo Wellcome (Research Triangle Park, N.C.). The 3TC-TP was a generous gift from Raymond Schinazi and Jean-Pierre Sommadossi. Lymphocyte separation medium (Ficoll Hypaque) was purchased from Organon Teknika Corp. (Durham, N.C.). Keyhole limpet hemocyanin (KLH) was purchased from Pierce (Rockford, Ill.). Type XA acid phosphatase, goat anti-rabbit precipitating complex, and other chemicals were purchased from Sigma Chemical Co. (St. Louis, Mo.). Tissue culture medium RPMI 1640, Hanks balanced salt solution, glutamine, nonessential amino acids, penicillin-streptomycin, and fetal calf serum were purchased from BioWhittaker (Baltimore, Md.). Sep-Pak C-18 and QMA anion-exchange cartridges were purchased from Waters Co. (Milford, Mass.). ZDV kits were purchased from INCSTAR (Stillwater, Minn.).

Production of 3TC antibodies.

Anti-3TC antibodies were raised in rabbits by immunizing animals with a 3TC-KLH conjugate. The 5′-hemisuccinate-3TC derivative (structure confirmed by mass spectrophotometry and nuclear magnetic resonance) was covalently coupled to KLH by reaction of the corresponding activated N-hydroxysuccinimide ester with primary amino groups of the protein, and rabbits were immunized by intradermal injections of 1 mg of immunogen as previously described (20). Booster injections were given 6 weeks later and every 2 months for 1 year. Rabbits were bled once a week during the 2 weeks following booster injections, and the antisera were kept at 4°C after addition of 0.1% sodium azide.

PMBC isolation and culture.

PBMCs were isolated as previously described by centrifugation over Ficoll-Hypaque (16). The resulting mononuclear cells were suspended to a density of 2 × 10⁶ cells per ml in fresh growth medium. The PBMC suspensions were kept overnight at 37°C in a CO₂ incubator, and the nonadherent cells (primarily lymphocytes) split into separate fractions and were incubated for 24 h with unlabeled 3TC (for RIA) and [³H]3TC (specific activities, 4035, 807, 404, and 81 dpm/pmol for 1, 5, 10, and 50 μM 3TC, respectively). Nucleotides were extracted from PBMCs with 200 μl of 70% methanol buffered to pH 7.4 with 0.015 M Tris (final concentration) per 10⁷ cells and kept at −20°C until analysis.

Drug administration and sample collection from HIV-infected subjects.

PBMCs were isolated from seven HIV-infected volunteers who were receiving drug treatment. These patients were receiving 300 mg of ZDV and 150 mg of 3TC combination therapy twice daily. The duration of previous treatment with this combination at the time of study ranged from 1 to 18 months. ZDV therapy prior to the time of this study ranged from 1 month to 5 years. Sixteen milliliters of venous blood was sampled in CPT Vacutainer tubes at a different time after the oral dosing. PBMCs were separated from erythrocytes by centrifuging them at 1,500 × g for 20 min. The mononuclear cell fraction was transferred to a centrifuge tube, and the cell count was determined. The cell suspension was pelleted by centrifugation, and a 1.5-ml aliquot of the supernatant was removed for determination of drug concentration in plasma. The pelleted cells were extracted with 200 μl of 70% methanol buffered to pH 7.4 per 10⁷ cells and stored at −20°C until analysis. All patient volunteers gave informed consent, and the Institutional Review Board of Methodist Hospital approved the protocol.

Intracellular triphosphate isolation.

Sep-Pak cartridges were used as previously described for separation of ZDV metabolites (16, 17) but with modifications for separation of 3TC-TP. The cell lysates (cultured or HIV-infected PBMCs) were loaded onto the cartridges at a flow rate of 3 ml/min, and sample fractions were collected in polypropylene tubes. 3TC and its mono-, and diphosphate derivatives were eluted from the cartridge with 8 ml of 95 mM KCl, and 3TC-TP was eluted with 5 ml of 300 mM KCl. Phosphate groups were cleaved by the addition of 1.5 U of acid phosphatase (sweet potato type XA) per ml for 60 min at 37°C after adjusting the nucleotide fractions to pH 4.5 with 1 M sodium acetate.

RIA for 3TC.

Two hundred microliters of the dephosphorylated cell or plasma samples diluted 1:1,000 with sample buffer (25 mM potassium phosphate [pH 7.4]) was combined with 100 μl of the anti-3TC antiserum (1:700 dilution with sample buffer) and 100 μl of [³H]3TC label (∼10,000 dpm). The mixture was incubated for 2 h at room temperature, and 500 μl of goat anti-rabbit secondary antibody was added to each tube (except total counts). After allowing the mixture to stand for an additional 30 min, the samples were centrifuged (30 min, 2,000 × g), the pellets were suspended in 600 μl of 0.1 N HCl, and the radioactivity in 500-μl samples was determined. The reference standard bound about 30% of the total counts, and a standard curve was fitted to the data by using a four-parameter logistic fit. A standard curve for the RIA was constructed by the use of eight independent concentrations of 3TC performed in triplicate. ZDV was measured by using RIA as described previously (16, 17).

RIA characterization studies.

The 3TC antiserum had an optimum binding of 25 to 30% when diluted 1:700. Specificity studies were performed with a variety of nucleoside analogs and natural nucleosides. A range of concentrations including 100, 500, 1,000, 5,000, and 10,000 ng/ml was used as samples, and the amount of binding was determined. The percentage of cross-reactivity was calculated with the following equation: (concentration of 3TC at 50% binding/concentration of analog at 50% binding) × 100. The limit of quantitation was determined by the lowest concentration whose six replicates had a coefficient of variation (CV) and error of less than 20%. Interassay variation was determined from triplicate control samples over 16 assays, and intraassay variation was determined in triplicate control samples over 4 assays. Plasma spike recovery studies were performed with plasma obtained from normal donors. The 1-ml samples were adjusted to a concentration of 1,000 ng of 3TC/ml and diluted 1:200, 1:1,000, and 1:2,000 with sample buffer, and the amount of 3TC was determined by RIA.

Cartridge-RIA validation.

The validity of the cartridge-RIA for measuring intracellular 3TC-TP levels was determined by comparison of the RIA with high-performance liquid chromatography (HPLC). PBMC samples incubated with 1, 5, 10, and 50 μM 3TC (as above) were compared by using HPLC and RIA. The radioactive HPLC samples were separated by anion-exchange chromatography (16).

RESULTS

Standard curve and RIA statistics.

Figure 1 shows a representative standard curve for 3TC determination which exhibited a concentration range from 0.097 to 12.5 ng/ml. The CV for the standard curve ranged from 2.4 to 12.1% and the error ranged from −2.2 to 6.0%. There was no obvious bias, as the errors were both positive and negative.

FIG. 1.

Standard curve for 3TC RIA. The standard curve for measuring 3TC concentration was constructed with standards corresponding to final concentrations of 0.097, 0.195, 0.39, 0.78, 1.56, 3.125, 6.25, and 12.5 ng/ml. The data are three determinations for each concentration.

The intraassay variability is shown in Table 1 and had CV values ranging from 1.5 to 15.1% and an error of −10.8 to 17.1%. Interassay variability was taken from the control samples with concentrations of 0.25, 1.00, and 10 ng/ml from 10 assays performed on 10 different days. The CV was 21% at 0.25 ng/ml and 12% at both 1.00 and 10 ng/ml. The deviation of the observed values from expected values was 5% or less. Sensitivity of the assay was determined and showed a limit of quantitation of 0.195 ng/ml with an average CV of 19.1% and an error (accuracy) of 11.3%.

TABLE 1.

Intraassay variation for the 3TC RIA^a

Study no.	Control sample 1			Control sample 2			Control sample 3
	Mean	SD	CV (%)	Mean	SD	CV (%)	Mean	SD	CV (%)
1	0.241	0.029	12.0	1.14	0.08	7.4	10.64	0.99	9.3
2	0.255	0.008	3.1	0.93	0.01	10.6	10.80	1.140	10.6
3	0.273	0.004	1.5	1.05	0.08	7.7	9.68	1.007	10.4
4	0.223	0.005	2.3	0.98	0.05	4.9	8.29	1.252	15.1

^aData are the statistics of the daily triplicate control samples run over 4 days at the indicated concentrations (means and standard deviations are given in nanograms per milliliter). The concentrations represent low, middle, and high regions of the assay. CV (coefficient of variation) was calculated as (standard deviation/mean) × 100. The target mean concentrations were 0.250, 1.00, and 10.00 for control samples 1, 2, and 3, respectively. 

The assay was quite specific, with less than 0.016% cross-reactivity with several compounds including 2′-deoxycytidine, cytidine, thymidine, deoxyadenosine, zidovudine, dideoxyinosine, and D4T. 3TC-TP and dideoxycytidine exhibited slightly higher values, with 0.5 and 0.3% cross-reactivity, respectively. Plasma recovery studies showed negligible background interference (matrix effect) from plasma. Plasma samples with known 3TC concentrations exhibited less than 21% differences and no measurable signal with untreated plasma blanks.

Determination of 3TC-TP levels in PBMCs.

Intracellular 3TC-TP levels in PBMCs from healthy donors were determined by our new cartridge-RIA and compared to levels determined by HPLC. PBMCs (from normal donors) were incubated for 24 h with a range of 3TC concentrations (1 to 50 μM) and assayed by cartridge-RIA. As shown in Table 2, 3TC-TP levels were proportionally higher with the increasing dose of the parent drug. A 50-fold increase in 3TC from 1 to 50 μM resulted in intracellular 3TC-TP concentrations of 1.94 to 9.22 pmol/10⁶ cells. As shown by the results the difference between the levels of 3TC-TP determined by cartridge-RIA and HPLC was less than 15%. Interestingly, there was less variability in the cartridge-RIA procedure than with HPLC. Sensitivity of the assay when used to measure intracellular 3TC metabolites corresponded to 0.212 pmol/10⁶ cells. A sample of only 3 × 10⁶ to 4 × 10⁶ cells was sufficient for the RIA.

TABLE 2.

Comparison of 3TC-TP determination by HPLC and RIA in quiescent PBMC^a

Concn of 3TC (μM)	Intracellular 3TC-TP concn as determined byb:		% Difference in values obtained by HPLC and RIA
	HPLC (pmol/106 cells)	RIA (pmol/106 cells)
1	1.94 ± 0.58	1.72 ± 0.05	11
5	4.24 ± 0.92	3.74 ± 0.20	12
10	4.63 ± 0.94	5.32 ± 0.38	15
50	9.22 ± 2.69	9.42 ± 0.88	2

^aData are the averages of two experiments in which quiescent PBMCs were incubated for 24 h with the indicated concentrations of 3TC. The intracellular 3TC-TP concentrations were determined by HPLC and the cartridge-RIA. 

^bData are concentrations ± the ranges of the two experiments. 

Patient samples.

Combining this method for measuring intracellular 3TC-TP with our previous method for ZDV metabolite measurements (16, 17), we set out to measure these metabolites in patients receiving ZDV and 3TC combination therapy. Subjects in this study were recruited from an area clinic, were in good health, and were receiving daily doses of 300 mg of 3TC and 600 mg of ZDV. We were able to use a single 16-ml blood draw and split the intracellular extract to determine 3TC-TP and ZDV metabolite concentrations. Extracts corresponding to 4 million cells were analyzed for 3TC-TP content while the remainder were analyzed for ZDV monophosphate (ZDV-MP) and ZDV triphosphate (ZDV-TP) concentrations. Aliquots of the plasma were analyzed by RIA for 3TC or ZDV, respectively. Table 3 shows the concentrations of 3TC in plasma, ranging from 0.34 to 9.40 μM, which were about eightfold higher than ZDV levels in plasma of 0.04 to 1.4 μM. As shown previously (14, 17), ZDV-MP was the major intracellular ZDV metabolite with concentrations 15 to 30 times the concentration of ZDV-TP. The intracellular 3TC-TP concentrations in these subjects were about 100-fold higher than that of intracellular ZDV-TP, with a range from 2.21 to 7.29 pmol/10⁶ cells, which compared to intracellular ZDV-TP concentrations of 0.01 to 0.07 pmol/10⁶ cells. It should be noted that the blood was drawn at different times after drug administration. This was due to variations in the times at which samples could be obtained from these subjects.

TABLE 3.

Drug concentrations in plasma and cellular extract in HIV-infected patients^a

Patient no.	3TC			ZDV
	Time of blood draw following drug administration (h:min)	Concn of 3TC in plasma (μM)	Concn of 3TC-TP in cellular extract (pmol/106 cells)	Time of blood draw following drug administration (h:min)	Concn of ZDV in plasma (μM)	Concn of ZDV-MP in cellular extract (pmol/106 cells)c	Concn of ZDV-TP in cellular extract (pmol/106 cells)
1	1:15	2.26	2.59	1:15	0.69	1.19	0.04
2	3:30	9.40	6.14	3:30	0.20	1.36	0.03
3	3:45	3.07	2.21	3:45	0.10	ND	0.01
4b	4:50	4.55	3.72	4:50	1.4	1.76	0.07
5	6:00	0.34	4.51	6:00	0.04	ND	0.02
6	7:00	4.78	7.29	7:00	1.0	1.55	0.05
7	7:30	0.94	3.48	7:30	0.08	0.53	0.03

^aPatients receiving combination therapy were assayed for levels of parent drug in plasma and intracellular levels of nucleoside analog triphosphate. 

^bPatient received indinavir in addition to 3TC and ZDV. 

^cND, none detected. 

DISCUSSION

The increasing use of multidrug therapy has dramatically improved the prognosis of HIV-infected subjects. Recently, a number of aggressive drug regimens that utilize triple combination of two nucleoside analogs with a protease inhibitor have been introduced. These protocols have substantially increased the number of individual subjects displaying sustained suppression of detectable viral RNA or infectivity in blood for up to 2 years (6, 21). The improvements in the prognosis of individuals treated with these therapies will almost inevitably result in the application of the therapies to an increased proportion of HIV-positive individuals to reduce the progression of the disease. The efficacy and toxicity of the nucleoside analogs in clinical use are related to their phosphorylation and the intracellular concentration of the respective triphosphate in the target cells. There are studies which show that a higher concentration of ZDV-TP in HIV-infected patients is associated with an increased reduction in viral load and some improvement in CD4 lymphocytes (2). Thus, knowledge of the intracellular metabolite concentration of these agents in the patients during therapy could provide a basis to optimize and individualize combination drug therapy.

3TC is being used extensively in the treatment of HIV, especially in combination therapies with ZDV and protease inhibitors (7, 10). However, there is currently no method available for determining the intracellular metabolites of this agent, or any antiretroviral agent other than ZDV, in patient cells. The RIA developed herein for 3TC was reliable and sensitive with an inter- and intraassay CV of less than 21% and a limit of quantitation of 0.195 ng/ml, which compared very favorably to a previously described RIA of 0.4 ng/ml (22). With the coupled cartridge-RIA procedure, we were able to measure 3TC-TP concentrations corresponding to a detection limit of 0.2 pmol/10⁶ cells. Our antibody showed a cross-reactivity of less than 0.016% with a variety of nucleosides, including ZDV, d4T, cytidine, and 2′-deoxycytidine. The new methodology can measure 3TC-TP produced in PBMCs incubated with subtherapeutic concentration of 1 μM 3TC and median recovery of the entire procedure of >75%. Table 2 shows that the minimal error due to sample processing was less than 15% difference between 3TC-TP levels measured by HPLC and the cartridge-RIA method. There was also less variability associated with the cartridge-RIA than HPLC, which may be due to minimal peak tailing as a result of the lower resin content and short length of the cartridges.

The cartridge-RIA was used to determine 3TC-TP concentrations in PBMCs in a small cohort of HIV-infected subjects receiving combination therapy. The RIA for 3TC-TP required fewer than 4 × 10⁶ cells (∼4 ml of blood). Thus, this method may also be suitable for measuring 3TC metabolite levels in pediatric patients, for whom blood volume is usually a limitation. Given the sensitivity and specificity of our RIAs, we were able to measure the levels of both 3TC and ZDV metabolites in the cells of the same patient. Samples were drawn at various times after drug administration, and these exhibited a wide range of drug concentrations. Even between subjects whose drug administration times were less than 30 min apart (subjects 2 and 3 and subjects 6 and 7), there was over 50% variability. Of the seven subjects studied, each of whom received 150 mg of 3TC and 300 mg of ZDV twice daily, the concentrations in plasma of 3TC and ZDV ranged from 0.34 to 9.4 μM and 0.04 to 1.4 μM, respectively. The levels of 3TC-TP and ZDV-TP in the PBMCs of these subjects ranged from 2.21 to 7.29 pmol/10⁶ cells and 0.01 to 0.07 pmol/10⁶ cells, respectively. This wide variation supports measurement of the intracellular triphosphates directly. As shown previously, ZDV-MP was the main ZDV metabolite produced in PBMCs of patients. This is apparently due to the inefficient phosphorylation of ZDV past the monophosphate step (3). To our knowledge, these values represent the first quantitation of both 3TC and ZDV metabolites in HIV-infected subjects receiving combination antiviral therapy. These results demonstrate that there are substantial differences in the extent of conversion of these antiviral agents to their respective triphosphate derivatives in patients.

In summary, we describe a combined cartridge-RIA for measuring levels of 3TC and intracellular 3TC-TPs. This assay is sensitive and specific and permits the analysis of a large number of patient samples. This assay, together with our previously developed cartridge-RIA for ZDV metabolites, provides the opportunity to further explore the relationship between intracellular phosphorylation of these drugs and their efficacy in HIV-infected patients.