Study Participants
We enrolled adults receiving oral PrEP (TDF + emtricitabine) and individuals not receiving any HIV medications at the Madison Clinic at Harborview Medical Center in Seattle. Exclusion criteria were age under 18 years, seropositivity for HIV or flavivirus (Zika, Dengue, West Nile, Yellow Fever), or previous enrollment in HIV or flavivirus vaccine study. We collected participant data on HIV status, sociodemographic characteristics, and body mass index (BMI). All study participants were enrolled and sampled in accordance with the University of Washington/Fred Hutch Center for AIDS Research (CFAR) Enhanced Data and Specimen Collection Service. All participants provided informed consent and samples were collected in association with study identifiers.
Blood Sample Collection and LC-MS/MS Measurement
Venous whole blood was collected from each study participant. Dried blood spot (DBS) cards were prepared using 25 µL of each whole blood sample. Whole blood tubes were stored on ice and analyzed by RESTRICT within 4 hours of sample collection. Matched whole blood and DBS samples were tested using the RESTRICT assay and LC-MS/MS. DBS cards were stored at -70˚C to -80˚C until analysis. TFV-DP concentrations were measured using a validated LC-MS/MS assay in accordance with the Clinical Pharmacology Quality Assurance and Quality Control Program validation guidelines.(23)
RESTRICT Assay Principle and Workflow
RESTRICT detects TFV-DP drug concentrations based on its mechanism of action on HIV RT.(22) Fluorescence output from in vitro DNA synthesis by recombinant HIV RT is used to estimate TFV-DP concentration in a patient’s blood. High fluorescence and high RT activity indicates low TFV-DP concentrations and vice-versa.
Reactions were carried out in a buffer containing: 60 mM Tris (77-86-1, Sigma, St. Louis, MO), 30 mM KCl (7447-40-7, Sigma, St. Louis, MO), 8 mM MgCl2 (7786‑30‑3, Sigma, St. Louis, MO), 10 mM dithiothreitol (20-265, Sigma, St. Louis, MO), 400 nM deoxynucleotide triphosphates (dNTPs) (D7295, Sigma, St. Louis, MO), 40 nM primer 16S rRNA Forward primer AGA GTT TGA TCC TGG CTC AG (51-01-19-06, Integrated DNA Technologies, Coralville, IA) and 4 nM DNA template buffered to pH 8.0 using HCl (7647-01-0, Acros Organics, Fair Lawn, NJ). Custom-designed DNA templates were synthesized in silico (Integrated DNA Technologies, Coralville IA) and consisted of a 20-nucleotide primer binding site followed by a 180-nucleotide detection T-rich region consisting of TTCA repeats to increase the likelihood of chain termination by TFV-DP (TTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCATTCACTGAGCCAGGATCAAACTCT). Recombinant RT was obtained through the NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH: HIV-1 RT Catalog #3555 from Dr. Stuart Le Grice and Dr. Jennifer T. Miller.(24)
RESTRICT was conducted in 5 steps (Fig.1). First, blood samples were collected from study participants at the CFAR. Next, venous blood was diluted to 8% volume in nuclease-free water (3098, Sigma-Aldrich, St. Louis, MO) and vortexed for 5 min to lyse red blood cells (RBCs), release intracellular TFV-DP, and reduce assay inhibition by blood components. Then, 5 µL of diluted whole blood was added to 30 µL of buffered master mix in flat-bottom polystyrene 384-well plates with nonbinding surfaces (3575, Corning, Corning, NY). 5 µL of HIV-1 RT at a final enzyme concentration of 100 nM was added as the last reagent to initiate DNA synthesis and incubated at 37˚C for 30 min in a microplate reader (SpectraMax iD3, Molecular Devices, San Jose, CA). Finally, PicoGreen™ dye (P7581, ThermoFisher Scientific, Waltham, MA) diluted 1:400 in 1×TE (10128-588, VWR, Radnor, PA) was added to stop the reaction and provide fluorescence output. Five replicates were tested for each sample.
A standard curve was generated with five aliquots of TFV-DP spiked into diluted blood (from participant 007, not on PrEP) at final concentrations corresponding to 8.9 to 58333 fmol/punch in 9-fold increments and spanning nearly two orders of magnitude above and below the PrEP adherence clinical range.
Statistical Analysis
Baseline correction was carried out by subtracting the fluorescence obtained from each sample without added RT enzyme from the endpoint assay fluorescence after 30 min RT incubation. The fluorescence intensity from each sample was normalized by dividing by the average fluorescence obtained from blood samples without detectable TFV-DP by LC-MS. We calculated the Pearson correlation coefficient between RESTRICT fluorescence and LC-MS/MS TFV-DP concentrations. We compared the fluorescence of samples at thresholds for adequate adherence (700 fmol/punch i.e., 4 doses/week) and perfect adherence (1250 fmol/punch i.e., 7 doses/week) among men who have sex with men receiving PrEP. We established a priori thresholds for fluorescence at 700 fmol/punch and 1250 fmol/punch by interpolating standard curves obtained by spiking known concentrations of TFV-DP in blood using GraphPad Prism.