Improved Detection of HIV Gag p24 Protein Using a Combined Immunoprecipitation and Digital ELISA Method

Abstract

Greater than 90% of HIV-1 proviruses are thought to be defective and incapable of viral replication. While replication competent proviruses are of primary concern with respect to disease progression or transmission, studies have shown that even defective proviruses are not silent and can produce viral proteins, which may contribute to inflammation and immune responses. Viral protein expression also has implications for immune-based HIV-1 clearance strategies, which rely on antigen recognition. Thus, sensitive assays aimed at quantifying both replication-competent proviruses and defective, yet translationally competent proviruses are needed to understand the contribution of viral protein to HIV-1 pathogenesis and determine the effectiveness of HIV-1 cure interventions. Previously, we reported a modified HIV-1 gag p24 digital enzyme-linked immunosorbent assay with single molecule array (Simoa) detection of cell-associated viral protein. Here we report a novel p24 protein enrichment method coupled with the digital immunoassay to further extend the sensitivity and specificity of viral protein detection. Immunocapture of HIV gag p24 followed by elution in a Simoa-compatible format resulted in higher protein recovery and lower background from various biological matrices and sample volumes. Quantification of as little as 1 fg of p24 protein from cell lysates from cells isolated from peripheral blood or tissues from ART-suppressed HIV participants, as well as simian-human immunodeficiency virus-infected non-human primates (NHPs), with high recovery and reproducibility is demonstrated here. The application of these enhanced methods to patient-derived samples has potential to further the study of the persistent HIV state and examine in vitro response to therapies, as well as ex vivo study of translationally competent cells from a variety of donors.

Keywords: HIV; SIV; Simoa; biomarker; immunoprecipitation; p24; p27; rectal biopsy.

FIGURE 1

Optimization of the immunoprecipitation (IP) Simoa assay. (A) Selection of IP capture antibody. Recovery of p24 using antibodies from various sources, conjugated to magnetic beads, showed the Capricorn antibody to yield the best capture and was similar to the input concentration after release and volume adjustment. (B) Optimization of bead concentration for p24 IP. Using a 10 mg/mL stock, adding >10 μL of antibody-conjugated beads to a lysate is sufficient for p24 capture. (C) Elution buffer selection. Comparing various elution buffers, 0.1% TFA had no significant difference from input solution after assay, indicating it efficiently releases p24 without destroying crucial epitopes. Other elution buffers resulted in poorer p24 recovery. (D) Time required for p24 IP capture. Recovery of p24 with 2- or 4-h incubation with antibody-conjugated beads is markedly lower than that of input solution, whereas recovery with 24 or 48 h gave recovery similar to that of the input solution (no IP). See Materials and Methods for statistical analysis.

FIGURE 2

Capture of low p24 concentrations. (A) Recombinant p24 at the lowest concentrations of the standard curve was completely recovered after IP, and (B) p24 concentration matches between direct Simoa and IP-Simoa after sample volume normalization. (C) The IP-Simoa assay’s lower limit of detection (LOD) is ≤0.005 pg/mL defined as the coefficient variation (CV) <20% for each p24 standard. This would equate to 1 fg of p24 concentrated from a given volume and released into 200 μL for assay.

FIGURE 3

Detection of p24 from HIV⁺, ART-suppressed peripheral blood CD4⁺ T cells. (A) p24 production was induced in CD4⁺ T cells following a 3-day stimulation with anti-CD3/anti-CD28 beads (n = 3 independent experiments, p < 0.05). p24 was enriched 5-fold with a 5-fold volume reduction after IP in treated group (n = 3, p < 0.001), but not in unstimulated group. (B) From samples that respond well to stimulus, p24 was enriched proportionally to volume reduction in five inducible donors (p < 0.01), and no detectable p24 was observed in their flow through after IP. (C) From samples that do not show robust p24 production after stimulation, IP enrichment was able to demonstrate that p24 is indeed produced but was not at the limit of detection without enrichment. Even with enrichment and bead stimulation, p24 was not detected from HIV-negative CD4⁺ T cells and shows significant difference (p < 0.001) between HIV⁺ and uninfected donors after IP. (D) p24 protein was detected following a 3-day treatment with 1 μM VOR in three of the five donors’ CD4⁺ T cells (4 × 10⁶ cells/condition). The signal was enriched significantly after p24 IP with 4-fold cell lysate volume reduction (p < 0.01). No p24 was detected from unstimulated cells. This opens the door for studying even weak LRAs’ effects using donor-derived cells. (E) Mimicking applications with kill-focused assays, p24 was not detected when including 1 μM staurosporine along with anti-CD3/anti-CD28 bead stimulation for those three inducible donors (p < 0.001). See Materials and Methods for statistical analysis.

FIGURE 4

Development of human CD4 protein assay and application to rectal biopsy lysate characterization. (A) Signal in the ELISA was proportional to the CD4 protein level over a broad standard curve concentration range from 0 to 250 pM. (B) Assay specificity was shown with CD4 signal linearly decreasing following rectal biopsy sample lysate serial dilution. (C) Twenty-4 h of biopsy soaking while mixing in lysis buffer yields complete CD4 protein release from rectal biopsy.

FIGURE 5

Detection of p24 protein from rectal pinch biopsies. (A) Recovery of recombinant p24 spiked into biopsy lysate at 0.02 and 0.1 pg/mL, respectively, is comparable to recovery from buffer as the matrix after IP concentration. (B) Recovery was comparable between buffer and uninfected rectal lysate matrix when using p24 protein recovered from a viremic donor-derived sample as the protein source for spike-in. (C) Cells isolated from a viremic rectal biopsy were lysed and run in standard p24 Simoa or IP-Simoa. The p24 was enriched by the volume reduction during IP and shows linear signal after dilution supporting signal specificity. (D) Reprocessing of rectal biopsies (n = 3) for p24 after lysis revealed that a single extraction yielded >96% of the p24 contained within the biopsy. There was no detectable p24 in solution after IP, indicating complete capture of p24 onto the beads. (E) Comparing measured CD4 protein from HIV-negative and HIV-positive, ART-suppressed rectal biopsies showed no significant difference in CD4 level, implying that similar numbers of CD4⁺ cells were present and lysed from each group. (F) Using the same lysates as in (E), p24 was measurable in several of the HIV⁺ donor biopsy samples (five of nine) after IP-Simoa and not in biopsies from HIV-negative donors (p < 0.05, n = 6). See Materials and Methods for statistical analysis.

FIGURE 6

IP-Simoa assay for SIV p27 protein measurement. (A) SIV p27 Simoa assay standard curve. Linear signal response for recombinant p27 from 0 to 28 pg/mL. (B) Recombinant p27 spiked into uninfected rhesus monkey rectal biopsy (prepared by soaking in lysis buffer) and run with IP Simoa showed enriched p27 signal proportional to the sample volume reduction (20-fold) and expected reduction of signal upon dilution indicating assay specificity. (C) p27 was enriched significantly from the lysate of viremic, SIV⁺ monkey rectal biopsy (p < 0.05, n = 4). No detectable p27 was observed in the flow through after IP demonstrating high efficiency of p27 IP from biopsy lysate matrix. See Materials and Methods for statistical analysis.