Spontaneous HIV-1 replication in a B-lymphoblastoid cell line obtained from an HIV-1-positive patient with undetectable plasma viral load

Abstract

Combined antiretroviral treatment (ART) has improved the survival rate and retarded progression to AIDS in HIV-1-infected patients by reducing active viral replication and allowing immune reconstitution [1]. Nevertheless, the source of re-emerging virus after ART interruption is difficult to establish. Reservoirs of viral persistence include latently infected CD4 T memory cells, tissue macrophages and pockets of continuous low key HIV-1 replication in different tissues [2,3]. Resting peripheral blood B lymphocytes are poor targets for HIV-1 infection under normal conditions, but an activated immunological environment could induce the expression of HIV-1 receptors CD4 and CXCR4 on B lymphocytes, increasing their susceptibility to infection [4,5]. We now report the occurrence of a spontaneous B lymphoblastoid cell line (B-LCL) producing sustained high levels of infectious HIV-1 (20 000 pg/ml p24 accumulated in the supernatant after 20 days) for over 34 months. This B-LCL overgrew a non-stimulated peripheral blood mononuclear cell (PBMC) culture [6] from a haemophilic HIV-1-positive patient receiving continuous zidovudine–lamivudine–indinavir treatment for 4 years preceding the appearance of the HIV-1-positive B-LCL (CDK) with an undetectable HIV-1 plasma viral load (< 50 HIV-1-RNA copies/ml). No HIV-1 was produced by monocytes/macrophages (M/M) or T lymphocytes in PBMC cultures during the 4–8 weeks of culture preceding the spontaneous overgrowth of Epstein–Barr virus (EBV)-positive B lymphoblasts. On previous occasions prolonged non-stimulated culture of PBMC from this patient resulted in the overgrowth of 19 different B-LCL, but in contrast to CDK, HIV-1 replication was not detected or it lasted for a short time at low levels in two out of 19 B-LCL. To date, the patient has had a satisfactory clinical evolution, with few complications and a stable CD4 cell count (240–500 CD4 T lymphocytes/μl). He was originally classified as belonging to the Centers for Disease Control and Prevention C2 group [7] after an episode of interstitial pneumonia of probable Pneumocystis carinii origin. Tuberculosis was detected and treated after 6 months of this. CDK B-LCL was characterized by flow cytometry, and its phenotype corresponded to that of EBV-positive B-LCL: CD19+, CD20+, CD23+, CD30+, CCR5−, less than 1% CD4+, 5% CXCR4+, EBV LMP-1+. The presence of HIV-1 in CDK B-LCL was demonstrated by p24 detection and confirmed by in-situ hybridization (Fig. 1a) and by electron microscopy. The continuous release of HIV-1 was assessed by p24 detection, and the presence of infectious HIV-1 virions in CDK B-LCL supernatants was confirmed by assays of infection using different cell targets: MT-2 cell line for X4 strains, primary non-stimulated cultures for M/M-tropic strains (R5) [8], and EBV-positive B-LCL for atypical non-X4/non-R5 strains. CDK culture supernatants preferentially infected EBV-positive B-LCL when compared with T-cell lines or primary PBMC (Fig. 1b). HIV-1 from CDK supernatants evidenced the non-nucleoside reverse transcriptase inhibitor mutation 103N conferring resistance to nevirapine, delavirdine and efavirenz (generously tested by Dr H. Salomon, CNRS, Facultad de Medicina, Universidad de Buenos Aires, using VircoTYPE HIV-1; Mechelen, Belgium).Fig. 1: Production of infectious HIV-1 by the HIV-1-positive B lymphoblastoid cell line. (a) In-situ hybridization using biotin-labeled probes and tyramide signal amplification was performed: for HIV-1 RNA detection two probes were used: Gag/5′ genome: 5′GGA TGT ACT CTA TCC CAT TCT GCA GCT TCC TCA 3′; Rev probe: 5′TCC TGC CAT AGG AGA TGC/CAG TCG CCG CCC CTC 3′. The signal was developed by adding the chromogenic indicator dye diaminobencidine, which is oxidized by the peroxidase enzymes. Dark brown precipitates at the hybridization site are shown in haematoxilin-stained slides of HIV-1-positive B lymphoblastoid cell line (CDK) cytocentrifuged cells. (b) In order to assay the infectivity of HIV-1 from CDK, HIV-1-negative Epstein–Barr virus (EBV)-positive B-lymphoblastoid cell line (B-LCL), 3 × 106 cells were suspended in 15 ml RPMI 10% fetal calf serum containing CDK supernatant (3000 pg p24/ml) (CDK-RPMI) (▪–▪); 5-day primary non-stimulated peripheral blood mononuclear cell (PBMC) cultures (2 × 106 cells; n = 8) (•–•) were infected with 2 ml CDK RPMI; the X4 target MT-2 T-cell line (3 × 106 cells) (○–○) was infected with 15 ml CDK RPMI. All cell suspensions were incubated for 18 h at 37°C, washed three times and cultured for 20 days; p24 was measured by enzyme-linked immunosorbent assay in cell-free supernatants (Vironostika, HIV-1 antigen; Biomerieux, Boxtel, the Netherlands). For the infection of HIV-1-negative B-LCL, one example of five experiments is shown; for PBMC and MT-2, the mean ± SEM of eight experiments is shown.It is known that HIV-1 can be adapted to replicate in EBV-transformed B-LCL [9] or in EBV-negative B cell lines [10]. We had previously demonstrated that the spontaneous outgrowth of EBV-positive B-LCL after non-stimulated PBMC culture occurred in 60% of HIV-1-positive patients compared with less than 4% of non-infected normal individuals, and in some of the B-LCL continuous HIV-1 replication was observed [5]. HIV replication in those B-LCL could be the result of previous in-vivo HIV-1 infection of B lymphocytes, giving rise to the outgrowth of the EBV-infected lymphoblasts. Alternatively, it could derive from in-vitro HIV-1 infection during the culture period that preceded B-LCL generation. We had also shown that the extent of HIV-1 infection of short-lived cells (PBMC-derived M/M) during non-stimulated culture reflected continuous HIV-1 replication after successful ART [8,11]. In this regard, the probability of detecting HIV-1-producing B-LCL was directly related to HIV-1 replication during the culture period that preceded B-LCL generation, and was inversely related to the extent of ART received [5]. This indicates that before achieving the suppression of the HIV-1 plasma viral load, HIV-1-positive B-LCL could be the result of in-vitro infection from other cell sources or EBV-dependent clonal expansion of in-vivo infected B lymphocytes. In contrast, in HIV-1-positive patients that had been treated for more than 3 years and had undetectable HIV-1 plasma viral loads (as CDK), it is likely that a clone of HIV-1-infected B lymphocytes was the source of HIV-1-positive CDK, suggesting that covert B-lymphocyte infection may persist after successful ART. B lymphocytes could thus act as long-term HIV-1 reservoirs that could trigger HIV-1 re-emergence upon treatment interruption. Sponsorship: This work was supported by grants from Fundacion Rene Baron, SECyT and CONICET, Buenos Aires, Argentina.