HTLV-1–infected T cells contain a single integrated provirus in natural infection

Human T lymphotropic virus type 1 (HTLV-1) infection results in lifelong carriage. More than 90% of infected persons remain asymptomatic carriers; the remainder develop either adult T-cell leukemia/lymphoma (ATLL) or a progressive neurologic disorder known as HTLV-1–associated myelopathy/tropical spastic paraparesis (HAM/TSP) or other inflammatory conditions, such as HTLV-1–associated polymyositis or uveitis.^1-3 

In each infected person, a stable proviral load (PVL), the number of proviral copies per 100 PBMCs appears to be maintained by proliferation chiefly of HTLV-1⁺ CD4⁺ T cells. The PVL remains relatively stable within 1 host but varies by > 1000-fold between hosts. A high PVL is associated with oligoclonal proliferation of HTLV-1⁺ T cells, with several clones reaching a high abundance in the peripheral blood. However, the dynamics of HTLV-1 replication in vivo remain poorly understood: there is no reliable estimate of the number of HTLV-1–infected T-cell clones, the number and abundance of infected T-cell clones, or the relative contributions of T-cell proliferation and infectious spread to the maintenance of the PVL. The reason for selective expansion of individual T-cell clones also remains unexplained.

Detection of a single genomic integration site by Southern blot is characteristic of malignant disease,^4,5  but this technique lacks the sensitivity to detect minor populations in polyclonal HTLV-1–infected PBMCs. Inverse PCR and linker–mediated PCR have been used to map integration sites in patients with malignant or nonmalignant HTLV-1 infection.^6-9  However, these techniques have 2 major biases that prevent accurate mapping and quantification. First, the use of restriction enzyme digestion leads to preferential detection of proviruses that lie near the restriction sites. Second, PCR amplification strongly favors short DNA fragments.^10,11  We therefore recently developed a novel high-throughput protocol to map and accurately quantify proviral integration sites in the host cell genome.¹²  The results demonstrated that the previous estimate of ∼ 100 proviral integrations in 1 host was grossly inaccurate: the true number is frequently > 10⁴. The question arises: how many distinct HTLV-1–infected T-cell clones are present in each host? That is, what is the average number of proviral copies per infected cell?

Here, we generated T-cell clones from CD4⁺CD25⁺ PBMCs of 10 persons with different clinical manifestations of HTLV-1 infection and used unbiased linker–mediated PCR and high-throughput sequencing to accurately quantify the abundance of unique integration sites within each T-cell clone.

Ten persons (7 female, 3 male; median age, 55 years) infected with HTLV-1 attending the National Center for Human Retrovirology (Imperial College Healthcare NHS Trust, London) provided blood samples with written consent in accordance with the Declaration of Helsinki. This study was approved by the United Kingdom National Research Ethics Service (NRES). The subjects represented each major clinical manifestation of HTLV-1 infection: 2 asymptomatic carriers (median PVL, 6.2%), 5 HAM/TSP patients (median PVL, 13.6%), 2 ATLL patients (median PVL, 57.5%), and 1 patient with polymyositis (PVL, 18.3%). PBMCs were isolated using standard techniques and cryopreserved in 90% FCS (Invitrogen) with 10% DMSO (Sigma-Aldrich).

Because cells expressing the activation/regulatory marker CD25 are more frequently infected than other PBMC subsets,¹³  CD4⁺25⁺ cells were isolated by magnetic activated cell sorting (CD4⁺CD25⁺ isolation kit, Miltenyi Biotec). Cells were subsequently cloned by limiting dilution in RPMI containing 10% human AB serum (Invitrogen) in the presence of 50 IU/mL IL-2 (Promocell), 1 μg/mL PHA (Sigma-Aldrich), 10μM raltegravir (Selleck Chemicals), and 0.5 × 10⁶/mL γ-irradiated feeder cells (mixed PBMCs from 3 uninfected donors). Clones were expanded with feeder cells and PHA every 14 days and fed with IL-2 twice weekly. The integrase inhibitor raltegravir was present throughout the in vitro culture, to minimize secondary infectious spread of HTLV-1.¹⁴  Clones were cultured for 4-6 weeks before genomic DNA extraction (DNeasy Blood and Tissue Kit; QIAGEN).

T-cell cultures were screened for the presence of HTLV-1 by PCR amplification of the tax gene using the primers: 5′-CGGATACCCAGTCTACGTGT-3′ and 5- GAGCCGATAACGCGTCCATCG-3′¹⁵  (Phusion blood direct PCR kit, Finnzymes). Cycling conditions were as follows: 98°C 5 minutes, 35 cycles each of 98°C 5 seconds, 65°C 5 seconds, 72°C 30 seconds; and final extension 72°C for 5 minutes. The PCR products were separated on a 2% agarose gel and visualized under UV light.

The high-throughput protocol was used as previously described.¹²  Fifty base-pair paired-end reads were acquired on an Illumina analyzer (either GAII or HiSeq 2000) and the insertion sites and shear sites deduced as described.¹² 

TCR gene rearrangements were analyzed in the Molecular Diagnostic Department, Imperial College Healthcare NHS Trust, London using the established BIOMED-2 protocol followed by analysis using GeneMapper 4.1 software (Applied Biosystems).

Twenty-seven HTLV-1–infected, CD4⁺ cultures were successfully expanded in vitro (Table 1). In 26 of these, a unique integration site constituted 99.9% (range, 95.5%-100%) of integrated provirus detected, with a median of 206 197 sequence reads (range, 6621-2 599 892), indicating that we had cultured 26 clones, each with a single dominant integration site. The remainder of the integration sites detected within each clone were of low frequency (median, 3 sequence reads): we attribute these rare integration events to the incomplete inhibition by raltegravir of HTLV-1 infectious spread during in vitro culture.

One HTLV-1⁺ culture, derived from asymptomatic carrier 2, contained cells with 2 equally abundant integration sites. This observation suggested either the presence of 2 HTLV-1 proviruses in a single T-cell clone or that 2 infected T-cell clones were cloned in the same well. Analysis of the TCRG gene rearrangement revealed 2 distinct gene rearrangements in the Vγ1-8 region (not illustrated), consistent with the presence of 2 T-cell clones.

The CD4⁺CD25⁺ T-cell clones isolated from the 2 patients with acute ATLL were distinct from the abnormally expanded, putatively malignant clone but were small clones contributing 0.12% and 1.6% of the PVL in PBMCs, respectively. This study was not designed to compare the mean proviral copy number between different disease states: such a study will require a larger number of persons with each respective condition.

In conclusion, the majority of nontransformed cells naturally infected with HTLV-1 contain a single copy of the HTLV-1 provirus in the host genome. In HIV infection, despite the high frequency of viral recombination observed and efficient cell-to-cell spread of multiple viruses, Josefsson et al¹⁷  recently estimated that > 85% of CD4⁺ HIV-1–infected cells contain a single integrated provirus. One possible explanation for the observed low mean proviral copy number is superinfection resistance (SIR). SIR is well documented in retroviruses,¹⁸  but the precise molecular mechanisms have not been unequivocally identified. In HIV-1 infection, down-regulation of CD4 from the infected cell surface by the HIV Nef protein may contribute to SIR. However, SIR becomes evident substantially earlier than CD4 down-regulation, suggesting that additional mechanisms may contribute to SIR in HIV-1 infection, including interference by host proteins with transport of the preintegration complex and blocking the HIV receptor.^18-20  The mechanisms of SIR in HTLV-1 infection have not been identified.

Because the number of distinct proviral integration sites in 1 host¹²  often exceeds 10⁴, the conclusion that each clone contains a single provirus implies that a typical host carries a very large number of HTLV-1–infected T-cell clones.

The authors thank the members of the Genomics Laboratory of the MRC Clinical Sciences Center, Hammersmith, London (Laurence Game, Adam Giess, Nathalie Lambie, and Ivan Andrew), the Molecular Diagnostic Department, Imperial College Healthcare NHS Trust (Mikel Valganon), and the patient donors in the HTLV-1 clinic at the National Center for Human Retrovirology, Imperial College Healthcare NHS Trust (St Mary's Campus).

This work was supported by Leukemia and Lymphoma Research and the Wellcome Trust.

Wellcome Trust

Contribution: L.B.C., A.G.R., and C.R.M.B. conceived and designed the research and wrote the manuscript; L.B.C. and A.G.R. performed the experiments and analyzed the data; A.M. developed essential bioinformatic data analysis tools; and G.P.T. recruited patients.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Charles R. M. Bangham, Section of Immunology, Department of Medicine, Imperial College London, St Mary's Campus, Wright-Fleming Institute, Norfolk Place, London W2 1PG, United Kingdom; e-mail: c.bangham@imperial.ac.uk.