The editorial by Bushman  on HIV DNA two long terminal repeat (2-LTR) circles concluded that they are not a valid or useful new surrogate marker for monitoring persistent viral replication in patients whose plasma HIV-1-RNA level has been reduced to undetectable levels by highly active antiretroviral therapy (HAART). The author cited experiments performed in his laboratory and in the collaborating laboratory of Siliciano using transformed cell lines, vectors and X4 laboratory adapted isolates of HIV-1 [2,3]. He concluded that 2-LTR circles are intrinsically stable and that the decline in 2-LTR circles when SupT1 cells are infected can be accounted for by syncytia formation and cell death. However, when a non-cytopathic HIV-based vector is used, the circles decline in proportion to their dilution during cell division. Therefore, different results were obtained depending upon whether a virus or a vector was used to perform the experiments.
There are important distinctions between his and our observations [4,5]. In his half-life experiments, 6–10 copies of 2-LTR circles/cell were generated using spinfection methods. In contrast, in our experiments there were 0.033 ± 0.005 copies of 2-LTR circles/cell (mean ± sem) when peripheral blood mononuclear cells (PBMC) were infected with primary patient-derived R5 virus in vitro, and only 0.0003 ± 0.0001 copies of 2-LTR circles/cell in PBMC infected in vivo. It is therefore possible that a ‘physiological’ copy number of 2-LTR circles can be degraded by the cell whereas artificially large numbers of 2-LTR circles cannot. In addition, the marked syncytium-induced cell death that is seen in Sup T1 cells is rarely seen in PBMC. Pierson et al.  used the HTLV-1 transformed MT-2 cell line infected with a monotropic X4 virus, and maintained these cultures in a logarithmic phase of growth throughout their experiments. These culture conditions are not representative of the much slower growth characteristics of primary cells. Finally, we have found that indinavir monotherapy, as used by Pierson et al.  to prevent secondary rounds of HIV-1 infection, does not completely abolish the synthesis of new RU5 complementary DNA transcripts in PBMC. Triple antiretroviral therapy was required .
In a retrospective study, Brussel et al.  observed that the 2-LTR circle copy number in the PBMC of a small group of patients on HAART remained stable over a period of several years. The authors did not present data on circle half-life in vitro. Although Bushman  concluded that their observations are different and contradictory to ours, we believe that the difference lies in the interpretation of our respective results. Table 1 summarizes and compares their methodology and results with ours. The overall observations are remarkably similar. Although Brussel et al.  concluded that the circle copy number in the PBMC of treated patients is stable over time (half-life 3.2 months to infinity), they did acknowledge that this could be caused by ongoing de novo infection continuously replenishing the 2-LTR circle copy number in PBMC, as well as the stability of 2-LTR circles in long-lived non-dividing cells. Our identification of a group of patients with a plasma HIV-RNA level of less than 50 copies/ml on HAART, who went from being circle positive to being circle negative, and who then remained circle negative for a period of many months, led us to conclude that 2-LTR circles must degrade over time in vivo. Furthermore, the long-term persistence of 2-LTR circles in other patients whose plasma HIV-RNA level fell to and was maintained at less than 50 copies/ml on HAART suggested that 2-LTR circles were being replenished by acute infection events that were taking place in extravascular tissue-based reservoirs .
We want to emphasize that most of our studies on 2-LTR circles have been performermed on cells from patients who were infected by HIV-1 in vivo, and who had been monitored for several years. The validity of the argument of Bushman  is based upon a one-week analysis in vitro. Consequently, we remain of the opinion that the validity and pathophysiological significance of a new surrogate marker for monitoring persistent viral replication in patients cannot be easily or simply established using the experimental systems referred to by Bushman .
1. Bushman F. Measuring covert HIV replication during HAART.AIDS
2. Butler SL, Johnson EP, Bushman FD. HIV cDNA metabolism: notable stability of two-long terminal repeat circles.J Virol
3. Pierson TC, Kieffer TL, Ruff CT, Buck C, Gange SJ, Siliciano RF. Intrinsic stability of episomal circles formed during HIV-1 replication.J Virol
4. Teo IA, Morlese J, Choi JW, Shaunak S. Reliable and reproducible LightCycler qPCR measurement of HIV DNA circles.J Immunol Methods
5. Morlese J, Teo IA, Choi JW, Gazzard B, Shaunak S. Identification of two mutually exclusive groups after long-term monitoring of HIV DNA 2-LTR circle copy number in patients on HAART.AIDS
6. Choi JW. Development and characterization of a new model of chronic HIV infection.
PhD thesis. University of London, London; 2002.
7. Brussel A, Mathez D, Broche-Pierre S, Lancar R, Calvez T, Sonigo P, Liebowitch J. Longitudinal monitoring of 2-LTR circles in the peripheral blood mononuclear cells from patients with chronic HIV-1 infection.AIDS
8. Teo I, Veryard C, Barnes H, An S, Jones M, Lantos P, et al
. Circular forms of unintegrated HIV-1 DNA and high levels of viral protein expression: association with dementia and multinucleated giant cells in the brains of patients with AIDS.J Virol