| |
A Case of Superinfection
|
| |
| |
"In-Depth Analysis of a Heterosexually Acquired HIV-1 Superinfection: Evolution, Temporal Fluctuation, and Intercompartment Dynamics from the Seronegative Window Period through 30 Months Postinfection"
Journal of Virology
September 2005
F. E. McCutchan,1* M. Hoelscher,2 S. Tovanabutra,1 S. Piyasirisilp,1 E. Sanders-Buell,1 G. Ramos,1 L. Jagodzinski,1 V. Polonis,1 L. Maboko,3 D. Mmbando,4 O. Hoffmann,2,5 G. Riedner,5 F. von Sonnenburg,2 M. Robb,1 and D. L. Birx1
U.S. Military HIV Research Program, 1600 E. Gude Drive, Rockville, Maryland 20850,1 Department of Infectious Diseases and Tropical Medicine, Ludwig-Maximilians-University, Leopoldstrasse 5, Munich, Germany,2 Mbeya Medical Research Programme, Mbeya, Tanzania,3 Mbeya Regional Medical Office, Ministry of Health, Mbeya, Tanzania,4 London School of Hygiene and Tropical Medicine, 50 Bedford Square, London WC1B 3DP, United Kingdom5
ABSTRACT
Human immunodeficiency virus type 1 (HIV-1) superinfection refers to the acquisition of another strain by an already infected individual. Here we report a comprehensive genetic analysis of an HIV-1 superinfection acquired heterosexually. The infected individual was in a high-risk cohort in Tanzania, was exposed to multiple subtypes, and was systematically evaluated every 3 months with a fluorescent multiregion genotyping assay. The subject was identified in the window period and was first infected with a complex ACD recombinant strain, became superinfected 6 to 9 months later with an AC recombinant, and was monitored for >2.5 years. The plasma viral load exceeded 400,000 copies/ml during the first 9 months of infection but resolved to the set point of 67,000 copies/ml by 3 months after superinfection; the CD4 cell count was 377 cells/µl at 30 months. Viral diversity was evaluated with techniques designed to fully sample the quasispecies, permitting direct observation of the evolution, temporal fluctuation, and intercompartment dynamics of the initial and superinfecting strains and recombinants derived from them. Within 3 months of superinfection, seven different molecular forms were detected in gag and six were detected in env. The proportions of forms fluctuated widely over time in plasma and peripheral blood mononuclear cells, illustrating how challenging the detection of dually infected individuals can be. Strain-specific nested PCR confirmed that the superinfecting strain was not present until the 9 month follow-up. This study further defines the parameters and dynamics of superinfection and will foster appropriate studies and approaches to gain a more complete understanding of risk factors for superinfection and its impact on clinical progression, epidemiology, and vaccine design.
Superinfection or cotransmission? Is it possible that the superinfecting strain was transiently present and/or present at an extremely low abundance from the beginning and that this case represents cotransmission rather than superinfection? This question was addressed by the development and application of a strain-specific nested PCR assay. Figure S7 in the supplemental material shows the design of the assay and its validation with clones from the two strains. Figure 6 shows the amplification of serial samples from PBMC and plasma. For both compartments, the initial strain was easily detected from the earliest time points, whereas the superinfecting strain was not amplified until visit 3.
AUTHOR DISCUSSION
This study contributes to a growing body of data suggesting that infection with HIV-1 does not necessarily protect against reinfection with another HIV-1 strain. It will be important to determine whether superinfection is the result of an immune system that is weakened or damaged by HIV-1 infection itself if HIV-1 infection in humans parallels recent work with the simian immunodeficiency virus (SIV) model system (28). An early and sometimes severe depletion of memory CD4+ T cells in the first weeks of infection may abrogate the defense against many pathogens, even those to which the host had established memory responses; HIV-1 superinfection may be part of this generalized immune impairment.
Alternatively, superinfection may reflect a failure of cross-protective immunity resulting from the genetic diversity of HIV-1. Superinfection might be expected to occur more readily when genetically diverse strains, such as different subtypes, challenge the already-infected host. Previous reports have documented superinfections both with strains of the same subtype and with strains of different subtypes; this report presents an intermediate case, with recombinant strains partly of the same subtype. A recent report by Yang et al. (41), like an earlier study (34), describes superinfection in an individual where T-cell responses to the initial strain were limited in the ability to recognize the superinfecting strain. However, superinfection can occur even in individuals with a broad T-cell response (2).
There is indirect evidence that participant 123 did subject the superinfecting strain to immune pressure. A key observation is the virtual replacement of the initial strain with a new recombinant in gp41/nef. The rapid turnover of the viral quasispecies in PBMC after superinfection could also reflect continuing immune pressure, or this could be the result of random expansion of latent viral variants. The fact that the superinfecting strain, which must have been vastly in the minority at the moment of superinfection, came to have a relatively high abundance in the viral quasispecies in plasma, along with recombinants derived from it, is also presumptive evidence of strong selection. Finally, participant 123 did gain control of an initially high viral load and was able to establish and maintain a set point; it is doubtful that this could have occurred without some immune control of the superinfection.
The potential public health importance of superinfection, with its capacity to rapidly generate new and potentially transmissible recombinant forms (10) and a higher viral load set point in dually infected individuals than in singly infected individuals (13), has already been noted. It appears that superinfection can generate not just one but many new recombinant forms, some of which persist at appreciable levels in the host for years. The link between dual infection and recombinant forms is now strongly forged by this and other studies, and given the established relationship between viral load and transmission (5, 32, 33), it may be that recombinants are generated in individuals who also develop higher viral loads and readily transmit them. Eventually, superinfection would be expected to increase the complexity of viral genotypes circulating in the population. Specific and concerted interventions in high-risk populations, who may be particularly susceptible to superinfection, may have a population-level as well as individual benefit.
Does HIV-1 superinfection have direct, negative implications for vaccine development? The most important data to answer this question may come from clinical trials of candidate HIV-1 vaccines in seronegative volunteers because this is the best opportunity to measure the protective effect of immune responses generated with a defined immunogen in the context of a healthy immune system. It will also be important to determine what proportion of individuals harboring two or more HIV-1 strains acquired them simultaneously at the time of first infection because these cases do not pertain to the issue of reinfection in the face of an ongoing anti-HIV-1 immune response. This report and others do establish that superinfection is possible, but additional research will be needed to interpret superinfection in the context of anti-HIV-1 immunity and vaccine development.
INTRODUCTION
At every replication cycle, human immunodeficiency virus type 1 (HIV-1) recombines the two genomic RNA molecules packaged in the virion through the mechanism of strand switching at reverse transcription (7, 19). HIV-1-infected cells harbor, on average, four separate proviruses, whose genomic RNA transcripts have the opportunity to assort at the packaging stage and are recombined during the next replication cycle (21). Recombination reshuffles point mutations arising in the infected individual (40) but gains new significance when individuals become infected with more than one HIV-1 strain, either from the same or from different subtypes. In this case, recombination provides for a series of adjacent mutations to be acquired (or discarded) simultaneously (36). Intersubtype recombinant HIV-1 strains, which are easier to detect than their intrasubtype counterparts, have been widely documented (29) and have become a major force in the pandemic (24). Sixteen circulating recombinant forms (CRFs) and hundreds of unique recombinant forms (URFs), the latter being isolated only from single individuals, have been identified. Some regional epidemics are dominated by a single CRF, while others have mostly URFs. In East Africa, URFs account for 30 to 50% of infections, while CRFs have been rare (3, 8, 15, 17). Epidemics in which CRFs dominate include those in West and West Central Africa, where CRF02_AG accounts for at least 50% of infections (22), and Southeast Asia, where CRF01_AE is the predominant strain (39).
Approaches for the molecular epidemiology of HIV-1 have been progressively adapted to better detect coinfections and recombination. The complete sequencing of HIV-1 genomes has led the effort to characterize recombinant strains (6, 30, 38). Multiregion hybridization assays (MHAs) provide an accurate, high-throughput, and less labor-intensive approach to the detection of recombinants and coinfected individuals. Three different MHAs are in development (16, 22; G. Kijak, S. Tovanabutra, et al., Abstr. XV Int. AIDS Conf., abstr. MoPeC3415, 2004), with each tailored to the mixture of subtypes in a specific geographic region. MHAs provide the capacity to accurately genotype hundreds of samples in a short time period, permitting comprehensive comparisons of different populations and epidemics. Recent work in Tanzania has shown an association between the risk for HIV-1 infection, HIV-1 prevalence, the proportion of strains that are URFs, and the fraction of individuals that may be coinfected with multiple subtypes (4; M. Hoelscher, M. Arroyo, et al., Abstr. AIDS Vaccine 04, abstr. 21, 2004; K.-H. Herbinger, M. Gerhardt, et al., submitted). Clearly, coinfection in populations is a potential source of rapid viral evolution by recombination and an increasing genetic complexity of strains, and some of the tools are in place to determine the impact of coinfection on HIV-1 epidemics and on infected individuals.
More recently, the parameters of HIV-1 coinfection and recombination at the level of the infected cell have been better defined. The demonstration that infected cells harbor multiple proviruses in their genome was an important milestone because it confirmed that the conditions for interstrain recombination are met at the cellular level (21). By the incorporation of fluorescent reporter genes into HIV-1 strains, it has become possible recently to directly measure rates of recombination, and the parameters that affect these rates, in an in vitro system (25). The number of crossovers per replication cycle is much higher than previously thought, estimated at 9 per replication round in T lymphocytes and up to 30 per round in macrophages. The same study showed that multiple infections of cells occur freely and that the generation of recombinants is proportional to the square of the infection rate.
The large proportion of unique recombinant forms in some mixed-subtype epidemics, which can exceed 50% of strains, is suggestive, if not prima facie, evidence that coinfection does occur at an appreciable frequency. A recent mathematical model suggests that even a low rate of coinfection would be sufficient to explain an accumulating, substantial fraction of recombinant strains in an epidemic (14). The major gaps in knowledge are the circumstances that permit coinfection to occur and the dynamics of viral evolution in coinfected individuals.
To date, there have been several well-documented HIV-1 superinfections, and because of their implications for antiviral immunity and vaccines, these reports have precipitated considerable analysis and debate (1, 10, 12, 26). The first report concerned an individual with MSM (men having sex with men) risk who was initially infected with CRF01_AE and later acquired subtype B (20). The interval between infection and superinfection was about 2 years. Two other MSM with superinfection were reported shortly thereafter, but both of these, initially infected with subtype B, acquired a second subtype B strain, one 4 months and the other 32 months after initial infection (2, 23). Two injecting drug users (IDUs) with superinfection were detected in Thailand (34), one of whom acquired subtype B within 3 months of CRF01_AE infection and the other of whom acquired CRF01_AE almost 3 years after an initial infection with subtype B. All of these superinfected individuals were at high risk for exposure to multiple HIV-1 strains or subtypes. The two cases of IDUs were the result of systematic surveillance in a cohort of 130 individuals, whereas those in MSM were identified because of clinical parameters such as an abrupt elevation in viral load during treatment or the emergence of new drug-resistant strains.
Here we report a comprehensive analysis of a superinfected individual, identified in Africa in the context of a well-designed and systematic cohort study of 600 individuals with heterosexual risk, the HIV Superinfection Study (HISIS). The approaches that permitted the successful development and retention of this cohort, as well as some of its social and behavioral characteristics, have been reported (18, 35). Other pending reports from HISIS include descriptions of HIV-1 infection in the cohort and its parameters (M. Hoelscher, O. Hoffmann, et al., submitted for publication) and of other coinfected individuals who were also studied in substantial molecular detail (11; S. Piyasirisilp, S. Tovanabutra, et al., unpublished data). Participant 123, who is the subject of this report, was ascertained in the seronegative window period, after HIV-1 infection but before seroconversion, and was identified as a potential superinfection case by MHAacd genotyping every 3 months for 30 months.
|
|
| |
| |
|
|
|
