The relevance of superinfection, that is, infection with a second strain, during the course of an established HIV-1 infection has been recently reconsidered1 after several reports about HIV-1-reinfected patients.2-6 These studies suggested that superinfection could be a frequent event in the natural history of HIV-1, particularly in individuals with high-risk practices.7 Almost all of the reported cases described superinfection a short time after primoinfection, indicating that only this period is susceptible to reinfection. However, Fang et al8 described a dual infection 11 years after primoinfection in a long-term nonprogressor.
In this work, we performed the genetic analysis in several plasma samples obtained during the clinical follow-up in an HIV-1-infected patient. This patient showed high-risk practices because of unsafe intravenous drug abuse and unprotected sexual contacts. Phylogenetic analysis of viral sequences obtained during the follow-up has permitted the identification of superinfection with 2 strains and the appearance of a recombinant virus in the viral population.
MATERIALS AND METHODS
The plasma samples analyzed were obtained from an HIV-1 patient diagnosed in 1987 who reported continuous intravenous drug use since 1981 and heterosexual contacts without protection. The patient follow-up in the AIDS unit of the hospital started in 1994, and since 1996, he began different antiviral regimens (Fig. 1). Voluntary treatment interruptions and poor adherence to treatment were assessed by direct interview and pharmacy records. The CD4+ T-cell count showed a continuous decrease from 616 cells per milliliter in the first sample analyzed in October 1994 to 55 cells per milliliter in February 2003. Viral load was more than the limits of detection in all but 3 samples studied, with values from 5.3 log to 1.7 log copies per milliliter. All these data are summarized in Figure 1.
Viral RNA was extracted from plasma samples following the method of Boom et al.9 It was reverse-transcribed using the Titan reverse transcriptase polymerase chain reaction (RT-PCR) kit (Roche). First PCR product was reamplified using the corresponding inner primers. Nucleic acid extraction, amplification, and cloning from each plasma sample were performed in individual experiments to exclude cross-contamination. To avoid population bottleneck, at least 200 copies of RNA were used to amplify by RT-PCR, except in the July 1999 sample because of the low viral load. All PCR amplifications were done with procedural safeguards and physical separation of sample processing and post-PCR handling steps. Moreover, contemporary laboratory control sequences were included in all phylogenetic analysis, and they never grouped with the patient sequence strains, discarding cross-contamination.
The emergence of drug-resistant mutations was studied by amplification of a fragment from the RT region using primers 191 RU (5′-GTTAAACAATGGCCATTGACAGAAGA-3′, 2610-2635 HXB2 position) and 41 RD (5′-AGCTGGACTGTCAATGACATACAGAA-3′, 3325-3300 HXB2 position) for the first PCR and 15 RU (5′-TAGATATCAGTACAATGTGCTTCCA-3′, 2975-2999) and 20 RD (5′-GCCAGAAAAAGACAGCTGGACTG-3′, 3287-3309 HXB2 position) for the nested. The C2 to C5 region in env gene was amplified as previously described.10
Polymerase chain reaction products were cloned following the TA cloning kit instructions (Invitrogen, Carlsbad, Calif). A mean of 16 clones per sample were sequenced in an ABI PRISM 377 automated sequencer (Perkin Elmer, Norwalk, Conn). Nucleotide sequences were analyzed using the Seqman II (DNAstar, Inc, Madison, WI) program, aligned using the Megalign (DNAstar Inc, Madison, WI) program (from the DNAstar package) and hand edited using the Bioedit (Carlsbad, CA) program.
Inasmuch as pol gene is under the pressure of the antiviral drug used in the therapy, the main resistant positions were excluded in the phylogenetic analysis. A first phylogenetic approximation was generated using the neighbor-joining method with the genetic distances estimated by the Kimura 2-parameter model (transition/transversion ratio, 2.0) in 1000 bootstrapped data sets as implemented in the MEGA (Kumar, Tamura, Nei, 2004) version 2.1 program. Afterward, the best-fit model for the phylogenetic reconstruction was estimated using a β test to compare up to 56 different models as implemented in the Modeltest 3.4 software.11 The derived parameters of the selected models for each region studied were the Hasega-Kishino-Yano (HKY85) + G model for the RT region and the general time reversible model + G for the env region. This model was used to reconstruct the phylogenetic trees by maximum likelihood (ML) as implemented in PHYML.12 The reliability of internal branches was tested using nonparametric bootstrap (1000 replicas). Recombination analyses were performed using the SimPlot version 3.5 program.13
Phylogenetic Analysis of the Samples in RT Region
Phylogenetic analysis of the RNA viral quasispecies derived from the RT region in pol gene was carried out in samples taken in October 1994 (10-94), February 1998 (2-98), April 1999 (4-99), July 1999 (7-99), June 2000 (6-00), and October 2003 (10-03) (Fig. 1). The composite ML tree with all the viral sequences is shown in Figure 2A. The samples from 1994, 1998, and April 1999 showed only 1 virus (represented by filled symbols), whereas 2 new clusters were defined in the July 1999 sample supported by 96% and 61% bootstrap values, respectively (Fig. 2A). The major group that was present in all the samples of the follow-up was designated cluster A. The other 2 clusters named B and C were detected only in the July 1999 sample, and clade B was also present in the next sample analyzed in June 2000 (Fig. 2B). These clusters did not group with any of the samples analyzed in the laboratory and marked with asterisks (Fig. 2A).
The mean genetic distance of the sequences between strains A and B was of 4% and with strain C of 5.4%. This genetic distance was very similar to the distance between the main virus and unrelated subtype B control viruses included in the phylogenetic analysis (4.1%), to other Spanish isolates (4,2%), 14 or to values obtained by Gonzales et al.15 Thus, phylogenetic analysis of several samples from an HIV-1-infected patient and the genetic distance between the sequences confirmed the presence of 3 different strains within the patient viral quasispecies. In summary, we propose that patient was primoinfected with strain A, and 12 years later (between April and July 1999), he was superinfected with 2 new strains (viruses B and C); this event is marked with an arrow in Figure 1.
To help in the differentiation of the viral populations, an analysis of resistance mutations and genetic polymorphisms in RT was performed. Clones belonging to the main A group did not show any resistance mutation except for the April 99 sample in which the mutation M184V was detected in all clones (Fig. 3A) after a 3-month period of lamivudine (3TC) + stavudine (d4T) + nelfinavir (NFV) treatment (Fig. 1). Virus B showed in all clones analyzed the azidothymidine (AZT) resistance mutation T215Y together with the M184V, 3TC-associated resistance mutation. The other superinfecting virus (strain C) presented the V179D polymorphism that was not present in the other 2 viruses. All these different sequences are shown in Figure 3A. The distinct mutation patterns displayed by the 3 viruses supported the identification of 3 different strains in the patient.
Phylogenetic Analysis of the Samples in env Gene
To confirm the existence of 3 different viruses, phylogenetic analysis of the C2 to C5 region in env gene was carried out in 4 samples (4-99, 7-99, 6-00, and 10-03). The ML phylogenetic tree showed the existence of 2 groups of sequences (named "a" and "b") and a small subcluster within the main group (marked with a question mark in Fig. 4A) linked to strain "a" with a high value of bootstrap (80%). SimPlot analysis of the sequences of this small subgroup showed that these viruses were recombinant between the main "a" group and a different strain (that we named "c"), with a crossing point around position 230 to 240 of the fragment (Fig. 4B). The phylogenetic analysis of these first 240 nucleotides showed 3 distinct groups designated a, b, and c separated by high values of bootstrap (99% for each group) (Fig. 4C). In the phylogenetic analysis of the 3` end of this fragment, only 2 groups were defined (data not shown). The amino acid sequence derived from the C2 to C5 region in env gene supported the presence of 3 strains in the patient because of the high number of differences among them (Fig. 3B).
It was interesting to focus on the genetic distance calculated in env gene between the different viruses detected in the study. The main group (a) showed a genetic distance of 18.2% to virus b and of 14.3% to recombinant virus (?). These genetic distances were very high compared with the mean value of 13.9% obtained in a set of 79 Spanish isolates.16 The values of genetic distances between the different viruses obtained in the patient also confirmed the presence of 3 distinct isolates.
The phylogenetic analysis of viral nucleotide sequences in the RT region in pol gene and in env gene has permitted the identification for the first time of a dual superinfection in an HIV-1 patient showing high-risk practices.2,4,5,17
Because of the sensitivity of the quasi-species analysis performed, which cannot detect variants that are less than 5% in the viral population, we could not exclude that the presence of 3 strains was not due to a triple coinfection. However, several findings supported superinfection over a triple primo-coinfection. On the one hand, if coinfection by 3 viruses had occurred during primoinfection, because of the long time between primoinfection and the date of the samples analyzed and the high frequency of genetic recombination in HIV-1, it would have been very difficult to observe distinct clades in the phylogenetic trees. Moreover, the extremely high genetic distance observed between the different clades (14.3%-18.2% in env gene) could not be explained as the result of the evolution of the virus within the patient because these values have been found only in coinfected or reinfected patients.18
The resistance mutations analysis gave a strong support for the exclusion of the multiple primoinfection hypothesis. It is extremely unlikely that the patient, primoinfected between 1981 and 1987, was infected with a virus such as strain B displaying the T215Y + M184V resistance mutations associated with the AZT + 3TC therapy. The general instauration of this regimen occurred in 1995 in Spain, many years after the patient primoinfection.
The existence of 3 different viruses in the patient viral quasispecies was studied by phylogenetic analysis of the sequences in the C2 to C5 region in env gene. The 3 distinct clades (a, b, and c) could only be confirmed in the first 230 nucleotides of this env fragment (Fig. 4C). The third virus was a recombinant genome (Fig. 4B) between one of the superinfecting strains (virus c) in the 5 end and the main (a) virus in the 3 end. These recombinant viruses were not the consequence of PCR recombination, frequently found in RT-PCR, because the same recombinant was obtained in the proviral DNA in a later sample (data not shown).
The data obtained in this study are compatible with patient being primoinfected with 1 strain and, 12 years later, being superinfected with 2 strains. We could not definitely identify if the presence of the 2 new strains in the June 1999 sample was due to 2 independent infection events or to a dual superinfection. However, the short time between the presence of only 1 strain in April 1999 and, 3 months later, the detection of 3 strains favors dual coinfection. This criterion was used in the definition of dual infections.19
It was thought that individuals were susceptible to reinfection only during the window period, which is a short time after primary infection. Superinfection in patients during chronic infection seems to be a rare event as suggested in longitudinal studies.15,20 However, in this work, we detected superinfection 12 years after the primary infection, during the AIDS phase. A similar result has been recently described in a nonprogressor patient.8
Two main consequences could be derived from this study. From a virological point of view, this work reports the first case of a dual superinfection that, together with a recombination event between the infecting strains, provides evidence for the extremely high viral genetic diversity and complexity that can be observed in HIV-1 in vivo infections. From the clinical perspective, we have demonstrated that superinfection (with >1 strain) could occur at any moment during the infection. These results highlight the importance of reducing risky behaviors in HIV-1-infected individuals.
The authors appreciate the cooperation of the patient as well as of the physicians in Subdireción General de Sanidad Penitenciaria and in Instituciones Penitenciarias in Valdemoro and Alama and the helpful comments of 2 anonymous reviewers. Work in Centro Nacional de Microbiologia was supported by grants MPY 1359/02, MPY 1028/04 by the Plan Nacional del SIDA and in part by the "Red Tematica Cooperativa de Investigación en SIDA (Red de grupos 173) del FISss."
1. Allen TM, Altfeld M. HIV-1 superinfection. J Allergy Clin Immunol
. November 2003;112(5):829-836.
2. Ramos A, Hu DJ, Nguyen L, et al. Intersubtype human immunodeficiency virus type 1 superinfection following seroconversion to primary infection in two injection drug users. J Virol
3. Altfeld M, Allen TM, Yu XG, et al. HIV-1 superinfection despite broad CD8+
T-cell responses containing replication of the primary virus. Nature
. November 28, 2002;420(6914):434-439.
4. Jost S, Bernard MC, Kaiser L, et al. A patient with HIV-1 superinfection. N Engl J Med
. September 5, 2002;347(10):731-736.
5. Koelsch KK, Smith DM, Little SJ, et al. Clade B HIV-1 superinfection with wild-type virus after primary infection with drug-resistant clade B virus. AIDS
. May 2, 2003;17(7):F11-F16.
6. Yang OO, Daar ES, Jamieson BD, et al. Human immunodeficiency virus type 1 clade B superinfection: evidence for differential immune containment of distinct clade B strains. J Virol
. January 15, 2005;79(2):860-868.
7. Blackard J, Ta KHM. HIV superinfection in the era of increased sexual risk taking. Sex Transm Dis
8. Fang G, Weiser B, Kuiken C, et al. Recombination following superinfection by HIV-1. AIDS
. January 23, 2004;18(2):153-159.
9. Boom R, Sol CJA, Salimans MMM, et al. Rapid and simple method for purification of nucleic acids. J Clin Microbiol
10. Bello G, Casado C, Garcia S, et al. Co-existence of recent and ancestral nucleotide sequences in viral quasispecies of human immunodeficiency virus type 1 patients. J Gen Virol
. February 2004;85(pt 2):399-407.
11. Posada D, Crandall KA. Selecting models of nucleotide substitution: an application to human immunodeficiency virus 1 (HIV-1). Mol Biol Evol
. June 2001;18(6):897-906.
12. Guindon S, Lethiec F, Duroux P, et al. PHYML online-a Web server for fast maximum likelihood-based phylogenetic inference. Nucleic Acids Res
. July 1, 2005;33:W557-W559. [Web server issue].
13. SimPlot for Windows 95/NT
[computer program]. Version 2.4. Baltimore, MD: Distributed by author (S. Ray); 1999.
14. Pariente N, Pernas M, de la Rosa R, et al. Long-term suppression of plasma viremia with highly active antiretroviral therapy despite virus evolution and very limited selection of drug-resistant genotypes. J Med Virol
. July 2004;73(3):350-361.
15. Gonzales MJ, Delwart E, Rhee SY, et al. Lack of detectable human immunodeficiency virus type 1 superinfection during 1072 person-years of observation. J Infect Dis
. August 1, 2003;188(3):397-405.
16. Casado C, Urtasun I, Martin-Walther MV, et al. Genetic analysis of HIV-1 samples from Spain. J Acquir Immune Defic Syndr
17. Yerly S, Jost S, Monnat M, et al. HIV-1 co/super-infection in intravenous drug users. AIDS
. July 2, 2004;18(10):1413-1421.
18. Zhu T, Wang N, Carr A, et al. Evidence for coinfection by multiple strains of human immunodeficiency virus type 1 subtype B in an acute seroconvertor. J Virol
. February 1995;69:1324-1327.
19. Gottlieb GS, Nickle DC, Jensen MA, et al. Dual HIV-1 infection associated with rapid disease progression. Lancet
. February 21, 2004;363(9409):619-622.
20. Tsui R, Herring BL, Barbour JD, et al. Human immunodeficiency virus type 1 superinfection was not detected following 215 years of injection drug user exposure. J Virol
. January 2004;78(1):94-103.