Longitudinal Analysis of Human Immunodeficiency Virus Type 1 nef/Long Terminal Repeat Sequences in a Cohort of Long-Term Survivors Infected from a Single Source

Abstract

We studied the evolution of human immunodeficiency virus type 1 (HIV-1) in a cohort of long-term survivors infected with an attenuated strain of HIV-1 acquired from a single source. Although the cohort members experienced differing clinical courses, we demonstrate similar evolution of HIV-1 nef/long-terminal repeat (LTR) sequences, characterized by progressive sequence deletions tending toward a minimal nef/LTR structure that retains only sequence elements required for viral replication. The in vivo pathogenicity of attenuated HIV-1 is therefore dictated by viral and/or host factors other than those that impose a unidirectional selection pressure on the nef/LTR region of the HIV-1 genome.

The nef gene is a major determinant of virulence in primate lentiviruses. Mutations in nef attenuate replication capacity and pathogenicity of simian immunodeficiency viruses (SIV) (3, 7, 10, 11, 13, 19) and may promote long-term survival of human immunodeficiency virus type 1 (HIV-1) infection in humans (6, 14, 15, 18, 20).

The largest described cohort of long-term survivors is the Sydney Blood Bank Cohort (SBBC), which consists of multiple individuals who became infected with an attenuated strain of HIV-1 via contaminated blood products from a common blood donor between 1981 and 1984 (6, 16, 17). Viral attenuation has been attributed to gross deletions in the nef/long terminal repeat (LTR) region of the HIV-1 genome (6). Despite being infected from a single source, SBBC members comprise slow progressors (SP) and long-term nonprogressors (LTNP) (2, 4, 17). The SBBC therefore provides an unprecedented opportunity to study the pathogenesis of nef/LTR-deleted HIV-1 infection in a naturally occurring human setting.

Subjects.

We undertook a longitudinal study of SBBC SP and LTNP. The clinical history of the study subjects from infection to 1998 has been described (17). Subjects C54, C98, C49, C64, and C135 were referred to previously as recipients 13, 7, 12, 9, and 4, respectively (17). Subjects C54 and C98 have since died from causes unrelated to HIV-1 infection. An update on the results of laboratory studies and clinical history of these two subjects and of the surviving SBBC members studied in this report is summarized in Table 1 and detailed below.

TABLE 1.

Subjects, laboratory studies, and clinical history

Subject, sex, and date of Birtha	Date transfuseda	Date of blood samplea	Yr since infection	CD4+ T-cell count (cells/μl)b	Viral load (RNA copies/ml)c	Antiretroviral drugsd	Clinical historye
D36, male, 6/4/1958	Not applicable; infected with HIV-1 via sexual transmission 12/1980 (18)	2/1996	15.2	609	1,100	ABC, AZT, NVP (1/1999-9/2004) (4, 5); ABC, NVP, 3TC (9/2004-present)	SP; diagnosed with HIVD 12/1998 (4)
		4/1996	15.3	504	7,800
		10/1997	16.8	336	4,400
		1/1999	18.1	210	9,900
		9/2000	19.8	391	BD
		4/2001	20.3	476	BD
		2/2003	22.2	624	BD
		2/2004	23.2	638	BD
C49, female, 6/9/1954	6/11/1984	2/1994	9.7	1,045	BD	None	LTNP; diagnosed with age-onset diabetes in 2004, managed by diet; chronic alcoholism
		5/1994	9.8	1,458	BD
		10/1996	12.1	1,134	BD
		12/1997	13.5	918	BD
		6/1999	15.0	605	BD
		11/2001	17.4	624	BD
		8/2002	18.2	468	BD
		3/2004	19.7	874	BD
C54, male, 2/17/1928	7/24/1984	7/1993	9.0	1,519	N/A	None	LTNP; IDDM; surgery for colon cancer in 1995; died 8/28/2001 from myocardial infarct; death not related to HIV-1
		6/1995	10.9	1,504	3,000
		3/1996	11.7	1,188	1,500
		9/1996	12.2	1,120	1,800
		5/1997	12.8	1,286	5,500
		8/1997	13.1	1,419	1,700
		3/2000	15.7	840	1,600
		5/2001	16.8	1,537	2,660
C64, female, 3/20/1926	5/4/1983	8/1996	13.3	925	BD	None	LTNP; hypertension; hypercholesterolemia
		2/1997	13.8	851	BD
		5/1997	14.0	1,050	BD
		8/1997	14.3	805	BD
		11/1997	14.5	936	BD
		4/1999	15.9	1,026	BD
		11/1999	16.5	1,332	BD
		5/2000	17.0	875	BD
C98, male, 7/11/1937	1/2/1982	7/1993	11.5	880	N/A	d4T, NVP, IND (11/1999-death)	SP; prednisone since 1995 for asthma; died 3/30/2001 from bronchial amyloidosis; death not related to HIV-1
		10/1995	13.8	576	670
		11/1996	14.8	646	690
		5/1997	15.3	527	760
		2/1998	16.1	627	1,100
		10/1998	16.8	429	1,500
		3/2000	18.2	684	BD
		3/2001	19.6	324	BD

^aDates shown are month/day/year. The dates refer to times when PBMC were collected for HIV-1 nef/LTR sequencing. The results of only those laboratory studies that correspond to these time points are shown.

^bCD4⁺ T-cell levels were measured by flow cytometry.

^cPlasma HIV-1 RNA was measured using COBAS AMPLICOR HIV-1 monitor version 1.0 (Roche Molecular Diagnostic Systems, Branchburg, N.J.) prior to July 1999 and version 1.5 after July 1999. HIV-1 RNA levels of <400 copies/ml (version 1) or <50 copies/ml (version 1.5) were considered below detection. BD, below detection; N/A, not available.

^dABC, abacavir; AZT, zidovudine; NVP, nevirapine; 3TC, lamivudine; d4T, stavudine; IND, indinavir.

^eHIVD, HIV-associated dementia; IDDM, insulin-dependent diabetes melitis.

SBBC subjects with slowly progressing HIV-1 infection include the donor (D36) and transfusion recipient C98. D36 commenced highly active antiretroviral therapy (HAART) in January 1999 following HIV-associated dementia that coincided with a fall in the CD4 cell count to <200 cells/μl and the presence of high plasma and cerebrospinal fluid (CSF) HIV-1 RNA levels (4, 5). As reported previously, C98 commenced prednisone for treatment of asthma in 1995 (17). C98 was diagnosed with pulmonary amyloidosis in 1998. C98 commenced HAART in November 1999 after experiencing a steady decline in his CD4⁺ T-cell count and a gradual increase in HIV-1 RNA from below-detectable levels to 1,500 RNA copies/ml. During 2001 his CD4⁺ T-cell count declined and fluctuated between 213 and 484 cells/μl despite the continuance of HAART and a viral load below detectable levels. He died at the age of 64 in March 2002 of amyloidosis, which was not HIV-1 related.

Nonprogressing SBBC subjects include transfusion recipients C49, C54, C64, and C135; these subjects have experienced steady CD4⁺ T-cell counts since infection was first identified, with median values of >900, >1,000, >900, and > 500 cells/μl, respectively. HIV-1 RNA levels have remained consistently low or below detectable levels in these subjects despite being infected for up to 20 years without antiretroviral intervention. Patient C54 died aged 73 from a myocardial infarct in September 2001. C135 was not included in the present study because of the lack of consistent amplification of HIV-1 DNA.

Methods.

Peripheral blood mononuclear cells (PBMC) were obtained at each of the times indicated in Table 1 according to guidelines endorsed by the Australian Red Cross Blood Service human ethics committee. The nef/LTR region of the HIV-1 genome was amplified from PBMC by nested PCR, as described previously (4, 20). The products of six independent PCRs were pooled and cloned into pGEM-T-Easy (Promega, Madison, WI), and the nucleotide sequences of multiple independent clones were determined using a SequiTherm EXCEL II DNA sequencing kit (Epicenter Technologies, Madison, WI) and a model 4000L LI-COR DNA sequencer (LI-COR, Lincoln, NE). Nucleotide sequences were aligned and analyzed using DNAMAN software (Lynnon, Quebec, Canada).

Results.

Intersubject nef/LTR sequences cloned from the earliest available PBMC samples were heterogenous but contained a number of common mutations (Fig. 1A): (i) deletions of various lengths in the amino terminus of the nef gene, (ii) at least one deletion in the nef/LTR overlap region, and (iii) duplication and/or rearrangement of nuclear factor-κB (NF-κB) and specificity factor 1 (Sp-1) binding sites in the LTR (Fig. 1B). The 3′-most deletion in the nef/LTR overlap region shared by all viruses is thought to have been present in the transmitted virus, as none of the subjects has antibodies to a peptide of the corresponding region in the Nef protein (9). None of the viruses are capable of encoding Nef, carrying either an in-phase termination codon (D36) or lacking the nef ATG. The duplicated or rearranged region of the LTR was unique to each of the cohort members and varied with respect to the number and arrangement of NF-κB and Sp-1 sites inserted (Fig. 1). Thus, despite the common origin of the viruses, the nef/LTR regions differed from subject to subject.

FIG. 1.

Analysis of SBBC nef/LTR sequence. (A) Schematic representation of nef/LTR sequence deletions of HIV-1 cloned from the earliest available PBMC samples. The data shown represent a consensus of at least 20 independent nef/LTR sequences cloned from each PBMC sample. The genomic structures are compared to wild-type HIV-1 (NL4-3). Numbers refer to nucleotide positions in NL4-3. Black blocks represent intact sequence, and gaps represent deletions. Gray blocks represent the sequence area containing duplicated and rearranged NF-κB and Sp-1 binding sites in the LTR. The dates shown represent the times when PBMC were collected for analysis. (B) More-detailed analysis of the LTR, depicting transcription factor binding sites. PPT, polypurine tract; NRE, negative regulatory element.

Sequential analysis of nef/LTR sequences spanning a 4- to 10-year period demonstrated a further loss of nef sequence that differed in magnitude between subjects (Fig. 2). A large deletion of 128 bp emerged in D36, effectively removing the entire nef gene with the exception of the region surrounding the nef start codon, the polypurine tract which contains terminal signals for HIV-1 integration, and a 90-bp region of the nef/LTR overlap region surrounding the negative regulatory element. The basal promoter and enhancer elements of the LTR were retained, but a 44-bp deletion appeared in the rearranged or duplicated NF-κB/Sp-1 motif. The pattern of nef sequence loss in C98 was remarkably similar to that which occurred in D36. The pattern of nef sequence loss in C54 was also similar but was less extensive than that observed in D36 and C98. However, the additional loss of nef sequence in C64 was comparatively minimal. Thus, viruses harbored by D36, C54, C98, and C64 appear to be evolving in a convergent fashion toward a highly deleted, minimal nef/LTR structure containing only sequence elements that are absolutely essential for HIV-1 replication. The convergent nature of the nef/LTR evolution is further illustrated in Fig. 3, where sequences from the earliest and most recent nef/LTR clones from each subject are compared. The convergent nature of the nef/LTR sequence changes implies the presence of strong selection pressures that maintain the ability of defective HIV-1 genomes to persist in vivo.

FIG. 2.

Evolution of SBBC nef/LTR sequence deletions. Schematic representation of nef/LTR sequence deletions of HIV-1 cloned from longitudinally collected PBMC samples from SBBC long-term nonprogressors (A) and slow progressors (B). The data shown represent a consensus of at least 20 independent nef/LTR sequences cloned from each PBMC sample. The genomic structures are compared to wild-type HIV-1 (NL4-3). Numbers refer to nucleotide positions in NL4-3. Black blocks represent intact sequence, and gaps represent deletions. Gray blocks represent the sequence area containing duplicated and rearranged NF-κB and Sp-1 binding sites in the LTR. The dates shown represent the times when PBMC were collected for analysis. PPT, polypurine tract; NRE, negative regulatory element.

FIG. 3.

Convergent evolution of SBBC nef/LTR sequences to a minimal nef/LTR structure. Comparisons of the genomic structures of nef/LTR sequences cloned from the earliest available and most recently obtained PBMC samples are shown. The genomic structures are compared to those of wild-type HIV-1 (NL4-3). Numbers refer to nucleotide positions in NL4-3. Black blocks represent intact sequence, and gaps represent deletions. Gray blocks represent the sequence area containing duplicated and rearranged NF-κB and Sp-1 binding sites in the LTR. The dates shown represent the times when PBMC were collected for analysis. PPT, polypurine tract; NRE, negative regulatory element.

The highly evolved nef/LTR sequences harbored by D36, C54, and C98 are strikingly similar to those that have remained dominant in C49 for at least 10 years (Fig. 2). The presence of Nef antibodies directed against peptides spanning the entire Nef protein (with the exception of the common, 3′-most deletion in the nef/LTR overlap region) in C49 (9) suggests that prior to 1994 a near-complete nef sequence existed and that the bulk of nef/LTR evolution occurred during the first 10 years after infection. C49 has had below-detectable HIV-1 RNA levels since February 1994 (Table 1) and persistent HIV-1 cytotoxic T-cell responses since monitoring of the SBBC began in 1992 (2, 8, 17). Thus, our findings suggest that the highly evolved nef/LTR structure is stable and, in the case of C49, does not increase pathogenicity.

Discussion.

In this study, we demonstrated a progressive loss of nef/LTR sequence in an epidemiologically-linked cohort of long-term survivors who have been infected with attenuated HIV-1 for up to 25 years. The intersubject evolution of nef/LTR sequence loss appears convergent, tending toward retention of only those sequence elements that are absolutely required for HIV-1 replication. The highly evolved, minimal nef/LTR structure appears stable, being present as the dominant strain in C49 for at least 10 years. Its persistence in this subject, who has consistently had a below-detectable viral load in the absence of HAART (Table 1), initially suggested that evolution to the minimal nef/LTR sequence may result in a further-attenuated, even less-pathogenic virus. However, we found no association between evolution toward the minimal nef/LTR sequence and the clinical status of the subject. For example, both progressing (D36, C98) and nonprogressing (C54, C49) subjects harbored highly evolved forms of nef/LTR, and a nonprogressing subject (C64) harbored a significantly less evolved nef/LTR. Thus, other viral and/or host factors are likely to contribute to modulating the in vivo pathogenicity of HIV-1 strains with nef/LTR deleted.

Reversion to pathogenicity by SIV with nef deleted has been associated with restoration of a truncated Nef protein (21), acquisition of further deletions in the nef/LTR overlap region (1), and/or duplications of NF-κB binding sites in the LTR (1). In contrast to the SIV studies, the in vivo evolution of HIV-1 with nef/LTR was unidirectional toward a smaller nef/LTR sequence and the majority of the additional sequence loss was within the nef-alone region. Furthermore, none of the clones were capable of encoding Nef. In addition, the presence of duplicated NF-κB binding sites in the LTR was not associated with the clinical status of the SBBC subjects. Therefore, it is likely that any viral factors that modulate the in vivo pathogenicity of HIV-1 with nef/LTR deleted will be distinct from those in SIV with nef deleted. Interestingly, the unidirectional evolution toward the minimal nef/LTR sequence observed in SBBC members is strikingly similar to the pattern of evolution in a slow progressor infected with a nef/LTR deletion variant of HIV-1 circulating recombinant form 01_AE (15). The convergent pattern of nef/LTR evolution among viruses harbored by SBBC members is therefore unlikely to be due to a unique property of the infecting strain but rather likely reflects an intrinsic instability of HIV-1 with nef/LTR defects that is common across clades.

The molecular mechanisms underlying the increased or acquired pathogenicity of nef/LTR-deleted HIV-1 harbored by slow progressors D36 and C98 remain to be determined, but possibilities include changes in LTR or Env function that may contribute to increasing replicative capacity or cytopathicity, respectively. In support of this hypothesis, enhanced transcriptional activity of LTR clones isolated from CSF of D36 was shown to contribute to high CSF HIV-1 RNA levels and the development of HIV-associated dementia (4). Moreover, enhanced coreceptor usage by HIV-1 isolated from D36 when CD4⁺ cell counts fell below 200/μl contributed to enhanced cell killing in ex vivo human tissue cultures (12). Further studies on the function of sequential LTR and Env clones from SBBC subjects are in progress to elucidate their role in the pathogenesis of nef/LTR-deleted HIV-1 infection.

In conclusion, while our studies affirm that nef is important for HIV-1 pathogenesis, convergent evolution of HIV-1 in vivo toward a minimal and apparently stable nef/LTR structure via extensive loss of nef sequence was not associated with HIV-1 progression. Factors other than nef sequence therefore contribute significantly to HIV-1 progression in the SBBC of long-term survivors.

Nucleotide sequence accession numbers.

The nef/LTR nucleotide sequences reported here have been assigned GenBank accession numbers DQ287272 to DQ287311.