| |
The Gut (Microbial translocation) is associated with sustained failure in CD4+ T-cell reconstitution in HIV-infected patients on long-term highly active antiretroviral therapy
[Research Letters]
|
| |
| |
AIDS:Volume 22(15)1 October 2008p 2035-2038
Marchetti, Giuliaa; Bellistri, Giusi Ma; Borghi, Elisab; Tincati, Camillaa; Ferramosca, Stefaniac; La Francesca, Mariab; Morace, Giuliab; Gori, Andread; Monforte, Antonella d'Arminioa aDepartment of Medicine, Surgery and Dentistry, Clinic of Infectious Diseases, 'San Paolo' Hospital, University of Milan, Milan, Italy bDepartment of Public Health, Microbiology, Virology, University of Milan, Milan, Italy cDepartment of Clinical Sciences, Chair of Infectious Diseases and Tropical Medicine, 'Luigi Sacco' Hospital, University of Milan, Milan, Italy dDivision of Infectious Diseases, 'San Gerardo' Hospital, Monza, Italy.
Abstract
Patients with inefficient CD4+ T-cell recovery on virogically suppressive highly active antiretroviral therapy constitute a major clinical hurdle given the threat of HIV/AIDS disease progression. We show heightened circulating lipopolysaccharide associated with plasma enterobacterial DNA and highly activated Ki67+CD4+CD8+ in 24 immunologic-nonresponders (CD4+ T-cell ≦ 200; HIV-RNA ≦ 50) compared with 11 full responders (CD4+ T-cell ≥ 400; HIV-RNA ≦ 50). These data provide novel insight into INRs pathogenesis, since they correlate augmented systemic translocation of microbial bioproducts with T-cell hyperactivation.
Impairment of CD4+ T-cell rescue despite HIV-viremia control is a significant issue in up to 30% of HIV-infected highly active antiretroviral therapy (HAART)-treated patients. These individuals, known as immunologic-nonresponders (INRs), or most appropriately haematologic-nonresponders, have an increased risk of HIV/AIDS progression [1]. Persistent T-cell hyperactivation proved to be a major feature of INRs, ultimately exhausting compensatory homeostatic pathways of T-cell recovery and jeopardizing functional competence and disease outcome [2,3].
In progressed HIV disease, chronic immune hyperactivation plays a crucial role in sustaining CD4+ T-cell depletion [4]. Most recently, a fascinating pathogenetic model of HIV-driven hyperactivation has been proposed that would establish a breakdown of the gastrointestinal mucosal barrier. The consequent increased translocation of luminal microbial bioproducts would exert a sustained trigger to immune activation [5]. Given its role as a marker of microbial translocation [6-8], circulating lipopolysaccharide (LPS) proved to be significantly raised in HIV-infected patients, correlating with immune hyperactivation. Moreover, successful HAART significantly reduced plasma LPS, yet this remained at more elevated levels than in uninfected controls [5].
We hypothesized that perpetuating microbial translocation despite long-term HAART might be relevant in hampering CD4+ T-cell reconstitution in INRs by sustaining T-cell hyperactivation.
We have previously demonstrated that INRs were characterized by increased levels of activated/proliferating CD38+/Ki67+CD4+ and CD8+ in peripheral blood versus full responders (FRs) [2]. In this cohort, we investigated the possible correlation between T-cell activation and microbial translocation in unselected 24 INRs (CD4+ T-cell count ≦200 cells/μl; HIV-RNA ≦50 cp/ml) and 11 FRs (CD4+ T-cell count ≥400 cells/μl; HIV-RNA ≦50 cp/ml in at least three consecutive determinations). As controls, we studied 12 HIV+ advanced naïve patients (CD4+ T cell count ≦250 cells/μl; HIV-RNA ≥5000 cp/ml).
Patients' characteristics at the time of study are presented in Fig. 1(a). No significant differences in demographic and HIV-related parameters were observed among groups. In particular, at the time of analysis, none of the patients included in the study had signs or symptoms of acute infection, gastrointestinal disease, opportunistic infections, or overt signs of bacteremia, cirrhosis, or end-stage liver disease.
Microbial translocation was evaluated in plasma samples stored at -80°C by LPS levels (Limulus Amebocyte Assay, Cambrex, Italy) and PCR detection and identification of the bacterial 16sRNA gene [9].
Both INRs and FRs displayed significantly lower mean LPS compared with advanced naïve patients (INRs: 45 ± 26, FRs: 29 ± 17, naïve: 50 ± 51 pg/ml, P = 0.01 and P = 0.04 for advanced naïve versus FRs and INRs, respectively) (Fig. 1b), supporting the highest degree of microbial translocation in untreated advanced HIV infection. Interestingly, despite a comparable HIV-viremia suppression, INRs presented a more elevated LPS than FRs (P = 0.03) (Fig. 1b). This supports a negative correlation between CD4+ T-cell recovery subsequent to HAART and LPS levels.
HAART-treated individuals as a whole presented a positive correlation between activated/proliferating Ki67+CD4+ and CD8+ and LPS levels (Ki67+CD4+, R = 0.601, P = 0.008; Ki67+CD8+, R = 0.575, P = 0.013) (Fig. 1c, d). Most interestingly, a positive correlation between LPS and activated Ki67+T-cell proportion was maintained only in INRs (Ki67+CD4+, R = 0.644, P = 0.024; Ki67+CD8+, R = 0.591, P = 0.043) (Fig. 1e, f), whereas no correlation was displayed by FRs (Ki67+CD4+, R = 0.904, P = 0.281; Ki67+CD8+, R = 0.892, P = 0.299).
Given the strictest association between LPS and T-cell activation in INRs, we investigated whether this could reflect an increased exposure to bacterial DNA fragments. Plasma samples from advanced naïve patients, and from a subgroup of seven unselected INRs and seven FRs, were examined for the presence and identification of DNA bacterial fragments using a broad-range 16S rRNA PCR amplification followed by HaeIII restriction enzyme analysis [9]. Interestingly, five out of seven (71%) INRs yielded a positive PCR amplification, whereas seven out of seven FRs were consistently negative. A positive PCR amplification was shown in six out of 12 (50%) HIV+ advanced naïve individuals. The restriction endonuclease digestion yielded a similar digestion pattern for all the amplicons, consistent with DNA from Enterobacteriaceae (Fig. 1g).
Subsequent sequencing analysis (ABI PRISM 3130 Genetic Analyzer; Applied Biosystems, Inc., Abilene, Texas, USA) confirmed that bacterial DNA belonged to Enterobacteriaceae in all six out of six advanced naïve patients (Serratia spp., n = 4; Rahnella spp., n = 2) and in five out of seven INRs (Serratia spp.) with an homology of 99-100%.
Although the correlative nature of this study does not establish definite causality, the strictest association between persisting T-cell hyperactivation in INRs and highest circulating LPS suggests an increased translocation of microbial products through the gastrointestinal mucosa. In turn, this finding indicates persistent HIV-driven alterations of the gastrointestinal barrier as a possible mechanism behind the lack of CD4+ T-cell recovery in INRs. However, heightened HIV-driven immune activation might also result in mucosal immune dysfunctions, thus favouring microbial translocation [10]. Similarly, a positive association between persistently elevated LPS and T-cell hyperactivation has been described in untreated HIV infection [11]; consistent with these data, our observations of reduced LPS in FRs might be a consequence of HAART, which is ineffective in INRs. Furthermore, although in untreated viremic individuals, highest LPS was associated with CD8+ T-cell hyperactivation [11], our finding of a positive association in INRs also with elevated CD4+ T-cell turnover is consistent with homeostatic CD4+ T-cell activation secondary to CD4+ T-lymphopenia [11,12], and with the correlation between CD4+ T-cell turnover and immune activation [13,14].
In addition, by showing an association between LPS and bacterial DNA fragments in plasma, our study is the first to directly demonstrate bacterial translocation in both naïve and HAART-treated HIV-infected patients through the direct demonstration of enterobacteria genome sequences.
Several nonmutually exclusive mechanisms might be involved in promoting an increase of bacterial translocation in INRs. The demonstration of enterobacterial DNA fragments in plasma seems to involve a mechanism of relative enterobacteria outgrowth and translocation, which correlates with the recent evidence of HIV-driven subversion of gut microbiota composition [15]. Furthermore, analogous to advanced HIV-infection [5], bacterial translocation might be favoured in INRs by reduced T-cell-mediated competence [2,16] failing to provide full immune control in mucosa and mesenteric lymph nodes, thus permitting peripheral egress and survival of bacteria [17,18].
Given the significant clinical burden of INR individuals, our findings provide relevant insights on elevated microbial translocation via a pathologic gut barrier in INRs, offering an appealing perspective to test novel, more targeted adjunctive treatments such as gut microflora interventions [5,19,20] and/or immune-modulants to interact with microbial-mediated immune activation.
|
|
| |
| |
|
|
|
