Plasma-derived HIV Nef+ exosomes persist in ACTG384 study participants despite successful virological suppression

Human Immunodeficiency Virus (HIV) accessory protein Negative factor (Nef) is detected in the plasma of HIV+ individuals associated with exosomes. The role of Nef+ exosomes (exNef) in HIV pathogenesis is unknown. We perform a retrospective longitudinal analysis to determine correlative clinical associations of exNef plasma levels in ARV-treated HIV+ patients with or without immune recovery. exNef concentration in a subset of AIDS Clinical Trial Group (ACTG) 384 participants with successful virological suppression and with either high (Δ >100 CD4 cell recovery/High Immunological Responders (High-IR) or low (Δ ≤100 CD4 cell recovery/ Low Immunologic Responders (Low-IR) immunologic recovery was measured and compared for study weeks 48, 96, and 144. CD4 recovery showed a negative correlation with exNef at study week 144 (r = −0.3573, *p=.0366). Plasma exNef concentration in high IRs negatively correlated with naïve CD4 count and recovery (r = −0.3249, *p = 0. 0348 (High-IR); r =0.2981, *p= #0.0513 (Low-IR)). However, recovery of CD4 memory cells positively correlated with exNef (r =.4534, *p=.0358) in Low-IRs but not in High-IRs. Regimen A (Didanosine, Stavudine, Efavirenz) lowered exNef levels in IRs by 2-fold compared to other regimens. Nef+ exosomes persist in ART-treated HIV+ individuals despite undetectable viral loads, negatively correlates with naive and memory CD4 T cell restoration and may be associated with reduced immunological recovery. Taken together, these data suggest that exNef may represent a novel mechanism utilized by HIV to promote immune dysregulation.


41
The prognosis for patients infected with HIV has improved significantly with the advent of 42 combination antiretroviral therapy (cART) which can lead to suppression of plasma viremia 43 usually associated with marked improvements in CD4+ T cell counts [1,2]. However, a subset of 44 patients do not experience a robust immune recovery despite viral suppression. While host 45 factors such as age and baseline CD4 + T-cells with a naïve phenotype have been observed to be 46 negatively associated with the magnitude of CD4 + T-cell count improvement [1,[3][4][5][6], virological 47 factors that contribute to these immunologic outcomes are less well defined. 48 Immune reconstitution can be defined as an increase in the number of peripheral CD4 T-49 cells to greater than 350-500 cells/dL after 4 years of effective cART [3]. Discordant responses 50 characterized by a lack of immune recovery despite viral suppression occur in 7-39% of 51 participants receiving cART [4][5][6][7]. The reasons for this phenomenon are not understood.

52
The HIV-encoded Negative factor (Nef) has been implicated in pathogenesis based on 53 studies examining HIV-infected long term non-progressors (LTNP) and elite controllers (EC). 54 LTNPs remain asymptomatic for years with stable CD4 counts while ECs control plasma HIV 55 RNA in the absence of ART [12]. Although several hypotheses have been presented to account 56 for the protective factors leading to delayed disease progression, replication defective virions/nef-57 defective virions have been associated with these populations in some cases [9][10][11]. Recent 58 reports have shown that the viruses within some LTNPs and ECs are actually replication 59 competent and control of HIV may be due to host and/or viral factors that maintain low levels of 60 viremia [11]. HIV-1 Nef may be integral to the control of viral replication and/or T-cell responses 61 in an HIV-infected individual. Early in vitro studies have shown that soluble Nef is cytotoxic to T-cells and is released from HIV-infected cells in plasma derived microvesicles that can be 63 detected in HIV-infected patients [13][14][15][16][17][18][19][20][21][22][23]. These studies suggest that Nef does not exist as a 64 soluble protein in vivo but instead non-virion associated Nef is found in microvesicles or 65 exosome-like microvesicles (exNef) [7,8]. 66 Several in vitro studies have identified pathogenic activities of soluble Nef including its 67 capacity to reduce surface expression of CD4 and MHC class II, to increase HIV infectivity, to 68 stimulate primary macrophages to release pro-inflammatory cytokines and chemokines, and Murine Leukemia Virus) [28][29][30][31][32]. Given the immunomodulatory functions of exosomes, we 84 sought to explore whether exNef may selectively impair CD4+ T-cells recovery during cART.

85
For this study, Nef concentration was determined in plasma-derived exosomes isolated 86 from a subset of the ACTG-384 cohort with and without immune recovery post-cART. We show 87 that Nef+ exosomes persist and can be detected in study participants with undetectable viral 88 loads even after 144 weeks of therapy. This suggests that exNef production is independent of 89 plasma viral load. Interestingly, we demonstrate that Low-IRs have significantly higher levels of   particulates. Microvesicles were pelleted from pre-cleared plasma via ultra-centrifugation for 1 117 hour at 300,000 x g and then re-suspended in 250 µl of phosphate buffered saline (PBS). Immune 118 complexes within the re-suspended exosomes were removed using acid-dissociation prior to Nef 119 measurement similar to p24 antigen measurements from plasma. Exosomes/high speed pellets 120 (100 µl) were treated with 100 µl of 0.3N hydrochloric (HCl, Sigma) and allowed to incubate for 1 121 hour at 37 C. The acid mixture was neutralized with 100 µl of 0.3N sodium hydroxide (NaCl, 122 Sigma) prior to assaying for Nef.

123
Nef Enzyme-linked Immunosorbent Assay (ELISA). Nef concentration in acid-dissociated 124 microvesicle/exosome preparations was measured using a commercially available anti-Nef 125 sandwich ELISA kit (Immunodiagnostics, Bedford, MA) according to the manufacturer's 126 instructions. Briefly, the neutralized microvesicle preparations were diluted 1:1 with sample 127 diluent (Component C) and added to ELISA plates coated with anti-Nef (Component A).

144
Baseline characteristics of ACTG384 sub-cohort. This sub-cohort consisted primarily of males 145 (88.4%) ( Table 1). Plasma HIV RNA levels at baseline were not significantly different between the three groups (range). Baseline CD4 counts were significantly different between groups (137 147 cells/mm3 in TF, 60 cells/mm3 in High IRs, 38 cells/mm3 in Low IRs) (as previously described 148 that CD4 count at the initiation of therapy may not play a role in immune recovery [3]. 149 Activated CD4 and CD8 T-cells were not significantly different between the High-and Low-IRs  suppression that had CD4 recovery above or below 300 cells/mm 3 at study week 144 (Fig 1A). At 184 study week 144, 40% of the sub-cohort exhibited discordant VL and CD4 cell recovery along with 185 increased exNef level. In the absence of detectable viral replication, exNef could be detected in 186 the plasma and the levels correlated with total CD4 recovery (Fig 1B).  Interestingly, exNef was detected in the plasma of these patients despite suppression of viral 200 replication (Fig 2A). A retrospective longtitudinal analysis shows that at the initiation of therapy,

201
Nef levels were not significantly different between the Low_IRs and High_IRs (Fig 2A, upper   202 panel). However, by study week 144, the Low-IRs had significantly higher levels of plasma exNef 203 than the High IRs (Fig 2A, lower panel). The median baseline plasma Nef level in the sub-cohort  (Fig 2B). This finding suggests that plasma exNef may play a role immunological recovery.  (Fig 3A). However, none of the Low-IRs exhibited increases in 220 CD4 T-cell count close to 350 cells/mm 3 by 144 weeks (Fig 3A). Although naïve T-cells do recover 221 in both High-or Low-IR the IRs have appreciably less naïve CD4 cells than the IRs 96-and 144 222 weeks post treatment initiation (Fig 3C). Notably the changes in both CD4 T-cell count and CD4

223
naïve T-cells negatively correlated with the Nef concentration in plasma-derived microvesicles 224 (Fig 3, B and D) suggesting that in vivo microvesicular Nef may be associated with immune 225 recovery/ CD4 T-cell rebound. appreciably more CD4 memory cells than the Low_IRs (Fig 4, panel A). The degree of CD4 236 increase directly correlated with exNef in the Low_IRs (Fig 4, panel B), suggesting that CD4 237 memory cells could be one of the sources of exNef during anti-viral suppression.  Table 2. Basically, two NRTIs zidovudine and lamuvidine or didanosine and 248 stauvidine followed by either efavirenz or nelfinavir were compared (Fig 5A, upper panel).  (Fig 5, lower panel).
This suggests that drug regimen may also dictate exNef levels and that PI-sparing regimens 259 reduce both viral load and exNef level.  Overall, these data also suggest that increased Nef levels maybe a double-edged sword -in 341 terms of viral suppression-high exNef is associated with decreased viral load but in regard to