Elevation of intact and proteolytic fragments of acute phase proteins constitutes the earliest systemic antiviral response in HIV-1 infection

Abstract

The earliest immune responses activated in acute human immunodeficiency virus type 1 infection (AHI) exert a critical influence on subsequent virus spread or containment. During this time frame, components of the innate immune system such as macrophages and DCs, NK cells, beta-defensins, complement and other anti-microbial factors, which have all been implicated in modulating HIV infection, may play particularly important roles. A proteomics-based screen was performed on a cohort from whom samples were available at time points prior to the earliest positive HIV detection. The ability of selected factors found to be elevated in the plasma during AHI to inhibit HIV-1 replication was analyzed using in vitro PBMC and DC infection models. Analysis of unique plasma donor panels spanning the eclipse and viral expansion phases revealed very early alterations in plasma proteins in AHI. Induction of acute phase protein serum amyloid A (A-SAA) occurred as early as 5-7 days prior to the first detection of plasma viral RNA, considerably prior to any elevation in systemic cytokine levels. Furthermore, a proteolytic fragment of alpha-1-antitrypsin (AAT), termed virus inhibitory peptide (VIRIP), was observed in plasma coincident with viremia. Both A-SAA and VIRIP have anti-viral activity in vitro and quantitation of their plasma levels indicated that circulating concentrations are likely to be within the range of their inhibitory activity. Our results provide evidence for a first wave of host anti-viral defense occurring in the eclipse phase of AHI prior to systemic activation of other immune responses. Insights gained into the mechanism of action of acute-phase reactants and other innate molecules against HIV and how they are induced could be exploited for the future development of more efficient prophylactic vaccine strategies.

Figure 1. Characterization of plasma panels obtained from HIV-1-infected subjects.

(A) Serial plasma HIV-1 RNA titers in the 19 US plasma donors acquiring HIV infection used in this study. The sample time courses are aligned relative to a common time origin, T₀, defined as the time point when the plasma viral load first reached 100 RNA copies ml⁻¹ as described in . (B) Scheme representing the strategy for plasma sample preparation and analysis by mass spectrometry.

Figure 2. Identification of plasma proteins present at elevated levels prior to or during acute HIV-1 viremia by mass spectrometry.

(A) Analysis of serial samples derived from plasma panel 64012 by MALDI-TOF mass spectrometry. Mass profiling reveals peaks (indicated by arrows in the insert) that are observed uniquely in HIV-1 RNA-positive (after T₀, indicated in red, data for time points +11 and +13 days are shown) as compared to HIV-1 RNA-negative plasma samples (before T₀, indicated in blue, data for time points −23 and −18 days are shown). (B) MALDI-TOF LIFT (MS/MS) spectrum of precursor ion mass 2178 Da [M+H]⁺ that was identified as the peptide 86–105 derived from human A-SAA (Swissprot accession nr. P02735). Identified b- and y- fragment ions are indicated. (C) MALDI-TOF-based mass profiling of three plasma panels (64012, 9018 and 9034) revealed an elevation of peaks 2178 Da (peptide 86–105 from A-SAA) and 2213 Da (peptide 960–979 from complement C3, see Fig. S1) before and during viremia. Viral load and T₀ are as described previously . Top panels: viral load; second and third panels: peak intensities of mass peaks 2178 Da and 2213 Da normalized to the corresponding peak intensities observed for the first time point of 64012; bottom panels: relative peak intensities of mass peaks 1555 and 2552 that were used as standards, illustrating use of comparable MS acquisition conditions for all time points. The arrows in the middle and bottom panels indicate viremic time points.

Figure 3. Kinetic analysis of A-SAA levels prior to and during HIV-1 viremia.

(A) A-SAA levels as measured by ELISA in plasma panels from 19 donors acquiring HIV-1 infection. In each graph, time is plotted relative to T₀, plasma HIV RNA levels (log₁₀ copies ml⁻¹) are indicated by the open circles and are as reported previously . The solid circles show A-SAA protein levels. Subject-specific background levels of A-SAA (dotted grey lines) and the 90% prediction interval (threshold for defining a significant elevation, black lines) were calculated as described in the methods section. 15 out of 19 subjects show a significantly elevated A-SAA level before the first detectable viral load. (B). A-SAA and C-SAA protein levels in sequential plasma samples from 10 donors acquiring HIV-1 infection as assessed using anti-A-SAA and anti-C-SAA immunoblotting. The two sets of blots are positioned so that the time point when plasma viral RNA was first detected is aligned (vertical line). Post-viremic time points are indicated by the horizontal arrow. (C) A-SAA and C-SAA protein levels in sequential plasma samples from 3 donors acquiring HCV infection (top panel) and 3 donors acquiring HBV infection (bottom panel). As above, the two sets of blots are positioned so that the time point when plasma viral nucleic acids were first detected is aligned (vertical line); and post-viremic time points are indicated by the horizontal arrow.

Figure 4. Tandem mass spectrometry-based detection of the antiviral peptide VIRIP, derived from the C-terminus of AAT, in plasma during AHI.

(A) MS/MS spectrum of precursor ion peptide 773.75 Da [M(ox)+3H]³⁺ identified as VIRIP 377–396 derived from AAT (Swissprot accession nr P01009) as detected in plasma panel 9012 (time point 5 d post-T₀). Identified b- and y- fragment ions are shown. (B) LC-MS/MS based detection of the VIRIP precursor ion 773.75 Da [M(ox)+3H]³⁺ in plasma panels 63229 and 9012. Mass peak ion counts are shown in black, and viremic time points are indicated by the grey arrow. (C) Quantitation of synthetic VIRIP peptide by LC-MS/MS indicates that the amount of VIRIP detected in plasma is in the low micromolar range.

Figure 5. Proteolytic processing of AAT by MMP-7 generates VIRIP in vitro.

(A) Potential MMP cleavage sites located in the C-terminal part of AAT (370–418) as predicted by the MEROPS database (http://merops.sanger.ac.uk/). (B) In vitro degradation of AAT by MMP-7 reveals distinct degradation intermediate fragments. Purified AAT was incubated with recombinant MMP-7 at the indicated concentrations and times, followed by SDS-PAGE separation on a 4–12% Bis-Tris gel and immunoblotting using an anti-VIRIP antibody. (C) Prolonged incubation of AAT with MMP-7 resulted in the generation of the VIRIP peptide as well as AAT C-terminal fragments; Upper panel: 16% Tris-Tricine SDS-PAGE separation followed by immunoblotting using an anti-VIRIP antibody; lower panel: LC-MS/MS analysis after 2 h digestion followed by database searching using the MASCOT algorithm identifies C-terminal AAT peptides including VIRIP (AAT 377–396) as proteolysis products.

Figure 6. Components of the acute phase response interfere with HIV-1 infection in vitro.

(A) Effect of AAT or its C-terminal proteolytic fragments 397–418 or 377–396 (VIRIP) on infection of PBMCs with R5- and X4-tropic viruses. HIV-1 replication was assessed by analysis of supernatant p24 levels 7 d post-infection, and results are expressed relative to p24 levels in control infected cells treated with vehicle only. R5 and X4 virus replication in PBMCs was significantly inhibited by VIRIP (p<0.0001, as determined by one-way ANOVA). (B) Effect of A-SAA on infection of PBMCs with R5 or X4 viruses, or infection of MDDCs with an R5 virus, assessed as in (A), although p24 levels in MDDC supernatants were evaluated after a 4 d rather than a 7 d infection period. Data from two MDDC infection experiments is shown, one using an A-SAA concentration range from 12.5µg–100µg/ml, and the other from 0.5µg–8µg/ml. R5 virus replication in MDDCs was significantly inhibited by A-SAA in both experiments (p<0.0001, as determined by one-way ANOVA). (C) Effect of CRP on infection of PBMCs with R5- or X4-tropic viruses, assessed as in (A). For all results, mean and STDEV values are shown for triplicates per experiment.