Cross-omics mapping of the human HIV restrictome across clinical states, immune-cell compartments, and protein responses

Reuben S Maghembe

doi:10.1371/journal.pone.0356591

. 2026 Oct 13;21(10):e0356591. doi: 10.1371/journal.pone.0356591

Cross-omics mapping of the human HIV restrictome across clinical states, immune-cell compartments, and protein responses

Reuben S Maghembe ^1,^2,^*

Editor: Carlos Manuel de Noronha³

PMCID: PMC13660893 PMID: 42868142

Abstract

Host restriction factors can block HIV at multiple stages, but they are often studied as individual genes or within a single molecular layer. I performed an evidence-structured computational reanalysis of three independent public datasets to determine how the human HIV restrictome varies across clinical states, immune-cell compartments, and protein responses. The analysis used bulk monocyte RNA sequencing (GSE157198), PBMC single-cell/CITE-seq data (GSE256089), and primary CD4 T-cell proteomics (PXD012263) within a prespecified restrictome-immunometabolism framework. In GSE157198, ART-naive monocytes had higher core-restrictome activity than healthy controls (difference = 1.154, p = 0.000746), and core scores were strongly correlated with interferon/innate-sensing scores (Spearman r = 0.800, p = 7.47 × 10^-11). Core-lipid/cholesterol correlation was weaker and was not independent of interferon after conditional analysis. In GSE256089, participant-paired core-restrictome scores were enriched in NK cells relative to B cells, monocytes, and T cells, and matched random-gene-set analyses supported enrichment beyond background expression structure. FCGR3A monocytes also had higher participant-paired core scores than CD14 monocytes. Exploratory sensing/death-gene analysis identified cell-type-specific expression patterns but did not establish pathway activation, pyroptosis, or apoptosis. In PXD012263, the historical q < 0.05 proteomic selection contained 19 curated genes and retained mixed response polarity; sensitivity thresholds q < 0.01 and q < 0.10 selected 10 and 25 genes, respectively. Together, the results show that HIV restrictome activity varies across clinical and immune-cell contexts and is embedded within broader interferon-linked and immunometabolic states. Because the datasets are independent and not subject matched, the findings are observational and hypothesis-generating rather than causal.

Introduction

Antiretroviral therapy has transformed HIV infection into a manageable chronic condition, but it does not eliminate persistent viral reservoirs or the inflammatory and immune disturbances that accompany infection [1]. HIV also evolves rapidly through error-prone reverse transcription, recombination, and immune selection, generating extensive diversity within and among infected people [2–7]. These features make HIV persistence a problem of both viral adaptation and host-cell state.

Intrinsic restriction factors are host proteins that block HIV at entry, reverse transcription, nuclear import, integration, transcription, assembly, release, or cell-to-cell spread [8–11]. Established systems include APOBEC3 enzymes, SAMHD1, BST2/tetherin, TRIM proteins, MX2, IFITM proteins, SERINC3/SERINC5, MOV10, ZC3HAV1, and wider interferon-stimulated effector programs [9,10,12–18]. Many were defined through viral countermeasures: Vif antagonizes APOBEC3 proteins, Vpu targets BST2, Vpx counteracts SAMHD1 in HIV-2 and related lentiviruses, and Nef excludes SERINC3/SERINC5 from virions [14,15,17–20]. The restrictome therefore reflects an evolutionary contest between host antiviral pressure and viral counter-restriction.

Restriction factors do not operate as an interchangeable antiviral list. Their abundance and activity vary with leukocyte lineage, interferon exposure, activation, and metabolism [16,21–26]. APOBEC3 expression differs among tissues and hematopoietic subsets [27], TRIM5alpha illustrates lineage-specific restriction pressure [28], IFITM proteins exert context-dependent effects at cellular membranes [12], and interferon-stimulated gene screens distribute antiviral activity across many effectors [16]. Chronic HIV infection is also accompanied by systemic inflammation [29] and major immunometabolic remodeling [25]. Glucose and glutamine metabolism can support early infection in CD4 T cells [21], and highly metabolic CD4 T cells are preferentially susceptible to reservoir seeding [26]. Comorbidities such as diabetes and tuberculosis further alter inflammatory and metabolic environments relevant to HIV host responses [20,30–32].

Public HIV omics datasets now permit this biology to be examined across molecular and cellular scales, although the most informative datasets were generated for different primary questions [33–35]. GSE157198 investigated CREM/ICER isoform expression in spontaneous HIV control [35], GSE256089 identified monocyte programs associated with HIV-1 reservoir size after acute treatment [33], and PXD012263 mapped proteome dynamics during HIV infection of primary CD4 T cells [34]. Reanalyzing these resources through a common biological framework can therefore test whether signals separated by study design converge on a shared organization.

Here I asked whether the human HIV restrictome behaves as a uniform interferon-induced program or instead varies with clinical state, immune-cell identity, and molecular context. I analyzed bulk monocyte RNA sequencing, PBMC single-cell/CITE-seq, and primary CD4 T-cell proteomics as three complementary public evidence layers rather than as a subject-matched multi-omics cohort. This evidence-structured computational reanalysis tests three related questions: whether clinical state is associated with coordinated restriction activity, which immune-cell compartments carry these programs, and how selected restriction-related proteins behave during HIV exposure. The data suggest that immune cells deploy context-specific combinations of antiviral defenses rather than a single uniform intrinsic antiviral program.

Results

A common framework separates clinical state, cell compartment, and protein response

The analysis integrated three independent public omics datasets within one prespecified biological framework (Fig 1). GSE157198 contributed clinical-state-resolved monocyte transcriptomes, GSE256089 contributed participant-level immune-cell localization, and PXD012263 contributed protein responses in HIV-exposed primary CD4 T cells [33–35]. Because these datasets differ in cohort, cell type, and experimental design, they were treated as complementary evidence layers rather than as subject-matched multi-omics measurements. The framework retained the eight prespecified modules spanning core HIV restriction, interferon/innate sensing, nucleotide metabolism, lipid/cholesterol metabolism, mitochondrial/redox biology, glycolysis/inflammatory metabolism, autophagy/proteostasis, and immune-cell state. Reviewer-requested analyses were used as explicit robustness or exploratory extensions and did not retrospectively redefine the prespecified modules.

Table 1 summarizes the role of each dataset, and Table 2 defines the eight biological modules. The design separated three questions that are often conflated: whether clinical state changes coordinated restriction programs, which immune-cell compartments carry those programs, and whether transcript-level organization is preserved or remodeled at the protein level.

Table 1. Public datasets used in the study.

Dataset	Omics layer	Biological context	Role in this study	Main result
GSE157198	Bulk RNA sequencing	Enriched monocytes from healthy controls, ART-treated participants, untreated participants, and elite controllers	Disease-state transcriptomic layer	Untreated monocytes showed the strongest IFN-coupled core restrictome state; elite-controller monocytes were not uniformly elevated.
GSE256089	Single-cell/CITE-seq	PBMCs from 38 HTO-demultiplexed A5354 participants during ART	Cell-type localization and participant heterogeneity	Participant-level analysis showed specific NK core-restrictome enrichment beyond matched random gene sets and higher core-restrictome scores in FCGR3A than CD14 monocytes; broad reservoir-category module differences remained non-significant after correction.
PXD012263	Proteomics	Primary human CD4 T cells across an HIV infection time course	Protein-level remodeling layer	Selected restriction factors were depleted in HIV-positive cells while other antiviral and IFN-linked proteins were induced.

Open in a new tab

Table 2. Prespecified restrictome-immunometabolism module framework.

Module	Number of curated genes	Representative genes
Core HIV restriction factors	37	APOBEC3A; APOBEC3B; APOBEC3C; APOBEC3D; APOBEC3F; APOBEC3G; APOBEC3H; BST2; CH25H; DDX58; DDX60; HERC5
Interferon/innate sensing	25	CGAS; IFNAR1; IFNAR2; IFNB1; IFNGR1; IFNGR2; IKBKE; IRF1; IRF3; IRF7; IRF9; JAK1
Nucleotide metabolism	14	CAD; DCK; DHFR; GMPS; IMPDH1; IMPDH2; MTHFD1; MTHFD2; RRM1; RRM2; RRM2B; SAMHD1
Lipid/cholesterol membrane metabolism	17	ABCA1; ABCG1; ACACA; CH25H; CPT1A; CPT2; FASN; HMGCR; HMGCS1; LDLR; NPC1; NPC2
Mitochondrial/redox biology	23	ATP5F1A; CAT; COX4I1; GCLC; GCLM; GPX1; GPX4; GSR; HMOX1; NDUFA1; NDUFB8; NFE2L2
Glycolysis/PPP/inflammatory metabolism	26	ACOD1; ALDOA; ARG1; ARG2; ENO1; G6PD; GAPDH; GPI; HIF1A; HK1; HK2; IDO1
Autophagy/proteostasis/ER stress	22	ATF4; ATF6; ATG12; ATG16L1; ATG3; ATG5; ATG7; BECN1; DDIT3; EIF2AK3; ERN1; HSP90AA1
Immune-cell state/activation/exhaustion	36	CCL2; CCL3; CCL4; CCR7; CD3D; CD3E; CD4; CD79A; CD8A; CD8B; CLEC9A; CTLA4

Open in a new tab

ART-naive monocytes show a robust interferon-linked restrictome elevation

GSE157198 contained enriched monocyte transcriptomes from healthy controls, ART-treated individuals, untreated individuals, and elite controllers [35]. All 37 genes in the core HIV restriction module were detected, and most immunometabolic modules had high coverage, supporting module-level comparison across clinical groups.

The ART-naive group showed the strongest elevation in the core-restrictome score relative to healthy controls (difference = 1.153725, p = 0.000745751; Fig 2D). Across all four groups, the core-restrictome score showed a substantial group effect (eta2 = 0.420447; Kruskal-Wallis p = 0.00279821). The core-restrictome score was strongly correlated with the interferon/innate-sensing score (Spearman r = 0.799718, p = 7.46724 × 10^-11; Fig 2E). The core score also correlated with the lipid/cholesterol score in the unadjusted analysis (Spearman r = 0.451163, p = 0.00211408; Fig 2F). This association remained detectable after removing CH25H, which is shared by the core and lipid gene sets (r = 0.424242, p = 0.00410444), showing that simple gene-set overlap was not sufficient to explain the unadjusted correlation. However, the lipid association was not independent of interferon. After conditioning on the interferon/innate-sensing score, the partial correlation fell to r = 0.059104 (p = 0.703132); after conditioning on both interferon and clinical group, it was r = −0.017597 (p = 0.909734). The lipid/cholesterol signal is therefore interpreted as part of a shared IFN-associated immunometabolic state rather than as an independent core-lipid coupling axis. Across three prespecified null schemes containing 1,000 random 37-gene sets each, the ART-naive-versus-healthy effect had empirical p = 0.000999001 under every scheme (BH-adjusted empirical p = 0.002997). The core-IFN correlation was also robust, with empirical p values ranging from 0.000999001 to 0.001998002 (BH-adjusted values 0.002997–0.004995). By contrast, empirical p values for the unadjusted core-lipid relationship ranged from 0.0239760 to 0.0989011, and the IFN-adjusted core-lipid association remained null (empirical p = 0.700300–0.715285). Thus, the disease-state and core-IFN signals were robust to random-set testing, whereas lipid specificity was weaker. In addition, the ART-naive-versus-healthy disease-effect direction remained positive in all 37 leave-one-core-gene-out analyses, indicating that the result was not dependent on a single dominant core-restrictome gene.

Fig 2 — (A) Module effects across clinical contrasts. (B) Strongest module shifts. (C) Concordance of module effects. (D) Core-restrictome scores across clinical groups. The ART-naive-versus-healthy difference was 1.154 (p = 0.000746), and the four-group comparison gave eta2 = 0.420 and Kruskal-Wallis p = 0.00280. Across three prespecified 1,000-set random-gene null schemes, the ART-naive-versus-healthy effect had empirical p = 0.000999 in each scheme. (E) Core-restrictome and interferon/innate-sensing scores were strongly associated (Spearman r = 0.800, p = 7.47 × 10^-11). (F) Core-restrictome and lipid/cholesterol scores were associated in the unadjusted analysis (Spearman r = 0.451, p = 0.00211) and after removing the overlapping CH25H gene from the lipid score (r = 0.424, p = 0.00410), but the association was not independent of interferon: partial Spearman r = 0.059, p = 0.703 after conditioning on IFN and r = −0.018, p = 0.910 after conditioning on IFN plus clinical group. Lipid/cholesterol activity is therefore interpreted as part of a shared IFN-associated architecture rather than an independent core-lipid relationship. The ART-naive disease-effect direction remained positive in all 37 leave-one-core-gene-out analyses.

Multiple antiviral families drive the monocyte state

Gene-level decomposition showed that the untreated-versus-healthy contrast was distributed across multiple antiviral families (Fig 3). The largest contributors included IFITM1, ISG15, RSAD2, MX1, IFITM3, HERC5, OAS3, LY6E, APOBEC3A, OAS2, MX2, DDX60, IFIH1, OAS1, IFITM2, TRIM22, PML, MOV10, and BST2.

No single factor dominated the monocyte state. Instead, IFITM, OAS, MX, APOBEC3, ubiquitin-like modification, viral sensing, and virion-release pathways contributed in parallel. The elite-controller contrast remained weaker along the dominant interferon-restrictome axis, reinforcing the distinction between coordinated antiviral activation and durable control.

The leave-one-gene-out analysis supports this distributed interpretation: removing each core-restrictome gene in turn preserved the direction of the ART-naive-versus-healthy effect in 37 of 37 analyses. Thus, the clinical-state signal is not attributable to one dominant gene such as a single highly responsive interferon-stimulated factor.

Participant-level single-cell analysis identifies specific NK enrichment and monocyte-subtype structure

GSE256089 provided HTO-demultiplexed PBMC profiles from 38 participants receiving ART in the A5354 cohort [33]. I used these data to localize the prespecified modules across broad immune-cell compartments and then linked participant-level scores to published reservoir categories.

Broad cell-type localization remained descriptive, but all inferential comparisons were performed using participants rather than individual cells as the independent units (Fig 4). Across participants, the mean core-restrictome score was 0.395507 in NK cells, 0.323249 in monocytes, 0.309693 in T cells, and 0.238208 in B cells. The participant-paired difference between NK cells and each participant’s mean across B cells, monocytes, and T cells was 0.105124 (dz = 3.8490; paired p = 5.551754 × 10^-24), and NK cells had the highest core score in 35 of 38 participants. Because lineage-specific baseline expression could generate an apparent NK-high pattern, I compared the observed enrichment with 1,000 random gene sets per prespecified matching scheme. These sets matched the core module for gene-set size and expression/detection structure. The observed NK-versus-reference enrichment remained unusual under the matched null distributions, with frozen empirical upper-tail p values of 0.001998, 0.007992, and 0.000999 across the relevant schemes. In contrast, simple counting of participants in which NK cells ranked highest was less specific under matched nulls. The result is therefore interpreted as quantitative enrichment of the curated restrictome beyond lineage background, not merely as an NK-high ranking. Higher-resolution monocyte analysis identified additional structure. Across 38 participants, 37,081 CD14 cells and 7,581 FCGR3A cells met the analysis criteria. Participant-paired core-restrictome scores were higher in FCGR3A than CD14 monocytes by 0.157650 on average (dz = 2.95165; paired p = 4.773127 × 10^-20), with FCGR3A exceeding CD14 in 37 of 38 participants. The participant-level relationship between monocyte composition and core-restrictome activity was also positive (Pearson r = 0.634, p = 1.92 × 10^-5; Spearman r = 0.614, p = 4.17 × 10^-5). The available single-cell annotations did not provide validated participant-linked CD56bright and CD56dim NK-cell labels. I therefore did not infer NK subtypes from unvalidated marker combinations.

Fig 4 — (A) Mean scores for the eight prespecified modules across broad PBMC cell-type labels provide a descriptive localization overview. (B) Participant-paired comparison of the core-restrictome score in NK cells with each participant’s mean score across B cells, monocytes, and T cells. NK enrichment was observed across 38 participants (mean paired difference 0.105); matched random-gene-set analyses supported specificity beyond gene-set size and expression/detection structure. (C) Ranked participant-level transcriptional heterogeneity, showing the eight lowest and eight highest values of the prespecified core-restrictome + IFN/innate + lipid/cholesterol composite z-score. Color encodes signed direction and magnitude around zero and does not represent an additional statistical test. (D) Participant-paired comparison of core-restrictome scores between CD14 and FCGR3A monocyte subtypes. FCGR3A scores exceeded CD14 scores in 37 of 38 participants, with a mean paired difference of 0.158. All inferential comparisons use participants, not cells, as the independent unit.

Monocyte composition does not explain the independent bulk clinical-state effect

Because FCGR3A and CD14 monocytes differed in restrictome activity, I tested whether variation in monocyte subtype balance could explain the GSE157198 clinical-state result. Three transcriptional proxies of CD14-FCGR3A balance were evaluated in the independent bulk-monocyte cohort. After adjustment, the ART-naive-versus-healthy core-restrictome effect remained approximately 1.14–1.17, with p values approximately 0.0006–0.0011. Monocyte subtype composition is therefore biologically associated with restrictome activity, but transcriptional proxies of CD14-FCGR3A balance did not account for the clinical-state-associated restrictome elevation observed in the independent GSE157198 bulk-monocyte cohort.

Exploratory sensing and cell-death genes show context-specific expression without establishing pathway activation

cGAS and STING were already represented in the prespecified interferon/innate-sensing framework through MB21D1/CGAS and TMEM173. In response to the reviewer, I additionally examined IFI16, TREX1, CASP1, and CASP3 as exploratory targets. These additions were analyzed separately from the prespecified module tests. In the GSE157198 bulk-monocyte analysis, TREX1 showed clear FDR-supported variation, whereas the bulk cGAS/MB21D1 and STING/TMEM173 contrasts were not FDR-supported. In the participant-level GSE256089 analysis, IFI16, STING/TMEM173, and CASP3 showed higher expression in NK cells, whereas TREX1 and CASP1 were higher in monocytes. cGAS/CGAS did not differ materially between NK cells and monocytes. These expression patterns identify cell-context differences but do not establish cGAS-STING pathway activation, inflammasome activation, pyroptosis, executioner-caspase activity, or apoptosis. Functional activation of these pathways remains a mechanistic hypothesis requiring direct perturbational or biochemical testing.

Primary CD4 T-cell proteomics shows mixed polarity that is stable to FDR-threshold sensitivity

PXD012263 provided a time-resolved proteomic map of HIV infection in primary human CD4 T cells [34]. I used the published source tables to ask whether proteins represented in the curated framework showed uniform induction or selective, bidirectional remodeling in HIV-positive cells.

The historical q < 0.05 WT-versus-mock selection remains the primary differential proteomic boundary. Among 137 curated genes with evaluable published q values, 19 met q < 0.05, reproducing the historical curated membership exactly (Fig 5C). The selected proteins showed mixed rather than uniformly induced polarity. This supports selective remodeling of restriction-related and interferon-linked proteins rather than a single-direction protein response. Sensitivity analysis using the same published q values selected 10 curated genes at q < 0.01, 19 at q < 0.05, and 25 at q < 0.10 (Fig 5D). Mixed polarity persisted, although the balance of increased, depleted, and stable/subtle proteins depended on threshold. The source-table time-course abundance summaries in Fig 5B and Fig 5E are descriptive overlays. They are not interpreted as independent time-course differential-significance tests. No quantitative protein-abundance imputation was identified in this reanalysis. Missing source-table abundance values were retained as missing. Replicate summaries used available nonmissing values, and a pseudocount of 1 was included in derived log2-ratio calculations. The pseudocount did not replace missing protein-abundance measurements and is distinct from missing-value imputation.

Fig 5 — (A) Curated protein detection and the subset represented in the published WT-versus-mock q < 0.05 table. (B) Median source-table abundance dynamics across module-level comparisons; these time-course summaries are descriptive and are not interpreted as time-course differential-significance tests. (C) Abundance polarity of the 19 curated proteins in the historical published q < 0.05 WT-versus-mock selection, which remains the primary proteomic selection. (D) Sensitivity across published q < 0.01, q < 0.05, and q < 0.10 thresholds. Among 137 q-evaluable curated genes, 10, 19, and 25 were selected, respectively. The q < 0.05 set reproduced the historical 19-gene membership exactly. Mixed polarity persisted, although the balance of increased, depleted, and stable/subtle proteins was threshold-dependent. (E) Descriptive 48-h HIV-positive-versus-mock abundance direction for the same q < 0.05-selected proteins. The q threshold derives from the published single-timepoint source table; 48-h abundance is shown descriptively rather than as an independent significance test.

Across the three layers, the evidence converged on a model in which the HIV restrictome is organized by clinical state, immune-cell compartment, and selective protein remodeling (Fig 6; Tables 3 and 4). The integrated architecture distinguishes transcriptional activation from cell-type localization and from the fate of individual proteins in infected cells.

Fig 6 — (A) Observed evidence from independent public datasets is summarized across three complementary layers: clinical-state-associated monocyte transcription in GSE157198, participant-level immune-cell localization in GSE256089, and protein responses in primary CD4 T cells in PXD012263. The datasets are not subject matched, and cross-layer convergence does not represent a within-person trajectory. (B) The observed patterns motivate noncausal, testable hypotheses in which interferon state, immune-cell lineage, monocyte subtype composition, lipid/cholesterol state, nucleotide availability, proteostasis/autophagy, and redox context may modulate restrictome expression or function. Hypothesis arrows represent proposed relationships for experimental testing rather than established causal paths. Candidate perturbational tests include manipulating cholesterol synthesis or trafficking, nucleotide availability, redox state, autophagy, or proteostasis while measuring restriction-factor abundance and antiviral function in defined immune-cell contexts. The framework therefore separates observed evidence, mechanistic hypothesis, and perturbational test and suggests that host-directed strategies should be evaluated in the cellular contexts in which the relevant restriction programs operate.

Table 3. Cross-layer evidence supporting the integrated model.

Evidence layer	Dataset	Main figure	Key support	Interpretation scope
Bulk monocyte RNA sequencing	GSE157198	Figs 2 and 3	Module- and gene-level contrasts identify an untreated high-IFN/high-restrictome monocyte state.	Associative monocyte host-state signal; not a direct infected-cell measurement.
Single-cell PBMC profiling	GSE256089	Fig 4	Restriction and immunometabolic modules localize to different PBMC cell types.	Reservoir-category differences were not robust after correction.
Primary CD4 T-cell proteomics	PXD012263	Fig 5	Curated proteins show both depletion and induction in HIV-positive cells.	Published source-table overlay; not raw mass-spectrometry reprocessing.
Integrated synthesis	All layers	Fig 6	Disease state, cell compartment, and protein polarity converge on one organizational model.	Cross-study convergence; not a causal or subject-matched trajectory.

Open in a new tab

Table 4. Genes and proteins supporting the integrated model.

Category	Representative genes/proteins	Primary source	Main figure/table
Bulk monocyte induced core/ISG drivers	IFITM1, ISG15, RSAD2, MX1, IFITM3, HERC5, OAS3, LY6E, APOBEC3A, OAS2, MX2, DDX60, IFIH1, OAS1	GSE157198	Fig 3; Table S4
Proteomic q < 0.05 HIV-positive-cell depleted restriction factors	APOBEC3C, APOBEC3G, APOBEC3F, APOBEC3D, BST2, SERINC5	PXD012263	Fig 5; Table S8
Proteomic q < 0.05 HIV-positive-cell induced proteins	TRIM22, OAS1, STAT2	PXD012263	Fig 5; Table S8
Proteomic lipid/cholesterol support	SQLE, SCD, HMGCS1	PXD012263	Fig 5; Table S8
High-dynamic non-q < 0.05 proteomic ISG/restrictome proteins	RSAD2, MX2, DDX60, ISG15, MX1, HERC5, OAS3, SP100, DDX58, IFIH1	PXD012263	Fig 5; Table S8

Open in a new tab

Discussion

This evidence-structured computational reanalysis shows that the human HIV restrictome varies across clinical states, immune-cell compartments, and protein-response contexts. The strongest bulk-monocyte signal occurred in ART-naive HIV and was tightly associated with interferon/innate-sensing activity. At single-cell resolution, the curated core restrictome was quantitatively enriched in NK cells beyond matched background gene sets, while monocyte subtype was also associated with restrictome activity. In primary CD4 T-cell proteomics, selected restriction-related and interferon-linked proteins showed mixed response polarity rather than uniform induction. These observations support a context-dependent view of intrinsic antiviral defense but do not establish a causal or subject-matched molecular trajectory across the three independent datasets.

The monocyte findings refine the interpretation of spontaneous HIV control. Luo and colleagues identified CREM/ICER isoform differences in the same cohort [35], whereas the present framework resolved a broader interferon-coupled restrictome state. That state was strongest in untreated infection, not in elite control. Durable control therefore appears distinct from maximal peripheral monocyte interferon activation, consistent with evidence that HIV controllers rely on coordinated immune mechanisms rather than generalized activation [28,36,37] and with the contribution of chronic inflammation to HIV pathogenesis [29].

The gene-level pattern also explains why a module perspective is more informative than a single-factor ranking. IFITM, OAS, MX, APOBEC3A, RSAD2, ISG15, HERC5, DDX60, IFIH1, TRIM22, PML, MOV10, and BST2 jointly shaped the untreated monocyte state, consistent with broad host-factor and interferon-effector screens [16,19,33–35,38–40]. These factors act at different stages of the viral life cycle, including entry and fusion, nucleic-acid sensing, reverse transcription, post-entry trafficking, and virion release [13–18,22], so their coordinated elevation represents a composite antiviral state rather than one mechanism.

The single-cell layer separated the location of antiviral programs from their association with reservoir size. The source study identified monocyte expression profiles associated with HIV-1 reservoirs [33], whereas the broad modules used here localized core restriction activity most strongly to NK cells and interferon-lipid activity to monocytes and DCs. This division of labor offers a biological explanation for disagreement among bulk studies: the same pathway score can reflect altered cell composition, cell-intrinsic activation, or shared cytokine exposure.

The NK-cell result also requires a distinction between localization and induction. NK cells are developmentally equipped for rapid antiviral responses, so a high broad antiviral score could arise from constitutive lineage biology rather than HIV-specific activation. The matched-null analysis addresses part of this concern: the curated core-restrictome enrichment exceeded that expected from gene sets matched for size and expression/detection structure. I therefore interpret the signal as constitutive/developmental antiviral readiness together with specific restrictome enrichment, not as evidence that HIV uniformly induced the program in NK cells. The monocyte subtype analysis reveals a different source of cellular structure. FCGR3A monocytes had higher participant-paired core-restrictome scores than CD14 monocytes, and subtype composition correlated with participant-level restrictome activity. However, adjustment for transcriptional proxies of CD14-FCGR3A balance did not remove the independent ART-naive clinical-state effect in GSE157198. Cell composition can therefore modulate a restrictome measurement without being sufficient to explain the bulk disease-associated elevation. The myeloid localization findings are consistent with the broader biological importance of dendritic-cell and monocyte antiviral programs [41]. The lack of corrected reservoir-category differences is important because it defines the boundary of the module framework. It does not contradict the source study’s reservoir-associated signatures [33]; rather, it shows that broad, prespecified biological modules are better suited to mapping cellular architecture than to reproducing a dataset-specific biomarker model. This distinction prevents cell-type localization from being misinterpreted as reservoir prediction.

The exploratory sensing/death analysis adds another boundary to the interpretation. cGAS and STING were already represented in the prespecified innate module, while IFI16, TREX1, CASP1, and CASP3 were examined as reviewer-requested exploratory targets. Their expression differed across clinical or cell-type contexts, with TREX1 showing the clearest bulk statistical support and several targets showing NK- or monocyte-enriched single-cell expression. These observations identify candidate cellular contexts for follow-up but cannot establish DNA-sensing pathway activation or a specific mode of CD4 T-cell death. Direct assays of second-messenger production, inflammasome assembly, caspase cleavage, membrane integrity, or related functional endpoints would be required to distinguish pathway activation, pyroptosis, apoptosis, or compensatory expression [42,43].

The proteomic results expose a further asymmetry between antiviral signaling and restriction-factor abundance. The source atlas mapped extensive host remodeling during HIV infection [34]. Within that response, interferon-linked proteins such as TRIM22, OAS1, and STAT2 increased while APOBEC3C/G/F/D, BST2, SERINC5, and CD4 decreased. Productively infected cells can therefore retain a strong antiviral-stress signature while losing selected proteins that directly constrain infection. This pattern is consistent with selective counter-restriction by lentiviral accessory proteins rather than global suppression of host defense [14,15,17–20].

TRIM22 provides one example of a restriction factor whose antiviral role includes transcriptional restriction [44], while broader ISG programs contain effectors with diverse mechanisms and potencies [45]. The immunometabolic modules connect this architecture to mechanism. SAMHD1 links restriction to intracellular deoxynucleotide availability and reverse transcription [14,46,47]. BST2 and SERINC5 act at membrane-dependent stages of release and virion infectivity [15,18–20], making lipid and cholesterol state biologically relevant rather than merely correlated. Autophagy, proteostasis, ER stress, and redox programs can further influence antiviral-protein turnover, antigen processing, inflammatory signaling, and infected-cell survival. These connections define experimentally testable edges between metabolism and intrinsic immunity.

CH25H nevertheless provides a biologically relevant example of how lipid metabolism can directly intersect antiviral restriction [48].

Previous genome-wide, proteomic, and single-cell studies established that HIV engages a broad host network [33–35,38–40] and that interferon effectors differ markedly in mechanism and antiviral potency [16]. The contribution here is to organize those observations across disease state, cell compartment, and protein polarity. This cross-layer view explains why RNA abundance alone cannot determine antiviral capacity and why apparently discordant datasets may be measuring different parts of the same system.

The same logic may extend to common comorbid states. Tuberculosis and diabetes alter myeloid activation, substrate use, redox balance, and inflammatory tone [30–32], each of which could shift the restrictome without changing the identity of its canonical factors. Testing these contexts would establish whether the architecture observed here is stable or remodeled across clinically important immune-metabolic environments.

This study has several limitations. The three datasets were generated in different cohorts, cell types, and experimental systems and were not subject matched. The cross-layer synthesis therefore cannot establish within-person molecular trajectories. Module scores depend on prespecified gene-set composition and dataset-specific feature coverage. Although participant-level analyses reduce pseudoreplication in the single-cell layer, the available annotations did not support a validated CD56bright/CD56dim NK-cell transcriptional comparison. The monocyte-subtype analysis demonstrates biologically meaningful composition effects but cannot reconstruct the exact cellular composition of the independent GSE157198 bulk samples. The sensing/death analyses measure expression rather than functional pathway activation. The proteomic analysis uses published source tables rather than raw-spectrum reprocessing, retains source missingness, and treats time-course abundance changes descriptively. Finally, all analyses remain observational; the mechanistic edges proposed in Fig 6 require experimental perturbation.

Fig 6 separates what is observed from what remains hypothetical. The observed layer consists of the GSE157198 clinical-state signal, participant-level cellular localization in GSE256089, and the PXD012263 protein-response patterns. These observations motivate, but do not prove, mechanistic relationships among interferon state, cell lineage, lipid/cholesterol metabolism, nucleotide availability, proteostasis/autophagy, redox biology, and restriction-factor abundance or function. The next step is therefore perturbational. Cholesterol synthesis or trafficking, nucleotide availability, redox state, autophagy, and proteostasis can be manipulated in defined immune-cell compartments while measuring both restriction-factor protein abundance and antiviral function. Such experiments would distinguish a correlated host state from a causal regulatory edge. Translationally, the present results argue that host-directed strategies should be evaluated in the cell type and molecular layer in which the relevant restriction program operates rather than assumed to exert uniform effects across immune compartments.

Materials and methods

Dataset selection and source-data processing

This study is an evidence-structured computational reanalysis of public, deidentified datasets and published source tables. The datasets were selected as complementary host-response evidence layers rather than as a subject-matched multi-omics cohort. GSE157198 contributed enriched monocyte RNA-sequencing profiles from healthy controls, ART-treated participants, ART-naive participants, and elite controllers [35]. GSE256089 contributed HTO-tagged PBMC single-cell/CITE-seq profiles from the ART time point of the A5354 cohort [33]. PXD012263 contributed published quantitative proteomic source data from HIV-exposed primary CD4 T cells [34]. Each dataset was analyzed within its own sampling and experimental structure; no cross-study subject matching or latent-factor integration was performed.

Prespecified module framework and scoring

Before dataset-specific analysis, I defined eight modules: core HIV restriction factors; interferon/innate sensing; nucleotide metabolism; lipid/cholesterol membrane metabolism; mitochondrial/redox biology; glycolysis/PPP/inflammatory metabolism; autophagy/proteostasis/ER stress; and immune-cell state/activation/exhaustion. Membership was based on established restriction-factor, interferon, and immunometabolic biology [9–11,16,21,25,26] and was not tuned to any dataset. Module coverage was calculated separately for each molecular layer. Scores were computed from normalized abundance values using only detected genes or proteins, and all comparisons were interpreted in relation to the available feature coverage.

Bulk monocyte RNA-sequencing analysis

For GSE157198, processed counts were parsed from the GEO supplementary workbook, mapped to gene symbols, and combined with curated sample metadata. Quality-control summaries included library size and detected-gene distributions. Module scores were calculated after coverage assessment, and prespecified contrasts compared untreated, ART-treated, elite-controller, and healthy-control groups. Benjamini-Hochberg adjustment controlled the false-discovery rate across module contrasts.

Within each module, gene-level contrasts were decomposed to identify the features contributing most strongly to group differences. Figs 2 and 3 were generated from the complete module- and gene-level contrast tables rather than from a post hoc subset selected for visualization.

Bulk-monocyte robustness and conditional-coupling analyses

The primary GSE157198 core-restrictome disease effect was evaluated against three prespecified null schemes, each containing 1,000 random 37-gene sets, and with 37 leave-one-gene-out scores, each omitting one member of the 37-gene core module. Sample-level relationships between the core-restrictome, interferon/innate-sensing, and lipid/cholesterol scores were quantified using Spearman correlations. To test whether the lipid relationship was caused by direct gene-set overlap, the core-lipid analysis was repeated after removing CH25H from the lipid score. To test whether lipid coupling was independent of interferon, partial Spearman correlations were calculated conditioning first on the interferon/innate-sensing score and then on interferon plus clinical group. The permutation/null analyses, jackknife outputs, observed effect estimates, and conditional-correlation results were frozen before manuscript integration and were not recomputed for figure preparation.

Single-cell analysis and reservoir metadata linkage

For GSE256089, 10x matrices were processed by sequencing lane, and gene-expression features were separated from antibody-derived and HTO features. Low-information cells were removed during preliminary quality control. Broad immune-cell identities were assigned using prespecified marker rules, and HTO labels were mapped to participant identifiers, recovering all 38 A5354 participants. Module scores were summarized at the cell, cell-type, participant, and participant-by-cell-type levels; participant-by-cell-type summaries were interpreted only when the cell-count thresholds documented in S5 Table were met.

Reservoir metadata were normalized before linkage. Participant identifiers were harmonized between module-score outputs and the published demographic and reservoir table. Participants were classified as undetectable, middle, or detectable according to the source categories. Module comparisons were performed at participant and participant-by-cell-type levels, with multiple-testing correction applied across the prespecified module tests.

Participant-level NK specificity and monocyte-subtype analyses

For inferential single-cell analyses, the participant rather than the cell was the independent unit. For each of the 38 participants, broad-cell-type core-restrictome scores were summarized separately. NK enrichment was quantified as the difference between the participant’s NK score and that participant’s mean score across B cells, monocytes, and T cells. Specificity was evaluated against 1,000 random gene sets per prespecified matching scheme. Null sets were matched for the size of the core-restrictome module and for gene-expression/detection structure. Empirical upper-tail probabilities were calculated by comparing the observed participant-level NK enrichment with the matched null distributions. Existing single-cell monocyte annotations supported participant-paired CD14 and FCGR3A comparisons. Core-restrictome scores were therefore summarized for each subtype within each participant and analyzed as paired participant-level values. No validated participant-linked CD56bright/CD56dim single-cell annotation was available, so no NK-subtype transcriptional comparison was generated.

Exploratory innate-sensing and cell-death targets

cGAS/MB21D1 and STING/TMEM173 were already represented in the prespecified interferon/innate-sensing framework. In response to peer review, IFI16, TREX1, CASP1, and CASP3 were added as explicitly exploratory targets without changing the membership or multiplicity structure of the original prespecified modules. Bulk-monocyte and participant-level single-cell expression of these targets was evaluated using the corresponding frozen dataset-specific analysis framework. These tests quantify transcript abundance and cell-context differences only. They were not interpreted as direct measurements of cGAS-STING activation, inflammasome activation, pyroptosis, caspase cleavage, or apoptosis.

Proteomic overlay and cross-layer integration

For PXD012263, I used the published eLife source-data workbooks rather than reprocessing raw mass-spectrometry spectra. Candidate rows were matched to the curated framework by gene identifier. The historical published q < 0.05 WT-versus-mock selection was retained as the primary differential proteomic boundary. No quantitative protein-abundance imputation was performed in this reanalysis. Missing abundance entries in the source tables remained missing. Replicate summaries used available nonmissing values. A pseudocount of 1 was included in derived log2-ratio calculations and did not replace missing protein-abundance measurements. The identified fill operations were confined to non-abundance derived fields: a zero fallback was used for a derived protein-priority score and 99 was used for categorical sorting order. For sensitivity analysis, the existing published q values were evaluated at q < 0.01, q < 0.05, and q < 0.10 without recomputing the source-study multiple-testing adjustment. The historical q < 0.05 membership remained primary. Time-course abundance summaries were treated as descriptive source-table overlays and were not analyzed as new time-course differential-significance tests.

Cross-layer integration was evidence structured rather than based on a latent-factor model because the datasets differed in cohort, cell type, and experimental design. Evidence was indexed by dataset, molecular layer, biological context, module, direction, and interpretation. Convergence was defined as complementary support for the same organizational principle; no subject-level trajectory or causal relation was inferred across studies.

Statistical analysis and reproducibility

All reported module contrasts were prespecified. Benjamini-Hochberg-adjusted P values were used for families of module comparisons, and both positive and null results were retained. Exact sample counts, feature coverage, effect estimates, P values, adjusted P values, cell-count thresholds, and source-to-figure mappings are provided in S1-S10 Tables. For the revision analyses, sample- or participant-level inference was used as appropriate to each dataset. Cells were not treated as independent replicates. Robustness analyses included permutation/null checks, leave-one-gene-out sensitivity, participant-paired comparisons, matched random-gene-set null distributions, partial correlations, monocyte-composition sensitivity, and FDR-threshold sensitivity using existing published q values. These analyses were used to test stability or specificity of existing claims rather than to redefine the prespecified biological modules. Author-generated analysis code and environment documentation underlying the reported findings are publicly available without restriction in Zenodo at https://doi.org/10.5281/zenodo.22396354.

Ethics statement

This study used publicly available, deidentified processed datasets and published source-data tables. No new participants were recruited, and no new human specimens or animal experiments were generated. The original studies reported their own ethics approvals and consent procedures.

Use of artificial intelligence tools

ChatGPT (OpenAI) was used to assist with code verification, manuscript organization, and language editing. Its outputs were checked against author-reviewed computational results, source-data documentation, and the cited literature. No primary data were generated, altered, or interpreted autonomously by the tool. The author reviewed and revised all affected code and text and is responsible for the final manuscript.

Supporting information

S1 Table. Dataset inventory and accession-level provenance.

(XLSX)

pone.0356591.s001.xlsx^{(9KB, xlsx)}

S2 Table. Curated restrictome-immunometabolism gene-set definitions.

(XLSX)

pone.0356591.s002.xlsx^{(11.4KB, xlsx)}

S3 Table. GSE157198 module scores and clinical-group contrasts.

(XLSX)

pone.0356591.s003.xlsx^{(53.4KB, xlsx)}

S4 Table. GSE157198 gene-level decomposition.

(XLSX)

pone.0356591.s004.xlsx^{(176.4KB, xlsx)}

S5 Table. GSE256089 cell-level and participant-level module summaries.

(XLSX)

pone.0356591.s005.xlsx^{(192.5KB, xlsx)}

S6 Table. GSE256089 reservoir-linked analyses.

(XLSX)

pone.0356591.s006.xlsx^{(100.1KB, xlsx)}

S7 Table. PXD012263 curated proteomic overlay.

(XLSX)

pone.0356591.s007.xlsx^{(108.5KB, xlsx)}

S8 Table. PXD012263 protein-polarity classifications.

(XLSX)

pone.0356591.s008.xlsx^{(61.8KB, xlsx)}

S9 Table. Cross-layer evidence synthesis.

(XLSX)

pone.0356591.s009.xlsx^{(23.8KB, xlsx)}

S10 Table. Figure-source and reproducibility register.

(XLSX)

pone.0356591.s010.xlsx^{(10.5KB, xlsx)}

Acknowledgments

The author thanks the investigators who generated and publicly released GSE157198, GSE256089, and PXD012263, and the repositories and journal source-data services that enabled this reanalysis.

Abbreviations

ART: antiretroviral therapy
CITE-seq: cellular indexing of transcriptomes and epitopes by sequencing
DC: dendritic cell
ER: endoplasmic reticulum
HTO: hashtag oligonucleotide
IFN: interferon
NK: natural killer
PBMC: peripheral blood mononuclear cell
PPP: pentose phosphate pathway

Data Availability

All data and analysis code underlying the findings reported in this study are publicly available without restriction. The relevant repositories and persistent identifiers are as follows: NCBI Gene Expression Omnibus, GSE157198: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE157198 NCBI Gene Expression Omnibus, GSE256089: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE256089 ProteomeXchange/PRIDE, PXD012263: https://www.ebi.ac.uk/pride/archive/projects/PXD012263 Dataset DOI: https://doi.org/10.6019/PXD012263 Author-generated analysis code supporting the reported findings is publicly available without restriction in Zenodo: https://doi.org/10.5281/zenodo.22396354 The curated gene sets, dataset inventories, processed module scores, statistical contrasts, gene-level results, single-cell summaries, reservoir-linked analyses, proteomic overlays, protein-polarity classifications, cross-layer evidence summaries, and figure-source data are provided in the Supporting Information files S1–S10. No new human participant data or biological specimens were generated in this study.

Funding Statement

The author(s) received no specific funding for this work.

References

1.UNAIDS. The urgency of now: AIDS at a crossroads. 2024.
2.Barré-Sinoussi F, Chermann JC, Rey F, Nugeyre MT, Chamaret S, Gruest J, et al. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science. 1983;220(4599):868–71. doi: 10.1126/science.6189183 [DOI] [PubMed] [Google Scholar]
3.Hemelaar J, Elangovan R, Yun J, Dickson-Tetteh L, Fleminger I, Kirtley S, et al. Global and regional molecular epidemiology of HIV-1, 1990–2015: A systematic review, global survey, and trend analysis. Lancet Infect Dis. 2019;19:143–55. doi: 10.1016/S1473-3099(18)30647-9 [DOI] [PubMed] [Google Scholar]
4.Rambaut A, Posada D, Crandall KA, Holmes EC. The causes and consequences of HIV evolution. Nat Rev Genet. 2004;5(1):52–61. doi: 10.1038/nrg1246 [DOI] [PubMed] [Google Scholar]
5.Roberts JD, Bebenek K, Kunkel TA. The accuracy of reverse transcriptase from HIV-1. Science. 1988;242(4882):1171–3. doi: 10.1126/science.2460925 [DOI] [PubMed] [Google Scholar]
6.Sharp PM, Hahn BH. Origins of HIV and the AIDS pandemic. Cold Spring Harb Perspect Med. 2011;1:a006841. doi: 10.1101/cshperspect.a006841 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Swanstrom R, Coffin J. HIV-1 pathogenesis: The virus. Cold Spring Harb Perspect Med. 2012;2:a007443. doi: 10.1101/cshperspect.a007443 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Duggal NK, Emerman M. Evolutionary conflicts between viruses and restriction factors shape immunity. Nat Rev Immunol. 2012;12(10):687–95. doi: 10.1038/nri3295 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Harris RS, Hultquist JF, Evans DT. The restriction factors of human immunodeficiency virus. J Biol Chem. 2012;287(49):40875–83. doi: 10.1074/jbc.R112.416925 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Malim MH, Bieniasz PD. HIV restriction factors and mechanisms of evasion. Cold Spring Harb Perspect Med. 2012;2:a006940. doi: 10.1101/cshperspect.a006940 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Simon V, Bloch N, Landau NR. Intrinsic host restrictions to HIV-1 and mechanisms of viral escape. Nat Immunol. 2015;16(6):546–53. doi: 10.1038/ni.3156 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Brass AL, Huang I-C, Benita Y, John SP, Krishnan MN, Feeley EM, et al. The IFITM proteins mediate cellular resistance to influenza A H1N1 virus, West Nile virus, and dengue virus. Cell. 2009;139(7):1243–54. doi: 10.1016/j.cell.2009.12.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Goujon C, Moncorge O, Bauby H, Doyle T, Ward CC, Schaller T, et al. Human MX2 is an interferon-induced post-entry inhibitor of HIV-1 infection. Nature. 2013;502(7472):559–62. doi: 10.1038/nature12542 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Laguette N, Sobhian B, Casartelli N, Ringeard M, Chable-Bessia C, Segeral E, et al. SAMHD1 is the dendritic- and myeloid-cell-specific HIV-1 restriction factor counteracted by Vpx. Nature. 2011;474(7353):654–7. doi: 10.1038/nature10117 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Neil SJD, Zang T, Bieniasz PD. Tetherin inhibits retrovirus release and is antagonized by HIV-1 Vpu. Nature. 2008;451(7177):425–30. doi: 10.1038/nature06553 [DOI] [PubMed] [Google Scholar]
16.Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P, et al. A diverse range of gene products are effectors of the type I interferon antiviral response. Nature. 2011;472(7344):481–5. doi: 10.1038/nature09907 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Sheehy AM, Gaddis NC, Choi JD, Malim MH. Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature. 2002;418(6898):646–50. doi: 10.1038/nature00939 [DOI] [PubMed] [Google Scholar]
18.Usami Y, Wu Y, Göttlinger HG. SERINC3 and SERINC5 restrict HIV-1 infectivity and are counteracted by Nef. Nature. 2015;526(7572):218–23. doi: 10.1038/nature15400 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Rosa A, Chande A, Ziglio S, De Sanctis V, Bertorelli R, Goh SL, et al. HIV-1 Nef promotes infection by excluding SERINC5 from virion incorporation. Nature. 2015;526(7572):212–7. doi: 10.1038/nature15399 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Van Damme N, Goff D, Katsura C, Jorgenson RL, Mitchell R, Johnson MC, et al. The interferon-induced protein BST-2 restricts HIV-1 release and is downregulated from the cell surface by the viral Vpu protein. Cell Host Microbe. 2008;3(4):245–52. doi: 10.1016/j.chom.2008.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Clerc I, Abba Moussa D, Vahlas Z, Tardito S, Oburoglu L, Hope TJ, et al. Entry of glucose- and glutamine-derived carbons into the citric acid cycle supports early steps of HIV-1 infection in CD4 T cells. Nat Metab. 2019;1(7):717–30. doi: 10.1038/s42255-019-0084-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Doyle T, Goujon C, Malim MH. HIV-1 and interferons: Who’s interfering with whom?. Nat Rev Microbiol. 2015;13:403–13. doi: 10.1038/nrmicro3449 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.O’Neill LAJ, Kishton RJ, Rathmell J. A guide to immunometabolism for immunologists. Nat Rev Immunol. 2016;16(9):553–65. doi: 10.1038/nri.2016.70 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Pearce EL, Pearce EJ. Metabolic pathways in immune cell activation and quiescence. Immunity. 2013;38(4):633–43. doi: 10.1016/j.immuni.2013.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Saez-Cirion A, Sereti I. Immunometabolism and HIV-1 pathogenesis: Food for thought. Nat Rev Immunol. 2021;21(1):5–19. doi: 10.1038/s41577-020-0381-7 [DOI] [PubMed] [Google Scholar]
26.Valle-Casuso JC, Angin M, Volant S, Passaes C, Monceaux V, Mikhailova A, et al. Cellular metabolism is a major determinant of HIV-1 reservoir seeding in CD4+ T cells and offers an opportunity to tackle infection. Cell Metab. 2019;29(3):611–626.e5. doi: 10.1016/j.cmet.2018.11.015 [DOI] [PubMed] [Google Scholar]
27.Koning FA, Newman ENC, Kim E-Y, Kunstman KJ, Wolinsky SM, Malim MH. Defining APOBEC3 expression patterns in human tissues and hematopoietic cell subsets. J Virol. 2009;83(18):9474–85. doi: 10.1128/JVI.01089-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Stremlau M, Owens CM, Perron MJ, Kiessling M, Autissier P, Sodroski J. The cytoplasmic body component TRIM5alpha restricts HIV-1 infection in Old World monkeys. Nature. 2004;427(6977):848–53. doi: 10.1038/nature02343 [DOI] [PubMed] [Google Scholar]
29.Deeks SG, Tracy R, Douek DC. Systemic effects of inflammation on health during chronic HIV infection. Immunity. 2013;39(4):633–45. doi: 10.1016/j.immuni.2013.10.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Brown TT, Cole SR, Li X, Kingsley LA, Palella FJ, Riddler SA, et al. Antiretroviral therapy and the prevalence and incidence of diabetes mellitus in the multicenter AIDS cohort study. Arch Intern Med. 2005;165(10):1179–84. doi: 10.1001/archinte.165.10.1179 [DOI] [PubMed] [Google Scholar]
31.Kwan CK, Ernst JD. HIV and tuberculosis: A deadly human syndemic. Clin Microbiol Rev. 2011;24(2):351–76. doi: 10.1128/CMR.00042-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Pawlowski A, Jansson M, Sköld M, Rottenberg ME, Källenius G. Tuberculosis and HIV co-infection. PLoS Pathog. 2012;8(2):e1002464. doi: 10.1371/journal.ppat.1002464 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Ehrenberg PK, Geretz A, Volcic M, Izumi T, Yum LK, Waickman A, et al. Single-cell analyses identify monocyte gene expression profiles that influence HIV-1 reservoir size in acutely treated cohorts. Nat Commun. 2025;16(1):4975. doi: 10.1038/s41467-025-59833-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Naamati A, Williamson JC, Greenwood EJ, Marelli S, Lehner PJ, Matheson NJ. Functional proteomic atlas of HIV infection in primary human CD4+ T cells. eLife. 2019;8. doi: 10.7554/elife.41431 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Luo Z, Li M, Li T-W, Lv Z, Ye Z, Cisneros WJ, et al. Differential expression of CREM/ICER Isoforms is associated with the spontaneous control of HIV infection. mBio. 2022;13. doi: 10.1128/mbio.01979-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Deeks SG, Walker BD. Human immunodeficiency virus controllers: mechanisms of durable virus control in the absence of antiretroviral therapy. Immunity. 2007;27(3):406–16. doi: 10.1016/j.immuni.2007.08.010 [DOI] [PubMed] [Google Scholar]
37.Migueles SA, Connors M. Success and failure of the cellular immune response against HIV-1. Nat Immunol. 2015;16(6):563–70. doi: 10.1038/ni.3161 [DOI] [PubMed] [Google Scholar]
38.Jäger S, Cimermancic P, Gulbahce N, Johnson JR, McGovern KE, Clarke SC, et al. Global landscape of HIV-human protein complexes. Nature. 2011;481(7381):365–70. doi: 10.1038/nature10719 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.König R, Zhou Y, Elleder D, Diamond TL, Bonamy GMC, Irelan JT, et al. Global analysis of host-pathogen interactions that regulate early-stage HIV-1 replication. Cell. 2008;135(1):49–60. doi: 10.1016/j.cell.2008.07.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Park RJ, Wang T, Koundakjian D, Hultquist JF, Lamothe-Molina P, Monel B, et al. A genome-wide CRISPR screen identifies a restricted set of HIV host dependency factors. Nat Genet. 2017;49(2):193–203. doi: 10.1038/ng.3741 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Martin-Gayo E, Yu XG. Role of dendritic cells in natural immune control of HIV-1 infection. Front Immunol. 2019;10:1306. doi: 10.3389/fimmu.2019.01306 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Doitsh G, Galloway NLK, Geng X, Yang Z, Monroe KM, Zepeda O, et al. Cell death by pyroptosis drives CD4 T-cell depletion in HIV-1 infection. Nature. 2014;505(7484):509–14. doi: 10.1038/nature12940 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Monroe KM, Yang Z, Johnson JR, Geng X, Doitsh G, Krogan NJ, et al. IFI16 DNA sensor is required for death of lymphoid CD4 T cells abortively infected with HIV. Science. 2014;343(6169):428–32. doi: 10.1126/science.1243640 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Kajaste-Rudnitski A, Marelli SS, Pultrone C, Pertel T, Uchil PD, Mechti N, et al. TRIM22 inhibits HIV-1 transcription independently of its E3 ubiquitin ligase activity, Tat, and NF-kappaB-responsive long terminal repeat elements. J Virol. 2011;85(10):5183–96. doi: 10.1128/JVI.02302-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Schoggins JW. Interferon-stimulated genes: What do they all do?. Annu Rev Virol. 2019;6:567–84. doi: 10.1146/annurev-virology-092818-015756 [DOI] [PubMed] [Google Scholar]
46.Hrecka K, Hao C, Gierszewska M, Swanson SK, Kesik-Brodacka M, Srivastava S, et al. Vpx relieves inhibition of HIV-1 infection of macrophages mediated by the SAMHD1 protein. Nature. 2011;474(7353):658–61. doi: 10.1038/nature10195 [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Lahouassa H, Daddacha W, Hofmann H, Ayinde D, Logue EC, Dragin L, et al. SAMHD1 restricts the replication of human immunodeficiency virus type 1 by depleting the intracellular pool of deoxynucleoside triphosphates. Nat Immunol. 2012;13(3):223–8. doi: 10.1038/ni.2236 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Liu S-Y, Aliyari R, Chikere K, Li G, Marsden MD, Smith JK, et al. Interferon-inducible cholesterol-25-hydroxylase broadly inhibits viral entry by production of 25-hydroxycholesterol. Immunity. 2013;38(1):92–105. doi: 10.1016/j.immuni.2012.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS One. 2026 Oct 13;21(10):e0356591. doi: 10.1371/journal.pone.0356591.r001

Author response to Decision Letter 0

Transfer Alert

This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present.

5 Aug 2026

PLoS One. doi: 10.1371/journal.pone.0356591.r002

Decision Letter 0

Carlos de Noronha

1 Sep 2026

PONE-D-26-39263Immune-cell compartmentalization, interferon-immunometabolic coupling, and proteomic polarity define the human HIV restrictomePLOS One

Dear Dr. Maghembe,

Thank you for submitting your manuscript to PLOS One. After careful consideration, we feel that it has merit but does not yet fully meet PLOS One’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Oct 31 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS One offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

As the corresponding author, your ORCID iD is verified in the submission system and will appear in the published article. PLOS supports the use of ORCID, and we encourage all coauthors to register for an ORCID iD and use it as well. Please encourage your coauthors to verify their ORCID iD within the submission system before final acceptance, as unverified ORCID iDs will not appear in the published article. Only the individual author can complete the verification step; PLOS staff cannot verify ORCID iDs on behalf of authors.

We look forward to receiving your revised manuscript.

Kind regards,

Carlos Manuel de Noronha, Sc.D.

Academic Editor

PLOS One

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In the online submission form, you indicated that your data will be submitted to a repository upon acceptance. We strongly recommend all authors deposit their data before acceptance, as the process can be lengthy and hold up publication timelines. Please note that, though access restrictions are acceptable now, your entire minimal dataset will need to be made freely accessible if your manuscript is accepted for publication. This policy applies to all data except where public deposition would breach compliance with the protocol approved by your research ethics board. If you are unable to adhere to our open data policy, please kindly revise your statement to explain your reasoning and we will seek the editor's input on an exemption.

3. Thank you for stating the following in the Competing Interests section:

“The author has declared that no competing interests exist.”

We note that one or more of the authors are employed by a commercial company: “Department of Omics and Computational Biology, AfroBiomics Co. Ltd.”

1. Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form.

Please also include the following statement within your amended Funding Statement.

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”

If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

2. Please also provide an updated Competing Interests Statement declaring this commercial affiliation along with any other relevant declarations relating to employment, consultancy, patents, products in development, or marketed products, etc.

Within your Competing Interests Statement, please confirm that this commercial affiliation does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests) . If this adherence statement is not accurate and there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include both an updated Funding Statement and Competing Interests Statement in your cover letter. We will change the online submission form on your behalf.

4. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This single-author study reanalyzes three public datasets (bulk monocyte RNA-seq GSE157198, PBMC single-cell/CITE-seq GSE256089, and primary CD4 T-cell proteomics PXD012263) through a prespecified eight-module framework linking HIV restriction factors to interferon and immunometabolic programs. The main claims are that untreated infection produces the strongest interferon-coupled monocyte restrictome while elite controllers do not, that restriction and interferon/lipid programs localize to different immune compartments (NK versus monocyte and DC), and that infected CD4 T cells show bidirectional protein-level remodeling, with APOBEC3C/D/F/G, BST2, SERINC5 and CD4 depleted while TRIM22, OAS1 and STAT2 are induced.

In conclusion, the author suggests that the restrictome is a compartmentalized cell-state-dependent program rather than a uniform ISG set. This conclusion is reasonable and is the main contribution presented in this work.

The organizing question is worthwhile and the cross-layer framing may have contribution to the field: it offers a potential explanation for why transcript abundance, cell-type localization and protein behavior diverge in chronic infection. The prespecified module design, retention of null results, and explicit limitations section are commendable. However, the work is currently all descriptive and the analysis does not yet support the strength of the conclusions. Three issues are material. First, the curated framework omits the innate sensing and cell-death axis most relevant to HIV pathogenesis and CD4 T cell depletion, including IFI16, STING1 and TREX1 sensing of reverse-transcribed products and the caspase-1/inflammasome versus caspase-3 death pathways that govern CD4 depletion; CGAS is listed but never appears in any result. Second, the central coupling claims lack supporting statistics, the lipid-restrictome association is not separated from the shared interferon axis, and the headline NK finding is not controlled for lineage-driven baseline expression or for monocyte subset composition in the bulk layer. Third, Figure 6A is helpful for understanding how the HIV restrictome is organized by cell identity and immunometabolic programs. However it presents an architecture without a mechanistic hypothesis or translational interpretation, which leaves the paper's most interesting result underexploited.

A general note, the article language in many parts (starting from the title) is unnecessarily complex and difficult to interpret. To avoid losing readers from the outset, I would recommend a simpler, clearer, and more elegant title.

For example:

HIV Restrictome Organizeation by Clinical State, Cell Identity, and Immunometabolic Programs

The HIV Restrictome Across Immune Cells and Clinical States

Mapping the HIV Restrictome Across Immune Cells

Similarly, the conclusion, as stated in the introduction uses un-necessary complex language: "Together, the results support a cross-layer model in 5 which intrinsic antiviral capacity emerges from cell-state-specific network organization rather than uniform induction of a fixed gene set".

A more simplified conclusion could be: The data suggest that cells employ cell-specific combinations of antiviral defense mechanisms rather than a uniform intrinsic antiviral response.

Reviewer #2: In this manuscript, the author presents a computational synthesis investigating whether the human HIV restrictome functions as a static list of antiviral effectors or an integrated network coupled to IFN signaling and immunometabolism. The author applied an eight-module framework across three public datasets: bulk RNA sequencing of monocytes across clinical states (GSE157198), single cell CITE seq profiling of PBMCs linked to viral reservoir categories (GSE256089), and time resolved CD4+ T cell quantitative proteomics (PXD012263). Key findings show that untreated HIV drives the highest core restriction factor activity and tightest coregulation with IFN signaling in monocytes, whereas elite controllers lack a baseline hyper restrictome state. Single cell profiling localizes core restriction activity primarily to NK cells, while IFN signaling and lipid/cholesterol metabolism concentrate in monocytes and DCs. Furthermore, broad module scores do not separate patient viral reservoir categories after Benjamini Hochberg false discovery rate (FDR) correction. Finally, infected CD4+ T cells display a bidirectional proteomic response where canonical targets (i.e., APOBEC3 family, BST2, etc.) are depleted by viral accessory proteins while unantagonized ISGs are strongly upregulated.

This manuscript aligns with PLoS ONE publication criteria, with emphasis on technical soundness, methodological rigor, open data, and proper interpretation. The analytical design utilizes prespecified biological modules to eliminate post hoc selection bias, applies appropriate Benjamini Hochberg FDR adjustments across module contrasts, and avoids overclaiming causality or within person trajectories across non-subject matched cohorts. Retaining null results, such as the lack of reservoir category separation after FDR adjustment, demonstrates scientific integrity. Additionally, relying on public repositories and committing to open code sharing satisfies journal reproducibility requirements.

The study possesses several strengths, most notably its multi layered resolution that captures post transcriptional viral degradation (i.e., Vif, Vpu, or Nef targets) undetectable by transcriptomics alone. Again, predefining the eight biological modules prevents post hoc bias when analyzing complex datasets. Resolving cell lineage division of labor (NK effector profiles versus myeloid metabolic and IFN signaling) helps clarify why bulk transcriptomic studies often yield discordant results. Finally, Instead of using complex math to force together data from completely different groups of people, the study neatly organizes the findings into a clear, side-by-side comparison table. This keeps the analysis scientifically honest and accurate.

The primary weaknesses stem from structural constraints inherent to reanalyzing public data, most of which are commendably acknowledged by the author in the Discussion. Because the datasets derive from separate cohorts and experimental systems, within-subject longitudinal trajectories and direct causal relationships cannot be established. As noted in the manuscript, mass spectrometry coverage limitations create scoring depth imbalances relative to bulk RNA sequencing, and the single-cell CITE-seq analysis relies on broad marker rules rather than fine subpopulation resolution. Lastly, the author appropriately bounds the study as observational and hypothesis-generating, transparently highlighting the absence of wet-lab functional perturbation assays to validate predicted immunometabolic interactions.

To strengthen technical soundness and ensure result stability, several computational and textual revisions are recommended. First, the author should perform permutation testing against 1,000 random gene sets of equivalent size to confirm that monocyte disease state separation and NK cell restriction enrichment reflect specific biological pathways rather than background transcriptomic noise. A ‘jackknife’ or ‘leave one out sensitivity’ analysis on the 37 gene core restrictome module should also be conducted to verify that monocyte contrasts are not driven by single dominant genes such as ISG15 or IFITM1. Where cell counts permit, single cell scores should be evaluated at higher resolution, comparing CD56 dim versus CD56 bright NK cells and CD14 positive versus CD16 positive monocytes. For dataset PXD012263, missing value imputation procedures should be clarified, and sensitivity across alternative FDR thresholds beyond q < 0.05 should be confirmed.

Textually, the manuscript should consistently define the work as an evidence structured computational reanalysis to avoid confusion with clinical PRISMA meta-analyses. The Discussion should be updated to note NK cells' constitutive antiviral priming. Four key references should be added to reinforce biological mechanisms: Liu et al. 2013 (PMID 23273844) for CH25H entry restriction, Kajaste-Rudnitski et al. 2011 (PMID 21345949) for TRIM22 transcriptional restriction, Martin-Gayo & Yu 2019 (PMID 31244850) for dendritic cell programs, and Schoggins 2019 (PMID 31283436) for ISG effector networks.

Finally, to prepare the figures for final publication, remove leftover working draft notes in Figures 2F and 6B. Eliminate text collisions and overlaps of Figures 2C, 4B, 4D, 5D and 6B. Center text within boxes in Figures 1. Ensure consistent font sizing and consider colorblind-safe color schemes.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures

You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation.

NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications.

PLoS One. 2026 Oct 13;21(10):e0356591. doi: 10.1371/journal.pone.0356591.r003

Author response to Decision Letter 1

15 Sep 2026

Response to Reviewers

Manuscript ID: PONE-D-26-39263

Revised title: Cross-omics mapping of the human HIV restrictome across clinical states, immune-cell compartments, and protein responses

Author: Reuben S. Maghembe

Dear Dr. Carlos Manuel de Noronha and Reviewers,

Thank you for the careful and constructive evaluation of the manuscript. I have substantially revised the analytical interpretation and prepared additional robustness, cell-subtype, sensing/death-axis, and proteomic sensitivity analyses in response to the comments. I have also simplified the title and prose, clarified that the work is an evidence-structured computational reanalysis of independent public datasets rather than a subject-matched multi-omics study or a clinical meta-analysis, and revised the figures and their interpretation.

The principal scientific changes are: formal quantification of core-restrictome coupling with interferon and lipid/cholesterol programs; conditional analysis showing that the apparent lipid association is not independent of interferon; leave-one-gene-out and permutation/null robustness checks; participant-level matched random-gene-set analysis of NK-cell enrichment; participant-paired CD14 versus FCGR3A monocyte analysis and an independent bulk-monocyte composition sensitivity analysis; exploratory analysis of cGAS/STING and reviewer-requested sensing/death genes; clarification of missing-value handling in PXD012263; and sensitivity analysis across q<0.01, q<0.05, and q<0.10. All inferential single-cell comparisons use participants rather than cells as independent units.

I respond point by point below.

Journal requirements

J01. PLOS ONE style requirements and file naming

Response

I have prepared the revision for PLOS ONE formatting rather than the earlier transfer format. The revised title is in sentence case and has been simplified to:

Cross-omics mapping of the human HIV restrictome across clinical states, immune-cell compartments, and protein responses

The short title is:

Mapping the human HIV restrictome

Figure citations and legends use the PLOS form “Fig 1,” “Fig 2,” and so forth. The final submission package will include the clean revised manuscript, a marked-up version showing changes, the Response to Reviewers, the six separately uploaded frozen main figures, and the required supporting-information files.

Zotero-preserving manuscript integration is complete. Final submission file names will be checked during final upload-set assembly; this response uses stable section, figure, and supporting-information identifiers rather than provisional line numbers.

J02. Data availability and repository deposition

Response

The manuscript already identifies the three public source datasets using their permanent accessions: GSE157198, GSE256089, and PXD012263. The processed module scores, gene-level contrasts, participant-level single-cell summaries, reservoir-linked analyses, proteomic overlays, evidence tables, and figure-source registers are supplied in the supporting information.

The revised manuscript now identifies the permanent Zenodo repository for author-generated analysis code and derived reproducibility materials: https://doi.org/10.5281/zenodo.22396354.

J03. Amended Funding Statement and Role of Funders

Response

I confirm that this study received no specific funding. My commercial affiliation with AfroBiomics Co. Ltd. is declared, but AfroBiomics Co. Ltd. provided no salary support for this study and no study-specific financial, material, computing, software, administrative, publication, or other support. AfroBiomics Co. Ltd. had no role in study design, data collection, analysis, interpretation, decision to publish, or manuscript preparation.

Funding Statement

The author received no specific funding for this work. Reuben S. Maghembe is affiliated with AfroBiomics Co. Ltd.; however, AfroBiomics Co. Ltd. provided no salary, financial, research-material, computing, software, administrative, publication, or other material support for this study and had no role in study design, data collection and analysis, interpretation, decision to publish, or preparation of the manuscript.

J04. Updated Competing Interests Statement

Response

I confirm my commercial affiliation as Chief Executive Officer of AfroBiomics Co. Ltd. AfroBiomics Co. Ltd. provided no financial or material support for this study and had no role in study design, data collection, analysis, interpretation, decision to publish, or manuscript preparation. I have no other competing interests to declare, and AfroBiomics Co. Ltd. imposes no restrictions on sharing the study's data, code, materials, or publication.

Competing Interests Statement

I am Chief Executive Officer of AfroBiomics Co. Ltd. I have no other competing interests to declare. This commercial affiliation provided no financial or material support for this study and had no role in study design, data collection, analysis, interpretation, decision to publish, or manuscript preparation. AfroBiomics Co. Ltd. imposes no restrictions on sharing the study's data, code, materials, or publication. This does not alter my adherence to PLOS ONE policies on sharing data and materials.

J05. Reviewer-recommended references

Response

I evaluated the four publications specifically recommended by Reviewer 2. They are relevant to the revised discussion of CH25H-mediated restriction, TRIM22-mediated restriction, dendritic-cell antiviral programs, and interferon-stimulated effector networks. They were incorporated through the protected Zotero-preserving reference-integration workflow at the relevant manuscript locations.

The requested publications are identified by PMIDs 23273844, 21345949, 31244850, and 31283436. The bibliographic record associated with PMID 21345949 also has a later correction record, PMID 39868830, which was handled during reference integration as bibliographic provenance without treating it as an additional reviewer-requested citation.

Reviewer 1

R1.01. Innate sensing and cell-death axis

Comment addressed

The reviewer noted that the original framework did not adequately expose the HIV-relevant sensing and cell-death axis, including IFI16, STING, TREX1, caspase-1/inflammasome biology, caspase-3, and the absence of cGAS from the Results.

Response

I agree that this axis required more explicit treatment. cGAS and STING were not actually absent from the prespecified framework: cGAS/MB21D1 and STING/TMEM173 were already represented in the interferon/innate-sensing module. I nevertheless added a focused analysis that makes these genes explicit and added IFI16, TREX1, CASP1, and CASP3 as clearly labeled reviewer-requested exploratory targets rather than retrospectively redefining the prespecified modules.

In GSE157198, TREX1 showed clear FDR-supported variation, whereas the bulk cGAS and STING contrasts were not FDR-supported. In GSE256089, IFI16, STING/TMEM173, and CASP3 were higher in NK cells, whereas TREX1 and CASP1 were higher in monocytes; cGAS/CGAS did not show a material NK-versus-monocyte difference.

I have explicitly bounded the interpretation. Transcript abundance does not demonstrate cGAS-STING pathway activation; CASP1 expression does not demonstrate inflammasome activation or pyroptosis; and CASP3 expression does not demonstrate executioner-caspase activation or apoptosis. The revised manuscript therefore presents these findings as expression-state observations and mechanistic hypotheses rather than functional cell-death measurements.

R1.02. Formal statistics for the coupling claims

Comment addressed

The reviewer requested quantitative statistical support for the core-restrictome coupling claims.

Response

I added formal sample-level coupling statistics in GSE157198. Core-restrictome and interferon/innate-sensing scores were strongly correlated (Spearman r=0.799718, p=7.46724×10^-11). Core-restrictome and lipid/cholesterol scores were also correlated in the unadjusted analysis (Spearman r=0.451163, p=0.00211408).

These values are now stated directly in the Results and Fig 2 legend rather than describing coupling qualitatively.

R1.03. Whether lipid coupling is independent of interferon

Comment addressed

The reviewer asked whether the core-restrictome/lipid association persists independently of the shared interferon axis.

Response

I performed two sensitivity analyses. First, removing CH25H, which overlaps the core and lipid/cholesterol modules, still gave an unadjusted association (Spearman r=0.424242, p=0.00410444). However, the association disappeared when interferon was controlled: partial Spearman r=0.059104, p=0.703132 after conditioning on IFN, and r=-0.017597, p=0.909734 after conditioning on both IFN and clinical group.

I therefore revised the interpretation throughout the manuscript. Lipid/cholesterol activity is no longer described as an independently coupled restrictome program. Instead, the data support a shared IFN-associated immunometabolic architecture in this cohort.

R1.04. NK-cell lineage baseline and specificity

Comment addressed

The reviewer noted that the original observation that NK cells had the highest score could reflect constitutive lineage-specific expression rather than specific restrictome enrichment.

Response

I agree that simple rank ordering was insufficient. I therefore moved the analysis to the participant level and compared each participant's NK core-restrictome score with that participant's mean score across B cells, monocytes, and T cells. Across 38 participants, the mean NK-minus-reference difference was 0.105124, with dz=3.8490 and paired p=5.551754×10^-24. NK cells had the highest score in 35 of 38 participants.

More importantly, I compared the observed effect with 1,000 random gene sets per prespecified matching scheme, controlling for gene-set size and expression/detection structure. The NK-versus-reference enrichment remained unusual under these matched nulls, with empirical upper-tail p values of 0.001998, 0.007992, and 0.000999 across the frozen matching schemes.

I have therefore changed the interpretation. The revised manuscript describes the NK pattern as constitutive/developmental antiviral readiness together with specific enrichment of the curated restrictome beyond matched background gene sets. It does not interpret the NK signal as uniform HIV-induced activation.

All inferential comparisons use participants, not cells, as independent units.

R1.05. Monocyte subset composition

Comment addressed

The reviewer requested evaluation of whether the bulk-monocyte clinical-state result could be driven by monocyte subset composition.

Response

I addressed this in two complementary ways.

First, in GSE256089, I compared participant-paired CD14 and FCGR3A monocyte subtypes. Across 38 participants, there were 37,081 CD14 cells and 7,581 FCGR3A cells. The core-restrictome score was higher in FCGR3A than CD14 monocytes by 0.157650 on average (dz=2.95165; paired p=4.773127×10^-20), with FCGR3A exceeding CD14 in 37 of 38 participants. Monocyte composition was therefore biologically associated with restrictome state.

Second, I tested three transcriptional proxies of CD14-FCGR3A balance in the independent GSE157198 bulk-monocyte cohort. Adjustment for these proxies left the ART-naive clinical-state effect essentially unchanged at approximately 1.14–1.17, with p values approximately 0.0006–0.0011.

The revised interpretation is:

Monocyte subtype composition is biologically associated with restrictome activity, but transcriptional proxies of CD14-FCGR3A balance did not account for the clinical-state-associated restrictome elevation observed in the independent GSE157198 bulk-monocyte cohort.

R1.06. Mechanistic and translational interpretation of Fig 6

Comment addressed

The reviewer considered the architecture useful but requested an explicit mechanistic hypothesis and translational interpretation.

Response

I revised Fig 6 around three deliberately separated levels:

1. observed evidence;

2. noncausal mechanistic hypotheses; and

3. perturbational tests.

The figure no longer presents arrows as established causal pathways. Instead, the observed transcriptomic, cell-localization, and proteomic findings motivate testable hypotheses concerning IFN state, immune-cell lineage, lipid/cholesterol state, nucleotide availability, proteostasis/autophagy, and redox context.

The revised Discussion specifies perturbational tests such as manipulating cholesterol synthesis, nucleotide availability, redox state, or autophagy/proteostasis while measuring both restriction-factor abundance and antiviral function in defined immune-cell compartments. The translational implication is correspondingly bounded: host-directed interventions should be evaluated in the cell type and molecular layer in which the relevant restriction program operates rather than assumed to act uniformly across immune compartments.

R1.07. Title and prose complexity

Comment addressed

The reviewer recommended a simpler title and clearer wording.

Response

I agree and have substantially simplified the title and manuscript-wide prose.

The revised title is:

Cross-omics mapping of the human HIV restrictome across clinical states, immune-cell compartments, and protein responses

I also replaced several abstract and Introduction formulations that could imply a stronger integrated or causal model than the data support. The revised framing states that the data suggest context-specific combinations of antiviral defenses across clinical states and immune-cell compartments rather than a uniform intrinsic antiviral program.

Reviewer 2

R2.01. Random-gene-set / permutation robustness

Comment addressed

The reviewer requested null testing to establish that both the monocyte disease-state separation and NK-cell enrichment were not generic transcriptomic background effects.

Response

I addressed the two parts separately because they involve different experimental structures.

Bulk monocyte disease-state signal

The GSE157198 robustness analysis used three prespecified null schemes, with 1,000 random 37-gene sets per scheme (3,000 total). The observed ART-naive-minus-healthy core-restrictome difference was 1.153725 (Welch p=0.000745751) and was more extreme than the null distribution in all three schemes (empirical p=0.000999001 in each; BH-adjusted empirical p=0.002997 in each).

The observed core-IFN correlation (Spearman r=0.799718) was also highly unusual under the null schemes, with empirical p values of 0.000999001, 0.001998002, and 0.000999001 and corresponding BH-adjusted values of 0.002997, 0.004995, and 0.002997. By contrast, the unadjusted core-lipid relationship was less stable under expression-matched nulls (empirical p=0.0239760, 0.0989011, and 0.0929071), and the IFN-adjusted core-lipid association remained null under all schemes (empirical p=0.709291, 0.700300, and 0.715285). These results support the specificity and robustness of the disease-state and core-IFN signals while reinforcing the revised conclusion that the lipid association is not independent of interferon.

NK-cell enrichment

For the NK analysis, I generated 1,000 size- and expression/detection-matched random gene sets per prespecified null scheme. The observed participant-level NK enrichment exceeded the matched background distributions, with frozen empirical upper-tail p values of 0.001998, 0.007992, and 0.000999 across the relevant matching schemes.

Thus, the revised manuscript no longer relies on the observation that NK cells simply rank highest.

R2.02. Leave-one-gene-out analysis of the 37-gene core restrictome

Comment addressed

The reviewer requested a jackknife analysis to determine whether one dominant gene drives the GSE157198 clinical-state effect.

Response

I performed 37 leave-one-core-gene-out analyses. The ART-naive-versus-healthy disease effect retained the same direction in 37/37 jackknife analyses. This supports a distributed module-level signal rather than an effect dependent on one dominant member of the core restrictome.

The result is now stated in the Results and Fig 2 legend.

R2.03. Higher-resolution NK and monocyte subtypes

Comment addressed

The reviewer suggested CD56dim versus CD56bright NK cells and CD14-positive versus CD16-positive monocytes where the data support such analysis.

Response

I performed the higher-resolution analysis only where valid labels existed.

For monocytes, the historical single-cell annotations supported CD14 and FCGR3A subtypes, enabling participant-paired analysis. FCGR3A core-restrictome scores exceeded CD14 scores in 37 of 38 participants, with a mean difference of 0.157650 and paired p=4.773127×10^-20.

For NK cells, the single-cell dataset did not contain validated participant-linked CD56bright and CD56dim single-cell labels. Although separate flow-cytometry metadata contained CD56dim information, it did not establish matched single-cell CD56bright/CD56dim assignments suitable for the requested transcriptomic comparison. I therefore did not manufacture an NK-subtype classification from marker assumptions that were not validated in the source dataset.

The limitation is now stated explicitly.

R2.04. Missing-value imputation in PXD012263

Comment addressed

The reviewer requested clarification of missing-value handling.

Response

I traced the complete time-course overlay and polarity workflow to the source tables and adjudicated every imputation-related operation. No quantitative protein-abundance imputation was identified in this reanalysis. Missing source-table abundance values were retained as missing.

Replicate summaries used available nonmissing source values, and a pseudocount of 1 was included in derived log2-ratio calculations. The pseudocount did not replace missing protein-abundance measurements and is therefore distinct from missing-value imputation. The identified fill operations were likewise non-abundance operations: fillna(0) was used for a derived protein-priority score, and fillna(99) was used for categorical sorting order.

I have revised the Methods to state this explicitly.

R2.05. FDR-threshold sensitivity

Comment addressed

The reviewer requested evaluation beyond q<0.05.

Response

I evaluated the existing published q values at q<0.01, q<0.05, and q<0.10. Among 137 q-evaluable curated genes, the thresholds selected 10, 19, and 25 genes, respectively. The q<0.05 analysis reproduced the historical 19-gene membership exactly.

Mixed protein-response polarity persisted across the thresholds, although the relative balance of increased, depleted, and stable/subtle proteins varied with threshold. I therefore retain historical q<0.05 as the primary proteomic analysis and present q<0.01 and q<0.10 strictly as sensitivity analyses.

The time-course abundance panels remain descriptive and are not presented as independent differential-significance tests.

R2.06. Terminology and study type

Comment addressed

The reviewer recommended distinguishing the study from a clinical PRISMA meta-analysis.

Response

I agree. The manuscript now consistently uses the term:

“evidence-structured computational reanalysis of public datasets”

The three datasets are described as complementary evidence layers. They are not described as a subject-matched multi-omics cohort, and the manuscript does not imply within-subject cross-platform trajectories.

R2.07. Constitutive antiviral priming of NK cells

Comment addressed

The reviewer asked that the Discussion acknowledge the constitutive antiviral readiness of NK cells.

Response

I added this distinction to the revised Discussion. The high NK core-restrictome score is interpreted as a combination of lineage/developmental antiviral readiness and specific curated-restrictome enrichment beyond size- and expression/detection-matched null gene sets.

I explicitly avoid interpreting the broad NK signal as evidence that HIV uniformly induced the program in NK cells.

R2.08. Recommended references

Comment addressed

The reviewer recommended four publications concerning CH25H, TRIM22, dendritic-cell programs, and ISG effector networks.

Response

I evaluated all four and consider them relevant to the revised biological interpretation. They were incorporated through the protected Zotero-preserving reference-integration workflow at the relevant manuscript locations:

- PMID 23273844 — CH25H and viral-entry restriction;

- PMID 21345949 — TRIM22-mediated HIV transcriptional restriction;

- PMID 31244850 — dendritic-cell antiviral programs;

- PMID 31283436 — interferon-stimulated effector networks.

The later correction record PMID 39868830 was handled as bibliographic provenance for PMID 21345949 rather than as a fifth reviewer-requested reference.

Final reference numbering was generated through Zotero during the protected reference-integration stage.

R2.09. Textual and editorial revisions

Response

I simplified the title, Abstract, Introduction conclusion, Results interpretation, Discussion, and figure legends. I also standardized the study description as an evidence-structured computational reanalysis and maintained single-author wording throughout.

The revised text avoids claims of subject-matched multi-omics, avoids causal language unsupported by the datasets, and distinguishes observed expression from pathway activation and cell-death mechanisms.

R2.11. Figure cleanup and accessibility

Comment addressed

The reviewer requested removal of working notes, correction of text collisions, improved box centering, consistent typography, and consideration of color accessibility.

Response

I addressed these presentation issues in the final frozen figure set. The revised versions remove the working annotations, resolve the identified collisions, improve text placement and typography, and retain the scientifically approved panel architecture.

No scientific model was refitted and no figure-level statistical analysis was rerun merely for presentation cleanup.

Closing response

I appreciate the reviewers' suggestions. The additional analyses materially strengthened the manuscript by distinguishing robust clinical-state and NK signals from background structure, separating lipid/cholesterol association from the shared interferon axis, identifying a monocyte-subtype contribution without attributing the independent bulk disease effect to composition, and more clearly bounding the mechanistic interpretation of the sensing/death and proteomic results.

The revised manuscript now presents the human HIV restrictome as a context-dependent pattern observed across independent public evidence layers, with explicit separation between observed associations, mechanistic hypotheses, and future perturbational tests.

Sincerely,

Reuben S. Maghembe

Attachment

Submitted filename: Response_to_Reviewers_PONE-D-26-39263_cross_omics.docx

pone.0356591.s012.docx^{(10.6KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0356591.r004

Decision Letter 1

Carlos de Noronha

22 Sep 2026

Cross-omics mapping of the human HIV restrictome across clinical states, immune-cell compartments, and protein responses

PONE-D-26-39263R1

Dear Dr. Maghembe,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Carlos Manuel de Noronha, Sc.D.

Academic Editor

PLOS One

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #2: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #2: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #2: (No Response)

**********

6. Review Comments to the Author

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

**********

PLoS One. doi: 10.1371/journal.pone.0356591.r005

Acceptance letter

Carlos de Noronha

PONE-D-26-39263R1

PLOS One

Dear Dr. Maghembe,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS One and supporting open access.

Kind regards,

PLOS One Editorial Office Staff

on behalf of

Dr. Carlos Manuel de Noronha

Academic Editor

PLOS One

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Dataset inventory and accession-level provenance.

(XLSX)

pone.0356591.s001.xlsx^{(9KB, xlsx)}

S2 Table. Curated restrictome-immunometabolism gene-set definitions.

(XLSX)

pone.0356591.s002.xlsx^{(11.4KB, xlsx)}

S3 Table. GSE157198 module scores and clinical-group contrasts.

(XLSX)

pone.0356591.s003.xlsx^{(53.4KB, xlsx)}

S4 Table. GSE157198 gene-level decomposition.

(XLSX)

pone.0356591.s004.xlsx^{(176.4KB, xlsx)}

S5 Table. GSE256089 cell-level and participant-level module summaries.

(XLSX)

pone.0356591.s005.xlsx^{(192.5KB, xlsx)}

S6 Table. GSE256089 reservoir-linked analyses.

(XLSX)

pone.0356591.s006.xlsx^{(100.1KB, xlsx)}

S7 Table. PXD012263 curated proteomic overlay.

(XLSX)

pone.0356591.s007.xlsx^{(108.5KB, xlsx)}

S8 Table. PXD012263 protein-polarity classifications.

(XLSX)

pone.0356591.s008.xlsx^{(61.8KB, xlsx)}

S9 Table. Cross-layer evidence synthesis.

(XLSX)

pone.0356591.s009.xlsx^{(23.8KB, xlsx)}

S10 Table. Figure-source and reproducibility register.

(XLSX)

pone.0356591.s010.xlsx^{(10.5KB, xlsx)}

Attachment

Submitted filename: Response_to_Reviewers_PONE-D-26-39263_cross_omics.docx

pone.0356591.s012.docx^{(10.6KB, docx)}

Data Availability Statement

[pone.0356591.ref001] 1.UNAIDS. The urgency of now: AIDS at a crossroads. 2024.

[pone.0356591.ref002] 2.Barré-Sinoussi F, Chermann JC, Rey F, Nugeyre MT, Chamaret S, Gruest J, et al. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science. 1983;220(4599):868–71. doi: 10.1126/science.6189183 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref003] 3.Hemelaar J, Elangovan R, Yun J, Dickson-Tetteh L, Fleminger I, Kirtley S, et al. Global and regional molecular epidemiology of HIV-1, 1990–2015: A systematic review, global survey, and trend analysis. Lancet Infect Dis. 2019;19:143–55. doi: 10.1016/S1473-3099(18)30647-9 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref004] 4.Rambaut A, Posada D, Crandall KA, Holmes EC. The causes and consequences of HIV evolution. Nat Rev Genet. 2004;5(1):52–61. doi: 10.1038/nrg1246 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref005] 5.Roberts JD, Bebenek K, Kunkel TA. The accuracy of reverse transcriptase from HIV-1. Science. 1988;242(4882):1171–3. doi: 10.1126/science.2460925 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref006] 6.Sharp PM, Hahn BH. Origins of HIV and the AIDS pandemic. Cold Spring Harb Perspect Med. 2011;1:a006841. doi: 10.1101/cshperspect.a006841 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref007] 7.Swanstrom R, Coffin J. HIV-1 pathogenesis: The virus. Cold Spring Harb Perspect Med. 2012;2:a007443. doi: 10.1101/cshperspect.a007443 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref008] 8.Duggal NK, Emerman M. Evolutionary conflicts between viruses and restriction factors shape immunity. Nat Rev Immunol. 2012;12(10):687–95. doi: 10.1038/nri3295 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref009] 9.Harris RS, Hultquist JF, Evans DT. The restriction factors of human immunodeficiency virus. J Biol Chem. 2012;287(49):40875–83. doi: 10.1074/jbc.R112.416925 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref010] 10.Malim MH, Bieniasz PD. HIV restriction factors and mechanisms of evasion. Cold Spring Harb Perspect Med. 2012;2:a006940. doi: 10.1101/cshperspect.a006940 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref011] 11.Simon V, Bloch N, Landau NR. Intrinsic host restrictions to HIV-1 and mechanisms of viral escape. Nat Immunol. 2015;16(6):546–53. doi: 10.1038/ni.3156 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref012] 12.Brass AL, Huang I-C, Benita Y, John SP, Krishnan MN, Feeley EM, et al. The IFITM proteins mediate cellular resistance to influenza A H1N1 virus, West Nile virus, and dengue virus. Cell. 2009;139(7):1243–54. doi: 10.1016/j.cell.2009.12.017 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref013] 13.Goujon C, Moncorge O, Bauby H, Doyle T, Ward CC, Schaller T, et al. Human MX2 is an interferon-induced post-entry inhibitor of HIV-1 infection. Nature. 2013;502(7472):559–62. doi: 10.1038/nature12542 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref014] 14.Laguette N, Sobhian B, Casartelli N, Ringeard M, Chable-Bessia C, Segeral E, et al. SAMHD1 is the dendritic- and myeloid-cell-specific HIV-1 restriction factor counteracted by Vpx. Nature. 2011;474(7353):654–7. doi: 10.1038/nature10117 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref015] 15.Neil SJD, Zang T, Bieniasz PD. Tetherin inhibits retrovirus release and is antagonized by HIV-1 Vpu. Nature. 2008;451(7177):425–30. doi: 10.1038/nature06553 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref016] 16.Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P, et al. A diverse range of gene products are effectors of the type I interferon antiviral response. Nature. 2011;472(7344):481–5. doi: 10.1038/nature09907 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref017] 17.Sheehy AM, Gaddis NC, Choi JD, Malim MH. Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature. 2002;418(6898):646–50. doi: 10.1038/nature00939 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref018] 18.Usami Y, Wu Y, Göttlinger HG. SERINC3 and SERINC5 restrict HIV-1 infectivity and are counteracted by Nef. Nature. 2015;526(7572):218–23. doi: 10.1038/nature15400 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref019] 19.Rosa A, Chande A, Ziglio S, De Sanctis V, Bertorelli R, Goh SL, et al. HIV-1 Nef promotes infection by excluding SERINC5 from virion incorporation. Nature. 2015;526(7572):212–7. doi: 10.1038/nature15399 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref020] 20.Van Damme N, Goff D, Katsura C, Jorgenson RL, Mitchell R, Johnson MC, et al. The interferon-induced protein BST-2 restricts HIV-1 release and is downregulated from the cell surface by the viral Vpu protein. Cell Host Microbe. 2008;3(4):245–52. doi: 10.1016/j.chom.2008.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref021] 21.Clerc I, Abba Moussa D, Vahlas Z, Tardito S, Oburoglu L, Hope TJ, et al. Entry of glucose- and glutamine-derived carbons into the citric acid cycle supports early steps of HIV-1 infection in CD4 T cells. Nat Metab. 2019;1(7):717–30. doi: 10.1038/s42255-019-0084-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref022] 22.Doyle T, Goujon C, Malim MH. HIV-1 and interferons: Who’s interfering with whom?. Nat Rev Microbiol. 2015;13:403–13. doi: 10.1038/nrmicro3449 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref023] 23.O’Neill LAJ, Kishton RJ, Rathmell J. A guide to immunometabolism for immunologists. Nat Rev Immunol. 2016;16(9):553–65. doi: 10.1038/nri.2016.70 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref024] 24.Pearce EL, Pearce EJ. Metabolic pathways in immune cell activation and quiescence. Immunity. 2013;38(4):633–43. doi: 10.1016/j.immuni.2013.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref025] 25.Saez-Cirion A, Sereti I. Immunometabolism and HIV-1 pathogenesis: Food for thought. Nat Rev Immunol. 2021;21(1):5–19. doi: 10.1038/s41577-020-0381-7 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref026] 26.Valle-Casuso JC, Angin M, Volant S, Passaes C, Monceaux V, Mikhailova A, et al. Cellular metabolism is a major determinant of HIV-1 reservoir seeding in CD4+ T cells and offers an opportunity to tackle infection. Cell Metab. 2019;29(3):611–626.e5. doi: 10.1016/j.cmet.2018.11.015 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref027] 27.Koning FA, Newman ENC, Kim E-Y, Kunstman KJ, Wolinsky SM, Malim MH. Defining APOBEC3 expression patterns in human tissues and hematopoietic cell subsets. J Virol. 2009;83(18):9474–85. doi: 10.1128/JVI.01089-09 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref028] 28.Stremlau M, Owens CM, Perron MJ, Kiessling M, Autissier P, Sodroski J. The cytoplasmic body component TRIM5alpha restricts HIV-1 infection in Old World monkeys. Nature. 2004;427(6977):848–53. doi: 10.1038/nature02343 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref029] 29.Deeks SG, Tracy R, Douek DC. Systemic effects of inflammation on health during chronic HIV infection. Immunity. 2013;39(4):633–45. doi: 10.1016/j.immuni.2013.10.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref030] 30.Brown TT, Cole SR, Li X, Kingsley LA, Palella FJ, Riddler SA, et al. Antiretroviral therapy and the prevalence and incidence of diabetes mellitus in the multicenter AIDS cohort study. Arch Intern Med. 2005;165(10):1179–84. doi: 10.1001/archinte.165.10.1179 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref031] 31.Kwan CK, Ernst JD. HIV and tuberculosis: A deadly human syndemic. Clin Microbiol Rev. 2011;24(2):351–76. doi: 10.1128/CMR.00042-10 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref032] 32.Pawlowski A, Jansson M, Sköld M, Rottenberg ME, Källenius G. Tuberculosis and HIV co-infection. PLoS Pathog. 2012;8(2):e1002464. doi: 10.1371/journal.ppat.1002464 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref033] 33.Ehrenberg PK, Geretz A, Volcic M, Izumi T, Yum LK, Waickman A, et al. Single-cell analyses identify monocyte gene expression profiles that influence HIV-1 reservoir size in acutely treated cohorts. Nat Commun. 2025;16(1):4975. doi: 10.1038/s41467-025-59833-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref034] 34.Naamati A, Williamson JC, Greenwood EJ, Marelli S, Lehner PJ, Matheson NJ. Functional proteomic atlas of HIV infection in primary human CD4+ T cells. eLife. 2019;8. doi: 10.7554/elife.41431 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref035] 35.Luo Z, Li M, Li T-W, Lv Z, Ye Z, Cisneros WJ, et al. Differential expression of CREM/ICER Isoforms is associated with the spontaneous control of HIV infection. mBio. 2022;13. doi: 10.1128/mbio.01979-21 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref036] 36.Deeks SG, Walker BD. Human immunodeficiency virus controllers: mechanisms of durable virus control in the absence of antiretroviral therapy. Immunity. 2007;27(3):406–16. doi: 10.1016/j.immuni.2007.08.010 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref037] 37.Migueles SA, Connors M. Success and failure of the cellular immune response against HIV-1. Nat Immunol. 2015;16(6):563–70. doi: 10.1038/ni.3161 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref038] 38.Jäger S, Cimermancic P, Gulbahce N, Johnson JR, McGovern KE, Clarke SC, et al. Global landscape of HIV-human protein complexes. Nature. 2011;481(7381):365–70. doi: 10.1038/nature10719 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref039] 39.König R, Zhou Y, Elleder D, Diamond TL, Bonamy GMC, Irelan JT, et al. Global analysis of host-pathogen interactions that regulate early-stage HIV-1 replication. Cell. 2008;135(1):49–60. doi: 10.1016/j.cell.2008.07.032 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref040] 40.Park RJ, Wang T, Koundakjian D, Hultquist JF, Lamothe-Molina P, Monel B, et al. A genome-wide CRISPR screen identifies a restricted set of HIV host dependency factors. Nat Genet. 2017;49(2):193–203. doi: 10.1038/ng.3741 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref041] 41.Martin-Gayo E, Yu XG. Role of dendritic cells in natural immune control of HIV-1 infection. Front Immunol. 2019;10:1306. doi: 10.3389/fimmu.2019.01306 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref042] 42.Doitsh G, Galloway NLK, Geng X, Yang Z, Monroe KM, Zepeda O, et al. Cell death by pyroptosis drives CD4 T-cell depletion in HIV-1 infection. Nature. 2014;505(7484):509–14. doi: 10.1038/nature12940 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref043] 43.Monroe KM, Yang Z, Johnson JR, Geng X, Doitsh G, Krogan NJ, et al. IFI16 DNA sensor is required for death of lymphoid CD4 T cells abortively infected with HIV. Science. 2014;343(6169):428–32. doi: 10.1126/science.1243640 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref044] 44.Kajaste-Rudnitski A, Marelli SS, Pultrone C, Pertel T, Uchil PD, Mechti N, et al. TRIM22 inhibits HIV-1 transcription independently of its E3 ubiquitin ligase activity, Tat, and NF-kappaB-responsive long terminal repeat elements. J Virol. 2011;85(10):5183–96. doi: 10.1128/JVI.02302-10 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref045] 45.Schoggins JW. Interferon-stimulated genes: What do they all do?. Annu Rev Virol. 2019;6:567–84. doi: 10.1146/annurev-virology-092818-015756 [DOI] [PubMed] [Google Scholar]

[pone.0356591.ref046] 46.Hrecka K, Hao C, Gierszewska M, Swanson SK, Kesik-Brodacka M, Srivastava S, et al. Vpx relieves inhibition of HIV-1 infection of macrophages mediated by the SAMHD1 protein. Nature. 2011;474(7353):658–61. doi: 10.1038/nature10195 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref047] 47.Lahouassa H, Daddacha W, Hofmann H, Ayinde D, Logue EC, Dragin L, et al. SAMHD1 restricts the replication of human immunodeficiency virus type 1 by depleting the intracellular pool of deoxynucleoside triphosphates. Nat Immunol. 2012;13(3):223–8. doi: 10.1038/ni.2236 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0356591.ref048] 48.Liu S-Y, Aliyari R, Chikere K, Li G, Marsden MD, Smith JK, et al. Interferon-inducible cholesterol-25-hydroxylase broadly inhibits viral entry by production of 25-hydroxycholesterol. Immunity. 2013;38(1):92–105. doi: 10.1016/j.immuni.2012.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Cross-omics mapping of the human HIV restrictome across clinical states, immune-cell compartments, and protein responses

Reuben S Maghembe

Roles

Abstract

Introduction

Results

A common framework separates clinical state, cell compartment, and protein response

Fig 1. Evidence-structured analysis framework for the human HIV restrictome.

Table 1. Public datasets used in the study.

Table 2. Prespecified restrictome-immunometabolism module framework.

ART-naive monocytes show a robust interferon-linked restrictome elevation

Fig 2. Bulk-monocyte restrictome and immunometabolic architecture in GSE157198.

Multiple antiviral families drive the monocyte state

Fig 3. Gene-level drivers of monocyte restrictome and immunometabolic polarity in GSE157198.

Participant-level single-cell analysis identifies specific NK enrichment and monocyte-subtype structure

Fig 4. Single-cell localization and participant-level restrictome structure in GSE256089.

Monocyte composition does not explain the independent bulk clinical-state effect

Exploratory sensing and cell-death genes show context-specific expression without establishing pathway activation

Primary CD4 T-cell proteomics shows mixed polarity that is stable to FDR-threshold sensitivity

Fig 5. Proteomic overlay, primary q < 0.05 polarity, and FDR-threshold sensitivity in primary CD4 T cells.

Fig 6. Evidence-structured working model and testable hypotheses for the human HIV restrictome.

Table 3. Cross-layer evidence supporting the integrated model.

Table 4. Genes and proteins supporting the integrated model.

Discussion

Materials and methods

Dataset selection and source-data processing

Prespecified module framework and scoring

Bulk monocyte RNA-sequencing analysis

Bulk-monocyte robustness and conditional-coupling analyses

Single-cell analysis and reservoir metadata linkage

Participant-level NK specificity and monocyte-subtype analyses

Exploratory innate-sensing and cell-death targets

Proteomic overlay and cross-layer integration

Statistical analysis and reproducibility

Ethics statement

Use of artificial intelligence tools

Supporting information

Acknowledgments

Abbreviations

Data Availability

Funding Statement

References

Author response to Decision Letter 0

Transfer Alert

Decision Letter 0

Carlos de Noronha

Roles

Author response to Decision Letter 1

Decision Letter 1

Carlos de Noronha

Roles

Acceptance letter

Carlos de Noronha

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases