TYPE Original Research PUBLISHED 03 April 2023 DOI 10.3389/fcimb.2023.1074847 Gene dysregulation in acute HIV-1 OPEN ACCESS infection – early transcriptomic EDITED BY Namal P. M. Liyanage, Wexner Medical Center, The Ohio State analysis reveals the crucial University, United States REVIEWED BY biological functions affected Fabio Romerio, Johns Hopkins Medicine, Johns Hopkins University, United States Erica Parker 1†, Melinda A. Judge 1*†, Lucia Pastor 2,3,4, Patrick L. Collins, The Ohio State University, United States Laura Fuente-Soro 2, Chenjerai Jairoce 4, Kim W. Carter 5, *CORRESPONDENCE Denise Anderson 5, Inácio Mandomando 4, Holly D. Clifford 5, Melinda A. Judge Denise Naniche 2,4‡ and Peter Neils Le Souëf 1,5‡ melindaajudge@gmail.com † 1 School of Medicine, University of Western Australia, Crawley, WA, Australia, 2 ISGlobal, Barcelona These authors contributed Institute for Global Health, Hospital Clinic–Universitat de Barcelona, Barcelona, Spain, 3 AIDS Research equally to this article and share Institute-IrsiCaixa, Institut Germans Trias i Pujol (IGTP), Hospital Germans Trias i Pujol, Universitat first authorship Autonoma de Barcelona, Badalona, Spain, 4 Centro de Investigação em Saúde da Manhiça (CISM), ‡ Maputo, Mozambique, 5 Telethon Kids Institute, Perth, WA, Australia These authors share senior authorship SPECIALTY SECTION This article was submitted to Microbes and Innate Immunity, a section of the journal Introduction: Transcriptomic analyses from early human immunodeficiency Frontiers in Cellular and virus (HIV) infection have the potential to reveal how HIV causes widespread Infection Microbiology and lasting damage to biological functions, especially in the immune system. RECEIVED 20 October 2022 ACCEPTED 14 March 2023 Previous studies have been limited by difficulties in obtaining early specimens. PUBLISHED 03 April 2023 Methods: A hospital symptom-based screening approach was applied in a rural CITATION Parker E, Judge MA, Pastor L, Mozambican setting to enrol patients with suspected acute HIV infection (Fiebig Fuente-Soro L, Jairoce C, Carter KW, stage I-IV). Blood samples were collected from all those recruited, so that acute Anderson D, Mandomando I, Clifford HD, Naniche D and Le Souëf PN (2023) Gene cases and contemporaneously recruited, uninfected controls were included. dysregulation in acute HIV-1 infection – PBMC were isolated and sequenced using RNA-seq. Sample cellular early transcriptomic analysis reveals the composition was estimated from gene expression data. Differential gene crucial biological functions affected. Front. Cell. Infect. Microbiol. 13:1074847. expression analysis was completed, and correlations were determined doi: 10.3389/fcimb.2023.1074847 between viral load and differential gene expression. Biological implications COPYRIGHT were examined using Cytoscape, gene set enrichment analysis, and © 2023 Parker, Judge, Pastor, Fuente-Soro, enrichment mapping. Jairoce, Carter, Anderson, Mandomando, Clifford, Naniche and Le Souëf. This is an open-access article distributed under the Results: Twenty-nine HIV infected subjects one month from presentation and 46 terms of the Creative Commons Attribution uninfected controls were included in this study. Subjects with acute HIV infection License (CC BY). The use, distribution or reproduction in other forums is permitted, demonstrated profound gene dysregulation, with 6131 (almost 13% of the genome provided the original author(s) and the mapped in this study) significantly differentially expressed. Viral load was correlated copyright owner(s) are credited and that with 1.6% of dysregulated genes, in particular, highly upregulated genes involved in the original publication in this journal is cited, in accordance with accepted key cell cycle functions, were correlated with viremia. The most profoundly academic practice. No use, distribution or upregulated biological functions related to cell cycle regulation, in particular, reproduction is permitted which does not comply with these terms. CDCA7 may drive aberrant cell division, promoted by overexpressed E2F family proteins. Also upregulated were DNA repair and replication, microtubule and spindle organization, and immune activation and response. The interferome of acute HIV was characterized by broad activation of interferon-stimulated genes with antiviral functions, most notably IFI27 and OTOF. BCL2 downregulation alongside upregulation of several apoptotic trigger genes and downstream effectors may contribute to cycle arrest and apoptosis. Transmembrane protein 155 (TMEM155) Frontiers in Cellular and Infection Microbiology 01 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 was consistently highly overexpressed during acute infection, with roles hitherto unknown. Discussion: Our study contributes to a better understanding of the mechanisms of early HIV-induced immune damage. These findings have the potential to lead to new earlier interventions that improve outcomes. KEYWORDS HIV, gene expression profiling, sub-Sahara Africa (SSA), Mozambique, acute HIV infection, host microbial interactions, innate immunity Introduction 358 differentially expressed genes (DEG) identified). Zhang et al. used RNA-seq and compared peripheral blood mononuclear cells Acute HIV infection encompasses the period after HIV (PBMC) from 3 early HIV infection subjects to 3 uninfected transmission and prior to seroconversion, typically 3-12 weeks controls and also identified many DEG (n=2049) (Zhang et al., (Fiebig et al., 2003), following an initial 8-10 day undetectable 2018). Hyrcza et al. reported transcriptomic changes in CD4+ and eclipse period (Konrad et al., 2017). The rate and severity of acute CD8+ T cells from 5 subjects with early infection and 10 with infection may be predictive of disease progression and long-term longer-term infection, and found that some changes persisted sequelae, reflecting the extent of immune damage and during chronic infection (Hyrcza et al., 2007). Unsurprisingly, hyperinflammatory state during this time (Lavreys et al., 2006). dysregulation centred on immune activation, innate immune Furthermore, the early failure of immune responses to curb the viral defenses, adaptive immunity (Li et al., 2009; Zhang et al., 2018) threat and the latent viral reservoirs established during the acute and interferon-stimulated genes (Hyrcza et al., 2007). A fourth stage set the challenging environment in which researchers seek a paper, by Kazer et al. provides a detailed description of gene therapeutic vaccine and cure (Johnston and Fauci, 2008; Davenport modules in PBMC of 4 individuals with acute HIV, but the focus et al., 2019). was more on temporal changes in module weighting over time, Enhanced understanding of the pathogenesis of acute HIV rather than describing differential gene expression against pre- infection may offer insights toward personalized therapies, infection samples (Kazer et al., 2020). Although these studies vaccine development or curative treatment (Chun et al., 2015). offered key insights, conclusions are hampered by variable Additionally, diagnostic testing that enables earlier detection may definitions of early infection (with sampling up to several months reduce onward transmission during this time of heightened risk due post-transmission), small sample sizes, varied tissue types, use of to peak viremia (Pilcher et al., 2004; Wawer et al., 2005; Powers superseded technologies, and participant bias in terms of sex, et al., 2011). Transcriptomic analyses can pinpoint specific geographical location, and HIV subtypes. molecular mechanisms and pathways altered by HIV. We recently This study aimed to use gold-standard RNA-seq transcriptomic reviewed research of gene expression during HIV infection, and this profiling to describe the human host response to acute HIV revealed that studies to date have provided fragmentary insight into infection, using PBMC from a large cohort of recently infected the host response to HIV, with conclusions limited by small sample individuals as compared with HIV-negative controls. Importantly, sizes, and variation in study designs and data reporting (Judge et al., the study aimed to recruit both men and women from Eastern and 2020). Case-control study designs comparing individuals with HIV Southern Africa (ESA) where HIV-1 subtype C is predominant, to healthy controls were a common approach to describing major reflective of the generalized epidemic. We hypothesized that transcriptomic changes attributable to infection. Despite examining a cohort early in HIV infection would reveal a acknowledgement that gene expression changes are most consistent pattern of gene dysregulation among crucial biological profound during acute infection (Li et al., 2009; Judge et al., functions, and highlight early “gateway” paths and processes 2020), and the potential benefits of such research, relevant leading to known HIV pathologies. information is scarce, largely due to the well-described clinical and resource-intensive screening challenges of identifying individuals with acute infection (Cohen et al., 2010; Miller Materials and methods et al., 2010). We identified only four published studies examining human Human ethics approval host gene expression during acute HIV infection. Li et al. (2009) used microarray to describe stage-specific changes in gene This research was approved by ethical review boards at expression in inguinal lymphoid tissue, with greatest Barcelona Clinic Hospital (2011/6264), the Ministry of Health of dysregulation observed during acute infection (n=9 samples with Mozambique (461/CNBS/12) and the University of Western Frontiers in Cellular and Infection Microbiology 02 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 Australia (2019/RA/4/1/6296). Written informed consent was cytokine assay (Biorad, California), and Human Magnetic obtained from all subjects. Luminex Screening Assay (R&D, Minnesota) (Pastor et al., 2017a). Sample collection Subtyping The present analysis is a sub-study of the Gastrointestinal Previous work has demonstrated that circulating virus in biomarkers in Acute HIV-infected Mozambican Adults (GAMA) Mozambique is almost exclusively HIV-1 Group M subtype C study. Procedures of recruitment for this cohort study have been (Engelbrecht et al., 1998; Bellocchi et al., 2005; Parreira et al., described elsewhere (Pastor et al., 2017a; Pastor et al., 2017b). In 2006). One acute HIV case was the subject of a published case brief, the cohort was enrolled in a semi-rural area of Southern report (Velasco et al., 2015) and had sequencing completed to Mozambique with an HIV prevalence of 39.9% (Gonzalez et al., confirm subtype C. 2012). Patients presenting with febrile illness or for voluntary HIV testing were invited to be screened for study eligibility using antibody-based fingerprick HIV rapid tests as per national RNA extraction diagnostic guidelines (Determine HIV 1/2 (Abbott Laboratories, Illinois) and Uni-Gold (Trinity Biotech Co., Ireland)). Those testing RNA was extracted using a modified protocol for the RNeasy negative or indeterminate had HIV viral load measured via RT-PCR MinElute Clean-up kit (Qiagen, Germany). Briefly, PBMC pellets (Abbott RealTime HIV-1 assay, Illinois). Viral load positive, were resuspended in 1ml of TRIzol, shaken vigorously, 200µl of seronegative or indeterminate patients were diagnosed with acute chloroform was added and shaken again. After centrifugation at HIV and invited for follow-up four weeks later. Time-matched 14000xg for 5 minutes at 4°C, the aqueous phase was added to an HIV-uninfected controls were selected from patients who tested equal volume of 70% ethanol. After mixing by pipetting, samples negative on rapid tests and viral load PCR, and also invited back were transferred to an RNeasy MinElute spin column and four weeks later. centrifuged at 800xg for 30 seconds. The columns were washed At this follow-up visit, clinical data and specimens for with 500µl of RPE buffer by centrifugation at 8000xg for 30 seconds, immunological and microbiological evaluation were collected. The then the process was repeated using 80% ethanol. An additional Manhiça Health Research Centre (CISM) is co-located with the 500µl of 80% ethanol was added and centrifuged at 8000xg for 2 Manhiça District Hospital and contains a well-equipped clinical minutes. Once the flow through was discarded, samples were again research laboratory (Sacoor et al., 2013). PBMC were isolated from centrifuged at 8000xg for 5 minutes. To elute the RNA, 20µl of peripheral blood on the day of collection, stored in RNAprotect RNase-free water was pipetted directly onto the column membrane (Qiagen, Germany) at -80°C, then shipped to Western Australia on and stood for 1 minute prior to centrifugation at maximum speed dry ice. for 2 minutes. Quality control was performed using nanodrop (Thermo Fisher Scientific, Massachusetts) and Bioanalyzer (Agilent, California); samples with a minimum concentration of Fiebig staging 20ng/ml (total input 250ng) and an RNA integrity number ≥ 8 were sequenced. HIV-specific antibodies were analysed by Western blot assay using INNO-LIA HIV I/II Score (Innogenetics, Belgium). Staging of primary HIV infection was completed according to Fiebig and RNA-seq colleagues, 2003 (Fiebig et al., 2003). Date since infection was estimated by adding the mean time since infection for those in Illumina 50 base-pair, single-end RNA-seq was completed, using Fiebig stages I-IV (Fiebig et al., 2003) to the number of days HiSeq2000 at the Australian Genome Research Facility, with between recruitment and sample collection. minimum 20 million reads/sample. Read quality was assessed using FastQC. Reads were aligned to human reference genome hg19 using HISAT. Alignment quality was assessed using SAMStat. T-cell and plasma protein determinations Quantification and organization of read counts was performed using summarizeOverlaps (GenomicAlignments package) and voom As previously described (Pastor et al., 2017b), CD4+ and CD8+ (limma package), both from Bioconductor. Sequencing data was T cell counts were determined using CD3, CD8, CD4 and CD45 functionally validated against protein levels in plasma. Data was fluorochrome-labelled antibodies on fresh whole blood using deposited in Gene Expression Omnibus (GSE199911). Trucount tubes and FaCSCalibur flow cytometry (BD Biosciences, New Jersey). Cytokine levels in plasma were determined using enzyme-linked immunosorbent assay (Orgentec, Germany; Hycult Statistical analysis Biotech, Pennsylvania; and Immunodiagnostik, Germany) and Luminex multianalyte profiling: Human Cytokine Magnetic 30- Sample cellular composition was estimated from gene plex panel (Invitrogen, California), Bio-Plex Pro Human Th17 expression data using CIBERSORT, an in silico gene expression Frontiers in Cellular and Infection Microbiology 03 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 deconvolution tool. Estimates of CD4+ and CD8+ T cell counts satisfactorily for 50 primary HIV samples and 46 HIV negative from CIBERSORT were examined for consistency against flow samples (Figure 1) with an average 23 million reads/sample, and cytometry counts (Spearman’s rank order correlation). Cohort 80% mapping accuracy to human reference genome hg19. To descriptive statistics were completed in R 3.4.2. Independent examine batch effect, three samples were sequenced on both samples T-tests were used to compare parametric data, Mann- RNA-seq runs, with highly similar results between batches Whitney U-tests and Spearman’s rank order correlations were used (Supplementary Figure 1). Given our objective was to study the for non-parametric data, and Chi-square or Fisher’s exact tests were earliest stages of infection, further down-selection of cases was used for categorical data. Comparison of acute HIV-infected versus completed based on Fiebig staging at enrolment (Fiebig et al., 2003), uninfected control clinical data was performed using Spearman’s with only the n=29 most acute cases (Fiebig I-IV) included in rank order correlation, with statistical significance at p<0.05. further analyses. Differential gene expression analysis was completed using At PBMC collection, participants enrolled as acute HIV cases TopHat and Cufflinks (Trapnell et al., 2012). Briefly, reads were had an estimated median time since infection of 8.0 weeks (range aligned by TopHat (using the BowTie engine), and transcripts were 5.9-10.6 weeks) and 83% of cases tested positive on HIV point-of- assembled using Cufflinks. The latter involved analysis of immature care rapid tests [which have median window periods of 19 and 32 transcriptomes and final transcriptome assembly with Cuffmerge; days, respectively (Delaney et al., 2017)], while the remainder tested mapping of reads with Cuffdiff to quantify genes and transcripts and negative or indeterminate. While these rapid tests cannot confirm determine statistical significance of gene expression between full seroconversion, they indicate initiation of an adaptive immune samples; comparison with reference databases using Cuffcompare response among most cases. Median Log10 HIV viral load at PBMC to determine presence of novel genes and transcripts; and collection was 5.03 (IQR 4.48-5.49) copies/mL; viral load at management, visualization and integration of Cuffdiff data with recruitment one month earlier had been higher, at median Log10 CummeRbund. Data were exported to R for further analyses. 6.92 (IQR 6.15-7.52) copies/mL. CD4+ T cell count, and CD4 Adjusted p values were calculated by Benjamini-Hochberg false +/CD8+ ratio were significantly lower in cases compared with discovery correction (5%). Genes with adjusted p values <0.05 were controls (Table 1); there were no other significant differences considered differentially expressed. between groups. Differential gene expression was adjusted for participant age and sex, as well as RNA extraction year (a proxy for two individual extractors) to mitigate plausible batch effect, using an empirical Differential gene expression in Bayes framework. Viral load correlations were assessed using acute HIV Spearman correlation with FDR adjustment (5%) using the psych package in R. Network Analyst generated a heatmap of raw Of 47,971 mapped genes, there were 6131 (12.8%) genes expression counts. Cytoscape v3.8.1 and v3.9.1 were used to significantly differentially expressed during acute HIV-1 infection. visualize known interactions between DEG, and between DEG Of these, 3307 genes (6.9%) were upregulated and 2824 (5.9%) were and upstream regulators, using the STRING database (Szklarczyk downregulated. Such high numbers of DEG have been previously et al., 2021) and CyTargetFinder with the ENCODE transcription described for other acute infections, such as influenza (Zhai et al., factor database. Gene Set Enrichment Analysis (GSEA) was 2015), dengue (Popper et al., 2012), and tuberculosis (Wang et al., completed in GSEA 4.1.0, using a list of DEG ranked by the 2018). The top 50 DEG by significance are presented in Table 2 (full negative Log10 of adjusted p value divided by sign of the fold list in Supplementary Table 1). Notably, upregulated genes tended change (Ziemann, 2014; Dębski et al., 2016), with identifiers from to have greater significance, with the first downregulated gene Human_ENSEMBL_Gene_ID_MSigDB.v7.0.chip and gene sets appearing at position 123 when ranked by adjusted p value. from the Gene Ontology database (c5.go.bp.v7.2.symbols.gmt). Upregulation was also more profound than downregulation, with Enrichment maps were created in Cytoscape v3.8.1 using GSEA 221 genes upregulated to Log2 fold change >1.5 (the highest at results, with clusters generated using Community Cluster (GLay) Log24.84) compared with 10 genes downregulated to Log2 fold Annotation Set in the AutoAnnotate plugin, with minor manual adjustments to improve language. Results Study population Primary HIV cases (n=85) and uninfected controls (n=58) were recruited among adults aged 18-60 of both sexes, with ethnicities FIGURE 1 comprised from Indigenous African subgroups. From these RNA-seq 50 base pair single-end sequencing was completed for PMBC from acute HIV cases and uninfected controls. PBMC, subjects, 59 cases and 52 controls returned for a clinical follow-up peripheral blood mononuclear cells; RNA, ribonucleic acid; QC, visit one month later, during which PBMC samples were collected quality control. for gene expression analyses. RNA sequencing was completed Frontiers in Cellular and Infection Microbiology 04 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 TABLE 1 Characteristics of study population one month from presentation. Characteristic Acute HIV cases HIV negative controls (n=46) P value (n=29) Sex, n (%) female 18 (62.1) 35 (76.1) 0.194 Age, median (IQR) 27.0 (19.0-35.0) 25.0 (20.7-35.5) 0.699 CD4 T cell count (flow cytometry), median (IQR) 599.0 (440.0-700.5) 995.5 (764.5-1156.0) <0.001 *** CD4:CD8 ratio (flow cytometry), median (IQR) 0.50 (0.27-0.69) 1.62 (1.24-2.36) <0.001 *** Fever in previous 24 hours (self-reported), n (%) 2 (6.9) 3 (6.5) 0.726 Diarrhoea in previous week (self-reported), n (%) 2 (6.9) 3 (6.5) 0.726 Malaria in previous month (self-reported), n (%) 5 (19.2) 9 (20.0) 0.308 Hepatitis B, n (%) 4 (14.3) 2 (4.3) 0.191 Syphilis, n (%) 2 (6.9) 3 (6.8) 0.990 Entamoeba histolytica, n (%) 1 (3.6) 3 (6.8) 0.826 Giardia lamblia, n (%) 2 (7.1) 5 (11.4) 0.083 Cryptosporidium spp., n (%) 0 (0.0) 0 (0.0) NA Clostridium difficile, n (%) 0 (0.0) 1 (2.3) 0.655 Strongyloides stercoralis, n (%) 1 (0.0) 1 (2.2) 0.424 Pregnancy, n (%) 1 (3.4) 0 (0.0) 0.259 ***p < 0.001. change <-1.5 (the lowest at Log2-1.89). The top 50 DEG were highly through G1, S and G2 phases through CCNA2 (Jiang et al., consistent among cases and controls, respectively (Figure 2). 2022). Notably, CCNA2 has been associated with aberrant cell cycle progression in several types of cancer, and has been posited as a potential therapeutic target that could be inhibited to repress cell Differential gene expression correlation proliferation (Ma, 2019; Yang et al., 2020). Heightened viral load with viral load may also be associated with sister chromatid cohesion during S phase via ESCO2 (Bender et al., 2020); increased microtubule The correlation between gene expression and Log10HIV viral activity [via KIF15-led motor activity and spindle assembly load was examined for the 29 acute HIV cases, with FDR (Tanenbaum et al., 2009), and attachment to kinetochores during adjustment for multiple comparisons (a=0.05). Among the top 50 mitosis via SKA3 (Abad et al., 2016)], as well as functions in DNA DEG, 42 showed positive association with viral load (8 significantly repair and transcriptional regulation (CDCA7 and KIAA0101) (Gill correlated). None were negatively correlated. The FDR-adjusted et al., 2013; Zhang et al., 2021). CDCA7 is notable as a Myc-sensitive correlation of viral load with the top ten upregulated and top five gene, through which apoptotic pathways and cellular proliferation downregulated genes is depicted in Figure 3. can be regulated (Gill et al., 2013; Jimenez et al., 2018). When all 6131 DEG were examined, 51 genes (0.85%) showed The genes that demonstrated the strongest negative correlation significant positive correlation with viral load, while 50 genes with viral load (Supplementary Table 2) also have key functional (0.82%) showed significant negative correlation (i.e. 1.67% of all overlaps. For example, PTPN13, SFMBT1, ARMCX1, ZNF10, and DEG correlate with viral load, Supplementary Table 2). As ZBTB18 are all involved in repression of transcription (Huang et al., mentioned, eight of the genes positively correlated with viral load 2008; Nishitsuji et al., 2015; Liu et al., 2020; Heng et al., 2022; Xie were also in the Top 50 over-expressed DEG by fold change et al., 2022). Downregulation of ZNF10 during high viremia is (Table 2). With the exception of SKA3, these eight genes have noteworthy as this protein has demonstrated functional capacity to established protein-protein interactions (Figure 4), with protein limit HIV-1 gene expression in vitro (Nishitsuji et al., 2015). Several activities implying that viral load is associated with numerous core negatively correlated genes, including FGF9, ARMCX1, PMP22, functions across the cell cycle process. Specifically, these core and ZBTB18 are involved in nervous system functions, such as glial functions include initiation of DNA replication [via MCM4-led cell growth, axon regeneration, myelin development and brain increased DNA helicase unwinding (Bochman and Schwacha, tumour suppression (Li et al., 2013; Cartoni et al., 2016; Deng 2009), and loading of DNA polymerase alpha onto chromatin et al., 2018; Heng et al., 2022). High viral load may therefore be through CDC45 and possibly GINS complex activity (Bochman associated with inhibition of actors that regulate transcription and and Schwacha, 2009)]; and positive regulation of progression protect both the central and peripheral nervous systems. Frontiers in Cellular and Infection Microbiology 05 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 TABLE 2 The top 50 genes differentially expressed during acute HIV infection, eight of which were positively correlated with viral load. Gene name HGNC Log2 fold adjusted viral load viral load change P value correlation correlation coefficient adj P value 1 transmembrane protein 155 TMEM155 3.267514 4.46E-35 2 cell division cycle associated 7 CDCA7 2.47209 9.42E-34 0.619** 0.038* 3 nucleolar and spindle associated protein 1 NUSAP1 1.979959 1.08E-26 4 KIAA0101 KIAA0101 2.784338 3.2E-26 0.589** 0.048* 5 thymidylate synthetase TYMS 2.473498 3.2E-26 6 non-SMC condensin I complex, subunit G NCAPG 2.334234 3.2E-26 7 cell division cycle 45 CDC45 2.836749 1.12E-25 0.594** 0.048* 8 CD8b molecule pseudogene CD8BP 1.570053 1.12E-25 9 cyclin A2 CCNA2 2.321098 3.88E-25 0.627** 0.034* 10 denticleless E3 ubiquitin protein ligase homolog (Drosophila) DTL 2.886901 6.18E-25 11 maternal embryonic leucine zipper kinase MELK 2.489211 6.18E-25 12 kinesin family member 11 KIF11 2.168095 6.18E-25 13 CD8b molecule CD8B 1.50107 3.77E-24 14 BUB1 mitotic checkpoint serine/threonine kinase B BUB1B 2.214816 6.3E-24 15 myosin, light chain 6B, alkali, smooth muscle and non-muscle MYL6B 1.562934 6.3E-24 16 marker of proliferation Ki-67 MKI67 2.577557 6.65E-24 17 anillin, actin binding protein ANLN 2.351843 1.37E-23 18 chemokine (C-C motif) ligand 5 CCL5 1.389217 1.54E-23 19 ribonucleotide reductase M2 RRM2 2.648566 1.78E-23 20 PDZ binding kinase PBK 2.758444 1.99E-23 21 ubiquitin-like with PHD and ring finger domains 1 UHRF1 1.923329 2.07E-23 22 claspin CLSPN 2.304573 2.77E-23 23 cell division cycle 6 CDC6 2.411704 4.19E-23 24 minichromosome maintenance complex component 10 MCM10 2.817904 5.57E-23 25 cell division cycle associated 5 CDCA5 2.485974 1E-22 26 topoisomerase (DNA) II alpha 170kDa TOP2A 2.114054 1E-22 27 apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like APOBEC3H 1.665933 1E-22 3H 28 fatty acid binding protein 5 (psoriasis-associated) FABP5 1.642369 1E-22 29 kinesin family member 15 KIF15 2.38185 1.25E-22 0.663** 0.027* 30 asp (abnormal spindle) homolog, microcephaly associated ASPM 2.263024 1.57E-22 (Drosophila) 31 BRCA1 interacting protein C-terminal helicase 1 BRIP1 1.402205 1.97E-22 32 spindle and kinetochore associated complex subunit 3 SKA3 2.784406 2.39E-22 0.612** 0.042* 33 GINS complex subunit 2 (Psf2 homolog) GINS2 2.33612 4.25E-22 34 kinesin family member 7 KIF7 2.730583 4.96E-22 35 E2F transcription factor 8 E2F8 2.679807 4.96E-22 36 exonuclease 1 EXO1 2.342272 4.96E-22 (Continued) Frontiers in Cellular and Infection Microbiology 06 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 TABLE 2 Continued Gene name HGNC Log2 fold adjusted viral load viral load change P value correlation correlation coefficient adj P value 37 kinesin family member 2C KIF2C 2.156903 4.96E-22 38 polymerase (DNA directed), theta POLQ 2.096836 4.96E-22 39 minichromosome maintenance complex component 4 MCM4 1.709862 4.96E-22 0.588** 0.049* 40 TPX2, microtubule-associated TPX2 2.327686 6.45E-22 41 solute carrier family 27 (fatty acid transporter), member 2 SLC27A2 2.142166 8.65E-22 42 cell division cycle 25A CDC25A 2.626844 9.19E-22 43 kinesin family member C1 KIFC1 2.118952 1.23E-21 44 establishment of sister chromatid cohesion N-acetyltransferase 2 ESCO2 2.401353 1.63E-21 0.654** 0.028* 45 E2F transcription factor 7 E2F7 2.296616 1.73E-21 46 centromere protein M CENPM 1.953828 3.66E-21 47 nei endonuclease VIII-like 3 (E. coli) NEIL3 2.569431 4.14E-21 48 CD8a molecule CD8A 1.587898 4.22E-21 49 MIR4435-1 host gene (non-protein coding) MIR4435- 1.136808 4.22E-21 1HG 50 RAD54-like (S. cerevisiae) RAD54L 1.814108 5.04E-21 HGNC, HUGO Gene Nomenclature Committee; adjusted P value = Benjamini-Hochberg false discovery rate adjusted p-value.; viral load correlation coefficient = Spearman coefficient (only shown when also significant) * = 0.25 - <0.5 (fair); ** = 0.5 - <0.75 (moderate); viral load correlation adj P value = significance adjusted for FDR, *p<0.05; **p<0.01. Cell populations HIV negative controls, at 37.8% (p<0.001), 5.0% (p<0.001), and 6.5% (p=0.028) respectively. Other cell populations were not Gene expression data were used to estimate constituent cell significantly different between groups. These estimated data were populations within PBMC samples using in silica deconvolution validated by correlating with CD4+ and CD8+ counts via flow tool CIBERSORT (Newman et al., 2015; Newman et al., 2015) cytometry, with moderate Spearman correlations of 0.645 and 0.705 (Figure 5A). Among PBMC from acute HIV cases, CD4+ T cells respectively (both p<0.001) (Figures 5B, C). contributed median 24.3% of cells, CD8+ T cells contributed 22.2%, and B cells contributed 4.5%. These were significantly different from Differential expression after adjustment for differences in cell populations Given the significant differences in cell populations between acute cases and HIV-negative controls, we investigated whether cell counts were a confounding variable adversely impacting differential gene expression interpretation. CD4+ T cell count (as a proportion of lymphocytes) was shown to be inversely correlated with CD8+ T cell count (rho = -0.815, p = <0.001) for all study participants. Furthermore, both CD4+ and CD8+ T cell counts were associated with viral load (rho = -0.518 and 0.480, and p = 0.005 and 0.010, respectively) among HIV cases. CD4+ T cell count was therefore considered a suitable proxy with which to examine the impact of shifts in cell populations on gene expression. When CD4+ T cell count (%) was adjusted for, 2530 fewer genes were recognized as differentially expressed, and 274 additional FIGURE 2 genes were recognized as differentially expressed. The impact on Heatmap of the top 50 most upregulated genes during HIV infection, demonstrating high consistency between individuals. the makeup of the list of most differentially expressed genes was Network Analyst was used to view a heatmap of the raw expression low, with only minor rearrangements to the top 50 DEG counts for each acutely infected (n=29) and HIV negative (n=46) individual. The top 50 most significantly DEGs were all upregulated (Supplementary Table 3). Adjustment of gene expression for during acute HIV. CD4+ T cell count reduced the range of fold changes of DEG, from between log2FC 4.84 and -1.89, to between 3.1 and -1.64. Frontiers in Cellular and Infection Microbiology 07 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 A B FIGURE 3 Of genes differentially expressed during acute HIV infection, 1.6% were correlated with viral load. (A) Spearman correlation of Log10 HIV viral load with mean mapped reads of the top ten upregulated genes; (B) Spearman correlation of Log10 HIV viral load with mean mapped reads of the top five downregulated genes. p<0.05; **p<0.01; ***p<0.001 (FDR adjusted). GSEA was used to examine the biological relevance of changes 11 gene sets potentially missed pertain to downregulation of cellular that may be attributable to PBMC cell population differences organization processes (macromolecule localization, and protein between cases and controls. The 2530 genes ‘lost’ as a result of complex subunit organization), intracellular signal transduction, adjustment contributed to 671 upregulated gene sets and 301 and nucleobase compound metabolic processes. downregulated gene sets. Supplementary Figure 2 demonstrates In summary, cell population frequencies were associated with the top 20 positively and negatively enriched processes, which are both independent variables (viral load) and outcome measures important pathways in the immune response, with many directly (gene expression) in this analysis, making them an effect modifier relevant to viral pathogenesis. rather than a confounder. Given that the primary objective of this The 274 additional DEG resulting from adjustment for CD4+ T analysis was to provide a global overview of host PBMC gene cell population contributed to 96 differentially enriched gene sets. expression and pathogenic pathways during acute infection, The implication is that by not adjusting for CD4+ T cell count, these inclusive of changes reflective of diversity in viral load and cell potentially important biological processes may have been missed. populations, all subsequent analyses of biological implication were However, GSEA analysis of these genes indicated that all 23 not further adjusted for cell populations. This approach is in upregulated gene sets, and all but 11 of the 73 downregulated keeping with the previous mixed cell gene expression studies gene sets, were captured by analyses without CD4+ adjustment. The during early infection that have employed a case-control study design (Li et al., 2009; Zhang et al., 2018). We continued to adjust analyses for essential confounders (age, sex, and RNA extraction personnel). We acknowledge that studies with repeat measures, or comparing different timepoints along the course of HIV infection, would necessitate cell population adjustment. Biological interactions and upstream regulators FIGURE 4 The eight genes that are both highly overexpressed during acute The STRING database identified 35,759 established biological HIV and correlated with viral load have core cell cycle functions. Generated using Cytoscape. Nodes were colored according to interactions between the 6,131 DEG. A closer look at the top 50 broad function as per the legend. Edges represent known protein- DEG revealed high connectivity and co-upregulation between genes protein interactions in the STRING database, with the arrow indicating direction of interaction (source to target), and the weight (Figure 6). The most highly connected genes included those encoding indicating confidence of association (minimum 80%). cell division-related proteins (CDC45, BUB1B, KIF2C, NCAPG, KIF11), cell cycle regulation-related proteins (CCNA2), and DNA Frontiers in Cellular and Infection Microbiology 08 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 A B FIGURE 6 Interaction network of top 50 genes differentially expressed (by adjusted P value) during acute HIV infection. Many of the top 50 most upregulated genes have known biological interactions and upstream regulators. Generated using the STRING database in Cytoscape, with all known upstream regulators linked via CyTargetFinder and the ENCODE transcription factor database; circular nodes represent differentially expressed genes; triangular nodes represent upstream transcription factors; edges, shown as grey lines, represent known biological interactions with the arrow indicating direction of interaction (source to target); node color for genes reflects log2 fold change value (more red = more significantly differentially expressed, all upregulated in this case); node size C represents degree of connectivity to other nodes in the network. replication-related proteins (RRM2 and TOP2A). Transcription factor E2F4 was identified as an upstream regulator of many genes highly over-expressed in this cohort of individuals with acute HIV infection. E2F4 was not found to be overexpressed during acute HIV in our analysis, though functionally-similar E2F1-3, as well as E2F7-8 were overexpressed. Notably, the most significantly over-expressed gene, TMEM155, had no known interactions with other DEG, and no established upstream regulators. Dysregulated gene sets GSEA was performed to determine over-represented Gene FIGURE 5 Ontology (GO) biological processes among DEG. 1421 gene sets Cell populations among PBMC were predicted by interrogation of were upregulated in acute HIV infection (405 reaching p<0.01), differential gene expression between acute HIV cases and uninfected controls. (A) Acute HIV was associated with predictable compared with 575 biological processes downregulated (150 declines in CD4+ T cell and B cell populations, and expansion of reaching p<0.01). Normalized enrichment scores of the top 20 CD8+ T cells. Both CD4+ T cell counts (B) and CD8+ T cell counts (C) estimated using gene expression data correlated well with cell most up- and down-regulated gene sets are presented in Figure 7, counts via flow cytometry. *p<0.05. ***p<0.001. and for all highly significant gene sets in Supplementary Table 4. The 20 most upregulated sets all related to cell cycle functions. The Frontiers in Cellular and Infection Microbiology 09 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 receptor heterodimers, only IFNLR1 (part of the IFNl receptor) was upregulated. Kinases TYK2, JAK1 and JAK2 were not differentially expressed, however signal transduction molecules STAT1, STAT2 and complex subunit IRF9 were all overexpressed compared with healthy controls. Many antiviral ISGs were upregulated in early stages of infection, with IFI27, OTOF, ISG15, MX1 and USP18 the most highly overexpressed. IFI27, in particular, was the single most overexpressed gene in individuals with HIV by fold change (+4.8). Only four of the ISGs were downregulated, namely PAK3, AKT3, CLEC4A and STING1. Most ISGs can be induced by multiple types of interferon, and the antiviral ISG analysis did not reveal patterns of upregulation specific to any type of interferon response. However, GSEA results indicated that Type 1 and Type 2 pathways were most recognisable among upregulated DEG overall. Enrichment mapping GSEA results were used to inform an unbiased enrichment map of all up- and down-regulated GO biological process gene sets (Figure 9). The largest gene set clusters related to DNA replication and repair, regulation of cell cycle, microtubule and spindle organization, and immune activation and response. These functional clusters also have a high degree of gene overlap and similarity. Conversely, acute HIV downregulated functions related FIGURE 7 The top 20 most upregulated and downregulated biological process to regulation of growth and development, as well as numerous other gene sets during acute HIV. Upregulated gene sets overlap to functions in comparatively discrete pathways. These include bone promote functions relating to cell cycle, DNA repair and replication, mineralization, spinal cord and retina development, and regulation and immune response. Normalized expression score reflects the degree to which a gene set is overrepresented at the top or bottom of systemic blood pressure, among others. of a ranked list of genes, adjusted for differences in gene set size and in correlations between gene sets and the expression dataset. Brackets following gene set names indicate the size of (i.e. number of genes within) an enriched gene set. Only significantly enriched Plasma protein validation gene sets (p<0.05) are featured. The relationship between gene expression and protein expression is complex, however any observed correlation validates the reliability top 20 downregulated categories relate to a variety of biological of sequencing data, as gene expression changes are observed to be processes, with commonalities around growth and differentiation, associated with functional protein changes. Of the 29 acute HIV cases and synaptic transmission. and 46 HIV negative controls in this analysis, n=15 and n=34 respectively had corresponding protein data from the same blood draw. Quantification of n=49 protein-based biomarkers in plasma The ‘interferome’ of acute HIV infection was completed for this cohort as previously reported (Pastor et al., 2017a). Correlations between proteins in peripheral blood and their Within the GSEA results, ten pathways relating to interferon respective genes in circulating PBMC were significant in 20.4% of all stimulation were prominent (beginning at position 187 by FDR- relationships tested (Supplementary Table 5). adjusted q value, normalized enrichment score 2.04). To explore this further, we collated from existing literature a list of interferon pathway genes, as well as interferon-stimulated genes (ISGs) with Discussion known antiviral activity (Liu et al., 2011; Schoggins and Rice, 2011; Schoggins et al., 2014; Kane et al., 2016; Yang and Li, 2020; Ullah This study describes the global PBMC gene expression profile of et al., 2021; Ding et al., 2022). We mapped our differential gene humans acutely infected with HIV. By using the largest number of expression results over this list to visualize the interferome pattern subjects for an early HIV transcriptomic study to date, we have been of acute HIV infection (Figure 8). able to describe the massive gene expression dysregulation taking Of all interferon types, only IFNg was overexpressed within place in circulating PBMC, detailing the top DEG, frequency of PBMC of individuals with acute HIV (noting that extracellular correlation between DEG and viral load, up- and down-regulated distribution is not captured in this type of study). Of the interferon gene sets, and enriched biological processes. Frontiers in Cellular and Infection Microbiology 10 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 FIGURE 8 Gene expression changes for all types of interferon, key components of the Type 1, 2 and 3 interferon signalling pathways, as well as 170 other interferon-stimulated genes (ISGs) with established antiviral functions. ISGs listed in alphabetical order. Previous studies of gene expression during early HIV sites of North America (Hyrcza et al., 2007; Li et al., 2009) and expression have had six notable limitations. First, the non-specific China (Zhang et al., 2018) are regions that represent minority HIV symptoms of acute HIV make prompt diagnosis difficult and this is subtypes (Hemelaar et al., 2019; Gartner et al., 2020), with low HIV reflected in the time of sample collection, ranging from 1-5 months burden and mainly a concentrated HIV epidemic (in comparison (Hyrcza et al., 2007) or within 6 months (Zhang et al., 2018) post- with the generalized epidemic present across countries in ESA infection, or within 4 months post-documented seroconversion (Li which bear two-thirds of the worldwide burden of HIV [The et al., 2009). Kazer et al. were able to collect earlier samples but this Joint United Nations Programme on HIV/AIDS (UNAIDS), required the resource intensive approach of monitoring a 2020)]. Fifth, each study used different sample types (T cells, prospective cohort of at-risk women (Kazer et al., 2020). Second, lymph node tissue and PBMC) limiting comparability and sample sizes were small (between 3 and 9 individuals) which meaningful conclusions. Finally, Li and Hyrcza utilized hinders interpretation and reliability. Third, almost all subjects microarray technology, now largely superseded by RNA-seq with HIV were male, although almost half of new infections (Zhao et al., 2014; Judge et al., 2020). worldwide occur in women (The Joint United Nations To address these methodological and external validity issues, Programme on HIV/AIDS (UNAIDS), 2020). Fourth, the study this study used RNA-seq profiling to describe gene expression Frontiers in Cellular and Infection Microbiology 11 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 FIGURE 9 Enrichment map of enriched biological processes. Enrichment map was created in Cytoscape using GSEA results; each node represents a gene set; node size reflects gene set size; node color reflects upregulation (red) or downregulation (blue); clusters were generated using Community Cluster (GLay) Annotation Set in the AutoAnnotate plugin, with minor adjustments to improve language; edges, shown as blue lines, reflect gene set overlap between nodes; edge width reflects the similarity coefficient. changes in a large cohort of recently infected individuals, compared attempts at cell regeneration triggered by CD4+ T cell loss with contemporaneously recruited, HIV-negative controls who had (Mehandru et al., 2007), and rapid expansion of CD8+ T cells presented with similar symptoms and were demographically and immature transitional B cells, required for cytotoxic and early indistinguishable. The study recruited men and women from ESA cellular immune responses. The quadruplicate evidence of E2F- where HIV-1 subtype C is predominant. PBMC were identified as sensitive CDCA7 ranking as the second most upregulated gene, the an ideal tissue for analysis, being readily collectable in resource- strong correlation of CDCA7 with viral load, the identification of an limited settings, and able to provide a comprehensive view of the E2F-family protein as a common upstream regulator of many other host response, encompassing both target cells of HIV, and other overexpressed genes during acute infection, and the over-expression cells directly and indirectly involved in the immune and of E2F1-3 among early HIV cases, points toward a key role of E2F inflammatory response to HIV. Although RNA-seq has the proteins in promoting CDCA7-led cell division during early HIV. capacity to generate read lengths and depth of coverage greater Lymphocyte survival during acute HIV infection is a very than those applied in our experiment, sequencing using single-end delicate balance of pro-apoptotic and pro-survival signals. 50 base pairs reads has been shown to produce high quality Pathways related to cell cycle entry are significantly upregulated differential gene expression results, with negligible difference in PBMC, perhaps reflecting premature halting of cell cycle, or the compared to up to 100 base paired reads (Chhangawala et al., posited ‘tap and drain’ hypothesis (Ho et al., 1995; Vidya Vijayan 2015) and therefore was used for this study to maintain both quality et al., 2017), as a mechanism for widespread cell death. The BCL2 and economy. Studies looking in detail at structural rearrangements family of proteins is known to protect cells from apoptosis (Schinzel to the genome, alternate splicing, or allele-specific expression may et al., 2004; Kale et al., 2018). In this study, BCL2 and BCL2L2 genes necessitate a greater depth of sequencing. were both downregulated among HIV cases, whereas pro-apoptotic Changes in gene expression during acute HIV infection drive genes BIM, BMAIP1 and BIK (involved in triggering the apoptosis early immune system dysregulation, setting the stage for lifelong cascade), as well as BAX and BAK (downstream apoptosis effector infection and progressive immunological decline. Our study molecules) were all upregulated. In short, two of the key genes that identified that processes relating to cell cycle and cell division are protect lymphocytes from apoptosis are downregulated and five of the most profoundly upregulated functions during acute infection those that promote apoptosis are upregulated. Gene set enrichment (Figure 9). By design, our study is likely to have captured urgent analysis systematically corroborated upregulation of pathways Frontiers in Cellular and Infection Microbiology 12 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 related to cellular stress responses, suggesting that cell death in these expression in PBMC contrasts previous work by Smith et al. early stages is predominantly pathological rather than physiological. (2009), whereby gene expression in mixed cell lymphatic tissue At this stage in the course of infection, the immune response is showed an overwhelmingly negative correlation with viral load characterized by upregulated interferon-driven inflammatory and (95% of associations) (Smith et al., 2010). antiviral pathways, complement activation, humoral immunity The genes consistently differentially expressed during acute (particularly upregulated antigen processing and presentation), HIV, but which are not correlated with viral load (even when and increased phagocytic activity (Supplementary Table 4). We conservative pre-FDR adjusted p value thresholds are used), are described the interferome of acute HIV infection, which was particularly interesting for exploration of biomarkers of recent characterized by a broad upregulation of antiviral ISGs, with infection. These candidates have the potential to remain stable IFI27, OTOF, ISG15, MX1 and USP18 the most overexpressed. during acute infection even as viral load fluctuates. Genes fitting this Of all DEG, the most upregulated gene by fold change was profile, warranting further investigation, include cell cycle interferon, alpha-inducible protein 27 (IFI27). High levels of transcription factor genes E2F7 and E2F8, cell cycle regulation expression of IFI27 have been reported previously, across gene DTL, DNA repair gene BRIP1, and antiretroviral different stages of infection (Liu et al., 2022; Palm et al., 2022), gene APOBEC3H. and the gene was even put forth as a ‘potential therapeutic target for TMEM155 (also known as SMIM43) is a notable addition to HIV infection’ based on its downregulation after ART this category. Uncorrelated with viral load, and not known to commencement (Huang et al., 2022). Among HIV cases in our interact with any other over-expressed genes, TMEM155 was analysis, IFI27 was not correlated with viral load. The interferome is unexpectedly found atop the list of most highly over-expressed of dual interest, as a potential cause of aberrant inflammation, and genes (in terms of significance, but also ranked 3rd in terms of fold also in understanding the mechanisms by which HIV evades these change of overexpression). Figure 2 highlights that upregulation of interferon effectors and establishes long-term infection. this gene was consistent among all individuals with acute HIV. We did not expect to find significant differential gene expression TMEM155 was identified as significantly differentially expressed (a relating to establishment of latent reservoirs. Although the reservoir is 1.7-fold increase) in Li et al., 2009, but was not highlighted further thought to exist from the earliest stages of infection, only small as an important gene. The chromosomal region encoding numbers (one per million) of memory and functional CD4+ T cells TMEM155 was found to be one of many differentially methylated go on to be latently infected (Kuo and Lichterfeld, 2018), and at this during HIV infection (Shiau, 2017). Outside of HIV, TMEM155 has stage of infection there is already a diminishing population of CD4+ been associated with hepatitis C liver cirrhosis (Ijaz et al., 2019) and T cells relative to total lymphocytes. Other cell types contributing to brain tumours (Wang et al., 2010; Grupenmacher et al., 2013), and the reservoir would comprise an even smaller proportion within our is possibly associated with essential tremor (Odgerel et al., 2019), samples (Kuo and Lichterfeld, 2018). Of four markers thought to be basal cell nevus syndrome fibroblasts (Phatak et al., 2019), overexpressed in infected, transcriptionally-active T cells (Descours nystagmus (Toyono et al., 2015), cerebellar atrophy with motor et al., 2017; Pardons et al., 2019), only MKI67 was upregulated; neuronopathies (Burns et al., 2018), and regulation of female FCGR2A and TNFRSF9 were not differentially expressed and IL2RA fertility (Barragá n et al., 2017). was downregulated. However, LAG3, TIGIT and HAVCR2 [three of The biological functions downregulated during acute HIV four immune checkpoint markers found to be enriched in cells with infection are remarkable for their resemblance to the longer-term chromosomally-integrated HIV-1 proviruses (Chew et al., 2016; consequences of HIV infection. Hypertension, retinopathy, Fromentin et al., 2016; Pardons et al., 2019)] were over-expressed osteopenia and muscular atrophy are well characterized sequelae among HIV cases in this analysis. All three genes have a broad array of HIV, and this study demonstrates that the gene sets potentially of roles in immune responses, so their presence likely has alternate predictive of these changes may well be disturbed within the first explanations. Gene expression analyses would be of benefit to the few weeks of infection (Figure 5B). ongoing investigation of reservoirs, though single-cell approaches Our study has a number of inherent limitations. Completing would be better equipped for this task than our global PMBC study. analyses on mixed cell samples (as opposed to single cell) does Acute HIV infection, particularly subtype C, is associated with introduce a level of effect modification which we have attempted to tremendously high viral load. Among all DEG, 0.8% positively describe using cell population deconvolution analysis to estimate correlated with viral load, including 8 of the top 50 most profoundly population frequencies, and describing the effects of CD4+ T cell upregulated genes (16%) which have core cell cycle functions. This population adjustment. As with all genome-wide association suggests that high viremia may drive key pathological processes studies, we cannot rule out an element of reverse causation; for seen during acute infection, particularly cell division disruption. example, there is a possibility that some genes were over-expressed Gene expression correlating with viral load has been previously as a by-product of heightened cell turnover, rather than because reported, with strong associations between expression of interferon- they contribute mechanistically to the turnover. Almost certainly stimulated genes and antiretroviral defense genes with viremia our DEG results reflect a combination of both. Others have (Rotger et al., 2010). An almost equal amount (0.8%) of genes attempted to untangle this issue in vitro for specific genes, differentially expressed in PBMC during acute HIV infection were arguing that individual genes are essential mediators of cell negatively correlated with viral load. The relatively balanced pattern proliferative signalling, rather than consequences of said of positive and negative correlation of viral load with gene proliferation (Jimenez et al., 2018; Zhang et al., 2021). Frontiers in Cellular and Infection Microbiology 13 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 Compared to previous studies, this study offers a large sample repositories and accession number(s) can be found below: https:// size of PMBC collected ex vivo from patients in the earliest stages of www.ncbi.nlm.nih.gov/geo/, GSE199911. HIV infection, prior to treatment. Although sequencing of samples from the recruitment visit would have been ideal, it was unfeasible for our rural laboratory to process more than 3000 blood samples Ethics statement collected at this time point, when only 2.7% of participants would be diagnosed with acute HIV infection (Pastor et al., 2017a). The studies involving human participants were reviewed and Additionally, most study participants presented with a non- approved by ethical review boards at Barcelona Clinic Hospital specific febrile illness, and recall at four weeks post-recruitment (2011/6264), the Ministry of Health of Mozambique (461/CNBS/ allowed time for resolution of other illnesses, thereby enabling 12) and the University of Western Australia (2019/RA/4/1/6296). comparisons with relatively healthy, time-matched controls. The The patients/participants provided their written informed consent burden of acute infections is high in ESA; however, Table 1 clearly to participate in this study. shows that there were no significant differences in infections other than HIV between the groups. In this cohort, gene expression was measured intracellularly in Author contributions PBMC, whereas protein biomarkers were measured in plasma (Pastor et al., 2017a). Despite the limitations of correlating Conceptualization, EP, MJ, and PS; Methodology, EP and MJ; intracellular gene expression with peripherally circulating Software, EP, MJ, KC, and DA; Formal Analysis, EP and MJ; proteins, 20% of measured proteins correlated with gene Investigation, all authors; Resources, EP, MJ, LP, LF-S, and CJ; expression, approximating previous reports (Chen et al., 2002; Du Data Curation, EP and MJ; Writing – Original Draft, EP and MJ; et al., 2014). Although RNA-seq does not require functional Writing – Review & Editing, all authors; Visualization, EP and MJ; validation in the classical sense (Coenye, 2021), these findings Supervision, DN and PS; Project Administration, EP and support the biological plausibility of our results. MJ; Funding Acquisition, HC, DN and PS. All authors contributed to the article and approved the submitted version. Conclusions Funding This work demonstrates that acute HIV severely disrupts circulating immune cells, profoundly upregulating many genes, This work was supported by Bill and Melinda Gates Foundation while downregulating others. The degree of dysregulation was [OPP1068252] to DN, The Spanish Ministry of Science (Mineco) frequently, but not universally, correlated with viral load. [SAF-2011-27901] to DN, the Australian Government Research Analysis of the impacted genes, gene sets and pathways reveals in Training Program Scholarship to EP and to MJ, the University of detail the mechanisms behind the shift toward a system of high cell Western Australia Safety Net Top-Up Scholarship to EP and to MJ, turnover, cell death and cell expansion, and unwitting viral and The Spanish Ministry of Health through the Institute of Health propagation. This work highlights a need for further investigation Carlos III (ISCIII) [FI12/00096] to LP. ISGlobal is a member of the of the role of E2F transcription factors in driving exhaustive cell CERCA Programme, Generalitat de Catalunya. The funders had no division during acute infection; the role of BCL2 family role in study design, data collection and analysis, decision to downregulation in permitting unchecked lymphocyte death, and publish, or preparation of the manuscript. the unknown role of highly over-expressed TMEM155 during acute infection. CCNA2, CDCA7, and IFI27 are being explored as therapeutic targets within oncology spheres, and heightened Acknowledgments expression of these transcripts in acute infection suggest that learnings may be relevant for application in HIV treatments. The authors are grateful for the continued support of the Additionally, the high degree of consistency in gene expression clinical staff at the Manhiç a District Hospital, as well as the among HIV cases hints at the possibility that transcriptomic research, laboratory and administrative staff of the Centro de biomarkers might be useful as indicators of recent incident HIV Investigação de Saú de de Manhiça (CISM). We would particularly infection. These findings illuminate the molecular mechanisms like to thank the study participants who contributed to this underlying the pathogenesis of acute HIV infection in vivo, and research. We also thank Julià Blanco and Jorge Carrillo for their have implications for designing therapeutic interventions aiming to work on plasma biomarkers which is referenced in this paper. prevent or halt this early immune damage. Conflict of interest Data availability statement The authors declare that the research was conducted in the The transcriptomic datasets presented in this study can be absence of any commercial or financial relationships that could be found in online repositories. The names of the repository/ construed as a potential conflict of interest. Frontiers in Cellular and Infection Microbiology 14 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 SUPPLEMENTARY TABLE 3 Publisher’s note Effect of adjustment for CD4+ T cell count (%) on Top 50 DEG. Pos. = Position ranked by adj.P.Val; HGNC = HUGO Gene Nomenclature Committee name All claims expressed in this article are solely those of the authors of the gene; logFC = Log2 fold change; AveExpr = average expression across all samples, in log2 counts per million reads; adj.P.Val = Benjamini- Hochberg and do not necessarily represent those of their affiliated organizations, false discovery rate adjusted p-value; no-adj pos. = comparative position in or those of the publisher, the editors and the reviewers. Any product the previous table, without adjustment for CD4+ T cell count (%). Yellow = that may be evaluated in this article, or claim that may be made by its top 50 DEG concordant in both analyses; orange = top 50 DEG discordant between analyses (though all within top 100). manufacturer, is not guaranteed or endorsed by the publisher. SUPPLEMENTARY TABLE 4 Full list of differentially expressed gene sets – acute HIV cases versus HIV Supplementary material negative controls as determined by gene set enrichment analysis (GSEA). (A) Overexpressed gene sets. (B) Underexpressed gene sets. # = Rank by NES; The Supplementary Material for this article can be found online at: Name = Gene Ontology name of the geneset; Size = the number of genes in the geneset; NES = Normalized Enrichment Score; NOM p-val = nominal P https://www.frontiersin.org/articles/10.3389/fcimb.2023.1074847/ value (statistical significance of the observed ES relative to null distribution); full#supplementary-material FDR q-val = False Discovery Rate Q value (the P value has been adjusted for FDR). SUPPLEMENTARY TABLE 1 Full list of differentially expressed genes – acute HIV cases versus HIV SUPPLEMENTARY TABLE 5 negative controls. # = Rank by adj.P.Val; EnsemblID = Ensembl Stable ID Intracellular gene expression changes correlated moderately with measured name of the transcript; HGNC = HUGO Gene Nomenclature Committee plasma proteins (significant results only, of forty-nine quantifiable name of the gene; logFC = Log2 fold change; AveExpr = average expression soluble biomarkers). across all samples, in log2 counts per million reads; adj.P.Val = Benjamini- Hochberg false discovery rate adjusted p-value. SUPPLEMENTARY FIGURE 1 Mean mapped reads of three case samples as sequenced on two different SUPPLEMENTARY TABLE 2 RNAseq batches. Full list of differentially expressed genes that correlate with viral load in acute HIV cases. # = Rank by rho; EnsemblID = Ensembl Stable ID name of the transcript; SUPPLEMENTARY FIGURE 2 HGNC = HUGO Gene Nomenclature Committee name of the gene; rho = Top 20 most up and down-regulated gene sets, associated with adjustment Spearman correlation coefficient; adj.P.Val = false discovery rate adjusted p-value. for cell populations. References Abad, M. A., Zou, J., Medina-Pritchard, B., Nigg, E. A., Rappsilber, J., Santamaria, A., Coenye, T. (2021). Do results obtained with RNA-sequencing require independent et al. (2016). Ska3 ensures timely mitotic progression by interacting directly with verification? Biofilm 3, 100043. doi: 10.1016/j.bioflm.2021.100043 microtubules and Ska1 microtubule binding domain. Sci. Rep. 6 (1), 34042. doi: Cohen, M. S., Gay, C. L., Busch, M. P., and Hecht, F. M. (2010). The detection of 10.1038/srep34042 acute HIV infection. J. Infect. Dis. 202 Suppl 2, S270–S277. doi: 10.1086/655651 Barragá n, M., Pons, J., Ferrer-Vaquer, A., Cornet-Bartolomé , D., Schweitzer, A., Dębski, K. J., Pitkanen, A., Puhakka, N., Bot, A. M., Khurana, I., Harikrishnan, K. N., Hubbard, J., et al. (2017). The transcriptome of human oocytes is related to age and et al. (2016). Etiology matters – genomic DNA methylation patterns in three rat models ovarian reserve. Mol. Hum. Reprod. 23 (8), 535–548. doi: 10.1093/molehr/gax033 of acquired epilepsy. Sci. Rep. 6 (1), 25668. doi: 10.1038/srep25668 Bellocchi, M. C., Forbici, F., Palombi, L., Gori, C., Coelho, E., Svicher, V., et al. Davenport, M. P., Khoury, D. S., Cromer, D., Lewin, S. R., Kelleher, A. D., and Kent, (2005). Subtype analysis and mutations to antiviral drugs in HIV-1-infected patients S. J. (2019). Functional cure of HIV: The scale of the challenge. Nat. Rev. Immunol. 19 from Mozambique before initiation of antiretroviral therapy: Results from the DREAM (1), 45–54. doi: 10.1038/s41577-018-0085-4 programme. J. Med. Virol. 76 (4), 452–458. doi: 10.1002/jmv.20382 Delaney, K. P., Hanson, D. L., Masciotra, S., Ethridge, S. F., Wesolowski, L., and Bender, D., Da Silva, E. M. L., Chen, J., Poss, A., Gawey, L., Rulon, Z., et al. (2020). Owen, S. M. (2017). Time until emergence of HIV test reactivity following infection Multivalent interaction of ESCO2 with the replication machinery is required for sister with HIV-1: Implications for interpreting test results and retesting after exposure. Clin. chromatid cohesion in vertebrates. Proc. Natl. Acad. Sci. 117 (2), 1081–1089. doi: Infect. Dis. 64 (1), 53–59. doi: 10.1093/cid/ciw666 10.1073/pnas.1911936117 Deng, B., Lv, W., Duan, W., Liu, Y., Li, Z., Song, X., et al. (2018). FGF9 modulates schwann Bochman, M. L., and Schwacha, A. (2009). The mcm complex: Unwinding the cell myelination in developing nerves and induces a pro-inflammatory environment during mechanism of a replicative helicase. Microbiol. Mol. Biol. Rev. 73 (4), 652–683. doi: injury. J. Cell Biochem. 119 (10), 8643–8658. doi: 10.1002/jcb.27105 10.1128/MMBR.00019-09 Descours, B., Petitjean, G., Loez-Zaragoza, J. L., Bruel, T., Raffel, R., Psomas, C., et al. Burns, D. T., Donkervoort, S., Müller, J. S., Knierim, E., Bharucha-Goebel, D., (2017). CD32a is a marker of a CD4 T-cell HIV reservoir harbouring replication- Faqeih, E. A., et al. (2018). Variants in EXOSC9 disrupt the RNA exosome and result in competent proviruses. Nature 543 (7646), 564–56+. doi: 10.1038/nature21710 cerebellar atrophy with spinal motor neuronopathy. Am. J. Hum. Genet. 102 (5), 858– 873. doi: 10.1016/j.ajhg.2018.03.011 Ding, H., Zhang, X., Zhu, Z., Wang, S., Xiong, Y., Shang, H., et al. (2022). Membrane protein OTOF is a type I interferon-induced entry inhibitor of HIV-1 in macrophages. Cartoni, R., Norsworthy, M. W., Bei, F., Wang, C., Li, S., Zhang, Y., et al. (2016). The mBio 13 (4), e0173822. doi: 10.1128/mbio.01738-22 mammalian-specific protein Armcx1 regulates mitochondrial transport during axon regeneration. Neuron 92 (6), 1294–1307. doi: 10.1016/j.neuron.2016.10.060 Du, H. W., Chen, G. Y., Yang, C. H., Li, C. B., Xun, Y. P., Liu, J., et al. (2014). Chen, G., Gharib, T. G., Huang, C.-C., Taylor, J. M. G., Misek, D. E., Kardia, S. L. R., Multiple correlations of mRNA expression and protein abundance in human cytokine et al. (2002). Discordant protein and mRNA expression in lung adenocarcinomas *. profile. Mol. Biol. Rep. 41 (10), 6985–6993. doi: 10.1007/s11033-014-3585-8 Mol. Cell Proteomics. 1 (4), 304–313. doi: 10.1074/mcp.M200008-MCP200 Engelbrecht, S., Koulinska, I., Smith, T.-L., Barreto, J., and Van Rensburg, E. J. (1998). Chew, G. M., Fujita, T., Webb, G. M., Burwitz, B. J., Wu, H. L., Reed, J. S., et al. Sequence note: Variation in HIV type 1 V3 region env sequences from Mozambique. AIDS (2016). TIGIT marks exhausted T cells, correlates with disease progression, and serves Res. Hum. Retroviruses 14 (9), 803–805. doi: 10.1089/aid.1998.14.803 as a target for immune restoration in HIV and SIV infection. PLoS Pathogens. 12 (1), Fiebig, E. W., Wright, D. J., Rawal, B. D., Garrett, P. E., Schumacher, R. T., Peddada, e1005349. doi: 10.1371/journal.ppat.1005349 L., et al. (2003). Dynamics of HIV viremia and antibody seroconversion in plasma Chhangawala, S., Rudy, G., Mason, C. E., and Rosenfeld, J. A. (2015). The impact of donors: implications for diagnosis and staging of primary HIV infection. Aids 17 (13), read length on quantification of differentially expressed genes and splice junction 1871–1879. doi: 10.1097/00002030-200309050-00005 detection. Genome Biol. 16 (1), 131. doi: 10.1186/s13059-015-0697-y Fromentin, R., Bakeman, W., Lawani, M. B., Khoury, G., Hartogensis, W., DaFonseca, S., Chun, T.-W., Moir, S., and Fauci, A. S. (2015). HIV Reservoirs as obstacles and et al. (2016). CD4+ T cells expressing PD-1, TIGIT and LAG-3 contribute to HIV persistence opportunities for an HIV cure. Nat. Immunol. 16 (6), 584–589. doi: 10.1038/ni.3152 during ART. PLoS Pathogens. 12 (7), e1005761. doi: 10.1371/journal.ppat.1005761 Frontiers in Cellular and Infection Microbiology 15 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 Gartner, M. J., Roche, M., Churchill, M. J., Gorry, P. R., and Flynn, J. K. (2020). signatures in HIV-1 infection. J. Immunol. 183 (3), 1975–1982. doi: 10.4049/ Understanding the mechanisms driving the spread of subtype c HIV-1. EBioMedicine jimmunol.0803222 53, 102682. doi: 10.1016/j.ebiom.2020.102682 Liu, M., Guo, S., Hibbert, J. M., Jain, V., Singh, N., Wilson, N. O., et al. (2011). CXCL10/ Gill, R. M., Gabor, T. V., Couzens, A. L., and Scheid, M. P. (2013). The MYC-associated IP-10 in infectious diseases pathogenesis and potential therapeutic implications. Cytokine protein CDCA7 is phosphorylated by AKT to regulate MYC-dependent apoptosis and Growth factor Rev. 22 (3), 121–130. doi: 10.1016/j.cytogfr.2011.06.001 transformation. Mol. Cell. Biol. 33 (3), 498–513. doi: 10.1128/MCB.00276-12 Liu, X., Lin, L., Lu, L., Li, X., Han, Y., Qiu, Z., et al. (2022). Comparative Gonzalez, R., Munguambe, K., Aponte, J., Bavo, C., Nhalungo, D., Macete, E., et al. transcriptional analysis identified characteristic genes and patterns in HIV-infected (2012). High HIV prevalence in a southern semi-rural area of Mozambique: A immunological non-responders. Front. Immunol. 13. doi: 10.3389/fimmu.2022.807890 community-based survey. HIV Med. 13 (10), 581–588. doi: 10.1111/j.1468- Liu, X., Simon, J. M., Xie, H., Hu, L., Wang, J., Zurlo, G., et al. (2020). Genome-wide 1293.2012.01018.x screening identifies SFMBT1 as an oncogenic driver in cancer with VHL loss. Mol. Cell Grupenmacher, A. T., Halpern, A. L., Bonaldo Mde, F., Huang, C. C., Hamm, C. A., de 77 (6), 1294–306.e5. doi: 10.1016/j.molcel.2020.01.009 Andrade, A., et al. (2013). Study of the gene expression and microRNA expression profiles of Ma, Q. (2019). MiR-219-5p suppresses cell proliferation and cell cycle progression in malignant rhabdoid tumors originated in the brain (AT/RT) and in the kidney (RTK). Child's esophageal squamous cell carcinoma by targeting CCNA2. Cell. Mol. Biol. Lett. 24 (1), nervous system ChNS 29 (11), 1977–1983. doi: 10.1007/s00381-013-2268-4 4. doi: 10.1186/s11658-018-0129-6 Hemelaar, J., Elangovan, R., Yun, J., Dickson-Tetteh, L., Fleminger, I., Kirtley, S., Mehandru, S., Poles, M. A., Tenner-Racz, K., Manuelli, V., Jean-Pierre, P., Lopez, P., et al. (2019). Global and regional molecular epidemiology of HIV-1, 1990-2015: A et al. (2007). Mechanisms of gastrointestinal CD4+ T-cell depletion during acute and systematic review, global survey, and trend analysis. Lancet Infect. Dis. 19 (2), 143–155. early human immunodeficiency virus type 1 infection. J. Virol. 81 (2), 599–612. doi: doi: 10.1016/S1473-3099(18)30647-9 10.1128/JVI.01739-06 Heng, J. I., Viti, L., Pugh, K., Marshall, O. J., and A, M. (2022). Understanding the Miller, W. C., Rosenberg, N. E., Rutstein, S. E., and Powers, K. A. (2010). Role of impact of ZBTB18 missense variation on transcription factor function in acute and early HIV infection in the sexual transmission of HIV. Curr. Opin. HIV neurodevelopment and disease. J. Neurochem. 161 (3), 219–235. doi: 10.1111/jnc.15572 AIDS. 5 (4), 277–282. doi: 10.1097/COH.0b013e32833a0d3a Ho, D. D., Neumann, A. U., Perelson, A. S., Chen, W., Leonard, J. M., and Newman, A., Alizadeh, A., and Liu, C. (2015). CIBERSORT (Stanford, CA: Stanford Markowitz, M. (1995). Rapid turnover of plasma virions and CD4 lymphocytes in University). Available at: https://cibersort.stanford.edu/. HIV-1 infection. Nature 373 (6510), 123–126. doi: 10.1038/373123a0 Newman, A. M., Liu, C. L., Green, M. R., Gentles, A. J., Feng, W., Xu, Y., et al. (2015). Huang, H., Lv, J., Huang, Y., Mo, Z., Xu, H., Huang, Y., et al. (2022). IFI27 is a Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods 12 (5), potential therapeutic target for HIV infection. Ann. Med. 54 (1), 314–325. doi: 10.1080/ 453–457. doi: 10.1038/nmeth.3337 07853890.2021.1995624 Nishitsuji, H., Sawada, L., Sugiyama, R., and T, H. (2015). ZNF10 inhibits HIV-1 Huang, W., Zhu, C., Wang, H., Horvath, E., and EA, E. (2008). The interferon LTR activity through interaction with NK-kB and Sp1 binding motifs. FEBS Lett. 589 consensus sequence-binding protein (ICSBP/IRF8) represses PTPN13 gene (15), 2019–2025. doi: 10.1016/j.febslet.2015.06.013 transcription in differentiating myeloid cells. J. Biol. Chem. 283 (12), 7921–7935. doi: 10.1074/jbc.M706710200 Odgerel, Z., Sonti, S., Hernandez, N., Park, J., Ottman, R., Louis, E. D., et al. (2019). Whole genome sequencing and rare variant analysis in essential tremor families. PloS Hyrcza, M. D., Kovacs, C., Loutfy, M., Halpenny, R., Heisler, L., Yang, S., et al. One 14 (8), e0220512–e. doi: 10.1371/journal.pone.0220512 (2007). Distinct transcriptional profiles in ex vivo CD4+ and CD8+ T cells are established early in human immunodeficiency virus type 1 infection and are Palm, A. A., Veerla, S., Lindman, J., Isberg, P.-E., Johansson, E., Biague, A., et al. (2022). characterized by a chronic interferon response as well as extensive transcriptional Interferon alpha-inducible protein 27 expression is linked to disease severity in chronic changes in CD8+ T cells. J. Virol. 81 (7), 3477–3486. doi: 10.1128/JVI.01552-06 infection of both HIV-1 and HIV-2. Front. Virol. 2. doi: 10.3389/fviro.2022.929053 Ijaz, B., Ahmad, W., Das, T., Shabbiri, K., Husnain, T., and Hassan, S. (2019). HCV Pardons, M., Baxter, A. E., Massanella, M., Pagliuzza, A., Fromentin, R., Dufour, C., infection causes cirrhosis in human by step-wise regulation of host genes involved in et al. (2019). Single-cell characterization and quantification of translation-competent cellular functioning and defense during fibrosis: Identification of bio-markers. Genes viral reservoirs in treated and untreated HIV infection. PloS Pathogens. 15 (2), Dis. 6 (3), 304–317. doi: 10.1016/j.gendis.2019.04.007 e1007619. doi: 10.1371/journal.ppat.1007619 Jiang, A., Zhou, Y., Gong, W., Pan, X., Gan, X., Wu, Z., et al. (2022). CCNA2 as an Parreira, R., Piedade, J., Domingues, A., Lobão, D., Santos, M., Venenno, T., et al. (2006). immunological biomarker encompassing tumor microenvironment and therapeutic Genetic characterization of human immunodeficiency virus type 1 from beira, Mozambique. response in multiple cancer types. Oxid. Med. Cell. Longev. 2022, 5910575. doi: 10.1155/ Microbes Infect. 8 (9-10), 2442–2451. doi: 10.1016/j.micinf.2006.05.005 2022/5910575 Pastor, L., Parker, E., Carrillo, J., Urrea, V., Fuente-Soro, L., Respeito, D., et al. Jimenez, R., Martin-Cortazar, C., Kourani, O., Chiodo, Y., Cordoba, R., Dominguez- (2017a). A cytokine pattern that differentiates preseroconversion from Franjo, M. P., et al. (2018). CDCA7 is a critical mediator of lymphomagenesis that postseroconversion phases of primary HIV infection. Jaids-J Acq Imm Def. 74 (4), selectively regulates anchorage-independent growth. Haematologica 103 (10), 1669– 459–466. doi: 10.1097/QAI.0000000000001272 1678. doi: 10.3324/haematol.2018.188961 Pastor, L., Urrea, V., Carrillo, J., Parker, E., Fuente-Soro, L., Jairoce, C., et al. (2017b). Johnston, M. I., and Fauci, A. S. (2008). An HIV vaccine — challenges and prospects. Dynamics of CD4 and CD8 T-cell subsets and inflammatory biomarkers during early New Engl. J. Med. 359 (9), 888–890. doi: 10.1056/NEJMp0806162 and chronic HIV infection in Mozambican adults. Front. Immunol. 8, 1925. doi: 10.3389/fimmu.2017.01925 Judge, M., Parker, E., Naniche, D., and Le Souëf, P. (2020). Gene expression: The key to understanding HIV-1 infection? Microbiol. Mol. Biol. Rev. 84 (2), e00080–e00019. Phatak, A., Athar, M., Crowell, J. A., Leffel, D., Herbert, B.-S., Bale, A. E., et al. (2019). doi: 10.1128/MMBR.00080-19 Global gene expression of histologically normal primary skin cells from BCNS subjects reveals "single-hit" effects that are influenced by rapamycin. Oncotarget 10 (14), 1360– Kale, J., Osterlund, E. J., and Andrews, D. W. (2018). BCL-2 family proteins: 1387. doi: 10.18632/oncotarget.26640 Changing partners in the dance towards death. Cell Death Differ. 25 (1), 65–80. doi: 10.1038/cdd.2017.186 Pilcher, C. D., Tien, H. C., Eron, J. J.Jr., Vernazza, P. L., Leu, S. Y., Stewart, P. W., et al. (2004). Brief but efficient: Acute HIV infection and the sexual transmission of Kane, M., Zang, T. M., Rihn, S. J., Zhang, F., Kueck, T., Alim, M., et al. (2016). HIV. J. Infect. Dis. 189 (10), 1785–1792. doi: 10.1086/386333 Identification of interferon-stimulated genes with antiretroviral activity. Cell Host Microbe 20 (3), 392–405. doi: 10.1016/j.chom.2016.08.005 Popper, S. J., Gordon, A., Liu, M., Balmaseda, A., Harris, E., and Relman, D. A. (2012). Temporal dynamics of the transcriptional response to dengue virus infection in Nicaraguan Kazer, S. W., Aicher, T. P., Muema, D. M., Carroll, S. L., Ordovas-Montanes, J., children. PLoS Negl. Trop. Dis. 6 (12), e1966. doi: 10.1371/journal.pntd.0001966 Miao, V. N., et al. (2020). Integrated single-cell analysis of multicellular immune dynamics during hyperacute HIV-1 infection. Nat. Med. 26 (4), 511–518. doi: 10.1038/ Powers, K. A., Ghani, A. C., Miller, W. C., Hoffman, I. F., Pettifor, A. E., Kamanga, s41591-020-0799-2 G., et al. (2011). The role of acute and early HIV infection in the spread of HIV and implications for transmission prevention strategies in Lilongwe, Malawi: A modelling Konrad, B. P., Taylor, D., Conway, J. M., Ogilvie, G. S., and Coombs, D. (2017). On study. Lancet 378 (9787), 256–268. doi: 10.1016/S0140-6736(11)60842-8 the duration of the period between exposure to HIV and detectable infection. Epidemics 20, 73–83. doi: 10.1016/j.epidem.2017.03.002 Rotger, M., Dang, K. K., Fellay, J., Heinzen, E. L., Feng, S., Descombes, P., et al. (2010). Genome-wide mRNA expression correlates of viral control in CD4+ T-cells Kuo, H. H., and Lichterfeld, M. (2018). Recent progress in understanding HIV from HIV-1-infected individuals. PLoS Pathogens. 6 (2), e1000781. doi: 10.1371/ r e s e r v o i r s . C u rr . O p i n . H I V A I D S . 1 3 ( 2 ) , 1 3 7 – 1 4 2 . d o i : 1 0 .1 0 9 7 / journal.ppat.1000781 COH.0000000000000441 Sacoor, C., Nhacolo, A., Nhalungo, D., Aponte, J. J., Bassat, Q., Augusto, O., et al. Lavreys, L., Baeten, J. M., Chohan, V., McClelland, R. S., Hassan, W. M., Richardson, (2013). Profile: Manhiça health research centre (Manhiça HDSS). Int. J. Epidemiol. 42 B. A., et al. (2006). Higher set point plasma viral load and more-severe acute HIV type 1 (5), 1309–1318. doi: 10.1093/ije/dyt148 (HIV-1) illness predict mortality among high-risk HIV-1–infected African women. Clin. Infect. Dis. 42 (9), 1333–1339. doi: 10.1086/503258 Schinzel, A., Kaufmann, T., and Borner, C. (2004). Bcl-2 family members: Intracellular targeting, membrane-insertion, and changes in subcellular localization. Biochim. Biophys. Li, J., Parker, B., Martyn, C., Natarajan, C., and G, J. (2013). The PMP22 gene and its Acta (BBA) - Mol. Cell Res. 1644 (2), 95–105. doi: 10.1016/j.bbamcr.2003.09.006 related diseases. Mol. Neurobiol. 47 (2), 672–698. doi: 10.1007/s12035-012-8370-x Schoggins, J. W., MacDuff, D. A., Imanaka, N., Gainey, M. D., Shrestha, B., Eitson, J. Li, Q., Smith, A. J., Schacker, T. W., Carlis, J. V., Duan, L., Reilly, C. S., et al. (2009). L., et al. (2014). Pan-viral specificity of IFN-induced genes reveals new roles for cGAS Microarray analysis of lymphatic tissue reveals stage-specific, gene expression in innate immunity. Nature 505 (7485), 691–695. doi: 10.1038/nature12862 Frontiers in Cellular and Infection Microbiology 16 frontiersin.org Parker et al. 10.3389/fcimb.2023.1074847 Schoggins, J. W., and Rice, C. M. (2011). Interferon-stimulated genes and their Wang, Z., Arat, S., Magid-Slav, M., and Brown, J. R. (2018). Meta-analysis of human antiviral effector functions. Curr. Opin. Virol. 1 (6), 519–525. doi: 10.1016/ gene expression in response to mycobacterium tuberculosis infection reveals potential j.coviro.2011.10.008 therapeutic targets. BMC Syst. Biol. 12 (1), 3. doi: 10.1186/s12918-017-0524-z Shiau, S. (2017). Early treatment initiation and outcomes in perinatally HIV-infected Wang, H.-W., Wu, Y.-H., Hsieh, J.-Y., Liang, M.-L., Chao, M.-E., Liu, D.-J., et al. infants and young children in Johannesburg, south Africa: Age, aging, and antiretrovirals (2010). Pediatric primary central nervous system germ cell tumors of different (New York City, NY: Columbia University). prognosis groups show characteristic miRNome traits and chromosome copy Smith, A. J., Li, Q., Wietgrefe, S. W., Schacker, T. W., Reilly, C. S., and Haase, A. T. number variations. BMC Genomics 11, 132. doi: 10.1186/1471-2164-11-132 (2010). Host genes associated with HIV-1 replication in lymphatic tissue. J. Immunol. Wawer, M. J., Gray, R. H., Sewankambo, N. K., Serwadda, D., Li, X., Laeyendecker, 185 (9), 5417–5424. doi: 10.4049/jimmunol.1002197 O., et al. (2005). Rates of HIV-1 transmission per coital act, by stage of HIV-1 infection, Szklarczyk, D., Gable, A. L., Nastou, K. C., Lyon, D., Kirsch, R., Pyysalo, S., et al. in rakai, Uganda. J. Infect. Dis. 191 (9), 1403–1409. doi: 10.1086/429411 (2021). The STRING database in 2021: Customizable protein–protein networks, and Xie, A., Wang, P., Chen, D., and Z, H. (2022). Aberrant ARMCX1 expression is an functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res. independent predictor of poor prognosis in gastric cancer. J. Oncol. 2022, 9348917. doi: 49 (D1), D605–DD12. doi: 10.1093/nar/gkab835 10.1155/2022/9348917 Tanenbaum, M. E., Macůrek, L., Janssen, A., Geers, E. F., Alvarez-Ferná ndez, M., Yang, R., Du, Y., Wang, L., Chen, Z., and Liu, X. (2020). Weighted gene co- and Medema, R. H. (2009). Kif15 cooperates with Eg5 to promote bipolar spindle expression network analysis identifies CCNA2 as a treatment target of prostate cancer assembly. Curr. Biol. 19 (20), 1703–1711. doi: 10.1016/j.cub.2009.08.027 through inhibiting cell cycle. J. Cancer. 11 (5), 1203–1211. doi: 10.7150/jca.38173 The Joint United Nations Programme on HIV/AIDS (UNAIDS) (2020). UNAIDS Yang, E., and Li, M. M. H. (2020). All about the RNA: Interferon-stimulated genes that data 2020 Vol. 2020 (Geneva: UNAIDS). Report No.: JC2997E. interfere with viral RNA processes. Front. Immunol. 11. doi: 10.3389/fimmu.2020.605024 Toyono, T., Usui, T., Yokoo, S., Taketani, Y., Nakagawa, S., Kuroda, M., et al. (2015). Zhai, Y., Franco, L. M., Atmar, R. L., Quarles, J. M., Arden, N., Bucasas, K. L., et al. Angiopoietin-like 7 is an anti-angiogenic protein required to prevent vascularization of (2015). Host transcriptional response to influenza and other acute respiratory viral the cornea. PLoS One 10 (1), e0116838. doi: 10.1371/journal.pone.0116838 infections – a prospective cohort study. PloS Pathogens. 11 (6), e1004869. doi: 10.1371/ Trapnell, C., Roberts, A., Goff, L., Pertea, G., Kim, D., Kelley, D. R., et al. (2012). journal.ppat.1004869 Differential gene and transcript expression analysis of RNA-seq experiments with Zhang, Q., Yuan, J., Liu, Y., Liu, X., Lv, T., Zhou, K., et al. (2021). KIAA0101 knockdown TopHat and cufflinks. Nat. Protoc. 7 (3), 562–578. doi: 10.1038/nprot.2012.016 inhibits cell proliferation and induces cell cycle arrest and cell apoptosis in chronic Ullah, H., Sajid, M., Yan, K., Feng, J. P., He, M., Shereen, M. A., et al. (2021). lymphocytic leukemia cells. Ann. Trans. Med. 9 (6), 487. doi: 10.21037/atm-21-626 Antiviral activity of interferon alpha-inducible protein 27 against hepatitis b virus Zhang, Y., Zhang, H., An, M., Zhao, B., Ding, H., Zhang, Z., et al. (2018). Crosstalk in gene expression and replication. Front. Microbiol. 12. doi: 10.3389/ competing endogenous RNA networks reveals new circular RNAs involved in the fmicb.2021.656353 pathogenesis of early HIV infection. J. Trans. Med. 16 (1), 332. doi: 10.1186/s12967- Velasco, C., Parker, E., Pastor, L., Nhama, A., Macuacua, S., Mandomando, I., et al. 018-1706-1 (2015). Case report: Rapid HIV progression during acute HIV-1 subtype c infection in Zhao, S., Fung-Leung, W. P., Bittner, A., Ngo, K., and Liu, X. (2014). Comparison of a mozambican patient with atypical seroconversion. Am. J. Trop. Med. Hygiene. 92 (4), RNA-seq and microarray in transcriptome profiling of activated T cells. PLoS One 9 (1), 681–683. doi: 10.4269/ajtmh.14-0554 e78644. doi: 10.1371/journal.pone.0078644 Vidya Vijayan, K. K., Karthigeyan, K. P., Tripathi, S. P., and Hanna, L. E. (2017). Ziemann, M. Genome spot - data analysis step 8: Pathway analysis with GSEA. Pathophysiology of CD4+ T-cell depletion in HIV-1 and HIV-2 infections. Front. Available at: http://genomespot.blogspot.com/2014/09/data-analysis-step-8-pathway- Immunol. 8 (580). doi: 10.3389/fimmu.2017.00580 analysis.html. Frontiers in Cellular and Infection Microbiology 17 frontiersin.org