Abstract

   Chikungunya virus (CHIKV) infection often results in a chronic joint
   condition known as Post-Chikungunya Chronic Inflammatory Joint Disease
   (pCHIKV-CIJD). This condition disrupts individuals’ daily lives and
   contributes to increased healthcare expenditure. This study
   investigated the molecular mechanisms underlying pCHIKV-CIJD
   development by analyzing RNA transcripts, including small RNAs, of
   whole blood from CHIKV-infected patients. By comparing patients who
   evolved to pCHIKV-CIJD with those who did not, we identified molecular
   signatures associated with chronification in acute and post-acute
   disease phases. These molecules were primarily associated with an
   altered immune response regulation. Notably, LIFR, an immune receptor
   that enhanced IL-6 transcription, was down-regulated in the acute phase
   of pCHIKV-CIJD patients, while its inhibitor, hsa-miR-98-5p, was
   up-regulated in these individuals. Other downregulated genes include
   members of immune mechanisms whose impairment can lead to a reduction
   in the first line of antiviral response, thereby promoting virus
   persistence for a longer period in these patients. Additionally,
   pCHIKV-CIJD patients exhibited reduced transcript levels of MMP8, LFT,
   and DDIT4, genes already implicated in the pathological process of
   other types of inflammatory arthritis and seemingly relevant for
   pCHIKV-CIJD development. Overall, our findings provide insights into
   the early molecular mechanisms involved in the chronification and
   highlight potential targets for further investigation.

   Keywords: CHIKV, Chikungunya, Post-Chikungunya Chronic Inflammatory
   Joint Disease, Total RNA sequencing, Small RNA sequencing, Chikungunya
   prognosis

   Subject terms: Computational biology and bioinformatics, Systems
   biology, Diseases, Pathogenesis, Rheumatology

Introduction

   CHIKV, the Chikungunya virus, was first isolated in 1952 during an
   epidemic of febrile illness that resembled dengue^[56]1. Autochthonous
   transmission of CHIKV was later confirmed in Brazil in 2014, and the
   country experiences the highest number of Chikungunya fever cases in
   the Americas^[57]2. The disease is characterized by fever, joint pain,
   headache, myalgia, and skin rash. While most patients recover within a
   week, some experience complications, including chronic musculoskeletal
   symptoms that can persist for months or even years^[58]3.

   Chikungunya fever is clinically divided into acute, post-acute, and
   chronic phases. The acute phase, lasting up to 14 days, is marked by
   symptoms such as fever, polyarthralgia, rash, and retro-ocular pain.
   The post-acute phase, from 14 to 90 days, is characterized by the
   persistence or recurrence of acute phase symptoms, with joint
   involvement including post-acute hypertrophic tenosynovitis. The
   chronic phase extends beyond 90 days and is associated with
   long-lasting arthralgia, inflammatory signs, and varying levels of
   severity. This condition, known as Post-Chikungunya Chronic
   Inflammatory Joint Disease (pCHIKV-CIJD) is the main determinant of
   morbidity in affected individuals^[59]4,[60]5.

   In a recent study, we observed, through a longitudinal CHIKV cohort,
   that symptomatic post-acute disease is a relevant predictor of
   evolution to chronic arthritis with synovitis, drawing attention to
   joint pain, edema, and multiple articular involvements including small
   hand and feet joints as risk factors for chronification beyond three
   months, especially in women^[61]5.

   Transcritptomic analysis elucidated the immune related pathways
   disregulated during the CHIKV infection. In the CHIKV mouse model
   mostup-regulated genes are Type-I interferon response genes. In the
   same work the granzyme A was identified as an important driver of
   arthritic inflammation^[62]6. Corroborating this data, whole blood
   RNA-seq analysis of pediatric patients also found strong up-regulation
   of interferon-associated genes, such as IFIs, MX1 and MX2, OAS genes,
   and RSAD2. Major histocompatibility complex transcripts were also
   associated with the disease’s severity, where HLA-B transcripts were
   more expressed in patients with less severe acute-phase symptoms^[63]7.
   By comparing peripheral blood samples from CHIKV patients to those from
   healthy controls, a 2019 study identified transcriptional changes
   associated with CHIKV infection. Specific gene signatures, particularly
   those involving Eukaryotic Initiation Factor (eIF) family genes and
   APOBEC3A, were found to play significant roles in the viral replication
   process. Additionally, the study revealed that CHIKV infection
   activates the NLRP3 inflammasome in macrophages, leading to the
   production of IL-1 beta, a key component of the immune response^[64]8.

   Advancements in understanding CHIKV infection have primarily focused on
   the acute phase, with limited insights into the molecular mechanisms
   driving the transition to chronic arthralgia. To address this knowledge
   gap, we conducted an in-depth analysis using advanced sequencing
   techniques to elucidate the molecular mechanisms underlying the
   development of pCHIKV-CIJD. Our study involved the examination of RNA
   transcripts, including both small and long non-coding RNAs, from whole
   blood samples collected from patients in a Brazilian cohort. We
   compared samples from individuals who progressed to pCHIKV-CIJD with
   those who did not, over a 90-day follow-up period. Our results revealed
   a significant downregulation of LIFR, an immune receptor implicated in
   IL-6 transcription, alongside upregulation of its inhibitory microRNA
   hsa-miR-98-5p in patients who developed chronic disease.

   Additionally, we observed a reduction in the expression of genes
   crucial for immune responses, which may contribute to diminished
   antiviral activity and prolonged viral persistence. Notably, reduced
   transcript levels of MMP8, LTF, and DDIT4 were found in patients with
   pCHIKV-CIJD, highlighting their potential involvement in the
   pathogenesis of chronic inflammatory arthritis. These findings enhance
   our understanding of the molecular pathways involved in CHIKV
   chronification and suggest potential targets for therapeutic
   intervention.

Results and discussion

Total RNA analysis reveals distinct gene expression profiles early associated
to chronification

   By comparing patients who progressed to Post-Chikungunya Chronic
   Inflammatory Joint Disease (pCHIKV-CIJD) with those who did not, we
   comprehensively analyzed their transcriptomes in acute and post-acute
   phases. First, Principal Component Analysis (PCA) revealed that the
   main differences in the dataset are driven by the infection phase
   rather than group classification (pCHIKV-CIJD vs. non-pCHIKV-CIJD)
   (Figure [65]S1). PC2 (24.28% variance) clearly separates the samples
   between the acute and post-acute phases, with acute-phase samples
   clustering predominantly lower on the axis and post-acute-phase samples
   higher. In contrast, PC1 (39.51% variance) captures additional
   variability but does not distinctly separate the groups. When
   performing the differential expression analysis we observed differences
   in gene expression profiles between the two groups. In the acute
   disease phase, we identified 83 differentially expressed genes (DEGs;
   Log2FoldChange ≥ |0.5| and FDR < 0.1), with 35 genes up-regulated and
   48 genes down-regulated in patients who progressed to pCHIKV-CIJD
   (Fig. [66]1A, Table [67]S1a).

Fig. 1.

   [68]Fig. 1
   [69]Open in a new tab

   Differences in gene expression between patients who evolve or not to
   pCHIKV-CIJD in different phases of CHIKV infection. (A, B) Volcano Plot
   representing the differentially expressed genes (DEGs) found in the
   comparison between patients pCHIKV-CIJD and non-pCHIKV-CIJD in the
   acute phase (A) and post-acute phase (B) of infection. Points on the
   plot represent genes, with the x-axis showing the log2FoldChange and
   the y-axis showing the -log10 of the FDR-adjusted p-value. Genes
   significantly regulated are highlighted in red (up-regulated) or blue
   (down-regulated), while genes not significantly regulated are shown in
   gray. (C, D) Classification of the DEGs found in patients pCHIKV-CIJD
   compared to those non-pCHIKV-CIJD in the acute phase (C) and post-acute
   phase (D) of infection. The DEGs are categorized into different
   functional classes based on their known or predicted biological
   function. (E, F) Hierarchical clustering heatmap depicting the global
   expression profile of DEGs between samples from patients pCHIKV-CIJD
   and non-pCHIKV-CIJD in the acute (E) and post-acute (F) phases of
   infection. The rows represent genes, and the columns represent samples.
   The color shades indicate gene expression levels, with lighter shades
   indicating lower expression and darker shades indicating higher
   expression.

   In contrast, during the post-acute phase, the transcriptomic analysis
   revealed an increase in DEGs, with 458 genes showing differential
   expression. Among these, 427 genes were up-regulated and 31 were
   down-regulated (Fig. [70]1B, Table [71]S1a). This substantial increase
   suggests a sustained and possibly intensified immune response in
   patients progressing to pCHIKV-CIJD, which could contribute to the
   persistence and exacerbation of inflammation and joint symptoms
   characteristic of chronic disease. Additionally, in a previous study
   conducted by our group with the same cohort of patients, it was
   observed that those who progressed to pCHIKV-CIJD exhibited more
   symptoms during the subacute phase of the disease, particularly signs
   of arthritis and joint pain. This observation was considered a risk
   factor for chronification^[72]5. This clinical correlation underscores
   the importance of our current molecular findings, as the increased
   symptomatology in the subacute phase may reflect the underlying
   molecular disruptions we have identified.

   Since the acute phase, patients who evolved to pCHIKV-CIJD presented
   many differentially expressed long non-coding RNAs (lncRNAs),
   representing 20.5% (17/83) of the set of DEGs in this stage.
   Interestingly, this number reaches 65.1% (298/458) in the post-acute
   phase of the disease (Fig. [73]1C,D). LncRNAs are more than 200 bp and
   represent the majority of non-coding RNAs in humans with 127,802 lncRNA
   transcripts sequenced and annotated^[74]9 have been increasingly
   associated with the control of many cell functions^[75]10 by several
   different mechanisms, such as regulation of gene transcription^[76]11;
   altering the standard function of proteins^[77]12; function as
   competing endogenous RNA (ceRNA) being able to bind to multiple
   microRNAs, which regulate their targets expression^[78]13; and
   participating in the antiviral immune response, directly and indirectly
   controlling the production of anti-viral molecules in addition to
   pro-inflammatory cytokines^[79]14 and interferons^[80]15. The
   significant presence of lncRNAs among the DEGs found here, particularly
   in the post-acute phase, highlights their potential regulatory roles in
   the pathological processes leading to pCHIKV-CIJD.

   Despite the number of studies increasing in recent years, there are
   still few lncRNA with a well-described function and more studies are
   needed to evaluate its role in the response to CHIKV infection and its
   progression. In 2022, Basavappa et al. employed RNAseq to analyze gene
   expression in HBMECs infected with CHIKV and ZIKV, discovering that a
   substantial portion of differentially expressed genes were lncRNAs,
   many of which were virus-specific. These findings highlight the
   pivotal, virus-dependent role of lncRNAs in the antiviral response.
   Among these lncRNAs, the authors detected the overexpression of ALPHA
   in response to CHIKV infection. They showed that ALPHA overexpression
   decreases CHIKV replication by interacting with viral RNA and
   preventing its translation, thus reducing viral protein production.
   Additionally, ALPHA enhances the host’s immune response by promoting
   interferon-stimulated genes, which is crucial for antiviral defense
   against chikungunya virus^[81]16.

   Here, we did not find differences in ALPHA expression in the
   pCHIKV-CIJD group that could explain increased viral persistence in
   these patients in the acute and post-acute phases. However, the high
   number of lncRNAs in the post-acute phase suggests they play a crucial
   role in regulating of immune and inflammatory responses, potentially
   influencing the progression to chronic disease. Although the role these
   lncRNAs play in these clinical conditions remains to be elucidated, our
   interactome analyses showed that some differentially expressed lncRNAs
   interacted with genes found in the enrichment core of altered pathways
   in post-acute phase, as Neutrophil degranulation and Class I MHC
   mediated antigen processing and presentation (Table [82]S2).

   To further understand gene expression patterns, we performed
   hierarchical DEGs clustering in the acute (Fig. [83]1E) and post-acute
   (Fig. [84]1F) phases. This analysis demonstrated a clustering of
   patients according to their prognosis, with those progressing to
   pCHIKV-CIJD showing distinct gene expression profiles. This consistency
   in gene expression patterns within patient groups underscores the
   molecular differences between those who develop chronic disease and
   those who do not.

   Gene Set Enrichment Analysis (GSEA) is a computational method used to
   analyze gene expression data, focusing on differences in expression and
   statistical significance. This method identifies coordinated changes in
   predefined gene sets related to biological pathways or processes,
   providing insights into the functional significance of molecular
   changes in disease states. GSEA allows us to determine if a particular
   class of genes is over-represented in a condition, in our case,
   pCHIKV-CIJD.

   Figure [85]2 shows the GSEA results for expressed genes, highlighting
   differences in expression and statistical significance to identify
   enriched pathways associated with Post-Chikungunya Chronic Inflammatory
   Joint Disease (pCHIKV-CIJD) across acute and post-acute phases. The top
   ten pathways, selected based on statistical significance and negative
   Normalized Enrichment Scores (NES), indicate a predominant
   down-regulation of core genes in pCHIKV-CIJD patients compared to
   non-pCHIKV-CIJD individuals (detailed statistics are in Table [86]S3).

Fig. 2.

   [87]Fig. 2
   [88]Open in a new tab

   Top 10 REACTOME pathways enriched in pCHIKV-CIJD patients. Functional
   enrichment analysis of expressed genes revealed the top 10 REACTOME
   pathways enriched in patients with Post-Chikungunya Chronic
   Inflammatory Joint Disease (pCHIKV-CIJD). Each dot represents a
   pathway, with the size indicating the number of enriched genes. The
   color gradient reflects the Negative Normalized Enrichment Score (NES)
   value, with warmer colors representing greater enrichment significance.
   Since all enriched pathways have negative NES values, warmer colors
   indicate pathways with stronger down-regulation in pCHIKV-CIJD
   patients.

   Through GSEA analyses, we observed a reduction in the expression of
   several genes involved in the immune “Signaling by Interleukin”
   REACTOME pathway early in the acute phase of pCHIKV-CIJD patients.
   Among these, CCR2, the gene encoding for an immune receptor is involved
   in inflammatory responses and the recruitment and activation of immune
   cells. Alterations in the CCR2 signaling pathway have been implicated
   in pathological conditions, including rheumatoid arthritis^[89]17. In
   this context, a 2014 study challenged CCR2 knockout mice with CHIKV
   infection. The authors observed that CCR2 deficiency impaired the
   immune response, reducing pro-inflammatory cytokine production. This
   imbalance in cytokine production may contribute to the chronic and
   dysregulated immune response observed in chikungunya virus-induced
   arthritis^[90]18. Here, although not significantly, it is possible to
   observe a reduction in CCR2 expression (Log2FC = − 0.42; FDR = 0.19;
   Table [91]S4) in patients who progressed to pCHIKV-CIJD, corroborating
   data found by others in vivo.

   Another gene in the “Signaling by Interleukin” enrichment core, LIFR ,
   shows a significant reduction in expression in pCHIKV-CIJD patients
   compared to non-pCHIKV-CIJD patients (Log2FC = – 2.12; FDR = 0.08,
   Table [92]S1a). LIFR encodes the leukemia inhibitory factor (LIF)
   receptor, which, in complex with gp130, activates a tyrosine kinase
   signaling cascade in response to its ligand LIF. This pathway is key in
   coordinating immune responses with both pro- and anti-inflammatory
   functions^[93]19. This signal transduction pathway dysregulation is
   associated with many pathological conditions^[94]20.

   Interestingly, a study published in 2017 describes an autocrine
   regulatory network orchestrated by LIF/LIFR. Briefly, LIFR/gp130 forms
   a transmembrane complex with JAK1 and TYK2, which promotes the
   phosphorylation of the STAT4 transcription factor (pSTAT4) in the
   presence of the LIF. Then, pSTAT4 enhances the transcription of key
   inflammatory mediators, including IL-6, IL-8, IL-1α, IL-1β, and
   granulocyte-colony stimulating factor (G-CSF)^[95]21. In our study,
   non-pCHIKV-CIJD patients exhibited a higher LIFR transcription in the
   acute phase compared to pCHIKV-CIJD individuals. We hypothesize that
   this expression may result in higher LIFR availability, leading to
   increased LIF binding and signaling cascade activation. This, in turn,
   could enhance the transcription of inflammatory mediators, potentially
   contributing to antiviral immune responses in the early stages of the
   disease. This improved antiviral activity could help resolve the
   infection, reducing the risk of the disease becoming chronic in these
   patients. In contrast, the low amount of LIFR transcripts in the acute
   phase of infection, as observed in the pCHIKV-CIJD group, could result
   in the opposite effect, contributing to viral persistence that would
   increase the risk of chronicity.

   During the acute phase, additional alterations were observed in
   pathways related to Signal Transduction, Metabolism, Immune System, and
   Development Biology. Notable differentially expressed genes (DEGs)
   within these pathways included Albumin (ALB), Leukemia Inhibitory
   Factor Receptor (LIFR), Necdin (NDN), and Zinc Finger and BTB Domain
   Containing 16 (ZBTB16). Furthermore, a widespread reduction in the
   transcriptional levels of genes associated with antiviral immune
   responses, interleukins and their receptors, and genes involved in
   immune cell recruitment and adaptive responses were evident in
   pCHIKV-CIJD patients. These findings are integral to the understanding
   of the dysregulated immune responses characteristic of chronic CHIKV
   infection.

   In the post-acute phase, GSEA revealed enriched pathways associated
   with Signal Transduction, RNA Metabolism, Homeostasis, DNA Repair, and
   Cell Cycle. Noteworthy DEGs within these pathways included Anaplastic
   Lymphoma Receptor Tyrosine Kinase (ALK) and Protein C Cofactor (PROS1),
   which play crucial roles in DNA repair and Platelet Activation
   Signaling and Aggregation, respectively. These pathways shed light on
   the ongoing molecular processes contributing to the pathogenesis of
   pCHIKV-CIJD in the post-acute phase.

   In both the acute and post-acute phases, two immune system pathways,
   Neutrophil Degranulation and Class I MHC Mediated Antigen Processing
   and Presentation, exhibited significant down-regulation in pCHIKV-CIJD
   patients.

   Neutrophil degranulation involves the release of biological mediators
   stored in intracellular granules, including enzymes, cytokines, and
   antimicrobial molecules. Scientific studies have demonstrated the
   importance of this pathway in the response to viral infections.
   Opasawatchai et al. investigated the immune response during Dengue
   Virus infection and observed that neutrophil degranulation was critical
   for viral clearance and modulation of the inflammatory response^[96]22.
   Chikungunya patients also increased the neutrophil degranulation
   pathway in the acute phase of the disease^[97]23,[98]24. Considering
   its crucial role in viral elimination, reducing genes that participate
   in degranulation activity can decrease viral clearance and raise the
   risk of viral dissemination, enabling wider systemic viral
   dissemination. Furthermore, an impaired degranulation process may
   decrease the release of pro-inflammatory cytokines, which can impact
   the activation of other immune cells and the coordination of the immune
   response.

   Class I MHC-mediated antigen Processing and Presentation is critical to
   the host defense against viral infections, including CHIKV^[99]25.
   Here, pCHIKV-CIJD patients showed reduced levels of ZBTB16 transcripts
   (Log2FC = − 1.15; FDR = 0.06; Table [100]S1a), encoded for PLZF
   (promyelocytic leukemia zinc finger), in the acute phase. PLZF is a
   transcriptional factor that directly influences the expression of a
   specific set of genes induced by interferon cytokines. It can also
   influence the immune system’s ability to control viral replication and
   contain the spread of the infection^[101]26. Furthermore, there is
   evidence that ZBTB16 is necessary for the proper differentiation,
   survival and function of human T cells^[102]27. Therefore, a reduction
   in the ZBTB16 transcription could impact the ability of T cells to
   eliminate virus-infected cells.

   From the acute to the post-acute phases, pCHIKV-CIJD patients have
   shown a reduction in the expression levels of class I Human Leukocyte
   Antigen (HLA), such as HLA-B, HLA-C, HLA-E, and HLA-F, which are
   essential for effective immune responses against pathogens^[103]28.
   HLA, also known as MHC, proteins are expressed on the surface of cells,
   and their main function is to present virus peptide fragments to immune
   system cells like T CD8 + lymphocytes. This antigen presentation is
   essential for properly activating the immune system, allowing it to
   identify and eliminate infected cells. Interestingly, the suppression
   of HLA expression was recently described as an evasion mechanism in
   viral infections^[104]29.

   This set of results seems to indicate that, early in the acute phase,
   alteration of at least three key mechanisms in pCHIKV-CIJD patients may
   impair their immune response, possibly leading to an increase in viral
   persistence (Fig. [105]6), already described as a risk factor for
   chronification^[106]30.

Fig. 6.

   [107]Fig. 6
   [108]Open in a new tab

   Hypothetical Mechanism of LIFR Expression and Viral Persistence.
   Proposed diagram of the molecular alterations found still in the early
   stages of CHIKV infection in patients who progressed to
   post-chikungunya chronic inflammatory joint disease (pCHIKV-CIJD).
   Higher expression of LIF receptor (LIFR) may enhance LIF binding and
   activate downstream signaling pathways, leading to increased
   transcription of inflammatory mediators. This boost in antiviral immune
   responses could aid in resolving the infection and reducing the risk of
   chronic disease. Conversely, reduced LIFR transcript levels in the
   acute phase of Post-Chikungunya Chronic Inflammatory Joint Disease
   (pCHIKV-CIJD) could impair antiviral responses, contributing to viral
   persistence and a higher likelihood of chronicity. E2 estradiol
   hormone. Created in BioRender. Sabino, E. (2025)
   [109]https://BioRender.com/w73d110

   A detailed breakdown of all genes within the enrichment core of the top
   ten REACTOME pathways, alongside their corresponding statistical
   parameters, can be found in Table [110]S3.

MicroRNA expression analysis reveals dysregulated regulatory networks in
pCHIKV-CIJD patients

   In this study, alongside total RNA sequencing, we conducted small RNA
   sequencing using the Illumina platform to assess the expression of
   microRNAs (miRNAs) in patients with Post-Chikungunya Chronic
   Inflammatory Joint Disease (pCHIKV-CIJD). Integrating miRNA expression
   data with total RNA sequencing offers valuable insights into
   post-transcriptional regulatory mechanisms and enhances our
   understanding of gene expression dysregulation in disease states.

   The differential expression analysis of miRNAs revealed significant
   dysregulation in pCHIKV-CIJD patients compared to non-pCHIKV-CIJD
   individuals (Table [111]S1b). In the acute phase, we identified 55
   differentially expressed miRNAs (DEMs), include 21 up-regulated and 34
   down-regulated miRNAs (Fig. [112]3A). Similarly, 73 DEMs were
   identified in the post-acute phase, comprising 35 up-regulated and 38
   down-regulated miRNAs (Fig. [113]3B). These dysregulated miRNAs may
   play critical roles in modulating gene expression patterns associated
   with the pathogenesis of pCHIKV-CIJD. Additionally, a heatmap
   illustrating the hierarchical clustering of DEMs in both the acute
   (Fig. [114]3C) and post-acute phases (Fig. [115]3D) provides further
   insights into the dynamics of miRNA expression during different stages
   of the disease. Although not wholly segregating, the Hierarchical
   clustering heatmap revealed discernible expression profiles among
   patients with the same outcomes, suggesting potential associations
   between miRNA expression patterns and disease severity or progression.
   The complete list of identified mature miRNAs, along with their count
   information for each library, is provided in Table [116]S5.

Fig. 3.

   [117]Fig. 3
   [118]Open in a new tab

   Differential expression analysis of microRNAs (miRNAs) in pCHIKV-CIJD
   patients. (A) Volcano Plot representing the differentially expressed
   microRNAs (DEMs) found in patients with Post-Chikungunya Chronic
   Inflammatory Joint Disease (pCHIKV-CIJD) compared to non-pCHIKV-CIJD
   individuals in the acute phase of infection. (B) Volcano Plot
   representing the DEMs found in pCHIKV-CIJD patients compared to
   non-pCHIKV-CIJD individuals in the post-acute phase of infection. (C) A
   hierarchical clustering heatmap depicts the global expression profile
   of DEMs between pCHIKV-CIJD and non-pCHIKV-CIJD samples in the acute
   phase of infection. (D) A hierarchical clustering heatmap depicts the
   global expression profile of DEMs between pCHIKV-CIJD and
   non-pCHIKV-CIJD samples in the post-acute phase of infection.

   The dysregulation of miRNAs observed in pCHIKV-CIJD patients provides
   insights into potential regulatory networks involved in the disease
   process. We conducted an interactome analysis to elucidate these
   networks, considering both DEMs and differentially expressed genes
   (DEGs) in the acute and post-acute phases. All identified regulatory
   networks can be visualized in Fig. [119]4, providing a comprehensive
   overview of the interactions between dysregulated miRNAs and their
   target genes in pCHIKV-CIJD patients.

Fig. 4.

   [120]Fig. 4
   [121]Open in a new tab

   Regulatory networks involving differentially expressed microRNAs (DEMs)
   and genes (DEGs) in acute and post-acute phases. This interactome
   illustrates the regulatory networks involving DEMs and DEGs that are
   differentially expressed in both acute and post-acute phases of
   post-Chikungunya chronic Inflammatory Joint Disease (pCHIKV-CIJD).
   MicroRNAs are represented by ellipses, while target genes are depicted
   as rectangles. The color gradient ranging from blue to red represents
   the Log2 Fold Change value, indicating the extent of differential
   expression. Lines connecting microRNAs and targets represent
   interactions between them. These regulatory networks provide insights
   into the complex interplay between microRNAs and their target genes,
   highlighting potential regulatory mechanisms involved in the
   pathogenesis of pCHIKV-CIJD across different phases of the disease.

   Five potential regulatory networks between miRNAs and their validated
   target genes were identified in the acute phase of the disease.

   One notable interaction identified was between the up-regulated miRNA
   hsa-miR-98-5p and the down-regulated target gene LIFR. The low levels
   of LIFR transcripts observed in the acute phase of pCHIKV-CIJD
   patients, as discussed earlier, can be partly explained by the
   increased expression of hsa-miR-98-5p in these patients, which is an
   experimentally validated regulator of LIFR^[122]31. Interestingly,
   hsa-miR-98-5p was already described as an estradiol (E2)-responsive
   miRNA. In vitro experiments using breast cancer cell culture showed
   that E2 induces 21 microRNAs, including hsa-miR98-5p^[123]32. E2 is
   described as the main female sex hormone and steroid which when binding
   to its receptors (ER), promotes the regulation of a series of
   biological processes, including the activation of gene
   expression^[124]33,[125]34. Sex hormones may play a crucial role since
   they can potentially regulate many cellular processes, including the
   transcription of microRNAs, which in turn regulate the translation of
   thousands of targets^[126]35. Indeed, gender differences in response
   and prognosis are evident in several viral diseases. For instance,
   during the COVID-19 pandemic, men experienced higher mortality rates
   compared to women^[127]36.

   Studies have shown that women have at least a 50% increase in the risk
   of developing post-chikungunya chronic disease^[128]5. Of the 29
   pCHIKV-CIJD patients from whom we obtained transcriptome data, 25 were
   women. Notably, non-pCHIKV-CIJD patients were selected for sequencing
   based on sex and age matching. Therefore, the differences in the
   expression of hormonally induced microRNAs and other related genes may
   be even greater in the general population.

   Another interesting interaction between the up-regulated miRNAs
   hsa-miR-4775 and hsa-miR-25-3p with the down-regulated target genes
   DDIT4. Notably, hsa-miR-4775 has been experimentally validated to
   target the DDIT4 gene, which exhibits a 68% reduction in expression
   earlier in chronic patients (Log2FC = − 0.75; FDR = 0.06; Table
   [129]S1). DDIT4 encodes the REDD1 protein, known for its regulatory
   role in cellular stress response and as a canonical mTOR
   regulator^[130]37. Furthermore, REDD1 has been implicated in the
   regulation of autophagy through a mTOR-independent mechanism, leading
   to the induction of reactive oxygen species and activation of the ATG4B
   gene, a crucial component of the autophagy machinery^[131]38.

   Previous studies have demonstrated high expression levels of REDD1 in
   cartilaginous tissue cells in both humans and rats. However, it has
   been observed that REDD1 expression significantly declines with age and
   in animal models of osteoarthritis, suggesting a potential involvement
   in the pathogenesis of this condition. The decreased expression of
   REDD1 in cartilage may contribute to enhanced mTOR activity and
   diminished autophagy, which are characteristic features of
   osteoarthritis. Remarkably, lower expression levels of REDD1 have been
   associated with more severe inflammatory conditions in
   osteoarthritis^[132]39.

   Similarly, other up-regulated miRNAs, including hsa-miR-196a-5p and
   hsa-miR-7-5p, were found to interact with down-regulated target genes
   such as RTL8C and ZP3, respectively. In contrast, down-regulated miRNAs
   hsa-miR-27a-3p showed interactions with up-regulated target gene AP3B2,
   and this regulatory interaction was identified both in the acute and
   post-acute phases, highlighting the dynamic nature of gene regulatory
   processes during the course of pCHIKV-CIJD.

   Two targets were found for the down-regulated hsa-miR-484. The gene
   Growth Regulating Estrogen Receptor Binding 1 Like (GREB1L) is involved
   in immune cell infiltration in lung adenocarcinoma, and its methylation
   level is related to immune response and cytolysis in gastric
   adenocarcinoma^[133]40. The other target LMOD3 belongs, to the
   Leiomodin protein family, and is associated with Nemaline myopathy,
   characterized by muscle weakness, hypotonia, and the presence of
   nemaline bodies in skeletal muscle fibers^[134]41.

   The interactome analysis provides us critical insights into the
   regulatory networks influenced by miRNAs in pCHIKV-CIJD. By revealing
   complex interactions between dysregulated miRNAs and their target
   genes, this approach enhanced our understanding of the molecular
   mechanisms underlying disease progression and offers a novel avenue for
   future studies in the field.

Gene co-expression modules found in pCHIKV-CIJD patients

   Gene co-expression analysis is a powerful approach for identifying
   correlated gene expression patterns across different samples or
   conditions. It enables the identification of functional gene modules
   that are co-regulated and play similar biological roles in cellular or
   pathological processes. In this study, we analyzed gene co-expression
   using total RNA sequencing data to identify gene co-expression modules
   associated with Post-Chikungunya Chronic Inflammatory Joint Disease
   (pCHIKV-CIJD). The results of this analysis provided insights into the
   molecular mechanisms underlying disease pathogenesis, as well as
   identify potential biomarkers or therapeutic targets.

   The co-expression network analysis detected thirteen functional modules
   with similar gene expression, ranging from 42 to 1095 genes in each
   module. These modules were classified and named based on the number of
   genes they comprised (as shown in Figure [135]S2). We observed that
   modules M1, M2, and M10 exhibited higher gene expression, while modules
   M4, M6, M7, and M11 showed lower gene expression in the acute phase of
   the disease compared to the post-acute phase.

   Three modules (M3, M12, and M13) exhibited distinct behavior between
   pCHIKV-CIJD and non-pCHIKV-CIJD patient groups. Module M3 showed higher
   activity in pCHIKV-CIJD and lower activity in non-pCHIKV-CIJD patients,
   encompassing 886 genes, including 5 hubs (Fig. [136]5A). The top
   pathways of module M3 in the over-representation analysis involved in
   RNA metabolism, such as the processing of tRNAs and rRNAs in
   mitochondria (Fig. [137]5B). Conversely, modules M12 and M13 showed
   less expression in pCHIKV-CIJD and greater expression in
   non-pCHIKV-CIJD patients in both acute and post-acute phases of the
   disease. Module M12 included 46 hub genes (Fig. [138]5C) and was
   associated with the HDMs demethylate histones pathway (Fig. [139]5D).
   In comparison, module M13 comprised 42 genes, including 5 hubs
   (Fig. [140]5E), and was associated with Neutrophil Degranulation,
   Antimicrobial Peptides, Activation of Matrix Metalloproteinases, and
   Collagen Degradation pathways (Fig. [141]5F).

Fig. 5.

   [142]Fig. 5
   [143]Open in a new tab

   Co-expression modules found in pCHIKV-CIJD patients. The three
   co-expression modules exhibited distinct patterns in patients with
   Post-Chikungunya Chronic Inflammatory Joint Disease (pCHIKV-CIJD). Gene
   networks of modules displaying the most connected genes (hubs) of M3
   (A), M12 (C), and M13 (E) modules are illustrated. Each dot represents
   a gene, with the size and color reflecting the degree of interaction
   within the module. Larger dots and warmer colors indicate genes with
   higher connectivity within the module. Additionally,
   over-representation analysis was performed to identify pathways
   impacted by the genes within each module. The pathways affected by the
   genes of M3 (B), M12 (D), and M13 (F) modules are presented. In both
   analyses, the yellow color indicates a trend of greater expression in
   the genes of this module, while the purple color indicates a trend of
   lower expression in the genes of this module.

   The set of genes co-expressed in the M13 module (Fig. [144]5E) had
   reduced activity in pCHIKV-CIJD patients in both disease phases. These
   genes are enriched in the activation of matrix metalloproteinases and
   collagen degradation pathways, already associated with the development
   and/or higher severity of other types of arthritis, such as rheumatoid
   arthritis and osteoarthritis^[145]42. One of the M13 hubs is MMP8
   (Fig. [146]5E), encoding matrix metalloproteinase 8, which was
   described as playing a protective role in the progression of
   inflammatory arthritis in mice. MMP8 deficiency led to a reduction in
   caspase 11 expression, involved in programmed cell death, which can
   result in a delay in neutrophil apoptosis, subsequently causing
   neutrophil hyper infiltration in joints, contributing to the
   persistence of joint inflammation and an exacerbation in the severity
   of inflammatory arthritis^[147]42.

   Another M13 hub is LFT, which encodes lactoferrin, a protein present in
   many biological fluids capable of performing various biological
   functions such as immune response modulation, antimicrobial activity,
   and anti-inflammatory effect^[148]43. More than 20 years ago, a study
   highlighted the potential of local administration of lactoferrin in
   reducing articular inflammation in mice autoimmune and infectious
   arthritis models^[149]44. LFT was described as having an anti-apoptotic
   effect on IL-1β-induced chondrocytes and thus may be a promising novel
   therapeutic agent for osteoarthritis (OA). Chondrocyte apoptosis is
   mainly responsible for cartilage degeneration in OA^[150]45. Indeed, a
   recent study even observed that, in mouse models of rheumatoid
   arthritis, supplementation with bovine LFT considerably reduced
   inflammation and disease progression^[151]46. Considering the low
   expression of LFT found here since the early stages of the pCHIKV-CIJD,
   in addition to the large number of studies showing the broad
   physiological role played by lactoferrin, we can consider
   supplementation with lactoferrin as a possible treatment that can be
   applied even in acute phase of the disease aiming to mitigate the risks
   of disease progression, thus reducing the impacts on the quality of
   life of patients in the long term and, consequently, reducing expenses
   for the public health system. However, further studies are needed to
   evaluate this phenomenon in more patients.

   These results provide valuable insights into the gene co-expression
   patterns associated with pCHIKV-CIJD, highlighting potentially relevant
   biological pathways and cellular processes for disease pathogenesis.
   Additionally, identifying distinct gene modules between patients with
   and without pCHIKV-CIJD suggests the existence of specific molecular
   signatures associated with disease progression and severity.

Conclusion

   Responses to CHIKV infections involve complex mechanisms with intricate
   regulation of processes. To better understand these mechanisms, we
   aimed to investigate the factors involved in the progression of
   patients to post-Chikungunya Chronic Inflammatory Joint Disease
   (pCHIKV-CIJD), a condition that affects a considerable proportion
   (25–50%) of symptomatic CHIKV fever patients. To achieve this, we
   performed total and small RNA sequencing from whole blood samples
   collected during the acute and post-acute phases of the disease in
   patients who either progressed or did not progress to pCHIKV-CIJD.
   These patients were part of a longitudinal cohort study that we
   recently published^[152]5.

   Overall, our findings provide critical insights into the molecular
   mechanisms underlying the progression to pCHIKV-CIJD. The
   identification of specific DEGs, particularly the high prevalence of
   lncRNAs, opens new avenues for understanding the pathogenesis of this
   chronic condition. These results highlight potential molecular targets
   for future research and therapeutic intervention to mitigate the
   chronic inflammatory processes associated with pCHIKV-CIJD.

   In summary, here we show that patients who evolve to the chronic phase
   of the disease have a reduction in interleukin signaling in the initial
   phases of the disease compared to those who do not evolve. The
   impairment observed in this pathway seems to be explained at least in
   part by the increased expression of hsa-miR-98-5p, a miRNA activated by
   estradiol. It is important to note that estradiol is the main hormone
   produced by women, the population most affected by pCHIKV-CIJD.
   Furthermore, other antiviral immune response mechanisms in acute and
   post-acute phases seem to be impaired in these patients, such as
   neutrophil degranulation and antigen presentation via the MHCI
   pathways. Together, these mechanisms can lead to a reduction in the
   first line of antiviral response, promoting the persistence of the
   virus for a longer period in these patients (Fig. [153]6). This lack of
   immune response and other mechanisms such as reduced levels of matrix
   metalloproteinases (MMP8) and DDIT4, which appear to contribute to
   developing different types of arthritis, are important mechanisms for
   developing the post-Chikv chronic joint disease.

   The study has limitations, including a small sample size, which may
   affect the generalizability of the findings, and the lack of data
   validation in an independent cohort due to the unavailability of
   samples.

   Despite these limitations, we set out to generate and analyze small and
   total RNA data in two phases of infection in nearly 60 CHIKV-infected
   patients who were followed for 90 days and objectively diagnosed as to
   the outcome. This study provides valuable insights that can help to
   find a way to mitigate the progression of the disease to a more severe
   prognosis. In Brazil, a tropical country where the Aedes aegypti
   mosquito is endemic, chikungunya poses a significant public health
   risk, especially without an effective vaccine. The present study helps
   to elucidate the mechanisms involved in the evolution of patients to
   pCHIKV-CIJD. This disease significantly reduces the quality of life of
   affected individuals and financially impacts public health systems.
   Further studies must confirm these findings in a larger sample size and
   with different methodological approaches.

Methods

Sample selection

   A prospective longitudinal cohort at the Evandro Chagas National
   Institute of Infectology—INI/FIOCRUZ—Rio de Janeiro-RJ was established
   in 2019, including patients who suspected CHIKV. The inclusion and
   exclusion criteria and clinical and epidemiological aspects of the
   cohort were described in a previous study published recently by Lazari
   and coworkers^[154]5. Briefly, patients were included with up to seven
   days of fever and classic arbovirus symptoms. Two follow-up visits were
   made, 21 (D21) and 90 days (D90) after inclusion (D0), where
   collections of biological material and clinical information about the
   pain of joint symptoms were carried out. Those patients who, at D90,
   remained with joint symptoms and signs of arthritis were referred for
   ultrasound imaging examination. The 29 cases that showed image
   alteration were classified as post-CHIKV Chronic Inflammatory Joint
   Disease (pCHIKV-CIJD). From 47 patients who completed the follow-up
   scheme and did not present pCHIKV-CIJD, 25 were selected as a
   comparison group. This pairing was made considering, as criteria, that
   there were no significant differences regarding the age (pCHIKV-CIJD
   age median = 49.5/non-pCHIKV-CIJD age median = 43; p = 0.31) and sex
   (pCHIKV-CIJD = 25 female and 4 male/non-pCHIKV-CIJD = 22 female and 11
   male; p = 0.08) between the groups. Finally, samples from these
   patients were sequenced in the acute and post-acute phases, according
   to the availability and quality of clinical material. The number of
   samples from each disease phase by group that were sequenced is shown
   in Fig. [155]7 and the clinical information of the patients whose
   samples were sequenced is provided in Table [156]S6.

Fig. 7.

   [157]Fig. 7
   [158]Open in a new tab

   Representative flowchart of the methods used in this study. pCHIKV-CIJD
   post-Chikungunya Chronic Inflammatory Joint Disease, non-pCHIKV-CIJD
   non-post-Chikungunya Chronic Inflammatory Joint Disease, n number of
   samples. Acute phase: Samples collected at inclusion (D0); post-acute
   phase: Samples collected until D21. Created in BioRender. Sabino, E.
   (2025) [159]https://BioRender.com/t78w537

Total and small RNA sequencing

   For total RNA sequencing, the TruSeq Stranded Total RNA Library Prep
   kit with RiboZero Gold (Illumina, San Diego, California, USA) was used
   to prepare the libraries with an input of 500 ng. The RNA samples had a
   RIN (RNA Integrity Number) greater than 7 and/or a DV[200] value
   greater than 70%. The libraries were sequenced using the NextSeq
   500/550 High Output Kit v2.5 (150 Cycles) in a paired-end mode,
   generating reads of 2 × 75 bp. For small RNA sequencing, the TruSeq
   Small RNA Library Preparation Kit (including the miR-456 depletion
   step^[160]47) was used, and sequencing was performed with the NextSeq
   500/550 High Output Kit v2.5 (75 Cycles), generating single-end reads
   of 75 bp. A total of 98 libraries were made and 12 pooled samples were
   run simultaneously in each round on an Illumina NextSeq 550 (Illumina).

   The Fastq files from the total and small RNA sequencing are submitted
   on the Gene Expression Omnibus (GEO) platform under the accession
   number PRJNA1001790 and the metadata of the individuals and the sample
   characteristics can be found in Table [161]S6.

   A summary of the study workflow is shown in Fig. [162]7.

Total RNA sequencing analysis

   After sequencing, all the Fastq files were analyzed for the quality
   control of the reads using the FASTQC software^[163]48. The sequences
   were mapped to the human genome (GRCh38) using STAR software
   (v2.7)^[164]49. Gene count matrices were generated with the
   featureCounts function from the Rsubread package (v2.4.3)^[165]50 in
   the R programming language (v4.0.5). Differentially expressed genes
   were identified using the edgeR package (v3.32.1)^[166]51, through a
   comparison of pCHIKV-CIJD vs. non-pCHIKV-CIJD. A pathway enrichment
   analysis was performed using p-value and gene expression signal to rank
   the gene list and compare it with the Reactome pathways (v76)^[167]52
   information to identify pathways enriched in the expressed gene pool.
   This analysis was performed using the fGSEA package (v1.16.0)^[168]53.
   Finally, to identify genes with similar expression profiles between the
   sample classes (chronic early, chronic late, non-chronic early,
   non-chronic late), a coexpression analysis was performed using the
   CEMITool package (v1.14.1)^[169]54.

Small-RNA sequencing analysis

   The read quality control was accessed as described above. The sequences
   were mapped to the mature miRNA sequences from miRBase release
   22^[170]55 using bowtie2^[171]56, which was chosen for its
   compatibility with updated software tools and efficient handling of
   small RNA reads. Gene count matrices were generated with the
   featureCounts^[172]57 in the R programming language (v4.0.5).
   Differentially expressed microRNAs were identified using the edgeR
   package (v3.32.1)^[173]51. To evaluate the microRNA target genes, the
   miRTarBase database release 8.0^[174]58 was used, which considers only
   those targets that have been experimentally validated.

Interactome

   Only the validated target mRNAs with inverted log2FC patterns from
   their miRNAs were considered to build the miRNA-target networks. We
   used the interaction information of lncRNAs and protein-coding genes
   from the NPInter v5.0 database^[175]59 to build a network for the genes
   found in this study. Only interactions with at least log2FC > |0.5| and
   adjusted p-value < 0.05 for one of the nodes were considered to create
   the network. All the networks in this study were created using the
   Cytoscape software^[176]60.

   The pipeline and associated scripts used for this study have been
   uploaded to GitHub and are accessible at
   [177]https://github.com/ftencaten/pchikv_cijd_transcriptomics.

Supplementary Information

   [178]Supplementary Figure S1.^ (122.2KB, pdf)
   [179]Supplementary Figure S2.^ (120.9KB, pdf)
   [180]Supplementary Table S1.^ (51.5KB, xlsx)
   [181]Supplementary Table S2.^ (17.1KB, xlsx)
   [182]Supplementary Table S3.^ (56.4KB, xlsx)
   [183]Supplementary Table S4.^ (51.4KB, xlsx)
   [184]Supplementary Table S5.^ (22.7KB, tsv)
   [185]Supplementary Table S6.^ (22.7KB, tsv)

Acknowledgements