Cytomegalovirus infection drives a distinct mucosal and systemic cytokine signature in ulcerative colitis

Abstract

Purpose

Ulcerative colitis (UC) is a chronic inflammatory bowel disease in which cytomegalovirus (CMV) reactivation has been linked to severe and steroid-refractory cases. However, the effect of CMV on cytokine regulation in UC remains unclear. This study investigated colonic and systemic cytokine profiles in CMV-positive and CMV-negative UC patients.

Methods

A total of 190 UC patients were enrolled, including 90 CMV-positive and 100 CMV-negative cases. CMV status was confirmed by molecular and histopathological tests. Colonic gene expression of cytokines (TNF-α, IL-1β, IL-6, IL-17 A, IFN-γ, TGF-β, IL-8, IL-12, and RANTES) was measured by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR), and serum concentrations were assessed using enzyme-linked immunosorbent assay (ELISA).

Results

Demographic and clinical features were comparable between groups. CMV-positive patients showed higher colonic mRNA expression and serum concentrations of TNF-α, IL-1β, IL-6, and IL-17 A compared to CMV-negative patients (all p < 0.05). In contrast, IFN-γ, TGF-β, IL-8, IL-12, and RANTES did not differ significantly. Notably, CMV tissue viral load was significantly higher in patients with moderate to severe disease (Mayo score ≥ 6) and showed a positive correlation with plasma viral load (r = 0.664, p < 0.01). Multivariable logistic regression analysis adjusted for Mayo score confirmed that elevated TNF-α, IL-1β, and IL-6 remained independently associated with CMV positivity. These findings indicate a selective pro-inflammatory cytokine pattern rather than generalized immune activation.

Conclusion

CMV infection in UC is associated with selective upregulation of pro-inflammatory cytokines at both mucosal and systemic levels. These findings highlight CMV as a potential driver of immune amplification in UC and support cytokine profiling as a tool for risk stratification and management of affected patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12985-026-03200-7.

Introduction

Ulcerative colitis (UC) is a chronic, relapsing inflammatory bowel disease (IBD) affecting the colon and rectum. It is characterized by continuous mucosal inflammation, leading to diarrhea, rectal bleeding, abdominal pain, and fatigue. The disease significantly impairs quality of life and carries long-term risks, including colectomy in severe cases [1]. Despite the introduction of biologic and immunomodulatory therapies, many patients remain refractory to treatment [2]. The incidence and prevalence of UC have been increasing worldwide, particularly in industrializing countries, underscoring the combined roles of environmental and genetic factors in disease pathogenesis [3–5].

The etiology of UC is multifactorial, involving a dysregulated immune response to luminal antigens in genetically predisposed individuals. Infectious triggers have long been considered as potential modulators of disease activity. Among them, cytomegalovirus (CMV), a β-herpesvirus with global prevalence, has received considerable attention [6]. Following primary infection, CMV establishes lifelong latency. While usually asymptomatic in immunocompetent hosts, the virus can reactivate during immunosuppression or chronic intestinal inflammation [7]. In UC, CMV reactivation is most commonly detected in patients with severe or steroid-refractory disease [8]. Clinical studies suggest that CMV infection is associated with poor outcomes, such as resistance to corticosteroid therapy, higher colectomy rates, and increased hospitalization [6, 8, 9]. However, controversy persists regarding whether CMV acts as a genuine pathogen worsening UC or merely reflects the severity of mucosal inflammation. European consensus statements emphasize the need for careful diagnostic evaluation before initiating antiviral therapy [9].

Cytokine imbalance plays a central role in UC pathogenesis. Pro-inflammatory mediators, including tumor necrosis factor alpha (TNF-α), interleukin (IL)-1β, IL-6, and IL-17 A, drive chronic mucosal injury, while deficiencies in regulatory pathways such as IL-10 further contribute to immune dysregulation [10–13]. The clinical success of cytokine-targeting biologics underscores the importance of these mediators, although variability in patient responses highlights the role of additional modifying factors [12]. Viral infections may represent such modulators. CMV has been proposed to alter cytokine networks, amplifying pro-inflammatory pathways and potentially diminishing responsiveness to therapy [14, 15]. Indeed, several studies have detected higher CMV prevalence in inflamed tissue alongside elevated cytokine activity, suggesting a pathogenic link [16, 17]. Nonetheless, the evidence remains limited and sometimes inconsistent, reflecting small sample sizes and methodological differences.

A major gap in current knowledge concerns the relationship between mucosal cytokine gene expression and systemic cytokine levels in CMV-positive versus CMV-negative UC patients. Clarifying this relationship is clinically important, as cytokine profiling could help identify biomarkers of severity, stratify patient risk, and guide therapeutic decisions. We deliberately selected a defined cytokine panel (TNF-α, IL-1β, IL-6, IL-17 A, IFN-γ, TGF-β, IL-8, IL-12, and RANTES) rather than using an unbiased omics-based approach, because these mediators collectively represent key pro- and anti-inflammatory pathways implicated in UC pathogenesis. Our goal was to test a targeted hypothesis of selective cytokine dysregulation rather than perform broad exploratory screening. To the best of our knowledge, this is the first study to simultaneously evaluate both mucosal gene expression and serum cytokine profiles in UC patients with molecularly confirmed active CMV infection, providing a parallel view of local and systemic immune alterations. Therefore, the present study was undertaken to evaluate the influence of CMV infection on cytokine regulation in UC. By analyzing both colonic cytokine gene expression and serum cytokine concentrations, we sought to provide insight into the immunological changes associated with CMV and its potential role in disease exacerbation. Such findings could contribute to more informed clinical decisions regarding the management of UC patients with concomitant CMV infection.

Materials and methods

Patients

One hundred ninety patients with UC participated in this case-control study and who had a colon biopsy sample taken at the time of initial diagnosis of the disease. Colonic mucosal biopsies were obtained during colonoscopy from endoscopically inflamed areas. The case group consisted of 90 individuals with CMV infection, while the control group included 100 individuals without CMV infection. Patients with active CMV colitis were identified based on quantitative molecular detection of CMV DNA in colonic biopsy specimens. Active CMV colitis was defined as a tissue CMV DNA load > 250 copies/mg. A tissue CMV DNA load > 250 copies/mg was used as a predefined threshold to indicate clinically significant active CMV involvement, according to previously published evidence associating higher tissue viral burden with adverse clinical outcomes in UC [18, 19]. In the context of UC, CMV involvement is typically confined to the intestinal mucosa. Consistent with current consensus recommendations, the diagnosis of CMV colitis relies primarily on tissue-based evidence, including histology or quantitative tissue polymerase chain reaction (PCR), whereas systemic markers such as CMV DNAemia do not constitute mandatory diagnostic criteria [9]. Therefore, patients with tissue CMV DNA above the predefined cut-off were classified as having CMV colitis with high tissue viral burden. The participants were recruited from February 2022 to November 2024 by convenience sampling. Patient recruitment was carried out at the IBD Clinic of Imam Reza Hospital, the primary referral center for IBD management in East Azerbaijan, Northwest of Iran. The diagnosis of UC was established using clinical evaluation, endoscopic and/or radiological imaging, and confirmed by pathological findings. Disease activity was assessed using the Mayo scoring system, which comprises four subscores: rectal bleeding (0–3), stool frequency (0–3), physician’s global assessment (0–3), and endoscopic findings (0–3). The total Mayo score was calculated as the sum of these subscores. For the purpose of this study, a score ≤ 5 was defined as remission and mild active, while a score ≥ 6 was indicative of moderate to severe UC. Patients were excluded if they had other inflammatory bowel diseases, such as Crohn’s disease, or concurrent infectious colitis. Individuals with malignancy, pregnancy, or systemic immune disorders were also excluded. Additionally, participants with incomplete clinical, endoscopic, pathological data, or inadequate biopsy samples, were not included in the study. Biopsies were frozen in liquid nitrogen immediately after collection and stored at -80 °C until the analysis. In every instance, the information required for analysis was extracted from patients’ medical records. These data included demographic characteristics (e.g., age and sex), clinical features and laboratory findings relevant to the study variables.

Detection and quantification of CMV DNA in colonic tissue

After histological and immunohistochemistry examination, molecular detection was performed. CMV DNA was extracted from intestinal biopsies using QIAamp Viral DNA Mini Kit (Qiagen, Hilden, Germany, ) according to the manufacturer’s instruction. CMV DNA load in colonic tissue was expressed as copies per milligram of tissue (copies/mg). At the final step, DNA was eluted using 100 µL of elution buffer, stored in − 20 °C. A quantitative real-time PCR was applied using RealStar CMV PCR Kit 1.2 (Altona Diagnostics, Hamburg, Germany) according to the manufacturer’s recommendations. The lower detection limit of the assay was determined to be < 20 copies/mg tissues. Real-time PCR reactions were run in duplicate, and mean values were used for statistical analyses.

CMV DNAemia and IgM serological assay

Peripheral blood samples were collected in ethylenediaminetetraacetic acid (EDTA) tubes. Viral DNA was extracted from plasma using the QIAamp Viral DNA Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. CMV DNAemia was quantified using a quantitative real-time PCR assay (RealStar^® CMV PCR Kit 1.2, Altona Diagnostics, Hamburg, Germany) and expressed as copies per milliliter (copies/mL). CMV IgM serology was performed in all 190 patients using a commercial enzyme-linked immunosorbent assay (ELISA) kit (Pishtaz Teb Zaman, Iran), according to the manufacturer’s instructions. Diluted serum samples were incubated in CMV-antigen–coated wells, followed by washing and addition of horseradish peroxidase (HRP)-conjugated anti-human IgM. After substrate development, absorbance was read at 450 nm using a microplate reader. All reactions were performed in duplicate, and the mean value was used for analysis. Test results were interpreted based on the cutoff index provided by the manufacturer.

RNA extraction and quantitative real-time reverse-transcriptase PCR (qRT-PCR)

Total RNA was extracted using the TRIZOL reagent (Ambion, USA) according to the protocol provided by the manufacturer. Any potential DNA contamination was removed by treating the extracted RNA with RNase-free DNase I (Qiagen, Germany). The extracted RNAs’ purity was assessed using spectrophotometric method (Boeco, Germany) with the calculation of 260/280 ratio. RNA samples with a calculated A 260/280 absorbance value ranging from 1.8 to 2.0 were selected to perform the reverse transcription reaction. RNA samples were reverse-transcribed into complementary DNA (cDNA) using QuantiTect Reverse Transcription Kit (Qiagen, Germany) according to the manufacturer’s guidelines. Generated cDNA products were stored at -70 °C for long-term storage.

QRT-PCR was carried out on the Corbett Rotor-Gene 6000 instrument, and all reactions were performed with 2X AddScript RT-PCR SYBR Master (AddBio, Korea) using specific gene primers (Metabion, Germany). The nucleotide sequences of the gene-specific primers are presented in Supplementary Table S1. Primer efficiency and correlation coefficients (R²) were determined from prior validation experiments conducted under the same qPCR conditions. Efficiency values ranged between 93% and 106% with R² > 0.99 (Supplementary Table S2). Laboratory analyses were performed without formal blinding to patients’ CMV status. No-template controls (NTCs) and no–reverse transcription controls (no-RT) were included in each run to exclude contamination. Amplification specificity was verified for reactions by melt curve analysis, which consistently showed a single distinct peak for each primer pair. Representative melt curves are provided in Supplementary Fig. S1. PCR program was as follows: 1 cycle of 95 °C for 2 min; 40 cycles of 95 °C for 5 s (Denaturation), and 60–65 °C (annealing/Extension) for 20 to 30 s. To normalize cytokine expression levels, the GAPDH gene was utilized as a reference for each sample. Differences in gene expression were analyzed using the threshold cycle (2^−ΔΔCT) method [20], normalized to the mean ΔCt of the CMV-negative group used as reference. Relative values of the gene of interest were represented as fold change to compare mRNA levels between groups. All qRT-PCR assays were performed in duplicate.

Cytokine assay

Blood samples were collected from all 190 studied patients. Serum was separated and cytokine concentrations were measured using ELISA. The serum levels of TNF-α, IL1β, IL-6 and IL-17 A were evaluated using a commercial kit from the Karmania Pars Gene Company, Kerman, Iran, according to the manufacturer’s instructions. The lower detection limit of all cytokines was 2 pg/mL. Optical density (OD) was detected using an ELISA reader (DANA-3200, Iran) at 450 nm. Each experiment was performed in duplicate, and mean values were used for statistical analysis to minimize technical variation.

Statistical analysis

Statistical analyses were performed using SPSS v.25.0 (USA) and GraphPad Prism v.10.3. Continuous variables were expressed as mean ± standard deviation (SD) or median with interquartile range (IQR), as appropriate, and categorical variables were presented as absolute numbers and percentages. The Shapiro–Wilk test was employed to evaluate variables’ normality of distribution. Student’s t-test and Mann–Whitney U test were used for comparisons between two groups. In addition, Fisher’s exact test and the chi-squared test were applied to assess the associations between categorical variables. Moreover, the area under the receiver operating characteristic (ROC) curves was calculated using statistical analysis to evaluate the diagnostic accuracy of the cytokines. Correlation analyses were performed using Spearman’s rank correlation coefficient. Multivariable logistic regression was used to assess the independent association between CMV status and cytokine levels after adjustment for Mayo score. Cytokine concentrations were log10-transformed prior to analysis, and adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. A p-value less than 0.05 considered statistically significant.

Results

Demographic and clinical characteristics of study participants

A total of 190 patients with UC were enrolled in this case-control study, including 90 patients with CMV-DNA positive status and 100 patients without CMV infection. The mean age of the CMV-positive group was 45.2 ± 17.8 years, compared to 43.3 ± 18.6 years in the CMV-negative group, with no statistically significant difference (p = 0.475). Regarding sex distribution, 43.3% of the CMV-positive group were male and 56.7% female, while in the CMV-negative group, 48% were male and 52% female. No significant association was observed between sex and CMV status (p = 0.561).

There were also no significant differences between groups in terms of C-Reactive Protein (CRP), albumin, erythrocyte sedimentation rate (ESR), and hemoglobin levels (all p > 0.05). Likewise, drug history did not differ significantly between CMV-positive and CMV-negative patients (p = 0.609). The median Mayo score did not differ significantly between the CMV-positive and CMV-negative groups (7 [IQR 5–9] vs. 6 [IQR 4–8], p = 0.410). The difference was not statistically significant (p = 0.410) (Table 1). However, CMV tissue viral load (copies/mg) was significantly higher in patients with Mayo score ≥ 6 compared to those with score ≤ 5 (median 8424 vs. 1960 copies/mg, p < 0.01). Additionally, a significant positive correlation was observed between tissue CMV viral load (copies/mg) and plasma CMV DNA load (copies/mL) (r = 0.664, p < 0.01). Patients with detectable CMV DNA in plasma (copies/mL) exhibited significantly higher tissue viral loads (copies/mg) compared to those without detectable plasma CMV DNA (p < 0.01).

Table 1.

Demographic and clinical characteristics of study participants

Variables	CMV-DNA positive (n = 90)	CMV-DNA negative (n = 100)	p-value
Age, meanSD	45.217.8	43.318.6	0.475
Sex, n (%)
Male  Female	39 (43.3) 51 (56.7)	48 (48) 52 (52)	0.561
CRP (mg/dL)	2.813.92	3.062.84	0.613
Albumin (mg/dL)	3.640.79	3.760.84	0.314
ESR (mm/h)	33.0925.14	32.2425.03	0.816
Hemoglobin (g/dL)	12.961.4	13.11.5	0.509
Drug history, n (%)
5-ASA  5-ASA + IM  Glucocorticoids  No drug	22 (24.5) 11 (12.2) 19 (21.1) 38 (42.2)	20 (20) 10 (10) 18 (18) 52 (52)	0.609
Mayo score, median (IQR)	7 (5–9)	6 (IQR 4–8)
Mayo score ≤ 5 Mayo score ≥ 6	40 (44.5) 50 (55.5)	50 (50) 50 (50)	0.410
Tissue CMV viral load (copies/mg), median [range]	4232 [766-131652] score < 5: 1960 [766–8451] score > 6: 8424 [2119-131652]	-	-
CMV DNAemia (copies/mL), median [range]	3729 [433-39684]
Positive Negative	39 (43.3) 51 (56.7)	-	-
CMV IgM, n (%)
Positive Negative	22 (24.4) 68 (75.6)	0 (0.0) 100 (100)	-

CRP: C-Reactive Protein; ESR: Erythrocyte sedimentation rate; 5-ASA: 5-aminosalicylic acid; IM: Immunomodulator; UC: Ulcerative colitis; CMV: Cytomegalovirus; SD: Standard deviation; IQR: Interquartile range; mg/dL: Milligrams per deciliter; mm/h: Millimeters per hour; Copies/mg: Copies per milligram; Copies/mL: Copies per milliliter

A total of 90 patients with tissue-proven CMV colitis were included in the study. Blood samples were analyzed for CMV DNA using quantitative real-time PCR, and CMV DNAemia was identified in 39 patients (43.3%). The quantitative measurement of CMV DNA in blood-positive cases revealed a mean viral load of 1274.64 copies/mL (SD = 579.16). Serologic testing demonstrated that serum CMV IgM was positive in 22 patients (24.5%).

Gene expression analysis

We next evaluated the expression of inflammatory cytokine genes in colonic biopsy samples. The expression of TNF-α, IL-1β, IL-6, and IL-17 A was significantly elevated in CMV-positive patients compared to CMV-negative controls (p = 0.019, 0.037, 0.029, and 0.033, respectively). In contrast, IFN-γ, RANTES, IL-8, and IL-12 showed slightly higher mean expression levels in CMV-positive patients compared with CMV-negative controls, whereas TGF-β demonstrated modestly lower values. However, none of these differences reached statistical significance (all p > 0.05). These findings suggest that CMV infection is associated with selective upregulation of pro-inflammatory cytokines in UC (Table 2; Fig. 1).

Table 2.

Relative expression of cytokines according to the CMV status

Gene	CMV-DNA positive (n = 90) Mean ± SD	CMV-DNA negative (n = 100) Mean ± SD	P-Value
TNF-α	0.092 ± 0.018	0.053 ± 0.017	0.009
IFN-γ	0.058 ± 0.016	0.050 ± 0.014	0.482
TGF-β	0.030 ± 0.010	0.034 ± 0.012	0.613
RANTES	0.115 ± 0.024	0.110 ± 0.023	0.537
IL-1β	0.141 ± 0.029	0.084 ± 0.024	0.037
IL-6	0.051 ± 0.015	0.032 ± 0.012	0.029
IL-8	0.265 ± 0.051	0.248 ± 0.049	0.414
IL-12	0.098 ± 0.021	0.092 ± 0.020	0.003
IL-17 A	0.086 ± 0.014	0.043 ± 0.011	0.033

TNF-α: Tumor necrosis factor alpha; IFN-γ: Interferon gamma; TGF-β: Transforming growth factor beta; RANTES: Regulated on activation, normal T-cell expressed and secreted (CCL-5); IL: Interleukin; GAPDH: Glyceraldehyde 3-phosphate dehydrogenase

Fig. 1.

Relative expression of TNF-α, IFN-γ, TGF-β, RANTES, IL-1β, IL-6, IL-8, IL-12 and IL-17 A in control group (CMV negative) and case group (CMV positive). *p < 0.05, ** p ˂ 0.01

Serum levels of cytokines

In addition to mRNA expression, serum cytokine concentrations were analyzed. Consistent with tissue gene expression results, serum levels of TNF-α, IL-1β, IL-6, and IL-17 A were significantly higher in CMV-positive UC patients compared to CMV-negative controls (p = 0.003, 0.007, 0.004, and 0.012, respectively). These results further support the role of CMV infection in amplifying systemic inflammatory responses in UC patients (Fig. 2). ROC curve analysis was performed to evaluate the diagnostic performance of serum inflammatory cytokines in distinguishing CMV-DNA–positive from CMV-DNA–negative UC patients. As shown in Fig. 3, the areas under the curve (AUC) were 0.91 (95% CI 0.85–0.97, p < 0.001) for TNF-α, 0.82 (95% CI 0.74–0.89, p < 0.001) for IL-1β, 0.79 (95% CI 0.70–0.86, p = 0.002) for IL-6, and 0.61 (95% CI 0.50–0.71, p = 0.14) for IL-17 A. Although IL-17 A showed statistical significance in group comparisons, its discriminative performance was limited, as reflected by a modest AUC value.

Fig. 2.

TNF-α, IL-1β, IL-6 and IL-17 A serum levels in control group (CMV negative) and case group (CMV positive). Serum levels of all cytokines significantly increased in the CMV positive UC patients in comparison to CMV negative UC patients, *P < 0.05, ** P˂ 0.01

Fig. 3.

Receiver operating characteristic (ROC) curve analysis of serum TNF-α, IL-1β, IL-6, and IL-17A for discriminating CMV-DNA–positive from CMV-DNA–negative ulcerative colitis patients. TNF-α showed the highest diagnostic performance (AUC = 0.91), followed by IL-1β (AUC = 0.82) and IL-6 (AUC = 0.79), whereas IL-17A demonstrated limited discriminative ability (AUC = 0.61)

Correlation of mucosal cytokine expression levels with tissue CMV viral load and systemic CMV DNAemia

To further evaluate the relationship between viral burden, inflammatory mediators, and clinical severity, Spearman correlation analyses were performed between cytokine levels, CMV viral load parameters, and Mayo score. Correlation analyses were primarily performed using mucosal cytokine expression levels, given the tissue-based definition of active CMV colitis. As shown in Table 3, tissue CMV viral load demonstrated significant positive correlations with TNF-α, IL-1β and IL-6 expression, while plasma viral load showed moderate associations primarily with TNF-α and IL-1β. In contrast, only TNF-α displayed a weak but statistically significant correlation with Mayo score. Other cytokines did not show significant associations with disease activity. These findings support a selective inflammatory amplification pattern rather than a global cytokine activation.

Table 3.

Spearman correlations between mucosal cytokine expression levels, tissue CMV viral load (copies/mg), plasma CMV DNA load (copies/mL), and Mayo score in UC patients

Gene*	Tissue CMV load (r, p) (n = 90)	Plasma CMV load (r, p) (n = 39)	Mayo score (r, p) (n = 190)
TNF-α	0.38, < 0.001	0.34, 0.028	0.16, 0.034
IL-1β	0.32, 0.002	0.30, 0.041	0.11, 0.12
IL-6	0.26, 0.011	0.22, 0.11	0.10, 0.15
IL-17 A	0.14, 0.18	0.09, 0.52	0.05, 0.48
IFN-γ	0.06, 0.57	0.07, 0.63	0.04, 0.62
TGF-β	-0.05, 0.63	-0.06, 0.58	-0.03, 0.71
IL-8	0.08, 0.44	0.10, 0.49	0.06, 0.39
IL-12	0.04, 0.71	0.05, 0.68	0.05, 0.47
RANTES	0.07, 0.49	0.09, 0.55	0.02, 0.81

^* Spearman correlation coefficients (r) are shown. Cytokine values represent relative mucosal gene expression (2−ΔΔCT fold change normalized to GAPDH). Tissue CMV viral load is expressed as copies/mg tissue and plasma CMV load as copies/mL plasma. Tissue viral load correlations were performed in CMV tissue-positive patients (n= 90). Plasma viral load correlations were performed in DNAemia-positive patients (n= 39). Mayo score ranged from 0 to 12

Among the 90 patients with tissue-confirmed CMV colitis, 39 (43.3%) had detectable CMV DNAemia, whereas 51 (56.7%) were plasma-negative. In stratified analyses, DNAemia-positive patients demonstrated significantly higher mucosal TNF-α expression compared with tissue-only positive cases (median 0.106 [IQR 0.088–0.129] vs. 0.082 [0.069–0.098], p = 0.002). A similar pattern was observed at the systemic level, where serum TNF-α concentrations were elevated in DNAemia-positive individuals (36.2 pg/mL [28.4–44.8] vs. 27.6 pg/mL [20.9–35.2], p = 0.001). IL-1β showed a modest increase in DNAemia-positive patients at both mucosal and serum levels; however, these differences did not reach conventional statistical significance (p = 0.051 and p = 0.057, respectively). No significant differences were observed for IL-6 or IL-17 A between subgroups (all p > 0.05).

Multivariable analysis adjusted for disease activity

To evaluate whether the association between CMV status and cytokine levels was independent of clinical disease activity, multivariable logistic regression analyses were performed with CMV status (positive vs. negative) as the dependent variable and Mayo score included as a covariate. Given their right-skewed distribution, cytokine concentrations were log10-transformed prior to analysis. After adjustment for Mayo score, higher serum TNF-α remained significantly associated with CMV positivity (adjusted OR = 2.75 per 1 log10 increase; 95% CI 1.60–4.85; p = 0.0003). IL-1β (adjusted OR = 2.10; 95% CI 1.25–3.55; p = 0.005) and IL-6 (adjusted OR = 1.75; 95% CI 1.10–2.85; p = 0.021) also retained statistically significant associations. In contrast, IL-17 A was not independently associated with CMV status following adjustment (adjusted OR = 1.30; 95% CI 0.85–2.05; p = 0.22). Mayo score was not significantly associated with CMV positivity in these models (OR = 1.05 per one-point increase; 95% CI 0.97–1.14; p = 0.24). Importantly, inclusion of Mayo score in the models did not materially alter the magnitude or direction of the associations between CMV status and the significantly elevated cytokines. Exploratory adjusted models including IFN-γ, TGF-β, IL-8, IL-12, and RANTES did not demonstrate statistically significant independent associations with CMV status.

Discussion

The present study provides evidence that CMV infection in patients with UC is associated with a distinct and selective upregulation of key pro-inflammatory cytokines, both at the mucosal level and systemically. Our findings demonstrate that CMV-positive UC patients exhibit significantly elevated colonic mRNA expression and corresponding serum concentrations of TNF-α, IL-1β, IL-6, and IL-17 A compared to their CMV-negative counterparts. Protein analysis was limited to these four cytokines based on their significant transcriptional changes and established relevance in UC pathogenesis, as well as practical considerations related to sample availability and assay feasibility. In contrast, the expression of other cytokines, including IFN-γ, TGF-β, IL-8, IL-12, and RANTES, remained unchanged. This pattern suggests that CMV reactivation does not induce a global, non-specific inflammatory cascade but rather orchestrates a specific immunomodulatory shift toward a selectively Th17-oriented and innate-amplified inflammatory pattern. The preferential upregulation of specific pro-inflammatory cytokines is consistent with virus-mediated immune modulation [21]. It is important to acknowledge that previously published studies evaluating CMV in ulcerative colitis have employed heterogeneous diagnostic methodologies, including immunohistochemistry, qualitative or quantitative PCR, viral culture, and antigenemia assays [7, 9]. Differences in assay sensitivity and viral load thresholds complicate direct comparisons across studies and may partly explain inconsistencies in reported prevalence and clinical impact. European consensus statements have further emphasized the need for standardized diagnostic criteria before attributing clinical significance to CMV detection [9].

Among upstream inflammatory pathways, nuclear factor kappa B (NF-κB) functions as a central regulator of pro-inflammatory cytokine transcription in chronic inflammatory disorders, including IBD [22]. CMV infection has been shown to activate the IκB kinase (IKK) complex and promote IκB degradation, resulting in sustained NF-κB activation. This activation drives the transcription of pro-inflammatory cytokines such as TNF-α, IL-6, and IL-1β [23]. Moreover, NF-κB directly transactivates the CMV major immediate-early promoter/enhancer (MIEP), thereby regulating viral immediate-early (IE) gene expression [23, 24]. This reciprocal interaction establishes a positive-feedback loop that sustains both viral replication and inflammatory signaling. Experimental data further indicate that CMV may activate the MyD88/NF-κB axis, reinforcing this amplification circuit in intestinal inflammation models [25]. Such a self-sustaining inflammatory circuit provides a biologically plausible explanation for the persistent and selective pro-inflammatory cytokine upregulation observed in CMV-positive UC patients and supports the concept that CMV may be functionally involved in inflammatory amplification in UC.

Because a tissue CMV DNA load > 250 copies/mg has been associated with clinically significant active CMV colitis in previous studies [18, 26], we assessed systemic markers (CMV DNAemia and CMV-IgM) alongside colonic CMV DNA detection to complement tissue-based evidence and evaluate the systemic footprint of viral activity. CMV DNAemia was detected in 43.3% of CMV-positive patients and CMV-IgM was positive in 24.4% of cases. Although systemic CMV activation was present in a subset of patients, quantitative tissue viral load confirmed active colitis in all CMV-positive cases. The discordance between tissue detection and systemic markers likely reflects compartmentalized mucosal reactivation rather than systemic infection, a phenomenon well described in IBD-associated CMV colitis [9]. Notably, in exploratory subgroup analyses, TNF-α levels were further elevated in patients with concomitant CMV DNAemia compared with tissue-only positive cases, suggesting that systemic viral dissemination may intensify TNF-driven inflammatory amplification. Importantly, in those patients who showed systemic evidence of CMV, the coexistence of tissue viral load and systemic markers-together with the selective elevation of TNF-α, IL-1β, IL-6 and IL-17 A-suggests that CMV may contribute to the amplification of inflammatory signaling in this setting. Previous studies have also associated systemic CMV markers (DNAemia, IgM or antigenemia) with more severe or steroid-refractory UC [18, 27].

The CMV immediate-early genes can directly activate the NF-κB pathway, a main regulator of TNF-α and IL-6 transcription [28, 29]. Beyond NF-κB activation, CMV is thought to influence cytokine signaling through additional pathways such as signal transducer and activator of transcription 3 (STAT3) and mitogen-activated protein kinases (MAPK), both of which are known to regulate IL-6 and IL-17 A expression. Such cross-talk could amplify the Th17 response and sustain mucosal inflammation, providing a plausible mechanistic basis for the selective cytokine upregulation observed in our study [30, 31]. In IBD, the significant elevation of TNF-α provides direct human molecular evidence for the pathway proposed by Pillet et al., wherein CMV infection potentiates TNF-α-driven pathology, thereby creating a self-perpetuating inflammatory loop that may underpin the steroid-refractory phenotype commonly seen in these patients [28]. The elevated TNF-α appears to be followed by a similarly elevated IL-6 level, consistent with previous observations linking IL-6 to CMV-associated exacerbation of chronic T-cell–mediated intestinal inflammation [29]. Notably, the significant increase in IL-1β we observed aligns with reports highlighting the role of the IL-1 family in neutrophilic infiltration and steroid-refractory inflammatory responses in IBD [32, 33].

The significant increase in IL-17 A, a cytokine pivotal for neutrophil recruitment and chronic inflammation, further underscores a potential mechanism for worsened disease severity [34, 35]. Recent studies have highlighted the role of viral infections in shaping the Th17 axis [36, 37], and our results are consistent with the hypothesis that CMV may amplify IL-17-mediated pathology in the inflamed colon. A study by Dhital et al.., strongly supports that CMV disrupts mucosal homeostasis to foster a pro-inflammatory Th17 response [38]. This collective evidence draws a picture of a coordinated Th1/Th17 response, a notion further refined by the insightful contrast provided by our IFN-γ data. No significant change in the level of certain cytokines reflects CMV’s sophisticated immune evasion strategies. The unchanged IFN-γ and IL-12 levels are particularly notable, as these cytokines orchestrate potent antiviral responses through Th1 differentiation.

The stability of TGF-β, IL-8 and RANTES levels in our study adds critical details, implying that CMV’s effect is targeted. The unchanged TGF-β parallels the findings of Lawlor et al.., indicating that this key regulatory pathway may remain intact, while the consistent IL-8 levels, as hinted in the study by Roblin et al.., suggest that robust neutrophil recruitment may be a general consequence of active UC, independent of CMV status [7, 18]. Crucially, the parallel elevation of TNF-α, IL-6, IL-1β and IL-17 A in the serum underscores that this is not a confined mucosal event but a systemic inflammatory spillover, a phenomenon verified by Heald-Sargent et al.., who linked systemic inflammation to mucosal viral activity in IBD [39]. Elevated TNF-α and IL-6 levels are well-recognized features of active ulcerative colitis and reflect baseline inflammatory activity inherent to the disease [10, 11]. The additional increase observed in our CMV-positive cohort therefore supports the concept of viral-associated inflammatory amplification beyond underlying UC-related immune activation. This systemic dimension provides a plausible molecular explanation for the stark clinical outcomes documented by Qin et al.., whose meta-analysis associated CMV positivity with significantly higher colectomy rates [40]. Thus, the cytokine tetrad we identified (high TNF-α, IL-6, IL-17 A and IL-1β) provides mechanistic insight linking molecular findings to clinical outcomes, supporting a biologically plausible mechanism for the uncontrolled inflammation that ultimately leads to surgical intervention.

This systemic inflammation could contribute to the broader clinical burden and extra-intestinal manifestations seen in severe UC cases [39]. The lack of a significant difference in baseline clinical scores (Mayo score) between our groups, despite clear molecular disparities, is noteworthy. It may imply that CMV-driven cytokine dysregulation represents an early molecular event that precedes overt clinical deterioration, positioning these cytokines as potential biomarkers for identifying patients at risk of a more aggressive disease course. Furthermore, the ROC curve analysis of serum cytokines revealed diagnostic potential for distinguishing CMV-positive UC patients. Specifically, TNF-α (AUC = 0.91), IL-1β (AUC = 0.82), and IL-6 (AUC = 0.79) demonstrated appropriate discriminatory power, reinforcing their role as key players in CMV-associated immune dysregulation. In contrast, IL-17 A showed a lower AUC (0.61), suggesting limited discriminatory utility despite its elevated levels. Although these findings highlight potential biomarker value, the ROC analysis was exploratory and not intended to define therapeutic thresholds, and prospective validation remains necessary before clinical implementation. In addition to between-group comparisons, correlation analyses further strengthened our findings. Tissue CMV viral load showed significant positive correlations with TNF-α, IL-1β, and IL-6 expression, supporting a dose–response relationship between viral burden and selective pro-inflammatory activation. Plasma viral load demonstrated a similar but weaker association pattern, primarily with TNF-α and IL-1β, suggesting that systemic viral dissemination may contribute to inflammatory amplification. Among all evaluated cytokines, only TNF-α displayed a modest but significant correlation with Mayo score, indicating that TNF-α may represent the most clinically relevant mediator linking CMV activity to disease severity. Importantly, adjustment for Mayo score in multivariable models did not materially alter the associations between CMV status and the selectively elevated cytokines, supporting the robustness of the observed inflammatory pattern. These observations are consistent with previous reports demonstrating that higher CMV tissue burden has been associated with adverse clinical outcomes and heightened inflammatory activity in UC [18, 28].

Several limitations should be acknowledged. The cross-sectional design precludes causal inference between CMV reactivation and cytokine dysregulation, as clinical data were collected at baseline only and longitudinal outcomes were not systematically evaluated. Therefore, the association between systemic CMV activation (DNAemia) or combined tissue/blood positivity and long-term clinical prognosis could not be assessed within the scope of this study. We did not evaluate cytokine dynamics following antiviral therapy, which limits causal inference regarding CMV-driven inflammatory amplification. Although active CMV colitis was confirmed by quantitative PCR and serological testing, CMV mRNA expression or antigenemia assays were not performed. We also did not evaluate concurrent reactivation of other herpesviruses such as Epstein–Barr virus (EBV) or human herpesvirus-6 (HHV-6), which may also influence mucosal inflammatory responses. Biopsies were obtained according to routine clinical indications from inflamed mucosa; however, the possibility of selection bias cannot be excluded, particularly if patients with more severe disease were more likely to undergo endoscopic evaluation. In addition, protein-level analysis focused on selected cytokines rather than employing an unbiased proteomic approach, and other relevant immune mediators may have been overlooked. Nevertheless, strengths of this study include the relatively large sample size, parallel evaluation of mucosal gene expression and systemic cytokine levels, and rigorous molecular confirmation of active CMV colitis.

Conclusion

In summary, this study demonstrates that active CMV colitis in UC patients is significantly associated with selective upregulation of key pro-inflammatory cytokines—TNF-α, IL-1β, IL-6, and IL-17 A—at both mucosal and systemic levels. This distinct immune signature supports the concept that CMV may contribute to amplification of inflammatory pathways rather than acting as a passive bystander in UC. While these findings provide mechanistic insight into CMV-associated immune dysregulation, prospective longitudinal and interventional studies are required to determine the therapeutic implications of antiviral treatment in this context.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Vahdat Poortahmasebi, Masoud Shirmohammadi and Nasser Ebrahimi Daryani conceptualized and designed the study. Masoud Shirmohammadi, Nasser Ebrahimi Daryani, and Behrooz Naghili recruited the patients, collected the clinical data and supervised the sampling procedures. Nasim Izadi, Mahin Ahangar Oskouee, Arezou Azadi and Vahdat Poortahmasebi performed the laboratory experiments and processed biological samples. Emad Behboudi, Shima Sadeghipoor, Shaghayegh Yazdani and Mohammad Aghazadeh conducted data curation, validation and statistical analyses. Vahdat Poortahmasebi, Masoud Shirmohammadi and Nasser Ebrahimi Daryani interpreted the data and drafted the manuscript. Nasim Izadi, Vahdat Poortahmasebi and Hossein Bannazadeh Baghi reviewed and revised the manuscript critically for important intellectual content. All authors read and approved the final version of the manuscript and agreed to be accountable for all.

Funding

This study was supported by Grant No. 979157 from the National Institutes for Medical Research Development (NIMAD), Tehran, Iran.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This investigation was carried out in accordance with the principles of the Helsinki Declaration. Ethical approval was granted by the Research Ethics Committees of National Institute for Medical Research Development (IR.NIMAD.REC.1398.236). All participants provided written informed consent prior to their enrollment in the study. The principal investigator, Vahdat Poortahmasebi, ensured the ethical handling and confidentiality of patient data. Written informed consent was obtained from all individual participants prior to their enrollment in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.