HCMV DNA viral load and serological results
We examined the HCMV viral load in 5,664 samples of peripheral blood leukocytes that collected from 3,986 patients by quantitative real-time PCR. Among which 145 whole blood samples were defined as viremia with the HCMV DNA titers upper than 500 IU/mL and others with HCMV DNA titers lower than 500 IU/mL. Among which, 120 samples from 120 patients with HCMV-IgG seropositive and HCMV-IgM seronegative including 60 samples with HCMV DNA titers over 500 IU/mL and 60 samples with HCMV DNA titers below 500 IU/mL were selected in our study (Fig. 1). The mean levels of HCMV DNA viral load for total 60 viremia patients were 1 925 (840.3, 6 503) IU/mL, 1 755 (1 048, 3 450) IU/mL, in the training cohort and 1 995 (795.5, 9 995) IU/mL, in the validation cohort, respectively (Table 1). For the concentrations of anti-HCMV IgG, there was no significant difference between the total HCMV viremia patients (n = 60) and the controls (n = 60) (t = 0.6013, P = 0.5488). Similar results were also observed in the training cohort and the validation cohort (t = 0.6351, P = 0.5285 and t = 1.245, P = 0.2174), respectively.
Expression profiles of HCMV-encoded miRNAs by RT-qPCR analysis
In the training cohort, 22 HCMV-encoded miRNAs were measured using RT-qPCR assay in individual serum samples from 24 patients with a HCMV DNA > 500 IU/mL (referred as case set) and 24 patients with a HCMV DNA < 500 IU/mL (referred as control set) (Fig. 1). Sixteen of the 22 HCMV-encoded miRNAs were up-regulated in case set when compared with control set (P < 0.05), among which 8 miRNAs were significantly up-regulated with a fold change > 2, and P < 0.01, including hcmv-miR-US25-2-3p, hcmv-miR-US4-5p, hcmv-miR-US25-2-5p, hcmv-miR-US25-1-3p, hcmv-miR-US25-1, hcmv-miR-UL36, hcmv-miR-UL148D and hcmv-miR-US29-3p (Table 2).
Table 2
Expression profile of HCMV-encoded miRNAs in the training set
HCMV encoded miRNAs
|
Viremia (n = 24)
|
Control (n = 24)
|
Fold change
|
P values†
|
US25-2-3p
|
6.41 ± 2.07 × 10− 2
|
2.66 ± 0.57 × 10− 2
|
2.41
|
0.0028
|
US4-5p
|
9.51 ± 6.55 × 10− 3
|
1.25 ± 0.33 × 10− 3
|
7.62
|
0.0035
|
US-25-2-5p
|
6.38 ± 1.69 × 10− 2
|
2.54 ± 0.69 × 10− 2
|
2.50
|
0.0037
|
US25-1-3p
|
1.98 ± 0.75 × 10− 2
|
6.71 ± 1.59 × 10− 3
|
2.95
|
0.0039
|
US25-1
|
2.46 ± 0.67 × 10− 3
|
9.53 ± 2.28 × 10− 4
|
2.58
|
0.005
|
UL36
|
2.19 ± 0.55 × 10− 2
|
9.97 ± 2.90 × 10− 3
|
2.20
|
0.0055
|
UL148D
|
1.96 ± 0.50 × 10− 1
|
7.73 ± 1.62 × 10− 2
|
2.54
|
0.0081
|
US29-3p
|
2.00 ± 0.57 × 10− 3
|
7.51 ± 1.67 × 10− 4
|
2.66
|
0.0091
|
UL69
|
1.34 ± 0.35 × 10− 3
|
6.09 ± 1.46 × 10− 4
|
2.20
|
0.0117
|
US5-1
|
1.69 ± 0.53 × 10− 2
|
8.74 ± 2.54 × 10− 3
|
1.93
|
0.0117
|
US22-5p
|
2.01 ± 0.66 × 10− 1
|
8.19 ± 1.80 × 10− 2
|
2.46
|
0.0132
|
US33-3p
|
2.29 ± 0.52 × 10− 3
|
1.07 ± 0.25 × 10− 3
|
2.15
|
0.0132
|
UL22a
|
1.12 ± 0.29 × 10− 2
|
4.79 ± 1.07 × 10− 3
|
2.34
|
0.0171
|
UL70-5p
|
1.20 ± 0.36 × 10− 2
|
5.29 ± 1.71 × 10− 3
|
2.28
|
0.0262
|
UL36*
|
2.32 ± 0.74 × 10− 3
|
6.76 ± 1.79 × 10− 4
|
3.45
|
0.0267
|
US22-3p
|
4.38 ± 1.29 × 10− 3
|
2.24 ± 0.62 × 10− 3
|
1.95
|
0.038
|
US5-2-3p
|
3.89 ± 1.09 × 10− 4
|
1.35 ± 0.35 × 10− 4
|
2.87
|
0.068
|
UL112
|
2.91 ± 0.82 × 10− 2
|
1.23 ± 0.34 × 10− 2
|
2.37
|
0.0889
|
UL22a*
|
4.18 ± 1.24 × 10− 3
|
1.53 ± 0.44 × 10− 3
|
2.72
|
0.1147
|
UL112-5p
|
4.53 ± 1.36 × 10− 3
|
1.95 ± 0.59 × 10− 3
|
2.32
|
0.1254
|
US4-3p
|
6.10 ± 1.96 × 10− 3
|
4.54 ± 1.87 × 10− 3
|
1.34
|
0.2399
|
UL59
|
2.56 ± 0.61 × 10− 2
|
1.40 ± 0.35 × 10− 2
|
1.83
|
0.2977
|
miRNAs are presented as mean ± SEM.†, Mann-Whitney U test. |
Confirmation of the up-regulated HCMV-encoded miRNAs
Subsequently, the 8 up-regulated HCMV-encoded miRNAs were confirmed in an additional cohort including 36 cases and 36 controls (refereed as validation cohort). The 8 miRNAs exhibited consistent alterations as the results from the training cohort (Fig. 2a-h). Moreover, when combined the results of the training set and validation set (Fig. 2i-p), consistent with our expectations, the concentrations of all the eight HCMV-miRNAs were significantly increased in the viremia patients as compared with control group.
HCMV miRNAs in the autoimmune disease patients
We next analyzed the 8 up-regulated hcmv-miRs in the only antoimmune disease patient subgroup, and found that seven HCMV-miRNAs were significantly up-regulated except for hcmv-miR-US29-3p. Notably, four miRNAs including hcmv-miR-US25-2-3p, hcmv-miR-US25-2-5p, hcmv-miR-US25-1-3p and hcmv-miR-UL148D were markedly increased with a P value of < 0.01 (Fig. 3a-g). Receiver operating characteristic curve (ROC) analysis on the seven selected HCMV-miRNAs yielded areas under ROC curve (AUCs) ranged from 0.74 to 0.79 (Fig. 3h). Using the optimal cutoff value, we obtained the following sensitivity and specificity values: hcmv-miR-US25-2-3p (AUC: 0.794, sen: 85.7%, spe: 72.2%, 95%CI: 0.629, 0.958), hcmv-miR-US4-5p (AUC: 0.766, sen: 85.7%, spe: 66.7%, 95%CI: 0.597, 0.935), hcmv-miR-US25-2-5p (AUC: 0.786, sen: 78.6%, spe: 72.2%, 95%CI: 0.628, 0.943), hcmv-miR-US25-1-3p (AUC: 0.794, sen: 78.6%, spe: 77.8%, 95%CI: 0.632, 0.955), hcmv-miR-US25-1 (AUC: 0.766, sen: 85.7%, spe: 77.8%, 95%CI: 0.589, 0.942), hcmv-miR-UL36 (AUC: 0.738, sen: 85.7%, spe: 66.7%, 95%CI: 0.562, 0.914) and hcmv-miR-UL148D (AUC: 0.794, sen: 85.7%, spe: 77.8%, 95%CI: 0.629, 0.958).
Viral miRNAs expression pattern in the blood of patients with HCMV disease
In combined samples of the training set and validation set of patients (n = 60), we found that 91.67% patients had detectable expression of at least one HCMV-miRNA (10% with one, 5% with two, 6.67% with three, 3.33% with four, 3.33% with five, 6.67% with six, 6.67% with seven and 50% with eight) (Fig. 4a). Analysis of individual HCMV-miRNA showed that hcmv-miR-US4-5p was the most commonly detected in 81.67% of the patient, followed by hcmv-miR-US29-3p with 76.67%, hcmv-miR-UL36 and hcmv-miR-US25-1-3p with 73.33%, hcmv-miR-UL148D with 61.67%, hcmv-miR-US25-2-5p with 58.33%, hcmv-miR-US25-2-3p and hcmv-miR-US25-1 with 56.67% (Fig. 4b).
Association of serum hcmv-miRNAs’ levels with clinical parameters
We subsequently wonder whether serum hcmv-miRNAs’ levels were correlated with clinical parameters. We evaluated the associations between the clinical features and miRNA abundance using Pearson correlation analysis in all of the studied individuals. Hcmv-miR-US25-1-3p levels were significantly correlated with HCMV DNA copies (r = 0.2970, P = 0.022) (Fig. 5a). HCMV DNA copies was significantly correlated with CRP (r = 0.3007, P = 0.0246) (Fig. 5b) but not with WBC count (r = 0.0497, P = 0.80). The concentrations of anti-HCMV IgG was significantly correlated with hcmv-miR-US25-1 (r = 0.3071, P = 0.018) (Fig. 5c) but not with other miRNAs. There was no significant difference in WBC count (t = 0.9167, P = 0.3612), CRP (t = 0.6005, P = 0.5494), PCT (t = 1.229, P = 0.2249), and ESR (t = 0.6939, P = 0.4908) between the viremia and the control group.
Independent cohorts or antiviral treatment
Five patients received antiviral therapy, including 2 patients with severe aplastic anemia, one with myelodysplastic syndromes, one with acute myeloid leukemia M2a and acute lymphoblastic leukemia of one case respectively which underwent bone marrow transplantation. 47 serum samples were collected at different time points during the antiviral therapy, and hcmv-miR-US25-1-3p levels was significantly correlated with HCMV DNA copies during the antiviral therapy (Fig. 6a-e).