Patients
Twenty-eight patients with secondary syphilis, 27 patients with serofast status, and 40 healthy volunteers from our hospital were enrolled in this study. There were no significant differences in the sex and age distributions among these three patient groups. All patients met the criteria for the diagnosis of syphilis according to Chinese national guidelines [22]. Pregnant or breast-feeding women and patients who were HIV antibody-positive or had neurosyphilis, tuberculosis, an autoimmune disease, cancer, chronic liver disease, kidney disease, or another systemic disease were excluded from enrollment. Among the enrolled participants, five subjects from each group were selected for further small RNA transcriptome sequencing (Table 1).
Overview of global miRNA expression profiles
Small RNA transcriptomes from the PBMCs of 15 subjects (i.e., five secondary syphilis patients, five serofast status patients, and five healthy volunteers) were sequenced using the Illumina platform. The average number of high-quality sequencing reads in these 15 samples was 26,713,263 with 78% of reads completely mapped to the human genome (Table 1). Among these reads, 73.32% were identified as miRNA transcripts and enriched in 22 bp (Fig. 1).
Using a cutoff of read counts N > 0, we identified a total of 1,875, 1,791, and 1,793 known miRNAs in secondary syphilis, serofast status, and healthy control subjects, respectively. Of these, 1,592 known miRNAs were expressed in samples from all three groups; the remaining 113, 70, and 76 miRNAs were uniquely expressed in the secondary syphilis, serofast status, and healthy control subjects, respectively (Fig. 2a). Interestingly, miR-548b was expressed only in serofast status patients; miR-548b was reported to inhibit the proliferation and invasion of malignant gliomas by targeting metastasis tumor-associated protein-2 [30]. Thus, we thought that miR-548b might have similar functions, inhibiting the proliferation and invasion of T. pallidum and leading to a low RPR titer.
There were 364, 361, and 364 novel predicted miRNAs, of which 6, 3, and 5 were uniquely expressed, discovered in the secondary syphilis, serofast status, and healthy control subjects, respectively (Fig. 2b). Among them, the novel predicted miRNA np-miR-210 (ACAATAGGGTTTACGACC) was specifically expressed at a high level in serofast status subjects.
To better evaluate the miRNA expression levels in different samples, the read counts were normalized by their TPM values and classified into five levels (Fig. 2cd). We noticed that the overall expression level was very low (enriched in 0–1), and only ~11% of known miRNAs and less than 0.5% of novel predicted miRNAs had a TPM value of more than 100. The distributions of samples across the five expression levels were similar for all three patient groups.
DEmiRNA analysis revealed a specific miRNA expression profile and potential markers of serofast status
An analysis of the DEmiRNAs among secondary syphilis, serofast status, and healthy control subjects was conducted. A total of 146 DEmiRNAs (54 upregulated and 92 downregulated), 111 DEmiRNAs (61 upregulated and 50 downregulated), and 142 DEmiRNAs (88 upregulated and 54 downregulated) was identified between secondary syphilis and healthy control subjects, serofast status and healthy control subjects, and serofast status and secondary syphilis subjects, respectively (Fig. 3). The finding that the amount of DEmiRNAs between the serofast status and healthy control subjects was the lowest indicates that the miRNA expression in the serofast status is closer to that in healthy control subjects than to that in secondary syphilis patients. In addition, clustering of the identified DEmiRNAs could distinctly classify the 15 samples into three groups (Fig. 4).
Some DEmiRNAs overlapped among the three groups. We identified 15 DEmiRNAs that were upregulated (miR-548bv, miR-889-5p, miR-3135a, miR-144-3p, miR-338-5p, miR-548j-3p, np-miR-194*, miR-196b-3p, miR-548bs, np-miR-210, np-miR-50, miR-3135b, np-miR-284*, miR-208b-3p, and np-miR-211) (Fig. 3e) and five that were downregulated (np-miR-66, miR-7641, miR-3150b-3p, np-miR-3, and miR-6868-3p) (Fig. 3f) in serofast status patients compared with both the healthy control and secondary syphilis subjects. These can be used as potential markers for the assessment of serofast status.
DEmiRNA target gene and function analysis revealed serofast status may be closely related to immunological function
Target genes of the DEmiRNAs were predicted based on the Sanger MicroRNA database. The Unigene database was further used for screening these target genes. A GO analysis of the DEmiRNA target genes revealed that: (1) for secondary syphilis vs. healthy control subjects, the target genes of miRNAs upregulated in secondary syphilis patients were enriched in the ncRNA catabolic process, regulation of double-strand break repair, and positive regulation of cellular carbohydrate, whereas the target genes of miRNAs downregulated in secondary syphilis patients were enriched in the pyruvate metabolic process, lymphocyte co-stimulation, and T-cell co-stimulation (Fig. 5a); (2) for serofast status vs. healthy control subjects, the target genes of miRNAs upregulated in serofast status patients were enriched in histone deacetylation, regulation of receptor activity, and positive regulation of receptor biosynthetic process, whereas the target genes of miRNAs downregulated in serofast status patients were enriched in DNA synthesis involved in DNA repair, the RNA catabolic process, and calcium activated cation channel activity (Fig. 5b); and (3) for serofast status vs. secondary syphilis patients, the target genes of miRNAs upregulated in serofast status patients were abundant in positive regulation of alpha-beta T-cell differentiation and calcium-activated potassium channel activity, whereas the target genes of miRNAs downregulated in serofast status patients were enriched in DNA replication, DNA metabolic process, and ion and transmembrane transporter activity (Fig. 5c).
Interestingly, compared with healthy control samples, DEmiRNAs targeted to genes involved in the regulation of double-strand break repair were upregulated in both the serofast status and secondary syphilis samples. Furthermore, the target genes of miRNAs upregulated in the serofast status samples compared with in the secondary syphilis and healthy control samples were mainly related to receptor biosynthesis or T-cell differentiation, which further implies that the pathogenesis of serofast status is closely related to immunological function.
A KEGG pathway enrichment of the DEmiRNA target genes was also conducted, and the results are shown in Table 2. Compared with healthy control samples, the pathways significantly enriched in patients with serofast status or secondary syphilis included metabolic pathways (e.g., histidine metabolism, fructose and mannose metabolism, tyrosine metabolism, sphingolipid metabolism, and ether lipid metabolism) and a cell growth pathway (Circadian entrainment). Additionally, DEmiRNAs upregulated in serofast status subjects but not in secondary syphilis subjects targeted the neurotrophin signaling pathway and calcium signaling pathway.
DEmiRNAs were further validated by qPCR
To confirm the expression levels of DEmiRNAs, qPCR was conducted. A total of 25 DEmiRNAs (fold-change of >2) in both secondary syphilis and serofast status patients compared with healthy control subjects were selected for further validation (miR-10b, miR-62, miR-122, miR-299, miR-451a, miR548-ar-3p, miR-548av-3p, miR-548az, miR-548j, miR-654, miR-1285f, miR-2478, miR-3135b, miR-3150b, miR-338-5p, miR-3897, miR-4732, np-miR-5, np-miR-73, np-miR-124, np-miR-128, np-miR-163, np-miR-166, np-miR-194, and np-miR-244) (Additional file: Table S3). Due to individual variability, clinical samples always show diverse expression levels. Of these, the qPCR results confirmed 13 miRNAs (miR-62, miR-299, miR-451a, miR-548j, miR-654, miR-3135b, miR-338-5p, np-miR-5, np-miR-73, np-miR-124, np-miR-128, np-miR-163, and np-miR-244) (Additional file: Table S4).
These 13 DEmiRNAs confirmed by qPCR were further verified in additional samples (23 secondary syphilis, 22 serofast status, and 40 healthy control subjects); these results validated five miRNAs (miR-338-5p, np-miR-5, np-miR-128, np-miR-163, and np-miR-244). Among these five miRNAs, np-miR-128 and np-miR-244 were downregulated in the PBMCs of secondary syphilis and serofast status patients; np-miR-163 was downregulated and np-miR-5 was upregulated in PBMCs from secondary syphilis patients; and the known miR-338-5p was upregulated in PBMCs from serofast status patients (Additional file: Table S4 and Fig. S1).
miR-338-5p, upregulated in serofast status patients, might be a potential biomarker of serofast status
In the present study, we also predicted the miR-338-5p target genes and conducted a GO function enrichment of these miR-338-5p target genes (Fig. 6). We found that the functions were mainly involved ion or protein transport, and many of these biological processes required RANBP17, XPO1, and XPO6. In microbial infection, XPO1 mediates the export by different viruses (e.g., HIV-1, HTLV-1, and influenza A) of unspliced or incompletely spliced RNAs out of the nucleus [31, 32]. RANBP17 and XPO6 are also involved in protein import to the nucleus or export from the nucleus. Therefore, we speculate that in the PBMCs of serofast status patients, miR-338-5p might play an important role in regulating the infection process of T. pallidum by targeting RANBP17, XPO1, or XPO6, thus reducing the deleterious effects of T. pallidum.