Abstract
Objective. SAPHO (synovitis, acne, pustulosis, hyperostosis, osteitis) syndrome is a type of rare chronic aseptic inflammation of unknown etiology. To date, no research to our knowledge has reported copy number variation (CNV) of genes that could affect predisposition to SAPHO syndrome. We investigated the association between CNV profile and SAPHO syndrome.
Methods. We used array comparative genomic hybridization (CGH) to screen for CNV in a nuclear family including 2 patients and a healthy control. We then validated the copy numbers of candidate genes found in the array CGH assay and other candidate genes by TaqMan real-time PCR in 360 case and control samples.
Results. Ten regions from 8 chromosomes were found to have abnormal gene copies in the nuclear family, so the CNV of candidate genes (ADAM5, CSF2RA, IL3RA, and 9 other genes) were tested by TaqMan PCR. Significant copy number loss of CSF2RA (p = 0.000) and NOD2 (p = 0.005), and significant copy number gain of MEGF6 (p = 0.002) and ADAM5 (p = 0.000) were seen in patients with SAPHO compared with controls at the a = 0.05 level. There were no differences in the other 8 candidate genes between patient and control samples (p > 0.05).
Conclusion. Our study established the first association between CNV in CSF2RA, NOD2, MEGF6, and ADAM5 and SAPHO syndrome. These findings may offer insight into the pathogenesis of SAPHO and provide the basis for improved diagnosis and treatment.
SAPHO (synovitis, acne, pustulosis, hyperostosis, osteitis) syndrome, a rare disease with an estimated prevalence of less than 1 in 10,0001, was first identified in 19872. SAPHO syndrome can occur at any age3, and its exact etiology is unknown. However, a large body of literature supports a pathogenesis involving a combination of genetic, infectious, and immunological components4. Experiments in mice indicate that SAPHO is associated with mutations in PSTPIP1, PSTPIP25, and NOD2 genes6,7, but the genetic basis of human SAPHO syndrome remains poorly understood.
Copy number variation (CNV) is widespread in human populations, with 5%–10% of the human reference genome showing CNV in healthy individuals8,9. Several studies have concluded that variation in DNA copy number may influence the expression of genes or even alter their structure, thus contributing to substantial phenotypic variation. Moreover, CNV of genes causes a variety of human genetic diseases10.
Little is known about CNV as a potential risk factor for SAPHO syndrome; we screened the CNV profile in a nuclear family using array comparative genomic hybridization (CGH). We then further confirmed the candidate CNV genes in a larger sample size (156 SAPHO cases and 204 healthy controls) using TaqMan real-time PCR. We hope to propose the potential application of CNV evaluation in individuals at high risk of developing SAPHO syndrome, and the use of CNV genes as candidates for gene therapy.
MATERIALS AND METHODS
In the first stage, a core family with SAPHO syndrome was identified [mother: age 51 yrs, with SAPHO syndrome, with palmoplantar pustolosis and osteoarticular symptoms (anterior chest wall, spine, peripheral skeleton), duration of disease 8 yrs; daughter: age 22 yrs, with SAPHO syndrome, with osteoarticular symptoms (anterior chest wall, spine), duration of disease 1.5 yrs; father: age 53 yrs, healthy control]. The second group comprised 156 patients with SAPHO syndrome (92 women, 64 men; average age 41.4 ± 0.08 yrs) and 204 controls (122 women, 82 men; average age 59.1 ± 0.08 yrs; Table 1). All individuals were enrolled from Beijing Chaoyang Hospital and were matched for ethnic and geographic characteristics. SAPHO syndrome was diagnosed according to the criteria of Kahn, et al11.
General characteristics of patients with SAPHO syndrome (n = 156) in this study.
The study was approved by the Ethics Committee of the National Research Institute for Family Planning (registration no. 2201512), and all participants provided written informed consent.
DNA isolation
Participants’ peripheral blood samples were collected into tubes containing EDTA. DNA extraction was carried out using the RelaxGene Blood DNA System (Tiangen Biotech) according to the manufacturer’s instructions. Array CGH analysis was performed using 4 × 180K commercial arrays (Capitalbio Technology Corp.). DNA labeling, hybridization, and washing were performed according to the manufacturer’s instructions. After hybridization, arrays were scanned in a dual-laser scanner, and images were extracted and analyzed using Agilent CytoGenomics software. Probes with fewer than 2 valid replicate values or probes that showed SD above 0.1 between replicate values were excluded. Chromosomal regions with a ratio between 1.11 and 1.41 were scored as “gained” and those with a ratio of 1.41 or greater were scored as “amplified.” Chromosomal regions with a ratio of between 0.84 and 0.73 were scored as “loss,” whereas a second threshold for loss was set for regions showing a ratio < 0.73.
TaqMan quantitative PCR (qPCR)
TaqMan qPCR was used to identify candidate genes. Primers and fluorescence-labeled probes are shown in Table 2. qPCR was performed in a final volume of 25 µl including 20 ng of extracted DNA and 20 µl of PCR mixture. All qPCR reactions were run for 45 cycles. The Ct value of each gene was recorded and 2–ΔCt (ΔCt = Ct target gene – Ct RNase P) was used to determine the CNV of each sample.
Candidate genes and primers used in this study.
Statistical analysis
Clinical data were described as mean ± SD. Comparison of cases and controls was by Pearson chi-square test or Fisher’s exact test. Statistical analyses were performed using SPSS version 13.0 (SPSS Inc.); p < 0.05 was considered statistically significant.
RESULTS
We conducted a 2-stage candidate gene study to test the association of CNV and SAPHO syndrome. First, through array CGH analysis, we identified several aberrant chromosome regions shared by 2 of the core family patients with SAPHO syndrome: q11 region of chromosome 11 was deemed as amplification, including OR4S2, OR4C6; 5 regions (q31.3 of chromosome 1, p11.22 of chromosome 8, p12 of chromosome 19, q11.23–q12.1 of chromosome 22, and q13.1 of chromosome 22) were deemed as gain, including ADAM5P, APOBEC3A, CFHR3, etc. Three regions were deemed as loss (q13 of chromosome 2, p22.3 of chromosome 9, p22.33 of chromosome X), including CSF2RA, MIR3690, IL3RA, etc. Detailed results are shown in Table 3.
General information on candidate genes in the nuclear family in this study.
In the validation stage, we chose ADAM5, CSF2RA, IL3RA, and 9 other genes associated with inflammation or autoimmune (AAGAB, HRVR, IGSF9, PSTPIP1, PSTPIP2, LPIN2, IRAK3, NOD2, MEGF6) as candidate genes. TaqMan qPCR was used to determine the variations of candidate gene copy numbers in 156 patients with SAPHO syndrome and 204 controls. Finally, 4 of 12 chosen genes were found to have significantly different copy numbers between patients with SAPHO syndrome and controls at the a = 0.05 level, of which CSF2RA (ratio SAPHO/HC = 0.597, p = 0.000) and NOD2 (ratio SAPHO/HC = 0.470, p = 0.002) had fewer gene copies, and MEGF6 (ratio SAPHO/HC = 1.407, p = 0.001) and ADAM5 (ratio SAPHO/HC = 1.713, p = 0.000) had more copies, compared with healthy controls. There were no differences in the other 8 candidate genes between patients with SAPHO syndrome and control samples (p > 0.05; Tables 4A and 4B; Figures 1 and 2). In subsequent analysis, patients with SAPHO syndrome were subgrouped according to osteoarticular and skin symptoms (Tables 4A and 4B). The results showed the gene copies of CSF2RA, NOD2, ADAM5, and MEGF6 were significantly different in palmoplantar pustulosis; meanwhile, CSF2RA was found to be strongly correlated with osteoarticular symptoms; the results achieved statistical differences in all subgroups. As for other symptoms, perhaps because of small sample sizes, the results showed inconsistencies among different subgroups, especially in the no-skin-manifestation samples, so the results should be treated cautiously.
CNV ratio of candidate genes in SAPHO patients and healthy controls (Part A). HC: healthy controls; CNV: copy number variations; SAPHO (synovitis, acne, pustulosis, hyperostosis, osteitis) syndrome.
CNV ratio of candidate genes in SAPHO patients and healthy controls (Part B). HC: healthy controls; CNV: copy number variations; SAPHO (synovitis, acne, pustulosis, hyperostosis, osteitis) syndrome.
Copy number variation ratio of candidate genes in patients with SAPHO syndrome and healthy controls.
Copy number variation ratio of candidate genes in patients with SAPHO syndrome and healthy controls.
DISCUSSION
SAPHO syndrome is a type of rare chronic aseptic inflammation, and its complete pathogenesis is unknown. Previous investigations showed that PSTPIP1, PSTPIP2, and LPIN2 were associated with SAPHO syndrome in mice5,12,13. Other research showed that dysregulation of interleukin 1 (IL-1) signaling caused sterile osteomyelitis in PSTPIP2-deficient mice10,14. However, genetic screening found no specific variants in PSTPIP1, PSTPIP2, NOD2, or LPIN2 in patients with SAPHO syndrome7. We used array CGH and TaqMan qPCR to determine the role of genetic factors in the development of SAPHO. For the first time, we demonstrated that copy number loss of CSF2RA and NOD2, and copy number gain of MEGF6 and ADAM5, contribute to SAPHO syndrome.
CSF2RA is located on Xp22.33 and Yp11.2, and is inherited in an autosomal recessive manner15. The CSF2RA protein functions as the alpha subunit of the heterodimeric receptor for colony-stimulating factor 2, a cytokine that controls the production, differentiation, and function of granulocytes and macrophages16,17. We demonstrated that the CSF2RA copy number was significantly decreased in patients with SAPHO syndrome, which would likely decrease its interaction with granulocyte-macrophage colony-stimulating factor (GM-CSF) and its receptor, resulting in an accumulation of GM-CSF in the peripheral blood of patients. Previous research showed GM-CSF played a pivotal role in regulating inflammatory networks, and aberrant expression of GM-CSF was found in some autoimmune diseases, including rheumatoid arthritis18,19 and synovitis20,21,22. In synovitis, abundant GM-CSF can promote joint inflammation by secreting large amounts of tumor necrosis factor-α (TNF-α) and IL-623. Indeed, elevated IL-6 and TNF concentrations have been observed in the serum of patients with SAPHO syndrome24, while the use of anti-TNF agents has proven a valid alternative for patients unresponsive to conventional treatments, including nonsteroidal antiinflammatory drugs, corticosteroids, disease-modifying antirheumatic drugs, and bisphosphonates25. GM-CSF also enhances IL-1β production by macrophages by regulating lipopolysaccharide (LPS)-mediated pro–IL-1β expression26. Excessive IL-1β can lead to osteitis, a major feature of SAPHO syndrome, and enhance mesenchymal cell differentiation in osteoblasts27. SAPHO syndrome has been shown to partly depend on genetically encoded overproduction of IL-1β27. Additionally, excessive IL-1β triggers an innate immune response in the skin28, including in vivo inflammatory responses to Propionibacterium acnes, another major phenotype of SAPHO syndrome, which is in accord with the acne phenotype of SAPHO syndrome. We therefore speculate that the excess GM-CSF induced by decreased CSF2RA is a major cause of SAPHO syndrome development.
NOD2 is located in chromosome 16, a region highly associated with some systemic autoinflammatory diseases. NOD2 was the first susceptibility gene identified in inflammatory bowel disease29. Previous results described genetic variants of NOD2 in a series of systemic autoinflammatory diseases, including Crohn disease, Blau syndrome, and Yao syndrome30. NOD2 is also associated with skin diseases31, and it plays a pivotal role in the immune response to intracellular bacterial LPS by recognizing the muramyl dipeptide and activating the NFKB protein31; and delayed and ineffective recognition of a localized bacterial infection in the absence of NOD2 protein can lead to increased bacterial load and a delayed, more severe local inflammatory response32. Given its important role in autoinflammatory diseases, much effort has been made to determine its relationship with SAPHO syndrome. However, no genetic variants were found: we observed that NOD2 copy number losses instead of genetic variants were involved in patients with SAPHO syndrome, and this represents significant progress in SAPHO research.
We also showed that the copy numbers of MEGF6 and ADAM5 were higher in patients with SAPHO syndrome than in healthy controls. Although we inferred that these differences may increase the risk of acne and pustulosis, little is known about MEGF6 and ADAM5 proteins, so further study is required.
We established the first association between CNV in CSF2RA, NOD2, MEGF6, and ADAM5 and SAPHO syndrome. These findings may provide insight into the pathogenesis of SAPHO syndrome, and perhaps a basis for diagnosis and treatment. However, further studies are required to elucidate the pathogenesis of SAPHO syndrome.
Acknowledgment
We thank all participants for their contributions to the research. We thank Sarah Williams, PhD, from Liwen Bianji, Edanz Group China, for editing the English text of a draft of this report.
Footnotes
Supported by grants 0102012DFB30130 from the International Science & Technology Cooperation Program of China; 2016YFC1000300, 20161000307, 2016YFC1000803, and 2017YFC1002000 from the National Key Research and Development Program; and 81501418 from the National Natural Science Fund.
- Accepted for publication September 27, 2019.
