Genome‐wide DNA methylation profile analysis in thoracic ossification of the ligamentum flavum

Abstract Thoracic ossification of the ligamentum flavum (TOLF) causes serious spinal canal stenosis. The underlying aetiology may relate to genetic and inflammatory factors. DNA methylation plays a critical role in osteogenesis and inflammation, whereas there is no genome‐wide DNA methylation analysis about TOLF. The two subtypes of TOLF (single‐level and multiple‐level) have distinct clinical features. Using micro‐computed tomography (micro‐CT), we showed the ossification arose from the joint between two vertebrae at one/both sides of ligament flavum. With Illumina Infinium Human Methylation 850 BeadChip arrays, genome‐wide DNA methylation profile was measured in ligament flavum of eight healthy and eight TOLF samples. Only 65 of the differentially methylated cytosine‐phosphate‐guanine dinucleotides were found in both subtype groups. Principal component analysis and heat map analysis showed a different methylation pattern in TOLF samples, and methylation patterns of two subtypes are also distinct. The Gene Ontology enrichment analysis was significantly enriched in differentiation and inflammation. Pyrosequencing analysis and quantitative real‐time polymerase chain reaction were performed to validate the arrays results and expression levels, to test six differentially methylated genes (SLC7A11, HOXA10, HOXA11AS, TNIK, homeobox transcript antisense RNA, IFITM1), using another independent samples (P < 0.05). Our findings first demonstrated an altered Genome‐wide DNA methylation profile in TOLF, and implied distinct methylated features in two subtypes.


| INTRODUC TI ON
Thoracic ossification of the ligamentum flavum (TOLF) is a type of pathological heterotopic ossification that can cause severe thoracic myelopathy in Asian populations. Epidemiological studies have demonstrated that TOLF frequently occurs in the thoracic spine, and over 70% of patients suffering from thoracic spinal stenosis have been diagnosed with TOLF. 1 Due to the insidious progressive nature of the ossification and lack of conservative treatment strategies, TOLF generally requires aggressive surgery. [1][2][3][4] The pathological process of TOLF involves the differentiation of fibroblasts into osteoblasts, which is a highly regulated development process and can be described histologically based on endochondral ossification. 5 Several investigations have suggested that the potential factors associated with TOLF mainly include mechanical effects, 6,7 inflammatory factors 8,9 and genetic factors. 10,11 Recently, studies have shown that inflammation and endochondral ossification play an important role in TOLF. 5,12 Several evidences have proven the involvement and possible mechanism of tumour necrosis factor α in TOLF, suggesting the contribution of inflammation in TOLF. 8,9 However, the mechanisms underlying the development of TOLF have not yet been clarified.
TOLF is usually classified into single-level and multiple-level TOLF based on the lesion distribution in the magnetic resonance imaging (MRI), with distinct clinical features. Single-level TOLF occurs mainly in the lower thoracic spine (T10-T12), 4 while multiple-level TOLF presents with a wide and serious ossification in immobile as well as the mobile segments that differs from single-level lesions in terms of disease progression and clinical outcomes. [13][14][15] One of our previous studies has shown that the differences between the osteogenic differentiation potency in single-and multiple-level TOLF may be related to differences in the disease pathogenesis and genetic backgrounds of patients. 7 However, the pathological differences between single-and multiple-level TOLF remain unknown. Therefore, it is important to figure out whether single-and multiple-level TOLF should be analysed separately, which would make TOLF research more reasonable and conducive in the future. DNA methylation, in which the cytosine residue in cytosine-phosphate-guanine dinucleotides (CpGs) acquires a methyl group, is an important epigenetic mechanism. Several multiple rheumatic and autoimmune diseases, such as ankylosing spondylitis (AS), osteoporosis and osteoarthritis, show abnormal DNA methylation. [16][17][18] However, few studies have reported abnormally methylated loci in patients with TOLF. A Japanese research found that decreased DNA methylation in the promoter regions of the WNT5A and GDNF genes may indirectly promote the osteogenicity of mesenchymal stem cells (MSCs) from patients with ossified spinal ligaments. 19 And genome-wide DNA methylation profiling of TOLF and healthy ligamentum flavum has not been performed. Meanwhile, the differences between single-and multiple-TOLF with regards to DNA methylation are yet to be elucidated.
In this study, we first evaluated the morphological characteristics of TOLF by micro-computed tomography (micro-CT) to investigate the ossification patterns in single-and multiple-level TOLF. Then we performed a comparative analysis of the genome-wide DNA methylation profiles of ligamentum flavum samples from patients with TOLF and healthy controls. Genome-wide DNA methylation was measured to assess the communalities and discrepancies of the ligamentum flavum methylome in patients with TOLF (single-and multiple-level TOLF) and healthy controls. Furthermore, this is the first study investigating the genome-wide epigenetic landscapes of the ligamentum flavum from patients with TOLF, and reveals the potential epigenetic differences between the two types of TOLF.

| Patient specimens
This study was approved by the Ethics Committee for Human Subjects of the Peking University Third Hospital in accordance with the Declaration of Helsinki (PUTH-REC-SOP-06-3.0-A27, #2014003).
Written consent was obtained from all subjects. The diagnosis of TOLF was made on the basis of clinical symptoms and radiological examination. We previously established a TOLF classification scheme based on the lesion distribution observed in the MRI analysis 7 as follows: single-level TOLF (ossification in two or fewer adjacent levels) and multiple-level TOLF (continuous or intermittent ossification in three or more levels). We investigated patients with TOLF who underwent decompressive laminectomy between January 2018 and January 2019, and a total of 32 patients were enrolled in this study.  Table 1. All ligament samples were aseptically obtained from patients during surgery, and were separated from non-ossified sites to avoid any possible contamination of osteogenic cells.

| Micro-CT evaluation
The lamina was resected integrally by piezosurgery to ensure the collection of the ossified ligamentum flavum samples without damage.

| DNA isolation
DNA was isolated using the QIAamp DNA Mini kit (250) (Qiagen, Hilden, North Rhine-Westphalia, Germany), according to the manufacturer's protocol. Data were normalized by subtracting the background value, which was determined by averaging the signals of built-in negative control bead types. The normalized data were then used to calculate the DNA methylation levels, which were displayed as β-values ranging from 0 to 1, corresponding to unmethylated and methylated sites, respectively.

| Pyrosequencing analysis
The DNA extraction protocol was the same as the 850 K array screening subjects. DNA was bisulphite-converted using the EpiTect Bisulfite Kit (Qiagen) according to the manufacturer's standard protocol. Bisulphite-converted DNA was amplified by polymerase chain reaction (PCR) using the PyroMark PCR Kit (Qiagen) in a total reaction volume of 25 μL, which contained sequencing primer (0.3 mmol/L) and 50 ng bisulphite-converted DNA, with PCR primers listed in Table S1. After purification, 20 μL PCR product was pyrosequenced using the PyroMark Gold Q96 Kit (Qiagen) and PyroMark Gold Q96 pyrosequencer (Qiagen) according to the manufacturer's instructions. Data were collected and analysed using the PyroMark Q96 software (version 2.5.8, Qiagen).

| Statistical analysis
Data were presented as the mean ± SEM. Statistical differences in continuous variables between two groups were compared by the Student's t test. P < 0.05 was considered to be statistically sig-

| Micro-CT scanning of ossified ligamentum flavum in single-and multiple-level TOLF
Sixteen patients were enrolled in this study, that is, eight patients each with single-and multiple-level TOLF. As shown in Figure 1,  Next, we overlapped the differentially methylated CpGs from multiple-healthy group and single-healthy group, only 65 differentially methylated CGs common to these two groups were identified ( Figure 2A). These data indicated distinct DNA methylation profiles

| GO and pathway enrichment analysis
We further identified differentially methylated genes (hypermethylated and hypomethylated) that were significantly enriched in the two TOLF types. GO terms and KEGG with P < 0.05 were considered significantly enriched by differential methylation loci-related genes. The full lists were shown in the Supporting Information. After reviewing the research articles about TOLF, we shortlisted the differentially methylated GO terms functionally related to bone development, osteogenesis and inflammatory processes ( Table 2).

| Pyrosequencing validation
The number of genes that were validated was limited by the remaining amount of DNA. We therefore selected several genes with a potential role in bone biology, osteogenesis and inflammation, and validated differential methylation by pyrosequencing, including in the single-healthy TOLF group, listed in Table 3. We selected six most significantly differentially methylated CpGs from all the differentially methylated CpGs of these genes, which were then validated by pyrosequencing, and all the CpG sites showed different methylation level in their groups (Figure 4).

| qRT-PCR validation
To further assess the functional relevance of the identified genes for TOLF, we compared the mRNA expression levels. We demonstrated significantly differential expressions of these genes from TOLF samples compared with those from non-TOLF samples ( Figure 5). This was consistent with the methylation arrays data.  we plan to carry out follow-up studies to identify the function of these targeted genes and discuss the mechanism underlying the regulation of ossification by the altered epigenetic patterns.

| D ISCUSS I ON
Collectively, these findings will contribute to a better understanding of TOLF.

ACK N OWLED G M ENTS
The authors would like to thank the patients and their families for their support and participation and the Central Laboratory of Peking University Third Hospital for the technical support. This work was supported by the National Natural Science Foundation of China (grant numbers: 81772381, 81572101 and 81700779). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

CO N FLI C T O F I NTE R E S T
The authors declared that they have no competing interests. Writing-review & editing (lead).

DATA AVA I L A B I L I T Y S TAT E M E N T
Data can be obtained from the corresponding author on request.