Introduction

Osteonecrosis of the femoral head (ONFH) is a very common disease in daily life. It is estimated that there are 20,000,000 ONFH patients in the world, of which alcohol-induced ONFH accounts for 20–45% of ONFH [1,2,3]. With alcohol into people’s lives, the incidence of alcoholic ONFH is still rising. The risk of alcohol-induced ONFH is higher with the high alcohol intake and the longer drinking time [4]. Alcohol-induced ONFH has brought great damage to the patient’s physiology and psychology and also brought a great burden to the society. A large number of literatures have reported that alcohol-induced ONFH is related to genetic factors. Okazaki et al. showed that TLR4 signaling pathway was involved in the pathogenesis of alcohol-induced ONFH [5]. Yoon BH et al. pointed that alcohol-induced ONFH was associated with individual genetic susceptibility [6]. However, the exact pathogenesis and molecular mechanisms leading to the onset of the disease are still unclear.

Malignant brain tumor (MBT) family is a large family with L3MBTL1 [7], L3MBTL2 [8], L3MBTL3, L3MBTL4, and other members. Studies have shown that L3MBTL1 and L3MBTL2 are related to bone tissue related diseases. L3MBTL3 (also known as MBT1) is located on the long arm of chromosome 6 and has 30 exons. It is a member of the transcriptional inhibitor family of chromatin interaction in MBT [9]. It can bind to methyl lysine in histone, further affecting the transcription and its abnormal regulation associated with many diseases. Seyede Zahra Nazari Mehrabani et al. showed that L3MBTL3 gene polymorphism was related to the susceptibility of multiple sclerosis [10]. Research by Satoko Arai et al. showed that MBT-1 specifically regulated the maturation and progression of myeloid progenitor cells, and MBT-1 also showed that it effected bone marrow cell production by transiently increasing the expression of p57KIP2 [11]. However, the association of L3MBTL3 with alcoholic osteonecrosis has not been reported.

Protein tyrosine phosphatase acceptor 9 (PTPN9), also known as PTP-MEG2, is an important member of the protein tyrosine phosphatase family, and its N-terminal contains a phosphoinositol binding Sec14 homologous domain. Previous studies have shown that PTPN9 is involved in the regulation of vesicle fusion and transport and then plays a crucial role in the physiological and pathological processes such as cell proliferation, differentiation, and migration [12]. Researches have shown that PTPN9 is associated with a variety of human cancers, including hepatocellular carcinoma [13], colorectal cancer [14], cervical cancer [15], and breast cancer [16]. It has also been shown that PTPN9 affects growth and expansion of erythroid cells [17] and embryonic development [18]. However, the relationship between PTPN9 gene and alcohol-related ONFH is not clear.

As mentioned above, we speculate that the L3MBTL3/PTPN9 gene is related to alcohol-induced ONFH. In this study, genotyping was used to explore the genetic association between L3MBTL3/PTPN9 polymorphism and alcohol-related ONFH risk in Chinese Han population.

Methods

Research subjects

In this study, 308 patients with alcohol-induced ONFH were recruited, and 425 healthy control population samples were collected from the physical examination center of the same hospital during the same period. This study was approved by the ethics committee. Prior to the questionnaire and blood collection, all subjects signed informed consents.

Diagnosis and exclusion criteria

The patient’s intake of pure ethanol for more than 6 months should be greater than 400 ml/week (320 g/week, any alcoholic beverages); the patient was diagnosed with ONFH within 1 year after alcohol intake and the patient should not have other risk factors [19]. The combined results of bone scan [20], x-ray, and magnetic resonance imaging (MRI) were diagnosed as alcoholic ONFH. Patients with a clear history of direct trauma or chronic illnesses or possibly multiple causes were excluded. All healthy controls had a drinking habit, drinking more than 400 ml (320 g/week, any type of alcoholic beverage) per week for more than 6 months; however, they had never been diagnosed with alcohol-induced ONFH, hyperlipidemia, rheumatoid arthritis, syringomyelia, osteoporosis, decompression disease, cardiovascular disease, steroid use, smoking, etc.

Questionnaire investigation

In this study, epidemiological questionnaire was used to investigate the subjects including general demographic characteristics (age, gender, height, weight, etc.), smoking, family genetic history, and other basic information.

Genotyping

The basis for our selection of candidate loci is based on haplotype data or genotype data [21, 22] and from the 1,000 Genome Projects (http://www.internationalgenome.org/) to select loci with a minor allele frequency (MAF) greater than 0.05 in the global population. By consulting a large number of documents and searching the database, we found that L3MBTL3/PTPN9 gene was related to a variety of human diseases. However, the three single-nucleotide polymorphisms (SNPs) (rs1125970, rs4897367, rs2068957) in the L3MBTL3 gene and the three SNPs (rs76107647, rs10851882, and rs75393192) in the PTPN9 gene have not been reported before, and we used them as candidate sites for this study.

In the fasting state of the test population in the morning, 5 ml of peripheral venous blood was drawn into an anticoagulation tube containing EDTA and stored in a refrigerator at − 20 °C for later use. Then, according to the whole blood genomic DNA extraction kit (GoldMag Co. Ltd, Xi′an, China) [23], we extracted DNA from venous blood and measured the DNA concentration and purity by nanodrop 2000 ultraviolet spectrophotometer. We used Agena’s online software (https://agenacx.com/online-tools/) to design the upstream and downstream primers and single-base extension primers for the selected six SNPs and used Agena MassARRAY system (Agena Bioscience, San Diego, CA, USA) for SNP genotyping, which was used in articles previously reported [24].

Statistical analyses

We used Microsoft Excel and SPSS 2.0 software to perform statistical analysis. Welch’s t test was used to analyze whether there was statistical difference in the mean age between the case group and the control group. Furthermore, in order to ensure the reliability of the experimental results, an accurate test was performed to analyze whether the distribution of genotype frequency of each SNP site in the control group satisfied Hardy–Weinberg equilibrium (HWE) [25]. Afterwards, in order to clarify the relationship between studied sites and the susceptibility of alcoholic ONFH, we performed an unconditional logistic regression model analysis under the genetic models (co-dominant, dominant, recessive, and log‐additive) through PLINK software (version 1.07) [26]. Then, we conducted stratification analysis on age, course of disease, necrotic site, and stage. We have also carried out linkage disequilibrium (LD) and haplotype analysis through the Haploview software package. Finally, to better understand the interaction of SNP–SNP in alcoholic ONFH risk, we conducted a multifactor dimensionality reduction (MDR) (version 3.0.2) analysis [27]. For all tests, a two-tailed p value < 0.05 is considered to have statistical significance. The sample size and the proportion between the experimental group and the control group were determined through G*Power 3.1.9.2 software [28]. After calculation, the total sample size of the experiment, the sample size of the control group, and the sample size of the case group were required to be greater than 400, 233, and 167, respectively. In our study, the total sample size, the sample size of the control group, and the sample size of the case group were 733, 308, and 425, respectively, so our sample size met the statistical requirements. At the same time, the calculated power of this study is 99.99%, which was fully in line with statistical significance.

Results

Subject characteristics

The basic information of all study participants is shown in Table 1. In this study, we recruited 308 alcoholic ONFH and 425 healthy controls. The average age of subjects was 43.4 ± 11.3 years in the case group and 42.7 ± 12.9 years in the control group. Three hundred eight patients presented with different distribution, according to necrotic sites [unilateral necrosis 65 (21%), bilateral necrosis 243 (79%)], phase [phase III + IV 218 (71%), phase I + II 90 (29%)], and course [greater than 22 months 103 (33%), less than or equal to 22 months 205 (67%)].

Table 1 Comparison of characteristics between control group and alcohol-induced ONFH
Full size table

Relationships of L3MBTL3/PTPN9 polymorphisms with alcoholic ONFH predisposition

The basic information related to the study site and alcoholic ONFH is described in Table 2, including position, alleles (A/B), p HWE, and MAF (case and control). We analyzed the correlation between L3MBTL3/PTPN9 polymorphisms and alcohol-induced ONFH risk in the allele model, and the results are shown in Table 2. However, we did not find any significant statistical difference between the SNPs and alcohol-induced ONFH risk (p > 0.05). The genetic model analyses showed that rs2068957 in the L3MBTL3 gene increased the risk of alcohol-induced ONFH by 2.18 times under the co-dominant model (OR = 2.18; 95% CI, 1.01–4.72; p = 0.048) and 2.34 times under the recessive model (OR = 2.34; 95% CI, 1.09–5.02; p = 0.030) before correction and only 2.31 times under the recessive model (OR = 2.31; 95% CI, 1.07–4.97; p = 0.032) after correction. However, there was no significant statistical significance between the other loci and the risk of alcohol-induced ONFH in the four genetic models (co-dominant, dominant, recessive, and log-additive models) (Table 3).

Table 2 Basic information of the 6 SNPs in L3MBTL3/PTPN9 gene in this study
Full size table
Table 3 Genetic model analyses of the SNP and the risk of alcohol-induced ONFH
Full size table

Stratified analysis

In order to better understand the relationship between the six loci studied and the susceptibility of alcohol-induced ONFH, we conducted four stratified analyses of age, stage, necrotic sites, and course of disease. In the age stratification analysis, we only found that rs75393192 in the PTPN9 gene was related to the reduction of alcohol-induced ONFH when it was over 40 years old under the co-dominant model (OR = 0.54; 95% CI, 0.29–1.00; p = 0.049). When it was less than or equal to 40 years old, the six loci studied were not related to the susceptibility of alcohol-induced ONFH (p > 0.05). In the stage stratification analysis, we only found that rs75393192 was related to the reduction of alcohol-induced ONFH under the allele model (OR = 0.49; 95% CI, 0.26–0.93; p = 0.025) (Table 4). In the necrotic site stratification analysis, we only found that rs2068957 was increased to the susceptibility of alcohol-induced ONFH under the co-dominant model (OR = 2.26; 95% CI, 1.01–5.05, p = 0.047) and recessive model (OR = 2.45; 95% CI, 1.1–5.43; p = 0.028) (Table 5). In the course of disease stratification analysis, there was no significant correlation between the six loci studied and the susceptibility of alcohol-induced ONFH in four genetic models.

Table 4 Stratified analysis of the age and stage on association between selected SNPs and alcohol-induced ONFH risk
Full size table
Table 5 Stratified analysis of the necrotic areas on association between selected SNPs and alcohol-induced ONFH risk
Full size table

The association between L3MBTL3/PTPN9 haplotypes and alcoholic ONFH susceptibility

In the study of the relationship between the L3MBTL3/PTPN9 haplotype and the susceptibility of alcoholic ONFH, we only found that the GC haplotype of rs76107647|rs10851882 in PTPN9 gene was related to the susceptibility of alcoholic ONFH (OR = 0.69; 95% CI, 0.49–0.98; p = 0.040 in crude analysis; OR = 0.67; 95% CI, 0.47–0.96, p = 0.029 in adjusted by age) (Table 6), and there was no statistical significance between the haplotype of rs1125970, rs4897367, and rs2068957 in L3MBTL3 gene and rs75393192 in PTPN9 gene and the susceptibility of alcoholic ONFH (p > 0.05).

Table 6 Haplotype analysis for the effect of L3MBTL3/PTPN9 gene haplotypes in course of disease stratified analysis on alcohol-induced ONFH risk
Full size table

MDR analysis for the effect of L3MBTL3/PTPN9 SNP–SNP interaction on alcoholic ONFH risk

In order to better understand the impact of L3MBTL3/PTPN9 SNP–SNP interaction on the risk of ONFH, we conducted MDR analysis. Unfortunately, we did not find the best model for the impact of L3MBTL3/PTPN9 SNP–SNP interaction on the risk of ONFH (Table 7, Supplementary Fig. 1).

Table 7 MDR analysis of SNP–SNP interactions in relation with alcohol-induced ONFH risk
Full size table

Discussion

In this case–control study, we found that rs2068957 in the L3MBTL3 gene was related to the susceptibility of alcoholic ONFH under the recessive model. After stratified analysis, we also found that rs75393192 in the PTPN9 gene and rs2068957 in the L3MBTL3 gene were relevant to the susceptibility of alcoholic ONFH. Haplotype “GC” had a certain relationship with reducing the susceptibility of alcoholic ONFH. These results suggest that the risk of alcoholic ONFH is associated with variants L3MBTL3/PTPN9 gene in Chinese Han populations.

Alcoholic ONFH is a common hip disease, and it is relatively frequently disabled. Although much progress has been made in clinical manifestations and surgical treatment, further research is still needed. Intravascular thrombosis, abnormal lipid metabolism, and steroid sensitivity are considered to be the main causes of ONFH [29]. The L3MBTL3 gene encodes a malignant brain tumor (MBT) family. Its function is a methyl lysine reader, which recognizes methyl lysine residues in the histone tail and is related to the suppression of gene expression. By consulting a large number of documents, we learned the encoded protein could regulate hematopoiesis [9, 30]. Therefore, we speculate that the abnormal expression of L3MBTL3 may cause abnormalities in regulating hematopoiesis and further lead to the formation of intravascular thrombosis and eventually the formation of alcoholic ONFH. Studies have shown that Notch signaling pathway plays an important role in bone development and homeostasis, and serious bone diseases can be attributed to changes in Notch signaling pathway [31]. As a general modulator of Notch signaling pathway [32], L3MBTL3 will cause abnormal expression of Notch signaling pathway, which will further affect bone development, osteoporosis, and alcoholic ONFH. Existing studies have shown that L3MBTL3 gene polymorphisms are associated with various diseases, such as breast cancer [33], multiple sclerosis [10], and children’s specific growth cycle [34]. However, the rs1125970, rs4897367, and rs2068957 in the L3MBTL3 gene we studied have not been reported before. Our results also show that the L3MBTL3 polymorphism is associated with the risk of alcoholic ONFH, which is also consistent with our previous hypothesis.

The protein encoded by PTPN9 belongs to the family of protein tyrosine phosphatases (PTP) [35]. PTPs are considered to be signaling molecules that regulate various cellular processes, including cell growth, differentiation, mitotic cycle, and oncogenic transformation [36]. PTPN9 is found to be activated by inositol polyphosphate peptides and is thought to be involved in the signaling events that regulates phagocytosis and has been reported to play an important role in promoting vesicle fusion of hematopoietic cells in endocrine fluid [35]. Studies have shown that PTPN9 controls the expansion of erythroid cells [17] and regulates embryonic development [18]. PTPN9 negatively regulates vascular endothelial growth factor receptor signaling and endothelial cell function, which is closely related to intravascular thrombosis [37]. Intravascular thrombosis is considered to be one of the main causes of ONFH [29]. Therefore, we speculate that the abnormality of PTPN9 will cause the abnormality of hematopoietic function and red blood cells and then cause the formation of intravascular thrombosis. There are only two articles about PTPN9 gene association analysis [38, 39], but three SNPs (rs76107647, rs10851882, and rs75393192) in the PTPN9 gene we studied have not been reported before. Our research results showed that rs75393192 in the PTPN9 gene was found to be associated with the risk of alcohol-reducing ONFH in a subgroup analysis older than 40 years and staging and the PTPN9 gene GC haplotype was also associated with the risk of reducing alcoholic ONFH.

There are some potential limitations in this study. Firstly, this study has limitations of sample selection range and number. The sample size of this study is small, and a larger sample size is needed to verify the impact of L3MBTL3/PTPN9 genetic polymorphism on risk of alcoholic ONFH. Secondly, the test population in this study is limited to the Chinese Han population, so other populations are needed to further verify the results of this study. Thirdly, this study only found that L3MBTL3/PTPN9 polymorphisms were related to risk based on DNA level. Despite these limitations, our study is the first to demonstrate that L3MBTL3/PTPN9 gene polymorphism is associated with alcoholic ONFH, which provides new insights into the association between L3MBTL3/PTPN9 gene and alcoholic ONFH. This study provides a theoretical basis for further research on the function of L3MBTL3/PTPN9 gene in the occurrence and development of alcoholic ONFH and provides a certain direction for screening high-risk groups of alcoholic ONFH.

Conclusion

In summary, our study showed that L3MBTL3/PTPN9 polymorphism was related to the susceptibility of alcohol-induced ONFH in Chinese Han population. This study provides new insights for the identification, diagnosis, and treatment of alcohol-induced ONFH high-risk groups.