Genetic Polymorphisms of Delta-Like 1 Homolog Influence the Susceptibility to Antituberculosis Drug-Induced Hepatotoxicity

Abstract

This study was designed to investigate the relationship between Delta-like 1 homolog (DLK1) polymorphisms and the occurrence of antituberculosis drug-induced hepatotoxicity (ATDH) in the Western Chinese Han population. A total of 746 tuberculosis patients including 118 ATDH cases and 628 non-ATDH cases were enrolled from West China Hospital of Sichuan University during 2016–2018. Ten single nucleotide polymorphisms (rs11160604, rs7149242, rs7141210, rs7155375, rs876374, rs57098752, rs2400940, rs12431758, rs4900472, and rs6575802) within DLK1 were studied by the improved multiplex ligation detection reaction method genotyping technology assay. It was found that G allele of rs11160604 was associated with an increased risk for ATDH (p = 0.001) and G allele of rs4900472 showed a protective effect for ATDH (p = 0.030). Recessive model and dominant model of rs11160604 were observed as a risk factor for ATDH predisposition, whereas the recessive model of rs4900472 was a protective one. Moreover, the interaction genetic model composed of rs11160604, rs57098752, and rs12431758 showed a combined effect for the occurrence of ATDH. Our finding was a novel one indicating that the G allele of DLK1 rs11160604 might serve as a hazard for the development of ATDH in the Western Chinese Han population.

Introduction

Tuberculosis (TB) is a contagious disease caused by Mycobacterium tuberculosis (MTB) infection, which is a serious disease to human health (Glaziou et al., 2018). Globally, an estimated 10.0 million people fell ill with TB in 2019, a number that has been declining very slowly in recent years. There were an estimated 1.2 million TB deaths among HIV-negative people in 2019, and an additional 208 000 deaths among HIV-positive people. China accounted for 8.4% of the global TB in 2019, ranking third among the 30 countries with the highest TB burden (WHO, 2020). The anti-TB treatment program, as the key effective means to block the spread of TB, is accompanied by a very serious adverse drug reaction (ADR). Antituberculosis drug-induced hepatotoxicity (ATDH), considered the most severe ADR, is reported to occur at the rate of 3%–28% (Ahmad et al., 2018; Tweed et al., 2018). It is the main cause of the interruption of anti-TB treatment, which could result in a negative impact including noncompliance to treatment, emergence of drug-resistant MTB strains, and even death.

The possible pathogenic mechanisms of ATDH include the production of toxic metabolites, oxidative stress, mitochondrial dysfunction, and immune response (Perwitasari et al., 2015; Metushi et al., 2016; Ramachandran et al., 2018). However, the exact mechanism of ATDH has not been fully clarified. The influence of single nucleotide polymorphisms (SNPs), which serve as third-generation genetic markers, have been found to contribute to susceptibility to ATDH. As reported, various genes such as cytochrome enzyme P450 2E1 (CYP2E1), GST enzymes (GSTM1 and GSTT1), N-acetyltransferase 2 (NAT2), pregnane X receptor (PXR), nuclear factor-kappa B (NF-κB), and superoxide dismutases (SODs) have been widely investigated to clarify their potential roles in ATDH (Khan et al., 2019; Santos et al., 2019; Wu et al., 2019).

It is worth noting that obesity and inflammation may be possible risk factors for drug-induced hepatotoxicity (Dickmann et al., 2012). Delta-like 1 homolog (DLK1), a member of the epidermal growth factor-like homeotic family, is associated with adipogenesis inhibition, hepatocytes differentiation, and closely related to the occurrence and development of hepatic carcinoma (Xu et al., 2012). In addition, DLK1 served as a new target and contributor to early liver fibrogenesis and cirrhosis (Pan et al., 2011). As the noncanonical Notch ligand, DLK1 may participate in the Notch signaling pathway to regulate the immune response on hepatocyte. Notch4 could negatively regulate inflammatory damage by inhibiting TAK1 activation to MTB infection (Zheng et al., 2018). Therefore, we speculated that the activated DLK1 be an indispensable element in the development of ATDH by regulating lipid metabolism and inflammatory response.

To investigate the possible relationship between DLK1 and ATDH susceptibility in TB patients, a set of 10 SNPs within DLK1 were selected to assess the role of gene variants with ATDH in a cohort of the Chinese Han population. This study may provide useful information of DLK1 signaling pathway on ATDH early identification of high-risk TB patients, to reduce the incidence of liver injury.

Materials and Methods

Study population

This study was approved by the Ethics Committee of West China Hospital [No. 198 (2014)]. As reported in the previous article, all samples in this study were from the Bio-Bank of resources of TB Researches in the Department of Laboratory Medicine, West China Hospital of Sichuan University in China. A total of 1060 highly suspected TB patients were enrolled in the study between December 2016 and April 2018 in West China hospital of Sichuan University. All patients were newly diagnosed with TB and none of them had previously received anti-TB treatment. Enrolled patients had typical TB symptoms and signs and conformed to at least one of the following: (1) more than two times of acid-fast staining smear positive or MTB culture positive or TB-DNA positive; (2) bacteriological evidence, such as smear or culture, was negative, but imaging examination, such as computerized tomography (CT), supported the typical manifestation of active TB; (3) the pathological diagnosis supports TB; and (4) more than two times of interferon-gamma release assay positive. The exclusion criteria were as follows: (1) respiratory diseases other than TB, such as pulmonary infection, chronic obstructive pulmonary disease, lung tumors, and so on; (2) autoimmune diseases, cancer, hepatitis virus infection, HIV infection, patients with immunosuppressants or enhancers, and pregnancy; (3) patients with definite abnormal liver function before anti-TB treatment or patients with severe liver function damage caused by recent drug use. Based on typical TB symptoms, microbiological and imaging findings, and exclusion criteria, 746 eligible TB patients were enrolled eventually.

All included TB patients in the study were in strict accordance with the first-line anti-TB regimen consisting of two consecutive stages of treatment: daily treatment of isoniazid, rifampin, and ethambutol, pyrazinamide for the first 2 months, followed by the daily treatment of isoniazid, rifampin, and ethambutol for at least 4 months; the total course is equal to or more than 6 months. Demographic and clinical data of the participants were obtained from their medical records. Before initiating anti-TB treatment, the baseline laboratory indicators were collected. After treatment, liver function indicators were performed twice a month in the first 2 months and monthly in the remaining 4 months. The peak values of laboratory indicators during the treatment were recorded to assess ATDH.

Ultimately, according to the definition of ATDH assigned by Common Toxicity Criteria for Adverse Events v. 4.0 (CTCAE v.4.0), 118 patients with ATDH, and 628 patients without ATDH were recruited. In detail, ATDH was defined as a level of alanine aminotransferase (ALT) and/or aspartate aminotransferase (AST) of at least three times the upper limited number (ULN) accompanied by symptoms of nausea, vomiting and abdominal pain, and/or a combined increase in total bilirubin (TBIL) greater than two times the ULN; elevation of serum AST and/or ALT level greater than five times the ULN with or without symptoms. The ULN for AST was 40 IU/L, 50 IU/L for ALT, and 28 μmol/L for TBIL.

Sample collection and DNA extraction

Peripheral blood 3–5 mL were collected from all subjects for DNA extraction and genomic DNA was extracted using the QIAamp DNA Blood Midi Kit (Qiagen, Germany) according to the manufacturer's protocol. The concentration and purity of extracted DNA samples were determined by NanoVue Plus microspectrophotometer at a wavelength of 260/280 nm. The DNA was stored at −80°C until further SNP genotyping analysis.

Candidate polymorphisms selection and genotyping

DLK1 gene, located on human chromosome 14q32, is an imprinted gene with paternal expression and maternal silencing. The principles of screening SNP candidates in the highly polymorphic DLK1 gene are through the following steps: (1) All the candidate SNPs were obtained from the dbSNP database. (2) Tag single nucleotide polymorphisms (tagSNPs) with a minor allele frequency (MAF) >0.05 in Southern Han Chinese, and then tagSNPs were filtered by pairwise tagger method (threshold: r ² > 0.8); (3) the functional regulatory SNPs in exon region, promoter region, or untranslated region are preferentially included.

Finally, 10 SNPs within DLK1 (rs11160604, rs7149242, rs7141210, rs7155375, rs876374, rs57098752, rs2400940, rs12431758, rs4900472, and rs6575802) were eventually included and genotyped.

Statistical analysis

Statistical analysis was performed using SPSS software (version 22.0; SPSS, Inc., Chicago, IL). Continuous variables for clinical characteristics and laboratory biomarkers of the ATDH group and the non-ATDH group were presented as mean ± SD or median with interquartile range as appropriate. Significant differences were analyzed using Student's t-test or Mann–Whitney U-test as appropriate for comparison between the two groups (two-tailed, p < 0.05).

The distributions of allele, genotype and genetic models of DLK1 SNPs in TB patients were analyzed by PLINK v1.90 software. Odds ratios (ORs) and corresponding 95% confidence intervals (CIs) were calculated to evaluate the ATDH risk. We constructed the linkage disequilibrium (LD) and haplotype construction to figure out whether there were additive associations among the selected SNPs at a threshold of pairwise r ² > 0.80 using Haploview v4.2 software. The Generalized Multifactor Dimensionality Reduction (GMDR) v0.9 software was adopted to explore potential interaction models associated with ATDH risk.

Results

Clinical characteristics of TB patients

A total of 746 active TB patients in Western Chinese Han population were included in this study. The demographic characteristics of 118 patients in ATDH group and 628 patients in non-ATDH group are given in Table 1. There were 69 men (58.47%) and 49 women (41.53%) in ATDH group. The mean age was 42.85 ± 18.44 years. In the non-ATDH group, there were 375 men (59.71%) and 253 women (40.29%). The mean age was 40.92 ± 15.72 years. There were no statistically significant differences in gender composition, average age, and distribution of smoking and drinking between the two groups. Although patients with ATDH were more susceptible to fever (p = 0.013) and weight loss (p = 0.024), there were no statistically significant differences in the distribution of night sweats, poor appetite, and fatigue between ATDH group and non-ATDH group (p > 0.05).

Table 1.

The Demographic, Clinical Characteristics and Laboratory Indicators of the Tuberculosis Patients

Characteristics	ATDH	Non-ATDH	p
Characteristics	(n = 118)	(n = 628)	p
Clinical features
Age, mean ± SD (years)	42.85 ± 18.44	40.92 ± 15.72	0.234
Sex, male/female	69/49	375/253	0.801
Smoking, n (%)	38 (32.20)	221 (35.19)	0.532
Drinking, n (%)	32 (27.12)	141 (22.45)	0.267
General symptoms, n (%)
Fever	68 (57.73)	284 (45.22)	0.013
Weight loss	49 (41.53)	254 (40.45)	0.024
Night sweat	36 (30.51)	261 (41.56)	0.394
Poor appetite	32 (27.12)	195 (31.05)	0.827
Fatigue	33 (27.97)	166 (26.43)	0.730
Laboratory results
Erythrocyte ( × 10¹²/L)	4.31 ± 0.74	4.28 ± 0.68	0.481
Hemoglobin (g/L)	122.87 ± 22.11	122.06 ± 20.58	0.717
Platelet ( × 10⁹/L)	236 (184–321)	232 (172–297)	0.134
Leukocyte ( × 10⁹/L)	7.24 ± 3.36	7.05 ± 2.90	0.761
Neutrophil ( × 10⁹/L)	5.23 ± 2.89	5.10 ± 2.73	0.631
Monocyte ( × 10⁹/L)	1.29 ± 0.79	1.26 ± 0.62	0.625
Lymphocyte ( × 10⁹/L)	0.55 ± 0.29	0.50 ± 0.25	0.099
TBIL (μmol/L)	10.05 (7.50–14.13)	8.70 (6.30–12.10)	0.002
ALT (IU/L)	28.00 (15.75–38.00)	15.00 (10.00–21.00)	<0.001
AST (IU/L)	27.00 (20.00–34.00)	19.50 (16.00–25.00)	<0.001
ALB (g/L)	38.64 ± 7.35	37.89 ± 6.90	0.248
ALP (IU/L)	85.50 (68.50–106.00)	79.00 (64.00–98.00)	0.021
GGT (IU/L)	42.50 (26.00–78.00)	29.00 (19.00–48.00)	<0.001
URIC (μmol/L)	291.29 ± 125.98	331.51 ± 155.30	0.008
C-reactive protein (mg/L)	9.74 (2.30–39.23)	12.25 (2.67–37.43)	0.961
Erythrocyte sedimentation rate (mm/h)	38.50 (20.50–63.00)	33.50 (14.75–64.00)	0.173

Data are given as mean ± SD or median with (Q_1–Q₃). Values in bold are significant.

ALB, albumin; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ATDH, antituberculosis drugs-induced hepatotoxicity; GGT, gamma-glutamyl transpeptidase; TBIL, total bilirubin; URIC, uric acid.

In terms of baseline analysis of laboratory indicators, several indicators related to liver function including TBIL, ALT, AST, alkaline phosphatase, and gamma-glutamyl transpeptidase were significantly higher in patients with ATDH than those in the non-ATDH group before the anti-TB treatment (p < 0.05), although both groups were in the normal ranges. During anti-TB treatment, the ALT and AST of ATDH patients were significantly increased, whereas the laboratory indicators of non-ATDH patients were not significantly changed or only slightly increased (Supplementary Table S1). Moreover, uric acid was decreased significantly in patients with ATDH (p = 0.008). No statistical differences were presented in the levels of erythrocyte, hemoglobin, platelets, leukocytes, C-reactive protein, and erythrocyte sedimentation rate in patients with ATDH compared with patients with non-ATDH.

Distribution of allelic and genotypic frequency of DLK1

All 118 patients in the ATDH group and 628 patients in the non-ATDH group were conducted for SNPs genotyping. The information on genotype SNPs is given in detail in Supplementary Table S2, according to chromosome location, molecular results, MAF, and a p-value of Hardy–Weinberg equilibrium (HWE) test. All the SNP genotype distributions followed HWE (p > 0.05 for all loci) (Supplementary Table S2).

The results for the rs11160604 SNP in DLK1 showed statistical significance. The proportion of the minor allele (G allele) was 21.61% in the ATDH group and 13.36% in the non-ATDH group compared with the A allele. G allele of rs11160604 was associated with higher susceptibility for ATDH (OR = 1.77, 95% CI = 1.25–2.51, p = 0.001). A parallel result was found in the frequency of genotype rs11160604 GG, which indicated a significant risk association (p = 0.005). However, G allele of rs4900472 showed a protective effect for ATDH (OR = 0.74, 95% CI = 0.56–0.98, p = 0.03). No significant difference was observed in the allele and genotype distributions of the other eight SNPs of the DLK1 (rs7149242, rs7141210, rs7155375, rs876374, rs57098752, rs2400940, rs12431758, and rs6575802) between the case and control groups. The distributions of the allelic and genotypic frequency of DLK1 for all 10 SNPs between cases and control are given in Table 2.

Table 2.

The Distribution of Allelic and Genotypic Frequency of DLK1 Between Cases and Controls

Gene	SNP	Group	Allele		Allele, n (%)		OR (95% CI)	p	Genotype, n (%)			p
Gene	SNP	Group	1	2	1	2	OR (95% CI)	p	11	12	22	p
DLK1	rs7149242	Case	A	C	99 (41.95)	137 (58.05)	0.92 (0.69–1.22)	0.560	21 (17.80)	57 (48.31)	40 (33.9)	0.841
		Control			550 (44.00)	700 (56.00)	0.92 (0.69–1.22)	0.560	121 (19.36)	308 (49.28)	196 (31.36)	0.841
	rs7141210	Case	G	A	68 (28.81)	168 (71.19)	0.86 (0.63–1.17)	0.337	7 (5.93)	54 (45.76)	57 (48.31)	0.400
		Control			401 (31.98)	853 (68.02)	0.86 (0.63–1.17)	0.337	62 (9.89)	277 (44.18)	288 (45.93)	0.400
	rs7155375	Case	G	A	85 (36.02)	151 (63.98)	0.95 (0.73–1.29)	0.828	16 (13.56)	53 (44.92)	49 (41.53)	0.950
		Control			461 (36.76)	793 (63.24)	0.95 (0.73–1.29)	0.828	85 (13.56)	291 (46.41)	251 (40.03)	0.950
	rs876374	Case	A	C	58 (24.58)	178 (75.42)	0.95 (0.69–1.31)	0.760	8 (6.78)	42 (35.59)	68 (57.63)	0.914
		Control			320 (25.52)	934 (74.48)	0.95 (0.69–1.31)	0.760	42 (6.70)	236 (37.64)	349 (55.66)	0.914
	rs57098752	Case	A	G	110 (46.61)	126 (53.39)	1.23 (0.93–1.61)	0.141	24 (20.34)	62 (52.54)	32 (27.12)	0.214
		Control			519 (41.45)	733 (58.55)	1.23 (0.93–1.61)	0.141	115 (18.37)	289 (46.17)	222 (35.46)	0.214
	rs11160604	Case	G	A	51 (21.61)	185 (78.39)	1.77 (1.25–2.51)	0.001	6 (5.08)	39 (33.05)	73 (61.86)	0.005
		Control			167 (13.36)	1083 (86.64)	1.77 (1.25–2.51)	0.001	12 (1.92)	143 (22.88)	470 (75.20)	0.005
	rs2400940	Case	G	A	57 (24.15)	179 (75.85)	1.34 (0.97–1.85)	0.065	8 (6.78)	41 (34.75)	69 (58.47)	0.196
		Control			237 (18.93)	1015 (81.07)	1.34 (0.97–1.85)	0.065	27 (4.31)	183 (29.23)	416 (66.45)	0.196
	rs12431758	Case	A	G	111 (47.03)	125 (52.97)	1.11 (0.84–1.46)	0.550	23 (19.49)	65 (55.08)	30 (25.42)	0.293
		Control			557 (44.42)	697 (55.58)	1.11 (0.84–1.46)	0.550	129 (20.57)	299 (47.69)	199 (31.74)	0.293
	rs4900472	Case	G	A	89 (38.03)	145 (61.97)	0.74 (0.56–0.98)	0.030	16 (13.68)	57 (48.72)	44 (37.61)	0.088
		Control			573 (45.69)	681 (54.31)	0.74 (0.56–0.98)	0.030	138 (22.01)	297 (47.37)	192 (30.62)	0.088
	rs6575802	Case	A	C	52 (22.22)	182 (77.78)	0.92 (0.65–1.29)	0.637	3 (2.56)	46 (39.32)	68 (58.12)	0.439
		Control			297 (23.65)	959 (76.35)	0.92 (0.65–1.29)	0.637	33 (5.25)	231 (36.78)	364 (57.96)	0.439

Values in bold are significant.

CI, confidence interval; OR, odd ratio; SNPs, single nucleotide polymorphisms.

Next, we used three genetic model analyses, including additive model, dominant model, and recessive model, to figure out the differences in genotype distribution of each SNP. As given in Table 3, recessive model and dominant model of rs11160604 were significantly associated with a higher risk for ATDH by an estimated OR of 2.74 (p = 0.049, 95% CI = 1.01–7.44) and 1.87 (p = 0.003, 95% CI = 1.24–2.83), respectively. We further discovered that the recessive model of rs4900472 had a decreased association with the occurrence of ATDH (recessive model: OR = 0.56; 95% CI = 0.32–0.98, p = 0.043). The genetic models of other SNPs (rs7149242, rs7141210, rs7155375, rs876374, rs57098752, rs2400940, rs12431758, and rs6575802) of DLK1 were not found to be related to the occurrence of ATDH.

Table 3.

Analysis of DLK1 Genetic Model in Relation to Antituberculosis Drug-Induced Hepatotoxicity Risk in Tuberculosis Patients

Group	Gene	SNP	Additive model		Recessive model		Dominant model
Group	Gene	SNP	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Patient vs. control	DLK1	rs7149242	0.97 (0.73–1.30)	0.860	0.90 (0.54–1.50)	0.692	0.89 (0.59–1.35)	0.587
		rs7141210	1.10 (0.98–1.40)	0.700	0.57 (0.26–1.29)	0.179	0.91 (0.61–1.35)	0.636
		rs7155375	1.03 (0.85–1.24)	0.760	1.00 (0.56–1.78)	0.999	0.94 (0.63–1.40)	0.762
		rs876374	0.91 (0.68–1.22)	0.530	1.01 (0.46–2.22)	0.974	0.92 (0.62–1.37)	0.693
		rs57098752	1.02 (0.83–1.25)	0.880	1.14 (0.69–1.86)	0.615	1.48 (0.95–2.29)	0.081
		rs11160604	0.97 (0.73–1.30)	0.860	2.74 (1.01–7.44)	0.049	1.87 (1.24–2.83)	0.003
		rs2400940	0.91 (0.68–1.22)	0.530	1.61 (0.71–3.64)	0.250	1.41 (0.94–2.10)	0.096
		rs12431758	1.10 (0.98–1.40)	0.700	0.93 (0.57–1.53)	0.789	1.36 (0.87–2.13)	0.174
		rs4900472	1.02 (0.83–1.25)	0.880	0.56 (0.32–0.98)	0.043	0.73 (0.49–1.10)	0.137
		rs6575802	1.03 (0.85–1.24)	0.760	0.47 (0.14–1.57)	0.223	0.99 (0.67–1.48)	0.975

Values in bold are significant.

LD analysis and haplotype construction

We further performed haplotype analysis on the target SNPs to explore the correlation between SNPs and the risk of ATDH with a threshold of pairwise r ² > 0.80. Unfortunately, no haplotype was successfully established using these SNPs analyzed in our study. The loci in the DLK1 gene within the LD block are given in Supplementary Figure S1.

Interaction model

The interaction models of 10 DLK1 polymorphic loci were analyzed by GMDR analysis for training/testing balance accuracy and cross-validation consistency. The model composed of three SNPs (rs11160604, rs57098752, and rs12431758) had a significant interaction effect, with testing balance accuracy of 0.550 and cross-validation consistency of 8/10 at p = 0.011 (Table 4).

Table 4.

Interactive Models of DLK1 Genetic Variants Adopting the Generalized Multifactor Dimensionality Reduction Method

Model	Training balance accuracy	Testing balance accuracy	Cross-validation consistency	p
rs11160604	0.567	0.559	10/10	0.055
rs57098752 rs12431758	0.605	0.557	9/10	0.055
rs57098752 rs11160604 rs12431758	0.643	0.550	8/10	0.011

Sign p-value was calculated by GMDR v0.9 software.

GMDR, generalized multifactor dimensionality reduction.

The high-risk and low-risk genotype combinations represented by the combination of three factors (rs11160604, rs57098752, and rs12431758) are given in Figure 1. Each box represented one pattern of three-locus interaction combination, where dark gray represented the high-risk factor combinations, and the light gray implicated the low-risk genotype combinations. Furthermore, we identified two statistically significant high-risk genotype combinations using the chi-square test, the ORs of which for the (GA)-(GA)-(AA) and (GA)-(GA)-(GA) were 3.43 (95% CI = 1.10–10.67) and 4.31 (95% CI = 2.24–8.30) in this three-loci (rs11160604–rs57098752–rs12431758) model (all p < 0.05).

FIG. 1.

Classification chart of the three SNPs (rs57098752, rs11160604, and rs12431758) combination with high risk or low risk for ATDH. Dark squares represent high-risk groups, light-colored squares represent low-risk groups. The bars on the left represent the positive score of the combination, and right-hand bars represent negative score of the combination. ATDH, antituberculosis drugs-induced hepatotoxicity; SNPs, single nucleotide polymorphisms.

Discussion

Considerable evidence showed that the genetic mutation of related genes may be involved in ATDH pathogenesis. As recent literature reports the close relationship between DLK1 and hepatocyte function, we aimed to figure out whether DLK1 plays an effective risk prediction marker for the occurrence and development of ATDH. We performed genotype analysis, haplotype construction, and gene interaction analysis of DLK1 SNPs to explore the association between SNPs and the occurrence risk of ATDH in Western Chinese Han population. The results showed that there were significant differences in the allele distribution of rs11160604 and rs4900472 in the DLK1 gene between the case group and the control group. G allele of rs11160604 was associated with an increased risk for ATDH (OR = 1.77, 95% CI = 1.25–2.51, p = 0.001), whereas G allele of rs4900472 played a protective role for ATDH (OR = 0.74, 95% CI = 0.56–0.98, p = 0.03). In addition, dominant model of rs11160604 showed high risk for ATDH predisposition (OR = 1.87, 95% CI = 1.24–2.83, p = 0.003). The results suggested that the G allele of DLK1 rs11160604 served as a hazard for the development of ATDH in the Western Chinese Han population.

Although there were no significant differences in the allele distribution of rs57098752 and rs12431758 in the DLK1 gene, the interaction genetic model composed of rs11160604, rs57098752, and rs12431758 showed a higher risk for occurrence of ATDH with an increased OR than those observed in rs11160604 analyses. Our results indicated that the three SNP loci have synergistic interaction; individuals with rs11160604 GA genotype, rs57098752 GA genotype, and rs12431758 GA genotype were at the highest risk for ATDH, which had significant clinical implications for the prediction of the risk of ATDH and the preventive intervention of ATDH.

To my knowledge, current reports on DLK1 gene polymorphism mainly focus on imprinting mechanism of tumors and animals' lipid metabolism, and there are no literature reports regarding the relationship between DLK1 polymorphism and human ATDH. An SNP at T852C (rs2295660) of DLK1 was discovered to be monoallelic specific, not biallelic in hepatocellular carcinoma (Huang et al., 2007). DLK1 can affect lipid metabolism in bovine and two SNPs, IVS3 + 478 C > T and IVS3 + 609 T>G, might be applied as genetic markers of meat quality traits for beef cattle breeding (Wang et al., 2020). In addition, polar overdominant inheritance of a DLK1 polymorphism is associated with growth and fatness in pigs (Kim et al., 2004). Our research was a novel one suggesting that the DLK1 rs11160604 might serve as a hazard, whereas DLK1 rs4900472 was a protective one for the development of ATDH in the Western Chinese Han population. These results would fill the gap of the correlation between polymorphisms in DLK1 genes and ATDH.

DLK1 is a noncanonical ligand of Notch signaling pathway that has a high homology with Notch ligand Delta. In Notch signaling pathway, DLK1 competitively binds to Notch receptor and plays a negative regulatory role. As reported, DLK1 is involved in the proliferation and differentiation of many kinds of cells and is closely related to the occurrence and development of tumors. In recent years, DLK1 has been further found to be closely related to adipogenesis (Espina et al., 2009), hematopoiesis (Schneider et al., 2016), and liver fibrosis (Qiu et al., 2014).

The occurrence and development of ATDH is closely related to abnormal lipid metabolism, liver fibrosis, and abnormal immune regulation, while DLK1 rightly happens to be involved with them.

As an inhibitor of adipogenesis, gene polymorphism of DLK1 is related to human obesity. DLK1 plays a direct role in reducing the size of all adipose tissues in transgenic mice with adipocyte-specific overexpression of DLK1 (Lee et al., 2003). In addition, DLK1 ^-/- mice were associated with an increased adiposity and development of fatty liver, which resulted in an increased expression of lipogenic enzymes, Fas and Scd1 (Moon et al., 2002). These reports showed that there was a relationship between DLK1 and the process of lipid metabolism and hepatocyte injury, suggesting a role for DLK1 as a promising molecular basis underlying ATDH onset and progress.

Moreover, DLK1-associated liver fibrosis may be involved in the abnormal hepatic function. DLK1 was strongly induced and initially expressed in hepatocytes and then released into the hepatic stellate cells (HSCs) in a paracrine manner during liver injury (Pan et al., 2011). In addition, DLK1 is expressed selectively in HSCs in the adult rodent liver and induced in liver fibrosis and regeneration (Zhu et al., 2012). Although the mechanism of DLK1 pathway in ATDH remains to be further elucidated, it can be assumed that DLK1 is really involved in mediating hepatocyte fibrosis through certain kinds of signaling pathway.

On the contrary, DLK1 may have a potential function in inflammatory response to the development and progress of ATDH. Notch ligand delta-like ligand 4 (DLL4) can influence T cell cytokine secretion in both humans and mice, and an overexpression of DLL4 was upregulated in early hematopoietic progenitors in response to chronic mycobacterial infection (Schaller et al., 2016). In addition, recombinant Delta-like ligand 4 protein (rDll4) ameliorated hepatocyte apoptosis, inflammation, and fibrosis by downregulating cytokine expression to protect liver function from inflammatory damage. Meanwhile, rDll4 significantly decreased lipopolysaccharide-stimulated chemokine expression in both Kupffer and HSCs (Shen et al., 2016). Furthermore, DLK1 can suppress the malignant behaviors of hepatocellular carcinoma cells by directly disrupting cancer stem/progenitor cells through the inhibition of siRNA (Xu et al., 2012). DLK1⁺ cell was found to have a stronger regeneration capacity and DLK1 was highly expressed in human HCC tissue compared with normal liver and paracancerous tissues (Luo et al., 2006). Therefore, it was suggested that DLK1 could play a vital part in the immune regulation of hepatocyte. Based on all the above information, our findings may provide the evidence of hepatocyte injury associated with DLK1in ATDH.

There are some limitations in our study. First, this study only analyzed 10 SNPs; more SNPs that may be related to the occurrence and development of ATDH can be further analyzed. Moreover, other relevant genes, environmental risks, and individual conditions should also be included for a more accurate and comprehensive evidence for the relationship between DLK1 gene polymorphism and ATDH. Second, the results were based on ATDH induced by first-line anti-TB regimens without the evaluation of the second-line drugs. The next step is to consider using larger population samples from other ethnic groups to confirm the current findings.

Conclusions

We found that genetic polymorphisms in rs11160604 in DLK1 were associated with susceptibility to ATDH in Western Chinese Han population. Our research revealed the correlation between DLK SNPs and ATDH, providing a certain basis for screening new genetic biomarkers for ATDH, and more experiments are needed to explore the mechanisms of DLK1 pathway in ATDH development and progression.

Consent

Informed consent was obtained from all individual participants included in the study.

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the National Natural Science Foundation of China [81472026, 81672095 and 81672096] and the National Science and Technology Pillar Program during the 13th Five-year Plan Period [2018ZX10715003].

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Figure S1

References

Ahmad

, Ahuja

S.D.

, Akkerman

O.W.

, Alffenaar

J.C.

, Anderson

L.F.

, Baghaei

, et al. (2018). Treatment correlates of successful outcomes in pulmonary multidrug-resistant tuberculosis: an individual patient data meta-analysis. Lancet, 392, 821–834.

Dickmann

L.J.

, VandenBrink

B.M.

, and Lin

Y.S.

(2012). In vitro hepatotoxicity and cytochrome P450 induction and inhibition characteristics of carnosic acid, a dietary supplement with antiadipogenic properties. Drug Metab Dispos, 40, 1263–1267.

Espina

A.G.

, Méndez-Vidal

, Moreno-Mateos

M.A.

, Sáez

, Romero-Franco

, Japón

M.A.

, et al. (2009). Induction of Dlk1 by PTTG1 inhibits adipocyte differentiation and correlates with malignant transformation. Mol Biol Cell, 20, 3353–3362.

Glaziou

, Floyd

, and Raviglione

M.C.

(2018). Global epidemiology of tuberculosis. Semin Respir Crit Care Med, 39, 271–285.

Huang

, Zhang

, Zhu

J.D.

, Zhang

Y.L.

, Lin

, et al. (2007). Up-regulation of DLK1 as an imprinted gene could contribute to human hepatocellular carcinoma. Carcinogenesis, 28, 1094–1103.

Khan

, Mandal

R.K.

, Elasbali

A.M.

, Dar

S.A.

, Jawed

, Wahid

, et al. (2019). Pharmacogenetic association between NAT2 gene polymorphisms and isoniazid induced hepatotoxicity: trial sequence meta-analysis as evidence. Biosci Rep, 39, BSR20180845.

Kim

K.S.

, Kim

J.J.

, Dekkers

J.C.

, and Rothschild

M.F.

(2004). Polar overdominant inheritance of a DLK1 polymorphism is associated with growth and fatness in pigs. Mamm Genome, 15, 552–559.

Lee

, Villena

J.A.

, Moon

Y.S.

, Kim

K.H.

, Lee

, Kang

, et al. (2003). Inhibition of adipogenesis and development of glucose intolerance by soluble preadipocyte factor-1 (Pref-1). J Clin Invest, 111, 453–461.

Luo

J.H.

, Ren

, Keryanov

, Tseng

G.C.

, Rao

U.N.

, Monga

S.P.

, et al. (2006). Transcriptomic and genomic analysis of human hepatocellular carcinomas and hepatoblastomas. Hepatology, 44, 1012–1024.

10.

Metushi

, Uetrecht

, and Phillips

(2016). Mechanism of isoniazid-induced hepatotoxicity: then and now. Br J Clin Pharmacol, 81, 1030–1036.

11.

Moon

Y.S.

, Smas

C.M.

, Lee

, Villena

J.A.

, Kim

K.H.

, Yun

E.J.

, et al. (2002). Mice lacking paternally expressed Pref-1/Dlk1 display growth retardation and accelerated adiposity. Mol Cell Biol, 22, 5585–5592.

12.

Pan

R.L.

, Wang

, Xiang

L.X.

, and Shao

J.Z.

(2011). Delta-like 1 serves as a new target and contributor to liver fibrosis down-regulated by mesenchymal stem cell transplantation. J Biol Chem, 286, 12340–12348.

13.

Perwitasari

D.A.

, Atthobari

, and Wilffert

(2015). Pharmacogenetics of isoniazid-induced hepatotoxicity. Drug Metab Rev, 47, 222–228.

14.

Qiu

, Zhou

, Zhai

, Jia

, and Zhou

(2014). Curcumin regulates delta-like homolog 1 expression in activated hepatic stellate cell. Eur J Pharmacol, 728, 9–15.

15.

Ramachandran

, Visschers

R.G.J.

, Duan

, Akakpo

J.Y.

, and Jaeschke

(2018). Mitochondrial dysfunction as a mechanism of drug-induced hepatotoxicity: current understanding and future perspectives. J Clin Transl Res, 4, 75–100.

16.

Santos

E.A.

, Gonçalves

J.C.S.

, Fleury

M.K.

, Kritski

A.L.

, Oliveira

M.M.

, Velasque

L.S.

, et al. (2019). Relationship of anti-tuberculosis drug-induced liver injury and genetic polymorphisms in CYP2E1 and GST. Braz J Infect Dis, 23, 381–387.

17.

Schaller

M.A.

, Allen

R.M.

, Kimura

, Day

C.L.

, and Kunkel

S.L.

(2016). Systemic Expression of notch ligand delta-like 4 during mycobacterial infection alters the T cell immune response. Front Immunol, 7, 527.

18.

Schneider

, Sellers

Z.P.

, and Ratajczak

M.Z.

(2016). Parentally imprinted genes regulate hematopoiesis-new evidence from the Dlk1-Gtl2 locus. Stem Cell Investig, 3, 29.

19.

Shen

, Liu

, Dewidar

, Hu

, Park

, Feng

, et al. (2016). Delta-like ligand 4 modulates liver damage by down-regulating chemokine expression. Am J Pathol, 186, 1874–1889.

20.

Tweed

C.D.

, Wills

G.H.

, Crook

A.M.

, Dawson

, Diacon

A.H.

, Louw

C.E.

, et al. (2018). Liver toxicity associated with tuberculosis chemotherapy in the REMoxTB study. BMC Med, 16, 46.

21.

Wang

, Jiang

, Yu

, Mi

, Bai

, Zhang

, et al. (2020). Analysis of the bovine DLK1 gene polymorphism and its relation to lipid metabolism in chinese simmental. Animals (Basel), 10, 923.

22.

WHO. Global Tuberculosis Report. 2020. Available at https://www.who.int/tb/publications/global_report/en/ Accessed October 21, 2020 .

23.

, Bai

, Zhao

, Wang

, Hu

, Jiao

, Q, et al. (2019). A prospective study on associations between superoxide dismutase gene polymorphisms and antituberculosis drug-induced liver injury in a Chinese Han population. J Gene Med, 21, e3121.

24.

, Liu

R.F.

, Zhang

, Huang

L.Y.

, Chen

, Fei

Q.L.

, and Han

Z.G.

(2012). DLK1 as a potential target against cancer stem/progenitor cells of hepatocellular carcinoma. Mol Cancer Ther, 11, 629–638.

25.

Zheng

, Liu

, Zhou

, Yan

, Chen

, Ma

, et al. (2018). Notch4 negatively regulates the inflammatory response to mycobacterium tuberculosis infection by inhibiting TAK1 activation. J Infect Dis, 218, 312–323.

26.

Zhu

N.L.

, Asahina

, Wang

, Ueno

, Lazaro

, Miyaoka

, et al. (2012). Hepatic stellate cell-derived delta-like homolog 1 (DLK1) protein in liver regeneration. J Biol Chem, 287, 10355–10367.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.04 MB

0.09 MB