Multiple genetic variants associated with posttransplantation diabetes mellitus in Chinese Han populations

Abstract

Objectives

Posttransplantation diabetes mellitus (PTDM) is a major complication after solid organ transplantation. This study is to investigate the association of nine genetic variant factors and PTDM in Chinese Han patients.

Methods

HLA‐DP (rs3077, rs9277535), HLA‐DQ (rs7453920), signal transducer and activator of transcription 4 (STAT4) (rs7574865), IL‐28B (rs12979860, rs8099917, and rs12980275), and IL‐18 (rs1946518 and rs187238) were investigated in 260 liver transplant recipients (PTDM vs non‐PTDM) by high‐resolution melting curve analysis. Serum interleukin (IL)‐1β, IL‐6, IL‐8, IL‐17, interferon‐γ, inducible protein‐10, monocyte chemoattractant protein‐1, and macrophage inflammatory protein‐1b were analyzed by a Bio‐Plex suspension array system (Bio‐Plex Multiplex Immunoassays, Bio‐Rad, Hercules, CA, USA).

Results

Signal transducer and activator of transcription 4 (rs7574865) T allele and IL‐18 (rs1946518) A allele increase the risk for insulin resistance and PTDM.

Conclusions

Recipients with STAT4 (rs7574865) T allele are associated with an increased concentration of IL‐1β, interferon‐γ, monocyte chemoattractant protein, and macrophage inflammatory protein‐1b. The genetic variants of STAT4 (rs7574865) and IL‐18 (rs1946518) may be new important markers for PTDM.

Abbreviations

AT

adipose tissue

BUN

urea nitrogen

CHOL

cholesterol

CNI

calcineurin inhibitor

GLU

glucose

GWAS

genome‐wide association studies

HbA1C

glycosylated hemoglobin A1C

HCC

hepatocellular carcinoma

HLA

human leukocyte antigen

HRM

high‐resolution melting

ICD

international classification of diseases

IL28B

interleukin 28B

IP‐10

interferon‐inducible protein 10

IRS

insulin receptor substrate

ISGs

IFN‐stimulated genes

MCP‐1

monocyte chemoattractant protein‐1

MFI

median fluorescence intensity

MIP‐1b

macrophage inflammatory protein‐1 beta

PTDM

posttransplantation diabetes mellitus

SA‐PE

streptavidin‐phycoerythrin

SNPs

single nucleotide polymorphisms

SOCS

suppressor of cytokine signaling

STAT4

signal transducer and activator of transcription 4

TG

triglyceride

1. INTRODUCTION

Posttransplantation diabetes mellitus (PTDM) is a frequent complication in patients treated with the immunosuppressive calcineurin inhibitor tacrolimus (Tac; FK506). Approximately 30%‐40% of liver transplant recipients have sustained diabetes beyond 6 months after transplant.1 The true prevalence of insulin resistance in patients after liver transplantation is unknown but presumed to be even greater. It is important to find biological markers that can predict the risk of PTDM; this could also provide new ideas for individualized therapies. Recent studies have identified single nucleotide polymorphisms (SNPs) associated with PTDM risk among Tac‐treated patients.2 It has been suggested that genotyping diabetes‐related polymorphisms could be a possible method of predicting a patient's risk for developing PTDM, and this information could be a valuable asset in the selection of appropriate immunosuppressive regimens.2

Posttransplantation diabetes mellitus is characterized by a combination of insulin resistance and insulin hyposecretion in liver transplant recipients. It has been reported that there is a particularly high prevalence of insulin resistance associated with viral epidemic diseases.3 Although the mechanism by which virus infection promotes insulin resistance is not fully understood, viral eradication has been shown to improve insulin sensitivity. Chronic hepatitis C virus (HCV) infection has been reported to be associated with a high prevalence of insulin resistance, both in the pre‐ and posttransplant setting. In addition, sustained virological response (SVR) to antiviral treatment has been shown to improve insulin sensitivity. It was demonstrated that genetic IL28B polymorphisms coding for IFN‐λ are strongly associated with SVR to pegIFN‐α treatment for chronic HCV infection. Because IFN‐λ, in analogy with IFN‐α, may act on IFN‐stimulated genes (ISGs) via suppressor of cytokine signaling (SOCS) and the JAK‐STAT pathway, IL28B polymorphisms might also be associated with development of diabetes mellitus (DM) in HCV patients during long‐term follow‐up after liver transplantation.4, 5

But in our previous study we did not find significant association between IL‐28B SNPs and hepatitis B virus (HBV) natural clearance, this made us to consider whether IL‐28B SNPs were associated with PTDM of HBV patients after liver transplantation (LT).6 As is known, the prevalence of hepatitis virus infection varies greatly based on region disparity. In North America and Europe, the prevalence of HBV infection is rather low, but in East Asia, this number soars up; the prevalence of HCV is rather opposite. This difference can be truly originated from race disparity since the degree of genetic variations differs among individuals of different races.7, 8 Therefore, we want to explore whether IL‐28B SNPs are associated with insulin resistance and PTDM in Chinese Han people.

Genome‐wide association studies have demonstrated SNPs of human leukocyte antigen (HLA) might be correlated with virus clearance after infection with HBV, especially polymorphisms on HLA‐DP/DQ loci.9 Another study conducted in the Chinese Han population substantiated that polymorphisms of signal transducer and activator of transcription 4 (STAT4) might correlate with progression of hepatocellular carcinoma (HCC) after HBV infection.10 Inflammatory factor IL‐18 was reported to take part in the antivirus reaction, but whether these factors correlate with expression of SOCS and insulin resistance still remains elusive. Our studies focus on the association between SNPs of HLA‐DP/DQ, STAT4, IL‐28B, IL‐18, and insulin resistance or PTDM. We are trying to clarify the importance of these host genetic factors on insulin resistance and to find some markers for PTDM.

2. METHODS AND MATERIALS

2.1. Subjects

The present study recruited Chinese Han people as subjects. All the subjects were recruited from West China Hospital of Sichuan University from June 2004 to May 2013. All recipients had received a calcineurin inhibitor–based immunosuppressive regimen (Tac + mycophenolate mofetil + steroids) for more than 1 year without interruption after liver transplantation. All of the recipients were in the stable period without rejection. Whole blood was collected before operation for DNA extraction, serum was collected both before and after operation, for glucose and liver function monitor. In the first month after operation, sample was collected each week, then collected each month since the second month after operation, then monitored every 3 months after half a year. There were 260 liver transplant recipients recruited; 55 recipients had PTDM and 205 recipients were without PTDM.

The clinical diagnosis of PTDM was defined as use of insulin or oral hypoglycemic agents, hemoglobin A1C >6.5% or fasting blood glucose >126 mg/dL on two separate occasions.11 The International Statistical Classification of Diseases and Related Health Problems, 10th revision, disease classification code is E09. All of the following conditions should be excluded: diabetes mellitus due to underlying condition (E08.‐), gestational diabetes (O24.4‐), neonatal diabetes mellitus (P70.2), postpancreatectomy diabetes mellitus (E13.‐), postprocedural diabetes mellitus (E13.‐), secondary diabetes mellitus NEC (E13.‐), type 1 diabetes mellitus (E10.‐), and type 2 diabetes mellitus (E11.‐).

This study was approved by the ethics committee of our hospital, and the methods were carried out in accordance with the approved guidelines. All patients signed an informed consent prior to inclusion in this study.

2.2. Genomic DNA extraction

Blood samples (3 mL) were collected in ethylenediaminetetraacetic acid tubes, and genomic DNA was isolated from whole blood samples using the whole blood DNA kit (Biotake Corporation, Beijing, China). DNA was extracted from 200 μL of the whole blood according to the manufacturer's protocol. The concentration of DNA was measured by a NanoDrop 2000c spectrophotometer (Thermo Scientific, Wilmington, DE, USA). DNA was diluted to 10 ng/μL for working solutions, and the isolated DNA was stored at −20°C.

2.3. Polymerase chain reaction

HLA‐DP (rs3077, rs9277535), HLA‐DQ (rs7453920), STAT4 (rs7574865) polymorphisms, IL‐28B (rs12979860, rs8099917, and rs12980275), and IL‐18 (rs1946518 and rs187238) were investigated in this study. Primers for the nine SNPs are shown in Table S1. All the nine SNPs were genotyped using polymerase chain reaction high‐resolution melting (HRM) analysis performed on a Light Cycler 480 (Roche Diagnostics, Penzberg, Bavaria, Germany). Single nucleotide polymorphism genotyping was performed in a 10 μL reaction system containing 5 μL Roche Master Mix (Roche Applied Science, Mannheim, Germany), which comprises FastStart Taq DNA Polymerase and the high‐resolution melting dye in a reaction buffer, 1.2 μL 25 mmol/L MgCl2, 0.1 μL 10 μmol/L forward Primer, 0.1 μL 10 μmol/L reverse Primer, 2.6 μL deionized water, and finally 1 μL DNA sample. The whole genotyping process encompasses four programs, namely predenaturation, amplification, HRM, and cooling. When finished, the results were analyzed by the corresponding Gene Scanning Software v1.2 (Roche Diagnostic).

2.4. High‐resolution melting curve analysis

The collected data were analyzed by the Light Cycler 480 Gene Scanning software v1.2 (Roche Diagnostics). All the samples with amplification were monitored by real‐time polymerase chain reaction. During the HRM stage of the melt curve, double‐stranded amplicons slowly denature, releasing bound dsDNA‐binding dye. The real‐time polymerase chain reaction instrument measures this decrease in fluorescence signal, and the HRM software plots fluorescence signal over temperature. The melt curves from the variant DNA can be distinguished from the many wild‐type samples in this view. Both melting temperature shifts and curve shape can be used to identify sequence differences. Homozygous allelic/sequence variants are typically characterized by the temperature (x‐axis) shift observed in an HRM melt curve, whereas heterozygotes are commonly characterized by a change in melt curve shape generated from base‐pairing mismatches as a result of destabilized heteroduplex annealing between some of the wild‐type and variant strands. HRM data are often plotted as difference curves to visually magnify differences between the melt profiles of different clusters within the same genotype. The exact same setting of the normalization was used for all experiments. The genotype of the subset was defined according to known genotypes of controls, which were determinate by gene sequencing. The genotypes of the SNPs in HLA‐DPA1 rs3077, HLA‐DPB1 rs9277535, HLA‐DQB2 rs7453920, STAT4 rs7574865, and IL‐18 rs1946518 G/T were showed in Figs S1‐S5. Sequencing results of IL‐18 rs1946518 were showed in Fig. S6. A dominant model was used in our study.

2.5. Cytokines and chemokines detection

Plasma was collected for cytokines and chemokines detection. IL‐1β, IL‐6, IL‐8, IL‐17, interferon (IFN)‐γ, inducible protein‐10, monocyte chemoattractant protein (MCP)‐1, and macrophage inflammatory protein (MIP)‐1b were analyzed by the Bio‐Plex suspension array system (Bio‐Plex Multiplex Immunoassays, Bio‐Rad, Hercules, CA, USA).

The principle of the Bio‐Plex pro bead‐based assays is similar to capture sandwich immunoassays. The capture antibody‐coupled beads are first incubated with the sample followed by incubation with biotinylated detection antibodies. After washing away the unbound biotinylated antibodies, the beads are incubated with a reporter streptavidin‐phycoerythrin conjugate. The beads are then passed through the Bio‐Plex array reader, which measures the fluorescence of the bound streptavidin‐phycoerythrin on each bead. Measurements are provided as the median fluorescence intensity for a given bead population. Bead populations are identified by fluorescence to determine the analyte being measured. All assay incubations are performed at room temperature as described in the instruction manual (Bio‐Rad part number 10024929). All washes were performed using a Bio‐Plex Pro Wash Station with cycles of 200 μL of wash buffer per well. Data acquisition was done using the Bio‐Plex Manager 6.1 software.

2.6. Logistic regression and statistical analysis

The Hard‐Weinberg Equilibrium (HWE) was determined using the goodness‐of‐fit chi‐square test to compare the observed frequency with the expected frequency in the subjects. The genotype and allele‐type association analyses were performed using the Pearson's chi‐square test. We tested two different genetic models including the dominant model and the recessive model. The better fitting model between the two models was the one with the smaller P value. A Student's t test was used to test for differences in continuous variables.

Statistical analysis was performed by SPSS 16.0 (SPSS Inc., Chicago, IL, USA). The Mann‐Whitney U test was used for comparisons of assays results between two groups. Data were expressed as mean ± standard deviation or median (range). The frequencies of genotype and allele were compared among patients and controls using Pearson χ² analysis. A P value <.05 indicated the statistical significance.

Logistic regression models were used for calculating odds ratios (95% confidence interval) and corresponding P values for each clinical and laboratory characteristics using SPSS. All of the variables were included for single factor regression, and then the parameters that had significant predict value were included for multivariate regression. Age (continuous value) and sex (male = 0, female = 1) were adjusted by inclusion in logistic analysis as covariates.

All of the experimental protocols were approved by the ethics committee of West China Hospital of Sichuan University.

3. RESULTS

The detailed clinical characteristics of the subjects were depicted in Table 1. No difference existed in the proportion of male vs female in both the PTDM and non‐PTDM groups. There were significant differences between PTDM and non‐PTDM recipients in their triglyceride, cholesterol, urea nitrogen, glucose, and glycosylated hemoglobin A1C; this may have resulted from abnormal glucose metabolism, which disrupts lipid metabolism and kidney function.

Table 1.

Demographic characteristics of allo‐liver recipients with or without posttransplantation diabetes mellitus (PTDM)

Characteristics	PTDM (n = 55)	Non‐PTDM (n = 205)	P
Age (yr)	45 (38,57)	47 (41,53)	.361
Gender (M:F)	37/18	154/51	.242
BMI	22.4 (19,25)	22.3 (18,24)	.71
Transplant time (mo)	42 (25,69)	47 (33,56)	.435
Tacrolimus (ng/mL)	4.6 (3.0,6.2)	4.9 (3.2,6.5)	.468
Concentration/dosage (ng/mL per mg/kg)	138.7 (84,235)	132.4 (89,197)	.664
Total protein (TP)	59.8 (52.8,65.6)	62.1 (55.5,67.1)	.35
Albumin (ALB)	46.7 (45.4,49)	46.5 (43.7,48.1)	.416
Triglyceride (TG)	1.58 (1.1,2.42)	1.07 (0.86,1.42)	.003a
Cholesterol (CHOL)	4.71 (3.96,5.5)	4.26 (3.57,4.91)	.04a
Total bilirubin (TBIL)	17.2 (11.3‐26.2)	16.5 (11‐28.3)	.542
Direct bilirubin (DB)	4.6 (3.4,9.4)	4.9 (3.6,7.6)	.985
Alanine aminotransferase (ALT)	40 (22,93)	38 (23,56)	.813
Aspartate aminotransferase (AST)	29 (21,55)	36 (24,52)	.563
Alkaline phosphatase (ALP)	94 (79,171)	93 (74,159)	.541
Gamma‐glutamyl transpeptidase (GGT)	62 (26,250)	42 (21,185)	.302
Lactate dehydrogenase (LDH)	205 (172,226)	190 (176,232)	.579
α‐Hydroxybutyrate dehydrogenase (HBDH)	165 (144,190)	158 (145,188)	.704
Urea nitrogen (BUN)	7.42 (6.01,8.99)	6.26 (5.22,7.49)	.01a
Creatinine (CREA)	86.6 (79.2,105)	82.6 (72.9,98)	.174
Cysteine C (CYS‐ C)	0.93 (0.84,1.01)	0.81 (0.75,1.03)	.057
Glucose (GLU)	7.68 (7.08,9.01)	5.41 (4.92,5.82)	.000a
Glycosylated hemoglobin A1C (HbA1C)	7.1 (5.8,7.9)	5.3 (4.8,5.5)	.000a

Values are shown as median (range). Data were analyzed by chi‐square analyses test and Mann‐Whitney U test.

^a P < .05.

All the genotype distributions in both groups of the SNPs were in Hardy‐Weinberg equilibrium (P > .05). The association of IL‐28B and HLA‐DP/DQ genotypes with these glucose metabolic problems was not adequately powered to reach statistical significance and may require further study; however, PTDM was associated with STAT4 (rs7574865) polymorphisms and IL‐18 (rs1946518) polymorphisms (Table 2). The most important observation in our current study is that STAT4 (rs7574865) T allele and IL‐18 (rs1946518) A allele may increase the risk for insulin resistance and PTDM. When the liver transplantation recipients were divided into two groups according to genotypes of STAT4 (rs7574865), the expression concentration of IL‐1β, IFN‐γ, MCP, and MIP‐1b were significantly different. Recipients with STAT4 (rs7574865) T allele are associated with an increased concentration of IL‐1β, IFN‐γ, MCP, and MIP‐1b (Table 3 and Figure 1).

Table 2.

Association of HLA‐DP/DQ, STAT4, IL‐28 and IL‐18 polymorphisms with PTDM

	PTDM n = 55	Non‐PTDM n = 205	χ2	Unadjust P	Bonferroni P
IL‐28B rs12979860
CC	47 (85.5)	185 (90.2)	1.035	.309	.8379
CT+TT	8 (14.5)	20 (9.76)
IL‐28B rs12980275
AA	48 (87.3)	189 (92.2)	1.303	.254	.8011
AG+GG	7 (12.7)	16 (7.8)
IL‐28B rs8099917
AA	49 (89.1)	186 (90.7)	0.134	.714	.9444
AC+CC	6 (10.9)	19 (9.27)
IL‐18 rs1946518
AA	20 (36.4)	47 (22.9)	4.093	.043a	.0484a
AC+CC	35 (63.6)	158 (77.1)
IL‐18 rs187238
GG	41 (74.5)	143 (69.8)	0.481	.488	.8379
GC+CC	14 (25.5)	62 (30.2)
HLA‐DP rs 3077
GG	32 (58.2)	117 (57.1)	0.022	.883	.9646
AA+AG	23 (41.8)	88 (42.9)
HLA‐DP rs 9277535
GG	34 (61.8)	111 (54.1)	1.035	.309	.8258
AA+AG	21 (38.2)	94 (45.9)
HLA‐DQ rs 7453920
GG	45 (81.8)	178 (86.8)	0.892	.345	.5013
AA+AG	10 (18.2)	27 (13.2)
STAT4 rs 7574865
GG	23 (41.8)	119 (58.0)	4.609	.032a	.0417a
GT+TT	32 (58.2)	86 (42.0)

^a P < .05.

Table 3.

Association between concentration of cytokines/chemokines and STAT4 polymorphisms

Cytokine/chemokine	STAT4 rs7574865 Genotype GT+TT n = 118	STAT4 rs7574865 Genotype GG n = 142	Z	P
IL‐6	7.69 (6.42,12.3)	7.57 (5.5,14.5)	−0.463	.664
IL‐8	14.2 (8.55,19.7)	13.4 (9.50,16.9)	−0.278	.781
IL‐1β	9.03 (4.13,13.2)	2.81 (1.98,7.66)	−3.309	.001a
IL‐17	23.9 (16.7,45.1)	28.8 (21.3,34.4)	−0.523	.601
IFN‐γ	48.3 (29.1,77.2)	23.3 (17.1,28.6)	−5.179	.000a
IP‐10	2006 (1416,3161)	1984 (1334,2656)	−0.837	.402
MCP‐1	122 (73.2,144)	71.8 (48.8,95.2)	−3.575	.000a
MIP‐1b	255.9 (167.7,359.4)	190.6 (140.9,270.3)	−6.653	.000a

^a P < .05.

Figure 1.

Expression of cytokines/chemokines with different genotypes of STAT4 polymorphisms are shown in Figure 1.

Because the sample size of PTDM was very small (only around 50), so the regression results was not very credible. Our logistic regression results indicated that Tac concentration, gamma‐glutamyl transferase, serum glucose, hemoglobin A1C, MCP, and STAT4 rs7574865 were associated with the risk of PTDM (Table 4). We divided all the subjects into subgroups according to genotype of different SNPs. For example, one group is CC genotype of IL‐28B rs12979860, the other is CT+TT, then logistic regression was analyzed to investigate whether the genotype of IL‐28B rs12979860 was associated with gender, age, BMI, transplant time, concentration of Tacrolimus, and PTDM. The same analysis was determined in all of the SNPs to explore the pleiotropy. The results were attached in the supplement materials.

Table 4.

Logistic regression of clinical and lab characteristics with PTDM

Assay	Sig	OR (95%CI)	Assay	Sig	OR (95%CI)
Age	0.762	1.76 (0.82,3.12)	CYSC	0.543	1.63 (0.73,3.47)
Gender	0.164	1.65 (0.56,4.25)	GLU	0.044a	0.41 (0.25,0.76)
BMI	0.547	0.83 (0.27,3.12)	HbA1C	0.037a	4.17 (1.94,7.36)
Time	0.215	1.04 (0.54,2.76)	IL‐6	0.215	3.04 (0.65,6.12)
Tac	0.029a	0.64 (0.19,0.83)	IL‐8	0.117	0.76 (0.22,2.37)
C/Da	0.130	3.27 (0.35,9.32)	IL‐1β	0.481	3.31 (0.87,9.81)
TP	0.347	1.65 (0.69,4.27)	IL‐17	0.27	1.77 (0.58,4.38)
ALB	0.216	1.03 (0.54,2.58)	IFN‐γ	0.353	1.21 (0.39,2.73)
TG	0.172	0.41 (0.25,1.76)	IP‐10	0.186	0.51 (0.13,2.04)
TC	0.141	4.17 (0.64,7.36)	MCP	0.027a	0.45 (0.23,0.68)
TB	0.139	1.88 (0.21,4.74)	MIP‐1b	0.139	1.38 (0.41,2.58)
DB	0.608	0.91 (0.43,2.46)	HLA‐DP rs3077	0.637	0.81 (0.32,1.95)
ALT	0.998	2.25 (0.91,5.63)	HLA‐DP rs9277535	0.547	2.13 (0.27,7.06)
AST	0.426	1.14 (0.65,2.87)	HLA‐DQ rs7453920	0.318	0.85 (0.42,1.91)
ALP	0.277	0.93 (0.46,1.93)	STAT4 rs7574865	0.038a	2.17 (1.62,4.68)
GGT	0.035a	0.48 (0.22,0.79)	IL‐28B rs12979860	0.466	1.58 (0.87,4.52)
LDH	0.269	0.71 (0.23,1.76)	IL‐28B rs12980275	0.541	1.03 (0.51,3.38)
HBDH	0.137	0.84 (0.24,3.68)	IL‐28B rs8099917	0.102	0.75 (0.39,2.27)
BUN	0.839	2.68 (0.32,5.45)	IL‐18 rs1946518	0.529	1.34 (0.66,6.78)
CREA	0.988	0.58 (0.29,1.17)	IL‐18 rs187238	0.311	0.93 (0.43,1.89)

^aC/D indicate concentration/dosage (ng/mL per mg/kg).

4. DISCUSSIONS

Liver transplant recipients are particularly at risk for developing PTDM as a consequence of immune suppression.12 It was reported that some host genetic factors such as SNPs of IL‐28B are associated with expression of SOCS and insulin resistance.13 IL‐28B is one of the antivirus factors in process of HCV infection, and its genetic variants are associated with virus clearance and treatment response.14, 15 Some other similar factors such as HLA‐DP/DQ, STAT4, and IL‐18 were newly reported to be associated with virus infection and progress of diseases; however, there has been no study about the relationship between insulin resistance and these factors.16, 17

Meanwhile, inflammation and activation of the immune system are emerging as key mechanisms associated with type 2 diabetes. Adipose tissue inflammation was recently identified as an early indicator of insulin resistance and type 2 diabetes and as a contributor to disease susceptibility and progression.18 Strong evidence from mouse models of obesity suggest that adipose tissue infiltration with proinflammatory macrophages, T cells, and natural killer cells leads to cytokine and chemokine production and free fatty acid release, which can induce pancreatic β‐cell dysfunction, insulin resistance, and atherosclerosis.19

Therefore, the concentration change of cytokine and chemokine induced by inflammation may be associated with insulin resistance and PTDM, so our study is focused on two items: the relationship between genetic variants and PTDM, and whether cytokines and chemokines are regulated by IL‐28B, HLA‐DP/DQ, STAT4, and IL‐18, which take part in the cellular immune system regulation and inflammation. We wanted to find some new target or marker associated with PTDM.

It was reported that IL‐28B polymorphisms are strongly associated with the development of diabetes mellitus during following liver transplant. HCV core protein downregulates insulin receptor substrate 1 and insulin receptor substrate 2, promoting insulin resistance. In addition, eradication of HCV has been convincingly shown to improve insulin sensitivity. Studies have consistently showed that the unfavorable genotypes (CT and TT for rs12979860 SNP, and GG and TG for rs8099917) are associated with an increased expression of IFN‐stimulated genes. It has been described that HCV can induce both SOCS‐1 and SOCS‐3—either directly through the core protein or indirectly through TNF‐α/IL‐6 induction—which impairs signaling of the insulin receptor pathway and causes insulin resistance.20

Our results of Asian people were different from those of Caucasians. We have not found significant differences in genotype of IL‐28B between PTDM and non‐PTDM. This difference may be originated from ethnic disparity since the degree of genetic variations differs among individuals of different ethnicities. There is no difference exists in the two groups for allele frequency distribution of HLA‐DP/DQ, whereas there was significant difference of frequency distribution of STAT4 in the two groups, recipients with STAT4 (rs7574865) T allele is associated with an increased risk for insulin resistance and PTDM. This has not been reported by other studies before, and these results indicate the cellular immune regulation mechanism mediated by STAT4 may play an important role in the development of insulin resistance and PTDM.

STAT4 is a transcription factor that regulates the immune response, and transmitted signals from IL‐12 and type I IFN induce IFN‐c production. It is required for the signal transduction of various proinflammatory cytokines such as IL‐12, IL‐15, and IL‐23. Therefore, STAT4 is crucial in the polarization and perpetuation of Th1 cell immune response. It has also been involved in the development of the newly discovered subset of Th17 cells, which display a dominant role in autoimmunity‐associated inflammation. STAT4 deficiency may impair Th1 lineage development in response to IL‐12 stimulation of T cells, reduce IFN‐g production, and display propensity toward the development of Th2 cells. At the same time, STAT4 limits the development of regulatory CD4+Foxp3+ cells, suggesting a role in peripheral immune tolerance.21

Signal transducer and activator of transcription 4 were expressed primarily in T cells and natural killer cells; it is downstream of the Janus kinase/tyrosine kinase. Upon phosphorylation in response to growth factor activation, it dimerizes and translocates to the nucleus, where it acts as transcription factors inducing the expression of genes involved in proliferation and differentiation of various hematopoietic and nonhematopoietic cells. The most notable effect of STAT4 deficiency was on the peripheral glucose clearance in response to an in vivo insulin challenge. STAT4(‐) adipocytes expressed less in SOCS3 and Janus kinase 2, which were causal for insulin resistance.22

Another interesting finding of our study is that the SNP of inflammation marker IL‐18 (rs1946518) was found to be associated with PTDM; recipients with A allele may increase the risk for insulin resistance and PTDM. This caused us to pay attention to the relationship between inflammation and PTDM; we presume genetic variants of STAT4 and IL‐18 may influence the immune regulatory effect of them, and then induce cellular cascade and concentration change of cytokines and chemokines, finally leading to PTDM. This may help us to find new molecular markers for PTDM.

A large number of human population studies have linked insulin resistance to systemic inflammation. It is known that TNF‐α and several other proinflammatory cytokines, including IL‐6, IL‐12, and MCP‐1, are expressed in adipocytes and adipose tissue and are increased in diabetes. TNF‐α, IL‐6, and IL‐12 contribute to systemic insulin resistance; IL‐6 promotes hypertriglyceridemia; IL‐12 promotes atherosclerosis; and MCP‐1 facilitates infiltration of inflammatory cells such as macrophages into adipose tissues.23 Adipose tissue inflammation may be due to a stress reaction of adipocytes to nutrient overload and consequently increased production of proinflammatory cytokines by the adipocytes. This in turn leads to migration of inflammatory cells (macrophages and T cells) into the adipose tissue, promotes a vicious cycle of escalating inflammation, thus be associated with insulin resistance. Unbalanced regulation of Th1/Th17 vs Th2/Treg as well as impaired M1 macrophage polarization and recruitment may induce protective effects on glucose and insulin homeostasis.24

In our study, we chose some cytokines and chemokines to reflect the expression of Th1/Th17 vs Th2/Treg. This study divided liver transplantation recipients into two groups according to genotypes of STAT4 (rs7574865); recipients with T allele are associated with an increased concentration of IL‐1β, IFN‐γ, MCP‐1, and MIP‐1b. This indicates an existing imbalance between dominant Th1 responses and reduced Th2 response. This disturbance was induced by the immune regulatory disorder. Signal transducer and activator of transcription 4 are involved in the regulation of several chemokine receptors that are preferentially expressed by the Th1 cells. It appears that Th1 cytokines or Th1 polarized immune cells promote insulin resistance, whereas Th2 signaling sustains metabolic homeostasis. The genetic variants of STAT4 (rs7574865) and IL‐18 (rs1946518) may be new important markers for PTDM. This conclusion may provide some new ideas for individualized treatment of PTDM.

4.1. Limitation

The sample size of liver transplantation recipients is not very big, let alone the sample size of PTDM, thus the results may be less persuasive and it could give us prompt significance. As many Chinese hospitals have transplantation centers, our next step is to consolidate the recipients and data together and collect more data of liver transplant recipients. Postoperative complications influence survival of transplant recipients significantly, so it is important to find new ideas (even if the conclusion is weak) to predict PTDM and other complications.

AUTHOR CONTRIBUTIONS

J.C. wrote the main manuscript; L.X.L. and Y.F.A. prepared the experiment. Yun Liao and Yi Li finished the statistic work. All authors reviewed the manuscript.

Supporting information

 

 

 

Chen J, Li L, An Y, et al. Multiple genetic variants associated with posttransplantation diabetes mellitus in Chinese Han populations. J Clin Lab Anal. 2018;32:e22308 10.1002/jcla.22308