The ACACB gene rs2268388 polymorphism is associated with nephropathy in Caucasian patients with diabetes: a meta-analysis

Abstract

To date, case–control studies on the association between a single-nucleotide polymorphism (SNP), rs2268388, in the acetyl-coenzyme A carboxylase beta (ACACB) gene and diabetic nephropathy have provided controversial results. To clarify the effect of rs2268388 on the risk of diabetic nephropathy, a meta-analysis of all case–control studies was performed. The fixed effects and random effects models showed that the C allele was associated with a decreased susceptibility risk of diabetic nephropathy compared with the T allele among Caucasian patients with diabetes. The contrast of the recessive model produced the same pattern of results as the allele contrast. Our pooled data suggest a significant association exists between rs2268388 and diabetic nephropathy among Caucasian patients with diabetes.

• ACACB
• diabetic nephropathy
• meta-analysis
• polymorphism
• susceptibility

Introduction

Diabetes mellitus (DM) and/or its complications, such as nephropathy, are a worldwide concern in terms of their high prevalence and serious socio-medical burden. Diabetic nephropathy (DN) is a major complication of diabetes mellitus and is the leading cause of end-stage renal disease (ESRD). Although significant improvements in DN diagnosis and treatment have been made over the past several decades, the etiology of most cases of DN remains unknown due to probable multifactorial mechanisms of pathogenesis. Many epidemiological studies have detected that genetic susceptibility plays an important role.1^,2

Acetyl-coenzyme A carboxylase beta (ACACB) catalyzes the synthesis of malonyl-CoA, the substrate for fatty acid synthesis, and is a regulator of fatty acid oxidation.3 Relative to the C allele, a 29-bp DNA fragment containing the T allele in intron 18 of ACACB gene (rs2268388; intron-18 +4 139 C > T) demonstrated greater enhancer activity in cultured human renal proximal tubular epithelial cells, indicating higher ACACB expression in risk allele carriers.4 Therefore, rs2268388 could contribute to the regulation of metabolism, alterations in lipids, and adiposity.5 A recent study revealed that the T allele is associated with DM-related nephropathy.4 A similar result was also observed by Tang et al.6 However, a recent meta-analysis on the role of the ACACB SNP rs2268388 in the development of DN have shown conflicting results (OR = 0.83; 95% CI = 0.62–1.11).7 The exact relationship between the ACACB SNP rs2268388 and DN susceptibility is not well established. Therefore, we performed a meta-analysis of all eligible studies to obtain a more precise estimation of the association between the ACACB SNP rs2268388 and DN. This work was performed in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines.8

Materials and methods

Publication search and data extraction

The electronic databases Pub Med, Embase, Web of Science, and CNKI (China National Knowledge Infrastructure) were searched for studies to include in the present meta-analysis using the terms ‘‘acetyl-coenzyme A carboxylase beta,’’ ‘‘ACACB,’’ ‘‘diabetic nephropathy,’’ ‘‘polymorphism,’’ and ‘‘rs2268388.’’ An upper date limit of 31 October 2014, was used, but no early date limit was applied. The search was conducted without any restrictions on language but focused on studies that had been conducted on human subjects. The reference lists were screened of all of the identified studies and of the comprehensive reviews in the field. The number of cases and controls for the rs2268388 genotype was extracted from each study. We did not define a minimum number of patients as a criterion for a study’s inclusion in our meta-analysis.

Statistical analysis

The association between rs2268388 polymorphism and DN risks was estimated by calculating pooled ORs and 95%CI in the allele model (C vs. T), the homozygous model (CC vs. TT), the dominant model (CC + CT vs. TT), and the recessive model (CC vs. CT + TT). The random effects model adjusts for the variability of results among trials and provides a more conservative estimate of an effect using a wider CI.9 However, a random effects analysis will give more weight to smaller trials, which it appears overestimate the benefit of treatment, leading to biased overall results.10 Therefore, the pooled OR estimate of each study was calculated by both the fixed-effects model (the Mantel–Haenszel method)11 and the random-effects model (the DerSimonian and Laird method).12 The heterogeneity between studies was tested using the Q statistic.13 Heterogeneity was considered statistically significant if p < 0.10. The Begg–Mazumdar test and the Egger test were performed to assess the publication bias in the literature. All calculations were performed using Review Manager 5.0 (RevMan Inc., New York, NY).

Results

Data were found in three articles,4^,6^,14 representing 11 studies (Table 1). Nine of the studies were case–control studies,4^,6 and two of them were cohort studies.14 Studies were conducted in various populations of different ethnicities: eight studies were conducted in Asian populations and three studies were conducted in Caucasian populations. The studies provided 3273 cases and 3242 controls for rs2268388. For case groups, the frequency of CC-homozygous individuals was 62.3%. However, 31.3% of CT-heterozygous individuals and 6.4% of TT-homozygous individuals displayed the rs2268388 polymorphism. In control groups, the frequencies of CC-homozygous individuals, CT-heterozygous individuals, and TT-homozygous individuals were 68.8%, 27.4%, and 3.8%, respectively. The C allele frequencies in the case and control groups were 78% and 82.5%, respectively. In the Japanese-4 study and Shah cohort studies of the rs2268388 polymorphism, the distributions of genotypes in the control groups were not in HWE (p < 0.05), indicating genotyping errors and/or population stratification. These three studies were excluded from this meta-analysis.

Table 1. Study characteristics.

						Case
Author	Year	Country	Ethnicity	Design	TD	Sample size	Definition
Maeda (Janpanese-1)	2010	Japan	East Asian	CC	2	737	Proteinuria
Maeda (Janpanese-2)	2010	Japan	East Asian	CC	2	31	Proteinuria
Maeda (Janpanese-3)	2010	Japan	East Asian	CC	2	69	Proteinuria
Maeda (Janpanese-4)^a	2010	Japan	East Asian	CC	2	298	ESRD
Maeda (Korean)	2010	korea	East Asian	CC	2	177	ESRD
Maeda (Singaporean)	2010	Singapo	East Asian	CC	2	199	Proteinuria
Maeda (European-1)	2010	USA	Caucasian	CC	2	47	Proteinuria
Maeda (European-2)	2010	USA	Caucasian	CC	2	473	ESRD
Maeda (DANE)	2010	Denmark	Caucasian	CC	1	448	ESRD
Tang	2010	China	East Asian	CC	2	294	ESRD
Shah (Cohort 1)^a	2012	India	Asian	Cohort	2	240	Proteinuria
Shah (Cohort 2)^a	2012	India	Asian	Cohort	2	260	Proteinuria
	Control			OR (95% CI)
Author	Sample size	Definition	HWE	Allele level
Maeda (Janpanese-1)	552	Normoalbuminuria	0.660	0.62 (0.51–0.75)
Maeda (Janpanese-2)	163	Normoalbuminuria	0.135	0.81 (0.43–1.54)
Maeda (Janpanese-3)	182	Normoalbuminuria	0.053	0.68 (0.44–1.06)
Maeda (Janpanese-4)^a	221	Normoalbuminuria	0.023	1.28 (0.95–1.73)
Maeda (Korean)	196	Normoalbuminuria	0.700	1.20 (0.85–1.69)
Maeda (Singaporean)	212	Normoalbuminuria	0.067	0.93 (0.68–1.28)
Maeda (European-1)	108	Normoalbuminuria	0.573	0.68 (0.37–1.23)
Maeda (European-2)	415	Normoalbuminuria	0.296	0.62 (0.47–0.82)
Maeda (Dane)	425	Normoalbuminuria	0.809	0.88 (0.68–1.14)
Tang	298	Normoalbuminuria	0.493	0.88 (0.67–1.15)
Shah (Cohort 1)^a	255	Normoalbuminuria	<0.001	0.39 (0.28–0.54)
Shah (Cohort 2)^a	215	Normoalbuminuria	0.003	0.38 (0.26–0.54)

HWE, Hardy–Weinberg equilibrium; CC, case–control; TD, type of diabetes; ESRD, end-stage renal disease.

^aThese studies were excluded from the meta-analysis, because they were not in HWE (p < 0.05).

All comparisons are listed in Table 2. Overall, the C allele was associated with a decreased risk of DN compared with the T allele (OR = 0.79; 95% CI = 0.68–0.93). The contrast of recessive model produced the same pattern of results as the allele contrast. Large heterogeneity (I² = 54%, p = 0.03) was detected among the nine studies. To eliminate heterogeneity, we divided the nine studies into subgroups. Subsequently, heterogeneity disappeared in subgroups of Caucasian subjects, which revealed that most of the studies could not be grouped based on ethnicity. In the analysis stratified by Caucasian ethnicity, a significant association was found between DN and the various genetic models. Moderate heterogeneity (I2 = 41%, p = 0.18) was detected among the three studies.

Table 2. Odds ratios (ORs) and heterogeneity results for the genetic contrasts of ACACB gene rs2268388 polymorphism for DN.

		OR
	Population	Fixed effects (95% CI)	p	Random effects (95% CI)	p	I2 (%)	p Value Q test
Alleles (C vs. T)	All	0.77 (0.70–0.85)	<0.0001	0.79 (0.68–0.93)	0.05	54	0.03
	Caucasian	0.74 (0.62–0.89)	0.001	0.73 (0.57–0.94)	0.02	41	0.18
	East Asian	0.78 (0.69–0.89)	0.003	0.83 (0.66–1.04)	0.33	64	0.02
CC to TT	All	0.54 (0.40–0.72)	0.001	0.56 (0.41–0.76)	0.09	6	0.39
	Caucasian	0.47 (0.25–0.89)	0.02	0.47 (0.19–1.14)	0.10	39	0.19
	East Asian	0.56 (0.40–0.78)	0.02	0.57 (0.40–0.80)	0.34	3	0.4
CC to (CT + TT)	All	0.77 (0.68–0.86)	0.0001	0.79 (0.65–0.97)	0.07	58	0.02
	Caucasian	0.74 (0.60–0.90)	0.003	0.73 (0.59–0.92)	0.006	13	0.32
	East Asian	0.78 (0.68–0.91)	0.008	0.83 (0.62–1.12)	0.37	69	0.006
(CC + CT) to TT	All	0.58 (0.43–0.77)	0.006	0.60 (0.44–0.80)	0.13	0	0.44
	Caucasian	0.51 (0.27–0.95)	0.03	0.51 (0.22–1.20)	0.12	34	0.22
	East Asian	0.60 (0.43–0.83)	0.04	0.61 (0.44–0.85)	0.42	0	0.44

The Begg–Mazumdar test and the Egger test were performed to assess the publication bias in the literature. All the studies investigating the C allele versus the T allele yielded Begg’s test score of p = 0.412 and Egger’s test score of p = 0.168, which did not indicate a potential for publication bias.

Discussion

The genetic susceptibility to cancer has been the focus of research in the scientific community. This meta-analysis summarized all the available data on the association between the ACACB SNP rs2268388 and DN, including a total of 2773 cases and 2772 controls. In Caucasians, a significant association was found for the genetic models examined. In addition, heterogeneity decreased when the population was viewed as separate groups, which suggested that the effect of the C allele on the risk of DN might differ based on ethnicity.

In the development of DN, obesity and an increase in free fatty acid levels are independent risk factors.3 ACACB is a rate-limiting enzyme involved in mitochondrial fatty acid oxidation and plays a key role in fatty acid metabolism.15–18 There is evidence implicating altered lipid metabolism in the pathogenesis of DN.19^,20 ACACB converts acetyl-CoA to malonyl-CoA, and its product is known to inhibit carnitine palmitoyl-CoA transferase, resulting in inhibition of mitochondrial beta-oxidation of fatty acids. ACACB is involved in fatty acid synthesis, and its activity results in accumulation of free fatty acid in cells and further lipotoxicity, including in renal cells.21 Epidemiological results reveal that ACACB SNP rs2268388 associates with body mass index (BMI) in general populations and with obesity in subjects with type 2 diabetes.22 ACACB could accelerate the progression of nephropathy from diabetes by affecting insulin sensitivity through the modulation of fatty acid metabolism.19 T allele demonstrated greater enhancer activity in cultured human renal proximal tubular epithelial cells, indicating higher ACACB expression.4

A previous meta-analysis addressing the association between rs2268388 and DN had several errors.7 First, it only included the data of five case–control studies, and the European 1 (Steno 2) study was not included in that analysis. Second, the OR of the contrast of alleles in the Dane study was 1.13, which was provided in the original report of Maeda et al. However, Mooyaart et al. calculated the wrong OR (0.99). These errors resulted in the wrong conclusion.

Considering the limitations of this meta-analysis, our results should be interpreted with caution. First, our results are based on unadjusted estimates. A more precise analysis should be conducted using individual data, which would allow researchers to adjust for covariates, including age, ethnicity, family history, lifestyle, and environmental factors. Second, only published studies were included in this meta-analysis. Although we did not find a potential for publication bias, non-significant or negative findings may not have been published. Third, only three Caucasian studies from two papers might be not sufficient for getting conclusive result or for evaluating heterogeneity or publication bias.

In conclusion, our pooled data suggest evidence for a major role of the ACACB SNP rs2268388 in DN among Caucasian populations.

Declaration of interest

The authors report no conflicts of interest.