Exploring the Role of Serum Lysyl Oxidase (LOX) Levels and LOX Gene Polymorphism (G473A) in Diabetic Nephropathy

Abstract

Diabetic Nephropathy (DN) is a leading microvascular complication of diabetes with poor prognosis. Extra cellular matrix accumulation (ECM) leading to renal fibrosis is the key manifestation of diabetic kidney. Lysyl oxidase (LOX) plays an important role in cross linking of collagen. Altered LOX expression and LOX depended cross links with Diabetes have been documented both in rat and human model. Evidence of LOX gene polymorphism (G473A) has been shown limited to Diabetic foot ulcer. Studies on association of single nucleotide polymorphism (SNP) (G473A) of LOX gene with Diabetic nephropathy is scarce. A pilot study was proposed with an objective to find out association of serum LOX and LOX gene polymorphism (G473A) in Diabetic Nephropathy. This cross-sectional study was conducted in the Dept. of Biochemistry, AIIMS Bhubaneswar. Serum LOX level, LOX gene polymorphism (G473A) by PCR-RFLP and its association with disease process was assessed in Diabetes as well as Diabetes Nephropathy cases. Results were analysed using SPSS-26 version. Serum LOX registered a significant rise having positive association with glycaemic status and inverse relationship with eGFR, whereas LOX gene polymorphism (G473A) did not reveal significant association with serum LOX and the disease process in DN cases. Increased serum LOX level might have contributed to disordered ECM with resultant lowering of kidney function in DN. LOX gene polymorphism (G473A) might not have influenced serum LOX level. A probable regulatory effect of hyperglycaemia through other mediators on LOX expression may exist that needs further exploration.

Introduction

Diabetic Nephropathy (DN) is a leading microvascular complication of diabetes with a prevalence of 39% in the diabetic population globally [1] With the existing management procedures, there still remains a risk of progression to End Stage Kidney Disease (ESKD) contributing to poor prognosis with heavy economic burden. Early detection of diabetic nephropathy has shown a considerable decrease in the progression of the disease pointing towards the need for its early diagnosis with new biomarkers for implementing better treatment modalities [2].

Diabetic Nephropathy kidney in its early stages shows thickening of both the glomerular basement membrane and tubular basement membrane leading to nodular as well as diffuse glomerulosclerosis [3]. Manifestations of a diabetic kidney are due to extracellular matrix accumulation that leads to renal fibrosis [3]. These major extracellular matrix changes in DN are suggested to be contributed by disordered collagen levels, glycation of collagen & defective crosslinking of collagen [4].

Lysyl Oxidase (LOX), an important Copper-containing enzyme participates in the process of cross-linking of collagen [5]. It modifies the side chain of lysyl residues in elastin and collagen, that leads to spontaneous cross-link formation [6]. Previous studies have documented altered LOX expression and LOX depended cross links in Diabetes cases both in rat model [5, 7, 8] and human model [6]. Single Nucleotide Polymorphisms (SNPs) of LOX gene (G473A) has been widely studied in malignancies. The association of LOX gene polymorphism has been shown with pathogenesis of diabetic foot ulcer [9]. LOX gene single nucleotide polymorphism (G473A) though has been studied in relation to malignancies, Diabetic foot ulcer, yet information available regarding its association with Diabetic Nephropathy is scarce. Hence the present study aimed to assess serum LOX levels, frequency of allele expression of LOX gene (G473A) polymorphism in Diabetic Nephropathy cases which would be of Diagnostic and therapeutic significance.

Materials and Methods

This pilot observational cross-sectional study was conducted in the Department of Biochemistry, AIIMS Bhubaneswar, Odisha during the period of November 2019–November 2021. Ethical clearance was approved by the Institutional Ethics Committee (IEC), AIIMS, Bhubaneswar, Ref number- IEC/AIIMS BBSR/PG Thesis/2019-20/78).

This study included 40 Diabetic Nephropathy patients, 40 Diabetes Mellitus patients with a control group of 40 Healthy individuals of either sex having age more than 18 years. Sample size of 120 was based on the data from previous study by Rajapurlar et al. [10]. The sample size was calculated by the formula,

$$n=\frac{Z^2\ast P\ast \left(1,-,P\right)}{d^2}$$

where n = total sample size,

P = Prevalence.

Z = Level of Confidence (for 95% = 1.96).

D = Precision (0.05).

Clinically confirmed cases of diabetic nephropathy based on history, clinical features, laboratory investigations (albumin creatinine ratio and eGFR) and attending Nephrology clinic, AIIMS, Bhubaneswar were enrolled in Diabetic Nephropathy group. Patients with a history of Diabetes mellitus for more than five years and without any complications attending NCD clinic of AIIMS, Bhubaneswar were included in diabetic group. Both these Diabetic groups were as per ADA Criteria. 40 age and sex matched attenders and/or relatives of patients with normal glycemic status and renal function were included as control after getting written consent.

Seriously ill, patients with severe liver disease, previous malignancy, pregnant and nursing women were excluded from the study. This study has been conducted as a pilot study as previous studies on humans on LOX gene SNPs were not found with enough data in Diabetes & Diabetic Nephropathy cases. A review of case reports was done for the collection of clinical data on these DM and DN cases.

5 ml of fasting venous blood was collected from each of the study subject. 2 ml was collected in EDTA vial and stored in − 20 °C for DNA Extraction and PCR–RFLP study. Rest 3 ml was collected in plain vacutainer tube, allowed to stand for 30 min for clot formation and centrifuged at 3500 rotations per minute for serum separation and stored in − 20 °C in labelled plastic microcentrifuge tube for estimation of LOX levels of subjects by ELISA technique.

Routine biochemical investigations like FBS, PPBS, HbA1_C, and renal functions were performed using AU5800 model Beckman Coulter Autoanalyzer. FBS and PPBS were assayed by hexokinase G-6-PDH method and HbA1c by measuring absolute HbA1c, total Hb values and HbA1c/Total Haemoglobin ratio being expressed as HbA1c percentage. Serum Urea, Serum Creatinine and Serum Uric acid were measured by enzymatic rate method (urease), kinetic modified Jaffe’s reaction, and modification of Fossati method respectively. Serum Total protein was measured by Biuret Method and eGFR was calculated using CKD EPI Formula [11].

Quantitative estimation of serum LOX was conducted using commercially available ELISA kits from Elabscience (United States) for LOX with sandwich ELISA technique as per manufacturer’s protocol (Fig. 1).

Fig. 1.

Standard curve for serial dilution of LOX for the assay of serum LOX by ELISA

LOX gene polymorphism (G473A) was analyzed by PCR- RFLP (Polymerase Chain Reaction-Restriction Fragment Length Polymorphism) with primer pair sequence such as Forward: 5′-CACTGGTTCCAAGCTGGCTA-3′ and Reverse: 5′-GGAAGTAGCCAGTGCCGTAT-3’. Restriction digestion was carried out using restriction enzyme Pst1. DNA band with sizes of 149 and 110 base pairs (bp) represented Homozygous A allele, an uncut fragment of 259 base pair for Homozygous G allele, and a combination of 259, 149 and 110 base pair fragments was considered as heterozygous genotypes (Fig. 2).

Fig. 2.

DNA gel electrophoresis after PCR–RFLP showing different genotypes of LOX gene polymorphism (G473A) in the study population

The gel picture obtained was interpreted and documented as allele profile of each subject. For confirmation of the obtained results, 20% of the total PCR amplified samples were selected randomly, and restriction digestion was repeated and analyzed by agarose gel electrophoresis with matching of results.

Statistical analysis was done using SPSS software (version 26). Descriptive statistics has been represented as mean ± standard deviation or as median. Continuous variables were analyzed by Kruskal Wallis test with Dunn- Bonferroni post hoc analysis. Serum LOX level was compared between groups using Mann–Whitney U test. Association of LOX gene polymorphism (G473A) with susceptibility to DM was assessed with odds ratio. Association of LOX gene polymorphism with serum LOX levels, glycemic status, and kidney function was calculated by bivariant logistic regression in the dominant model. Diagnostic accuracy of Serum LOX for Diabetic Nephropathy was determined by ROC Curve analysis. Correlation of serum LOX levels with disease process (glycemia and eGFR) has been evaluated by Spearman correlation test. P value less than 0.05 was taken as significant.

Results

Demographic and clinical characteristics of the study groups (Table 1) revealed no significant difference in age, sex distribution among the 3 groups. A marked difference was noted in BMI among the study groups (p = 0.006). Duration of diabetes between subjects in DN group and DM group was found to be significant. All participant with Diabetic Nephropathy and Diabetes were taking Oral Hypoglycemic Agents (OHA). Significantly low hemoglobin was documented in DN group compared to DM group and healthy controls (p value < 0.001). Analysis of glycemic status revealed marked rise in FBS, PPBS, and HbA1C in the DN group in comparison to DM and control group (p value < 0.001). Other biochemical parameters like serum urea, creatinine, and uric acid were also found to be significantly raised in DN group compared to DM and healthy controls (p < 0.001). Serum total protein concentration was markedly low in DN group followed by DM and healthy controls.

Table 1.

Demographic and clinical characteristics of diabetic nephropathy patients, diabetic patients, and healthy controls

Characteristics	Diabetic nephropathy cases (n = 40)	Diabetic cases (n = 40)	Controls (n = 40)	p value
Age (years)	63.87 ± 6.094	57.23 ± 5.470	60.20 ± 6.673	0.083
Males	25	22	26	0.321
Females	15	18	14	0.132
BMI (kg/m2)	27.44 ± 4.86	25.91 ± 4.14	24.49 ± 2.75	0.006
Duration of DM (years)	22.78 ± 5.48	8.75 ± 3.08	00	< 0.001
Treatment History (OHA)	40(100)	40(100)	NA	1a
Haemoglobin (gm%)	10.2 ± 0.9	12.2 ± 1.2	12.8 ± 0.7	< 0.001
FBS (mg%)	170 ± 36	159 ± 38	83 ± 6	< 0.001
PPBS (mg%)	240 ± 62	235 ± 58	109 ± 10	< 0.001
HbA1C (%)	9.5 ± 1.9	7.1 ± 1.3	5.7 ± 0.2	< 0.001
Urea (mg%)	65 ± 6.4	29.4 ± 5.23	13.1 ± 4.5	< 0.001
Serum Creatinine (mg%)	3.1 ± 0.9	1.0 ± 0.2	0.9 ± 0.12	< 0.001
Uric acid (mg%)	9.1 ± 2.3	5.8 ± 2.1	5.5 ± 1.7	< 0.001
Total Protein (Gm%)	4.5 ± 1.9	7.0 ± 1.8	6.9 ± 2.1	< 0.001
eGFR (ml/min/1.73m2)	22.14 ± 9.18	87.4 ± 13.35	103.24 ± 18.73	< 0.001

Results represented as Mean ± standard deviation or Number of Subjects (Percentage)

P value < 0.05 is considered statistically significant

^aP value compared between patients with Diabetic Nephropathy and patients with Diabetes Mellitus as it was not applicable for Healthy controls

The median expression of serum LOX level was found to be 1.28 ng/dl, 2.78 ng/dl, and 9.01 ng/dl in controls, DM and DN groups respectively, with a significant rise in DN group in comparison to DM and healthy controls (P < 0.001). Serum LOX registered significant positive association with FBS, PPBS, HbA1c respectively (p < 0.001), showing that people having higher FBS, PPBS, and HbA1C levels have increased serum LOX levels.

The Spearman’s rho correlation of serum LOX with eGFR (Fig. 3) revealed significant negative association with a r_s value of − 0.793 (P value < 0.001). People with raised serum LOX levels had lowered eGFR revealing that raised serum LOX level is associated with reduced glomerular filtration.

Fig. 3.

Correlation between serum LOX levels and eGFR Serum LOX expressed in ng/dl, eGFR expressed in mL/min/1.73 m²

To explore further the diagnostic role of serum LOX levels for diabetic kidney disease, a ROC analysis was done. The Area Under the Curve in the ROC analysis was 0.847 (P value < 0.001) for serum LOX levels. The cut off value of serum LOX in diagnosing DN was found to be 4.3 ng/dL with a sensitivity of 75% and specificity of 80%.

Frequency distribution of LOX gene polymorphism (G473A) in three study groups were analyzed (Table 2). The frequency of GA genotype (heterozygous) in Diabetic nephropathy group was 35, compared to 27.5 in diabetic group and 25 in healthy controls. The frequency of AA genotype was 2.5 in DN, 5 in DM, and 2.5 in healthy controls respectively. The genotypic frequencies in the three groups were in Hardy Weinberg equilibrium. No significant association was noted between LOX genetic polymorphism and serum LOX level.

Table 2.

Frequency distribution of LOX gene polymorphism in the study group

Genotype	Diabetic nephropathy (n = 40) (%)	Diabetes (n = 40) (%)	Healthy controls (n = 40) (%)
GG	25 (62.5)	27 (67.5)	29 (72.5)
GA	14 (35)	11 (27.5)	10 (25)
AA	1 (2.5)	2 (5)	1 (2.5)
Minor allele (A) frequency	0.20	0.18	0.15
HW χ2 P value	0.298	0.934	0.765

Data represented as N (%), HW, hardy–Weinberg equilibrium

Association of different genotypes of LOX gene polymorphism as dominant model with glycemic status and eGFR was analyzed as shown in Table 3. The association of GA + AA genotype with glycemic status (i.e., FBS, PPBS, Hb1c) was seen not to be significant in all the study groups. Similarly, the odds ratio for association of GA + AA genotype with eGFR also did not reveal any statistical significance (Table 3).

Table 3.

Association of LOX gene polymorphism (G473A) with glycemic status and kidney function in study groups

Study group	Geno types	N	FBS		PPBS		HbA1C		eGFR
			OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value
DN (40)	GG	25	1 (ref)
	GA + AA	15	1.004 (0.98–1.02)	0.744	0.99 (0.98- 1.01)	0.790	0.98 (0.69–1.41)	0.944	0.96 (0.89–1.04)	0.405
DM (40)	GG	27	1 (ref)
	GA + AA	12	0.99 (0.97–1.01)	0.742	1.001 (0.98- 1.01)	0.743	1.05 (0.59–1.87)	0.850	0.98 (0.93–1.03)	0.567
Control (40)	GG	29	1 (ref)
	GA + AA	11	0.91 (0.79- 1.04)	0.175	0.99 (0.92–1.06)	0.873	1.88 (0.10–33.8)	0.668	1.01 (0.97- 1.05)	0.432

Discussion

The present study has been conducted to evaluate the association of serum lysyl oxidase (LOX) levels and LOX gene polymorphism (G473A) in patients with Diabetic Nephropathy.

Mean duration of diabetes in DN patients were 22.7 years compared to 8.75 years in diabetes Mellitus cases. Diabetics and DN group revealed a marked rise in glycemic parameters indicating poor glycemic control. As serum urea, creatinine, uric acid, and total protein levels get deranged with progression of kidney involvement, these parameters documented marked alteration in Diabetes Nephropathy patients, though Diabetes Mellitus subjects did not register marked difference with healthy controls (Table 1). Post hoc analysis of these parameters i.e., FBS, PPBS, HbA1c, serum urea, creatinine, uric acid and total protein revealed significant difference between Diabetes with Nephropathy (DN) in comparison to DM patients showing severity of hyperglycaemia and renal damage in Diabetic Nephropathy cases.

Lysyl oxidase being an extracellular enzyme responsible for covalent cross-link formation in collagen fibrils, its contributory role in pathogenesis of various conditions like cancer, fibrotic diseases and inflammatory processes has been shown. The group of researchers in their study on ovarian cancer have documented significant increase in serum lysyl oxidase levels in patients compared to controls, showing its contributory role [12]. Studies conducted in fibrotic conditions like liver fibrosis, systemic sclerosis and kidney fibrosis have also documented increase in levels of LOX revealing its role in fibrogenesis [13–15]. Present study registered markedly raised serum LOX levels in DN group compared to DM group and healthy controls with p value < 0.001, in consistence with above findings suggesting that diabetic kidney disease being a fibrotic condition with impaired extracellular matrix accumulation, LOX mediated collagen cross-linking might have played a crucial role in its pathogenesis by affecting ECM stability and vascular remodeling [16].

Presence of mature LOX has been shown both in the nucleus [17–21] and cytoplasm [17–20]. Nuclear LOX is protected and remains functionally active, which is of great interest. A study has also shown that mature LOX can translocate into the cytosol from the extracellular space, then cluster in the nucleus, [22] potentially mediated by interaction of the catalytic domain of LOX with p66β, a transcription repressor [23]. Thus, the internalization of LOX into the intracellular and nuclear space takes place; even though its purpose or functionality remains unknown. A study conducted on retinal endothelial cells in rats have noted that hyperglycemia (HG) promotes the binding of LOX with ECM proteins, collagen IV and Fibronectin extracellularly and this could interfere in LOX internalization resulting in a reduction of intracellular LOX level, which creates a negative feedback signaling and contributes partially to LOX overexpression associated with Diabetic Retinopathy (DR) [8, 24]. Significant association of serum LOX with FBS, PPBS and HbA1c (p < 0.001) in the present study reveals the probable effect of hyperglycaemia in modifying LOX level pointing towards the similar mechanism cited by Song et al. and Nguyen et al. [8, 24].

A decrease in eGFR is a sign of impaired kidney function. As LOX is an enzyme involved in ECM remodelling, and fibrogenesis it may have a contributory role in the pathogenesis of chronic kidney damage. Present study also documented similar results with significantly higher levels of LOX in DN patients with a prominent negative correlation between serum LOX levels and eGFR among the study population. People with higher serum LOX levels had low eGFR registering that increase in serum LOX is associated with reduced glomerular filtration as a result of deranged kidney function similar to observation by Zhang et al. [15], and Stangenberg et al. [5].

Area under curve in ROC analysis adds to strengthening of association of LOX with the Nephropathy in Diabetes cases with a cut of value of 4.3 ng/dL, having sensitivity and specificity of 75% and 80% respectively.

Previous studies on LOX gene polymorphism (G473A) have focused mostly on cancers and fibrotic conditions. Study by Kumari et al. and Wang et al. found that frequency of LOX 473AA genotype and A allele were significantly higher in patients with ovarian cancer compared to controls [12, 25]. When Mukherjee et al. registered that LOX G473A polymorphism confers increased risk of oral submucosal fibrosis [26], study by Thorawat et al. did not reveal LOX G473A polymorphism in oral submucous fibrosis patients as well as controls [27].

Despite the fact that some genetic alterations take place in DN, the complexity of this disease prevents placing its biology into a simple conceptual framework. Studies on the effect of LOX gene polymorphism on susceptibility to diabetes and its complications are limited. Pichu et al. in their study on diabetic foot ulcer have shown that LOX gene polymorphism (G473A) had a significant association with diabetes and diabetic foot ulcer compared to healthy controls [9]. Information available regarding effect of genetic polymorphisms of LOX gene in making a diabetic individual susceptible to developing DN is scarce. Present study (Table 2) evaluated the frequency of LOX gene polymorphism (G473A) in the three study groups DN, DM and healthy control and observed GG genotype had the highest frequency followed by GA and AA in all three groups. LOX gene polymorphism (G473A) also revealed no impact on serum LOX level. The difference in frequency distribution among the study groups were statistically insignificant indicating no association of LOX gene polymorphism (G473A) towards increased susceptibility for diabetes and diabetic nephropathy. Analysis to find out the association of LOX gene polymorphism (G473A) with glycemic status and eGFR (Table 3) also did not register any significance stating that LOX gene polymorphism (G473A) might not have contributed towards increased serum LOX level for affecting kidney function. This difference of opinion of present study from observation by Pichu et al. points towards probable role of other factors for deranged LOX level in Diabetes Nephropathy [28, 29], that needs further exploration.

Strength of the Study

1. DM and DN groups were assessed individually for analysing the contributory role of LOX in disease progression

2. First study of Gene Polymorphism of LOX in DN in Humans

Limitations of the Study

1. It is a Cross- sectional study, etiological association could not be studied

2. Small sample size for genetic study and staging of Diabetic Nephropathy

Conclusion

Diabetes may have influence on serum LOX level affecting kidney function. Genetic polymorphism of LOX (G473A) may not have a contributory role in altered LOX level in both DM and DN cases. Future prospective studies with large sample size may enlighten on the role of LOX gene polymorphism (G473A) as well as probable role of other mediators of Diabetes on LOX upregulation affecting collagen crosslinking and kidney function in Diabetic Nephropathy.

Author Contribution

GD and MM framed research question, designed the study. GD, SK, MM performed the experiment and analysed the data, SKP recruited the patient. GD, MM, SK, SKP and GKS discussed the result and wrote the manuscript. MM will act as guarantor. All authors have discussed the result, and commented on the manuscript.

Funding

This work was supported by the Indian Council of Medical Research (ICMR) Grant Number- 3/2/Dec-2019/PG-Thesis-HRD (27).

Declarations

Conflict of interest

The authors have no relevant financial or non-financial competing interests to disclose.

Ethical Approval

This study was performed in line with the principles of the Declaration of Helsinki. Ethical clearance was approved by the Institutional Ethics Committee (IEC), AIIMS, Bhubaneswar, Ref number- IEC/AIIMS BBSR/PG Thesis/2019-20/78).

Consent to Participate

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