Study of the Association of -667 Aquaporin-2 (AQP-2) A/G Promoter Polymorphism with the Incidence and Clinical Course of Chronic Kidney Disease in Korea

Abstract

Background. Impaired urinary concentration is uniformly present with advanced disease in chronic renal failure. Aquaporin-2 (AQP-2) is known to be expressed in the renal collecting duct cells and participates in urinary concentration in response to vasopressin. Recently, the study of AQP expression in various forms of chronic kidney disease (CKD) demonstrated a reduction in AQP-2 expression associated with a loss of nephrons and the presence of chronic interstitial fibrosis. No information on aquaporin genetic variations in CKD is available to date. The aim of our study was to evaluate the possible impact of aquaporin-2 genotype on the development and clinical course of CKD. Methods. Blood samples from 259 patients with CKD and 106 ethnicity-, age-, and sex-matched healthy controls were collected, and genomic DNA was extracted. AQP-2 -667 genotype was assessed by PCR, followed by restriction fragment length polymorphism analysis. Results. There were no significant differences in genotype and allele frequencies between the patients and healthy controls (p = 0.3936, p = 0.2941, respectively). In all, 79 (30.5%) patients had the AQP-2 -667 wild-type A/A, 123 (47.5%) were heterozygous for the G allele, and 57 (22.0%) patients showed homozygosity. After subclassification of CKD according to underlying disease, no significant differences were observed between those patients and controls (p = 0.72 for diabetic nephropathy, p = 0.52 for hypertensive nephropathy, p = 0.27 for chronic glomerulonephritis, and p = 0.80 for unknown etiology). Genotype and allele frequencies of the AQP-2 gene polymorphism (rs3759126) of hypertensive patients in pre-ESRD did not show a noticeable difference compared with normal blood pressure patients in pre-ESRD (p = 0.50). No correlation was found to exist between the AQP-2 gene polymorphism (rs3759126) and serum electrolyte levels in pre-ESRD patients (p = 0.38 for serum sodium level and p = 0.44 for serum potassium level). Conclusion. Our data indicate that no association exists between the -667 AQP-2 A/G polymorphism and susceptibility to CKD or its clinical course.

Keywords:

• aquaporin-2
• chronic kidney disease
• polymorphism

INTRODUCTION

In chronic renal failure (CRF), the limited capacity to concentrate the urine often correlates with other measures of impaired renal function. Isosthenuria is therefore an almost universal finding when the glomerular filtration rate (GFR) falls below 25 mL/min. At this level of GFR and below, urine osmolality does not rise even when supraphysiologic doses of arginine vasopressin (AVP) are administered, suggesting that the concentrating defect relates to impaired concentrating capacity in surviving nephrons.[1]

Aquaporin-2 (AQP-2) is known to be expressed in the renal collecting duct cells and participates in urinary concentration in response to AVP.[2],[3] The urinary excretion of AQP-2 is decreased in accordance with the development of nephropathy in diabetes mellitus and some experimental animal models.[4–7] An impaired urinary concentration is uniformly present with advanced disease in chronic renal failure. In a 5/6 nephrectomy rat model of chronic renal failure, several defects have been identified, including near absence of V₂ receptor mRNA, as well as diminished AQP-2 and AQP-3.[8] More recently, immunohistochemical studies of biopsy samples from a wide range of human renal diseases revealed that there was judged qualitatively to be a reduction in the amounts of AQP-2 and AQP-3 expression, especially in lesions with substantial interstitial fibrosis and nephron loss, as compared with a healthy region of normal kidneys.[9]

To our knowledge, no information on aquaporin genetic variations in chronic kidney disease (CKD) is available to date. The involvement of aquaporins in renal pathophysiological processes prompted us to investigate the expression of aquaporin-2 genetic variations in CKD by genotyping, and to study the possible impact of aquaporin-2 genotype on the development and clinical course of CKD.

MATERIALS AND METHODS

The study was approved by the University Ethics Committee, and informed consent was obtained from all patients and controls. The study group consisted of 259 CKD patients followed up at the nephrology department, Inje University, Busan, Korea. Their GFR was less than 90mL/min, so their stage of CKD was only from stage II to stage V. The control subjects were 106 healthy individuals who underwent a 2003 heath examination at the Kyung Hee University Medical Center, Seoul, Korea. There was no clinical evidence for kidney disease or any other disorder in any of the controls. The characteristics of the groups are shown in Tables 1 and 2.

Table 1 Demographic characteristics of studied groups

	Chronic kidney disease (n = 259)	Controls (n = 108)
Age (years)	64.1 (46–78)	49.3 (13–82)
Gender
M	147 (56.8%)	61 (56.5%)
F	112 (43.2%)	47 (43.5%)

Table 2 Clinical characteristics of the chronic kidney disease patients

PreESRD vs. ESRD	46 (17.8%) vs. 213 (82.2%)
Etiology of CKD^†	Diabetes 64 (24.7%)
	Hypertension 81 (31.3%)
	Chronic glomerulonephritis 60 (23.2%)
	Tubulointerstitial disease 3 (1%)
	Polycystic kidney disease 6 (2%)
	Unknown etiology 34 (13.1%)
	Chronic obstructive uropathy 2 (1%)
	Renovascular disease 4 (1%)
	Systemic lupus erythematosus 4 (1%)
	Alport's syndrome 1
Renal replacement therapy in ESRD	Kidney transplantation 36 (16.9%)
	Hemodialysis 150 (70.4%)
	Peritoneal dialysis 27 (12.7%)
Percent of hypertension in preESRD	38/46 (82.6%)
Serum electrolyte level in preESRD
Sodium (mEq/L)	140.4 ± 3.2
Potassium (mEq/L)	4.9 ± 0.7
Creatinine (mg/dL)	2.6 ± 1.5

*End stage renal disease, ^†chronic kidney disease.

Polymerase chain reaction and restriction fragment length polymorphism (PCR-RFLP) was performed to detect genotypes in the DNA sequence variants of the AQP-2 promotor (-667 G/A polymorphism, rs3759126). Genomic DNA was prepared from heparinized venous blood sampling using a Core-one™ blood genomic DNA isolation kit (Corebiosystem, Seoul, Korea). The forward primer was 5′-GTG TGC AAA GAC CTG TTC TCC T‐3′, and reverse primer was 5′-TGA CTC AAA CTG TTG ACC AAG GC-3′. A total of 20 μl of the reaction mixture contained 200 ng of genomic DNA, 10 pmol of each primer, 200 μM of each dNTP, and buffers recommended by the manufacturer. AQP-2 promotor (from ‐916 to −352) of DNA samples was amplified by GeneAmp PCR system 9600 (Applied Biosystems, Foster City, California, USA) under the following conditions: 35 cycles of 30 sec at 95°C, 30 sec at 62°C, and 30 sec at 72°C. The extension time of the last cycle was set at 7 min. PCR products were digested by the corresponding restriction enzyme Sma I (New England BioLab, Beverly, Massachusetts, USA) for 16 h at 25°C. Digestive products were separated from 2.0% agarose gels stained with ethidium bromide. Samples exhibiting 565-bp band were assigned as AA, samples revealing two bands of 316- and 249-bp were typed as GG, and samples illustrating three bands of 565-, 316- and 249-bp were assigned as AG (see Figure 1).

Figure 1. Electrophoresis of Sma I polymorphism in AQP-2 gene. Lane 1 is AG genotype; Lane 2 is GG genotype; Lane 3 is AA genotype; Lane M is size marker, 300bp.

Allele frequencies were estimated from genotype distribution. Observed genotype frequencies were tested against Hardy-Weinberg equilibrium using a χ²-test of fit. Differences in clinical parameters between different genotypes in the AQP-2 gene were tested with either a χ²- test or Fisher's exact test. Correlation analysis of continuous variables was carried out according to the Kruskal-Wallis test. A p value of less than 0.05 was considered to be statistically significant for all the analyses, all of which were performed using SAS 8.1 for Windows.

RESULTS

The genotype and allele frequencies of the AQP-2 gene polymorphism are presented in Table 3. There were no significant differences in genotype and allele frequencies between the patients group and the healthy control group (p = 0.3936, p = 0.2941, respectively). After subclassification of CKD according to underlying disease, the genotype and allele frequencies were reanalyzed. No significant differences were observed between any of the groups and the control group (see Table 4 p = 0.72 for diabetic nephropathy, p = 0.52 for hypertensive nephropathy, p = 0.27 for chronic glomerulonephritis, and p = 0.80 for unknown etiology).

Table 3 Genotype and allele frequencies of AQP-2 gene polymorphism in studied groups

AQP-2 (Sma I)	Genotype distribution (%)			Allele frequency (%)
	AA	AG	GG	A	G
Control (n = 108)	36 (33.33)	55 (50.94)	17 (15.74)	127 (58.80)	89 (41.20)
CKD (n = 259)	79 (30.50)	123 (47.50)	57 (22.00)	281 (54.25)	237 (45.75)
HWP^†	Case: 0.481, control: 0.5958
p value	0.3936			0.2941

*Chronic kidney disease, ^†Hardy-Weinberg equilibrium.

Table 4 Genotype and allele frequencies of AQP-2 gene polymorphism in subgroups of chronic kidney disease

	Genotype distribution (%)			Allele frequency (%)		p value (genotype)(allele)
AQP-2 (Sma I)	AA	AG	GG	A	G
Controls (n = 108)	36 (33.33)	55 (50.94)	17 (15.74)	127 (58.80)	89 (41.20)
Diabetic nephropathy (n = 64)	21 (32.81)	30 (46.88)	13 (20.31)	72 (56.25)	56 (45.75)	(0.7242) (0.6531)
Hypertension (n = 81)	24 (29.63)	39 (48.15)	18 (22.22)	87 (53.70)	75 (46.30)	(0.5196) (0.3462)
CGN^† (n = 60)	15 (25.0)	30 (50.0)	15 (25.0)	60 (50.0)	60 (50.0)	(0.2743) (0.1366)
Unknown^‡ (n = 34)	10 (29.41)	17 (50.0)	7 (20.59)	37 (54.41)	31 (45.59)	(0.8042) (0.5742)

*Hypertensive nephropathy, ^†chronic glomerulonephritis, ^‡unknown etiology.

We focused on the association of the AQP-2 gene polymorphism (rs3759126) with hypertension only in pre-end stage renal disease (pre-ESRD) patients because multiple variable factors associated with the development of hypertension exist in ESRD patients. Genotype and allele frequencies of the AQP-2 gene polymorphism of hypertensive patients in pre-ESRD did not show any noticeable difference compared with normal blood pressure patients in pre-ESRD (see Table 5; p = 0.50).

Table 5 Genotype and allele frequencies of AQP-2 gene polymorphism according to blood pressure level in pre-end stage renal disease patients

AQP-2 (Sma I)	Genotype distribution (%)			Allele frequency (%)
	AA	AG	GG	A	G
Normal BP groups (n = 8)	2 (25.0)	4 (50.0)	2 (25.0)	8 (50.0)	8 (50.0)
Hypertensive groups (n = 38)	14 (36.84)	20 (52.63)	4 (10.53)	48 (63.16)	28 (36.84)
p value	0.4958			0.4849

*Blood pressure.

We also focused on the correlation of the AQP-2 gene polymorphism (rs3759126) with serum sodium and potassium level only in pre-ESRD patients because multiple variable factors associated with the electrolyte imbalance exist in ESRD patients. No correlation was found to exist between the AQP-2 gene polymorphism (rs3759126) and serum electrolyte levels in pre-ESRD patients (p = 0.38 for serum sodium level and p = 0.44 for serum potassium level). We present the serum electrolyte levels, according to the AQP-2 gene polymorphism, in Table 6.

Table 6 Serum electrolyte levels according to AQP-2 gene polymorphism in pre-ESRD

AQP-2 genotype	Number of patients	Electrolyte	Mean ± SD	Minimum	Maximum
AA	16	Na^†	141.00 ± 2.13	136.00	145.00
		K^‡	4.86 ± 0.61	4.00	6.20
		Cr^§	2.44 ± 1.58	0.90	6.40
AG	24	Na^†	140.54 ± 2.80	135.00	146.00
		K^‡	4.86 ± 0.82	3.70	6.50
		Cr^§	2.53 ± 1.44	1.00	6.10
GG	6	Na^†	138.83 ± 6.15	127.00	144.00
		K^‡	5.20 ± 0.55	4.70	6.20
		Cr^§	3.02 ± 1.82	1.20	6.00

*End stage renal disease, ^†sodium, ^‡potassium, ^§creatinine.

DISCUSSION

Impaired urinary concentration capacity occurs in both acute and chronic renal failure. This concentrating defect does not respond to AVP, therefore defining a nephrogenic diabetes insipidus state. In a renal artery clamp model of ischemic acute renal failure, the major structural abnormalities occurred in the proximal nephron, with the collecting ducts generally spared from apparent structural damage. However, it has been shown in this model that not only AQP-1, but also collecting duct AQP-2, -3, and -4 are decreased.[10] In addition to these water channel defects, it seems highly likely that components of the countercurrent concentrating mechanism, which generates the hyperosmotic driving force for water transport through the AQP water channels, are also impaired. The same is no doubt true in chronic renal failure, in which an impaired urinary concentration is uniformly present with advanced disease. In ischemia-induced renal failure and chronic renal failure induced by partial nephrectomy, there are lower levels of aquaporin-1 and −2.[11] However, all of these experiments involved rats, and recently, aquaporin expression has been studied in benign human renal diseases in two reports.[9],[12] In the first report,[9] immunohistochemical studies of biopsy samples from a wide range of renal diseases revealed a substantial and striking upregulation of AQP-1 in the glomeruli of most diseased kidneys. AQP-1 expression remained prominent in proximal tubules in all lesions. In contrast, there was judged qualitatively to be a reduction in the amounts of AQP-2 and AQP-3 expression, especially in lesions with substantial interstitial fibrosis and nephron loss, as compared with a healthy region of normal kidneys. The results were quantitatively confirmed by real-time reverse transcriptase-PCR. In the other report,[12] there was less aquaporin-2 immunoreactivity in human renal explants diagnosed with chronic renal allograft dysfunction (CRAD), starting from the medullary region. In explants with no CRAD and viable collecting ducts, or in normal controls, aquaporin-2 immunoreactivity remained unchanged.

Another study,[4] undertaken to determine whether progression of renal dysfunction affects urinary excretion of AQP-2 in diabetic nephropathy, had 8 control subjects and 14 patients with type 2 diabetes who were classified into two groups according to serum creatinine level (cut-off point; 1.5 mg/dL). After an 8-hour water deprivation, both urinary osmolality and urinary excretion of AQP-2 significantly decreased in the diabetic patients with chronic renal failure as compared to two other subjects (p < 0.0001, p < 0.05, respectively). After a water load (10 mL/kg), no differences were found in plasma osmolality, AVP levels, and urine osmolality, whereas urinary excretion of AQP-2 significantly decreased in the patients with chronic renal failure as compared to the control subjects (p < 0.05). These results indicate that the decreased urinary excretion of AQP-2 in diabetic nephropathy is due to the impaired cellular signaling of AVP in collecting duct cells, which may be partly involved in the urinary concentrating defect in renal failure.

On the other hand, the dysregulation of AQP-2 release is associated with hyponatremia, hypokalemia, or water retention in human disease.[13] Thus, he involvement of aquaporins in renal pathophysiological processes prompted us to investigate the possible impact of AQP-2 genetic variations on the development of CKD or specific renal disease and clinical outcome (i.e., the development of hypertension or electrolyte imbalance). However, the present study revealed that AQP-2 polymorphism (rs 3759126) does not affect the development and clinical outcome of CKD in Korea. We think that an association study between V2 receptor gene polymorphism and CKD may be needed because a decreased excretion of AQP-2 in chronic renal failure might be linked to the downregulation of V2 receptor in previous study.[14] In order to evaluate the impact of aquaporin-2-667 genotype on the specific cause of CKD, further incorporation of additional cases may be needed for subset analysis. However, we conclude that the sample size, although small, created no critical limitations, because the main objective of this study was to elucidate the effect of aquaporin-2-667 genotype on the overall incidence of CKD.

After comparing the genotype distribution (rs3759126) in healthy controls of the present study with the genotype distribution (rs3759126) of the International HapMap project (http://www. hapmap.org), it was found to be very similar to that reported in Chinese population. The fact that it differs greatly from those observed in Japanese and African populations, may indicate possible ethnic or racial differences in genetic background (see Table 7). It also suggests that positive findings in the other populations may be possible in an association study of AQP-2 gene polymorphism and CKD. Only a single polymorphism in the promoter region of the gene was analyzed in this study. An association study between other AQP2 gene polymorphism and CKD may be needed in the future.

Table 7 Comparison of genotype distribution in -667 AQP-2 in control populations

AQP-2 (Sma I)		Genotype distribution (%)
	Ethnic	AA	AG	GG
This study	Korean	33.3%	50.9%	15.8%
IHmP	Chinese	27.3%	54.5%	18.2%
IHmP	Japanese^†	47.7%	36.4%	15.9%
IHmP	Nigerian^‡	55.0%	38.3%	6.7%

Abbreviation: *IHmP = International HapMap project.

^†p = 0.0075, ^‡p < 0.0001, as compared with this study.

There are several mechanisms for regulation of AQP-2 expression. First, vasopressin induces cAMP phosphorylation and subsequent binding of the cAMP response element in the promoter region of the aquaporin-2 gene.[15] These events initiate aquaporin-2 mRNA and protein synthesis. Second, the osmolality and solute composition of the renal interstitium are potent non-vasopressin-dependent regulators of aquaporin-2. A putative tonicity responsive element, identified in the human aquaporin-2 gene promoter, is thought to initiate increased aquaporin-2 expression during elevated effective osmolality.[11] To actually reveal the association between DNA polymorphisms and phenotype, it is identified that single nucleotide polymorphism (SNP) affects the transcription of the gene or protein expression. However, we did not test it through western blot of urinary AQP-2 by three genotypes in CKD subjects. This is another limitation in the present study.

To the best of our knowledge, this is the first report of aquaporin-2 gene polymorphism in CKD. Notwithstanding the limitations noted, the present study might serve to stimulate further world-wide interest in genetics of kidney disease for more detailed investigations.

In summary, our data suggest that -667 AQP-2 A/G gene polymorphism is not associated with the incidence and clinical course of chronic kidney disease in the Korean population.

ACKNOWLEDGMENT

The authors thank MS Hong, EY Ha and Y You for their excellent technical help as well as HS Sohn for help with statistical analyses.

Notes

This work was supported by the 2005 Inje University Grant for Research Institute.

This study was presented in part at ERA-EDTA Congress, July 15–18, 2006, Glasgow, United Kingdom, and has been published in abstract form.