Association of Urinary Activity of MMP-2 with Microalbuminuria in an Isolated Sample of Subjects Living in High Altitude Rural Locations in México

Abstract

Hernández-Hernández, Magda Elena, Jaime Morales-Romero, Clara Luz Sampieri, Diego Jesús Luna Lozano, Isidra del Carmen Valencia Lezama, Mónica Janett Muñoz Contreras, and Arturo Rodríguez Hernández. Association of urinary activity of MMP-2 with microalbuminuria in an isolated sample of subjects living in high altitude rural locations in México. High Alt Med Biol. 18:209–218, 2017.—Matrix metalloproteinases (MMP) are implicated in remodeling of the renal extracellular matrix. In a cross-sectional study we evaluated renal impairment in general population of high-altitude rural locations in México. Multivariable analysis was performed to identify the association between MMP-2 and MMP-9 and microalbuminuria. Twenty-eight (20.9%) subjects with renal impairment (WRI) and 106 (79.1%) without renal impairment were included. No differences were found relating to sex, location, marital status, current habits, weight, height, body mass index, waist size in males, creatinine in males, and uric acid. In contrast, differences were found among age, level of education, waist size in general and in females, creatinine in general and in females, urinary albumin, urea, glucose, total cholesterol, and triglycerides. Proportions of hypertension, type 2 diabetes mellitus, central abdominal obesity, hypertriglyceridemia, and hypercholesterolemia were greater in the group WRI. Presence of urinary MMP-2 or of both urinary gelatinases and arbitrary unit (AU) values ≥P₉₀ were associated with microalbuminuria. We conclude that AU values ≥P₉₀ of urinary MMP-2 (OR = 20.1, p = 0.002) is associated with microalbuminuria.

Introduction

Humans that are born and raised in regions of high altitude (HA), defined as ≤2400 m above sea level (asl), appear to have a substantial advantage over those born and raised at sea level, given the adaptation of the former group to survive under conditions of hypoxia (Moore, 2001). Changes in renal function at HA arise from the direct effects of hypoxia on the kidney, as well as from multiple compensatory adaptations, including cardiac output, ventilation, oxygenation of blood, and erythropoiesis (Hurtado et al., 2012). The kidneys play a central role in human adaptation to HA through a hypoxic diuretic response, reducing the concentration of antidiuretic hormones, renin and aldosterone, and increasing levels of natriuretic hormones, epinephrine, norepinephrine, and adrenomedullin, causing a reduction in total body water and plasma volume and increased hemoglobin values (Haditsch et al., 2007).

While little is known regarding risks in subjects with renal damage or human chronic kidney disease (CKD), there is a concern that HA may accelerate progression (Luks et al., 2008). Acute hypoxia has been associated with an increase in the excretion of albumin (Hansen et al., 1994) and proteinuria (Winterborn et al., 1987). The mechanisms behind this increase are not fully understood; however, it may be the result of an increase in glomerular permeability or a decrease in tubular reabsorption (Winterborn et al., 1987).

The absence of sclerosis and the integrity of the glomerular and tubular basement membranes are keys to appropriate glomerular filtration rate (GFR) (Mason and Wahab, 2003). Imbalances between the synthesis and degradation of the renal extracellular matrix are thought to play important roles in the progression of nephropathy (Han et al., 2006). Matrix metalloproteinases (MMPs) are a family of zinc-dependent proteins that together have the capacity to break down virtually all components of the extracellular matrix. Currently, there are 28 human MMPs (Dimas et al., 2013). In addition to their role in extracellular matrix turnover, MMPs have been shown to activate several growth factors, such as TGF-β, TNF-α, and IGF, implicated in renal scarring and fibrosis, renal hypertrophy, and tubular cell proliferation (Thrailkill et al., 2009).

The MMP-9 (gelatinase-B) mRNA has been proposed as a marker of epithelial mesenchymal transition, which correlates with progression of diabetic nephropathy (Zheng et al., 2011). The MMP-2 (gelatinase-A) degrades basement membrane fibrillar collagen and gelatin (Sanders et al., 2004) and induces the transformation of renal tubular epithelium to the myofibroblastic phenotype, a critical step in renal interstitial fibrosis (Cheng et al., 2003). In rats with experimental early stage proteinuria, MMPs appear to contribute to the breakdown of the glomerular basement membrane, whereas at later stages, they contribute to removal of the extracellular matrix associated with scarring and fibrosis (Ronco and Chatziantoniou, 2008). It has been proposed that MMP-2 reflects the degradation of the glomerular basement membrane, whereas MMP-9 reflects degradation of the mesangial matrix (Nagano et al., 2009). Increased urinary MMP-9 levels have been detected in humans with type 2 diabetes mellitus (T2DM) and diabetic nephropathy, and correlate with the degree of albuminuria (Tashiro et al., 2004; Lauhio et al., 2008).

In the present study, we measured MMP-2 and MMP-9 activities in urine samples of two HA rural communities in México to determine their performance as predictors of albuminuria, using risk factors and common and economical clinical determinations included in urine test strips as covariates.

Materials and Methods

Design and study population

A cross-sectional study was conducted (August–July 2012) in the general population of two HA rural communities located in the state of Veracruz, México. The communities were “El Conejo,” with 1044 inhabitants at 3241 m asl, and “Los Pescados” with 1555 inhabitants at 2981 m asl (separated by 3.2 km). Both rural communities have similar geology, vegetation, climate, land use, solar radiation characteristics, and customs. Criteria for inclusion in the study were subjects of age ≥20 years and who fulfilled the clinical requirements for calculation of GFR according to the guide KDOQI (NFK, 2002). Subjects with body temperature ≥37.5°C, undergoing menstruation, and with hematuria were excluded.

Sampling

Random sampling of households was performed geographically. In cases where the selected household did not contain anyone who fulfilled the criteria of selection, or where the occupants declined to participate, another household was chosen at random.

Ethics

The research protocol was approved by the Instituto de Salud Pública of the Universidad Veracruzana. This study was conducted according to the principles of the Declaration of Helsinki. Those who agreed to participate signed an informed consent.

Clinical, biochemical, and biomarker assessments

Personnel took the blood samples, following 8 hours of fasting, as well as taking anthropometric and vital sign measurements. Serum was obtained directly in the rural communities. In serum samples, glucose, urea, creatinine, cholesterol, triglycerides, uric acid, and urinary albumin in serum (ERBA XL-200 BioSystems Laboratories) were determined. Presence of blood in the urine was determined with Multistix 10 SG^® Urine Reagent Strip.

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis

Urine samples were centrifuged at 2000 g for 10 minutes at 4°C (Altemtam et al., 2012). Carrying buffer with β-mercaptoethanol was then added and sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) was performed using 5% polyacrylamide stacking gel and 10% resolving gel. Albumin (Bio Basic, Inc.), a molecular weight marker and a sample of urine with a diagnostic of normoalbuminuria were simultaneously loaded onto the gel as controls. Gels were run in standard Tris glycine-SDS running buffer. Following electrophoresis, the protein bands were then made visible using Coomassie Blue and Sliver staining (Marshall and Williams, 1993).

Quantitative zymography in gelatin

Quantitative zymography in gelatin was conducted as reported previously (Sampieri et al., 2010). Electrophoresis was performed using 5% polyacrylamide stacking gel and 10% resolving polyacrylamide gel copolymerized with 1 mg/mL gelatin. MMP-2/MMP-9 standards were simultaneously loaded onto the gel. Gelatin gels were washed overnight at room temperature and incubated in 50 mmol/L Tris-HCl pH 7.5, 5 mmol/L CaCl₂ or 10 mmol/L EDTA for 18 hours. Densitometry was quantified using ImageJ 1.47v software. Total MMP-9 was calculated by adding 250, 110, 92, and 82 kDa bands, whereas total MMP-2 by adding 72 and 62 kDa bands (Sampieri et al., 2010). Arbitrary Units (AU) were determined for each individual band from a standard curve (Sampieri et al., 2010).

Analysis of protein patterns

A courtesy license for GelCompar II 6.6 (Applied Maths NV) was used for the hierarchical clustering analysis and calculation of molecular weights. Groups of patterns generated were classified as: group A—low risk of presenting an abnormal pattern of urinary proteins and group B—high risk of presenting an abnormal pattern of urinary proteins (Kshirsagar and Wiggins, 1986).

Variables

The GFR was estimated according to the KDOQI guide, with classification as “no renal impairment” (NRI): GFR ≥60 mL/(min ·1.73 m²) and ≤2.9 mg/dL urinary albumin, and “with renal impairment” (WRI): GFR ≥60 mL/(min ·1.73 m²) and ≥3.0 mg/dL urinary albumin or with GFR ≤59 mL/(min ·1.73 m²), regardless of the level of urinary albumin. Hypertension: systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg, use of antihypertensive medication, or report of medical diagnosis. T2DM status: use of insulin or hypoglycemic agents or ≥126 mg/dL serum glucose under fasting, or report of medical diagnosis. Hypercholesterolemia: ≥200 mg/dL total serum cholesterol under fasting. Hypertriglyceridemia: ≥150 mg/dL total serum triglycerides under fasting. Hyperuricemia: ≥6.1 mg/dL serum uric acid under fasting for females and ≥7.1 mg/dL for males. Central obesity (CO): ≥81 cm waist size for females and ≥91 cm for males. Overweightness (OW): body mass index (BMI) ≥25 kg/m². Normoalbuminuria: ≥2.9 mg/dL urinary albumin. Microalbuminuria: 3–29 mg/dL urinary albumin. Macroalbuminuria: ≥30 mg/dL urinary albumin.

Statistical analysis

Categorical variables were summarized using absolute frequencies and percentages; continuous variables of non-normal distribution were expressed through medians and the interquartile range. Proportions were compared by chi-square test, or Fisher's exact test, and medians were compared by Mann–Whitney U test. MMP-2 and MMP-9 were dichotomized as present (>0 AU) or absent (0 AU) and according to their 90th percentile (P₉₀). Age, BMI, waist size, urea, and total cholesterol were compared between six urinary protein patterns using one-way analysis of variance and MMP-2 AU, MMP-9 AU, glucose, GFR, creatinine, triglycerides, uric acid, and urinary albumin were compared using the Kruskal–Wallis test. Chi-square for trends was performed to compare presence or absence of MMPs between subjects with and without microalbuminuria and between groups A and B. Multivariable analysis was performed to identify the association between MMP-2 and MMP-9 and microalbuminuria. Five models were constructed, in which the possible covariates were: age, sex, leukocyturia, nitrituria, bilirubinuria, glycosuria, T2DM, hypertension, hypercholesterolemia, hypertriglyceridemia, and OW. The ORs and CI95% were calculated by logistic regression. A p < 0.05 was considered statistically significant. EpiDat version 3.1 and IBM-SPSS Statistics, version 22.0 were used for analyses.

Results

A total of 194 subjects were weighed and measured; 174 subjects fulfilled the clinical criteria for calculation of GFR, 5 were excluded for presenting a BMI <18.5 kg/m² and 15 for presenting a BMI >35 kg/m²; in addition, 32 subjects were excluded for hematuria, fever, or for not providing samples. The nonresponse rate was 5.5%. Subjects included in the study were 134. All of the samples were subjected to clinical serum and urine analyses; 133 urine samples were analyzed with SDS-PAGE and 132 urine samples were analyzed by quantitative zymography.

General characteristics

Of the 134 subjects included, 28 corresponded to the group WRI (20.9%) and 106 (79.1%) to that of NRI. On comparison of the sociodemographic characteristics and current consumption of tobacco and alcohol, no differences were found in sex, location, marital status, and habits between the groups. In contrast, differences were found in the average age and educational level. Within the population in general, a low percentage (≤24%) reported having remunerated occupations, such as farmer or employee (Table 1).

Table 1.

General Characteristics of 134 Participants According to Renal Impairment Status

	General n = 134	No renal impairment n = 106	With renal impairment n = 28
Characteristic	n (%)	n (%)	n (%)	p value
Sex
Female	104 (77.6)	83 (78.3)	21 (75.0)	0.71
Male	30 (22.4)	23 (21.7)	7 (25.0)
Location
“El Conejo”	69 (51.5)	52 (49.1)	17 (60.7)	0.27
“Los Pescados”	65 48.5	54 (50.9)	11 (39.3)
Age (years)	44.2 [14.5]	41.2 [13.7]	55.9 [11.5]	<0.001
Marital status
With partner	118 (88.1)	94 (88.7)	24 (85.7)	0.74
Without partner	16 (11.9)	12 (11.3)	4 (14.3)
Occupation
Housewife	101 (75.4)	80 (75.5)	21 (75.0)	NA
Farmer	31 (23.1)	24 (22.6)	7 (25.0)
Student	1 (0.7)	1 (0.9)	0 (0.0)
Employee	1 (0.7)	1 (0.9)	0 (0.0)
Education
Lower than primary	61 (45.5)	43 (40.6)	18 (64.3)	0.03
Primary or above	73 (54.5)	63 (59.4)	10 (35.7)
Currently smoking tobacco
No	119 (88.8)	92 (86.8)	27 (96.4)	0.19
Yes	15 (11.2)	14 (13.2)	1 (3.6)
Currently drinking alcohol
No	107 (79.9)	87 (82.1)	20 (71.4)	0.21
Yes	27 (20.1)	19 (17.9)	8 (28.6)

Values for categorical variables are given as frequency (percentage); age as mean [±standard deviation]. All percentages are calculated according to column.

NA, not applicable.

Anthropometric and biochemical measurements

The groups WRI and NRI were compared in terms of their anthropometric and biochemical characteristics. No differences were found between the groups in weight, height, BMI, waist size in males, creatinine in males, and uric acid in general and within each sex. In contrast, general waist size, waist size in females, general creatinine, creatinine in females, urinary albumin, urea, glucose, total cholesterol, and triglycerides were higher in the group WRI, whereas GFR was greater in the group NRI (Table 2).

Table 2.

Anthropometric and Biochemical Variables of 134 Participants According to Renal Impairment Status

Characteristic	General n = 134		No renal impairment n = 106		With renal impairment n = 28		p value
Weight (kg)	65.8	[10.3]	65.5	[10.2]	67.2	[11.0]	0.44
Height (cm)	155.0	150.0, 160.0	155.0	150.0, 160.0	154.5	149.5, 159.5	0.73
BMI (kg/m²)	27.2	[3.8]	27.0	[3.9]	28.0	[3.6]	0.23
WS general (cm)	91.0	84.0, 100.0	90.0	82.0, 98.0	97.0	90.5, 102.0	0.003
WS female (cm)	90.0	83.5, 100.0	88.0	81.2, 98.8	96.0	89.5, 101.0	0.02
WS male (cm)	94.0	88.0, 103.0	91.0	88.0, 97.8	103.0	95.0, 103.0	0.08
SC general (mg/dL)	0.9	0.82, 1.0	0.9	0.8, 0.8	1.0	0.8, 1.1	0.01
SC female (mg/dL)	0.9	0.8, 1.00	0.8	0.8, 0.9	1.0	0.8, 1.1	<0.01
SC male (mg/dL)	1.1	0.9, 1.2	1.1	0.9, 1.2	1.1	0.9, 1.4	0.45
UAlb (mg/dL)	0.7	0.1, 1.3	0.6	0.1, 1.1	3.3	0.0, 5.1	<0.01
SUA general (mg/dL)	5.5	4.6, 6.6	5.3	4.6, 6.6	5.9	4.6, 6.7	0.43
SUA female (mg/dL)	5.1	4.4, 5.9	5.0	4.4, 5.8	5.7	4.5, 6.1	0.40
SUA male (mg/dL)	7.4	6.3, 8.6	7.4	6.5, 8.5	7.8	6.3, 9.0	0.85
GFR mL/(min ·1.73 m²)	76.6	67.4, 89.6	79.2	70.7, 91.0	59.7	56.6, 80.9	<0.001
SU (mg/dL)	26.5	20.5, 33.2	25.3	19.8, 31.9	31.2	25.4, 38.2	0.01
SG (mg/dL)	92.7	83.6, 104.3	92.0	83.6, 101.2	111.3	84.6, 219.0	0.02
SC (mg/dL)	193.0	171.0, 219.0	190.0	164.0, 212.0	216.0	196.5, 244.0	<0.001
ST (mg/dL)	181.5	125.0, 258.0	170.0	122.0, 248.0	210.5	159.0, 312.5	0.03

Values are given as mean [± standard deviation] or median and interquartile range percentile 25, percentile 75. In the group of females, 83 subjects with no renal impairment are compared with 21 subjects with renal impairment. In the group of males, 23 subjects with no renal impairment are compared with 7 with renal impairment.

BMI, body mass index; WS, waist size; SC, serum creatinine; UAlb, urinary albumin; SUA, serum uric acid; GFR, glomerular filtration rate; SU, serum urea, SG, serum glucose, SC, serum cholesterol, ST, serum triglycerides.

Prevalence of conditions of impact on renal function

The presence of conditions related to renal function was compared between the groups WRI and NRI: hypertension, T2DM, CO, OW, microalbuminuria, macroalbuminuria, hyperuricemia, hypertriglyceridemia, and hypercholesterolemia. Proportions of hypertension, T2DM, CO, hypertriglyceridemia, and hypercholesterolemia were higher in the group WRI. In contrast, there were no differences in the frequency of OW and hyperuricemia between the two groups. By definition, only the subjects of the group WRI presented microalbuminuria and none of the subjects included presented macroalbuminuria (Table 3). In addition, the presence of T2DM was associated with microalbuminuria (chi-square 11.3, p < 0.001).

Table 3.

Conditions That Impact on the Renal Function of 134 Participants According to Renal Impairment Status

	General n = 134	No renal impairment n = 106	With renal impairment n = 28
Condition	n (%)	n (%)	n (%)	p value
Hypertension	50 (37.3)	33 (31.1)	17 (60.7)	0.004
T2DM	22 (16.4)	10 (9.4)	12 (42.9)	<0.0001
Central obesity	103 (76.9)	76 (71.7)	27 (96.4)	0.003
Over weight	99 (73.9)	76 (71.7)	23 (82.1)	0.26
Microalbuminuria^a	15 (11.2)	0 (0.0)	15 (54.0)	NA
Macroalbuminuria	0 (0.0)	0 (0.0)	0 (54.0)	NA
Hyperuricemia	41 (30.6)	31 (29.2)	10 (35.7)	0.51
Hypertriglyceridemia	87 (64.9)	63 (59.4)	24 (85.7)	0.01
Hypercholesterolemia	50 (37.3)	33 (31.1)	17 (60.7)	0.004

Values are given as frequencies (percentage). All percentages are calculated according to column.

By definition, only the subjects of the group with renal impairment presented microalbuminuria.

T2DM, type 2 diabetes mellitus.

Quantitative zymography

Zymography was used to analyze 26 samples of the group WRI and 106 of the group NRI; in each sample, AU were determined for MMP-2 and MMP-9 and, based on these values, the samples were categorized according to P₉₀. The individual or simultaneous absence or presence of MMP-2 and MMP-9 and AU values of ≥P₉₀ were compared among subjects with and without T2DM, hypertriglyceridemia, hypercholesterolemia, CO, OW, and hyperuricemia, as well as between the groups WRI and NRI. The presence of MMP-2 was associated to a greater extent (p = 0.02) with hyperuricemia (19/74 [20.4%]), compared with those without hyperuricemia (2/37 [5.1%]). The presence of both gelatinases was associated with a value bordering significance (p = 0.05), with the group WRI (3/23 [11.5%]), compared with the group NRI (2/104 [1.9%]).

The presence of MMP-2 or of both gelatinases and AU values ≥P₉₀ were associated with subjects with microalbuminuria, compared with those with normoalbuminuria (Table 4). The separate presence of MMP-9 and AU values ≥P₉₀ were not associated with subjects with microalbuminuria (Table 4).

Table 4.

Arbitrary Units and Interquartile Range of Enzymatic Activity of MMP-2 and MMP-9 of 132 Participants According to Urinary Albuminuria Status

	General n = 132	NAlb^a status n = 119	MAlb^a status n = 13
MMP status	n (%)	n (%)	n (%)	p value
MMP-2 and MMP-9 absent	98 (74.2)	91 (76.5)	7 (53.8)	0.01
MMP-2 or MMP-9 present	29 (22.0)	26 (21.8)	3 (23.1)
MMP-2 and MMP-9 present	5 (3.8)	2 (1.7)	3 (23.1)
MMP-2 and MMP-9 absent	98 (95.1)	91 (97.8)	7 (70.0)	0.002
MMP-2 and MMP-9 present	5 (4.9)	2 (2.2)	3 (30.0)
MMP absent or present any	127 (96.2)	117 (98.3)	10 (76.9)	0.002
MMP-2 and MMP-9 present	5 (3.8)	2 (1.7)	3 (23.1)
MMP-2 absent	111 (84.1)	104 (87.4)	7 (53.8)	0.006
MMP-2 present	21 (15.9)	15 (12.6)	6 (46.2)
MMP-9 absent	114 (86.4)	104 (87.4)	10 (76.9)	0.54
MMP-9 present	18 (13.6)	15 (12.6)	3 (23.1)
MMP-2 <P₉₀ AU^a	119 (90.2)	111 (93.3)	8 (61.5)	0.002
MMP-2 ≥P₉₀ AU	13 (9.8)	8 (6.7)	5 (38.5)
MMP-9 <P₉₀ AU^b	119 (90.2)	107 (89.9)	12 (92.3)	0.83
MMP-9 ≥P₉₀ AU	13 (9.8)	12 (10.1)	1 (7.7)

Values are given as frequencies (percentage). All percentages are calculated according to column.

Percentile 90 = 0.05 AU.

Percentile 90 = 0.03 AU.

NAlb, normoalbuminuria; MAlb, microalbuminuria; MMP, matrix metalloproteinases; P, percentile; AU, arbitrary units.

Patterns of urinary proteins

Twenty-seven samples of the group WRI and 106 of the group NRI were analyzed using SDS-PAGE. The patterns of urinary proteins with silver dye were examined using hierarchical cluster analysis (Fig. 1), distinguishing six different patterns. Pattern I: No dyeing n = 9 (7%); Pattern II: Dyeing in one band only that comigrates with albumin, n = 12 (9%); Pattern III: Dyeing in the band that comigrates with albumin and a signal around 100 kDa, n = 16 (12%); Pattern IV: Dyeing in the band that comigrates with albumin, one signal around 100 kDa, and one or two signals in the zone of 37–25 kDa, n = 25 (19%); Pattern V: Dyeing in the band that comigrates with albumin, no signal or one or two signals in the zone of 50–37 kDa, and one or two signals in the zone of 37–25 kDa, n = 50 (38%); Pattern VI: Dyeing in the band that comigrates with albumin, no signal or one or two signals in the zone of 75–100 kDa, one or two signals in the zone of 50–37 kDa; one or two signals in the zone of 37–25 kDa, with a greater number of bands than in pattern IV, n = 21, (16%).

FIG. 1.

Hierarchical clustering of urinary protein patterns from 133 subjects. Each row corresponds to a urine sample. Black lines are representing protein bands. +, higher molecular weight in comparison with urine albumin; −, lower molecular weight in comparison with urine albumin.

Comparisons were made in terms of the anthropometric, biochemical, MMP-2, and MMP-9 variables among the patterns of proteins I–VI. Differences were found (p = 0.008) in total cholesterol between pattern II (mean ± SD = 167.7 ± 23.0) and pattern III (mean ± SD = 214.3 ± 2.2). Differences were found (p < 0.001) for urinary albumin between pattern I (median = 0, P₂₅ = 0, P₇₅ = 0.05) and pattern II (median = 0.95, P₂₅ = 0.6, P₇₅ = 1.3), pattern V (median = 0.9, P₂₅ = 0.4, P₇₅ = 1.5) and pattern VI (median = 1.2, P₂₅ = 0.5, P₇₅ = 4.1).

The patterns of urinary proteins with silver dye I–VI were classified into: group A—patterns that appeared normal: I and II, n = 21, (16%), and group B—patterns that appeared abnormal: III, IV, V, and VI, n = 112 (84%). Between groups A and B, a comparison was made among the frequencies in the results of the variables: reagent strips for urinalysis; urinary albumin; groups NRI and WRI; diseases that impact on renal function; absence/presence and AU values of ≥P₉₀ of MMP-2 and MMP-9. No differences were found in any of these variables (data not presented).

Multivariate analysis of factors associated with microalbuminuria

Five multivariate models were constructed of factors that can predict microalbuminuria, introducing up to nine independent covariates. In all of the models, values ≥P₉₀ of MMP-2 were associated with microalbuminuria. In model 5, after adjusting for MMP-9 (levels ≥P₉₀), age, sex, presence of leukocyturia, nitrituria, bilirubinuria, glycosuria, T2DM, hypertension, hypercholesterolemia, hypertriglyceridemia, and OW, values ≥P₉₀ of MMP-2 (OR = 20.1, p = 0.002) remained associated with microalbuminuria (Table 5).

Table 5.

Multivariate Model of Predictors of Microalbuminuria

Model/covariates	OR	95% CI	p value
Model 1
MMP-9 ≥P₉₀	0.8	0.09–7.31	0.84
MMP-2 ≥P₉₀	8.6	2.29–32.61	0.001
Model 2
MMP-9 ≥P₉₀	0.5	0.04–7.04	0.62
MMP-2 ≥P₉₀	16.1	3.18–81.29	0.001
Age >45 years	7.8	1.62–37.72	0.01
Sex (female)	2.0	0.40–9.64	0.40
Model 3
MMP-9 ≥P₉₀	0.6	0.04–10.00	0.74
MMP-2 ≥P₉₀	16.6	3.23–85.32	0.001
Age >45 years	8.1	1.63–39.93	0.01
Sex (female)	1.9	0.38–9.78	0.43
Leukocyturia	0.6	0.04–9.41	0.72
Nitrituria	1.3	0.35–5.13	0.67
Bilirubinuria	0.8	0.12–5.24	0.79
Model 4
MMP-9 ≥P₉₀	0.9	0.05–15.48	0.93
MMP-2 ≥P₉₀	17.6	3.34–92.89	0.001
Age >45 years	5.7	1.10–29.66	0.04
Sex (female)	1.8	0.34–9.92	0.49
Leukocyturia	0.8	0.05–13.20	0.86
Nitrituria	1.4	0.36–5.74	0.61
Bilirubinuria	1.1	0.15–7.98	0.93
Glycosuria	4.6	0.87–24.11	0.07
Model 5
MMP-9 ≥P₉₀	1.03	0.03–32.77	0.99
MMP-2 ≥P₉₀	20.1	3.0–134.96	0.002
Age >45 years	5.3	0.74–38.49	0.10
Sex (female)	1.3	0.22–8.32	0.75
Leukocyturia	0.59	0.02–17.12	0.76
Nitrituria	1.1	0.23–5.62	0.88
Bilirubinuria	1.2	0.16–9.82	0.84
Glycosuria	1.3	0.10–16.56	0.84
T2DM	6.3	0.73–53.99	0.10
Hypertension	0.5	0.07–3.20	0.44
Hypercholesterolemia	3.1	0.61–15.60	0.17
Hypertriglyceridemia	1.1	0.13–8.39	0.96
Overweight	1.2	0.20–6.76	0.86

Dependent variable: Microalbuminuria was defined as >3 mg/dL of urinary albumin in a urinary sample. The OR was obtained by logistic regression using the method Enter. The covariates were categorical (with and without the characteristic of interest).

Discussion

This study revealed that AU values ≥P₉₀ for urinary MMP-2 are associated with microalbuminuria and that T2DM, hypertension, CO, hypertriglyceridemia, and hypercholesterolemia are associated with the group WRI in a HA rural population in México. To our knowledge, this is the first study to investigate gelatinolytic activity in urinary samples taken from a HA rural population, for which it is difficult to make comparisons. In addition to the fact that they did not feature determination of MMPs, the few studies conducted on HA populations have involved a reduced number of subjects (Luks et al., 2008).

The study by Chen et al. (2011) is possibly equivalent to the present study, although it differs in terms of the inclusion of subjects from HA populations that present hematuria. Chen et al. (2011) found 12.4% users of tobacco, 38.8% hypertension, and 29.1% hyperuricemia; percentages that were similar to those of this study, at 11.2%, 37.3%, and 30.6%, respectively. Apart from the fact that in this study subjects with BMI <18.5 and >35 kg/m² were excluded, since there may have been clinical situations in which clearance measures would be required to estimate GFR (NFK, 2002), the 37.3% prevalence of hypertension in this study is greater than the national prevalence of 31.5% (Instituto Nacional de Salud Pública, 2012). Given that hypertension is an ailment associated with obesity and that this study excluded subjects with BMI >35, it may be that this variable was underestimated.

Chen et al. (2011) found 14.4% microalbuminuria, whereas this study revealed 11.2%. There are several potential mechanisms that could explain the increased frequency of albuminuria and hypertension in HA residents. Hypoxia per se could be responsible for renal injury; indeed, hypoxia has been reported to stimulate extracellular matrix synthesis (Norman et al., 2000). Another possible mechanism could be driven by polycythemia, which may contribute to an increase in the frequency of proteinuria and hypertension (Chen et al., 2011), although polycythemia was not measured in this study. A further possibility is that hyperuricemia could play a role in renal injury (Jefferson et al., 2002). In this study, a frequency of hyperuricemia was found that was similar to that previously reported (Chen et al., 2011), although it was not associated with the group WRI.

The prevalence of 2.9% of T2DM (Chen et al., 2011) differs from the 16.4% found in this study, and the 41.2% of CO (Chen et al., 2011) differs from the 76.9% found in this study. Prevalence of CO and OW in this study exceeds the national values of 74.0% and 71.3%, respectively (Instituto Nacional de Salud Pública, 2012). The high prevalence of CO, OW, and T2DM found in this study reflect the epidemiological panorama of México, in recent decades an epidemic of obesity and diabetes has been documented. On the 1st of November 2016 the government issued for the first time two epidemiological alerts for noncommunicable diseases: EE-3-2016 epidemiological emergency of OW and obesity, and EE-4-2016 epidemiological emergency of diabetes (Centro Nacional de Programas Preventivos y Control de Enfermedades). These epidemics in México are the consequence of recent socioeconomic and nutritional transitions that introduced access to inexpensive, high-energy dense food, high rates of sedentary lifestyles, a lack of safe open spaces, socioeconomic vulnerability, reduced healthy meal options at schools, false or misunderstood sociocultural traditional beliefs, low rates of breastfeeding, and inequities in the healthcare system (Rivera et al., 2002; Aceves-Martins et al., 2016). Moreover, a polarization phenomenon has been documented in México: in the northern states epidemiological profiles similar to those of developed countries exist, whereas the southern states, with greater socioeconomic vulnerability, present pretransitional epidemiological profiles (Rivera et al., 2002). In this context, our area of study is located in the southern states.

We found an association of T2DM with the group WRI, because diabetic nephropathy results from interactions between metabolic and hemodynamic factors (Forbes et al., 2007) and hypertension is consistently reported in HA residents (Hurtado et al., 2012), therefore, a constant monitoring of arterial pressure and glycemia in this population is vital. Moreover, CO, greater levels of cholesterol and triglycerides were also associated with the group WRI, these findings differ from the literature that indicate that HA residents have a decreased frequency of T2DM (Chen et al., 2011; Hurtado et al., 2012) and obesity (Hurtado et al., 2012). The alarming public health situation in our country could probably have worse consequences in HA residents, given the potential combined effect of hypoxia and obesity that can drive obesity-related glomerulopathy, accompanied by lesions of focal and segmental glomerulosclerosis and eventually lead to microalbuminuria (Navarro Diaz, 2016).

Chen et al. (2011) reported stages of CKD, although this variable is not comparable to our results, since GFR was only estimated on one occasion in our study, it is noteworthy that we found a prevalence of 20.9% of the group WRI, which may be comparable to the reported prevalence of 19.1% of CKD. Although there are few studies of CKD at sea level compared with HA, authors have proposed some tentative conclusions regarding the potential outcomes due of chronic exposure to HA. Hemodialysis in patients at HA may present an increased risk for volume overload, which may predispose to pulmonary edema, since most of the usual compensatory mechanisms required to adjust to the hypoxic environment are impaired or even missing (Mairbaurl et al., 1989). In end-stage renal disease patients, chronic exposure to HA has been associated with low use of recombinant erythropoietin and higher hematocrit levels, given that HA residents produce more endogenous erythropoietin and respond more efficiently to erythropoietin compared with patients at sea level (Brookhart et al., 2008). Some mechanisms have been proposed to explain how chronic hypoxia contributes to the progression of CKD patients who are long-term HA residents, including decreased oxygen diffusion from peritubular capillaries to tubular and interstitial cells as a result of renal fibrosis; decreased peritubular capillary blood flow as a result of an imbalance of vasoactive substances; inappropriate energy use as a result of oxidative stress and decrease of oxygen delivery as a result of anemia (Nangaku, 2006).

In this study, only 16% of the subjects had a urine pattern that appeared normal analyzed by SDS-PAGE, all presented normoalbuminuria, and MMP-2 was absent; to our knowledge, there are no studies to compare these data.

We found an association between urinary MMP-2 ≥P₉₀ AU and microalbuminuria in univariate and multivariate analysis, and association between T2DM and microalbuminuria in univariate analysis, but not in multivariate analysis. It is known that MMP-2 plays an important role in the vasculature and mediates epithelial mesenchymal transition in the kidneys, leading to interstitial fibrosis, as well as being upregulated in glomerular inflammation (Sanders et al., 2004). MMP-2 activity correlates positively with tubulointerstitial damage and negatively with creatinine clearance (Sanders et al., 2007). In a mouse model for T1DM, urinary MMP-2 activity is increased compared with nondiabetic mice and urinary MMP-2 and NGAL/MMP-9 activities are detected at earlier stages of T1DM before albuminuria (McKittrick et al., 2011). Moreover, urine MMP-2 concentration correlates with renal hyperfiltration and the presence of microalbuminuria in T1DM (Thrailkill et al., 2007), but this correlation has not been examined in a HA rural population.

Apart from its role in renal fibrosis and epithelial mesenchymal transition, MMP-2 is also involved in renal repair, where an intricate balance exists between extracellular matrix synthesis and degradation; the major physiological regulator of extracellular matrix degradation in the glomerulus is MMP (Lenz et al., 2000). Matrix accumulation in diabetic nephropathy is the result of increased matrix synthesis coupled with poor degradation (Lenz et al., 2000). Interestingly, MMP downregulation has been associated with the progression of hypertensive nephrosclerosis (Singhal et al., 1996; Trachtman et al., 1996), chronic experimental unilateral ureteral obstruction (Engelmyer et al., 1995), and chronic cyclosporine nephropathy (Duymelinck et al., 1998). Moreover, diabetic nephropathy is accelerated in Mmp2 knockout mice (Takamiya et al., 2013).

One approach to the MMP is to consider them as compensatory factors before the onset of microalbuminuria and as accelerating factors associated with breakdown of the glomerular basement membrane, renal scarring and fibrosis during the progression of diseases. In this context, given the magnitude of detected risk factors associated with CDK, urinary MMP-2 activity could be a useful biomarker of incipient renal damage achieved through a noninvasive and cheap test, in our experience the cost of one well of zymogram gel is around $6.4 USD (excluding the basic laboratory equipment).

Conclusions

In the population studied, apart from hypoxia, important risk factors exist that affect renal function; urinary MMP-2 activity could constitute an early and economical biomarker of kidney damage.

Footnotes

Acknowledgments

The authors would like to acknowledge a postgraduate degree scholarship to M.E. Hernandez-Hernandez (National Council of Science and Technology, CONACyT 258304/319031), a grant for a laboratory assistant (CONACyT 86575 authorized to C.L. Sampieri) to D.J. Luna Lozano, and a grant from the Dirección General de Investigaciones for Bachelors study to M.J. Muñoz Contreras (authorized to C.L. Sampieri). The authors would like to thank V. Celis Arellano and B.L. López-Santos. The authors are thankful to the subjects who participated in this study and to the anonymous reviewers of this journal for their suggestions. Support: “Instrumentos y Equipos Falcón, S.A. de C.V” (authorized to A. Rodríguez) and provision of fully featured license of GelCompar II 6.6 Software from Applied Maths NV, Belgium.

Author Disclosure Statement

No competing financial interests exist.

References

Aceves-Martins

, Llaurado

, Tarro

, Sola

, and Giralt

. (2016). Obesity-promoting factors in Mexican children and adolescents: Challenges and opportunities. Glob Health Action, 9:29625.

Altemtam

, Nahas

, and Johnson

. (2012). Urinary matrix metalloproteinase activity in diabetic kidney disease: A potential marker of disease progression. Nephron Extra, 2:219–232.

Brookhart

, Schneeweiss

, Avorn

, Bradbury

, Rothman

, Fischer

, Mehta

, and Winkelmayer

. (2008). The effect of altitude on dosing and response to erythropoietin in ESRD. J Am Soc Nephrol, 19:1389–1395.

Centro Nacional de Programas Preventivos y Control de Enfermedades. Available at www.cenaprece.salud.gob.mx (last accessed February 12, 2017).

Chen

, Liu

, Wang

, Johnson

, Dong

, Li

, Ba

, Tan

, Luo

, Liu

, He

, and Yu

. (2011). Prevalence and risk factors of chronic kidney disease: A population study in the Tibetan population. Nephrol Dial Transplant, 26:1592–1599.

Cheng

, and Lovett

. (2003). Gelatinase A (MMP-2) is necessary and sufficient for renal tubular cell epithelial-mesenchymal transformation. Am J Pathol, 162:1937–1949.

Dimas

, Iliadis

, and Grekas

. (2013). Matrix metalloproteinases, atherosclerosis, proteinuria and kidney disease: Linkage-based approaches. Hippokratia, 17:292–297.

Duymelinck

, Deng

, Dauwe

, De Broe

, and Verpooten

. (1998). Inhibition of the matrix metalloproteinase system in a rat model of chronic cyclosporine nephropathy. Kidney Int, 54:804–818.

Engelmyer

, van Goor

, Edwards

, and Diamond

. (1995). Differential mRNA expression of renal cortical tissue inhibitor of metalloproteinase-1, −2, and −3 in experimental hydronephrosis. J Am Soc Nephrol, 5:1675–1683.

10.

Forbes

, Fukami

, and Cooper

. (2007). Diabetic nephropathy: Where hemodynamics meets metabolism. Exp Clin Endocrinol Diabetes, 115:69–84.

11.

Haditsch

, Roessler

, and Hinghofer-Szalkay

. (2007). Renal adrenomedullin and high altitude diuresis. Physiol Res, 56:779–787.

12.

Han

, Jee

, Han

, Kang

, Kim

, Han

, Kim

, and Cha

. (2006). An imbalance between matrix metalloproteinase-2 and tissue inhibitor of matrix metalloproteinase-2 contributes to the development of early diabetic nephropathy. Nephrol Dial Transplant, 21:2406–2416.

13.

Hansen

, Olsen

, Feldt-Rasmussen

, Kanstrup

, Dechaux

, Dubray

, and Richalet

. (1994). Albuminuria and overall capillary permeability of albumin in acute altitude hypoxia. J Appl Physiol (1985), 76:1922–1927.

14.

Hurtado

, Escudero

, Pando

, Sharma

, and Johnson

. (2012). Cardiovascular and renal effects of chronic exposure to high altitude. Nephrol Dial Transplant, 27 Suppl 4:iv11–iv16.

15.

Instituto Nacional de Salud Pública. (2012). Centro de Investigación en Evaluación de Encuestas. Encuesta Nacional de Salud y Nutrición. Available at http://ensanut.insp.mx (last accessed January 10, 2016).

16.

Jefferson

, Escudero

, Hurtado

, Kelly

, Swenson

, Wener

, Burnier

, Maillard

, Schreiner

, Schoene

, Hurtado

, and Johnson

. (2002). Hyperuricemia, hypertension, and proteinuria associated with high-altitude polycythemia. Am J Kidney Dis, 39:1135–1142.

17.

Kshirsagar

, and Wiggins

. (1986). A map of urine proteins based on one-dimensional SDS-polyacrylamide gel electrophoresis and Western blotting using one microliter of unconcentrated urine. Clin Chim Acta, 158:13–22.

18.

Lauhio

, Sorsa

, Srinivas

, Stenman

, Tervahartiala

, Stenman

, Gronhagen-Riska

, and Honkanen

. (2008). Urinary matrix metalloproteinase −8, −9, −14 and their regulators (TRY-1, TRY-2, TATI) in patients with diabetic nephropathy. Ann Med, 40:312–320.

19.

Lenz

, Elliot

, and Stetler-Stevenson

. (2000). Matrix metalloproteinases in renal development and disease. J Am Soc Nephrol, 11:574–581.

20.

Luks

, Johnson

, and Swenson

. (2008). Chronic kidney disease at high altitude. J Am Soc Nephrol, 19:2262–2271.

21.

Mairbaurl

, Schobersberger

, Hasibeder

, Knapp

, Hopferwieser

, Humpeler

, Loeffler

, Wetzels

, Wybitul

, Baumgartl

, et al. (1989). Exercise performance of hemodialysis patients during short-term and prolonged exposure to altitude. Clin Nephrol, 32:31–39.

22.

Marshall

, and Williams

. (1993). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of urine: Concentration of urinary proteins by precipitation with coomassie blue. Clin Chem, 39:2314–2318.

23.

Mason

, and Wahab

. (2003). Extracellular matrix metabolism in diabetic nephropathy. J Am Soc Nephrol, 14:1358–1373.

24.

McKittrick

, Bogaert

, Nadeau

, Snell-Bergeon

, Hull

, Jiang

, Wang

, Levi

, and Moulton

. (2011). Urinary matrix metalloproteinase activities: Biomarkers for plaque angiogenesis and nephropathy in diabetes. Am J Physiol Renal Physiol, 301:F1326–F1333.

25.

Moore

. (2001). Human genetic adaptation to high altitude. High Alt Med Biol, 2:257–279.

26.

Nagano

, Fukami

, Yamagishi

, Ueda

, Kaida

, Matsumoto

, Yoshimura

, Hazama

, Takamiya

, Kusumoto

, Gohara

, Tanaka

, Adachi

, and Okuda

. (2009). Circulating matrix metalloproteinase-2 is an independent correlate of proteinuria in patients with chronic kidney disease. Am J Nephrol, 29:109–115.

27.

Nangaku

. (2006). Chronic hypoxia and tubulointerstitial injury: A final common pathway to end-stage renal failure. J Am Soc Nephrol, 17:17–25.

28.

Navarro Diaz

. (2016). Consequences of morbid obesity on the kidney. Where are we going?. Clin Kidney J, 9:782–787.

29.

NFK. (2002). K/DOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Am J Kidney Dis, 39:S1–S266.

30.

Norman

, Clark

, and Garcia

. (2000). Hypoxia promotes fibrogenesis in human renal fibroblasts. Kidney Int, 58:2351–2366.

31.

Rivera

, Barquera

, Campirano

, Campos

, Safdie

, and Tovar

. (2002). Epidemiological and nutritional transition in México: Rapid increase of non-communicable chronic diseases and obesity. Public Health Nutr, 5:113–122.

32.

Ronco

, and Chatziantoniou

. (2008). Matrix metalloproteinases and matrix receptors in progression and reversal of kidney disease: Therapeutic perspectives. Kidney Int, 74:873–878.

33.

Sampieri

, de la Pena

, Ochoa-Lara

, Zenteno-Cuevas

, and Leon-Cordoba

. (2010). Expression of matrix metalloproteinases 2 and 9 in human gastric cancer and superficial gastritis. World J Gastroenterol, 16:1500–1505.

34.

Sanders

, Huitema

, Hanemaaijer

, van Goor

, Kallenberg

, and Stegeman

. (2007). Urinary matrix metalloproteinases reflect renal damage in anti-neutrophil cytoplasm autoantibody-associated vasculitis. Am J Physiol Renal Physiol, 293:F927–F1934.

35.

Sanders

, van Goor

, Hanemaaijer

, Kallenberg

, and Stegeman

. (2004). Renal expression of matrix metalloproteinases in human ANCA-associated glomerulonephritis. Nephrol Dial Transplant, 19:1412–1419.

36.

Singhal

, Sagar

, and Garg

. (1996). Simulated glomerular pressure modulates mesangial cell 72 kDa metalloproteinase activity. Connect Tissue Res, 33:257–263.

37.

Takamiya

, Fukami

, Yamagishi

, Kaida

, Nakayama

, Obara

, Iwatani

, Ando

, Koike

, Matsui

, Nishino

, Ueda

, Cooper

, and Okuda

. (2013). Experimental diabetic nephropathy is accelerated in matrix metalloproteinase-2 knockout mice. Nephrol Dial Transplant, 28:55–62.

38.

Tashiro

, Koyanagi

, Ohara

, Ito

, Saitoh

, Horikoshi

, and Tomino

. (2004). Levels of urinary matrix metalloproteinase-9 (MMP-9) and renal injuries in patients with type 2 diabetic nephropathy. J Clin Lab Anal, 18:206–210.

39.

Thrailkill

, Bunn

, Moreau

, Cockrell

, Simpson

, Coleman

, Frindik

, Kemp

, and Fowlkes

. (2007). Matrix metalloproteinase-2 dysregulation in type 1 diabetes. Diabetes Care, 30:2321–2326.

40.

Thrailkill

, Clay Bunn

, and Fowlkes

. (2009). Matrix metalloproteinases: Their potential role in the pathogenesis of diabetic nephropathy. Endocrine, 35:1–10.

41.

Trachtman

, Futterweit

, Garg

, Reddy

, and Singhal

. (1996). Nitric oxide stimulates the activity of a 72-kDa neutral matrix metalloproteinase in cultured rat mesangial cells. Biochem Biophys Res Commun, 218:704–708.

42.

Winterborn

, Bradwell

, Chesner

, and Jones

. (1987). The origin of proteinuria at high altitude. Postgrad Med J, 63:179–181.

43.

Zheng

, Lv

, Ni

, Li

, Ma

, and Liu

. (2011). Urinary podocyte-associated mRNA profile in various stages of diabetic nephropathy. PLoS One, 6:e20431.