MMP-2, MMP-3, TIMP-1, TIMP-2, and TIMP-3 Protein Levels in Human Aqueous Humor: Relationship With Axial Length

Abstract

Purpose.: We measured the aqueous humor levels of matrix metalloproteinases (MMP) MMP-1, MMP-2, MMP-3, and tissue inhibitors of matrix metalloproteinases (TIMP) TIMP-1, TIMP-2, and TIMP-3 in patients with myopia or cataract, and investigated the relationship between their levels and axial length (AL).

Methods.: We measured MMP/TIMPs levels with the Luminex xMAP Technology by using commercially available Milliplex xMAP Kits. A total of 65 aqueous humor samples was collected from patients with myopia or cataract during cataract or clear lens extraction surgery. According to the AL, the samples were divided into three groups: group A, AL ≤ 24 mm; group B, AL 24 to 26 mm; and group C, AL ≥ 26 mm.

Results.: Levels of MMP-2, MMP-3, TIMP-1, TIMP-2, and TIMP-3 could be detected in the aqueous humor. The levels of MMP-2, TIMP-1, TIMP-2, and TIMP-3 were positively correlated with AL. The differences of the levels of these MMPs/TIMPs among the three groups were statistically significant. The MMP-3 levels were not correlated with AL and there was no significant difference in MMP-3 levels among these three groups. Levels of MMP-1 could not be detected in the aqueous humor samples.

Conclusions.: Elevated levels of aqueous MMP-2, TIMP-1, TIMP-2, and TIMP-3 were found in the eyes with elongated axis.

Introduction

Myopia is one of the major causes of visual impairment worldwide. The most important contribution of ocular parameters to myopia is the excessive elongation of the axial length (AL). ^1,2 The pathologic changes, such as glaucoma, ³ posterior scleral staphyloma, ⁴ retinal detachment, ⁵ and macular hole, ⁶ may be caused by elongation of the AL. 

The axial elongation rate of the eye is determined primarily by the rate of expansion of the scleral shell, which is a dense extracellular matrix (ECM) connective tissue produced by scleral fibroblast cells. The ECM in the sclera is comprised largely of collagen fibrils (predominantly fibrillar type I collagen), elastin, proteoglycans, glycoproteins, hyaluronan, and other proteins. During the development stage of myopia, there is a loss of ECM proteins, hyaluronan, and other glycosaminoglycans (GAGs). This is called scleral remolding process, which can cause a weakness of the sclera, and results in elongation of the AL and the development of myopia. ⁷  

Matrix metalloproteinases (MMPs) are a family of enzymes that include at least 28 members, most of which degrade ECM. ^8–13 The role of MMPs in the pathogenesis of myopia has been studied in experimental myopia models and in vitro. ^14–27 The expression and activities of MMP-2 are increased during the development of myopia and decreased during the recovery stage in various experiment myopia models. ^14–20 Study in cultured human scleral fibroblasts suggested that mechanical strain stimulated the activation of MMP-2 by scleral fibroblasts. ²⁷ All of these findings could be related to the degradation of ECM in the sclera, which can result in scleral remolding and development of myopia. 

The tissue inhibitors of matrix metalloproteinases (TIMPs) provide a negative control of MMPs activity. ^8–11,13 Experimental animal studies showed various results in the changes of TIMPs levels during the development and recovery of myopia. ^{14,18,20,25,26} Some studies demonstrated a decrease in TIMP expression (mRNA or protein) during myopia development and an increase during recovery while others did not detect any changes. ^14,18,20,25  

Measurement of levels of MMPs and TIMPs in the aqueous humor have been reported in patients with various ocular diseases, including diabetic retinopathy, various types of glaucoma and uveitis, and compared to those in the cataract patients (as the control). ^28–32 To our best knowledge, the levels of various MMPs and TIMPs in eyes with different AL have not been reported. Therefore, we studied the aqueous levels of MMPs/TIMPs in the cataract and high myopia patients with different AL, and investigated the relationship between MMPs/TIMPs levels and axial elongation. 

Methods

Patients and Samples

This study was an observational study in 65 patients, with high myopia or cataract. Exclusion criteria included prior intraocular surgery and patients with hypermature cataract. All patients were relatively healthy with no systemic diseases, such as kidney diseases, hematologic diseases, immune diseases, or history of drug use. Aqueous samples were collected from these patients during cataract or clear lens extraction surgery. This study was performed in accordance with the Declaration of Helsinki. All patients signed an informed consent and the Institutional Review Board at the Shanghai Ninth People's Hospital Affiliated Shanghai Jiaotong University School of Medicine approved the study. 

The AL was measured with a Zeiss IOL-Master laser interferometer (Optical Biometry, IOL Master; Carl Zeiss Meditec AG, Jena, Germany). According to the AL, the 65 eyes were divided into three groups: group A (AL ≤ 24 mm), group B (AL between 24 and 26 mm), and group C (AL ≥ 26 mm). 

The aqueous humor (0.1–0.2 mL) was collected at the beginning of operation. After a clear corneal paracentesis, a 30-gauge needle on a tuberculin microsyringe was used to aspirate aqueous humor from the central anterior chamber. The aqueous humor samples were stored immediately under −80°C until analysis. 

Measurement of Levels of MMPs and TIMPs

All samples were assayed for total protein levels of MMP-1, MMP-2, MMP-3, TIMP-1, TIMP-2, and TIMP-3 with the Luminex system (Luminex xMap Technology; Bio-Rad, Hercules, CA, USA) by using commercially available Milliplex xMAP Kits (Millipore Corporation, Billerica, MA, USA). ^33,34 This technology uses multiplexed microsphere-based immunoassays that apply flow cytometric resolution to measure spectrally distinct microspheres coupled with capture molecules and reporter fluorochromes bound to detection antibodies. 

The assays were performed following to the manufacturer's guidelines and measured on a Bio-Plex 200 system (Bio-Rad). Each sample was measured in duplicate. The amount of MMPs/TIMPs (pg/mL) was calculated from the standard curves for each MMP/TIMP according to the manufacturer's instructions. The same aqueous humor sample was measured in each plate for an interassay control. ^35,36 Interassay coefficients of variations (CVs) were 8.4% for MMP-1, 18.0% for MMP-2, 10% for MMP-3, ≤10% for TIMP-1, ≤10% for TIMP-2, and ≤10% for TIMP-3. The intraassay CVs were 2.6% for MMP-1, 5.4 % for MMP-2, 5% for MMP-3, ≤10% for TIMP-1, ≤10% for TIMP-2, and ≤10% for TIMP-3. The range of detection was 8.2 to 20,000 pg/mL for MMP-1, 260.0 to 50,000 pg/mL for MMP-2, 32.0 to 50,000 pg/mL for MMP-3, 13.6 to 20,000 pg/mL for TIMP-1, 37.4 to 50,000 pg/mL for TIMP-2, and 92.3 to 100,000 pg/mL for TIMP-3. ^35,36  

Statistics

The original data were not normally distributed by the Kolmogorov-Smirnov test. Therefore, the results were expressed as median and range (25th and 75th percentiles) in continuous variables that are not normally distributed, such as the levels of MMPs/TIMPs, or as mean (SD) for normal distributed continuous variables, such as age. Nonparametric test (Kruskal-Wallis test) was used to compare the levels of aqueous humor MMPs/TIMPs among different groups. The relationship between AL and MMPs/TIMPs or between the age and MMP/TIMPs was analyzed with Spearman's correlation test. The Mann-Whitney U test was used to evaluate the difference of MMP/TIMPs levels between the males and females. Comparison of sex distribution in three different groups was analyzed by the χ² test. A 1-way ANOVA was used for comparison among age distributions in three different groups. The SPSS 22.0 software for windows (SPSS, Inc., Chicago, IL, USA) was used to do these analyses. A 2-tailed P < 0.05 was considered to be statistically significant for the analysis. 

Results

Aqueous humor samples were obtained from 65 eyes from 65 subjects, including 27 eyes with AL ≤ 24 mm (group A), 8 eyes with AL at 24 to 26 mm (group B), and 30 eyes with AL ≥ 26 mm (group C). The mean AL was 26.7 mm in the whole groups (Table 1). Mean (± SD) diopters (D) were −0.8 ± 2.1, −9.3 ± 5.3, and −17.6 ± 6.8 D in groups A to C, respectively. Their age averaged 67.0 ± 11.7 years, and there were 29 men and 36 women (Table 1). Among the 65 subjects, 30 eyes had cataracts and underwent cataract extractions; 35 eyes were myopic (all in the stationary stage). In the myopia group, 14 eyes underwent clear lens extractions and 21 eyes underwent cataract extractions because of the presence of myopia and cataract simultaneously (Table 1). There were 11, 5, and 13 males, and 16, 3, and 17 females in groups A to C, respectively. There was no statistically significant difference in sex distribution (χ² = 1.220, P = 0.543) or ages (P > 0.05) among these three groups. 

Table 1

 

Demographic Characteristics of 65 Patients With Cataract or Myopia

Variant	Total	Group A	Group B	Group C
N	65	27	8	30
Age, y, mean ± SD	67.0 ± 11.7	74.3 ± 7.3	68.1 ± 7.4	60.0 ± 11.8
Mean D ± SD	−10.3 ± 9.4	−0.8 ± 2.1	−9.3 ± 5.3	−17.6 ± 6.8
Sex, M/F	29/36	11/16	5/3	13/17
AL, mm	26.7 ± 3.8	23.1 ± 0.5	25.0 ± 0.4	30.4 ± 2.3
High myopia, underwent clear lens extraction	14	0	2	12
High myopia with cataract, underwent cataract extraction	21	0	3	18
Cataract, underwent cataract extraction	30	27	3	0

Levels of MMP-1, -2, and -3

The MMP-1 could not be detected in the aqueous specimens. Levels of MMP-2 could be detected in the aqueous samples. The aqueous humor MMP-2 levels in group C were the highest and that in group A were the lowest among the three groups. The difference of MMP-2 levels in the three groups was statistically significant (H = 7.891, P = 0.019, Table 2, Fig. 1). In addition, there was a significantly positive correlation between MMP-2 level and AL (r ² = 0.338, P = 0.006, Fig. 2A). 

Figure 1

 

Aqueous MMP-2/TIMP-2 levels in cataract and myopic eyes with different ALs. The MMP-2/TIMP-2 levels in groups A to C are shown. The Kruskal-Wallis test demonstrated that the differences of MMP-2 and TIMP-2 levels in the three groups were statistically significant (MMP-2, P = 0.019; TIMP-2, P = 0.012).

Figure 2

 

Correlation between AL and aqueous MMP/TIMP levels. Light colored plots represent nonmyopic eyes, the dark circle represents myopic eyes, and eyes with cataract and myopia at the same time. (A) Correlation between AL and MMP-2. (B) Correlation between AL and MMP-3. (C) Correlation between AL and TIMP-1. (D) Correlation between AL and TIMP-2. (E) Correlation between AL and TIMP-3. Spearman correlation test showed that MMP-2, TIMP-1, TIMP-2, and TIMP-3 were correlated positively with AL (P < 0.05).

Table 2

 

Aqueous MMPs/TIMPs Levels in Cataract and Myopic Eyes With Different ALs

MMPs/TIMPs	Group A, n = 27	Group B, n = 8	Group C, n = 30
MMP-2	4,003 (3,150, 5,679)	4,303 (3,105, 5,791)	5,179 (4,116, 7,082)
MMP-3	341 (182, 723)	178 (80, 496)	213 (82, 596)
TIMP-1	5,694 (3,791, 9,805)	6,540 (4,143, 7,903)	10,907 (6,708, 16,345)
TIMP-2	7,447 (5,630, 11,746)	9,972 (6,168, 10,882)	11,241 (9,383, 15,400)
TIMP-3	1,732 (927, 2,818)	1,831 (1,431, 2,847)	3,028 (1,548, 4,513)

 Data of MMPs/TIMPs were expressed as pg/mL: P₅₀ (P₂₅, P₇₅).

Levels of MMP-3 also could be detected in aqueous humor samples. The difference in the levels of MMP-3 in all groups was statistically nonsignificant (H = 4.534, P = 0.104, Table 2, Fig. 3). Besides, the levels of MMP-3 in the aqueous were not correlated with AL (P = 0.176). 

Levels of TIMP-1, -2, and -3

According to the AL, the aqueous humor levels of TIMP-1 were divided into groups A to C, each was higher than its predecessor (Table 2). The difference of levels of TIMP-1 in the three groups was statistically significant (H = 12.087, P = 0.002, Fig. 4). There was a significantly positive correlation between TIMP-1 level and AL (r ² = 0.427, P = 0.000, Fig. 2C). 

The TIMP-2 levels in group C were the highest and those of the group A were the lowest in the three groups as shown in Figure 1. Moreover, there was significant difference of TIMP-2 levels among the three groups (H = 8.865, P = 0.012). There was a significantly positive correlation between TIMP-2 level and AL (r ² = 0.359, P = 0.003, Fig. 2D). 

The TIMP-3 levels in group C were the highest and those of the group A were the lowest in the three groups. The difference among these three groups was statistically significant (H = 7.579, P = 0.023, Table 2, Fig. 3). There was a significant positive correlation between TIMP-3 level and AL (r ² = 0.266, P = 0.032, Fig. 2E). 

Figure 3

 

Aqueous MMP-3/TIMP-3 levels in cataract and myopic eyes with different ALs. The Kruskal-Wallis test demonstrated that TIMP-3 levels in the in the three study groups were significantly different (TIMP-3, P = 0.023). The differences in the MMP-3 among all groups were not statistically significant (P = 0.104).

Figure 4

 

Aqueous TIMP-1 levels in cataract and myopic eyes with different AL. The Kruskal-Wallis test demonstrated that TIMP-1 levels in the in the three study groups were significantly difference (TIMP-1, P = 0.002).

Relationship Between MMPs/TIMPs and Age

No significant correlation was found between the age and MMPs/TIMPs levels (MMP-2, P = 0.066; MMP-3, P = 0.116; TIMP-1, P = 0.069; TIMP-2, P = 0.186; TIMP-3, P = 0.646). 

Males Versus Females

There was no significant difference of aqueous MMPs/TIMPs levels between the males and females with an exception in MMP2, which had a very little, but significant difference (MMP-2, P = 0.046; MMP-3, P = 0.070; TIMP-1, P = 0.303; TIMP-2, P = 0.501; TIMP-3, P = 0.687). 

Discussion

In this study, we have found that, apart from MMP-1, levels of MMP-2, MMP-3, TIMP-1, TIMP-2, and TIMP-3 could be detected and measured in the aqueous specimens. Schloetzer-Schrehardt et al. ²⁷ measured the levels of various MMPs/TIMPs in the aqueous from cataract patients and compared them to levels in glaucoma patients. In the cataract groups, MMP-1 was not detected in both studies. However, MMP-2, MMP-3, TIMP-1, and TIMP-2 could be detected in both studies; and the TIMP-2 levels were the highest in all tested MMP/TIMPs and MMP-3 levels were the lowest in both studies. ³³ This finding suggests that the results of the present study are consistent with previous reports. 

Except MMP-3, the differences of the MMPs/TIMPs level in the three groups with various ALs were statistically significant. In addition, aqueous levels of MMP-2, TIMP-1, TIMP-2, and TIMP-3 were correlated significantly with AL. 

Gellatinase A (MMP-2) has an important role in the final degradation of fibrillar collagens after they have first been cleaved by collagenases. In addition, MMP-2 also can degrade various collagens (type I, IV, V, VII, X), proteoglycans, fibronectin, laminin, and elastin. ^9,13,37 The increased MMP-2 activity degrades various components of the ECM in the sclera, which can lead to the weakness of the sclera, and results in the elongation of AL and the development of myopia. ^14,17,18,21  

In the present study, the aqueous MMP-2 protein levels in group C (with marked elongation of AL) were significantly greater than in group A (with AL less than 24 mm). There was a significant positive correlation between the levels of MMP-2 in the aqueous with the AL. All of these findings were consistent with the changes of MMP-2 levels in experimental animals and in vitro studies. ^{14,15,17–21,23}  

Little is known on the relationship between MMP-1 and −3 and the development of myopia. Gene polymorphism studies on the association of these genes in myopic patients showed conflicting and mainly negative results. ^38–40 Recently, Schache et al. ⁴¹ studied the association of various MMPs gene polymorphisms (MMP-1, -2, -3, -8, -9, -10, -11, and -13) with myopia and found that polymorphisms in MMPs genes do not have a major role in myopia and various refractive indexes. Although there is strong evidence that these genes are involved in the sclera remodeling process that accompanies myopia, the investigators proposed that the role of MMP is not driven by single nucleotide polymorphisms, but, instead, is influenced by other genetic changes that might include copy number changes, epigenetic changes, or alterations in the regulatory elements of the genes, all of which may results in changes in expression. ⁴¹ The changes of MMP-1 and MMP-3 in experimental myopia remain poorly understand. Because MMP-3 is known to degrade proteoglycan core proteins, it has been expected that MMP-3 might increase during the development of myopia and decrease during recovery. ²¹ However, the expression of MMP-3 mRNA levels during the development of myopia or recovery does not show significant changes as expected. ^18,21 In the present study, there was no significant difference in aqueous MMP-3 protein levels among the three groups with different AL and the level of MMP-3 did not correlate with the AL. These findings are consistent with the previous results obtained from experimental animal studies. ^18,21  

The TIMPs are the major endogenous regulators of MMP activities in the tissue, and four homologous TIMPs (TIMP-1, -2, -3, and -4) have been identified to date. ^8–11,13 The TIMP-1 is capable of inhibiting the activities of all known MMPs except membrane type-matrix metalloproteinases (MT-MMPs). The TIMP-2 and -3 inhibit the activity of MT-MMPs, in addition to the inhibition of other MMPs. ^8–11,13  

The relationship between TIMPs levels in the sclera and the development of myopia is not as consistent as in the changes of MMP-2. The TIMP-1 mRNA levels in the sclera of form-deprivation myopia (FDM) in guinea pigs increased in the recovery of myopia, but did not show significant changes in the development of myopia. ¹⁸ Expression of TIMP-2 mRNA in chicken FDM was decreased in the development of myopia, but not increased in the recovery. ¹⁹ In the lens-induced myopia in guinea pigs, TIMP-2 mRNA was increased during the recovery, but did not change during the development of myopia. ¹⁷  

In the present study, the aqueous TIMP-1, -2, and -3 protein levels in group C (with marked elongation of AL) were greater than in group A (with AL less than 24 mm) and group B (with AL between 24 and 26 mm). Besides, there was a significant positive correlation between the levels of these three TIMPs and AL. These unexpected results may be due to several different reasons. 

First, although the main effect of TIMPs is the inhibition of the activities of various MMPs, the relationship between TIMP and MMP is not simple and straightforward. The TIMPs are multifunctional proteins; they have various biological effects that extend beyond their role as inhibitors of MMP activity. ^8–11,13 The TIMPs can affect cell morphology, cell growth, apoptosis, and the production of growth factors. All of these functions may influence the development and recovery of myopia not related to the changes of MMP activities. ^8–11,13  

In experimental myopia models, myopia is caused mainly by environmental factors (form-deprivation or minus lens compensation). However, in human myopia, it could be caused by environmental and genetic factors, and some cases might even be caused mainly by genetic factors (gene mutation), especially in high myopia with marked elongation of AL, such as in group C in the present study. ^42,43 Therefore, it is possible that the changes of TIMPs in human myopia might be different from those in the experimental myopia models. Furthermore, in experimental animal study, tested subjects are young animals at the developmental or early recovery stage. The subjects in the present study were old persons and all myopes were at the stationary stage. This, again, could be the reason for the different results in TIMPs levels between experimental animal studies and the present study. This difference also may influence the validity of the link between experimental animal studies and the present human study. 

Third, in the experimental animal study, the mRNA levels of MMP/TIMPs were measured in the sclera, which represents the secretory abilities of these factors by scleral fibroblasts. ^14–27 However, the proteins of MMP/TIMPs in the aqueous humor were not only originated from the scleral fibroblast. It has been reported that uveal melanocytes, ⁴⁴ ciliary muscle cells, ^45–47 and ciliary nonpigment epithelial cells, ⁴⁸ trabecular cells, ^49,50 and retinal pigment epithelial cells ^51–53 are able to produce and secrete the MMP/TIMPs directly to or in exchange with the aqueous humor. The MMP/TIMP levels in the aqueous humor may be the response by various cells to the changes in the sclera as a hemostatic mechanism. Therefore, the TIMPs levels in the aqueous may or may not be in the same direction that occurred in the sclera. 

In conclusion, this study suggested that the protein levels of several key MMPs and TIMPs were increased in the eyes with a longer axis. However, the studied individuals were old persons and all myopes were at the stationary stage. Therefore, caution should be maintained in the application of these results to the role of MMPs/TIMPs in the development of myopia. 

Acknowledgments

Supported by the National Nature Science Foundation of China, Foundation Number 81371050, and the Shanghai Jiaotong University School of Medicine Innovation Fund (YZ1022). The authors alone are responsible for the content and writing of the paper. 

Disclosure: Y. Jia, None; D.-N. Hu, None; D. Zhu, None; L. Zhang, None; P. Gu, None; X. Fan, None; J. Zhou, None