Changes in Extracellular Matrix Gene and Protein Expressions in Human Trabecular Meshwork Cells in Response to Mechanical Fluid Flow Stimulation

Purpose To investigate the changes in extracellular matrix (ECM) gene and protein expressions in human trabecular meshwork (HTM) cells in response to mechanical fluid flow stimulation. Methods HTM cells were cultured on a glass plate and exposed to shear stress (0, 0.2, and 1.0 dyne/cm2) for 12 hours. Changes in gene expressions were evaluated using microarray analysis. The representative genes related to ECM metabolism underwent real-time reverse-transcriptase polymerase chain reaction. Fibronectin (FN) and collagen (COL) IV levels in the supernatant were evaluated using immunoassays. Rho-associated coiled-coil-containing protein kinase (ROCK) activity also was investigated. Results After stimulation, transforming growth factor β2 mRNA levels were significantly (p < 0.01) lower than that of the static control (0 dyne/cm2 for 12 hours). Matrix metalloproteinase 2 mRNA levels were significantly (p < 0.05) higher than the static control. COL type 1 alpha 2 mRNA, COL type 4 alpha 2 mRNA, and COL type 6 alpha 1 mRNA levels were significantly (p < 0.05, < 0.01, and < 0.05, respectively) higher than the static control. The mean ± standard deviation FN levels (ng/mL) in the supernatant after stimulation (0, 0.2, 1.0 dyne/cm2) were 193.7 ± 7.6, 51.5 ± 21.8, and 34.9 ± 23.6, respectively (p < 0.01). The FN and COL IV levels and ROCK activity were significantly (p < 0.01 and < 0.05, respectively) lower than the static control. Conclusions Changes in gene and protein expressions related to ECM metabolism occurred in HTM cells after stimulation. Specifically, the suppression of FN and COL IV production might explain the importance of mechanical fluid flow stimulation on maintenance of the normal aqueous humor outflow.


27
HTM cells were cultured on a glass plate and exposed to shear stress (0, 0.2, and 1.0 28 dyne/cm 2 ) for 12 hours. Changes in gene expressions were evaluated using microarray 29 analysis. The representative genes related to ECM metabolism underwent real-time 30 reverse-transcriptase polymerase chain reaction. Fibronectin (FN) and collagen (COL) IV 31 levels in the supernatant were evaluated using immunoassays. Rho-associated coiled-coil-32 containing protein kinase (ROCK) activity also was investigated.

34
After stimulation, transforming growth factor 2 mRNA levels were significantly (p < 35 0.01) lower than that of the static control (0 dyne/cm 2 for 12 hours). Matrix 36 metalloproteinase 2 mRNA levels were significantly (p < 0.05) higher than the static 37 control. COL type 1 alpha 2 mRNA, COL type 4 alpha 2 mRNA, and COL type 6 alpha 1 38 mRNA levels were significantly (p < 0.05, < 0.01, and < 0.05, respectively) higher than Introduction 58 Intraocular pressure (IOP) is maintained through a proper function of the aqueous 59 humor, which is produced by the ciliary body [1]. About 70% to 95% of the aqueous 60 humor drains through the conventional outflow pathway [2].Therefore, normal function of 61 this outflow component is important to the IOP homeostasis and prevention of glaucoma 62 [3]. Increased aqueous outflow resistance in this component is the main cause of glaucoma 63 accompanied by elevated IOP [4,5]. 64 The trabecular meshwork (TM), juxtacanalicular meshwork, and Schlemm's canal, the 65 collector channels, and the episcleral veins comprise conventional outflow pathway. 66 Among those, extracellular matrix (ECM) in TM tissue, which are composed of collagen 67 (COL) or fibronectin (FN) [6], is critical for the homeostatic maintenance of the normal 68 outflow resistance [7]. Of note, recent studies revealed that ECM turnover is regulated by 69 mechanical stress, at least in part [8][9][10]   Controls, Tokyo, Japan). In the current study, the HTM cells were exposed to the

138
The HTM cells were collected with a scraper, and the total RNA was extracted using a  Table 1. For all amplifications, the cycling conditions were as 144 follows: an initial denaturation period for 5 minutes at 95°C, followed by 45 cycles of 10 145 seconds at 95°C, 30 seconds at 60°C, and 1 second at 72°C. The quantification of each 146 gene expression signal was normalized with respect to the signal for the glucose-6-9 147 phosphate dehydrogenase (G6PDH) gene. The relative fold changes in the expression of 148 each gene were determined using the 2 -Ct method.  and examined using LAS-3000 Imager (Fujifilm, Tokyo, Japan).

179
The cell lysates were collected from the HTM cells with or without exposure to shear 180 stress for 12 hours. The total protein concentration was measured using a NanoDrop of the static control to those exposed to shear stress (0.2 dyne/cm 2 ) for 12 hours. The gene categories, names, and fold changes are shown in Table 2. Those genes listed then were 196 analyzed by real-time RT-PCR to ascertain the results obtained by microarray analysis. As 197 a result, transforming growth factor (TGF)-2 mRNA levels were significantly lower than 198 that of the static control (0.2 dyne/cm 2 , 0.65-fold vs. static control, p < 0.01) (Fig. 1). The control, p<0.05) levels were significantly higher than the static control. Although the FN1 207 mRNA levels were higher than the static control (1.0 dyne/cm 2 , 1.87-fold vs. static 208 control), the differences were not significant (p = 0.085) (Fig. 3).

215
2 mRNA levels in the HTM cells exposed to shear stress (0.2 dyne/cm 2 ) for 12 hours are 216 significantly lower than the static control. The data are expressed as the means ± standard 217 deviation (n = 5). *p < 0.01. The mean (± standard deviation) FN levels (ng/mL) in the medium exposed to shear stress 235 (dyne/cm 2 ) of 0, 0.2, and 1.0 for 1 hour were 31.6 ± 19.6, 30.0 ± 18.7, and 29.5 ± 13.6, 236 respectively. The values for 12 hours were 193.7 ± 7.6, 51.5 ± 21.8, and 34.9 ± 23.6, 237 respectively. After 12 hours, the FN levels were significantly (p < 0.01) lower at 1.0 238 dyne/cm 2 than the static control (Fig. 4). Representative images of each band obtained by Quantitative assessment of the intensity of each band determined by densitometry. The FN   252 and COL IV levels are significantly lower compared to the static control after stimulation.

ROCK activity in HTM cells after exposure to shear stress 257
To examine the ROCK activity, we carried out the ELISA technique. Our experiment 258 showed that the ROCK activity was significantly lower than the static control after 259 exposure to shear stress (0.2 dyne/cm 2 , 0.78-fold vs. static control, p < 0.05) (Fig. 6). the HTM cells, a previous report also investigated the responses of HTM cells exposed to 289 shear stress. Ashpole et al. [24] reported that when shear stress (10 dyne/cm 2 ) was applied 290 to human SC endothelial (SCE) cells, they responded similarly to vascular endothelial 291 cells, i.e., shear stress also triggered nitric oxide (NO) production on human SCE cells.

292
Those authors concluded that increased shear stress to SCE cells during SC collapse in the 293 presence of elevated IOP may partly mediate IOP homeostasis by NO production.

294
Although, in that study, they also investigated the responses of HTM cells exposed to a 295 shear stress (10 dyne/cm 2 ), they only focused on NO production. In the current study, we 296 applied shear stress to HTM cells, but the strength was very weak (0.2 or 1.0 dyne/cm 2 ).
297 Therefore, the current study differed from their research in that we focused on the gene or 298 protein changes related to ECM metabolism in response to very weak shear stress.

299
In our microarray analysis, we focused on the gene changes related to the TGF-β 300 superfamily and ECM components and remodeling. decreases in IOP when the drug is used over the long term.

323
Although, in the current study, the HTM cells were stimulated by steady fluid flow 324 using our shear stress experimental system, it is presumed that the fluid flow in the TM is 325 probably not a laminar or steady flow in normal eyes. A previous report that simulated the 326 wall shear stress in each component of the conventional outflow pathway reported that the 327 shear stress on the inner wall of the SC was about 0.01 dyne/cm 2 [32]. Therefore, in the 328 JCT, which is the principal tissue of the ECM metabolism in the TM, the shear stress 329 might be equal to or less than that in the inner wall of the SC (i.e., much weaker 330 stimulation than the current experiment). Further, it also has been reported that the TM 331 tissue pulsates in conjunction with the heart rate [33,34]. Taken together, considering this 332 histologic feature of the meshwork, the aqueous outflow in the TM may be turbulent.

333
Further investigation is needed to determine if the results of the current shear stress 334 experiment also occur in vivo.

335
In conclusion, gene and protein changes related to ECM metabolism were observed