Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Aug 1.
Published in final edited form as: J Vasc Surg. 2008 May 23;48(2):447–456. doi: 10.1016/j.jvs.2008.03.004

Prolonged Increases in Vein Wall Tension Increase Matrix Metalloproteinases and Decrease Constriction in Rat Vena Cava. Potential Implications in Varicose Veins

Joseph D Raffetto 3, Xiaoying Qiao 1, Vera V Koledova 1, Raouf A Khalil 1,2
PMCID: PMC2575039  NIHMSID: NIHMS42197  PMID: 18502086

Abstract

Background

Increased venous hydrostatic pressure plays a role in the pathogenesis of varicose veins. Increased expression of matrix metalloproteinases (MMPs) has been identified in varicose veins. Also, we have shown that MMP-2 inhibits venous contraction. However, the relation between venous pressure, MMP expression and venous dysfunction is unclear. The purpose of this study was to test the hypothesis that prolonged increases in vein wall tension cause overexpression of MMPs and decreased contractility, which in turn promote venous dilation.

Methods

Circular segments of inferior vena cava (ICV) were isolated from male Sprague-Dawley rats, and suspended between two wires in Krebs solution. Preliminary vein wall tension-contraction relation showed maximal KCl (96 mmol/L) contraction at 0.5g basal tension, which remained steady with increases in tension up to 2g. Vein segments were subjected to either control (0.5g) or high (2g) basal tension for short (1 hr) or long duration (24 hr). Isometric contraction in response to phenylephrine (Phe, 10−5 mol/L), angiotensin II (AngII, 10−6 mol/L), and KCl was measured. The veins were frozen to determine the expression and localization of MMPs using immunoblots and immunohistochemistry.

Results

In IVC segments subjected to 0.5g tension for 1 hr Phe and AngII produced significant contraction. At higher 2g basal tension for 24 hr, both Phe and AngII contractions were significantly reduced. Reduction in KCl contraction was also observed at high 2g basal tension for 24 hr, suggesting that the reduction in vein contraction is not specific to a particular receptor, and likely involves inhibition of a post-receptor contraction mechanism. In vein segments under 2g tension for 24 hr and treated with TIMP-1, Phe, AngII, and KCl contractions were partially restored, suggesting the involvement of MMPs. IVC immunoblot analysis demonstrated prominent bands corresponding to MMP-2 and MMP-9 protein. High 2g wall tension for 24 hr was associated with marked increase in the amount of MMP-2 and -9 relative to the housekeeping protein actin. There was a correlation between MMP expression and decreased vein contraction. Also, significant increases in MMP-2 and -9 immunostaining were observed in IVC segments subjected to high 2g tension for 24 hr. Both MMP-2 and MMP-9 caused significant inhibition of Phe contraction in IVC segments.

Conclusions

In rat IVC, increases in magnitude and duration of wall tension is associated with reduced contraction and overexpression of MMP-2 and -9. In light of our findings that MMP-2 and -9 promote IVC relaxation, the data suggest that protracted increases in venous pressure and wall tension increase MMPs expression, which in turn reduce venous contraction and lead to progressive venous dilation.

CLINICAL RELEVANCE

Varicose veins is a major vascular disease that affect more than 25 million adults in the United States. The fundamental changes in varicose veins involve dilation, tortuousity and structural changes in the vein wall, including elastin fragmentation and disorganized smooth muscle. Venous hypertension is often associated with varicose veins and advanced venous disease. Also, MMPs are overexpressed in all layers of the varicose vein wall. The present study examined possible relation between venous pressure, MMP expression and venous tissue dysfunction in rat IVC. Although differences in the physiologic behavior and structure of rat IVC and human veins make it difficult to extrapolate the response of rat IVC to human veins, the results indicate that the magnitude and duration of vein wall tension have significant impact on venous function and MMP expression. The data make it tempting to speculate that protracted increases in venous pressure and wall stretch increase MMPs expression, which in turn produces venous dilation and further increases in venous pressure, leading to a recalcitrant cycle and progressive venous dilation.

INTRODUCTION

Varicose veins is a common vascular disease of the lower extremity, and is characterized by excessive dilation and tortuousity of the venous system.1–3 The primary cause of varicose vein formation is not clear; however, both vein valve dysfunction and hydrostatic venous pressure appear to play a critical role in the initiation and progression of the disease.4–6 Although valve-reflux may precede vein-dilatation,7 there is a significant body of evidence supporting the view that vein dilation can precede venous reflux, and that valvular dysfunction may be an epiphenomenon of vein wall dilation.4,8–10 Clinical studies have demonstrated that venous insufficiency can occur in varicosities without axial reflux of the superficial, deep or perforator veins,11 and that an imbalance in extracellular matrix proteins may cause connective tissue changes prior to valvular insufficiency.12 Also, studies have shown increases in matrix metalloproteinases (MMPs) in the venous wall and the plasma of patients with varicose veins. Specifically, the expression and/or activity of MMP-1, -2, -3, -9, -12 and -13 are increased in varicose veins.13–17 Varicose veins are associated with increased expression of MMP-1 and -9 not only in the adventitial fibroblasts, but also in the medial smooth muscle and in endothelial cells.15 In varicose veins complicated with thrombophlebitis the content/activity of MMP-1, -2 and -9 in the vein wall are significantly elevated, which may cause extensive extracellular matrix remodeling and influence the mechanical properties of the vein wall, and thereby predispose to further progression of the disease.17

Previous studies have shown that mechanical stretch/pressure in human endothelial cells from muscle and in lung tissue subjected to high volume are associated with overexpression of MMP-2.18–19 Also, the expression of MMP-2, -9, and -14 is increased in fibroblasts, endothelial cells, and smooth muscle cells from human arteries and veins subjected to mechanical stretch.20–22 We have recently demonstrated that MMP-2 induces significant relaxation of rat inferior vena cava (IVC), possibly through smooth muscle cell hyperpolarization and activation of large conductance Ca2+-dependent K+ channel (BKca), a novel mechanism of MMP that may play a role in the early stages of venous dilation and varicose vein formation.23–24 Although increases in both venous pressure and MMPs have been implicated in the pathogenesis of varicose veins, little is known regarding the link between mechanical stretch of the vein wall, the MMPs expression in venous tissue, the contractile function, and the venous dilation associated with varicose veins. The objective of this study was to test the hypothesis that an increase in venous pressure/wall tension is associated with upregulation of specific MMPs in the vein wall, and venous smooth muscle dysfunction, leading to progressive venous dilation. We used rat IVC segments to test: 1) whether increases in magnitude and/or duration of vein wall tension are associated with decreased vein contraction. 2) Whether increased vein wall tension is associated with increased expression of specific MMPs. 3) Whether the specific MMPs expressed during increases in wall tension cause inhibition of vein contraction.

MATERIAL AND METHODS

Solutions and drugs

Physiological Krebs solution contained (in mmol/L): NaCl 120, KCl 5.9, NaHCO3 25, NaH2PO4 1.2, dextrose 11.5, CaCl2 2.5, MgCl2 1.2, bubbled with 95% O2 and 5% CO2 and maintained at pH 7.4. High KCl (96 mmol/L) was prepared as Krebs solution with equimolar substitution of NaCl with KCl. Tested compounds included phenylephrine (Phe), acetylcholine (Ach), and angiotensin II (AngII) (Sigma, St. Louis, MO), MMP-2 and MMP-9 (recombinant human, active form, Biomol, Plymouth Meeting, PA), and tissue inhibitor metalloproteinase-1 (TIMP-1) (recombinant human, Calbiochem, San Diego, CA). All other chemicals were of reagent grade or better.

We examined the effects of recombinant human MMPs and TIMP-1 on rat tissues. MMP-2 and -9 and TIMP-1 are highly conserved in human and rat tissues. MMPs (except MMP-23) are highly homologous Zn2+-dependent proteinases that share the same conserved cysteine switch, and have a catalytic domain with a Zn2+ binding motif that is conserved in different species. Similarly, TIMPs have conserved sequences in tissues from different species.25,26 In our previous studies on rat aorta we used recombinant human MMP-2 and -9.27 The commercial source (Biomol) makes human recombinant MMP-2 and -9 and purified mouse MMP-2. We opted to continue to use human MMP-2 and -9 and human TIMP-1 because that would allow us to compare the present results to our previous findings with MMPs. Also, using human MMP-2 and -9 and TIMP-1 enhances the clinical relevance of the vasodilator effects of MMPs. Additionally, our future goal is to test the effects of MMPs on human veins, and it would be advantageous to have reference data regarding their minimal and maximal effective dose and time course in rat tissues.

Animals and tissues

Male Sprague-Dawley rats (12 weeks, 250–300 g, Charles River Lab, Wilmington, MA) were euthanized by inhalation of CO2. Euthanasia was judged by cessation of breathing and heart beats. The abdominal cavity was opened, and the inferior vena cava (IVC) was rapidly excised, placed in Krebs solution, and carefully dissected and cleaned of connective tissue under microscopic visualization. The IVC was portioned into four 3 mm wide rings in preparation for isometric contraction experiments. All procedures followed the guidelines of the Institutional Animal Care and Use Committee.

Isometric contraction

Each of the 4 IVC segments was suspended between two tungsten wire hooks, one hook is fixed to a glass rod at the bottom of the tissue bath and the other hook is connected to a Grass force displacement transducer (FT03, Astro-Med Inc., West Warwick, RI). The 4 vein segments were mounted on the wire hooks in different tissue baths at no specific order to minimize the effects of variability in tissue size on the observed contractile response. Unless indicated otherwise, vein segments were stretched under 0.5g of basal tension and allowed to equilibrate for 1 hr in a temperature controlled, water-jacketed tissue bath, filled with 50 mL Krebs solution continuously bubbled with 95% O2 5% CO2 at 37°C. The changes in isometric contraction were recorded on a Grass polygraph (Model 7D, Astro-Med). The sensitivity of the contraction set-up was adjusted to detect changes in vein contraction between 20mg to 2g. As in many vascular beds, the diameter of vascular segments tends to be greater in the proximal segments closer to the heart. To correct for the effect of vein segment size and diameter on the magnitude of contractile response, the contraction was corrected for the tissue mass and presented in mg/mg tissue weight. Using these settings, the maximal control KCl (96 mmol/L)-induced contraction ranged between 80.6 and 260.8 (mean=181.0±17.2 mg/mg tissue, n=12). At the end of vein function experiments the tissues were gently padded dry and weighed, and stored at −80°C for immunoblot analysis and immunohistochemistry.

Basal tension-contraction relation

IVC segments incubated in normal Krebs were subjected to increasing basal tension of 0.0625, 0.125, 0.25, 0.5, 1, 2, and 3 g. Following equilibration of IVC at the specific basal tension for 30 min, the tissue bath solution was changed to 96 mmol/L KCl and contraction was measured. The veins were washed 3 times, 10 min each in Krebs solution prior to the next increase in basal tension and KCl contraction. We chose KCl to determine the tension-contraction relation because KCl (96 mmol/L) produces maximum and reproducible vein contraction that can be repeated multiple times. Also, KCl-induced contraction is receptor-independent, i.e. not dependent on α-adrenergic or AngII receptors, thus avoiding possible receptor saturation or desensitization with repeated agonist stimulation which could significantly affect the maximum vein contraction at the specific basal tension.

Prolonged basal tension and the KCl, Phe, and AngII contraction

Vein segments were stretched under 0.5g of basal tension and allowed to equilibrate for 1 hr. The control contraction/relaxation properties of the IVC segments were first determined. To determine the control contraction, IVC segments were stimulated twice with 96 mmol/L KCl solution, followed by washing in Krebs solution 3 times 10 min each. To test the IVC relaxation function, the tissues were stimulated with Phe 10−5 mol/L to achieve a steady-state contraction, then treated with acetylcholine (Ach, 1−5 mol/L) to confirm the presence of intact endothelium. The IVC segments were then subjected to 0.5g basal tension for 24 hr, or 2g for 24 hr, or 2g for 24 hr plus TIMP-1 (0.18–0.24 μg/mL). The vein segments were randomly assigned for a given experimental condition to reduce any potential bias. After 24 hr incubation at the specific basal tension the tissues were stimulated with 96 mmol/L KCl, Phe (10−5 mol/L), and AngII (10−6 mol/L), and the maximum contraction was measured.

The venous pressure corresponding to the low 0.5g and high 2g basal tension was calculated. LaPlace’s law states that tension (force) is proportional to pressure and as the tension increases, the pressure increases. The pressure generated was calculated using the formula P=F/A, where P=pressure expressed as gram-force/cm2, F=force, A=area, and assuming an average IVC diameter of 1.5 mm as detected by histology. At 0.5g tension the pressure generated is 28.4 gram-force/cm2 or 20.8 mmHg (1 torr=1 mmHg=19.337×1E-3 pound-force/inch2), and at 2g tension the pressure generated is 113.6 gram-force/cm2 or 83.4 mmHg.

Effect of MMPs on Phe contraction

IVC segments were subjected to 0.5g tension for 1 hr. The tissues were stimulated with Phe (10−5 mol/L) to elicit a contraction. When Phe-induced contraction reached steady-state, MMP-2 or -9 was added and its effect was observed. Our previous dose-response and time course experiments in rat aorta demonstrated that MMP-2 and -9 at 1 μg/mL caused maximum aortic relaxation in 30 min.27 Also, our previous experiments on rat IVC have shown time-dependent MMP-2 induced relaxation that reached steady-state in 30 min.23 These MMP-2 and -9 concentrations are physiologically consistent with the plasma and vein tissue levels in human which range between 1000 ng/g tissue (~1 mcg/ml) and 100 mcg/g tissue (~0.1 mcg/ml).24,28 Therefore, in all experiments 1 μg/mL MMP-2 or MMP-9 was used and the continuous trace of the effect of MMP on vein contraction was recorded for at least 30 min. To compare quantitatively the magnitude of Phe contraction in MMP-treated and non-treated veins, the contraction trace was analyzed for the magnitude of the response at selected time points at 5 min intervals for 30 min.

Immunoblot Analysis

Vein segments were homogenized in a 2 mL homogenization buffer at 4°C. The homogenate was centrifuged at 10,000g for 2 min, the supernatant collected and protein concentration was determined. Tissue homogenate was subjected to electrophoresis on 8% SDS polyacrylamide gel then transferred electrophoretically to nitrocellulose membranes (Bio-Rad, Hercules, CA). The membranes were incubated in 5% dried non-fat milk in PBS-Tween buffer for 1 hr, then incubated in the antibody solution (MMP-2 [1:500] or MMP-9 [1:200] rabbit polyclonal IgG, Santa Cruz Biotechnology) at 4°C for 24 hr. Actin was used as internal control to standardize loading and band intensities, and was detected using anti-α-smooth muscle actin monoclonal antibody (1:500,000, Sigma). The membranes were washed in PBS then incubated in horseradish peroxidase-conjugated secondary antibody (1:1000) for 1.5 hr. The membrane blots were washed with PBS, and visualized with ECL detection system (Amersham, Arlington Heights, IL). The reactive bands corresponding to the specific MMP were quantified by optical densitometry and analyzed using ImageJ software.

Immunohistochemistry

Cryosections of IVC (6 μm) were fixed in ice-cold acetone for 10 min. Endogenous peroxidase was quenched in 1.5% H2O2 solution and nonspecific binding blocked in 10% horse serum. Sections were treated with the specific MMP antibody. After rinsing with PBS, sections were incubated with biotinylated-secondary antibody, rinsed in PBS, and then incubated with avidin-labeled peroxidase (VectaStain Elite ABC Kit, Vector Lab, Burlingame, CA). Positive labeling was visualized using diaminobenzadine. Images were acquired on a Nikon microscope with digital camera mount and analyzed using Metamorph Imaging software.

Statistical Analysis

The data from 4 to 12 IVC segments from 3 to 6 rats were analyzed and presented as means±SEM. Data were analyzed using ANOVA followed by Scheffe’s F test for comparison of multiple means. Student’s t-test for unpaired and paired data was used for comparison of two means. Differences was considered significant if p<0.05.

RESULTS

Tension-contraction relationship

IVC segments subjected to step-wise increases in basal tension for short duration (30 min) demonstrated incremental contraction to 96 mmol/L KCl that reached a maximum at 0.25 to 0.5g basal tension. Increases in basal tension to 1g and 2g did not show any further increase in KCl contraction compared to 0.5g. When the basal tension was increased to 3g the KCl contraction was reduced significantly (n=8, p<0.001) (Fig. 1). Based on these results, we selected 0.5g as the normal control basal tension that produces maximum contraction. We also selected 2g as the maximum basal tension that still produces maximum contraction without causing excessive tissue stretch.

Fig. 1.

Fig. 1

Open in a new tab

Tension-KCl contraction relationship. IVC segments were subjected to increasing basal tensions (0.0625 to 3 g) for 30 min, then stimulated with 96 mmol/L KCl. KCl contraction was not significantly different at 0.5 to 2g basal tension. Further increase in wall tension to 3g was associated with significant decrease in contraction (p<0.01). Data points represent means±SEM of measurements from 8 experiments.

Effect of prolonged tension on Phe contraction

IVC segments subjected to 0.5g basal tension for 1 hr produced significant maximal contraction (127.4±11.6 mg/mg tissue) to the α-adrenergic agonist Phe (10−5 mol/L) that was maintained at 73.4±11.6 mg/mg tissue for 30 min. When the tissues were placed under 0.5g basal tension for 24 hr, a reduction in Phe contraction was observed (57.3±17.8 mg/mg tissue, p=0.03) as compared to IVC under 0.5g tension for 1 hr. In tissues subjected to 2g basal tension for 24 hr, the Phe contraction was further reduced to 40.8±7.3 mg/mg tissue (Fig. 2A). In tissues subjected to 2g basal tension for 24 hr but simultaneously treated with TIMP-1, the Phe contraction was partially restored to (68.8±17.6 mg/mg tissue, p=0.047), suggesting the involvement of MMPs.

Fig. 2.

Fig. 2

Open in a new tab

Effect of increases in basal tension on Phe- and AngII-induced contraction. IVC segments were subjected to 0.5g tension for 1 h, or 2g tension for 24 hr in the absence or presence of TIMP-1. The tissues were stimulated with Phe (10−5 mol/L) (A) or AngII (10−6 mol/L) (B) and the contractile response was recorded. Data points represent means±SEM of measurements from 4 experiments.

* Significantly different (p<0.05) from contraction at 0.5g for 1 hr basal tension.

§ Significantly different (p<0.05) from contraction at 2g for 24 hr basal tension.

Effect of prolonged tension on AngII contraction

To test whether the decreased IVC contraction with prolonged tension is related to specific changes in α-adrenergic receptors, we examined the effects of prolonged tension on another receptor-mediated contraction in response to AngII. In IVC segments under 0.5g basal tension for 1 hr, AngII (10−6 M) produced a transient response that reached a peak of 122.5±13.0 mg/mg tissue (Fig. 2B). In tissues under 0.5g basal tension for 24 hr, a reduction in AngII contraction was observed (48.0±14.1 mg/mg tissue, p=0.008) as compared to tissues under 0.5g tension for 1 hr. The AngII contraction was further reduced in tissues subjected to 2g basal tension for 24 hr (24.0±4.9 mg/mg tissue), and was partially restored in tissues simultaneously treated with TIMP-1 (79.5±8.2 mg/mg tissue, p=0.002), further suggesting the involvement of MMPs.

Effect of prolonged tension on KCl contraction

To further test whether the decreased IVC contraction with prolonged tension is related to changes in specific receptors, we examined the effects of prolonged tension on a receptor-independent contraction in response to membrane depolarization by KCl. KCl (96 mmol/L)-induced contraction was not significantly different in IVC segments subjected to 2g tension as compared to those subjected to 0.5g basal tension for 1 hr (Fig. 1). Experiments were performed on vein segments that are identically incubated under no tension or 0.5g tension for 24 hr, and the results were compared with vein segments incubated under 2g tension for 24 hr. KCl-induced contraction in tissues incubated under no tension for 24 hr (162.4±26.3 mg/mg tissue) was not significantly different from the initial control KCl-induced contraction (181.0±17.2 mg/mg tissue). The KCl induced response was reduced in vein segments incubated under 0.5g tension for 24 hr (86.7±22.0 mg/mg tissue) compared to control tissues under 0.5g tension for 1 hr. In comparison, the KCl induced response was significantly reduced (p<0.05) in tissues incubated under 2g tension for 24 hr (61.4±10.0 mg/mg tissue) (Fig. 3). These data demonstrate that the magnitude of KCl contraction is reduced with increasing amount of basal tension, and confirm that the applied tension is the important variable in determining the changes in contractile response. The KCl contraction appeared to be partially restored in tissues under 2g of tension for 24 hr and simultaneously treated with TIMP-1 (114.8±25.8 mg/mg tissue), but the changes only approached statistical significance (p=0.050) (Fig. 3).

Fig. 3.

Fig. 3

Open in a new tab

Effect of increases in basal tension on KCl-induced contraction. IVC segments were subjected to 0.5g tension for 1 hr, or 2g tension for 24 hr in the absence or presence of TIMP-1. The tissues were stimulated with KCl (96 mmol/L) and the contractile response was recorded. Data points represent the means±SEM of measurements from 4 experiments.

* Significantly different (p<0.05) from KCl contraction at 0.5g 1 hr basal tension.

Immunoblot analysis for MMPs

Immunoblot analysis in rat IVC subjected to 0.5g basal tension for 1 hr revealed measurable amounts of both MMP-2 and MMP-9. When normalized to actin, the amount of MMP-2 in the IVC appeared to be greater than that of MMP-9. In tissues subjected to 2g tension for 24 hr, the amount of MMP-2 and MMP-9 was markedly enhanced (Fig. 4). The relation between MMP expression and the Phe- and KCl-induced contraction at 0.5g for 1 hr and 2g for 24 hr basal tension was constructed. A correlation between MMP expression and the vein wall response to Phe- and KCl was observed (Fig. 5). Control experiments indicated that in tissues under 0g tension changing the duration of the experiment alone (1 hr versus 24 hr) was not associated with significant changes in MMP-2 expression as measured by immunoblots.

Fig. 4.

Fig. 4

Open in a new tab

Immunoblot analysis for MMPs in IVC. IVC segments were subjected to 0.5g tension for 1 hr or 2g tension for 24 hr. The tissues were rapidly frozen and prepared for Western blot analysis using specific anti MMP-2 (1:500) and anti MMP-9 antibody (1:200). The optical density of the immunoreactive bands was normalized to the house-keeping protein actin. Data points represent means±SEM of 8 measurements.

* Significantly different (p<0.05).

Fig. 5.

Fig. 5

Open in a new tab

Correlation between MMPs expression and Phe and KCl contractile response. IVC segments were subjected to 0.5g tension for 1 hr or 2g tension for 24 hr. After eliciting the contractile response to KCl (96 mmol/L) and Phe (10−5 mol/L), the vein segments were frozen and prepared for Western blot analysis using anti MMP-2 (1:500) and anti MMP-9 antibody (1:200). The optical density of the immunoreactive bands was normalized to actin. The relation between the amount of MMP-2 (A) or MMP-9 (B) and the contractile response to KCl and Phe was then constructed. Data points represent 15 measurements for MMP-2 and 8 measurements for MMP-9.

Some studies have shown that mechanical stretch can affect actin expression in venous tissues in culture.29 However, these studies examined the effect of stretch on contractile differentiation, and concluded that stretch stabilizes the contractile smooth muscle phenotype. Also, in these studies, the increase in stretch load caused an increase in total actin protein over 3 to 7 days. In contrast, the present experiments examined the changes in protein expression over a 24 hr period. Also, the present actin immunoblots did not reveal detectable variations in actin expression as detected by the naked eye and the optical densiometry measurements. These observations make it less likely that variations in the levels of actin with stretch could have biased the measurements of MMP/actin ratio.

Immunohistochemistry for MMPs

Immunohistochemistry in sections of IVC suggested that both MMP-2 and MMP-9 are expressed in the three layers of the IVC wall (Fig, 6). Although immunohistochemistry is largely a qualitative method, we attempted to measure the average intensity of MMP staining to make it easier to discern the differences in the MMP images (Table 1). Significant increases in MMP-2 and -9 immunostaining were observed in IVC subjected to 2g basal tension for 24 hr as compared to tissues under 0.5g basal tension. Also, a relative increase in the MMP immunostaining in the smooth muscle media could be observed in tissues under 2g tension for 24 hr (Fig. 6, Table 1). Control experiments indicated that in tissues under 0g tension changing the duration of the experiment alone (1 hr versus 24 hr) was not associated with significant changes in MMP-2 expression as measured by immunohistochemistry.

Fig. 6.

Fig. 6

Open in a new tab

Immunohistochemistry for MMPs in IVC. IVC segments at rest or subjected to 2g tension for 24 hr were rapidly frozen, and cryosections were prepared for immunohistochemical staining using anti MMP-2 or MMP-9 antibody (1:500). The amount of brown positive staining in the tunica intima (I), media (M) and adventitia (A) was measured. Hematoxylin and eosin staining was performed to test for integrity of the vessel wall. Data represent measurements from 4–5 pictures of tissue sections. Total magnification 400.

Table 1.

Amount and distribution of MMP-2 and MMP-9 in rat IVC incubated at rest or under 2g basal tension for 24 hr.

MMP-2
MMP-9
Control 2g 24 hr p Control 2g 24 hr p
Amount (% Total Area) 8.98±0.85 11.57±1.05 0.104 13.86±1.49 21.68±2.18 0.018 *
Distribution (pixels/μm2)
Intima 3.61±1.02 6.96±0.85 0.045 * 5.45±0.32 7.35±0.48 0.011 *
Media 2.92±0.32 4.54±0.92 0.147 3.99±0.68 6.59±0.86 0.045 *
Adventitia 0.46±0.06 0.64±0.12 0.228 1.07±0.23 0.94±0.12 0.630
Open in a new tab

Data represent means±SEM of measurements in 4 to 5 pictures of tissue sections.

*

Indicates significant difference (p<0.05).

Effect of MMPs on IVC contraction

In a previous study we demonstrated that MMP-2 (1 μg/ml) caused significant relaxation of Phe contraction to 39.5±10.5% of control contraction in 30 min (p=0.029).23 In the present experiments, MMP-9 (1 μg/ml) caused similar time-dependent relaxation of Phe-induced contraction in IVC segments (Fig. 7). Comparison of MMP-2 and MMP-9 induced relaxation demonstrated no significant difference in the time course or magnitude of IVC relaxation between MMP-2 and MMP-9 (p=0.32).

Fig. 7.

Fig. 7

Open in a new tab

Effects of MMPs on Phe contraction of IVC segments. IVC segments were subjected to 0.5g tension for 1 hr and either nontreated or pretreated with TIMP-1. The tissues were stimulated with Phe (10−5 mol/L). When the Phe contraction reached a steady state, the tissues were treated with MMP-2 or MMP-9 (1 μg/ml) and the continuous trace of the effect of MMP on vein contraction was recorded for at least 30 min. To compare quantitatively the magnitude of Phe contraction in MMP and TIMP-1 treated and non-treated veins, the contraction trace was analyzed for the magnitude of the response at selected time points at 5 min intervals for 30 min. Data points represent means±SEM of measurements from 3–5 experiments.

* Significantly different (p<0.05) from control.

We also tested the effect of TIMP-1, inhibitor of MMP-2 activity, on tissues under 0.5g load and found that TIMP-1 alone did not cause any contractile response. On the other hand, the inhibitory effects of MMP-2 on Phe-induced contraction were prevented in IVC segments pretreated with TIMP-1 (Fig. 7), confirming that the effects are likely due to MMP activity and thereby supporting the specificity of the MMP effects.

DISCUSSION

The main findings of the present study are: 1) IVC segments subjected to high 2g basal tension for prolonged 24 hr show significant reduction in Phe and AngII contraction, 2) Phe and AngII contractions were partially restored in vein segments under 2g tension for 24 hr and treated with TIMP-1, 3) High 2g vein wall tension for 24 hr was associated with increased amount of MMP-2 and MMP-9, and 4) Similar to our previous observation with MMP-2, MMP-9 caused significant inhibition of Phe contraction in IVC segments.

We investigated the relation between venous pressure/wall tension, MMP expression and venous tissue dysfunction. We first examined whether increases in the magnitude of basal tension cause changes in venous contraction. We found that 0.5g basal tension for 30 min produced maximal IVC contraction to KCl, and therefore 0.5g was used as the control basal tension. Also, increases in basal tension up to 2g did not change the maximal IVC contraction to KCl, and therefore 2g was used as the maximal basal tension that could be applied to the veins without causing significant tissue damage/fatigue.

We then tested whether prolonged increases in vein wall tension affect venous contraction. Phe activates α-adrenergic receptors,30 and AngII activates angiotensin type 1 receptors.31 We found that prolonged increases in basal tension are associated with reduction in Phe-induced IVC contraction, which may not be specific to α-adrenergic mediated responses as AngII-induced contraction was similarly reduced under these conditions. These findings suggest that the decrease in vein contraction following prolonged stretch is independent of the type of receptor stimulated, and may involve inhibition of a common signaling pathway downstream from receptor activation.

To determine whether the decreased vein contraction with prolonged increases in vein wall tension involves increases in MMP activity, we tested the effects of MMP inhibitors. We have previously tested whether the inhibitory effects of MMP on vascular contraction can be reversed by MMP-2/MMP-9 Inhibitor IV and α-2 macroglobulin.27 Unfortunately, these MMP inhibitors exert an inhibitory effect on vascular contraction, possibly because they bind Zn2+ and other divalent cations such as Ca2+, a major determinant of vascular contraction, and hence would confound the results.27 We therefore tested the effects of TIMP-1, inhibitor of gelatinases A and B.24,25 The observation that the reduced IVC contraction in tissues subjected to prolonged basal tension was reversed in tissues treated with TIMP-1, suggest possible involvement of MMP-2 and -9. However, at least 26 MMPs have been identified and classified into major subgroups including interstitial collagenases, gelatinases, stromelysins, and membrane-type MMPs (MT-MMPs). Other MMP subgroups include matrilysins, enamelysin, and macrophage metalloelastase.24–26 Although TIMP-1 inhibits MMP-2 and -9, it could also bind to and inhibit membrane type-1 MMP (MT1-MMP), and perhaps other MMPs. Future experiments should test the effects of selective MMP monoclonal antibodies and siRNA and thereby further elucidate the specific MMP(s) involved in the decreased contraction associated with increased vein wall tension.

The present immunoblot analysis in IVC suggested increased expression of MMP-2 and -9 during prolonged increases in basal tension. Also, immunohistochemical staining suggested localization of MMP-2 and -9 in the three layers of the vein wall. Additionally, prolonged increases in wall tension were associated with relative increases in MMP-2 and -9 in the vicinity of the smooth muscle layer, suggesting an effect on the contractile cells. Our data in the rat IVC are consistent with previous reports that capillaries within the skeletal muscle subjected to mechanical stretch show increased MMP-2 mRNA transcription.19 Also, cultured vascular smooth muscle cells from mice and exposed to cyclic mechanical stretch demonstrate an increase in MMP-2 mRNA.21 The observation that the decreased IVC contraction during prolonged increases in basal tension was reversed in tissues pretreated with TIMP-1 suggests that the increased MMP amount is associated with increased MMP activity. Future zymography experiments would confirm whether the decreased contraction in venous tissues exposed to prolonged increases in tension reflect increases not only in the amount but also the activity of MMPs.

The question remains as of how MMP expression and changes in venous contraction could be interrelated. We have previously shown that treatment of aortic strips with MMP-2 and -9 did not cause significant changes in the blood vessel structure or morphometry as determined by H&E staining. Also, MMPs did not cause changes in the elastin architecture as determined by Verohoff Von Giesson, suggesting minimal elastolytic effects of MMPs within the acute time frame tested.27 The present data also showed that the IVC was still responsive to Phe despite prolonged tension, suggesting that α-adrenergic receptors were still intact. Also, the observation that prolonged tension was associated with reduced contraction not only to Phe but also to AngII indicates that the reduction in contraction is not specific to a particular receptor. Furthermore, the Phe and AngII contraction were restored in veins treated with TIMP-1, suggesting that the reduction in contraction was not due to reduction in the Phe or AngII receptors, but rather due to increased MMP. It could be argued that MMPs may proteolyze the membrane receptors. This is unlikely because prolonged stretch was also associated with reduced contraction to 96 mM KCl which stimulates receptor-independent Ca2+ entry, supporting the contention that the reduced contraction in response to prolonged increases in tension involves a post-receptor mechanism. We have previously shown that MMP-2 and -9 inhibit Phe-induced Ca2+ influx in rat aortic segments.27 We have also shown that MMP-2 causes significant inhibition of Phe- and AngII-induced IVC contraction likely through a post-receptor mechanism involving activation of plasmalemmal K+ channels, membrane hyperpolarization, and inhibition of Ca2+ influx.23 The present findings confirm our previous observations with MMP-2 and demonstrate that MMP-9 cause significant inhibition of IVC contraction.

Several studies have demonstrated the presence of MMPs in varicose veins.14,15 We found that MMP-9, similar to MMP-2, caused time-dependent relaxation of Phe induced contraction in rat IVC, suggesting that the effects of MMPs on IVC relaxation are shared within the gelatinase family of MMPs. However, the present study does not exclude the possibility that the presence/activation of MMP-2 may lead to activation of other MMPs.32,33 Also, while TIMP-1 can inhibit multiple MMPs, it may have some selectivity for some MMPs over others.23,25 This may explain the present observation that TIMP-1 only partially restored IVC constriction in tissues subjected to prolonged increases in basal tension.

The present data suggest that MMP-2 and -9 have acute venodilatory effect in addition to their known effects on the extracellular matrix.24–26 We have previously shown that the MMP-2 and MMP-9 induced inhibition of Phe contraction in vascular segments is dose-dependent.27 Also, our previous 23 and present results demonstrate that the venodilator effects of MMP-2 and -9 are time-dependent, within a total experimental time period of 30 min. Whether the acute venodilator effects of MMPs could cause prolonged and progressive venous dilation in vivo remains to be examined. Interestingly, a recent study has evaluated morphologic changes in venous valves and MMP expression in the vein wall in an in vivo rat model of venous hypertension induced by a femoral artery-vein fistula.34 The study demonstrated an increase in the femoral vein pressure (96±9 mmHg) and progressive reflux 42 days after induction of arterio-venous fistula. The valves distal to the fistula demonstrated increased diameter and fibrosis in the media and adventitia. Also, MMP-2 and -9 were significantly elevated after 21 and 42 days of venous hypertension.34 The present experiments have shown increased MMP expression and decreased constriction in veins exposed to high basal tension for 24 hr, suggesting that biochemical and functional changes in the veins occur in as early as one day of increased vein wall tension.

In the present study, we examined the effects of prolonged and consistent increases in vein wall tension on MMP expression and vein function. Whether cyclic increases in vein wall tension would differentially affect the MMP expression profile and vein function remains to be examined. Interestingly smooth muscle cells cultured from human saphenous vein and subjected to static strain have demonstrated increased MMP-2 and -9 mRNA expression as well as proMMP-2 and -9 protein, while cyclic strain (stretch) decreased the expression of MMPs.22 These data in human saphenous vein cells indicate that constant stretch is required for the overexpression of MMPs, which is consistent with the increased venous hydrostatic pressure seen in patients with chronic venous insufficiency.35

Although the IVC is less prone to varicosities, we used the rat IVC because it can be rapidly dissected and consistently responds to vasoconstrictors and vasodilators in a reproducible fashion.23 Future studies on iliac and femoral vein would increase the relevance to venous disease of the lower extremity. While inherent differences in the physiologic behavior and structure of rat veins and human veins make it difficult to extrapolate the response of rat IVC to human veins, the results indicate that the magnitude and duration of vein wall tension have significant impact on venous function and MMP expression. We do not wish to equate the results in rat IVC to human veins. On the other hand, studies in the rat make it easier to obtain the veins without delay and measure vein function rapidly. Also, using the same rat strain, age and gender avoids potential confounding factors and heterogeneity observed in discarded samples of human veins. The availability of discarded human veins with minimal surgical manipulation is critical to test the effects of increases in human vein wall tension on contraction and MMP expression, and to compare vein function and MMP expression in “normal” and varicose veins.

It is important to note that the 0.5g and 2g basal tension used in this study correspond to 20.8 mmHg and 83.4 mmHg venous pressure, respectively. In normal individuals of average 5-foot 10-inch height in the standing position the column of blood in the venous system reflects a venous pressure at the ankle of 90–100mm Hg. Thus, the pressure generated at high 2g basal tension approaches those seen in the lower extremity veins of humans in the upright position. However, in varicose veins as the perforating veins develop valve malfunction, the high pressures developed in the deep veins with muscle contractions up to 250 mmHg are transmitted to the superficial veins. Applying the actual pressure generated in human varicose veins would likely damage the IVC segments. We found that increases in basal tension above 2g caused significant and irreversible reduction in IVC contraction (see Fig 1), suggesting structural changes and irreversible vein damage. Therefore, high basal tensions above 2g were not used to avoid overstretch-induced structural changes that could complicate the interpretation of the tension-MMP expression relationship. Future studies in human vein specimens would allow examination of tensions similar to those observed in patients with varicosities/venous hypertension.

In conclusion, in rat IVC, increases in magnitude and duration of wall tension are associated with reduced contraction and overexpression of MMP-2 and -9. Also, MMP-9 induced IVC relaxation is consistent with our previous findings with MMP-2. The present data lay the ground for further experiments to test the possibility that protracted increases in venous pressure/wall stretch increase MMP-2 and MMP-9 expression, which in turn produces venous dilation and further increases in venous pressure, leading to a recalcitrant cycle and progressive venous dilation.

Acknowledgments

This work was supported by grants from National Heart, Lung, and Blood Institute (HL-65998, and HL-70659).

List of Abbreviations

AngII

angiotensin II

IVC

inferior vena cava

MMP

matrix metalloproteinase

Phe

phenylephrine

TIMP-1

tissue inhibitor of metalloproteinase-1

REFERECES

  • 1.Aunapuu M, Arend A. Histopathological changes and expression of adhesion molecules and laminin in varicose veins. Vasa. 2005;34:170–5. doi: 10.1024/0301-1526.34.3.170. [DOI] [PubMed] [Google Scholar]
  • 2.Goldman MP, Weiss RA, Bergan JJ. Diagnosis and treatment of varicose veins: a review. J Am Acad Dermatol. 1994;31:393–413. doi: 10.1016/s0190-9622(94)70202-0. [DOI] [PubMed] [Google Scholar]
  • 3.Goldman MP, Fronek A. Anatomy and pathophysiology of varicose veins. J Dermatol Surg Oncol. 1989;15:138–45. doi: 10.1111/j.1524-4725.1989.tb03020.x. [DOI] [PubMed] [Google Scholar]
  • 4.Raffetto JD, Khalil RA. Mechanism of varicose vein formation. Valve dysfunction and wall dilation. Phlebology. 2008 doi: 10.1258/phleb.2007.007027. in press. [DOI] [PubMed] [Google Scholar]
  • 5.Bergan JJ. Surgical management of primary and recurrent varicose veins. In: Gloviczki P, Yao JST, editors. Handbook of Venous Disorders: Guidelines of the American Venous Forum. Chapman & Hall; New York: 1996. pp. 394–415. [Google Scholar]
  • 6.Pascarella L, Schmid Schonbein GW. Causes of telengiectasias, reticular veins, and varicose veins. Semin Vasc Surg. 2005;18:2–4. doi: 10.1053/j.semvascsurg.2004.12.004. [DOI] [PubMed] [Google Scholar]
  • 7.Schultz-Ehrenburg U, Weindorf N, Matthes U, Hirche H. An epidemiologic study of the pathogenesis of varices. The Bochum study I–III. Phlebologie. 1992;45(4):497–500. [PubMed] [Google Scholar]
  • 8.Gandhi RH, Irizarry E, Neckman GB, Halpern VJ, Mulcare RJ, Tilson MD. Analysis of the connective tissue matrix and proteolytic activity of primary varicose veins. J Vasc Surg. 1993;20:814–20. [PubMed] [Google Scholar]
  • 9.Kirsch D, Wahl W, Bottger T, Junginger T. Primary varicose veins--changes in the venous wall and elastic behavior. Chirurg. 2000;71:300–5. doi: 10.1007/s001040050049. [DOI] [PubMed] [Google Scholar]
  • 10.Kirsch D, Dienes HP, Kuchle R, et al. Changes in the extracellular matrix of the vein wall--the cause of primary varicosis? Vasa. 2000;29:173–7. doi: 10.1024/0301-1526.29.3.173. [DOI] [PubMed] [Google Scholar]
  • 11.Labropoulos N, Kang SS, Mansour MA, Giannoukas AD, Buckman J, Baker WH. Primary superficial vein reflux with competent saphenous trunk. Eur J Vasc Endovasc Surg. 1999;18:201–6. doi: 10.1053/ejvs.1998.0794. [DOI] [PubMed] [Google Scholar]
  • 12.Gandhi RH, Irizarry E, Neckman GB, Halpern VJ, Mulcare RJ, Tilson MD. Analysis of the connective tissue matrix and proteolytic activity of primary varicose veins. J Vasc Surg. 1993;20:814–20. [PubMed] [Google Scholar]
  • 13.Jacob MP, Cazaubon M, Scemama A, Prie D, Blanchet F, Guillin MC, Michel JB. Plasma matrix metalloproteinase-9 as a marker of blood stasis in varicose veins. Circulation. 2002;106:535–8. doi: 10.1161/01.cir.0000027521.83518.4c. [DOI] [PubMed] [Google Scholar]
  • 14.Gillespie DL, Patel A, Fileta B, Chang A, Barnes S, Flagg A, Kidwell M, Villavicencio JL, Rich NM. Varicose veins possess greater quantities of MMP-1 than normal veins and demonstrate regional variation in MMP-1 and MMP-13. J Surg Res. 2002;106:233–8. doi: 10.1006/jsre.2002.6455. [DOI] [PubMed] [Google Scholar]
  • 15.Woodside KJ, Hu M, Burke A, Murakami M, Pounds LL, Killewich LA, Daller JA, Hunter GC. Morphologic characteristics of varicose veins: possible role of metalloproteinases. J Vasc Surg. 2003;38:162–9. doi: 10.1016/s0741-5214(03)00134-4. [DOI] [PubMed] [Google Scholar]
  • 16.Kosugi I, Urayama H, Kasashima F, Ohtake H, Watanabe Y. Matrix metalloproteinase-9 and urokinase-type plasminogen activator in varicose veins. Ann Vasc Surg. 2003;17:234–8. doi: 10.1007/s10016-003-0005-2. [DOI] [PubMed] [Google Scholar]
  • 17.Kowalewski R, Sobolewski K, Wolanska M, Gacko M. Matrix metalloproteinases in the vein wall. Int Angiol. 2004;23:164–9. [PubMed] [Google Scholar]
  • 18.Haseneen NA, Vaday GG, Zucker S, Foda HD. Mechanical stretch induces MMP-2 release and activation in lung endothelium: role of EMMPRIN. Am J Physiol Lung Cell Mol Physiol. 2003;284:L541–7. doi: 10.1152/ajplung.00290.2002. [DOI] [PubMed] [Google Scholar]
  • 19.Milkiewicz M, Haas TL. Effect of mechanical stretch on HIF-1{alpha} and MMP-2 expression in capillaries isolated from overloaded skeletal muscles: laser capture microdissection study. Am J Physiol Heart Circ Physiol. 2005;289:H1315–20. doi: 10.1152/ajpheart.00284.2005. [DOI] [PubMed] [Google Scholar]
  • 20.Zhou D, Lee HS, Villarreal F, Teng A, Lu E, Reynolds S, Qin C, Smith J, Sung KL. Differential MMP-2 activity of ligament cells under mechanical stretch injury: an in vitro study on human ACL and MCL fibroblasts. J Orthop Res. 2005;23:949–57. doi: 10.1016/j.orthres.2005.01.022. [DOI] [PubMed] [Google Scholar]
  • 21.Grote K, Flach I, Luchtefeld M, Akin E, Holland SM, Drexler H, Schieffer B. Mechanical stretch enhances mRNA expression and proenzyme release of matrix metalloproteinase-2 (MMP-2) via NAD(P)H oxidase-derived reactive oxygen species. Circ Res. 2003;92:e80–6. doi: 10.1161/01.RES.0000077044.60138.7C. [DOI] [PubMed] [Google Scholar]
  • 22.Asanuma K, Magid R, Johnson C, Nerem RM, Gslis ZS. Uniaxial strain upregulates matrix-degrading enzymes produced by human vascular smooth muscle cells. Am J Physiol Heart Circ Physiol. 2003;284:H1778–84. doi: 10.1152/ajpheart.00494.2002. [DOI] [PubMed] [Google Scholar]
  • 23.Raffetto JD, Ross RL, Khalil RA. Matrix metalloproteinase-2 induced venous dilation via hyperpolarization and activation of K+ channels: Relevance to varicose vein formation. J Vasc Surg. 2007;45:373–80. doi: 10.1016/j.jvs.2006.10.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Raffetto JD, Khalil RA. Matrix metalloproteinases and their inhibitors in vascular remodeling and vascular disease. Biochem Pharmacol. 2008;75(2):346–59. doi: 10.1016/j.bcp.2007.07.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Visse R, Nagase H. Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function, and biochemistry. Circ Res. 2003;92:827–39. doi: 10.1161/01.RES.0000070112.80711.3D. [DOI] [PubMed] [Google Scholar]
  • 26.Nagase H, Visse R, Murphy G. Structure and function of matrix metalloproteinases and TIMPs. Cardiovasc Res. 2006;69:562–73. doi: 10.1016/j.cardiores.2005.12.002. [DOI] [PubMed] [Google Scholar]
  • 27.Chew DKW, Conte MS, Khalil RA. Matrix metalloproteinase-specific inhibition of Ca2+ entry mechanisms of vascular contraction. J Vasc Surg. 2004;40:1001–10. doi: 10.1016/j.jvs.2004.08.035. [DOI] [PubMed] [Google Scholar]
  • 28.Kowalewski R, Sobolewski K, Wolanska M, Gacko M. Matrix metalloproteinases in the vein wall. Int Angiol. 2004;23:164–9. [PubMed] [Google Scholar]
  • 29.Zeidan A, Nordström I, Albinsson S, Malmqvist U, Swärd K, Hellstrand P. Stretch-induced contractile differentiation of vascular smooth muscle: sensitivity to actin polymerization inhibitors. Am J Physiol Cell Physiol. 2003;284(6):C1387–96. doi: 10.1152/ajpcell.00508.2002. [DOI] [PubMed] [Google Scholar]
  • 30.Murphy JG, Khalil RA. Gender-specific reduction in contractility and [Ca2+]i in vascular smooth muscle cells of female rat. Am J Physiol Cell Physiol. 2000;278:C834–44. doi: 10.1152/ajpcell.2000.278.4.C834. [DOI] [PubMed] [Google Scholar]
  • 31.Murphy JG, Herrington JN, Granger JP, Khalil RA. Enhanced [Ca2+]i in renal arterial smooth muscle cells of pregnant rats with reduced uterine perfusion pressure. Am J Physiol Heart Circ Physiol. 2003;284:H393–403. doi: 10.1152/ajpheart.00247.2002. [DOI] [PubMed] [Google Scholar]
  • 32.Norgauer J, Hildenbrand T, Idzko M, Panther E, Bandemir E, Hartmann M, et al. Elevated expression of extracellular matrix metalloproteinase inducer (CD147) and membrane-type matrix metalloproteinases in venous leg ulcers. Br J Dermatol. 2002;147:1180–6. doi: 10.1046/j.1365-2133.2002.05025.x. [DOI] [PubMed] [Google Scholar]
  • 33.Dozier S, Escobar GP, Lindsey ML. Matrix metalloproteinase (MMP)-7 activates MMP-8 but not MMP-13. Med Chem. 2006;2:523–6. doi: 10.2174/157340606778250261. [DOI] [PubMed] [Google Scholar]
  • 34.Pascarella L, Schmid-Schonbein GW, Bergan JJ. An animal model of venous hypertension: the role of inflammation in venous valve failure. J Vasc Surg. 2005;41:303–11. doi: 10.1016/j.jvs.2004.10.038. [DOI] [PubMed] [Google Scholar]
  • 35.Neglen P, Raju S. Ambulatory venous pressure revisited. J Vasc Surg. 2000;31:1206–13. doi: 10.1067/mva.2000.105669. [DOI] [PubMed] [Google Scholar]

RESOURCES