Cystamine Treatment Fails to Prevent the Development of Pulmonary Hypertension in Chronic Hypoxic Rats

Effect of Cystamine on Transamidase Activity in vitro and ex vivo

To measure transamidase activity in vitro, the Specific Tissue Transglutaminase Colometric Microassay Kit TG2-CovtTest (Zedira, Darmstadt, Germany) was used. This test is based on the incorporation of a biotinylated substrate specific for TG2 (Biotin-pepT26) into an amine donor (spermine)-coated strip [26^, 27]. Protein extracts from normoxic rat liver containing 250 μg of total protein (measured by a modified Lowry method) were incubated for 20 min with different doses of cystamine (25–2,500 μmol/L), and the assay was then performed according to the manufacturer’s indications. Assays were also carried out as described above in protein extracts from normoxic rat lungs containing 5 μg of total protein, but due to reduced sample amount, the Tissue Transglutaminase Colometric Picoassay Kit (Zedira, Darmstadt, Germany) was used instead, which is less specific as it also detects the activity of TG3 and factor XIII but presents higher sensitivity. This picoassay kit is similar to the TG2-Covtest but it uses biotinyl-cadaverine as a substrate and HA-HMS as the amino donor coating the strips.

To measure transglutaminase activity in more physiologically relevant conditions, we used an assay based on the incorporation of the tranglutaminase substrate 5-(biotinamido)pentylamine (BP) to structural proteins was used as previously described [19]. Sixteen-week-old female C57BL/6 mice were exposed to hypobaric hypoxia while others were left as the normoxic control group. After 2 weeks of exposure to normoxia or hypoxia, the animals were weighed and euthanized by cervical dislocation. Immediately after the mice were sacrificed, heart, lungs, and liver were removed and kept in cold (5°C) physiological saline solution (PSS). The heart was isolated, and both atria were removed. The free wall of the right ventricle (RV) was separated from left ventricle and septum (LV + S). Each part was weighed and the ratios of right ventricle to left ventricle plus septum (RV/(LV + S)) were calculated and used to assess right ventricular hypertrophy. Lung tissue contains a large amount of vessels with lumen diameters of 20–80 μm, where the remodeling contributing to increased pulmonary pressure takes place in chronic hypoxic mice and rats [28^, 29]. Left lung was then divided and incubated in PSS-containing BP (1 mmol/L) and an increased concentration of Ca²⁺ (2.5 mmol/L) at 37°C for 4 h, in presence and absence of cystamine (1.5 and 15 μmol/L). During incubation, PSS was continuously bubbled with 5% CO₂-75% N₂-20% O₂ to maintain pH (7.4). Then, unreacted BP was washed by rinsing the arteries with PBS before tissue homogenization and protein extraction. Then 10 μg of total proteins were subjected to Western blot using a stain-free gel (Bio-Rad). After transfer to the membrane, total protein was visualized using stain-free technology by exposing the membranes to UV light and the images were captured using a PXi 4 Touch image analysis system (Syngene). The membrane was blocked in 5% bovine serum albumin (BSA) for 2 h and then incubated with HRP-conjugated streptavidin (Amersham Bioscience; dilution 1:10,000 in 0.5% BSA) for 2 h to detect BP incorporation. The membrane was then developed using an ECL-Plus kit (General Electric [GE] Health care, Copenhagen, Denmark). The images were captured by a luminescence camera using a PXi 4 Touch image analysis system (Syngene). Blots were quantified using GeneTools 4 software (Syngene) and normalized to the total protein. For each experimental group, the average activity in control conditions was set to 100% and the results are expressed as percentage of transamidase activity normalized to control conditions.

Chronic Hypoxic Rat Model

All animal experiments followed the revised NIH publication no. 86–23, entitled ”Principle of laboratory animal care,” and they were approved by the National Danish Animals Experiments Inspectorate (Permission 2006/561-1160 and Permission 2011/561-2011). Nine-week-old male Wistar rats (n = 80) were divided into a normoxic group and 4 groups exposed to hypobaric hypoxia and the following treatments: (1) hypoxic group treated with vehicle (H); (2) hypoxic group treated with cystamine (HC); (3) hypoxic group treated with sildenafil (HS); (4) hypoxic group treated with cystamine and sildenafil (HCS).

Sildenafil was dissolved in the drinking water with dosage of 25 mg/kg/day adjusted to water intake that was similar in sildenafil and vehicle-administered animals. Cystamine was administered subcutaneously in mini-osmotic pumps (Model 2002; Alzet, Cupertino, CA, USA) while vehicle-treated animals were sham operated with pumps filled with distilled water. We have previously found that a dosage of 80 mg/kg/day resulted in plasma cystamine concentrations of ∼2.5 μM and lowering of the systemic blood pressure [10], while 40 mg/kg/day inhibited small arterial remodeling without changing the systemic blood pressure [9]. Therefore, the animals were treated with a dosage of 40 mg/kg/day.

The mini-osmotic pumps were implanted in the neck of the rats under subcutaneously administrated anesthesia with midazolam 0.4125 mg/100 g and fentanyl citrate 0.0381 mg/100 g plus fluanisone 0.825 mg/100 g (rodent mixture). The rats were given postoperational analgesia with Temgesic 0.13 mL. After 2 days, they were exposed to hypobaric hypoxia or normoxia. They were stored in a 12-h light:dark cycle at 20°C ventilated room with free access to water and chow.

In the hypobaric hypoxic chamber, the pressure was decreased to 560 mbar corresponding to an oxygen tension of 10–12%. The chamber was opened 3 times a week for about 30 min for cleaning the cages and renewing of water and chow.

After 2 weeks of exposure to normoxia or hypoxia, the animals were weighed and anesthetized by inhalation of isoflurane followed by subcutaneous injection of rodent mixture. Anesthesia was supplied during the procedure, and it was assessed by checking deep reflexes.

Hemodynamic Measurements

The animal was deposited on its back on a heat plate and with a rectal probe to measure the core temperature (Physitemp Instruments Inc., Clifton, NJ, USA). A pressure catheter (2F MicroTip Pressure Catheter; Millar Instruments Inc., Houston, TX, USA) was inserted into the RV through the right jugular vein. The catheter was connected to a QUAD Bridge connected to Powerlab 4/20 (AD Instruments, Oxfordshire, UK), and the pressure profile was registered by Chart 5.5 software (AD Instruments). After measurement of the RVSP and heart rate, the catheter was inserted into the left carotid artery for measurement of systemic blood pressure. Pressure profiles were recorded for 3 min. Right ventricular end-diastolic pressure (RVEDP), changes in the pressure over time during isovolumetric contraction and relaxation (dP/dt_max and dP/dt_min, respectively) were calculated directly from the pressure curves.

Cardiac output (CO) was measured in rats not used for RVSP measurements using the thermodilution technique in anaesthetized animals as described previously [24]. Briefly, the rat was anesthetized as described previously and a catheter for delivery of 0.2 mL saline, 22–24°C, inserted in the right jugular vein. The thermodilution curve was recorded with a T-type ultrafast thermocouple (AD Instruments, Oxfordshire, England) placed in the aorta, and CO was calculated as earlier described [30] using Chart 7 Cardiac Output Software (AD Instruments, Oxfordshire, England). Systemic vascular resistance and PVR were calculated as mean pressures divided by CO.

Organ Weights and Muscularization of Pulmonary Arteries

The animals were sacrificed by decapitation. The heart, lungs, and liver were removed and stored in cold PSS. Right ventricular hypertrophy was evaluated as (RV/LV + S). Liver and lung weights were obtained as well.

For evaluation of pulmonary arterial muscularization, the left lung was isolated; a catheter inserted into the pulmonary artery, and 0.1 mL of heparin 100 IE/mL was injected. The lung was perfused with PSS-containing albumin (2%) followed by PSS with papaverine (10⁻⁴M) and finally with formalin 4% at a perfusion pressure of 20 mm Hg. Sections of the lung were stored in formalin buffer 4%, embedded in paraffin and stained with antibody against human smooth muscle actin (1:200; Dako, Glostrup, Denmark) and human von Willebrand factor (1:4,000; Dako, Denmark). At ×400 magnification small pulmonary vessels of each animal ranging from 20 to 80 μm in internal diameter were counted by a blinded observer and noted as muscular, partially muscular, or nonmuscular. The numbers are expressed as percentage of the total number of small arteries.

Sirius Red Staining for Collagen

Collagen content was assessed using picrosirius red staining [31]. The sections of the left lung were fixated with formaldehyde and imbedded with paraffin and cut in 10 μm transverse sections. After deparaffinization, the sections were incubated with Picro-Sirius Red solution (Abcam, UK). The sections were rinsed twice in a 0.5% acetic acid solution, followed by dehydration in absolute alcohol.

In a blinded manner, a minimum of 6 images per animal were randomly obtained and analyzed under circularly polarized white light. The images were processed in ImageJ (image analysis software, US National Institutes of Health, Bethesda, MA, USA). The red-stained collagen was identified in grayscale images, using adjusted threshold color. The resulting data on collagen content in lung tissue are given as percent collagen area of the tissue area.

Localization and Immunoblotting of TG2

Pieces from the RV from normoxic and hypoxic rats were stored in formalin 4% until they were embedded into paraffin. Slices of approximately 3 μm were made from each piece. Removing paraffin from the sections was done by washing in different solutions each for 5 min following the schedule: 2x xylene (xylene: mixture of isomers [AppliChem, Germany]), 2 × 99% ethanol, 96% ethanol, 50% ethanol, and water. The slices were incubated in 3% H₂O₂ for 10 min and afterward washed 2 × 5 min in Coon’s buffer (Na₂H₂PO₄ [2H₂O], NaH₂PO₄ [H₂O], NaCl – ad. 10,000 mL H₂O). The sections were treated with either citrat buffer (10 mM [Tri-sodium citrate dihydrate 5 mM, dinatriumhydrogencitrat 5 mM, pH adjusted to 6.0]) or TEG buffer (Tris 10 mM, EGTA 0.5 mM, pH adjusted to 9.0), then for 2 × 5 min in a microwave at 650 W and afterward washed in Coon’s buffer. The serum was removed and the primary antibody to TG2 (1:1,000) diluted in 1% BSA was added and incubated overnight at 4°C in a moisture chamber.

The sections were washed in Coon’s buffer and then incubated with secondary antibody (Bionylated Link – Universal LSAB^TM 2 Kit/HRP, Rabbit/Mouse; Dako, CA, USA) for 20 min at 20°C in a moisture chamber, then the samples were washed with Coon’s buffer and incubated with streptavidin (Streptavidin; Dako) for 20 min at 20°C in a moisture chamber. After washing with Coon’s buffer, DAB (DAB Chromogen tablets dissolved in Coon’s buffer; Dako) was added for 5 min. The sections were incubated with Mayer’s hematoxylin for 1 min, washed in H₂O and incubated for 5 min in each solution: 50% ethanol, 96% ethanol, and 99% ethanol. Negative controls were made exactly identical but without incubation with primary antibody.

The lungs for immunoblotting were perfused with heparin 100 IE/mL and then homogenized in 600 μL of lysis buffer in a homogenizer (Precellys 24; Bertin Technologies, France). Heart tissue was homogenized in 300 μL lysis buffer using the same technique. Western blotting was performed as previously described [32]. Samples containing an equal amount of protein were loaded on 4–12% polyacrylamide Bis-Tris criterion gel. Proteins were blotted onto a polyvinylidine difluoride (PVDF membrane, which was incubated with primary antibodies (TG 2 Ab-2) overnight and with secondary antibody (HRP-conjugated goat anti-mouse IgG) for 2 h. The membranes were developed using an enhanced chemiluminescence detection system (ECL-Plus; Amersham Biosciences, Buckinghamshire, UK) and expressed on X-ray films. Therefore, the blots were analyzed by optical density measurements (Image Quant TL; Amersham Biosciences) and the densitometric analysis of TG2 and N-[Latin Small Letter Open E](γ-glutamyl)lysine was given in relation to protein content and also normalized to panactin. Hypoxia regulates expression of glyceraldehyde 3-phosphate dehydrogenase [33], and therefore we chose to normalize protein expression to panactin, despite it was increased to the same level in all hypoxic groups compared to tissue from normoxic control animals.

Drugs and Solutions

Cystamine dihydrochloride (Sigma-Aldrich, St. Louis, MO, USA), sildenafil citrate (Pfizer Inc., Groton, CT, USA), Dormicum (midazolam 5 mg/mL; F. Hoffmann-La Roche AG, Basel, Switzerland), Hypnorm (fentanyl citrate 0.315 mg/mL, fluanisone 10 mg/mL; VetaPharma Ltd, Leeds, UK), Temgesic (buprenorphinum 0.3 mg/mL; Schering-Plough, Belgium), isofluran (4.5%, oxygen flow 2 L/min, FORENE®; Abbott Scandinavia AB, Solna, Sweden), and bovine albumin (Sigma-Aldrich) were used. PSS consisted of the following (mM): NaCl 119 mM, KCl 4.7 mM, MgSO₄ (7H₂O) 1.17 mM, NaHCO₃ 25 mM, KH₂PO₄ 1.18 mM, EDTA 0.026 mM), glucose 5.5 mM). Formalin 4% (Bie & Berntsen, Herlev, Denmark), papaverine (Sigma-Aldrich), U46619 (Sigma-Aldrich), acetylcholine (Sigma-Aldrich), lysis buffer (20 mM Tris/HCL, 5 mM EGTA, 150 mM NaCl, 20 mM glycorophosphate, 10 mM NaF, 1% triton X-100, 0.1% tween-20, adjusted to pH 7.5 and added 10 μL orthovanadate, 4 μL PMSF, 100 μL PIM per mL of lysis buffer), human smooth muscle actin (Clone 1A4) (Dako, Denmark), human Von Willebrand factor (Dako), TG 2 Ab-2 (clone TG100) (Thermo Scientific, CA, USA), and HRP-conjugated goat anti-mouse IgG (Zymed, San Francisco, CA, USA) were used.

Statistical Analysis

N indicates the number of animals used for each experiment. For the statistical analysis, the computer software GraphPad Prism 7.0 (San Diego, CA, USA) was used. The animal data were analyzed by use of one-way analysis of variance, and in case of significance, Bonferroni’s posttest was made to compare it to the vehicle-treated hypoxic group (H). The assumptions of the models were investigated by inspecting Q-Q plots, and data were logarithmically transformed when necessary in order to generate a Gaussian-distributed data set. In case of clustered data, the data were analyzed by use of a nonparametric rank test. All values are presented as mean ± SD and considered significant when p < 0.05.

Effect of Cystamine on TG2 Activity

First, we examined the presence of TG1-TG7 mRNA by qualitative RT-PCR and sequence analysis. In rat pulmonary arteries, we found expression of TG1 and TG2. Lung tissue samples showed expression of TG1, TG2, and TG5, while in rat liver there was expression of TG1, TG2, and TG5 (see online suppl. Fig. 1; for all online suppl. material, see http://www.karger.com/doi/10.1159/000515511).

To investigate the effect of cystamine on transamidase activity, a test based on the incorporation of a specific TG2 substrate (Biotin-pepT26) was performed, and incubation with the irreversible selective TG2 inhibitor, VA5, inhibited transamidase activity in the liver suggesting that the transamidase activity can be attributed to TG2 activity. The effect of increasing concentrations of cystamine was examined in rat liver and showed that relatively high cystamine concentrations (>30 μM) inhibited transamidase activity in the liver (Fig. 1a) and even higher concentrations were required in the lung (Fig. 1b). However, in this assay the homogenized tissue is treated with DTT leading to full tissue activation of transamidase activity. Therefore, we examined the effect of cystamine on transamidase activity in lung tissue by measuring 5-BP incorporation (Fig. 1c). Lung tissue from control and chronic hypoxic mice was used. The chronic hypoxic mice had right ventricular hypertrophy (Fig. 1d), and there was a tendency to increased transamidase activity (Fig. 1e). Cystamine (1.5 and 15 μM) significantly inhibited the 5-BP incorporation, while there was no effect of vehicle (Fig. 1f).

Open multimedia modal

Effect of Cystamine and Sildenafil in Chronic Hypoxic Rats

Table 1 shows the organ and body weights of the animals by the end of the study. Fourteen days after exposure to chronic hypoxia, the body weight was decreased by 14%, and the heart to body weight, and lung to body weight ratios were markedly increased by 16 and 48%, respectively, while liver weights were unaltered. Treatment with cystamine, sildenafil, or the combination of cystamine and sildenafil did not change the organ or body weights in the hypoxic rats (Table 1).

Open multimedia modal

The degree of right ventricular hypertrophy was determined as the weight of the right ventricular wall to the weight of RV/LV + S. RV hypertrophy increased by 72% in the chronic hypoxic vehicle-treated rats and was unaltered by treatment with cystamine or sildenafil but significantly decreased in the group treated with the combination of cystamine and sildenafil (Table 1).

Hemodynamic Measurements and Cardiac Hypertrophy

In vehicle-treated chronic hypoxic rats, RVSP markedly increased compared to normoxic rats (Fig. 2). Treatment with cystamine failed to prevent the increase in RVSP in chronic hypoxic rats. RVSP was decreased in chronic hypoxic rats treated with sildenafil compared to vehicle-treated hypoxic rats and decreased in the rats treated with the combination of sildenafil and cystamine compared to the chronic hypoxic rats treated with cystamine (Fig. 2a). In chronic hypoxic rats, the CO was not changed by the treatments with cystamine or sildenafil (Fig. 2b). The estimated PVR was increased in vehicle-treated hypoxic compared to normoxic animals. Within the hypoxic groups, PVR was unaltered in the cystamine-treated chronic hypoxic rats, while sildenafil lowered PVR compared to the cystamine-treated chronic hypoxic rats (Fig. 1c).

Open multimedia modal

In addition to RVSP being higher in the chronic hypoxic rats, RVEDP was increased in the chronic hypoxic rats with exception of the sildenafil-treated animals (Table 2). Chronic hypoxic-induced pulmonary hypertensive rats exhibited a marked elevation in RV dP/dt_max and dP/dt_min. The increased RV dP/dt_max indicates RV systolic dysfunction, whereas the increased RV dP/dt_min indicates RV diastolic dysfunction. In the hypoxic rats treated with cystamine, the diastolic function was significantly worse, while treatment with sildenafil improved the systolic heart function in the hypoxic rats. The heart rate was similar in the normoxic and vehicle-treated hypoxic animals, and there were no differences in systemic systolic and diastolic blood pressure or systemic vascular resistance in any of the 5 groups (Table 2).

Open multimedia modal

Effects of Cystamine and Sildenafil on Muscularization and Fibrosis in Pulmonary Arteries in Chronic Hypoxic Rats

As observed in previous studies [29], lung sections stained for Von Willebrand factor showed only positive immunoreaction in the endothelial cell layer (Fig. 3a), while α-actin stained the smooth muscle layer (Fig. 3b). In lung sections from the chronic hypoxic compared to normoxic rats, the number of nonmuscularized arterioles (20–80 μm in diameter) was significantly decreased, and this was also the case in hypoxic rats treated with cystamine (Fig. 3c). Only the combination of cystamine and sildenafil partially prevented the decrease in nonmuscularized arterioles (Fig. 3c). The number of partially muscularized arteries was not significantly changed in hypoxic versus normoxic lung sections (Fig. 3d), while the number of fully muscularized arteries was significantly increased (Fig. 3e). Monotherapy with cystamine failed to change the muscularization of the pulmonary arteries in chronic hypoxic rats, and only the combined treatment with cystamine and sildenafil significantly reduced the muscularization of the pulmonary arterioles in chronic hypoxic rats (Fig. 3e).

Open multimedia modal

Picrosirius red staining was used to evaluate lung fibrosis. Examining the sections in polarized light revealed that the staining tended to be decreased in lung sections of all drug-treated rats (Fig. 4a–g), but the decrease was only significant in lung sections of cystamine-treated hypoxic rats (Fig. 4f).

Open multimedia modal

Effect of Cystamine and Sildenafil on TG2 Immunoreaction in Heart and Lung Tissue

Expression of TG2 in sections was visualized in the RV and lungs by immunostaining showing that TG2 is localized in the cardiomyocytes of the RV (Fig. 5a, b), and in the lungs TG2 was localized to the endothelium and the smooth muscle layer of the arteries (Fig. 5c, d). As reported previously, immunoblotting for human TG2 full-length protein gave a band at 82 kDa [34]. The content of TG2 evaluated by immunoblotting was increased several hundred percent in the RV by hypoxia [17] (Fig. 5e). Although less pronounced, TG2 expression was also significantly increased in the lung (p < 0.05, Students t test) from hypoxic animals (Fig. 5f). In the RV of the chronic hypoxic rats, the loading control panactin was also markedly increased compared to the RV from normoxic rats, but there was no difference in panactin among the hypoxic groups (online suppl. Fig. 2). Expressing the TG2 immunoreaction relative to panactin showed treatment with cystamine, sildenafil, or the combination did not decrease TG2 immunoreaction in the RV or lung of the chronic hypoxic rats (Fig. 5g, h).

Open multimedia modal

Effect of Cystamine and Sildenafil on the RV in Pulmonary Hypertension

In previous microarray studies confirmed by qPCR and immunoblotting of the RV, we observed upregulation of TG2 in relation to the increase in RVSP in chronic hypoxic rats [17]. In agreement with these studies and those by others [15], TG2 protein was expressed in the myocardium and was markedly increased in the RV of hypoxic rats in the present study. Overexpression of TG2 has been shown to result in left ventricular hypertrophy and interstitial fibrosis [13]. However, in the present study cystamine treatment by itself did not inhibit RV hypertrophy in the hypoxic rats. Moreover, endothelin-1 expression and plasma levels are increased by hypoxia [35^, 36], and both endothelin-1 and aldosterone have been shown to be associated with elevated TG2 expression and/or activity in cardiomyocytes [18]. Therefore, a likely explanation is that high-pressure load in the RV [17] and/or increased circulating levels of endothelin-1 [35] leads to upregulation of TG2 which may contribute to the adaptation rather than the development of right ventricular hypertrophy in the hypoxic rats.

Cystamine was reported to inhibit the development of right ventricular hypertrophy in chronic hypoxic rats and monocrotaline-induced pulmonary arterial hypertension [18^, 21]. In contrast, in the present study, treatment with cystamine failed to prevent the development of RV hypertrophy measured as RV/LV + S. One of the main differences is the drug administration. In previous studies, cystamine was either administrated twice daily by intraperitoneal administration [18] or by oral gavage [21], while in the present study cystamine was administered in mini-osmotic pumps. In contrast to intraperitoneal and oral administration yielding high plasma peak concentrations, administration by mini-osmotic pumps results in lower steady plasma concentrations. In previous studies in our laboratory, where the plasma concentration of cystamine was measured in rats treated with twice the dose and the same administration route as used in the present study, a concentration of 1–2 μM cystamine was detected [10].

In platform assays, 73 μM cystamine inhibits TG2; cystamine concentrations above 100–200 μM may also inhibit TG1 and caspase-3, and in isolated cells up to 500 μM only reduced transamidase activity with 70% [37]. In agreement with these observations, we observed that millimolar cystamine concentrations were required to inhibit the transamidase activity in liver and lung tissue from hypoxic rats, but in this assay TG2 is maximally activated by DTT. Another approach used to measure in vivo activity of TG2 is the incorporation of BP, and this activity is inhibited in the presence of a competitive irreversible inhibitor of the open state of the enzyme [20^, 38]. Incubation of lung sections from hypoxic mice revealed in the present study that 1.5–15 μM cystamine significantly inhibits 5-BP incorporation. Based on previous results of plasma-cystamine concentrations [9], these results support that cystamine in vivo may inhibit TG2, while it is unlikely with the dose applied in the present study that it will inhibit other potential targets. Therefore, the relation of plasma cystamine to pharmacodynamic effect on right ventricular hypertrophy suggest that TG2 inhibition is insufficient to prevent RV hypertrophy in chronic hypoxic rats. Thus, inhibition of other signal pathways by higher concentrations of cystamine may contribute in inhibition of RV hypertrophy in studies where cystamine was administered by intraperitoneal administration or oral gavage [18^, 21]. In support of our observations, a selective irreversible TG2 and TG1 inhibitor ERW1041E failed to prevent the development of RV hypertrophy in mice with pulmonary hypertension induced by a combination of hypoxia plus SU-5416 [20].

In the same mouse pulmonary hypertension model, however, ERW104E significantly reduced RVSP [20]. In contrast, in the present study, cystamine failed to alter RVSP. However, in agreement with our previous studies showing that sildenafil reduces RVSP in hypoxic rats [29^, 39], sildenafil alone or in combination with cystamine also significantly reduced RVSP in the hypoxic rats in the present study. Moreover, systolic ventricular function, measured as dP/dt_max, decreased by treatment with sildenafil compared to the vehicle-treated hypoxic rats, whereas cystamine failed to improve systolic ventricular function, and actually appeared to accelerate the deterioration in RV diastolic function, measured as an increase in dP/dt_min, compared to vehicle-treated rats. In small systemic arteries, cystamine was found to inhibit both inward and outward remodeling [9]. Moreover, TG2 null mice had increased mortality and enhanced ventricular dilation following left ventricular pressure overload suggesting maladaptation to high pressure [40]. Although further investigation will be required, it cannot be excluded that cystamine actually inhibits the adaptation of the RV to high RVSP and RVEDP. This is supported in the present study by the observation that RV weight expressed relative to RVSP is significantly less pronounced in cystamine-treated animals compared to hypoxic controls (online suppl. Fig. 3). Although other approaches are required to confirm our observations, our results may suggest that cystamine delays the remodeling process in the RV in response to an increased afterload.

Effects of Cystamine and Sildenafil on Muscularization in Pulmonary Arteries from Chronic Hypoxic Rats

TG2 is expressed in the lung endothelium and epithelium in mice [41]. In the present study, we found expression of TG1 and TG2 in small pulmonary arteries, while in the lung TG1, TG2, and TG5 expression was evident. Using the same primers, we have previously detected TG4 expression, confirmed by sequencing, in the rat ear and testes [34], but this variant was not detected in the peripheral rat lung in the present study. TG4 is expressed in larger arteries [34], and therefore we cannot exclude that it will also be expressed in samples including proximal large pulmonary arteries. In agreement with previous studies [19], we found that chronic hypoxia upregulates TG2 expression in the lung, and the expression remained high in hypoxic rats treated with cystamine and sildenafil.

Chronic hypoxia markedly upregulates the PVR due to increased vascular tone, hematocrite, muscularization, fibrosis [42], and rarefaction often reflected by pulmonary arterial occlusive lesions [43]. Sildenafil is a potent vasodilator in the pulmonary circulation [39] and decreased the PVR. In contrast to sildenafil, cystamine is a less potent vasodilator in mesenteric small arteries [34^, 44], and very high concentrations (>100 μM) of cystamine were required to relax pulmonary small arteries (online suppl. Fig. 4), and in agreement with these findings cystamine treatment did not decrease the PVR in the chronic hypoxic rats.

Cystamine was found to reduce arterial wall thickness in pulmonary arteries from chronic hypoxic rats [18], and recently TG2 was found to play an important role in stimulation of pulmonary vascular smooth muscle proliferation when lowering oxygen from 21 to 3% in cell cultures [19]. We found that chronic hypoxia increased the muscularization of the pulmonary arteries, while the treatment with cystamine alone failed to reduce pulmonary arterial muscularization. In agreement with previous studies [29^, 45], treatment with sildenafil decreased the number of fully muscularized arterioles in the lung of chronic hypoxic rats. However, the most remarkable effect of the treatments was that the number of nonmuscularized arterioles decreased in vehicle and cystamine-treated animals, and this effect was prevented in rats treated with sildenafil. These findings agree with our recent observations that the pulmonary vasodilator, levosimendan, is also associated with less pulmonary arterial occlusive lesions [43]. Thus, our findings suggest that in contrast to cystamine, the vasodilator effect of sildenafil maintains an open lumen in pulmonary small arteries in chronic hypoxic rats, and this contributes to the lowering of the PVR and pressure.

Collagen accumulates in lung tissue and pulmonary arteries, and it is important for the stiffening in the large proximal pulmonary arteries [46^-48], and in distal pulmonary arteries it plays an important role for right ventricular afterload in chronic hypoxic rats [49^, 50]. In the hypoxic rats, fibrosis measured by picrosirius red staining was increased in the lung tissue and was decreased in the lung tissue of rats treated with cystamine. The mechanism involved in cystamine-induced decrease in collagen content is unclear. However, increased TG2 activity has also been suggested to lead to platelet-derived growth factor and transforming growth factor activation [51^, 52] and hence induction of collagen formation. There was no additive effect of cystamine and sildenafil treatment on lung fibrosis suggesting that cystamine treatment would not provide additional effect compared to monotherapy with sildenafil.

Limitations

Cystamine is considered a pan transglutaminase inhibitor as half-maximal inhibition (IC₅₀) for human TG2 and TG1 is, respectively, 73 and 264 μM, and above 100 μM for inhibition of caspase-3 [37]. Therefore, it could perhaps have been an advantage to the test the hypothesis of the present study in TG2 knockout mice. However, the global TG2 knockout mouse phenotype resembles that of the maturity onset of diabetes [53]. Due to the less pronounced development of pulmonary hypertension in chronic hypoxic mice compared to rats [28^, 39^, 54], the presence of diabetes may have larger influence on the pulmonary circulation and heart than pulmonary hypertension. Therefore, as in our previous studies, the hypobaric hypoxic rat was chosen and found to have increased pulmonary pressure, PVR, muscularization of the pulmonary arteries, and RV hypertrophy [17^, 39^, 54].