Pulmonary hypertension (PH) is characterized by sustained vasoconstriction, with subsequent extracellular matrix (ECM) production and smooth muscle cell (SMC) proliferation. Changes in the ECM can modulate vasoreactivity and SMC contraction. Galectin-1 (Gal-1) is a hypoxia-inducible β-galactoside-binding lectin produced by vascular, interstitial, epithelial, and immune cells. Gal-1 regulates SMC differentiation, proliferation, and apoptosis via interactions with the ECM, as well as immune system function, and, therefore, likely plays a role in the pathogenesis of PH. We investigated the effects of Gal-1 during hypoxic PH by quantifying 1) Gal-1 expression in response to hypoxia in vitro and in vivo and 2) the effect of Gal-1 gene deletion on the magnitude of the PH response to chronic hypoxia in vivo. By constructing and screening a subtractive library, we found that acute hypoxia increases expression of Gal-1 mRNA in isolated pulmonary mesenchymal cells. In wild-type (WT) mice, Gal-1 immunoreactivity increased after 6 wk of hypoxia. Increased expression of Gal-1 protein was confirmed by quantitative Western analysis. Gal-1 knockout (Gal-1−/−) mice showed a decreased PH response, as measured by right ventricular pressure and the ratio of right ventricular to left ventricular + septum wet weight compared with their WT counterparts. However, the number and degree of muscularized vessels increased similarly in WT and Gal-1−/− mice. In response to chronic hypoxia, the decrease in factor 8-positive microvessel density was similar in both groups. Vasoreactivity of WT and Gal-1−/− mice was tested in vivo and with use of isolated perfused lungs exposed to acute hypoxia. Acute hypoxia caused a significant increase in RV pressure in wild-type and Gal-1−/− mice; however, the response of the Gal-1−/− mice was greater. These results suggest that Gal-1 influences the contractile response to hypoxia and subsequent remodeling during hypoxia-induced PH, which influences disease progression.
- vascular remodeling
- right ventricular failure
- immune response
- extracellular matrix
chronic pulmonary arterial hypertension (PH) can lead to progressive right ventricular (RV) failure and death. The basis for the pulmonary vascular remodeling is smooth muscle cell (SMC) contraction, increased pressure, and hypoxia-induced injury (26, 45, 54, 57). Endothelial cell injury is followed by the migration of medial SMC into the intimal vessel layer, subsequent SMC proliferation, and matrix deposition (1, 2, 15, 19, 26, 33, 34, 39, 43, 45, 54, 57). Additional evidence suggests that changes in composition of the extracellular matrix (ECM) contribute to changes in SMC phenotype, contractility, and promotion of pulmonary arterial (PA) SMC remodeling (2, 33, 34, 39, 41, 43). Galectin-1 (Gal-1) is produced by vascular, interstitial, epithelial, and immune cells (2, 34, 41, 43). Gal-1 regulates SMC differentiation and proliferation and cell-cell and cell-ECM interactions. It therefore likely plays a role in the pathogenesis of PH.
Gal-1 is a prototypical member of the galectin family of lectins, a highly conserved family of soluble proteins that specifically bins β-galactoside derivatives, such as oligosaccharide ligands of glycoproteins, on cell membranes as well as other ECM components (3, 4, 8–10, 14, 21, 22, 48–50, 61). Because Gal-1 is a divalent 14.5-kDa protein, it can form lattice-like complexes with receptor partners participating in cell-matrix recognition (9, 48–50, 55). Extracellularly Gal-1 regulates cell-cell and cell-matrix interactions, the immune response, apoptosis, and neoplastic transformation. Intracellular functions include cell cycle regulation, RNA splicing, and transcriptional regulation (8, 37, 46, 47, 52, 55).
Despite evidence for a variety of functions for Gal-1 in vitro and in vivo, Gal-1 is not required for development or survival, and null mice are viable (50). During embryogenesis, Gal-1 expression in the lung is restricted to mesodermal cell types and is absent in epithelial cells (48–50). However, in adults, it is produced by a majority of tissues (8, 48–50). In the vasculature, Gal-1 is produced by SMC and endothelial cells and is regulated by proinflammatory mediators and cell activation, including changes in vascular cells that allow them to participate in the immune response (10, 30, 51). In vitro, Gal-1 regulates SMC attachment to laminin and fibronectin in a dose-dependent manner, which subsequently regulates proliferation, migration, adhesion, and spreading of SMCs by means of β1-integrins and focal adhesion kinase signaling (8, 9, 10, 23, 37, 44). Increases in laminin and fibronectin have been colocalized to remodeled vasculature in animal models of PH (27, 34, 59). Decreasing levels of laminin parallel a reduction in the number of muscularized arterioles and PA pressure (27, 34, 43), illustrating the regulatory role for the ECM in SMC remodeling characteristic of PH.
In the present study, we examined the role of Gal-1 in the pathogenesis of hypoxia-induced PH. We found increased expression of Gal-1 mRNA in isolated pulmonary mesenchymal cells in response to hypoxia in vitro. This in vitro finding correlated with increased tissue expression and localization of Gal-1 protein in SV129 wild-type mice after 6 wk of hypoxia. Exposure of Gal-1-knockout (Gal-1−/−) mice to chronic hypoxia resulted in decreased RV pressure (RVP) and RV-to-left ventricle (LV) + septum wet weight ratio (RV/LV + S) compared with SV129 wild-type mice. In response to chronic hypoxia, increases in muscularization of microvessels, decreases in microvessel density, and fixed RVP in response to acute hypoxia were similar in both groups. Our data demonstrate that, at baseline, mice have similar RVP. In response to acute hypoxia, Gal-1−/− mice have increased vasoreactivity compared with wild-type mice. However, overall Gal-1−/− mice exhibit a decreased physiological response to chronic hypoxia-induced PH, suggesting that this difference is likely in part due to an altered hypoxia-sensing mechanism and a subsequent ECM remodeling-dependent mechanism.
MATERIALS AND METHODS
All procedures and protocols were approved by Institutional Animal Care and Use Committees at the University of Colorado. At 131 days of gestation, Columbia-Rambouillet sheep were euthanized with a lethal dose of pentobarbital sodium, and the lungs were harvested for pulmonary mesenchymal cells. Wild-type and Gal-1−/− mice bred into the same genetic background (SV129) were obtained from Francoise Poirier and bred in our animal facility for three generations.
In vitro model to screen for hypoxia-induced gene expression.
Initially, we screened the easily accessible, well-characterized population of ovine pulmonary mesenchymal cells (38). Briefly, at 131 days of gestation, fetal lambs were isolated from pregnant dams, and fetal lungs were harvested. Explant tissue from the distal lung was sectioned and placed in medium on plastic culture dishes for 48 h under normal culture conditions (α-MEM, 20% fetal calf serum, penicillin-streptomycin; Invitrogen Life Technologies, Carlsbad, CA). After 48 h the tissue was removed from the plates and additional medium was added. Mesenchymal cells that had migrated from the explant cultures were visible at this time and required an additional 2 wk of culture. Cells were characterized by immunocytochemistry for smooth muscle α-actin and platelet-derived growth factor receptor-β (mesenchymal markers), the absence of desmin (a pericyte marker), pan cytokeratin (an epithelial marker; DAKO clone AE1/AE3), and factor 8 (an endothelial marker) (38).
Ovine mesenchymal cells were cultured to 80% confluence and exposed to 21% normoxic or 3% relative hypoxic conditions for 48 h under otherwise standard tissue culture conditions (see above). Cells were harvested, and RNA was isolated using the RNeasy kit (Qiagen). RNA was quantitated, and first-strand synthesis was performed using 2 μg of total RNA for each condition following the recommended protocol for the Super SMART PCR cDNA synthesis kit (Clontech). Second-strand synthesis was carried out to obtain double-stranded cDNA, and the remaining subtraction was performed using the PCR-Select subtraction kit (Clontech). The cDNA sample corresponding to 21% O2 (normoxia) was used as the driver, and that corresponding to 3% O2 was used as the tester. RNAs expressed at higher levels in the 3% O2 sample were selected for amplification and further analysis. Resulting cDNAs were cloned into the Topo-TA cloning vector (Clontech) and transformed into Escherichia coli, and colonies were grown on LB-amp-X-Gal plates. White colonies were procured with a sterile pipette tip and directly amplified by PCR, and products were sequenced.
In vivo mouse model of chronic hypoxia-induced PH.
Age-matched (8- to 12-wk-old) wild-type SV129 and Gal-1−/− mice were exposed to ambient pressure (630 mmHg barometric pressure, 5,260 ft) or hypobaric hypoxia (∼400 mmHg barometric pressure, 16,000 ft) for 6 wk. For Gal-1 measurements, lung tissue was obtained from 14- to 18-wk-old mice after 6 wk of exposure. Adult mice were euthanized by ketamine overdose.
Gal-1 protein and mRNA expression.
Frozen lung tissue samples randomly selected from normoxic and hypoxic mice were cut into small pieces, homogenized in lysis buffer, and centrifuged. Equal amounts (10 μg) of protein from the supernatant were loaded onto 16% Tris-glycine gels for electrophoresis. Protein was transferred to a polyvinylidene difluoride membrane, blocked with 5% nonfat milk, incubated with primary Gal-1 antibody (catalog no. S-14, lot no. I1703, Santa Cruz Biotechnology), and then incubated with secondary antibody. A chemiluminescence system (ECL Plus, Amersham International) was used to generate autoradiographs of the membranes, and densitometry was performed to quantify the signal that was obtained. Equal protein loading was confirmed by Coomassie blue staining of gels and β-actin (catalog no. A5441, Sigma) immunostaining on the same blot. Three independent analyses (n = 3) were performed, with replicates of one or two per group. Quantitative PCR was performed with total cellular RNA extracted using the RNeasy kit (Qiagen). RNA was quantitated using a NanoDrop spectrophotometer and reverse transcribed using Superscript II (Invitrogen) according to the recommended protocol. cDNA was diluted 1:5 and amplified by real-time PCR (Applied Biosystems) using primers directed to Gal-1 and SYBR green (Perkin-Elmer) as the detection probe. Amplicons were normalized to the housekeeping gene hypoxanthine phosphoribosyltransferase and plotted relative to expression in room air-exposed tissue. Analysis was performed using three individual mice per 1- or 6-wk treatment (n = 3 each for normoxia and hypoxia). Primer sequences were as follows: 5′-CGCCATGTAGTTGATGGCC-3′ and 5′-CTGCCAGACGGACATGAATTC-3′.
Hemodynamic phenotyping and vasoreactivity measurements.
At 14–18 wk of age, mice (20–32 g body wt) were exposed to normoxia or chronic hypoxia for 6 wk and then anesthetized with intraperitoneal injections of ketamine (200 mg/kg) and xylazine (10 mg/kg). All measurements were recorded within 2 h of removal of the animals from chronic hypoxia chambers. Mice were positioned supine on a heated operating table, and baseline measurements of RVP and cardiac output were recorded in room air. Subsequent to room air measurements, vasoreactivity was quantified by measurement of RVP in response to acute hypoxic exposure in baseline mice (previously exposed to room air or chronic hypoxia) breathing from a cannula delivering constant 10% O2 for 2 min (n = 3–5). RVP and cardiac output were directly measured and analyzed as previously described (62). Two independent experiments were pooled for the hemodynamic measurements. Independent mice were analyzed for each measurement: n = 6 wild-type and n = 6 Gal-1−/− (room air) and n = 9 wild-type and n = 11 Gal-1−/− (hypoxia). Hearts were then collected for analysis of RV hypertrophy.
Isolated perfused mouse lungs.
The isolated perfused mouse lung (IPML) model consisted of an intact heart-lung block perfused with modified salt solution, as previously described (35). Briefly, after anesthesia, mouse tracheas were intubated and lungs were ventilated with 21% O2-5% CO2-74% N2 at 60 breaths/min. After 20 min of equilibration during ventilation with 21% O2, vascular responses were measured. A baseline perfusion pressure was established, and the lung was ventilated with three challenges of severe hypoxia (0% O2) for 5 min followed by 5 min of normoxia. After the hypoxic response was measured, increasing concentrations of 5-hydroxytryptamine (5-HT, 1 and 10 mM; Sigma) followed by KCl (20–60 mM) were added to the perfusate at 12-min intervals. Ventilation was maintained at 21% O2. Change in PA pressure was recorded and expressed in mmHg (35).
Measurement of RV hypertrophy.
RV hypertrophy was assessed as an index of PH in 14- to 18-wk-old Gal-1−/− and wild-type mice. Hearts were removed and dissected to isolate the free wall of the RV from the LV + S. RV/LV + S was used as an index of RV hypertrophy generated as a result of hypoxia-induced PH.
Histological analysis, including Gal-1 lung qualitative localization and vessel-to-alveolar density ratio, was performed. After euthanasia, a thoracotomy was rapidly performed, and the thorax was exposed by dissection. PBS was injected into the RV to flush the lungs. After tracheostomy, a stainless steel syringe was inserted into the trachea and the lungs were inflated with 4% PBS-buffered formalin, as previously described (35). The lungs were subsequently removed, fixed overnight, and then placed in 70% ethanol. Paraffin-embedded 10-μm sections were stained with hematoxylin and eosin, pentachrome, factor 8, and smooth muscle α-actin. Gal-1 immunohistochemistry was also performed on normoxic and hypoxic sections from wild-type SV129 mice. Gal-1−/− tissue was used as a negative control for background reactivity (Gal-1; catalog no. S-14, lot no. I1703, Santa Cruz Biotechnology). Vessel density (n = 4) and muscularization (n = 3–5) of 20- to 200-μm vessels were quantitated using factor 8- or smooth muscle α-actin-stained slides. Alveoli were counted on pentachrome-stained slides (n = 4) by a blinded observer in 8–10 randomly selected high-power fields of distal lung. Fields including bronchial or conducting airways or main PA vessels were excluded. All vessels and alveoli in a ×200 high-power field were counted and measured. For vessel counting, five high-power fields per animal were used. High-power fields were used for measurements of muscularization. For each group, 30–35 total fields of view containing at least one actin-positive vascular structure were counted (6–8 per animal).
Data analysis and statistics.
Values are means ± SE. The JMP 5.0 statistical package (SAS, Cary, NC) was used for statistical analysis. One-way ANOVA followed by Student’s t-test (2 groups) was used for comparisons between groups. Fisher’s protected least significant difference test (>2 groups) was used to determine whether there were statistical differences between treatment groups at a significance level of 0.05.
Subtractive hybridization suggests that Gal-1 may be important in adaptive responses of murine lung to chronic hypoxia.
To identify specific genes differentially expressed in pulmonary mesenchyme in response to hypoxia, we performed subtractive hybridization. This allowed the identification of genes expressed at significantly greater levels in cells in response to 3% O2 (hypoxia) than in response to 21% O2 (normoxia). After creation of the subtractive library and PCR amplification of bacterial colonies, we selected colonies for screening based on criteria that included product size >400 bp and the amplification of a single band. Approximately 30% of the colonies had been transformed, with more than one plasmid representing multiple cDNAs. In addition, 25 potential amplicons were identified for further evaluation by sequence analysis. Of these 25 sequences, 5 were confirmed to be Gal-1. On the basis of these results, we that believed that Gal-1 may have functional significance in pulmonary vascular responses to hypoxia.
Gal-1 protein levels increase in hypoxia-induced PH.
To further evaluate physiologically relevant changes in Gal-1 expression in response to in vivo hypoxic exposure, we used murine models including wild-type SV129 mice and a Gal-1−/− transgenic model of chronic hypoxia-induced PH. After 6 wk of exposure to hypobaric hypoxia, Gal-1 expression was increased in hypoxic wild-type mice compared with normoxic animals (Fig. 1, A–D). Expression of Gal-1 was diffusely distributed throughout the lung interstitium and adjacent to the basement membrane of vessels and airways in normoxia- and hypoxia-exposed mice. However, the intensity of Gal-1 staining was increased in hypoxia-exposed mice. No immunoreactivity was detected in Gal-1−/− tissue (Fig. 1, E and F).
Western blot analysis was performed to detect and quantitate Gal-1 levels. The Gal-1 bands were visualized at 14.3 and 30 kDa in the monomeric (P < 0.13) and the in vivo noncovalent dimeric (P < 0.05) forms, respectively. Gal-1 dimeric (P < 0.05) band intensity was greater in the hypoxia- than in the normoxia-exposed animals (Fig. 2, C and D). At 1 or 6 wk after exposure to room air or hypoxia, Gal-1 mRNA levels were measured using quantitative PCR (Fig. 3). At 1 wk there was a decrease in Gal-1 mRNA compared with control (P = 0.05), and at 6 wk no significant difference was evident.
Gal-1−/− mice develop attenuated PH in response to chronic hypoxia relative to their wild-type counterparts.
To further evaluate the importance of Gal-1 in the development of PH, cardiovascular hemodynamic parameters were measured in Gal-1−/− and wild-type mice after 6 wk of chronic exposure to hypobaric hypoxia. RV systolic pressures (RVP) were measured as an indication of PA pressure. RVP increased in wild-type and Gal-1−/− mice after 6 wk of hypoxic exposure, with a greater increase in the wild-type animals (Fig. 4A). In response to hypoxia, RVP increased from 32.7 to 43 mmHg (P < 0.0003) in the wild-type animals, characteristic of PH; the increase was less extreme (from 33.0 to 37.5 mmHg, P < 0.026) in the Gal-1−/− mice. The resulting RVP values were significantly different between the two groups after hypoxia (P < 0.0061). As an indication of RV hypertrophy, RV/LV + S was measured in wild-type and Gal-1−/− mice (Fig. 4B). RV hypertrophy was confirmed in the wild-type animals (P < 0.005) by an increase in RV/LV + S after hypoxia (Fig. 4B). No significant increase in RV/LV + S (P < 0.07) was detected in Gal-1−/− mice. RV/LV + S was significantly lower in Gal-1−/− than in wild-type mice (P < 0.05). There was no significant difference in body weight or heart rate between the groups (Table 1). There was no significant difference in cardiac output in the normoxia-exposed groups (P < 0.09). In response to hypoxia, cardiac output did not change significantly in Gal-1−/− mice (P < 0.26) but decreased from 6,743.5 to 2,772.1 μl/min (P < 0.0005) in wild-type animals (Table 1). A decrease in cardiac output in wild-type mice has previously been reported in other rodent models of PH (28).
Wild-type and Gal-1−/− mice exhibit increased arterial muscularization and microvessel loss after chronic hypoxia.
To determine whether Gal-1 expression after chronic hypoxia affects SMC infiltration of the microvasculature, we quantified numbers of totally and partially muscularized vessels vs. the number of fully muscularized vessels on the basis of smooth muscle α-actin immunoreactivity (Figs. 5 and 6). In wild-type and Gal-1−/− animals, the number of total (P < 0.0001 and P < 0.0001), partially (P < 0.0007 and P < 0.02), and fully muscularized vessels (P < 0.009 and P < 0.0019) increased. However, the increase in partially muscularized vessels was greater in wild-type than in Gal-1−/− mice (P < 0.04). The number of total and fully muscularized vessels was not significantly different (P < 0.4 and P < 0.2). There were no differences between the normoxia-exposed groups (P < 0.9, P < 0.4, and P < 0.5).
Pulmonary alveoli and pulmonary vessels were counted to measure changes in alveolar and vessel density (Fig. 7). There was no change in alveolar density in wild-type (P < 0.36) or Gal-1−/− (P < 0.33) mice after hypoxia (Fig. 7A). Analysis of microvessel density in lung tissue from wild-type and Gal-1−/− mice illustrated a decrease in <200-μm-diameter factor 8-positive microvessels (P < 0.041 and P < 0.00045), with corresponding decreases in vessel-to-alveolar density ratio after hypoxic exposure (Fig. 7, B and G). There was no significant difference between the normoxic groups (P < 0.29). Taken together, these results suggest that Gal-1 expression is not required for the increased muscularization of vasculature and loss of microvessels associated with the exacerbation of pathology in the murine chronic hypobaric hypoxia model of PH.
Wild-type and Gal-1−/− mice exhibit similar vasoreactivity in room air and after chronic hypoxia.
Vasoreactivity was tested in vivo after exposure to room air and chronic hypoxia by quantitation of RVP in response to acute hypoxia, i.e., 10% O2 for 2 min (Fig. 8). RVP at baseline was similar in the room air-exposed groups (P < 0.88), which responded to acute hypoxia with increases in RVP above baseline (P < 0.07 and P < 0.0009). However, RVP was significantly higher in the Gal-1−/− than the wild-type mice (P < 0.004). In response to chronic hypoxia, baseline RVP was significantly different between wild-type and Gal-1−/− groups (Figs. 5A and 8A; P < 0.006). When exposed to acute hypoxia, the wild-type and Gal-1−/− mice exhibited fixed vascular beds, demonstrated by the absence of additional increases in RVP (P < 0.7 and P < 0.23). After chronic hyperoxia and acute hypoxia, there was a difference in RVP between wild-type and Gal-1−/− (P < 0.005) groups, similar to the baseline differences.
To further evaluate the effect of Gal-1 deletion on pulmonary vasoreactivity in response to acute hypoxia, we used a closed system, the IPML model (Fig. 8C). Both groups responded to acute hypoxic stimuli with increased PA pressure; however, the response was dramatically greater in Gal-1−/− than in wild-type lungs (2.14 ± 0.21 mmHg, P < 0.0025). In addition to acute hypoxic stimuli, wild-type and Gal-1−/− lungs were exposed to vasoconstrictors such as 5-HT and KCl (Fig. 8D). Maximal contraction was achieved after administration of these agents, and no significant difference in PA pressure was detected between the groups.
Changes in ECM can modulate vasoreactivity and SMC contraction. Our data demonstrate a previously undescribed mechanism by which Gal-1 modulates pulmonary vasoreactivity in response to hypoxia and attenuates the physiological responses characteristic of PH. Gal-1 message expression increased in pulmonary mesenchymal cells in response to hypoxia in vitro. This observation supports in vivo findings of increased protein after 6 wk of exposure of SV129 wild-type mice to chronic hypoxia. We then exposed the Gal-1−/− transgenic mice to chronic hypoxia. In response to hypoxia, the RVP and RV hypertrophy increased more significantly in the wild-type animals and were attenuated in the Gal-1−/− mice. Increases in microvessel muscularization and decreases in factor 8-positive microvessel density were similar in both groups. The vasoreactivity of both room air-exposed groups was similar at baseline and increased to significantly different peak levels in response to an acute hypoxic stimulus. In contrast, RVP was fixed in animals exposed to chronic hypoxia. Our data demonstrate similar vasoreactivity in Gal-1−/− and wild-type mice, whereas Gal-1−/− mice exhibit a decreased physiological response to PH induced by chronic hypoxia.
The hypoxic regulation of Gal-1 at mRNA and protein levels has been demonstrated in tumor biology and used as a prognostic marker of malignancy (32). Because Gal-1 is a leaderless protein secreted by vesiculation, it can accumulate intracellularly and be induced posttranscriptionally (10, 32, 53). Our data demonstrate that in vitro Gal-1 mRNA was regulated in lung mesenchyme by acute hypoxia at the transcriptional level; however, in vivo the changes in protein were more significant. Because Gal-1 has intra- and extracellular functions, the quantifiable change in protein production has a sustainable impact on vascular remodeling associated with hypoxia-induced PH.
Interestingly, a less substantial increase in RVP after chronic hypoxic exposure was exhibited in Gal-1−/− than in wild-type SV129 mice. This difference in RVP between wild-type and Gal-1−/− mice may be partially attributed to decreased pulmonary vascular resistance and afterload. Interestingly, our findings illustrated similar increases in muscularized microvessels and microvessel loss in response to chronic hypoxia in wild-type and Gal-1−/− mice. To address these paradoxical results, we assessed RVP in vivo as a measure of vasoreactivity in response to acute hypoxic stimuli. After acute hypoxic stimuli, the vasoreactivity of both room air-exposed groups significantly increased. However, in contrast to the differences in RVP after chronic hypoxia, peak RVP was lower in wild-type than in Gal-1−/− mice: 38 ± 0.33 vs. 40 ± 0.31 mmHg. After chronic hypoxia, the wild-type and Gal-1−/− mice exhibited no increase in RVP in response to acute hypoxic stimuli, suggesting that the vascular beds had been remodeled and fixed and that the differences in RVP at this point were not mediated by differences in vascular tone.
As a comparative study, we used a closed IPML system to further evaluate the vasoreactivity of wild-type and Gal-1−/− lungs in response to acute hypoxia as well as vasoconstrictive agents. The wild-type and Gal-1−/− lungs responded to acute hypoxia in this system; however, the response was twofold greater in Gal-1−/− than in wild-type lungs. Maximal constriction was achieved with no significant differences between the groups in response to 5-HT or KCl. These findings illustrate differences between the groups in the SMC contractile response to hypoxia as opposed to vascular tone in wild-type and Gal-1−/− mice before exposure to chronic hypoxia. Therefore, the significant differences in pulmonary vascular resistance after chronic hypoxia may be attributed to varying susceptibility to chronic hypoxia as well as adaptive changes between wild-type and Gal-1−/− mice. The adaptive changes may be mediated by subtle differences in SMC phenotype and contractile responses, ECM deposition, and vascular remodeling. Subtle differences in muscle cell phenotypes in Gal-1−/− mice have previously been described in skeletal muscle during development and in response to injury (9, 10, 18, 61). These differences include cell size and increased accumulation of ECM, which were suspected to alter muscle cell function and regulate tissue regeneration (61). The initial diameter values for lung SMC and medial thickening in the PA also determine the variability of physiological responses to hypoxia-induced PH across species (56) and may also play a role in the physiological responses of wild-type and Gal-1−/− groups.
Increased RVP or PA pressure affects SMC contraction, which results in ECM production (14, 33, 37, 39, 55, 57–59) and subsequently influences SMC differentiation and proliferation (3, 4). Changes in the ECM of hypertensive vessels can modify the contractile responses of SMC (33). SMC phenotypes in large vessels, such as the PA, consist of heterogenous subpopulations, which may exhibit differential proliferative and matrix-producing responses to the stimuli associated with hypoxia-induced PH (15). In the lung, vascular remodeling in the Gal-1−/− mice may be distinct from that in the wild-type mice in terms of composition. One characteristic of PH is increased deposition of laminin and fibronectin in the subendothelial and medial basement membrane of the pulmonary vasculature (34, 41, 42, 57, 59). Gal-1 regulates SMC interaction with fibronectin and negatively regulates SMC interaction with laminin (9, 10, 44), affecting cell migration, spreading, and proliferation (44). Gal-1 also regulates interstitial fibroblast differentiation to SMC (17). Furthermore, alterations of the ECM have been correlated with changes in SMC phenotype and both SMC hyperplasia and hypertrophy during PH, as well as in animal models of PH that have been allowed to recover (30, 34, 59). After the animals recover from PH to a baseline state, laminin and fibronectin levels return to those of untreated animals (59). These studies illustrate a potential role for Gal-1 in modulating the interaction of SMC with ECM components regulating SMC phenotypes and remodeling associated with PH. Functional organization of laminin in the ECM by Gal-1 may also be necessary for SMC responses characteristic of chronic hypoxia-induced PH.
In addition to these studies in which Gal-1 protein is increased in a hypoxia-induced model of PH, Gal-1 has also been characterized as a target of monocrotaline alkylation in rodent models of chemically induced PH, which may influence inflammatory responses to the affected areas (31). Monocrotaline damages the endothelium in arterioles and remote pulmonary vasculature (7, 12, 26, 27, 31) and, subsequently, results in a dramatic immune response and vascular remodeling (12, 19, 54, 57). Gal-1 is capable of attenuating acute and chronic inflammation (5, 6, 30) and functions to regulate immune cell trafficking, Th1 vs. Th2 responses, and apoptosis (5, 6, 11, 20, 30, 32, 33, 40, 51). These studies suggest that, in addition to regulating SMC interactions with the ECM and the response to hypoxia, Gal-1 may also change vascular ECM remodeling by affecting the host immune response, recruitment of monocytes, and/or cytokine production in chronic hypoxia-mediated PH (11).
Chronic hypoxia did not affect alveolar density in wild-type or Gal-1−/− mice. The normal epithelial architecture and alveolar density in the Gal-1−/− mice were likely due to functional redundancy within the lectin family (13, 24, 60). However, microvessel density was decreased in wild-type and Gal-1−/− mice after exposure to chronic hypoxia. Recent studies challenge the hypothesis that decreases in microvessel density result in increased vascular resistance in PH (54). Although a small change in vessel radius can be attributed to increased pulmonary vascular resistance, the effect of microvessel loss is mathematically less significant than the luminal narrowing or obliteration of a large PA, because the resistance is much greater in the main PA than in the microvessel (39). Also, the PA is characteristically the site of significant vascular remodeling in murine models of PH (29, 39). Alternatively, although controversial, loss of microvessels in the lungs, given their extensive surface area, could contribute significantly to increases in the overall pulmonary resistance. Remodeling in Gal-1−/− mice may still be distinct from that in wild-type mice but not necessarily result in luminal narrowing and increased vascular resistance, as described in chronically infected rat lungs and chronic obstructive pulmonary disease (54).
We did not address the potential for developmental adaptations to result in differences in vascular remodeling or vessel compliance. More detailed characterization of lung morphogenesis in the Gal-1−/− mice, as well as strain testing of the vasculature, is necessary. Because of the complexity of the tissue systems under investigation, these studies do not determine the specific cellular mechanisms of PH disease. Gal-1 may function in regulating SMC differentiation from interstitial fibroblasts (9, 17, 18). SMC contact with the ECM, contractility, migration, and proliferation could influence these processes. In vitro studies using isolated cells and titrated matrix-Gal-1 components are warranted to clarify this issue.
We have demonstrated that Gal-1−/− mice had a decreased physiological response to chronic hypoxia-induced PH. In the absence of Gal-1, SMC contraction and vascular remodeling occur. Gal-1 may therefore regulate the response of SMC to hypoxia, vasoreactivity, and subsequent associated vascular remodeling necessary for increased vascular resistance. Gal-1 affects cell adhesion to the ECM, cell cycle regulation, regulation of the immune response, RNA splicing, and transcriptional regulation (2, 7, 9, 13, 16, 23, 25, 48–50, 55). All these processes likely contribute to PH, specifically at the level of SMC-mediated vascular remodeling. Understanding how Gal-1 contributes to these processes is critical to understanding the underlying mechanisms of vascular remodeling and resistance in PH.
This work was funded in part by American Heart Association Grant SDG-0335052N, the American Physiological Society Giles Filley Award, and National Institute of Diabetes and Digestive and Kidney Diseases Grant 5 T32 DK-007496 (to S. M. Majka), grants from Ligue contre le Cancer, Comité de Paris, and ARC Foundations (to F. Poirier), National Heart, Lung, and Blood Institute Grants R01 HL-71596-01A1 and K08-HL-74512 (to J. West and M. Saavedra), and American Heart Association Grant 0575003N (to C. Ivester).
We thank Drs. N. Chessler, K. Fagen, and I. McMurtry for critical review of the manuscript and K. Fox for technical assistance.
↵* Both authors contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society