Lysophosphatidic acid receptors 1 and 2 play roles in regulation of vascular injury responses, but not blood pressure

Mice

All procedures conformed to the recommendations of “Guide for the Care and Use of Laboratory Animals” (Department of Health, Education, and Welfare publication number NIH 78-23, 1996) and were approved by the Institutional Animal Care and Use Committee. The production and initial characterization of mice-deficient in LPA receptors 1 and 2 has previously been described ^{27, 28}. The mice were backcrossed for >10 generations to the BalbC background. Mice were housed in cages with HEPA-filtered air in rooms on 12-h light cycles and fed Purina 5058 rodent chow ad libitum. Systolic blood pressure and heart rate were measured for five consecutive days in conscious mice using the Blood Pressure Analysis tail cuff system (Hateras Systems, Apex, NC) daily after training for one week. Mean intraartrial pressure was measured by placement of a 1.4 Fr Millar catheter in the carotid artery of isoflurane-anesthetized mice.

Vascular Injury

At various intervals after carotid surgery^{29, 30}, 5 mm of aorta proximal to the suture were removed and processed for RNA analysis by qualitative PCR (qPCR) or analysis of protein markers by immunoblotting. Neointimal formation along the length of the vessel was assessed at four weeks after surgery using computer assisted morphometry as has been previously described ³⁰. Digital images were taken with a high performance digital camera (resolution 3840 × 3072 pixels) attached to a Nikon 80i microscope with a 10× (NA = 0.3) or 20× (NA = 0.5) objective and analyzed with Metamorph software. Femoral artery denudation injury was performed and analyzed at four weeks as previously described ^{31, 32}.

Isolation of SMCs

Mouse aortic SMCs were obtained from thoracic aortas by removing the adventitia and endothelium by digestion with collagenase type II (Worthington; 175 units/ml). The media were further digested in solution containing collagenase type II (175 units/ml) and elastase (Sigma; 0.5 mg/ml), which yielded ≈100,000 cells per aorta. Cells were grown in DMEM containing 0.5 ng/ml EGF, 5 μg/ml insulin, 2 ng/ml bFGF, 10% FBS, 100 units/ml penicillin, and 100 μg/ml streptomycin, and incubated at 37°C with 5% CO₂/95% air. SMC lineage was confirmed by the presence of immunoreactivity for α-actin (Sigma) in >99% of the cells. Experiments involving SMCs were performed by using cells with a passage number ≤5.

Details of methods for SMC experiments can be found in Online Supplemental methods.

Statistical analysis

All results were expressed as mean ± standard error of the mean (SEM). In vitro studies were repeated a minimum of three times and results analyzed by student t-test or ANOVA. Statistical significance within strains was determined using ANOVA with multiple pair-wise comparisons. Statistical analysis was performed using Sigma-STAT software, version 3.5 (Systat Software, Inc.). A p-value of less than 0.05 was considered significant.

Upregulation of autotaxin (ATX) and LPA receptors following vascular injury

Following ligation injury, we observed a time-dependent increase in vessel-associated levels of ATX protein that was ∼2.5 fold after 1 - 3 days (p<0.05 at 1 day and 3 days versus uninjured; Figure 1A). ATX is a secreted lysophospholipase D responsible for generation of biologically-active extracellular LPA. Increasing plasma levels of ATX elevates circulating LPA levels ²², therefore, extracellular LPA production along the vessel wall likely increases following injury. In many systems, LPA elicits cellular effects via G-protein-coupled receptor signaling. We therefore assessed the level of gene expression of the five known LPA receptors (LPA 1 – 5) following injury and observed that expression of LPA1 and LPA2 increased following carotid injury by 2.1 fold and 6.2 fold, respectively (Figure 1B).

LPA1^-/-2^-/- mice are partially protected from development of intimal hyperplasia in vascular injury models while LPA1^-/- mice exhibit positive remodeling characterized by an enhanced neointimal response

Mice that are singly or doubly deficient in LPA1 and LPA2 (LPA1^-/-, LPA2^-/- and LPA1^-/-2^-/- mice) are viable and fertile ^{27, 28}, allowing us to use them to investigate a role for these LPA signaling receptors in the regulation of vascular SMC responses. Carotid ligation was performed in wild-type (WT) mice, LPA1^-/-, LPA2^-/- and LPA1^-/-2^-/- double knock-out (DKO) mice. Four weeks after carotid ligation, neointima developed along the length of the carotid in WT mice (Figure 2A and 2B). The extent of intimal area and the intima/media ratio was similar in mice lacking LPA2 (Figure 2B and 2D). In contrast LPA1^-/- mice displayed an ∼ 3-fold enhanced intimal area (Figure 2B), an increase in intima/media ratio (p <0.017 versus WT; Figure 2D), a larger lumen area (Figure 2E), and a resulting increase in the area within the external elastic lamina, consistent with positive remodeling (i.e maintenance of luminal area) in the presence of heightened neointimal formation. Unlike either the LPA2^-/- or LPA1^-/- mice which displayed normal and enhanced neointima, respectively, LPA1^-/-2^-/- DKO were partially protected from the development of intimal hyperplasia with an ∼4-fold lower intimal areas (p<0.05; Figure 2B) and intima/media ratios after carotid ligation (Figure 2D). Medial areas were similar in all the mice (Figure 2C).

To determine if these results were model specific, we denuded the endothelium of the mouse femoral artery to elicit the development of intimal hyperplasia and observed an enhanced injury response that was broadly similar to that observed in the carotid artery ligation model in LPA1^-/- mice (see Online Figure 1). Additional experiments were not performed with the other strains of LPA receptor deficient mice using this femoral artery model because of a high incidence of acute arterial thrombosis in WT Balb/C.

Vascular injury elicits a characteristic sequence of events that includes inflammatory cell recruitment and SMC dedifferentiation, proliferation, and migration. To investigate the mechanism(s) by which LPA receptors might regulate vascular injury responses, we measured markers for these events after ligation. Following injury, there was an increase in ERK activation, as measured by the ratio of phosphoERK to total ERK, and the increase in ERK activity returned to baseline by five days in LPA1^-/-2^-/- mice and by seven days in LPA1^-/- mice (Figure 3A). However, expression of the proliferation marker PCNA was upregulated to similar extents in WT, LPA1^-/-and LPA1^-/-2^-/- vessels (Figure 3B), suggesting that cell proliferation following injury was similar in all genotypes. Likewise, lack of LPA1 and LPA2 did not affect SMC dedifferentiation following injury because down-regulation of SMC markers occurred in all genotypes of mice examined (Figure 3B). Expression of IL-6 and the neutrophil and monocyte derived- inflammatory marker myeloperoxidase (MPO) increased following injury and was also unaffected by the absence of LPA1 and LPA2 (Figure 3C). As detected by immunoblotting, vessel-associated CD68, MPO, and P-selectin, a marker of platelet activation, increased following injury in all genotypes (see Online Figure 2). Taken together, these results indicate that vascular inflammation, SMC dedifferentiation, and cell proliferation are similar in WT, LPA1^-/-and LPA1^-/-2^-/- arteries following injury. Since these markers of proliferation and differentiation were unaltered by combined LPA1 and LPA2 deficiency we considered the possibility that the attenuation of injury induced intimal hyperplasia in LPA1^-/-2^-/- mice involves a requirement for these receptors as regulators of SMC migration which, as discussed below, we evaluated using primary cultures of mouse SMC.

LPA1^-/-2^-/- vascular SMC exhibit attenuated proliferative responsiveness in vivo and in vitro

Although LPA responses of cultured SMCs from humans and rats have been investigated in detail these types of studies have not been reported with mouse SMCs. To identify the mechanistic basis for vascular injury response phenotypes observed in the LPA receptor deficient mice, we examined proliferation, dedifferentiation, and migration responses to LPA of isolated mouse vascular SMCs. No significant differences in growth of WT and LPA1^-/- SMC in serum were observed, but LPA1^-/-2^-/- SMC grew more slowly (Figure 4A). This difference appeared to be dependent on the LPA component of serum because there was no difference between PDGF-induced growth of wild type and LPA1^-/-2^-/- SMC (Figure 4B). LPA regulates cell proliferation through ERK activation in many cell systems. Therefore, we examined LPA-induced ERK responses in isolated SMC. LPA treatment of isolated SMC increased ERK activation maximally at 5 – 10 min (Figure 4C). This response was concentration-dependent with maximal activation observed with 1 μM LPA (data not shown). A similar increase in LPA-induced ERK activity in SMCs cultured from LPA2^-/- mice was observed (data not shown). In contrast, there was an approximately two-fold reduction in the initial burst of LPA-induced ERK activation in LPA1^-/- SMC and a five-fold reduction in LPA1^-/-2^-/- SMC (Figure 4C).

LPA1^-/-2^-/- SMCS exhibit decreased migration in response to LPA, while LPA1^-/- SMCs exhibit enhanced migration due to upregulation of the LPA3 receptor

Although the reduced proliferative responses observed in LPA1^-/-2^-/- SMC in vitro are consistent with the protection of LPA1^-/-2^-/- mice from the formation of neointima in response to injury, the lack of a change in proliferation markers following injury in LPA1^-/-2^-/- vessels does not support the idea that protection from the development of intimal hyperplasia is a direct result of attenuated cell proliferation. In many systems, LPA is a potent stimulus for migration, therefore, we compared migration responses to LPA in SMC derived from wild type and LPA receptor deficient mice (Figure 5). LPA, serum, and PDGF stimulated 4.7 ± 0.94, 6.44 ± .046, and 11.5 ± 0.79 fold increases respectively in migration of SMC (Figure 5). The migration of LPA1^-/-2^-/- cells was lower at baseline (0.39 ± 0.01 that of WT cells; p <0.01), and the cells did not display any increase in migration to LPA and serum (Figure 5) although migration to PDGF was preserved (10.5 ± 1.7 fold increase). In contrast, LPA1^-/- cells were hypermigratory at baseline (7.9 ± 1.8 fold higher than WT cells; p = 0.003) and displayed a further exaggerated migration response to LPA and serum (Figure 5).

To determine if expression of an alternate LPA receptor might be responsible for the hyper-migratory LPA response observed in the LPA1^-/- cells, the assays were conducted in the presence of VPC32301, a selective antagonist of the LPA3 receptor³³. LPA3 antagonism had a modest ∼30% inhibition of WT migration (22,978 ± 3507 versus 33,680 ± 6329 μm² in antagonist-treated and vehicle-treated cells, respectively), and inhibited LPA-induced ERK activation in WT cells by 10 – 15%. LPA3 receptor antagonism reduced both baseline and LPA-induced migration in LPA1^-/- cells to levels observed in WT cells (Figure 5) and inhibited LPA-induced ERK activation in LPA1^-/- cells by ∼35%. We found that LPA3 receptor expression was elevated 3.7 ± 0.2 fold in LPA1^-/-, but not LPA2^-/- or LPA1^-/-2^-/- cells as compared to WT cells (see Online Table 1). To determine if upregulation of LPA3 expression alone could account for the enhanced migration observed in LPA1-/- cells, LPA3 was overexpressed in WT SMC using recombinant lentivirus vectors which resulted in enhanced migration towards LPA and serum phenocopying the behavior of LPA1-/- cells (Figure 5).

LPA1 and LPA2 are not required for Rho activation or SMC dedifferentiation

Migration of SMC to LPA was attenuated by ERK and Rho kinase inhibitors (see Online Figure 3). Therefore, we examined LPA-induced Rho responses in WT, LPA1^-/- and LPA1^-/-2^-/- SMC. In two of three independent SMC cultures, LPA1^-/- cells had higher active Rho at baseline, but no differences in LPA-induced Rho activation were observed in WT, LPA1^-/- and LPA1^-/-2^-/- SMC (Figure 6A). These results suggest that lack of ERK activation, rather than changes Rho activation, account for the reduced migration in the LPA1^-/-2^-/- mice.

Both LPA and Rho have been implicated as regulators of the differentiation state of vascular SMC ³. However, our results following carotid ligation suggest that neither LPA 1 nor LPA2 is required for injury-induced SMC dedifferentiation in vivo. To determine if either of these receptors plays role in dedifferentiation of isolated SMC, we examined expression patterns of the differentiation marker SMC myosin heavy chain. Expression of SMC myosin heavy chain was high in confluent SMC deprived of serum for 72 hours (Figure 6B). Exposure of these cells to 10% serum or 1μM LPA for twenty four hours down-regulated SMC myosin heavy chain expression to a similar extent in WT, LPA1^-/- and LPA1^-/-2^-/- cells (Figure 6B). Taken together, these results suggest that LPA1 and LPA2 are not essential for LPA-triggered dedifferentiation of SMC in vitro.

LPA1 and 2 do not mediate acute blood pressure responses to LPA or regulate systemic blood pressure

The above results suggest that LPA regulates the vascular injury response by modulating migratory properties of SMC. In other species, intravascular administration of LPA alters blood pressure which likely results from direct effects on vascular smooth muscle cell contractility. We sought to determine if the same receptors that regulate vascular SMC migration responses to LPA also contribute to the effects of LPA on blood pressure. Intravenous infusion of BSA conjugated-LPA transiently increased mean arterial blood pressure in anesthetized mice, with 10 pmoles of LPA being the lowest dose that reproducibly increased mean arterial blood pressure above vehicle by 17 ± 6 mm Hg in WT mice. A similar response was observed in mice lacking either LPA1 or LPA2 or both, suggesting that neither receptor is responsible for acute increase in blood pressure elicited by LPA (Table 1). Likewise, there was no difference in blood pressure or heart rate in conscious WT, LPA1^-/-, LPA2^-/- or LPA1^-/-2^-/- mice (Table 2).