Signaling Properties and Functions of Two Distinct Cardiomyocyte Protease-Activated Receptors
Abstract—Previous studies have established that cardiomyocytes express protease-activated receptor (PAR)-1, a high-affinity receptor for thrombin, which is also activated by the tethered-ligand domain sequence (SFLLRN) and which promotes inositol trisphosphate accumulation, stimulates extracellular signal–regulated protein kinase, and modulates contractile function. A single previous report identified PAR-1 as a hypertrophic stimulus, but there have been no subsequent investigations of the mechanism. This study reveals the coexpression of PAR-1 and PAR-2 (a second PAR, which is activated by trypsin/tryptase but not thrombin) by Northern blot analysis and compares their signaling properties in neonatal rat ventricular cardiomyocytes. SFLLRN and SLIGRL (an agonist peptide for PAR-2) promote inositol trisphosphate accumulation, stimulate mitogen-activated protein kinases (extracellular signal–regulated protein kinase and p38-mitogen-activated protein kinase), elevate calcium concentration, and increase spontaneous automaticity. SFLLRN (but not SLIGRL) also activates c-Jun NH2-terminal kinase and AKT. In keeping with their linkage to pathways that have been associated with growth and/or survival, SFLLRN and SLIGRL both induce hypertrophy. However, PAR agonists promote cell elongation, a morphology that is distinct from the uniform increase in cell dimension induced by α1-adrenergic receptor activation. These studies provide novel evidence that cardiomyocytes coexpress 2 functional PARs, which link to a common set of signals that culminate in changes in contractile function and hypertrophic growth. PAR actions may assume clinical importance in the border zone surrounding an infarction, where local proteolysis of PARs by serine proteases generated during inflammatory or thrombogenic pathways would elevate calcium concentration (setting the stage for arrhythmias), promote hypertrophic growth, and/or influence cardiomyocyte survival.
Thrombin is a serine protease formed at the site of vascular injury with well-characterized roles in hemostasis, inflammation, and proliferative processes. Cellular responses to thrombin are mediated by a family of G-protein–coupled protease-activated receptors (PARs), of which PAR-1 is the prototype. PAR-1 is activated on cleavage of its N-terminal exodomain by thrombin, thus exposing a new amino terminal sequence (SFLLRN) that functions as a tethered peptide ligand.1 PAR-1 is also activated, independent of proteolysis, by SFLLRN.
Additional PAR family members have been identified. PAR-2 has 30% sequence identity to PAR-1 and is detected by Northern blot analysis in several tissues, including (at low levels) the heart.2 PAR-2 is activated by limited proteolysis of its amino-terminal exodomain by trypsin (but not thrombin) or by SLIGRL, the PAR-2 tethered-ligand sequence. Cleavage by trypsin likely represents the relevant PAR-2 activation mechanism in the gastrointestinal tract and in airway epithelium (where PAR-2 potently inhibits bronchoconstriction3 ). At sites not exposed to trypsin, serine proteases generated during inflammation, fibrinolysis, or thrombosis may be more relevant. Mast cell tryptase readily cleaves cell surface PAR-2 (but not PAR-1) and may be a pathophysiologically important activator of PAR-2 in the vasculature, where PAR-2 induces endothelium-dependent relaxation of arterial rings and mitogenesis.1 2 Mast cell infiltration and elevated tissue tryptase levels also are characteristic of certain cardiac syndromes,4 5 but a potential role for PAR-2 in the heart has never been considered. PAR-3 and PAR-4 are more recently identified PAR family members. PAR-3 and PAR-4 have been studied exclusively in the context of platelet aggregation; potential functions in other tissues remain completely unexplored.
Studies during the past several years have identified cardiomyocytes as targets for the action of thrombin. In rat ventricular myocytes, thrombin stimulates phosphoinositide hydrolysis, activates the extracellular signal–regulated protein kinase (ERK), induces atrial natriuretic factor expression, modulates calcium homeostasis, increases automaticity, and hastens recovery from an imposed acid load by activating Na+-H+ exchange.6 7 8 9 Collectively, these (and likely other signaling events) profoundly alter electrophysiological properties and contractile behavior and induce cardiomyocyte hypertrophy. On the basis of the identification of PAR-1 mRNA and responses to SFLLRN, PAR-1 has been implicated in the actions of thrombin.7 8 9 However, SFLLRN typically elicits biochemical responses that are more robust than those elicited by thrombin.7 This could result from the actions of SFLLRN at both PAR-1 and PAR-2, because SFLLRN activates both, whereas thrombin is selective for PAR-1.10 Accordingly, the present study uses SLIGRL (a synthetic PAR-2 hexapeptide agonist, which activates PAR-2 but not PAR-110 ) to test the hypothesis that incremental responses to SFLLRN over thrombin are due to the copresence of PAR-1 and PAR-2. The results provide novel insights into the signaling and growth-regulatory properties of individual PARs in cardiomyocytes.
Materials and Methods
Neonatal rat ventricular myocyte (NRVM) cultures were prepared according to methods described previously.7 For biochemical assays and measurements of calcium, myocytes were plated at a density of 3×104 cells/cm2 and cultured in DMEM supplemented with 10% FCS. This was followed by starvation in serum-free 1:1 DMEM/F-12 medium for 24 hours before assays on day 5 of culture. Cells were plated at 1.5×104 cells/cm2 for measurements of growth responses. Preliminary studies established that biochemical and calcium responses are not altered in a qualitative manner by differences in plating density. Cardiac fibroblast cultures were obtained from cells adherent to culture dishes during preplating.11
Northern blot hybridization was carried out with poly(A+)-enriched RNA and a fragment encompassing nucleotides 1 to 2123 of the human PAR-1 coding sequence12 or polymerase chain reaction products containing nucleotides 226 to 657 or 145 to 1364 of the murine PAR-2 or the human PAR-3 coding sequence.13 14 Probes were 32P-labeled by using the RadPrime labeling system (Life Technologies); hybridizations were carried out in Rapid-hyb buffer (Amersham) for 3 hours at 62°C. Wash stringency was 2× SSC and 0.1% SDS 2 times at 62°C, followed by 0.2× SSC and 0.1% SDS 2 times at 62°C.
Methods for the measurement of inositol phosphate (IP) accumulation and calcium in fura 2–loaded cultured NRVMs are published.7 Assessments of ERK and c-Jun NH2-terminal kinase (JNK) activities were by “in-gel” kinase assays with the use of 0.5 mg/mL myelin basic protein or glutathione S-transferase (GST)-c-Jun (purified by glutathione-Sepharose chromatography from Escherichia coli driven to express recombinant GST-c-Jun; plasmid was provided by Dr Peter Sugden, Imperial College School of Medicine, London, UK) as substrate, as described previously.7 Activation of p38-mitogen-activated protein kinase (p38-MAPK) was detected by immunoblot analysis with an antibody selective for the dually phosphorylated (activated) enzyme, according to the manufacturer’s instructions (New England Biolabs). Immunocomplex kinase assays for AKT were performed with histone H2B as substrate15 ; equal protein loading was validated by reprobing blots with anti-polyclonal AKT C-terminal.
Cell surface area was measured by digitized image analysis. After exposure to agonist for 72 hours, 8 to 10 frames per dish were captured at ×40 magnification; 30 to 50 cells were analyzed per treatment. For measurements of protein content, triplicate dishes were stimulated in serum-free medium with agonists (or vehicle) for 48 hours at 37°C. The medium was supplemented with [14C]phenylalanine (0.1 μCi/mL) plus 0.3 mmol/L nonradioactive phenylalanine during the final 24 hours of stimulation. Cells were rinsed with PBS and incubated in ice-cold 10% trichloroacetic acid for 30 minutes. Precipitates were washed twice with ice-cold 10% trichloroacetic acid and solubilized in 1% SDS at 37°C for 1 hour. Duplicate aliquots from each sample were assayed for radioactivity, protein, and DNA content. One hundred to 150 cells per treatment group were scored (at ×63 magnification) for sarcomeric organization after incubation for 72 hours with the indicated agonists and immunostaining with monoclonal anti–α-sarcomeric actinin.
Cardiomyocytes Coexpress PAR-1 and PAR-2
Figure 1⇓ shows that a characteristic size transcript for PAR-1 (3.5 kb) is readily detected in cardiomyocytes and fibroblasts cultured from the neonatal ventricle. The 3.5-kb PAR-1 band is also detected in the adult rat ventricle, where it must be resolved by short exposure times from a heterogeneous signal spanning from ≈1 to 5 kb (identical results were obtained with multiple RNA preparations and 3 different PAR-1 random-primed cDNA probes as well as a uniformly labeled cRNA probe). The diffuse nature of much of the PAR-1 signal is not the result of a bulk mobility artifact, because PAR-2 and β-actin are detected as discrete bands in this sample. PAR-2 transcripts are detected as a ≈3.0-kb band in adult ventricle and NRVMs; PAR-2 transcripts are at the limits of detection in cardiac fibroblasts. PAR-3 is abundant in megakaryocytes (included as controls), but it is not detected in cardiomyocytes or fibroblasts.
PAR-1 and PAR-2 Stimulate Phosphoinositide Hydrolysis
Consistent with previous studies, thrombin and SFLLRN induce rapid and transient increases in inositol trisphosphate (IP3) and inositol bisphosphate (IP2), which are followed by a more sustained accumulation of inositol monophosphate (IP1). IP1 accumulation in response to SFLLRN at 30 minutes is 2.3±0.2-fold greater than the response to thrombin (P<0.05, n=6; Figure 2A⇓). To determine whether the incremental stimulation of phosphoinositide hydrolysis by SFLLRN might be due to its combined actions at PAR-1 and PAR-2, responses to trypsin and SLIGRL were examined in parallel. Trypsin and SLIGRL promote the sequential accumulation of IP3, IP2, and IP1. Responses to SFLLRN and SLIGRL display similar kinetics and EC50 values (≈50 μmol/L; Figure 3B⇓ and Reference 7 ). However, responses to SFLLRN are greater in magnitude that those elicited by SLIGRL. FLLRN (500 μmol/L), a control peptide lacking the N-terminal serine (does not activate PAR-1), does not promote IP accumulation at 2 or 30 minutes (data not shown). This result establishes that the stimulatory actions of SFLLRN and SLIGRL are specific. Figure 2B⇓ demonstrates that SFLLRN activates phospholipase C in cardiac fibroblasts but that SLIGRL does not. Agonist peptide actions are consistent with the Northern analysis of mRNAs from cardiac fibroblasts (Figure 1⇑) and indicate that SLIGRL-induced biochemical responses cannot be attributed to robust PAR-2 signaling by a minor contaminating fibroblast population in cardiomyocyte cultures.
PAR Activation of MAPKs
The next experiments examined whether SLIGRL stimulates MAPKs. In most experiments, ERK, JNK, and p38-MAPK activities were assayed in parallel, on material from the identical culture preparation. This permitted a rigorous analysis of the balance of ERK versus JNK versus p38-MAPK activation by individual PARs, in line with our goal to identify potential quantitative/qualitative differences in the intensity of PAR signaling through individual MAPK cascades, which might translate into distinct functional consequences.
Figure 3A⇑ shows that SLIGRL induces transient ERK activation; both 42- and 44-kDa species of ERK are detectably activated by 2 minutes and are maximally activated by 5 minutes. The rapid and transient kinetics of this response are similar to the kinetics previously described for SFLLRN.7 ERK activation by SLIGRL is concentration dependent (EC50 50 μmol/L). The concentration-response relationships for SLIGRL-dependent activation of ERK and IP accumulation are similar (Figure 3B⇑). The magnitude of ERK activation by SFLLRN is greater than the responses to equimolar SLIGRL or thrombin (Figure 3C⇑). FLLRN (500 μmol/L for 5 minutes) does not activate ERK, establishing the specificity of SFLLRN and SLIGRL responses (data not shown).
To determine whether PARs activate JNK, lysates from cardiomyocytes stimulated with PAR agonists for variable intervals (2 to 30 minutes) were subjected to “in-gel” kinase assays with GST-c-Jun as substrate (with sorbitol, which induces a ≈10-fold increase in JNK activity, as control). Figure 4A⇓ shows that JNK is activated by thrombin and SFLLRN (PAR-1). These responses display kinetics that are rapid (peak at 5 to 10 minutes, data not shown) and contrast with the more gradual and sustained activation of JNK by other G-protein–coupled receptors (GPCRs) in cardiomyocytes. The magnitude of JNK activation by thrombin and SFLLRN is similar. JNK is not significantly activated by SLIGRL (PAR-2).
p38-MAPK becomes phosphorylated (activated) in cells treated with SFLLRN, thrombin, and SLIGRL (Figure 4B⇑); in each case, the kinetics of p38-MAPK activation is relatively rapid (responses detectable at 2 minutes and maximal by 5 minutes, data not shown). p38-MAPK activation by SFLLRN tended to be somewhat greater than activation by thrombin; both responses are much more robust than the response to SLIGRL. Collectively, these studies identify distinct patterns of MAPK activation by PARs. Both PARs activate ERK, but only PAR-1 couples effectively to the activation of the stress-activated protein kinases (JNK and p38 MAPK); activation of stress-activated protein kinases by PAR-2 is weak (p38-MAPK) or undetectable (JNK).
AKT Is Activated by PAR-1 but Not PAR-2
There is limited recent evidence for GPCR activation of the phosphatidylinositol 3-kinase/AKT survival pathway in noncardiomyocytes.15 Figure 4C⇑ shows that AKT is activated by thrombin and SFLLRN, whereas activation by SLIGRL is not detected. Nevertheless, these responses are relatively modest compared with the 9-fold increase in AKT activity induced by insulin-like growth factor (IGF)-1.
PAR-1 and PAR-2 Accelerate Automaticity and Elevate Calcium Concentration
SFLLRN (300 μmol/L) accelerates the automatic beating rate of cultured NRVMs (53±5%, from 45±4 to 69±8 bpm at the peak of the response, which occurs 15 to 20 seconds after bolus administration of peptide agonist; n=6). Parallel studies with equimolar SLIGRL establish that PAR-2 activation also induces a positive chronotropic response with identical kinetics. However, the magnitude of the SLIGRL-dependent positive chronotropic response is more modest (28±4%, from 43±3 to 55±3 bpm; n=6).
To determine whether SLIGRL also influences calcium concentration, independent of any effect on contractile rate, these experiments were repeated in fura 2–loaded cells electrically driven at 1 Hz (to maintain a constant beating rate during agonist exposure). Figure 5⇓ shows that SLIGRL mimics the effect of SFLLRN to rapidly increase calcium (results are summarized in Table 1⇓). For both agonists, the increase in calcium was dose dependent (maximum at 300 μmol/L) and was completely reversed during 3 to 4 minutes of washout. However, the increase in calcium evoked by SFLLRN was larger than that elicited by SLIGRL.
PAR-1 and PAR-2 Promote Cardiomyocyte Hypertrophy
A single report by Glembotski et al8 in 1993 demonstrated that NRVMs grown in the presence of thrombin enlarge, display highly organized sarcomeres, and express atrial natriuretic factor. That study also reported that SFLLRN promotes atrial natriuretic factor expression; its effects on other indices of hypertrophy were not examined. Because SFLLRN and SLIGRL both activate mechanisms implicated in hypertrophic signaling, we examined their effects on various indices of cell growth. Table 2⇓ demonstrates that SFLLRN and SLIGRL both induce significant increases in [3H]phenylalanine incorporation, protein content, cell size, and myofibrillar organization. Growth-stimulatory effects of SFLLRN are more pronounced than those elicited by SLIGRL (by all parameters measured). However, the morphological features of cells treated with SFLLRN and SLIGRL are similar and quite distinct from the morphology of cells grown in the presence of norepinephrine (Figure 6⇓). α1-Adrenergic receptor (α1-AR) activation by norepinephrine leads to a uniform increase in cell dimensions (length and width, typical of the morphological changes observed in the setting of pressure-overload hypertrophy). In contrast, PAR activation leads to cell elongation, with less of an increase in cell width (Table 2⇓). This morphologically distinct form of hypertrophy is more characteristic of the changes induced by the mechanical stimulus of volume overload or, in the in vitro culture model, cardiotrophin-1 (an interleukin-6 family member that signals through gp13017 ). Thus, the qualitative/quantitative differences in signaling molecule activation by individual PARs may contribute to differences in the degree of cell enlargement, but they do not impart specificity with respect to cell morphology. Rather, the distinct morphological phenotypes appear to be dictated by differences in signaling pathways recruited by PARs versus α1-ARs.
The profound cardiac actions of thrombin were first identified almost a decade ago. In the ensuing years, there has been considerable progress defining signaling pathways that contribute to the cardiomyocyte hypertrophic growth program. However, the initial observations on the actions of thrombin were not extended in the context of current concepts of cardiomyocyte hypertrophic signaling mechanisms, and the research completely neglected to consider whether cardiomyocytes also express other PARs with similar or overlapping signaling properties. The present study demonstrates that cardiomyocytes express PAR-2; PAR-2 mRNA is detected by Northern blot analysis, and the PAR-2 agonist peptide SLIGRL activates a spectrum of biochemical signaling pathways and functional responses (phospholipase C, ERK, p38-MAPK, calcium, spontaneous automaticity, and hypertrophic growth). Responses to SLIGRL report the specific actions of PAR-2; there is no evidence that SLIGRL can activate any other known PAR.10 The present study also demonstrates for the first time that PAR-1 activates JNK, p38-MAPK, and AKT in cardiomyocytes and that PARs (PAR-1 and PAR-2) promote a form of cardiomyocyte hypertrophy that is morphologically distinct from that elicited by the prototypical α1-adrenergic GPCR.
The observation that prolonged stimulation of either PAR-1 or PAR-2 culminates in hypertrophic growth is consistent with the literature implicating many of the signaling mechanisms common to PAR-1 and PAR-2 in cardiomyocyte hypertrophy. For example, PAR-1 and PAR-2 both effectively activate the ERK cascade. This mechanism originally was implicated in hypertrophic signaling and more recently has been identified as a mediator of cardiomyocyte survival.16 18 PAR-1 also activates the stress-activated protein kinases, whereas PAR-2 does not increase JNK activity and stimulates p38-MAPK weakly. Because PAR-2 induces cardiomyocyte growth, these results argue that the stress-activated protein kinase pathways are not absolutely required for a growth response (although the absence of these additional signals could underlie the more modest hypertrophic growth response induced by PAR-2, relative to PAR-1).
PAR-1 and PAR-2 both couple to increases in spontaneous automaticity and elevations in calcium. These responses would tend to predispose to arrhythmias, but they also are predicted to promote hypertrophic growth, at least in part through mediators such as the δB isoform of calmodulin kinase II19 and the calcium/calmodulin-dependent phosphoprotein phosphatase, calcineurin.20 Of note, PARs and α1-ARs promote morphologically distinct forms of hypertrophy; these GPCRs also differ in their ability to modulate calcium in NRVMs. PARs induce a robust increase in calcium, whereas calcium is not detectably modulated by α1-ARs (data not shown). Although increases in calcium that are localized or are below the limits of detection with the use of fura 2 as indicator cannot be excluded, the failure to detect prominent calcium responses to α1-AR agonist activation is consistent with published literature.21 These results suggest that calcium-dependent signals might play a particularly prominent role in hypertrophic signaling by PARs, relative to α1-ARs. The relative contribution(s) of calcium versus other signaling mechanisms (Rho/Rho kinase, Src) in the induction of morphologically distinct forms of hypertrophy by PARs and α1-ARs requires further study.
Results reported in the present study demonstrate that AKT is activated by PAR-1 in cardiomyocytes; to our knowledge, this represents the first evidence that a GPCR activates AKT in cardiomyocytes. The significance of the relatively low level of PAR-1–dependent AKT activation (relative to IGF-1) is uncertain. Known downstream targets of AKT (that also are activated by PARs) should be the focus of future investigations. These would include S6 kinase (which has been implicated in cardiomyocyte hypertrophy22 ) and endothelial NO synthase (which has emerged as a pivotal regulator of cardiomyocyte contractile function and growth responses23 ). AKT is a critical mediator of growth factor–dependent survival in neurons; a similar process may impact on cell loss during injury/infarction in the heart.
The present study establishes that PAR-1 and PAR-2 can mediate protease-dependent alterations in cardiomyocyte function. Ultimately, the significance of these findings rides on the identification of natural activators of endogenous cardiomyocyte PARs. The actions of thrombin are predicted to be important in the setting of hemorrhagic infarction, in which the endothelial barrier is broken and myocytes come into direct contact with blood-borne substances. Although cardiomyocytes typically do not encounter trypsin, 5 to 10 homologues of trypsin have been identified in the mammalian genome; one of these trypsin-like enzymes is expressed ubiquitously, including in the heart.24 Its activity as a proteolytic agonist for PARs must be tested. Mast cell tryptase is another protease that activates PAR-2.25 Mast cells can be identified between muscle fibers in normal ventricles and (in increased density) in idiopathic and dilated cardiomyopathies.5 Mast cells degranulate and release biologically relevant concentrations of tryptase during acute insults. Thus, it is reasonable to speculate that at least some of the derangements during myocardial infarction could result from PAR-2 activation by tryptase. Cardiomyocyte PAR-1 and/or PAR-2 activation in border zone regions adjacent to areas of myocardial necrosis could constitute critical signals that influence gene expression, promote hypertrophy/apoptosis, and/or change contractile function.
This study was supported by US Public Health Service, National Heart, Lung, and Blood Institute grant HL-49537. We thank Ema Stasko for preparing myocyte cultures.
- Received November 24, 1999.
- Accepted March 30, 2000.
- © 2000 American Heart Association, Inc.
Coughlin SR. How the protease thrombin talks to cells. Proc Natl Acad Sci U S A. 1999;96:11023–11027.
Dery O, Corvera CU, Steinhoff M, Bunnett NW. Proteinase-activated receptors: novel mechanisms of signaling by serine proteases. Am J Physiol. 1998;274:C1429–C1452.
Konanen PT, Kaartinen M, Paavonen T. Infiltrates of activated mast cells at the site of coronary atheromatous erosion or rupture in myocardial infarction. Circulation. 1995;92:1084–1088.
Patella V, Marino I, Arbustini E, Lamparter-Schummert B, Verga L, Adt M, Marone G. Stem cell factor in mast cells and increased mast cell density in idiopathic and ischemic cardiomyopathy. Circulation. 1998;97:971–978.
Steinberg SF, Robinson RB, Lieberman HB, Stern DM, Rosen MR. Thrombin modulates phosphoinositide metabolism, cytosolic calcium, and impulse initiation in the heart. Circ Res. 1991;68:1216–1229.
Jiang T, Kuznetsov V, Pak E, Zhang HL, Robinson RB, Steinberg SF. Thrombin receptor actions in neonatal rat ventricular myocytes. Circ Res. 1996;78:553–563.
Glembotski CC, Irons CE, Krown KA, Murray SF, Sprenkle AB, Sei CA. Myocardial α-thrombin receptor activation induces hypertrophy and increases atrial natriuretic factor gene expression. J Biol Chem. 1993;268:20646–20652.
Yasutake M, Haworth RS, King A, Avkiran M. Thrombin activates the sarcolemmal Na+-H+ exchanger: evidence for a receptor-mediated mechanism involving protein kinase C. Circ Res. 1996;79:705–715.
Lerner DJ, Chen M, Tram T, Coughlin SR. Agonist recognition by proteinase-activated receptor 2 and thrombin receptor: importance of extracellular loop interactions for receptor function. J Biol Chem. 1996;271:13943–13947.
Nystedt S, Larsson AK, Aberg H, Sundelin J. The mouse proteinase-activated receptor-2 cDNA and gene: molecular cloning and functional expression. J Biol Chem. 1995;270:5950–5955.
Sheng Z, Knowlton K, Chen J, Hoshijima M, Brown JH, Chien KR. Cardiotrophin 1 (CT-1) inhibition of cardiac myocyte apoptosis via a mitogen-activated protein kinase-dependent pathway. J Biol Chem. 1997;272:5783–5791.
Wollert KC, Taga T, Kishimoto T, Glembotski CC, Vernallis AB, Heath JK, Pennica D, Wood WI, Chien KR. Cardiotrophin-1 activates a distinct form of cardiac muscle cell hypertrophy. J Biol Chem. 1996;271:9535–9545.
Sugden PH. Signaling in myocardial hypertrophy: life after calcineurin? Circ Res. 1999;84:633–646.
Ramirez MT, Zhao XL, Schulman H, Brown JH. The nuclear δB isoform of Ca2+/calmodulin-dependent protein kinase II regulates atrial natriuretic factor gene expression in ventricular myocytes. J Biol Chem. 1997;272:31203–31208.
Olson EN, Molkentin JD. Prevention of cardiac hypertrophy by calcineurin inhibition: hope or hype? Circ Res. 1999;84:623–632.
Boluyt MO, Zheng JS, Younes A, Long X, O’Neill L, Silverman H, Lakatta EG, Crow MT. Rapamycin inhibits α1-adrenergic receptor–stimulated cardiac myocyte hypertrophy but not activation of hypertrophy-associated genes: evidence for involvement of p70 S6 kinase. Circ Res. 1997;81:176–186.
Drexler H. Nitric oxide synthases in the failing human heart: a double-edged sword? Circulation. 1999;99:2972–2975.
Ohmura K, Kohno N, Kobayashi Y, Yamagata K, Sato S, Kashiwabara SI, Baba T. A homologue of pancreatic trypsin is localized in the acrosome of mammalian sperm and is released during acrosome reaction. J Biol Chem. 1999;274:29426–29432.
Molino M, Barnathan ES, Numerof R, Clark J, Dreyer M, Cumashi A, Hoxie JA, Schechter N, Woolkalis M, Brass LF. Interactions of mast cell tryptase with thrombin receptors and PAR-2. J Biol Chem. 1997;272:4043–4049.