Expression, Release, and Biological Activity of Parathyroid Hormone–Related Peptide From Coronary Endothelial Cells
Abstract—Ventricular cardiomyocytes have previously been identified as potential target cells for parathyroid hormone–related peptide (PTHrP). Synthetic PTHrP peptides exert a positive contractile effect. Because systemic PTHrP levels are normally negligible, this suggests that PTHrP is expressed in the ventricle and acts as a paracrine mediator. We investigated the ventricular expression of PTHrP and its expression in cultured cells isolated from the ventricle, studied the release of PTHrP from hearts and cultures, and investigated whether this authentic PTHrP mimics the biological effects previously described for synthetic PTHrP on ventricular cardiomyocytes. We found PTHrP expressed in ventricles of neonatal and adult rat hearts. In cells isolated from adult hearts, we found PTHrP expression exclusively in coronary endothelial cells but not in cardiomyocytes. The latter, however, are target cells for PTHrP. PTHrP was released from isolated perfused hearts during hypoxic perfusion and from cultured coronary endothelial cells under energy-depleting conditions. This PTHrP was biologically active; ie, it exerted a positive contractile and lusitropic effect on cardiomyocytes. Authentic PTHrP was glycosylated and showed a slightly higher potency than synthetic PTHrP. These results suggest that PTHrP is an endothelium-derived modulator of ventricular function.
Parathyroid hormone–related peptide (PTHrP) has been found in various mammalian tissues, including the myocardium.1 2 This peptide hormone shares multiple biological activities with parathyroid hormone (PTH) (reviewed in Reference 3 ). Because of the strong homology in the N-terminal part of both peptide hormones, they bind to a common receptor on classical target cells, such as bone cells.4 Cardiac cells are not among classical target cells but nevertheless respond to PTHrP. PTHrP, but not PTH, was found to exert a positive inotropic effect on isolated rat hearts.5 It was demonstrated with the use of isolated cardiomyocytes that this contractile effect is indeed a direct effect of PTHrP on the cardiomyocytes.6 This positive contractile effect seems due to activation of adenylate cyclase in cardiomyocytes. In contrast to the functional similarities between PTHrP and PTH on classical target cells, PTH does not initiate the described contractile effect of PTHrP on cardiomyocytes.6
Plasma levels of PTHrP are normally very low, suggesting a paracrine or autocrine role of PTHrP. In cardiac tissue, PTHrP is locally expressed. The highest levels of expression are found in atria and vessels.7 In ventricular myocardium, the average level of its expression is low in general, and it is as yet unclear whether cardiomyocytes or other cell types represent the predominant site of expression. It is also not yet known whether cells of ventricular myocardium can release PTHrP in a biologically active form.
In the present study, the expression and release in biologically active form of PTHrP in 2 myocardial cell types, ie, ventricular cardiomyocytes and coronary endothelial cells, was investigated. Previous studies on different endothelial preparations showed that hypoxia induces the release of peptides with contractile effects, of which endothelin is now the one best known.8 9 We wondered whether PTHrP is released in a similar manner and analyzed whether PTHrP is released from isolated perfused hearts and endothelial cells under basal or energy-depleting conditions. The effect of authentic PTHrP, found to be released from endothelial cells in culture, on contractile function of cardiomyocytes was analyzed and compared with a biologically active synthetic PTHrP partial peptide, PTHrP(1–34).
Materials and Methods
Male Wistar rats (250 to 300 g) were used for all experiments on adult tissues and cells. Neonatal tissues were obtained from 3-day-old male Wistar rats. Ventricular cardiomyocytes were isolated from ventricles of adult male rats as described before.10 Coronary endothelial cells were isolated as described before11 and grown to confluence before use.12
Perfusion conditions for hearts and determination of enzyme release have been described before.13
Polymerase Chain Reaction (PCR)
PCR for expression of PTHrP was performed as described before.14
Total RNA from ventricles of adult rats were prepared from frozen tissues. Total RNA was electrophoresed in formaldehyde gels and subsequently transferred onto Hybond N (Amersham). Filters were fixed under UV irradiation. PCR amplificates of PTHrP were labeled with [α-32P]dATP and used to probe for PTHrP transcripts.
Cells were treated with lysis buffer as described before.6 PTHrP was also detected in perfusates of hearts and supernatants of cells. These fluids were treated on ice first with deoxycholate and then with trichloroacetic acid and centrifuged, and the pellet was resuspended in Laemmli buffer. Samples were loaded on 12.5% SDS-PAGE and blotted onto membranes as described before.6 Blots were incubated first with a monoclonal mouse antibody directed against the amino acid residues 38 to 64 of PTHrP (antibody GF08, Oncogene Research Products) and second with an anti–mouse Ig antibody coupled to alkaline phosphatase.
Glycosylation of PTHrP
Western blots of PTHrP were washed with PBS (pH 6.5), oxidized for 20 minutes by addition of sodium metaperiodate (10 mmol/L in sodium acetate buffer, pH 5.5), washed 3 times with PBS, and incubated for 1 hour with digoxigenin (DIG)–O-3-succinyl-ε-aminocaproic acid hydrazide (1 μmol/L in sodium acetate buffer, pH 5.5). Thereafter, blots were washed again with PBS and incubated with an anti-DIG antibody coupled to alkaline phosphatase.
Cell contraction of adult ventricular cardiomyocytes was investigated on isolated cells and paced at a constant frequency of 0.5 Hz, as described before.15 Accumulation of cAMP in cardiomyocytes was investigated as described before.15
Protein pellets from supernatants of energy-depleted coronary endothelial cells were resuspended in immunoprecipitation buffer and incubated with a monoclonal antibody directed against PTHrP amino acids 38 to 64. Protein A Sepharose was added to these samples. After centrifugation (3000g, 3 minutes), pellets were washed 3 times with buffer containing 50 mmol/L sodium phosphate and 0.1% (vol/vol) NP-40. Protein A Sepharose was removed by incubation with dissolving buffer.
Quantitative results are expressed as mean±SEM. In experiments in which more than 2 groups were compared with each other, ANOVA was used, with Student-Newman-Keuls test for post hoc analysis. In cases in which 2 groups were compared, conventional t tests were performed. P<0.05 was used as level of significance.
An expanded Materials and Methods section is available online at http://www.circresaha.org.
Ventricular Expression of PTHrP
The expression of PTHrP was investigated in hearts of neonatal and adult rats and, for comparison, in kidneys of adult rats. The primers used to detect the expression of PTHrP in these tissues amplify a 161-bp fragment. PTHrP was found to be expressed in the hearts of neonatal rats (Figure 1A⇓, lane 2), in the atrium and ventricle of hearts of adult rats (Figure 1A⇓, lanes 1 and 3, respectively), and in the kidney of adult rats (Figure 1A⇓, lane 4). No amplification products could be detected in the absence of reverse transcriptase (Figure 1A⇓, lane 5). To further characterize ventricular expression of PTHrP, total RNA from the ventricle was hybridized with a 32P-labeled PTHrP fragment, which corresponds to the fragment amplified by reverse transcription-(RT)–PCR. This Northern blot analysis detected 1 single PTHrP transcript, of ≈1.4 kb (Figure 1B⇓). The results of hybridization with GAPDH demonstrate integrity of RNA and give a reference for loading of the lanes.
The expression of PTHrP was also analyzed in 2 purified cell populations derived from ventricular tissue of adult rats, namely ventricular cardiomyocytes and coronary endothelial cells. PTHrP was found to be strongly expressed in endothelial cells, in which a 161-bp fragment corresponding to PTHrP was detected by RT-PCR (Figure 2A⇓, lanes 2 and 4). Abundant expression of PTHrP in these cells was confirmed on the protein level. In 4 analyzed cultures of coronary endothelial cells, a single band of ≈50 was detected on Western blots prepared with a monoclonal antibody against PTHrP (Figure 2B⇓, lanes 1 through 4). In contrast, in ventricular cardiomyocytes, no PTHrP transcript was detected by RT-PCR (Figure 2⇓, lanes 1 and 3), and Western blots for PTHrP also remained negative (Figure 2B⇓, lanes 5 through 8).
The apparent molecular weight of PTHrP expressed in coronary endothelial cells is higher than the calculated molecular weight for a 141–amino acid peptide. Thus, we investigated further whether post-translational modification by glycosylation may cause the shift in the apparent versus the calculated molecular weight. Therefore, PTHrP released from coronary endothelial cells was purified by immunoprecipitation. In a DIG-glycan detection assay, indicating glycosylation of proteins, we found that PTHrP is indeed modified in this way (Figure 3⇓, lanes 5 through 7).
Release of PTHrP
It was investigated whether a factor with immunoreactivity to PTHrP antibodies is released from adult cardiac tissue. Isolated rat hearts were perfused under normoxic conditions for up to 30 minutes, and aliquots of the perfusate were collected every 15 minutes. A small basal release of PTHrP immunoreactive material was observed. Subsequently, hearts were perfused for 60 minutes under hypoxic conditions. Again, samples from the perfusate were collected every 15 minutes and analyzed for PTHrP release. PTHrP was released from hypoxic perfused rat hearts as illustrated in Figure 4A⇓. On average, PTHrP release from perfused hearts increased during hypoxic perfusion, reaching a maximum after 30 minutes. PTHrP release exceeds prehypoxic values (Figure 4B⇓). Release of creatine kinase was determined in parallel to monitor unspecific protein release that may result from cardiac damage. During hypoxic perfusion, creatine kinase release did not exceed the low basal values seen before hypoxic perfusion (Figure 4B⇓).
To evaluate whether the observed release of PTHrP from hypoxic perfused rat hearts is caused by the release from coronary endothelial cells under energy-depleting conditions, such conditions were simulated in cultures of coronary endothelial cells with use of metabolic inhibitors (cyanide and deoxyglucose), and the release of PTHrP was determined. Without energy depletion, cultured coronary endothelial cells did not release PTHrP. On energy depletion, PTHrP was released as illustrated in Figure 5A⇓. PTHrP release from energy-depleted coronary endothelial cells was significantly elevated as early as 40 minutes after addition of potassium cyanide (KCN) and deoxyglucose (Figure 5B⇓). Release of lactate dehydrogenase did not occur during 1 hour of energy depletion. This indicates that the early PTHrP release is not due to loss of membrane integrity.
Biological Activity of PTHrP Produced by Coronary Endothelial Cells
It was investigated whether the PTHrP released by energy-depleted coronary endothelial cells is biologically active on cardiomyocytes. The previously described positive contractile effect of PTHrP on cardiomyocytes was monitored. The supernatants of metabolically inhibited endothelial cell cultures were collected, the proteins were precipitated by trichloroacetic acid treatment and resolved in buffer, and PTHrP was subsequently immunoprecipitated. The immunoprecipitated material showed a single band on silver-stained gels (Figure 6⇓, lane 1) that cross-reacted with PTHrP antibodies (Figure 6⇓, lanes 2 and 3). The precipitated protein was dissolved in water, and the protein content was determined. On average, ≈5.5 pg PTHrP/mg total protein was released from metabolically inhibited endothelial cells within 120 minutes of incubation. The concentration of PTHrP was calculated on the basis of the apparent molecular weight of the single protein band (see Figure 2B⇑).
The activity of endothelium-derived PTHrP was compared with that of a synthetic partial peptide of PTHrP(1–34) composing the adenylate cyclase-activating domain of PTHrP. Both endothelium-derived authentic PTHrP and synthetic PTHrP increased dose dependently the cell shortening of electrically paced cardiomyocytes, by 68% and 65%, respectively. The dose-response curve for endothelium-derived PTHrP is shifted 1 order of magnitude to the left compared with the synthetic PTHrP fragment (Figure 7⇓). In comparison, immunoprecipitates using an anti–mouse IgG antibody not directed against PTHrP did not exert a positive contractile effect (twitch amplitude as percentage of diastolic cell length, 4.63±0.23% versus 4.43±0.92%, n=12 cells; NS). The effects of synthetic and endothelium-derived PTHrP on cardiomyocytes were antagonized by preincubation of cardiomyocytes with PTH(1–34) (Figure 8⇓), which is known to act as a functional receptor antagonist of PTHrP on cardiomyocytes. These results indicate that both forms of PTHrP exert their effects via a common receptor. These experiments, investigating the positive contractile effect of endothelium-derived PTHrP, were also performed in the presence of the β-adrenoceptor antagonist propranolol (1 μmol/L). Whereas propranolol significantly attenuated the response to isoprenaline (100 nmol/L) from 12.24±2.98% to 5.46±1.85% (n=8 cells, P<0.01), it did not antagonize that of endothelium-derived PTHrP (9.10±1.16% versus 9.21±0.06%). The biological activity of endothelium-derived PTHrP was further demonstrated by a dose-dependent positive lusitropic effect on cardiomyocytes. PTHrP increased the maximal relaxation velocity from 58.2±3.1 to 87.3±5.4 μm/s (n=12, P<0.05) and reduced the time from peak contraction to 90% relaxation from 76.4±5.2 to 53.8±3.2 ms (n =12, P<0.05). These functional data were confirmed on the biochemical level by measuring the effect of endothelium-derived PTHrP on adenylate cyclase of cardiomyocytes. As an indicator for the adenylate cyclase activation, the accumulation of cAMP in the presence of a phosphodiesterase inhibitor, isobutyl methylxanthine, was determined. The observed effect with authentic PTHrP (30 nmol/L) was comparable with that achieved by isoprenaline (control, 0.3±0.2 pmol/mg protein; PTHrP, 4.2±1.3 pmol/mg protein, P<0.05 versus control; isoprenaline, 5.1±0.4 pmol/mg protein, P<0.05 versus control; NS versus PTHrP, each n=4).
It is the main and novel finding of the present study that coronary endothelial cells are able to express and release biologically active PTHrP, which activates adenylate cyclase and exerts a positive contractile and positive lusitropic effect on adult ventricular cardiomyocytes. The results indicate that PTHrP is an endothelium-derived modulator of ventricular function.
This study confirms previous findings on fetal and adult human hearts showing expression of PTHrP in heart tissue.1 2 7 In the present study, 1 isoform of PTHrP was found on the RNA level, corresponding to the 1.4-kb isoform previously found in human hearts as the major transcript. In human hearts, expression of 2 additional isoforms has been reported.2 The expression of only 1 single isoform in rat hearts is consistent with previous studies, showing the expression of a single PTHrP isoform in the rat tissue, corresponding to a 141-amino acid isoform on the protein level.16 We found expression of PTHrP in ventricular tissue as a whole, but not in cardiomyocytes isolated from the ventricles. In contrast to ventricular cardiomyocytes investigated here, cardiomyocytes from the atrium have been shown to express and secrete PTHrP.7 This expression pattern resembles that of atrial natriuretic factor, which suggests a similarity in transcriptional regulation. In contrast to ventricular cardiomyocytes, PTHrP is abundantly expressed in coronary endothelial cells, another constituent cell of myocardial tissue neighboring cardiomyocytes. Expression of PTHrP was found previously in several types of endothelial cells, including those isolated from human umbilical veins (HUVECs).17 Because endothelial cells do not express PTH/PTHrP receptors,17 they do not represent target cells for PTHrP. PTHrP seems, therefore, to represent a paracrine factor in the heart.
Coronary endothelial cells do not release PTHrP under basal culture conditions. Hypoxia has been shown to induce the release of peptides with contractile effects from the vasculature, of which endothelin is now the best known.8 9 We wondered whether PTHrP is released from endothelial cells in a similar manner. We found that PTHrP is released from hypoxic perfused hearts. Because PTHrP is expressed in the endothelial cell fraction, we also investigated whether PTHrP is released from isolated coronary endothelial cells under energy-depleting conditions. This was indeed the case. The findings indicate that the release of PTHrP from the hypoxic whole-heart preparation is mainly due to the release from energy-depleted endothelial cells. The mechanism by which energy depletion provokes the release was not investigated in this study and needs further analysis.
We used antibodies directed against a mid-regional part of PTHrP to immunoprecipitate the hormone, because they are generally believed to detect intact PTHrP that is not cleaved by proteases. Indeed, a single protein band was detected by the PTHrP antibody. This single protein band had an apparent molecular weight of ≈50. This is much higher than the expected molecular weight of ≈16 for a 141-amino acid protein. The difference may indicate post-translational modifications. We indeed demonstrated glycosylation of PTHrP released from the coronary endothelial cells. Glycosylation has also been reported for PTHrP secreted from epidermal keratinocytes.18 In view of such modifications, it was important to characterize the biological activity of this authentic PTHrP. All previous studies on functional effects of PTHrP on adult ventricular cardiomyocytes were performed using synthetic PTHrP peptides without such modifications.6 19 From these studies, it is known that PTHrP peptide fragments covering the N-terminal part of the molecule, eg, PTHrP(1–34), contain a binding and an adenylate cyclase–activating domain. Even though these synthetic peptides are useful to identify potential target cells of PTHrP, they cannot be used to predict the potency of authentic PTHrP. Authentic PTHrP isolated from coronary endothelial cells mimicked the known biological effects on cardiomyocytes. Both the authentic PTHrP and the synthetic partial peptide activate adenylate cyclase activity, increase the contractile activity, and exert a positive lusitropic effect on ventricular cardiomyocytes. Moreover, the functional effects of either PTHrP variant could be antagonized by PTH, which has no positive contractile effects on ventricular cardiomyocytes or in isolated perfused rat hearts.6 15 5 The dose-response curve of the authentic PTHrP was shifted in leftward direction compared with PTHrP(1–34), indicating a higher potency. The immunoprecipitated material was pure, as it gave a single band on silver-stained gels. However, we cannot rule out that minor amounts of other proteins may be included. Because the calculations for the concentration of the immunoprecipitated material were based on protein quantification of the samples, we may have overestimated the protein content of PTHrP in the immunoprecipitated samples. Considering these impurities, the dose-response curve would shift to the left. Another potential error may arise from the calculation of the concentrations by the use of the apparent molecular weight, because we do not know the exact molecular weight without clarifying the form of glycosylation. In regard to the sharp bands detected, we expect that PTHrP is modified by O-glycosylation. On average, one may assume 5 to 7 glucose moieties of ≈100 Da. Thus, the real molecular weight may be higher than that expected from the amino acid composition but lower than the apparent molecular weight. A lower molecular weight would shift the dose-response curve to the right, in this case by a factor of 2 to 3. This leaves the estimated dose-response curve of authentic PTHrP still left of the dose-response curve of synthetic PTHrP. Therefore, although we cannot determine accurately the molecular weight for endothelium-derived PTHrP, the main conclusion from our experiments, that endothelium-derived PTHrP is more potent than the previously used synthetic peptides, is well supported.
To date, little is known on the physiological and pathophysiological roles of PTHrP in the cardiovascular system. The finding of the present study that coronary endothelium releases PTHrP under hypoxia and energy-depleting conditions suggests that PTHrP influences vascular and contractile function of the heart during an early phase of reperfusion. With its vasodilatory effect, PTHrP may contribute to reactive hyperemia. Its positive inotropic effect may attenuate the extent of postischemic stunning.
In conclusion, the findings of this study indicate that in adult ventricular myocardium cardiomyocytes are physiological target cells for a paracrine action of endothelium-derived authentic PTHrP.
This study was supported by the Deutsche Forschungsgemeinschaft, Sonderforschungsbereich 547, project A1.
- Received October 14, 1999.
- Accepted February 8, 2000.
- © 2000 American Heart Association, Inc.
Deftos LJ, Burton DW, Brandt DW. Parathyroid hormone-like protein (PLP) is a secretory product of atrial myocytes. J Clin Invest. 1993;92:727–735.
Bui TD, Shallal A, Malik AN, Al-Mahdawi S, Moscoso G, Bailey MES, Burton PJB, Moniz C. Parathyroid hormone related peptide gene expression in human fetal and adult heart. Cardiovascular Res. 1993;27:1204–1208.
Schlüter K-D. PTH and PTHrP: Similar structures but different functions. News Physiol Sci. 1999;14:243–249.
Gaine SP, Hales MA, Flavahan NA. Hypoxic pulmonary endothelial cells release a diffusible contractile factor distinct from endothelin. Am J Physiol. 1998;274:L657–L664.
Kourembanas S, Marsden PA, McQuillan LP, Faller DV. Hypoxia induces endothelin gene expression and secretion in cultured human endothelium. J Clin Invest. 1991;88:1054–1057.
Piper HM, Spahr R, Mertens S, Krützfeld A, Watanabe H. Microvascular endothelial cells from heart. In: Piper HM, ed. Cell Culture Techniques in Heart and Vessel Research. Heidelberg: Springer; 1990:158–177.
Noll T, Muhs A, Besselmann M, Watanabe H, Piper HM. Initiation of hyperpermeability in energy-depleted coronary endothelial monolayers. Am J Physiol. 1995;268:H1462–H1470.
Schlüter K-D, Schwartz P, Siegmund B, Piper HM. Prevention of the oxygen paradox in hypoxic-reoxygenated hearts. Am J Physiol. 1991;261:H416–H423.
Schlüter K-D, Weber M, Piper HM. Parathyroid hormone induces protein kinase C but not adenylate cyclase in adult cardiomyocytes and regulates cyclic AMP levels via protein kinase C-dependent phosphodiesterase activity. Biochem J. 1995;310:439–444.
Schlüter K-D, Wingender E, Tegge W, Piper HM. Parathyroid hormone-related peptide antagonizes the action of parathyroid hormone on adult cardiomyocytes. J Biol Chem. 1996;271:3074–3078.