Isolation and Characterization of Coenzyme A Glutathione Disulfide as a Parathyroid-Derived Vasoconstrictive Factor
Background—Coenzyme A glutathione disulfide (CoA-SSG) was recently isolated from bovine adrenal glands and was shown to be a renal vasoconstrictor. The identification of CoA-SSG in human parathyroid glands and its action on cultured vascular smooth muscle cells (VSMCs) are described here.
Methods and Results—After purification to homogeneity by several chromatographic steps, CoA-SSG was identified by matrix-assisted laser desorption/ionization mass spectrometry and enzymatic analysis. The dose-dependent growth-stimulating effect of CoA-SSG on VSMCs, measured by the [3H]thymidine method, is characterized by a threshold of 10−8 mol/L and a maximum effect of 10 μmol/L, increasing VSMC proliferation 254±21% above control. A dose of 10 μmol/L methylmalonyl-CoA and 10 μmol/L CoA increased the rate of proliferation of VSMCs only by 178±43% and 50±42% above control, respectively. Glutathione has no proliferative effect on VSMCs. The growth-stimulating effect of CoA-SSG (1 μmol/L) was decreased by the antagonists 3,7-dimethyl-1-propargylxanthine (DMPX; 11 μmol/L) (38% compared with CoA-SSG without antagonist) and pyridoxal-phosphate-6-azophenyl-2,4-disulfonic acid (PPADS; 10 μmol/L) (48% compared with CoA-SSG without antagonist; each P<0.05 versus control), indicating that the effect is mediated partly via A2 and partly via P2Y1 and/or P2Y4 receptor.
Conclusions—CoA-SSG may play a regulatory role in VSMC growth as a progression factor and thereby could play an important role in development of hypertension.
The parathyroid glands have long been supposed to contain a yet unidentified vasoactive hormone. Pang et al1 and others2 showed that extracts of parathyroid glands from spontaneously hypertensive rats elicited vasoconstriction in several experimental models.
However, the so-called “parathyroid hypertensive factor” (PHF) has not yet been successfully identified. Interestingly, the PHF was shown to contain both peptide3 and nonpeptide components. Because convincing parathyroid-derived factor had been presented by those authors, we were interested in whether human parathyroid glands contain a recently identified vasoconstrictive substance, coenzyme A glutathione disulfide (CoA-SSG). In an earlier study, we isolated CoA-SSG from bovine adrenal glands and showed that this substance is a potent vasoconstrictor.4 In addition to its vasoactive action, the substance potentiates the effects of angiotensin II on vascular tone.
In the present study, we demonstrate that CoA-SSG is also found in human parathyroid glands. There is some probability that CoA-SSG is identical with the PHF, the existence of which had been shown only indirectly in earlier studies.
Furthermore, we extend our knowledge of the physiological action of CoA-SSG with respect to vascular smooth muscle cell (VSMC) growth. This may be of interest, because disturbances in VSMC growth are known to play an important role in cardiovascular disorders such as hypertension and atherosclerosis.5 6 The pathogenesis of hypertension is a process that involves both enhanced vasoconstriction5 and the remodeling of the arterial wall,7 which is characterized by hypertrophy and proliferation of VSMCs. Furthermore, it appeared likely that CoA-SSG exhibits not only vasoconstrictive but also proliferative actions, because several vasoconstrictors, such as angiotensin II,8 thromboxane,9 and ATP,10 are known to be potent VSMC growth factors as well.
All reagents were purchased from Sigma Chemical Co unless otherwise specified. Human parathyroid glands were obtained from patients undergoing parathyroidectomy for parathyroid adenoma or hyperplasia due to secondary or primary hyperparathyroidism. In all patients with secondary hyperparathyroidism, end-stage renal disease was present. The study was approved by the local ethical committee, and written consent was obtained from the patients. For isolation of CoA-SSG, only macroscopically intact tissue was used.
Purification of CoA-SSG From Human Parathyroid Gland Tissue
After excision from the patients, human parathyroid gland tissue (≈2 g) was immediately placed in ice-cooled physiological saline solution and processed within 30 minutes.
The following isolation procedure was designed to exclusively isolate CoA-SSG from human parathyroid glands. The human parathyroid gland tissue was cut into small pieces (≈1 cm3), frozen in liquid nitrogen, and stored at −80°C for 1 night. Then the tissue was lyophilized and powdered (step 1). The powder was suspended in 20 mL 0.6 mol/L ice-cold perchloric acid and homogenized at 25 000 rpm 10 times for 30 seconds. The homogenate was ultracentrifuged at 30 000 rpm for 60 minutes at 4°C. The supernatant was adjusted with KOH to pH 8.5 and stored at 4°C for 30 minutes to precipitate KClO4. After centrifugation at 4000 rpm for 10 minutes at 4°C, the supernatant was titrated to pH 6.5 with HCl and centrifuged again as above (step 2).
Preparative Reversed-Phase Chromatography
Extract of the human parathyroid gland tissue was concentrated on a C18 reversed-phase column (LiChroprep, 310×65 mm, 65 to 40 μm, Merck) with 40 mmol/L aqueous triethylammonium acetate (TEAA) in water (flow rate 2 mL/min). After nonbinding substances had been removed with aqueous 40 mmol/L TEAA, the remaining substances were eluted with 20% acetonitrile in water (flow rate 2 mL/min) (step 3). The elution was detected by UV absorption at 254 nm. The eluate was lyophilized and stored frozen at −80°C.
High-Performance Liquid-Displacement Chromatography
The lyophilized eluate of the preparative reversed-phase chromatography was dissolved in aqueous 40 mmol/L TEAA solution and injected onto a C18 reversed-phase column (Supersphere, 2.1×100 mm, 4 μm, Merck; step 4) that was equilibrated with aqueous 40 mmol/L TEAA as carrier as before. The carrier was pumped through the system at a flow rate of 50 μL/min during injection of the sample.
After the injection was finished, n-butanol (100 mmol/L) in 40 mmol/L TEAA was used as displacer (flow rate 50 μL/min). The displacement chromatography was monitored by UV absorption at 254 nm. The fraction size was 100 μL. Every fraction of the displacement chromatography was lyophilized and then further separated by anion-exchange chromatography.
Lyophilized fractions of the displacement chromatography were dissolved in 1 mL 20 mmol/L K2HPO4 in water (pH 8) (eluent A) and chromatographed by use of an anion exchanger (step 5) (column: Mono Q HR 5/5, 50×5 mm, 10 μm, Pharmacia Biotech; eluent B: 20 mmol/L K2HPO4 and 1 mol/L NaCl [pH 8] in water; gradient: 0 to 10 minutes: 0% to 5% B; 10 to 115 minutes: 5% to 40% B; 115 to 120 minutes: 40% to 100% B; flow rate: 0.1 mL/min; UV absorption wavelength: 254 nm).
Each peak of the anion-exchange chromatography with a retention time comparable to the retention time of authentic CoA-SSG was desalted by high-performance reversed-phase C18 liquid chromatography (step 6). The reversed-phase column (Supersphere, 250×4 mm, 4 μm, Merck) was equilibrated with 40 mmol/L aqueous TEAA in water. The sample dissolved in the same eluant was pumped with a flow rate of 0.5 mL/min onto the column. After the column had been washed with 10 mL 40 mmol/L TEAA in water, the substances were eluted with 35% acetonitrile in water at a flow rate of 0.5 mL/min. The resulting fractions were lyophilized.
Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry
The lyophilized fractions with a significant UV absorption at 254 nm from the reversed-phase chromatography were examined by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) and postsource decay (PSD)-MALDI-MS. A reflector-type time-of-flight mass spectrometer (Reflex III, Bruker) was used according to Hillenkamp and Karas.11 The sample was mounted on an x, y, z movable stage allowing irradiation of selected sample areas. In this study, a nitrogen laser (VSL-337 ND, Laser Science) with an emission wavelength of 337 nm and 3-ns pulse duration was used. The laser beam was focused to a diameter of typically 50 μm at an angle of 45° to the surface of a target. Microscopic sample observation was possible. Ten to 20 single spectra were accumulated for a better signal-to-noise ratio. In MALDI-MS, large fractions of the desorbed analyte ions undergo PSD during flight in the field free drift path. With a reflector-type time-of-flight setup, sequence information from PSD fragment ions of precursors produced by MALDI were obtained.12 Sample preparations for MALDI and PSD-MALDI experiments were identical. The concentrations of the analyzed substances were 1 to 10 μmol/L in bidistilled water. Analyte solution (1 μL) was mixed with 1 μL of matrix solution (50 mg/mL 3-hydroxy-picolinic acid in water). To this mixture, cation exchange beads (AG 50 W-X12, 200 to 400 mesh, Bio-Rad) equilibrated with H+ as counterion were added to remove Na+ and K+ ions. The mixture was gently dried on an inert metal surface before introduction into the mass spectrometer. The mass accuracy was in the range of ≈0.01%.
Enzymatic Analysis and Reduction With Mercaptoethanol
To test for the presence of disulfide bonds, 5 μL of the fraction was incubated with 1 μL glutathione reductase (EC 18.104.22.168, Type VII, Sigma; 10 mU/L) at room temperature for 30 minutes. In addition, 5 μL of the fraction was boiled with 5 μL mercaptoethanol for 20 minutes. The split products were identified by their molecular masses with the use of MALDI-MS and PSD-MALDI-MS, respectively.
Effect of CoA-SSG on Rate of Proliferation of VSMCs
Aortic VSMCs from normotensive Wistar-Kyoto rats were subcultured in 96-well dishes (Falcon) at a density of 5×104 cells/mL and kept in culture medium containing 10% FCS to reach a subconfluent monolayer. After 24 hours, the cells were growth-arrested in 0.1% FCS for 48 hours without affecting cell adherence to culture wells or viability as checked by trypan blue vital dye exclusion.
Quiescent VSMCs were then exposed to fresh culture medium with 0.1% FCS with CoA-SSG, CoA, methylmalonyl–coenyzme A, and glutathione for a 48-hour incubation period. Moreover, in the presence of CoA-SSG, the receptor antagonist pyridoxal-phosphate-6-azophenyl-2,4-disulfonic acid (PPADS, 10 μmol/L) and as a potent and selective A2 antagonist, 3,7-dimethyl-1-propargylxanthine (DMPX, 11 μmol/L)13 14 were added to the culture medium. Cell proliferation was measured by use of the [3H]thymidine incorporation rate as described elsewhere.15
Control Experiment: Incubation of CoA and Glutathione in the Presence of Perchloric Acid
CoA 1 μmol and glutathione 1 μmol were dissolved in 20 mL water. Perchloric acid was added to a final concentration of 0.6 mol/L. The sample was treated like the tissue extract described above. The supernatant passed the same chromatographic steps as described above. Fractions from the last chromatographic step with a significant UV absorption were analyzed by MALDI-MS.
The data were expressed as mean±SEM. The data of the proliferation experiments were analyzed by ANOVA. Statistical significance was assumed to be 2-tailed if P<0.05.
Figure 1A⇓ shows the chromatogram of a reversed-phase displacement chromatography with TEAA as an ion-pair reagent of the human parathyroid gland extract after extraction with perchloric acid and concentration on a reversed-phase chromatography gel. CoA-SSG eluted within 88 to 90 minutes, labeled in the figure by an arrow and a gray bar.
Each fraction of the displacement chromatography with a significant UV absorption at 254 nm underwent anion-exchange chromatography (step 5; Figure 1B⇑). Each fraction of the anion-exchange chromatography with a retention time comparable to that of CoA-SSG was desalted by reversed-phase chromatography (step 6). This lyophilized eluate of the reversed-phase chromatography fraction was subjected to the analytical procedures described below.
First, the molecular mass of the substance underlying the peak of the reversed-phase chromatography with a retention time comparable to that of authentic CoA-SSG was determined by MALDI-MS as 1073.4 Da (1074.6 Da=[M+H]+).
Figure 2⇓ shows the corresponding PSD-MALDI mass spectrum of the desalted fraction of anion-exchange chromatography, labeled by an arrow in Figure 1B⇑. The fragmentation pattern of the isolated substance obtained by this PSD-MALDI mass spectrum was identical to that from commercially available CoA-SSG.
Figure 3⇓ shows the MALDI mass spectra of the isolated fraction before (Figure 3A⇓) and after (Figure 3B⇓) incubation with glutathione reductase, indicating the presence of glutathione in the molecule and a disulfide bond of CoA and glutathione. Boiling with mercaptoethanol yields an identical MALDI spectrum.
To prove whether perchloric acid may oxidize CoA and glutathione, resulting in CoA-SSG, both substances were incubated with perchloric acid. The mixture passed the same purification procedure as the tissue extract. With MALDI-MS, no CoA-SSG was detectable.
CoA-SSG induced a dose-dependent increase in DNA synthesis in VSMCs, as determined by [3H]thymidine uptake (Figure 4⇓). The threshold of the growth-stimulating effect of CoA-SSG was 10−8 mol/L. The maximum effect was obtained at CoA-SSG concentrations of 10 μmol/L, which increased VSMC proliferation ≈254±21% above control. Methylmalonyl-CoA 10 μmol/L as well as CoA 10 μmol/L increased the rate of proliferation to 178±43% and 50±42%, respectively, above control. As shown in Figure 4⇓, significant effects on the proliferation rate of VSMCs were observed with 10 nmol/L CoA-SSG. Glutathione had no effect on the rate of proliferation of VSMCs. The threshold concentration for the direct growth-stimulatory effect of CoA-SSG and methylmalonyl-CoA was 10−8 mol/L, and for that of CoA 10−7 mol/L. The EC50 value (in log mol/L) of CoA-SSG was −8.48±0.15, that of methylmalonyl-CoA −8.15±4.3, and that of CoA −7.41±0.48.
The growth-stimulatory effect of CoA-SSG was inhibited by 2 different purinoceptor antagonists (Figure 5⇓). In the presence of 11 μmol/DMPX blocking the A2 receptor, the rate of proliferation decreased ≈38% compared with CoA-SSG alone (P<0.05 versus control). The growth-stimulating effect induced by CoA-SSG was, moreover, significantly inhibited by the P2 antagonist PPADS (rate of proliferation [% above control], CoA-SSG versus CoA-SSG+PPADS: 469±59 versus 244±71 [P<0.05 versus control]) (Figure 5⇓).
In the parathyroid gland, several known and unknown vasoactive hormones exist. The parathyroid hormone (PTH) causes dose-dependent transient vasodilatation in various vascular beds.16 Rat parathyroid cells were shown to synthesize and secrete endothelin I.17 In many studies, parathyroidectomy was demonstrated to decrease blood pressure.18 19 20 In patients who were hypertensive before operation, both the systolic and diastolic blood pressure significantly decreased after parathyroidectomy.21 Therefore, the existence of a vasopressor in human parathyroid glands possibly responsible for the development of primary hypertension is discussed. Pang et al1 suggested a parathyroid gland–derived hypertensive factor (PHF) according to the observation that parathyroidectomy decreased mean arterial blood pressure (MAP) in spontaneously hypertensive rats (SHR) and resulted in the disappearance of the hypertensive factor from the plasma. Transplantation of SHR parathyroid glands into Sprague-Dawley rats resulted in an increase in MAP and the appearance of the factor in the plasma, whereas transplantation of Sprague-Dawley parathyroid glands into SHR had exactly the opposite effect. Sprague-Dawley to Sprague-Dawley transplantation had no effect on MAP or hypertensive factor activity.22 In a further study, it was assumed that parathyroid gland can express PHF in humans and that such expression may be responsible for a proportion of the high reported incidence of hypertension in primary hyperparathyroidism.23 After partial characterization, the structure of the PHF was suggested to be a peptide linked to a lysophospholipid.3
In this study, we were able to isolate and identify CoA-SSG from human parathyroid glands, which was identified earlier from bovine adrenal glands.4 There is some probability that the substance isolated from human parathyroid glands by Schlüter et al24 in 1992 may be attributed to CoA-SSG.
From experiments in adrenal glands, it is known that CoA-SSG can be released by stimulation with carbachol4 or A23187.25 From these findings, it can be inferred that both an increase in cytosolic free Ca2+ concentration and cholinergic stimulation of adrenal medulla are important mechanisms of CoA-SSG release. For parathyroid gland tissue, however, these mechanisms can only be inferred by analogy at present.
Vascular actions of CoA-SSG have been demonstrated in previous studies. CoA-SSG has a direct effect on the renal vascular resistance. Intra-aortic injections of CoA-SSG increase blood pressure in intact animals.4 The vasopressor effect of CoA-SSG was demonstrated in both renal and mesenteric vasculature. CoA-SSG contracts renal vasculature, increases systemic blood pressure, and increases [Ca2+]i in VSMCs. Furthermore, with respect to the vasculature, an interesting interaction with angiotensin II was noted: The vasoconstrictor effect of angiotensin II on renal vasculature was potentiated in the presence of CoA-SSG.4
Is the PHF identical with CoA-SSG? Obviously, this question cannot be decided definitively at present. There are several similarities between the 2 agents, such as the peptide and nonpeptide moiety.26 In contrast, the hypertensive action of PHF is much more delayed and prolonged than that of CoA-SSG. However, the indirect effects of CoA-SSG, potentiating the vasoconstrictive effects of angiotensin II, show a latency similar to the vascular effect of PHF. With respect to both similar and different properties of the 2 agents, it may also be considered whether PHF is a derivative of CoA-SSG. Furthermore, it must be remembered that rat and human parathyroid glands may produce similar, but not identical, vasoactive factors. PHF has not been demonstrated in human parathyroid glands, and conversely, we did not examine rat parathyroid tissue.
At present it is not known whether CoA-SSG is produced or secreted in increased amounts in parathyroid adenomas or hyperplastic parathyroid glands. Several clinical observations may support a role of CoA-SSG in human pathology. First, primary hyperparathyroidism is accompanied by an increase in blood pressure, although PTH is a vasodilator.27 Given that part of the vascular changes in primary hyperparathyroidism are due to elevated serum Ca2+ levels, an increased secretion of CoA-SSG may present an alternative explanation for this clinical finding unexplained by PTH. Second, parathyroid hyperplasia in renal failure could lead to an increased release of CoA-SSG. Although PTH has been postulated to act as a uremic toxin,28 several features of uremic toxicity, such as neurotoxicity and the cardiovascular uremic changes, could never be thoroughly explained by the physiological actions of PTH. Therefore, it may be speculated that CoA-SSG could also play a role in secondary hyperparathyroidism, although at present, firm data are lacking.
In addition to the vasopressor properties of CoA-SSG, the results of the proliferation assay show that CoA-SSG has a direct effect on the rate of proliferation of VSMCs. The dose-dependent growth-stimulating effect of CoA-SSG is obviously determined by the CoA component of CoA-SSG (Figure 4⇑) and not by the glutathione component. Glutathione amplifies the growth-stimulating effect of the CoA component of CoA-SSG. The stimulating effect of CoA-SSG is mediated by a receptor that recognizes CoA.
As shown in Figure 5⇑, the growth-stimulating effect of CoA-SSG is partially inhibited by the potent and selective A2 antagonist DMPX13 as well as by the P2 antagonist PPADS.29 Because the P2X receptor is not involved in cell proliferation and, moreover, this receptor subtype is lost in cultured VSMCs,10 it seems reasonable that the mitogenic effect of CoA-SSG is mediated by the P2Y receptor. Because only the P2Y1 and P2Y4 subtypes of the P2Y receptor are inhibited by PPADS,30 31 it seems likely that the growth-stimulating effect of CoA-SSG is mediated by these P2 receptor subtypes. Both the sigmoidal concentration-dependency of the effects and the specific blockade by purinoceptor antagonists make a nonspecific proliferative effect rather unlikely.
The role of CoA-SSG in the regulation of vasomotor tone and growth is still largely speculative. First, the adequate stimulus to release CoA-SSG from an endocrine organ has not been defined. With respect to adrenal medulla, sympathetic stimulation may be one mechanism of release, because CoA-SSG has been found in exocytotic granules together with catecholamines.4 In parathyroid gland, the stimulus of CoA-SSG release is much more difficult to define, because we do not know whether calcium or phosphate homeostasis can affect CoA-SSG release in a manner similar to that of PTH. Nevertheless, some data indicate a role of parathyroid glands in vascular control. For example, calcium intake is known to affect blood pressure.32 This observation, among others, led to the concept of a PHF.33 From its biological properties, CoA-SSG can clearly be regarded as a parathyroid-derived hypertensive substance.
In summary, the findings show that human parathyroid glands contain a vasoconstrictive factor, as was suggested from studies in rat parathyroid tissue. In addition to its vasoconstrictive action, CoA-SSG also acts as a growth factor in VSMCs.
This work was supported by the Deutsche Forschungsgemeinschaft Schl 406/1–2. We thank A. Pacha for valuable technical assistance.
- Received March 21, 2000.
- Revision received May 23, 2000.
- Accepted June 8, 2000.
- Copyright © 2000 by American Heart Association
Pang PK, Lewanczuk RZ, Benishin CG. Parathyroid hypertensive factor. J Hypertens. 1990;8(suppl):S155–S159.
Kaneko T, Ohtani R, Lewanczuk RZ, et al. A novel cell type in the parathyroid glands of spontaneously hypertensive rats. Am J Hypertens. 1989;2:549–552.
Benishin CG, Lewanczuk RZ, Shan J, et al. Purification and structural characterization of parathyroid hypertensive factor. J Cardiovasc Pharmacol. 1994;23(suppl 2):S9–S13.
Schlüter H, Meissner M, van der Giet M, et al. Coenzyme A glutathione disulfide: a potent vasoconstrictor derived from the adrenal gland. Circ Res. 1995;76:675–680.
Griendling KK, Ushio-Fukai M, Lassegue B, et al. Angiotensin II signaling in vascular smooth muscle: new concepts. Hypertension. 1997;29:366–373.
Sachinidis A, Flesch M, Ko Y, et al. Thromboxane A2 and vascular smooth muscle cell proliferation. Hypertension. 1995;26:771–780.
Müller CE, Hipp J, Knoblauch B, et al. DMPX (3,7-dimethyl-1-propargylxanthine) derivatives: structure-activity relationships of potent selective A2a-adenosine receptor antagonists. Drug Dev Res. 1996;37:112.
Heidenreich S, Tepel M, Schlüter H, et al. Regulation of rat mesangial cell growth by diadenosine phosphates. J Clin Invest. 1995;95:2862–2867.
De Feo ML, Bartolini O, Orlando C, et al. Natriuretic peptide receptors regulate endothelin synthesis and release from parathyroid cells. Proc Natl Acad Sci U S A. 1991;88:6496–6500.
Kuczec A, Groβ I, Breuer E, et al. Coenzyme A glutathione disulfide released by A23187. Biol Chem Hoppe-Seyler. 1995;376:S123.
Benishin CG, Labedz T, Guo DD, et al. Identification and purification of parathyroid hypertensive factor from organ culture of parathyroid glands from spontaneously hypertensive rats. Am J Hypertens. 1993;6:134–140.
Ralevic V, Burnstock G. Receptors for purines and pyrimidines. Pharmacol Rev. 1998;50:413–492.
McCarron DA, Morris CD, Henry HJ, et al. Blood pressure and nutrient intake in the United States. Science. 1984;224:1392–1398.