This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hwang, K.-C. Articles by Im, M.-J. Search for Related Content PubMed PubMed Citation Articles by Hwang, K.-C. Articles by Im, M.-J.

(Circulation. 1996;94:718-726.)
© 1996 American Heart Association, Inc.

Articles

₁-Adrenergic Receptor Coupling With G_h in the Failing Human Heart

Ki-Chul Hwang, PhD; Caroline D. Gray, MS; Wendy E. Sweet, BS; Christine S. Moravec, PhD; Mie-Jae Im, PhD

the Department of Molecular Cardiology and Center for Anesthesiology Research (W.E.S., C.S.M.), Research Institute, the Cleveland (Ohio) Clinic Foundation.

Correspondence to Mie-Jae Im, PhD, Department of Molecular Cardiology (FFB-37), Research Institute, the Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Background We recently demonstrated that G_h, which transfers the signal from the ₁-adrenergic receptor to the 69-kD phospholipase C, is the previously identified tissue-type transglutaminase (TGase II). The ₁-adrenergic receptor mediates actions of the sympathetic nervous system, including cardiac, arteriolar, and smooth muscle contractions. In human cardiac tissue, the expression of the ₁-adrenergic receptor is increased under pathophysiological conditions, but changes in the physiological response are small. Therefore, it has been suggested that the other components involved in the ₁-adrenergic receptor–mediated signaling pathway are probably altered.

Methods and Results Immunological and biochemical studies with nonfailing and failing human heart tissues revealed that the GTP-binding and TGase activities of human heart TGase II (hhG_h) are downregulated in both ischemic and dilated cardiomyopathic human heart. In ischemic cardiomyopathy, the ₁-adrenergic receptor number increased twofold (27.0 fmol/mg) compared with the nonfailing (12.8 fmol/mg) and the dilated cardiomyopathic (15.6 fmol/mg) heart tissues, but the coupling of hhG_h with the ₁-adrenergic receptor did not increase. The intrinsic activity of hhG_h was greatly decreased in membrane fractions, whereas the cytosolic TGase activity was not changed. In the dilated cardiomyopathic human heart, these intrinsic enzyme activities of hhG_h were also downregulated in the membrane fraction, whereas the amount of hhG_h protein was greatly increased (2.8-fold) compared with the nonfailing heart.

Conclusions The results of the present study clearly demonstrate that the ₁-adrenergic receptor in human heart couples with G_h (TGase II) and indicate that downregulation of hhG_h activity is associated with human cardiac failure but that the mechanism differs between ischemic and dilated cardiomyopathies.

Key Words: heart failure • receptors, adrenergic, alpha • transglutaminase II • proteins • signal transduction

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Activation of cardiac ₁-adrenergic receptors leads to rapid changes in phosphoinositide breakdown, glycogenolysis, and positive inotropic effects, with an increase of free Ca²⁺ and of responsiveness of the myofilaments to Ca²⁺.^{1 2 3} Recent studies also suggest that ₁-agonists have long-term effects on cardiac structure and function by enhancing expression of several proto-oncogenes, c-myc, c-fos, and c-jun,^{4 5 6 7} as well as increasing the cellular contents of the cardiac myofibrillar proteins myosin light chain-2 and cardiac -actin.^{8 9 10 11} It has been postulated that in heart, under normal conditions, the ß-adrenergic receptors are the primary mediators of the effect of catecholamines, whereas the ₁-adrenergic receptors play a role during pathological development, such as ischemia, possibly acting as a reserve receptor system to maintain cardiac function.¹ Studies have also demonstrated that the ₁-adrenergic receptor density and its cellular response vary from species to species.^{1 2 12 13 14 15 16 17 18 19 20 21} Thus, in animals, the density of ₁-adrenergic receptor is severalfold higher than in humans and is correlated with a physiological responsiveness in both NF and failing models.¹⁵ In the human heart, however, especially in failing heart, these responses are not correlated with the numbers of receptors.^{14 17 18 19 20 21} Thus, in NF human hearts, ₁-adrenergic receptor stimulation increases the force of contraction more than twofold. In the failing human heart, such as hearts with end-stage ischemic heart disease, the ₁-receptor–mediated positive inotropic effect is usually smaller, although the absolute number of the ₁-adrenergic receptors did not change and actually even increased during the development of cardiac failure.^{1 12 17} Therefore, it has been suggested that components other than the receptor change in function or expression in the ₁-adrenergic receptor signaling system of the failing human heart.

The ₁-adrenergic receptors are members of the GTP-binding regulatory protein (G protein)–coupled receptor superfamily and mediate a variety of intracellular responses, including activation of PLC, A₂, and D as well as activation of calcium channels and Na⁺-H⁺ and Na⁺-Ca²⁺ exchange.^{1 12 22} To date, three subtypes of ₁-receptors have been identified.²³ Unlike other receptors, expression of the ₁-receptors varies markedly in distribution in different tissues and species.^{12 22} For example, in rat heart, the _1B-receptor is predominant, with minor amounts of other subtypes of the ₁-receptor, whereas in human heart, the _1C-receptor (now _1A-type, see Reference 23) is predominant, with minor amounts of the _1B- and _1D-subtypes, which have been identified by mRNA level.²⁴ Expression of the ₁-receptor subtypes in eukaryotic cells revealed that the same subtype of the ₁-receptor uses different G proteins depending on the cell type, suggesting that cross talk between the receptor and G proteins exists in the ₁-adrenergic receptor transmembrane signaling²⁵ (see also Reference 22).

G_h is a new class of GTP-binding regulatory protein (G protein) that was initially identified by the functional formation of ₁-agonist receptor–G_h ternary complex in rat liver.^{26 27} Reconstitution studies with ₁-receptor, G_h or G_h7, and 69-kD PLC demonstrated that _1B-adrenergic receptor activates 69-kD PLC through G_h.²⁸ Coexpression of the _1B-receptor with G_h or G_q also increased phosphoinositide breakdown,^{29 30} confirming previous observations. Formation of the ₁-receptor/G protein complex and coimmunoprecipitation studies using membranes from heart tissues of different species showed that the majority of ₁-receptors were coupled with G_h.³¹ Supporting this notion, Braun and Walsh³² also observed that the rat cardiac ₁-receptor coupled to 75-kD G_h.

Amino acid sequence and cloning of rat G_h revealed that the -subunit (G_h) of G_h is a previously identified tissue-type TGase (TGase II).²⁹ These findings imply that the classic TGase II has two important intrinsic enzyme activities, namely GTPase and TGase, including a signal transfer role. Thus, G_h binds and hydrolyzes GTP and functions as a signal mediator in hormone signal transduction, which is similar in function to the classic heterotrimeric G proteins.^{26 27 28 29 30 31 32} The TGase II is a calcium- and thiol-dependent enzyme that catalyzes an acyl-transfer reaction between the -carboxamide group of glutamine residues in proteins and primary amines of polypeptides.³³ The physiological functions of TGase II are yet to be explored. However, several studies have suggested that TGase II plays a role in the stabilization of cell-cell interaction^{34 35 36 37 38} as well as the regulation of cell growth and differentiation.^{39 40 41 42 43} In addition, TGase II activates cytosolic phospholipase A₂⁴⁴ and modifies contractile proteins, such as myosin heavy chain and -actin, in smooth muscles.⁴⁵ The multifunctional nature of G_h provides new insight into hormone signal transduction and the diversity of cellular functions.

To understand the multifunctional properties of human heart TGase II (hhG_h) and the ₁-adrenergic receptor–mediated signal transduction in NF and failing human cardiac tissues, we studied the functional coupling ability of the ₁-adrenergic receptor with hhG_h and the GTP-binding and TGase activities of hhG_h, including the hhG_h protein level. For the first time, we demonstrate here a possible linkage between human heart G_h and human cardiac failure.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Human Heart Tissue
Failing human heart tissue was obtained from the explanted hearts of cardiac transplant recipients at the Cleveland Clinic Foundation. Each heart was obtained in the operating room and stored in a cold cardioplegic solution containing (in mmol/L) NaCl 147.2, MgCl₂ 16, KCl 20, NaHCO₃ 10, and CaCl₂ 2.25. The heart in the cold cardioplegic solution was transferred to the laboratory. Time from explant until arrival in the laboratory ranged from 30 to 105 minutes. The heart was separated by chamber and immediately frozen in liquid nitrogen. Heart tissue was stored at -80°C until use. NF human heart tissue was obtained from the hearts of unmatched organ donors. Each heart was obtained in the operating room and stored in the same cardioplegic solution as described above. The heart in the cold cardioplegic solution was transferred from the donor hospital to the Cleveland Clinic Foundation. Time from explant until arrival in the laboratory ranged from 60 to 120 minutes. The heart tissue was immediately separated by chamber and frozen in liquid nitrogen. Heart tissue was stored at -80°C until use. For these studies, only left ventricular tissue was used. Random tissue samples from NF (9 subjects), ICM (10 subjects), and DCM (10 subjects) were used. The patient characteristics are shown in Table 1.

View this table: [in this window] [in a new window]   Table 1. Characteristics of Hearts From Patients and Donors

Separation of Pellets and Cytosol
The tissue was placed in a 20 mmol/L HEPES buffer (pH, 7.4) containing 1 mmol/L EDTA, 1 mmol/L EGTA, 1 mmol/L DTT, 10% glycerol, and protease inhibitors (PMSF 10 µg/dL, bacitracin 2 µg/dL, benzamidine 100 µg/dL, leupeptin 2 µg/dL, pepstatin A 2 µg/dL, soybean trypsin inhibitor 2 µg/dL, and antipain 20 µg/dL) and was briefly washed once with the same buffer. The tissue was minced and homogenized in an ice bath with a mechanical homogenizer, filtered through five layers of gauze, and then further homogenized in a Dounce homogenizer. Pellets and cytosol were separated by centrifugation at 100 000g for 1 hour. The pellets were washed once and resuspended in the same buffer. The pellet and cytosol fractions were divided into aliquots and stored at -80°C until use. Protein concentrations in the pellets and cytosol were measured with the D_C protein assay (Bio-Rad) with BSA as a standard.

Measurement of ₁-Agonist–Mediated Inhibition of TGase Activity of hhG_h
Pellets were washed once with HSD buffer and preincubated with or without 5x10^-6 mol/L (-)-epinephrine in the presence of ₂- and ß-receptor blockers, rauwolscine (10^-7 mol/L), and propranolol (10^-6 mol/L) at 30°C for 30 minutes. Pellets were also preincubated in the presence of the ₁-agonist and 10^-4 mol/L phentolamine under the same conditions. The ₁-agonist-mediated inhibition of the TGase activity was measured in the pellets (150 µg per assay) in HSD buffer in the presence of 0.35 mmol/L CaCl₂, 0.1 mmol/L GTP, 1 mmol/L MgCl₂, 0.1 mmol/L [³H]putrescine, and 1% N,N'-dimethylated casein at 20°C for 30 minutes in 100 µL final volume. Under the same conditions, Ca²⁺-mediated activation and GTP-mediated inhibition of the TGase activity were determined with 0.35 mmol/L CaCl₂ or 0.35 mmol/L CaCl₂ plus 0.1 mmol/L GTP in the presence of 1 mmol/L MgCl₂, respectively. After incubation, the TGase activity was analyzed by determination of the incorporation of radiolabeled putrescine into N,N'-dimethylated casein, as described below. All samples were simultaneously subjected to the experiments to minimize variability, and all experiments were repeated three times.

Measurement of TGase Activity
The TGase activity in human heart pellets and cytosol (50 µg per assay) was determined by evaluating the incorporation of [³H]putrescine (0.1 mmol/L) into N,N'-dimethylated casein (1%) at 30°C for 30 minutes in HSD buffer containing 0.05% sucrose monolaurate in 100 µL final volume.^{33 36} The calcium concentration was adjusted by the addition of CaCl₂. After incubation, the samples were transferred to an icewater bath, and the reaction was stopped by the addition of ice-cold 50% trichloroacetic acid (100 µL) containing 0.1% putrescine. The precipitates were trapped on GF/F glass fiber filters (Whatman). The GTP-mediated inhibition of the TGase activity was determined under the same conditions in the presence of 0.5 mmol/L GTP and 2 mmol/L MgCl₂. Changes are specified in the legends.

Measurement of ₁-Adrenergic Receptor Density and GTP Binding
The ₁-adrenergic receptor density in the pellets was measured according to the methods of Im and Graham.²⁶ The pellets (150 µg protein) were incubated with 4x10^-9 mol/L [³H]prazosin (specific activity, 71.6 Ci/mmol) in the presence or absence of 10^-4 mol/L phentolamine in HSD buffer containing 2 mmol/L MgCl₂ at 30°C for 45 minutes. The reaction volume was 250 µL. The radioligand-bound ₁-receptors were trapped on GF/F glass filters that were incubated with the same buffer containing 100 µg BSA for 2 hours. The GTP-binding activity of hhG_h was determined by the method previously described.^{26 27 28 29 31} Pellets and cytosol preparations (100 µg per assay) were incubated with 10 to 20 µCi [-³²P]GTP in the presence of 2 mmol/L MgCl₂ in HSD buffer containing 0.05% sucrose monolaurate. After incubation at room temperature for 20 minutes, the samples were subjected to UV irradiation in an icewater bath for 10 minutes. The GTP binding of hhG_h was then analyzed with a densitometer (Fotodyne, Inc), after SDS-PAGE and autoradiography. The intensity of the radiolabeled hhG_h was converted into protein concentration on the basis of the known concentration of the radiolabeled GPL G_h.

Western Blotting
Immunoblotting was performed by the method of Baek et al³¹ (see also References 29 and 30). Human heart pellets and cytosol were subjected to 8% to 10% SDS-PAGE. The resolved proteins were then transferred to Immobilon-P by transverse electrophoresis in a buffer (Tris-HCl-methanol) at 0.5 A/h for 2 hours. The membranes were incubated in a blocking buffer (50 mmol/L Tris-HCl, pH 8.0, 2 mmol/L CaCl₂, 80 mmol/L NaCl, and 0.02% sodium azide) containing 0.2% Nonidet P-40 and 5% nonfat dry milk. After a 2-hour incubation at room temperature, the membranes were further incubated in the same buffer containing specific G_h7 antibody for 1 hour at room temperature and then washed three times (10 minutes per wash) with the same buffer containing 2% Nonidet P-40. The membranes were incubated with the secondary antibody (anti-rabbit Ig, horseradish peroxidase–linked whole antibody, Amersham) for 1 hour at room temperature. After being washed with high-detergent buffer, the membranes were treated with chemiluminescence reagent (DuPont NEN). The dried membranes were subjected to Kodak XAR-5 film. To minimize the variability of the experiments, all samples were analyzed at the same time, along with a known concentration of GTP-agarose–purified GPL G_h, and exposure time was also fixed (5 minutes). The bands on the films were then analyzed by densitometry.

Immunoprecipitation
For immunoprecipitation studies, G_h7 antibody–protein A–agarose was prepared. The antisera (1 mL) were mixed with the same amount of protein A–agarose suspended in 10 mL of 0.1 mol/L borate buffer (pH 8.2) for 45 minutes at room temperature. The immunoaffinity resins were then washed intensively with HSD buffer and equilibrated with the same buffer. Immunoprecipitation of hhG_h by G_h7 antibody was performed with the immunoaffinity resins. The pellets (5 mg/mL protein) from NF, ICM, and DCM were extracted with 0.4% sucrose monolaurate at 4°C for 1 hour. The membrane extracts (200 µg protein) were incubated with 20 µCi [-³²P]GTP and 2 mmol/L MgCl₂ at 30°C for 30 minutes and then irradiated in an icewater bath with 254-nm UV light for 8 minutes.³¹ The samples were incubated with the G_h7 antibody–protein A–agarose (50 µL) with gentle rotation at 4°C for 2 hours. The reaction mixture was 250 µL final volume. The samples were centrifuged at 2000 rpm for 10 minutes. The pellets were washed three times with HSD buffer containing 0.05% sucrose monolaurate and subjected to SDS-PAGE (8% gel) and autoradiography overnight.

Statistical Analysis
Data were analyzed in three ways by a professional statistician from the Department of Biostatistics at the Cleveland Clinic Foundation: (1) A general analysis was done to obtain values of mean and SD. (2) Data were then subjected to ANOVA to compare groups on each of the sample measurements of interest. Results were considered statistically significant at a value of P.05. (3) For significant differences between NF, ICM, and DCM, Tukey's multiple comparison procedure was used as a follow-up analysis for identification of pairwise differences.

Materials
GPL G_h II, GTP-agarose, and N,N'-dimethylated casein were obtained from Sigma. Nucleotides were purchased from Boehringer Mannheim. Sucrose monolaurate (SM-1200) was a gift from the Mitsubishi-Kasei Co. [³H]Putrescine (44.4 Ci/mmol), [-³²P]GTP (3000 Ci/mmol), and [³H]prazosin (76 Ci/mmol) were purchased from DuPont NEN. Other chemical and biochemical materials were used as described previously.^{26 27 28 29 30 31}

	Results

Top Abstract Introduction Methods Results Discussion References

 
₁-Adrenergic Receptor Density and Its Coupling Ability With hhG_h
The ₁-adrenergic receptor density in pellets from NF, ICM, and DCM was ascertained by measurement of the ability of prazosin, a specific ₁-adrenergic receptor ligand, to bind the receptor. As shown in Fig 1, the ₁-adrenergic receptor density in NF was 12.77±3.70 fmol/mg protein. A significant increase in the receptor number was observed in ICM, with 27.0±6.79 fmol/mg protein, whereas DCM showed no change, with 15.6±5.79 fmol/mg protein of the ₁-receptor compared with NF. An increase in the ₁-adrenergic receptor density in ICM is consistent with the findings in animal and human heart tissue by others.^{1 12 13 14} In addition, it is of interest to note that the receptor density obtained from our experiments with all human heart tissues was approximately threefold higher than previous observations with the same tissues.¹⁴ The reason for the increase in the receptor number in all human hearts is most likely the stability of the proteins achieved by the rapid freezing of freshly excised heart tissue.

View larger version (18K): [in this window] [in a new window]   Figure 1. The 1-adrenergic receptor density in failing and NF human hearts. Number of 1-adrenergic receptors was determined in the pellets as detailed in "Methods." To measure the nonspecific binding, 10-4 mol/L phentolamine was used. Data represent means of triplicate experiments for each sample. Solid bars indicate the average of the 1-receptor density of the subjects in NF, ICM, and DCM. **Group that is statistically significantly different among groups, P<.001 by ANOVA after Tukey's comparison.

Activation of ₁-adrenergic receptor facilitates GTP binding to G_h by accelerating GDP/GTP exchange,²⁶ and the TGase activity of G_h is inhibited on binding of GTP to G_h or activation of the ₁-receptor.^{29 30} Therefore, the coupling ability of hhG_h to the ₁-adrenergic receptor can be evaluated by measuring the ₁-agonist–mediated inhibition of TGase activity. To assess functional interaction of hhG_h with the ₁-adrenergic receptor, the pellets were preincubated with or without epinephrine and epinephrine plus phentolamine. As presented in Fig 2A and 2B, the Ca²⁺-mediated TGase stimulation was effectively inhibited in all samples in the presence of GTP alone or by activation of the ₁-adrenergic receptor. Thus, the GTP-mediated inhibition of Ca²⁺-stimulated TGase activity was 24% in NF, 15% in ICM, and 17.5% in DCM, showing the GTP binding to hhG_h. Activation of the ₁-adrenergic receptor resulted in further inhibition of TGase activity, by 52.5% with NF, 38.1% with ICM, and 47.2% with DCM, demonstrating that on activation of the ₁-receptor, formation of the GTP-bound hhG_h was accelerated. These data clearly showed that coupling of hhG_h to the ₁-receptor effectively occurred. Moreover, the ₁-agonist–mediated inhibition of TGase activity was reversed in all samples by incubation with the ₁-antagonist phentolamine up to the level of GTP alone, demonstrating that the inhibition of the TGase activity is specifically mediated by the ₁-adrenergic receptor.

View larger version (31K): [in this window] [in a new window]   Figure 2. Coupling of human heart TGase II to the 1-adrenergic receptor. The 1-agonist–mediated inhibition of Ca2+-stimulated TGase activity was determined in the pellets (150 µg per assay) at 22°C for 30 minutes, as detailed (see "Methods"). A, Individual values of the subjects in each group. Data presented are the average values of three independent experiments carried out in duplicate. The value of Ca2+-mediated stimulation of TGase II activity in each sample was taken as 100% and compared with the value of the GTP-, 1-agonist–, and 1-antagonist–mediated inhibition of TGase II activity. , GTP-mediated inhibition of the TGase activity; , (-)-epinephrine (Epi)–mediated inhibition of the enzyme activity; and , TGase activity in the presence of epinephrine, GTP, and phentolamine. B, Mean±SD of data shown in A: TGase activity in the presence of GTP (open bars), (-)-epinephrine plus GTP (solid bars), and (-)-epinephrine plus phentolamine plus GTP (hatched bars). **Group that is statistically significantly different among groups, P=.001. Arrow indicates (-)-epinephrine–mediated inhibition of TGase activity with ICM.

When the specific ₁-agonist–mediated inhibition of TGase was compared, the inhibition in ICM was 27.4% and 19.3% lower than those in NF and DCM, respectively, suggesting that the coupling of hhG_h with the ₁-receptor is impaired in this failing heart. These results also suggested that the amount of functionally active hhG_h is decreased in ICM, since the ₁-antagonist effectively reversed the agonist-mediated TGase inhibition with all samples up to the level of the GTP-mediated TGase inhibition. The GTP-mediated TGase inhibition in ICM and DCM was not significantly different between groups.

Levels of hhG_h Protein in Pellets and Cytosol
Protein levels of hhG_h were determined in pellets and cytosol by immunoblotting with a polyclonal anti-G_h7 antibody.³¹ Fig 3 presents immunoblotting data that show the concentration dependence of GPL G_h (Fig 3A and 3B) and typical immuno–cross-reactivity of the 80-kD hhG_h in pellets and cytosol from NF and failing heart tissue (Fig 3C and 3D). The intensity of the autoradiogram was linear in a protein concentration–dependent manner, showing that the immunoblotting measurements are quantitative. The immunoblotting data from NF and failing heart tissue showed different patterns in pellets and cytosol from individual subjects. The accuracy of the protein level measurements by immunoblotting was further assessed by comparing TGase activity with the hhG_h level in NF. Fig 4A shows the correlation of the amounts of hhG_h with the TGase activity in pellets and cytosol fractions from NF. The amounts of hhG_h protein and the TGase activity showed a good correlation in the cytosolic fraction. The data also showed that the levels of hhG_h and the TGase activity were significantly higher in cytosol than in the pellets. Fig 4B presents the relative levels of hhG_h protein from NF, ICM, and DCM. The actual hhG_h levels (expressed as nanograms per milligram protein, mean±SD) in pellets from ICM and DCM were increased by 88.9% (544.4±104.4 ng/mg) and 181.0% (809.9±132.1 ng/mg), respectively, compared with those in NF (288.2±57.9 ng/mg). Conversely, in the cytosol from ICM, the hhG_h level was decreased substantially, by 56.9% (303.5±59.3 ng/mg) compared with that in NF (704.1±175.5 ng/mg) and by 51.1% compared with that of DCM (619.9±112.3). The hhG_h level in the cytosol from DCM was similar to the level in NF. Moreover, when the protein concentrations in pellet and cytosolic fractions were measured, the total amounts of protein in both failing hearts were not significantly changed compared with that in NF. Thus, the protein concentration in the pellets from DCM was increased slightly, by 2% to 4%, but in the cytosol the protein concentration was decreased slightly, by 8%. In ICM, the protein concentrations in pellets and cytosol were similar (2% to 5% increase) to those of the NF heart tissues. Taken together, these results clearly demonstrated that cellular distribution of hhG_h was changed in both groups of failing human hearts and that the expression level of hhG_h was increased in DCM.

View larger version (37K): [in this window] [in a new window]   Figure 3. Measurement of hhGh protein in NF and failing human heart tissues by polyclonal anti-Gh7 antibody. A, Autoradiogram of an immunoblot of GPL Gh in a concentration-dependent manner using Gh7 antibody. Various concentrations of the GTP-agarose–purified GPL Gh were applied to immunoblotting after SDS-PAGE (8% gel). The immuno–cross-reactivity was visualized by autoradiography after the blots were treated with chemiluminescence, as detailed in "Methods." B, Immunoblotting with various concentrations of Gh. Autoradiogram of the immunoblotting in Fig 3A was analyzed by measurement of optical density with a densitometer. C and D, 80-kD hhGh protein in pellets (C) and cytosol (D). Randomly chosen samples (100 µg per lane) from NF, ICM, and DCM were subjected to immunoblotting with Gh7 antibody after SDS-PAGE (8% gel). Cross-reactivity of the antibody was determined as described in "Methods."

View larger version (42K): [in this window] [in a new window]   Figure 4. Level of hhGh in pellets and cytosol from NF and failing human heart tissues. A, Comparison of the level of hhGh with the TGase activity in pellets and cytosol from NF. The level of hhGh was determined by immunoblotting with Gh7 antibody, and the TGase activity was measured as described in "Methods." The TGase activity is presented as 100 µg protein per assay to compare the level of hhGh in 100 µg protein of pellet and cytosol fractions. B, hhGh protein levels in pellets and cytosol from NF, ICM, and DCM were determined by immunoblotting with a polyclonal anti-Gh7 antibody.29 31 The intensity of the band on the film was analyzed by densitometry. The optical density (OD) is the intensity of hhGh in 100 µg pellet or cytosolic proteins. Data are mean values of three independent experiments for each sample. Solid bars are the averaged value of samples from NF, ICM, or DCM. *Group that is statistically significantly different compared with NF, P<.001; **group that is statistically significantly different among groups, P=.001.

GTP-Binding Activity of hhG_h
Our laboratory first applied the photoaffinity labeling method to identify the GTP-binding protein, G_h, that coupled to the ₁-adrenergic receptor in rat liver membranes.²⁶ Although the efficiency of direct labeling of G_h with [-³²P]GTP is low (10% to 20%), the labeling of G_h was concentration dependent (Fig 5A and 5B; see also Reference 27) and is thus a powerful tool for measuring the GTP-binding activity of specific GTP-binding proteins. Moreover, this method allows specific G proteins to be determined in tissues in which other methods, such as [³⁵S]GTPS binding, would be difficult. Therefore, we have applied direct photoaffinity labeling to evaluate the GTP-binding activity of hhG_h. Fig 6A and 6B presents typical [-³²P]GTP photoaffinity labeling of GTP-binding proteins in pellets and cytosol from randomly selected NF and failing heart tissues. In the pellets (Fig 6A), six distinct GTP-binding proteins were clearly visualized by photolabeling, including an 80-kD GTP-binding protein. Identity of the 80-kD protein was evaluated by immunoprecipitation studies using a G_h7 antibody. Thus, protein A–agarose–bound G_h7 antibody was incubated with [-³²P]GTP–labeled membrane extracts. Coimmunoprecipitation of the radiolabeled 80-kD protein was evident (Fig 6C), demonstrating that the 80-kD GTP-binding protein was indeed the 80-kD GTP-binding hhG_h. In the cytosol from NF, ICM, and DCM, hhG_h was also detected, together with several other GTP-binding proteins (Fig 6B). The radiolabeling of hhG_h was higher overall (2- to 5-fold) in pellets than in the cytosol. The reason for the difference in [-³²P]GTP–labeling efficiency between the cytosol and pellets is not clear yet. However, a conformational change may occur in association with plasma membranes that leads hhG_h to exhibit high affinity for GTP. In turn, hhG_h functions as a signal mediator. It is also possible that in the cytosol, hhG_h may associate with GTP-binding inhibitory proteins to function as TGase.

View larger version (29K): [in this window] [in a new window]   Figure 5. A, Concentration-dependent photoaffinity labeling of the purified GPL Gh with [-32P]GTP. The GTP-agarose–purified GPL TGase II was incubated with 10 µCi [-32P]GTP, 2 mmol/L MgCl2, and 0.1 mmol/L adenylylimidodiphosphate at 30°C for 20 minutes and subjected to UV irradiation (250 nm) for 8 minutes in an icewater bath. The radiolabeled GPL Gh was exposed on a Kodak X-OMAT film overnight after SDS-PAGE (8% gel). B, Concentration-dependent increase of [-32P]GTP photoaffinity labeling. Optical density of the autoradiogram of Fig 5A was measured by densitometry.

View larger version (70K): [in this window] [in a new window]   Figure 6. Photoaffinity labeling of the GTP-binding proteins in pellets (A) and cytosol (B) from NF, ICM, and DCM. The pellets and cytosol fractions (100 µg per lane) were incubated with 20 µCi [-32P]GTP in the presence of 2 mmol/L MgCl2 and 0.2 mmol/L adenylylimidodiphosphate at 30°C for 20 minutes and subjected to UV irradiation (250 nm) for 8 minutes in an ice bath. Overnight autoradiographs for the radiolabeled GTP-binding proteins followed SDS-PAGE (8% gel) as detailed in "Methods." Arrows show hhGh. C, Autoradiogram of immunoprecipitation of the [-32P]GTP–labeled 80-kD hhGh in pellets from NF, ICM, and DCM by Gh7 antibody. The membrane extracts (200 µg protein/50 µL) of pellets from randomly arranged NF, ICM, and DCM were incubated with 20 µCi [-32P]GTP and 2 mmol/L MgCl2 at 30°C for 30 minutes and then irradiated in an ice bath under 254 nm UV for 8 minutes, as detailed in "Methods." The samples were incubated with the Gh7 antibody–protein A–agarose (50 µL) with gentle rotation at 4°C for 2 hours. After immunoprecipitation, the pellets were washed three times and then subjected to SDS-PAGE (8% gel) after autoradiography overnight.

Fig 7 presents the specific GTP-binding activity of hhG_h in pellets and cytosol from NF, ICM, and DCM. The GTP-binding activity of hhG_h in the pellets from NF was 5.6-fold higher than in the cytosol. The GTP-binding activity in the pellets from ICM and DCM tissues was significantly decreased compared with NF (172.7±33.4). Decreases in GTP binding by hhG_h were 51.8% (83.2±39.7) in ICM and 59.1% (70.6±28.2) in DCM. These data clearly revealed that the GTPase function of hhG_h was decreased in pellets of both failing human heart tissues. Conversely, the GTP-binding activity in the cytosol was increased 41% in ICM (50.7±21.1) compared with that of NF (35.3±8.4) and DCM (35.9±11.1). The binding activity in DCM was similar to that in NF.

View larger version (20K): [in this window] [in a new window]   Figure 7. The GTP-binding activity of hhGh was determined by photoaffinity labeling of hhGh with 20 µCi [-32P]GTP as described in the legend of Fig 6. Autoradiograms were analyzed by densitometry. Data shown for the specific GTP binding of hhGh were calculated by dividing the optical density value of the [-32P]GTP–labeled hhGh band by the optical density value of the immunoblotting data of hhGh in 100 µg per lane of pellet and cytosol fractions. Solid bars are an averaged value for all samples from NF, ICM, or DCM. *Group that is statistically significantly different compared with NF, P<.001; **group that is statistically significantly different among groups, P=.04.

TGase Activity of hhG_h in Pellets and Cytosol
The other function of hhG_h, the TGase activity, was then evaluated. To optimize Ca²⁺ concentration, Ca²⁺ concentration–dependent TGase activity was measured by use of 50 µg of protein from the pellets and cytosol of three samples from NF, ICM, and DCM. The TGase activities in both pellet and cytosol fractions increased as Ca²⁺ concentration increased up to 1 mmol/L Ca²⁺ and subsequently reached maximal activity and sustained the activity up to 5 mmol/L Ca²⁺. A further increase of Ca²⁺ concentration inhibited the TGase activity. Based on the calcium titration curve, the TGase activity in pellets and cytosol was measured by use of 2 mmol/L CaCl₂. To determine whether the TGase activity observed originated from hhG_h, GTP-mediated inhibition of the TGase activity was also evaluated with all samples. The results showed complete inhibition of enzyme activity in the presence of 0.5 mmol/L GTP, indicating that all TGase in human heart is the 80-kD hhG_h (data not shown). Fig 8 presents the specific TGase activity in pellets and cytosol from NF, ICM, and DCM. The enzyme activity in the pellets from both groups of failing heart tissues was again significantly decreased, as observed with the GTP-binding activity of hhG_h (see Fig 7). Thus, the decrease in the TGase activity was 38.0% in ICM (75.1±21.7 µmol·mg^-1·min^-1) and 50.4% in DCM (60.1±19.7 µmol·mg^-1·min^-1) compared with that in NF (121.2±21.2 µmol·mg^-1·min^-1). However, the TGase activity in the cytosol from both groups of failing hearts (145.0±23.0 µmol·mg^-1·min^-1 in ICM and 152.6±21.5 µmol·mg^-1·min^-1 in DCM) was similar to that in the cytosol from NF (139.7±11.5 µmol·mg^-1·min^-1), demonstrating that the cytosolic hhG_h is intact in its TGase function in both groups of failing hearts. These results showed that the TGase function of membrane-associated hhG_h only was downregulated in both groups of failing heart tissues.

View larger version (18K): [in this window] [in a new window]   Figure 8. Specific TGase activity in the pellets and cytosol from NF, ICM, and DCM. The TGase activity was measured with 50 µg protein from the pellets and the cytosol in the presence of 2 mmol/L CaCl2 at 30°C for 30 minutes as detailed in "Methods." Specific activity is expressed as µmol·mg-1·min-1. Data presented for the specific TGase activity were obtained by normalizing to amounts of hhGh in 1 mg of protein in pellet or cytosol fractions. *In pellets: P<.001. The actual amount of hhGh in each sample was calculated from known concentrations of GPL Gh that were simultaneously subjected to immunoblotting, as described in the legend of Fig 3. *Group that is statistically significantly different compared with NF, P<.001.

To determine changes of hhG_h in pellets and cytosol, the ratio of hhG_h protein, GTPase, and TGase activities of hhG_h in membrane versus cytosol from both failing heart groups was compared with those of NF as shown in Table 2. The ratio of the protein levels was increased 6.6-fold in ICM and 3.2-fold in DCM, indicating that the cellular distribution of hhG_h was altered in both failing heart groups. Conversely, the ratio of the TGase activity was decreased 41.4% with ICM and 55.2% with DCM. The ratio of GTP-binding activity of hhG_h was also greatly altered in both failing heart groups. These results clearly demonstrated that changes in cellular distribution of hhG_h and downregulation of intrinsic hhG_h functions in membranes occurred in ICM and DCM.

View this table: [in this window] [in a new window]   Table 2. Comparison of hhGh Protein and Its Activity in Pellet vs Cytosolic Fraction From NF, ICM, and DCM

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
₁-Adrenergic receptors mediate a variety of sympathetic nervous system responses, such as vascular smooth muscle and cardiac contractions, and thus are involved in circulatory homeostasis. The ₁-receptor signaling also plays an important role in heart failure, affecting vasoconstriction, ventricular filling pressures, and systemic vascular resistance.^{46 47} These effects may lead to development of hypertrophy, necrosis, and fibrosis, which are often seen with human heart failure and in animal models of norepinephrine-induced cardiomyopathy.⁴⁸

In this study, we focused on a new class of GTP-binding protein, G_h, that possesses GTPase and TGase activities, including its functional coupling ability with the ₁-adrenergic receptor. To understand and dissect the pathogenic involvement of G_h, changes in hhG_h functions were evaluated in pellet and cytosolic fractions from NF and failing human heart tissues. The TGase activity of hhG_h exhibited a good correlation with the level of hhG_h protein in the pellets and cytosol from NF (Fig 4A), showing that data obtained from failing human hearts are valid. The results from quantification of GTP-binding activity with [-³²P]GTP show that [-³²P]GTP labeling of GPL G_h was increased in a protein concentration–dependent manner, indicating that the data obtained from the studies were also reliable (Fig 5).

Our studies clearly demonstrated that in human heart, G_h couples to ₁-adrenergic receptors, as demonstrated by inhibition of the TGase activity of hhG_h in response to activation of the ₁-adrenergic receptor (Fig 2). These results confirm our previous findings that the majority (70%) of ₁-receptors in human hearts couple with G_h, as shown by induction of ₁-agonist receptor–G_h ternary complex and coimmunoprecipitation of the ₁-receptor with G_h by the G_h7 antibody.³¹

Interesting findings from the studies are that multiple changes in hhG_h functions were observed with failing human heart tissues compared with the NF. In ICM, the ability of hhG_h to couple to the ₁-adrenergic receptors was lower, although the number of ₁-receptors and the level of hhG_h protein in the pellets were substantially increased (Figs 1, 2, and 4B). An increase in ₁-adrenergic receptors in this failing heart was observed with animal models^{15 16} and with human heart tissues.^{14 18} In the animal models, WB4101-sensitive _1A-receptor subtype was increased but not chlorethylclonidine-sensitive _1B-receptor subtype.⁴⁸ In the failing human heart, changes in the level of ₁-receptor subtype are completely unknown. However, it has been reported that in human heart, the message level of _1A(C)-receptor was higher than other ₁-receptor subtypes,^{24 49} which suggests that hhG_h may also couple to other ₁-receptor subtypes. It has been shown that the _1A(C)-receptor activation also resulted in phosphoinositide breakdown.⁴⁹ The GTP-binding activity of hhG_h was downregulated in the pellets. Since the GTPase functionof hhG_h is directly responsible for the ₁-receptor signaling, the downregulation probably affects the ₁-receptor–hhG_h coupling and hhG_h-PLC interaction. It has been demonstrated that the ₁-receptor number in human ischemic heart failure was increased but did not correlate with the physiological response.¹⁸ Downregulation of the intrinsic activities of hhG_h, specifically the GTP-binding activity, is probably a contributing factor to the lack of the ₁-receptor response in this type of heart failure. Moreover, hhG_h protein level was greatly increased in the pellets, accompanying a decrease of the protein level in cytosol, suggesting that hhG_h was translocated to the membranes. Although the reason for the translocation and downregulation of the hhG_h is not clear, our findings lead us to hypothesize that the decrease of functional activity of hhG_h, specifically the GTPase activity, in the membranes causes translocation of cytosolic hhG_h to membranes and an enhancement of ₁-adrenergic receptor expression as a compensatory mechanism.

The DCM human hearts are similar to NF in numbers of ₁-receptors and coupling ability of hhG_h to the ₁-receptors. However, significant changes in the protein level and the enzyme activities of hhG_h occurred in the pellets. Thus, the amount of hhG_h was increased 181% and 48.7% compared with the NF and ICM, respectively, while the GTPase and TGase activities of hhG_h were downregulated. The increase in protein level occurred only in the pellets, without changing the level in the cytosol; that differed from the observation with the ICM. These results also indicated that the expression of hhG_h was increased in DCM, since total protein level of human heart tissues was unchanged in both failing heart tissues compared with the NF. Although the mechanism of increase in hhG_h protein and downregulation of hhG_h in membranes of this failing heart is entirely unknown, it is possible that the downregulation of the intrinsic hhG_h functions may reflect the translocation of the protein and the expression of hhG_h that is probably associated with the feedback response.

Taken together, the results from studies with NF and ICM and DCM tissues clearly demonstrate that hhG_h is the signal mediator of the ₁-adrenergic hormone signaling in human heart and is associated with human heart failure by different mechanisms.

In addition, similar findings regarding cellular redistribution and decrease or increase of the TGase II activity have been reported in various tumors and transformed cells.^{50 51 52 53 54} The TGase II activity in hepatoma was greatly reduced in cytosol compared with the normal rat liver.⁵⁰ Moreover, the enzyme activity did not correlate with protein level, nor did it change during cellular proliferation. Reduction in TGase activity was accompanied by a significant shift in subcellular distribution to the particulate fraction.^{51 52} These findings are similar to our findings with the human ICM. On the other hand, induction of TGase expression by retinoic acid or overexpression results in an increase of the TGase activity, cell adhesion, and morphological changes, including blebbing and cellular fragmentation in BALB/c 3T3 cells.^{52 53 54} These changes are similar to the human DCM. The morphological change could be a result of prior dysfunction of hhG_h, since samples used for the studies are of end-stage heart failure. The mechanism involved in the redistribution and gain or loss of the TGase activity as well as downregulation of GTP binding of TGase II is entirely unknown. It is also possible that other associated proteins, such as the 50-kD ß-subunit of G_h, have been altered (see References 27 and 31) and that expression of an unknown inhibitory factor may be induced in the failing human hearts.

	Acknowledgments

 
This work was supported by National Institutes of Health grants GM-45985 (Dr Im) and SCOR HL-33713 (Core IV–human tissue core) (Dr Moravec). We deeply appreciate the cooperation of the following individuals at the Cleveland Clinic Foundation in making human heart tissue available for research: Dr Robert W. Stewart, Head of Cardiac Transplantation, Dr Norman Ratliff, Head of Cardiac Pathology, the cardiac transplant coordinators, the cardiac transplant team, and the residents in the departments of Pathology and Thoracic and Cardiovascular Surgery. We thank Life Banc of Northeast Ohio for the supply of NF tissue. We thank Michelle Secic for careful statistical analysis, Lui-Fu Rui for excellent technical assistance, and JoAnne Holl for typing.

	Selected Abbreviations and Acronyms

 

DCM = dilated cardiomyopathic hearts DTT = dithiothreitol GPL Gh = guinea pig liver transglutaminase hhGh = human heart TGase II HSD buffer = 20 mmol/L HEPES, pH 7.4, 150 mmol/L NaCl, 1 mmol/L DTT ICM = ischemic cardiomyopathic hearts NF = nonfailing hearts PLC = phospholipase C TGase = transglutaminase

Received December 18, 1995; revision received February 8, 1996; accepted February 17, 1996.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Terzic A, Puceat M, Vassort G, Vogel SM. Cardiac ₁-adrenoceptor: an overview. Pharmacol Rev. 1993;45:147-175.[Medline] [Order article via Infotrieve]
   
2. Fedida D, Braun AP, Giles WR. ₁-Adrenoceptor in myocardium: functional aspects and transmembrane signaling mechanisms. Physiol Rev. 1993;73:469-487.[Free Full Text]
   
3. Minneman KP. ₁-Adrenergic receptor subtypes, inositol phosphates, and sources of cell Ca²⁺. Pharmacol Rev. 1988;40:87-119.[Medline] [Order article via Infotrieve]
   
4. Simpson PC, Kariya K, Karns LR, Long CS, Karlinker JS. Adrenergic hormone and control cardiac myocyte growth. Mol Cell Biochem. 1991;104:35-43.[Medline] [Order article via Infotrieve]
   
5. van Bilsen M, Chien KR. Growth and hypertrophy of the heart: towards an understanding of cardiac specific and inducible gene expression. Cardiovasc Res. 1993;27:1140-1149.[Medline] [Order article via Infotrieve]
   
6. Iwaki K, Sukhatme VP, Shubeita HE, Chien KR. - and ß-Adrenergic stimulation induces distinct patterns of immediate early gene expression in neonatal rat myocardial cells. J Biol Chem. 1990;265:13809-13817.[Abstract/Free Full Text]
   
7. Starksen NF, Simpson PC, Bishopric N, Coughlin SR, Lee WM, Escobedo JA, Williams LT. Cardiac myocyte hypertrophy is associated with c-myc protooncogene expression. Proc Natl Acad Sci U S A. 1986;83:8348-8350.[Abstract/Free Full Text]
   
8. Kariya K-I, Karns LR, Simpson PC. An enhancer core element mediates stimulation of the rat ß-myosin heavy chain promoter by an ₁-adrenergic agonist and activated ß-protein kinase C in hypertrophy of cardiac myocytes. J Biol Chem. 1994;269:3775-3782.[Abstract/Free Full Text]
   
9. Lee HR, Henderson SA, Reynolds R, Dunnmon P, Yuan D, Chien KR. ₁-Adrenergic stimulation of cardiac gene transcription in neonatal rat myocardial cells. J Biol Chem. 1988;263:7352-7357.[Abstract/Free Full Text]
   
10. Long CS, Ordahl CP, Simpson PC. ₁-Adrenergic receptor stimulation of sarcomeric actin isogene transcription in hypertrophy of cultured rat heart muscle cells. J Clin Invest. 1989;83:1078-1082.
    
11. Lanzberg JS, Parker JD, Gauthier DF, Colucci WS. Effects of myocardial ₁-adrenergic receptor stimulation and blockade on contractility in humans. Circulation. 1991;84:1608-1614.[Abstract/Free Full Text]
    
12. Hwa J, De Young MB, Perez DM, Graham MG. Autonomic control of myocardium: alpha-adrenoceptor mechanisms. In: Burnstock E, ed. The Autonomic Nervous System, Vol VIII: The Nervous Control of the Heart. Shepherd J, Vatner SF, volume editors. Cambridge, Mass: Harvard Academic Press; 1996.
    
13. Mukhergee A, Haghani Z, Brandy J, Bush L, McBride W, Buja LM, Willerson JT. Differences in myocardial ₁- and ß-adrenergic receptor numbers in different species. Am J Physiol. 1983;245:H947-H961.
    
14. Steinfath M, Chen Y-Y, Lavicky J, Magnussen O, Nose M, Rosewag S, Schmitz W, Scholz H. Cardiac ₁-adrenoceptor densities in different mammalian species. Br J Pharmacol. 1992;107:185-188.[Medline] [Order article via Infotrieve]
    
15. Corr PB, Shayman JA, Kramer JB, Kipnis RJ. Increased -adrenergic receptors in ischemic cat myocardium: a potential mediator of electrophysiological derangements. J Clin Invest. 1981;67:1232-1236.
    
16. Dillon JS, Gu XH, Nayler WG. Alpha₁-adrenoceptor in the ischemic and reperfused myocardium. J Mol Cell Cardiol. 1988;20:725-735.[Medline] [Order article via Infotrieve]
    
17. Bristow MR, Minobe W, Rasmussen R, Hershberger RE, Hoffman BB. Alpha₁-adrenergic receptors in the non-failing and failing human heart. J Pharmacol Exp Ther. 1988;247:1039-1045.[Abstract/Free Full Text]
    
18. Steinfath M, Danielsen W, von der Leyen H, Mende U, Meyer W, Neumann J, Nose U, Reich T, Schmitz W, Scholz H, Starbatty J, Stein B, Doring V, Kalmar P, Haverich A. Reduced ₁- and ß₂-adrenoceptor-mediated positive inotropic effects in human end-stage heart failure. Br J Pharmacol. 1992;105:463-469.[Medline] [Order article via Infotrieve]
    
19. Landzberg JS, Parker JD, Gauthier DF, Colucci WS. Effects of myocardial ₁-adrenergic receptor stimulation and blockade on contractility in humans. Circulation. 1991;84:1608-1614.
    
20. Boehm M, Diet F, Feiler G, Kemkes B, Erdmann E. -Adrenoceptors and -adrenoceptor-mediated positive inotropic effects in failing human myocardium. J Cardiovasc Pharmacol. 1988;12:357-364.[Medline] [Order article via Infotrieve]
    
21. Jakob H, Nawrath H, Rupp J. Adrenoceptor-mediated changes of action potential and force of contraction in human isolated ventricular heart muscle. Br J Pharmacol. 1988;94:584-590.[Medline] [Order article via Infotrieve]
    
22. Garcia-Sainz JA. ₁-Adrenergic action: receptor subtypes, signal transduction and regulation. Cell Signal. 1993;5:539-547.[Medline] [Order article via Infotrieve]
    
23. Hieble JP, Bylund DB, Clarke DE, Eikenburg DC, Langer SZ, Lefkowitz RJ, Minneman KP, Ruffolo RR Jr. International union of pharmacology, X: recommendation for nomenclature of ₁-adrenoceptors: consensus update. Pharmacol Rev. 1995;47:267-270.[Medline] [Order article via Infotrieve]
    
24. Schwinn DA, Lomasney JN, Lorenz W, Szklut JP, Fremeau RT, Yang-Feng TL, Caron MG, Lefkowitz RJ, Cotecchia S. Molecular cloning and expression of cDNA for a novel ₁-adrenergic receptor subtype. J Biol Chem. 1990;265:8183-8189.[Abstract/Free Full Text]
    
25. Perez DM, DeYoung MB, Graham RM. Coupling of expressed _1B- and _1D-adrenergic receptors to multiple signaling pathways is both G protein and cell type specific. Mol Pharmacol. 1993;44:784-795.[Abstract]
    
26. Im M-J, Graham RM. A novel guanine nucleotide-binding protein coupled to the ₁-adrenergic receptor, I: identification by photolabeling of membrane and ternary complex preparations. J Biol Chem. 1990;265:18944-18951.[Abstract/Free Full Text]
    
27. Im M-J, Riek PR, Graham RM. A novel guanine nucleotide-binding protein coupled to the ₁-adrenergic receptor, II: purification, characterization and reconstitution. J Biol Chem. 1990;265:18952-18960.[Abstract/