Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

In the heart, -adrenergic stimulation increases force of contraction and enhances relaxation. The underlying biochemical mechanism involves the generation of cAMP and the subsequent activation of cAMP-dependent protein kinase. The cAMP-dependent protein kinase phosphorylates regulatory proteins. This triggers the inotropic, clinotropic, and relaxant effects of -adrenergic stimulation in the mammalian heart (Simmerman and Jones, 1998). Targets for the cAMP-dependent protein kinase have been identified in the sarcolemma and in the sarcoplasmic reticulum of the heart. This approach identified a major phosphoprotein in the sarcoplasmic reticulum called phospholamban and another major substrate of apparent molecular weight of 15,000 in the sarcolemma later called phospholemman (Jones et al., 1979; Palmer et al., 1991). -Adrenergic stimulation led to an increased phosphorylation state of phospholamban and phospholemman (Lindemann et al., 1983; Presti et al., 1985a). Other work focused on the function of these proteins and their regulation by phosphorylation. Disruption of the phospholamban gene enhanced basal contractility and hastened relaxation (Luo et al., 1994). Moreover, the positive inotropic effect of -adrenoceptor stimulation was greatly attenuated (Luo et al., 1994). Unphosphorylated phospholamban inhibits the activity of the SR-Ca²⁺-ATPase 2a. Thus, less Ca²⁺ is pumped into the sarcoplasmic reticulum. Phosphorylation relieves this inhibition. The activity of SR-Ca²⁺-ATPase 2a is enhanced, and more Ca²⁺ is pumped into the sarcoplasmic reticulum. This is thought to hasten relaxation (reviewed in Simmerman and Jones, 1998).

Recent evidence suggests that phospholemman can act as an ion channel for chloride or taurine (Moorman et al., 1992, 1995). How its function is regulated by phosphorylation remains to be elucidated.

Previous work indicates that serine/threonine phosphatases of type 1, 2A, 2B, and 2C are present in the heart (Cohen, 1989; DePaoli-Roach et al., 1994). More than 90% of phosphatase activity in the heart is contributed by phosphatase 1 and 2A (Cohen 1989, MacDougall et al., 1991). Studies in rabbit hearts indicate that phospholamban from cardiac membranes can be dephosphorylated by phosphatases of type 1 and type 2A, whereas type 2B and 2C are relatively inactive (MacDougall et al., 1991). Cantharidin and okadaic acid are naturally occurring compounds that inhibit cardiac type 1 and even more potently type 2 A phosphatases in vitro (Li et al., 1993; Neumann et al., 1993, 1995). Moreover, these phosphatase inhibitors increased the phosphorylation state of phospholamban in isolated cardiomyocytes (Neumann et al., 1993, 1994, 1995) and increased membrane currents (Hescheler et al., 1988). Thus, it is likely that phospholamban is dephosphorylated in the intact heart by type 1 and/or type 2 A phosphatases.

Evidence using cell membrane permeant phosphatase inhibitors like okadaic acid, cantharidin, and calyculin A suggests that phosphatases can alter cardiac function by changing the phosphorylation state of cardiac proteins independently of receptor activation. Indeed, phosphatase inhibitors exert positive inotropic effects in nonhuman cardiac preparations and increase the phosphorylation state of phospholamban (Neumann et al., 1993, 1994, 1995). Moreover, we have shown that cantharidin can increase the force of contraction in isolated electrically driven human cardiac preparations (Linck et al., 1996a). This indicates that phosphatase inhibition in the human myocardium can cause a positive inotropic effect and underscores the physiological importance of phosphatases even in the human heart.

However, it is not known whether phosphatases of type 1 or type 2A or both dephosphorylate phospholamban in the human heart. Moreover, it has not yet been reported which phosphatases dephosphorylate phospholemman in any species or tissue. Hence, we studied whether type 1 and type 2A phosphatases are present in the failing human heart and whether they are capable of dephosphorylating key phosphoproteins present in the sarcoplasmic reticulum and in the sarcolemma.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Preparation of Human Cardiac Tissue. Samples were taken from left ventricles of failing human hearts that were explanted in the course of replacement surgery. All patients were male, suffered from idiopathic dilated cardiomyopathy, and their ages ranged from 45 to 57 years. Patients gave informed consent, and the study was approved by the local ethics committee. Medication was comprised of digitalis, diuretics, and angiotensin converting enzyme inhibitors. There is no evidence that these drugs inhibit phosphatase activity as measured in our assay (see below). Moreover, during purification of the catalytic subunits of phosphatases (see below), these drugs are likely lost and should not copurify with the enzyme. Macroscopically visible blood vessels, fatty tissue, and endo- and epicardium were removed from the samples. Samples were then frozen in liquid nitrogen (in most cases within 5 min after explantation). Thereafter, one sample from each heart was homogenized in liquid nitrogen and aliquots (stored at 80°C) were use in additional analysis. The data shown are those from three (Table 1) or from three to four (Table 2) individual hearts. The same hearts were used in Tables 1 and 2. One additional heart was used in Table 2.

Preparation of Homogenates. Human left ventricular myocardium was pulverized in a mortar precooled in liquid nitrogen. The following steps were carried out at 4°C. Five volumes of 50 mM Tris-buffer (pH 7.4) were added to the frozen, pulverized tissue. The powdered tissue was then homogenized three times for 30 s with a Polytron PT-10 (Neumann et al., 1993) at top speed.

Expression of Recombinant Proteins. Recombinant canine phospholamban and phospholemman were expressed in insect cells using a baculovirus expression vector system as described (Reddy et al., 1995; Chen et al., 1998). The final concentration of recombinant phospholamban and phospholemman were 1.1 and 0.3 mg/ml, respectively, in a buffer containing 1% (v/v) Triton X-100, 86 mM triethylamine, 80 mM 4-morpholinepropanesulfonic acid, 18 mM glycine, 5 mM dithiothreitol, and 1% (v/v) n-octyl--D-glucopyranoside (pH 7.2). Recombinant proteins were diluted 200-fold in 50 mM Tris, 0.1 mM EDTA, and 0.1% (v/v) -mercaptoethanol for phosphatase studies.

Preparation of Membrane Vesicles. The membrane vesicles were prepared as described (Ahmad et al., 1989). One g of frozen myocardium was mechanically pulverized in liquid nitrogen. The following steps were carried out at 4°C. The pulverized myocardium was homogenized in 5 ml of buffer 1 containing -mercaptoethanol 1% (v/v), 4 mM EDTA (at pH 7.4) three times for 30 s with a Polytron PT 10 (Kinematica AG, Luzern, Switzerland) at maximum speed. The homogenate was then sedimented for 20 min at 14,000g in a precooled centrifuge (Centricon T-2170; Kontron Instruments, Milano, Italy). The resulting supernatant was sedimented for 30 min at 45,000g. The sediment obtained from this step was resuspended in buffer 1 containing 0.6 M KCl and sedimented again for 30 min at 45,000g. This step was repeated once. The resulting sediment, obtained after the second sedimentation, was resuspended in 220 µl of buffer 2 containing 1% (v/v) -mercaptoethanol, 50 mM Tris, and 0.1 mM EDTA, homogenized and stored at 80°C.

Preparation of Radioactively Labeled Substrates. Phosphorylation with 0.1 mCi [-³²P]ATP was carried as published (Zimmermann et al., 1996). To eliminate endogenous enzyme activities, the membrane vesicles were heat treated for 30 min at 65°C. Two hundred µl of diluted phospholamban, phospholemman, or membrane vesicles were included in a total volume of 430 µl of the phosphorylation mixture, resulting in final concentrations of 59 U/ml cAMP-dependent protein kinase A, 23 mM Tris, 1 mM Mg²⁺, 0.003 mM dl-dithiothreitol, and 0.16% (v/v) -mercaptoethanol. The reaction was started by adding [-³²P]ATP and was allowed to proceed at 30°C on a thermomixer (Eppendorf). After a 60-min incubation at 30°C, the reaction was stopped on ice. Radioactivity not incorporated into phospholamban or phospholemman was removed by dialyzing the proteins three times for 6 h in 500 ml of a buffer containing 50 mM Tris, 5 mM EDTA, and 0.1% (v/v) Triton X-100. Radioactivity not incorporated into membrane vesicles was removed by sedimenting the proteins for 5 min at 14,000g and resuspending them in buffer 2. The supernatant was discarded, and the pellet was resuspended in buffer 2. This sedimentation and resuspension was repeated until the radioactivity in the supernatant was <1% of the radioactivity in the pellet. The final pellet was resuspended in 220 µl of buffer 2. Preparation of [³²P]phosphorylase a followed a published method (Neumann et al., 1991).

Gel Electrophoresis. Incubation was stopped by adding stop solution, which consisted of 62.5 mM tris(hydroxymethyl)amino-methane, 10% (w/v) sodium dodecyl sulfate, 10% (v/v) glycerol, 0.6% (w/v) dl-dithiothreitol, and a trace of bromophenol blue; pH was adjusted to 6.8. Samples were frozen at 20°C. Sodium dodecyl sulfate polyacrylamide gel electrophoresis was performed using 10% polycrylamide separating gels with 4% stacking gels according to Neumann et al. (1993). Samples were thawed and then, unless otherwise stated, heat-treated for 10 min at 95°C. Gels were stained with Coomassie blue R-250, destained, and dried. For autoradiography, dried gels were exposed to medical imaging film at 20°C. Apparent molecular weights were determined using a low-molecular-weight calibration kit (Pharmacia, Piscataway, NJ).

Purification of Phosphatases. The catalytic subunits of type 1 and type 2A phosphatases were purified from failing human ventricles in slight modification of methods used previously by Erdödi et al. (1985) and Neumann et al. (1995) for rabbit liver and guinea pig myocardium, respectively. Unless stated otherwise, all procedures were carried out at 4°C. Samples (2 g) of human left ventricular myocardium were homogenized three times 30 s with a Polytron at top speed in a buffer 1 containing 20 mM Tris, 5 mM EDTA, 2 mM EGTA, 1 mM benzamidine (in 1 ml of ethanol), 0.5 mM phenylmethylsulfonyl fluoride, and 1% (v/v) -mercaptoethanol. The homogenate was sedimented at 3000g for 20 min. Ammonium sulfate (351 mg/ml) was added to the resulting supernatant, and the mixture was stirred for 30 min. This was followed by additional centrifugation at 3000g for 20 min. The resulting pellet was resuspended in 10 ml of buffer 1, and 50 ml of ethanol were added. The suspension was stirred for 30 min (at room temperature) and then sedimented at 27,000g for 20 min. The resulting pellet was resuspended in buffer 1 and sedimented again at 27,000g for 20 min. The supernatant was kept. The resulting supernatant was combined with the supernatant from the previous step and dialyzed for 3 h against a 10-fold volume of buffer 2 [20 mM Tris, 5 mM EDTA, 2 mM EGTA, and 1% (v/v) -mercaptoethanol]. The dialysate was then applied to a column containing heparin-Sepharose equilibrated in buffer 1. Fractions of 2-ml volume were collected of the flowthrough and from 100 ml of a linear gradient of 0-0.5 M NaCl in buffer 1. Phosphatase activity was assayed in the fractions as described below.

Protein Phosphatase Assay. The phosphatase assays were performed in slight modification of a method described by Neumann et al. (1993). Homogenates were diluted 1:100 in a buffer containing 50 mM Tris (pH 7.4). Twenty µl of okadaic acid or cantharidin in 1:20 (v/v) dimethylsulfoxide were added to 10 µl of the diluted homogenates to give final concentrations (in -log M) of (6, 7, 8, 9, 10, 11, and 12) for okadaic acid and (4, 5, 6, 7, 8, 9, and 10) for cantharidin. The homogenates were then preincubated for 10 min at 30°C. The incubation was stopped on ice. Twenty µl of [-³²P]ATP-labeled substrates (see above) in an incubation mixture composed of 12.5 mM caffeine, 2.5 mM Tris, 0.25 mM EDTA, and 0.25% (v/v) -mercaptoethanol were added to the preincubated phosphatases. After incubating for 45 min at 30°C, the assay was terminated on ice by adding 20 µl of trichloroacetic acid and 30 µl of 1 mg/ml bovine serum albumin to precipitate the proteins. The precipitate was sedimented by centrifugation, and released radioactivity in an aliquot of the supernatant was measured in a liquid scintillation counter. Phosphorylase a phosphatase activity was measured as described above for phospholamban and phospholemman phosphatase activity, except for a shorter incubation time of 10 min after the addition of phosphorylase a to the assay. Protein was measured as before (Neumann et al., 1991).

Materials. Compounds used were [-³²P]ATP (Amersham Buchler, Braunschweig, Germany), okadaic acid (Biotrend, Cologne, Germany), and cantharidin (LC Laboratories, Woburn, MA). All materials for SDS-polyacrylamide gel electrophoresis were purchased from Bio-Rad (Munich, Germany). All other chemicals were of analytical quality or best commercial grade available. Deionized and twice destilled water was used throughout.

Statistics. Data shown are means ± S.E.M. or with 95% confidence intervals in parentheses. Statistical significance was estimated with Student's t test for unpaired observations; P < .05 was considered significant.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

First recombinant phospholamban, recombinant phospholemman, and membrane vesicles from human hearts, which are known to contain phospholamban and phospholemman (Presti et al., 1985a,b), were phosphorylated by cAMP-dependent protein kinase in the presence of radioactive ATP. Samples were separated using gel electrophoresis, and ³²P-labeled proteins were visualized by autoradiography. An autoradiogram (Fig. 1B) shows that phosphoproteins of the expected molecular weights were detectable in membrane vesicles from human hearts. Of note, two bands probably corresponding to phospholamban and phospholemman were phosphorylated in human membrane. The tentative identification of these bands as phospholamban and phospholemman was supported by the fact that boiling of samples before electrophoresis reduced the apparent molecular weight of phospholamban but not of phospholemman, as noted before in rat and canine tissue (Presti et al., 1985a,b). Moreover, we have identified phospholamban before in human hearts using Western blots and specific antibodies (Linck et al., 1996b; Neumann et al., 1997). This finding extends data from canine heart studies that phospholamban and phospholemman are substrates for cAMP-dependent protein kinase in human cardiac membrane vesicles. These data are compatible with the view that phospholemman expression in human heart is substantially lower than phospholamban expression. Recombinant phospholamban also exhibited the expected molecular weight change (reviewed in Simmerman and Jones, 1998) upon boiling (Fig. 1A). Purified phospholemman was also phosphorylated by cAMP-dependent protein kinase. An autoradiogram is shown in Fig. 2. Apparently, recombinant phospholemman is an excellent substrate for cAMP-dependent protein kinase, as expected (Presti et al., 1985a,b; Palmer et al., 1991). For comparison, the classical substrate phosphorylase b was phosphorylated by phosphorylase kinase to phosphorylase a. After autoradiography, a single band at the expected molecular weight was detected (Fig. 2). To validate the phosphatase assays, ³²P-radiolabeled phospholemman, phospholamban, and phosphorylase a were dephosphorylated by phosphatases from diluted homogenates of human left ventricular tissue, subjected to gel electrophoresis, and autoradiographed (Fig. 2). All three substrates were dephosphorylated by human cardiac preparations (Fig. 2). In initial experiments, we tried to dephosphorylate human membrane vesicles by human cardiac phosphatases. This was very problematic. The radioactive phosphate incorporation was sufficient for detection on autoradiograms (Fig. 1) but too low for routine scintillation counting of radioactivity released by exogenously added human cardiac phosphatases. Tissue limitation also excluded human membrane vesicles as routine sources for phospholamban or phospholemman. However, recombinant phospholamban and phospholemman turned out to be excellent substrates and could be used like phosphorylase a for routine assays. Using these substrates, phosphatase activity in human homogenates (Table 1) or after column separation of the catalytic subunits of phosphatases type 1 and 2A (Fig. 3) and on peak fractions from column separations could be measured (Table 2).

View larger version (37K): [in this window] [in a new window]   Fig. 1.   A, autoradiogram of a polyacrylamide gel used to identify recombinant [32P]phospholamban (PLB). Samples marked (+) were heat treated for 10 min at 95°C to convert PLB into a low molecular weight form. On the left, molecular weight standards are indicated. B, autoradiogram of a polyacrylamide gel used to identify [32P]phospholamban (PLB) and [32P]phospholemman (PLM) in radioactively labeled membrane vesicles from failing human hearts. Samples marked (+) were heated for 10 min at 95°C before electrophoresis to convert PLB to a low molecular weight form. Note the time dependence of the phosphorylation by the cAMP-dependent protein kinase.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Autoradiograms showing the dephosphorylation of 32P-labeled phosphorylase a (Phos a), phospholamban (PLB), and phospholemman (PLM) by phosphatases from human left ventricular tissue. All samples were heat treated for 10 min at 95°C. On the left and the right, molecular weight standards are indicated in thousands. Note the different scales of standards. Samples were preincubated with phosphatases as indicated (+).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Phospholemman (PLM), phospholamban (PLB), and phosphorylase a (Phos a) phosphatase activity eluted from a heparin-Sepharose column. The column was first washed with 4 column volumes of homogenization buffer (fractions 11-43) as described in Experimental Procedures. Thereafter, a NaCl gradient from 0 to 0.5 M was started, ending at fraction 98. Totals counts were 7753 cpm (A), 48376 cpm (B), and 7648 cpm (C). Fraction size was 2 ml. Ten µl of each fraction were assayed for phosphatase activity as described in Experimental Procedures.

The dephosphorylation of all substrates was inhibited in a concentration-dependent manner by okadaic acid and cantharidin (Table 1 for phospholamban, phospholemman, and phosphorylase as substrates). Okadaic acid inhibited phosphorylase, phospholamban, and phospholemman phosphatase activity in human cardiac homogenates in a concentration-dependent manner starting at 1 nM.

Likewise, cantharidin inhibited dephosphorylation of phosphorylase a, phospholamban, and phospholemman in human cardiac homogenates in a concentration-dependent manner but starting at 0.1 µM. The IC₅₀ values for okadaic acid and cantharidin are provided as Table 1. Okadaic acid and cantharidin were equieffective, but okadaic acid was more potent than cantharidin. The inhibition curves of the phosphatase activities in homogenates were shallow. This result and our own previous work on guinea pig cardiac phosphatases (Neumann et al., 1995) indicated the presence of several phosphatases that were differently sensitive to these inhibitors. Hence, the phosphatase activities were further purified, using an ethanol precipitation step for separation of regulatory from catalytic subunits and a heparin-Sepharose column to distinguish type 1 from type 2A phosphatase catalytic subunits as described for guinea pig heart (Herzig et al., 1995; Neumann et al., 1995). The elution profiles for phospholamban, phospholemman, and phosphorylase phosphatase activities are shown in Fig. 3. The profiles were all very similar and exhibited two major peaks (peak 1 and peak 2) of phosphatase activity.

The first peak of phosphatase activity, noticed in the flowthrough, consisted of activities that did not bind to the column. The second peak was eluted by a NaCl gradient. In previous studies, type 2A phosphatases did not bind to the column, whereas type 1 did (Neumann et al., 1995). Therefore, peak 1 and peak 2 were tentatively identified as type 2A and type 1 phosphatase activities, respectively. Inhibition experiments with okadaic acid an cantharidin, as described below, are consistent with these assumptions.

Okadaic acid concentration dependently inhibited type 1 and type 2A phosphatase activities from human heart (Table 2). Similar inhibition curves were obtained for cantharidin (Table 2). Okadaic acid and cantharidin were equieffective, but okadaic acid was more potent than cantharidin. The IC₅₀ values of the inhibition experiments with okadaic acid and cantharidin are summarized in Table 2. Okadaic acid inhibited type 2A phospholamban, phospholemman, phosphorylase phosphatase activities 47, 129, and 133 times more potently than the respective type 1 phosphatase activities. Cantharidin inhibited type 2A phospholamban, phospholemman, and phosphorylase phosphatase activities 20, 25, and 36 times more potently than type 1 phosphatase activities. Comparing the IC₅₀ values of okadaic acid and cantharidin, okadaic acid inhibited type 2A phospholamban, phospholemman, and phosphorylase phosphatase activities 33, 48, and 38 times more potently than cantharidin, whereas type 1 phospholamban, phospholemman, and phosphorylase phosphatase activities were inhibited 14, 9, and 10 times more potently by okadaic acid than by cantharidin. Hence, as reported in guinea pig hearts (Neumann et al., 1995), cantharidin is less selective than okadaic acid for type 2A phosphatase activity. Nevertheless, both inhibitors clearly distinguished between type 1 and type 2A phosphatases.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The main finding of the present study is that two important, membrane-localized phosphoproteins in the heart, phospholamban and phospholemman, can be dephosphorylated by the type 1 and type 2A cardiac phosphatases.

Using recombinant phospholamban and phospholemman as substrates, we have extended our previous work. Phospholamban phosphorylated in rabbit membranes could be dephosphorylated by phosphatases from rabbit skeletal muscle (MacDougall et al., 1991). Recombinant phospholamban can be dephosphorylated by phosphatases from the guinea pig heart (Zimmermann et al., 1996). However, here we report for the first time that recombinant phospholamban can be dephosphorylated by phosphatases from the human heart. Moreover, dephosphorylation is mediated by both the catalytic subunits of type 1 and type 2A phosphatases. Dephosphorylation of phospholamban by human cardiac homogenates and by purified catalytic subunits of type 1 and type 2A phosphatases from human hearts can be concentration dependently inhibited by cantharidin and okadaic acid, which has not been reported before. For comparison, phosphorylase a dephosphorylation was measured. Cantharidin inhibited dephosphorylation of phosphorylase a in homogenates from human hearts with an IC₅₀ of ~170 nM (this report) and with an IC₅₀ of 540 nM in guinea pig ventricular homogenates (Neumann et al., 1995), which is comparable. In guinea pig preparations, cantharidin inhibited type 1 and type 2A phosphorylase phosphatase activity with IC₅₀ values of 2.7 µM and 130 nM, respectively. In human tissue, type 1 and 2A phosphorylase phosphatase activities were inhibited by cantharidin with IC₅₀ values of 410 and 11 nM, respectively. In guinea pig studies, we reported that okadaic acid inhibited type 1 and type 2A phosphorylase phosphatase activity with IC₅₀ values of 120 and 0.7 nM, respectively (Neumann et al., 1995), whereas in human tissue type 1 and 2A phosphorylase phosphatase activities were inhibited with IC₅₀ values of 40 and 0.3 nM, respectively (this report). In agreement with our previous data (Neumann et al., 1995), we found that cantharidin inhibited human cardiac phosphatases equieffectively but less potently than okadaic acid. It may further be concluded that both cantharidin and okadaic acid inhibited type 2A human cardiac phosphatases about 10 and 100 times, respectively, more potently than type 1 phosphatases.

Inhibition of phospholamban phosphatase activities for okadaic acid and cantharidin has not been reported before. It turns out that the IC₅₀ for phospholamban and phosphorylase phosphatase activity are very comparable for human cardiac phosphatases. This validates the work of our group and others that routinely used phosphorylase as a model substrate in the past.

The other new finding of the present work is the characterization of phospholemman dephosphorylation, which has previously been undetectable because of the low expression of phospholemman in tissues. Therefore, no comparable data on human or nonhuman preparations are available in the literature. Phospholemman was dephosphorylated by purified catalytic subunits of type 1 and/or type 2A phosphatases in human cardiac homogenates under our assay conditions. More specifically, phospholemman was dephosphorylated by type 1 and type 2A phosphatases. Both cantharidin and okadaic acid inhibited these dephosphorylations equieffectively, but okadaic acid was more potent than cantharidin. The inhibition of phospholamban and phospholemman dephosphorylation by the phosphatase inhibitors studied (okadaic acid and cantharidin) was very similar. Hence, one would predict that concentrations of okadaic acid or cantharidin that increase the phosphorylation state of phospholamban should in parallel increase phospholemman phosphorylation. However, in our previous studies on isolated cardiomyocytes, we have not unambiguously identified phosphatase inhibitor-induced phospholemman phosphorylation, probably because of its low expression (Neumann et al., 1993).

In fact, the present data support the hypothesis that -adrenoceptor-stimulated phosphorylation of phospholamban and phospholemman in the heart is mediated at least in part by inhibition of phosphatase type 1, conceivably via altered phosphorylation of phosphatase inhibitor 1 (Ahmad et al., 1989; Neumann et al., 1991). Moreover, it can be predicted that should type 2A inhibition also occur, which has not been unequivocally shown, this would also contribute to the effect of -adrenoceptor stimulation on protein phosphorylation in the heart.

Another interpretation of our data is also warranted. The present functional data confirm and extend previous biochemical data from our group. We have identified the mRNA coding for type 1 and type 2A and 2A catalytic unit subtypes and type 1, 1, and 1 in total RNA isolated from human atria and ventricles (Neumann et al., 1997; Klein-Wiele et al., 1998). The present work shows that these RNA data are compatible with the activities of the catalytic subunits of type 1 and 2A phosphatases measurable in the human heart.

There is evidence that type 1 mRNA expression (catalytic subunit) and total activity of protein phosphatase is increased in human heart failure (Neumann et al., 1997). We have not yet differentiated whether type 1 or type 2A phosphatase activity is elevated in human heart failure.

Preliminary evidence suggests that in some animal models of impaired ventricular function, alterations of phosphatases do occur. Increased phosphatase activity was noted after infarction (Huang et al., 1997) after chronic -adrenergic stimulation (which led to hypertrophy; Bokník et al., 1997) and chronic ischemia (Gupta et al., 1997).

In summary, we extend the physiological and functional work of our group by characterizing type 1 and type 2A phosphatase activity in the human heart using substrates that mediate the -adrenergic inotropic effects in the heart.