Search

It is widely accepted that biopharmaceutical companies have, in recent times, failed to deliver large numbers of new medicines to patients and have simultaneously failed to deliver large financial returns to their investors. We argue that addition of different business constructs with wider stakeholder ownership and/or control offers a way to improve returns from the great advances in medical science and drug discovery processes. Governments and other payers for medicines, the academic institutions engaged in bioscience knowledge creation, patient advocacy groups, venture philanthropists and charitable foundations can come together with commercial profit-centred businesses to develop corporate constructs that mutually benefit all of the stakeholders. A rebalancing of the social and financial motives in medicines research can arrest recent productivity decreases of the sector.

1. The calcium sensitivity of force production of cardiac muscle fibres is altered by certain drugs. The sites of action of three such compounds (pimobendan, sulmazole, isomazole) within the myofibril have been investigated. Calmodulin antagonists, perhexilene and bepridil, which have been shown to alter the calcium dependence of myofibrillar ATPase activity and oxmetidine, an H2-receptor antagonist which binds to calmodulin, were also studied. 2. The rates of dissociation of calcium from both the regulatory and high affinity sites on bovine isolated cardiac troponin C (cTnC) were measured in a stopped-flow fluorimeter. The rates of dissociation were found to be 136.5 +/- 16 s-1 and 1.3 +/- 0.20 s-1 (mean +/- s.e.mean, n = 11 determinations; conditions: 100 mM KCl, 10 mM MOPS, 3 mM MgCl2, 0.1 mM dithriothreitol, pH 7.0, 15 degrees C). Sulmazole, isomazole and perhexiline (final concentration of 50 microM) had no effect on the rate of Ca2+ dissociation from the regulatory Ca2+ site, indicating that these compounds do not act on cTnC directly. 3. The rate of dissociation of Ca2+ from the regulatory site was slightly reduced (approximately 20%) by pimobendan (50 and 100 microM) and was somewhat increased by oxmetidine (28% at 100 microM). 4. Bepridil (25 microM) reduced the rate of dissociation by 50%, indicating a direct effect of bepridil on TnC. 5. Sulmazole, isomazole, perhexiline, pimobendan (50 microM) and bepridil (25 microM) were without effect on the rate of dissociation of Ca2+ from the high affinity Ca2+/Mg2+ sites. Oxmetidine caused 24% decrease in the rate of Ca2+ dissociation from these sites. 6. The rate of dissociation of Ca2+ from the regulatory site on the complex of troponin-tropomyosin (TnTm) was measured. Sulmazole and pimobendan (50 microM) were without effect on the rate of dissociation of Ca2+ from the regulatory site in the protein complex, and isomazole (50 microM) caused only a slight reduction (23%). Perhexiline (50 microM) or bepridil (10 microM) reduced the rate of Ca2 dissociation by about 50%. The rate of dissociation of Ca2+ from the high affinity Ca2 +/Mg2 + sites was not altered by sulmazole, isomazole, or pimobendan (50 microM), but was decreased - 35% by perhexiline (50 microM) or bepridil (10 microM).

1H NMR spectroscopy at 360 MHz has been used to study the interactions between the calmodulin function inhibitor calmidazolium (R24571) and (i) calmodulin (CaM) and (ii) skeletal muscle troponin C (sTnC). One equivalent of racemic calmidazolium binds tightly to CaM and perturbs a number of protein signals, corresponding to residues in both dicalcium-binding domains, in a manner characteristic of slow exchange. Calmidazolium binds with lower affinity to sTnC but still induces widespread perturbations in both domains. Extensive spectral overlap precludes definite assignment of intermolecular nuclear Overhauser effect (NOEs) although intraprotein NOEs do indicate the nature of some drug-induced conformational changes. Relaxation enhancements induced by two spin-labeled calmidazolium analogues demonstrate that several methionine residues of CaM, significantly immobilized by calmidazolium binding, are in fact located at or near its binding sites. These and other residue-specific broadening effects have enabled low resolution models to be constructed of the predominantly hydrophobic drug-binding sites on each domain of CaM. The hydrophobic portions of calmidazolium itself, and its analogues, contact side chains of Ala-15, Leu-18, Phe-19, Val-35, Met-36, Leu-37, Leu-39, Met-51, Met-71, Met-72, and Met-76 in the N-terminal domain of calmodulin, and Ala-88, Val-91, Phe-92, Val-108, Met-109, Leu-112, Phe-141, and Met-145 in its C-terminal domain. The model, and an analogous one of sTnC, can be used to rationalize drug-induced changes in intraprotein NOEs. Issues pertaining to the possible simultaneous binding of calmidazolium to both globular domains of the proteins are discussed in terms of the experimental results and the overall structures of each protein.

1. The cyclic AMP-dependent protein kinase activity-ratio assay was investigated by comparing histone and a synthetic peptide, malantide [Malencik & Anderson (1983) Anal. Biochem. 132, 32-40], as substrates. 2. In several tissues the activity ratio was higher when assayed with histone as the substrate; this result was obtained in control tissues and also in those incubated with agents known to increase cyclic AMP. The effect of these agents to increase the activity ratio was more clearly demonstrated with malantide. 3. The higher activity ratios observed with histone are due to: (a) measurement of phosphorylation not catalysed by cyclic AMP-dependent protein kinase; (b) activation of cyclic AMP-dependent protein kinase by histone during the assay. 4. When tissues were homogenized in buffers without NACl, lower activity ratios were found, owing to the catalytic subunit being artifactually removed from the supernatant. 5. We conclude that the measured activity ratio more faithfully reflects that in the tissue when NaCl is included in the homogenization buffer and malantide is used in the assay. This was confirmed in experiments where cyclic AMP-dependent protein kinase was added to the tissue before homogenization, and no dissociation of the exogenous enzyme was observed.

1. The effects of siguazodan (SK&F 94836) a selective phosphodiesterase (PDE) inhibitor with inotropic and vasodilator activity, were studied on human platelets. 2. Siguazodan selectively inhibited the major cyclic AMP-hydrolysing PDE in human platelet supernatants. The inhibited enzyme has been variously termed cyclic GMP-inhibited PDE or PDE-III. 3. In platelet-rich plasma (PRP), siguazodan inhibited U46619-induced aggregation more potently than that induced by ADP and collagen. Treatment of the PRP with aspirin had no effect on the potency of siguazodan. 4. In washed platelets, siguazodan increased cyclic AMP levels and reduced cytoplasmic free calcium [( Ca2+]i). ADP decreased the ability of siguazodan to raise cyclic AMP and this may explain its lower potency in inhibiting responses to ADP. 5. Siguazodan has anti-platelet actions over the same concentration range that it is an inotrope and vasodilator.

Continuous perfusion of rat hearts with concentrations of forskolin between 0.1 and 12 microM resulted in transient increases in tension after 45 s, followed by a return to the control value after 5 min. In contrast, the content of cyclic AMP increased linearly with time over this period, reaching values up to 35 times control after 5 min. Increases in contractile force, intracellular cyclic AMP concentration and the proportion of phosphorylase in the a form were dependent on the concentration of forskolin when measured 45 s and 120 s after initiation of perfusion. In hearts perfused for 45 s with various concentrations of forskolin, the measured cyclic AMP-dependent protein kinase activity ratio and phosphorylase a content for a given measured intracellular cyclic AMP concentration were both much less than the corresponding values in hearts perfused for 30 s with various concentrations of isoprenaline. The phosphorylation of the contractile proteins troponin-I and C-protein also showed a concentration-dependent increase in hearts perfused with forskolin. There was a strong correlation between the cyclic AMP-dependent protein kinase activity ratios and the phosphorylation of the contractile proteins under all perfusion conditions. These results suggest that cyclic AMP is compartmented in perfused rat heart, and that much of the cyclic AMP produced in response to forskolin is unavailable to activate cyclic AMP-dependent protein kinase.

Volatile general anaesthetic agents exert negative inotropic effects on the heart during anaesthesia. Inhibition by anaesthetics of the myofibrillar ATPase responsible for energy transduction during cardiac contraction has been suggested as the primary site of these effects. In this study we have examined the actions of three anaesthetics with differing cardiac depressant potencies (enflurane, isoflurane and halothane) on the activity of isolated myofibrillar ATPase. The most powerful in vivo cardiac depressant, enflurane, was without effect on the ATPase but halothane and isoflurane were inhibitory to the enzyme.

Phosphorylation of phospholamban in cardiac sarcolemma is implicated in the increased influx of Ca2+ through the slow calcium channel induced by catecholamines. A method is described for the preparation of highly purified sarcolemmal vesicles from rat heart, and this has been used to examine the phosphorylation of phospholamban in 32Pi-perfused rat hearts. Phospholamban phosphorylation is increased 3-fold after 30 s of perfusion with 0.1 μM isoprenaline. The time course of this increase precedes the inotropic response by 5–10 s.

(i) A new, rapid method for the measurement of [γ-32P]ATP specific radioactivity in tissue extracts in the presence of other 32P-containing compounds is described. The deproteinized extract is incubated with phosphorylase b and phosphorylase kinase, and the incorporation of 32P into protein from [γ-32P]ATP is measured by precipitation on filter paper in trichloroacetic acid. No separation of ATP or other treatment of the extracts is required for the assay. (ii) 32Pi uptake in perfused rat heart was found to be a relatively slow process, with a Km of 0.084 m m , whereas equilibration between intracellular 32Pi and [γ-32P]ATP occurred rapidly.

1. Cyclic GMP-dependent protein kinase phosphorylates purified phospholamban. It also phosphorylates phospholamban present in vesicles of cardiac sarcoplasmic reticulum and smooth muscle microsomal fractions, and in transformants of Escherichia coli which contain a plasmid into which a gene encoding phospholamban has been inserted. 2. In vitro the phospholamban present in cardiac sarcoplasmic reticulum membranes is a better substrate for cyclic GMP-dependent protein kinase than for cyclic AMP-dependent protein kinase. 3. Studies using [32P]Pi to label the cellular ATP in intact cardiac or smooth muscle failed to demonstrate that phosphorylation of phospholamban occurs in response to stimuli which increase intracellular cyclic GMP. Possible reasons for this functional separation between increased cyclic GMP and phosphorylation of phospholamban are discussed.

1. The concentrations of free ATP, phosphocreatine (PCr), Pi, H+ and ADP (calculated) were monitored in perfused rat hearts by 31P n.m.r. before and during positive inotropic stimulation. Data were accumulated in 20 s blocks. 2. Administration of 0.1 microM-(-)-isoprenaline resulted in no significant changes in ATP, transient decreases in PCr, and transient increases in ADP and Pi. However, the concentrations of all of these metabolites returned to pre-stimulated values within 1 min, whereas cardiac work and O2 uptake remained elevated. 3. In contrast, in hearts perfused continuously with Ruthenium Red (2.5 micrograms/ml), a potent inhibitor of mitochondrial Ca2+ uptake, administration of isoprenaline caused significant decreases in ATP, and also much larger and more prolonged changes in the concentrations of ADP, PCr and Pi. In this instance values did not fully return to pre-stimulated concentrations. Administration of Ruthenium Red alone to unstimulated hearts had minor effects. 4. It is proposed that, in the absence of Ruthenium Red, the transmission of changes in cytoplasmic Ca2+ across the mitochondrial inner membrane is able to maintain the phosphorylation potential of the heart during positive inotropic stimulation, through activation of the Ca2+-sensitive intramitochondrial dehydrogenases (pyruvate, NAD+-isocitrate and 2-oxoglutarate dehydrogenases) leading to enhanced NADH production. 5. This mechanism is unavailable in the presence of Ruthenium Red, and oxidative phosphorylation must be stimulated primarily by a fall in phosphorylation potential, in accordance with the classical concept of respiratory control. However, the full oxidative response of the heart to stimulation may not be achievable under such circumstances.

The inhibitory subunit of troponin (troponin-I) from rabbit skeletal muscle has been shown to be phosphorylated by both phosphorylase kinase and cyclic 3’, 5’ AMP dependent protein kinase [l-3] . Both kinases also phosphorylate the 37 000 dalton component (troponin-T) [3-51, and whole.troponin from cardiac muscle [6] . A hypothesis has been proposed that in vivo phosphorylation of troponin could modify its properties so as to allow an increase in the contractility of muscle [2,5] . The phosphorylation could be initiated by hormonal stimulation of the muscle, in particular by catecholamines. This may be of special importance in cardiac muscle, where both contractility and cyclic 3’, 5’ AMP are increased by adrenaline [7,8]. However, before this hypothesis is acceptable as a mechanism for enhancing contractility, several criteria have to be met [9]. In particular, it has to be shown that phosphorylation occurs in vivo under conditions where contractility is increased. This paper presents results which show that when perfused rat heart is stimulated by 1 O6 M adrenaline there is a time dependent increase in phosphorylation of troponin-I which coincides with the increase in force of contraction. A direct correlation between contraction and phosphorylation is also seen in experiments where the concentration of adrenaline is varied from 3 X 10” M to 3 X 10d M.

Phosphorylase phosphatase from rabbit skeletal muscle, when purified 700-fold, dephosphorylated the 22,000 molecular weight subunit of troponin (TNI). The endogenous phosphate in TNI and that incorporated by phosphorylase kinase were both totally removed at identical rates. Five criteria were established to show that dephosphorylation of TNI was catalyzed by phosphorylase phosphatase and not a contaminant in the preparation: (a) co-purification of TNIand phosphorylase-phosphatase activities; (b) identical patterns of the activities on polyacrylamide gel electrophoresis; (c) parallel heat denaturation and (d) fluoride inhibition of the activities; (e) competitive effects when both TNI and phosphorylase a were present. The Km for phosphorylase a was 6.8 µm, and 21.2 µm for TNI. The Vmax with TNI was approximately 4-fold greater than with phosphorylase a.

The inotropic and metabolic effects of isoproterenol in heart are now thought to be mediated via changes in the concentration of cyclic adenosine 3′,5′-monophosphate (cAMP) [ 14 ], followed by phosphorylation of specific proteins [ 15 ]. There is evidence for an increase in cAMP in heart muscle following the onset of total ischaemia or anoxia [see 6 ], although this was not observed in isolated, ischaemic hearts [ 1 ]. The observed increase in cAMP presumably reflects an activation of adenylate cyclase by endogenous cardiac catecholamines released during acute ischaemia. Recently Drummond and Sordahl [ 2 ] reported a progressive decline in the stimulation of adenylate cyclase by catecholamine in dog hearts after 20 to 60 min of total ischemia, whereas an opposite effect was observed by Mori [ 10 ] following 15 min of ischaemia. Both these studies measured adenylate cyclase activity in homogenates prepared from ischaemic cardiac tissue. In the present study the isoproterenol-induced increase in cAMP in the isolated rat heart was used to investigate the catecholamine responsiveness of adenylate cyclase in intact ischaemic tissue. The results show that after 5 min of ischaemia there was an abolition of the increase in cAMP induced by isoproterenol. Corresponding to this there was a decreased response to isoproterenol of contractile force and phosphorylase activation.

The effect of the proteases trypsin, thermolysin and papain on the cardiac membrane protein phospholamban was examined before or after phosphorylating the protein with the catalytic subunit of cyclic AMP-dependent protein kinase. The sensitivity of phospholamban to digestion by trypsin and thermolysin was greatly reduced by phosphorylation, suggesting that phospholamban undergoes a conformational change upon phosphorylation. It is suggested that this change in conformation is the mechanism by which phospholamban phosphorylation relieves its inhibition of the sarcoplasmic reticulum Ca2+-ATPase pump.

Two contractile proteins, troponin-I and C-protein, are rapidly phosphorylated in perfused heart in response to stimulation by catecholamines. Both are phosphorylated only in response to increases in intracellular cyclic AMP, and are good substrates for cyclic AMP-dependent protein kinase in vitro. The phosphorylation of troponin-I decreases the affinity of the troponin complex for Ca2+ by increasing the off-rate for Ca2+ dissociation. This is probably part of the mechanism whereby catecholamines increase the rate of relaxation in heart. C-protein is phosphorylated in vivo and in vitro on 4–5 sites, but the function of both C-protein and its phosphorylation are unknown. Myosin P-light chain (phosphorylated by a Ca2+-dependent kinase) is approximately 50 % phosphorylated in control perfusions, and this is unchanged during acute inotropic interventions. However, perfusions with 32Pi indicate an active kinase/phosphatase couple, which may be fully active under control conditions.

The roles of cyclic AMP, cyclic AMP-dependent protein kinase and the phosphorylation of specific proteins in the regulation of cardiac contractility are briefly reviewed. Criteria for determining whether changes in cyclic AMP and protein phosphorylation are involved in a physiological response are discussed. Although cyclic AMP-dependent phosphorylation of the voltage-operated Ca2− channel, phospholamban, troponin-I and C-protein have all been implicated in the response of the heart to inotropic agents which elevate cyclic AMP, none of these phosphorylations satisfy all of the criteria completely. Evidence is presented that there are compartments of cyclic AMP in heart which are coupled to different functional responses.

Catecholamines in the heart cause increases in the rate and extent of tension development during systole and in the rate of relaxation during diastole. These changes may in large part be brought about by cyclic AMP-induced phosphorylations of contractile and membrane proteins. Catecholamines cause an increase in the concentrations of Ca*+ in the cytoplasm during systole, probably by phosphorylating proteins in the sarcolemma and sarcoplasmic reticulum which stimulate Ca2+ transport across these membrane systems (Katz et al., 1975; Will et al., 1973; Wollenberger & Will, 1978). In the sarcoplasmic reticulum a protein of M , 11 OOO, called phospholamban, is phosphorylated by cyclic AMPand Ca*+-dependent protein kinases (Katz et al., 1975; Le Peuch et al., 1980). In the sarcolemma a number of proteins have been reported to be phosphorylated, although there is now good evidence that phospholamban, or a very similar protein, is also present as a major phosphoprotein in this membrane fraction. In our laboratory a mixed sarcoplasmic reticulum and sarcolemma membrane fraction from rat hearts perfused with )*Pi contained one major phosphoprotein which co-migrated with authentic phospholamban. Phosphorylation was increased lG15-fold by perfusion with catecholamine, and the increase in phosphorylation just preceded the increase in contractile force. A highly purified sarcoplasmic reticulum fraction prepared from this mixed membrane preparation surprisingly contained phospholamban in a dephosphorylated form, although phosphorylated phospholamban was present in a less pure sarcolemma preparation. We tentatively conclude that in rat heart, at least part of the sarcoplasmic reticulum contains phospholamban which is not phosphorylated in response to catecholamines. In contrast, the sarcolemma contains phospholamban or a very similar protein which is phosphorylated, under these conditions. These results imply that much of the regulation of cytoplasmic Ca2+ in the presence of catecholamine may occur at the sarcolemma. There are two major contractile proteins in heart muscle that are phosphorylated in response to catecholamines and other agents which elevate cyclic AMP, namely troponin-I and C-protein (England, 1975, 1976; Jeacocke & England, 1980~). Both of these proteins when isolated are good substrates for cyclic AMP-dependent protein kinase (Cole & Perry, 1975; Ray & England, 1976). Recent studies in our laboratory have shown that C-protein may be rapidly phosphorylated in vi tro or in vivo on four to five sites, and that C-protein is phosphorylated by cyclic AMP-protein kinase at a 2-3-fold higher rate than is troponin-I. Both of these proteins are phosphorylated to a low extent in control perfused hearts, and following catecholamine administration they are both phosphorylated over a time course similar to that of the increase in contraction. Phosphorylation of troponin decreases the Ca*+-sensitivity of either isolated cardiac myofibrils (Ray & England, 1976) or skinned cardiac fibres (Mope et al., 1980), and increase the rate of dissociation of Ca2+ from troponin (Robertson et al., 1982). The function of troponin-I phosphorylation is therefore most probably involved in the increased rate of relaxation of heart muscle caused by catecholamines. The function of both C-protein itself, and its phosphorylation, are at present unknown. Following a pulse of catecholamine both C-protein and troponin-I remain phosphorylated for many minutes, although the contractile response rapidly returns to normal (England, 1976; Stull et al., 1981). In contrast, phosphorylase a is rapidly dephosphorylated under these conditions. We have examined