Abstract The phosphatidylinositol (PtdIns) turnover pathway in intact heart tissue differs from that in most cell types in that products of the inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] kinase pathway are not detected in 3H-labeling studies. In contrast, Ins(1,4,5)P3 kinase products are detected in isolated neonatal cardiomyocytes. To understand the basis for the observed properties of the cardiac pathway, a detailed study of inositol phosphate (InsP) release has been undertaken by using isolated adult rat left atria. Addition of norepinephrine to 3H-labeled atria caused a slow increase in 3H-labeled Ins(1,4,5)P3 and a more rapid increase in 3H-labeled Ins(1,4)P2, its immediate dephosphorylation product. The mass of Ins(1,4,5)P3 was high in unstimulated atria (13.5±1.1 pmol/mg tissue, mean±SEM, n=4) and did not change with stimulation. Measurements of the specific activities of Ins(1,4,5)P3 and PtdIns(4,5)P2 provided an estimate of the turnover rate of Ins(1,4,5)P3 that was 20- to 40-fold lower than the rate of accumulation of 3H label in InsP1 and InsP2. In agreement with this, specific activities of InsP1 and InsP2 were higher than the specific activity of InsP3 in both control and stimulated atria. Neomycin (5 mmol/L) did not inhibit the accumulation of 3H-labeled InsP1 and InsP2 in left atria, even though it reduced the accumulation of 3H label in Ins(1,4,5)P3, providing evidence that InsP1 and InsP2 do not derive primarily from Ins(1,4,5)P3. Stimulation with norepinephrine for 20 minutes resulted in a parallel decrease in 3H-labeled Ins(1,4,5)P3 and in Ins(1,4,5)P3 mass, demonstrating that atria do not contain two different pools of Ins(1,4,5)P3. In contrast with findings in atria, the mass content of Ins(1,4,5)P3 was low in unstimulated cardiomyocytes and increased with norepinephrine stimulation. Furthermore, neomycin (5 mmol/L) inhibited the accumulation of 3H label in all of the InsPs, demonstrating that they are derived from Ins(1,4,5)P3. The data obtained when using left atria cannot be explained by models proposing either receptor-stimulated breakdown of Ins(1,4,5)P3 or the presence of more than one pool of Ins(1,4,5)P3 with different specific activities and are most readily explained by a model in which stimulation of receptors causes primarily the hydrolysis of PtdIns(4)P to Ins(1,4)P2. The high and unchanging content of Ins(1,4,5)P3 may be related, in some way, to its poor Ca2+-mobilizing activity in heart tissue.
- inositol 1,4,5-trisphosphate
- inositol 1,4-bisphosphate
- phosphatidylinositol 4-monophosphate
- left atria
- neonatal cardiomyocytes
The phosphatidylinositol (PtdIns) turnover pathway is a complex signaling system initiated by stimulation of phospholipase C specific for PtdIns(4,5)P2 and resulting in the generation of two well-established second messengers sn-1,2-diacylglycerol (DAG) and inositol 1,4,5-trisphosphate [Ins(1,4,5)P3]. DAG activates protein kinase C and Ins(1,4,5)P3 initiates release of Ca2+ from specific intracellular stores.1 In addition, the pathway is associated with the activation of a number of other phospholipases,2 lipid kinases,3 and protein kinases.4 The overall pattern of the response varies between different cell types5 and depends, to some extent, on the nature of the stimulus,6 providing a complex flexible control system that can initiate and sustain a wide variety of cellular responses. The inositol phosphate (InsP) response itself is complex and variable. In addition to metabolism by dephosphorylation, released Ins(1,4,5)P3 is metabolized by phosphorylation to Ins(1,3,4,5)P4,7 which may itself have a signaling function8 and is the precursor of the isomers of InsP5 as well as other isomers of InsP4.7 Some of these compounds probably also have a signaling function.
Previous studies in our laboratory have shown that this complex InsP response to stimulation is not observed when experiments are performed using intact 3H-labeled heart tissue preparations.9 Instead, 3H-labeled Ins(1,4,5)P3 appears to be broken down almost exclusively by dephosphorylation to [3H]Ins(1,4)P2 and [3H]Ins(4)P1.10 Other features of the heart pathway are also unusual. Heart tissue contains mainly Ca2+-independent isoforms of protein kinase C11 and has unusual fatty acid substitutions on its phosphorylated inositol phospholipids.12 Studies from a number of laboratories have shown that heart tissue is relatively insensitive to Ins(1,4,5)P3 in terms of Ca2+ release. Addition of high concentrations of Ins(1,4,5)P3 causes a slow leakage of Ca2+ rather than a rapid quantal release, as typically observed in other cell types.13 14 15 The BC3H-1 myocyte, in its differentiated form,16 and the longitudinal smooth muscle of the guinea pig intestine17 also are insensitive to Ins(1,4,5)P3 in terms of Ca2+ release. In addition, both of these cell types show InsP profiles that indicate low activity of the phosphorylation pathway of Ins(1,4,5)P3 metabolism.18 19 The finding of three different cell types that have a slow Ca2+ response to Ins(1,4,5)P3 and that have reduced ability to phosphorylate Ins(1,4,5)P3 suggests the possibility of a relation between Ins(1,4,5)P3 phosphorylation and its ability to release Ca2+. However, in addition to the apparent absence of phosphorylation products, the Ins(1,4,5)P3 response in heart tissue is unusual in that peaks of [3H]Ins(1,4,5)P3 are generally large compared with peaks of [3H]InsP2 or [3H]InsP1 and the increase in 3H label in Ins(1,4,5)P3 is minimal even though accumulations of 3H label in InsP2 and InsP1 can be quite large when the tissue is stimulated. The pattern of InsP accumulation in intestinal longitudinal smooth muscle cells20 appears to be similar. It is possible that the absence of phosphorylation products of Ins(1,4,5)P3 reflects another unusual property of the InsP response in these tissues.
The possibility that Ins(1,4,5)P3 metabolism was, in some way, related to its effectiveness as a Ca2+-mobilizing agent prompted us to make a detailed examination of the InsP pathway in heart tissue. Intact heart tissue cannot be labeled to equilibrium with [3H]inositol; therefore, these studies involved the use of combined labeling studies and mass measurements so that changes in specific activity could be quantified. Isolated adult rat left atria were used for the mechanistic studies reported in this article. However, the pathway in right atria and in ventricle appears to be similar in terms of InsP isomers, mass measurements, and effects of inhibitors (data not shown). Where appropriate, comparative experiments were performed with isolated neonatal cardiomyocytes, which, we have previously demonstrated, have a “normal” pattern of InsP metabolism.21
Materials and Methods
Generation of InsPs in Intact Adult Rat Left Atria
Adult Sprague-Dawley rats were killed by decapitation. Hearts were removed immediately and chilled in ice-cold HEPES-buffered Krebs’ medium constantly gassed with 95% O2/5% CO2. Atria were dissected and mounted in 3-mL organ baths filled with HEPES-buffered Krebs’ medium at 32°C and constantly gassed.10 After 15 minutes of equilibration, inositol phospholipids were labeled by incubating in medium containing [3H]inositol (40 μCi /mL) for 5 hours. Labeled atria were washed with nonradioactive medium containing 5 mmol/L mercaptoethanol (to inhibit oxidation of norepinephrine). LiCl (10 mmol/L) was added for 10-minute preincubation together with propranolol (1 μmol/L). Norepinephrine was then added to a final concentration of 100 μmol/L. At the end of the incubation time, atria were rapidly frozen in liquid N2 and subsequently extracted. Control atria were treated similarly, except norepinephrine was omitted. Norepinephrine was used for these studies instead of an α1-selective compound, such as phenylephrine, because such compounds are partial agonists in studies of InsP accumulation in heart and other tissues.22 We have previously shown that stimulation of β-adrenergic receptors does not alter the InsP response to α1-adrenergic receptor stimulation in rat heart.23
Extraction of InsPs
The experiments described required the use of different assay procedures for the InsPs, which, in turn, necessitated the use of different extraction procedures described below.
Identification and Quantification of the Isomers of the 3H-Labeled InsPs by Anion-Exchange High-Performance Liquid Chromatography
Frozen atria were homogenized in 2 mL of ice-cold 5% trichloroacetic acid (TCA) containing 5 mmol/L EDTA and 5 mmol/L phytic acid using a Polytron homogenizer. Two 10-second passes were used separated by 15-second chilling on ice. The extract was then further homogenized by use of a Potter-Elvejheim homogenizer, sonicated, and centrifuged to remove TCA-insoluble material. The TCA pellet was reextracted (1 mL), and the pooled supernatants were subsequently extracted with a 1:1 mixture of Freon and tri-n-octylamine (0.75 mL per 1 mL TCA) by using vigorous vortexing followed by low-speed centrifugation.10 24 Extracts were prepared for high-performance liquid chromatography (HPLC) by passing through a 1-mL column of Dowex-50 resin (4% cross-linked, 4 to 400) and eluting with 1 mL of water.25 Samples were then lyophilized before chromatography.
Mass Assay of Ins(1,4,5)P3 by the Binding Protein Method
Atria were incubated in medium for 5 hours as described above, except [3H]inositol was omitted when the atria were to be used for mass assay. The extraction procedure used was also similar to that described above, except phytic acid was replaced by ATP because phytic acid cross-reacts in the mass assay. Specific activity measurements involved experiments in labeled and unlabeled atria performed in parallel.
Mass Assay of InsPs (and Specific Activity Measurement) by Use of Fluorescence Measurements
The fluorescence assay of InsPs requires multiple enzyme reactions and is sensitive to interference by compounds present in the extraction media as well as compounds present in the tissue.26 The large volume of homogenization solution required to extract atria precluded the use of either TCA or perchloric acid, which were found to interfere with the fluorescence assay. Addition of either EDTA or phytic acid was avoided because of inhibitory effects on the subsequent assay. Frozen atria were extracted in 1 mL of CHCl3:CH3OH (1:2), and the tissue was homogenized as described above. Water (1 mL) was added, and the phases were separated by low-speed centrifugation. The aqueous phase was removed, and the organic phase together with the interface was reextracted twice with water. Extracts were passed through 1-mL columns of Dowex-50 resin and eluted with 1 mL of water.
Fluorimetric Assay of InsPs
Extracts prepared from 3H-labeled atria were separated into InsP1, InsP2, and InsP3 fractions with SepPak Accel Plus columns.26 Each fraction was divided into three portions. To one was added 100 pmol of the appropriate InsP standard: Ins(1)P1, Ins(1,4)P2, or Ins(1,4,5)P3. One of the remaining samples formed a blank with inositol dehydrogenase omitted during processing. Mass analysis of InsP1, InsP2, and InsP3 was carried out exactly as described by Maslanski and Busa,26 except fractions were incubated overnight with alkaline phosphatase at pH 7.5 instead of for 2 hours at pH 9.0. Calculation of mass was achieved by using the spiked sample to estimate recovery and the blank sample to assess inherent fluorescence of the extract as follows: mass of InsPx=(sample RFU−blank RFU)×pmol of spike/(spike RFU−sample RFU), where RFU indicates relative fluorescence units. Samples (1 mL) were counted in a β-counter after the fluorescence measurements had been made.
Analysis of InsP Isomers by Anion-Exchange HPLC
Analysis of the isomers of the InsPs was carried out by anion-exchange HPLC using a Whatman Partisil 10-μm SAX column (1 cm×10 cm) in a Waters radial compression unit. InsPs were eluted with a complex gradient of ammonium phosphate, pH 3.8, as described in detail elsewhere.27 The gradient program involved the use of gradients from 0 to 0.08 mol/L over 22 minutes, from 0.2 to 0.28 mol/L over 30 minutes, and from 0.5 to 0.56 mol/L over 25 minutes separated by steep gradients over 1 minute. The concentration of ammonium phosphate was increased from 0.56 to 2 mol/L over 3 minutes and maintained at this concentration for 30 minutes to remove tightly bound material. The column was reintroduced to water over 5 minutes and maintained in water for 60 minutes before further analysis. 3H-labeled compounds were detected and quantified by using an on-line β-counter (model CR, Radiomatic Instruments). This provided retention times and integrated peak values for each of the isomers of the InsPs. The identities of the various isomers were established by use of appropriate standards. Nucleotide standards ATP, ADP, and AMP (20 μg each) were added to each sample before injection to monitor any change in chromatographic performance.
Analysis of 3H-Labeled Inositol Phospholipids
TCA pellets, prepared as described above, were extracted into chloroform:methanol:HCl (200:100:1 [vol/vol/vol]) followed by evaporation under N2. The dried lipids were deacylated with methylamine:methanol:butanol (42:47:9 [vol/vol/vol]) at 50°C for 45 minutes.28 After evaporation, the residue was redissolved in water and reextracted with butanol:petroleum ether:ethyl formate (20:40:1 [vol/vol/vol]). The organic phase was washed once, and the combined aqueous phases containing the deacylated lipids were analyzed by HPLC, as described above.
Mass Measurements of Ins(1,4,5)P3 and PtdIns(4,5)P2 by Binding Protein Assay
For the mass measurement of Ins(1,4,5)P3, TCA extracts of atria were prepared as described above. The concentration of Ins(1,4,5)P3 was determined by competitive binding assay using a commercially available kit.29 For the measurement of PtdIns(4,5)P2, the lipids were extracted and deacylated as described above. Glycerol moieties were removed by treating with periodate (50 mmol/L) for 30 minutes. Reaction was terminated using 10% (vol/vol) ethylene glycol for 15 minutes, followed by 0.3% (vol/vol) dimethylhydrazine (pH 7) for 3 hours.29 The resulting Ins(1,4,5)P3 was assayed by competitive binding assay.
Calculation of Turnover Rate of Ins(1,4,5)P3
The specific activities of Ins(1,4,5)P3 and PtdIns(4,5)P2 were measured by using 3H-labeled and unlabeled atria in parallel experiments. The rise in specific activity of Ins(1,4,5)P3 was calculated over the first 1 minute of norepinephrine stimulation. This provided an estimate of the mass of PtdIns(4,5)P2 (high specific activity) required to be hydrolyzed to increase the specific activity of [3H]Ins(1,4,5)P3 by this amount over 1 minute. Since there was no increase in mass of Ins(1,4,5)P3, the mass of Ins(1,4,5)P3 metabolized must equal the mass of PtdIns(4,5)P2 hydrolyzed over 1 minute. The turnover rate of Ins(1,4,5)P3 was then converted to counts per minute per milligram tissue by using the calculated range of specific activity.
Measurement of Mass (and Specific Activity) of PtdIns(4)P
TCA pellets prepared as described above were extracted with CHCl3:CH3OH:HCl as described elsewhere.30 Mass assay of PtdIns(4)P was performed exactly as described,30 except PtdIns(4)P kinase was prepared from human platelet cytosol. [32P]PtdIns(4,5)P2 was separated by using Whatman silica gel 60A (LK6D) thin-layer chromatography plates and quantified by using a phosphoimage analyzer (BAS-1000, Fuji).
Preparation of Neonatal Cardiomyocytes
Cardiomyocytes were prepared from 1- to 2-day-old rats essentially as described elsewhere.21 After removal of the atria, great vessels, and pericardium, tissue was repeatedly digested in trypsin (0.05%) containing 0.01% deoxyribonuclease. Cells were harvested by low-speed centrifugation and resuspended in medium 199 containing 10% fetal calf serum, 2% chick embryo extract, and 100 U/mL each of penicillin and streptomycin. The cell preparation was then subjected to preplating for 30 minutes to remove any fibroblasts present. Bromodeoxyuridine (10 mmol/L) was then added to inhibit fibroblast growth, and the cells were plated onto wells in 24-well dishes at a density of 700 cells per square millimeter. Bromodeoxyuridine was omitted from the medium after 2 days in culture. The yield of viable cells averaged 106 per neonatal heart.
Measurement of Ins(1,4,5)P3 Phosphatase Activity
Tissue was homogenized in buffered salts solution containing (mmol/L) KCl 110, KH2PO4 10, MgSO4 5, EGTA 1, HEPES 20, and NaCl 10, along with Ca2+ to give a free Ca2+ concentration of 10 μmol/L. Homogenization was carried out using a Polytron homogenizer followed by a Potter-Elvejheim homogenizer and brief sonication. Homogenates were filtered through gauze and assayed immediately. Assay tubes contained glucose-6-phosphate and LiCl (10 mmol/L), in addition to the buffered salts solution, together with [3H]Ins(1,4,5)P3 (0.1 to 1 μmol/L, 20 to 50 cpm/pmol). Glucose-6-phosphate was added to inhibit nonspecific phosphatases. Reaction was stopped by adding HCOOH to 0.1 mol/L. Samples were diluted 1:10 and applied to 1-mL Dowex-1 formate columns. Ins(1,4,5)P3 hydrolysis products were eluted with 10 mL of 400 mmol/L HCOONH4 and 100 mmol/L HCOOH. The product of the reaction was identified as primarily Ins(1,4)P2 by HPLC analysis.
Protein concentration was measured by the method of Lowry et al30 using bovine serum albumin as standard.
Assay kits for measurement of Ins(1,4,5)P3, [3H]inositol, [3H]Ins(1,4,5)P3, and unlabeled Ins(1,4,5)P3 were obtained from the Radiochemical Centre Amersham. [3H]Ins(1,4)P2 and [3H]Ins(4)P1 were supplied by DuPont, New England Nuclear. Ins(1)P1 and Ins(1,4)P2 were from Sigma Chemical Co. All 3H-labeled InsPs were checked for purity by HPLC before use. Tissue culture media and enzymes were supplied by the Commonwealth Serum Laboratories. Sep Pak columns were from Millipore Waters. All other chemicals were analytical reagent grade, and reagents were dissolved in Milli Q water.
Accumulation of 3H-Labeled InsPs in Left Atria
Atria were labeled with [3H]inositol and subsequently stimulated with norepinephrine (100 μmol/L) in the presence of LiCl. 3H-labeled InsPs were extracted and analyzed. HPLC profiles of left atrial extracts demonstrated the presence of [3H]Ins(1,4,5)P3 and its dephosphorylation products [3H]Ins(1,4)P2 and [3H]Ins(4)P1. [3H]Ins(1/3)P1 also was present, but Ins(1,3,4,5)P4, Ins(1,3,4)P3, and the 1,3- and 3,4-isomers of InsP2 were not present at detectable levels in either stimulated or unstimulated preparations (Fig 1⇓). The time course of accumulation of the various isomers of the InsPs is shown in Fig 2⇓. Norepinephrine caused an increase in labeled InsPs, the most rapid rise being in Ins(1,4)P2, where significant rises were detectable after 15 seconds. This was followed by rises in Ins(1,4,5)P3 and Ins(4)P1. There was also a small rise in the 1 (or 3) isomer of InsP1.
Mass Changes in Ins(1,4,5)P3
The relatively small and slow rise in 3H-labeled Ins(1,4,5)P3 prompted measurement of the increase in mass on stimulation. Mass measurements by binding protein assay were made at 1- and 5-minute stimulation with norepinephrine, because the increase in [3H]label is maximal at these times. Addition of norepinephrine either for 1 or 5 minutes did not alter the mass of Ins(1,4,5)P3 as shown in Table 1⇓.
As a comparison, experiments were performed with neonatal cardiomyocytes, which have a “normal” profile of InsP metabolism.21 Measurements of mass of Ins(1,4,5)P3 were made after 15 seconds of stimulation, because increases detected in 3H-labeling studies were maximal at this time.21 Addition of norepinephrine to neonatal cardiomyocytes for 15 seconds produced a rise in Ins(1,4,5)P3 from 8.5±2.1 to 18.1±5 pmol/mg protein (mean±SEM, n=3, P<.01, paired t test). This is approximately equal to a rise from 0.5±0.14 to 1.2±0.3 pmol/mg tissue. The Ins(1,4,5)P3 response in neonatal cardiomyocytes differed from that in adult atria both in the lower amount of Ins(1,4,5)P3 detected in the tissue and in the mass increase under norepinephrine stimulation.
The finding of substantial accumulations of 3H-labeled InsP1 and InsP2 in the absence of any detectable increase in mass of Ins(1,4,5)P3 suggested unusual properties of the InsP pathway in atria. The following three models were proposed to account for these findings, and experiments were performed to evaluate their validity: (1) receptor-stimulated Ins(1,4,5)P3 breakdown, (2) two or more pools of Ins(1,4,5)P3 of different specific activities, and (3) receptor-stimulated PtdIns(4)P breakdown. The use of intact atria means that all experiments are unpaired. In all cases, each data point is derived from four atria, unless stated otherwise.
Receptor-Stimulated Ins(1,4,5)P3 Breakdown
This model proposes that breakdown of Ins(1,4,5)P3 balances release so that no mass increase is observed. The model assumes that there is one pool of Ins(1,4,5)P3 from which all detected InsPs derive. The observed increase in 3H label in Ins(1,4,5)P3 could be explained by this model as being due to nonequilibrium labeling. The model requires that the rate of turnover of Ins(1,4,5)P3 equals the total rate of accumulation of InsP1 plus InsP2 and that the specific activities of all InsPs are similar because they are all derived from a single pool of Ins(1,4,5)P3. Specific activities of Ins(1,4,5)P3 and PtdIns(4,5)P2 were calculated from parallel experiments using 3H labeling and mass analysis (Table 1⇑). The turnover rate of Ins(1,4,5)P3 over the first 1 minute of norepinephrine stimulation, calculated as described in “Materials and Methods,” was between 16 and 32 cpm/mg tissue. Over this 1-minute stimulatory period, total counts per minute per milligram tissue accumulating in InsP1 and InsP2 was 672 cpm/mg in the presence of LiCl, which inhibits the dephosphorylation of all isomers of InsP1 and partially inhibits the dephosphorylation of Ins(1,4)P2.31 Although these calculations are approximations, the magnitude of the difference between the turnover rate of Ins(1,4,5)P3 over the stimulatory period and the accumulations of Ins(1,4)P2 and Ins(4)P1 fails to support the proposed model. These calculations clearly assume the presence of only one pool of both Ins(1,4,5)P3 and PtdIns(4,5)P2. The possibility of more than one pool of Ins(1,4,5)P3 is discussed below.
Slow turnover of Ins(1,4,5)P3 suggests low activity of the metabolizing enzymes. Total activities of enzymes hydrolyzing Ins(1,4,5)P3 were measured in atrial extracts and compared with activities in homogenates of neonatal cardiomyocytes. Vmax values in left atrial extracts averaged 87±3 compared with 389±47 pmol/min per milligram protein in neonatal cell homogenates (mean±SEM, n=3, P<.001, unpaired t test). As a comparison with noncardiac tissue, the activity in extracts of cerebral cortex was measured and found to be similar to the activity in neonatal cardiomyocytes (454±74 pmol/min per milligram protein).
A second prediction of the receptor-stimulated Ins(1,4,5)P3 breakdown model is that InsP2, InsP1, and Ins(1,4,5)P3 should have similar specific activities at any time of stimulation. Specific activities of the different InsPs were quantified after norepinephrine stimulation by using a sensitive fluorescence measurement to assess mass and measuring the 3H label in the same extracts. This allowed specific activities to be measured for each InsP fraction within individual atrial extracts. Specific activities of the InsPs increased with stimulation. There was no significant difference between the specific activities of InsP2 and InsP1 at any time of stimulation, but at all stimulation times, these compounds had higher specific activities than InsP3 (Table 2⇓). These findings do not support receptor-stimulated Ins(1,4,5)P3 breakdown as a model of atrial PtdIns turnover.
Taken together, results from these three different experiments argue against the model that proposes receptor-stimulated breakdown of Ins(1,4,5)P3 as an explanation for the lack of change in mass of Ins(1,4,5)P3 under norepinephrine stimulation. However, apparent differences in turnover rates and in specific activities between Ins(1,4,5)P3 and the lower InsPs could be explained by compartmentalization of Ins(1,4,5)P3 or by generation of InsPs directly from PtdIns(4)P. Both of these possibilities are discussed below.
Two Pools of Ins(1,4,5)P3
This model proposes that atria contain a pool of unlabeled Ins(1,4,5)P3 that is not influenced by receptor stimulation and that all 3H-labeled InsPs derive from a small high-specific-activity pool of Ins(1,4,5)P3, which is released from PtdIns(4,5)P2 under norepinephrine stimulation.
If this model is valid, then the accumulation of all 3H-labeled InsPs should be sensitive to agents that inhibit [3H]Ins(1,4,5)P3 release. At high concentrations, neomycin binds to PtdIns(4,5)P2, producing a partial inhibition of release of Ins(1,4,5)P3 and reducing the accumulation of metabolites of Ins(1,4,5)P3.32 33 Therefore, the effects of neomycin on the accumulation of the various isomers of the InsPs was assessed in left atria. Labeled atria were incubated with neomycin (5 mmol/L) for 10 minutes before and during 20 minutes of stimulation with norepinephrine in the presence of LiCl. 3H-labeled InsPs accumulating over this period were extracted and quantified by HPLC. As is shown in Table 3⇓, addition of neomycin caused a 20% reduction in 3H label in Ins(1,4,5)P3 (P<.05, unpaired t test) but did not inhibit the accumulation of any of the other InsPs. Similar experiments were performed with isolated neonatal cardiomyocytes where neomycin (5 mmol/L) inhibited the accumulation of all isomers of the InsPs by >50%. These data indicate that 3H-labeled InsP2 and InsP1 are not derived from 3H-labeled Ins(1,4,5)P3 and are therefore not in agreement with the “two-pool” model.
A second prediction of the two-pool model is that any decrease in 3H label in Ins(1,4,5)P3 will not be accompanied by a decrease in mass, since the bulk of the Ins(1,4,5)P3 is functionally separate from the 3H-labeled compound.
Norepinephrine stimulation of 3H-labeled left atria increased 3H label in Ins(1,4,5)P3 from 121±10 to 217±5 cpm/mg tissue (mean±SEM, n=4, P<.01) at 1 minute, but continual stimulation for 20 minutes decreased the label to 94±6.7 cpm/mg tissue (P<.01 relative to 1-minute stimulation). The mass of Ins(1,4,5)P3 was unchanged after 1-minute stimulation (12±1.1 compared with 11.2±1.1 pmol/mg tissue) but decreased in parallel with the 3H value between 1 and 20 minutes to 5.5±1.3 cpm/mg tissue (P<.01). The parallel decrease in mass and 3H label demonstrates that atria do not contain two pools of Ins(1,4,5)P3 of different specific activities.
Taken together, these experiments do not support the concept of two different pools of Ins(1,4,5)P3 in atria.
Receptor-Stimulated PtdIns(4)P Hydrolysis
This model proposes that most of the accumulated InsP2 and InsP1 derive from phospholipase C cleavage of PtdIns(4)P to Ins(1,4)P2. Some breakdown of PtdIns(4,5)P2 also occurs, as demonstrated by the increase in label in Ins(1,4,5)P3, but this is quantitatively less important than PtdIns(4)P breakdown. This model accommodates all of the data presented above.
The model requires that the specific activity of PtdIns(4)P is similar to that of Ins(1,4)P2, at least after stimulation of InsP release. Specific activity of PtdIns(4)P was measured by using parallel mass and 3H-labeling experiments, as described in “Materials and Methods.” Atria contained 2.9±0.7 pmol/mg tissue of PtdIns(4)P (mean±SEM), and the calculated specific activity was 76 cpm/pmol (calculated from the average value for mass and 3H labeling). There was no significant change in mass or specific activity with norepinephrine stimulation for 1 minute. The PtdIns(4)P mass averaged 4.4±0.9 pmol/mg, and the specific activity was 70 cpm/pmol. Thus, the specific activity of PtdIns(4)P is higher than that of PtdIns(4,5)P2 (Table 1⇑) and similar to that of InsP2 after norepinephrine stimulation (Table 2⇑). These data support the proposed model of receptor-stimulated PtdIns(4)P breakdown.
Taken together, these experiments exclude receptor stimulation of Ins(1,4,5)P3 breakdown and two pools of Ins(1,4,5)P3 as possible explanations for the unusual properties of the cardiac PtdIns turnover pathway. However, all data are compatible with the model in which activation of PtdIns turnover in left atria involves primarily the stimulation of release of Ins(1,4)P2 from PtdIns(4)P.
Left atria were found to contain a high concentration of Ins(1,4,5)P3, which did not change appreciably after norepinephrine stimulation over a 5-minute period, even though there was substantial accumulation of InsP1 and InsP2. Three models were considered to explain these findings.
The first model proposes that Ins(1,4,5)P3 hydrolysis balances release. A number of experimental findings were inconsistent with this model. First, the total rate of accumulation of InsP1 and InsP2 was 20 to 40 times higher than the turnover rate of Ins(1,4,5)P3 calculated from specific activity measurements. In agreement with this, the activity of Ins(1,4,5)P3 phosphatase was low in atria compared with activity in neonatal cardiomyocytes or cerebral cortex. Second, the specific activities of [3H]InsP1 and [3H]InsP2 were shown to be 10 times higher than the specific activity of [3H]InsP3. In addition, if InsP1 and InsP2 derive from Ins(1,4,5)P3, their accumulations would be expected to be inhibited by neomycin, which was found not to be the case.
The second model hypothesizes the presence within atria of compartmentalized pools of low-specific-activity Ins(1,4,5)P3 that are not in equilibrium with the Ins(1,4,5)P3 released after norepinephrine stimulation. This model accommodates the differences in the turnover rates and specific activities of the different InsPs discussed in the previous paragraph. However, it is not consistent with the lack of effect of neomycin on the accumulations of the isomers of InsP1 and InsP2 and does not explain the slow increase in 3H label in Ins(1,4,5)P3 relative to the increase in Ins(1,4)P2. Furthermore, it is not compatible with the finding that at 5 minutes, the specific activities of both InsP1 and InsP2 were higher than the specific activity of PtdIns(4,5)P2. Most important, the compartmentalization model requires that any decrease in [3H]Ins(1,4,5)P3 will not be paralleled by decreases in mass, whereas parallel decreases in mass and label were observed.
The third model requires that receptor stimulation activates primarily the hydrolysis of PtdIns(4)P to Ins(1,4)P2, which then breaks down primarily to Ins(4)P1. This model accommodates all of the data obtained in the present studies. Additionally, the specific activity of PtdIns(4)P was shown to be higher than that of PtdIns(4,5)P2 and similar to that of Ins(1,4)P2 after norepinephrine stimulation. Thus, our data suggest that PtdIns(4)P is more important than PtdIns(4,5)P2 as the precursor of the InsPs in left atria. Clearly, a small amount of breakdown of PtdIns(4,5)P2 also occurs in atria because a small rise in labeling of Ins(1,4,5)P3 was observed after norepinephrine stimulation, at least over the first 1 minute. The pathway in heart in comparison with that in other tissues is shown in Fig 3⇓.
The different pathway in intact heart tissue was not due to the presence of nonmuscle cells, because the specificity of the stimulation of the pathway was not compatible with other cell types present, as described previously.23 Furthermore, the pathway in isolated adult cardiomyocytes (which are more differentiated than the neonatal cells) is essentially similar to the pathway in intact tissue.34
We have recently reported that the isolation of neonatal cardiomyocytes causes a reappearance of phosphorylation products of Ins(1,4,5)P3 such that InsP metabolism in these cells appears to be similar to that described in nonmuscle cell types.21 Compared with adult left atria, these cells contain low concentrations of Ins(1,4,5)P3 (0.5±0.14 pmol/mg tissue compared with 13.5±1.1 pmol/mg tissue in atria), and the mass of Ins(1,4,5)P3 was found to increase on stimulation. In addition, the accumulation of InsPs was sensitive to neomycin, and the activity of Ins(1,4,5)P3 phosphatase was higher in myocytes than in atria. Thus, in neonatal cardiomyocytes a large percentage of the InsPs generated under norepinephrine stimulation derive from Ins(1,4,5)P3 and therefore from cleavage of PtdIns(4,5)P2. This shows that the differences between the pathway in neonatal cardiomyocytes and in adult left atria cannot be explained simply by a quantitative difference in Ins(1,4,5)P3 phosphorylation. Instead, the low level of kinase products most likely reflects the generally slow metabolism of Ins(1,4,5)P3 and its low specific activity. Thus, the primary change induced by the isolation of neonatal cardiomyocytes most likely involves a switch in the primary lipid substrate from PtdIns(4)P to PtdIns(4,5)P2, together with an activation of Ins(1,4,5)P3 phosphatase.
The proposed model explains our previous finding of low levels of phosphorylation products of [3HIns(1,4,5)P3 in intact heart tissue.9 10 21 Any metabolites of Ins(1,4,5)P3 in heart, including both phosphorylation and dephosphorylation products, will have low specific activity and would be difficult to detect in labeling studies unless they were present at relatively high concentrations. In agreement with this, a recent study using mass analysis of inositol polyphosphates in ventricle showed low concentrations of the isomers of InsP4 and InsP5 relative to a very high concentration of Ins(1,4,5)P3.35 These are unlike profiles reported from other tissues, in which Ins(1,3,4,6)P4 and Ins(1,3,4,5,6)P5 are present at much higher concentrations than Ins(1,4,5)P3.36 37
The discovery that Ins(1,4,5)P3 (but not the isomers of InsP2 or InsP1) induced Ca2+ release led to the concept that PtdIns(4,5)P2 was the primary substrate of receptor-stimulated phospholipase C.38 Subsequently, it was suggested that breakdown of PtdIns and possibly PtdIns(4)P might be important during prolonged stimulation to provide a continued source of DAG.39 Subsequent studies have questioned these ideas on the basis that the DAG formed after prolonged stimulation often derives from phosphatidylcholine rather than the inositol phospholipids.40 However, recent studies have reopened the debate by demonstrating rapid breakdown of PtdIns in islets where both PtdIns and PtdIns(4,5)P2 serve as substrates for receptor-stimulated phospholipase C activity33 apparently under two different control mechanisms.41 In left atria, PtdIns(4)P hydrolysis appears to predominate over either PtdIns or PtdIns(4,5)P2. The turnover rate of Ins(1,4,5)P3 compared with the accumulations of InsP1 and InsP2 argues against a major contribution from PtdIns(4,5)P2, as does the insensitivity to neomycin. It is likely that some breakdown of PtdIns also occurs, but this contribution must be small because the major isomer of InsP1 was Ins(4)P1, which can only be formed from Ins(1,4)P2. At high concentrations of Ins(1,4)P2, some hydrolysis to Ins(1)P1 occurs.7 Thus, it is not possible to estimate the contribution of direct PtdIns breakdown.
Several laboratories have reported that heart tissue is relatively insensitive to Ins(1,4,5)P3 in terms of Ca2+ release. Addition of Ins(1,4,5)P3 causes a slow leakage of Ca2+ rather than the rapid release seen in most tissues.13 14 Skeletal muscle, BC3H-1 myocytes, and longitudinal smooth muscle cells of guinea pig intestine also are insensitive to Ins(1,4,5)P3.16 17 It is of interest that both of these two latter cell types also show a low percentage of Ins(1,4,5)P3 kinase products in response to stimulation of the PtdIns turnover pathway.18 19 The finding of three different cell types (heart, BC3H-1 myocytes, and longitudinal smooth muscle), which are insensitive to Ins(1,4,5)P3 and apparently lack kinase products, suggests a relation between these two properties. At least in heart tissue, the lack of 3H-labeled kinase products is most likely related to the slow generation and metabolism of Ins(1,4,5)P3, together with its low specific activity. This raises the question as to whether the PtdIns turnover pathway in these other muscle preparations also involves breakdown of PtdIns(4)P, although the proportions of PtdIns(4)P to PtdIns(4,5)P2 breakdown may be different from that in atria. There is some evidence for this in the case of longitudinal smooth muscle cells. First, compared with the response in circular smooth muscle cells, which are Ins(1,4,5)P3 sensitive, the rise in mass of Ins(1,4,5)P3 was extremely small, even though the total accumulation of InsPs was similar in the two cell types.20 Second, breakdown of PtdIns(4)P but not PtdIns(4,5)P2 was obvious in longitudinal muscle, whereas both decreased in the cells from the circular muscle.20 Thus, it is likely that this unusual form of the PtdIns turnover pathway, involving primarily PtdIns(4)P breakdown and high concentrations of Ins(1,4,5)P3, occurs in some other muscle tissues as well as in heart. It is of interest that a recent study has demonstrated very high concentrations of Ins(1,4,5)P3 in skeletal muscle, similar to concentrations in left atria.42
A clear role for the PtdIns turnover pathway in the control of cardiac contractility has not been established. Under physiological conditions, sympathetic stimulation operates largely by activation of β-adrenergic receptors acting via increases in cAMP content.43 Activation of α1-adrenergic receptors can increase contractility, but the effects are small compared with the response to β-receptor stimulation.44 An involvement of InsPs in the inotropic response to α1-receptor stimulation also has not been established. There is no obvious correlation between the InsP response and the inotropic response in different species.45 In addition, Ca2+ released by Ins(1,4,5)P3 in heart tissue does not stimulate Ca2+-induced Ca2+ release, the mechanism that ultimately controls the contraction cycle.15 The heart contains only low concentrations of Ins(1,4,5)P3 receptors, and these have been reported to be mainly on the conduction fibers46 rather than on the cardiomyocytes, and those that are present on the myocytes appear to be on the intercalated disks rather than on the sarcoplasmic reticulum.47 On these grounds, Ins(1,4,5)P3 is unlikely to be a major controller of cardiac contractility under physiological conditions. It seems more likely that activation of protein kinase C, possibly in association with activation of the Na+-H+ exchanger,48 is the mechanism relating PtdIns turnover to cardiac inotropic responses. In this case, a change in lipid substrate from PtdIns(4,5)P2 to PtdIns(4)P would not jeopardize the release of DAG and consequent simulation of protein kinase C.
The question remains as to the mechanism that determines that PtdIns(4)P rather than PtdIns(4,5)P2 will be hydrolyzed after stimulation of intact heart tissue. It is possible that a specific “heart-type” phospholipase C exists that shows preference for PtdIns(4)P as substrate. Alternatively, some feature of the structure or composition of the cardiac sarcolemma might result in a change in the specificity of a phospholipase C isoform that would otherwise hydrolyze PtdIns(4,5)P2. It is of interest that the fatty acid substitutions on PtdIns(4)P and PtdIns(4,5)P2 in rat heart are different from those in other tissues.12 The changes that occur when neonatal cardiomyocytes are isolated suggest that the switch from PtdIns(4)P to PtdIns(4,5)P2 as preferred substrate is accompanied by an activation of Ins(1,4,5)P3 metabolism. Furthermore, the spectrum of isoforms of protein kinase C in neonatal cardiomyocytes is different from that in adult heart.11 Taken together, these findings suggest the involvement of a coordinate control mechanism. Any such control system might also be responsible for the suppression of Ca2+ responses to Ins(1,4,5)P3. Further studies are required to establish these relations and their importance in cardiac physiology and pathophysiology.
This study was supported by the National Heart Foundation of Australia and by the Australian National Health and Medical Research Council. K. Anderson is the recipient of a Dora Lush Bio-Medical Post-Graduate Scholarship. We would like to thank Dr Robert Andrews (Vascular Biology Laboratory, Baker Medical Research Institute) for help in preparing PtdIns(4)P kinase.
- Received May 13, 1994.
- Accepted October 6, 1994.
- © 1995 American Heart Association, Inc.
Nishizuka Y. Intracellular signaling by hydrolysis of phospholipids and activation of protein kinase-C. Science. 1992;258:607-614.
Wahl MI, Jones GA, Nishibe S, Rhee SG, Carpenter G. Growth factor stimulation of phospholipase-C-gamma1 activity: comparative properties of control and activated enzymes. J Biol Chem. 1992;267:10447-10456.
Rosenberg SM, Berry GT, Yandrasitz JR, Grunstein MM. Maturational regulation of inositol 1,4,5-trisphosphate metabolism in rabbit airway smooth muscle. J Clin Invest. 1991;88:2032-2038.
Plevin R, Macnulty EE, Palmer S, Wakelam M. Differences in the regulation of endothelin-1-stimulated and lysophosphatidic-acid-stimulated Ins(1,4,5)P3 formation in Rat-1 fibroblasts. Biochem J. 1991;280:609-615.
Shears SB. Metabolism of inositol phosphates produced upon receptor activation. Biochem J. 1989;260:313-324.
Bogoyevitch M, Parker P, Sugden P. Characterization of protein kinase C isotype expression in adult rat heart: protein kinase C-ε is a major isotype present, and it is activated by phorbol esters, epinephrine, and endothelin. Circ Res. 1993;72:757-767.
Kentish JC, Barsotti RJ, Lea TJ, Mullican IP, Patel JR, Ferenczi MA. Calcium release from cardiac sarcoplasmic reticulum induced by photorelease of calcium or Ins(1,4,5)P3. Am J Physiol. 1990;258:H610-H615.
De Smedt H, Parys JB, Himpens B, Missiaen L, Borghgraef R. Changes in the mechanism of Ca2+ mobilization during the differentiation of BC3H1 muscle cells. Biochem J. 1991;273:219-224.
Murthy KS, Grider JR, Makhlouf GM. InsP3-dependent Ca2+ mobilization in circular but not longitudinal muscle cells of intestine. Am J Physiol. 1991;261:G937-G944.
Ambler SK, Thompson B, Solski PA, Brown JH, Taylor P. Receptor-mediated inositol phosphate formation in relation to calcium mobilization: a comparison of two cell lines. Mol Pharmacol. 1987;32:376-383.
Salmon D, Boulton T. Early events in inositol phosphate metabolism in longitudinal smooth muscle from guinea pig intestine stimulated with carbachol. Biochem J. 1988;254:553-557.
Murthy KS, Makhlouf GM. Phosphoinositide metabolism in intestinal smooth muscle: preferential production of Ins(1,4,5)P3 in circular muscle cells. Am J Physiol. 1991;261:G945-G951.
Woodcock EA, Tanner JK, Fullerton M, Kuraja IJ. Different pathways of inositol phosphate metabolism in intact neonatal rat hearts and isolated cardiomyocytes. Biochem J. 1992;281:683-688.
Woodcock EA, White LBS, Smith AI, McLeod JK. Stimulation of phosphatidylinositol metabolism in the isolated, perfused rat heart. Circ Res. 1987;61:625-631.
Downes C, Hawkins P, Irvine R. Inositol 1,3,4,5-tetrakisphosphate and not phosphatidylinositol 3,4-bisphosphate is the probable precursor of inositol 1,3,4-trisphosphate in agonist-stimulated parotid gland. Biochem J. 1986;238:501-506.
Maslanski J, Busa W. A sensitive and specific mass assay for myo-inositol and inositol phosphates. In Irvine RF, ed. Methods in Inositide Research. New York, NY: Raven Press Publishers; 1990:113-126.
Auger K, Serunian L, Cantley L. Separation of novel polyphosphoinositides. In Irvine R, ed. Methods in Inositide Research. New York, NY: Raven Press Publishers; 1990:159-166.
Divecha N, Banfic H, Irvine RF. The polyphosphoinositide cycle exists in the nuclei of Swiss 3T3 cells under the control of a receptor (for IGF- I) in the plasma membrane, and stimulation of the cycle increases nuclear diacylglycerol and apparently induces translocation of protein kinase-C to the nucleus. EMBO J. 1991;10:3207-3214.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265-275.
Hughes PJ, Drummond AH. Formation of inositol phosphate isomers in GH3 pituitary tumour cells stimulated with thyrotropin-releasing hormone: acute effects of lithium ions. Biochem J. 1987;248:463-470.
Biden TJ, Prugue ML, Davison A. Evidence for phosphatidylinositol hydrolysis in pancreatic islets stimulated with carbamoylcholine: kinetic analysis of inositol polyphosphate metabolism. Biochem J. 1992;285:541-549.
Scholz J, Troll U, Sandig P, Schmitz W, Scholz H, Esch J. Existence and alpha1-adrenergic stimulation of inositol polyphosphates in mammalian heart. Mol Pharmacol. 1992;42:134-140.
Mayr GW. A novel metal-dye detection system permits picomolar-range h.p.l.c. analysis of inositol polyphosphates from non-radioactively labeled cell or tissue specimens. Biochem J. 1988;254:585-591.
Balla T, Hunyady L, Baukal A, Catt K. Structures and metabolism of inositol tetrakisphosphates and inositol pentakisphosphates in bovine adrenal glomerulosa cells. J Biol Chem. 1989;264:9386-9390.
Griendling K, Rittenhouse S, Brock T, Ekstein L, Gimbrone GM Jr, Alexander RW. Sustained diacylglycerol formation from inositol phospholipids in angiotensin II-stimulated vascular smooth muscle cells. J Biol Chem. 1986;261:5901-5906.
Lassegue B, Alexander RW, Clark M, Akers M, Griendling KK. Phosphatidylcholine is a major source of phosphatidic acid and diacylglycerol in angiotensin-II-stimulated vascular smooth-muscle cells. Biochem J. 1993;292:509-517.
Biden TJ, Davison A, Prugue ML. Regulation by membrane potential of phosphatidylinositol hydrolysis in pancreatic islets. J Biol Chem. 1993;268.11065-11072.
Hidalgo C, Jorquera J, Tapia V, Donoso P. Triads and transverse tubules isolated from skeletal muscle contain high levels of inositol 1,4,5-trisphosphate. J Biol Chem. 1993;268:15111-15117.
Scholz H, Bruckner R, Mugge A, Repucke C. Myocardial alpha- adrenoceptors and positive inotropy. J Mol Cell Cardiol. 1986;18:79-87.
Endoh M, Hiramoto T, Ishihata A, Takanashi M, Inui J. Myocardial α1-adrenoceptors mediate positive inotropic effect and changes in phosphatidylinositol metabolism: species differences in receptor distribution and the intracellular coupling process in mammalian ventricular myocardium. Circ Res. 1991;68:1179-1190.
Gorza L, Schiaffino S, Volpe P. Inositol 1,4,5-trisphosphate receptor in heart: evidence for its concentration in Purkinje myocytes of the conduction system. J Cell Biol. 1993;121:345-353.
Kijima Y, Saito A, Jetton TL, Magnuson MA, Fleischer S. Different intracellular localization of inositol 1,4,5-trisphosphate and ryanodine receptors in cardiomyocytes. J Biol Chem. 1993;268:3499-3506.
Fedida D. Modulation of cardiac contractility by α1 adrenoceptors. Cardiovasc Res. 1993;27:1735-1742.