INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE SIGNAL-TRANSDUCTION events that stimulate functional responses in many cell systems are associated with activation of phospholipase C (PLC) and the hydrolysis of phosphatidylinositol 4,5-diphosphate (PIP₂), resulting in the production of inositol 1,4,5-trisphosphate (IP₃) and sn-1,2-diacylglycerol (DAG). IP₃ causes the release of Ca²⁺ from intracellular stores, producing a Ca²⁺-calmodulin complex, which (in smooth muscle) activates myosin light-chain kinase and causes phosphorylation of the 20,000-Da myosin light chain, resulting in contraction. DAG causes activation of protein kinase C (PKC), a family of isoenzymes implicated in signal transduction, which induces sustained contraction of smooth muscle cells (4).

Early studies had focused on PLC-mediated phosphoinositide hydrolysis as an intracellular signaling system for DAG generation. The establishment of DAG as a second messenger was confirmed by the measurement of the increase in the mass of DAG after stimulation, and the enhancement of stearic and arachidonic acids in the incremental fraction of DAG supported the involvement of phosphoinositide hydrolysis (27). More recently, it has become increasingly apparent that nonphosphoinositide sources of DAG, especially arising from agonist-stimulated PLC- and PLD-activated phosphatidylcholine (PC) breakdown, exist (11, 27). DAG coproduced from hydrolysis of other phospholipid sources, such as PC, is important in the PKC-mediated pathway for contraction (3, 9, 11). The activation of PKC by DAG appears to play a pivotal role in signal- transduction events involved in such processes as growth, differentiation, and contraction (25). Evidence has been cited in support of a PKC-mediated pathway as an alternative signaling mechanism for regulating excitation-contraction coupling in smooth muscle and explains phorbol-induced contraction of smooth muscle without increases in cytosolic Ca²⁺ or in myosin light- chain phosphorylation (17). It was of interest to us, therefore, to examine the contribution of nonphosphoinositide DAG sources in isolated smooth muscle cells of the rabbit rectosigmoid. For this purpose, the use of benzoylated DAG (-DAG) derivatives and their analysis by HPLC, known to be both sensitive and reliable (18, 19), was adopted as the preferred method of analysis. This technique allows the quantitative measurement of the actual new DAG mass formed, the determination of which may be somewhat complicated with the use of radiolabeling approaches. Another significant advantage of using this chromatographic method over a popular radiometric assay, which relies on enzymatically phosphorylating DAG by DAG kinase with [-³²P]ATP and quantitating the resulting [³²P]phosphatidic acid, is that it allows a profiling, or fingerprinting, of the different molecular species of DAG (18).

We focused on the time course of formation of the molecular species of newly generated DAG to 1) deduce the phospholipid origin of the new DAG formed during bombesin-induced contraction and 2) measure the approximate relative contributions of these phospholipid DAG precursors. The data indicate that PC appears to be the major phospholipid source for newly generated DAG during both short- and long-term stimulation with bombesin. In addition, our data support the hypothesis that PKC plays an important role as a mediator of this contractile response in rabbit rectosigmoid smooth muscle cells. This in turn not only supports similar findings made by others but also reaffirms our knowledge of the importance of nonphosphoinositide phospholipids as participants in second-messenger generation during the early stages of upstream signaling at the cell membrane.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Materials

HPLC assay. HEPES, Tris, phospholipid standards, Triton X-100, 1,2-distearoyl-rac-glycerol (DSG), 14% (wt/vol) boron trifluoride-methanol, 4-pyrrolidinopyridine, benzoic anhydride, B. cereus PLC, bombesin, 1-(5-isoquinolinylsulfonyl)-2-methyl-piperazine, and cyclohexane (redistilled) were from Sigma Chemical. Ham's F-12 medium, DMEM, and N-2 supplement were obtained from GIBCO (Grand Island, NY). PUFA-2 mix fatty acid methyl ester standard for gas liquid chromatography was from Matreya (Pleasant Gap, PA). HPLC grade solvents were from Mallinckrodt Specialty Chemicals (Paris, KY) or Burdick & Jackson (Baxter Healthcare, Muskegon, MI). An HPLC solvent-delivery system consisted of two Waters M-45 pumps (Milford, MA) coupled to a Rheodyne model 7125 syringe-loading rotary sample-injector valve equipped with a 50-µl sample loop (Rheodyne Cotati). Analytic Silica gel 60-precoated glass TLC plates were from Merck (Darmstadt, Germany). All other reagents were purchased from Sigma Chemical.

Methods

Isolation of smooth muscle cells from rabbit rectosigmoid. The internal anal sphincter, consisting of the most distal 3 mm of the circular muscle layer, ending at the junction of skin and mucosa, was removed by sharp dissection. A 5-cm length of the rectosigmoid orad to the junction was dissected and digested to yield isolated smooth muscle cells. Cells were isolated as previously described (4, 5). The tissue was incubated for two successive 60-min periods at 31°C in 15 ml of HEPES buffer (pH 7.4) containing 0.1% (wt/vol) collagenase (150 U/mg, Worthington CLS type II) and 0.01% (wt/vol) soybean trypsin inhibitor. At the end of the second enzymatic incubation period, the medium was filtered through 500-µm Nitex mesh. The partially digested tissue left on the filter was washed four times with 50 ml of collagenase-free buffer solution. The tissue was then transferred into 15 ml of fresh buffer solution and incubated for 30 min to allow the cells to disperse spontaneously. After a hemocytometric cell count, the harvested cells were resuspended in maintenance medium, pH 7.4, containing Ham's F-12 medium, DMEM, and N-2 supplement (50:50:1, vol/vol) and aliquoted into glass scintillation vials in 1-ml portions. These cells were allowed to equilibrate for 90 min at 37°C in a humidified 95% air-5% CO₂ (vol/vol) incubator atmosphere before either control, inhibitor, or agonist-stimulated treatments were initiated. Each rectosigmoid yielded 15-20 × 10⁶ cells.

Biosynthesis and extraction of DAG. Smooth muscle cells at a concentration of 1.5 × 10⁶ cells/ml were dispensed in 200-µl aliquots in 16 × 100 mm glass tubes. The tubes were incubated in a humidified CO₂ incubator at 37°C and allowed to equilibrate for 2 h. An aliquot of cells was examined under the microscope, and cell measurement was performed at this stage. We then added 22-µl aliquots of DMEM or agonists at 10 times the desired concentration, and the reaction was stopped at the appropriate times by adding 0.6 ml ice-cold 1% perchloric acid and placing the tubes in a melting ice bath. We then added 3 ml ice-cold chloroform-methanol (1:1, vol/vol), and the lipids were extracted for 15 min. To separate the phases, 1 ml of 1 M NaCl and 1 ml of CHCl₃ were added to each tube and vortexed. The tubes were then centrifuged at 1,500-2,000 g for 5 min. Taking care not to disturb the interface, we then aspirated and discarded part of the aqueous top layer. A 1.5-ml aliquot of the organic lower layer was then removed from each tube, evaporated to dryness under N₂, and capped immediately under N₂. The extracted lipids were assayed within 24 h.

DAG (phosphatidic acid) assay. To each cell extract and/or diolein standard, 20 µl of micelles prewarmed to 37°C were added, and the mixture was vortexed and sonicated for 10 min at room temperature. An assay buffer was prepared for each sample (50 µl of 2× assay buffer, 10 µl of 20 mM DTT, 10 µl DAG kinase, and 5 µl of 20 mM fresh ATP), and the mixture was allowed to incubate at room temperature for 20 min. Then 1 µCi [-³²P]ATP per sample was added, and 80 µl of this mixture were added to each sample and allowed to incubate for 30 min at 25°C. The reaction was stopped with 0.7 ml of cold 1% HClO₄. The lipids were then extracted by adding 3 ml CHCl₃/MeOH (1:2, vol/vol) and vortexing. Phases were then separated by adding 1 ml of 1 M NaCl and 1 ml CHCl₃, followed by vortexing and centrifugation. We then evaporated 1 ml of the organic layer to dryness under N₂. The dried lipid was reconstituted in 5% MetOH CH₃OH/MeOH in CHCl₃ and applied to preactivated TLC plates (Silica gel 60). The plates were placed in a preequilibrated chromatography chamber containing a mixture of CHCl₃, MeOH, acetone, acetic acid, and water (30:9.8:11.3:9:5.3, vol/vol) for 30 min and then exposed to X-ray film (X-OMat, XAR-2). The positions of phosphatidic acid spots were determined on the film, and then they were scraped from the TLC plate and quantified by liquid scintillation counting in 5 ml Ecolume (ICN Radiochemicals, Costa Mesa, CA).

Preparation of whole lipid extracts. After an appropriate incubation period with or without agonist at 37°C, 3.4 ml of 1:2 (vol/vol) chloroform-methanol quenching solution was added to each 1-ml cell suspension (10-20 × 10⁶ cells), and the mixture was shaken vigorously. We then added 2-3 nmol of DSG internal standard to each sample and mixed them well. The samples were sonicated for 20 s in an ultrasonic bath and then centrifuged at 10,000 g for 10 min. The supernatants were collected, and the pellets were reextracted with 2 ml of 1:1 (vol/vol) chloroform-methanol, sonicated, and centrifuged as described above. With the combined supernatants from each sample, 1 ml of chloroform was mixed in well and then 1.4 ml of 0.9% (wt/vol) NaCl was added and mixed by vortexing. The aqueous and organic layers were resolved by centrifuging at 10,000 g for 5 min. After the upper layer was removed, the lower layer was dried under blowing N₂ and redissolved with 200-500 µl of 2:1 (vol/vol) chloroform-methanol. The DAGs were immediately purified from this extract by TLC and then quickly benzoylated as described below. Otherwise, the extract was stored under N₂ at 20°C for isolation and analysis of phospholipids.

Purification and quantitation of phospholipids. Individual major phospholipid bands [i.e., PC, phosphatidylserine (PS), phosphatidylinositol (PI), CL, phosphatidylethanolamine (PE), and sphingomyelin (SM)] were purified from neutral lipid extracts by applying ~50-100 nmol of the phospholipids as 2-cm bands on 20 × 20-cm Silica gel 60 glass TLC plates (0.25-mm layer thickness) and doubly developing up to 1 cm from the top using the chloroform-methanol-0.25% (wt/vol) KCl-ethyl acetate-isopropanol (150:45:30:90:125, vol/vol) solvent system of Hedegaard and Jensen (14). The bands were visualized by spraying them with 0.001% (wt/vol) primulin in 4:1 (vol/vol) acetone-water and circling them with a pencil under long-wave ultraviolet (365 nm) light. Then we scraped each band and equivalent blank areas as blanks. The amount of each phospholipid was determined by washing the bands with Mg(NO₃)₂ over a flame and measuring the liberated total P_i with acid-molybdate reagent by the method of Ames (2). Using a standard curve prepared simultaneously with potassium dihydrogen phosphate as a standard, we determined the amount of phospholipid from the absorbance measured at 660 nm after removal of silica by centrifuging at 1,000 g for 5 min.

PLC treatment of phospholipids or whole lipid extracts. Individual phospholipid bands purified by TLC [chloroform-methanol-0.25% (wt/vol) KCl-ethyl acetate-isopropanol (150:45:30:90:125, vol/vol)] as described above were extracted with 2 × 3 ml of 2:1 (vol/vol) chloroform-methanol. For PI, 2 × 3 ml of 2:1 (vol/vol) chloroform-methanol containing 0.25% (wt/vol) HCl were used instead, and the HCl was removed by shaking with 1.5 ml of water, centrifuging at 1,000 g for 5 min to separate the phases, and collecting the lower layer. After the extracts were pooled from several bands and dried under N₂, they were treated as described below for the preparation of DAGs derived from whole lipid extracts. Dried whole lipid extracts (20 × 10⁶ cells) or purified phospholipid extracts (as described above) were sonicated for 1 min with 0.3 ml of 0.25% (wt/vol) Triton X-100 in 50 mM Tris · HCl buffer (pH 7.5) containing 10 U of B. cereus PLC and incubated overnight at 37°C under N₂ in sealed glass test tubes. The liberated DAGs were extracted with 5 ml of 2:1 (vol/vol) chloroform-methanol and 1 ml of 0.9% (wt/vol) NaCl; the lower layer collected after centrifuging at 1,000 g for 5 min was dried under N₂, redissolved with 200 µl (600 µl for whole lipid extracts) of 2:1 (vol/vol) chloroform-methanol, and then purified by TLC and benzoylated as described below.

DAG purification and benzoylation protocol. Samples of 200 µl of whole lipid extract or phospholipid from PLC treatment were applied as 2-cm bands on a Silica gel 60 20 × 20 cm glass plate (0.25-mm layer thickness) and developed in a mixture of toluene, ether, and methanol (80:10:10, vol/vol) up to 10 cm from the origin. The DAG band [retention factor (R_f) = 0.58], corresponding with the internal standard, DSG (R_f = 0.58), was visualized by spraying with 0.001% (wt/vol) primulin in acetone-water (4:1, vol/vol) and then circling the bands with a pencil under long-wave ultraviolet (365 nm) light. The DAG bands were scraped into test tubes, extracted once with 2 ml ether (distilled), and centrifuged at 1,000 g for 5 min. The pellet was then reextracted twice with 2 ml ether each time. The combined extracts were dried under blowing N₂. To each tube of dried DAGs were added 100 µl of dry benzene, 10 µl (6.8 µmol) of 0.68 M 4-pyrrolidinopyridine in dry benzene, and 10 µl (5.5 µmol) of 0.55 M benzoic anhydride in dry benzene. The contents were mixed well and the tubes were gassed with N₂ and incubated overnight at room temperature. The next morning, the reaction was stopped by adding 100 µl of methanol to each tube. The tubes were allowed to stand for 5 min after methanol was added. (A similar control reaction mixture was run in a separate tube containing no DAG.)

Purification and reverse-phase HPLC analysis of -DAGs. The benzoylation reactions were applied in toto to a 20 × 20 cm glass plate of Silica gel 60 (0.25-mm layer) as a 2-cm band along with the control reaction. After development of the plate in a mixture of benzene, hexane, ether, and 30% (wt/wt) ammonia (50:45:5:1, vol/vol) up to 1 cm from the top, the -DAG (R_f = 0.34) band and corresponding benzoylated internal standard, -DSG (R_f = 0.34), were visualized with primulin after 10-15 min air-drying and scraped into clean glass tubes. The -DAGs were extracted with 2 ml of ether, centrifuged at 1,000 g for 5 min, and reextracted from the pellets twice more with 2 ml ether each time. The combined ether extracts were dried under blowing N₂ and taken up with 200 µl of distilled cyclohexane-ether (95.5:4.5, vol/vol) after filtering through a 0.45-µm Teflon syringe filter. The sample was stored under N₂ at 20°C for further purification by normal-phase HPLC before analysis on a reverse-phase HPLC column (Ultrasphere ODS, 5 µm, 4.6 mm × 25 cm; Beckman, San Ramon, CA). For cleanup on a normal-phase column (Microsorb Si, 5 µm, 4.6 mm × 25 cm; Rainin Instrument, Woburn, MA), the sample, in normal-phase solvent, was concentrated under blowing N₂ to ~50-100 µl before injecting ~50 µl onto the column and then equilibrated with cyclohexane-ether (95.5:4.5, vol/vol) at 1.0 ml/min. -DAG (retention time = ~10 min) was monitored at 228 nm with a Spectroflow 773 absorbance detector (Kratos Analytical Instruments, Westwood, NJ). Then the effluent was collected and stored under N₂ at 20°C or dried under blowing N₂ and redissolved with ~200 µl of reverse-phase solvent and either stored at 20°C under N₂ or concentrated to ~50-100 µl with blowing N₂ before injection of ~50 µl for reverse-phase analysis on a column equilibrated at 1.0 ml/min with acetonitrile-isopropanol (70:30, vol/vol). -DAG peaks detected on-line at 228 nm were quantitated by electronic integration of peak areas with a Macintosh SE personal computer interfaced with an on-line data acquisition system (MacIntegrator II, Rainin Instrument). To calculate actual amounts present in the original sample, we multiplied areas relative to the -DSG internal standard by the fixed amount of internal standard added at the beginning of sample workup, assuming equal and linear detector responses for all -DAG.

Formation and purification of fatty acid methyl esters from phospholipids and DAGs. -DAGs isolated from reverse-phase HPLC or individual phospholipid bands (~50-100 nmol) purified by Silica gel TLC [chloroform-methanol-0.25% (wt/vol) KCl-ethyl acetate-isopropanol (150:45:30:90:125, vol/vol)] from neutral lipid extracts (i.e., PC, PS, PI, CL, PE, and SM) were placed in 16 × 100 mm Teflon-lined screw cap glass test tubes. A known quantity of triheptadecanoyl glycerol (5-10 nmol) was added to each tube, and the contents were dried under blowing N₂. The glycerolipids were then methanolyzed in 0.5 ml of 14% (wt/vol) BF₃-methanol (Sigma Chemical) in sealed tubes for 1 h at 60°C after flushing the tubes with N₂. SM was methanolyzed under the same conditions, except that 0.5 ml chloroform (for solubilization) and 0.5 ml of methanol-concentrated HCl (5:1, vol/vol) were used for 20 h.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Phospholipid Composition of Isolated Rabbit Rectosigmoid Smooth Muscle Cells

Molecular Species Composition of DAGs from Whole Lipid Extracts of Isolated Smooth Muscle Cells

View larger version (20K): [in this window] [in a new window]   Fig. 1.   HPLC profile of sn-1,2-diacylglycerol (DAG) from isolated cells of rabbit smooth muscle. Major species of DAG found by reverse-phase HPLC analysis of lipid extracts prepared from isolated smooth muscle cells of rabbit colon are shown. Whole lipid extracts were prepared as described in Methods. DSG, 1,2-distearoyl-rac-glycerol.

Phospholipid Fingerprinting: Molecular Species Composition of DAGs of Phospholipids Purified from Whole Lipid Extracts of Isolated Rabbit Smooth Muscle Cells

View larger version (26K): [in this window] [in a new window]   Fig. 2.   HPLC profile of DAGs from isolated cells of rabbit smooth muscle. Molecular species of DAG found by reverse-phase HPLC analysis of benzoylated DAGs (-DAGs) obtained after phospholipase C (PLC) treatment of phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylinositol (PI), and phosphatidylserine (PS) purified from whole lipid extracts of isolated cells of rabbit colon smooth muscle are shown. Sample was analyzed on a 4.6-mm × 25-cm Beckman 5-µm ODS column eluted isocratically at 1.0 ml/min with acetonitrile-isopropanol (70:30, vol/vol). Peaks were detected and quantitated by on-line monitoring of absorbance at 228 nm with a Kratos variablewavelength spectrophotometer coupled to a Rainin MacIntegrator II interface system for digitization and on-line integration of peak areas on a Macintosh SE personal computer. Details of analysis are described in Methods.

Effect of Bombesin Stimulation on DAG Levels in Isolated Rabbit Colon Smooth Muscle Cells

View larger version (27K): [in this window] [in a new window]   Fig. 3.   DAG production by smooth muscle cells from the rabbit rectosigmoid in response to bombesin (BB) (106 M) (not shown) and 12-O-tetradecanoylphorbol 13-acetate (TPA) (1 µM). Cells were isolated, and 1-ml aliquots of relaxed cell suspensions were treated for indicated periods with BB (1 µM). Reaction was terminated by addition of 3.4 ml of chloroform-methanol (1:2, vol/vol), followed by addition of 2-3 nmol of DSG internal standard to each sample. Samples were quantitated for DAG by reverse-phase HPLC as described in detail in Methods. Data are expressed as means ± SE of 3 experiments. Newly formed DAG increased significantly (0.01 > P > 0.001) by 29.2 ± 4.1% (n = 3) at 30 s and by 81.5 ± 9.6% (n = 3) at 4 min above control levels. TPA (1 µM) induces an increase in newly formed DAG both at the early 30-s and late 4-min stimulations, which corresponds with the sustained contraction induced by TPA. DAG increase was inhibited by preincubating cells with the protein kinase C inhibitor calphostin C (1 µM).

Effect of TPA Stimulation on DAG Levels and Inhibition of TPA-Induced DAG Production by PKC Inhibitor Calphostin C in Isolated Rabbit Colon Smooth Muscle Cells

Origin of DAG Production Stimulated by Bombesin

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Comparison of new DAGs induced by BB with control DAGs in isolated cells of rabbit smooth muscle. Graph shows comparison of the DAG profile for newly formed BB-induced DAG (data from Table 4) relative to the corresponding profile in unstimulated controls (data from Table 2). Data (in mol/100 mol) are from reverse-phase HPLC analysis of DAGs present in control samples from means ± SE of 5 separate experiments and of new DAG production from means ± SE of 3 experiments. There is a significant change in contribution of certain DAG molecular species after BB stimulation, shown here by an increase in the molar fraction for peaks 2, 4, and 6, but a decrease in the molar fraction of peaks 9 and 11. These changes indicate that new DAG must be coming from a new source/process different from the one contributing to basal DAG levels and are a result of BB-induced glycerophospholipid hydrolysis.

If one were to consider the possibility that the new DAG produced in response to bombesin might be due solely to phosphoinositide hydrolysis, then the pattern of elicited DAGs should reflect those that were obtained for PI (Fig. 2). Figure 5 compares the profile of new DAG in response to bombesin stimulation with that of PI (mol/100 mol). These new DAG and PI profiles do not match, and therefore we must consider alternatives to account for the experimentally observed composition of new DAG molecules in response to bombesin stimulation. Comparisons made with other phospholipid DAG profiles (Fig. 5) also indicate that no single phospholipid can uniquely determine the experimentally observed new DAG pattern formed in response to bombesin stimulation. The remaining consideration is that matching DAG patterns can arise if many phospholipids contribute to the production of new agonist-induced DAGs. This requires a calculation of the expected distribution of DAGs that would match the observed profile by assuming various values for the fractional contributions made by each contributing phospholipid. In this manner, we are able to give an approximation of the relative contributions of PI, PC, PS, and PE consistent with the profile of newly formed DAGs, using the data from Table 3 for the various major DAG peaks obtained from each phospholipid HPLC profile.

View larger version (22K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   View larger version (22K): [in this window] [in a new window]   View larger version (25K): [in this window] [in a new window]   Fig. 5.   Comparison of new DAGs induced by BB with phospholipid-derived DAGs in isolated smooth muscle cells from the rabbit rectosigmoid. New DAG and phospholipid DAG values (mol/100 mol) obtained by reverse-phase HPLC analysis are compared. The new DAG and PI profiles (A) do not match, and therefore PI alone cannot account for the experimentally observed composition of new DAG molecules. Similar comparisons made with other phospholipid DAG profiles (B, C, and D) also reveal that no single phospholipid can uniquely match the experimentally observed new DAG values, indicating that matching DAG patterns most likely arise from a multiplicity of phospholipids contributing to production of new agonist-induced DAGs. Comparisons were plotted using values from Table 3 for various major DAG peaks obtained from each phospholipid HPLC profile and new DAG values from Table 4. DAGs were analyzed as outlined in legends to Fig. 1 and Table 3.

Approximation of Relative Contributions of PI, PC, PS, and PE Most Consistent with Profile of Newly Formed DAGs

(1)

After the PI fraction values from several (n = 3 and and n = 5) experiments were calculated for each peak and then averaged, the major peaks having the lowest SE for those values were found to also yield values that were consistent, or close to each other. It was also noted that these peaks corresponded to peaks where there was a large difference in the DAG values (mol/100 mol) for PI and PC. Peaks 2, 4, and 6 gave the most consistent values, because the differences in the PI and PC values (mol/100 mol) were greatest for these peaks, and were deemed to be the most reliable ones for determining which PI fraction values fit the data most closely.

For each value of the fixed variable, these PI fraction values were then averaged and a standard deviation was calculated to determine their closeness. The calculation for a given fixed variable value showing the greatest consistency in PI fraction values, i.e., having the lowest standard deviation, was selected as the one yielding the best fractional values. Thus the calculated values of DAG (mol/100 mol) would most closely match the experimental data for DAG (mol/100 mol) in response to bombesin stimulation.

The logic in these calculations is that if our calculated mean PI fraction value is a "best-fit" value for the PI fraction value and is actually a good approximation to some true hypothetical mean PI fraction value for the observed data, then there should be a minimum spread, measured by the standard deviation, in the computed PI fraction values among peaks. In other words, two calculated PI fraction values that are close together (low standard deviation) will probably give a better approximation to the mean value than two that are further apart. Because each peak should ideally give the same hypothetical mean PI fraction value, then deviations in calculated PI fraction value among peaks at a given value of fixed variable reflect the variation in the observed new DAG value (mol/100 mol) from some true hypothetical mean new DAG value (mol/100 mol) for those peaks.

The validity of using this approach was tested by using hypothetical data, which showed that the calculated PI fraction values varied more from their true value as the new DAG value (mol/100 mol) varied more from its true value, for a given value of fixed variable. Also, from this hypothetical data, the spread of calculated PI fraction values was found to increase as the fixed variable (either the PS fraction value or the PC fraction value) was allowed to vary from its true value, and the approximation to the true mean PI fraction value was poorer as well.

Having found our approximate solution to the fractional values for PI, PC, PS, and PE, their goodness of fit to the experimentally observed data was then verified by substituting those values back into the equation

(2)

where fPI, fPC, and fPS are the respective fractional values of PI, PC, and PS contributing to new DAG formation and PI, PC, and PS are their corresponding DAG values in moles per 100 mol. Calculating the new DAG moles per 100 mol values for each peak and determining how well these calculated new DAG values (mol/100 mol) matched the experimental ones were done by calculating the deviation (as an absolute or standard deviation relative to the calculated or observed new DAG mol/100 mol value) between observed and calculated new DAG values (mol/100 mol) for the major observed new DAG peaks (peaks 4, 6, and 8). The fractional values giving the lowest spread (calculated as another standard deviation) in these relative errors were accepted as a best fit, as the relative error was found to be more evenly distributed among the major peaks.

Figure 6 shows the comparison between observed and calculated new DAG values (in mol/100 mol, at 30 s poststimulation with bombesin) using the original calculated fractional values that were supposed to give a good fit to the experimental data. There was good agreement for peak 6, but for the other two major peaks, peaks 4 and 8, there was a rather large relative difference between calculated and observed values. However, by increasing the relative contributions from PC we found a better fit for peaks 4 and 8, i.e., the relative error, or deviation, between calculated and observed DAG values (mol/100 mol) was now more uniform.

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Comparison of 30-s new DAGs induced by BB with calculated new DAGs in isolated cells of rabbit smooth muscle. Observed new DAG values at 30 s are compared with calculated DAG values. As indicated by 2 statistical analysis, the best fit, i.e., the least significant difference measured by the lowest calculated 2 value, between observed and calculated new DAG values is obtained for PI = 0.30 and PC = 0.70. Comparisons are based on data from Tables 3 and 4 and on Eq. 1 for calculating expected new DAG. Open bars are means ± SE for observed new DAG from 3 experiments.

At 30 s after stimulation with bombesin, it appears that a good approximation for the fractional value from PI is somewhere in the range of 0.1-0.4, with the remainder of the new DAG produced apparently coming from either PC (0.66-0.70) and PS (0-0.10) or a combination of PC (0.45-0.50) and PE (0.1-0.45). A similar analysis suggested that PI = 0.3 ± 0.1 and PC = 0.7 ± 0.1 would be a best estimate for the data at 4 min (Fig. 7). As shown in Fig. 8, preincubation of the cells with the PC-specific PLC inhibitor D609 resulted in inhibition of DAG production in response to bombesin, confirming that the main source of DAG is from hydrolysis of PC. Comparable results were also obtained for the new DAG elicited by stimulation with TPA, which revealed that, at both 30 s and 4 min (Fig. 9), there are contributions of ~20 ± 10% from PI and 80 ± 10% from PC. The data clearly indicate that PI cannot account solely for the newly formed DAG, during both short-term phasic (30-s) and long-term sustained (4-min) contraction, and that PC appears to be the major contributor for DAG production in both of these stages.

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Comparison of 4-min new DAGs induced by BB with calculated new DAGs in isolated cells of rabbit smooth muscle. Shown is the initial comparison between observed new DAG at 4 min and calculated new DAG using original calculated fractional values of PI, PS, PC, or PE that were supposed to give a good fit to the experimental data. There is good agreement for the 3 major peaks (4, 6, and 8), with a rather uniform relative difference between calculated and observed values. A 2 statistical analysis indicated that the best fit, i.e., the least significant difference as measured by the lowest calculated 2 value, between observed and calculated new DAG is obtained for PI = 0.30 and PC = 0.70. Comparisons were derived from data in Tables 3 and 4, using Eq. 1 to calculate expected new DAG values. Open bars are means ± SE for observed new DAG from 3 experiments.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Effect of preincubation of smooth muscle cells with phospholipase inhibitor D609. Preincubation of smooth muscle cells with the PC-specific phospholipase inhibitor D609 for 20 min resulted in inhibition of BB-stimulated DAG production.

View larger version (28K): [in this window] [in a new window]   Fig. 9.   Comparison of 30-s (A) and 4-min (B) new DAGs induced by TPA with calculated new DAGs in isolated cells of rabbit smooth muscle. Shown is a comparison based on results of 3 experiments for which new DAG values were determined from new DAG elicited by stimulation with 106 M TPA, which revealed that at 30 s and 4 min there are contributions of ~20 ± 10% from PI and 80 ± 10% from PC. A 2 statistical analysis indicated that the best fit, i.e., the least significant difference as measured by the lowest calculated 2 value, between observed and calculated new DAG values is obtained for PI = 0.20 and PC = 0.80. Open bars are means ± SE (error bars) for observed new DAG values from 3 experiments.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

In many types of smooth muscle cells, agonist-induced contraction has been associated with a rise in intracellular Ca²⁺ concentration (6, 21, 24). The rise in intracellular Ca²⁺ concentration activates the calmodulin-dependent enzyme, myosin light-chain kinase (1, 13, 29), which results in phosphorylation of the 20-kDa myosin light chain. This leads to interaction between myosin and actin, which, in turn, initiates the contractile response.

Agonist-induced production of IP₃ and DAG is not always equivalent. In Swiss 3T3 mouse fibroblasts (7), bombesin stimulates the release of both IP₃ and DAG. The increase in IP₃ is rapid and transient, returning to basal levels by 30 s. In contrast, the increase in DAG is biphasic: the first phase (0-30 s) mirrors the transient IP₃-mediated response, whereas the second phase is sustained (up to 60 min) and occurs in the absence of elevated IP₃, at a time when IP₃ generation is not occurring. The possibility that PC serves as an alternative source for DAG production is suggested by a strong temporal correlation between bombesin- and PMA-stimulated [³H]choline release (8) and by a correlation between PMA-stimulated DAG production (30) and the second phase of bombesin-stimulated DAG production.

It is, therefore, possible that PIP₂ hydrolysis may not be the sole source of bombesin-stimulated DAG generation, but that a second source may be present, possibly PC. In addition to bombesin, other peptides such as endothelin, a contractile peptide of vascular smooth muscle, activate with unique selectivity and potency the DAG/PKC pathway (20). Thus the differential activation of IP₃ and PKC may depend on differential activation of distinct phospholipases, resulting in different mixes of second messengers such as phosphoinositides and/or DAG, depending on the agonist employed.

Phosphoinositides represent ~6% of the phospholipids present in the plasmalemma of total cellular phospholipid of rectosigmoid smooth muscle cells. Ligand binding causes PLC-induced PI hydrolysis and results in equimolar production of IP₃ and DAG. As suggested by others (8), there is much evidence that PC hydrolysis is a major source for sustained DAG production in cellular control mechanisms requiring prolonged activity of PKC and that this role is well suited to PC. PC is the most abundant cellular phospholipid, accounting for ~50%, and thereby can ensure an ample source of DAG without dangerously altering the structural/functional integrity of the membrane (3, 11). We have investigated the relevance of this hypothesis in smooth muscle cells and the source of DAG and its relationship to sustained contraction. It was crucial to have a reliable quantitative measurement of DAG. This is often done radiometrically by an enzymatic phosphorylation method that, using commercially obtained DAG kinase contained in a crude membrane preparation from Escherichia coli, phosphorylates DAG with [-³²P]ATP and then quantitates the resulting [³²P]phosphatidic acid after TLC. Lee and Hajra (19) noted difficulty in using this commercial preparation, as the presence of other hydrolytic enzymes (e.g., phosphatases or phospholipases) in this crude preparation could result in artifactual formation of higher amounts of DAG from phospholipids present in the cellular lipid extracts being analyzed, or possibly even those present in the enzyme/phospholipid reagent mixture. It is possible that even the lipid extracts being analyzed might have contained some of these contaminating enzymes. The radiometric technique does not provide a qualitative distinction of the source of the new DAG produced. Available evidence indicates that the rapid increase in DAG at 30 s and 4 min after stimulation with bombesin is all due to hydrolysis. The biosynthetic rate of DAG is too slow to contribute a significant amount to the DAG pool during this time period. For these reasons, the established techniques (18, 19) using -DAG derivatives and their analysis by HPLC were adopted as the preferred method of analysis.

Examination of the different phospholipid classes present in smooth muscle cells from rabbit colon confirmed that PC is the most prominent phospholipid, accounting for ~40% of the total cellular phospholipid, whereas PI contributes only 6% (Table 1). Other prominent phospholipids in our system are PE and SM, each constituting ~20% of the total cellular phospholipid pool. The finding of a large amount of SM is rather striking and may be related to our recent discovery of its role as a precursor source for ceramide, an important second messenger and PKC activator in PKC-mediated MAP kinase activation in smooth muscle contraction (28). Reverse-phase analysis of DAGs liberated after individual treatment of the major glycerophospholipids (PC, PE, PS, and PI) with PLC revealed characteristic molecular species differences (Table 3, Fig. 2), which allowed us to determine the source of DAG induced in bombesin-stimulated cells.

Stimulation of contraction in isolated rabbit colon smooth muscle cells with 1 µM bombesin led to a significant (0.01 > P > 0.001) 29.2 ± 4.1% (n = 3) increase in DAG production above the control levels within 30 s of agonist addition (Fig. 3), this period being associated with the transient phase of contraction. In the later stages of the response beyond 30 s, representing sustained tonic contraction, the DAG remained significantly elevated (0.01 > P > 0.001) above the control after 4 min, at a level 81.5 ± 9.6% (n = 3) above control. We have compared the profile of endogenous DAGs present in controls before stimulation with that of DAGs elicited by bombesin. A comparison (Fig. 4) of the newly formed DAG profile in response to stimulation with bombesin (Table 4) with the corresponding profile in unstimulated controls (Table 2) reveals that there was a significant change in the relative moles per 100 mol contribution of certain DAG molecular species after stimulation. The pattern of elicited DAGs in response to bombesin stimulation is different and does not mirror those that were obtained from the unstimulated control samples. The profile indicates the presence of DAG, in specific peaks, where it is not present in the control (unstimulated) cells. These changes are important and indicate that the new profile represents an increase in total DAG.

By comparing the molecular species profiles for newly formed DAGs, arising from bombesin-stimulated contraction (Table 4), with those profiles obtained for the major cellular glycerophospholipids (PC, PE, PS, and PI), we were able to calculate that ~70% of the DAGs formed at both 30 s (Fig. 6) and 4 min (Fig. 4) are due to PC breakdown and that a PLC/PLD-mediated hydrolysis of PC plays a major role in both short- and long-term bombesin-stimulated smooth muscle contraction. The data reaffirm previous reports (3, 11) that in many cellular systems phosphoinositide hydrolysis is not the only source for DAG production and that PC breakdown plays a more prominent role in generating DAG.

It has been reported that stimulation of many cell types by tumor-promoting phorbol esters promotes PC hydrolysis, which may or may not be mediated by PKC, as shown by the use of inactive phorbol ester analogs, PKC inhibitors, or the phenomenon of phorbol ester-induced downregulation of PKC under prolonged exposure (3, 11). This hydrolysis exclusively produces DAG from PC via both PLC- and PLD-mediated pathways and occurs without provoking phosphoinositide hydrolysis. Therefore, stimulation with a phorbol ester such as TPA would be expected to give rise to DAG molecular species that have arisen predominantly, if not exclusively, from PC hydrolysis and provides us with a reference response with which to compare and assess the reliability of our conclusion. Our analysis of the new DAG species induced in TPA-stimulated smooth muscle cells confirmed this expectation. The distribution of new TPA-induced DAG species at both 30 s and 4 min (Fig. 9) was very similar to that obtained with bombesin, except that in the TPA-induced DAG there was a smaller contribution from peak 6, which represents an 18:0-20:4 DAG molecular species (Table 2) containing stearic and arachidonic acids and is the single most prominent characteristic of phosphoinositides.

Using calculations similar to those explained above for bombesin-stimulated new DAG, our data were consistent with ~80% of the TPA-generated DAG being derived from PC hydrolysis, with the remainder coming from phosphoinositides. PC hydrolysis was almost completely blocked by a specific PKC inhibitor, calphostin C. This is also consistent with the knowledge that phorbol esters induce contraction in smooth muscle by a process dependent on PKC, because it can be blocked by PKC inhibitors (17). Despite the shortcomings of PKC inhibitors (31), the ability to mimic the action of a physiological ligand (bombesin) with an exogenous activator of PKC (TPA) and to block its action with calphostin C suggests that the natural contractile response to bombesin may be a PKC-mediated process and validates our conclusion that DAG production indeed arises primarily from increased PC breakdown.

Our findings are consistent with proposed models for receptor-mediated PC hydrolysis (3, 11). Bombesin appears to induce a sustained contraction by stimulating PC hydrolysis through a PKC-mediated activation of either PLC- or PLD-mediated pathways or both. In these models, it is reported that G proteins are involved in mediating PC hydrolysis by activating PLC and/or PLD in some cell systems, as shown by the use of nonhydrolyzable GTP analogs such as GTPS, which in addition to potentiating an agonist-induced response, by itself can activate both phospholipase activities. In some systems, activation of both phospholipases appears to be operative and is not dependent on G proteins. For example, in rat vascular smooth muscle cells the platelet-derived growth factor (PDGF)-activated formation of phosphatidic acid, an important mitogen in these cells, occurs both indirectly by the combined but sequential action of PC-PLC and DAG kinase and also directly by activation of PLD (15). Furthermore, PDGF-stimulated phosphatidic acid synthesis could occur independently of receptor tyrosine kinase or PKC activity through activation of a DAG kinase pathway, suggesting an important independent role for DAG kinase in signal transduction (16). However, in endothelin-induced vascular smooth muscle contraction, which occurs through a G protein-coupled pathway, activation of PLD appears to be more important for the continued activation of PKC and maintenance of sustained entry of Ca²⁺ through PKC-dependent activation of L-type Ca²⁺ channels, implying that phosphorylation of Ca²⁺ channels may have an important modulatory function, although evidence for regulation of these channels by PKC phosphorylation was not reported (32).

These results add to the growing body of evidence that DAG production can occur in more than one phase and simultaneously from more than one source. In the case of smooth muscle, it could be derived from a dual source, PI and PC. The activation of distinct PLC- isozymes by bombesin has been shown in pancreatic acinar cells (26), and in smooth muscle cells receptor activation of distinct G proteins (23) and isozymes of PLC (22) has been demonstrated. Our results also support our earlier findings of the important role of PKC in mediating calmodulin-independent sustained contraction of GI smooth muscle.