Role of Protein Kinase C, Ca²⁺/Calmodulin-dependent Protein Kinase II, and Mitogen-activated Protein Kinases in Volatile Anesthetic–induced Relaxation in Newborn Rabbit Pulmonary Artery

Thymol-free halothane was supplied by Peter Haines, B.S. (director of medical marketing, Halocarbon Laboratories, Hackensack, NJ) and isoflurane from OHMEDA Inc. (Liberty Corner, NJ). The peptide of the inhibitor protein (CKIINtide) 11for CaMKII was supplied by Debra Brickey, Ph.D. (research scientist), and Thomas Soderling, Ph.D. (professor), of Vollum Research Institute (Portland, OR). Inhibitors of PKC (Gö69769 12and bisindolylmaleimide hydrochloride 13), CaMKII (KN-93 14), p38 (SB203580 hydrochloride 15), and MEK/ERK1/2 (PD98059 16) were purchased from Calbiochem (La Jolla, CA), RBI (Natick, MA), or Cell Signaling Technology (New England Biolabs, England, UK). Chemicals were analytical or reagent grade. Stock solutions of Gö6976 and PD98059 were made in 100% dimethyl sulfoxide, and the final concentrations of the inhibitors contained 0.1% dimethyl sulfoxide. The same concentrations of dimethyl sulfoxide were used in parallel for time controls.

The method of preparing skinned arterial strips has been described previously. 4–7,17Briefly, newborn New Zealand White rabbits (9–12 days old) were killed with a decapitator. This method of euthanasia has been approved by University of Washington Animal Care Committee. The procedure for making the sarcolemma of pulmonary arterial strips to be permeable (300 μg/ml saponin in relaxing solution for 4 min) and the composition of the bathing solution (pCa buffer) were as described previously, 2except that 10 mm EGTA was used. Isometric tension of the skinned strips was recorded on a computer (Tower Macintosh G3; Apple Computer, Inc., Cupertino, CA) with a customized LabVIEW software program interfaced with a multifunction I/O board with 16-bit resolution (NB-MIO-16XL; National Instrument, Austin, TX). The experiments were performed at room temperature (23 ± 2°C).

The experimental procedure was the same as described previously. 4–7The skinned strips were activated by Ca²⁺buffered with 10 mm EGTA (pCa 6.3 or 6.0 buffer). When the force reached steady state (control force), a fresh solution with the same ionic composition containing one of the concentrations of halothane (1, 2, or 3%) (fig. 1) or isoflurane (1, 3, or 5%) was then administered, and the results were observed for up to 30 min (test force). At the end of each experiment, the maximum force of the strips was tested with pCa 4.0 buffer, which was used to estimate the control force as a percentage of the maximum force. The inhibitors were preincubated in a relaxing solution and were present in the subsequent contracting solutions. Time-control experiments (without anesthetic and inhibitor, or at one of the concentrations of anesthetic without inhibitor) were performed in parallel. The test results were analyzed at 15 or 30 min and were expressed as a percentage of the control.

The height from baseline to steady-state force (control) and force at various time intervals (test) after administration of the test solutions were analyzed. The test conditions included (1) vehicle only (no inhibitor or anesthetic) as time control for anesthetic, (2) one of the anesthetic concentrations plus vehicle (no inhibitor) as time control for inhibitors, and (3) one of the anesthetic concentrations plus one of the inhibitor concentrations. The test results were expressed as a percentage of the control. Means and standard deviations of the mean were obtained from at least three arterial strips and three animals.

The StatVIEW software program (BrainPower, Inc., Calabasas, CA) was used for statistical analysis. Results from the strips treated with the inhibitors and the time controls for the inhibitors were compared by Student t  test. Two-factorial ANOVA was used to compare among anesthetic concentrations or various concentrations of the inhibitors. A value of P < 0.05 was considered statistically significant. 18 

We found that pCa 6.3 (or 0.5 μm free Ca²⁺) generated approximately [mean ± SD (n)] 13.7 ± 3.7 mg (3) or 43.0 ± 9.3% (3) of the maximum force generated by pCa 4.0 [31.3 ± 1.2 mg (3)], and pCa 6.0 (or 1 μm free Ca²⁺) generated 20.1 ± 2.8 mg (3) or 63.2 ± 5.5% (3) of the maximum force from three different arterial strips and three newborn rabbits. We found that isoflurane caused greater relaxation in pCa 6.3– than pCa 6.0–activated force, but the relaxation induced by halothane was not significantly different between these pCa values. Therefore, pCa 6.0 was chosen to study the effects of halothane in skinned strips, which generated a force lower than 10 mg at pCa 6.3.

As shown in figures 1 and 2, isoflurane and halothane decreased the pCa-induced force. The force decreased by 3 and 5% isoflurane was statistically significantly different from that of 0% isoflurane (fig. 2, A ) [% of time control as 100% (0%, fig. 2, A ); 96.3, 76.7, and 68% for 1, 3, and 5% isoflurane, respectively]. However, the relaxation induced by isoflurane was not statistically significantly different between 3 and 5% isoflurane. The pCa-induced force was dose-dependently decreased by halothane (fig. 2, B ) [82.5, 45.2, and 24% of the time control as 100% (0%, fig. 2, B ) for 1, 2, and 3% halothane, respectively]. In contrast to the observations in the adult rabbits, 6,7however, the anesthetics did not induce a transient increase in force.

The relaxation induced by 3% isoflurane was prevented by bisindolylmaleimide (3 and 30 nm) in a dose-dependent manner and was completely prevented by 30 nm bisindolylmaleimide (3%, fig. 3, A ), which is similar to the concentration to inhibit 50% of Ca²⁺-dependent PKC (cPKC) activity in vitro . 13In contrast, the relaxation induced by 5% isoflurane was not prevented but rather enhanced by 30 nm bisindolylmaleimide (5%, fig. 3, A ). Conversely, the relaxation induced by halothane was partially prevented by higher concentrations of bisindolylmaleimide and in a dose-dependent manner (0.3 and 1 μm, fig. 3, B ) for both 2% and 3% halothane.

The role of cPKC in the anesthetic-induced relaxation was examined further by use of a specific cPKC inhibitor, Gö6976. 12We found that the relaxation induced by isoflurane was also prevented by Gö6976 up to 0.01 μm in a dose-dependent manner and was completely prevented by 0.01 μm Gö6976 for both 3 and 5% isoflurane (fig. 4, A ). The relaxation induced by halothane was partially prevented by a higher dose range of Gö6976 (0.001–0.1 μm), and the maximum effective concentration of Gö6976 was 0.1 μm for both 2 and 3% halothane (fig. 4, B ).

Because halothane-induced relaxation was partially prevented by the PKC inhibitors, we hypothesized that the halothane-induced relaxation was mediated in part by a CaMKII signaling pathway, as shown in the adult rabbits. 7We used the CaMKII inhibitor KN-93 by competing with the calmodulin binding site 14or the peptide of the inhibitor protein (CKIINtide) by noncompetitive binding to the catalytic site. 11 

We found that the isoflurane-induced relaxation was again completely prevented by 10 nm CKIINtide for 3 and 5% isoflurane (fig. 5, A ) and that of halothane was partially prevented by CKIINtide (10 and 30 nm for 2% halothane and 30 nm for 3% halothane;fig. 5, B ). In contrast, KN-93 dose-dependently and completely prevented the relaxation induced by 3% (0.1 μm KN-93) and 5% (1 μm KN-93) isoflurane (fig. 6, A ). However, 2 and 3% halothane-induced relaxation was partially prevented by 0.01 μm KN-93 for 2% halothane and 0.01, 0.1, and 0.3 μm KN-93, reaching a plateau at 0.01 μm (fig. 6, B ).

In cultured vascular smooth muscle cells from rabbit femoral arteries, we have shown that isoflurane may activate PKC and CaMKII via  one of the MAP kinases, extracellular signal–regulated kinases (ERK1/2). 19Another MAP kinase, p38 MAP kinase, has also been implicated in angiotensin II–induced vascular smooth muscle contraction. 20Accordingly, we examined further the role of ERK1/2 and p38 in the anesthetic-induced relaxation using a selective inhibitor of MEK/ERK1/2 (PD98059, 15noncompetitive with respect to ATP binding to MEK) or p38 (SB203580 16).

As shown in figure 7, PD98059 at 1 μm completely prevented 3% isoflurane–induced relaxation (3%, fig. 7, A ) but not that of 5% isoflurane (5%, fig. 7, A ). The relaxation induced by 2 and 3% halothane was partially prevented by PD98059, which was statistically significant at 0.1 and 1 μm PD98059 for 2% halothane and at 1 μm PD98059 for 3% halothane (fig. 7, B ).

An inhibitor of p38 mitogen-activated kinase, SB203580, fully prevented the relaxation induced by isoflurane (0.03 and 0.3 μm SB203580 for 3 and 5% isoflurane, respectively;fig. 8, A ) and partially prevented that of halothane (3 and 10 nm SB203580 for 2% halothane peak at 3 nm SB203580; 10 and 30 nm SB203580 for 3% halothane peak at 30 nm SB203580;fig. 8, B ).

The exciting findings and implications from this study in the skinned pulmonary arterial strip from the newborn rabbit under low-Ca²⁺–clamped condition are that (1) isoflurane and halothane induce dose-dependent late relaxation; (2) the inhibitors of PKC, CaMKII or MAP kinases ERK1/2 and p38 completely prevents the relaxation caused by isoflurane but partially prevents that by halothane; (3) a similarity of the pattern of effective inhibitor concentrations of bisindolylmaleimide or Gö6976 and PD98059 with respect to anesthetic concentrations suggests that ERK1/2 is a downstream effector of PKC, 19,20possibly cPKC; (4) the effective inhibitor concentrations of KN-93 (for isoflurane), CKIINtide (for halothane), or SB20358 as a direct function of the anesthetic concentrations suggest that p38 is a downstream effector of CaMKII 21,22; (5) the isoflurane-induced relaxation completely prevented by the inhibitors suggests that there is cross talk between CaMKII and PKC, 23–25between PKC and p38, 23between CaMKII and ERK1/2, 26–28and between ERK1/2 and p38 29; and (6) a direct correlation between concentrations of CKIINtide and halothane suggests that CaMKII also acts via  myosin light chain kinase. 30Thus, the anesthetic-induced relaxation is via  PKC, CaMKII, and MAP kinases in a Ca²⁺-, 31activity-, isozyme-, and time-dependent 32manner.

The degree of relaxation induced by isoflurane or halothane (except at 1%) shown in this study in newborn rabbits is comparable to that observed in the adult rabbit 6,7but without the transient increase in force. This lack of biphasic effects of the anesthetics in the newborn rabbit could be a result of low Ca²⁺in the sarcoplasmic reticulum to be released by the anesthetics, resulting in lower activation of the PKC and CaMKII. Further investigation with respect to the Ca²⁺release mechanisms would confirm the above speculation.

The effective concentrations of the inhibitors of PKC in preventing the relaxation, however, are not consistent between newborn and adult rabbits with respect to isoflurane and halothane. At 3% isoflurane, the relaxation prevented by the PKC inhibitor bisindolylmaleimide (0.03 μm, fig. 9, A ) is lower than that observed in the adult (0.3 μm 6). In contrast, the halothane-induced relaxation prevented by bisindolylmaleimide is higher in the newborn (1 μm, fig. 9, B ) than in the adult (0.01 μm 7). Using the effective inhibitor concentrations as an estimate of the enzyme activity, PKC would be activated by isoflurane to a lesser degree and by halothane to a greater degree in the newborn than the adult, suggesting that PKC plays a more important role in the halothane-induced relaxation in the newborn than the adult.

The pattern of the concentrations of inhibitors of PKC and MEK/ERK1/2 with respect to isoflurane (fig. 9, A ) or halothane (fig. 9, B ) suggests that MEK/ERK1/2 is a downstream effector of PKC resulting in anesthetic-induced relaxation. However, the lack of changes in effective inhibitor concentrations of PKC with respect to anesthetic concentrations [0.01 μm Gö6976 for 3 and 5% isoflurane (fig. 9, A ); 0.1 μm Gö6976 or 1 μm bisindolylmaleimide for 2 and 3% halothane (fig. 9, B )] suggests activation of cPKC resulting in relaxation that is independent of PKC activity measured under our experimental conditions. The facts that relaxation induced by 5% isoflurane is enhanced by higher concentrations of bisindolylmaleimide [a cPKC/Ca²⁺-independent (c/nPKC) inhibitor] and prevented by 0.01 μm Gö6976 (a cPKC inhibitor) suggest that nPKC plays a role in the contraction induced by isoflurane. Whether high isoflurane concentration (5%) has a direct inhibitory effect on PKC, like that reported in halothane, 33remains to be examined. Thus, PKC could lead to relaxation via  MEK/ERK1/2 (fig. 9, C ) 19,20or contraction possibly via  CPI-17 (fig. 9, D ). 34A direct measurement of activity of PKC and MEK/ERK1/2 would confirm this speculation. The effectiveness of the inhibitors of PKC suggests that phospholipids (such as diacylglycerol) and Ca²⁺(fig. 9C and D) contribute to activation of PKC. 35 

The effective concentrations of KN-93 (competing with the calmodulin site of CaMKII 14) but not that of CKIINtide (binding to the catalytic site of autophosphorylated CaMKII 11) as a direct function of isoflurane suggest that CaMKII is activated by Ca²⁺/calmodulin to a lower activity (0.01 μm CKIINtide in the autophosphorylated state;fig. 10, A ). In contrast, a positive correlation between effective concentrations of CKIINtide (0.1 and 1 μm) and halothane (2 and 3%) but not that of KN-93 suggests that higher Ca²⁺is released by halothane, which activates CaMKII, resulting in higher activity (0.03 μm CKIINtide in the autophosphorylated state;fig. 10, B ).

The absence of increasing effectiveness of CKIINtide for isoflurane and that of KN-93 for halothane can be speculated on as follows. At low Ca²⁺levels (released by isoflurane), CaMKII is initially activated by binding of Ca²⁺/calmodulin, which is blocked by KN-93 by competing for the calmodulin binding site 14in a dose-dependent manner, whereas CaMKII undergoes autophosphorylation 10in a low-activity state blocked by CKIINtide (fig. 10, A ). In contrast, at higher Ca²⁺levels (released by halothane), the initial Ca²⁺/calmodulin binding and activation of CaMKII (blocked by KN-93;fig. 10, B , open circles ) result in a higher autophosphorylated state (or a higher activity) that increases with increases in Ca²⁺levels (blocked by CKIINtide;fig. 10, B, filled circles ).

In contrast to PKC, the increased effective concentrations of the inhibitors of CaMKII or p38 as a direct function of anesthetic concentration (KN-93 for isoflurane, CKIINtide for halothane) observed in this study have also been reported in the adult rabbits with halothane. 7Thus, a role for CaMKII/p38 signaling 21,22in modulation of vascular smooth muscle relaxation (fig. 10, C ) is further confirmed. The complete prevention of isoflurane-induced relaxation by the inhibition of PKC, CaMKII, p38, and ERK1/2 suggests cross talk among these protein kinases, possibly CaMKII/PKC, 23–25PKC/p38, 23CaMKII/ERK1/2, 19,26–28and between ERK1/2 and p38 29but not PKC/CaMKII. 36The halothane-induced relaxation partially prevented by the inhibitors of CaMKII suggests that CaMKII/MLCK/MLC signaling (fig. 10, D ) 30may also play a role.

The effectiveness of CaMKII inhibitors suggests that Ca²⁺and calmodulin play a role in activation of CaMKII. Whether the anesthetics induce Ca²⁺release from the sarcoplasmic reticulum in newborn pulmonary artery, as shown in the adult, 5remains to be studied. Further investigations are needed as to factors such as phospholipids and Ca²⁺levels, 37,38Ca²⁺release mechanisms by isoflurane and halothane, localization, or translocation between the membrane and the cytosol for PKC 32and CaMKII resulting in differential effectiveness of the inhibitors and anesthetic- and time-dependent relaxation.

The similar degree of anesthetic-induced relaxation induced by the volatile anesthetics between newborn and adult rabbits suggests that the clinical impression of greater vasodilatation in the newborn than the adult may be attributed more to factors (such as endothelium, sensitivity of receptor on the plasma membrane, etc.) regulating vascular smooth muscle contraction than by the direct effects on the contractile proteins via  intracellular signaling pathways shown in this study.

In summary, PKC, CaMKII, and the downstream effectors MAP kinases play a role in the anesthetic-induced relaxation in the pulmonary artery of newborn rabbits. CaMKII, in an activity-dependent manner, modulates concentration-dependent relaxation induced by both anesthetics, and PKC, depending on activity, modulates both relaxation and contraction in skinned pulmonary arterial strips of newborn rabbits.

The authors thank Debra Brickey, Ph.D. (Research Scientist), and Thomas Soderling, Ph.D. (Professor), of Vollum Research Institute (Portland, Oregon) for the generous supply of the inhibitor protein; Peter Haines, B.S. (Director of Medical Marketing of Halocarbon Laboratories, Hackensack, New Jersey), for thymol-free halothane; and OHMEDA Inc. (Liberty Corner, New Jersey) for isoflurane.