Lidocaine and bupivacaine impair wound healing, but the mechanism of this side effect has not been determined. The phospholipid messenger lysophosphatidate is released from activated platelets and induces fibroblast and smooth muscle proliferation. Because it may play a role in wound healing, the authors studied the effects of local anesthetics on lysophosphatidate signaling in Xenopus oocytes.
Defolliculated Xenopus oocytes expressing endogenous G protein-coupled lysophosphatidate receptors were voltage clamped and studied in the presence or absence of lidocaine or bupivacaine. Lysophosphatidate-induced Ca(2+)-activated Cl- currents (IC1(Ca)) were measured. To determine the site of action of the local anesthetics on the signaling pathway, the authors studied 1) the effects of local anesthetics on signaling induced by intracellular injection of the second messenger inositoltrisphosphate, and 2) the effects of local anesthetics on functioning of recombinantly expressed angiotensin II receptor signaling through the same pathways as the lysophosphatidate receptor.
Lysophosphatidate signaling was inhibited in the presence of local anesthetics. The half maximal inhibitory concentration (IC50S) for lidocaine and bupivacaine were 29.6 mM and 4.7 mM, respectively. Neither responses induced by inositoltrisphosphate injection nor angiotensin signaling were influenced by local anesthetics.
Lysophosphatidate signaling is inhibited by the extracellular application of lidocaine or bupivacaine. In contrast, inositoltrisphosphate or angiotensin signaling was not affected by local anesthetics. Therefore local anesthetics have a specific, extracellular effect on lysophosphatidate receptor functioning. As the local anesthetic concentrations used were similar to those observed after injection around surgical wounds, LP inhibition may play a role in the observed detrimental effects of local anesthetics on wound healing.
Local anesthetics impair wound healing in various models. For example, the tensile strength of wounds is decreased when lidocaine is injected in wounds for to reduce pain after operation.  Lidocaine, with or without epinephrine, also caused a 13-day delay in surgical wound healing in rats.  Although long-acting local anesthetics are becoming ever more important to provide pain relief after outpatient surgery, their effects on wound healing have not been reported.
The mechanisms responsible for the adverse effects of local anesthetics on wound healing are unclear. Suppression of mucopolysaccharide synthesis has been postulated.  In tissue cultures of fibroblasts, slices of newborn rat skin or in vivo granuloma tissue induced by subcutaneous implantation of steel cylinders in rats, application of lidocaine or bupivacaine inhibited synthesis of collagen to a greater extent than that of noncollagenous proteins. Glycosaminoglycan synthesis was found defective as well.  Nevertheless, the specific site of local anesthetic interference in the wound healing process is not known.
Lysophosphatidate is a phospholipid released by activated platelets and injured fibroblasts. [4,5] The compound activates a specific membrane receptor and, through several coupled G proteins, activates a number of intracellular signaling cascades. [6–9] Its effects include platelet activation, vasoconstriction, and induction of mitogenesis, particularly in fibroblasts. [6,10,11] Thus the compound may be released after injury and play a role in wound healing. Therefore we hypothesized that interference of local anesthetics with the lysophosphatidate signaling pathway might explain this phenomenon. We investigated the effects on lysophosphatidate signaling of lidocaine and bupivacaine, two local anesthetics of different lipid solubility and duration of action.
As a model, we used oocytes of the African clawed toad, Xenopus laevis. These cells express lysophosphatidate-specific G protein-coupled membrane receptors, the activation of which results in a Cl sup - current induced by intracellular Ca2+ release (Figure 1(A)).  To localize the site of local anesthetic action, we expressed the angiotensin receptor, which responds to a structurally different agonist but couples to the same intracellular signaling pathway as does the lysophosphatidate receptor. In addition, we determined the effects of the local anesthetics on signaling induced by microinjection of inositoltrisphosphate (IP3). Specifically, we tried to answer the following questions:
1. Does lidocaine or bupivacaine interfere with lysophosphatidate-induced Ca2+-activated Cl sup - currents; if so, are the calculated half maximal inhibitory concentration (IC50s) comparable to concentrations reached after subcutaneous injection?
2. Are these effects reversible?
3. Where in the signaling pathway are these effects localized?
Materials and Methods
The study protocol was approved by the Animal Research Committee of the University of Virginia. Adult female Xenopus laevis frogs were obtained from Xenopus I (Ann Arbor, MI) and housed in an established frog colony. To obtain oocytes, frogs were anesthetized with 1% tricaine (3-aminobenzoic acid ethyl ester) until they became unresponsive, and they were operated on ice. Oocytes were harvested through a 5-mm incision in the lower lateral abdominal wall and immediately placed in modified Barth's solution (containing 88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO3, 0.41 mM CaCl2, 0.82 mM MgSO4, 0.3 mM Ca(NO3)2, 15 mM HEPES, and 0.1 mM mg/ml gentamicin, with pH adjusted to 7.6). Frogs were returned to the main tank after recovering for 2 days in isolation. Oocytes (Dumont's stage 5 and 6) were defolliculated manually using microforceps and incubated in modified Barth's solution at 18 degrees Celsius for 24 h because this results in more consistent lysophosphatidate responses.
Endogenous lysophosphatidate receptors were used to evaluate the influence of local anesthetics on lysophosphatidate signaling. The angiotensin receptor was expressed recombinantly. A 1.2-kilobase pair cDNA in the CDM8 vector (Invitrogen, San Diego, CA), encoding the rat AT sub 1A angiotensin II receptor, was a gift from Dr. K. R. Lynch, of the University of Virginia. The construct was linearized with the nuclease Xho I and transcribed in vitro by T7 RNA polymerase in the presence of a capping analog. Oocytes were injected with 50 nl mRNA (5 ng) in water, using an automated microinjector (Nanoject; Drummond Scientific, Broomall, PA). Slight ballooning of the oocyte confirmed the adequacy of injection. The cells were cultured for 72 h at 18 degrees Celsius in modified Barth's solution before study.
Lysophosphatidate or angiotensin receptor activation in the oocyte leads to G protein-induced phospholipase C activity and release of IP3from phosphoinositolbisphosphate (Figure 1(A)). Inositoltrisphosphate then binds to its specific receptor on intracellular Ca2+ stores to release Ca2+, which in turn opens an endogenous Ca2+-activated Cl sup - channel. The resulting current can be measured using a voltage clamp, and this Ca2+-activated Cl sup - current (ICl(Ca)), integrated and expressed in microcoulombs (micro Coulomb), is a measure of intracellular Ca2+ release. [8,12–14] All experiments were performed at room temperature (approximately 22 degrees Celsius).
Microelectrodes were pulled in one stage from capillary glass (BBL with fiber; World Precision Instruments, Sarasota, FL) on a micropipette puller (model 700C; David Kopf Instruments, Tujunga, CA). Tips were broken to a diameter of approximately 10 mM, providing a resistance of 1–3 M Ohm, and filled with 3 M KCl. A single oocyte was placed in a perfusable bath containing 3 ml Tyrode's solution (consisting of 150 mM NaCl, 5 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 10 mM dextrose, and 10 mM HEPES, with the p3H adjusted to 6.4 with NaOH) with or without local anesthetic. Two microelectrodes were inserted into the oocyte, and a holding potential of -70 mV was applied to the membrane. The voltage clamp amplifier (OC725A; Warner Corp., New Haven, CT) was connected to a data acquisition and analysis system running on an IBM-compatible personal computer. The acquisition system was a DAS-8 A/D conversion board (Keithly-Metrabyte, Taunton, MA), and analysis was performed with OoClamp software.  Occasional cells that did not show a stable holding current less than 1 mA during a 1-min equilibration period were excluded from analysis. Membrane current was sampled at 125 Hz and recorded for 5 s before and 55 s after application of the agonist (lysophosphatidate or angiotensin, 10 sup -7 M). Agonists were delivered in 30-ml aliquots during 1 or 2 s using a micropipettor positioned approximately 1 mm from the oocyte. Responses were quantified by integrating the current trace by quadrature and are reported as microcoulombs because this reflects intracellular Ca2+ release better than peak current does.  Oocytes were preincubated with local anesthetics at least 2 min before agonists were applied. In pilot experiments we found that inhibition develops within 60 s after exposure to the local anesthetic.
For intracellular injections of IP3, a third microelectrode connected to an automated microinjector (the Nanoject) was inserted into the oocyte. Cells were injected under voltage clamp, and the adequacy of injection was verified by observing the small increase in cell size on injection.
Results are reported as means +/- standard error of the mean. Because variability among batches of oocytes is common, responses were at times normalized to controls for each batch. Differences between treatment groups were analyzed using analysis of variance and Student's unpaired t test. Probability values less than 0.05 were considered significant.
Molecular biology reagents were obtained from Promega (Madison, WI), and other chemicals were obtained from Sigma Chemical Company (St. Louis, MO). Lidocaine HCl and bupivacaine HCl (Sigma Chemical Co.) were diluted in Tyrode solution. Because of bupivacaine's low solubility at pH values greater than 6.4, all solutions were adjusted to pH of 6.4, and pilot experiments showed that lysophosphatidate and angiotensin signaling were not influenced by this lower value. Consecutive dilutions for dose-response calculations were made with pH-adjusted Tyrode solution. Lysophosphatidate (1-oleoyl-2-hydroxy-sn-glycero-3-phosphate [sodium salt]; Avanti Polar Lipids, Alabaster, AL) was prepared as a 10 sup -4 M stock solution in 1% delipidated bovine serum albumin (ICN Biomedicals, Aurora, OH). Consecutive dilutions to 10 sup -7 M were made in Tyrode solution. Before experiments, lysophosphatidate solutions were sonicated for 5 s using a Branson Sonifier 250 (Misonix, Farmingdale, NY) to ensure that the compound was fully in solution. All other chemicals were obtained from Sigma Chemical Company.
Lysophosphatidate Responses in Xenopus Oocytes
Oocytes responded to 30 ml 10 sup -7 M lysophosphatidate with a transient inward current (Figure 1(B)). The vehicle bovine serum albumin had no effect. The current developed after 1–3 s and consisted of a fast component and a gradual return to baseline in 10 to 20 s, which is the typical response to G protein-coupled receptor stimulation in Xenopus oocytes. The average control response was 2.4 +/- 0.2 micro Coulomb.
Local Anesthetic Effects on Lysophosphatidate Signaling
Lysophosphatidate signaling was inhibited when oocytes were exposed to lidocaine or bupivacaine at concentrations used clinically for injection around wounds (Figure 2(A)). The inhibitory effect was concentration dependent; the IC50was 29.9 +/- 8.2 mM for lidocaine (Figure 2(B)) and 4.5 +/- 1.2 mM for bupivacaine (Figure 2(C)). Hill coefficients were 1.5 and 1.6, respectively. At high local anesthetic concentrations, virtually complete inhibition of lysophosphatidate signaling was observed. In the presence of local anesthetics, a modest but consistent slowing of the response was noted, but we did not investigate the mechanism any further.
Reversibility of Local Anesthetic Effects on Lysophosphatidate Signaling
The reversibility of the local anesthetic inhibitory effect on lysophosphatidate signaling was investigated in three separate groups of oocytes because we were concerned that multiple lysophosphatidate applications might desensitize the lysophosphatidate receptor (Figure 3). In the first group, responses to 10 sup -7 M lysophosphatidate were recorded in Tyrode solution. In the second group, responses were determined in the presence of 42.8 mM lidocaine (1%) or 15.4 mM bupivacaine (0.5%); inhibitions to 26 +/- 7% and 11 +/- 3%, respectively, of control values were seen. In the third group, oocytes were first placed in 42.8 mM lidocaine or 15.4 mM bupivacaine for 5 min, which was then completely washed out with Tyrode solution. After 5 min of washing, responses to lysophosphatidate were recorded. Responses after bupivacaine washout were 115 +/- 17% of control, and thus inhibition by bupivacaine was completely reversible. Responses after lidocaine washout were 58 +/- 11% of control. Although this was not significantly different from control values, it suggests a prolonged and possibly irreversible effect of lidocaine.
Site of Action: Effects on Angiotensin sub 1A Signaling
To determine that the effect of lidocaine and bupivacaine on lysophosphatidate signaling was specific and not due to nonspecific membrane interactions or inhibition of the Cl sup - channel, we expressed the angiotensin1Areceptor, which like the lysophosphatidate receptor is a member of the G protein-coupled receptor superfamily. The angiotensin1Areceptor, when expressed in oocytes, also induces ICl(Ca) through a mechanism that, apart from receptor and possibly G protein, is considered to be the same as that for lysophosphatidate signaling. Lack of local anesthetic inhibition on angiotensin signaling would make unlikely a nonspecific membrane effect or an effect on the signaling pathway downstream of the lysophosphatidate receptor (including the Cl sup - channel). Angiotensin II (10 sup -7 M) induced no responses in uninjected oocytes (data not shown) but did induce currents in angiotensin1A-injected oocytes that were indistinguishable from those induced by lysophosphatidate (Figure 4(A); the average control response was 1.6 micro Coulomb). Lidocaine and bupivacaine were without effect on angiotensin signaling. ICl(Ca) in response to 10 sup -7 M angiotensin were 111.1 +/- 22% of control in the presence of lidocaine (42.7 mM) and 92.1 +/- 38.1% of control in the presence of bupivacaine (15.5 mM)(Figure 4). Therefore bupivacaine and lidocaine do not appear to affect the oocyte membrane, intracellular pathways, or the Ca2+-activated Cl sup - channel.
Site of Action: Intracellular Inositoltrisphosphate Injection
To confirm that local anesthetics do not interfere with intracellular Ca2+ release or the Ca2+-activated Cl sup - channel, we determined the effects of lidocaine and bupivacaine on ICl(Ca) induced by microinjection of IP3into the oocyte. Previously we showed that heparin, an IP3receptor antagonist, inhibits lysophosphatidate-induced ICl(Ca) completely, and thus that lysophosphatidate-induced Ca2+ release is mediated solely by IP3. [6,12,14]
Intracellular injection of 50 nl of 100 mM IP3generated ICl(Ca) s in the oocyte indistinguishable from those induced by extracellular lysophosphatidate (Figure 5(A)). However, in contrast to currents induced by lysophosphatidate application, size and kinetics of these responses were not affected by the presence of extracellular local anesthetic. Average responses to lysophosphatidate were 3.5 +/- 0.9 micro Coulomb in the control group, 3.3 +/- 0.8 micro Coulomb in the presence of 42.7 mM lidocaine, and 3.0 +/- 0.6 micro Coulomb in the presence of 15.5 mM bupivacaine (Figure 5(B)). These findings confirm that lysophosphatidate signaling inhibition by local anesthetics occurs early in the signaling pathway.
Our data show that the local anesthetics lidocaine and bupivacaine inhibit lysophosphatidate signaling in a concentration-dependent manner, with appreciable inhibition occurring at concentrations similar to those observed after injection around wounds. Lack of effect on signaling by angiotensin or microinjected IP3suggests that the lysophosphatidate receptor environment (receptor molecule, its lipid environment, or coupled G protein) is the site of action. These findings indicate that local anesthetics can interact significantly with lipid mediator receptors. More specifically, they may explain some observed effects of local anesthetics on the wound healing process. As bupivacaine is, and ropivacaine likely will be, used commonly to provide pain relief after surgery, the interactions of these long-acting local anesthetics with wound healing deserve to be studied in detail.
Several potential limitations of our model should be considered. First, we have only studied a single form of lysophosphatidate signaling (Cl sup - currents induced by intracellular Ca2+ release), whereas several intracellular signaling cascades are activated by lysophosphatidate: decreases in cAMP, activation of Rho, prostaglandin synthesis, and ras activation, to mention just a few. No data suggest that these effects are transduced through different subtypes of receptor. Instead they are transduced through several G proteins, presumably coupling to a single, as yet uncloned, receptor subtype.  Given the importance of intracellular Ca2+ signaling, we chose phospholipase C activation and IP3release as a pathway to investigate. Second, we used Xenopus oocytes to study this pathway, and consequently we performed our experiments at room temperature. This raises the question if the Xenopus lysophosphatidate receptor might behave differently from its mammalian counterpart. However, there is no molecular biological or pharmacologic evidence to suggest significant functional differences between lysophosphatidate receptors in various tissues. We have shown that the oocyte model closely replicates findings in other systems. [11,14,16] The angiotensin1Areceptor derives from rat and therefore was expressed at a lower temperature than it normally functions in. However, this has not been shown to affect its signaling properties appreciably. Temperature differences in oocytes primarily affect latency of response of expressed receptors.  Therefore we believe that the flexibility of the oocyte model outweighs these minor disadvantages.
The role of lysophosphatidate in wound healing is not known. However, because the compound is released primarily from activated platelets  and injured fibroblasts,  its presence in appreciable quantities around wounds seems likely. We have found in preliminary experiments that lysophosphatidate is absent in healthy skin but released in significant quantities after injury. Because its most prominent effects are fibroblast [9,19] and smooth muscle cell  proliferation, vasoconstriction,  platelet activation, [22,23] and induction of changes in cell motility, [24–26] it seems unlikely that its presence in wounds would have no effect on the wound healing process. Therefore we postulate that lysophosphatidate plays a physiologic role in wound healing, and that interference of local anesthetics with lysophosphatidate signaling can adversely affect the healing process.
Adverse effects of local anesthetics on wound healing have been well described, [1–3,27] but the mechanisms remain unclear. Although local anesthetics are present in the wound for a short time only, their effects are noticeable for longer periods. For example, fibroblast proliferation and collagen deposition are impaired after application of local anesthetics  and might explain subsequent decreases in wound breaking strength.  Although collagen deposition does not occur until 3–5 days after wounding,  a time when the local anesthetic is unlikely to be present still in appreciable concentrations, local anesthetic-induced delays in fibroblast proliferation could explain this. Transient inhibition of lysophosphatidate signaling in the early stages of wound healing may disturb the subsequent wound healing cascade. Alternatively, an irreversible component of lidocaine block of lysophosphatidate signaling, as suggested by our findings, might play a role. Clinically, signs of inflammation and occasional necrosis have been observed after local anesthetic were administered around wounds. [1,27–29] Inhibition of proliferation and motility of immune system cells by local anesthetic block of lysophosphatidate signaling might explain some of these findings. On the other hand, inhibition of lysophosphatidate signaling by local anesthetics might have beneficial effects as well. Lysophosphatidate-induced vasoconstriction and platelet aggregation may be important immediately after wounding but will decrease blood flow in the regenerating wound bed. Local anesthetics might reverse this and thereby provide a better healing environment. Indeed local anesthetics have been shown to inhibit platelet aggregation, [30–32] and reverse vasospasm, [33–36] by mechanisms that have not been identified yet.
The exact site of interaction between the local anesthetics and the lysophosphatidate signaling pathway is important. Although the main site of action of local anesthetics is the Na sup + channel,  secondary activities exist. For example, local anesthetics inhibit Ca2+ entry through voltage-gated channels, by direct interference with the channel molecule.  In our study, we could rule out an effect of the anesthetics on the Ca2+-activated Cl sup - channel. As G protein-coupled pathways signal over many more intermediates than channels do, localization of the site of anesthetic action is relevant. The lack of effect on lysophosphatidate signaling after intracellular microinjection of IP3suggests that the site of action is extracellular. This is confirmed by the lack of local anesthetic effect on angiotensin signaling. The latter finding, in addition, makes a nonspecific membrane effect unlikely. Therefore, the most likely site of action is the lysophosphatidate receptor itself. The lipid moiety of the local anesthetic molecule might interact with the presumed hydrophobic agonist binding site on the receptor. It is important that the more lipophilic local anesthetic bupivacaine (octanol-buffer partition coefficient 3420 ) inhibited lysophosphatidate signaling at significantly lower concentrations than did lidocaine (partition coefficient 366 ). (The agonist binding site might allow other anesthetic interactions, as we have shown recently that the lipophilic anesthetic propofol similarly interferes with lysophosphatidate signaling. ) Although the sigmoidal shape of the concentration-inhibition relation is compatible with a receptor site of action, the variability associated with high lysophosphatidate concentrations [7,14] makes it difficult to show competitive interactions conclusively. One other potential site of action deserves further study, namely a direct interaction with the lysophosphatidate molecule, or with lysophosphatidate binding to albumin, its physiologic carrier.
Our data indicate that commonly used local anesthetics can interfere significantly with lipid mediator signaling. Various lipid mediators (platelet-activating factor, prostaglandins, leukotrienes) are relevant in the perioperative context, making it important to understand their interactions with commonly used local anesthetics. This is of particular relevance to the newer, long-acting compounds.