The Effects of Extracellular pH with and without Bicarbonate on Intracellular Procaine Concentrations and Anesthetic Effects in Crayfish Giant Axons 

Abdominal nerve bundles from crayfish (weight, 20–64 g; n = 45) were isolated immediately before each experiment in HEPES - van Harreveld's solution (195 mM NaCl, 5.4 mM KCl, 13.5 mM CaCl², 2.6 mM MgCl², 7.5 mM NaHCO³, and 5 mM HEPES) adjusted to pH 7.60 by 2 N NaOH to achieve a bicarbonate-containing solution with a pH of 7.6 (experimental group bicarb/7.6). The four giant axons in each nerve bundle were exposed by removing the nerve sheath between the third and fifth ganglia from the distal end. The nerve bundle was treated for 30 min with 1 mg/ml collagenase (Wako Pure Chemical, Osaka, Japan) to ease microelectrode insertion. (In our previous study, a 30-min collagenase treatment did not affect the generation and propagation of action potentials.) Then the nerve bundle was fixed in a recording chamber and perfused with bicarb/7.6 at 10 ml/min (Figure 1). One of four axons in a new nerve bundle was used for each measurement in all experiments. Bicarbonate-free HEPES - van Harreveld's solutions (nonbicarb group) also were prepared by equimolar (7.5 mM) replacement of NaHCO³with NaCl and adjusted to either pH 7.6 (experimental group nonbicarb/7.6) or 8.0 (experimental group nonbicarb/8.0) with NaOH. All experiments were performed at room temperature (24 - 28 [degree sign]C).

The pH-sensitive microelectrode was constructed using the techniques of Ammann et al. [11]and Kubota et al. [12]A glass capillary was pulled into a micropipette (tip diameter < 1 [micro sign]m), baked at 200 [degree sign]C for 60 min, and then made surface hydrophobic by exposure to vapor of di-butyl-chlorosilane (Chisso, Tokyo, Japan) at 200 [degree sign]C for 30 min, with excess saline removed by additional baking. The tip of the glass micropipette was beveled by grinding and the micropipette was filled, moving from the end toward the tip, with a small amount of hydrogen ion exchanger (pH sensor; IE-010, WPI, Sarasota, FL), then with the internal solution (1 M KCl), through an extended fine polyethylene tube. This apparatus was connected to a high-impedance amplifier (FD-233; WPI) via the silver/silver chloride junction of the electrode holder. The absence of bubbles was confirmed by microscopic examination. Voltage responses to the various pH levels achieved by adding NaOH or HCl were measured in an intracellular-like solution (100 mM KCl, 10 mM NaCl, 5 mM HEPES) and recorded in a thermal array recorder (RTA-1200M; Nihonkoden, Tokyo, Japan) to obtain the standard curve. The slope of the standard curve from seven independent experiments ranged from -51 to -64 mV/pH.

The procaine-sensitive microelectrode was constructed according to a similar method. [13–16]We chose procaine for these experiments because it exists primarily in cationic form within physiologic pH levels because of a high dissociation constant (pK^a), [17]and it can be measured readily by specificity designed microelectrodes. The procaine sensor contained 98 weight/% onitrophenyl octyl ether (Sigma Chemical Co., St. Louis, MO), 2 weight/% polyvinylchloride (Wako Pure Chemical, Osaka, Japan), 5 mM Na tetraphenylborate (Sigma Chemical), and 5 mM procaine HCl (Wako Pure Chemical). These compounds were dissolved in the same volume of tetrahydrofurane (Wako Pure Chemical), and the solvent was evaporated at room temperature. To apply the procaine sensor as a liquid ion exchanger, we used polyvinylchloride (an agent known to improve the responsiveness and stability of the sensor for a small-tipped microelectrode [12,18]) in an amount less than that previously reported for the membrane-type procaine sensor. [13]The procaine sensor was filled similarly as the pH sensor. The microelectrode was back-filled with 4 M KCl) as an internal solution, and its surface, excluding the tip, was coated with electroconductive silver paint [19](Dotite S-1; Fujikurakasei, Tokyo, Japan) to overcome high circuit capacitance and the resultant slow response. The procaine-sensitive microelectrode was calibrated with procaine dissolved in an intracellular-like solution within the concentration range of 0.01 to 5 mM (pH 7.2;Figure 2).

The procaine-sensitive microelectrodes exhibited linear response over the concentration range of 0.2 to 5 mM procaine. The slope of the electrode responses were determined as 51.6 +/- 2.53 mV/log[procaine](means +/- SD) by regression analysis of the linear portion of calibration curves. To evaluate the reproducibility, the electrode potentials were measured repeatedly with increasing and decreasing procaine concentrations, [13]and the coefficient of variance of our procaine-sensitive microelectrode ranged from 0.53% to 0.99%(mean, 0.77%). The influence of pH on the response of the procaine-sensitive microelectrode was investigated. The electric response was measured in 1 mM procaine containing intracellular-like solution with changing pH values by adding NaOH or HCl. The change in pH did not affect the response of the procaine-sensitive microelectrode at pH 8.0 or lower (Figure 3). Selectivity coefficients (K^A,B^pot) of the procaine-sensitive microelectrode were determined. The K^A,B^potvalues were the ability of the sensor electrode to distinguish between different ions in the same solution and were evaluated by the fixed interference method (mixed solution method). [20]The K^A,B^potvalue for Na⁺, K⁺, Mg²⁺, and Ca²⁺were 7.19 x 10^-5, 1.92 x 10^-4, 2.47 x 10^-5, and 4.52 x 10^-5, respectively.

When measuring the intracellular concentration, the potential of the pH-sensitive microelectrode (E^ph) is the sum of the membrane potential (E^m) and the potential produced by hydrogen ion. [21]Similarly, the potential in the procaine-sensitive microelectrode (E^proc) is the sum of the E^mand potential produced by procaine cation. Therefore, a pH-sensitive or a procaine-sensitive microelectrode was inserted into the same axon (one of four) with a conventional glass E^mmicroelectrode filled with only 3 M KCl, with a distance <or= to 2 mm between electrodes. Then E^mwas subtracted from the sensor electrode potentials.

The E^pHand E^mvalues were monitored continuously using a storage oscilloscope (VC-11; Nihonkoden, Tokyo, Japan) and the thermal array recorder described before. After two microelectrodes were inserted into the giant axon, we perfused the preparation for about 30 min with bicarb/7.6. When the potentials became stable, the perfusion medium was switched from bicarb/7.6 to nonbicarb/7.6 or nonbicarb/8.0, or it was unchanged, and E^mand E^pHwere measured continuously for 15 min. Intracellular pH was calculated according to the following equation: intracellular pH = 7.6 +(E (^pH)- E^m)/S, when s is the slope of the standard curve.

E^procand E^mwere monitored continuously using the same equipment described to measure intracellular pH. After the electrodes were inserted, the axon preparation was perfused with bicarb/7.6 for about 30 min. When the potentials became stable, the perfusion medium was switched from bicarb/7.6 to 1 mM procaine dissolved in bicarb/7.6 or nonbicarb/7.6 or nonbicarb/8.0, and the axon was perfused for 15 min while E^procand E (^m) were recorded. Each measurement was performed in an axon of a new nerve bundle.

As shown in Figure 2, the potential of the procaine-sensitive microelectrode deviated from the linear response toward the logarithm of procaine concentration and decreased the slope at <0.2 mM. In a general property, the potential of an ion-selective electrode is expressed in the modified Nernst equation (Nicolsky-Eisenmann equation), and the potentials converge to a constant under the detection limit. [20]The potential is not zero even when procaine is absent. We shifted the calibration curves to make the potential 0 mV at 0 mM and expressed the linear portions by the following equation: E = S log¹⁰[proc]+ C, where E is relative E^procversus 0 mM procaine, s is slope of the linear portion, [proc] is procaine concentration expressed in millimoles, and C is relative E^procat 1 mM procaine from 0 mM. Because no procaine presented in axoplasm before exposure to procaine, intracellular procaine concentration was calculated by the following equation:[procⁱ]= 10 [Lambda]{(Delta E - C)/S}, where [procⁱ] is intracellular procaine expressed in millimoles, and Delta E is the change in the difference between E^procand E^mduring exposure to procaine.

To minimize possible damage to the nerve membrane, only an E^melectrode was used to measure action potential inhibition. Action potentials were evoked by a square wave pulse of supramaximal electric intensity (0.2 ms at 0.1 Hz) using an electric stimulator (model SEN-3301, Nihonkoden) and converted to dV/dt by a differentiator (model DDT-A; WP1). After electrode insertion, the perfusion protocol replicated the protocol to measure intracellular procaine concentration, and action potentials and the dV/dt of action potentials were recorded continuously. Because the positive phase of dV/dt is proportional to the inward ionic current, the maximum (peak) value of dV/dt (dV/dt^max) is a more sensitive index than action potential amplitude for evaluating local anesthetic action. [7]In general, the anesthetic effect is evaluated with a voltage-clamping method, but we measured dV/dt^maxfor two reasons:(1) the giant axons of crayfish are smaller in diameter than those of squid, and the voltage clamp is too difficult for accurate measurement, and (2) the dV/dt^maxcan evaluate local anesthetic actions with the same experimental system used to measure intracellular ion (hydrogen or procaine) concentration. The control value was calculated as the mean of dV/dt^maxfor 1 min just before the switch from bicarb/7.6 to a solution containing 1 mM procaine. The percentage decrease in dV/dt^maxfrom control was calculated as the ratio of the decrease in dV/dt^maxfrom control after procaine perfusion.

All values are expressed as means +/- SD. Comparisons among the three solutions were made by one-way analysis of variance, followed by the Scheffe test. Probability values <0.05 were considered significant.

(Figure 4) shows a typical recording of E^pHand E^m. After E^pHand E^mbecame stable in the bicarb/7.6 group, axons were perfused with nonbicarb/8.0 for 15 min, during which time E^pHdecreased from -61 to -68 mV and E^mremained unchanged at -88 mV. Thus the difference between E^pHand E^mdecreased from 27 to 20 mV in the nonbicarb/8.0 solution, indicating that intracellular pH increased from 7.16 to 7.28. The mean intracellular pH after 15 min of perfusion was significantly lower in bicarb/7.6 (7.16 +/- 0.07, n = 17) solution than in the nonbicarb/7.6 (7.33 +/- 0.09, n = 18) and nonbicarb/8.0 (7.33 +/- 0.12, n = 14) groups (P < 0.01) but did not differ between the nonbicarb/7.6 and nonbicarb/8.0 groups (Figure 5).

As shown in Figure 6, changing the perfusion medium from bicarb/7.6 without procaine to nonbicarb/8.0 containing 1 mM procaine increased E^procfrom -59 to +17 mV within 15 min but had little effect on E^m(-90 to -92 mV). The difference between E^procand E^mincreased from 33 to 107 mV, indicating that the intracellular procaine concentration reached 0.46 mM after 1 mM procaine in the nonbicarb/8.0 solution. The mean intracellular concentration of procaine was significantly higher in the bicarb/7.6 (0.53 +/- 0.08 mM, n = 7)(P < 0.05) and the nonbicarb/8.0 (0.58 +/- 0.13 mM, n = 9) groups (P < 0.01) than in the nonbicarb/7.6 (0.32 +/- 0.14 mM, n = 8) solution (Figure 7). No difference in procaine concentration, however, was observed between the bicarb/7.6 and the nonbicarb/8.0 solution.

As shown in Figure 8, when compared to dV/dt^maxin the bicarb/7.6 group without procaine, perfusion of 1 mM procaine in the nonbicarb/8.0 group for 15 min decreased dV/dt^maxby 15% from 826 to 701 V/s. The mean percentage decrease in dV/dt^maxwas significantly greater in the bicarb/7.6 (23 +/- 5%, n = 11) group (P < 0.01) and the nonbicarb/8.0 (21 +/- 8%, n = 10) group (P < 0.05) than in nonbicarb/7.6 (13 +/- 3%, n = 9) group (Figure 7). No difference in percentage decrease in dV/dt^maxwas observed between the bicarb/7.6 and nonbicarb/8.0 groups.

Our results show that intracellular pH increased reversibly and intracellular concentration of ionized procaine was lower in the absence of sodium bicarbonate (as defined by the comparison of the bicarb/7.6 and nonbicarb/7.6 groups). The change in intracellular concentration of ionized procaine showed a similar tendency with the degree of dV/dt^maxinhibition, and the coefficient of correlation was 0.85 for linear regression. These findings imply that extracellular bicarbonate-related changes in intracellular pH affect anesthetic action by variation of intracellular concentration of ionized procaine, because the extracellular concentration of uncharged procaine is equal in both solutions. However, alkalinization of extracellular solution, without difference of intracellular pH, increased intracellular concentration of ionized procaine and enhanced anesthetic effect (as demonstrated by the comparison of the nonbicarb/7.6 and nonbicarb/8.0 groups). These findings indicate that the amount of membrane-permeable (unionized) procaine increases in response to an increase in extracellular pH, making more anesthetic available to traverse the neural membrane and increase the intracellular anesthetic concentration. [22–24]Combined, it is likely that the presence of sodium bicarbonate affects the intracellular concentration of ionized procaine and anesthetic action by changing both intra- and extracellular pH.

When the pH difference across the membrane (i.e., cross-membrane pH gradient) increases, especially with intracellular pH lower than extracellular pH, both the increase of base traversing into the axoplasm increased the amount of cation in axoplasm (i.e., ion trapping). [9,25,26]Intracellular acidification and extracellular alkalinization increase the cross-membrane pH gradient, thereby enhancing the effect of ion trapping and increasing the concentration of cation. In the absence of ion trapping, the intracellular procaine concentration would depend only on the extracellular procaine base concentration. If so, the extracellular concentration of un-ionized procaine was equal in the bicarb/7.6 and nonbicarb/7.6 solutions, leaving only the ratio of ionization at the intracellular pH to determine intracellular concentration of ionized procaine. Using the dissociation Equation withpK^avalue (9.06), [17]this ratio is 98.8% and 98.2%, respectively, in the bicarb/7.6 and nonbicarb/7.6 solutions, a difference of only 0.6%. Our finding indicated a difference of 66% in the intracellular concentration of ionized procaine (bicarb/7.6 = 66% > nonbicarb/7.6). Because the pK^avalue of procaine was much greater than intracellular pH, the greater part of un-ionized procaine would ionize after permeating into the cytoplasm. The increase in the cross-membrane pH gradient might increase both the amount of permeation and intracellular concentration of ionized procaine; that is, ion trapping might be an important part of the mechanism for bicarbonate potentiation of the local anesthetic effect. On the other hand, the un-ionized (and ionized) local anesthetics are reported to block the Na (⁺) channel via the hydrophobic pathway. [17,27]As the concentration of local anesthetic in the membrane is equal to the product of intracellular concentration and membrane - buffer partition coefficient, [17]the change in extracellular or intracellular (or both) pH would affect the procaine concentration in the membrane and the anesthetic action. As an index of hydrophobicity, the octanol - buffer partition coefficient for un-ionized local anesthetic is much higher than for the ionized form; however, the same parameters for membrane - buffer partitioning have little difference because the membrane is heterogeneous in physicochemical properties. [17]The calculated ratio of un-ionized form was <8% in any phase of our experiments. The change in extracellular or intracellular pH might have little effect on the hydrophobic pathway, except for a change in the intracellular concentration of ionized procaine. With other local anesthetics, using the dissociation equation, the higher the pK^athe greater will be the effect of extracellular alkalinization. However, a higher pK^aalso will increase intracellular proton consumption with the ionization of local anesthetic, potentially decreasing the effect of intracellular acidification. Thus the addition of sodium bicarbonate to other local anesthetics may potentiate the anesthetic effect by the same mechanism as procaine, but the magnitude of ion trapping is unknown without studies for each local anesthetic.

The cross-membrane pH gradient was greater in the nonbicarb/8.0 solution (8.0 - 7.33 = 0.67) than in the bicarb/7.6 solution (7.6 - 7.16 = 0.44), resulting in no difference in the intracellular procaine concentration of these two solutions. Thus the cross-membrane pH gradient appears to decrease by intracellular proton consumption occurring with procaine ionization in axoplasm. This interpretation is consistent with that of previous investigations showing that 3.75 mM procaine increased intracellular pH by 0.45 over pH 8.2 in frog sartorius muscle, [23]and 5 mM procaine increased intracellular pH by 0.25 over pH 7.5 in snail subesophageal neurons. [28]Because the H²CO^3/HCO3-system is an important buffer of the axon, in the nonbicarb groups, the buffering power of axoplasm would decrease with the reduction in the intracellular concentration of bicarbonate ion by the efflux of carbon dioxide, [21]resulting in more variable intracellular pH in the nonbicarb/8.0 solutions than in the bicarb/7.6 solution with procaine dissociation. From these points, the change in intracellular pH might affect the modification of the local anesthetic effect more strongly than extracellular pH. However, we could not measure intracellular pH in the presence of procaine, because the pH sensor was not selective to organic cations.

In desheathed sciatic nerve from the frog, Wong et al. [29]reported that a reduction in the amplitude of the compound action potential was 3.9 times greater with procaine containing 2.5 mM bicarbonate at pH 7.2 than with procaine without bicarbonate at the same pH. Our current results show that the decrease in dV/dt^maxwas 1.6 times greater in bicarb/7.6 solution than in the non-bicarb/7.6 solution. The difference in potentiating magnitude between our results and theirs may be due to a difference in the bicarbonate and carbon dioxide concentrations. Ackerman et al. [30]suggested that carbon dioxide may be important in achieving rapid onset of action with epidural anesthesia with bicarbonate-alkalinized local anesthetic solutions. In addition, carbon dioxide alone directly affects the function of axons, [2,9]as indicated by an increase in the threshold for firing or a decrease in action potential amplitude in bicarbonate-carbon dioxide solution without local anesthetic. In contrast, the extremely low carbon dioxide concentration might participate in attenuating the anesthetic effect in the nonbicarb groups.

We measured the concentration of intracellular ionized procaine using procaine cation-sensitive microelectrodes. Although the cation:base ratio of procaine changes when the surrounding pH is altered, procaine exists primarily in cationic form at physiologic pH. The procaine-sensitive microelectrodes were calibrated in fixed pH (pH 7.2), and the ratio of ionized procaine was 98.6%, as calculated with the pK^a(9.06). However, intracellular pH values were not constant, resulting in variable ratios of cation:base, which could produce errors in measurement. For example, measured intracellular pH ranged from 7.04 to 7.54, the calculated ratio of ionized procaine ranging from 97.1%(at pH 7.54) to 99.1%(at pH 7.04). Although the range of ionizing ratio between the variation and calibration was within 1.5%, the measuring error of our procaine-sensitive microelectrode system was near 1%, and the degree of change in the ionization ratio would negligibly affect measurement of the intracellular concentration of ionized procaine. We could not use a bicarbonate-containing solution (bicarb/8.0). Because the van Harreveld's solution has a high calcium concentration, bicarbonate addition or pH elevation precipitate calcium carbonate. However, a change in calcium concentration might affect anesthetic action, [31]so we tried to avoid reducing calcium for bicarbonate addition or pH elevation.

In conclusion, extracellular reduction of bicarbonate decreased the cross-membrane pH gradient by intracellular alkalinization and decreased both the intracellular concentration of ionized procaine and the local anesthetic effect. Extracellular alkalinization increased the pH gradient, the intracellular concentration of ionized procaine, and local anesthetic effect. Ion trapping due to an increased cross-membrane pH gradient is suggested as an important mechanism of the potentiation of the local anesthetic effect by the addition of sodium bicarbonate.

The authors thank Winifred von Ehrenburg and Professor Akihiko Wada (Department of Pharmacology, Miyazaki Medical College) for their valuable review and advice on the manuscript.