Freshly dissociated myocytes from nonpregnant, pregnant, and postpartum rat uteri have been studied with the tight-seal patch-clamp method. The inward current contains both INa and ICa that are vastly different from those in tissue-cultured material. INa is abolished by Na+-free medium and by 1 μM tetrodotoxin. It first appears at ∼−40 mV, reaches maximum at 0 mV, and reverses at 84 mV. It activates with a voltage-dependent τ of 0.2 ms at 20 mV, and inactivates as a single exponential with a τ of 0.4 ms. Na+ conductance is half activated at −21.5 mV, and half inactivated at −59 mV. INa reactivates with a τ of 20 ms. ICa is abolished by Ca2+-free medium, Co2+ (5 mM), or nisoldipine (2 μM), and enhanced in 30 mM Ca2+, Ba2+, or BAY-K 8644. It first appears at ∼−30 mV and reaches maximum at +10 mV. It activates with a voltage-dependent τ of 1.5 ms at 20 mV, and inactivates in two exponential phases, with τ's of 33 and 133 ms. Ca2+ conductance is half activated at −7.4 mV, and half inactivated at −34 mV. ICa reactivates with τ's of 27 and 374 ms. INa and ICa are seen in myocytes from nonpregnant estrus uteri and throughout pregnancy, exhibiting complex changes. The ratio of densities of peak INa/ICa changes from 0.5 in the nonpregnant state to 1.6 at term. The enhanced role of INa, with faster kinetics, allows more frequent repetitive spike discharges to facilitate simultaneous excitation of the parturient uterus. In postpartum, both currents decrease markedly, with INa vanishing from most myocytes. Estrogen-enhanced genomic influences may account for the emergence of INa, and increased densities of INa and ICa as pregnancy progresses. Other influences may regulate varied channel expression at different stages of pregnancy.
Smooth muscles comprise a diverse group of involuntary excitable tissues, which are dispersed widely throughout the body, subserving important physiological functions. Because individual smooth myocytes are small, their ionic channel functions have not been studied in earnest until the introduction of the tight-seal patch-clamp method (Hamill et al., 1981), and successful regimens of dissociating individual myocytes from tissue assemblies (Bagby et al., 1971; Fay and Delise, 1973; Momose and Gomi, 1980). However, the usefulness of enzyme-dissociated smooth myocytes as physiological models has not been tested. The uterine smooth muscle is singularly suitable for addressing this issue because it has been studied as multicellular preparations (Anderson, 1969; Mironneau, 1974; Kao and McCullough, 1975), as dissociated cells (references below), and in tissue culture (e.g., Mollard et al., 1986; Amedee et al., 1987; Toro et al., 1990; Rendt et al., 1992).
Because the structure and function of uterine smooth muscle change markedly under the influence of ovarian hormones and during pregnancy (see Kao, 1989), it is also suitable for studying hormonal regulation of ionic channels. Although several papers on freshly dissociated uterine myocytes have appeared (Ohya and Sperelakis, 1989; Inoue et al., 1990; Inoue and Sperelakis 1991; Miyoshi et al., 1991; Piedras-Renteria et al., 1991), each dealt with a limited aspect of myometrial function. In this and another paper (Wang, S.Y., M. Yoshino, and C.Y. Kao, manuscript submitted for publication), we will provide a wider coverage of ionic channel functions of uterine myocytes in the nonpregnant state and throughout pregnancy. There are significant changes in these functions during pregnancy that not only profoundly influence the ultimate physiological functions, but also illustrate some hormonal influence on ionic channels. In this paper, we address the inward currents and changes in them during pregnancy; in the other paper, we will deal with the outward currents. Preliminary accounts have been given (Suput et al., 1989; Kao et al., 1989; Yoshino et al., 1989, 1990; Wang and Kao, 1993; Wang et al., 1996).
Female rats (Sprague-Dawley; Harlan Sprague Dawley Inc., Indianapolis, IN) were mated individually, and pregnancy was dated from the morning when cervical plugs were found. Dating was confirmed at the experiment from fetal size (Witschi, 1956). Uteri from days 2–22 (term) of pregnancy and postpartum were used. For nonpregnant rats, the estrus status was ascertained by cytological examinations of vaginal washing.
Isolation of Uterine Myocytes
The procedures for cell isolation are similar to those used for taenia coli myocytes (Yamamoto, et al., 1989). Briefly, strips (∼2 × 15 mm) of endometrium-free longitudinal myometrium were incubated in 0.1% collagenase (Wako Bioproducts, Richmond, VA) in a Ca2+-free modified Krebs solution for 30 min. They were then washed free of collagenase before being triturated. Use of higher concentrations of collagenase or other additional proteolytic enzymes were associated with more cells with large leakage conductance and poor ionic currents, whereas longer incubations were associated with higher incidences of myocytes coated with a layer of optically transparent material that interfered with cell settling and seal formation with electrodes. In practice, concentrations of 200–300 cells/μl (by hemocytometer counting) have been obtained (see also Fig. 1 A). The myocytes, kept in a high K+, Ca2+-free medium (“KB” medium, Isenberg and Klöckner, 1982) at 5–8°C, retained good electrophysiological properties for 3 d. Most data were obtained within 6 h after isolation. All myocytes used for this work were relaxed and adhered to the bottom of the plastic or glass chamber with no additional substrate. More than 90% remained relaxed when exposed to Ca2+; after visual screening, 70–80% of the cells selected for study would have good ionic currents.
The methods are similar to those used for the work on taenia myocytes (Yamamoto et al., 1989), in which the quality of the voltage clamp and the isopotentiality of the whole cell were demonstrated experimentally. Adequacy of control of the uterine myocytes can be surmised from the gradual current–voltage (I-V)1 relations in the negative resistance region. For this work, the range of capacitance cancellation in the amplifier (EPC-7; List Electronic, Darmstadt, Germany) was extended to cover 200 pF to accommodate the larger pregnant uterine myocytes. Because of the presence of fast INa, series-resistance compensation and capacitance cancellation were essential, as these procedures introduced some positive feedback, accelerating the charging of the membrane capacity (Sigworth, 1986). For the usual recording, the settling time was <800 μs. For small myocytes under optimal compensation, the transient artifact could be reduced to <150 μs. In selecting results for inclusion, we looked for a state of zero current between the end of the transient and the beginning of INa (e.g., Fig. 3 B). Leakage currents and residual transient artifacts after hardware correction were corrected with a p/4 protocol.
All experiments were conducted at room temperature (∼22°C).
The bath contained (mM): 135 NaCl, 5.4 KCl, 3 CaCl2, 1 MgCl2, 5 glucose, 10 HEPES, pH adjusted to 7.25 with NaOH. Ca2+ was sometimes increased (as specified), with an equimolar reduction in Na+. The pipette solution contained (mM): 140 KCl, 1 EGTA, 2 Na2ATP, 10 HEPES, pH adjusted to 7.25 with KOH. Osmolarity of all solutions was routinely maintained at 275–290 mosM. For studying the inward current, the K+ in the pipette solution was replaced with Cs+, and the bath contained 30 mM tetraethylammonium chloride (TEA) and 5.4 mM CsCl with equimolar reduction of Na+. Run-down of ICa, which was a problem only in small myocytes of nonpregnant or early pregnant uteri, could be obviated by incorporating 5 mM each of potassium salts of pyruvate, oxaloacetate, and succinate in the pipette solution (Klöckner and Isenberg, 1985).
Because the uterine myocyte hypertrophies during pregnancy, cellular morphometric data at different stages of pregnancy are needed to correlate with physiological observations. Although freshly dissociated uterine myocytes were usually long, slender, and fusiform, pleomorphism was common, especially in late pregnancy when cells with irregularly swollen central regions or terminal arms (Fig. 1) constituted ∼10% of the population. In the present study, to avoid possible inadequate space clamp in cells with such complex geometry, we used only long and relaxed myocytes, assuming that the pleomorphic cells shared similar basic electrophysiological properties.
Morphometric data at three stages were collected: nonpregnant (also representing early pregnancy from days 1–8), midpregnancy (14 d, also representing days 9–16), and late pregnancy (days 17–21). Table I shows that during pregnancy, the maximum diameter of the individual myocyte increased 2.8-fold from 8 to 22 μm, and the length increased 1.7-fold from 130 to 225 μm. Such changes led to a fourfold increase of surface area to 7,600 μm2, and an eightfold increase of cell volume to 21 pl in the term myocyte.
In the present experiments, the seal resistance ranged from 5 to 40 GΩ. When studied with Cs+-filled pipettes, the input resistance of the whole myocyte ranged from 0.5 to 3 GΩ. Resting and action potentials were not routinely measured because the Cs+-pipette solution caused a significant depolarization. However, myocytes for this work were selected for their low leakage conductance. If the average input resistance were taken as 1 GΩ, then the specific membrane resistance for the averaged size late-pregnant myocyte would be 76 kΩ-cm2.
Consistent with hypertrophy, the cell capacitance increased as pregnancy progressed (Table II). In early pregnancy, the average cell capacitance remained ∼30 pF, slightly higher than that of the nonpregnant myocyte (25 pF). In midpregnancy, capacitance increased markedly, possibly stimulated by fetal growth and stretch of the uterus. In late pregnancy, capacitance stabilized at ∼110 pF, because there were no statistically significant differences among the values listed for days 18–21. Within 18-h postpartum, there were no significant changes in the cell capacitance.
As the amount of caveolae in myocytes at different stages of pregnancy is not known, estimation of specific membrane conductance is based on the morphometric surface. Taking the average of 108 pF as the cell capacitance for the late-pregnant myocyte with an average surface area of 7,600 μm2 (Table I), the specific membrane capacitance works out to be 1.42 μF/cm2. For the nonpregnant myocyte, based on a surface area of 1,928 μm2, the specific capacitance works out to be 1.30 μF/cm2.
Coexistence of INa and ICa
The inward current consists of two distinct components: a fast activating and inactivating component, followed by a more slowly activating and more sustained component (Fig. 2). Although the peak magnitudes of the two components and the degree of overlap vary considerably from cell to cell, the slow component is seen in all myocytes, and the fast component in half of the myocytes from nonpregnant estrus uterus or early pregnancy, and >90% of myocytes from late pregnancy.
The nature of each component is readily identifiable by ion substitution and by selective blocking agents. In a Na+-free medium, the fast component was abolished, whereas the slow component and its associated tail current were unchanged (Fig. 3,A). The fast component was fully blocked by tetrodotoxin (TTX) at 1 μM concentration (Fig. 3,B, also Ohya and Sperelakis, 1989). In a Ca2+-free medium, the slow component was abolished, leaving the fast component (Fig. 3,C). It was also abolished when nisoldipine (2 μm) was added to the bath (Fig. 4 C). Thus, the fast component can be identified as INa and the slow component as ICa.
INa first appeared at ∼−40 mV, reached a maximum at 0–10 mV, and reversed at 80–84 mV (Fig. 3 D), in good agreement with the expected ENa. The magnitude of INa varied widely, in part because of differences in cell size, and in part with the stage of pregnancy (see below). When normalized to cell capacitance, and adjusted for the morphometric surface areas, peak INa works out to be 2.77 μA/cm2 for nonpregnant myocytes, and 5.10 μA/cm2 for late-pregnant myocytes.
In 3 mM [Ca2+]o, ICa was first seen at ∼−30 mV, and reached a maximum at +10 mV (Fig. 4,A). The magnitude of ICa varied because of cell size and stage of pregnancy. When normalized to cell capacitance and the morphometric surface, peak ICa (in 3 mM Ca2+) works out to be 5.67 μA/cm2 for the nonpregnant myocyte and 3.43 μA/cm2 for the late-pregnant myocyte. In 30 mM Ca2+, the activation of ICa shifted positive by ∼15 mV. The maximum current increased, and the voltage at which it was attained also shifted positive by ∼20 mV (Fig. 4 A), as expected from a screening effect on surface negative charges (Frankenhauser and Hodgkin, 1957).
When 3 mM Ba2+ replaced Ca2+ in the bath solution, the inward current activated as rapidly, reached approximately the same maximum but slightly later, and the I-V relation was shifted ∼10 mV to the negative. The inactivation was markedly slower (Fig. 4 B), probably because of interference with Ca2+-mediated Ca2+ inactivation (see below). The Ba2+ current, as well as ICa, could be fully blocked by 5 mM Co2+ (not illustrated).
BAY K 8644, a dihydropyridine compound, increased the magnitude of the ICa without shifting the I-V relations (Fig. 4,D). In contrast to Ba2+, it did not slow the inactivation, and the larger current was inactivated at an appreciably faster rate. Nisoldipine (2 μM) readily blocked the ICa of pregnant (Fig. 3,B, legend) and nonpregnant myocytes (Fig. 4 C), similar to those reported for 10 μM nifedipine (Miyoshi et al., 1991) and on Ba2+ current (Ohya and Sperelakis, 1989).
INa and ICa in Relation to Stages of Pregnancy
ICa was recordable in all uterine myocytes. INa was additionally seen in 19 of 38 myocytes taken from nonpregnant estrus, metestrus, or diestrus, but not from proestrus, animals. INa was seen in 11 of 22 myocytes from 2-d pregnant uterus, and almost all myocytes from day 5 of pregnancy to term.
Table III summarizes the INa and ICa at different stages of pregnancy. The data are based on currents that were stable for at least 15–20 min. For comparing cellular properties, only myocytes with both INa and ICa were included. Several features are apparent in the data, (a) from baseline values at the beginning of pregnancy, the densities of both INa and ICa increased during the first trimester; (b) they then declined during the second trimester; (c) whereas INa reached its lowest density on the 17th d and began to recover by the 18th d, ICa continued to decline and reached its lowest value on the 18th d; (d) from days 2 to 17, peak INa density was less than peak ICa density, but in late-pregnant myocytes, this relation was reversed; and (e) the net changes over the course of pregnancy are that the density of peak INa increased by 1.8-fold, while that of ICa decreased by 1.7-fold.
In 12 myocytes from uteri that were 14–18 h postpartum, ICa was observed in every myocyte, but INa was observed in only one. This myocyte was large, with a cell capacitance of 200 pF. Both the average peak ICa density and the peak INa density of the single myocyte were lower than the least values recorded during pregnancy. The decline in the current densities occurred even as the cell capacitance remained the same (Table II). Also, peak INa/ICa in this single myocyte was 0.57, similar to that in nonpregnant myocytes.
activation and inactivation Kinetics of Activation and Inactivation
To avoid possible errors in time-to-peak measurements caused by current inactivation, kinetics of activation was examined by curve fitting the early part of the ionic currents. With optimum compensation, INa can be distinctly separated from the capacitive current transient (e.g., Figs. 2 and 3). At ∼22°C, INa reached a peak in ∼1 ms. The best fit was obtained with a fourth-power function (Fig. 5,A), with voltage-dependent τ's varying between 0.39 ms at −20 mV and 0.18 ms at 20 mV (Fig. 5 B).
For ICa, inactivation was more complex. In a small number of myocytes, a small fraction (<5%) of ICa remained even at the end of a 400-ms step (Fig. 6,C, inset). The inactivation time course can be fitted with two exponential phases, with τ's of ∼32 ms (τf) and 133 ms (τs) (Fig. 6,C). τf was strongly voltage dependent, showing a U-shaped relation with the fastest rate at 10 mV (usually <40 ms) and significantly slower rates at either more negative or more positive voltages (Fig. 6 D). τs may also be voltage dependent (175 ms at −10 mV, 81 ms at +20 mV, 289 ms at +40 mV), but there was too much variability in the small sampling to support any statistically significant differences.
Deactivation of both INa and ICa follow single exponential time constants. In five myocytes, in the presence of a high concentration of nisoldipine (20 μM), INa, when repolarized from 0 to −80 mV, decreased with a τ of 0.87 ± 0.10 ms. In another group of five myocytes in which INa had been abolished by depolarizing steps of >10 ms, ICa (repolarizing from +10 to −80 mV) deactivated with a τ of 2.71 ± 0.23 ms.
Steady State Activation and Inactivation
Fig. 7 shows the steady state activation and inactivation of INa and ICa. These observations were made on four to five different cells from 17-, 18-, and 20-d pregnant uteri that show both INa and ICa. To facilitate the experiments, one current was blocked so that the other could be studied.
For both INa and ICa, the activation relations followed Boltzmann distributions. For INa (Fig. 7,A), significant conductance was first seen at ∼−40 mV. Half activation occurred at −21 mV, full activation at ∼+20 mV. The slope factor was 5 mV. For ICa (Fig. 7 B), the first detectable conductance appeared at ∼−30 mV; half activation at −8 mV, full activation at ∼+30 mV. The slope factor was 6.6 mV.
For the steady state inactivation curves, the relations also followed Boltzmann distributions well. For INa (Fig. 7,A), half inactivation occurred at −59 mV, and the slope was 8.7 mV. For ICa (Fig. 7 B), half inactivation occurred at −34 mV, and the slope was 5.4 mV. Therefore, at the usual resting potential of ∼−50 mV, 24% of INa and 95% of ICa are available.
For both INa and ICa, there was a small overlap of the activation and inactivation relations (window current), with that for INa peaking to ∼5% of the maximum current at −35 mV, and that for ICa peaking to 10% of the maximum at −25 mV.
Recovery from Inactivation
Because the spontaneous electrical activity of the myometrium consists typically of bursts of action potentials that lead to contractions (see Kao, 1989), the influence of an action potential on those following it can be important. Fig. 8 shows the recovery of INa from inactivation. The data are closely clustered even though the myocytes came from different stages of pregnancy. They are also well fitted by a single exponential curve with a time constant of 20 ms.
Fig. 9 shows the recovery of ICa from inactivation. The recovery is best fitted by two exponential components; the smaller component (12%) has a τ of 27 ms, and the larger component (86%) has a τ of 374 ms.
Viewed differently, for INa, 50% recovery is attained in <15 ms and 80% recovery (possibly needed to generate propagated action potentials) is attained in ∼30 ms. For ICa, 50% recovery is attained in ∼200 ms, and 80% recovery in ∼600 ms. Since the rates of recovery from inactivation are important determinants of the rates of repetitive action potentials, the widely different reactivation of the two types of channels must affect their respective contributions to myometrial excitability.
Dissociated Smooth Myocytes as Physiological Models
The tight-seal patch-clamp method and improved cell-isolation procedures have fostered a surge of recent studies on dissociated single smooth muscle cells, which are unencumbered by complex intercellular connections, ion accumulation in extracellular clefts, and insurmountable cleft resistance encountered in multicellular preparations. Although the procedures of enzyme-aided dissociation are conceded as potentially injurious (Bolton et al., 1985; Sanders, 1989), the suitability of such myocytes as models has rarely been tested. The pregnant rat myometrium provides an opportunity for a critical scrutiny of this issue, because it possesses some unique tissue-specific properties that can serve as markers, and because it has been studied in a wide spectrum of organization levels and in tissue culture. The small multicellular preparations, against which dissociated myocytes can be compared, were subjected only to being dissected free from the uterus but not to any enzymes or artificial intracellular milieu. Hence, their ionic channel functions are likely to be as nearly physiological as any isolated tissue or cells can be.
Using new information obtained on dissociated uterine myocytes (specific capacitance, 1.42 μF/cm2; surface:volume ratio, 0.36 μm−1; average cell capacitance, 108 pF; see results), the active region of a previous study (with a total capacitance of 0.13 μF; Kao and McCullough, 1975) can be estimated to contain ∼1,000 (867–1,204) cells. In such preparations, the presence of a Na+ and a Ca2+ component in the inward current had already been shown (also Anderson et al., 1971; Kao, 1978), as had an increasing relative contribution of the Na+ component in late pregnancy as term approached (Nakai and Kao, 1983). Phenotypic expression of a Na+ channel is rare among visceral smooth muscles. Confirmation in dissociated uterine myocytes not only of that rare phenotype but also of its participation in highly tissue-specific functional changes demonstrate that such cells, if properly prepared, do not deviate substantially from physiological norms.
Because of such similarities, the freshly dissociated uterine myocytes can also serve as a basis for assessing some tissue-cultured preparations as physiological models (e.g., Mollard et al., 1986; Amedee et al., 1987; Toro et al., 1990; Rendt et al., 1992). Such a comparison reveals so many differences as to suggest that the observed channel functions pertain to different cells.
Some of the key differences are: (a) most fresh myocytes used in this work had a readily recordable INa, which could be elicited from holding potentials of −90 to −60 mV (also Ohya and Sperelakis, 1989). It was fully blocked by 1 μM TTX (Fig. 3,B), which had a Kd of 27 nM (Ohya and Sperelakis, 1989), characterizing the TTX receptor as a high affinity type (references in Kao and Levinson, 1986). In cultured cells, with the exception of the human uterine myocyte (Young and Herndon-Smith, 1991), none showed any Na+ action potentials (Mollard et al., 1986) or any INa, all the inward current being attributed to ICa (Amedee et al., 1987; Rendt et al., 1992). Some cultured cells had no inward currents at all (Toro et al., 1990). In cultured cells, INa could be elicited only after interference with Na+ inactivation; when elicited, the Kd of TTX blockade was 2 μM (Amedee et al., 1986), characterizing the receptor as a low affinity type. The different TTX sensitivities could reflect different amino acid compositions of the channel molecules. (b) In fresh (Fig. 4,C), but not cultured (Rendt et al., 1992) nonpregnant myocyte, ICa was susceptible to blockade by dihydropyridines. (c) In fresh myocytes, half activation of the Ca2+ conductance in 3 mM of Ca2+ was at −8 mV (Fig. 7,B). In cultured cells, half activation in 10 mM Ca2+ was at −14 mV (Amedee et al., 1987), or at −7 mV (Rendt et al., 1992), in spite of a known screening effect of 10 mM Ca2+ on surface negative charges (Frankenhauser and Hodgkin, 1957; Yamamoto et al., 1989; Sui and Kao, 1997; and Fig. 4 A), which should have caused a significant positive displacement. (d) In fresh myocytes, the steady state ICa inactivation was almost all voltage dependent, with half inactivation at −34 mV. In cultured cells, 10–12% of the ICa did not inactivate, and the remainder had a half inactivation voltage of −17 (Amedee et al., 1987) or −55 (Rendt et al., 1992) mV.
Hence, the cultured myometrial cells described so far possess none of the unique cell-specific functional properties of the freshly dissociated uterine myocytes, and therefore are less likely to provide physiologically relevant information on ionic channel functions of uterine myocytes.
Charge Carriers and Channels for Inward Currents
The density of Na+ channels in uterine myocytes (3–5 μA/cm2) is two to three orders of magnitude lower than those in peripheral nerves, skeletal, or cardiac muscles. Kinetic data suggest the presence of a single class of Na+ channels in uterine myocytes.
The Ca2+ channels in rat uterine myocytes are mostly of the L-type (also Ohya and Sperelakis, 1989; Miyoshi et al., 1991), resembling the case in other visceral smooth myocytes (urinary bladder, Klöckner and Isenberg, 1985; ureter, Sui and Kao, 1997; taenia coli, Yamamoto et al., 1989). However, in human uterine myocytes, T-type Ca2+ channels have been described (Inoue et al., 1990; Young et al., 1993). Inactivation of ICa involves both voltage-dependent and Ca2+-mediated mechanisms. A voltage-dependent mechanism is demonstrated in the steady state voltage-inactivation relation, whereas a Ca2+-mediated mechanism is seen in the U-shaped relation between τf of inactivation and ICa (Fig. 6,D), in a slowing of the rate when Ba2+ replaced Ca2+ (Fig. 4,B), and in an increased rate when ICa was enhanced by BAY-K 8644 (Fig. 4 D).
The density of ICa in uterine myocytes (3–11 μA/cm2) is approximately midway in a spectrum among different visceral smooth myocytes (urinary bladder, 20 μA/cm2, Klöckner and Isenberg, 1985; taenia coli, 20 μA/cm2, Yamamoto et al., 1989; ureter, 3 μA/cm2, Sui and Kao, 1997). In taenia coli myocytes (Yamamoto et al., 1989) and urinary bladder myocytes (Sui et al., 1993), Ca2+ influx during an action potential is capable of discharging the membrane capacity and raising [Ca2+]i to 8–13 μM. Thus, the influx is potentially adequate for initiating various physiological functions. In the non- and early-pregnant uterine myocytes, the situation is similar and the same conclusion may apply. In the late-pregnant myocyte, the situation is more complex. Although the cell capacitance is larger, the combined influx of Na+ and Ca2+ may still discharge it to initiate the action potential. However, the cell volume is much larger while the density of ICa is lower (Tables II and III). Under such circumstances, whether influx of Ca2+ alone is adequate to supply the needed Ca2+ or whether internal sources become more important are problems that require further study.
Changing Densities of INa and ICa in Pregnancy and Their Implications
Using small multicellular preparations from 16–21-d pregnant rat uterus, Nakai and Kao (1983) described a changing ratio of the Na+ and Ca2+ components of the inward current, such that the ratio of peak INa/peak ICa increased as term approached. Those observations are now confirmed on single uterine myocytes, and extend to cover the entire pregnancy.
Although Inoue and Sperelakis (1991) have made a similar confirmation, they found no INa in any day-5 myocyte, and constant densities of INa and ICa in individual myocytes from day 9 to term (their Fig. 5,A). They also found that the fraction of myocytes expressing INa increased with the progression of pregnancy (their Fig. 5,B). By interpreting the frequency of their observing INa as a probability of INa occurrence, and by averaging data from all cells, including all day-5 cells with no INa, they concluded that the density of INa increased towards term (their Fig. 5 C).
Our observations are quite different. We found INa in myocytes from all stages of pregnancy, from day 2 to term, and also in those from nonpregnant uteri under estrogen stimulation. Although only half of the day-2 myocytes had recordable INa, the finding of any INa indicates that conditions for the phenotypic expression of Na+ channels were already in place. Whereas a specific ionic current indicates both the presence and the expression of that channel, its absence only indicates the nonexpression of the channel and not its nonexistence. Because of this consideration, and because of our belief that population phenomena should be supported by more inclusive random sampling than can be provided by the usual limited sampling of whole-cell patch-clamp studies, we focused our interest on the properties of individual myocytes and compared only those myocytes in which both INa and ICa were recorded. Our data show that, in individual myocytes, densities of both INa and ICa change during the course of pregnancy.
Two broad categories of regulation by ovarian hormones can account for the observed changes: genomic and nongenomic influences. Considering the fourfold increase of the surface area in the hypertrophied uterine myocyte during pregnancy, new channel proteins must be synthesized at a rate exceeding that needed for replacement, because the densities of INa and ICa are always more than a quarter of the original level in the nonpregnant myocyte. In the simplest sense, these increased current densities might be attributed to estrogen-enhanced transcription, leading to more copies of Na+ and Ca2+ channels (genomic influence). However, the regulation may be more complex, and may include significant contributions from nongenomic influences. For Na+ channels, it involves bringing forth a previously unexpressed phenotype (as in the estrus–diestrus phase of nonpregnant uteri), and then rapidly extinguishing it (as in postpartum uteri). The decline in the densities of INa and ICa in midpregnancy may involve progesterone, which is known to rise at that time and has an antiestrogen effect on the expression of a K+ channel (Yang et al., 1994). Whatever the regulatory mechanism might be, the observed changes are probably physiological, because the envelope of their effects has already been observed in multicellular preparations (Nakai and Kao, 1983) that contained enough individual myocytes to provide a reasonably inclusive random sampling.
The increased role of INa in late pregnancy subserves well the physiological functions of the parturient uterus. The more intense sodium currents could contribute to a faster spread of the electrical impulse throughout the parturient uterus. More importantly, the faster inactivation and reactivation of INa would permit more frequent repetitive spike discharges than would be possible with the slower ICa. Furthermore, because of a concomitant “de-expression” of some large-conductance Ca2+-activated K+ channels (Kao et al., 1989; and Wang, S.Y., M. Yoshino, and C.Y. Yao, manuscript submitted for publication), the membrane conductance is lowered, and the membrane potential less negative than otherwise. Perhaps, through a combination of these processes, the excitability of the parturient uterus is enhanced and coordinated across large areas of the whole organ to facilitate contraction throughout.
This work was supported in part by grants from the National Institutes of Health (HD-00378 and DK-39371).
Abbreviations used in this paper: I-V, current–voltage; TEA, tetraethylammonium chloride; TTX, tetrodotoxin.
Address correspondence to Dr. C.Y. Kao, Department of Pharmacology (Box 29), SUNY Health Science Center, 450 Clarkson Ave., Brooklyn, NY 11203. Fax: 718-270-3309.