Mechanisms Underlying Greater Sensitivity of Neonatal Cardiac Muscle to Volatile Anesthetics

All procedures involving animal use were approved by the Institutional Animal Care and Use Committee of the Mayo Clinic (Rochester, MN). Hearts from 3–4 Sprague-Dawley pups (neonates, < 3 days postpartum; mean body weight, 7 ± 2 g) were combined for one experiment (total 50 dissociations). A total of 42 male rats (at least 12 weeks of age; 250–300 g body weight) formed the adult group.

Animals were anesthetized with intramuscular ketamine (60 mg/kg) and xylazine (2.5 mg/kg). The hearts were excised and rapidly immersed in oxygenated (95% O², 5% CO²) ice-cold Krebs solution (118.3 mm NaCl, 4.7 mm KCl, 1.2 mm KH²PO⁴, 1.2 mm MgSO⁴, 1.0 mm CaCl², 11 mm glucose; pH 7.40).

Adult cardiac myocytes were enzymatically dissociated using a Langendorf perfusion-based technique. 18The heart was sequentially perfused with Dulbecco minimum essential medium (Joklik modification with 10 mm KCl and 10 mm HEPES; Sigma Chemicals, St. Louis, MO) containing 1 mm Ca²⁺, 0 Ca²⁺, and 80 mg/ml Type I collagenase (Worthington Chemicals, Lakewood, NJ) with 1% bovine serum albumin (Sigma) in 0 Ca²⁺. During collagenase perfusion, Ca²⁺was added back in graded steps to 1 mm. The tissue was gently triturated to obtain single myocytes, which were finally washed in Joklik medium with 1% albumin and 1 mm Ca²⁺.

In pilot studies on 3–4 animals, we found that aortic cannulation in neonates was technically difficult and unreliable. In further studies on 3–4 animals, we used a commercially available neonatal myocyte isolation kit (Worthington), compared resting [Ca²⁺]ⁱand responses to 5 mm caffeine in cells isolated using the two techniques, and found the responses to be less than 10% different. Accordingly, we used the isolation kit for dissociation of neonatal myocytes for the remainder of the study. Ventricles from 3–4 animals were placed directly in ice-cold Ca²⁺and Mg²⁺free Hanks balanced salt solution (Sigma), minced with scissors, washed, and incubated overnight at 4°C with 50 μg/ml trypsin. A soybean trypsin inhibitor was then added, and the tissue was oxygenated for 1 min. Subsequently, the samples were warmed to 37°C and incubated further for 1 h with 50 mg/ml collagenase. The cells were finally washed and processed in the same manner as adult myocytes.

Myocytes were plated on laminin-coated glass coverslips and kept at 37°C for a maximum of 5 h, within which time daily experiments were completed. Myocyte viability and Ca²⁺tolerance were evaluated as described previously. 18Before data analyses, the following exclusion criteria for cells were applied: (1) greater than 50 nm variation in resting [Ca²⁺]ⁱat the initiation of a protocol; (2) greater than five spontaneous [Ca²⁺]ⁱtransients per minute (noted in ∼15% of cells); (3) greater than 10% change in resting [Ca²⁺]ⁱduring intervening washes; and (4) greater than 10% dye bleaching over a 5-min period of laser exposure.

We previously published detailed descriptions of the real-time confocal imaging techniques for cardiac myocytes. 18Myocytes incubated in 5 μm fluo-3/AM (Molecular Probes, Eugene, OR) were imaged using a Noran Odyssey XL real-time confocal system (Noran Instruments, Middleton, WI) on a Nikon Diaphot inverted microscope (Nikon Instruments USA, New York, NY). Normal Tyrode medium containing 145 mm NaCl, 4 mm KCl, 1 mm MgCl², 1 mm CaCl², 10 mm glucose, and 10 mm HEPES (pH 7.4; 25°C) was used initially. A fiberoptic fluid-level controller (ALA Instruments, Westbury, NY) was used to minimize the influence of the fluid level on NCX activation, threshold for electrical stimulation, and agonist concentration. Images (640 × 480 pixels) were acquired at 30 frames/s at 400× magnification (Olympus 40X/1.3 oil-immersion lens; Olympus USA, Los Angeles, CA) and an optical section thickness of 1 μm. The [Ca²⁺]ⁱresponses were obtained using individual software-defined regions of interest for two to three cells.

Empirical calibration of fluo-3 fluorescence levels for [Ca²⁺]ⁱwas performed using previously described techniques, 18based on sequential exposure to known Ca²⁺concentrations (0 nm to 1.25 μm; Molecular Probes Calcium Calibration Buffer Kit; Molecular Probes, Eugene, OR), and 10 μm A-23187 (Ca²⁺ionophore) to allow equilibration of [CA²⁺]ⁱand extracellular Ca²⁺concentration ([Ca²⁺]^o). An empirical calibration curve was then generated and compared with measurements by other investigators using the following equation:

to calculate [Ca²⁺]ⁱfrom fluorescence values (F), where F^minis the fluorescence at minimal [Ca²⁺]ⁱ(0 nm in this study) and F^maxat saturating concentrations, determined using a buffer and ionophore technique as above. 19Using the F^min, F^max(expressed as gray levels ranging from 0 to 254), we estimated the apparent K^dfor fluo-3 to be 522 ± 101 nm in adults and 504 ± 89 nm in neonates. Although these values are somewhat higher than the 400 nm used in previous studies, 19it must also be recognized that the K^dfor fluo-3 can differ in vivo  versus  in vitro . 20 

As in previous studies, 18a calibrated online vaporizer was used to add halothane (Wyeth-Ayerst Laboratories, Philadelphia, PA) and sevoflurane (Abbott Laboratories, Deerfield, IL) to the aerating gas mixture. Aqueous anesthetic concentrations equivalent to 1 and 2 adult rat minimum alveolar concentration (MAC) at room temperature (25°C) were measured for halothane and were determined by gas chromatography and an electron capture detector (Hewlett-Packard 5880A; Hewlett-Packard, Sunnyvale, CA) 21and for sevoflurane using a flame ionization detector. Halothane concentrations were 0.32 ± 0.11 mm for 1 MAC and 0.50 ± 0.11 mm for 2 MAC, and sevoflurane concentrations were 0.48 ± 0.10 mm and 0.70 ± 0.11 mm for 1 and 2 MAC, respectively.

Myocytes perfused with Joklik medium containing 1 mm Ca²⁺were electrically stimulated at 0.25 Hz, 5 ms using two platinum wires within the microscope chamber. Stimulation was maintained for 2 min to ensure [Ca²⁺]ⁱstability. However, to minimize dye bleaching, recordings were made only every 30 s for two to three stimulations. Cells were subsequently perfused with Joklik medium containing 1 or 2 MAC volatile anesthetic, and the [Ca²⁺]ⁱresponses were recorded for 3 min, after which recovery from anesthetic was evaluated for 1 min. In control experiments, cells were electrically stimulated for 5 min with no anesthetic exposure.

Electrically stimulated myocytes were exposed for 3 min to 1 μm BayK8644 (Sigma), a selective activator of L-type Ca²⁺influx channels. In the presence of BayK8644, cells were perfused with anesthetics for an additional 2 min, followed by washout for 1 min. Control cells were only exposed to BayK8644 for 5 min.

In a second set of experiments, myocytes were preexposed to 10 μm ryanodine (Sigma), 5 μm thapsigargin (Sigma), and 10 μm KBR 7943 (Tocris Corp., Ellisville, MO). The combination of ryanodine and thapsigargin ensured inhibition of SR Ca²⁺release as well as reuptake, thus functionally isolating the plasma membrane. KBR 7943 is a novel drug that has been reported to inhibit NCX. 22During these conditions, the contribution of Ca²⁺influx channels to the [Ca²⁺]ⁱresponse to electrical stimulation could be determined. Myocytes were electrically stimulated for an additional 2 min, exposed to anesthetics for 2 min, and washed for 1 min. Control cells were stimulated for 5 min.

In a third set of experiments, myocytes preexposed to ryanodine, thapsigargin, and KBR 7943 were exposed to 1 μm BayK8644 for 2 min. Lastly, cells were perfused with anesthetics for an additional 2 min, and their response to electrical stimulation was recorded.

Myocytes were exposed to 5 mm caffeine in the presence of 2 mm extracellular Ca²⁺, and [Ca²⁺]ⁱresponses were recorded. Cells were washed for 10 min to replenish SR Ca²⁺stores and then reexposed to caffeine in the presence of anesthetic. The length of time for SR refilling was kept constant across all experiments.

In a second set of experiments, myocytes were preexposed to zero extracellular Ca²⁺, low (4 mm) Na⁺, 1 mm lanthanum chloride (La³⁺), and 10 μm KBR 7943. This combination functionally isolated the SR by blocking Ca²⁺influx channels, NCX, and Ca²⁺efflux via  the plasma membrane Ca²⁺adenosine triphosphatase (PMCA; inhibited by La³⁺). Cells were preexposed for 5 min to 10 μm cyclopiazonic acid (CPA; Sigma), a selective and reversible inhibitor of SERCA. Subsequently, SR Ca²⁺release was induced by 5 mm caffeine. The rise time of the [Ca²⁺]ⁱresponse (normalized for amplitude) was taken to represent the rate of SR Ca²⁺release, and the amplitude as an index of SR Ca²⁺content. Cells were washed for 15 min to allow SR replenishment and were then reexposed to CPA in the absence (control) or presence of volatile anesthetics for 5 min, followed by caffeine in the continued presence of CPA and anesthetic.

Myocytes preexposed to zero extracellular Ca²⁺, low Na⁺, La³⁺, and KBR 7943 were exposed to 5 mm caffeine. The fall time of the [Ca²⁺]ⁱresponse was taken to represent the rate of Ca²⁺reuptake via  SERCA. The contribution of Ca²⁺efflux via  PMCA and NCX was assumed to be minimal in the presence of La³⁺and low extracellular Na⁺. The cells were then washed for 15 min, and the protocol was repeated in the absence (control) or presence of volatile anesthetics.

In a separate set of animals, cardiac ventricles were carefully separated, washed, stretched and pinned on cork, and rapidly frozen in melting isopentane cooled by liquid nitrogen. Cross-sections (10 μm) cut using a cryostat (Reichert-Jung Model 2000E, Germany; −20°C) were placed on slides and incubated for 10 min in 1 μm Bodipy-tetramethylrhodamine conjugated ryanodine (Molecular Probes) in 0.1 m phosphate buffer to stain the SR. Sections were then fixed in 2% paraformaldehyde in phosphate buffer, washed, and coverslipped. Sections for staining control were directly fixed without exposure to ryanodine. The samples were visualized at 400× using a BioRad MRC500 laser confocal microscope (BioRad Instruments, Hercules, CA) equipped with an Ar-Kr laser (568-nm line for excitation of rhodamine, 590-nm emission filter), at a resolution of 0.3 μm/pixel. Areas of fluorescence above a background threshold (determined from control samples) were considered to be areas of SR. This measured area was expressed as a fraction of myocyte area. At least 20 myocytes were sampled from each animal.

The technique for evaluating the effect of volatile anesthetics on influx mode of NCX was recently published 18and is also illustrated in figure 1. Cells with stable, resting [Ca²⁺]ⁱwere “Na-loaded” with Tyrode solution containing 0 Ca²⁺(5 mm EGTA) and normal Na⁺, 5 μm CPA and 10 μm ryanodine (Sigma), a specific RyR channel inhibitor, thus functionally isolating the plasma membrane. Because the major mechanisms of Ca²⁺influx across the plasma membrane were blocked by 0 Ca²⁺, namely, Ca²⁺influx channels and NCX, [Ca²⁺]ⁱwas expected to remain stable. After Na⁺loading cells for 1 min, perfusion was rapidly switched (< 300 ms) to Tyrode solution containing 0 Na⁺, normal Ca²⁺, CPA, and ryanodine, selectively activating the influx mode of NCX, reflected by the rate of increase of [Ca²⁺]ⁱ. After 1 min, both CPA and ryanodine were washed out with normal Tyrode solution. No attempt was made to determine whether ryanodine was actually washed out because it was not expected to adversely affect subsequent manipulations. The protocol was then repeated in the absence of (control) or with preexposure to volatile anesthetic before rapid NCX activation.

To examine the effect of volatile anesthetics on the efflux mode of NCX (fig. 1), cells initially perfused with normal Tyrode solution were switched to solution containing 0 Na⁺, 0 Ca²⁺(with 5 mm EGTA), and 5 μm CPA. Thus, the major mechanisms for decreasing [Ca²⁺]ⁱ(SERCA and NCX) as well as Ca²⁺influx were inhibited. Accordingly, [Ca²⁺]ⁱwas expected to increase because of continued SR Ca²⁺“leak.” A potential mechanism for Ca²⁺extrusion not inhibited was the PMCA. However, in our experience the contribution of PMCA to [Ca²⁺]ⁱduring this protocol is fairly insignificant and can be ignored (see Discussion). After [Ca²⁺]ⁱwas allowed to increase for approximately 1–2 min (and more or less stabilized), efflux mode NCX was rapidly (< 300 ms) and selectively activated by Tyrode solution containing 0 Ca²⁺(with EGTA), normal Na⁺, and CPA. The rapid rate of decrease in [Ca²⁺]ⁱwas recorded as an index of NCX efflux rate. When [Ca²⁺]ⁱreached baseline, the CPA was washed out for 5 min, allowing SR replenishment. The protocol was then repeated in the presence or absence of 1 or 2 MAC volatile anesthetic in the 0 Na⁺–0 Ca²⁺and 0 Ca²⁺–normal Na⁺solutions.

The effect of volatile anesthetics on the relation between Na⁺gradient and influx mode of NCX was examined by Na⁺-loading cells with Tyrode solution containing 0 Ca²⁺, CPA, ryanodine, and one of three Na⁺concentrations (35, 70, or 145 mm). The same cell was used for all three concentrations. Cells were exposed to volatile anesthetics as in the previous protocol. The effect of volatile anesthetics on the relation between Na⁺concentration and efflux mode of NCX was examined by introducing Tyrode solution containing 0 Ca²⁺and one of three different concentrations of Na⁺(35, 70, or 145 mm) during rapid activation of NCX in the presence or absence of anesthetic. Cell viability was tested at the end of the protocol by repeating the protocol with 145 mm Na⁺.

At least 15 cells were analyzed for each protocol. These cells were obtained from at least five animals. Data from each animal were averaged, and statistical comparisons were made across animals (repeated measures). The specific numbers of animals analyzed in each protocol are provided in the results. Within an age group, there were no significant differences between vehicle control groups for either anesthetic. Accordingly, the data from these control groups were pooled. Data were compared using analysis of variance with repeated measures with age, experimental condition, anesthetic, and concentration as grouping variables. Multiple comparisons were performed using Bonferroni and Scheffépost hoc  analysis. These post hoc  analyses were performed across age and experimental condition. P < 0.05 was considered significant (two-tailed). Significant differences between 1 and 2 MAC anesthetic are not indicated. All data are expressed as mean ± SD.

Neonatal and adult cardiac myocytes did not differ in basal [Ca²⁺]ⁱ(95 ± 8 nm vs.  107 ± 11 nm, respectively). There were no significant differences between age groups, anesthetics, or anesthetic concentration in basal [Ca²⁺]ⁱ. The amplitude of the [Ca²⁺]ⁱresponses to electrical stimulation in neonatal myocytes (389 ± 49 nm) was significantly smaller compared with adults (891 ± 95 nm;P < 0.05). Both anesthetics significantly decreased the amplitude of the [Ca²⁺]ⁱresponse, especially in neonates (fig. 2;P < 0.05; n = 5), but sevoflurane had a smaller effect compared with halothane.

BayK8644 potentiated [Ca²⁺]ⁱresponses to electrical stimulation to a greater extent in neonates (figs. 3A and B;P < 0.05; n = 6). Inhibition of the SR and NCX by exposure to ryanodine, thapsigargin, and KBR 7943 decreased the amplitude of the [Ca²⁺]ⁱresponse to electrical stimulation, but to a lesser extent in neonates (fig. 4; n = 5). In the absence of SR and NCX, BayK8644 significantly increased [Ca²⁺]ⁱresponse to electrical stimulation in neonatal myocytes compared with adults (figs. 5A and B;P < 0.05; n = 5 for each age). These data highlight the role of Ca²⁺influx channels in neonates. Subsequent exposure to anesthetic still resulted in significantly greater decrement in the amplitude of the response in neonates compared with adults (figs. 3A and 4;P < 0.05; n = 6). However, at either age, BayK8644 blunted the inhibitory effect of anesthetic. The effects of sevoflurane were generally smaller than those of halothane.

The transient [Ca²⁺]ⁱresponse in neonatal myocytes (n = 11) was smaller and slower than that of adult myocytes (n = 12;fig. 6, table 1;P < 0.05). Even when corrected for SR volume density, the amplitude of the neonatal response (1,983 ± 193 normalized units) was significantly smaller than that of adults (2,678 ± 339 normalized units;P < 0.05).

Compared with vehicle controls, exposure to anesthetics, especially halothane, decreased the amplitude of the [Ca²⁺]ⁱresponse in both age groups (P < 0.05), but to a significantly lesser extent in neonates compared with adults (P < 0.05;fig. 6and table 2). Halothane significantly prolonged increase times in both age groups, but less so in neonates (P < 0.05). Compared with the effects on increase time, halothane had considerably greater effects on the decrease time of the [Ca²⁺]ⁱresponse to caffeine, with the effects being several-fold smaller in neonates compared with adults. In general, sevoflurane had minimal to insignificant effects on the [Ca²⁺]ⁱresponses in neonates.

Exposure to zero extracellular Ca²⁺, low Na⁺, La³⁺, CPA, and KBR 7943 did not significantly elevate basal [Ca²⁺]ⁱ. Subsequent exposure to caffeine induced a brisk elevation in [Ca²⁺]ⁱ(more so in adults) that was more or less sustained. The amplitude of the [Ca²⁺]ⁱresponse was smaller in neonates (P < 0.05). Exposure to anesthetics significantly slowed the increase time and blunted the amplitude of the [Ca²⁺]ⁱresponse to caffeine (P < 0.05;figs. 7A and B, respectively), but to a greater extent in adults.

During conditions of zero extracellular Ca²⁺, low Na⁺, La³⁺, and KBR 7943 (no CPA), the response to caffeine was qualitatively similar to that in figure 6. The decrease time, representing SERCA activity, was significantly slower in neonates (P < 0.05). Preexposure to anesthetic significantly slowed the decrease time of subsequent caffeine responses, especially in the adult (fig. 7C).

Cross-sections stained for RyR channels displayed a relatively less-organized SR in neonate compared with adult cardiac muscle, as expected. Accordingly, the SR volume density of the neonatal muscle, estimated using either cross-sections or single cells, was significantly less than that of adult muscle (6.3 ± 2.3%vs.  28.4 ± 3.7% of the cell area;P < 0.05).

Na⁺loading did not significantly change resting [Ca²⁺]ⁱ. Rapid reintroduction of [Ca²⁺]^o, with simultaneous removal of extracellular Na⁺concentration ([Na⁺]^o), resulted in a rapid and monotonic increase in [Ca²⁺]ⁱ. Introduction of normal Tyrode solution resulted in a plateau. The influx rate of NCX (calculated from the slope of the steepest portion of the ascending curve) ranged from 74.4 to 132.3 nm/s in neonates compared with 32.7 to 78.3 nm/s in adults (pooled control data; n = 10 for each age). Both halothane and sevoflurane slowed NCX-mediated Ca²⁺influx to a greater extent in neonatal myocytes (fig. 8;P < 0.05; n = 5 for each concentration and age). In adults, slowing of influx was comparable between the two anesthetics.

Simultaneous inhibition of SERCA and NCX resulted in a rapid increase in [Ca²⁺]ⁱfollowed by a slow decline (most likely because of the PMCA, which was not specifically inhibited in this protocol). Rapid reintroduction of [Na⁺]^oresulted in a steep decrease in [Ca²⁺]ⁱ. The efflux rate of NCX (calculated from the steepest portion of the descending curve) ranged from 55.4 to 95.2 nm/s in neonates versus  18.5 to 36.2 nm/s in adults (pooled control data; n = 10). In the presence of anesthetic, reactivation of NCX resulted in significantly slower efflux in neonatal myocytes, both in absolute terms (fig. 9A;P < 0.05; n = 5) and when normalized for “peak”[Ca²⁺]ⁱ(fig. 9B;P < 0.05 18). Compared with the neonate, anesthetic effects on efflux in adults were smaller (figs. 9C and D;P < 0.05). Overall, sevoflurane had a smaller effect on efflux rates in both age groups (P < 0.05).

Because the time for Na⁺loading was fixed, the extent of Na⁺loading was assumed to be proportional to [Na⁺]^o. The actual Na⁺concentration in the cell was not determined. Decreasing Na⁺loading resulted in comparable slowing of influx in both neonates and adults (fig. 10;P < 0.05). Exposure to halothane resulted in a further, concentration-dependent slowing of NCX-mediated Ca²⁺influx, especially in neonates (figs. 10A and B;P < 0.05; n = 5), such that 2 MAC halothane completely inhibited influx at 35 mm Na⁺. Sevoflurane produced less slowing of influx compared with halothane, at least in neonates (figs. 10C and D;P < 0.05 for sevoflurane effects only; n = 5).

Decreasing [Na⁺]^oslowed NCX-mediated efflux to a greater extent in neonates (fig. 11;P < 0.05). The confounding influence of different “peak”[Ca²⁺]ⁱvalues on efflux rate was avoided by making measurements only in cells displaying comparable “peak”[Ca²⁺]ⁱvalues. Halothane further slowed efflux at each Na⁺concentration, especially in neonates (figs. 11A and B;P < 0.05; n = 5). Overall, sevoflurane produced less slowing of NCX-mediated efflux in both age groups for every Na⁺concentration (figs. 11C and D;P < 0.05; n = 5).

The current study demonstrates that volatile anesthetics decrease [Ca²⁺]ⁱto a greater extent in neonates by greater inhibition of Ca²⁺channels and NCX, rather than inhibition of SR Ca²⁺release and reuptake. Given a greater role for Ca²⁺influx channels and NCX in [Ca²⁺]ⁱregulation of the neonatal heart, inhibition of these mechanisms by volatile anesthetics is likely to have a greater impact on cardiac function at that age.

A potential concern with separate techniques for dissociating cardiac myocytes (Langendorf vs.  trypsin-based mechanical dissociation) is whether the techniques themselves explain some of the age-related differences in volatile anesthetic effects. However, given the lack of difference in basal [Ca²⁺]ⁱand the response to caffeine, it is unlikely that these issues, even if present, were relevant in the current study. “Freshly” dissociated neonatal cardiac myocytes express only 25% of the total Ca²⁺current density of that in “freshly” dissociated adult myocytes. 8Our data in 1-day-old neonatal myocytes are consistent and suggest that the dissociation technique alone cannot account for the observed age-related differences in [Ca²⁺]ⁱregulation.

Although fluo-3 is a nonratiometric [Ca²⁺]ⁱindicator, in several studies 18,19we found relatively small variations in basal [Ca²⁺]ⁱacross myocytes from different coverslips and animals, underlining the reliability of the empirical Ca²⁺calibration technique using this dye. A potential limitation of fluo-3, especially relating to adult cardiac myocytes, is its relative lack of sensitivity at extremely high [Ca²⁺]ⁱvalues. Although it is possible that peak [Ca²⁺]ⁱresponses may have been underestimated, the results and conclusions of the study are unlikely to be affected because comparisons were made within the same cell before and during anesthetic exposure.

Pharmacologic techniques rather than electrical stimulation were used to inhibit or activate NCX during defined conditions of [Ca²⁺]ⁱ, [Ca²⁺]^o, [Na⁺]ⁱ, and [Na⁺]^obecause several studies, including the current one, have demonstrated that anesthetics inhibit both Ca²⁺channels 12,14,23and SR Ca²⁺release–reuptake, 15–17which would have influenced the [Ca²⁺]ⁱresponse to electrical stimulation. Furthermore, to isolate anesthetic effects on NCX using electrical stimulation, agents such as nifedipine or ryanodine would only have abolished the [Ca²⁺]ⁱresponses.

In the NCX efflux protocol, we did not specifically inhibit the PMCA, which may have accounted for the slow decline in [Ca²⁺]ⁱobserved even in the absence of [Na⁺]^o. However, the contribution of PMCA to the overall results are unlikely to be significant, because the PMCA has a 10-fold higher affinity for Ca²⁺but a considerably slower turnover rate compared with NCX 24(evidenced by large change in slope of the [Ca²⁺]ⁱprofile after reintroduction of [Na⁺]^o).

The current study used MAC values from adult rats for comparison across age groups. Similar to clinical observations, it is possible that MAC values of halothane and sevoflurane in neonatal rats may be increased compared with adult rat values. Regardless, the fact that the lower adult MAC values used in the current study resulted in greater inhibition of [Ca²⁺]ⁱregulatory mechanisms only highlights greater anesthetic sensitivity of the neonatal myocardium.

Consistent with previous results in adult hearts, 12–14we found that halothane and sevoflurane both inhibit the opening of L-type channels by BayK8644. However, studies with halothane and isoflurane have found no differences in potency in inhibition of Ca²⁺influx. We found that, unlike isoflurane, which has effects comparable to halothane, 12sevoflurane had a much smaller impact on Ca²⁺influx in adult cardiac myocytes compared with halothane. Differences in relative anesthetic potency may be related to individual lipid solubilities and potential changes in channel protein configuration. Furthermore, volatile anesthetics may differ in their relative impact on other mechanisms that regulate Ca²⁺influx channels, such as cyclic nucleotides and protein kinases.

Inasmuch as Ca²⁺influx plays a greater role in [Ca²⁺]ⁱregulation of the neonatal heart, this is a potential mechanism underlying greater myocardial depression produced by volatile anesthetics. Inhibition of Ca²⁺influx channels can reduce [Ca²⁺]ⁱdirectly by reducing the contribution to total [Ca²⁺]ⁱand indirectly by preventing Ca²⁺-induced Ca²⁺release (CICR). Previous studies have suggested that the contribution of Ca²⁺influx per se  to total [Ca²⁺]ⁱin the neonatal heart may not be substantial, because Ca²⁺channel protein and peak current densities are actually lower. 9,11,25,26However, the surface area of the neonatal myocyte is also smaller than that of the adult. Accordingly, even though absolute Ca²⁺current may be small, it may have a significant contribution to total [Ca²⁺]ⁱ. Regardless, inhibition of Ca²⁺influx by volatile anesthetics will decrease its contribution to total [Ca²⁺]ⁱand may further interfere with CICR, leading to the same overall effect, i.e. , greater reduction in [Ca²⁺]ⁱ.

Previous studies have demonstrated that the neonatal SR is functionally immature. 27In accordance, the current study found that, compared with the adult, neonatal cardiac myocytes have significantly smaller SR Ca²⁺stores that are released during activation. During these conditions, volatile anesthetics (halothane more so than sevoflurane) inhibited SR Ca²⁺release to a greater extent in adult cardiac myocytes, rather than neonatal myocytes, indicating that greater sensitivity of neonatal myocardium to anesthetics cannot be explained by greater effects at the level of the SR.

In adult cardiac myocytes, the pronounced effects of halothane on the [Ca²⁺]ⁱresponse to caffeine are consistent with several previous studies. 15–17The general consensus has been that anesthetics increase SR Ca²⁺“leak” through RyR channels, thus depleting SR Ca²⁺. Only a few studies have examined age-related differences in volatile anesthetic effects at the level of the SR and have also interpreted the results as greater SR Ca²⁺leak in the neonate. 1Nonetheless, given the smaller and immature SR, it is still unlikely that increased SR Ca²⁺leak can explain the greater sensitivity of neonatal cardiac myocytes to volatile anesthetics.

Volatile anesthetics have also been shown to decrease SR Ca²⁺content, independent of an effect on RyR channels. 28Our results using halothane are generally consistent with these previous reports. In addition, we found that sevoflurane, chemically closer to isoflurane, produces a smaller effect on the [Ca²⁺]ⁱresponse to caffeine, a finding consistent with the study by Davies et al.  29in adult rat heart. Therefore, it is possible that differences in volatile anesthetic effects on SR Ca²⁺content underlie potency differences for myocardial depression. However, as with SR Ca²⁺leak, the smaller SR makes it unlikely that decreased SR Ca²⁺content can explain greater myocardial depression in neonates.

Whether volatile anesthetics have an effect on SERCA is a particularly relevant issue in the neonatal heart, where relaxation is already limited by the lower rate, number, and efficiency of SERCA protein. 30Some studies have reported a decrease in reuptake, 31whereas others have reported either no effect 32or increased reuptake. 33A potential problem with directly attributing the effects of volatile anesthetics to SERCA is that these effects are critically dependent on both pH and adenosinde triphosphate concentration. Furthermore, even when pH is controlled, 34it is difficult to isolate the effects of NCX from SERCA in decreasing [Ca²⁺]ⁱ. However, in the current study, we inhibited NCX-mediated Ca²⁺efflux by using low extracellular Na⁺and still found that halothane had a profound effect on the decrease time of the [Ca²⁺]ⁱresponse to caffeine, at least in the adult. Accordingly, these data clearly demonstrate that volatile anesthetics inhibit SERCA. The overall effect of this inhibition on [Ca²⁺]ⁱregulation will then depend on the relative role of SERCA and may thus be of less consequence in the neonate.

It is now well-recognized that NCX-mediated Ca²⁺efflux plays a key role in relaxation of the adult heart, 4,5,35albeit with species differences in its relative contribution to [Ca²⁺]ⁱregulation. 36Some studies in the rat 36,37have suggested a role for NCX-mediated Ca²⁺influx in CICR. The results of the current study, as well as those from a recently published study, 18show that NCX-mediated influx during conditions of inhibited Ca²⁺channels and SR, and NCX-mediated efflux during conditions of inhibited SERCA, are both demonstrable in adult rat cardiac myocytes.

In the current study, we found that the rate of NCX-mediated Ca²⁺influx is higher in the neonate. Based on the finding that, at peak action potential, Ca²⁺influx can occur even when Ca²⁺channels are inactivated, Wetzel et al.  10suggested that NCX-mediated Ca²⁺influx likely plays a physiologic role in the neonatal heart. Furthermore, sarcolemmal vesicles isolated from newborn rabbit ventricle have been shown to exhibit greater Na⁺-dependent Ca²⁺uptake 38as well as NCX mRNA 39and protein content 38(determined immunologically) compared with adult animals. Artman et al.  40determined that NCX current density in neonatal rabbit myocytes is greatest perinatally and reaches lower adult values by the third postnatal week. These results support a key role for NCX in neonatal cardiac [Ca²⁺]ⁱregulation. Furthermore, NCX-mediated Ca²⁺efflux may then be key to relaxation of the neonatal heart, evidenced by a higher rate of NCX-mediated Ca²⁺efflux.

Studies have shown that the direction of Ca²⁺flux via  NCX depends on membrane potential itself, [Na⁺]ⁱ, [Na⁺]^o, [Ca²⁺]^o, or [Ca²⁺]ⁱ. 4,41,42We previously demonstrated the relation between [Na⁺]ⁱand rate of NCX-mediated Ca²⁺influx and that between [Na⁺]^oand Ca²⁺efflux in adult rat cardiac myocytes. 18In this study, we found no significant age-related differences in the relation between Ca²⁺influx or efflux and [Na⁺]ⁱand [Na⁺]^o.

Detailed studies on anesthetic and NCX interactions have been limited by the lack of specific inhibitors, as well as concurrent anesthetic effects on other [Ca²⁺]ⁱregulatory mechanisms. Interactions between volatile anesthetics and NCX need to be interpreted with caution. During contraction, NCX inhibition should decrease the inward Ca²⁺current and produce negative inotropy, adding to the decrease in [Ca²⁺]ⁱproduced by inhibition of Ca²⁺influx channels. On the other hand, inhibition of NCX during diastole should decrease the outward Ca²⁺current, delay Ca²⁺efflux, and produce positive inotropy. Accordingly, the net effect of NCX inhibition would depend on several factors: (1) the relative role of NCX in elevating and decreasing [Ca²⁺]ⁱ; (2) the duration of NCX activation during a single cardiac cycle, anesthetic effects on this duration, and thus the overall contribution to total [Ca²⁺]ⁱ; (3) the degree of compensation by other mechanisms involved in [Ca²⁺]ⁱregulation. These issues may need to be addressed in studies examining NCX-mediated Ca²⁺currents during the cardiac cycle using electrophysiologic techniques.

There are no published studies on age-related differences on volatile anesthetic interactions with NCX. In adult rat cardiac myocytes, halothane, isoflurane, and enflurane all completely inhibit NCX. 43In a recently published study on adult rat cardiac myocytes, we found that both halothane and sevoflurane inhibit NCX-mediated Ca²⁺influx as well as efflux. 18In the current study, we found that both anesthetics produce greater inhibition of NCX in neonatal cardiac myocytes. Such an effect may decrease [Ca²⁺]ⁱto a greater extent in neonates that are already more dependent on this mechanism.

In the recently published study on adult rat cardiac myocytes, 18we were the first to report that both halothane and sevoflurane blunt the relation between NCX-mediated Ca²⁺influx and [Na⁺]ⁱ, as well as between NCX-mediated Ca²⁺efflux and [Ca²⁺]ⁱand [Na⁺]^o. In the current study, we found that the extent of blunting of these relations by anesthetics is greater in neonatal myocytes, especially for Ca²⁺influx. These data also support the idea that inhibition of NCX is one mechanism by which anesthetics produce greater myocardial depression in neonates.

The little data on the inhibitory effects of volatile anesthetics on influx mode NCX in cardiac muscle are limited to the adult. 43In adult rat cardiac myocytes exposed to zero [Na⁺]^o, Blanck et al.  44found no effect of halothane on NCX and speculated that NCX sensitivity to volatile anesthetics was reduced during these conditions. On the other hand, in our previous study, we also examined NCX-mediated Ca²⁺influx during conditions of zero [Na⁺]^o18and found that volatile anesthetics have a proportionately greater inhibitory effect on influx mode of NCX compared with efflux. This discrepancy may be related to the fact that our results represent anesthetic effects on the maximum rates of NCX-mediated influx or efflux. It is possible that during the cardiac cycle, combined anesthetic effects on NCX rate as well as on duration of NCX activation (if any) may lead to a different scenario in terms of NCX contribution to total [Ca²⁺]ⁱ.

In the current study, we further demonstrate that NCX-mediated Ca²⁺influx is decreased to a greater extent in neonates by anesthetics. Given the smaller role of the SR and the relatively slow kinetics of Ca²⁺influx channels, NCX-mediated influx may be key to neonatal cardiac contraction. Accordingly, anesthetic inhibition may have a disproportionately greater effect on contractility in the neonate. We also found that anesthetics produce considerable alterations in the relation between [Na⁺]ⁱand NCX-mediated influx, especially in neonates. Previous studies 41,42have shown that [Na⁺]ⁱactivation of NCX-mediated influx is sigmoidal with a half-maximal activation of 15–25 mm, easily achieved with Na⁺loading in our protocols. It is possible that volatile anesthetics interfere with Na⁺binding, which is regulated by [Ca²⁺]ⁱ, 24perhaps to a greater extent in neonates. However, even if comparable between age groups, such inhibition may have greater functional impact in the neonate.

The efflux mode of NCX is arguably more important in cardiac function, especially in the neonate. There are currently very little published data on anesthetic effects on this aspect of NCX. 18We had reported that anesthetics decrease efflux, blunt the normal positive correlations between [Ca²⁺]ⁱand efflux rate, as well as [Na⁺]^oand efflux. 4,42We now demonstrate that volatile anesthetics produce significantly greater rightward shift in the Na⁺dependence of NCX-mediated efflux in neonatal myocytes compared with the adult. In a recent review, Blaustein and Lederer 4speculated that the binding of three Na⁺ions to two binding sites is essential for proper functioning of NCX-mediated Ca²⁺efflux. As with influx, age- and anesthetic-related differences in efflux at different [Na⁺]^ovalues may be related to competitive interference with Na⁺binding. Another potential factor is the energy level of the cell, which is known to influence the Na⁺affinity of NCX. 4 

The overall effect of volatile anesthetics on the [Ca²⁺]ⁱprofile in cardiac myocytes will be determined by the relative contributions of different regulatory mechanisms. If Ca²⁺influx channels and NCX play greater roles in the neonate for elevating [Ca²⁺]ⁱ, their inhibition may well overcome any positive inotropy produced by simultaneous inhibition of efflux mode NCX and SERCA. Volatile anesthetics also affect compensatory neural mechanisms of cardiac control, 45which may be particularly important in neonates, where the sensitivity of the myocardium to neural stimulation may already be lower (e.g. , β-adrenoceptor sensitivity and adenylyl cyclase activities are lower 46).

Compared with humans and some other species such as rabbits, the adult rat heart is more dependent on SR Ca²⁺for contraction, whereas the neonatal rat heart is clearly dependent on plasma membrane Ca²⁺fluxes. The relative roles of these mechanisms in neonates of other species are still under investigation. It is possible that even with comparable quantitative effects on Ca²⁺regulatory mechanisms, age-related differences in anesthetic sensitivity may arise purely from the fact that different mechanisms are relatively more important. However, we demonstrated that anesthetic effects, especially on plasma membrane Ca²⁺fluxes, are both qualitatively and quantitatively different between neonates and adults. Accordingly, even in species such as humans, where plasma membrane Ca²⁺fluxes are more important, age-related differences in anesthetic sensitivity may still arise from greater inhibition of L-type Ca²⁺channels and NCX, consistent with our hypothesis.