Removal of sialic acid alters both T- and L-type calcium currents in cardiac myocytes BERNARD Department
FERMINI
AND
RICHARD
D. NATHAN
of Physiology, Texas Tech University Health Sciences Center, Lubbock, Texas 79430
FERMINI, BERNARD, AND RICHARD D. NATHAN. Removal of sialic acid alters both T- and L-type calcium currents in cardiac myocytes. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): H735H743, 1991.-The whole cell configuration of the patch-clamp technique was used to test the hypothesis that the presence of sialic acid residues influences both T- and L-type Ca2+ currents (Ica,T and 1c& in cultured pacemaker cells isolated from the rabbit sinoatrial node. Removal of these anionic sugar moieties by neuraminidase (1.0 U/ml for 5-20 min) increased lca,T in five of nine cells (by a factor of 2.2-5.1) and lca L in three of six cells (by a factor of 1.2-1.6). In cells that did not exhibit such an increase, the enzyme reduced &,T but had no significant an increase in &T, effect on lca,L. In cells that exhibited exposure to neuraminidase also shifted the activation curve to more negative potentials and increased the slope of the inactivation curve. The enzyme did not influence the gating of lca,L or the rates of inactivation of either &,T or lca,~. The enhancement of &a,T and lca,L could not be mimicked by including neuraminidase in the patch pipette or by adding a contaminant of the enzyme preparation, phospholipase C, to the bath. When external Ca2+ was replaced by Ba2’, neither I&T nor ~c~,L was increased significantly by neuraminidase. It is proposed that by removing sialic acid residues neuraminidase might directly alter the gating of T-type Ca2+ channels. On the other hand, the increased amplitudes of Ica,T and lca,L might be due to a rise in intracellular Ca2+.
neuraminidase;
calcium
channels;
sinoatrial
node
SIALIC ACID (N-acetylneuraminic acid), which forms the terminal end of N-linked complex oligosaccharides of glycoproteins and glycolipids, is commonly found on the external surface of cardiac myocytes (1, 9, 32). Removal of these anionic residues by neuraminidase, an enzyme that selectively catalyzes the hydrolysis of sialic acid’s glycosidic linkage (4), results in a five- to sixfold increase in 45Cauptake (9), a rise and then fall in the plateau of the action potential (23), a negative shift of the action potential threshold (23), and arrhythmic firing as a result of depolarizing afterpotentials (13) or spontaneous fluctuations in the membrane potential (2). In the sinoatrial node, neuraminidase increased spontaneous rate by 27% after 36 min of perfusion through the nodal artery (32). Then after another 60-90 min, the rate slowed and became arrhythmic, concomitant to an apparent loss of intercellular coupling. At that time, the duration of the action potential had increased, but the maximum diastolic potential, maximum upstroke velocity, and conduction velocity all had declined. A recent study has shown that some of these effects might be mediated by 0363-6135/91
$1.50 Copyright
an increase in low-threshold transient (T-type) Ca” current (34). Sialic acid residues account for 8 and l2%, respectively, of the molecular weights of the a-subunits of rat brain (25) and eel electroplax (15) Na+ channels. Removal of these residues by neuraminidase had no effect on the binding of saxitoxin or a-scorpion toxin but eventually led to the appearance of multiple reduced conductance states (15, 24) and positive shifts of the steady-state activation curve (15). Although cardiac Ca2’ channels have yet to be tested, the apparent sizes of both the a2- and y-subunits of skeletal muscle L-type Ca2’ channels were reduced by neuraminidase, confirming that sialic acid is indeed present (29); the al-subunit appears to be “lightly” glycosylated (11, 27), but the presence of sialic acid has not been investigated. The goal of the present study was to test the hypothesis that the presence of sialic acid residues influences both T- and L-type Ca2+currents (lca,T and lca,L) in pacemaker cells of the sinoatrial node. Preliminary reports of our results have been published (7, 8). METHODS
Cell isolation and culture. After the heart was removed from a male New Zealand White rabbit, the right atrium was excised, pinned to the bottom of a Sylgard-coated Petri dish, and bathed (at room temperature) in an oxygenated N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (HEPES) -buffered balanced salt solution (HBSS) that contained 50 PM Ca? The sinoatrial node was removed, cut into pieces (2-3 mm), and transferred to a 25-ml Erlenmeyer flask that contained 5 ml of an enzyme solution (see below for composition) and a 3 x lo-mm magnetic stir bar. The following procedure is a simplification of the one originally used to isolate single pacemaker cells (19). After the solution was allowed to stand for 10 min (37”C), the supernatant was discarded and replaced by 5 ml of fresh enzyme solution. This solution was stirred at 200 rpm for 15 min, and the supernatant (which contained freed cells) was transferred to a refrigerated centrifuge tube that contained 10 ml of HBSS with 0.1% bovine serum albumin (BSA; A2153, Sigma). After this step had been repeated 4 times, any remaining tissue was triturated lo-20 times, and the supernatant was transferred to the centrifuge tube. The cell suspension was then centrifuged at 180 g for 10 min. After discarding the supernatant, we resuspended the cells in 0.5 ml of a culture medium that contained 1.8
0 1991 the American
Physiological
Society
H735
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
H736
REMOVAL
OF
SIALIC
ACID
mM Ca2+ (see below for composition) and counted them on a hemacytometer (1.0-3.5 x lo5 cells). Between 2 and 4 X lo* cells were plated in each of 3-10, 35mm plastic culture dishes (GIBCO) and maintained in a carbon dioxide incubator (humidified atmosphere of 95% air-5% COZ, 37°C). All of the experiments were performed on type II pacemaker cells (19) after periods of 2-5 days in vitro. Electrophysiology. The whole cell configuration of the patch-clamp technique was used to record Ca”+ currents (lca) from single myocytes. Patch pipettes were prepared from l.O-mm-OD borosilicate capillary glass (Sutter Instrument) on a Flaming and Brown horizontal puller (Sutter). The pipettes were filled with (in mM) 120 CsCl, 1 MgCl,, 5 MgATP, 5 Na-creatine phosphate, 10 ethylene glycol-bis(P-aminoethyl ether)-N,N,N’,N’-tetraacetic acid (EGTA), 20 tetraethylammonium chloride, and 5 HEPES. The pH was adjusted to 7.4 with CsOH. Assuming that the contaminant Ca2+ concentration was ~50 PM, we calculated the pCa of this solution to be >12.5 (5). Electrode resistances ranged from 5 to 10 MQ. For each electrode, we corrected for the liquid-junction potential between the pipette solution and the bath (-3 to -5 mV), but not the (variable) liquid-junction potential between the pipette and cytoplasm. The latter potential was not a problem, since the contents of the pipette appeared to equilibrate with the cytoplasm before the experiments were begun. A Dagan model 8900 patchclamp amplifier with a lo- GQ feedback resistor was used to record the currents. Series resistance compensation was used routinely to maximize the rate of decay of the capacitive transient. Experiments were performed at temperatures between 36 and 37°C. Data analysis. Whole cell currents were filtered at 1 kHz, digitized by an 80-kHz Labmaster analog-to-digital board (Scientific Solutions) and stored on the hard disk of an IBM PC/AT computer with the help of pClamp software (Axon Instruments). Ca2+ current was taken to be the difference between the peak inward current and the steady-state current at the end of the voltage step. Such measurements did not differ from th .ose obtained when lca was corrected for “leakage current” (i.e., the current remaining after I Ca had been blocked by 2 mM NiC12). These currents were normalized for cell surface areas (A), which were estimated from microscopic measurements of the major and minor axes of rotation (2a and 2b), assuming that each cell could be approximated bY a.n oblate spheroid A = 2ra” + rb2 In [(1 + c)/(l - E)]/E
(1)
where
For 38 pacemaker cells, the mean dimensions were 19.8 t 0.5 X 24.4 t 0.7 pm, and the mean cell surface area was 1,679 t 87 pm2. Activation-voltage relationships for 1ca, d( V), were estimated from normalized conductance-voltage curves d
z
GCa/GCa,max
=
ICa/[
GCa,max(
Vm
-
Erw)
1
(3)
where Gca is the chord conductance at membrane poten-
ALTERS
cA2+
CURRENTS
is the maximum chord conductance, ( VIII), GCa,max and Rev is the reversal potential, which was estimated from a linear fit of the last three values of the currentvoltage relationship. This equation is only approximate, because it is based on the assumption that the inactivation parameter (f) is 1.0 at the time 1oa reaches its maximum (12). Steady-state inactivation-voltage relationships for 1cawere measured in two-pulse experiments in which a variable-amplitude prepulse, 450 ms in duration and long enough to produce complete inactivation at each potential, was followed by a fixed-amplitude test pulse, 250 ms in duration. The test-pulse voltage was chosen to maximize the current of interest (-40 to -50 mV for JC,a,Tand +10 to +30 mV for 1ca,L). The dual pulses were presented once every 5 s, and f(V) was determined from normalized currents recorded during the test pulse. Activation and inactivation curves, respectively, were calculated from the following relationships tial
GCa/GCa,max
=
1
- l/( 1 +
=
l/(1 + exp [( V -
exp
[(V
-
K,2)/4j
(4)
and ICa/lCa,rnax
V1j2>/41
(5)
where values for the half-activation or half-inactivation potential ( Vli2) and slope factor (k) were determined from a nonlinear least-squares curve fitting program written for the IBM-PC by W. N. Goolsby (Dept. of Anatomy and Cell Biology, Emory University). Data are expressed as means t SE. Student’s t tests for paired or grouped data were used to evaluate the statistical significance of differences between means. Values of P < 0.05 were considered to indicate significance. Solutions. The bathing solution consisted of (mM) 126 NaCl, 5.4 KCI, 2.5 CaC12 (or 2.5 BaC12), 0.8 MgC12, 0.3 NaH2POd, 5.5 dextrose, and 5.0 HEPES (pH adjusted to 7.4 with NaOH). In addition, 20 mM CsCl, 30 PM tetrodotoxin (TTX), and 4 mM 4-aminopyridine were added to block interfering currents (19). The enzyme solution used to isolate single pacemaker cells, which was prepared the morning of the culture, contained 0.2% collagenase (type CLS II; Worthington Biochemical) and 0.1% BSA in a nominally Ca2+- and Mg2+-free salt solution composed of (mM) 116 NaCl, 5.4 KCI, 0.4 NaH2P04, 1.0 Na2HP04, and 5.5 dextrose (pH adjusted to 7.4 with NaOH). The HBSS contained (mM) 126 NaCl, 1.8 KCl, 0.8 MgCl,, 5.5 dextrose, and 25 HEPES (pH adjusted to 7.4 with NaOH). The culture medium contained 20% medium 199 (GIBCO), 4% fetal bovine serum (GIBCO), 2% horse serum (GIBCO), 0.5% gentamicin sulfate (GIBCO), and 73.5% of a balanced salt solution composed of (mM) 116 NaCl, 1.8 CaC12, 0.8 MgSO4, 0.9 NaH2P04, 26.0 NaHC03, and 5.6 dextrose. Neuraminidase (type X, Sigma) was dissolved in distilled water to yield a stock solution of 100 U/ml, aliquoted, and then frozen. When needed, 10 ~1 were thawed and added to the bathing solution (1.0 ml). The amount of sialic acid released by the enzyme was unaffected by such freezing and thawing (17). Nifedipine (Sigma) was dissolved in ethanol.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
ALTERS
cA2+
CURRENTS
H737
As illustrated in Fig. lC, the maximum amplitudes of ka,T and IC a,Lwere about the same: Ioa,T/Ioa,L = 1.2 t 0.2 Cdcium currents. The results to be described below (n = 16). This ratio is much larger than the one measured were unaffected by the time the cells remained in culture. previously by Hagiwara et al. (0.2 t 0.01; n = 10; Ref. Neither the current densities (at -40 mV for Ica T or at lo), possibly because Ica T was sixfold greater in the +20 mV for I caL), half-activation potentials nor half- present study and I ca,L w’as larger in the former study inactivation potentials for Ica,T and Ica~ varied signifiwhere adenosine 3’,5’-cyclic monophosphate (CAMP) cantly with time in vitro between 2 and’5 days. was included in the patch pipettes. Also, the threshold As demonstrated previously in single pacemaker cells and maximum value of the I ca,T-voltage relationship were freshly isolated from the rabbit sinoatrial node (lo), two 10 and 20 mV more negative, respectively, in the present Ca”’ currents could be separated by adjusting the holding study. potential. When the holding potential was -80 or -90 Figure 2 illustrates activation- and inactivation-voltmV, Ica,T could be recorded at potentials between -60 age relationships for Ioa,T and Ioa,L. The activation paand -40 mV (Fig. 1A); however, when the holding potenrameters for Ica,T are VI,2 = -38.3 & 1.4 mV and k = 5.5 tial was -50 or -40 mV, this current was eliminated t 0.3 mV (n = 11); for activation of 1ca,L, V1,2= -20.3 t (Fig. 1B). The current that remained, 1ca,L, could be 0.8 mV and k = 4.5 t 0.2 mV (n = 10). The inactivation blocked by 4 PM nifedipine (7 expts). In contrast to the parameters for IoaT are VI,2 = -58.7 t 1.3 mV and k = previous study (lo), we were unable to block IcaT with 3.8 t 0.2 mV (n L 12); for inactivation of 1ca,L, VI/z = 40 PM Ni2+, and higher concentrations reduced both -35.4 t 1.2 mV and k = 5.3 t 0.3 mV (n = 13). These currents. When both Ica T and Ica L were blocked by Ni2+ results differ from previous data on nodal pacemaker and nifedipine, we observed no transient inward current cells (10); 1) our half-activation potentials for both Ica T at any potential (holding potential = -80 mV; 4 cells). and 1ca,L are more negative; 2) our half-inactivatio’n This confirms that Na+ current (INa) was eliminated by potential for I oa,T is more positive and that for Ica~ is 30 PM TTX in the bath and that 1Nadid not interfere more negative; 3) the separation of inactivation curves with the measurement of IcaT. This is also evidenced by for ICa,T and ICa,L is only 23.3 mV compared with 50 mV the much slower decay of ILaT when Ca2+ was replaced in the previous study; 4) the area of overlap of IcaT by Ba2+ (Fig. 7A). Current-voltage relationships for Ica T activation and inactivation curves (i.e., “window curwere obtained by subtracting 1ca,L from the total 1La rent”) is much greater in the present study; and 5) the recorded when the holding potential was -80 or -90 mV. activation and inactivation curves for both Ica ,T and Ica 7L RESULTS
I
n cu 5
1OOms
-10
I1
II
-60
I
-40
Membrane
I
-20
I
I
0
Potential
11
11
20
FIG. 1. Separation of T- and L-type Ca2’ currents. A: both IcaT and lca L are elicited by voltage steps from a holding potential (HP) of -80 mV. Horizontal arrow is 0 pA. B: only lca,L remains when HP = -50 mV (no transient inward current flows at -40 mV as it did in A). C: current-voltage relationships for total transient inward current (Ica; HP = -90 mV; A), lca,L (HP = -50 mV; w), and lca,T 15 or 16. Peak curuca - ka.L; a; n = rents were measured with respect to steady-state current at end of 530-ms voltage steps and normalized by cell surface areas. Values shown here and in subsequent figures are means t SE. Extracellular [Ca”‘] = 2.5 mM.
1
40
(mV>
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
H738
5 c-l
-E
REMOVAL
OF
SIALIC
ACID
ALTERS
cA2+
0.4
40
J“N
--N
0.2
-is ;s z
0.0’
CURRENTS
’
-80
’
-e
.*.- -&... A \ .....** .** o ’
2. Activationand inactivation-voltage relationships, d( V) and f( V), for lctl,T (dotted curves) and I ca,L (solid curves). d( V) was approximated by normalized chord conductance (G/G,,,), and f( V) was determined by using a double-pulse protocol (inset) and normalizing peak ) during test pulses to -40 or -50 mV for lca,T and currents &alka,max to +lO or +30 mV for &J~. Voltage steps were applied once every 5 s. Averaged data were fit by Eqs. 4 and 5 (text) and are characterized by half-activation and half-inactivation potentials and slope factors listed in text. Holding potentials were -80 or -90 mV for lca,T and -40 or -50 mV for 1c+. FIG.
are steeper in the present study. Effect of neuraminidase. Addition of neuraminidase to the bathing solution (1.0 U/ml) lead to an increase in &,T in five of nine cells and an increase in lca,L in three of six cells. For example, loa,T was enhanced by 50% within 5 min (Fig. 3A) and IcaL by 18% within 10 min (Fig. 3B) of adding the enzyme.’ Time-independent components of these currents were also potentiated. In these two examples, the holding current was reduced after the addition of neuraminidase; however, this was not a consistent finding. After a lo-min exposure to neuraminidase, the increase in the time-dependent component varied considerably from cell to cell, ranging from 91 to 200% for lca,T and from 18 to 77% for IcaL. Changes in lca,T in comparison with “rundown” were maximal within 20 min (Fig. 4A), but were statistically significant only at 10 min; those for I ca,L were maximal within 10 min (Fig. 4B) but were significant only at 10 and 20 min. These effects eventually declined, but they did not differ significantly from the maximum by the end of the 30min recording periods (Fig. 4). In fact, when cells were exposed to neuraminidase for 1 h and then whole cell recordings were obtained, we compared the treated cells with controls (a different group of untreated cells) and still found a significant increase in Ica, ranging from 91 to 119% at potentials between -10 and -30 mV. After l-h exposures, augmentation of IcaL ranged from 23 to 54% at potentials between -20 and +40 mV, but these changes were not statistically significant. Such changes were significant (P < O.Ol), however, when cells were exposed to neuraminidase for only lo-20 min (Fig. 4B). To investigate the possibility that the effects of neuraminidase were mediated by the hydrolysis of intracellular sialic acid, we added the enzyme to the patch pipette solution (1.0 U/ml). When we compared untreated cells
FIG. 3. Neuraminidase (1.0 U/ml) enhances both Ica,T (A) and lca,L (B). A: current at -50 mV before (C) and 5 min after neuraminidase (N); holding currents were -80 pA (C) and -67 pA (N) at -90 mV. B: current at +lO mV before (C) and 10 min after neuraminidase (N); holding currents were -77 pA (C) and -59 pA (N) at -40 mV. Traces were superimposed so that holding currents overlapped. Peaks of capacitive currents are not illustrated.
with cells that had been exposed to neuraminidase via the pipette, we found no significant differences for either I Ca,T (Fig. u) (Fig. 4B). Finally, control experiOr Ica,L ments were performed with phospholipase C, a contaminant often found in neuraminidase preparations (1, 9). We used 0.02 U/ml (type XIV, Sigma), a concentration that was twice the contaminant level found in Worthington type NEUA neuraminidase (9). We showed previously that the Worthington enzyme preparation is much less pure than the one used in the present study (1). Phospholipase C failed to increase either Ica T or Ica L. In fact, this enzyme reduced both IcaT (Fig. 4& and’Io,r,, (Fig. 4B) in comparison with untreated cells. Because neuraminidase enhanced 1oain some cells but not others, we treated these two groups separately in analyzing their current-voltage relationships. For the five cells that responded with an increase in Ica,T (Fig. 5A), the changes were statistically significant only at -60 and -40 mV, potentials at which IcaL was unlikely to be activated. For the four cells that did not respond with an increase (Fig. 5C), there was a significant decrease at potentials between -30 and 0 mV, where both IcaT and 1ca,i,would have been activated. Interestingly, there was essentially no effect of the enzyme at potentials positive to 0 mV, where most of the channels would have been open. In contrast, with the holding potential at -40 mV to eliminate I ca,T, neuraminidase increased IC,a,Lin three cells (Fig. 5B) and decreased Ica L in three cells (Fig. 50), all at positive potentials. Although not significantly different from I oa,Lrecorded before exposure to neuraminidase, the amplitudes of IC,a,L10 or 20 min after addition
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
A 800
I
I
t t L- 600 t c3
-n
L
400
t
I
CI0 b ?
?OO1 y 0.05). To begin to investigate the mechanisms that underlie the changes in I ca,T and 1ca,L described above, we analyzed the effects of neuraminidase on their activationand inactivation-voltage relationships. Figure 6 illustrates data for cells that showed an increase in current amplitude. For the activation of Ica T (Fig. 6A), VI,2 was shifted from -36.1 t 4.1 to -43.4 i 3.1 mV (P c 0.02; n = 5),
ALTERS
CA’+
CURRENTS
H739
but k was not changed significantly. For the inactivation (Fig. 6A), k was reduced from 4.2 t 0.4 to 3.3 t of ICa,T 0.1 mV (P < 0.05; n = 4), but V1iz was not changed significantly. None of the neuraminidase-induced changes in 1c,a,L activation or inactivation (Fig. 6B) were statistically significant. This was true regardless of whether curves measured 10 min after the addition of neuraminidase were compared with those measured before its addition or with those measured in untreated cells 10 min after initiating the whole cell recording. In those cells that exhibited a decrease in Ica,T (not illustrated), the only significant changes were an increase in the slope factor for the activation curve from 5.1 t 0.3 to 5.9 t 0.3 mV (P < 0.02; n = 4) and a shift of the inactivationcurve V1iz from -56.3 t 0.8 to -47.8 t 0.4 mV (P < 0.01; n = 3). In cells that did not exhibit an increase in 1ca,L, none of the changes were statistically significant. In untreated cells, we found no significant changes in the gating parameters for either Ica T or Ica,T,, even after lo20 min of rundown. Finally, time constants for the inactivation of Ica T at -40 mV (holding potential = -80 or -90 mV) or for the inactivation of IcaL at +10 mV (holding potential = -40 mV) were not affected significantly by neuraminidase, both in cells that exhibited an increase in calcium current and those that did not (Table 1) . Role of external Ca”+. To determine whether the increases in I ca,T and Ica,L were due to a direct effect on Tand L-type Ca2+ channels or to the greater influx of Ca2+ induced by neuraminidase (9, 23), we replaced the external Ca2+ with 2.5 mM Ba2’. Cells were rinsed twice with the Ba2+-containing bathing solution before 1.0 ml of that solution was added. Atomic absorption spectroscopy indicated that Ca2+ concentration was
FIG. 5. Effects of neuraminidase (1.0 U/ml) on current-voltage relationships for Ica (A and C) and lca,L (23 and 0). Controls currents (0, dotted curves) were measured just before addition of enzyme. Currents influenced by neuraminidase (m, solid curves) were measured between 5 and 12 min after its addition to bath. A: increases in Ica were significant at -60 and -40 mV (P < 0.05 and 0.02; n = 5). B: increases in I ca,~ were not significant at any potential (n = 3). C: reductions of I ca,T were significant at potentials between -30 and 0 mV (P < 0.05, 0.01, 0.01, and 0.02; n = 4). D: reductions of I ca,~ were not significant at any potential (n = 3). Paired t tests were used in all comparisons. Holding potentials were -80 or -90 mV in A and C and -40 mV in B and D.
ence of the gating of I ca,T;the rates of inactivation of the two currents were unchanged. These findings differ considerably from those obtained in guinea pig ventricular myocytes exposed to neuraminidase (34). In that study, only IcaT was affected; it was increased in 2550% of the cells, and the time constant for its inactivation was reduced. Two factors could have contributed to less hydrolysis of sialic acid residues in the former study. Yee and colleagues (34) used a lower concentration of neuraminidase (0.25 vs. 1.0 U/ml in the present study) and lower enzyme incubation temperatures (23-25 vs. 3637°C in the present study). Another important difference might be the greater number of ,&adrenergic receptors on nodal vs. ventricular myocytes, assuming that neuraminidase influences I ca,L indirectly by acting on these receptors (see below). Neuraminidase-induced changes in amplitudes of Ica,T At least three mechanisms could account for and &Lthe enhancement of I ca,T and Ica,L: 1) a direct effect of the enzyme on T- and L-type Ca2+ channels, such as changes in their conformations; 2) an indirect effect in which increased phosphorylation or dephosphorylation of the channels results from the rise in intracellular Ca2+ concentration ([ Ca2+];); and 3) an indirect effect me-
diated by other second messengers, such as CAMP or guanine nucleotide binding proteins (G proteins), following hydrolysis of sialic acid-containing receptors. A fourth mechanism, an increase in local extracellular Ca2+ concentration due to the release of Ca2+ bound to sarcolemma1 sialic acid, seemsunlikely, since such an increase would have reduced the negative external surface potential and thereby shifted the activation or inactivation curves to more positive potentials (20). The presence of sialic acid on P-cells of the sinoatrial node has been confirmed by the use of neuraminidase to remove ruthenium red, a cationic stain (32). But because of the small yield of cells from the rabbit sinoatrial node (1.0-3.5 x lo5 cells), we were unable to quantify the amount of sialic acid released by neuraminidase. As a rough estimate, we found previously that 1.0 U/ml of Sigma type X neuraminidase (36°C; 1 h) released 84.3 t 3.6% (n = 7) of the sarcolemmal sialic acid from embryonic chick ventricular myocytes (17). Also, Yee et al. (34) found that 0.12 U/ml of Worthington neuraminidase released 20% of the cellular sialic acid from guinea pig ventricular myocytes in 20 min. Still, three critical questions remain to be answered: 1) are sialic acid residues present on the glycosylated subunits of T- and L-type
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
ALTERS
CA’+
H741
CURRENTS
1. Time constants for inactivation of Ica,T and Ica,L are not influenced by neuraminidase TABLE
Neuraminidase
Control Condition
of Cells ms
Increased Ica,T NO increased Ica,T Increased 1ca,L NO increased 1ca,L
0.2
0.0
-80
-60
-40
-20
0
5 4
43.4t3.6
3
49.3t9.6
3
R
ms
n
R
0.96
13.2t2.9
5
0.98
0.98 1.00 0.99
13.1t3.7 38.522.9 54.8t6.4
4 3 3
0.96 1.00 0.96
Values are means t SE of time constants calculated by pClamp for single exponential fits to the first 100 ms of the decay of Ica,T at -40 mV (holding potential = -80 or -90 mV) and 1ca,L at +lO mV (holding potential = -40 mV). Currents were recorded before or after 10 min of exposure to 1.0 U/ml of neuraminidase. R indicates the goodness of fit by a single exponential, with 1.0 being a perfect fit. As determined by Student’s t test for paired variates, none of the differences were statistically significant (P > 0.05).
A
B
10.7t0.9 8.1zH.5
n
I
0.8
c
L ZJ
0.6
B 0.0
Membrane
Potent
ial
(mV>
k+&
6. Effects of neuraminidase (1.0 U/ml) on activation and inactivation of lca,~ (4 as described in Fig. and L,L (B). M easurements, 2, were made before (0, dotted curves) or 5-25 min after addition of neuraminidase (H, solid curves). Only cells that exhibited an increase in current (Fig. 5, A and B) were included. A: significant effects were a negative shift of lca,T activation curve from -36.1 to -43.4 mV (P < 0.02; n = 5) and a reduction of ilz for I ca,T inactivation curve from 4.2 to 3.3 mV (P < 0.05; n = 4). B: there were no significant effects of neuraminidase on gating of 1ca,L (n = 3). FIG.
Ca2+ channels; 2) did neuraminidase reach these sites; and 3) did hydrolysis of these sites alter channel function? Although cardiac Ca2+ channels have yet to be analyzed, skeletal muscle L-type Ca2+ channels do possesssialic acid residues on both the cy2-and y-subunits (29). The al-subunit, which constitutes the pore, is only lightly glycosylated (11,27), and its content of sialic acid has not been examined. A cardiac-specific isoform of the al-subunit has two additional potential extracellular Nglycosylation sites (26), but, again, the presence of sialic acid has yet to be investigated. Although it is possible that neuraminidase had a direct effect upon T- and L-type Ca2+channels in our cultured pacemaker cells (see below), it is unlikely that such an effect was responsible for the enhancement of IcaT , and Ica,L, since these currents were not increased when external Ca2’ was replaced by Ba2+ (Figs. 7 and 8), which also moves readily through T- and L-type Ca2+ channels. On
FIG. 7. Neuraminidase (1.0 U/ml) fails to enhance Ica,T (A) or lca,L (B) when external Ca2’ is replaced by Ba? A: current at -40 mV before (C) and 10 min after neuraminidase (N); holding currents were -82 (C) and -45 pA (N) at -90 mV. B: current at -5 mV before (C) and 10 min after neuraminidase (N); holding currents were -52 (C) and -35 pA (N) at -40 mV. Traces were superimposed so that holding currents overlapped. Peaks of capacitive currents are not illustrated.
the other hand, these experiments do suggest that the enhancement of I Ca,Tand Ica,L may depend upon external Ca2+. We and others have shown that when neuraminidase is used to remove sialic acid from the sarcolemma, 45Cauptake is enhanced in cultured ventricular cells (9, 23), and canine Purkinje fibers display depolarizing afterpotentials (13), spontaneous voltage fluctuations (22), and triggered activity (13, 32), processes that are promoted by elevated [ Ca2+];. Although its mode of entry is unknown, such an increase in [Ca2+]; could modulate the conductance of T- and/or L-type Ca2+ channels. Both Ca2+-induced potentiation (16) and Ca”‘-induced inhi-
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
H742
REMOVAL
OF
SIALIC
ACID
r
H
0 ti
20
‘.
.. .. ..
.. ..
0 1
0
I
1
Time
I
I
20
10
(mid
FIG. 8. Changes in Ica at -30 mV (HP = -90 mV; A) and lca,L at -15 mV (HP = -40 mV; B) when external Ca2+ is replaced by Ba? Cells were exposed to neuraminidase (1.0 U/ml) for times indicated (m), or currents were allowed to rundown without exposure to neuraminidase (0; elapsed time since initiation of whole cell recording). For cells exposed to neuraminidase, only those that exhibited an increase in current (3 cells) were included in the averages. Mean values were also determined for 2-4 controls. Individual results for each of 3 cells after 20 min (values without error bars) are also included. No changes induced by neuraminidase were significantly different from those due to rundown.
bition of IcaL (14, 31) have been observed following increases in lCa2+];, and we have seen the latter process in our cultured pacemaker cell preparation (21). Consistent with the apparent biphasic effects of neuraminidase on lca,T (Figs. 4 and 5, A and C), Tseng and Boyden (31) recently found that when [Ca2+]; was increased by dialysis of canine ventricular or Purkinje cells with EGTA and Ca2+,lca,T first increased and then decreased or lca,T increased in some cells and decreased in others. L-type Ca2+channels are also regulated by CAMP (30) and by a G protein, G, (33), both of which are able to mediate the stimulatory effects of agonist binding to ,& adrenergic receptors. Sialic acids contribute significantly to the carbohydrate content of ,0 receptors (28), which are known to be concentrated on nodal pacemaker cells. Thus the changes in I ca,Linduced by neuraminidase could have arisen indirectly from the removal of these sialic
ALTERS
CA"
CURRENTS
acid residues. Such removal could have altered the conformation of the receptors and resulted in stimulation or inhibition of adenylate cyclase and/or G,. This would seem to be unlikely, however, since neuraminidase treatment of guinea pig hearts did not alter the positive inotropic effect of isoproterenol, a ,&agonist (6). Moreover, coupling between these receptors and T-type Ca2+ channels has not been demonstrated in the sinoatrial node (10). Neuraminidase-induced changes in gating of Ica,T. A negative shift of the I ca,T activation curve (Fig. 6A ) may be partly responsible for the increased amplitude of IcaT following neuraminidase treatment (Fig. 5A ). Moreover, the increased overlap of the activation and inactivation curves (Fig. 6A), i.e., window current, might account for the rise in steady inward current following exposure to the enzyme (Fig. 3A). On the other hand, it is unlikely that the shift of the activation curve resulted from an increase in I ca,T or [Ca2+];. In theory, binding of Ca2+ to negative fixed charges on the inner surface of the sarcolemma could have accounted for the negative shift (20); however, this would have been unlikely because this (at most micromolar) Ca2’ would have had to compete with almost millimolar concentrations of Mg2+ already present in the cytosol(18). It is also unlikely that the negative shift occurred because sialic acid residues constitute a portion of the membrane’s external surface charge, since removal of that charge would have shifted the activation curve toward positive potentials (20). Taken together, these results suggest that neuraminidase may directly influence the gating of T-type Ca2+ channels. This idea is supported by the fact that neuraminidase altered the slopes of both inactivation (Fig. 6A) and activation (not illustrated) curves (for cells in which IcaT was increased or decreased, respectively) but had no’ effect on their half-inactivation and half-activation potentials. Thus channel components involved in gating, rather than membrane surface charge, appear to have been modified by the enzyme. In summary, the removal of sialic acid residues by neuraminidase altered the gating of IcaT and the amplitudes of both Ica T and Ica L. These effects appear to have been the result of both direct and indirect actions of the enzyme on T- and L-type Ca”+ channels. Additional experiments will be required to prove this hypothesis. We thank Theresa E. Redington for preparing the cultured pacemaker cells. This study was supported by National Heart, Lung, and Blood Institute Grant HL-20708. B. Fermini held a fellowship from the Canadian Heart Foundation. Present address of B. Fermini: Montreal Heart Institute, 5000 East Belanger, Montreal, Quebec HlT lC8, Canada. Address reprint requests to R. D. Nathan. Received
18 December
1989; accepted
in final
form
17 October
1990.
REFERENCES 1. BARRON, E. A., R. R. MARKWALD, AND R. D. NATHAN. Localization of sialic acid at the surface of embryonic myocardial cells. J. Mol. Cell. Cardiol. 14: 381-395, 1982. 2. BHATTACHARYYA, M. L., R. D. NATHAN, AND V. L. SHELTON. Release of sialic acid alters the stability of the membrane potential in cardiac muscle. Life Sci. 29: 1071-1078, 1981.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
3. CASSIDY, J. T., G. W. JOURDIAN, AND S. ROSEMAN. The sialic acids. VI. Purification and properties of sialidase from CZostridium perfringens. J. BioZ. Chem. 240: 3501-3506, 1965. 4. DRZENIEK, R. Substrate specificity of neuraminidases. Histochem. J. 5: 271-290,1973. 5. FABIATO, A. Computer programs for calculating total from specified free or free from specified total ionic concentrations in aqueous solutions containing multiple metals and ligands. Methods Enzymol. 157: 378-417, 1988. 6. FAWZI, A. B., AND J. H. MCNEILL. Effect of neuraminidase treatment on the inotropic response to ouabain, isoproterenol and calcium in the guinea pig heart. Eur. J. Pharmacol. 112: 295-311, 1985. 7. FERMINI, B., AND R. D. NATHAN. Do sialic acid residues account for surface charge modulation of T- and L-type calcium currents in cultured pacemaker cells from the sino-atria1 node? (Abstract). Biophys. J. 55: 299A, 1989. 8. FERMINI, B., AND R. D. NATHAN. Removal of sialic acid alters both T- and L-type Ca currents in cardiac myocytes (Abstract). Biophys. J. 57: 519A, 1990. 9. FRANK, J. S, G. A. LANGER, L. M. NUDD, AND K. SERAYDARIAN. The myocardial cell surface, its histochemistry, and the effect of sialic acid and calcium removal on its structure and cellular ionic exchange. Circ. Res. 41: 702-714, 1977. 10. HAGIWARA, N., H. IRISAWA, AND M. KAMEYAMA. Contribution of two types of calcium currents to the pacemaker potentials of rabbit sino-atria1 node cells. J. Physiol. Lond. 395: 233-253, 1988. 11. HOSEY, M. M., J. BARHANIN, A. SCHMID, S. VANDAELE, J. PTASIENSKI, C. O’CALLAHAN, C. COOPER, AND M. LAZDUNSKI. Photoaffinity labelling and phosphorylation of a 165 kilodalton peptide associated with dihydropyridine and phenylalkylamine-sensitive calcium channels. Biochem. Biophys. Res. Commun. 147: 11371145,1987. 12. ISENBERG, G., AND U. KL~CKNER. Calcium currents of isolated bovine ventricular myocytes are fast and of large amplitude. Pfluegers Arch. 395: 30-41, 1982. 13. KIMURA, S., S. HAZAMA, S. FUJII, H. NAKAYA, AND M. KANNO. Delayed afterdepolarisations and triggered activity in canine Purkinje fibres treated with neuraminidase. Cczrdiouasc. Res. 18: 294301,1984. 14. KOKUBUN, S., AND H. IRISAWA. Effects of various intracellular Ca ion concentrations on the calcium current of guinea-pig single ventricular cells. Jpn. J. Physiol. 34: 599-611, 1984. 15. LEVINSON, S. R., W. B. THORNHILL, D. S. DUCH, E. RECIO-PINTO, AND B. W. URBAN. The role of nonprotein domains in the function and synthesis of voltage-gated sodium channels. In: Ion ChunneZs, edited by T. Narahashi. New York: Plenum, 1990, vol. 2, p. 33-64. 16. MARBAN, E., AND R. W. TSIEN. Enhancement of calcium current during digitalis inotropy in mammalian heart: positive feed-back regulation by intracellular calcium? J. Physiol. Lond. 329: 589-614, 1982. 17. MCDONAGH, J. C., AND R. D. NATHAN. Sialic acid and the surface charge of delayed rectifier potassium channels. J. MOL. CeZZ. Curdiol. 22: 1305-1316,199O. 18. MURPHY, E., C. C. FREUDENRICH, L. A. LEVY, R. E. LONDON, AND M. LIEBERMAN. Monitoring cytosolic free magnesium in cultured chicken heart cells by use of the fluorescent indicator Furap-
ALTERS
CA2+ CURRENTS
H743
tra. Proc. NutZ. Acud. Sci. USA 86: 2981-2984, 1989. 19. NATHAN, R. D. Two electrophysiologically distinct types of cultured pacemaker cells from rabbit sinoatrial node. Am. J. Physiol. 250 (Heart Circ. Physiol. 19): H325-H329, 1986. 20. NATHAN, R. D. Negative surface charge: its identification and regulation of cardiac electrogenesis. In: Cardiac Muscle: The Regulation of Excitation and Contraction, edited by R. D. Nathan. Orlando, FL: Academic, 1986, p. 55-86. 21. NATHAN, R. D. Voltage and calcium regulation of calcium channels in pacemaker cells from the sinoatrial node (Abstract). Federation Proc. 45: 511, 1986. 22. NATHAN, R. D., AND M. L. BHATTACHARYYA. Membrane sialic acid and electrophysiology of cardiac Purkinje fibers (Abstract). Biophys. J. 33: 34A, 1981. 23. NATHAN, R. D., S. J. FUNG, D. M. STOCCO, E. A. BARRON, AND R. R. MARKWALD. Sialic acid: regulation of electrogenesis in cultured heart cells. Am. J. Physiol. 239 (Cell Physiol. 8): Cl97-C207, 1980. 24. SCHEUER, T., L. MCHUGH, F. TEJEDOR, AND W. CATTERALL. Functional properties of neuraminidase-treated rat brain sodium channels (Abstract). Biophys. J. 53: 54lA, 1988. 25. SCHMIDT, J. W., AND W. A. CATTERALL. Palmitylation, sulfation, and glycosylation of the cy subunit of the sodium channel: role of post-translational modifications in channel assembly. J. BioZ. Chem. 262: 13713-13723,1987. 26. SLISH, D. F., D. B. ENGLE, G. VARADI, I. LOTAN, D. SINGER, N. DASCAL, AND A. SCHWARTZ. Evidence for the existence of a cardiac specific isoform of the a1 subunit of the voltage dependent calcium channel. FEBS Lett. 250: 509-514,1989. 27. SOLDATOV, N. M. Purification and characterization of dihydropyridine receptor from rabbit skeletal muscle. Eur. J. Biochem. 173: 327-338,1988. 28. STILES, G. L., J. L. BENOVIC, M. G. CARON, AND R. J. LEFKOWITZ. Mammalian /3-adrenergic receptors: distinct glycoprotein populations containing high mannose or complex type carbohydrate chains. J. BioZ. Chem. 259: 8655-8663, 1984. 29. TAKAHASHI, M., M. J. SEAGAR, J. F. JONES, B. F. X. REBER, AND W. A. CATTERALL. Subunit structure of dihydropyridine-sensitive calcium channels from skeletal muscle. Proc. NutZ. Acud. Sci. USA 84: 5478-5482,1987. TRAUTWEIN, W., A. CAVALI~, V. FLOCKERZI, F. HOFMANN, AND D. PELZER. Modulation of calcium channel function by phosphorylation in guinea pig ventricular cells and phospholipid bilayer membranes. Circ. Res. 61, SuppZ. I: 1-17-I-23, 1987. 31. TSENG, G.-N., AND P. A. BOYDEN. Different effects of intracellular Ca and a phorbol ester on cardiac T and L Ca channels (Abstract). Biophys. J. 57: 314A, 1990. 32. WOODS, W. T., K. IMAMURA, AND T. N. JAMES. Electrophysiological and electron microscopic correlations concerning the effects of neuraminidase on canine heart cells. Circ. Res. 50: 228-239, 1982. 33. YATANI, A., J. CODINA, Y. IMOTO, J. P. REEVES, L. BIRNBAUMER, AND A. M. BROWN. A G protein directly regulates mammalian cardiac calcium channels. Science Wash. DC 238: 1288-1292, 1987. 34. YEE, H. F., JR., J. N. WEISS, AND G. A. LANGER. Neuraminidase selectively enhances transient Ca2’ current in cardiac myocytes. Am. J. Physiol. 256 (CeZZ Physiol. 25): Cl267-Cl272, 1989.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
FERMINI
AND
RICHARD
D. NATHAN
of Physiology, Texas Tech University Health Sciences Center, Lubbock, Texas 79430
FERMINI, BERNARD, AND RICHARD D. NATHAN. Removal of sialic acid alters both T- and L-type calcium currents in cardiac myocytes. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): H735H743, 1991.-The whole cell configuration of the patch-clamp technique was used to test the hypothesis that the presence of sialic acid residues influences both T- and L-type Ca2+ currents (Ica,T and 1c& in cultured pacemaker cells isolated from the rabbit sinoatrial node. Removal of these anionic sugar moieties by neuraminidase (1.0 U/ml for 5-20 min) increased lca,T in five of nine cells (by a factor of 2.2-5.1) and lca L in three of six cells (by a factor of 1.2-1.6). In cells that did not exhibit such an increase, the enzyme reduced &,T but had no significant an increase in &T, effect on lca,L. In cells that exhibited exposure to neuraminidase also shifted the activation curve to more negative potentials and increased the slope of the inactivation curve. The enzyme did not influence the gating of lca,L or the rates of inactivation of either &,T or lca,~. The enhancement of &a,T and lca,L could not be mimicked by including neuraminidase in the patch pipette or by adding a contaminant of the enzyme preparation, phospholipase C, to the bath. When external Ca2+ was replaced by Ba2’, neither I&T nor ~c~,L was increased significantly by neuraminidase. It is proposed that by removing sialic acid residues neuraminidase might directly alter the gating of T-type Ca2+ channels. On the other hand, the increased amplitudes of Ica,T and lca,L might be due to a rise in intracellular Ca2+.
neuraminidase;
calcium
channels;
sinoatrial
node
SIALIC ACID (N-acetylneuraminic acid), which forms the terminal end of N-linked complex oligosaccharides of glycoproteins and glycolipids, is commonly found on the external surface of cardiac myocytes (1, 9, 32). Removal of these anionic residues by neuraminidase, an enzyme that selectively catalyzes the hydrolysis of sialic acid’s glycosidic linkage (4), results in a five- to sixfold increase in 45Cauptake (9), a rise and then fall in the plateau of the action potential (23), a negative shift of the action potential threshold (23), and arrhythmic firing as a result of depolarizing afterpotentials (13) or spontaneous fluctuations in the membrane potential (2). In the sinoatrial node, neuraminidase increased spontaneous rate by 27% after 36 min of perfusion through the nodal artery (32). Then after another 60-90 min, the rate slowed and became arrhythmic, concomitant to an apparent loss of intercellular coupling. At that time, the duration of the action potential had increased, but the maximum diastolic potential, maximum upstroke velocity, and conduction velocity all had declined. A recent study has shown that some of these effects might be mediated by 0363-6135/91
$1.50 Copyright
an increase in low-threshold transient (T-type) Ca” current (34). Sialic acid residues account for 8 and l2%, respectively, of the molecular weights of the a-subunits of rat brain (25) and eel electroplax (15) Na+ channels. Removal of these residues by neuraminidase had no effect on the binding of saxitoxin or a-scorpion toxin but eventually led to the appearance of multiple reduced conductance states (15, 24) and positive shifts of the steady-state activation curve (15). Although cardiac Ca2’ channels have yet to be tested, the apparent sizes of both the a2- and y-subunits of skeletal muscle L-type Ca2’ channels were reduced by neuraminidase, confirming that sialic acid is indeed present (29); the al-subunit appears to be “lightly” glycosylated (11, 27), but the presence of sialic acid has not been investigated. The goal of the present study was to test the hypothesis that the presence of sialic acid residues influences both T- and L-type Ca2+currents (lca,T and lca,L) in pacemaker cells of the sinoatrial node. Preliminary reports of our results have been published (7, 8). METHODS
Cell isolation and culture. After the heart was removed from a male New Zealand White rabbit, the right atrium was excised, pinned to the bottom of a Sylgard-coated Petri dish, and bathed (at room temperature) in an oxygenated N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (HEPES) -buffered balanced salt solution (HBSS) that contained 50 PM Ca? The sinoatrial node was removed, cut into pieces (2-3 mm), and transferred to a 25-ml Erlenmeyer flask that contained 5 ml of an enzyme solution (see below for composition) and a 3 x lo-mm magnetic stir bar. The following procedure is a simplification of the one originally used to isolate single pacemaker cells (19). After the solution was allowed to stand for 10 min (37”C), the supernatant was discarded and replaced by 5 ml of fresh enzyme solution. This solution was stirred at 200 rpm for 15 min, and the supernatant (which contained freed cells) was transferred to a refrigerated centrifuge tube that contained 10 ml of HBSS with 0.1% bovine serum albumin (BSA; A2153, Sigma). After this step had been repeated 4 times, any remaining tissue was triturated lo-20 times, and the supernatant was transferred to the centrifuge tube. The cell suspension was then centrifuged at 180 g for 10 min. After discarding the supernatant, we resuspended the cells in 0.5 ml of a culture medium that contained 1.8
0 1991 the American
Physiological
Society
H735
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
H736
REMOVAL
OF
SIALIC
ACID
mM Ca2+ (see below for composition) and counted them on a hemacytometer (1.0-3.5 x lo5 cells). Between 2 and 4 X lo* cells were plated in each of 3-10, 35mm plastic culture dishes (GIBCO) and maintained in a carbon dioxide incubator (humidified atmosphere of 95% air-5% COZ, 37°C). All of the experiments were performed on type II pacemaker cells (19) after periods of 2-5 days in vitro. Electrophysiology. The whole cell configuration of the patch-clamp technique was used to record Ca”+ currents (lca) from single myocytes. Patch pipettes were prepared from l.O-mm-OD borosilicate capillary glass (Sutter Instrument) on a Flaming and Brown horizontal puller (Sutter). The pipettes were filled with (in mM) 120 CsCl, 1 MgCl,, 5 MgATP, 5 Na-creatine phosphate, 10 ethylene glycol-bis(P-aminoethyl ether)-N,N,N’,N’-tetraacetic acid (EGTA), 20 tetraethylammonium chloride, and 5 HEPES. The pH was adjusted to 7.4 with CsOH. Assuming that the contaminant Ca2+ concentration was ~50 PM, we calculated the pCa of this solution to be >12.5 (5). Electrode resistances ranged from 5 to 10 MQ. For each electrode, we corrected for the liquid-junction potential between the pipette solution and the bath (-3 to -5 mV), but not the (variable) liquid-junction potential between the pipette and cytoplasm. The latter potential was not a problem, since the contents of the pipette appeared to equilibrate with the cytoplasm before the experiments were begun. A Dagan model 8900 patchclamp amplifier with a lo- GQ feedback resistor was used to record the currents. Series resistance compensation was used routinely to maximize the rate of decay of the capacitive transient. Experiments were performed at temperatures between 36 and 37°C. Data analysis. Whole cell currents were filtered at 1 kHz, digitized by an 80-kHz Labmaster analog-to-digital board (Scientific Solutions) and stored on the hard disk of an IBM PC/AT computer with the help of pClamp software (Axon Instruments). Ca2+ current was taken to be the difference between the peak inward current and the steady-state current at the end of the voltage step. Such measurements did not differ from th .ose obtained when lca was corrected for “leakage current” (i.e., the current remaining after I Ca had been blocked by 2 mM NiC12). These currents were normalized for cell surface areas (A), which were estimated from microscopic measurements of the major and minor axes of rotation (2a and 2b), assuming that each cell could be approximated bY a.n oblate spheroid A = 2ra” + rb2 In [(1 + c)/(l - E)]/E
(1)
where
For 38 pacemaker cells, the mean dimensions were 19.8 t 0.5 X 24.4 t 0.7 pm, and the mean cell surface area was 1,679 t 87 pm2. Activation-voltage relationships for 1ca, d( V), were estimated from normalized conductance-voltage curves d
z
GCa/GCa,max
=
ICa/[
GCa,max(
Vm
-
Erw)
1
(3)
where Gca is the chord conductance at membrane poten-
ALTERS
cA2+
CURRENTS
is the maximum chord conductance, ( VIII), GCa,max and Rev is the reversal potential, which was estimated from a linear fit of the last three values of the currentvoltage relationship. This equation is only approximate, because it is based on the assumption that the inactivation parameter (f) is 1.0 at the time 1oa reaches its maximum (12). Steady-state inactivation-voltage relationships for 1cawere measured in two-pulse experiments in which a variable-amplitude prepulse, 450 ms in duration and long enough to produce complete inactivation at each potential, was followed by a fixed-amplitude test pulse, 250 ms in duration. The test-pulse voltage was chosen to maximize the current of interest (-40 to -50 mV for JC,a,Tand +10 to +30 mV for 1ca,L). The dual pulses were presented once every 5 s, and f(V) was determined from normalized currents recorded during the test pulse. Activation and inactivation curves, respectively, were calculated from the following relationships tial
GCa/GCa,max
=
1
- l/( 1 +
=
l/(1 + exp [( V -
exp
[(V
-
K,2)/4j
(4)
and ICa/lCa,rnax
V1j2>/41
(5)
where values for the half-activation or half-inactivation potential ( Vli2) and slope factor (k) were determined from a nonlinear least-squares curve fitting program written for the IBM-PC by W. N. Goolsby (Dept. of Anatomy and Cell Biology, Emory University). Data are expressed as means t SE. Student’s t tests for paired or grouped data were used to evaluate the statistical significance of differences between means. Values of P < 0.05 were considered to indicate significance. Solutions. The bathing solution consisted of (mM) 126 NaCl, 5.4 KCI, 2.5 CaC12 (or 2.5 BaC12), 0.8 MgC12, 0.3 NaH2POd, 5.5 dextrose, and 5.0 HEPES (pH adjusted to 7.4 with NaOH). In addition, 20 mM CsCl, 30 PM tetrodotoxin (TTX), and 4 mM 4-aminopyridine were added to block interfering currents (19). The enzyme solution used to isolate single pacemaker cells, which was prepared the morning of the culture, contained 0.2% collagenase (type CLS II; Worthington Biochemical) and 0.1% BSA in a nominally Ca2+- and Mg2+-free salt solution composed of (mM) 116 NaCl, 5.4 KCI, 0.4 NaH2P04, 1.0 Na2HP04, and 5.5 dextrose (pH adjusted to 7.4 with NaOH). The HBSS contained (mM) 126 NaCl, 1.8 KCl, 0.8 MgCl,, 5.5 dextrose, and 25 HEPES (pH adjusted to 7.4 with NaOH). The culture medium contained 20% medium 199 (GIBCO), 4% fetal bovine serum (GIBCO), 2% horse serum (GIBCO), 0.5% gentamicin sulfate (GIBCO), and 73.5% of a balanced salt solution composed of (mM) 116 NaCl, 1.8 CaC12, 0.8 MgSO4, 0.9 NaH2P04, 26.0 NaHC03, and 5.6 dextrose. Neuraminidase (type X, Sigma) was dissolved in distilled water to yield a stock solution of 100 U/ml, aliquoted, and then frozen. When needed, 10 ~1 were thawed and added to the bathing solution (1.0 ml). The amount of sialic acid released by the enzyme was unaffected by such freezing and thawing (17). Nifedipine (Sigma) was dissolved in ethanol.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
ALTERS
cA2+
CURRENTS
H737
As illustrated in Fig. lC, the maximum amplitudes of ka,T and IC a,Lwere about the same: Ioa,T/Ioa,L = 1.2 t 0.2 Cdcium currents. The results to be described below (n = 16). This ratio is much larger than the one measured were unaffected by the time the cells remained in culture. previously by Hagiwara et al. (0.2 t 0.01; n = 10; Ref. Neither the current densities (at -40 mV for Ica T or at lo), possibly because Ica T was sixfold greater in the +20 mV for I caL), half-activation potentials nor half- present study and I ca,L w’as larger in the former study inactivation potentials for Ica,T and Ica~ varied signifiwhere adenosine 3’,5’-cyclic monophosphate (CAMP) cantly with time in vitro between 2 and’5 days. was included in the patch pipettes. Also, the threshold As demonstrated previously in single pacemaker cells and maximum value of the I ca,T-voltage relationship were freshly isolated from the rabbit sinoatrial node (lo), two 10 and 20 mV more negative, respectively, in the present Ca”’ currents could be separated by adjusting the holding study. potential. When the holding potential was -80 or -90 Figure 2 illustrates activation- and inactivation-voltmV, Ica,T could be recorded at potentials between -60 age relationships for Ioa,T and Ioa,L. The activation paand -40 mV (Fig. 1A); however, when the holding potenrameters for Ica,T are VI,2 = -38.3 & 1.4 mV and k = 5.5 tial was -50 or -40 mV, this current was eliminated t 0.3 mV (n = 11); for activation of 1ca,L, V1,2= -20.3 t (Fig. 1B). The current that remained, 1ca,L, could be 0.8 mV and k = 4.5 t 0.2 mV (n = 10). The inactivation blocked by 4 PM nifedipine (7 expts). In contrast to the parameters for IoaT are VI,2 = -58.7 t 1.3 mV and k = previous study (lo), we were unable to block IcaT with 3.8 t 0.2 mV (n L 12); for inactivation of 1ca,L, VI/z = 40 PM Ni2+, and higher concentrations reduced both -35.4 t 1.2 mV and k = 5.3 t 0.3 mV (n = 13). These currents. When both Ica T and Ica L were blocked by Ni2+ results differ from previous data on nodal pacemaker and nifedipine, we observed no transient inward current cells (10); 1) our half-activation potentials for both Ica T at any potential (holding potential = -80 mV; 4 cells). and 1ca,L are more negative; 2) our half-inactivatio’n This confirms that Na+ current (INa) was eliminated by potential for I oa,T is more positive and that for Ica~ is 30 PM TTX in the bath and that 1Nadid not interfere more negative; 3) the separation of inactivation curves with the measurement of IcaT. This is also evidenced by for ICa,T and ICa,L is only 23.3 mV compared with 50 mV the much slower decay of ILaT when Ca2+ was replaced in the previous study; 4) the area of overlap of IcaT by Ba2+ (Fig. 7A). Current-voltage relationships for Ica T activation and inactivation curves (i.e., “window curwere obtained by subtracting 1ca,L from the total 1La rent”) is much greater in the present study; and 5) the recorded when the holding potential was -80 or -90 mV. activation and inactivation curves for both Ica ,T and Ica 7L RESULTS
I
n cu 5
1OOms
-10
I1
II
-60
I
-40
Membrane
I
-20
I
I
0
Potential
11
11
20
FIG. 1. Separation of T- and L-type Ca2’ currents. A: both IcaT and lca L are elicited by voltage steps from a holding potential (HP) of -80 mV. Horizontal arrow is 0 pA. B: only lca,L remains when HP = -50 mV (no transient inward current flows at -40 mV as it did in A). C: current-voltage relationships for total transient inward current (Ica; HP = -90 mV; A), lca,L (HP = -50 mV; w), and lca,T 15 or 16. Peak curuca - ka.L; a; n = rents were measured with respect to steady-state current at end of 530-ms voltage steps and normalized by cell surface areas. Values shown here and in subsequent figures are means t SE. Extracellular [Ca”‘] = 2.5 mM.
1
40
(mV>
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
H738
5 c-l
-E
REMOVAL
OF
SIALIC
ACID
ALTERS
cA2+
0.4
40
J“N
--N
0.2
-is ;s z
0.0’
CURRENTS
’
-80
’
-e
.*.- -&... A \ .....** .** o ’
2. Activationand inactivation-voltage relationships, d( V) and f( V), for lctl,T (dotted curves) and I ca,L (solid curves). d( V) was approximated by normalized chord conductance (G/G,,,), and f( V) was determined by using a double-pulse protocol (inset) and normalizing peak ) during test pulses to -40 or -50 mV for lca,T and currents &alka,max to +lO or +30 mV for &J~. Voltage steps were applied once every 5 s. Averaged data were fit by Eqs. 4 and 5 (text) and are characterized by half-activation and half-inactivation potentials and slope factors listed in text. Holding potentials were -80 or -90 mV for lca,T and -40 or -50 mV for 1c+. FIG.
are steeper in the present study. Effect of neuraminidase. Addition of neuraminidase to the bathing solution (1.0 U/ml) lead to an increase in &,T in five of nine cells and an increase in lca,L in three of six cells. For example, loa,T was enhanced by 50% within 5 min (Fig. 3A) and IcaL by 18% within 10 min (Fig. 3B) of adding the enzyme.’ Time-independent components of these currents were also potentiated. In these two examples, the holding current was reduced after the addition of neuraminidase; however, this was not a consistent finding. After a lo-min exposure to neuraminidase, the increase in the time-dependent component varied considerably from cell to cell, ranging from 91 to 200% for lca,T and from 18 to 77% for IcaL. Changes in lca,T in comparison with “rundown” were maximal within 20 min (Fig. 4A), but were statistically significant only at 10 min; those for I ca,L were maximal within 10 min (Fig. 4B) but were significant only at 10 and 20 min. These effects eventually declined, but they did not differ significantly from the maximum by the end of the 30min recording periods (Fig. 4). In fact, when cells were exposed to neuraminidase for 1 h and then whole cell recordings were obtained, we compared the treated cells with controls (a different group of untreated cells) and still found a significant increase in Ica, ranging from 91 to 119% at potentials between -10 and -30 mV. After l-h exposures, augmentation of IcaL ranged from 23 to 54% at potentials between -20 and +40 mV, but these changes were not statistically significant. Such changes were significant (P < O.Ol), however, when cells were exposed to neuraminidase for only lo-20 min (Fig. 4B). To investigate the possibility that the effects of neuraminidase were mediated by the hydrolysis of intracellular sialic acid, we added the enzyme to the patch pipette solution (1.0 U/ml). When we compared untreated cells
FIG. 3. Neuraminidase (1.0 U/ml) enhances both Ica,T (A) and lca,L (B). A: current at -50 mV before (C) and 5 min after neuraminidase (N); holding currents were -80 pA (C) and -67 pA (N) at -90 mV. B: current at +lO mV before (C) and 10 min after neuraminidase (N); holding currents were -77 pA (C) and -59 pA (N) at -40 mV. Traces were superimposed so that holding currents overlapped. Peaks of capacitive currents are not illustrated.
with cells that had been exposed to neuraminidase via the pipette, we found no significant differences for either I Ca,T (Fig. u) (Fig. 4B). Finally, control experiOr Ica,L ments were performed with phospholipase C, a contaminant often found in neuraminidase preparations (1, 9). We used 0.02 U/ml (type XIV, Sigma), a concentration that was twice the contaminant level found in Worthington type NEUA neuraminidase (9). We showed previously that the Worthington enzyme preparation is much less pure than the one used in the present study (1). Phospholipase C failed to increase either Ica T or Ica L. In fact, this enzyme reduced both IcaT (Fig. 4& and’Io,r,, (Fig. 4B) in comparison with untreated cells. Because neuraminidase enhanced 1oain some cells but not others, we treated these two groups separately in analyzing their current-voltage relationships. For the five cells that responded with an increase in Ica,T (Fig. 5A), the changes were statistically significant only at -60 and -40 mV, potentials at which IcaL was unlikely to be activated. For the four cells that did not respond with an increase (Fig. 5C), there was a significant decrease at potentials between -30 and 0 mV, where both IcaT and 1ca,i,would have been activated. Interestingly, there was essentially no effect of the enzyme at potentials positive to 0 mV, where most of the channels would have been open. In contrast, with the holding potential at -40 mV to eliminate I ca,T, neuraminidase increased IC,a,Lin three cells (Fig. 5B) and decreased Ica L in three cells (Fig. 50), all at positive potentials. Although not significantly different from I oa,Lrecorded before exposure to neuraminidase, the amplitudes of IC,a,L10 or 20 min after addition
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
A 800
I
I
t t L- 600 t c3
-n
L
400
t
I
CI0 b ?
?OO1 y 0.05). To begin to investigate the mechanisms that underlie the changes in I ca,T and 1ca,L described above, we analyzed the effects of neuraminidase on their activationand inactivation-voltage relationships. Figure 6 illustrates data for cells that showed an increase in current amplitude. For the activation of Ica T (Fig. 6A), VI,2 was shifted from -36.1 t 4.1 to -43.4 i 3.1 mV (P c 0.02; n = 5),
ALTERS
CA’+
CURRENTS
H739
but k was not changed significantly. For the inactivation (Fig. 6A), k was reduced from 4.2 t 0.4 to 3.3 t of ICa,T 0.1 mV (P < 0.05; n = 4), but V1iz was not changed significantly. None of the neuraminidase-induced changes in 1c,a,L activation or inactivation (Fig. 6B) were statistically significant. This was true regardless of whether curves measured 10 min after the addition of neuraminidase were compared with those measured before its addition or with those measured in untreated cells 10 min after initiating the whole cell recording. In those cells that exhibited a decrease in Ica,T (not illustrated), the only significant changes were an increase in the slope factor for the activation curve from 5.1 t 0.3 to 5.9 t 0.3 mV (P < 0.02; n = 4) and a shift of the inactivationcurve V1iz from -56.3 t 0.8 to -47.8 t 0.4 mV (P < 0.01; n = 3). In cells that did not exhibit an increase in 1ca,L, none of the changes were statistically significant. In untreated cells, we found no significant changes in the gating parameters for either Ica T or Ica,T,, even after lo20 min of rundown. Finally, time constants for the inactivation of Ica T at -40 mV (holding potential = -80 or -90 mV) or for the inactivation of IcaL at +10 mV (holding potential = -40 mV) were not affected significantly by neuraminidase, both in cells that exhibited an increase in calcium current and those that did not (Table 1) . Role of external Ca”+. To determine whether the increases in I ca,T and Ica,L were due to a direct effect on Tand L-type Ca2+ channels or to the greater influx of Ca2+ induced by neuraminidase (9, 23), we replaced the external Ca2+ with 2.5 mM Ba2’. Cells were rinsed twice with the Ba2+-containing bathing solution before 1.0 ml of that solution was added. Atomic absorption spectroscopy indicated that Ca2+ concentration was
FIG. 5. Effects of neuraminidase (1.0 U/ml) on current-voltage relationships for Ica (A and C) and lca,L (23 and 0). Controls currents (0, dotted curves) were measured just before addition of enzyme. Currents influenced by neuraminidase (m, solid curves) were measured between 5 and 12 min after its addition to bath. A: increases in Ica were significant at -60 and -40 mV (P < 0.05 and 0.02; n = 5). B: increases in I ca,~ were not significant at any potential (n = 3). C: reductions of I ca,T were significant at potentials between -30 and 0 mV (P < 0.05, 0.01, 0.01, and 0.02; n = 4). D: reductions of I ca,~ were not significant at any potential (n = 3). Paired t tests were used in all comparisons. Holding potentials were -80 or -90 mV in A and C and -40 mV in B and D.
ence of the gating of I ca,T;the rates of inactivation of the two currents were unchanged. These findings differ considerably from those obtained in guinea pig ventricular myocytes exposed to neuraminidase (34). In that study, only IcaT was affected; it was increased in 2550% of the cells, and the time constant for its inactivation was reduced. Two factors could have contributed to less hydrolysis of sialic acid residues in the former study. Yee and colleagues (34) used a lower concentration of neuraminidase (0.25 vs. 1.0 U/ml in the present study) and lower enzyme incubation temperatures (23-25 vs. 3637°C in the present study). Another important difference might be the greater number of ,&adrenergic receptors on nodal vs. ventricular myocytes, assuming that neuraminidase influences I ca,L indirectly by acting on these receptors (see below). Neuraminidase-induced changes in amplitudes of Ica,T At least three mechanisms could account for and &Lthe enhancement of I ca,T and Ica,L: 1) a direct effect of the enzyme on T- and L-type Ca2+ channels, such as changes in their conformations; 2) an indirect effect in which increased phosphorylation or dephosphorylation of the channels results from the rise in intracellular Ca2+ concentration ([ Ca2+];); and 3) an indirect effect me-
diated by other second messengers, such as CAMP or guanine nucleotide binding proteins (G proteins), following hydrolysis of sialic acid-containing receptors. A fourth mechanism, an increase in local extracellular Ca2+ concentration due to the release of Ca2+ bound to sarcolemma1 sialic acid, seemsunlikely, since such an increase would have reduced the negative external surface potential and thereby shifted the activation or inactivation curves to more positive potentials (20). The presence of sialic acid on P-cells of the sinoatrial node has been confirmed by the use of neuraminidase to remove ruthenium red, a cationic stain (32). But because of the small yield of cells from the rabbit sinoatrial node (1.0-3.5 x lo5 cells), we were unable to quantify the amount of sialic acid released by neuraminidase. As a rough estimate, we found previously that 1.0 U/ml of Sigma type X neuraminidase (36°C; 1 h) released 84.3 t 3.6% (n = 7) of the sarcolemmal sialic acid from embryonic chick ventricular myocytes (17). Also, Yee et al. (34) found that 0.12 U/ml of Worthington neuraminidase released 20% of the cellular sialic acid from guinea pig ventricular myocytes in 20 min. Still, three critical questions remain to be answered: 1) are sialic acid residues present on the glycosylated subunits of T- and L-type
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
ALTERS
CA’+
H741
CURRENTS
1. Time constants for inactivation of Ica,T and Ica,L are not influenced by neuraminidase TABLE
Neuraminidase
Control Condition
of Cells ms
Increased Ica,T NO increased Ica,T Increased 1ca,L NO increased 1ca,L
0.2
0.0
-80
-60
-40
-20
0
5 4
43.4t3.6
3
49.3t9.6
3
R
ms
n
R
0.96
13.2t2.9
5
0.98
0.98 1.00 0.99
13.1t3.7 38.522.9 54.8t6.4
4 3 3
0.96 1.00 0.96
Values are means t SE of time constants calculated by pClamp for single exponential fits to the first 100 ms of the decay of Ica,T at -40 mV (holding potential = -80 or -90 mV) and 1ca,L at +lO mV (holding potential = -40 mV). Currents were recorded before or after 10 min of exposure to 1.0 U/ml of neuraminidase. R indicates the goodness of fit by a single exponential, with 1.0 being a perfect fit. As determined by Student’s t test for paired variates, none of the differences were statistically significant (P > 0.05).
A
B
10.7t0.9 8.1zH.5
n
I
0.8
c
L ZJ
0.6
B 0.0
Membrane
Potent
ial
(mV>
k+&
6. Effects of neuraminidase (1.0 U/ml) on activation and inactivation of lca,~ (4 as described in Fig. and L,L (B). M easurements, 2, were made before (0, dotted curves) or 5-25 min after addition of neuraminidase (H, solid curves). Only cells that exhibited an increase in current (Fig. 5, A and B) were included. A: significant effects were a negative shift of lca,T activation curve from -36.1 to -43.4 mV (P < 0.02; n = 5) and a reduction of ilz for I ca,T inactivation curve from 4.2 to 3.3 mV (P < 0.05; n = 4). B: there were no significant effects of neuraminidase on gating of 1ca,L (n = 3). FIG.
Ca2+ channels; 2) did neuraminidase reach these sites; and 3) did hydrolysis of these sites alter channel function? Although cardiac Ca2+ channels have yet to be analyzed, skeletal muscle L-type Ca2+ channels do possesssialic acid residues on both the cy2-and y-subunits (29). The al-subunit, which constitutes the pore, is only lightly glycosylated (11,27), and its content of sialic acid has not been examined. A cardiac-specific isoform of the al-subunit has two additional potential extracellular Nglycosylation sites (26), but, again, the presence of sialic acid has yet to be investigated. Although it is possible that neuraminidase had a direct effect upon T- and L-type Ca2+channels in our cultured pacemaker cells (see below), it is unlikely that such an effect was responsible for the enhancement of IcaT , and Ica,L, since these currents were not increased when external Ca2’ was replaced by Ba2+ (Figs. 7 and 8), which also moves readily through T- and L-type Ca2+ channels. On
FIG. 7. Neuraminidase (1.0 U/ml) fails to enhance Ica,T (A) or lca,L (B) when external Ca2’ is replaced by Ba? A: current at -40 mV before (C) and 10 min after neuraminidase (N); holding currents were -82 (C) and -45 pA (N) at -90 mV. B: current at -5 mV before (C) and 10 min after neuraminidase (N); holding currents were -52 (C) and -35 pA (N) at -40 mV. Traces were superimposed so that holding currents overlapped. Peaks of capacitive currents are not illustrated.
the other hand, these experiments do suggest that the enhancement of I Ca,Tand Ica,L may depend upon external Ca2+. We and others have shown that when neuraminidase is used to remove sialic acid from the sarcolemma, 45Cauptake is enhanced in cultured ventricular cells (9, 23), and canine Purkinje fibers display depolarizing afterpotentials (13), spontaneous voltage fluctuations (22), and triggered activity (13, 32), processes that are promoted by elevated [ Ca2+];. Although its mode of entry is unknown, such an increase in [Ca2+]; could modulate the conductance of T- and/or L-type Ca2+ channels. Both Ca2+-induced potentiation (16) and Ca”‘-induced inhi-
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
H742
REMOVAL
OF
SIALIC
ACID
r
H
0 ti
20
‘.
.. .. ..
.. ..
0 1
0
I
1
Time
I
I
20
10
(mid
FIG. 8. Changes in Ica at -30 mV (HP = -90 mV; A) and lca,L at -15 mV (HP = -40 mV; B) when external Ca2+ is replaced by Ba? Cells were exposed to neuraminidase (1.0 U/ml) for times indicated (m), or currents were allowed to rundown without exposure to neuraminidase (0; elapsed time since initiation of whole cell recording). For cells exposed to neuraminidase, only those that exhibited an increase in current (3 cells) were included in the averages. Mean values were also determined for 2-4 controls. Individual results for each of 3 cells after 20 min (values without error bars) are also included. No changes induced by neuraminidase were significantly different from those due to rundown.
bition of IcaL (14, 31) have been observed following increases in lCa2+];, and we have seen the latter process in our cultured pacemaker cell preparation (21). Consistent with the apparent biphasic effects of neuraminidase on lca,T (Figs. 4 and 5, A and C), Tseng and Boyden (31) recently found that when [Ca2+]; was increased by dialysis of canine ventricular or Purkinje cells with EGTA and Ca2+,lca,T first increased and then decreased or lca,T increased in some cells and decreased in others. L-type Ca2+channels are also regulated by CAMP (30) and by a G protein, G, (33), both of which are able to mediate the stimulatory effects of agonist binding to ,& adrenergic receptors. Sialic acids contribute significantly to the carbohydrate content of ,0 receptors (28), which are known to be concentrated on nodal pacemaker cells. Thus the changes in I ca,Linduced by neuraminidase could have arisen indirectly from the removal of these sialic
ALTERS
CA"
CURRENTS
acid residues. Such removal could have altered the conformation of the receptors and resulted in stimulation or inhibition of adenylate cyclase and/or G,. This would seem to be unlikely, however, since neuraminidase treatment of guinea pig hearts did not alter the positive inotropic effect of isoproterenol, a ,&agonist (6). Moreover, coupling between these receptors and T-type Ca2+ channels has not been demonstrated in the sinoatrial node (10). Neuraminidase-induced changes in gating of Ica,T. A negative shift of the I ca,T activation curve (Fig. 6A ) may be partly responsible for the increased amplitude of IcaT following neuraminidase treatment (Fig. 5A ). Moreover, the increased overlap of the activation and inactivation curves (Fig. 6A), i.e., window current, might account for the rise in steady inward current following exposure to the enzyme (Fig. 3A). On the other hand, it is unlikely that the shift of the activation curve resulted from an increase in I ca,T or [Ca2+];. In theory, binding of Ca2+ to negative fixed charges on the inner surface of the sarcolemma could have accounted for the negative shift (20); however, this would have been unlikely because this (at most micromolar) Ca2’ would have had to compete with almost millimolar concentrations of Mg2+ already present in the cytosol(18). It is also unlikely that the negative shift occurred because sialic acid residues constitute a portion of the membrane’s external surface charge, since removal of that charge would have shifted the activation curve toward positive potentials (20). Taken together, these results suggest that neuraminidase may directly influence the gating of T-type Ca2+ channels. This idea is supported by the fact that neuraminidase altered the slopes of both inactivation (Fig. 6A) and activation (not illustrated) curves (for cells in which IcaT was increased or decreased, respectively) but had no’ effect on their half-inactivation and half-activation potentials. Thus channel components involved in gating, rather than membrane surface charge, appear to have been modified by the enzyme. In summary, the removal of sialic acid residues by neuraminidase altered the gating of IcaT and the amplitudes of both Ica T and Ica L. These effects appear to have been the result of both direct and indirect actions of the enzyme on T- and L-type Ca”+ channels. Additional experiments will be required to prove this hypothesis. We thank Theresa E. Redington for preparing the cultured pacemaker cells. This study was supported by National Heart, Lung, and Blood Institute Grant HL-20708. B. Fermini held a fellowship from the Canadian Heart Foundation. Present address of B. Fermini: Montreal Heart Institute, 5000 East Belanger, Montreal, Quebec HlT lC8, Canada. Address reprint requests to R. D. Nathan. Received
18 December
1989; accepted
in final
form
17 October
1990.
REFERENCES 1. BARRON, E. A., R. R. MARKWALD, AND R. D. NATHAN. Localization of sialic acid at the surface of embryonic myocardial cells. J. Mol. Cell. Cardiol. 14: 381-395, 1982. 2. BHATTACHARYYA, M. L., R. D. NATHAN, AND V. L. SHELTON. Release of sialic acid alters the stability of the membrane potential in cardiac muscle. Life Sci. 29: 1071-1078, 1981.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
REMOVAL
OF
SIALIC
ACID
3. CASSIDY, J. T., G. W. JOURDIAN, AND S. ROSEMAN. The sialic acids. VI. Purification and properties of sialidase from CZostridium perfringens. J. BioZ. Chem. 240: 3501-3506, 1965. 4. DRZENIEK, R. Substrate specificity of neuraminidases. Histochem. J. 5: 271-290,1973. 5. FABIATO, A. Computer programs for calculating total from specified free or free from specified total ionic concentrations in aqueous solutions containing multiple metals and ligands. Methods Enzymol. 157: 378-417, 1988. 6. FAWZI, A. B., AND J. H. MCNEILL. Effect of neuraminidase treatment on the inotropic response to ouabain, isoproterenol and calcium in the guinea pig heart. Eur. J. Pharmacol. 112: 295-311, 1985. 7. FERMINI, B., AND R. D. NATHAN. Do sialic acid residues account for surface charge modulation of T- and L-type calcium currents in cultured pacemaker cells from the sino-atria1 node? (Abstract). Biophys. J. 55: 299A, 1989. 8. FERMINI, B., AND R. D. NATHAN. Removal of sialic acid alters both T- and L-type Ca currents in cardiac myocytes (Abstract). Biophys. J. 57: 519A, 1990. 9. FRANK, J. S, G. A. LANGER, L. M. NUDD, AND K. SERAYDARIAN. The myocardial cell surface, its histochemistry, and the effect of sialic acid and calcium removal on its structure and cellular ionic exchange. Circ. Res. 41: 702-714, 1977. 10. HAGIWARA, N., H. IRISAWA, AND M. KAMEYAMA. Contribution of two types of calcium currents to the pacemaker potentials of rabbit sino-atria1 node cells. J. Physiol. Lond. 395: 233-253, 1988. 11. HOSEY, M. M., J. BARHANIN, A. SCHMID, S. VANDAELE, J. PTASIENSKI, C. O’CALLAHAN, C. COOPER, AND M. LAZDUNSKI. Photoaffinity labelling and phosphorylation of a 165 kilodalton peptide associated with dihydropyridine and phenylalkylamine-sensitive calcium channels. Biochem. Biophys. Res. Commun. 147: 11371145,1987. 12. ISENBERG, G., AND U. KL~CKNER. Calcium currents of isolated bovine ventricular myocytes are fast and of large amplitude. Pfluegers Arch. 395: 30-41, 1982. 13. KIMURA, S., S. HAZAMA, S. FUJII, H. NAKAYA, AND M. KANNO. Delayed afterdepolarisations and triggered activity in canine Purkinje fibres treated with neuraminidase. Cczrdiouasc. Res. 18: 294301,1984. 14. KOKUBUN, S., AND H. IRISAWA. Effects of various intracellular Ca ion concentrations on the calcium current of guinea-pig single ventricular cells. Jpn. J. Physiol. 34: 599-611, 1984. 15. LEVINSON, S. R., W. B. THORNHILL, D. S. DUCH, E. RECIO-PINTO, AND B. W. URBAN. The role of nonprotein domains in the function and synthesis of voltage-gated sodium channels. In: Ion ChunneZs, edited by T. Narahashi. New York: Plenum, 1990, vol. 2, p. 33-64. 16. MARBAN, E., AND R. W. TSIEN. Enhancement of calcium current during digitalis inotropy in mammalian heart: positive feed-back regulation by intracellular calcium? J. Physiol. Lond. 329: 589-614, 1982. 17. MCDONAGH, J. C., AND R. D. NATHAN. Sialic acid and the surface charge of delayed rectifier potassium channels. J. MOL. CeZZ. Curdiol. 22: 1305-1316,199O. 18. MURPHY, E., C. C. FREUDENRICH, L. A. LEVY, R. E. LONDON, AND M. LIEBERMAN. Monitoring cytosolic free magnesium in cultured chicken heart cells by use of the fluorescent indicator Furap-
ALTERS
CA2+ CURRENTS
H743
tra. Proc. NutZ. Acud. Sci. USA 86: 2981-2984, 1989. 19. NATHAN, R. D. Two electrophysiologically distinct types of cultured pacemaker cells from rabbit sinoatrial node. Am. J. Physiol. 250 (Heart Circ. Physiol. 19): H325-H329, 1986. 20. NATHAN, R. D. Negative surface charge: its identification and regulation of cardiac electrogenesis. In: Cardiac Muscle: The Regulation of Excitation and Contraction, edited by R. D. Nathan. Orlando, FL: Academic, 1986, p. 55-86. 21. NATHAN, R. D. Voltage and calcium regulation of calcium channels in pacemaker cells from the sinoatrial node (Abstract). Federation Proc. 45: 511, 1986. 22. NATHAN, R. D., AND M. L. BHATTACHARYYA. Membrane sialic acid and electrophysiology of cardiac Purkinje fibers (Abstract). Biophys. J. 33: 34A, 1981. 23. NATHAN, R. D., S. J. FUNG, D. M. STOCCO, E. A. BARRON, AND R. R. MARKWALD. Sialic acid: regulation of electrogenesis in cultured heart cells. Am. J. Physiol. 239 (Cell Physiol. 8): Cl97-C207, 1980. 24. SCHEUER, T., L. MCHUGH, F. TEJEDOR, AND W. CATTERALL. Functional properties of neuraminidase-treated rat brain sodium channels (Abstract). Biophys. J. 53: 54lA, 1988. 25. SCHMIDT, J. W., AND W. A. CATTERALL. Palmitylation, sulfation, and glycosylation of the cy subunit of the sodium channel: role of post-translational modifications in channel assembly. J. BioZ. Chem. 262: 13713-13723,1987. 26. SLISH, D. F., D. B. ENGLE, G. VARADI, I. LOTAN, D. SINGER, N. DASCAL, AND A. SCHWARTZ. Evidence for the existence of a cardiac specific isoform of the a1 subunit of the voltage dependent calcium channel. FEBS Lett. 250: 509-514,1989. 27. SOLDATOV, N. M. Purification and characterization of dihydropyridine receptor from rabbit skeletal muscle. Eur. J. Biochem. 173: 327-338,1988. 28. STILES, G. L., J. L. BENOVIC, M. G. CARON, AND R. J. LEFKOWITZ. Mammalian /3-adrenergic receptors: distinct glycoprotein populations containing high mannose or complex type carbohydrate chains. J. BioZ. Chem. 259: 8655-8663, 1984. 29. TAKAHASHI, M., M. J. SEAGAR, J. F. JONES, B. F. X. REBER, AND W. A. CATTERALL. Subunit structure of dihydropyridine-sensitive calcium channels from skeletal muscle. Proc. NutZ. Acud. Sci. USA 84: 5478-5482,1987. TRAUTWEIN, W., A. CAVALI~, V. FLOCKERZI, F. HOFMANN, AND D. PELZER. Modulation of calcium channel function by phosphorylation in guinea pig ventricular cells and phospholipid bilayer membranes. Circ. Res. 61, SuppZ. I: 1-17-I-23, 1987. 31. TSENG, G.-N., AND P. A. BOYDEN. Different effects of intracellular Ca and a phorbol ester on cardiac T and L Ca channels (Abstract). Biophys. J. 57: 314A, 1990. 32. WOODS, W. T., K. IMAMURA, AND T. N. JAMES. Electrophysiological and electron microscopic correlations concerning the effects of neuraminidase on canine heart cells. Circ. Res. 50: 228-239, 1982. 33. YATANI, A., J. CODINA, Y. IMOTO, J. P. REEVES, L. BIRNBAUMER, AND A. M. BROWN. A G protein directly regulates mammalian cardiac calcium channels. Science Wash. DC 238: 1288-1292, 1987. 34. YEE, H. F., JR., J. N. WEISS, AND G. A. LANGER. Neuraminidase selectively enhances transient Ca2’ current in cardiac myocytes. Am. J. Physiol. 256 (CeZZ Physiol. 25): Cl267-Cl272, 1989.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on November 18, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
