Electrophysiological demonstration of Na⁺/Ca²⁺ exchange in bovine articular chondrocytes

Abstract

Altered fluxes of Ca²⁺ across the chondrocyte membrane have been proposed as one pathway by which mechanical load can modulate cartilage turnover. In many cells, Na⁺/Ca²⁺ exchange (NCX) plays a key role in Ca²⁺ homeostasis, and recent studies have suggested it is operative in articular chondrocytes. In this study, an electrophysiological characterisation of NCX in articular bovine chondrocytes has been performed, using the whole-cell patch clamp technique, and the effects of inhibitors and the transmembrane electrochemical gradients of Na⁺ and Ca²⁺ on NCX function have been assessed. A Ni²⁺-sensitive current (I_NCX) which exhibited outward rectification, was elicited by a voltage ramp protocol. The current was also attenuated by the NCX inhibitors benzamil and KBR7943, without significant differences between the effect of these two compounds upon outward and inward currents. The Ni²⁺-sensitive current was modulated by changes in extracellular and pipette Na⁺ and Ca²⁺ in a manner characteristic of I_NCX. Measured values for the reversal potential differed significantly from those predicted for an exchanger stoichiometry of 3Na⁺: 1Ca²⁺, implying that accumulation of intracellular Ca²⁺ (from influx or release from stores) or more than one transport mode is occurring. These results demonstrate the operation of NCX in articular chondrocytes and suggest that changes in its turnover rate, as might occur in response to mechanical load, may modify cell composition and thereby dictate cartilage turnover.

1. INTRODUCTION

Chondrocytes are solely responsible for the synthesis and degradation of the articular cartilage matrix, [24, 35] and changes in their intracellular composition can modify turnover of the matrix [3, 18, 19, 23, 46]. Alterations in chondrocyte composition can be elicited in response to fluctuations in the physicochemical environment, which arise from variations in mechanical load during joint movements [47, 49]. In particular, these types of variables have been shown to modify the intracellular free calcium concentration ([Ca²⁺]_i) in chondrocytes (for example, [10, 11, 16, 28, 43, 50]).

In vertebrate cells, the regulation of [Ca²⁺]_i is achieved by a number of processes comprising membrane transport, cytoplasmic buffering and mobilisation from, and uptake into, intracellular stores. The operation of Na⁺/Ca²⁺ exchange (NCX) at the plasma membrane has been identified in both excitable cells (most notably, cardiac myocytes [6, 45]) and non-excitable cells as one of the principal mechanisms participating in this regulation. To date, three mammalian NCX isoforms have been identified, the ubiquitous NCX 1, and NCX 2 and NCX 3, confined to brain and skeletal muscle [6, 38, 40]. The exchanger can function in two ways, either moving Ca²⁺ out of cells (Ca²⁺ efflux or direct mode) or moving Ca²⁺ into cells (Ca²⁺ influx or reverse mode), depending on the prevailing electrochemical driving forces (determined by the Ca²⁺ and Na⁺ concentration gradients, and by the membrane potential [6, 40]). NCX most commonly mediates the exchange of three Na⁺ ions for one Ca²⁺ ion [4, 6, 9, 15, 22], thereby generating a current [45].

There is evidence for the Ca²⁺ homeostasis by NCX in porcine articular chondrocytes [39] and NCX has been shown to mediate the effects of extracellular alkalinisation on [Ca²⁺]_i in the C-20/A4 chondrocyte cell line [8]. Moreover, NCX mediates Ca²⁺ influx in bovine articular chondrocytes subjected to hyperosmotic challenge, and is the principal mechanism responsible for Ca²⁺ extrusion during the recovery of Ca²⁺ levels following the osmotic challenge [44]. Despite evidence for its operation, to date there have been no demonstrations of NCX protein expression in chondrocytes.

In the present study, NCX activity has been investigated in bovine articular chondrocytes, using electrophysiological techniques. The effects of classical inhibitors of this exchanger and of changes to intracellular and extracellular Ca²⁺ and Na⁺ concentrations have been determined. The findings demonstrate an electrogenic, Ni⁺-sensitive pathway modulated by Ca²⁺ and Na⁺, which is consistent with NCX activity in chondrocytes, and supports the role of NCX as an important determinant of [Ca²⁺]_i in these cells.

2. MATERIALS AND METHODS

2.1 Chemicals

Ouabain, verapamil, benzamil, KBR7943, tetraethylammonium chloride (TEA), 1,2-bis(2-aminophenoxy)ethane-N,N,N’,N’-tetraacetic acid (BAPTA), ethyleneglycol-bis(β-aminoethyl)-N,N,N’,N’-tetraacetic acid (EGTA), N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (HEPES), N-methyl-D-glucamine (NMDG) and Dulbecco’s modified Eagle medium (DMEM) were obtained from Sigma-Aldrich (Poole, UK). With the exception of NiCl₂, stock solutions of inhibitors were prepared in dimethylsulfoxide (when used, the final concentration of solvent was < 1% and had no detectable effects on membrane currents).

2.2 Isolation of chondrocytes

Cells were isolated from bovine articular cartilage, using methods described previously [7]. Briefly, cells were isolated by type I collagenase digestion (2000 U/ml in DMEM supplemented with 2 mM glutamine, 0.1 mM penicillin and 0.1 mM streptomycin for 18 h at 37 °C) from the metacarphophalangeal joint of 18-36 month-old cattle obtained immediately following slaughter. After filtration of the isolation solution, cells were stored (< 2h) in fresh supplemented DMEM at room temperature prior to experimentation. It is worth noting that chondrocytes isolated in the manner employed here represent cells harvested from different zones of cartilage, and that origin-dependent differences in cell behaviour will not be apparent using such pooled cell isolates.

2.3 Electrophysiological recording

Washed bovine articular chondrocytes were placed in a recording chamber attached to an inverted microscope (Nikon, Tokyo, Japan) and standard whole-cell tight-seal voltage-clamp protocols were used to record membrane currents [20]. All experiments were performed at room temperature (20-22 °C). Patch pipettes (Clark PG150T glass, from Harvard Apparatus (Kent, UK)) were pulled and polished on a DMZ-Universal Puller (Zeit-Instrumente, Augsburg, Germany) to resistances of 4-7 MΩ. After seal formation (5-15 GΩ), brief, strong suction was applied to the pipette interior to rupture the membrane patch and attain the whole-cell configuration.

Standard bath solution contained (mM): NaCl 140, CsCl 5, CaCl₂ 2, MgCl₂ 1, HEPES 15, glucose 10 (pH 7.4). Verapamil (100 μM) and ouabain (20 μM) were added to the bath solutions to inhibit voltage-activated Ca²⁺ channels and Na⁺/K⁺ ATPase, respectively. Standard pipette solution contained (mM): CsCl 110, NaCl 20, CaCl₂ 4, MgCl₂ 4, HEPES 10, TEA 20, glucose 5, BAPTA 20 (pH 7.1). Where appropriate, the free Ca²⁺ concentration within the pipette solution ([Ca²⁺]_p) was changed by altering the CaCl₂/BAPTA concentration ratio and [Ca²⁺]_p calculated using MAXCHELATOR software (Dr C. Patton, John Hopkins University, USA). The standard pipette solution had a calculated [Ca²⁺]_p of 102 nM. In some experiments, NaCl was replaced by NMDG or LiCl in equimolar concentrations.

Recordings were made using an Axopatch 200B amplifier and a CV203BU headstage, with voltage-commands generated and currents analysed using the pCLAMP software suite (Version 9, Axon Instruments Inc., USA). Current-voltage (I-V) relations were measured using experimental protocols similar to those developed previously [37]. Briefly, at 10 s intervals the cell was first depolarized from a holding potential (V_H) of 0 mV to +80 mV and then the membrane potential was ramped to −120 mV over the next 2 s, before returning to V_H. Cell capacitance was recorded before each experiment using the Membrane Test routine in pCLAMP and found to be 18 ± 2.5 pF (mean ± SEM, n = 117).

2.4 Statistics

Results are presented as mean ± SEM, where n denotes the number of individual cells tested. Each experimental observation was repeated in at least four cells from isolation batches of cartilage from at least three different animals. Differences between sample means were tested using the Student’s t-test (unpaired, two-tailed).

3. RESULTS

3.1 Ni²⁺-sensitive currents in bovine articular chondrocytes

Figure 1 shows the results of a typical whole-cell patch-clamp experiment on an isolated bovine articular chondrocyte, using the standard bath (external Na⁺ and Ca²⁺ concentrations, ([Na⁺]_o, [Ca²⁺]_o) of 140 mM and 2 mM, respectively) and pipette (having an Na⁺ concentration ([Na⁺]_p) of 20 mM and [Ca²⁺]_p of 102 nM) solutions. When the bath solution is deficient in K⁺ and to which inhibitors of voltage-activated Ca²⁺ channels and Na⁺/K⁺ ATPase have been added (see Methods), application of a hyperpolarising ramp from +80 to −120 mV (over 2 s) evoked a membrane current at both positive and negative potentials (a, Fig. 1A). The I-V curve generated by currents responses to this ramp command pulse rectified in the outward direction, with a measured reversal potential (E_r) near −40 mV (a, Fig. 1B).

Fig. 1.

Effect of Ni²⁺ on whole-cell currents in bovine articular chondrocytes. A: Typical current traces in the absence (i) and in the presence (ii) of 5 mM Ni²⁺ and the Ni²⁺-sensitive current obtained (iii) by subtracting (ii) from (i). B: I-V curve derived from the data in A.

Addition of 5 mM Ni²⁺ (a classical inhibitor of NCX; [30]) to the bath solution, abolished virtually all recorded current (b, Fig. 1A and B) except for a small residual outward component at positive membrane potentials. The I-V curves for control and Ni²⁺-sensitive currents were therefore almost identical (c, Fig. 1A and B), with essentially the same E_r values (see Table 1 for averaged Ni²⁺-sensitive current E_r data). Compared to the effects of Ni²⁺, Benzamil (500 μM) and KBR7943 (50 μM) only partially attenuated both inward and outward currents, again with no effect on E_r values (Table 1). At +80 mV, benzamil and KBR7943 inhibited outward current by 70 ± 2% and 73 ± 1%, respectively, while at −120 mV, benzamil and KBR7943 inhibited inward current by 81 ± 4% and 82 ± 3% (mean ± SEM; n = 4; no significant differences).

Table 1.

The effect of altering internal and external Ca²⁺ and Na⁺ concentrations and inhibitors on the E_r values of Ni²⁺-sensitive currents in bovine articular chondrocytes.

[Ca2+]p (nM)	[Ca2+]o (mM)	[Na+]p (mM)	[Na+]o (mM)	Measured Er (mV)	Drug	n	Predicted Er (mV)
102	2	20	140	−38 ± 3		12	−102
102	2	20	140	−37 ± 1	BZ	4	−102
102	2	20	140	−39 ± 2	KB	4	−102
102	2	20	140	−62 ± 5	TG	5	−102
0	2	20	140	*		4	†
102	2	20	140	−38 ± 3		12	−102
334	2	20	140	−39 ± 5		4	−72
540	2	20	140	−43 ± 4		4	−60
102	0	20	140	*		5	†
102	1	20	140	−34 ± 4		4	−85
102	2	20	140	−38 ± 3		12	−102
102	5	20	140	−66 ± 4		4	−125
102	2	0	140	*		4	†
102	2	5	140	−11 ± 2		4	+3
102	2	10	140	−41 ± 3		4	−50
102	2	20	140	−38 ± 3		12	−102
102	2	40	140	−84 ± 2		4	−154
102	2	20	0	*		7	†
102	2	20	25	‡		4	−232
102	2	20	50	−116 ± 6		4	−180
102	2	20	75	−92 ± 3		4	−149
102	2	20	100	−76 ± 2		4	−127
102	2	20	140	−38 ± 3		12	−102
102	2	5	0	*		4	†
102	2	5	25	−73 ± 1		4	−127
102	2	5	50	−39 ± 3		4	−75
102	2	5	75	−38 ± 2		4	−44
102	2	5	100	−31 ± 4		4	−23
102	2	5	140	−11 ± 2		4	+3

E_r values are shown as the mean ± SEM.

^*The E_r value could not be derived because the current was negligible.

^†The E_r calculation was not possible.

^‡The E_r measurement was not possible because the value was more negative than −120 mV and could therefore not be established. BZ, benzamil. KB, KBR7943. TG, thapsigargin. The last column shows predicted E_r values (assuming the current is produced by NCX with a 3Na⁺: 1Ca²⁺ stoichiometry).

3.2 Effect of Ca²⁺ on Ni²⁺-sensitive currents in bovine articular chondrocytes

If the Ni²⁺-sensitive, outwardly rectifying conductance described above is due to NCX, it would be expected to be modulated by alterations in internal and external Ca²⁺ concentrations. In particular, as reported in other cell types, if intracellular Ca²⁺ is buffered to very low levels, then exchanger turnover is abolished [2]. Figure 2 shows the effect of changing [Ca²⁺]_p on Ni²⁺-sensitive currents, while maintaining [Na⁺]_p, [Na⁺]_o and [Ca²⁺]_o at values of 20, 140 and 2 mM, respectively. The experiment shown in Fig. 2A shows that when a cell was dialysed with a pipette solution to which no additional Ca²⁺ had been added ([Ca²⁺]_p< 1 nM) only a very small Ni²⁺-sensitive-current was recorded. As [Ca²⁺]_p was increased to 102, 334 or 540 nM, inward and outward currents of increasing amplitudes were evoked. At [Ca²⁺]_p values of 102 nM and above, the outward currents saturated at potentials positive to about +20 mV (Fig. 2A and B), but inward currents increased continually with increases in [Ca²⁺]_p at all negative potentials tested (Fig. 2A and C). The effects of [Ca²⁺]_p on E_r are shown in Table 1.

Fig. 2.

Effect of [Ca²⁺]_p on the Ni²⁺-sensitive currents in bovine articular chondroytes. A: Typical I-V curves obtained at [Ca²⁺]_p of 0, 102, 334 and 540 nM as indicated. B: Relationship between [Ca²⁺]_p and the outward Ni²⁺-sensitive current measured at +80 mV. C: Relationship between [Ca²⁺]_p and the inward Ni²⁺-sensitive current measured at −120 mV. Data points represent mean ± SEM (n = 4).

Figure 3 illustrates the effects of changing [Ca²⁺]_o, while maintaining [Na⁺]_p and [Na⁺]_o at values of 20 and 140 mM, respectively, and [Ca²⁺]_p at 102 nM. With no Ca²⁺ added to the bath (chelating residual Ca²⁺ with 0.1 mM EGTA to < 10 nM) only a small Ni²⁺-sensitive inward current was recorded (Fig. 3A). As [Ca²⁺]_o was increased to 1, 2 or 5 mM, both inward and outward currents were evoked. Outward currents also increased with increasing [Ca²⁺]_o (Fig. 3A and B) and for [Ca²⁺]_o > 1 mM, inward currents saturated at potentials negative to −100 mV (Fig. 2A and C). The E_r values became increasingly negative as [Ca²⁺]_o was raised (Table 1).

Fig. 3.

Effect of [Ca²⁺]_o on the Ni²⁺-sensitive currents in bovine articular chondroytes. A: Typical I-V curves obtained at [Ca²⁺]_o of 0, 1, 2 and 5 mM as indicated. B: Relationship between [Ca²⁺]_o and the outward Ni²⁺-sensitive current measured at +80 mV. C: Relationship between [Ca²⁺]_o and the inward Ni²⁺-sensitive current measured at −120 mV. Data points represent mean ± SEM (n = 4).

3.3 Effect of Na⁺ on Ni²⁺-sensitive currents in bovine articular chondrocytes

Clearly, the Ni²⁺-sensitive currents recorded in bovine chondrocytes are modulated by changes in the trans-membrane Ca²⁺ gradient. In order to test whether the same is true for Na⁺, we conducted experiments of the type shown in Figure 4, where effects of changing [Na⁺]_p were investigated, while maintaining [Na⁺]_o and [Ca²⁺]_o at 140 mM and 2 mM, respectively, and [Ca²⁺]_p at 102 nM. When pipette Na⁺ was replaced completely by NMDG, virtually no membrane current was recorded (Fig. 4A). However, outward Ni²⁺-sensitive currents rose progressively as [Na⁺]_p was increased from nominally 0 mM to 10, 20 or 40 mM, with little or no saturation (Fig. 4A and B). At [Na⁺]_p values of 5 mM and above, the inward currents were saturated at potentials negative to −80 mV (Fig. 4A and C). The E_r values became increasingly more negative as [Na⁺]_p was raised (Table 1). Similar results were obtained when Li⁺ was substituted for Na⁺ (data not shown).

Fig. 4.

Effect of [Na⁺]_p on the Ni²⁺-sensitive currents in bovine articular chondroytes. A: Typical I-V curves obtained at [Na⁺]_p of 0, 10, 20 and 40 mM as indicated. B: Relationship between [Na⁺]_p and the outward Ni²⁺-sensitive current measured at +80 mV. C: Relationship between [Na⁺]_p and the inward Ni²⁺-sensitive current measured at −120 mV. Data points represent mean ± SEM (n = 4).

Figures 5A and B show the effects of changing [Na⁺]_o, while maintaining [Na⁺]_p and [Ca²⁺]_o at values of 20 and 2 mM, with [Ca²⁺]_p still at 102 nM. In the absence of external Na⁺, an outward Ni²⁺-sensitive current was still evoked, which increased and then saturated (at potentials positive to +50 mV) as [Na⁺]_o was increased to 50 mM and above. The E_r values became increasingly negative as [Na⁺]_o was reduced (Table 1). In an effort to elicit greater inward currents while altering [Na⁺]_o, a further set of experiments was performed as above, but with a [Na⁺]_p of 5 rather than 20 mM. Figures 5C and D show that in the absence of external Na⁺, no inward Ni²⁺-sensitive inward current could be measured in chondrocytes. However, inward Ni²⁺-sensitive currents rose progressively as [Na⁺]_o was increased to 25, 50, 75, 100, or 140 mM, with little or no saturation. Once again, the E_r values became increasingly negative at lower [Na⁺]_o (Table 1).

Fig. 5.

Effect of [Na⁺]_o on the Ni²⁺-sensitive currents in bovine articular chondroytes. A: Typical I-V curves obtained at [Na⁺]_o of 0, 75 and 140 mM as indicated. B. Relationship between [Na⁺]_o and the outward Ni²⁺-sensitive current measured at +80 mV. C: Typical I-V curves obtained at [Na⁺]_o of 0, 50, 100 and 140 mM as indicated, using a [Na⁺]_p of 5 mM. The latter enabled inward Ni²⁺-sensitive currents at negative potentials to be measured, which was not possible when using the standard [Na⁺]_p of 20 mM. D. Relationship between [Na⁺]_o and the inward Ni²⁺-sensitive current measured at −120 mV. Data points represent mean ± SEM (n = 4).

4. DISCUSSION

There is increasing evidence that the modulation of chondrocyte composition is a key route by which physicochemical changes induced by joint loading could exert their effects on cartilage turnover [46, 49]. A number of studies have illustrated that [Ca²⁺]_i can be altered by fluid flow, osmotic challenges, hydrostatic pressure and the like, by changes in Ca²⁺ transport pathway activity [28, 43, 44, 50]. In previous work, indirect evidence has been provided that altered NCX activity is in part responsible for the changes in [Ca²⁺]_i observed in response to hyperosmotic challenges [44]. In this study, electrophysiological recordings are presented of a trans-membrane current in chondrocytes, which has the distinct characteristics to be expected from that evoked by an electrogenic NCX transporter. It is likely that the currents which we have measured here undertestimate NCX activity: first, experiments were performed at room temperature, and NCX in other cells has been shown to be temperature-sensitive [30]; second, the ionised [Ca²⁺] in matrix interstitial fluid is likely to be higher (around 10mM) than the concentration (2mM) employed here.

The application of a hyperpolarising ramp from +80mV to −120mV elicited an outwardly-rectifying current with an E_r around −40mV, which could be almost entirely abolished by the application of Ni²⁺ (5mM). The I-V curve generated from these data exhibited a profile which was very similar to that observed for Na⁺/Ca²⁺ exchange current in other cell types (in heart, for example, [15, 29, 30]). Moreover, benzamil [26] and KBR7943 [25, 48], two inhibitors of NCX, also attenuated this current. These drugs inhibited both inward and outward components of the current, with no marked differences between their effects on positive and negative currents. This is in contrast to the notion that KBR7943 is a selective inhibitor of NCX Ca²⁺ influx mode [25, 48], but consistent with reports that have challenged this idea [31, 32]. In the heart, non-specific actions of KBR7943 have also been reported [42].

Under the conditions employed in the experiments presented here, measured E_r values of the Ni²⁺-sensitive currents deviated markedly from those to be expected from an exchanger operating with a stoichiometry of 3Na⁺:1Ca²⁺ [6, 36]. These discrepancies cannot be due to the effects of either additional Ni²⁺-insensitive membrane conductances or a significant background current, as residual currents following Ni²⁺ application were very small. Deviations can arise from the use of V_H values positive to the exchanger E_r. Under these conditions, the exchanger will be generating outward current and intracellular Ca²⁺ will accumulate, resulting in a measured E_r value that is more positive than the theoretical value calculated from the ionic content of the bath and pipette solutions [15, 37]. In addition, NCX stoichiometry may vary with [Na⁺]_i and [Ca²⁺]_i [17]. The situation may be even more complex in chondrocytes if, together with a major 3:1 transport mode, NCX imports 1 Na⁺ ion and 1 Ca²⁺ ion that defines a Na⁺-conducting mode exporting 1 Ca²⁺ ion, and also has an electroneutral Ca²⁺ influx mode that exports 2 Na⁺ ions, as has been reported in heart [27]. Finally, given that intracellular Ca²⁺ may increase at holding potentials positive to E_NCX, despite dialysis by the recording pipette, it is possible that raised cytosolic Ca²⁺ will induce release from intracellular Ca²⁺ stores. The observation that for cells pre-incubated for one hour in solutions containing 1 μM thapsigargin, experimental E_r was −96 ± 4 mV in 4 cells held at −100 mV - a value not significantly different from the calculated E_NCX - supports this notion.

The current exhibited an outward component dependent on [Ca²⁺]_o and [Na⁺]_i, which was increased with elevation in the concentrations of either ion. This response is typical of NCX and has been considered in detail in previous studies using other cell types [30, 33, 34]. The outward current was, however, also dependent on intracellular Ca²⁺ levels to some degree. It has been reported that Ca²⁺ influx mode (outward current) is dependent on non-transported intracellular Ca²⁺ ions in squid axons [12-14] and cardiac myocytes [21], although the [Ca²⁺]_i required for activation of NCX remains unclear: some studies have reported that it is in the low micromolar range [12, 21], while others have shown that NCX currents can be elicited at [Ca²⁺]_i in the nanomolar range [4, 34, 37]. In the experiments presented here, the measured currents were recorded when [Ca²⁺]_p was as low as 50 nM, but it was nevertheless entirely abolished when [Ca²⁺]_p was < 1 nM. The very small current which persisted is most likely to be the result of Ca²⁺ released from intracellular stores (as suggested by the action of thrapsigargin), which increases [Ca²⁺]_i above [Ca²⁺]_p. Activation in the low mM range is not a surprising finding because, under physiological conditions, [Ca²⁺]_i in articular chondrocytes remains at nM levels, even during responses to osmotic challenges [43, 44].

The data reported here showed that reducing [Na⁺]_o as low as 50 mM had no effects on outward currents. However; further reductions did decrease outward currents although, even in the nominal absence of external Na⁺, an outward current was still recorded. In cardiac myocytes, decreases in [Na⁺]_o correlate with increases in outward currents [34], except when the fall in [Na⁺]_o is extreme. Based on these findings, competitive inhibition between Na⁺ and Ca²⁺ ions at the external binding site of the exchanger has been proposed [12, 34, 41, 51]. In chondrocytes, no evidence was found for such an interaction between external Na⁺ and external Ca²⁺, since increases in [Na⁺]_o were not associated with increases in the outward component of the generated current and increases in [Ca²⁺]_o did not evoke increases in the inward component. Baker et al. [1] and Miura & Kimura [34] have shown that outward current generated by NCX is also reduced at low [Na⁺]_o in cardiac myocytes; these workers have suggested that either the Na⁺ substitutes employed (Li⁺ and NMDG⁺) facilitate Ca²⁺/Ca²⁺ exchange through NCX or that low levels of external Na⁺ have activating effects on the exchanger.

Within the range of [Na⁺]_o and [Ca²⁺]_p used in the present study, it was observed that increased concentrations of either ion produced proportional increases in the inward component of the generated current. These effects of [Na⁺]_o and [Ca²⁺]_p are characteristic of NCX and have been considered by a number of authors [6, 33, 34]. However, in the present study, the inward component was also dependent on the presence of Na⁺ in the pipette, although at [Na⁺]_p as low as 5 mM the current was apparent and was not increased as [Na⁺]_p was raised. It has been reported that intracellular Na⁺ activates NCX in squid axons at low concentrations [5, 14], which would be consistent with the findings here, although in chondrocytes the [Na⁺]_i required for activation would be expected to be lower than that in squid axons, given the finding that values as low as 5 mM elicit complete activation by intracellular Na⁺ of the inward component. In contrast, variations in [Ca²⁺]_o had no effect on the magnitude of this inward component, maximal currents being similar at the different [Ca²⁺]_o tested, a result which is also in agreement with the findings of DiPolo & Beaugé [14] in squid axon.

In conclusion, this study is the first electrophysiological demonstration of NCX operation in articular chondrocytes. Given that intracellular ionic composition is one determinant of cartilage matrix turnover, sensitivity of Ca²⁺ regulators such as NCX and the Ca²⁺-ATPase to components of mechanical load constitutes a potential mechanotransduction pathway and warrants further characterisation.