The Na+/H+ Exchanger Is Constitutively Activated in P19 Embryonal Carcinoma Cells, but Not in a Differentiated Derivative* RESPONSIVENESS TO GROWTH FACTORS AND OTHER STIMULI*

We have examined the functional properties and growth factor responsiveness of the plasma membrane Na+/H+ exchanger in pluripotent P19 embryonal car- cinoma (EC) cells and in a differentiated mesodermal derivative (MES-1) by analyzing the recovery of cy- toplasmic pH (pHi) from an acute acid load under bi-carbonate-free conditions. In the absence of exogenous growth factors, the mean steady-state pHi of undifferentiated P19 cells (7.49 f 0.03) is 0.55 unit higher than the value of differentiated MES-1 cells (6.94 f 0.01). In both cell types, recovery of pHi from an NHt-induced acid load follows an exponential time course and is entirely mediated by the amiloride-sen-sitive Na+/H+ exchanger in the plasma membrane. Ki- netic analysis indicates that the higher steady-state pHi in P19 EC cells is due to an alkaline shift in the pHi sensitivity of the Na+/H+ exchange rate, as compared to that in MES-1 cells. The Na+/H+ exchanger of MES-1 cells is responsive to epidermal growth factor, platelet-derived growth factor, serum, phorbol esters, and diacylglycerol, as shown by a rapid amiloride-sensitive rise in pHi of 0.15-0.35 unit. This mitogen-induced alkalinization is attributable

The Na+/H+ Exchanger Is Constitutively Activated in P19 Embryonal Carcinoma Cells, but Not in a Differentiated Derivative* RESPONSIVENESS TO GROWTH FACTORS AND OTHER STIMULI* (Received for publication, March 10, 1987) Arjo J. Bierman, Leon G . J. Tertoolen, Siegfried W. de  We have examined the functional properties and growth factor responsiveness of the plasma membrane Na+/H+ exchanger in pluripotent P19 embryonal carcinoma (EC) cells and in a differentiated mesodermal derivative (MES-1) by analyzing the recovery of cytoplasmic pH (pHi) from an acute acid load under bicarbonate-free conditions. In the absence of exogenous growth factors, the mean steady-state pHi of undifferentiated P19 cells (7.49 f 0.03) is 0.55 unit higher than the value of differentiated MES-1 cells (6.94 f 0.01). In both cell types, recovery of pHi from an NHt-induced acid load follows an exponential time course and is entirely mediated by the amiloride-sensitive Na+/H+ exchanger in the plasma membrane. Kinetic analysis indicates that the higher steady-state pHi in P19 EC cells is due to an alkaline shift in the pHi sensitivity of the Na+/H+ exchange rate, as compared to that in MES-1 cells.
The Na+/H+ exchanger of MES-1 cells is responsive to epidermal growth factor, platelet-derived growth factor, serum, phorbol esters, and diacylglycerol, as shown by a rapid amiloride-sensitive rise in pHi of 0.15-0.35 unit. This mitogen-induced alkalinization is attributable to an alteration in the pHi sensitivity of the exchanger. In contrast, the Na+/H+ exchanger of P19 EC cells fails to respond to any of these stimuli. Similarly, hypertonic medium rapidly activates the Na+/H+ exchanger in MES-1, but not in P19 EC cells . We conclude that the Na+/H+ exchanger in undifferentiated P19 EC stem cells is maintained in a fully activated state which is unaffected by extracellular stimuli, as if signal pathways normally involved in growth factor action are constitutively operative.
The molecular mechanisms underlying embryonic growth and development are largely unknown, although it seems likely that polypeptide growth factors and their receptorlinked signal pathways have a major role in mammalian embryogenesis (1). A suitable approach to studying embryonic growth control is to use in uitro model systems such as murine EC' cells, the undifferentiated stem cells of teratocarcinomas, * This work was supported by the Organization for the Advancement of Pure Research (ZWO) and by the Netherlands Cancer Foundation (Koningin Wilhelmina Fonds). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
which share many properties with the inner cell mass cells of the pre-implantation embryo (2,3). Some EC cells are able to differentiate, either in vivo or in vitro, into a variety of nontumorigenic cell types that derive from the three primitive germ layers (2, 3).
Differentiation of EC cells is accompanied by drastic changes in growth properties, as has been demonstrated for the pluripotent mouse P19 EC cell line. Undifferentiated P19 EC cells exhibit a highly transformed phenotype and grow rapidly in the complete absence of exogenous polypeptide growth factors (4, 5). Upon differentiation these cells lose their transformed properties and their ability to proliferate autonomously. The differentiated derivatives show anchorage-dependent, contact-inhibited growth and are dependent on the supplementation of exogenous growth factors for sustained proliferation (5, 6). Thus, the P19 EC cell line and its derivatives provide a useful system for studying (autonomous) growth control mechanisms, likely to be analogous to those in early mouse embryogenesis, and in particular for identifying the cellular signaling mechanisms involved.
As a first step in elucidating such potential signal pathways, we have characterized and compared the properties of the plasma membrane Na+/H+ exchanger in both P19 EC cells and in a mesodermal derivative (MES-1) in bicarbonate-free media. It is generally accepted that the Na+/H+ exchanger not only has a major housekeeping role in pHi regulation but also may function as a signal transducer in the action of growth factors in that it mediates a rapid and persistent rise in pHi in stimulated cells (7). There is increasing evidence that this mitogen-induced cytoplasmic alkalization has a permissive effect on the initiation of protein and DNA synthesis, at least in fibroblastic cells maintained in bicarbonate-free media (8-10) . We report here that, while both P19 EC and MES-1 cella have a normally functioning Na+/H+ exchanger in terms of pHi regulation, the exchanger of MES-1 cells is highly responsive to growth factors and other agonists, whereas the Na+/H+ exchanger in P19 EC cells is in a permanently activated state, resulting in a much higher steady-state pHi value and in a complete lack of responsiveness to extracellular stimuli. Our results suggest that certain signal pathways normally utilized by mitogens to activate the Na+/H+ exchanger in responsive somatic cells are constitutively active in autonomously growing EC stem cells. This could represent a critical mechanism involved in early embryonic growth control. Sweden). Other agents were obtained from the following sources: dimethylamiloride from Merck Sharp and Dohme; monensin from Lilly; nigericin, phorbol esters, and diacylglycerol from Sigma.
Electrophysiology-Cells were grown in 35-mm dishes (gelatinized) and electrophysiological methods identical with those described previously were used (14,15  At a constant pH, of 7.35, the mean steady-state pH, of P I 9 EC cells, maintained in serum-free HEPES-buffered DMEM for at least 10 h, is estimated at 7.49 f 0.03 (means f S.E.; n = 7). By contrast, the steady-state pH; value of MES-1 cells under identical conditions is substantially more acidic (6.94 f 0.01; n = 31). Since the mean membrane potential of both P19 EC and MES-1 cells is in the 50-60 mV range (interior negative), the steady-state pHi in both cell types is well above the electrochemical equilibrium value of - 6.4 predicted by the Nernst equation, which demonstrates the presence of an H+-extruding mechanism that regulates 0, pHi. Further insight into the nature of this pHi-regulating system can be obtained by studying the recovery of pHi from a sudden cytoplasmic acidification (16, 17).

pH, Recovery from Acidification Is Mediated by Na+/H+ Exchange-
To examine the pH,-regulating mechanism(s) in both P19 EC and MES-1 cells, the cells were acid-loaded by the NH: prepulse method (12,16,17) under HCO;/C02-free conditions. Addition of 15 mM NH&1 causes an immediate rise in pHi of 0.4-0.6 unit, due to influx and subsequent protonation of the weak base NH3. The passive entry and dissociation of the less permeant weak acid NH: follows, causing pHi to decrease slowly. Removal of external NH: then evokes an acute cytoplasmic acidification, as NH3 leaves the cell immediately and an excess of protons is trapped intracellularly (Fig. 2). The rapid fall in pHi is followed by a spontaneous, exponential recovery which is 90% complete within 4-5 min in both P19 EC ( Fig. 2A)  ; i l-/-c ( Fig. 2 I ) ) . It is seen that in P19 EC cells, there is an initial, rather slow phase of pHi recovery, which lasts for about 30 s; this initial slowing is not observed in MES-1 cells (cf. Fig. 2, A and D). We have not investigated the mechanism(s) underlying this phenomenon, but we note that the initial portion of the pH, recovery curve is unavoidably contaminated by the continuing efflux of NHs from the cells, which tends to slow the recovery process. Since P19 EC cells probably have a smaller surface:volume ratio than MES-1 cells (6), it seems likely that complete washout of NHs from P19 EC cells requires more time, which would explain the initial slowing of the pH, recovery in these cells' (Fig. 2A). When external Na' is replaced with choline, pHi does not recover in either cell type (Fig. 2, B and E ) . Furthermore, pHi recovery is rapidly and reversibly blocked by the Na'/H+ exchange inhibitor amiloride (1 mM) and its more potent analogue &methylamiloride (50 PM) (Fig. 2, C and F ) .
From these results we conclude that pH; recovery from cytoplasmic acidification in P19 EC and MES-1 cells, as in other vertebrate cells (16-18), is mediated by a Na+/H+ exchange system in the plasma membrane.
Intracellular Buffering Capacity-From the initial shifts in pHi produced by exposing the cells to NH,Cl, an estimate can * In curve fitting procedures, we chose to disregard the pHi recovery time points within -1 min of NH, washout. The initial slowing of pH, recovery is not observed in the Na+ re-addition experiment of Fig. ZB, nor when pH, recovers from an acetate-induced acid load (not shown).
be obtained of the intracellular buffering capacity (p), defined as the amount of OH-that has to be added intracellularly to raise pHi by 1 unit (1, 3). In 12 such experiments, the value of /3 was calculated to be -10 mM/pH for P19 EC and 17 mM/ pH for MES-1 cells, in reasonable agreement with previously measured values in mammalian cells (16, 17).
Kinetic Analysis of pHi Recovery-To obtain further insight into the kinetic properties of the Na+/H+ exchanger in both cell types, we calculated the rate constants of exponential pHi recoveries using an iterative curve fitting procedure to fit the pHi time courses to the equation where k is the rate constant, t is the time, and pHi(m) is pHi at the new asymptotic steady state. Typical exponential curve fittings are illustrated in Fig. 3, A and B. The mean rate constants of exponential recovery were calculated to be 0.57 & 0.02 min" (mean f S.E.; n = 4) for P19 EC cells and 0.72 zk 0.04 min" ( n = 9) for MES-1 cells. As expected from the marked differences in the resting pHi of both cell types, the value of pH,(m) of P19 EC cells is -0.5 unit more alkaline than that of MES-1 cells.
. As in other systems (13,16,17,19) and in agreement with the exponential time courses of pHi recovery, the relationship between the rate of pHi recovery, d(pHi)/dt, and the value of pHi was found to be linear for both P19 EC and MES-1 cells (Fig. 3C). As is seen in Fig. 3C, the function d(pHi)/dt versus pHi for P19 EC cells is markedly shifted along the pHi axis Initial points of pH; recovery were disregarded to avoid artifacts from NHs efflux. Dashed curues are single exponentials obtained from an iterative curve fitting procedure. C, dependence of pH, recovery rate on pH; in both cell types as indicated. Data points were obtained from 32 different experiments in which pHi was allowed to recover from NH;-induced acid loads. At constant external pH (7.35) values of ApHJmin were calculated from the net rise in pH, over 10-s intervals. The data points were fit to straight lines by the linear regression method. Recovery rates extrapolate to zero at the normal resting pHi of the cells (6.94 for MES-1 and 7.50 for P19 EC cells).
towards alkaline values when compared to that for MES-1 cells. In other words, since Na+/H+ exchange activity is proportional to d(pHi)/dt, the Na+/H+ exchanger in P19 EC cells appears to be more sensitive to cytoplasmic H+ than in MES-1 cells. The Na+/H+ exchange rates extrapolate to 0 at pHi 6.97 0.03 for MES -1 cells and at pHi 7.46 k 0.03 for P19 EC cells, the normal resting pHi of these cells. Fig. 3C also shows that the relationship between recovery rate and pHi for MES-1 cells has a steeper slope than that for P19 EC cells. This difference cannot simply be ascribed to differences in buffering power (which is lower in P19 EC than in MES-1 cells), but may reflect differences in V, , , for Na+/H+ exchange in both cell types.
Effects of Growth Factors and Protein Kinase C Activators on pH, in MES-1 Cells-Growth factors and agents that stimulate protein kinase C, such as phorbol esters and cellpermeable diacylglycerols, induce a rapid activation of the Na+/H+ exchanger in their target cells leading to an increase in the steady-state pH;. We measured the effects of various growth factors and known kinase C activators on the pHi of P19 EC and MES-1 cells using the fluorimetric technique. As shown in Fig. 4, addition of FCS, EGF, phorbol esters, or diacylglycerols to MES-1 cells leads to a substantial cytoplasmic alkalinization, up to 0.3 pH unit above the initial steady-state pH;. In most cases, there is a lag of 30-60 s, followed by a pH; rise to a new steady state within 8-10 min. A similar pH, response was observed with PDGF (25 ng/ml), although a small initial acidification was consistently observed (not shown). Table I (second column) summarizes the mean increases in pH; for MES-1 cells in response to various stimuli.
The induced alkalinizations are not observed in Na+-free media (see Fig. 5 for a typical example with EGF and TPA) and the pH; shifts are completely inhibited by 1 mM amiloride (data not shown). Fig. 5 illustrates that stimulation of the cells in the absence of Na' followed by a shift to normal medium results in alkalization without a detectable lag, suggesting that the events that take place during the lag period are independent of Na+ influx into the cells. From these results we conclude that the increases in steady-state pHi are mediated by the Na+/H+ exchanger, as in other stimulated cells.
P19 EC Cells Fail to Raise Their Resting pH; in Response to Extracellular Stimuli-In marked contrast to MES-1 cells, the P19 EC cell line fails to increase its pHi above the resting value when treated with growth factors or with phorbol esters and diacylglycerols (Table I, first column). The lack of effect of EGF on pHi probably has a trivial explanation, since P19 EC cells have very few EGF receptors (-900 receptors per cell; Ref. 6). However, the number of specific phorbol ester binding sites in P19 EC cells is relatively high and comparable with that of MES-1 cells and other differentiated derivatives of P19 EC cells (20). Yet, TPA and diacylglycerol are completely ineffective in raising pH; in P19 EC cells, in contrast to the findings with these agents in MES-1 cells (Table I).
In many cells, the Na+/H+ exchanger can also be activated by exposing the cells to hypertonic medium (21)(22)(23). This osmotic activation of the Na+/H+ exchanger appears to play an important role in volume regulation and occurs via an as yet unidentified pathway, apparently not involving protein kinase C (22). In view of the finding that the Na+/H+ exchanger of P19 EC cells remains unresponsive to activators of kinase C, we undertook pH; measurements to test whether the exchanger of these cells can be stimulated by hypertonic media. For comparison, pHi was also measured in osmotically activated MES-1 cells.
Increasing the osmolarity of the medium from 300 to 450 mosM by addition of sucrose or KC1 results in a rapid activation of the Na+/H+ exchanger in MES-1 cells, as indicated by an amiloride-sensitive rise in pHi of -0.25 unit; again, no pHi shift is observed in P19 EC cells, even when the osmolarity of the medium is increased by 200 mosM (Table I). Yet, the inability of P19 EC cells to raise their pH; in response to any extracellular stimulus is apparently not due to a limited driving force for the Na+/H+ exchanger (given by the magnitude of the transmembrane Na' and H+ gradients), since the steady-state pH, in P19 EC cells can readily be increased by addition of the Na+/H+ ionophore monensin (5 PM; Table I).
Effects of Lowering Extracellular pH;-Considering the relatively high steady-state pHi value of P19 EC cells and in view of the lack of response to external stimuli, it is conceiv- in Na+-free medium until pH; stabilized at a new resting value. EGF (100 ng/ml) and TPA (100 ng/ml) were added where indicated. Na+free medium was replaced by normal medium ( 140 mM Na+) by rapid perfusion.
able that activation of Na'/H' exchange in these cells does take place, but fails to evoke a detectable alkaline pHi shift simply because the resting cytoplasmic H' concentration is too low. In order to test this possibility, we lowered the steadystate pHj of both P19 EC and MES-1 cells by -0.3 unit by exposing the cells to a pH 6.80 medium; subsequently we monitored the pH; of such acidified cells in response to EGF, TPA, and hypertonicity. Fig. 6, A and B , illustrates the response of MES-1 cells to EGF and TPA, at a constant pH, of 6.80. As can be seen, the alkalization induced at pH 6.80 is somewhat larger than that recorded at pH, 7.35 (cf. Fig. 4) /ml) ( B ) . C, effect of pH, 6.80 on pHi in P19 EC cells and lack of pHi response to treatment with TPA (100 ng/ml).
quantitatively similar result was obtained in two experiments using 200 mosM sucrose as a stimulus (not shown).
In contrast, P19 EC cells remain completely unresponsive to TPA (Fig. 6C) and hypertonic medium (not shown) when the initial pHi value is lowered. Thus, there is no evidence that activation of Na+/H+ exchange in P19 EC cells is undetectable because of the high resting pHi. Instead, our observations consistently suggest that, unlike MES-1 cells, P19 EC cells have a permanently activated Na+/H+ exchanger that cannot be stimulated further. Kinetic Analysis of pH; Recovery in Activated MES-1 Cells-In order to gain insight into the possible mechanism(s) by which growth factors and phorbol esters activate Na+/H+ exchange in MES-1 cells, we analyzed the kinetics of pHi recovery from an NHZ-induced acid load. Typical time courses of pHi recovery from acidification ( Fig. 7) show that stimulated MES-1 cells have a significantly higher recovery rate than untreated control cells. Furthermore, the value at which pHi stabilizes, i.e. the new resting pHi, is increased by at least 0.2 pH unit. For cells treated with TPA, we determined the relationship between pHi recovery rate (d(pHi)/dt) and pHi; in unstimulated cells, this relationship is linear over the pHi range 6.2-7.0 (Fig. 7, open symbokr; see also Fig. 3), with the recovery rate extrapolating to zero at the resting pHi of about 6.97. When cells are stimulated with TPA, the dependence of the pHi recovery rate on pHi is shifted to the right, implying that the H+ extruding mechanism, i.e. the Na+/H+ exchanger, has acquired a greater sensitivity to cytoplasmic H+. As shown in Fig. 7, the TPA-induced shift towards more alkaline pHi values is not simply a parallel one: the increased pHi recovery rate manifests itself as an increase in the slope of the linear regression line by about 50% (from 0.78 for control cells to 1.24 after TPA treatment). Similarly, growth stimulants such as EGF and FCS imitate the effects of TPA in that they induce not only a shift in the pHi dependence of the exchanger, but also an increase in the slope of the pHi dependence curve (by 27 and 72%, respectively; data not shown).

DISCUSSION
In this study we have examined and compared the properties of the Na+/H+ exchanger in mouse P19 embryonal carcinoma cells, being undifferentiated, phenotypically transformed pluripotent stem cells, and in the differentiated mesodermal derivative MES-1 which has a nontransformed phe- notype. While the rapidly proliferating P19 EC stem cells have no requirements for exogenous growth factors, the P19derived MES-1 cells are growth factor-dependent showing a significant mitogenic response to serum, EGF, PDGF (6) and TPA? We, therefore, also compared the responsiveness of the Na+/H+ exchanger in both cell types to various extracellular stimuli.
The results of the present study demonstrate that the Na+/ H' exchanger is present in both cell types and functions normally in terms of pHi regulation, independent of the state of differentiation. However, the most important observation is that the exchanger in P19 EC cells cannot be activated by a variety of extracellular stimuli; by contrast, in MES-1 cells the exchanger is highly sensitive to growth factors, phorbol esters, and hypertonic medium as previously described for many somatic cells (reviewed in Refs. 7 and 24). Taken together, our observations strongly suggest that the exchanger in P19 EC cells is already in a fully activated state, as if the C. L. Mummery, unpublished results. intracellular signal pathways normally involved in Na+/H+ exchange activation are constitutively turned on. The following lines of evidence lend support to this notion: ( a ) the steady-state pHi in PI9 EC cells is relatively high (-7.50) when compared to the value in MES-1 cells (-6.95); (6) pH, fails to decrease even after prolonged deprivation of exogenous growth factors; ( c ) the relationship between Na+/H+ exchange activity (plotted as pH, recovery rate in Fig. 3) and pHi is shifted to more alkaline values, in a manner similar to that observed in MES-1 cells when activated by growth factors or phorbol esters (cf. Fig. 7C); that is, the pH; sensitivity of the exchanger in PI9 EC cells corresponds to that of activated MES-1 cells; ( d ) pHi cannot be raised further by any of several extracellular stimuli tested (serum, phorbol esters, hypertonicity, etc.); and ( e ) artificially lowering the resting pHi potentiates stimulus-induced alkalinization in MES-1 cells, but not so in PI9 EC cells.
Our results further show that the MES-1 cell line is a convenient model system to investigate growth factor regulation of Na+/H+ exchange. We and others have reported that extracellular stimuli activate the Na+/H+ carrier through an apparent increase in the affinity for cytoplasmic H+ to a more alkaline pH, without changing the apparent Na' affinity or Vmax (12,19,21,25). In other studies of Na+/H+ exchange in myoblasts (26), serum stimulates exchange activity by increasing both the internal H+ affinity and the Vmax for Na+ uptake. Our present data on the kinetics of pHi recovery in MES-1 cells (Fig. 7), demonstrating both an alkaline shift in pHi dependence and an accelerated recovery at acidic pHi, suggest that it is likely that mitogens raise steady-state pHi not only by increasing the internal H+ affinity of the Na+/H+ exchanger but also by enhancing its V,,,. Isotopic Na+ uptake studies should help to test this hypothesis.
A major question, to which we have no answers yet, is what mechanism underlies the apparent constitutive activation of the Na+/H+ carrier in P19 EC cells. There is good evidence that protein kinase C can activate, either directly or indirectly, the Na+/H+ exchanger in most cell types (27-29). It is therefore conceivable that kinase C is permanently activated in P19 EC cells, thus keeping the Na+/H+ exchanger in its fully activated state. However, it has recently been shown that treatment of P19 EC cells with phorbol esters such as TPA drastically reduce kinase C activity in the soluble cell fraction, with a substantial increase in the particulate fraction when compared to untreated control cells in a manner very similar to that observed in TPA-stimulated MES-1 cells and many other cell types.4 Furthermore, TPA is an inducer of c-myc protooncogene expression4 and of morphological alterations in P19 EC cells (20). Together, these results make it less likely that kinase C in P19 EC cells is constitutively activated, suggesting that a different pathway is responsible for the fully activated state of the Na+/H+ exchanger in PI9 EC cells. Interestingly, alternative modes of activation, not involving kinase C, appear to be involved in the action of EGF and hyperosmolarity on Na+/H+ exchange in fibroblasts (30) and lymphocytes (31), respectively. The biochemical nature of these activation pathways remains to be determined.
It remains possible, of course, that the observed permanent activation of Na+/H+ exchanger in P19 EC stem cells is attributable to the endogenous production and secretion of growth factors acting via membrane receptors on the signal pathway(s) that lead(s) to Na+/H+ exchange activation. Such autocrine mechanisms might be essential in early embryogen-' W . Kruijer, Habets; G., van Genesen, S., Feijen, A., and Mummery, C., manuscript in preparation. esis, where rapid proliferation of pluripotent cells is required. Whether the present results on P19 EC cells can be explained by the "autocrine secretion" hypothesis is yet to be established.
Finally, it is noteworthy that constitutive activation of Na+/ H+ exchange is not a typical characteristic of autonomously growing tumor cells in general: the exchanger in epidermoid carcinoma cells (32), HeLa cells, and neuroblastoma cells (27), Ehrlich cells (33), and virally transformed fibroblasts' can be activated to some extent by various agonists. Future studies should reveal whether a permanently activated Na+/ H+ exchanger and an elevated resting pHi are hallmarks of pluripotent embryonal carcinoma and embryonic stem cells in general. It seems plausible to assume that by keeping Na+/ H+ exchange permanently activated, cells maintain their steady-state pHi in a range permissive for protein and DNA synthesis (8, 34). Another challenge for further research is to investigate whether constitutive activation of the Na+/H+ exchanger and, hence, an alkaline resting pHi has a regulatory role in the growth and differentiation characteristics in germ cell tumors and early embryogenesis.