Vol. 16, No. 1, 1992 HEPATOLOGY Elsewhere 271

PUMPING IONS: REGULATION OF INTRACELLULAR pH IN HEPATOCYTES

Benedetti A, Strazzabosco M , Corasanti JG, Haddad P, GrafJ, Boyer JL. C 1 ~ -HCO, exchanger in isolated rat hepatocytes: role in regulation of intracellular pH. Am J Physiol 1991;261:G512-G522.

ABSTRACT

In rat hepatocytes, basolateral Na+-H + exchange and Na+-HC0,- cotransport function as acid ex- truders. To assess mechanisms of acid loading, intra- cellular pH (pHi) recovery from an alkaline load was analyzed in short-term cultured rat hepatocyte mono- layers using the pH-sensitive dye BCECF. Electrophys- iological techniques were also used to assess the role of the membrane potential (V,). Cells were alkaline loaded by suddenly reducing external CO, and HC0,- (from 10% and 50 mM, respectively, to 5% and 25 mM) at constant pH. After this maneuver, pH, rapidly rose by 0.13 k 0.03 pH units (p&) and recovered to baseline at an initial rate of 0.026 -t 0.009 pHJmin. Intracel- lular buffering power was estimated from the depen- dence of pH, on [NH,-l, and varied between 70 and 10.5 mM/p& in a pH, range of 6.5-7.6. Initial pH, recovery corresponded to a rate of OH- efflux (JoH) of 1.76 k 0.71 mM/min and was blocked by 0.5 mM DIDS (0.003 & 0.002; JOH = 0.18 f 0.06) or by 1 mM H,DIDS (0.001 f 0.002; JoH = 0.26 f 0.08) and by removal of [Cl-1, (0.003 k 0.007; JOH = 0.28 k 0.07). The depen- dence of JOH on [Cl-1, exhibited saturation kinetics with an apparent K, for [Cl-1, of 5.1 mM. pHi recovery was Na+ independent and was not inhibited by substitution of Na+ with NMDG (0.045 -C 0.09; JoH = 2.94 & 0.59). During an alkaline load, cell V, hyperpolarized from -33.4 +- 1.8 to -43.4 -+ 2.8 mV, mainly due to an increase in K+ conductance by a factor of 2.8 k 0.3. Ba2+ blocked these changes and depolarized V, by 12.1 2 1.2 mV but had no effect on pH, recovery (0.025 2 0.007; JOH = 1.46 f 0.60), ex- cluding that V, functions as a mqjor force for HC0,- extrusion after an alkaline load. The rate of JoH was directly proportional to the pH, reached after the alkaline load and varied fourfold over the pH, range of 7.26-7.46. These data indicate that pH, recovery from an alkaline load in rat hepatocytes is mediated by an electroneutral Na+ -independent Cl--HCO,- ex- changer.

COMMENTS The plasma membrane of hepatocytes contains

several specialized proteins that transport H + and HCO, - into and out of the cell to meet rapidly changing cellular demands. These proteins contribute to a variety of functions, including regulation of cell volume and bile secretion, but they appear to be most important for the regulation of intracellular pH (pHi). Under normal conditions a continuously varying rate of intracellular H + production occurs as a result of changes in cell metabolism and uptake of organic acids, but pHi remains constant between -7.20 and -7.30. Mainte- nance of pHi in this range is critical because many key enzymes are pH sensitive. A fall of only 0.10 pH unit inhibits synthesis of glucose by -70% and urea by

Na+ H +

,&+ -

FIG. 1. Model for regulation of pHi in hepatocytes. Cellular transport mechanisms that appear to contribute to regulation of pHi in hepatocytes include ATP-dependent Na+/K+ pump (11, Na/H+ exchange (21, Na + IHCO, - cotransport (31, C1- /HCO, - exchange (41 and membrane K + (5) and C1- (6) conductances as described in the text.

- 50% (1). Consequently, intracellular H + generation must be matched on a minute-to-minute basis by the efflux of H + or the uptake of HC0,- to prevent progressive intracellular acidification.

Most previous studies in hepatocytes have focused on the mechanisms responsible for recovery from intracel- lular acidosis. Distribution of H + ions across the membrane according to electrochemical equilibrium would result in a pHi of - 6.90. Consequently, transport of H + out of the cell to maintain pHi in the physiological range represents an “uphill” or thermodynamically unfavorable process. H + and HC0,- transport pro- teins are energy transducers that capture the potential energy available in transmembrane Na + or C1- gra- dients to drive H + and HCO, - transport. Functionally, efflux of H + or influx of HCO, - are equivalent in their effects on pHi, but binding sites on transporters are highly specific.

One model for regulation of pHi in hepatocytes is shown in Figure 1. In the basolateral (sinusoidal) membrane, continuous turnover of the Na + /K + pump driven by hydrolysis of ATP maintains intracellular Na + - 10 mmolb and contributes to the generation of the interior negative membrane potential difference. This results in the large electrochemical gradient of - 105 mV (based on extracellular Na + concentration of 140 mmoVL, intracellular Na+ concentration of - 10 mmoVL and membrane potential difference of - 40 mV) favoring influx of Na+ ions. Two Na+-dependent transporters, Na + iH + exchange and Na + /HCO, - cotransport, are also in the basolateral membrane and play an important role in the regulation of pHi. Although the molecular mechanisms involved are not known, these proteins presumably span the plasma membrane. Na+/H+ exchange involves binding of Na+ to an extracellular site and H + to an intracellular site,

272 HEPATOLOGY Elsewhere HEPATOLOGY

followed by a conformational change that translocates Na + in and H + out of the cell. Similarly, Na + /HCO, - cotransport involves binding of Na+ and HC0,- to extracellular sites followed by inward translocation of both ions. In each case, the energy available in the Na + gradient is used to drive “uphill” movement of H + or HC0,- and alkalinize the cell interior.

In the absence of HCO, - , Na + /H + exchange is the principal mechanism responsible for recovery from an acid load (2) and exhibits an apparent set point near - 7.20 (3). Na+/H + exchange is inactive under basal conditions but increases after an acid challenge, re- sulting in the restoration of pHi toward this set point. In the presence of HCO, - , Na + /HCO, - cotransport also contributes to recovery from an acid load (4). However, comparatively little is known about this transporter. Unlike Na + /H + exchange, which is electroneutral, Na + /HCO, - cotransport appears to be electrogenic with coupling of more than one HCO, - to each Na + (5). Consequently, each transport cycle results in net movement of negative charge. HCO, - transport through this pathway may be regulated in part by the membrane potential difference, but this has not yet been established experimentally. In addition, the capacity of this mechanism relative to Na + /H + exchange is un- known. However, HCO, - associated Na + influx ap- pears to be active at physiological values of pHi where Na+/H+ exchange is inactive (61, and HC0,- de- pendent recovery of pHi after intracellular acidosis equals or exceeds that attributable to Na+/H+ ex- change (4, 7). Indeed, coupled movement of Na+ and HCO, - represents a major determinant of Na + influx in hepatocytes. Because Na+/K+ pump activity is sensitive to intracellular Na + concentration, the rise in intracellular Na+ associated with HCO, - influx represents a challenge to intracellular homeostasis. About half of hepatic Na+/K+ pump activity ap- pears dedicated to recycling Na+ entering in con- junction with HC0,- to maintain intracellular Na+ concentrations in the physiological range (6). Although these transporters are driven primarily by thermo- dynamic ion gradients, examination of other cell types suggests that transporter activity is also likely to be regulated by intracellular messengers such as Ca2+, cyclic AMP and related kinases, resulting in alterations in the number, turnover or set point for transport.

This interesting and careful study by Benedetti and coworkers evaluates the cellular response to intracel- lular alkalosis, introduces an important role for C1- /HCO, - exchange and adds several new pieces to the puzzle of hepatocyte pH regulation (8). C1- /HCO, - exchangers have been widely recognized in many cell types, where they contribute to regulation of pHi, cell volume and HCO, - secretion, depending on the tissue studied. C1 -/HCO, - exchange previously has been demonstrated in the liver only in membrane vesicles isolated from the canalicular region (9). These studies demonstrate for the first time that electroneutral C1-/HCO,- exchange is active in intact hepatocytes

and mediates recovery from an intracellular alkaline load.

For these studies, pHi was measured using the pH-sensitive fluorochrome 2’ ,7’-bis(carboxyethyl)-5(6’)- carboxyfluorescein under control conditions and during recovery from an alkaline load. A variety of ion substi- tutions were used to determine the ionic dependence of changes in pHi. The authors provide compelling evi- dence for electroneutral C1- /HCO, - exchange in that recovery is C1- dependent, Na + independent, inhibited by the stilbene derivative 4,4’-diisothiocyanostilbene- 2,2’-disulfonic acid in concentrations known to inhibit C1-/HCO,- exchange and increased by higher intra- cellular HCO, - concentrations. These studies further illustrate that exchange shows half-maximal activity at - 5 mmoUL C1- and functions to acidify the cytoplasm through the efflux of HCO, - .

One important conclusion of these studies is that C1- /HCO, - exchange activity is closely coupled to pHi. At basal pHi, C1- /HCO, - exchange was inactive, but as pHi increased to 7.46 a fourfold increase occurred in HC0,- efflux consistent with the activation of transport by intracellular alkalinization. This is an interesting parallel to Na + /H + exchange, which is also quiescent at basal pHi but increases with intracellular acidification. Increased C1- /HCO, - exchange is pre- sumably driven by increased intracellular HCO, - con- centration, but a decrease in intracellular C1- concen- tration could have similar effects. Because intracellular C1- concentration is near electrochemical equilibrium, membrane hyperpolarization associated with intracel- lular alkalosis (as demonstrated in these and other studies) would be expected to decrease intracellular C1- concentration (10) so that an increased inward C1- gradient and outward HC0,- gradient might work in parallel to speed recovery from alkalosis. However, inhibition of membrane K + conductance by Ba2 + had no apparent effect, implying that transport activity is more sensitive to changes in intracellular HCO,- concentration. Whether C1- /HCO, - exchange is also regulated by intracellular messengers as has been demonstrated in other cell types is not established.

Another important contribution of this paper is the detailed evaluation of intracellular buffering capacity, an important but overlooked component of the cellular defense against challenges to pHi. Buffering capacity, expressed in micromoles per pH unit, represents the H + equivalents required to produce a defined change in pHi in the absence of membrane H + or HCO, - transport, and cells with high buffering capacity are relatively resistant to changes in pHi. The authors demonstrate that buffering capacity is strongly dependent on both pHi and intracellular HC0,- concentration. In the absence of HCO, - , buffering capacity increased markedly from - 12 mmol/L/pH unit at pHi - 7.30 to - 60 mmoVL/pH unit at pHi - 6.50. In the presence of HCO, - , buffering capacity showed a more complex dependence on pHi, but it was generally stable at -57 mmol/L/pH unit at pHi -7.30.

In aggregate, these studies provide a convincing

Vol. 16, No. 1, 1992 HEPATOLOGY ElseLuhtw 273

demonstration that C1- IHCO, - exchange activity is present in intact hepatocytes and may be the principal mechanism responsible for recovery from intracellular alkaline challenge. These studies also provide a strong foundation for future investigation of related questions regarding the overall contribution of C1- IHCO, - ex- change to other cellular functions including biliary HCO, - secretion and bile formation. Unlike Na + /H + exchange and Na + MCO, - cotransport, C1- /HCO, - exchange is located in the canalicular membrane. These studies predict that intracellular alkalosis would be followed by rapid alkalinization of canalicular bile and that transport activity would be sensitive to C1- in the canalicular rather than the extracellular space. Reg- ulation by factors other than C1- and HC0,- con- centration that alter the “set point” or rate of transport including H + ions and Ca2’ - or cyclic AMP-dependent signaling will also be of great interest. Finally, it will be important to determine how regu- lation of C1-MC0,- exchange is integrated with other mechanisms of H + and HCO, - transport. It is attractive to speculate that a balance exists between HCO, - influx through Na + /HCO, - cotransport and efflux through C1- MCO, - exchange. Because these transporters are located in separate membrane do- mains, coregulation might contribute to transcellular movement of HC0,- from the sinusoidal to the can- alicular space and the regulation of pHi.

GREG Frm, M.D. Department of Medicine Duke University Medical Center Durham, North Carolina 27710

REFERENCES

1. Kashiwagura T, Deutsch CJ, Taylor J, Erecinsha M, Wilson DF. Dependence of gluconeogenesis, urea synthesis, and energy me- tabolism of hepatocytes on intracellular pH. J Biol Chem 1984;

2. Henderson RM, Graf J, Boyer JL. Na ‘ -H ’- regulated intracellular pH in isolated rat hepatocyte couplets. Am J Physiol 1987;252:

3. Renner EL, Lake JR, Scharschmidt BF. Na--H+ exchange activity in hepatocytes: its role in regulation of intracellular pH. Am J Physiol 1988;256:G44-G52.

4. Gleeson 0, Smith ND, Boyer JL. Bicarbonate dependent and independent pH regulatory mechanisms in rat hepatocytes. J Clin Invest 1989;84:312-321.

5. Fitz JG, Persico M, Scharschmidt BF. Electrophysiologic evidence for Na+ -coupled HCO, - transport in cultured rat hepatocytes. Am J Physiol 1989;256:G491-G500.

6. Fitz JG, Lidofsky SD, Weisiger RA, Xie M-H, Cochran M, Grotmol T, Scharschmidt BF. HC0,-coupled Na+ influx is a major determinant of Na + turnover and Na + /K + pump activity in rat hepatocytes. J Membr Biol 1991;122:1-10.

7. Fitz JG, Lidofsky SD, Xie M-H, Cochran M, Scharschmidt BF. Plasma membrane H+/HCO,- transport in rat hepatocytes: a principal role for Na+ -coupled HCO, - transport [in press]. Am J Physiol 1991;24.

8. BenedettiA, StrazzaboscoM, Corasanti J , Haddad P, GrafJ, Boyer JL. CI--HCO- exchanger in isolated rat hepatocytes: role in regulationofintracellularpH. Am J Physiol1991;261:G512-G622.

9. Meier PJ, Knickelbein R, Moseley RH, Dobbins JW, Boyer JL. Evidence for a carrier-mediated chloride-bicarbonate exchange in

269:237-243.

G109-G113.

canalicular rat liver plasma membrane vesicles. J Clin Invest

10. Fitz JG, Scharschmidt BF. Intracellular C1- activity in intact rat liver: relationship to membrane potential and bile flow. Am J Physiol 1987;252:G699-G706.

1985;83: 1225-1235.

PROVOCATIVE GENE THERAPY STRATEGY FOR THE TREATMENT OF

HEPATOCELLULAR CARCINOMA Huber BE, Richards CA, Krenitsky TA. Retroviral- mediated gene therapy for the treatment of hepatocel- lular carcinoma: an innovative approach for cancer therapy. Proc Natl Acad Sci USA 1991;88:8039-8043.

ABSTRACT

An approach involving retroviral-mediated gene therapy for the treatment of neoplastic disease is described. This therapeutic approach is called “virus- directed enzymelprodrug therapy” (WEPT). The WEPT approach exploits the transcriptional differ- ences between normal and neoplastic cells to achieve selective killing of neoplastic cells. We now describe development of the WEPT approach for the treatment of hepatocellular carcinoma. Replication-defective, amphotrophic retroviruses were constructed con- taining a chimeric varicella-zoster virus thymidine kinase (VZV TK) gene that is transcriptionally regu- lated by either the hepatoma-associated wfetoprotein or liver-associated albumin transcriptional regulatory sequences. Subsequent to retroviral infection, ex- pression of VZV TK was limited to either a-fetoprotein- or albumin-positive cells, respectively. VZV TK metabolically activated the nontoxic prodrug 6-methoxypurine arabinonucleoside (araM), ulti- mately leading to the formation of the cytotoxic anab- olite adenine arabinonucleoside triphosphate (araATP). Cells that selectively expressed VZV TK became selectively sensitive to araM due to the VZV TK-dependent anabolism of araM to araATP. Hence, these retroviral-delivered chimeric genes generated tissue-specik expression of VZV TK, tissue-specific anabolism of araM to araATp, and tissue-specific cyto- toxicity due to araM exposure. By utilizing such retro- viral vectors, araM was anabolized to araAW in hep atoma cells, producing a selective cytotoxic effect.

COMMENTS The rapid advance in gene transfer technology has

fueled speculation with regard to future human appli- cations. This article defines a gene therapy strategy for the treatment of HCC in patients and then thoroughly explores its feasibility in vitro. The strategy exploits two well-described phenomena, which are as follows: (a) the ability of tissue-specific regulatory elements to selec- tively drive gene expression in different types of cells and (b) the ability of the varicella-zoster virus thymidine kinase (TK) enzyme to metabolically activate the prodrug 6 methoxypurine arabinonucleoside (araM) to adenine arabinonucleoside triphosphate (ardTP). The prodrug, araM, is not toxic, but araATP is toxic to the