Induction of Glycinebetaine Uptake into Xenopus Oocytes by Injection of Poly(A)+ RNA from Renal Cells Exposed to High Extracellular NaCl*

Madin-Darby canine kidney (MDCK) cells accumu- late glycinebetaine via Na+-dependent transport in response to hypertonic stress. When extracellular tonic-ity is increased by the addition of NaCl, V,,, for gly- cinebetaine transport increases without an associated change in K,, consistent with an increase in the number of functioning transporters. To test whether increased transport activity results from increased gene expression, we injected poly(A)+ RNA (mRNA) from MDCK cells into Xenopus oocytes and assayed for glycinebetaine uptake in ouo. RNA-induced Na+-depend- ent uptake is observed in oocytes injected with mRNA from cells exposed to high extracellular NaC1, but not in oocytes injected with either water or mRNA from cells maintained in isotonic medium. Unfractionated mRNA induces glycinebetaine uptake in ouo at a rate which is -3-fold higher than in water-injected controls. Size-fractionated mRNA (median size 2.8 kilo- bases) induces uptake at a rate which is -7-fold higher than controls. Such RNA-induced transport activity in ouo is consistent with heterologous expression of Na+/ glycinebetaine cotransporters encoded by


Glycinebetaine' accumulates in renal cells both in vivo
and in vitro (Nakanishi et al., 1988) when they are exposed to high extracellular NaC1. Accumulated glycinebetaine can function as both 1) a compatible or nonperturbing solute which can balance high extracellular salt concentration and 2) a counteracting osmolyte which can offset the destabilizing effects of urea on macromolecular function (Yancey et al., 1982;Yancey and Burg, 1990). Its * 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. ll To whom reprint requests should be addressed Bldg. IO, Rm.
The term betaines actually refers to a class of compounds consisting of fully N-methyl-substituted amino acids (see Wyn Jones and Storey, 1981). By popular usage, however, the terms betaine and glycinebetaine (2-(N,N,N-trimethylamino)acetate) are frequently used interchangeably. Older names for glycinebetaine such as lycin, oxyneurin, and glykokollbetaim are no longer employed. accumulation has been characterized in several cultured renal cell lines including Madin-Darby canine kidney (MDCK)* cells (Nakanishi et al., 1988).
MDCK cells in hypertonic culture medium accumulate high levels of intracellular glycinebetaine by sodium-dependent uptake (Nakanishi et al., 1988). This uptake occurs via both high-and low-affinity Na+/glycinebetaine cotransporter sites which exhibit apparent Michaelis constants (K,) of 0.12 and 5.6 mM, respectively . When medium osmolality is increased by the addition of NaC1, glycinebetaine transporters exhibit an increase in maximum transport velocity (V,,,) but no apparent change in K,,,. One possible explanation for these changes is that the absolute number of transporters is increased. This could result from translation of elevated numbers of transcripts encoding transporters, provided mRNA levels increase in response to hypertonicity. Transcriptional control of osmoregulation has been demonstrated for the accumulation of glycinebetaine in other systems as phylogenetically diverse as enteric bacteria (see Csonka, 1989) and spinach (Weretilnyk and Hanson, 1990), and there is evidence supporting such control for other osmolytes found in mammalian renal medullary tissue. For example, sorbitol accumulation in renal medullary cells both in vivo (Yancey and Burg, 1989) and in vitro (Bagnasco et al., 1987) in response to high extracellular NaCl results from increased aldose reductase gene expression Moriyama et al., 1989;Cowley et al., 1990;Smardo et al., 1990). Similar osmoregulation has been suggested for myo-inositol accumulation by Na+/myo-inositol cotransport in cultured renal cells (Kwon et al., 1991).
To test whether hypertonicity might increase glycinebetaine transport via induction of mRNAs specific for glycinebetaine transporters, we measured Na+-dependent glycinebetaine uptake in oocytes from Xenopus laevis (the South African clawed toad) after injection with heterologous poly(A)+ RNA from MDCK cells grown in either isotonic or high salt medium. We chose this system because Xenopus oocytes can translate heterologous mRNA from a variety of sources (Gurdon et al., 1971) and can carry out the posttranslational modifications and targeting necessary for the active expression of native function in ouo (Soreq, 1985). Xenopus oocytes have been employed to express a variety of Na+-dependent membrane transporters encoded by exogenous mRNA (Hediger et al., 1987;Aoshima et al., 1988;Longoni et al., 1988;Sigel et al., 1988;Hagenbuch et al., 1990). Oocytes have also been used to demonstrate changes in gene expres-* The abbreviations used are: MDCK, Madin-Darby canine kidney; kb, kilobase(s1. sion for the rat pituitary thyrotropin-releasing hormone receptor (Oron et al., 1987), the rat thyroid Na+/I-symporter (Vilijn and Carrasco, 1989), rat cortical neurotrophic factor (Duchemin et al., 1990), and the canine renal Na+/myoinositol cotransporter (Kwon et al., 1991).

Lack of Native Na+-dependent Glycinebetaine Uptake in
Ovo- [ l-l4C]Glycinebetaine uptake by Xenopus oocytes was measured in the presence of either 100 mM NaCl (Na+containing uptake buffer) or 100 mM LiCl (Na+-free uptake buffer). In preliminary experiments, glycinebetaine uptake into uninjected Xenopus oocytes increased linearly for 60 min in the presence of sodium and was not significantly different from that observed in the absence of sodium (data not shown). The rate of this uptake was relatively low, with some seasonal variation, and did not differ from that observed in waterinjected controls (data not shown). Mean basal uptake by water-injected oocytes in 45 experiments performed between July 1988 and September 1990 was 1.39 +-0.74 pmol/oocyte/ h (mean +-S.D.).

Poly(A)+ RNA from Cells Exposed to High Extracellular NaCl (but Not from Cells Maintained in Isotonic Medium) Induces Increased Glycinebetaine Uptake When Injected into
Oocytes-RNA was isolated from MDCK cell monolayers 21-22 h after medium osmolality was acutely raised from 310 to 515 mOsm via the addition of NaC1. Parallel cultures maintained in isotonic medium (310 mOsm) were extracted at the same time. Polyadenylated RNA was isolated from each sample (mRNA pair A), and oocytes were microinjected with 50 nl of ultrapure water containing either 0 or 20 ng of poly(A)+ RNA. Four days after injection, [l-14C]glycinebetaine uptake by oocytes was measured in the presence of 100 mM NaC1. When injected with poly(A)+ RNA from osmotically stressed MDCK cells (515 mOsm), oocytes from three different toads exhibited a marked increase in glycinebetaine uptake which was 2-or 3-fold higher than that observed in either waterinjected controls or oocytes injected with poly(A)+ RNA from cells maintained in isotonic medium (Fig. lA). Poly(A)+ RNA was isolated from three additional pairs of cultured cell monolayers exposed to either isotonic or hypertonic medium, and paired poly(A)+ RNA samples (mRNA pairs B-D) were injected into oocytes from a single animal. We observed RNAinduced glycinebetaine uptake in all oocytes injected with RNA from cells grown in hypertonic medium. No such activity was observed in oocytes injected with RNA from cells maintained in isotonic medium (Fig. 1B).

RNA-induced Glycinebetaine Uptake in Ovo Is Na+-dependent-
The Na+ dependence of RNA-induced glycinebetaine uptake in oocytes was evaluated by substitution of equimolar amounts of LiCl for NaCl in the uptake buffer. RNA-induced glycinebetaine uptake was seen only in oocytes injected with poly(A)+ RNA from cells grown in hypertonic medium and was noted only in the presence of sodium (Fig. 2). In the absence of sodium, uptake in oocytes injected with RNA from cells exposed to high extracelluar NaCl was not significantly different from that observed in oocytes injected with either water or RNA from cells grown in isotonic medium. Neither the presence nor the absence of sodium had any effect upon '' Portions of this paper (including "Materials and Methods" and Figs. A-F) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press. the basal glycinebetaine uptake seen in water-injected controls or in oocytes injected with RNA from cells grown in isotonic medium. oocytes injected with 1.7-4.9-kb size-fractionated poly(A)+ RNA had higher uptake rates than oocytes injected with other fractions or water (Robey et al., 1990). Similarly, we sizefractionated the paired poly(A)+ RNA used in Fig. L4 (i.e. mRNA pair A) using a sucrose gradient and tested individual fractions for their ability to induce uptake in ovo. Oocytes were injected with 50 nl of ultrapure water containing either 0 or 20 ng of poly(A)+ RNA from individual fractions, and [l-14C]glycinebetaine uptake into individual oocytes was measured four days later (Fig. 3, closed circles). A fraction with a median size of approximately 2.8 kb was found to induce the greatest uptake rate, approximately 7-fold higher than that found in water-injected controls (12.70 & 0.87 versus 1.80 & 0.08 pmol/oocyte/h). In contrast, identical size-fractionated poly(A)+ RNA from cells maintained in isotonic medium did not result in an in ouo glycinebetaine uptake rate significantly different from that observed in water-injected controls (1.83 f 0.38 versus 1.80 k 0.08 pmol/oocyte/h). We have also observed expression of transport activity in oocytes injected with size-fractionated poly(A)+ RNA of similar size from a renal papillary cell line, PAP-HT25 cells (data not shown), when grown under hypertonic conditions shown previously to induce the accumulation of glycinebetaine (Moriyama et al., 1990).

RNA Samples Enriched for Poly(A)+ RNA Species between 1.7 and 4.9 kb in Length (Median Size 2.8 kb) Contain the Message(s) Responsible for the RNA-induced Increase in Glycinebetaine Uptake Observed in
RNA-induced Uptake of Both Glycinebetaine and myo-lnositol Can Be Observed in the Same Oocytes, but These Uptakes Are Induced by RNA Species of Slightly Different Size-To test the ability of our poly(A)+ RNA to induce myo-inositol uptake, we simultaneously measured [1-l4C]glycinebetaine and my0-[2-~H]-inositol uptake rates in oocytes four days after microinjection with size-fractionated poly(A)+ RNA from mRNA pair A. RNA-induced myo-inositol uptake (Fig.   3, closed circles) was observed in oocytes injected with several RNA fractions noted to induce glycinebetaine uptake (Fig. 3, open circles). Maximal myo-inositol uptake, however, was noted in oocytes injected with a slightly larger RNA fraction than that associated with peak glycinebetaine uptake (median size approximately 3.5 kb for myo-inositol versus 2.8 kb for glycinebetaine uptake).

DISCUSSION
High extracellular salt induces glycinebetaine accumulation in the mammalian renal medulla as well as in a variety of phylogenetically diverse organisms including prokaryotes (Csonka, 1989), higher plants (Wyn Jones and Storey, 1981), and marine invertebrates (Yancey et al., 1982). In all systems studied thus far, this accumulation occurs by one of two mechanisms: 1) via uptake from the environment or 2) via the oxidation of choline.
In species which synthesize glycinebetaine, choline is oxidized by choline dehydrogenase (EC 1.1.99.1) and betainealdehyde dehydrogenase (EC 1.2.1.8). Although choline oxidase activity in MDCK and PAP-HT25 cells has not been studied, these cells do not accumulate glycinebetaine under hypertonic conditions when grown in choline-containing glycinebetaine-free medium Moriyama et aL, 1990). This suggests that glycinebetaine transport, and not synthesis, plays a major role in the glycinebetaine accumulation observed in these cells when exposed to high extracellular salt.
Choline oxidase activity has been demonstrated in mammalian kidneys (Wirthensohn and Guder, 1982;Grossman and Hebert, 1989), but there is no evidence that renal medullary choline oxidase is osmoregulated. Na+-dependent glycinebetaine transport has been demonstrated in membrane vesicles prepared from both rabbit renal cortex (Wright and Wunz, 1989) and rat renal medulla (Seifter et al., 1990), but these transporters have not been sufficiently characterized to determine their relatedness to each other or to those observed in cultured renal cells. Given the apparent lack of osmoregulated choline oxidase and the presence of Na+-dependent glycinebetaine uptake in the medulla, it is likely that osmoregulated medullary accumulation of glycinebetaine occurs via increased uptake in vivo.
The osmoregulated accumulation of glycinebetaine in renal tissue shares several characteristics with that of sorbitol. For example, both osmolytes accumulate in response to the same extracellular solutes (e.g. in response to elevated raffinose4 or NaCl, but not to elevated glycerol4 or urea; see Moriyama et al., 1990). Also, both osmolytes appear to be interchangeable to some extent (Moriyama et al., 1990;Moriyama et al., 1991). Changes in intracellular ionic strength serve as a signal for the accumulation of sorbitol in renal cells (Uchida et al., 1989) T. Moriyama

Glycinebetaine Transporter Expression in
and, by analogy, may trigger glycinebetaine accumulation in the kidney. In Escherichia coli, however, elevated potassium glutamate, rather than increased ionic strength, per se, may serve as the signal for glycinebetaine accumulation via the ProU transport system, a well described prokaryotic glycinebetaine transport system which requires the presence of a specific periplasmic binding protein and, hence, the direct hydrolysis of ATP for its activity (Prince and Villarejo, 1990). Nakanishi et al. (1990) established the sodium dependence of MDCK cell glycinebetaine transporters by substituting equimolar amounts of LiCl for NaCl in the uptake buffer. The corresponding activity in ouo exhibits similar discrimination between lithium and sodium. Choline-for-sodium substitutions in ouo yield identical results (data not shown) but were not employed in the present study because of chemical similarities between choline and glycinebetaine which might ultimately complicate interpretation of the data.
It is likely that oocytes translate functional transporters from injected renal cell mRNA. In principle, however, translation of mRNAs encoding regulators or stimulators of native oocyte transporters could also account for our findings. Two lines of evidence argue against such a possibility: 1) The baseline glycinebetaine uptake observed in uninjected oocytes is Na'-independent and is consistent with passive diffusion down its concentration gradient. 2) Native transport activity in ouo is not stimulated by increasing the osmolality of the oocyte uptake buffer from 200 to 300 mOsm by the addition of NaCl (data not shown).
A small amount of Na+-dependent glycinebetaine transport activity is detected in MDCK cells maintained in isotonic medium , and yet there is little or no expression of transporter activity in oocytes injected with the corresponding mRNA. This could reflect either an absence of glycinebetaine transporter mRNA or mRNA levels in the isotonic condition which are below the lower limit of detection for our bioassay system. In either case, it is possible that transporter activity resides in a stable pool of membrane transporters with little turnover and, hence, minimal requirements for additional translational template under isotonic conditions.
Oocytes injected with size-fractionated RNA enriched for 1.7-4.9-kb species consistently took up more [l-14C]glycinebetaine than those injected with other fractions or water. The greatest uptake activity was seen in oocytes injected with an RNA fraction with a median size of approximately 2.8 kb.
Myo-Inositol uptake was also observed in these oocytes, but peak uptake was observed in oocytes injected with a slightly larger fraction of the same RNA sample. Peak myo-inositol uptake was associated with a poly(A)+ RNA fraction having a median size of approximately 3.5 kb in the present study and approximately 4 kb in a previous study (Kwon et al., 1991). These results suggest that the mRNAs encoding Na+/ glycinebetaine cotransporters may be smaller than those encoding Na+/myo-inositol cotransporters.
We conclude that the induction of glycinebetaine accumulation in renal cells by high extracellular salt probably involves increased gene expression, most likely of a gene or group of genes responsible for specific mRNAs (approximately 3 kb in size) encoding Na"dependent glycinebetaine transporters. The exact nature of this osmoregulated process remains to be elucidated.

Supplemalal M a d to "lndlrtionOlGlyciMbaaineUplakelnloXmoprrOmeylaBy In)ccte(lonOfPoly(A)'RNAFmmRatalCellr
ExpasedTo High E I m I I u I a r NPC'I.'