Proc. Nati. Acad. Sci. USAVol. 88, pp. 7820-7824, September 1991Medical Sciences

Accumulation of glycerophosphocholine (GPC) by renal cells:Osmotic regulation of GPC:choline phosphodiesterase

(NaCI/urea/phospholipase A2/MDCK cells/osmolytes/"glyceryl phosphoryl choline")

K. ZABLOCKI*, S. P. F. MILLERt, A. GARCIA-PEREZ*, AND M. B. BURG**Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung and Blood Institute, and tDevelopmental and Metabolic Neurology Branch,National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD 20892

Contributed by M. B. Burg, May 21, 1991

ABSTRACT Although GPC has long been recognized as adegradation product of phosphatidylcholine, only recently isthere wide appreciation of its role as a compatible and coun-teracting osmolyte that protects cells from osmotic stress. GPCis osmotically regulated in renal cells. Its level varies directlywith extracellular osmolality. Cells in the kidney medulla invivo and in renal epithelial cell cultures (MDCK) accumulatelarge amounts of GPC when exposed to high concentrations ofNaCI and urea. Osmotic regulation of GPC requires choline inthe medium, presumably as a precursor for synthesis of GPC.Choline transport into the cells, however, is not osmoregulated.The purpose of the present studies was to use MDCK cellcultures as a defined model to distin h whether osmoticallyinduced accumulation of GPC results from increased GPCsynthesis or decreased GPC disappearance. The rate of incor-poration of 14C from [14C]choline into GPC, the steady-stateGPC synthesis rate, and the activity ofphospholipase A2 (whichcan catalyze a step in the synthesis of GPC from phosphati-dylcholine) are not increased by high NaCl and urea. In fact allare decreased by approximately one-third. Therefore, we findno evidence that high NaCI and urea increases the GPCsynthesis rate. On the other hand, the rate coefficient forcellular GPC disappearance and the activity of GPC:cholinephosphodiesterase (EC 3.1.4.2), which catalyzes degradationof GPC, are decreased by approximately two-thirds by highNaCl and urea. We conclude that high NaCl and urea increasethe level of GPC by inhibiting its enzymatic degradation.

Cells in the renal medullas ofmammals contain large amountsof organic osmolytes [namely, GPC (1, 2), sorbitol, inositol,and glycine betaine (for review, see ref. 3)]. When renalmedullary osmolality rises (as, for example, during antidi-uresis), the concentration of these organic osmolytes inmedullary cells increases slowly over several days (4). Therenal medullary organic osmolytes accumulate in response tothe high extracellular concentrations ofNaCl and urea in thisregion of kidney and vary with the NaCl and urea concen-trations. The organic osmolytes help balance the high os-motic pressure of the extracellular NaCl. The utilization ofthese compounds, rather than inorganic salts, for this pur-pose is attributed to the organic osmolytes being compatiblesolutes (5, 6) that can accumulate to high levels in cellswithout perturbing intracellular macromolecules, whereashigh concentrations of inorganic ions are perturbing. GPC isthought also to be a counteracting solute (5-7) that protectsintracellular macromolecules from being denatured by thehigh concentration of urea in the renal medulla.GPC is believed to be derived from phosphatidylcholine

(PC) (8). PC synthesis in mammalian kidney cells occurs byway of the Kennedy pathway through diacylglycerol andCDP-phosphocholine (9, 10). The synthesis ofGPC from PC

entails the removal of fatty acids by the activity of phospho-lipase Al (EC 3.1.1.32), phospholipase A2 (EC 3.1.1.4), andlysophospholipase (EC 3.1.1.5). The GPC that results fromthese enzymatic deacylations is degraded by hydrolysis toglycerol 3-phosphate and choline in a reaction catalyzed byGPC:choline phosphodiesterase (EC 3.1.4.2).

Ullrich (11) originally proposed that GPC:choline phos-phodiesterase controls the level of GPC in renal medullarycells. He reasoned this from two of his observations. (i) GPCdiesterase activity is higher in the renal cortex than in themedulla, consistent with the higher level of GPC in themedulla. (it) Addition of NaCl and urea to homogenates ofcortical tissue inhibits GPC diesterase activity. Thus, heconjectured that the high level of urea and NaCl in the renalmedulla during antidiuresis inhibits GPC diesterase and,thereby, elevates GPC. However, this theory was not sup-ported by the observation that renal medullary GPC di-esterase activity did not differ significantly between diureticand antidiuretic rats (12).A problem with studying renal medullary organic os-

molytes in vivo is that it is difficult to control and measure theinterstitial concentrations of NaCl and urea, which are be-lieved to control accumulation of the osmolytes. In thepresent study, we have examined osmotic regulation of GPCunder carefully controlled conditions in a defined cell culturemodel. MDCK (dog renal epithelial) cells (13, 14) and PAP-HT25 (rabbit renal medullary epithelial) cells (15) accumulatelarge amounts of GPC when exposed to medium containinghigh levels of NaCl and urea. Tissue culture medium ordi-narily contains choline, which was present in these experi-ments. If the choline is removed at the same time that theosmolality is increased, cell GPC still rises slightly by thesecond day, but it then falls back to the baseline level (14). Incontrast, if choline is present, the level of GPC continues torise. Thus, GPC presumably is synthesized from cholinetaken up from the medium and, even when there is nocholine, there must be enough of it or some intermediate,possibly PC, already in the cells to provide a transientincrease in GPC. Choline is taken up into MDCK cells bysodium-independent transport, but this transport is not af-fected by the osmolality of the medium. Therefore, osmo-regulation of GPC accumulation seems to involve changes,not in choline transport, but at some step(s) in the netsynthesis of the GPC (14).The purpose of the present studies was to distinguish to

what extent induction of GPC synthesis versus inhibition ofits degradation accounts for the elevation of GPC caused inMDCK cells by high NaCl plus urea. We find that hyperos-motic medium increases accumulation of GPC by MDCKcells through the inhibition of GPC:choline phosphodi-esterase and consequent decrease in GPC degradation. Noincrease in GPC synthesis is involved.

Abbreviations: GPC, glycerophosphocholine; PC, phosphatidylcho-line; mosmol, milliosmole(s).

7820

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 24

, 202

0 D

ownl

oade

d by

gue

st o

n F

ebru

ary

24, 2

020

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 24

, 202

0

Proc. Natl. Acad. Sci. USA 88 (1991) 7821

METHODSChemicals. GPC:choline phosphodiesterase, choline oxi-

dase, NADH, and ATP were purchased from Sigma. Phos-pholipase C, pyruvate kinase, lactate dehydrogenase, cholinekinase, and phosphoenolpyruvate were from BoehringerMannheim. [methyl-14C]Choline, [2-acyl-14C]dipalmitoylphosphatidylcholine, and [3H]raffinose were from Amer-sham. [methylcholine-'4C]GPC was prepared from [methyl-choline-14C]dipalmitoylphosphatidylcholine by saponifica-tion with LiOH (16). Media for cell culture were purchasedfrom Irvine Scientific. All other reagents were analyticalgrade.

Cell Culture. MDCK cells, purchased from the AmericanType Culture Collection, were studied in passages 62-78.They were grown in a defined medium (17). Inositol is 118 ,uMand D-glucose is 5 mM in this medium and its osmolality is 315milliosmoles (mosmol)/kg. In addition, 65 ,uM glycine be-taine was routinely added. Confluent cultures were switched,as indicated, to otherwise identical medium, made hyperos-motic (715 mosmol/kg) by the addition of equal osmolalconcentrations of NaCl and urea. All cultures were main-tained at 37°C and in 5% C02/95% air.

Cell GPC and PC. GPC was assayed in perchloric acid cellextracts by a combined enzymatic and chemical method thatemploys a specific enzyme to catalyze production of cholinefrom GPC and then measures the choline by using cholineoxidase, peroxidase, phenol, and 4-aminoantipyrine to pro-duce a "red dye" (18). The method was modified to replacephospholipase D by GPC:choline phosphodiesterase frommold to increase the sensitivity. With the modification thesensitivity is 1 nmol and linearity extends beyond 80 nmol.PC was determined by an enzymatic method that utilizes

phospholipase C, alkaline phosphatase, choline kinase, pyru-vate kinase, and lactate dehydrogenase (19). The cells werescraped and homogenized in a methanol/water mixture, 5:4(vol/vol) and then lipids were extracted with chloroform[chloroform/methanol/water, 25:15:12 (vol/vol)] (20). Thelower lipid-containing phases were dried under nitrogen andsonicated in 1.5 ml of assay medium [66 mM glycine-NaOH,pH 8.0/3.6 mM Mg2+/Triton X-100 (5.8 g/liter)]. The extractwas incubated for 30 min with phospholipase C and alkalinephosphatase, and then the reaction was stopped by heating inboiling water. After cooling, the mixture was centrifuged, andthe supernatant was transferred to a spectrophotometercuvette containing ATP, phosphoenolpyruvate, and NADH.Lactate dehydrogenase and pyruvate kinase were added tothe cuvette, followed in 10 min by choline kinase to start thereaction. The amount of choline released from PC wasdetermined from the decrease of NADH absorbance at 340nm.

[14C]Choline Incorporation into GPC and PC. Cell mono-layers were rinsed with growth medium lacking choline andthen incubated for up to 180 min in growth medium containing[methyl-14C]choline (3.6 ,uCi/ml, 65 ,tM; 1 Ci = 37 GBq). Theradioactive medium was decanted, and then the monolayerswere rinsed four times with ice-cold phosphate-bufferedsaline (PBS) containing NaCI and urea, as needed to providethe same osmolality as the growth medium. The cells wereextracted twice with 1 ml of7% (vol/vol) perchloric acid, firstfor 60 min and then for 30 min. The combined extracts wereneutralized with K2CO3 and centrifuged. The supernatantswere lyophilized and the pellets were dissolved in 0.25 ml ofwater. [14C]GPC was separated by TLC on microcrystallinecellulose sheets (Kodak), with the solvent system ethanol/2.7 M ammonium acetate, pH 5.0, 7:3 (vol/vol). GPC spotswere visualized with iodine vapor, identified by comparisonto standards, scraped, and extracted with water. GPC radio-activity was measured by liquid scintillation counting.

Radioactive PC was extracted from the cells, as describedabove for its chemical assay, and then chromatographed,using silica gel sheets (Kodak or Merck) with the solventsystem methanol/chloroform/ammonium hydroxide, 35:65:5(vol/vol). PC spots were visualized with iodine vapor, iden-tified by comparison to standards, scraped, and extractedwith chloroform/methanol, 2:1 (vol/vol). PC radioactivitywas measured by liquid scintillation counting.

Cellular [14C]GPC Disappearance. The cells were loadedwith [14C]GPC by incubation in medium hyposmotic to theexperimental medium and containing [14C]GPC (1.5 ,uM, 2mCi/mmol). Those cells that were originally in medium at 315mosmol/kg were loaded in medium at 180 mosmol/kg (re-duced NaCl), and cells originally in medium at 715 mos-mol/kg were loaded in medium at 315 mosmol/kg containing150 mM inositol and 50 mM glycine betaine, substitutedisosmotically for NaCl. After 10 min, the hypotonic loadingsolution was replaced by the original medium (that containedno GPC). The cells were then incubated for up to 7 days inthe original medium, and their content of [14C]GPC wasmeasured in acidic extracts, as described above. In oneexperiment [3H]raffinose was added in place of [14C]GPC toevaluate the contribution of extracellular contamination (seeresults).GPC:Choline Phosphodiesterase Activity. The tissue cul-

tures were rinsed twice with PBS that had been adjusted (byaddition of NaCl and urea) to the same osmolality as themedium. The wash solution was decanted and then the cellswere scraped into 0.25 M sucrose and homogenized, first ina Potter-Elvehjem apparatus with a motorized Teflon pestleand then with a glass Dounce homogenizer. Nuclei andunbroken cells were removed by centrifugation at 600 x g.Sodium deoxycholate was added to a final concentration of0.38% (21). Duplicate samples were incubated for 1-3 h in 50mM glycine-NaOH (pH 9.0, final volume 0.3 ml) containing10 mM [14C]GPC (12). [14C]Choline released from the[14C]GPC was separated by TLC, visualized with iodinevapor, identified by comparison to standards, and scrapedinto scintillation vials, and radioactivity was measured asabove.

Phospholipase A2 Activity. The tissue cultures were rinsed,scraped, homogenized, and centrifuged as above, except that0.05 M Tris-HCI (pH 8.6) was used. The assay mixturecontained 0.4 ml of homogenate, 10 mM CaC12, 7.2 ,AM1-palmitoyl-2-[1-14C]palmitoyl phosphatidylcholine (0.1,Ci), and 0.1% sodium deoxycholate in a total volume of0.505 ml. After 2 h at 37°C under air with constant shaking,the reaction was stopped by addition of chloroform/methanol/water, 2:2:1.8 (vol/vol) (22). The organic phasewas dried under N2 and taken up in 1 ml of chloroform/methanol, 2:1 (vol/vol). The radioactivity in 50-,l sampleswas measured to determine extraction efficiency; the remain-der of the sample was concentrated and chromatographed onSilica Gel 60 plates (Merck) in chloroform/methanol/water,65:25:4 (vol/vol). The lipids were visualized under iodinevapor and identified by comparison to standards, then thefatty acid spots were scraped, and radioactivity was mea-sured (23).

RESULTS AND DISCUSSIONHigh NaCl plus Urea Increases Cell GPC but not PC. In

agreement with earlier results (14), MDCK cells switched tomedium made hyperosmotic (by addition of NaCl plus urea)accumulated GPC (Table 1). Cell GPC doubles by 2 days afterswitching and is even higher by 7 days. PC apparently is aprecursor in the synthesis of GPC (8). The amount of PC inMDCK cells, however, is not affected by increased mediumNaCl and urea (Table 1).

Medical Sciences: Zablocki et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 24

, 202

0

7822 Medical Sciences: Zablocki et al.

Table 1. GPC and PC content of MDCK cells

Medium osmolality, Cell GPC, Cell PC,mosmol/kg nmol/kg of protein nmol/kg of protein

315 30.2 ± 7.9 (7) 96.2 ± 1.6 (3)715 (2 days) 78.2 ± 10.1 (13) 98.7 ± 12.1 (3)715 (7 days) 123.9 ± 12.5 (3)

Cells were grown in isosmotic medium (315 mosmol/kg) and thenthe osmolality was increased for 2 or 7 days. Results are mean ± SD.The number of tissue cultures analyzed individually is shown inparentheses.

Incorporation of 14C from ['4C]Choline into PC and GPC.To determine whether increased synthesis of GPC accountsfor its rise in cells exposed to hyperosmotic medium, wemeasured the incorporation of [14C]choline into PC and GPCas a function of medium osmolality. After addition of[14C]choline to the medium, the specific activity ofGPC riseslinearly with time, regardless of medium osmolality (Fig. 1).The specific activity of PC, on the other hand, rises hyper-bolically. Further, the specific activity of GPC greatly ex-ceeds that ofPC in isosmotic medium (Fig. lA) and is higherthan that of PC at 30 min in hyperosmotic medium. Thisapparently differs from the result of experiments in which[32P]phosphate was injected into rats (8). The rise in specificactivity of 32p in the PC in rat liver precedes that in GPC,which led to the conclusion that PC is a precursor of GPC,rather than vice versa. The contrary results in Fig. 1 do notnecessarily exclude this possibility for MDCK cells. Cell PChas been reported to be heterogeneous in rat liver (24, 25) andin murine neuroblastoma NlE-115 cells and human fibro-blasts (26). Also, there is evidence of "channeling" in PC andGPC synthesis from choline, such that the intermediates arenot part ofgeneral cellular pools (27). Thus, it is possible that

_

._

.A_

Time after I14C]choline addition, min

FIG. 1. Specific activity ofMDCK cell PC and GPC after additionof ["'Cicholine to the medium. Specific radioactivity of the cholineadded to the medium was 99,000 cpm/nmol. (A) Cells grown inisosmotic medium. Cell GPC was 31.8 + 0.2 mmol/kg of protein (n= 4). (B) Cells switched to hyperosmotic medium for 2 days. CellGPC was 74.9 9.0 mmol/kg of protein.

0

xaE0.m

00

1-

"O 12 24 36 48 60 72 84Time after [14C]GPC introduction, h

FIG. 2. Incorporation into MDCK cell GPC of 14C from[14C]choline added to the medium. GPC was 28.6 ± 7.5 mmol/kg ofprotein (n = 3) in cells kept in isosmotic medium and 105 ± 21mmol/kg of protein (n = 5) 2 days after switching to hyperosmoticmedium.

MDCK cell PC also is heterogeneous and that synthesis ofGPC involves a minor pool of PC that rapidly exchanges withextracellular choline. That would explain how the rise inspecific activity of GPC can precede that of the bulk of PCand have a very different time course without postulating asource of GPC other than from PC.

Since the incorporation of 14C from [(4Cjcholine into GPCis linear with time, we could calculate an apparent rate ofGPC synthesis from choline, using the slope of this line andthe specific activity of [14C]choline in the medium. Theapparent rate, calculated in this fashion, is less in cellsswitched for 2 days to hyperosmotic medium than in cells thatremain in isosmotic medium (Fig. 2). In three experiments,including that in Fig. 2, the mean apparent rate of 14Cincorporation from choline into GPC in cells in hyperosmoticmedium was 66.4 ± 7.5% of the rate in isosmotic medium(Table 2). By assuming that this is a valid measure of GPCsynthesis rate, the elevation of MDCK cell GPC in hyper-osmotic medium must be due to decreased disappearance,rather than increased synthesis, because synthesis is re-

duced, not increased, by hyperosmolality.However, there could well be an intermediate precursor

pool between choline in the medium and GPC in the cells, forexample, a pool ofPC that is separate from other cellular PCpools. Not being able to establish with certainty the specificactivity of the direct precursor pool, we could not be abso-lutely sure of the rate of GPC synthesis and, therefore, weperformed the following additional experiments.Rate of Disappearance of Intracellular [14CJGPC. In pre-

liminary experiments we attempted to measure GPC disap-pearance by pulse labeling with ["4C]choline for 1 h, followedby a chase with nonradioactive choline for 30 min, 3 h, and24 h. We found that the 14C labeling ofGPC continues to rise

Table 2. Apparent GPC synthesis rate

GPC synthesis rate, nmol perh per mg of protein

Isosmotic Hyperosmotic0.12 0.080.20 0.120.22 0.15

Cells were grown in isosmotic medium (315 mosmol/kg) and thenswitched for 2 days to hyperosmotic medium (715 mosmol/kg) orkept in isosmotic medium. Apparent synthesis rate was calculated asthe rate of increase in ['4C]GPC (e.g., in Fig. 2) divided by thespecific activity of [14C]choline added to the medium. The results arefrom three experiments.

Hyperosmotic0 medium

0

Isosmotic a8 %

__medium 0__

0

Proc. Natl. Acad. Sci. USA 88 (1991)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 24

, 202

0

Proc. Natl. Acad. Sci. USA 88 (1991) 7823

during the chase after the radioactive choline is replaced bythe nonradioactive choline. Presumably, during the chasethere is continued incorporation from labeled precursors, asobserved previously for PC in a similar experiment with C6glioma cells (27). At the time when a pulse ends and the chasestarts, the [14C]GPC is higher in cells in isosmotic medium(e.g., Table 2 and Fig. 2). After 3 h of chase with nonradio-active choline, however, the [14C]GPC increases more inhyperosmotic medium and becomes higher than in the isos-motic medium (data not shown). The less rapid increase in[14C]GPC under isosmotic conditions relative to hyperos-motic conditions during the chase is consistent with thehypothesis that hyperosmolality causes MDCK GPC to riseby inhibiting its disappearance. The interpretation is compli-cated, however, because the chase results in increasing,rather than decreasing, [14C]GPC, so that disappearanceevidently cannot be the only factor determining the result.To avoid continued labeling of GPC during the chase, we

introduced [14C]GPC directly into the MDCK cells withoutlabeling the metabolic precursors of GPC. Brief incubationwith a high concentration of [14C]GPC results in the uptakeof too few counts to be useful in a pulse-chase experiment(Table 3). However, hypotonicity results in loss of organicosmolytes, including GPC, from cells (14, 15). For sorbitol,there is evidence for an increase in permeability, resulting in['4C]sorbitol uptake by the cells while the bulk of the non-radioactive intracellular sorbitol is leaving the cells (28).Therefore, we tested the effect of hypotonicity on [14C]GPCuptake and found that it is accelerated by hypotonicity (Table3). Thus, the efflux of GPC from cells exposed to hypotonicmedium apparently involves increased permeability, as doesthe sorbitol efflux.The amounts of organic osmolytes (inositol, glycine be-

taine, and GPC) in MDCK cells are relatively small when thecells are grown in isosmotic medium. Thus, only smallamounts of them are available to exit from the cells during ahypotonic shock and cell composition is little affected. Incontrast, renal cells exposed to hyperosmotic medium con-tain large amounts of these organic osmolytes, and a majorportion of these osmolytes can exit the cells during a 10-minhypotonic shock (15, 29). To minimize changes in cell inositoland glycine betaine, their concentrations, calculated to equalthose in the cells, were added to the medium during the 10 minthat the cells were switched from hyperosmotic to isosmoticmedium. We did not increase the medium GPC concentra-tion, but this should not be a serious drawback because theefflux of GPC during hypotonic shock is considerably lessthan efflux of the other organic osmolytes (14, 15, 29).

In preliminary pulse-chase experiments, we found a largedecrease in tissue [14C]GPC during the first 15 min of chase,followed by a much slower and sustained rate of decrease.Raffinose is a trisaccharide that does not penetrate readilyinto cells. By using [3H]raffiinose as an extracellular marker,we found that the initial fall in [14C]GPC can be explained by

Table 3. [14C]GPC uptake by MDCK cells

Osmolality,mosmol/kg Cell [14C]GPC

During uptake, cpm/mgInitial loading of cell protein

315 315 168 ± 23315 180 1290 ± 31715 715 136 ± 16715 315 927 ± 74

Cells were grown in medium at 315 or 715 mosmol/kg and then

rapid efflux of ['4C]GPC associated with extracellular spaces(data not shown). Therefore, in subsequent experiments wefocused on the slower sustained rate of decrease of [14C]GPCthat occurs later in the chase.Between 6 and 80 h of chase, [14C]GPC decreased loga-

rithmically in MDCK cells exposed to either hyperosmotic orisosmotic medium (Fig. 3). This apparently represents con-version of GPC to another chemical form rather than effluxfrom the cells since we were unable to detect [14C]GPC in themedium during this interval. The half-life of [14C]GPC was

much longer in hyperosmotic than in isosmotic medium (212h versus 61 h, respectively). The corresponding rate coeffi-cients are 0.0114 h-1 in isosmotic medium and 0.0033 h-1 inhyperosmotic medium. Thus, hyperosmolality reduces therate coefficient for GPC disappearance by 71%, consistentwith inhibition of its disappearance.As expected, the total cellular pool of unlabeled GPC was

higher in the hyperosmotic medium (163 mmol/kg of protein)than in the isosmotic medium (62 mmol/kg of protein), and itdid not vary significantly with time from 6 to 72 h ofthe chase.The steady-state turnover rate ofGPC can be calculated fromthe product of the amount in the cells times the rate coeffi-cient for disappearance. It was 0.71 mmol per kg of proteinper h in isosmotic medium and 0.53 mmol per kg of proteinper h in hyperosmotic medium. Since cell GPC was constant,this represents both the rate of synthesis and the rate ofdisappearance, which are equal in a steady state. Thus,hyperosmolality inhibited GPC synthesis by 26%. Evidently,it also reduced the absolute rate of GPC disappearance by26%. It should be noted that the slower disappearanceoccurred in the face of a markedly higher concentration ofintracellular GPC, which by itself might be expected toincrease, rather than decrease, the absolute GPC disappear-ance rate.

Phospholipase A2 Activity. To test further whether in-creased synthesis of GPC might be involved in its osmoticregulation, we measured phospholipase A2 activity in ho-mogenates from MDCK cells maintained in isosmotic me-dium (315 mosmol/kg) and 2 days after they were switchedto hyperosmotic medium (715 mosmol/kg with added NaCland urea). The activity is 0.70 0.05 pmol of free fatty acidper mg of protein per min (n = 4) from cells maintained inisosmotic medium and 0.48 ± 0.04 (n = 4) in hyperosmoticmedium. Thus, hyperosmolality reduces phospholipase A2

Time after [14C]choline addition, min

FIG. 3. Effect of NaCl and urea on the rate of disappearance ofintracellular [14C]GPC. MDCK cells were grown in isosmotic me-dium (315 mosmol/kg) and then switched to hyperosmotic medium(715 mosmol/kg with added NaCl and urea) or maintained in theisosmotic medium, as a control. Seven days after the switch,[14C]GPC was loaded into the cells by hypotonic shock. Then, cell[14C]GPC was measured at the intervals shown. GPC was 62 + 10mmol/kg of protein (n = 7) in cells kept in isosmotic medium and 163± 22 (n = 7) in hyperosmotic medium. Cell GPC did not changesignificantly from 6 to 72 h in either condition.

exposed for 10 min to medium containing ['4C]GPC (1.5 mM, 1ACi/ml) with or without lowering the osmolality by removing someof the NaCl and all of the urea (when it was present) from themedium.

Medical Sciences: Zablocki et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 24

, 202

0

7824 Medical Sciences: Zablocki et al.

Table 4. GPC:choline diesterase activityActivity, nmol per mg of

protein per min

Isosmotic Hyperosmotic

0.35 0.130.21 0.060.26 0.09

Cells were grown in isosmotic medium (315 mosmol/kg) and thenswitched for 2 days to hyperosmotic medium (715 mosmol/kg) orkept in isosmotic medium. The results are the means of duplicate ortriplicate measurements in three experiments.

activity by 31%, comparable to the earlier findings of a 34%decrease in the rate of 14C incorporation from choline intoGPC and 26% decrease in the steady-state turnover rate forGPC (also a measure of synthesis rate). Thus, there is noevidence that hyperosmolality raises renal cell GPC byincreasing GPC synthesis. To the contrary, hyperosmolalityapparently inhibits GPC synthesis by approximately one-quarter to one-third.GPC:Choline Phosphodiesterase Activity. GPC:choline

phosphodiesterase catalyzes the degradation of GPC tocholine and glycerol 3-phosphate. [GPC:cholinephosphatephosphodiesterase has also been reported in renal tissue (12),but we found no evidence for this activity in MDCK cells(data not shown).] To test further whether degradation ofGPC is involved in its osmotic regulation, we measuredGPC:choline phosphodiesterase activity in homogenatesfrom MDCK cells maintained in isosmotic medium and 2days after switching to hyperosmotic medium. The meanactivity is 0.27 nmol of choline per mg of protein per min inhomogenates from cells maintained in isosmotic medium and0.09 in otherwise identical homogenates from cells in hyper-osmotic medium (Table 4). Thus, hyperosmolality inhibitsGPC:choline phosphodiesterase activity by 66%, consistentwith our observation that the rate coefficient for GPC dis-appearance is markedly decreased under the same condi-tions.As part of the last experiment in Table 4, GPC:choline

diesterase activity and cell GPC were measured at intervalsfrom 2 to 48 h after the osmolality was increased by addingNaCl and urea to the medium (Fig. 4). The GPC:cholinediesterase activity decreased substantially within 2 h and

Or . -

'o Ro

la ° 0.20

o 0

&.-

A= £ 0.10s &

U0 r. n nni

60 *r0

4 m

0

20 E

0 12 24 36 48 U

Time after increase in osmolality, h

FIG. 4. Effect of NaCl and urea on GPC:choline diesteraseactivity and cell GPC. MDCK cells were grown in isosmotic medium(315 mosmol/kg) and then switched to hyperosmotic medium (715mosmol/kg) for 2-48 h. The results in the isosmotic medium areshown at zero time.

decreased even more by 48 h. In contrast, cell GPC did notchange at 2 h and increased gradually over 48 h. Thus, thedecrease in GPC:choline diesterase activity precedes the risein cell GPC, consistent with the hypothesis that hyperosmo-lality elevates renal cell GPC by inhibiting its degradation.

In summary, we find direct evidence that high extracellularNaCl and urea increase MDCK cell GPC by inhibiting itsdegradation. Also consistent with this view is a preliminary[14C]choline pulse-chase study of rat inner medullary col-lecting ducts in vitro from which it was concluded thatturnover of GPC is reduced in hyperosmolal medium, prob-ably because of inhibition of GPC degradation (30).

We acknowledge the generous help of Vincent Manganiello,(Laboratory of Cellular Metabolism, National Heart, Lung andBlood Institute) who gave us many useful suggestions.

1. Ullrich, K. J. (1956) Pflugers Arch. 262, 551-561.2. Schimassek, H., Kohl, D. & Bucher, T. (1959) Biochem. Z. 331,

87-97.3. Garcia-Perez, A. & Burg, M. (1991) J. Membr. Biol. 119, 1-13.4. Blumenfeld, J., Hebert, S., Heilig, C., Balschi, J., Stromski, M.

& Gullans, S. (1989) Am. J. Physiol. 256, F916-F922.5. Yancey, P., Clark, M., Hand, S., Bowlus, R. & Somero, G.

(1982) Science 217, 1214-1222.6. Balaban, R. S. & Knepper, M. A. (1983) Am. J. Physiol. 245,

C439-C444.7. Somero, G. (1986) Am. J. Physiol. 251, R197-R213.8. Dawson, R. M. C. (1955) Biochem. J. 59, 5-8.9. Wilgram, G. F. & Kennedy, E. P. (1963) J. Biol. Chem. 238,

2615-2619.10. Toback, F. G., Smith, P. D. & Lowenstein, L. M. (1974) J.

Clin. Invest. 54, 91-97.11. Ullrich, K. J. (1959) Biochem. Z. 331, 98-102.12. Kanfer, J. N. & McCartney, D. G. (1989) FEBS Lett. 257,

348-350.13. Nakanishi, T., Balaban, R. S. & Burg, M. B. (1988) Am. J.

Physiol. 255, C181-C191.14. Nakanishi, T. & Burg, M. B. (1989) Am. J. Physiol. 257,

C795-C801.15. Moriyama, T., Garcia-Perez, A. & Burg, M. B. (1990) Am. J.

Physiol. 259, F847-F858.16. Spanner, S. & Ansell, G. B. (1987) Neurochem. Res. 12,

203-206.17. Taub, M., Chuman, L., Saier, M. H., Jr., & Sato, G. (1979)

Proc. Natl. Acad. Sci. USA 76, 3338-3342.18. Takayama, M., Itoh, S., Nagasaki, T. & Tanimizu, I. (1977)

Clin. Chim. Acta 79, 93-98.19. Diedrich, K., Hepp, S., Welker, H., Krebs, D., Beutler, H.-O.

& Michal, G. (1979) Geburtshilfe Frauenheilkd. 39, 849-856.20. Stem, P. H. & Vance, D. E. (1987) Biochem. J. 244, 409-415.21. Baldwin, J. J. & Cornatzer, W. E. (1968) Biochim. Biophys.

Acta 164, 195-204.22. Bligh, E. G. & Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37,

911-917.23. Limas, C. & Limas, C. J. (1983) Biochim. Biophys. Acta 753,

314-323.24. Vance, J. E. & Vance, D. E. (1986) J. Biol. Chem. 261,

4486-4491.25. Vance, J. E. (1991) J. Biol. Chem. 266, 89-97.26. Morash, S., Cook, H. & Spence, M. (1988) Biochim. Biophys.

Acta 961, 194-202.27. George, T. P., Morash, S. C., Cook, H. W., Byers, D. M.,

Palmer, F. B. S. C. & Spence, M. W. (1989) Biochim. Biophys.Acta 1004, 283-291.

28. Siebens, A. & Spring, K. (1989) Am. J. Physiol. 257, F937-F946.

29. Nakanishi, T. & Burg, M. B. (1989) Am. J. Physiol. 257,C964-C970.

30. Bauernschmitt, H. & Kinne, R. (1991) Pflugers Arch. 418, R68(abstr.).

Enzyme activity

Proc. Natl. Acad Sci. USA 88 (1991)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 24

, 202

0