Biochemical and Molecular Roles of Nutrients

Diet and Disease Alter Phosphoinositide Compositionand Metabolism in Murine Polycystic Kidneys1»2»3

HAROLD M. AÃœKEMA, TAM1O YAMAGÃœCHI, KOJI TOMOBE,DIANA J. PHILBRICK, ROBERT S. CHAPKIN,*HISAHIDE TAKAHASHr AND BRUCE J. HOLUB4

Department of Nutritional Sciences, University of Guelph, Guelph, Canada NIG 2W1;*Faculty of Nutrition, Molecular and Cell Biology Group, Texas A&M University,College Station, TX 77843-2471; and tLaboratory Animal Center,Fujita Health University, Toyoake, Japan

ABSTRACT Because diet can affect the progression ofpolycystic kidney disease (PKD) and because renalphosphoinositide metabolism is altered in mice withPKD, the effects of diet and disease on phosphoinositidecomposition and metabolism were examined in kidneysof mice with PKD. The phosphatidylinositol-phosphate(PIP) to phosphatidylinositol (PI) molar ratio was higher(0.034 ±0.003 vs. 0.023 ±0.001, P < 0.01) and the PI-bisphosphate (PIP¿)to PIP molar ratio was lower (0.70±0.08 vs. 1.19 ±0.10, P < 0.05) in kidneys of mice withPKD [DBA/2FG-pci/ (pcy)] compared with normal controls (DBA/2J). When initial incorporation (reflectingsynthesis) of pH]inositol into renal phosphoinositides inmice injected with [3H]inositol was measured, the[3H]PIP to [3H]PI ratio was higher in the diseasedkidneys compared with normal kidneys (0.016 ±0.001vs. 0.013 ±0.001, P < 0.05), whereas the [3H]PIP2 to[3H]PIP ratio was not significantly different. In a studyusing dietary manipulations that alter the progression ofPKD in pcy mice (6 vs. 25% casein and sunflower seedoil vs. fish oil in a 2 x 2 design), animals were injected¡ntraperitoneally with [3H]inositol 5 h before killing. Inthese animals, the [3H)PIP2 to [3H]PIP ratio seemed to

be the best indicator of disease progression. In addition,kidney weight (as altered by diet) was positively correlated (r = 0.62, P = 0.02) with the level of the [3H]PI-3-P isomer relative to total [3H]PIP in the kidney. Theseresults demonstrate that alterations in dietary proteinlevel and lipid composition can modulate renal phosphoinositide signal transduction in mice with PKD. J.Nutr. 125: 1183-1191, 1995.

INDEXING KEY WORDS:

•phosphoinositide •dietary protein•polycystic kidney disease•dietary lipid •mice

Polycystic kidney disease (PKD)5 encompasses avariety of kidney cystic disorders that are characterized by abnormal growth and the development of

renal cysts. In this disease, an excessive proliferationof the tubular epithelial cells in the kidney plays animportant role in the progressive development offluid-filled renal cysts and ultimately in renal failure(Gabow 1993, Grantham 1993). Abnormal oncogeneexpression (Calvet 1993, Herrera 1991), epidermalgrowth factor metabolism (Calvet 1993, Lakshmananand Eysselein 1993, Wilson and Sherwood 1991) andrenal polyphosphoinositide metabolism (Aukema etal. 1992a) have been implicated in the etiology of thisdisease, suggesting that there are biochemical alterations in select intracellular signaling systems in thisdisease state.

We demonstrated previously that the incorporationof [3H]inositol into renal phosphoinositides is alteredin DBA/2FG-pcy (pcy) mice with PKD compared withnormal (DBA/2J) mice (Aukema et al. 1992a). The

Presented in part at the Annual Meeting of the Federation ofAmerican Societies for Experimental Biology, April 1991, Atlanta,GA. [Aukema, H. M., Philbrick, D. ]., Yamaguchi, T., Tomobe, K.,Takahashi, H. & Holub, B. J., Phosphoinositide composition andmetabolism in murine polycystic kidney disease, FASEB J. 5: A1409(abs.ll.

^Supported by a grant to BJH from the Natural Sciences and

Engineering Research Council of Canada. HMA was supported byan Ontario Graduate Scholarship.

^The costs of publication of this article were defrayed in part by

the payment of page charges. This article must therefore be herebymarked "advertisement" in accordance with 18 USC section 1734

solely to indicate this fact.4To whom correspondence should be addressed.5Abbreviations used: DBA/2J, normal control mice; FO, fish oil;

HP, high protein; inositol-3,4,5-P3, inositol trisphosphate; LP, lowprotein/ pcy, DBA/2FG-pcy, mice with PKD; PI, phosphatidylinositol; PI-3-P, PI-4-P, PIP, phosphatidylinositol phosphate; PI-3,4-P2, PI-3,5-P2, PI-4,5-P2, PIP2, phosphatidylinositolbisphosphate; PI-3,4,5-P3, phosphatidylinositol trisphosphate; PKD,polycystic kidney disease, SO, sunflower seed oil.

0022-3166/95 $3.00 ©1995 American Institute of Nutrition.Manuscript received 29 July 1994. Initial review completed 23 September 1994. Revision accepted 10 November 1994.

1183

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

1184 AUKEMA ET AL.

phosphoinositides are a class of phospholipids involved in intracellular signal transduction and consistprimarily of phosphatidylinositol (PI) and its phos-phorylation products, Pi-phosphate (PIP), PI-bisphosphate (PIPi), and PI-trisphosphate (PIPs). Thepredominant isomers of these polyphosphoinositidesare PI-4-P and PI-4,5-P2- Receptor-activated degradation of PI-4,5-P2 by phospholipase C results in theproduction of second messengers associated with cellactivation and proliferation, namely, sn-l,2-diacyl-glycerol and inositol trisphosphate (inositol-3,4,5-P3)

(Rana and Hokin 1990). Diacylglycerol is a physiologic activator of the serine/threonine kinase, proteinkinase C (Nishizuka 1992), whereas the generation ofinositol-3,4,5-P3 results in the mobilization of intracellular Ca2+ (Berridge and Irvine 1989). In kidneysof pcy mice, the level of [3H]PIP2 following in-traperitoneal administration of [3H]inositol was 50%

lower than the level in the kidneys of normal controlmice, suggesting that this signaling pathway is alteredin PKD (Aukema et al. 1992a). Because the [3H]PIP

levels were not significantly altered, the lower[3H]PIP2 levels could be due to either reduced PIP

kinase activity or increased phospholipase C activity,which would result in a potential decrease or increase, respectively, in the formation of the secondmessengers, diacylglycerol and inositol-3,4,5-P3. Inthis regard, renal diacylglycerol levels are elevated inpcy mice compared with normal mice (Aukema et al.1992c).

Polyphosphoinositides phosphorylated on the D-3position of the inositol ring (e.g., PI-3-P, PI-3,4-P2, PI-3,5-P2, PI-3,4,5-P3) have also been implicated in cellsignaling and activation. Although these polyphosphoinositides are found only in small amountsrelative to PI-4-P and PI-4,5-P2 and are not brokendown by phospholipase C to liberate diacylglyceroland inositol-3,4,5-P3, the formation of the 3-phos-phorylated polyphosphoinositides has been associatedwith a host of growth factors and oncogenes (Parkerand Waterfield 1992, Varticovski et al. 1994). Therelatively higher levels of [3H]PI-3-P in the more dis

eased kidneys from pcy mice (Aukema et al. 1992a)suggest that the 3-phosphorylated polyphosphoinositides may be important in the regulation of signalingevents related to cell proliferation in PKD.

Several studies have demonstrated that phos-phoinositide metabolism can be modified in vivo byaltering dietary constituents. Diets enriched in (n-3)fatty acids have been shown to alter platelet phos-phoinositide metabolism in humans (Skeaff 1988) andrabbits (Medini et al. 1990). Dietary protein restriction in young rats results in reduced renal PIPlevels (Ananth et al. 1986). Phosphoinositide turnoverin cultured vascular muscle cells and in isolatedplatelets has also been reported to be affected byincubation with (n-3) fatty acids (Gaudette and Holub

1990, Locher et al. 1989).

We recently showed that diet can significantly altercyst development and disease progression in pcy mice(Aukema et al. 1992b and 1992e). Reduction of dietary protein to a low yet adequate level for normalgrowth resulted in smaller cyst and kidney size andsignificantly longer survival in pcy mice (Tomobe etal. 1994). When dietary fish oil (FO) was substitutedfor sunflower seed oil (SO) for 90 d, increases inkidney and cyst size were not large enough to bestatistically significant. In the long term, however, FOsupplementation resulted in shorter life spans andearlier onset of proteinuria. The effects of these dietson the phosphoinositide signaling pathway in PKD,however, are not known. In the present study,therefore, we examined renal phosphoinositide composition and metabolism (including the 3-phosphorylated polyphosphoinositides) in normal miceand pcy mice with PKD, and the effect of diet on thephosphoinositides. In addition to disease-relateddifferences, dietary manipulations in pcy mice werealso associated with changes in phosphoinositide metabolism. The alterations in phosphoinositide metabolism in diseased kidneys by diet demonstrate itspotential to modulate this signaling pathway in PKD.

MATERIALS AND METHODS

The animals used for these studies were DBA/2FG-pcy (pcy) mice, which develop PKD, and theirnormal counterparts, DBA/2 J mice. As previouslydescribed, the pcy mouse has a form of renal cysticdisease that appears similar in many respects to thatseen in autosomal dominant PKD (Takahashi et al.1991). Mice were housed individually in an environment with controlled temperature (23°C),hu

midity (60-65% relative humidity) and light (12 hlight: 12 h dark). Male mice weaned at 30 d were usedin all studies. The animal experimental protocolswere in accordance with the Canadian Council forAnimal Care guide and were approved by the AnimalCare Committee, University of Guelph.

In the first study, weanling animals were given freeaccess to nonpurified diet (Purina mouse chow, no.5015, Purina Mills, St. Louis, MO) until they reached120 d of age, when they were killed by carbon dioxideoverexposure. The kidneys and livers were rapidlyexcised, immediately frozen in liquid nitrogen,weighed, and stored at -70°C until analyzed. The

frozen kidneys were then dropped into 3.75 volumesof chloroform-methanol-concentrated HCl (20:40:1, v/v/v) containing 0.05% butylated hydroxytoluene andhomogenized immediately (Allan and Micheli 1978).Phase separation was achieved by adding 1.25volumes each of chloroform and distilled water, followed by centrifugation at 1000 x g for 25 min. Thelower phase containing the phosphoinositides was removed (twice), dried under nitrogen and reconstituted

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

DIET AND PHOSPHOINOSITIDES IN POLYCYSTIC KIDNEYS 1185

in chloroform-methanol (2:1, v/v). A known aliquot ofthe extract was taken to purify PI, using two-dimensional TLC on silica gel G plates (Merck, British DrugHouse, Toronto, Ontario, Canada). The mobile phasesused in this step were chloroform-methanol-14.8 mol/L ammonium hydroxide (65:35:5.5, v/v/v) and chlo-roform-methanol-88% formic acid-water (55:28:5:1, v/v/v/v) (Thomas and Holub 1991). Total PIP and PIP2were purified from the remaining part of the extractby one-dimensional TLC, using the mobile phasechloroform-methanol-14.8 mol/L ammoniumhydroxide-water (45:35:8:4, v/v/v/v). The resolvedbands were then visualized under UV light afterspraying the plates with 0.1% 8-anilino-l-naphthalene-sulfonic acid (Sigma Chemical, St. Louis,MO) in water (Vinson and Hooyman 1977). Thesebands were scraped into test tubes containing appropriate amounts of monopentadecanoate (15:0) asinternal standard. Purified PI, PIP and PI?2 were tran-sesterified by adding tetrahydrofuran and 0.5 mol/Lsodium methoxide to the tube and incubating for 15min at 50°C(Christie 1982). Following addition of

acetic acid and water, the derived fatty acid methylesters were extracted twice with hexane, and purifiedby TLC using toluene as the mobile phase. The purified fatty acid methyl esters were visualized, scrapedfrom the plates and extracted from the silica gel withpetroleum ether. Fatty acid methyl esters were analyzed by gas-liquid chromatography as described(Holub and Skeaff 1987), using a DB-225 megaborecolumn (Chromatographie Specialties, Brockville, Ontario, Canada) at a temperature of 205°C.To correct

for background contamination, blank regions of theTLC plate corresponding to the lipid bands weretreated exactly as the samples. The peak areas obtained by gas-liquid chromatography for the blankswere then subtracted from the sample peak areas.Phosphoinositides were quantified by dividing theamount (nanomoles) of fatty acid for each fraction bythe number of fatty acids per molecule.

In the short-term labeling study, 120-d-old maleDBA/2J and pcy mice were injected intraperitoneallywith 6.66 GBq of [3H]inositol (Amersham, Missis-

sauga, Ontario, Canada) in 75 /*L of 9 g/L NaCl andkilled 12 min later. Preliminary kinetic analysis indicated that radiolabel accumulation at this early timewas predominantly due to phosphoinositide synthesis. Kidneys and livers derived from these micewere treated as described for the first study. Thephosphoinositides of interest were separated fromeach other by the one-dimensional TLC system asdescribed above because it is predominantly the in-ositol-containing lipids (PI, lyso PI, PIP and PIPi) thatare labeled (Holub and Kuksis 1972). The fractions ofinterest were scraped from the plates and the radioactivity determined using a Beckman 3801 liquid scintillation counter (Beckman, Irvine, CA).

In the dietary study, weanling pcy mice were randomly divided into four groups that were fed semipu-

rified diets containing either a high (HP, 25% casein)or low (LP, 6% casein) level of protein, and either aSO rich in the (n-6) fatty acid, linoleic acid [18:2(n-6)],or a FO concentrate rich in (n-3) fatty acids[eicosapentaenoic acid, 20:5(n-3) + docosahexaenoicacid, 22:6(n-3)j, in a 2 x 2 design. The FO-enricheddiet contained 1% SO to supply adequate amounts ofthe essential fatty acid 18:2(n-6). The composition ofthe diets is given in Table 1. At 120 d of age, the micewere injected intraperitoneally with 0.5 mL of physiologic saline containing 66.6 GBq of myo-[2-3H]inositol(Amersham). Earlier time studies (1-24 h) indicatedthat the in vivo formation of renal [3H]phospha-tidylinositol-phosphate ((3H]PIP) and [3H]phospha-tidylinositol-bisphosphate ((3H]PIP2) reached a plateau

by 5 h (Aukema et al. 1992a). The mice were thereforekilled 5 h after isotope injection. The phosphoinositide fractions were purified from kidneys andprocessed as described above for the labeling study.

To determine the radioactivity ratios in tissuesfrom these mice, a small aliquot of each extract was

TABLE 1

Composition of experimental diets

Dietary group1

Ingredient HP-SO HP-FO LP-SO LP-FO

Casein,vitamin-freeDL-Methionine2Corns

tarchSucroseSunflower

seedoilMaxEPAoil3Mineralmix4Choline

chlorideVitaminmix5Fiber6Antioxidant7250.3253010__5.50.213.90.005g/100250.32530195.50.213.90.005g

diet60.09443010_5.50.214.10.00560.094430195.50.214.10.005

'Abbreviations used: FO, fish oil; HP, high protein; LP, low

protein; SO, sunflower seed oil.2Added as part of the vitamin mix.3R. P. Scherer (Windsor, Ontario, Canada).4Supplied the following (mg/100 g diet): calcium carbonate, 720;

cupric sulfate, 5.3; calcium diphosphate, 1130; ferric citrate (5H2O),63.54; magnesium sulfate, 230; manganese sulfate (H2O), 15.4;potassium chloride, 730; potassium iodide, 0.3; zinc carbonate, 5.3;sodium diphosphate, 600; chromium acetate, 1.0; zinc sulfate, 8.4;magnesium chloride, 29.3; manganese chloride 2.7; sodiumselenite, 0.2 /ig.

5Supplied the following (/ig/g diet): thiamine-HCl, 6; riboflavin,6; pyridoxine-HCl, 7; nicotinic acid, 30; DL-calcium pantothenate,32; folie acid, 2; D-biotin, 0.2; vitamin B-12, 10; i-inositol, 1000;retinyl palmitate, 8; ergocalciferol, 2.5; all-rac-a-tocopheryl acetate,200; menadione, 0.5.

6Alfa floe, Lee Chemical (Toronto, Ontario, Canada). Some fiber

was included in the vitamin mix (0.6 g/100 g HP diet, 0.8 g/100 gLP diet).

7Ethoxyquin, Monsanto Chemical (Toronto, Ontario, Canada).

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

1186 AUKEMA ET AL.

TABLE2

Body and organ weights in DBA/2J and pcy mice fed nonpurified diet1

MouseMass

studyDBA/2J |n = 4)pcy (n -4)Isotope

studyDBA/21 (n = 8)pcy (n - 6)Body

weight32.7

±1.825.3 ±0.5**30.2

±1.227.2 ±1.1Kidney2g0.29

±0.010.76 ±0.07***0.28

±0.010.75 ±0.06***Liver1.38

±0.081.09 ±0.03*1.32

±0.041.14 ±0.04**Kidney/

Liver/body weight bodyweightg/100

g0.88

±0.02 4.22 ±0.133.02 ±0.29*** 4.30 ±0.020.94

±0.03 4.39 ±0.072.76 ±0.20* ** 4.20 ±0.08Kidney/

liverweightg/g0.21

±0.000.71 ±0.07***0.21

±0.010.66 ±0.05***

Abbreviations used: DBA/2J, normal control mice; pcy, DBA/2FG-pcy, mice with polycystic kidney disease.2Data for right kidney only. 'P < 0.05, "P < 0.01, ***P < 0.001, significantly different from corresponding DBA/21 value.

used. Most of each lipid extract was used for isomeranalysis of [3H]PIP and [3H]PIP2. These fractions werepurified by one-dimensional TLC as described above,scraped into test tubes, and the glycerophos-phoinositide isomers analyzed as described (Aukemaet al. 1992a, Gaudette et al. 1993). Briefly, phos-phoinositides were deacylated by incubation inmethanol-40% methylamine-butanol (4:5:1, v/v/v) forl h at 53°C.The samples were then dried down in

vacuo, redissolved in distilled water, and extractedtwice with butanol-hexane-ethylformate (20:40:1, v/v/v) to remove the cleaved fatty acyl groups. Thelower, aqueous phase was then dried down in vacuoand reconstituted in 0.005 mol/L (NH4)2HPO4 (pH3.8). The derived glycerophosphoinositides were separated by HPLC as described (Aukema et al. 1992a,Gaudette et al. 1993), using a Whatman Partisil 5 SAXRAG II (0.46 x 10 cm) column and a stepwise gradientof (NH4)2HPO4 from 0.005 mol/L to 1.4 mol/L, with aflow rate of 1 mL/min. To confirm the identity of theglycerophosphoinositides, several deacylated fractionswere deglycerated (Aukema et al. 1992a), and theresulting inositol phosphates were separated using thesame HPLC system. The retention times of the derived inositol phosphates were compared with commercially labeled inositol-1-P, inositol-4-P, inositol-1,4-P2, inositol-l,3,4-P3, and inositol-1,4,5-Pa standards (Du Pont NEN Products, Boston, MA).

All data are expressed as means ±SEM. Significantdifferences between data obtained from pcy and DBA/2J mice in the first two studies were determined usingStudent's unpaired t test. For the dietary study, data

were analyzed by two-way ANOVA, followed by theprotected least significant difference test when interactions were present (SAS, version 6.08, SAS Institute, Cary, NC). Multivariate ANOVA was used todetermine the significance of the correlation between[3H]PI-3-P and total kidney weight (SAS, version 6.08,

SAS Institute). The models used included effects forprotein level, lipid composition and litter. Differences

were considered significant at P < 0.05 and interactions at P < 0.10.

RESULTS

Kidneys from diseased (pcy) mice were significantlylarger than those from normal DBA/2J mice whenexpressed in absolute terms or relative to body orliver weight (Table 2). Liver weights were slightlylower in pcy mice due to smaller body weights inthese animals; hence liver weight to body weightratios were not different. Because livers are not affected by the disease in this model of PKD (Aukemaet al. 1992b, Takahashi et al. 1991), they were used asan internal standard to compare kidney size in normaland diseased mice. In the dietary study, kidneys weresignificantly enlarged in absolute terms as well asrelative to body weight and liver weight in the pcymice fed the HP diets compared with those fed LPdiets; kidney size was not significantly altered,however, by lipid composition of the diet (datareported in Aukema et al. 1992b).

The mass levels of the phosphoinositides weredifferent in kidneys from pcy mice compared withthose from DBA/2J mice (Table 3). Both PI and PIP2were lower (by 32 and 41%, respectively) in kidneysof pcy mice compared with kidneys of normal mice,whereas the levels of PIP were comparable. As aresult, the PIP/PI molar ratio was 48% higher and thePIP2/PIP molar ratio was 41 % lower in kidneys of pcymice.

The fatty acid compositions of the renal phosphoinositides in DBA/2J and pcy mice are given inTable 4. All major fatty acids were altered in the PIfraction in kidneys of pcy mice compared with DBA/2J mice. In general, the fatty acid compositions of therenal PIP and PIP2 fractions were similar to that of PI.Differences in polyunsaturated fatty acid compositions in these polyphosphoinositides in normal and

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

DIET AND PHOSPHOINOSITIDES IN POLYCYSTIC KIDNEYS 1187

TABLE 3

Kidney PI, PIP, PIP2 mass and molar ratios in DBA/2/and pcy mice fed nonpurified diet1-2

DBA/21 pcy

nmol/kidneyPIPIPPIP253412.114.3±

27±0.7±

1.236512.18.4±

43*±

1.0±1.2*

mol/mol

PIP/PIPIP2/PIP

0.023 ±0.0011.19 ±0.10

0.034 ±0.003*0.70 ±0.08*

'Values are means ±SEMfor 3-4 determinations. *"P < 0.01, "P

< 0.05, significantly different from corresponding DBA/2J value.Abbreviations used: DBA/2J, normal control mice; pcy, DBA/

2FG-pcy, mice with polycystic kidney disease; PI, phos-phatidylinositol; PIP, phosphatidylinositol phosphate; PIP2, phos-phatidylinositol bisphosphate.

diseased kidneys generally were similar to the differences seen in PI, in which the levels of 18:2(n-6), 20:3(n-6), 22:5(n-6) and 22:6(n-3) were lower and 20:4(n-6) and 22:4(n-6) were higher in pcy mouse kidneyscompared with normal kidneys. In contrast to thepolyunsaturated fatty acids, the differences in thesaturated fatty acid composition (i.e., 18:0 and 16:0)in PI between normal and diseased kidneys were notreflected in the polyphosphoinositides.

In mice that had been injected with [3H]inositol 12min prior to killing, radioactivity in the upperaqueous phase (consisting predominantly of free[3H]inositol| of the kidney lipid extraction mixtureswas 28% higher in those derived from pcy mice com

pared with DBA/2J mice (Table 5). On the other hand,radioactivity recovered in the phosphoinositides (i.e.,PI, PIP, PIP2 and lysoPI) obtained from the lowerphase of the lipid extraction mixtures was 57% lowerfor pcy mouse kidneys than for DBA/2J kidneys.Overall, total (upper and lower phase) radioactivitywas higher in kidneys from pcy mice than from DBA/2J mice. In the livers of these mice, total radioactivityand radioactivity in total phosphoinositides were significantly lower (by 27 and 53%, respectively) in pcymice compared with DBA/2J mice (data not shown).The higher level of total radioactivity in the kidneysbut not in the livers of the pcy animals comparedwith normal animals suggests that [3H]inositol accumulated in cyst fluid. In addition, the fact that thelevel of radioactivity associated with the phosphoinositides in both the kidneys and liver was lowerin pcy mice suggests that the [3H]inositol delivery tothe tissues was altered in the animals with PKD. Thiscould have been caused by a dilution of the isotope byelevated levels of endogenous inositol, because renaldisease is often associated with increased plasma inositol concentrations (Clements and Diethelm 1979).We also found that serum inositol concentrationswere elevated in older pcy mice (>120 d) comparedwith normal mice (283 ±50 vs. 76+1 ¿onol/L,n = 2,P < 0.05).

The [3H]PIP/[3H]PIratio was also elevated in pcymouse kidneys, a change similar to that observed forthe mass data. In contrast, the [3H]PIP2/[3H]PIPratiodid not differ in normal vs. diseased mouse kidneys(Table 5). Because the 3H was incorporated into thephosphoinositides in the form of [3H]inositol, theradioactivity ratios reflect a molar ratio, because eachmole of PI, PIP or PIP2 has exactly 1 mol of inositol.hi livers obtained from these mice, [3H]PIP/[3H]PIand

TABLE 4

Kidney fatty acid comparisons of PI, PIP, PIP¿in DBA/2J and pcy mice fed nonpurified diet1'2

Fatty acidDBA/2JPIpcyPIPDBA/2JpcyPIP2DBA/21pcymol/

100 mol fattyacids16:018:018:118:2(n-6)20:3(n-6)20:4(n-6)22:4(n-6|22:5(n-6|22:5(n-3)22:6(n-3)5.9

±0.242.5±0.63.1±0.25.8±0.44.0±0.229.8±0.40.3±0.00.9±0.00.4±0.05.3±0.35.6

±0.1*44.7±0.2*4.4±0.4*3.6±0.4**2.4±0.4*34.7±1.0**0.5±0.0*0.6

±0.1*0.2±0.0***3.2±0.5**5.9

±0.648.3±1.15.9±0.46.0±0.43.6±0.325.1±1.00.3±0.10.6±0.10.2±0.03.3±0.38.1

±0.844.4±0.9*5.7

±0.44.4±0.82.6±0.430.5±1.6*0.6

±0.10.3±0.10.2±0.02.0±0.5*6.8

±0.347.0±1.84.8±0.24.4±0.13.8±0.427.4±1.10.4±0.10.6±0.10.2±0.03.7±0.16.2

±1.047.2±2.05.3±0.94.1±0.22.4±0.4*30.4

±1.20.6±0.10.3±0.10.2±0.12.2±0.5*

'Values are means ±SEMfor 3-4 determinations. Only the major fatty acids are presented. '"P < 0.001, **P < 0.01, *P < 0.05, significantlydifferent from corresponding DBA/2J value; *P = 0.06.

Abbreviations used: DBA/2J, normal control mice; pcy, DBA/2FG-pcy, mice with polycystic kidney disease; PI, phosphatidylinositol; PIP,phosphatidylinositol phosphate; PIP2, phosphatidylinositol bisphosphate.

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

1188 AUKEMA ET AL.

TABLE 5

Radioactivity distribution and ratios of kidneyphosphoinositides in nonpurified diet-fed DBA/2/

and pcy mice1' '

DBA/21PcyUpper

phase (aqueous)PhosphoinositidesTotal13H)PIP/[3H]PI

[3H]PIP2/[3H]PIPkBq/kidney873

±42 1113 ±62*47 ± 1 21 ± 3"*920 ±42 1133 ±65*Bq/Bq0.013

± 0.001 0.016 ± 0.001'

0.561 ± 0.048 0.527 ± 0.038

'Values are means ±SEMfor 5-8 determinations. 'P < 0.05, "'P

< 0.001, significantly different from corresponding DBA/2J values.^Mice were injected intraperitoneally with 6.66 GBq of

[3H]inositol 12 min prior to killing. Phosphoinositides are composed of [3H]PI, [3H]lyso PI, [3H]PIP and [3H]PIP2.

3Abbreviations used: DBA/2J, normal control mice; pcy, DBA/2FG-pcy, mice with polycystic kidney disease; PI, phos-phatidylinositol; PIP, phosphatidylinositol phosphate; PIP2, phos-phatidylinositol bisphosphate.

[3H]PIP2/[3H]PIPwere not different in DBA/2J compared with pcy mice (data not shown). This indicatesthat although the incorporation of [3H]inositol intothe liver was impaired in pcy mice, phosphoinositidemetabolism was not affected in the livers (which arenormal and free of cysts) of these animals. The alterations in phosphoinositide ratios in kidneys of pcymice, therefore, seem to be specifically related to thepresence of PKD.

Both dietary protein level and lipid compositionaltered the amount of [3H]inositol incorporated intorenal phosphoinositides (Table 6). The LP (comparedwith HP) and SO (compared with FO) diets were

associated with a lower incorporation of label intorenal phosphoinositides. Radioactivity in the total extraction mixture, as well as radioactivity in the upperphase, was lower in the kidneys of mice fed LP dietscompared with HP diets. In contrast, radioactivity inthe total extraction mixture and in the upper phasewas not significantly different in mice fed SO compared with mice fed FO diets. Replacing dietary SOwith FO resulted in lower [3H]PIP/[3H]PI,and proteinrestriction resulted in higher [3H]PIP/[3H]PI in pcymouse kidneys. There was an interactive effect ofprotein level and lipid composition on kidney[3H]PIP2/[3H]PIP.In the SO-fed mice, [3H]PIP2/[3H]PIPwas higher in the protein-restricted animals. Therewas no significant effect of lipid composition at eitherprotein level on this ratio.

The [3H]PI-3-Pwas the only 3-phosphorylated poly-phosphoinositide that was detected in mouse kidneysat a level of 2-3% of total [3H]PIP, as we have previously shown (Aukema et al. 1992a). Although therewas sufficient incorporation of radioactivity intothese kidneys to easily detect [3H]PI-3,4-P2, [3H]PI-3,5-P2 and [3H]PI-3,4,5-P2at a level of <0.5% of total[3H]PIP2,these isomers were not found to be present.Although differences in the levels of [3H]PI-3-P as apercentage of total [3H]PIP were not statistically significant in the kidneys from pcy mice fed differentdiets (2.95 ±1.13, 3.45 ±0.98, 1.92 ±0.80 and 2.28 ±0.93 for mice fed the HP-SO, HP-FO, LP-SO and LP-FO diets, respectively), there was a significant correlation (r = 0.62, P = 0.02) between the amount of[3H]PI-3-P(relative to total [3H]PIP)and total kidneyweight.

DISCUSSION

The data demonstrate that renal phosphoinositidecomposition and metabolism are altered in the pcy

TABLE 6

The effect of dietary protein level and lipid composition on phosphoinositide radioactivity ratios in pcy mouse kidneys1'2

HP-SOHP-FOLP-SOLP-FOEffectMBq/kidneyUpper

phase (aqueous)PhosphoinositidesTotal[3H]PIP/[3HJPI

[3HJPIP2/[3H1PIP14.4

1.6016.00.026

0.87±

1.5±0.13±1.6±

0.002±0.0817.4

1.7719.20.023

1.02±

±±±

±1.4

0.121.5Bq/Bq0.002

0.0812.7

0.9113.60.032

1.15±

1.4±0.12±1.4±

0.002±0.07a11.2

1.312.50.025

0.99±

±±±

±1.4

0.131.50.002

0.08...

ti•

*.

tt

in t

'Abbreviations used: FO, fish oil; HP, high protein; LP, low protein; pcy, DBA/2FG-pcy, mice with polycystic kidney disease; PI,

phosphatidylinositol; PIP, phosphatidylinositol phosphate; PIP2, phosphatidylinositol bisphosphate; SO, sunflower seed oil.2Mice were injected intraperitoneally with 66.6 GBq [3H]inositol 5 h prior to killing. Values are means ±SEM for 6-7 animals.

Phosphoinositides are composed of [3H]PI, [3Hjlyso PI, [3HJPIP and [3H]PIP2. *P < 0.05, **P < 0.01, ***P < 0.001, main effect of protein level; f,P < 0.05,tf, P < 0.01, main effect of oil type; int, interaction between effects of protein level and lipid type (P < 0.10); aP < 0.05, protein effect

SO diet.

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

DIET AND PHOSPHOINOSITIDES IN POLYCYSTIC KIDNEYS 1189

mouse, suggesting that changes in this signalingsystem play an important role in the pathogenesis ofPKD. The phosphoinositide abnormality that seemsto be most affected in pcy mice is the lower level ofPI?2 in diseased kidneys compared with normalkidneys. We have previously shown that [3H]PIP2islower in pcy mouse kidneys compared with normalmouse kidneys in mice that had been injected in-traperitoneally with [3H]inositol (Aukema et al.1992a). These animals were killed at a later timepoint at which the label was equilibrated into thepolyphosphoinositides, suggesting a lower mass levelof PIP2 in pcy compared with normal mouse kidneys.The present mass analysis has confirmed this finding.

The lower PI-4,5-P2 as well as PI mass in pcy micecompared with DBA/2J mice could be due todecreased synthesis or increased degradation of thesephosphoinositides. In the short-term labeling study,the radioactivity in the phosphoinositides waspredominantly due to synthesis because the micewere killed at an early time after isotope administration. In pcy compared with DBA/2J mice, renal[3H]PIP2/[3H]PIPwas similar, but the correspondingmolar ratio (mass analysis), as well as the mass ofPIP2, was lower in pcy mice. This suggests that synthesis of PI?2 (via PIP kinase activity) is not differentin kidneys of pcy and DBA/2J mice, but that degradation of this phosphoinositide (possibly via phos-pholipase C) may be elevated in the diseased kidneys.In this regard, we previously found renal diglyceridelevels to be elevated in pcy mice (Aukema et al.1992c). The higher PIP/PI radioactivity and molarratios in pcy compared with DBA/2f mouse kidneyssuggest that the PI kinase enzyme activity may beelevated in the diseased kidneys. Increased PIP/PIratios in ras transformed cells (Van der Kaay et al.1990) and elevated PI kinase activity in erb B transformed cells in culture (Kato et al. 1987) have beenreported. Several of the enzymes involved in phosphoinositide turnover, including phospholipase C andPI kinase, are activated in epidermal growth factor-mediated signaling events (Bjorgeet al. 1990, Pike andHakes 1987). Epidermal growth factor metabolism hasbeen reported to be altered in PKD (Lakshmanan andEysselein 1993, Wilson and Sherwood 1991). In thisregard, the c-erb B-2 proto-oncogene has been foundto be overexpressed in several forms of PKD (Herrera1991). Members of this proto-oncogene family encodea receptor-like protein that closely resembles theepidermal growth factor receptor. The alterations inphosphoinositide metabolism may therefore reflectalterations in epidermal growth factor metabolismand/or oncogene expression. In this regard, [3H]lysoPI/[3H]PIwas also elevated in pcy compared with DBA/2J mouse kidneys (short-term labeling study, data notshown), suggesting a higher level of phospholipaseAi/2 activity in these kidneys. Activation of phospholipase A2 has also been reported to be associated

with epidermal growth factor-mediated events (Bon-venture et al. 1990, Goldenberg et al. 1990).

The lower incorporation of [3H]inositol into thephosphoinositides in the short-term labeling studymay have been a result of a lower rate of synthesis of[3H]PI in pcy compared with DBA/2J mouse kidneys.This lower [3H]inositol incorporation into the phosphoinositides, however, is also consistent with a dilution of the isotope by elevated plasma inositol concentrations in pcy mice compared with DBA mice.Renal disease is often associated with increasedplasma inositol concentrations, because the kidney isthe major regulator of inositol in the body (Clementsand Diethelm 1979). A dilution of [3H]inositol byhigher plasma inositol concentrations could also explain the lower levels of [3H]inositol in the livers ofpcy mice compared with normal mice. The highertotal incorporation of [3H]inositol into kidneys frompcy mice compared with DBA/2} mice may have beendue to accumulation of [3H]inositol in cyst fluid. Inthe dietary study, the larger kidneys derived frommice fed HP diets compared with LP diets also accumulated more [3H]inositol, indicating that labelmay accumulate in cyst fluid.

The data demonstrate that in vivo renal phosphoinositide metabolism in pcy mice can be alteredby dietary protein and lipid manipulation. In the SO-fed animals, renal [3H]PIP2/[3H]PIP(derived from pcymouse kidneys that had been labeled with[3H]inositol to equilibrium) was lower in mice fed theHP diets relative to those fed the LP diets. We haveshown that LP diets compared with HP diets significantly reduce the progression of cystic disease in pcymice (Aukema et al. 1992b). Lower renal [3H]PIP2/[3H]PIPin animals fed the HP diet rather than the LPdiet is consistent therefore with the idea that abnormalities in this ratio are directly related to thedisease. Lower PIP2/PIP ratios were associated withboth the presence of disease (mass data) and a greaterprogression of disease (isotopie data in mice fed HPcompared with LP diets) in SO-fed mice. In the dietary FO groups, however, renal [3H]PIP2/[3H]PIPwasnot lower in mice fed HP diets compared with thosefed LP diets. This may have been due to an interaction between FO treatment and dietary proteinlevel with an effect on phosphoinositide turnover.Over the long term, dietary FO results in a shortenedlife span and earlier onset of proteinuria in male pcymice (Aukema et al. 1992e). The effects of dietaryprotein level on renal phosphoinositide metabolismin the FO-fed animals may therefore have beenmasked by detrimental effects of dietary FO on theseanimals. Effects of (n-3) fatty acids on phosphoinositide turnover have previously been determined in activated platelets obtained from humansand rabbits (Medini et al. 1990, Skeaff 1988), as wellas cultured vascular muscle cells and isolated humanplatelets incubated with FO or a purified (n-3) fatty

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

1190 AUKEMA ET AL.

acid, 22:6(n-3) (Gaudette and Holub 1990, Locher etal. 1989). These studies indicated that dietary alterations have the potential to influence phosphoinositidemetabolism.

There were also differences in [3H]PIP/[3H]PI in

kidneys from mice fed different diets. If the labelingof the phosphoinositides to equilibrium is analogousto the molar ratios obtained from the mass analysis,the mice fed the HP diets would be expected to havehad higher [3H]PIP/[3H]PI, because kidneys of these

mice are more cystic (Aukema et al. 1992b). This wasnot the case, however, suggesting that dietary effectson [3H]PIP/[3H]PI may be different from those effectsthat are associated with the presence of disease.Nevertheless, it is evident that in vivo phosphoinositide metabolism can be modulated by diet.

The predominant renal polyphosphoinositideisomers in the DBA/2J and pcy mice are PI-4-P and PI-4,5-P2, with the only 3-phosphorylated isomer detected being [3H]PI-3-P at a level of 2-3% of total[3H]PIP (this study and Aukema et al. 1992a). Al

though there were no significant differences in renal[3H]PI-3-P levels in mice fed the different diets, there

was a positive correlation between the amount of[3H]PI-3-P as a percentage of total PIP and totalkidney weight. Kidney weight has been shown to becorrelated with the progression of PKD in this model(Aukema et al. 1992b, Takahashi et al. 1991). Thiscorrelation between [3H]PI-3-P and renal cyst en

largement is in agreement with our previous study(Aukema et al. 1992a), which demonstrated the possible importance of this isomer in vivo in cellulargrowth. Interestingly, the enzyme that produces PI-3-P, PI-3-kinase, has also been shown to be associatedwith epidermal growth factor metabolism (Bjorge etal. 1990).

The renal phosphoinositides in both DBA/2J andpcy mice contain predominantly 18:0 and 20:4(n-6),

which is characteristic of phosphoinositides in othertissues. The lower 22:6(n-3) levels in the phosphoinositides are consistent with lower 22:6(n-3)levels in other phospholipids (except diphosphatidyl-

glycerol) in pcy compared with DBA/2J mousekidneys (Aukema et al. 1992d). The importance of thelower 22:6(n-3) levels in pcy mouse kidney phosphoinositides for cell signaling and/or membranefluidity-related enzyme activities in PKD remains tobe elucidated.

In conclusion, in vivo phosphoinositide composition and metabolism are altered in a murine modelof PKD. In addition, changing the composition ofdiets (protein level and lipid composition) fed to pcymice resulted in alterations in phosphoinositide metabolism. This study demonstrates the potential ofdiet to influence intracellular signal transductionsystems associated with the pathogenesis of PKD.Further studies of the effects of diet on biochemicalevents downstream of the phosphoinositide pathway

are required to delineate the molecular mechanismsby which nutrients exert their biological effects.

ACKNOWLEDGMENT

The authors wish to thank Christopher A. Jolly forhis excellent technical assistance.

LITERATURECITED

Allan, D. & Micheli, R. H. (1978) A calcium-activated polyphosphoinositide phosphodiesterase in the plasma membrane ofhuman and rabbit erythrocytes. Biochim. Biophys. Acta 508:277-286.

Ananth, U. S., Ramakrishnan, C. V. & Hauser, G. (1986) Effects ofpre- and post weaning undemutrition on polyphosphoinositidepools in rat kidney. Lipids 21: 226-229.

Aukema, H. M., Chapkin, R. S., Tomobe, K., Takahashi, H. &Holub, B. J. (1992a) In vivo formation of polyphosphoinositideisomers and association with progression of murine polycystickidney disease. Exp. Mol. Pathol. 57: 39-46.

Aukema, H. M., Ogborn, M. R., Tomobe, K., Takahashi, H., Hibino,T. & Holub, B. J. (1992b| Effects of dietary protein restrictionand oil type on the early progression of murine polycystickidney disease. Kidney Int. 42: 837-842.

Aukema, H. M., Tomobe, K., Takahashi, H. & Holub, B. J. (1992c)Altered renal diacylglycerol mass and fatty acid compositions inmurine polycystic kidney disease: dietary effects. J. Nutr.Biochem. 3: 491^95.

Aukema, H. M., Yamaguchi, T., Takahashi, H., Celi, B. & Holub, B.J. (1992d) Abnormal lipid and fatty acid compositions of kidneysfrom mice with polycystic kidney disease. Lipids 27: 429-435.

Aukema, H. M., Yamaguchi, T., Takahashi, H., Philbrick, D. J. &Holub, B. J. (1992e) Effects of dietary fish oil on survival andrenal fatty acid composition in murine polycystic kidneydisease. Nutr. Res. 12: 1383-1392.

Berridge, M. J. & Irvine, R. F. (1989) Inositol phosphates and cellsignaling. Nature (Lond.) 341: 197-205.

Bjorge, J. D., Chan, T.-O., Antczak, M., Kung, H.-J. & Fujita, D. J.(1990) Activated type I phosphatidylinositol kinase is associatedwith the epidermal growth factor (EGF) receptor following EGFstimulation. Proc. Nati. Acad. Sci. U.S.A. 87: 3816-3820.

Bonventure, ]. V., Gronich, J. H. & Nemenoff, R. A. (1990)Epidermal growth factor enhances glomerular mesangial cellsoluble phospholipase A2 activity. }. Biol. Chem. 265:4934-4938.

Calvet, J. P. (1993) Polycystic kidney disease: primary extracellularmatrix abnormality or defective cellular differentiation? KidneyInt. 43: 101-108.

Christie, W. W. (1982) Lipid Analysis, 2nd éd.,pp. 52-56. PergamonPress, Toronto, Canada.

Clements, R. S., Jr. & Diethelm, A. G. (1979) The metabolism ofmyoinositol by the human kidney. J. Lab. Clin. Med. 93:210-219.

Gabow, P. A. (1993) Autosomal dominant polycystic kidneydisease. N. Engl. J. Med. 329: 332-342.

Gaudette, D. C., Aukema, H. M., Jolly, C. A., Chapkin, R. S. &Holub, B. J. (1993) Mass and fatty acid composition of the3-phosphorylated phosphatidylinositol bisphosphate isomer instimulated human platelets. J. Biol. Chem. 268: 13773-13776.

Gaudette, D. C. & Holub, B. J. (1990) Effect of albumin-bound DHAon phosphoinositide phosphorylation in collagen stimulatedhuman platelets. Thromb. Res. 58: 435-444.

Goldenberg, H. J., Viegas, M. M., Margolis, B. L., Schlessinger, J. &.Skorecki, K. L. (1990) The tyrosine kinase activity of the

by guest on March 14, 2015

jn.nutrition.orgD

ownloaded from

DIET AND PHOSPHOINOSITIDES IN POLYCYSTIC KIDNEYS 1191

epidermal-growth-factor receptor is necessary for phospholipaseA2 activation. Biochem. J. 267: 461-465.

Grantham, J. J. (1993)Fluid secretion, cellular proliferation, and thepathogenesis of renal epithelial cysts. J. Am. Soc. Nephrol. 3:1843-1857.

Herrera, G. A. (1991)C-erb B-2amplification in cystic renal disease.Kidney Int. 40: 509-513.

Holub, B. f. & Kuksis, A. (1972) Further evidence for the interconversion of monophosphoinositides in vivo. Lipids 7: 78-80.

Holub, B. J. & Skeaff, C. M. (1987)Nutritional regulation of cellularphosphatidylinositol. In: Methods in Enzymology (Conn, P. &Means, A. R., eds.), vol. 141, pp. 234—244.Academic Press, NewYork, NY.

Kato, M., Kawai, S. & Takenawa, T. (1987) Altered signal trans-duction in erbB-transformed cells. J. Biol. Chem. 262:5696-5704.

Lakshmanan, J. & Eysselein, V. (1993)Hereditary error in epidermalgrowth factor prohormone metabolism in a rat model of au-tosomal dominant polycystic kidney disease. Biochem. Biophys.Res. Commun. 197: 1083-1093.

Locher, R., Sachinidis, A., Steiner, A., Vogt, E. & Vetter, W. (1989)Fish oil affects phosphoinositide turnover and thromboxane Ametabolism in cultured vascular muscle cells. Biochim. Biophys. Acta 1012: 279-283.

Medini, L., Colli, S., Mosconi, C., Tremoli, E. M. & Galli, C. (1990)Diets rich in (n-9), (n-6) and (n-3) fatty acids differentially affectthe generation of inositol phosphates and of thromboxane bystimulated platelets, in the rabbit platelet. Biochem. Pharmacol.39: 129-133.

Nishizuka, Y. (1992) Intracellular signaling by hydrolysis of phos-pholipids and activation of protein kinase C. Science 258:607-614.

Parker, P. J. & Waterfield, M. D. (1992) Phosphatidylinositol3-kinase: a novel effector. Cell Growth & Diff. 3: 747-752.

Pike, L. J. & Eakes, A. T. (1987)Epidermal growth factor stimulatesthe production of phosphatidylinositol monophosphate and thebreakdown of polyphosphoinositides in A431 cells. J. Biol.Chem. 262: 1644-1651.

Rana, R. S. & Hokin, L. E. (1990) Role of phosphoinositides intransmembrane signaling. Physiol. Rev. 70: 115-164.

Skeaff, C. M. (1988) Fish Oil Supplementation in Humans: Effectson Platelet Responses, Phospholipid Composition and Metabolism. Doctoral thesis, University of Guelph, Guelph, Canada.

Takahashi, H., Calvet, J. P., Dittemore-Hoover, D., Yoshida, K.,Grantham, I. ]. & Gattone, V. H. (1991) A hereditary model ofslowly progressive polycystic kidney disease in the mouse. J.Amer. Soc. Nephrol. 1: 980-989.

Thomas, L. M. & Holub, B. J. (1991) Eicosanoid-dependent and-independent formation of individual [14C]stearoyl-labeledlysophospholipids in collagen-stimulated human platelets. Biochim. Biophys. Acta 1081: 92-98.

Tomobe, K., Philbrick, D., Aukema, H. M., Clark, W. F., Ogbom,M. R., Parbtani, A., Takahashi, H. & Holub, B. J. (1994) Earlydietary protein restriction slows disease progression andlengthens survival in mice with polycystic kidney disease. J.Am. Soc. Nephrol. 5: 1355-1360.

Van der Kaay, J., Draijer, R. & Van Haastert, P.J.M. (1990)Increasedconversion of phosphatidylinositol to phosphatidylinositol phosphate in Dictyostelium cells expressing a mutated ras gene.Proc. Nati. Acad. Sci. U.S.A. 87: 9197-9201.

Varticovski, L., Harrison-Findik, D., Keeler, M. & Susa, M. (1994)Role of PI 3-kinase in mitogenesis. Biochim. Biophys. Acta 1226:1-11.

Vinson, f. A. & Hooyman, J. E. (1977)Sensitive fluorogenic visualization reagent for the detection of lipids on thin layer chromato-grams. J. Chromatogr. 135: 226-228.

Wilson, P. D. & Sherwood, A. C. (1991) Tubulocystic epithelium.Kidney Int. 39: 450-463. by guest on M

arch 14, 2015jn.nutrition.org

Dow

nloaded from