Effects of Vitamin B-12 Deprivation on Phospholipid Fatty Acid Patterns in Liver and Brain of Rats Fed High and Low Levels of Linoleate in Low Methionine Diets1 JAMES J. PEIFER 2 ANDRICHARD D. LEWIS 3 Department of Foods and Nutrition, University of Georgia, Athens, Georgia 30602 ABSTRACT Experiments were conducted to determine how 20 weeks of deprivation of vitamin B-12 (B-12) would influence polyunsaturated fatty acid (PUFA) patterns in cerebral and liver phospholipids of female rats. All experimental diets included 24% soybean protein and either 3.5 or 0.35% of linoleate (18:2). Deprivation of B-12 and restricted intakes of 18:2 significantly inhibited growth of the rats. Rats depleted of B-12 had 6.7 times more methylmalonic acid in their liver, and significantly more odd-numbered acids ( 15 :0 + 17:0 ) were present in phosphatidyl choline (PC) of their liver and cerebrum. Deprivation of B-12 promoted 63.3 to 97.3% increases of 18:2 in PC and phosphatidyl ethanolamine (PE) of the liver. Rats deprived of B-12 had significantly less arachidonate (20:4w6) and 22:5w6 in their liver PC, as well as smaller amounts of 20:4«>6and 22:6o)3 in their cerebral PC. Some of the observed changes appear to be due to a significant interaction between the effects of B-12 and linoleate. Changes in PUFA patterns of cerebral and hepatic PC suggest that B-12 deprivation may interfere with the conversion of 18 :220 :4o>6— » 22 :.lofi. Some of the observed changes may also be related to an inhibited ability to convert PE to PC. J. Nutr. 109: 2160-2172, 1979. INDEXING KEY WORDS vitamin B-12 •linoleate •phospholipidfatty acids •brain •liver Several reports (1, 2) have suggested that development of neurological lesions of pernicious anemia may be partially due to the progressive accumulation of odd-numbered and branched-chain saturated fatty acids into cerebrosides and phospholipids of neural tissues. Similar fatty acid isomers have been found in rat glial cells grown in vitamin B-12 (B-12) depleted culture medium (3) and in liver lipids of B-12 deprived baboons (4). These altered satu rated fatty acid patterns occur in response to
. . c an interference
... With
. , propionate
, , metab-
bolic block tissue levels of propionyl CoA and MMCoA are greatly increased, and this promotes greater synthesis and accumulation of odd-number and branchedchain, 15 and 17 carbon, fatty acid isomers (2, 6, 7). Such acids accounted for 13% of the total fatty acids in phosphatidyl choline ( PC ) of myelinated nerves of an infant with severe methylmalonic aciduria (1). Any increased incorporation of unusual Received for publication September 18, 1978. , Supported in part by general research funds from
olism caused by a block of methylmalonyl CoA
.
(MMCoA)-B-12 j ,
.
dependent * r ......
mutase .
required1 /-« forA inconversion of MMCoA to K o\ \XT-LL. L SUCCinyl
CoA
(Z, 5, O).
With
SUCh a meta-
2160
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
périmentStation-Regional S-87 Project. 2To whom requests for reprints should be ad-
dressed. 'Includes fulfillment
data from thesis of R. Lewis as partial of the requirements for the M.S. degree.
VITAMIN B-12, LINOLEATE AND TISSUE PHOSPHOLIPIDS
fatty acids into cerebrosides and phospho lipids could conceivably interfere with the formation and integrity of myelin, neural and other biological membranes (8-10). Longer chain polyunsaturated fatty acids (LC-PUFA) are obligatory phospholipid components in neural and other tissues ( 10, 11), and a progressive depletion of B-12 could conceivably induce significant altera tions in PUFA patterns of such phospholipids. In B-12 deprived rats the elevated tissue levels of MMCoA inhibits biosynthesis of malonyl CoA required for elongation of linoleate (18:2) 4 to form arachidonate (20:4) and other longer chain PUFA (5, 6, 12). Alterations in phospholipid PUFA Eatterns could also occur as a result of a lock of the B-12 dependent methyltransferase required for conversion of homocysteine to methionine (13-15). When B-12 deprived rats receive limited dietary supplies of methionine, a block of the methyltransferase reaction could limit the available supply of S-adenosylmethionine (SAM) (15). A limited availability of SAM would interfere with the conversion of some PUFA-rich phosphatidyl ethanolamine (PE) to PUFA-rich PC (13, 14). Rats deprived of B-12 may also be capable of more rapid catabolism of linoleate ( 16), and some reported data (17) suggest that high intakes of linoleate may promote more rapid depletion of B-12 from tissues of the rat. Any one, or combinations, of these sug gested B-12 links to PUFA metabolism could be expected to significantly affect PUFA patterns found in major phospholipids of liver and neural tissues. Support for such an interrelationship is suggested by the report of Noble and Moore (18) who found B-12 depleted chick embryos to have altered PUFA patterns in their liver phospholipids. The present study was designed to de termine how 20 weeks of B-12 deprivation might affect growth, PUFA and other fatty acid patterns in major phospholipids in the liver and cerebrum of female rats. Groups of rats were fed 3.5 and 0.35% levels of linoleate. Should B-12 deprivation accelerate catabolism of PUFA, this should be most apparent in those rats receiving the low, but adequate (19, 20), levels of
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
2161
linoleate. The 24% level of soybean pro tein provided at least marginal amounts of methionine and other amino acids needed for the growing rat5 ( 21 ). Measurements of methylmalonic acid and odd-number acids in phosphatidyl choline served as biochemical markers of the relative deple tion of B-12 from tissues of the rats.6 MATERIALS
AND METHODS
Nutritional procedures. Sprague-Dawley strain, female rats,7 35 to 50 g in weight, were initially housed in groups of three in clear plastic cages. These cages, measuring 35 X 31 X 16 cm, had stainless-steel screen tops and wire mesh bottoms which raised the rats one-inch from a moisture absorb ing corrugated cardboard liner.8 During the last 13 weeks of the study, each rat was individually housed in galvanized hanging, screen bottom cages. The animal facility was maintained at 26°with a 12-hour daynight lighting schedule. Food and tap water were allowed ad libitum, and fresh water and food were supplied every 2 to 3 days. Fresh diets were prepared every 2 to 3 weeks and refrigerated until needed. On arrival at the laboratory, all rats were fed the control diet (diet HLS in table 1) for 1 week before being divided into groups having equivalent body weights. All diets included 24yc soy protein, 20%> total fat, 3.5 or 0.35% linoleate, 3% sodium propionate and 0.01% thyroid powder (table 1). In addition to supplying at least mar ginal amounts of methionine and other amino acids needed for growth of B-12 supplemented rats 5 (21), the 24% level of soy protein should permit maintenance of 1Fatty acids are designated by their carbon chain length and number of double bonds. Heptadecanolc acid is shown as 17:0 or 17:0br to designate either the linear or branched form of this saturated acid. Omega (in) designation is used to identify members of the linoleate (iu6), linolenate (w3 or oleate (w9) families of PUFA. An eicosatrienoic acid (20:3u9), formed through elongation and desaturation of oleate, accumulates in tissue llplds with the onset of an essential fatty acid deficiency (39). 5Related studies at our laboratory Indicate that growth of rats fed 24% soy protein diets Is 85 to 90% of that obtained with 24% casein diets when both diets include B-12, folie acid and adequate amounts of other nutrients. 6One of a series of ancillary studies concerned with dietary parameters likely to affect nutritional health of préadolescents ; a Georgia Experiment Sta tion—Regional S-87 Project. 'Charles River Breeding Laboratories, Wilmington, Massachusetts. 8Deotized corrugated cage boards, Upjohn Co., Kalajnazoo, Michigan.
2162
JAMES J. PEIFER
AND RICHARD
D. LEWIS
tion of tissue levels of B-12 from those groups not receiving supplements of the vitamin (24-26). 3.5% 0.35% Isolation of tissue lipids. Rats were fasted Linoleate Linoleate overnight (12 to 15 hours) and exsan IngredieEts (HL-) diets (LL-) diets guinated while under ether anesthesia. The protein1SucroseCellulose1Mineral Soybean liver and cerebrum of each rat were quickly excised, weighed, sealed in plastic bags and frozen until needed for lipid mixture2Sodium analysis. In this report cerebrum refers to propionate1Thyroid the section of rat's brain in front of the powder1L-cystine3Linoleate4Saturated cerebellum-forebrain junction except for the olfactory nerve tracts and pituitary triglycérides6Vitamin gland. Cerebral and liver tissue homogemixture6(B-12, „g)724.043.34.03.53.00.010.253.516.52.0(0.0 or 5.0)24.043.34.03.53.00.010.250.3519.652.0(0.0 or 5.0) nates were dispersed in a minimum of 20 volumes of chloroform-methanol (2:1) 9 1Nutritional Biochemicals—ICN Pharmaceuti and total lipids were isolated in chloro cals, Inc., Cleveland, Ohio. Cellulose purchased as form according to procedures of Folch, alphacel. 2Supplied as mg/100 g diet: CaCO3, 725; CaHPO4,1,130; CuSO4, 5.0; MgSO4, 230; KC1, Lee and Sloane (27). 730; ferric citrate-5H2O, 59.0; MnSO4-H2O, 15.4; Tissue levels of methylmalonic acid KIOj, 0.32; ZnCOj, 5.5; Na2HPO4, 600; chromium (MMA) serve as a metabolic marker of a acetate, 0.05; sodium selenite, 0.01; ammonium B-12 deficiency condition in the rat (5, 6). molybdate, 0.05. 3Fisher Scientific Co., Fair Lawn, New Jersey. *Contributed by the inclusion Our initial attempts to measure tissue of 7.0 or 0.70% Mazóla corn oil (Best Foods, Englelevels of MMA suggested that freshly pre wood Cliffs, New Jersey). The 7.0% dietary level of pared homogenates were best suited for corn oil included 3.5% linoleate, 2.0% oleate, 0.05% this analysis. Three control females and linolenate, 0.9% palmitate and 0.2% stéarate.The 0.70% level of corn oil supplied one-tenth levels of three B-12 deprived females (diets HLS these same acids. 6Saturated triglycérides—sup and HLD, respectively) were randomly plied by hydrogenated coconut oil (HCO) (Proctor selected at the end of 20 weeks from a and Gamble Co., Cincinnati, Ohio) and the nonsecond, longer term experiment. These rats linoleate components of corn oil. This saturated triglycéridemixture in HL-diets included the follow were killed by decapitation, their liver ing: 5.6% laurate (12:0), 2.4% myristate (14:0), excised and immediately homogenized in 2.2% palmitate (16:0), 2.1 stéarate (18:0), 2.0% KH,,PO4 buffer (pH 7.0). An aliquot of oleate (18:1) and 1.0% of shorter chain acids (8:0 and 10:0). *Vitamins, mixed in powdered sucrose, the homogenate was immediately analyzed supplied as mg/100 g diet: thiamin-HCL, 0.05; for both methylmalonic and succinic acids. riboflavin, 0.5; niacin, 4.0; pyridoxine-HCl, 0.2; To this aliquot, representing one-fifth of inositol, 10.0; calcium pantothenate, 4.0; folie acid, the total liver homogenate, was added an 0.2; menadione, 0.2; biotin, 0.05; choline chloride, 100; vitamin B-12 (when supplied), 0.005. Fat internal standard (1.0 mg) of orthohysoluble vitamins, dissolved in the hydrogenated droxyphenylacetic acid 10 ( O-HPAA ). Fol coconut oil of the diet, supplied as IU/100 g diet : lowing hydrolysis of their coenzyme A retinol, 825; cW-a-tocopherol, 10; ergocalciferol, 80. esters, MMA, succinic acid and O-HPAA 'Values in parenthesis are the amounts of vitamin B-12 included (0.0 or 5.0 ^g) in 100 g of HL- and were extracted with ethyl ether using mod LL-diets. ifications of methods described by Smith et al. (28) and Gibbs et al. (29). normal phospholipid levels in their liver Analytical methods. Hepatic and cere and plasma (22). The 20% level of fat pro bral total lipids were fractionated into vides 40c/c of the total calories, supports phosphatidyl choline (PC) and phosphatidyl good growth of rats (23), and it permits ethanolamine (PE) by a micro-thin layer considerable flexibility for varying compo chromatography (micro-TLC) technique sition of the dietary fat (i.e., varying lin (30). This included use of 8 X 10 cm silica gel oleate levels) without altering total fat G1^coated microchromatoplates and a decontent of the diet. Corn oil triglycérides served as a source of linoleate. Sodium pro *Ratios given for solvent mixtures are volumetric. 10ICX Pharmaceuticals, Inc.. Plainview, New York. pionate and thyroid powder were added to "Applied Science Laboratories Ltd., State College, all diets in an attempt to accelerate deple Pennsylvania. TABLE I Percentage composition of experimental diets
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
VITAMIN B-12, LINOLEATE AND TISSUE PHOSPHOLIPIDS
2163
veloping solvent system of chloroformTABLE 2 methanol-water (80:25:3).9 These experi Growth response to B-1% deprivation and low intakes of linoleate1 mental conditions permit relatively rapid separation of PE and PC into two distinct Heart TLC-bands that are separated by more Diet' Weight gain» Liver ventricle Cerebrum than 2 cm. Location of the PE and PC g/SO weeks bands were done by the use of standards 242.1 ± 9.9° 9.3 ±0.6 0.84 ±0.03 1.38 ±0.02 HLS prepared from egg yolk. Micro-TLC plates 179.8±11.9' HLD 8.8±0.5 0.77 ±0.10 1.27 ±0.10 containing PC and PE standards were de 194.7 ±11.0' LLS 9.9 ±0.6 0.82 ±0.03 1.37 ±0.02 170.6 ± 5.1» 11.3 ±0.0 0.73 ±0.03 1.35 ±0.02 veloped simultaneously with those spotted LLD with the cerebral or hepatic lipids. DichloAnalysis of variance (P-values)—growth data1 B-12 Linoleate Interaction rofluorescein and ninhydrin (for PE band) sprays were used to identify PC and PE NS 0.001 0.01 bands, and the respective phospholipids were eluted from the TLC-band scrapings 1Means ±BEMof nine rata in HLS, LLS, %LLD ; eight rate in with chloroform-methanol (1:1).°Fatty acid HLD. «HLS = high linoleate (3.5%) + B-12; HLD - high linoleate—B-12; LLS - low linoleate (0.35%) + B-12; LLD components of these phospholipids were = low linoleate—B-12. ' Means not sharing a common converted to methyl esters by transesterifisuperscript letter differ significantly at P < 0.05. « Signifi cation using sodium methoxide catalyst cant differences (P < 0.001 and 0.01) attributable to the effect«of B-12 or linoleate. NS indicates absence of a ( 31 ). Methyl esters of the fatty acids were dietary significant interaction between B-12 and linoleate. analyzed by gas liquid chromatography ( GLC ) using an instrumentl- equipped and MM A were identified on the basis of with dual hydrogen flame detectors, quartz their retention times relative to O-HPAA burner jets and facilities for linear tem the use of reference standards of pure perature programming. Glass columns and succinic10 and methylmalonic10 acids. (2 m X 6 mm id) were packed with 10% Quantitative analyses of these acids were (w/w) diethyleneglycol succinate poly ester11 coated on 100 to 120 mesh Gas based on the following relationships : Chrom Q J1 treated with phosphoric acid (32) and operated isothermally at 182° A. Peak areao-HpAA _ Peak mg O-HPAA mg MMA with nitrogen as the carrier gas. Identifica tion and quantitative analysis of specific B. mgof MMA = pe,a areaMMA X2.51. fatty acid components were based on com parative retention time data with those of Peak areao_HPAA known standards,13-14comparisons of equiv „ c „ Peak area sA *, 0 no X 2.03. alent chain length (ECL) values (33) and C. mg of SA = o—¡— Peak areao-HPAA related GLC-techniques described pre viously (34). A reference mixture of In order to equate GLC peak response methyl esters of fatty acids of rat testicular per mg of pure methylmalonate (MMA) lipids, which is particularly rich in 22:5c»6 with that of the internal standard, (35), was used for further confirmation of O-HPAA, it was necessary to introduce a some of the longer chain polyunsaturated correction factor, FMMA.Under our experi fatty acids (LC-PUFA). mental conditions FMMAwas 2.51, and the Ether extracts, containing methylmalonic corresponding correction factor for methyl and succinic acids and the O-HPAA in succinate, FSA,was 2.03. ternal standard, were methylated and silanated simultaneously by use of a com '- Packard-Becker, model No. 419, Donner» Grove. mercially available Tri Syl BSA reagent.15 Illinois. 13Supelco, Inc.. Bellefonte, Pennsylvania. Purchases The resulting methyl ester-silyl derivatives included GLC-PUPA mixtures (PUFA Nos. 1 and 2) containing 20:5w3, 22:4uG and 22:0w3. 14Nu Chek Prep, Inc., Elyslan, Minnesota. GLCwere analyzed by linear temperature pro grammed (80° to 210°at a rate of 6°/ »tandards Nos. 10A and 14A included 10 different straight chain fatty acid methyl esters including minute ) GLC12 using glass columns pentadecanoate (15:0) and heptadecanoate (17:0). GLC-standards Nos. 1A, 2B, 3A1, 6A and 7A Included packed with 3% (w/w) OV-1 " on 100 to most of the other fatty acid-methyl esters identified components of PC and PE. 120 mesh Chromosorb W.11 Succinic acid as «Pierce Chemical Co., Rockford, Illinois.
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
2164
JAMES J. PEIFER
AND RICHARD
D. LEWIS
TABLE 3 Methylmalonic acid and odd-numbered fatly acid concentrations related to intakes of B-lê and linoleate1 acids'15:0Trace PC-odd-numbered
acids6-615:0Trace PC-odd-numbered
Diet¡ILS malonic1mg/100 g '100 g fatty acid liver g fatty acid 1.3 ±0.2« 3.7 ±1.0°(3) 0.56 ±0.10 0.58 ±0.06« 0.31 ±0.04 0.36 ±0.04 8.7 ±0.9»Buccinate4MMAliver 0.7 ±0.1» (3)Cerebral HLD 0.29 ±0.04 0.74 ±0.14 1.12 ±0.24» 0.24 ±0.04 0.40 ±0.03 0.63 ±0.06 (8) (8) 0.44 ±0.06 0.54 ±0.05°(7) Trace LLS 0.12 ±0.03 0.27 ±0.08 0.39 ±0.09 (») 0.64 ±0.08 1.06 ±0.17»(7) LLDIMethyl 0.39 ±0.0817:0Totalgl 0.25 ±0.10 0.57 ±0.15(8) (8)Liver 0.32 ±0.0417:0Totala/100 (8) Analysis of variance—effe-ta of Õ.P)on total odd-numbered acids7 Cerebral PC Liver PC B-12
Linoleate
0.005
NS
Interaction NS
B-12
Linoleate
0.05
NS
Interaction
1 Means ±SEM of fatty acids isolated from the cerebrum and liver of the numbers of rats shown in the parenthesis. Means not sharing a common superscript letter in the same column differ significantly at P < 0.05. Number of rata per group is shown in parenthesis. 'HLS - high linoleate (3.5%) + B-12; HLD - high linoleate—B-12; LLS = low linoleate (0.35%) + B-12; LLD = low linoleate—B-12. 'Methylmalonic acid (MMA) and succinic acid analysis measured in fresh liver homogenates prepared from each of three rats fed HLS and three rats fed HLD. • Ratios of succinate to MMA (Succinate/MMA) in liver homogenates of rats. 'PC= phosphatidylcholine. Pentadecanoic (15:0) and heptadecanoic (17:0) acids were the only major odd numbered acids detected in these phospholipids. Trace = less than 0.10% of total fatty acids in PC. 'Total odd numbered acids (15:0 + 17:0) in liver PC of B-12 deprived rats (HLD + LLD) were significantly greater than those of B-12 supplemented groups (HLS + LLS) at P < 0.05. 'Significant differences (P < 0.005 or 0.05) attributable to dietary B-12. NS indicates absence of significant differences for effects of linoleate or an interaction of B-12 and linoleate on PC odd-numbered acids.
Statistical analysis. All data are presented as means ±SEM. Where comparisons were limited to two groups, significant differ ences on Stu dent's between f-test (36).means Data were on thebased comparative effects of altered intakes of B-12 and linole ate were subjected to a two-way analysis of variance (ANOVA). Where significant F ratios were obtained for the effects of B-12 or linoleate, or an interaction between them, significant differences ( P < 0.05 ) between appropriate were identified using Tukey's test asmeans described by Steel and Torrie (36) and Linton and Gallo (37). The minimal level of statistical sig nificance accepted was P < 0.05. All ANO VA calculations were based on data obtained from equal numbers of rats per group ( 7 or 8 ). Where groups included an extra rat, tables of random numbers were used to select specific rat data to be elimi nated from the ANOVA calculations. RESULTS
The female rats gained significantly less weight when they were deprived of B-12 or fed diets containing a limited level (0.35%) of linoleate (table 2). Those rats fed the 3.5% level of linoleate (18:2o>6) 4 and de
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
prived of B-12 (HLD) weighed 25.77, less than their B-12 supplemented controls ( HLS ). Limiting dietary levels of linoleate to only 0.357o, with or without B-12 (LLS or LLD), had a similar growth inhibitory effect on the rats. Results of analysis of variance showed that growth was signifi cantly affected by the dietary effects of B-12 and linoleate but not an interaction between them. Growth differences among the groups were not accompanied by com parable differences in weights of the liver, heart ventricle or cerebrum. Rats deprived of B-12 (HLD) for 20 weeks had significantly more methylmalonic acid (MMA) in their liver (table 3). The HLD rats had MMA concentra tions of 8.7 ±0.9 mg/100 g liver and this was 6.7 times as much as that found in their controls (HLS). Ratios of succinate to MMA were significantly lower in rats deprived of B-12. The onset of a significant metabolic block of the B-12 dependent mutase system could be expected to pro mote the observed changes in MMA and ratios of succinate: MM A (15). Small, but significantly greater, amounts of odd-numbered and branched-chain acids have been reported to accumulate in phos-
VITAMIN
B-12, LINOLEATE
phatidyl choline (PC) of neural tissues of a human subject having a blocked mutase system (1). In our study, rats deprived of B-12 accumulated significantly greater amounts of odd-numbered acids in PC of their cerebrum and liver (table 3). Pentadecanoic acid (15:0) * usually accounted for only trace amounts (less than 0.10%) of the total fatty acids in cerebral or liver PC of B-12 supplemented rats (HLS and LLS). The same PC components in B-12 deprived rats (HLD or LLD) contained two to three, or more, times as much 15:0. Increased amounts of heptadecanoic acid (17:0) 4 were also apparent in PC of the cerebrum, but not liver, of rats deprived of B-12. Cerebral PC of B-12 depleted rats contained 93.1 to 96.3% more odd-num bered acids (15:0+17:0) than their B-12 supplemented controls (HLD versus HLS, LLD versus LLS). Analysis of variance of the data showed that the concentrations of odd-numbered acids ( 15:0 + 17:0 ) in cerebral and liver PC were significantly af fected by the dietary effects of B-12, but not linoleate or an interaction between
AND TISSUE PHOSPHOLIPIDS
2165
B-12 and linoleate. Comparatively greater amounts of these odd-numbered acids ap peared to accumulate in cerebral PC. In rats deprived of B-12, cerebral PC con tained 77.8 to 86.0% more of these acids than was present in liver PC. Detectable amounts of branched-chain acids were not apparent in any of the phospholipid frac tions. Rats deprived of B-12 also experienced some major changes in PU FA and other fatty acid patterns of their liver PC (table 4). Significantly more 18:2w6, but less 20:4(,;6 and 22:5o>6, were present in liver PC of B-12 deprived rats fed the diet con taining 3.5% linoleate (HLD). The 12.9 ±1.07c of 18:20,6 in liver PC of HLD rats was 63.3% more than that in liver PC of their B-12 supplemented controls (HLS). Rats deprived of B-12 also accumulated significantly more palmitic ( 16:0 ) * and oleic (18:l620:3«922:020:4o,622:4w622:53LCPUFA in PE of the liver. Results of ANOVA indicate that the levels of
. . c an interference
... With
. , propionate
, , metab-
bolic block tissue levels of propionyl CoA and MMCoA are greatly increased, and this promotes greater synthesis and accumulation of odd-number and branchedchain, 15 and 17 carbon, fatty acid isomers (2, 6, 7). Such acids accounted for 13% of the total fatty acids in phosphatidyl choline ( PC ) of myelinated nerves of an infant with severe methylmalonic aciduria (1). Any increased incorporation of unusual Received for publication September 18, 1978. , Supported in part by general research funds from
olism caused by a block of methylmalonyl CoA
.
(MMCoA)-B-12 j ,
.
dependent * r ......
mutase .
required1 /-« forA inconversion of MMCoA to K o\ \XT-LL. L SUCCinyl
CoA
(Z, 5, O).
With
SUCh a meta-
2160
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
périmentStation-Regional S-87 Project. 2To whom requests for reprints should be ad-
dressed. 'Includes fulfillment
data from thesis of R. Lewis as partial of the requirements for the M.S. degree.
VITAMIN B-12, LINOLEATE AND TISSUE PHOSPHOLIPIDS
fatty acids into cerebrosides and phospho lipids could conceivably interfere with the formation and integrity of myelin, neural and other biological membranes (8-10). Longer chain polyunsaturated fatty acids (LC-PUFA) are obligatory phospholipid components in neural and other tissues ( 10, 11), and a progressive depletion of B-12 could conceivably induce significant altera tions in PUFA patterns of such phospholipids. In B-12 deprived rats the elevated tissue levels of MMCoA inhibits biosynthesis of malonyl CoA required for elongation of linoleate (18:2) 4 to form arachidonate (20:4) and other longer chain PUFA (5, 6, 12). Alterations in phospholipid PUFA Eatterns could also occur as a result of a lock of the B-12 dependent methyltransferase required for conversion of homocysteine to methionine (13-15). When B-12 deprived rats receive limited dietary supplies of methionine, a block of the methyltransferase reaction could limit the available supply of S-adenosylmethionine (SAM) (15). A limited availability of SAM would interfere with the conversion of some PUFA-rich phosphatidyl ethanolamine (PE) to PUFA-rich PC (13, 14). Rats deprived of B-12 may also be capable of more rapid catabolism of linoleate ( 16), and some reported data (17) suggest that high intakes of linoleate may promote more rapid depletion of B-12 from tissues of the rat. Any one, or combinations, of these sug gested B-12 links to PUFA metabolism could be expected to significantly affect PUFA patterns found in major phospholipids of liver and neural tissues. Support for such an interrelationship is suggested by the report of Noble and Moore (18) who found B-12 depleted chick embryos to have altered PUFA patterns in their liver phospholipids. The present study was designed to de termine how 20 weeks of B-12 deprivation might affect growth, PUFA and other fatty acid patterns in major phospholipids in the liver and cerebrum of female rats. Groups of rats were fed 3.5 and 0.35% levels of linoleate. Should B-12 deprivation accelerate catabolism of PUFA, this should be most apparent in those rats receiving the low, but adequate (19, 20), levels of
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
2161
linoleate. The 24% level of soybean pro tein provided at least marginal amounts of methionine and other amino acids needed for the growing rat5 ( 21 ). Measurements of methylmalonic acid and odd-number acids in phosphatidyl choline served as biochemical markers of the relative deple tion of B-12 from tissues of the rats.6 MATERIALS
AND METHODS
Nutritional procedures. Sprague-Dawley strain, female rats,7 35 to 50 g in weight, were initially housed in groups of three in clear plastic cages. These cages, measuring 35 X 31 X 16 cm, had stainless-steel screen tops and wire mesh bottoms which raised the rats one-inch from a moisture absorb ing corrugated cardboard liner.8 During the last 13 weeks of the study, each rat was individually housed in galvanized hanging, screen bottom cages. The animal facility was maintained at 26°with a 12-hour daynight lighting schedule. Food and tap water were allowed ad libitum, and fresh water and food were supplied every 2 to 3 days. Fresh diets were prepared every 2 to 3 weeks and refrigerated until needed. On arrival at the laboratory, all rats were fed the control diet (diet HLS in table 1) for 1 week before being divided into groups having equivalent body weights. All diets included 24yc soy protein, 20%> total fat, 3.5 or 0.35% linoleate, 3% sodium propionate and 0.01% thyroid powder (table 1). In addition to supplying at least mar ginal amounts of methionine and other amino acids needed for growth of B-12 supplemented rats 5 (21), the 24% level of soy protein should permit maintenance of 1Fatty acids are designated by their carbon chain length and number of double bonds. Heptadecanolc acid is shown as 17:0 or 17:0br to designate either the linear or branched form of this saturated acid. Omega (in) designation is used to identify members of the linoleate (iu6), linolenate (w3 or oleate (w9) families of PUFA. An eicosatrienoic acid (20:3u9), formed through elongation and desaturation of oleate, accumulates in tissue llplds with the onset of an essential fatty acid deficiency (39). 5Related studies at our laboratory Indicate that growth of rats fed 24% soy protein diets Is 85 to 90% of that obtained with 24% casein diets when both diets include B-12, folie acid and adequate amounts of other nutrients. 6One of a series of ancillary studies concerned with dietary parameters likely to affect nutritional health of préadolescents ; a Georgia Experiment Sta tion—Regional S-87 Project. 'Charles River Breeding Laboratories, Wilmington, Massachusetts. 8Deotized corrugated cage boards, Upjohn Co., Kalajnazoo, Michigan.
2162
JAMES J. PEIFER
AND RICHARD
D. LEWIS
tion of tissue levels of B-12 from those groups not receiving supplements of the vitamin (24-26). 3.5% 0.35% Isolation of tissue lipids. Rats were fasted Linoleate Linoleate overnight (12 to 15 hours) and exsan IngredieEts (HL-) diets (LL-) diets guinated while under ether anesthesia. The protein1SucroseCellulose1Mineral Soybean liver and cerebrum of each rat were quickly excised, weighed, sealed in plastic bags and frozen until needed for lipid mixture2Sodium analysis. In this report cerebrum refers to propionate1Thyroid the section of rat's brain in front of the powder1L-cystine3Linoleate4Saturated cerebellum-forebrain junction except for the olfactory nerve tracts and pituitary triglycérides6Vitamin gland. Cerebral and liver tissue homogemixture6(B-12, „g)724.043.34.03.53.00.010.253.516.52.0(0.0 or 5.0)24.043.34.03.53.00.010.250.3519.652.0(0.0 or 5.0) nates were dispersed in a minimum of 20 volumes of chloroform-methanol (2:1) 9 1Nutritional Biochemicals—ICN Pharmaceuti and total lipids were isolated in chloro cals, Inc., Cleveland, Ohio. Cellulose purchased as form according to procedures of Folch, alphacel. 2Supplied as mg/100 g diet: CaCO3, 725; CaHPO4,1,130; CuSO4, 5.0; MgSO4, 230; KC1, Lee and Sloane (27). 730; ferric citrate-5H2O, 59.0; MnSO4-H2O, 15.4; Tissue levels of methylmalonic acid KIOj, 0.32; ZnCOj, 5.5; Na2HPO4, 600; chromium (MMA) serve as a metabolic marker of a acetate, 0.05; sodium selenite, 0.01; ammonium B-12 deficiency condition in the rat (5, 6). molybdate, 0.05. 3Fisher Scientific Co., Fair Lawn, New Jersey. *Contributed by the inclusion Our initial attempts to measure tissue of 7.0 or 0.70% Mazóla corn oil (Best Foods, Englelevels of MMA suggested that freshly pre wood Cliffs, New Jersey). The 7.0% dietary level of pared homogenates were best suited for corn oil included 3.5% linoleate, 2.0% oleate, 0.05% this analysis. Three control females and linolenate, 0.9% palmitate and 0.2% stéarate.The 0.70% level of corn oil supplied one-tenth levels of three B-12 deprived females (diets HLS these same acids. 6Saturated triglycérides—sup and HLD, respectively) were randomly plied by hydrogenated coconut oil (HCO) (Proctor selected at the end of 20 weeks from a and Gamble Co., Cincinnati, Ohio) and the nonsecond, longer term experiment. These rats linoleate components of corn oil. This saturated triglycéridemixture in HL-diets included the follow were killed by decapitation, their liver ing: 5.6% laurate (12:0), 2.4% myristate (14:0), excised and immediately homogenized in 2.2% palmitate (16:0), 2.1 stéarate (18:0), 2.0% KH,,PO4 buffer (pH 7.0). An aliquot of oleate (18:1) and 1.0% of shorter chain acids (8:0 and 10:0). *Vitamins, mixed in powdered sucrose, the homogenate was immediately analyzed supplied as mg/100 g diet: thiamin-HCL, 0.05; for both methylmalonic and succinic acids. riboflavin, 0.5; niacin, 4.0; pyridoxine-HCl, 0.2; To this aliquot, representing one-fifth of inositol, 10.0; calcium pantothenate, 4.0; folie acid, the total liver homogenate, was added an 0.2; menadione, 0.2; biotin, 0.05; choline chloride, 100; vitamin B-12 (when supplied), 0.005. Fat internal standard (1.0 mg) of orthohysoluble vitamins, dissolved in the hydrogenated droxyphenylacetic acid 10 ( O-HPAA ). Fol coconut oil of the diet, supplied as IU/100 g diet : lowing hydrolysis of their coenzyme A retinol, 825; cW-a-tocopherol, 10; ergocalciferol, 80. esters, MMA, succinic acid and O-HPAA 'Values in parenthesis are the amounts of vitamin B-12 included (0.0 or 5.0 ^g) in 100 g of HL- and were extracted with ethyl ether using mod LL-diets. ifications of methods described by Smith et al. (28) and Gibbs et al. (29). normal phospholipid levels in their liver Analytical methods. Hepatic and cere and plasma (22). The 20% level of fat pro bral total lipids were fractionated into vides 40c/c of the total calories, supports phosphatidyl choline (PC) and phosphatidyl good growth of rats (23), and it permits ethanolamine (PE) by a micro-thin layer considerable flexibility for varying compo chromatography (micro-TLC) technique sition of the dietary fat (i.e., varying lin (30). This included use of 8 X 10 cm silica gel oleate levels) without altering total fat G1^coated microchromatoplates and a decontent of the diet. Corn oil triglycérides served as a source of linoleate. Sodium pro *Ratios given for solvent mixtures are volumetric. 10ICX Pharmaceuticals, Inc.. Plainview, New York. pionate and thyroid powder were added to "Applied Science Laboratories Ltd., State College, all diets in an attempt to accelerate deple Pennsylvania. TABLE I Percentage composition of experimental diets
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
VITAMIN B-12, LINOLEATE AND TISSUE PHOSPHOLIPIDS
2163
veloping solvent system of chloroformTABLE 2 methanol-water (80:25:3).9 These experi Growth response to B-1% deprivation and low intakes of linoleate1 mental conditions permit relatively rapid separation of PE and PC into two distinct Heart TLC-bands that are separated by more Diet' Weight gain» Liver ventricle Cerebrum than 2 cm. Location of the PE and PC g/SO weeks bands were done by the use of standards 242.1 ± 9.9° 9.3 ±0.6 0.84 ±0.03 1.38 ±0.02 HLS prepared from egg yolk. Micro-TLC plates 179.8±11.9' HLD 8.8±0.5 0.77 ±0.10 1.27 ±0.10 containing PC and PE standards were de 194.7 ±11.0' LLS 9.9 ±0.6 0.82 ±0.03 1.37 ±0.02 170.6 ± 5.1» 11.3 ±0.0 0.73 ±0.03 1.35 ±0.02 veloped simultaneously with those spotted LLD with the cerebral or hepatic lipids. DichloAnalysis of variance (P-values)—growth data1 B-12 Linoleate Interaction rofluorescein and ninhydrin (for PE band) sprays were used to identify PC and PE NS 0.001 0.01 bands, and the respective phospholipids were eluted from the TLC-band scrapings 1Means ±BEMof nine rata in HLS, LLS, %LLD ; eight rate in with chloroform-methanol (1:1).°Fatty acid HLD. «HLS = high linoleate (3.5%) + B-12; HLD - high linoleate—B-12; LLS - low linoleate (0.35%) + B-12; LLD components of these phospholipids were = low linoleate—B-12. ' Means not sharing a common converted to methyl esters by transesterifisuperscript letter differ significantly at P < 0.05. « Signifi cation using sodium methoxide catalyst cant differences (P < 0.001 and 0.01) attributable to the effect«of B-12 or linoleate. NS indicates absence of a ( 31 ). Methyl esters of the fatty acids were dietary significant interaction between B-12 and linoleate. analyzed by gas liquid chromatography ( GLC ) using an instrumentl- equipped and MM A were identified on the basis of with dual hydrogen flame detectors, quartz their retention times relative to O-HPAA burner jets and facilities for linear tem the use of reference standards of pure perature programming. Glass columns and succinic10 and methylmalonic10 acids. (2 m X 6 mm id) were packed with 10% Quantitative analyses of these acids were (w/w) diethyleneglycol succinate poly ester11 coated on 100 to 120 mesh Gas based on the following relationships : Chrom Q J1 treated with phosphoric acid (32) and operated isothermally at 182° A. Peak areao-HpAA _ Peak mg O-HPAA mg MMA with nitrogen as the carrier gas. Identifica tion and quantitative analysis of specific B. mgof MMA = pe,a areaMMA X2.51. fatty acid components were based on com parative retention time data with those of Peak areao_HPAA known standards,13-14comparisons of equiv „ c „ Peak area sA *, 0 no X 2.03. alent chain length (ECL) values (33) and C. mg of SA = o—¡— Peak areao-HPAA related GLC-techniques described pre viously (34). A reference mixture of In order to equate GLC peak response methyl esters of fatty acids of rat testicular per mg of pure methylmalonate (MMA) lipids, which is particularly rich in 22:5c»6 with that of the internal standard, (35), was used for further confirmation of O-HPAA, it was necessary to introduce a some of the longer chain polyunsaturated correction factor, FMMA.Under our experi fatty acids (LC-PUFA). mental conditions FMMAwas 2.51, and the Ether extracts, containing methylmalonic corresponding correction factor for methyl and succinic acids and the O-HPAA in succinate, FSA,was 2.03. ternal standard, were methylated and silanated simultaneously by use of a com '- Packard-Becker, model No. 419, Donner» Grove. mercially available Tri Syl BSA reagent.15 Illinois. 13Supelco, Inc.. Bellefonte, Pennsylvania. Purchases The resulting methyl ester-silyl derivatives included GLC-PUPA mixtures (PUFA Nos. 1 and 2) containing 20:5w3, 22:4uG and 22:0w3. 14Nu Chek Prep, Inc., Elyslan, Minnesota. GLCwere analyzed by linear temperature pro grammed (80° to 210°at a rate of 6°/ »tandards Nos. 10A and 14A included 10 different straight chain fatty acid methyl esters including minute ) GLC12 using glass columns pentadecanoate (15:0) and heptadecanoate (17:0). GLC-standards Nos. 1A, 2B, 3A1, 6A and 7A Included packed with 3% (w/w) OV-1 " on 100 to most of the other fatty acid-methyl esters identified components of PC and PE. 120 mesh Chromosorb W.11 Succinic acid as «Pierce Chemical Co., Rockford, Illinois.
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
2164
JAMES J. PEIFER
AND RICHARD
D. LEWIS
TABLE 3 Methylmalonic acid and odd-numbered fatly acid concentrations related to intakes of B-lê and linoleate1 acids'15:0Trace PC-odd-numbered
acids6-615:0Trace PC-odd-numbered
Diet¡ILS malonic1mg/100 g '100 g fatty acid liver g fatty acid 1.3 ±0.2« 3.7 ±1.0°(3) 0.56 ±0.10 0.58 ±0.06« 0.31 ±0.04 0.36 ±0.04 8.7 ±0.9»Buccinate4MMAliver 0.7 ±0.1» (3)Cerebral HLD 0.29 ±0.04 0.74 ±0.14 1.12 ±0.24» 0.24 ±0.04 0.40 ±0.03 0.63 ±0.06 (8) (8) 0.44 ±0.06 0.54 ±0.05°(7) Trace LLS 0.12 ±0.03 0.27 ±0.08 0.39 ±0.09 (») 0.64 ±0.08 1.06 ±0.17»(7) LLDIMethyl 0.39 ±0.0817:0Totalgl 0.25 ±0.10 0.57 ±0.15(8) (8)Liver 0.32 ±0.0417:0Totala/100 (8) Analysis of variance—effe-ta of Õ.P)on total odd-numbered acids7 Cerebral PC Liver PC B-12
Linoleate
0.005
NS
Interaction NS
B-12
Linoleate
0.05
NS
Interaction
1 Means ±SEM of fatty acids isolated from the cerebrum and liver of the numbers of rats shown in the parenthesis. Means not sharing a common superscript letter in the same column differ significantly at P < 0.05. Number of rata per group is shown in parenthesis. 'HLS - high linoleate (3.5%) + B-12; HLD - high linoleate—B-12; LLS = low linoleate (0.35%) + B-12; LLD = low linoleate—B-12. 'Methylmalonic acid (MMA) and succinic acid analysis measured in fresh liver homogenates prepared from each of three rats fed HLS and three rats fed HLD. • Ratios of succinate to MMA (Succinate/MMA) in liver homogenates of rats. 'PC= phosphatidylcholine. Pentadecanoic (15:0) and heptadecanoic (17:0) acids were the only major odd numbered acids detected in these phospholipids. Trace = less than 0.10% of total fatty acids in PC. 'Total odd numbered acids (15:0 + 17:0) in liver PC of B-12 deprived rats (HLD + LLD) were significantly greater than those of B-12 supplemented groups (HLS + LLS) at P < 0.05. 'Significant differences (P < 0.005 or 0.05) attributable to dietary B-12. NS indicates absence of significant differences for effects of linoleate or an interaction of B-12 and linoleate on PC odd-numbered acids.
Statistical analysis. All data are presented as means ±SEM. Where comparisons were limited to two groups, significant differ ences on Stu dent's between f-test (36).means Data were on thebased comparative effects of altered intakes of B-12 and linole ate were subjected to a two-way analysis of variance (ANOVA). Where significant F ratios were obtained for the effects of B-12 or linoleate, or an interaction between them, significant differences ( P < 0.05 ) between appropriate were identified using Tukey's test asmeans described by Steel and Torrie (36) and Linton and Gallo (37). The minimal level of statistical sig nificance accepted was P < 0.05. All ANO VA calculations were based on data obtained from equal numbers of rats per group ( 7 or 8 ). Where groups included an extra rat, tables of random numbers were used to select specific rat data to be elimi nated from the ANOVA calculations. RESULTS
The female rats gained significantly less weight when they were deprived of B-12 or fed diets containing a limited level (0.35%) of linoleate (table 2). Those rats fed the 3.5% level of linoleate (18:2o>6) 4 and de
Downloaded from https://academic.oup.com/jn/article-abstract/109/12/2160/4770673 by University of Wyoming Libraries user on 20 June 2018
prived of B-12 (HLD) weighed 25.77, less than their B-12 supplemented controls ( HLS ). Limiting dietary levels of linoleate to only 0.357o, with or without B-12 (LLS or LLD), had a similar growth inhibitory effect on the rats. Results of analysis of variance showed that growth was signifi cantly affected by the dietary effects of B-12 and linoleate but not an interaction between them. Growth differences among the groups were not accompanied by com parable differences in weights of the liver, heart ventricle or cerebrum. Rats deprived of B-12 (HLD) for 20 weeks had significantly more methylmalonic acid (MMA) in their liver (table 3). The HLD rats had MMA concentra tions of 8.7 ±0.9 mg/100 g liver and this was 6.7 times as much as that found in their controls (HLS). Ratios of succinate to MMA were significantly lower in rats deprived of B-12. The onset of a significant metabolic block of the B-12 dependent mutase system could be expected to pro mote the observed changes in MMA and ratios of succinate: MM A (15). Small, but significantly greater, amounts of odd-numbered and branched-chain acids have been reported to accumulate in phos-
VITAMIN
B-12, LINOLEATE
phatidyl choline (PC) of neural tissues of a human subject having a blocked mutase system (1). In our study, rats deprived of B-12 accumulated significantly greater amounts of odd-numbered acids in PC of their cerebrum and liver (table 3). Pentadecanoic acid (15:0) * usually accounted for only trace amounts (less than 0.10%) of the total fatty acids in cerebral or liver PC of B-12 supplemented rats (HLS and LLS). The same PC components in B-12 deprived rats (HLD or LLD) contained two to three, or more, times as much 15:0. Increased amounts of heptadecanoic acid (17:0) 4 were also apparent in PC of the cerebrum, but not liver, of rats deprived of B-12. Cerebral PC of B-12 depleted rats contained 93.1 to 96.3% more odd-num bered acids (15:0+17:0) than their B-12 supplemented controls (HLD versus HLS, LLD versus LLS). Analysis of variance of the data showed that the concentrations of odd-numbered acids ( 15:0 + 17:0 ) in cerebral and liver PC were significantly af fected by the dietary effects of B-12, but not linoleate or an interaction between
AND TISSUE PHOSPHOLIPIDS
2165
B-12 and linoleate. Comparatively greater amounts of these odd-numbered acids ap peared to accumulate in cerebral PC. In rats deprived of B-12, cerebral PC con tained 77.8 to 86.0% more of these acids than was present in liver PC. Detectable amounts of branched-chain acids were not apparent in any of the phospholipid frac tions. Rats deprived of B-12 also experienced some major changes in PU FA and other fatty acid patterns of their liver PC (table 4). Significantly more 18:2w6, but less 20:4(,;6 and 22:5o>6, were present in liver PC of B-12 deprived rats fed the diet con taining 3.5% linoleate (HLD). The 12.9 ±1.07c of 18:20,6 in liver PC of HLD rats was 63.3% more than that in liver PC of their B-12 supplemented controls (HLS). Rats deprived of B-12 also accumulated significantly more palmitic ( 16:0 ) * and oleic (18:l620:3«922:020:4o,622:4w622:53LCPUFA in PE of the liver. Results of ANOVA indicate that the levels of
