Linoleic acid controls neonatal tissue-specific stearoyl-CoA desaturase mRNA levels

291

Biochimica et Biophysics Acta, 1170 (1993) 291-295 0 1993 Elsevier Science Publishers B.V. All rights reserved 000%2760/93/$06.00

BBALIP 54265

Linoleic acid controls neonatal tissue-specific stearoyl-CoA desaturase mRNA levels James W. DeWille ‘yby*and Steven J. Farmer a a Department of Veterinary Pathob~~o~, 1923 Coffey Road, Columb~, OH 43210-1093 fUSAl and b The Ohio State Uniuersity Biochemistry Program, The Ohio State University, Columbus, OH 43210 (USA) (Received 21 June 1993)

Key words: Essential fatty acid; Stearoyl-CoA desaturase; CCAAT/enhancer-binding

protein; Development; mRNA

The mouse genome contains two stearoyl-CoA desaturase (SCD) structural genes WDl and SCD2) that are expressed in a tissue-specific manner. Brain SCD2 mRNA levels are about 2-fold higher in pups nursed by mothers fed a control diet (5% corn oil (CO), (essential fatty acid (EFA) adequate)), compared with brain SCD2 mRNA levels in pups nursed by mothers fed an EFA-deficient @FAD) diet (5% coconut oil (COCO)). In contrast to brain, control pup hepatic SCDl mRNA levels are reduced to < 1.0% of the EFAD pup hepatic SCDl mRNA levels. EFA status does not alter SCDl or SCD2 transcription initiation sites. CCAAT/enhancer-binding proteins (C/EBP) have been implicated in the transcriptional control of key genes in energy metabolism. Both the SCDl and SCD2 gene promoters contain C/EBP transcription factor consensus-binding sites. Neonatal mouse liver expresses C/EBP-CX, C/EBP-J3 and C/EBP-S mRNAs. In contrast, neonatal mouse brain expresses high levels of C/EBP-&, but little C/EBP-~U or C/EBP+ mRNA. EFA intake has no effect on tissue-specific C/EBP isoform mRNA levels suggesting that C/EBP isoform function is controlled at the translational or post-translational level.

Introduction Stearoyl-CoA desaturase @CD) catalyzes the A’-desaturation of stearoyl-CoA-forming oleic acid (CM: l(pt - 9)) [1,2]. Oleic acid is the major fatty acid in myelin phospholipids and adipose tissue triacylglycerols [2-41. The mouse genome contains two SCD structural genes (SCDl and SCD2) that are highly homologous at the nucleotide and amino acid level [5,6]. The 5’ flanking regions of the two mouse SCD structuraI genes differ resulting in divergent tissuespecific expression [5-71. SCDl gene expression is markedly induced in liver by short-term fasting/high carbohydrate refeeding and essential fatty acid deficiency (EFAD) [.5-71. In contrast, the SCD2 gene is not expressed in liver [6,7]. The SCD2 gene is developmentally induced in brain during the neonatal myelinating period and constitutively expressed in later (adult) life [6,71. CCAAT/ enhancer-binding proteins CC/ EBP) are a family of highly conserved ‘leucine zipper’ containing DNA-binding proteins that are expressed in a tissuespecific manner [S-14]. C/EBP isoforms bind to spe-

* ~rresponding

author.

cific DNA sequences and have been implicated in the transcriptional control of phosphoenolpyruvate carboxykinase (PEPCK) [11,141, stearoyl-CoA desaturase 1 (SCDl) [5,6,10,111, adipocyte 422 (aP2) ElO,ll] and the insulin-responsive glucose transporter 1111. This has led to speculation that C/EBPs may function as central regulators of carbohydrate and lipid metabolism [HI. Both the SCDl and SCD2 gene promoters contain C/EBP consensus binding sites and C/ EBP-(w has been linked to the transcriptional control of the SCDl gene in 3T3-Ll adipocytes [6,101. The initial aim of this study was to assess the influence of diet and development on SCD mRNA levels, which are divergently expressed in neonatal liver and brain [7]. Under control (EFA-adequate) dietary conditions neonatal hepatic SCDl mRNA levels are virtually undetectable by Northern blot analysis 171.In contrast, brain SCD2 mRNA levels are highly induced during neonatal life, presumably to provide oleic acid for incorporation into newly synthesized myelin [7]. The second aim was to identify the liver and brain SCDl and SCD2 transcription initiation sites to facilitate analysis of SCD upstream transcriptional control elements. Finally, because of the postulated role of C/ EBP isoforms in the transcriptional control of genes encoding enzymes in energy metabolism 19-111 and the documented role of C/EBP-(r as a transa~tivator of

292

the SCDZ gene [lo], the influence of diet on tissuespecific C/ EBP expression was assessed. Materials and Methods Female Balb/c mice were fed isocaloric diets providing O%, 5% or 25% of total calories from fat (coconut oil or corn oil) (Teklad, Madison, WI, USA) or a standard pellet diet (Purina, St. Louis, MO) beginning on day one of lactation. The composition of the diets has been previously described [7]. Litters were standardized to 8 pups. The maternal diets do not alter neonatal pup body weights [7]. Tissues from 4-8 mice were pooled for each experiment and representative blots from 3-4 experiments are shown. RNA was isolated by the guanidium isothiocyanate/ cesium chloride method and poly A + RNA selected by oligo-dt chromatography as previously described [7,15]. Northern blots were carried out by electrophoresis through a 1.2% agarose/ formaldehyde gel followed by direct transfer to nylon filters. Probes for Northern blots were labeiled by the random primer method 1151.Following hybridization in a formamide containing buffer at 42°C the filters were washed in 2 X SSC plus 0.5% SDS for 15 min at room temperature followed by 1-3 washes with 0.5 X SSC for 15 min at 5O’C. Filters were exposed to X-ray film and X-ray films were scanned with a laser scanning densitometer [7,15]. Since the protein encoding regions of the SCDl and SCD2 genes are highly homologous [5,63 an SCD probe was developed by the reverse transcriptase-polymerase chain reaction CRT-PCR) to permit detection of both SCD mRNAs with a single probe [Iti]. The SCD primers ((sense) 5’-TATCAGGATGATGAG-3’; (antisense) S’-CATGCAATCGATGAAGAA-3’) (Oligos, Wilsonville, OR) were homologous to published SCD cDNA sequences 1561 and amplified an 807 base pair fragment derived from a RT reaction. Mouse brain poly A + RNA was used as the starting material. PCR amplification was carried out under optimized conditions and the amplified SCD product was verified by restriction digest and hybridization analysis and subcIoned (TA Cloning vector, Invitrogen, San Diego, CA). The C/EBP isoform CC,/EBP-a, 8, -6) cDNA probes were generously provided by Dr. Steven McKnight (Tularik, South San Francisco, CA). A mouse cyclophilin cDNA clone was produced by RT-PCR and used as a constitutive probe. For primer extension assays a synthetic oligonucleotide primer (5’-GATCTCTTGGAGCATGTGGGCCGGCAT-3’) (Ohgos, Wilsonville, OR) complemental to nucleotides 279-153 of the published mouse SCDl cDNA sequence [61 was end-labelled using T4 polynucleotide kinase and [y-32PlATP (specific activity > 7000 Ci/mmol). The end-labelled 27 nucleotide product was isolated from the unincorporated free nu-

cleotides by centrifugation using a molecular weight cutoff membrane of 3000 kDa (Amicon, Beverly MA). 2 pg of poly A Jr RNA were annealed to 10 ng of oligonucleotide primer at 30°C in 80% formamide-containing hybridization buffer [ 151. AnneaIed products were precipitated and then extended using 50 units of Maloney Murine Leukemia Virus (MO-MI_V~ Reverse Transcriptase for 90 min at 42°C. One ,ul of 0.5 M EDTA (pH 8.0) and 1 pi of DNAase-free RNAase (1 mg/ml) were included in the final 30 min of the incubation. The primer extension products were separated by electrophoresis through an 8% polyacryiamide/7 M urea sequencing gel. The gels were dried and exposed to X-ray film, Results and Discussion To investigate the effects of EFA intake on SCD mRNA expression in neonates lactating mothers were fed a 5% coconut (COCO) or a 5% corn oil (CO) diet. Since coconut oil is a poor source of EFA (2% hnoleic acid (C18:2(pz - 6)) the 5% COCO diet is EFA-deficient ( < 0.1% of total calories from linoleic acid). This results in a milk n - 6 fatty acid content of < 1%. providing an EFA-deficient diet for the nursing pups [17]. In contrast, corn oil is an excellent source of EFA (60% finoleic acid) and the milk derived from mothers fed the control diet (5% CO) provides about 3% of total calories as n - 6 fatty acids. This is sufficient to meet neonatal EFA requirements for growth and myelination [?,I 71. To directly compare neonatal brain and liver SCD mRNA content we used a single SCD coding region probe developed by RT-PCR (see Materials and Methods). NeonataI brain and liver exhibit a divergent SCD mRNA response to dietary EFA (Fig. 1A and B). Brain SCD2 mRNA leveis are > 2-fold higher in control (EFA adequate) vs. EFAD pups (Fig. 2, lane 1 vs. 2). Developmental induction of brain SCD2 mRNA levels coincides with the major myehn-specific proteins: proteolipid protein (PLP) and myelin basic protein (MBP) and all three mRNAs are reduced by neonatal EFAD [?I. This is consistent with earlier reports indicating that myelin synthesis is inhibited by EFAD during neonatal life 1181.In contrast, adult brain SCD2 mRNA levels are unaffected by EFAD (data not shown) or fasting followed by refeeding a high carbohydrate diet

1-W. The presence of EFA in the maternal diet (5% CO) dramaticaily reduces neonatal hepatic SCDl mRNA levels to < 1.0% of pups nursed by mothers fed the 5% COCO diet (Fig. 2, lane 3 vs. 4). This indicates that hepatic SCDl gene expression is induced by EFAD and inhibited when EFA are present in the diet, even at intake levels approximating EFA requirements (about 2% of total calories) [17].

293 B.

NEONATAL BRAIN AND HEPATIC STEAROYL COA DESATURASE mFtNA LEVELS

A.

4 HOUR

EXPOSURE

285

B. 24 HOUR EXPOSURE

123 505

%CO

4 0

Fig. 1. Fatty acid intake influences tissue-specific stearoyi-CoA desaturase (SCD) mRNA levels in 21 day old mice. Day 21 neonates were nursed by mothers fed either a 5% corn oil (CO) (linoleic acid adequate) or a 5% coconut oil (COCO) (essential fatty acid-deficient) diet. Data are representative of 4 experiments. Each experiment used pooled tissue from 4-8 mice/treatment (10 pg of poly A+ RNA/lane). Filter probed with 32P-labelled SCD coding region probe. Lanes: (1) Brain 5% CO, (2) brain 5% COCO, (3) liver 5% co, (4) 5% coca.

Primer extension assays were carried out to further assess the effects of dietary EFA on SCD mRNA levels and to identify in vivo SCD transcription initiation sites. SCDl mRNA was detected in livers from 21 day old pups nursed by mothers fed a standard pellet diet, but not 7 and 14 day old pup livers (Fig. 3A, lanes l-3). Since the pellet diet consumed by the lactating mothers provides about 3% of total calories as linoleic acid the milk n - 6 fatty acid content is about 2.5% of total calories this level of IZ- 6 fatty acid intake ap-

DAY 21: BRAIN AND LIVER SCD RAW #

1,600 1,400 1,200 1.000 800 600 400 200 0 12

3

4

Fig. 2. Laser densitometric scan of dietary linoleic acid effects on neonatal brain and liver stearoyl-CoA desaturase mRNA levels (Fig. 1). Lanes: (1) Brain 5% CO, (2) brain 5% COCO, (3) liver 5% CO, (4) 5% coca.

123456

123

Fig. 3. Primer extension analysis of developmental, dietary and tissue-specific effects on stearoyl-CoA desaturase @CD) mRNA levels and transcription initiation sites. Each lane represents the product of a primer extension assay containing 50 units of MO-MLV reverse transcriptase, 2 Kg of poly A+ RNA and 10 ng (approx. 5OOooOcpm) of a primer homologous to the first 27 nucleotides of the mouse SCDl coding sequence 151.Data shown is representative of 3 experiments. Each experiment used pooled tissue from 4-8 mice/treatment. (A) Developmental and dietary effects on neonatal hepatic SCD mRNA levels. Lanes: (1) day 7, pellet diet, (2) day 14, pellet diet, (3) day 21, pellet diet, (4) day 21, 5% corn oil (CO), (5) day 21, 5% coconut oil (COCO), (6) adult 25% CO. (B) Dietary effects on neonatal brain SCD2 mRNA levels. Lanes: (1) brain, 5% CO, (2) brain, 5% COCO, (3) tRNA.

pears to be adequate to suppress hepatic SCDl mRNA levels during the early neonatal period. This inhibition diminishes at weaning (day 21). Hepatic SCDl mRNA was not detected in livers from 21 day old pups nursed by mothers fed the 5% CO diet (Fig. 3A, lane 4) or in adult mice fed the 25% CO diet (15% of calories from linoleic acid) (Fig. 3A, lane 6).

294 Hepatic SCDl mRNA levels are induced in 21 day old mice nursed by mothers fed the 5% COCO diet (Fig. 3A, lane 5). The fat source consumed by the lactating mother alters the fatty acid profile of the milk but does not alter milk volume, fat, protein or caloric content [19,201. Therefore, the limiting linoleic acid content of the milk produced by mothers consuming the 5% COCO diet ( < 0.1% of total calories as linoleic acid) induces pup hepatic SCDl mRNA levels. The size of the primer extension product (approx. 180 nucleotides) detected from mouse liver RNA is unaffected by n - 6 fatty acid intake. The hepatic SCDl primer extension product is essentially identical to that published from RNA isolated from differentiated mouse 3T3-Ll cells [5,6]. The promoter regions of the mouse SCDl and SCD2 structural genes differ significantly but the protein encoding regions are highly homologous [5,6]. The protein encoding region homologous to the oligonucleotide used in the SCDl primer extension assay differs in only two internal bases between SCDl and SCD2 [5,61. Therefore, this oligonucleotide was used to assess the effects of the diets on day 21 brain SCD2 RNA levels and to determine the transcription start site for SCD2 in mouse brain. In agreement with Northern blot results, day 21 brain SCD RNA levels were approx. 2-fold higher in pups nursed by mothers fed the control vs. the EFAD diet (Fig. 3B, lane 1 vs. lane 2). The length of the SCD2 primer extension product was approx. 330 nucleotides (Fig. 3B, lane 1 and 2). This is essentially identical to the 5’ end of the mouse SCD2 mRNA expressed in differentiated mouse 3T3-Ll cells [5]. Although annealing was carried out under low stringency (30’0, no SCD2 extended products were detected in liver (Fig. 3A) and no SCDl extended products were detected in brain (Fig. 3B). This further documents the high degree of tissue specificity of SCDl and SCD2 gene expression. C/ EBP isoform expression is highly tissue-specific [9,12,13]. This suggests that the amount and/or proportion of the individual C/EBP isoforms is crucial to carry out unique tissue-specific functions, but this has not been extensively investigated in vivo. Injecting rats with dibutyryl CAMP increases hepatic CAMP levels and induces C/EBP-P and PEPCK mRNA levels [14]. This suggests a mechanism by which glucagon mediated increases in hepatic CAMP levels may induce PEPCK gene transcription during fasting [14]. The influence of variations in EFA intake on tissue-specific C/EBP isoform mRNA levels has not been previously reported. Sequential reprobing of Fig. 1 with C/EBP-a, C/EBP-/3 and C/EBP-S cDNA probes indicated that high levels of C/EBP-a and C/ EBP-/3 mRNA are detectable in neonatal liver (Fig. 4). In contrast, neonatal brain expresses little C/ EBP-a or C/ EBP-/3 mRNA (Fig. 4). C/EBP-6 mRNA is

C/EBP

alpha

-

18s

beta

-

18s

delta

-

cyclophilin

1 IJ Brain

2

3

4 Liver

Fig. 4. Neonatal brain and liver C/EBP isoform mRNA levels. Day 21 neonates were nursed by mothers fed either a 5% corn oil (CO) (linoleic acid adequate) or a 5% coconut oil (COCO) (essential fatty acid-deficient) diet. Data is representative of 4 experiments. Each experiment used pooled tissue from 4-8 mice/treatment (10 pg of poly A+ RNA/lane). Filter was sequentially probed with C/EBP-c~ (A), C/EBP$ (B), C/EBPd (0, and mouse cyclophilin (a constitutive probe) (D). Lanes: (1) Brain 5% CO, (2) brain 5% COCO, (3) liver 5% CO, (4) 5% COCO.

expressed in both neonatal brain and liver (Fig. 4). This indicates that EFA intake, which alters SCD mRNA levels in brain and profoundly affects SCDl mRNA levels in liver, has no major effects on C/EBP isoform mRNA levels in either tissue. The results from neonatal brain (Figs. l-3) are the first evidence of a positive effect of a dietary fatty acid on the expression of a gene encoding a lipogenic enzyme. The results for brain are consistent with reports indicating that EFA are required for the expression of genes encoding myelin-specific structural proteins and biosynthetic enzymes during the neonatal myelinating period [6,11]. Only the neonatal brain is sensitive to dietary n - 6 fatty acid intake as EFA deficiency has no effect on adult SCD2, PLP or MBP mRNA levels (data not shown). This indicates that during this period the myelinating oligodendrocyte is transiently ‘com-

295 petent’ to respond to inducing signals associated with the availabili~ of n - 6 fatty acids [21]. The nature of these signals and the mechanism by which f~ - 6 fatty acids control SCD2 gene expression during myelination is unknown. In the liver, limiting EFA intake induces SCDl gene expression, increases hepatic SCDl enzyme activity and increases hepatic oleic acid (n - 9 fatty acids) levels [1,7]. This provides PE- 9 fatty acids which serve as partial biological substitutes for the limiting n - 6 essential fatty acids [1,71. The mechanism by which n - 6 fatty acids control hepatic Iipogenic gene expression is unknown, but may involve the activation or assembly of transcription factors on select gene promoters, or possibly effects on mRNA stability [221. C/EBP transcription factors have been directly linked to the transcriptional control of key genes in carbohydrate and lipid metabolism [lo-141. The expression of high levels of C/EBP+ and C/EBP+ mRNA and Iower levels of C/ EBP-6 mRNA in neonatal liver parallels results in adult mice [12,131. In contrast to liver, neonatal brain expresses high levels of C/EBP-6 mRNA, but little C/EBP-a! or C/EBP-@. Previous studies using adult animals had detected little or no C/EBP-cq $3 or -6 mRNA in brain 112,131. Our results suggest that C/EBP-6 may be induced in neonatal brain, possibly to transactivate the SCD2 gene during the neonatal myelinating period. The absence of a significant diet effect on tissue-specific C/EBP isoform expression suggests that translation or posttranslational modifications (i.e., phosphorylation, dimerization) may control C/ EBP isoform function under varying metabolic conditions in vivo. Current studies are aimed at further defining the role of EFA and C/EBP isoforms in the tissue-specific expression of SCDl and SCD2 genes. References 1 Jeffcoat, R. and James, A.T. (1984) in Fatty Acid Metabolism and Its Regulation (Numa, S., ed.); pp. 85-112, Elsevier, Amsterdam.

2 3 4 5

Ntambi, J.M. (1992) J. Biol. Chem. 267, 10925-10930. Sun, G.Y., Go, J. and Sun, A.T. (1974) Lipids 9, 450-454. Kasturi, R. and Joshi, V.C. (1982) J. Biol Chem. 257,12224-12230. Kaestner, K.H., Ntambi, J.M., Kelly, T.J. and Lane, M.D. (1989) J. Biol. Chem. 264, 14755-14761. 6 Ntambi, J.M., Buhrow, S.A., Kaestner, K.H., Christy, R.J., Sibley, E., Kelly, T.J. and Lane, M.D. (1988) J. Biol. Chem. 263, 1729117300. 7 DeWille, J.W. and Farmer, S.J. (1992) Dev. Neurosci. 14, 61-68. 8 Lamb, P. and M&night, S.L. (1991) Trends in Biochem. 16, 417-422. 9 Birkenmeier, E.H., Gwynn, B., Howard, S., Jerry, J., Gordon, J.I., Landschultz, W.H. and M&night, S.L. (1989) Genes Dev. 3, 1146-1156. 10 Christy, R.J., Yang, V.W., Ntambi, J.M., Geiman, D.E. Landschultz, W.H., Friedman, A.D., Nakabeppu, Y., Kelly, T.J. and Lane, M.D. (1989) Genes Dev. 3, 1323-1335. 11 M&night, S.L., Lane, M.D. and Gluecksohn-Waelsch, S. (1989) Genes Dev. 3: 2021-2024. 12 Cao, Z., Umek, R.M. and M&night, S.L. (1991) Genes Dev. 5, 1538-1552. 13 Williams, SC., Cantwell, CA. and Johnson, P.F. (1991) Genes Dev. 5, 1553-1567. 14 Park, E.A., Gurney, A.L., Nizielski, S.E., Hakim, P., Cao, Z., Moorman, A. and Hanson, R.W. (1993) J. Biol. Chem. 268, 613-619. 1.5 Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning, pp. 7.2-7.87, Cold Spring Harbor Press, Cold Spring Harbor. 16 Kawasaki, E.S. (1990) in PCR Protocols: A Guide to Methods and Applications (Innis, M.A., Gelfand, D.H., Sninsky, J.J., White, T.J., eds.), pp. 21-27, Academic Press, San Diego, CA. 17 Bourre, J.M., Piciotti, M., Dumont, O., Pascal, G. and Durand, G. (1990) Lipids 25, 465-472. 18 Berkow, S.E. and Campagnoni, A.T. (1981) J. Nutr. 111,886-894. 19 Wickwire, M.A., Craig-Schmidt, M.C., Weete, J.D. and Faircloth, S.A. (1987) J. Nutr. 117, 232-241. 20 Ross, A.C., Davila, M.E. and Cleary, M.P. (1985) J. Nutr. 115, 1488-1497. 21 Gluecksohn-Waelsch, S. and DeFranco, D. (1991) Bioessays 13, 557-561. 22 Clarke, S.D. and Abraham, S. (1992) FASEB J. 6, 3146-3152.