Early Embryonic Lethality in Mice with Targeted Deletion of the CTP:Phosphocholine Cytidylyltransferase {alpha} Gene (Pcyt1a)

CTP:phosphocholine cytidylyltransferase (CCT) catalyzes a rate-controllingstep in the biosynthesis of phosphatidylcholine (PtdCho). MultipleCCT isoforms, CCT ${alpha}$ , CCTß2, and CCTß3, areencoded by two genes, Pcyt1a and Pcyt1b. The importance of CCT ${alpha}$ in mice was investigated by deleting exons 5 and 6 in the Pcyt1agene using the Cre-lox system. Pcyt1a^–^/^– zygotesfailed to form blastocysts, did not develop past embryonic day3.5 (E3.5), and failed to implant. In situ hybridization inE11.5 embryos showed that Pcyt1a is expressed ubiquitously,with the highest level in fetal liver, and CCT ${alpha}$ transcripts aresignificantly more abundant than transcripts encoding CCTßor phosphatidylethanolamine (PtdEtn) N-methyl transferase, twoother enzymes capable of producing PtdCho. Reduction of theCCT ${alpha}$ transcripts in heterozygous E11.5 embryos was accompaniedby upregulation of CCTß and PtdEtn N-methyltransferasetranscripts. In contrast, enzymatic and real-time PCR data revealedthat CCTß (Pcyt1b) expression is not upregulated tocompensate for the reduction in CCT ${alpha}$ expression in adult liverand other tissues from Pcyt1a^+/^– heterozygous mice. PtdChobiosynthesis measured by choline incorporation into isolatedhepatocytes was not compromised in the Pcyt1a^+/^– mice.Liver PtdCho mass was the same in Pcyt1a^+/+ and Pcyt1a⁺^/^–adult animals, but lung PtdCho mass decreased in the heterozygousmice. These data show that CCT ${alpha}$ expression is required for earlyembryonic development, but that a 50% reduction in enzyme activityhas little detectable impact on the operation of the CDP-cholinemetabolic pathway in adult tissues.

INTRODUCTION

Top

Introduction

Phosphatidylcholine (PtdCho) is the most abundant phospholipidin mammalian cells and plays an important structural role incellular membranes and lipoprotein metabolism (8, 20). The CDP-cholinepathway is the primary route to PtdCho, and CTP:phosphocholinecytidylyltransferase (CCT) is the rate-controlling enzyme inthis three-step pathway. CCT activity is tightly regulated byits reversible interaction with lipid regulators and proteinphosphorylation (8, 20). A second route to PtdCho is the phosphatidylethanolamine(PtdEtn) N-methyltransferase (PEMT) reaction; however, thispathway is thought to be only of significant importance in theliver (19, 21). PEMT^–/– mice have a mild phenotypicdefect in hepatic lipoprotein metabolism (15) but die when cholinestarvation is used to eliminate the CDP-choline pathway forPtdCho synthesis in the liver (21), illustrating that PEMT isan important alternate route to PtdCho in the liver. Esko etal. (6) developed a temperature-sensitive cell line that wasconditionally defective in CCT activity, and when this cellline is shifted to the nonpermissive temperature PtdCho synthesisceases, cell growth halts, and apoptosis ensues (4). Similarresults are obtained using chemical inhibitors of CCT (2, 16).These results led to the hypothesis that CCT is essential formammalian cell survival.

Complexity was introduced into the field with the discoveryof a second CCT isoform, called CCTß (13, 14). CCT ${alpha}$ and CCTß are highly related proteins. Both proteinshave very similar catalytic and membrane interaction domains.An amphipathic helix in the membrane interaction domain regulatesenzyme activity by controlling the reversible interaction ofthe protein with phospholipid bilayers (8). However, CCT ${alpha}$ hasa nuclear localization signal in the amino-terminal domain andis found predominantly in the nucleus (22, 23), while CCTßlacks this signal and localizes with the endoplasmic reticulumand in the cytosol (13). The CCT ${alpha}$ isoform is the product of thePcyt1a gene and is ubiquitously expressed in cell lines andanimal tissues, whereas the CCTß2 and CCTß3isoforms are the products of the mouse Pcyt1b gene and are mosthighly expressed in the brain (9, 11, 13). Disruption of CCTß2expression results in gonadal dysfunction in mice, but the miceare otherwise normal, including the histology of the brain (9),illustrating that CCT ${alpha}$ and CCTß3 can substitute forCCTß2 function in most tissues. A mouse conditionalknockout model to study the role of CCT ${alpha}$ in specific tissueswas developed (24). The selective deletion of CCT ${alpha}$ expressionin cells of the macrophage lineage showed a surprisingly mildphenotype (24). The mice had normal numbers of macrophages,and macrophage development was supported in these cells by theupregulation of Pcyt1b gene expression. Similarly, the liver-specificknockout of CCT ${alpha}$ yielded viable mice with visually normal livers(10). Although these animals had reduced PtdCho levels bothin their liver and in serum lipoproteins, it is clear that CCT ${alpha}$ is not essential for liver development, due to the expressionof PEMT or perhaps CCTß. These two experiments withconditional knockouts suggested that the hypothesis that CCT ${alpha}$ is essential for mammalian cell survival may not be correctand prompted us to determine if one could derive mice that lackthe expression of the Pcyt1a gene in all tissues and to investigatewhether CCTß expression was sufficient. This workshows that the inhibition of CCT ${alpha}$ expression has an early embryoniclethal phenotype, with the embryos failing to develop past embryonicday 3.5 (E3.5) and implant, and that CCTß expressioncannot compensate at this stage of development.

MATERIALS AND METHODS

Top

Materials and Methods

Targeted disruption of the mouse Pcyt1a gene. Homozygous Pcyt1a^flox mice were derived previously (10, 24)and crossed with transgenic strain FVB/N-TgN(ACTB-Cre)2Mrt (TheJackson Laboratory, Bar Harbor, Maine) expressing Cre recombinasedriven by the human ß-actin gene promoter. The floxedPcyt1a locus in the resulting heterozygous embryos was deletedby the blastocyst stage of development, and the heterozygouspups carrying a wild-type and a deleted Pcyt1a gene were identifiedby PCR-mediated genotyping of tail DNA as described below. Theheterozygous mice were intercrossed and anticipated to yieldembryos and pups which were wild-type, heterozygous, and homozygousfor a deleted Pcyt1a allele in a 1:2:1 ratio. Littermates wereused as comparison controls.

Genotyping. The Pcyt1a genotypes were assessed by PCR analysis of genomicDNA extracted from mouse tails or embryos by incubation overnightat 55°C in 50 µl of lysis buffer made up of 20 mMNaCl, 50 mM Tris-HCl, 1 mM disodium EDTA (Na₂EDTA), 1% sodiumdodecyl sulfate, and 1 mg of proteinase K/ml. Aliquots of thedigestion mixture were used in a PCR with REDTaq genomic DNApolymerase (Sigma). Two sets of primers, illustrated in Fig.1, detected the wild-type or the Pcyt1a^–/^– allele.The sequence of forward primer 1 (FP1) was 5'-AATGTCTTCCAGGCTCCATAGG;FP2 was 5'-GAACTTAGGGCCTTATTCAAAGC; RP1 was 5'-TTGCCTGGTTTGTGTTTATGC.Primers FP1 and RP1 yielded a 255-bp product specific for thewild-type allele, and primers FP2 and RP1 yielded a 255-bp fragmentspecific for the deleted Pcyt1a allele. The PCR consisted ofan initial incubation at 94°C for 2 min, then five cyclesat 94°C for 30 s and 65°C for 30 s; and another fivecycles at 94°C for 30 s, 63°C for 30 s, and 72°Cfor 1 min followed by 35 cycles at 94°C for 30 s, 60°Cfor 30 s, and 72°C for 5 min. The amplified DNA fragmentswere distinguished by agarose gel electrophoresis. Preimplantationembryos were collected from pregnant dams on E3.5, where themorning of the day on which a vaginal plug was detected wasdesignated day E0.5. The uterine horns were flushed with HEPES-bufferedM2 culture medium and transferred into individual Eppendorftubes containing 5 µl of lysis buffer (0.05% sodium dodecylsulfate and 0.035 N NaOH) by using a mouth pipette assembly.The samples were boiled for 3 min, and 2.5 µl of thismixture was used in the PCR with either primer set FP1 and RP1or primer set FP2 and RP1. The PCR products were separated ona 1% agarose gel. Due to their low yield, they were detectedby Southern blot hybridization with radiolabeled probes correspondingto the sequences amplified by the primers FP1 and RP1 or FP2and RP1, as appropriate. Blots were visualized using a Typhoon9200 PhosphorImager (Molecular Dynamics). The probe templateshad previously been cloned into the pCR2.1 vector, and the DNAsequences were confirmed by the Hartwell Center for Biotechnologyat St. Jude Children's Research Hospital.

View larger version (83K):
[in this window]
[in a new window]
FIG. 1. CCT ${alpha}$ , CCTß, and PEMT gene expression in day E11.5 embryos. Embryos (day E11.5) were obtained following timed matings. (A to C) In situ hybridization of fixed and frozen sagittal sections from wild-type mice. The CCT ${alpha}$ gene (A) is widely expressed throughout the mouse embryo at E11.5 and is particularly predominant in liver; CCTß expression (B) is lower; PEMT expression (C) is lower. (D) Real-time PCR quantification of relative abundance of CCT isoform and PEMT transcripts in day E11.5 embryos using primers and probes listed in Table 1. Wild-type and Pcyt1a^+/– embryos were also genotyped in an independent assay. Both assays are described in Materials and Methods. The amount of target RNA (2^–^CT) was normalized to the endogenous GAPDH reference ( ${Delta}$ C_T) and related to the amount of target CCT ${alpha}$ in day E11.5 embryos from Pcyt1a^+/+ mice ( ${Delta}$ ${Delta}$ C_T), which was set as the calibrator at 1.0.

In situ hybridization. Riboprobe templates were generated by PCR from plasmid DNA containingthe cDNA from Pcyt1a (NM_00981.2) nucleotides (nt) 1088 to 1662,Pcyt1b (NM177546) nt 1358 to 1944, and Pemt (BC026796.1) nt26 to 651. The 3'-untranslated regions of Pcyt1a and Pcyt1bwere selected as templates to achieve sufficient sequence diversitybetween these two transcripts, and the cDNA sequences were verifiedby the Hartwell Center for Biotechnology. The templates werecloned into pBluescript, and T7 RNA polymerase was used forin vitro transcription reactions to generate [ ${alpha}$ -³³P]UTP antisenseradiolabeled riboprobes for in situ hybridization. E11.5 embryoswere removed from the mother by cesarean section and immersionfixed in 4% paraformaldehyde in 0.1 M phosphate-buffered saline(PBS) for 24 h. Embryos were cryoprotected in 30% sucrose (inPBS) and embedded in Triangle Biosciences freezing medium andfrozen on dry ice. Sixteen-micrometer tissue sections in thesagittal plane were cut using a cryostat, and tissues were mountedonto Superfrost Plus microscope slides. Several slides withE11.5 tissue were processed identically as the experimentalslides, except that they were not hybridized with probe, todocument the background signal from the tissues. Gene expressionwas evaluated by in situ hybridization as described previously(17, 18). Microscope slides were dipped in NBT2 liquid emulsion,and autoradiography proceeded for 9 days at 4°C. NBT2 emulsion-treatedslides were developed in Kodak D19 developer and fixer beforeslides were dehydrated and coverslips were attached with Permount.All treatments, hybridizations, and wash conditions were identicalfor the probes used in this assay. A nearby section in the tissueblock used was stained with cresyl violet as a reference forlocation of anatomical structures in the mouse. Micrographswere photographed with a Nikon Coolsnap ES camera mounted ontoa Zeiss Stemi 11 Stereo microscope. All photographs were capturedat 50 ms, using identical settings for all images includingthe negative control.

Detection of CCT isoform mRNAs. Real-time PCR was used to determine the relative expressionlevels of the intact CCT ${alpha}$ , CCTß2, and CCTß3isoform transcripts, encoded by the Pcyt1a and Pcyt1b genes,and PEMT mRNA, derived from the Pemt gene, in embryonic andadult tissues. Total RNA was isolated using TRIzol (Invitrogen)according to the manufacturer's instructions. Pelleted RNA wasresuspended in nuclease-free water, digested with DNase I toremove contaminating genomic DNA, aliquoted, and reprecipitatedin ethanol and stored at –20°C. Total RNA (0.5 µg)was reverse transcribed with random primers and SuperScriptII RNase H^– reverse transcriptase (Invitrogen). Quantitativereal-time PCR was carried out with 10% of the reverse transcriptionproduct in a 30-µl reaction volume of Taqman UniversalPCR master mixes (Applied Biosystems), using the Applied Biosystems7300 sequence detection system and software version 1.2.1. Specificprimers and probes are listed in Table 1. The Taqman rodentglyceraldehyde-3-phosphate dehydrogenase (GAPDH) control reagent(Applied Biosystems) was the source of the primers and probefor quantifying the control GAPDH mRNA. RNA was isolated fromthe liver, brain, heart, kidney, and lungs of three mice ofeach gender individually, and each RNA sample was quantifiedin quintuplicate. All of the real-time values for each tissuewere averaged and compared using the C_T method, where the amountof target RNA (2^–^CT) was normalized to the endogenousGAPDH reference ( ${Delta}$ C_T) and related to the amount of target CCT ${alpha}$ in liver from Pcyt1a^+/+ mice ( ${Delta}$ ${Delta}$ C_T), which was set as the calibratorat 1.0. Standard deviations from the mean ${Delta}$ C_T values were <10%.

View this table:
[in this window]
[in a new window]
TABLE 1. Real-time PCR primer-probe sets

CCT enzyme activity. CCT activities in murine tissues of adult Pcyt1a^+/+ and Pcyt1a^+/–mice were determined essentially as described previously (12).Selected organs from Pcyt1a^+/+ and Pcyt1a^+/^– mice werelysed in a Dounce homogenizer and cold 10 mM Tris-HCl (pH 7.4),5 mM Na₂EDTA, 1 µM Na₃VO₄, 5 mM NaF, 2% aprotinin, 10µg of leupeptin/ml, and 1 mM phenylmethylsulfonyl fluoride.The lysate was centrifuged at 2,000 x g for 20 min at 4°Cto remove large debris. Protein concentration was determinedaccording to the method of Bradford using gamma globulin asa standard (3). The standard assay mixture contained 150 mMbis-Tris-HCl (pH 6.5), 10 mM MgCl₂, 4 mM CTP, 100 µM lipidvesicles (PtdCho:oleic acid, 1:1), and 1 mM phospho[methyl-¹⁴C]choline(American Radiolabeled Chemicals; specific activity, 4.5 mCi/mmol)in a final volume of 50 µl. The reaction mixture was incubatedat 37°C for 10 min. The reaction was stopped by the additionof 5 µl of 0.5 M Na₂EDTA, and the tubes were vortexedand placed on ice. Next, 40 µl of each sample was spottedonto a preadsorbent silica gel G thin-layer plate (Analtech),which was developed in 2% ammonium hydroxide, 95% ethanol (1:1,vol/vol). CDP-[methyl-¹⁴C]choline was identified by comigrationwith authentic radiolabeled standard (American RadiolabeledChemicals). CCT activity was calculated from a series of assaysthat were linear with time and protein.

Metabolic labeling. Fresh livers were harvested from adult Pcyt1a^+/+ and Pcyt1a^+/^–mice, rinsed in buffer containing 66.7 mM NaCl, 6.7 mM KCl,100 mM HEPES (pH 7.6), and 36 mM glucose, and chopped with aMcllwain tissue chopper (The Vibratome Company, O'Fallon, Mo.).The minced tissue was digested in the same buffer plus 5 mMCaCl₂ · 2H₂O, collagenase (0.5 mg/ml), and DNase (6 µg/ml)for 20 min at 37°C three times successively. The debriswas allowed to settle each time, and the supernatant containingthe cell suspension was collected. After filtering the combinedcell suspension through 41-µm Spectra/Mesh nylon (FisherScientific), the hepatocytes in the filtrate were collectedby centrifugation at 100 x g for 2 min and washed three timeswith buffer containing 137 mM NaCl, 5 mM KCl, 0.65 mM Mg₂SO₄· 7H₂O, 1.2 mM CaCl₂ · 2H₂O, 10 mM HEPES (pH 7.4),and 15 mg of bovine serum albumin/ml. The cells were resuspendedin 10 ml of Dulbecco's PBS containing 5 mM glucose and 3.3 mMpyruvate plus 0.5% bovine serum albumin. Cells were countedusing a hemocytometer, and viability was determined using trypanblue exclusion. Cells were incubated at 37°C with aerationand gentle mixing in the same buffer with 3 µCi of [methyl-³H]choline/mlfor up to 6 h. After incubation, cells were transferred to iceand washed three times with PBS, and the cell pellets were extractedusing a two-phase system (1) to separate the chloroform-solublemetabolites. The amount of radiolabel incorporated into PtdCho(>95% of the total) in the organic phase was determined byscintillation counting.

Lipid determinations. Flash-frozen liver or lung tissue was thawed and weighed, andapproximately 50 mg was extracted by the method of Bligh andDyer (1). The organic phase containing lipid was concentratedunder nitrogen and resuspended in 400 µl of chloroform-methanol(2:1). A 1-µl aliquot was loaded onto a thin-layer silicagel rod, developed first in ether, dried, and then developedin chloroform-methanol-acetic acid-water (50:25:8:3) for phospholipiddeterminations or hexane-ethyl ether-acetic acid (80:20:0.5)for neutral lipid determinations. Lipid mass was detected byflame ionization using an Iatroscan instrument (Iatron Laboratories,Inc., Tokyo, Japan) with PEAK SIMPLE software (SRI Instruments),and peaks were identified by comigration with authentic standards.PtdCho mass was calculated using a standard curve prepared withegg PtdCho (Matreya, Inc.).

RESULTS

Top

Results

Targeted disruption of the Pcyt1a gene results in early embryonic lethality. Pcyt1a^+/^– mice were intercrossed to obtain Pcyt1a^–^/^–mice, and we investigated the phenotype associated with homozygousCCT ${alpha}$ deficiency. The genotypes of the offspring were determinedby PCR analysis of tail DNA at age 3 weeks. From 120 pups total,40 were wild type and 80 were heterozygous, but no homozygousmutant mice were obtained (Table 2) and no neonatal deaths occurred.These data suggested that Pcyt1a gene deficiency caused embryoniclethality. To determine the time of embryonic death, timed-pregnantmice were euthanized at day E11.5, embryos were removed, andembryonic DNA was analyzed for the presence of wild-type anddeleted Pcyt1a genes. Of 60 embryos analyzed, 19 were wild typeand 41 were heterozygous for the Pcyt1a deletion, but no Pcyt1a^–^/^–embryos were identified (Table 2). Thus, Pcyt1a^–^/^–embryos appeared to die before day E11.5. To identify furtherthe time of embryonic death, timed matings were done and dayE3.5 preimplantation embryos were isolated by flushing the uterinehorns. Analysis of the embryonic DNA by PCR followed by Southernblotting demonstrated the presence of 10 Pcyt1a^–^/^–embryos out of 44 total. We noted that all of the homozygousPcyt1a^–^/^– embryos exhibited retarded developmentat day E3.5. Whereas the vast majority of both wild-type andheterozygous embryos were blastocysts composed of a hollow trophoectodermsurrounding a fluid-filled cavity, with a small group of internalcavity mass cells as expected, 100% of the homozygous mutantembryos failed to form blastocysts. These results indicatedthat targeted disruption of both Pcyt1a alleles caused delayedor abnormal embryonic development prior to implantation. Todetermine whether homozygous knockout embryos became implantedin the uterus, we isolated day E8.5 embryos with surroundingyolk sacs from pregnant Pcyt1^+/^– dams and found neitherhomozygous mutant embryos nor evidence of uterine resorptionof embryos. The data support the conclusion that Pcyt1a^–^/^–embryos did not develop successfully beyond the preimplantationstage.

View this table:
[in this window]
[in a new window]
TABLE 2. Genotypes of offspring from intercrosses of Pcyt1a^+/– mice

Pcyt1a gene expression is dominant at day E11.5. RNA in situ hybridization of frozen sagittal sections at dayE11.5 of murine gestation showed widespread expression of CCT ${alpha}$ encoded by the Pcyt1a gene throughout the mouse embryo (Fig.1A), and a particularly high level of CCT ${alpha}$ expression was observedin the liver. There was more restricted expression of CCTß(Fig. 1D) encoded by the Pcyt1b gene compared to that of CCT ${alpha}$ .CCTß2 and -ß3 expression could not be distinguishedfrom each other using this technique, because the optimum probesize necessitated that the 3'-untranslated portion of the transcriptsbe used to signal the ${alpha}$ and ß isoforms and this regionis identical for the ß2 and ß3 species.PEMT transcripts were also investigated, since this enzyme catalyzesa reaction significant for PtdCho biosynthesis in liver (21)and was found to be expressed at a very low level in E11.5 embryos(Fig. 1C). The negative control (background) showed no signalon the tissue, whereas the PEMT showed barely detectable levelsabove background. Using a different approach, quantitative real-timePCR showed that CCT ${alpha}$ transcripts in wild-type E11.5 embryos constituted77%, whereas CCTß2 made up approximately 11%, CCTß3transcripts were 2%, and PEMT made up 10% of the four transcripts(Fig. 1B). These data support the in situ hybridization results,demonstrate that Pcyt1a gene expression is dominant at day E11.5in wild-type embryos, and support a role for CCT ${alpha}$ expressionduring embryonic development beyond the blastocyst stage.

Disruption of one Pcyt1a allele adjusts CCTß and PEMT expression. The heterozygous Pcyt1a^+/– embryos and adult mice didnot exhibit an apparently different phenotype compared to wild-typelittermate controls. Therefore, we investigated the relativeexpression of the CCT transcripts in day E11.5 embryos (Fig.1D) as well as multiple adult tissues (Fig. 2) by quantitativereal-time PCR to determine if expression of alternate genescompensated for the Pcyt1a reduced gene dosage. The value forCCT ${alpha}$ expression in the Pcyt1a^+/+ embryos was set at 1.0, andin the Pcyt1a^+/– embryos CCT ${alpha}$ was expressed at approximately0.7 times the amount found in the wild type. The expressionof CCTß2, CCTß3, and PEMT increased at leasttwofold in the heterozygous embryos in response to reduced CCT ${alpha}$ expression. On the other hand, reduction of CCT ${alpha}$ expression inPcyt1a^+/– adult liver, lung, heart, brain, or kidney didnot elicit an increase in the expression of the alternativetranscripts (Fig. 2), suggesting that the lower CCT ${alpha}$ expressionin adult heterozygotes was sufficient for maintaining PtdChoproduction via the CDP-choline pathway. The general distributionof transcripts in wild-type adult animals was consistent withpreviously measured values (9). CCTß2, CCTß3,and PEMT transcripts were considerably lower overall in theadult animals (Fig. 2) compared to their relative expressionin embryos (Fig. 1D), suggesting that these genes play a largerrole in embryo development, with the exception of the PEMT inadult liver. We cannot evaluate how much the slight, and perhapsinsignificant, elevation of CCTß expression contributesto total PtdCho production in adult tissues, but the data suggestthat CCT ${alpha}$ expression is sufficient in the Pcyt1a^+/– animals,since there are no overt phenotypic differences compared towild type.

View larger version (36K):
[in this window]
[in a new window]
FIG. 2. Relative abundance of the CCT isoforms and PEMT transcripts in tissues from adult wild-type and Pcyt1a^+/– mice. Selected organs were removed from Pcyt1a^+/+ and Pcyt1a^+/– mice. RNA was isolated from liver, brain, heart, kidney, and lung of three mice of each gender individually, and each RNA sample was quantified in quintuplicate. Primers and probes for expression of CCT ${alpha}$ (A), CCTß2 (B), CCTß3 (C), and PEMT (D) are listed in Table 1. All of the real-time values for each tissue were averaged and compared using the C_T method, where the amount of target RNA (2^–^CT) is normalized to the endogenous GAPDH reference ( ${Delta}$ C_T) and related to the amount of target CCT ${alpha}$ in liver from Pcyt1a^+/+ mice ( ${Delta}$ ${Delta}$ C_T), which was set as the calibrator at 1.0.

PtdCho biosynthesis in tissues. The biochemical consequence of disruption of one Pcyt1a alleleon CCT ${alpha}$ protein and enzyme activity was determined by measuringthe CCT specific activity in tissue lysates (Fig. 3A) and inintact hepatocytes isolated from wild-type and heterozygousmice (Fig. 3B). Within each tissue, the activity from Pcyt1a^+/–mice decreased significantly, by about 50% in most tissues comparedwith Pcyt1a^+/+ littermate controls (Fig. 3A). Whereas the lunghad a higher CCT ${alpha}$ transcript abundance than the liver, the CCTenzyme specific activity was higher in liver compared to lung,probably due at least in part to normalization of the enzymaticvalues to total protein content. Real-time PCR values, on theother hand, were normalized to total RNA. Liver is one of thefew tissues where CCTß is not expressed at detectablelevels (Fig. 2) (9), and so the reduction in CCT activity, measuredusing the same optimal conditions, from 0.466 ± 0.034in wild-type liver to 0.198 ± 0.026 pmol/min/µgof protein (P < 0.0001) in heterozygous liver was truly dueto reduction in CCT ${alpha}$ protein. On the other hand, the total CCTactivity in brain did not decrease significantly (P = 0.07)in the Pcyt1a^+/– mice, suggesting that the relativelyhigh level of CCTß expression in this organ (9) compensatedfor the loss of one active CCT ${alpha}$ allele. Livers from a total ofthree mice of each sex and each genotype were included in thedetermination. There was no difference between males and femaleswithin a genotype, and so the values were combined and averaged.

View larger version (21K):
[in this window]
[in a new window]
FIG. 3. Effect of targeted disruption of the Pcyt1a gene on hepatic PtdCho synthesis in wild-type and heterozygous mice. (A) CCT enzyme specific activity was determined in adult mouse tissue lysates and normalized to protein content as described in Materials and Methods. (B) Radiolabeling of PtdCho in hepatocytes derived from Pcyt1a^+/+ and Pcyt1a^+/– mice. Hepatocytes (2 x 10⁶ per sample per animal) were incubated in Dulbecco's PBS containing glucose and pyruvate for 6 h with [methyl-³H]choline (3 µCi/ml). Radiolabeled PtdCho was extracted from cells and quantified as described in Materials and Methods. The results shown are the average values obtained using hepatocytes from six animals of each genotype.

Liver did not express detectable levels of CCTß, andso we determined whether the significant reduction of CCT ${alpha}$ proteinwould alter the rate of PtdCho synthesis in primary hepatocytesisolated from heterozygous animals. Hepatocytes were isolatedfrom Pcyt1a^+/– and Pcyt1a^+/+ mice and incubated with [³H]cholineto estimate the rate of PtdCho synthesis in vivo (Fig. 3B).Incorporation into PtdCho was linear from 2 to 6 h and, surprisingly,there was no statistical difference between the heterozygouslivers and those from wild-type animals, i.e., 4.0 x 10⁴ cpmversus 5.0 x 10⁴ per 10⁶ viable cells after 6 h of incubation.The data certainly did not reflect the approximate 50% differencein lysate specific activity noted above (Fig. 3A), but the resultsexplained the lack of phenotypic distinction between heterozygousand wild-type animals. There was no difference in the metaboliclabeling results between the sexes, and thus the values arethe averages of a total of six independent determinations fromeach genotype, i.e., hepatocytes from three males and threefemales. These results indicated that the amount of CCT activityin wild-type cells was more than sufficient to maintain physiologicalPtdCho production and that CCT activity was biochemically regulatedin cells to maintain PtdCho synthesis at homeostatic levels.Tissue lipids were quantified in Pcyt1a^+/– and Pcyt1a^+/+mice and did not exhibit a substantive change in PtdCho contentin liver (32.5 ± 4.8 and 35.57 ± 1.69 µg/mg[wet weight], respectively), consistent with the metabolic labelingresults. The highest relative CCT ${alpha}$ expression was previouslyfound in lung, where CCTß expression was roughly 40times lower (9), and so we examined if the PtdCho content ofthis organ was more sensitive to the reduction in CCT ${alpha}$ gene expression.There was a small but significant reduction in the PtdCho contentof Pcyt1a^+/– animals (19.18 ± 2.54 in Pcyt1a^+/^–compared to 23.19 ± 1.74 µg/mg [wet weight] inPcyt1a^+/+; P < 0.01), but the other lipids, i.e., PtdEtn,triglyceride, and cholesterol, did not change.

DISCUSSION

Top

Discussion

Deletion of CCT ${alpha}$ expression in mice revealed its essential rolein early embryonic development. The ability of the knockoutpreimplantation embryos to progress through the first few celldivisions is likely due to the expression of maternal CCTßthat is present at high levels in mature ova (9). Retarded developmentof the CCT ${alpha}$ -deficient embryos is seen at the morula stage, indicatingthat cell division arrests at an earlier stage. CCT plays acentral role in membrane PtdCho synthesis (8) and, althoughembryonic cell size decreases during preimplantation development,there is an increase in total membrane surface area which requiresthe activity of CCT ${alpha}$ plus CCTß. These data are consistentwith observations in the liver-specific CCT ${alpha}$ knockout, wherethe organ contains fewer hepatocytes than control liver, suggestingthat the hepatocytes lacking CCT ${alpha}$ do not produce enough PtdChofor normal cell division (10). CCT ${alpha}$ expression dominates overCCTß and PEMT expression during embryogenesis, andCCT ${alpha}$ is found throughout the different cell types, with highestexpression in liver (Fig. 1). These data are consistent withthe very low level of PEMT activity observed in prenatal ratlivers (5) and the lack of detectable PEMT pathway activityin E9.5 embryos with impaired CCT activity (7). Although CCT ${alpha}$ is reduced in heterozygous embryos, the CCTß isoformsand PEMT expression both are increased, and so it is difficultto judge whether the level of CCTß expression alonewould be sufficient to support the normal development that isobserved.

In contrast to the essentiality of CCT ${alpha}$ in embryonic development,CCT ${alpha}$ expression is not critical for the function of some adulttissues. Significant reduction of CCT ${alpha}$ in heterozygous adultliver does not invoke a compensatory increase in the transcriptlevels of the CCTß proteins or PEMT (Fig. 2), noris there a significant difference in the PtdCho mass comparedto that in wild type. Despite a >50% reduction in total CCTenzyme specific activity in heterozygous liver lysates, themetabolic activity of the CDP-choline pathway in isolated hepatocyteswas the same as in wild-type hepatocytes (Fig. 3). Furthermore,it was reported previously that homozygous deletion of CCT ${alpha}$ inliver does not have a severe phenotype, probably because thePEMT enzyme activity increases in response to CCT ${alpha}$ deletion (10).The mild phenotype is found primarily in female mice lackingliver CCT ${alpha}$ expression, which have reduced PtdCho and increasedtriglyceride content (10). In the macrophage-specific CCT ${alpha}$ knockout,homozygous deletion of CCT ${alpha}$ is accompanied by increased expressionof CCTß, whose activity is sufficient for macrophagedevelopment (24). The phenotypic similarity between the heterozygousand wild-type adult mice in this report suggests that the amountof CCT ${alpha}$ enzyme in wild-type liver is more than sufficient tomaintain PtdCho-associated functions. Reduction of CCT ${alpha}$ expressionin the lungs of heterozygous animals results in reduced PtdChocontent, in contrast to the livers from the same animals. CCT ${alpha}$ expression in wild-type lung is comparatively the highest ofthose tissues examined thus far, with the exception of testis(9), and the lung is a very lipogenic organ, secreting largeamounts of disaturated PtdCho as a major component of airwaysurfactant. Taken together, these data suggest that homozygousdeletion of CCT ${alpha}$ expression in lung would be very deleterious,and future investigation will determine whether upregulationof CCTß expression can compensate for the lack ofCCT ${alpha}$ .

ACKNOWLEDGMENTS

We thank Chris Gunter and Daren Hemingway for expert technicalassistance in enzyme assays and lipid analyses, Pam Jacksonfor primer and probe designs, and Chuck Rock for discussionand interest.

This work was supported by National Institutes of Health grantGM 45737 (to S.J.), Cancer Center (CORE) support grant CA 21765,and the American Lebanese Syrian Associated Charities.

FOOTNOTES

* Corresponding author. Mailing address: St. Jude Children's Research Hospital, 332 N. Lauderdale, Memphis, TN 38105-2794. Phone: (901) 495-3494. Fax: (901) 495-3099. E-mail: suzanne.jackowski{at}stjude.org.

REFERENCES

Top