Inactivation of the Murine Pyruvate Dehydrogenase (Pdha1) Gene and Its Effect on Early Embryonic Development

Molecular Genetics and Metabolism 74, 293–302 (2001) doi:10.1006/mgme.2001.3249, available online at http://www.idealibrary.com on

Inactivation of the Murine Pyruvate Dehydrogenase (Pdha1) Gene and Its Effect on Early Embryonic Development Mark T. Johnson,* ,† Saleh Mahmood,‡ Susannah L. Hyatt,* Hsin-Sheng Yang,‡ Paul D. Soloway,§ Richard W. Hanson,* and Mulchand S. Patel‡ ,1 ‡Department of Biochemistry, School of Medicine and Biomedical Sciences, State University of New York at Buffalo, Buffalo, New York 14214; *Department of Biochemistry and †Department of Genetics, School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106; and §Department of Molecular and Cellular Biology, Roswell Park Cancer Institute, Buffalo, New York 14263 Received June 7, 2001; published online October 25, 2001

early postimplantation period of prenatal development. © 2001 Academic Press Key Words: pyruvate dehydrogenase complex deficiency; embryonic development; X chromosome; Pdha1 gene disruption; mouse; Cre/loxP.

A deficiency of pyruvate dehydrogenase complex (PDC) in humans results in lactic acidosis and neurological dysfunction that frequently results in death during infancy. Using gene targeting technology, a silent mutation was introduced into the murine X-linked Pdha1 gene that encodes the ␣ subunit of the pyruvate dehydrogenase or E1 component of the complex. Two loxP sequences were introduced into intronic sequences flanking exon 8 to generate the Pdha1 flox8 allele. In vitro studies in embryonic stem cells demonstrated that deletion of exon 8 ablated PDC activity. Homozygous Pdha1 flox8 females were bred with male mice carrying a wild-type Pdha1 allele and a transgene that ubiquitously expresses the Cre recombinase to produce progeny with a deletion in exon 8, Pdha1 ⌬ex8. The majority of progeny were found to be mosaic with the presence of both the flox and deleted alleles, and there were no apparent phenotypic effects associated with the null allele. The mosaic mice were interbred to increase the degree of mosaicism for the Pdha1 ⌬ex8 allele in the subsequent generation, resulting in a significantly smaller litter size (54% reduction). Embryos carrying predominantly the Pdha1 ⌬ex8 allele were found to be globally delayed in development by 9.5 days postcoitus, with resorption occurring over the following several days. These findings demonstrate an essential role for oxidative metabolism of glucose during the

Pyruvate dehydrogenase complex (PDC) is a mitochondrial multienzyme complex that catalyzes the irreversible oxidative decarboxylation of pyruvate to form acetyl-CoA (1). The enzymatic cascade catalyzed by PDC is essential for the metabolism of carbohydrates and select amino acids. The first catalytic component of the complex, referred to as pyruvate dehydrogenase or the E1 component, exists as a heterotetramer of two ␣ and two ␤ subunits. In mammals, two isoforms of the E1␣ subunit have been identified that are encoded by separate genes. The X-linked gene, abbreviated as PDHA1 in humans and Pdha1 in mice, codes for the E1␣ subunit protein present in all somatic tissues. An autosomal, intronless gene, denoted as PDHA2 in human and Pdha2 in mice, is expressed only in the testis (2,3). Genetic defects of PDC are reported to be the most commonly identified cause of primary congenital lactic acidosis in humans. The clinical spectrum of PDC deficiency is extremely broad, ranging from death in the neonatal period caused by lactic acidosis and severe neurologic dysfunction to milder presentations of neurologic dysfunction that include ataxia and mental retardation. The fact that the brain is most susceptible to PDC deficiency underscores its dependence on glucose as a primary energy source.

1 To whom correspondence and reprint requests should be addressed at Department of Biochemistry, School of Medicine and Biomedical Sciences, State University of New York at Buffalo, 140 Farber Hall, 3435 Main Street, Buffalo, NY 14214. Fax: (716) 829-2725. E-mail: [email protected].

293 1096-7192/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

294

JOHNSON ET AL.

PDC deficiency has been reported in more than 150 patients, with approximately 90% of cases involving deficiency of the E1 component. All mutations thus far identified that affect E1 component function have been localized to PDHA1 (4). The development of a line of mice carrying a null mutation in the Pdha1 gene may help to elucidate the cellular and metabolic consequences of this deficiency and ultimately serve as a model for developing novel therapeutic interventions. MATERIALS AND METHODS Generation of the Pdha1 targeting vector. To construct a targeting vector, genomic sequences from the murine Pdha1 gene were obtained by screening a ␭ bacteriophage library (129SV ␭ fix II genomic library, Stratagene). A probe for screening the library was generated by amplifying an approximately 300-bp intronic sequence that corresponds to intron 7 of the human PDHA1 gene. The sense primer (Pdha1S1) annealing to exonic sequence 5⬘ to the intron and antisense primer (Pdha1AS1) annealing to exonic sequence 3⬘ to the intron were synthesized with the following respective sequences: 5⬘ AGCAGCCAGCACGGACTACT 3⬘ and 5⬘ TGTTGCCTCTCGGACGCACAAGATA 3⬘. Thermocycling conditions were 30 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 2 min. An isolated clone containing the 3⬘ half of the Pdha1 gene was then used to construct a targeting vector with approximately 7 kb of homology (Figs. 1B, 1C). To circumvent the problems of targeting a potentially lethal X-linked gene using any of the commonly available embryonic stem (ES) cell lines which are derived from male embryos, a strategy using the Cre–loxP system was devised to introduce a silent mutation into ES cells that could then be induced to produce a null mutation in vivo (5). The initial step in construction of the targeting vector was to insert a 0.3-kb MscI–SalI genomic fragment containing exon 8 into the ploxP vector that has two loxP sites in direct repeat orientation (gift from Dr. R. Jaenisch, MIT). A 3.7-kb EcoRI–MscI genomic fragment containing flanking upstream sequence from the Pdha1 gene was then cloned adjacent to the 5⬘ loxP sequence. A cassette containing both a 3 kb SalI–EcoRV downstream genomic fragment and the diphtheria toxin fragment A (6) was cloned into a SalI site downstream of the 3⬘ loxP site. Another cassette containing a neomycin phosphotransferase gene (neo) driven by the pgk-1 promoter and lacking a polyadenylation signal

(7), a pMC1-tk thymidine kinase gene (tk) (8), and a third loxP site was cloned into a ClaI site located between the 3⬘ genomic sequence and the downstream loxP site. A NotI site located within the vector just 5⬘ to the region of homology was used to linearize the construct. Gene targeting, Cre-mediated recombination, and analysis of the Pdha1 ⌬ex8 allele in ES cells. The E14.1 ES cell line was electroporated with the targeting construct, selected for integration and karyotyped as described by Johnson et al. (9). Clonal ES cell lines that were resistant to G418 selection were screened by polymerase chain reaction (PCR) analysis to identify the presence of the neo/tk cassette in intron 8, referred to as the Pdha1 neo/tk8 allele (Figs. 1D, 1E). An antisense primer, Pdha1AS2, that anneals to an intron 8 downstream of the SalI cloning site (5⬘ GCAGCCAAACAGATTACACC 3⬘), and the previously described sense primer, Pdha1S1, were used for amplification with thermocycling conditions as previously described. A pair of primers amplifying a region of the murine Dld gene were included as a control for amplification: sense primer 5⬘ CACTAAGCTCCATCTTCAGCCATGAG 3⬘ and antisense primer 5⬘ GGTCTGTTTTTATCTTTAGACAGAGCCAAAAAA 3⬘. Clones that did not yield amplification products using this primer pair were further analyzed by Southern blot analyses using the intron 7 probe following digestion with several restriction endonucleases including EcoRI, ScaI, and PvuII, which resulted in different sizes of restriction fragments in the wild-type and neo/tk alleles (Figs. 1D, 1E). Several ES cell lines that had undergone homologous recombination were then electroporated with supercoiled plasmid pMC-cre that expresses the Cre recombinase (10). Two days after electroporation, the cells were passaged and then subjected to selection with gancyclovir at a concentration of 1 ␮M on the following day. Viable colonies were picked and analyzed by PCR using the Pdha1S1 and Pdha1AS2 primers. Cell lines carrying the Pdha1 flox8 allele resulting from recombination between the two proximal loxP sites (Fig. 1F) and cell lines carrying the Pdha1 ⌬ex8 allele resulting from recombination between the distal two loxP sites (Fig. 1G) were subjected to Western blot analysis according to Ho et al. (11). Cells were homogenized and sonicated without and then with Triton X-100. Supernatants were collected following separation by microcentrifugation and frozen at ⫺70°C. Aliquots were size-fractionated by SDS–PAGE and examined by standard im-

TARGETED DISRUPTION OF THE MURINE Pdha1 GENE

295

FIG. 1. Schematic representation of the Pdha1 gene, the targeting vector, and the various alleles produced in mice. (A) Map of the Pdha1 gene. Exons are indicated above the line with locations based on the human PDHA1 gene (32) and restriction endonuclease recognition sites are located below the line. (B) Region of homology within the Pdha1 gene. The exons are indicated as unfilled boxes with restriction sites below the line. The intron 7 fragment used for screening of the genomic library and for various Southern blot analyses is indicated as a filled-in box above the map. (C) Targeting vector. Three loxP sites are indicated as bars with overlying asterisks. The neomycin phosphotransferase gene (neo) and the thymidine kinase gene (tk) with flanking loxP sites have been inserted into the SalI site. New restriction endonuclease recognition sites that are contained within inserted sequences are denoted below. (D–G) Expanded region surrounding exons 7 and 8 of various Pdha1 alleles and results of amplification with primers Pdha1S1 and Pdha1AS2 (arrowheads above the sequence). The sizes of amplification products using these primers are listed to the right of the figures.

munoblotting techniques with a polyclonal antiPDC rabbit primary antibody. To determine enzymatic activity in ES cell lines carrying the Pdha1 flox8 or Pdha1 ⌬ex8 allele, cells were washed twice in phosphate buffered saline and resuspended in KCl–Mops buffer containing protease

inhibitors. Following centrifugation, samples were freeze–thawed, dephosphorylated with recombinant rat phosphopyruvate dehydrogenase phosphatase, and stored at ⫺80°C. Total PDC and E1 component activities were assayed by measuring the production of 14CO 2 from [1- 14C]pyruvate (12). The activity of

296

JOHNSON ET AL.

dihydrolipoamide dehydrogenase, the third catalytic component of PDC, was measured spectrophotometrically (13). Generation of Pdha1 flox8 animals. ES cells were injected into C57Bl/6 blastocysts to generate chimeric embryos that were transferred to the uteri of pseudopregnant CH3B1/6 females. The resulting chimeric progeny were then bred with Black Swiss females to assess for germ-line transmission by the presence of agouti progeny. Agouti pups were genotyped by previously described PCR techniques to detect transmission of the Pdha1 flox8 allele. Germline-transmitting chimeras were then bred with 129/J females to maintain the targeted allele in an inbred 129/J genetic background. Progeny carrying the Pdha1 flox8 allele were then interbred to produce a line of animals that carried only the Pdha1 flox8 allele. Generation and analysis of Pdha1 flox8: Pdha1 ⌬ex8 mosaic animals. To induce deletion of exon 8 in vivo, homozygous Pdha1 flox8 female mice were mated with male mice from the EIIaCre transgenic line that is is derived from the FBV/N inbred background and carries an autosomally integrated Cre transgene. The transgene has an adenoviral EIIa promoter that expresses Cre enzyme in many embryonic tissues including the germ line (14). Progeny were genotyped as previously described using Pdha1S1 and Pdha1AS2 primers to distinguish the three Pdha1 alleles (Figs. 1D, 1F, 1G). The Cre transgene was detected by PCR using a primer pair that anneals sequences within the coding region: sense primer 5⬘ CGTACTGACGGTGGGAGAAT 3⬘ and antisense primer 5⬘ CCCGGCAAAACAGGTAGTTA 3⬘. To increase the degree of Pdha1 ⌬ex8 mosaicism in the F4 generation, the F3 mice carrying the Pdha1 ⌬ex8 allele and Cre transgene were intercrossed. The litter size from the mating was compared with that of mice carrying the Pdha1 flox8 alleles. Genomic DNA isolated from the tails of newly weaned animals was digested with SacI and BglII and hybridized with the intron 7 probe to quantitate more accurately the degree of deletion of the flox allele. To examine the prenatal development of mosaic animals, females from the F3 interbreeding were sacrificed at 8.5 to 12.5 days postcoitus (dpc). Embryos were dissected from the decidua with removal of Reichert’s membrane, photographed under a dissecting microscope, and genotyped by PCR as previously described. The degree of deletion of exon 8 within the embryos was also assessed by Southern

blot analysis in which genomic DNA was isolated from 9.5- to 12.5-dpc embryos, digested with SacI and BglII restriction enzymes, and probed with the 300-bp intron 7 probe. Embryos were prepared for histologic analysis by fixing in Z-fix solution (formaldehyde, ionized zinc, and buffer), dehydrating in increasing concentrations of ethanol, and embedding in paraffin. A series of 4-␮m sections were stained with hematoxylin– eosin prior to light microscopic examination. RESULTS Targeting of the Pdha1 allele and Cre-mediated recombination in ES cells. Following electroporation of the targeting vector and G418 selection, 36 of 300 (12%) ES cell colonies analyzed were shown to have undergone homologous recombination as determined by PCR and Southern blot analysis (Fig. 2A). Several of the ES cell lines carrying the loxP– neo/tk–loxP cassette (Pdha1 neo/tk8) were then electroporated with a supercoiled plasmid to express transiently the Cre recombinase. The Cre-catalyzed recombination between loxP sites could produce either of two different types of deletions that would eliminate the thymidine kinase selectable marker. Following electroporation and selection with gancyclovir, 8 cell lines were identified to have the Pdha1 flox8 allele and 10 were found to have the Pdha1 ⌬ex8 allele (Figs. 1F, 1G, 2B). Western blot analysis of cell lines carrying the Pdha1 neo/tk8, Pdha1 flox8, and Pdha1 ⌬ex8 alleles showed that the ES cell lines carrying the neo/tk and flox alleles had normal levels of E1␣ and E1␤ proteins whereas the ⌬ex8 allele had dramatically reduced levels of both the E1␣ and E1␤ proteins (Fig. 2C). Enzymatic analysis of extracts from ES cells containing the Pdha1 ⌬ex8 allele demonstrated significantly reduced activities of both total PDC and E1 component, whereas the E3 component activity was normal compared with cell lines carrying the wildtype or Pdha1 flox8 alleles (Table 1). Generation of chimeras and a mouse line carrying the Pdha1 flox8 allele. Three clonal ES cell lines carrying the Pdha1 flox8 allele were injected into recipient blastocysts to generate chimeric animals. The progeny of germ line chimeras carrying the Pdha1 flox8 allele did not show any abnormalities in growth or fecundity (data not shown). Females heterozygous for the Pdha1 flox8 allele were intercrossed with males carrying the same allele. Of the 30 prog-

TARGETED DISRUPTION OF THE MURINE Pdha1 GENE

297

FIG. 2. Analysis of ES cells following Cre-mediated recombination. (A) Amplification of control and several clonal ES cell lines following electroporation with the targeting vector. The template DNA used is as follows: W, wild type; C1, C2, and C3, three ES clonal lines with introduced targeting vector; N, no DNA. Addition of the Pdha1 primers (both S1 and AS2) and Dld primers is indicated. The Pdha1 primers produce a 700-bp product whereas the Dld primers produce an 800-bp product. Clonal lines 1 and 3 show loss of the Pdha1 amplification product due to integration of the neo/tk cassette. (B) PCR amplifications following transient Cre expression in ES cells carrying the Pdha1 neo/tk8 allele. Recombination between loxP sites flanking the neo/tk cassette resulted in Pdha1 flox8 or Pdha1 ⌬ex8 alleles with correspondingly sized amplification products. (C) Western blot analysis of ES cells carrying the four types of Pdha1 alleles. All alleles had normal levels of E1␣ and E1␤ with the exception of the Pdha1 ⌬ex8 allele, which showed dramatically reduced levels of both proteins.

eny genotyped from the F2 generation, 4 (13.3%) showed only amplification of the Pdha1 allele, 21 (70%) amplified only the Pdha1 flox8 allele, and 5 (16.6%) demonstrated the presence of both Pdha1 and Pdha1 flox8 alleles. This ratio did not deviate significantly from the expected 1:2:1 Mendelian proportions expected for an X-linked gene, indicating that the Pdha1 flox8 allele in the homozygous or hemizygous state did not compromise viability. Furthermore, animals carrying the Pdha1 flox8 allele(s) did not have any noticeable physical differences compared with their littermates. As seen in the ES cell enzymatic studies, total PDC activity was similar in the livers of littermates carrying the Pdha1 or Pdha1 flox8 allele (5.99 ⫾ 1.57 and 5.33 ⫾ 0.33 mU of

total PDC/mg of protein) in Pdha1 and Pdha1 flox, respectively.

TABLE 1 Pyruvate Dehydrogenase Activities of ES Cells

Breeding of mosaic Pdha1 flox8/Pdha1 ⌬ex8 animals and analysis of progeny. Interbreeding of F3 mosaic animals resulted in a significantly reduced litter size (3.5 pups per litter, P ⫽ ⬍0.01) compared with the offspring of animals carrying the wild-type or flox alleles (Table 2). In comparing the F3 to F4 generations, there was a more pronounced decrease in females, which would be expected since all females in the F3 generation carried one Pdha1 allele whereas only half of the females in the F4 genera-

Genotype

PDC a

E1 a

E3 a

Pdha1 Pdha1 neo/tk8 Pdha1 flox8 Pdha1 ⌬ex8

4.80 ⫾ 0.56 5.33 ⫾ 0.21 4.95 ⫾ 0.25 0.004 ⫾ 0.005

0.14 ⫾ 0.01 0.12 ⫾ 0.01 0.15 ⫾ 0.04 0.012 ⫾ 0.01

79.8 ⫾ 4.1 85.1 ⫾ 2.8 74.4 ⫾ 2.2 83.4 ⫾ 3.1

a Expressed as mU/mg protein, assays performed in quadruplicate. The results are means ⫾ SD.

Generation of mosaic animals carrying the Pdha1 ⌬ex8 allele. Homozygous Pdha1 flox8 female mice were mated to EIIaCre transgenic males to generate progeny carrying deletion of exon 8. The average litter size was noted to be smaller, although not significantly than for the F2 generation (Table 2). The combined litters had fewer males as compared with females, although the difference was not statistically significant. All animals that carried the Pdha1 flox8 allele showed some degree of Cre-catalyzed deletion of exon 8 as assayed by PCR genotyping. Southern blot analysis of several of the animals confirmed the PCR results (Fig. 3). The mosaic animals did not have any abnormal physical features or growth patterns.

298

JOHNSON ET AL.

TABLE 2 Genotypes of Pups from Different Mating Pairs a Progeny genotype

Pdha1/ Pdha1 flox8

Pdha1/ Pdha1 flox8/ Pdha1 ⌬ex8

Pdha1 flox8/ Pdha1 ⌬ex8

Pdha1/ Pdha1 ⌬ex8

Mating type Parental generation

Female

⫻

Male

Pdha1 flox8/Y

No. of pups

Litter size

Sex

Pdha1

Pdha1

flox8

F1

Pdha1/ Pdha1 flox8

30

7.5

M 14

—

9

5

—

—

—

F2

Pdha1 flox8/ Pdha1/Y; 26 Pdha1 flox8; Cre⫹/⫹ Cre⫺/⫺ Pdha1/ Pdha1 flox8: 14 Pdha1 flox8: Pdha1 ⌬ex8/Y; Pdha1 ⌬ex8; Cre⫹/⫺ Cre⫹/⫺

6.5

F 16 M8 F 18

4 — —

12 — —

— — —

— — 16

— 8 —

— — 2

M9 F5

9 —

— —

— 4

— —

— —

— 1

F3

3.5

a

Seventy tail DNA samples were amplified by PCR and separated on 1% agarose gel. The distribution of genotypes from each mating is shown.

tion contained a wild-type allele. In the F4 generation, the numbers of Pdha1 hemizygous males and heterozygous females would be anticipated to be the same based on the Mendelian distribution. The lower number of females in the F4 generation was not statistically significant. Of note, there were no progeny in the F4 generation that carried the only detectable Pdha1 ⌬ex8 allele. To investigate whether the reduced number of animals, particularly those carrying the Pdha1 ⌬ex8 allele, was caused by reduced fertility or prenatal

FIG. 3. Southern blot analysis of genomic DNA from mouse tail and embryos. Genomic DNA was digested with SacI and BglII and probed with a 300-bp PCR-amplified intron 7 probe. Pdha1, Pdha1 flox8, and Pdha1 ⌬ex8 alleles were identified as 1000-, 1100-, and 650-bp fragments, respectively. (A–E) Tail DNA: (A) wild-type animal (Pdha1 allele), (B) heterozygous F2 female (Pdha1/Pdha1 flox8 alleles), (C) homozygous F3 female (Pdha1 flox8/ Pdha1 flox8 alleles), (D) mosaic F3 female (Pdha1/Pdha1 flox8: Pdha1 ⌬ex8 alleles), (E) heterozygous with high-degree mosaic F4 female (Pdha1/Pdha1 ⌬ex8 alleles). (F) Combination of three pooled abnomal embryos from 9.5 to 12.5 dpc (Pdha1 ⌬ex8 allele). (G) Size markers.

lethality, embryos from the F3 intercross were collected at 7.5 to 12.5 dpc, photographed, and genotyped by PCR analysis and Southern blot analysis (Table 2, Fig. 4A). All embryos collected at 7.5 dpc were phenotypically indistinguishable. Genotypic analysis at this stage was not possible due to difficulty separating embryonic from maternal tissue. One day later, all of the embryos had developed to the eight-somite stage without showing any gross phenotypic abnormalities. In collecting the embryos, two decidua were found to lack embryos. This number of failed embryos did not differ significantly from the background rate of postimplantation lethality in mice, which has been estimated to be approximately 10% (15). Genotypic analysis of one litter at 8.5 dpc identified two embryos that carried the Pdha1 ⌬ex8 allele in the majority of their cells. By 9.5 dpc, approximately 50% of the embryos were smaller and delayed in development by approximately 1 day (Fig. 4A). Genotyping of the smaller embryos showed that these embryos carried the Pdha1 ⌬ex8 allele in most cells, although precise quantitation could not be determined from the PCR assays. Southern blot analysis on pooled embryos (three to five embryos with abnormal phenotype from 9.5 to 12.5 dpc) confirmed that the majority of cells in these embryos had only the Pdha1 ⌬ex8 allele (Fig. 3, lane G). Comparing the size of the mutant embryos between 9.5 and 12.5 dpc revealed little overall growth which was consistent with the increasing size discrepancy noted between mutant and normal

TARGETED DISRUPTION OF THE MURINE Pdha1 GENE

299

9.5 to 12.5 dpc, 20 (26%) amplified only a Pdha1 allele; 29 (39%) amplified Pdha1, Pdha1 flox8, and Pdha1 ⌬ex8 alleles, indicating mosaic female heterozygotes; and 26 (35%) amplified both Pdha1 flox8 and Pdha1 ⌬ex8 alleles, identifying mosaic homozygotes or hemizygotes. This distribution was not significantly altered from the expected 1:2:1 Mendelian ratio. To examine the mutant phenotype in more detail, embryos collected and dissected at 7.5 to 12.5 dpc were examined histologically. The 7.5-dpc embryos did not show any difference in size or morphology. By 10.5 dpc, embryos carrying the Pdha1 ⌬ex8 allele were similar in size to 8.5-dpc wild-type embryos with no signs of additional somites or organ formation. Areas of necrosis suggested that these embryos were in the process of deterioration (Fig. 4B). Over the next 2 days of gestation, embryos carrying the deleted allele showed increasing signs of resorption with invasion of maternal macrophages and lymphocytes. DISCUSSION

FIG. 4. Analysis of mouse embryos carrying the Pdha1 ⌬ex8 allele. (A) Mosaic males (Pdha1 flox8:Pdha1 ⌬ex8/Y; Cre⫹/⫺) were mated with mosaic females (Pdha1/Pdha1 flox8:Pdha1 ⌬ex8; Cre⫹/⫺) and embryos were collected at 9.5–12.5 dpc, photographed, and genotyped. The genotypes are noted above the photographs. (B) Embryos collected at 10.5 dpc were sectioned in the parasagittal plane. The genotypes are noted below images. A, anterior; D, dorsal; P, posterior; V, ventral.

embryos over this period. In looking at the number of resorption sites during the period of embryologic analysis, the rate approximately doubled from about 10 to 20% over this period, suggesting that some of the mutant embryos had already been resorbed. Of the 75 embryos analyzed by PCR over the period of

In the more than 150 patients with deficiency of PDC, none thus far has been identified with a complete loss of enzymatic activity (16). To gain insight into the null phenotype, a gene targeting strategy was developed to generate mice carrying a null mutation in the Pdha1 gene. The potential problems associated with the X-linked location of Pdha1 and the conventional use of male ES cell lines were circumvented through the use of a two-step Cre–loxP strategy. A plan was devised to delete in vivo exon 8, a region of the E1␣ protein that has been postulated to be involved in subunit–subunit interactions (17). The consequences of both the insertion of loxP sites and the deletion of exon 8 were examined in ES cells prior to making chimeric animals. Deletion of exon 8 in ES cells resulted in a dramatic reduction of E1␣ protein and PDC activity (Table 1, Fig. 2C). A higher residual activity detected by E1 component assay (8%) as compared with the PDC assay (0.01%) could be explained by either a decreased affinity of the altered E1 component for the complex or a technical limitation of the E1 assay since E1 activity is less than 5% of total PDC activity in the decarboxylation assay, making it more susceptible to variations in the background. The significant loss of PDC activity was interpreted as the creation of a null mutation. Interestingly, Western blot analysis showed dramatic reductions in the amounts of both E1␣ and E1␤ subunit proteins. The reduced level of E1␤ pro-

300

JOHNSON ET AL.

tein is most likely due to mutually stabilizing E1␣– E1␤ interactions that occur within the E1 heterotetramer. A similar effect on E1␤ subunit abundance has been described in several cases of human E1␣ deficiency (11,18). The ability to maintain ES cells carrying the Pdha1 deletion in culture also demonstrated that ablation of PDC activity did not result in a cell autonomous lethal. During maintenance of the Pdha1 ⌬ex8 cell lines, the dividing time of the Pdha1 ⌬ex8 cells was grossly similar to that of wildtype ES cells, although the pH of the culture medium was more rapidly acidified presumably due to increased production of lactic acid as a result of increased glycolysis (data not shown). Progeny from the F1 generation that carried the Pdha1 flox8 allele in the hemizygous or heterozygous state were phenotypically indistinguishable from wild-type littermates. Homozygous Pdha1 flox8 females were then bred to males from the EIIaCre transgenic line. The progeny from this mating produced animals in which exon 8 had been partially deleted in various tissues. The mosaicism was not unexpected considering the breeding scheme in which the Cre transgene was transmitted via the male gamete. The absence of maternally encoded Cre protein would be expected to result in delayed expression of Cre protein until the two-cell stage or later based on studies of expression of proteins encoded by the zygotic genome (19). Previous studies of expression of the EIIa promoter revealed that this promoter is not active in the testis but would be activated on formation of the zygote (20). Several reports including the initial report of development of the EIIaCre transgenic line have noted mosaicism due to incomplete recombination between loxP sites when the Cre was contributed by either the maternal or the paternal germ line (14,21–23). The generation of mosaic PDC-deficient animals also demonstrates that PDC deficiency does not cause cell autonomous lethality in vivo. Two phenotypically normal female F3 progeny that appeared to be completely heterozygous for the Pdha1 ⌬ex8 alleles reinforce the notion that females may carry a null mutation without dramatic phenotypic consequences as has been previously observed in human females (24). Interbreeding of mosaic animals resulted in smaller litter sizes with no newborn animals carrying solely the Pdha1 ⌬ex8 allele. The most plausible explanation was that PDC deficiency resulted in prenatal lethality, an explanation previously proposed for humans. In this study, the reduction in the

numbers of both male and female mosaic mice in the F4 generation suggested a portion of carrier females may also die prenatally. The early embryonic lethality of some of the female carriers may be due either to skewed inactivation of the X chromosome carrying the wild-type Pdha1 allele in critical tissues or failure of cells with a functional allele to complement E1␣-deficient cells in vivo when the proportion of wild-type cells falls below a certain threshold. To date, there have been no documented cases of prenatal lethality due to PDHA1 mutations in heterozygous human females. Embryologic analysis of the progeny from interbreeding of mosaic animals has revealed that embryos carrying the Pdha1 ⌬ex8 allele begin to fall behind in growth and development by 9.5 dpc. These embryos appear to be normally formed, suggesting that there is a uniform slowing of cellular proliferation rather than damage to selected cell types. This observation is consistent with in vitro studies that have shown there to be little differences in glucose utilization in different embryonic tissues of the 8.5dpc embryo (25). By 10.5 dpc, the abnormal embryos carrying the Pdha1 ⌬ex8 allele are necrotic and have not developed significantly beyond the 8.5-dpc stage. The exact timing of the onset of developmental impairment in embryos carrying the Pdha1 ⌬ex8 allele cannot be precisely determined from this analysis since the embryos could potentially be mosaic and have residual PDC activity. However, the embryologic analysis presented here provides convincing proof for a requirement of PDC in the early postimplantation period. Corroborating data are provided in a recent report describing the prenatal lethality of male mice carrying the radiation-induced mutation stripey. Molecular analysis has now revealed stripey to be caused by an approximately 5-megabase deletion on the X chromosome that includes the Pdha1 gene (26). Although there are many additional genes that are deleted in stripey mice, this study suggests that the deletion of Pdha1 may play a significant role in this phenotype. Embryos deficient in PDC cease to develop at a time of dynamic change for the embryo. The early postimplantation period is a time of rapid cellular proliferation and differentiation during which the embryo completes gastrulation and initiates organogenesis. Thus, embryos impaired in their ability to oxidatively metabolize glucose would be expected to have difficulty during this period of increased biosynthetic activity. Previous work in our laboratory creating mice deficient in dihydrolipoamide dehy-

TARGETED DISRUPTION OF THE MURINE Pdha1 GENE

drogenase (E3 component) suggested that the early postimplantation embryo was susceptible to impairments in oxidative metabolism. The E3 component functions as an NAD ⫹ oxidoreductase not only for PDC but also for three other mitochondrial multienzyme complexes: ␣-ketoglutarate dehydrogenase complex, branched-chain ␣-keto acid dehydrogenase complex, and the glycine cleavage system. The E3 null phenotype manifests with delay in developmental approximately 2 days before that of PDC deficiency, during a period when the embryo is initiating gastrulation (27). The earlier stage of death in the case of E3 deficiency is presumably related to the combined impairment of PDC and the Krebs cycle due to a deficiency in ␣-ketoglutarate dehydrogenase complex. The fact that the E1-deficient embryos do not survive much longer than the E3 embryos indicates that embryos at this stage of development are not capable of using alternative energy sources of acetyl-CoA such as fatty acids to bypass the block in pyruvate metabolism. The reason for the lack of utilization of fatty acids may be either ineffective uptake or ineffective catabolism of fatty acids in the pathway prior to entry into the Krebs cycle. Further support for the role of oxidative metabolism comes from recent studies examining embryonic metabolism in vitro and targeting of other genes involved in intermediary metabolism. Houghton et al. (28) have shown, based on the studies of oxygen utilization and consumption/production of metabolites, that early postimplantation embryos are dependent on both oxidative phosphorylation and anaerobic glycolysis. Recent knockouts of the Tfam gene, a nuclear encoded transcription factor involved in regulation of mitochondrial DNA transcription and replication, and the Cyt c gene, encoding the cytochrome c component of the electron transport chain, both show significant developmental delay by 8.5 dpc. These knockout animals emphasize the importance of mitochondrial oxidative metabolism during this period of development (29,30). PDC-deficient embryos survive until the early postimplantation period due to either a lack of requirement for this enzyme during earlier stages of development or the reliance of embryos on maternally derived PDC. There is precedence for the persistence of maternally encoded proteins into the postimplantation period based on studies of the phosphoglucoisomerase isozymes (31). In vitro studies investigating the metabolism of preimplantation embryos indicate that these embryos use oxidative

301

phosphorylation at stages prior to implantation, thereby providing some support for the latter hypothesis (28). The development of a line of mice carrying an inducible null mutation in the Pdha1 gene should be valuable for further studies investigating the importance of oxidative metabolism of glucose in early embryos and postnatal animals. Furthermore, this PDC-deficient murine line may aid both in testing currently existing therapies and in developing new therapies for pyruvate dehydrogenase complex deficiency. ACKNOWLEDGMENTS This work was supported by U.S. Public Health Service Grants DK20478 (M.S.P.) and DK22541 (R.W.H.). We thank Dr. Douglas S. Kerr, M.D., Ph.D., and Marilyn Lusk for their assistance in enzymatic assays of embryonic stem cells. We also thank Gerald Goldberg for the maintenance of the mouse colony used in this study.

REFERENCES 1.

Patel MS, Roche TE. Molecular biology and biochemistry of pyruvate dehydrogenase complexes. FASEB J 4:3224 –3233, 1990.

2.

Brown RM, Dahl HHM, Brown GK. Pyruvate dehydrogenase E1␣ subunit genes in the mouse: Mapping and comparison with human homologues. Somat Cell Mol Genet 16:487– 492, 1990.

3.

Dahl HHM, Brown RM, Hutchison WM, Maragos C, Brown GK. A testis-specific form of the human pyruvate dehydrogenase E1 alpha subunit is coded for by an intronless gene on chromosome 4. Genomics 8:225–232, 1990.

4.

Kerr DS, Wexler ID, Tripatara A, Patel MS. Human defects of the pyruvate dehydrogenase complex. In Alpha-Keto Acid Dehydrogenase Complexes (Patel MS, Roche TE, Harris RA, Eds.). Basel: Birkhauser Verlag, pp. 249 –269, 1996.

5.

Gu H, Marth JD, Orban PC, Mossmann H, Rajewsky K. Deletion of a DNA polymerase beta gene segment in T cells using cell type-specific gene targeting. Science 265:103–106, 1994.

6.

Tomasiewicz H, Ono K, Yee D, Thompson C, Goridis C, Rutishauser U, Magnuson T. Genetic deletion of a neural cell adhesion molecule variant (N-CAM-180) produces distinct defects in the central nervous system. Neuron 11:1163– 1174, 1993.

7.

Rudnicki MA, Braun T, Hinuma S, Jaenisch R. Inactivation of MyoD in mice leads to up-regulation of the myogenic HLH gene Myf-5 and results in apparently normal muscle development. Cell 71:383–390, 1992.

8.

Mansour SL, Thomas KR, Capecchi MR. Disruption of the proto-oncogene int-2 in mouse embryo-derived stem cells: A general strategy for targeting mutations to nonselectable genes. Nature 336:348 –352, 1988.

9.

Johnson MT, Yang HS, Patel MS. Targeting E3 component

302

JOHNSON ET AL. of alpha-keto acid dehydrogenase complexes. Methods Enzymol 324:465– 476, 2000.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Gu H, Zou YR, Rajewsky K. Independent control of immunoglobulin switch recombination at individual switch regions evidenced through Cre-loxP-mediated gene targeting. Cell 73:1155–1164, 1993. Ho L, Hu CWC, Packman S, Patel MS. Deficiency of the pyruvate dehydrogenase component in pyruvate dehydrogenase complex-deficient human fibroblasts. J Clin Invest 78: 844 – 847, 1986. Kerr DS, Ho L, Berlin CM, Lanoue KF, Towfighi J, Hoppel CL, Lusk MM, Gondek CM, Patel MS. Systemic deficiency of the first component of the pyruvate dehydrogenase complex. Pediatr Res 22:312–318, 1987. Patel MS, Vettakkorumakankav NN, Liu TC. Dihydrolipoamide dehydrogenase: Activity assays. Methods Enzymol 252:186 –195, 1995. Lakso M, Pichel JG, Gorman JR, Sauer B, Okamoto Y, Lee E, Alt FW, Westphal H. Efficient in vivo manipulation of mouse genomic sequences at the zygote stage. Proc Natl Acad Sci USA 93:5860 –5865, 1996. Bradford GE. Genetic control of ovulation rate and embryo survival in mice. I. Response to selection. Genetics 61:905– 921, 1969. Dahl HHM, Brown GK, Brown RM, Hansen LL, Kerr DS, Wexler ID, Patel MS, De Meirleir L, Lissens W, Chun K. Mutations and polymorphisms in the pyruvate dehydrogenase E1 alpha gene. Hum Mut 1:97–102, 1992.

22.

23.

24.

25.

26.

27.

28.

Wexler ID, Hemalatha SG, Patel MS. Sequence conservation in the alpha and beta subunits of pyruvate dehydrogenase and its similarity to branched-chain alpha-keto acid dehydrogenase. FEBS Lett 282:209 –213, 1991.

29.

Huq AH, Ito M, Naito E, Saijo T, Takeda E, Kuroda Y. Demonstration of an unstable variant of pyruvate dehydrogenase protein (E1) in culture fibroblasts from a patient with congenital lactic acidemia. Pediatr Res 30:11–14, 1991.

30.

Latham KE, Garrels JI, Chang C, Solter D. Quantitative analysis of protein synthesis in mouse embryos. I. Extensive reprogramming at the one- and two-cell stages. Development 112:921–932, 1991.

31.

Dooley T, Miranda M, Jones N, de Pamphilis M. Transactivation of the adenovirus E1A protein is restricted to mouse oocytes and preimplantation embryos. Development 107: 945–956, 1989.

32.

Liu JL, Grinberg A, Westphal H, Sauer B, Accili D, Karas M,

LeRoith D. Insulin-like growth factor-I affects perinatal lethality and postnatal development in a gene dosage-dependent manner: Using Cre/loxP system in transgenic mice. Mol Endocrinol 12:1452–1462, 1998. Holzenberger M, Lenzner C, Leneuve P, Zaoui R, Hamard G, Vaulont S, Le Bouc Y. Cre-mediated germline mosaicism: A method allowing rapid generation of several alleles of a target gene. Nucleic Acids Res 28:E92, 2000. Keller P, Tremml G, Rosti V, Bessler M. X inactivation and somatic cell selection rescue female mice carrying a Piganull mutation. Proc Natl Acad Sci USA 96:7479 –7483, 1999. Dahl HHM, Hansen LL, Brown RM, Danks DM, Rogers JG, Brown GK. X-linked pyruvate dehydrogenase E1 alpha subunit deficiency in heterozygous females: Variable manifestation of the same mutation. J Inher Metab Dis 15:835– 847, 1992. Wales RG, Martin KL, Leese HJ. Glucose utilization by components of the mouse conceptus during early embryogenesis. J Reprod Fert 104:125–132, 1995. Blair HJ, Gormally E, Uwechue IC, Boyd Y. Mouse mutants carrying deletions that remove the genes mutated in Coffin– Lowry syndrome and lactic acidosis. Hum Mol Genet 7:549 – 555, 1998. Johnson MT, Yang HS, Magnuson T, Patel MS. Targeted disruption of the murine dihydrolipoamide dehydrogenase gene (Dld) results in perigastrulation lethality. Proc Natl Acad Sci USA 94:14512–14517, 1997. Houghton FD, Thompson JG, Kennedy CJ, Leese HJ. Oxygen consumption and energy metabolism of the early mouse embryo. Mol Reprod Dev 44:476 – 485, 1996. Larsson NG, Wang J, Wilhelmsson H, Oldfors A, Rustin P, Lewandoski M, Barsh GS, Clayton DA. Mitochondrial transcription factor A is necessary for mtDNA maintenance and embryogenesis in mice. Nat Genet 18:231–236, 1998. Li K, Li Y, Shelton JM, Richardson JA, Spencer E, Chen ZJ, Wang X, Williams RS. Cytochrome c deficiency causes embryonic lethality and attenuates stress-induced apoptosis. Cell 101:389 –399, 2000. West JD, Leask R, Green JF. Quantification of the transition from oocyte-encoded to embryo-encoded glucose phosphate isomerase in mouse embryos. J Embryol Exp Morphol 97:225–227, 1986. Maragos C, Hutchison WM, Hayasaka K, Brown GK, Dahl HHM. Structural organization of the gene for the E1 alpha subunit of the human pyruvate dehydrogenase complexes. J Biol Chem 264:12294 –12298, 1989.