Targeting E3 Component of α-Keto Acid Dehydrogenase Complexes

[42]

TARGETINGE3 OF PDC/KDC/BCKDC

465

[42] Targeting E3 Component of ~-Keto Acid Dehydrogenase Complexes By M A R K

T. JOHNSON, HSIN-SHENG YANG, a n d MULCHAND S. PATEL

Introduction Branched-chain a-keto acid dehydrogenase complex (BCKDC) is a mitochondrial multienzyme complex that catalyzes a series of reactions that form the first irreversible step in the catabolism of the essential branched-chain amino acids: leucine, isoleucine, and valine. After transamination of these amino acids to their respective ot-keto acids, BCKDC catalyzes the irreversible oxidative decarboxylation and subsequent formation of acyl-CoAs. BCKDC is estimated to be 4-5 million Da in size and consists of multiple copies of three enzymatic and two regulatory components. The three enzymatic components of BCKDC are abbreviated as E1b, E2b, and E3 components, indicating their sequential roles in the enzymatic cascade. The E3 component, also referred to as dihydrolipoamide dehydrogenase (EC 1.8.1.4), is shared among two other ot-keto acid dehydrogenase complexes, namely pyruvate dehydrogenase and ot-ketoglutarate dehydrogenase complexes, as well as with the glycine cleavage system. 1'2 E3 functions in these complexes by oxidizing the dihydrolipoyl moiety of the transacylase components and transferring the electron pair to a recipient NAD + molecule. Deficiency of E3 3-8 results in an inborn error of metabolism referred to as a variant form of maple syrup urine disease (MSUD). The presence of branched-chain keto acids and lactate in the urine of patients with MSUD imparts a distinctive odor. Branched-chain keto acids in the urine are a N. N. Vettakkorumakankav and M. S. Patel, Ind. J. Biochem. Biophys. 33, 168 (1996). 2 L. J. Reed, Acc. Chem. Res. 7, 40 (1976). 3 B. H. Robinson, J. Taylor, and W. G. Sherwood, Pediatr. Res. 11, 1198 (1977). 4T. C. Liu, H. Kim, C. Arizmendi, A. Kitano, and M. S. Patel, Proc. Natl. Acad. Sci. U.S.A. 90, 5186 (1993). 5 j. p. Bonnefont, D. Chretien, P. Rustin, B. Robinson, A. Vassault, J. Aupetit, C. Charpentier, D. Rabier, J. M. Saudubray, and A. Munnich, J. Pediatr. 212, 255 (1992). 6 N. Guffon, C. Lopez-Mediavilla, R. Dumoulin, B. Mousson, C. Godinot, H. Carrier, J. M. Collombet, P. Dirty, M. Mathieu, and P. Gubaud, J. Inher. Metab. Dis. 16, 821 (1993). 7 A. Kohlschutter, A. Behbehani, U. Langenbeck, M. Albani, P. Heidemann, G. Hoffmann, J. Kleineke, W. Lehnert, and U. Wendel, Eur, J. Pediatr. 138, 32 (1982). 8 j. H. Cross, A. Connelly, D. G. Gadian, B. E. Kendall, G. K. Brown, R. M. Brown, and J. V. Leonard, Pediatr. Neurol. 10, 276 (1994).

METHODS IN ENZYMOLOGY.VOL. 324

Copyright© 2000by AcademicPress All rightsof reproductionin any formreserved. 0076-6879/00$30.00

466

DETECTION AND CONSEQUENCESOF GENETICDEFECTS

[421

indicative of a systemic problem; these keto acids accumulate, causing an organic acidemia and neurologic damage. There are several phenotypic categories of MSUD that, if untreated, all result in varying degrees of physical and mental retardation within the first 2 years of life.9,1° The E3deficient category manifests with a more complex organic acidosis, including elevated levels of lactate, and has a phenotype similar to the severe or classic form of MSUD, in which there is rapid neurologic deterioration that results in neonatal death, n This chapter discusses the generation of E3 knockout mice and the characterization of a null allele of the dihydrolipoamide dehydrogenase (Did) gene. Materials and Reagents

Bacteriophage libraries: h EMBL-3 bacteriophages containing M b o I partial digests of DBA/2J genomic DNA (Clontech, Palo Alto, CA) and h Fix II bacteriophages containing Sau3A partially digested 129SVJ genomic DNA (Stratagene, La Jolla, CA) Bacterial strains: LE392 for EMBL-3, XL-1 Blue MRA for Fix II, DH5a for plasmid cloning (BRL-Life Technologies, Gaithersburg, MD) Polymerases: Klenow fragment of DNA polymerase I, Taq polymerase Zeta-Probe GT blotting membrane (Bio-Rad, Hercules, CA) Plasmids: pUC 19 (BRL-Life Technologies),/3-actin promoter-driven neomycin phosphotransferase gene and lacking a polyadenylation signal, diptheria toxin fragment minigene cassette 12 Mouse strains: C57BL (Taconic, Germantown, NY), 129 SVJ (Jackson Laboratory, Bar Harbor, ME), MTKNeo213 Buffers and Media

Church hybridization solution: 500 mM sodium phosphate (pH 6.8), 1 mM EDTA, 7% (w/v) sodium dodecyl sulfate (SDS), and bovine serum albumin (BSA; 0.5%, w/v) TAE buffer: 40 mM Tris-acetate (pH 7.5) and 1 mM EDTA SSC solution: 150 mM NaCI and 15 mM sodium citrate Embryonic stem (ES) cell growth medium: Dulbecco's modified 9D. T. Chuangand V. E. Shih,in "The Metabolicand MolecularBase of InheritedDisease" (C. R. Scriver, A. L. Beaudet,W. S. Sly, and D. Vane, eds.), p. 1239. McGraw-Hill,New York, 1995. 10F. Peinemannand D. J. Danner, J. Inher. Metab. Dis. 17, 3 (1994). 11p. A. W. Harper, P. J. Healy,and J. A. Dennis,Acta Neuropathol. 71, 316 (1986). 12H. Tomasiewicz,K. Ono, D. Yee, C. Thompson,C. Goridis,U. Rutishauser,and T. Magnuson, Neuron 11, 1163 (1993). 13S. J. Abbondanzo,I. Gadi, and C. L. Stewart,Methods Enzymol. 225, 808 (1993).

[42]

TARGETINGE3 OF PDC/KDC/BCKDC

467

Eagle's medium (DMEM) with 4.5 g of glucose per liter (Life Technologies) supplemented with 15% (v/v) ES cell-tested fetal calf serum (JRH Biosciences, Lenexa, KS), 0.1 mM 2-mercaptoethanol (Sigma, St. Louis, MO), penicillin (100 U/ml), streptomycin (100 U/ml) (Life Technologies), and E S G R O (1000 U/ml; Life Technologies) Trypsin solution: 0.25% (w/v) trypsin, 0.02% (v/v) EDTA (JRH Biosciences) Phosphate-buffered saline (PBS) buffer: 137 mM NaCI, 0.3 mM KC1, 10 mM Na2HPO4, and 0.18 mM KH2PO4, pH 7.4 I,ow-TE electroporation solution: 10 mM Tris-HC1 (pH 7.5), 0.1 mM EDTA (3418100x stock solution: Dissolve 180 mg of G418 in 10 ml of DMEM, filter sterilize, and store aliquots at - 2 0 ° Freezing solution (2X): 60% (v/v) DMEM, 20% (v/v) fetal calf serum, 20% (v/v) dimethyl sulfoxide (DMSO), stored at 4 ° Genomic D N A extraction buffer: 10 mM Tris-HC1 (pH 7.5), 10 mM EDTA, 10 mM NaC1, 0.5% (w/v) sarcosyl, and proteinase K added immediately before use to final concentration of 1 mg/ml D N A precipitation solution: 1.5 /zl of 5 M NaC1 added to ice-cold 100% ethanol immediately before use Embryo lysis buffer: 100 mM Tris-HCl (pH 8.0), 2 mM MgC12, 0.01% (w/v) gelatin, 0.45% (v/v) Tween 20, 0.45% (v/v) Nonidet P-40 (NP40), and proteinase K (500/zg/ml)

Cloning of Mouse D/d Gene The mouse D l d gene is cloned from a A EMBL-3 mouse genomic D N A library prepared from the DBA/2J strain (Clontech). 14 The full-length human E3 cDNA 15 is random-prime labeled to incorporate approximately 1 × 10s cpm//~g 16 and used to screen approximately 1 × 106 recombinant bacteriophages. Hybridizations are performed in Church buffer overnight at 65 ° and then serially washed in progressively lower salt concentrations ranging from 1× SSC and 0.1% (w/v) SDS down to 0.1X SSC at 65°. The washed filters are then exposed to Kodak (Rochester, NY) XAR-5 film overnight at - 8 0 °. Positive plaques are picked with Pasteur pipettes and purified by several rounds of replating and rehybridization until a pure clonal population of hybridizing plaque-forming units (PFU) is identified. 14M. Johnson, H. S. Yang, G. L. Johanning, and M. S. Patel, Genomics 41, 320 (1997). 15G. Pons, C. Raefsky-Estrin,D. J. Carothers, R. A. Pepin, A. A. Javed, B. W. Jesse, M. K. Ganapathi, D. Samols, and M. S. Patel, Proc. Natl. Acad. Sci. U.S.A. 85, 1422 (1988). 16A. P. Feinberg and B. Vogelstein,Anal. Biochem. 132, 6 (1983).

468

D E T E C T I O N A N D C O N S E Q U E N C E S OF G E N E T I C DEFECTS

[42]

The phage DNA is then purified, digested with restriction enzymes EcoRl, BamHI, and Sail, and subjected to Southern hybridizations with the same probe and under similar hybridization conditions. The DNAs from the positive phage clones are isolated and subcloned into pUC19 (Life Technologies) or pBluescript KS + (Stratagene) vectors. Restriction enzymes mapping and partial nucleotide sequencing are performed to identify the intron and exon locations of the mouse Did gene. 14

Construction of D/d Gene Targeting Vector To create an isogenic construct, approximately 1.8 × 106 clones from an isogenic 129SV genomic library are screened with two labeled fragments from the DBA/2J genomic clones (a 0.7-kb SaII-SphI fragment containing exon 9 and a 0.8-kb HindII-Sau3A fragment within intron 4). In our experiments one clone isolated from this screen contained a 13-kb insert extending approximately from exon 3 to exon 1I. A targeting constuct is then designed by using the positive-negative selection strategy, 17 in which a positively selectable marker, a neomycin phosphotransferase driven by the fl-actin promoter (neo), is utilized as a sequence to disrupt the reading frame and a negative marker, the diptheria toxin fragment A (DT-A), 12 is included to select against random insertions of the vector within the ES cell genome. A 2-kb EcoRI-BgIII DT-A cassette with the/3-actin promoter is inserted by blunt-ended ligation into an EcoRI site of the pUC19 vector (this vector is named pUC/DT), thereby destroying the EcoRI site. A 7-kb Xbal fragment including the 3' half of the A insert is subcloned into the Xbal site of the pUC/DT vector to serve as the region of homology in the targeting vector, with roughly equal-sized arms of homology on either side of an EcoRI site within exon 10 at codon 301 (Fig. i). From structure-function studies, it has been shown that residues downstream of the EcoRI site, such as H452 and E457, are essential for function. TMThe Did gene is disrupted by cloning a neomycin phosphotransferase gene driven by the fl-actin promoter lacking a polyadenylation signal (neo) into the EcoRI site. neo is inserted in the same transcriptional orientation as the endogenous E3 in order to utilize the endogenous polyadenylation signal of the E3 gene following a homologous recombination event. The lack of a polyadenylation signal confers an additional power of selection against random insertions, as many of these events will not be conveniently located upstream of a polyadenylation signal. A NotI site from the A polylinker, which is carried along with 17S. L. Mansour,K. R. Thomas,and M. R. Capecchi,Nature (London) 336, 348 (1988). 18H. Kim and M. S. Patel, J. Biol. Chem. 167, 5128 (1992).

[42]

A

TARGETING E3 OF P D C / K D C / B C K D C

I.uc

'l°'l

8

9

n

n

Targeted Vector

I

11

~

n

12 I

~ /

6

7

8

n

n

n/ \n

n /

/

A I\

~.

469

//

/ ~ 9

10

///I

11

,/n/

~. I~,,'s K

,LI

13 14

n n

"Y-

WT Allele B B Probe A

B B Probe B <

;

•

B (4 kb)

~

<

EN (6 kb)

~

<

B

~

6

7

8

9

~_~/n

n

n

n

11

fl-~n

A + K (7 kb) S (11 kb)

12 13 14

nn

n//_

Targeted Allele <

L y

4 Y

B (6 kb) EN (6 kb) A + K (5.5 kb) S (9 kb)

FIG. 1. Targeting of the Dld gene. (A) Linearized targeting vector is composed of the pUC19 vector, the/3-actin-diphtheria toxin A chain gene (DT-A) negative maker, 7 kb of homology, and the/3-actin-neomycin phosphotransferase gene (neo). The wild-type Dld allele is represented by numbered boxes indicating exons with restriction enzyme sites listed below. The two probes used for distinguishing the wild-type from the targeted alleles are designated by the solid boxes. The sizes of the various restriction fragments and the enzymes used are shown below the restriction map. (B) The targeted allele after a recombination event is shown with the integrated neo and the altered restriction map. A, ApaI; B, BstXI; ER, EcoRI; EN, EcoNl; K, KpnI; S, SalI; X, XbaI.

470

DETECTION AND CONSEQUENCES OF GENETIC DEFECTS

[42]

the genomic fragment at its 3' end, is utilized as a site for linearizing the construct before electroporation. Targeted Disruption of D/d Gene in Embryonic Stem Cells The El4.1 embryonic stem cell line derived from the 129/O1a strain a9 is cultured under routine conditions similar to those described by RamirezSolis e t al. 2° Briefly, monolayers of primary embryonic feeders are prepared from day 13 embryos, 13 using the MTKNeo2 transgenic line that carries a neomycin resistance gene. Feeder cells are maintained for up to five passages in ES cell growth medium lacking ESGRO. To prepare a monolayer of growth-arrested feeders, the primary embryonic feeders are trypsinized, y irradiated with 1000 rads, and plated at a density of 1 x 106 cells per 60mm dish at least 6 hr prior to ES cell passaging. ES cells are maintained on the feeder layer in a 37° incubator with 5% CO2 with passaging every 2-3 days, in which 6 x 105 ES cells are plated per 60-mm dish after trypsinization. For electroporation, 7 X 106 ES cells and linearized targeting construct (25/zg/ml) are suspended in 750/zl of low-TE solution in 0.4-cm cuvettes and pulsed at settings of 270 V and 650/zF, using a Gene Pulser electroporator (Bio-Rad). The cells are incubated at room temperature for 10 min and then plated at 106 cells per 100-mm plate in ES cell growth medium for 24 hr, followed by the addition of selection medium containing G418 (125/zg/ml). After a 10-day selection period, G418-resistant clones are individually transferred to separate wells in gelatinized 96-well plates lacking feeder cells, using a stereoscope and pulled Pasteur pipettes. The frequency of production of G418-resistant colonies is quite low, with each electroporation of 7 X 106 cells producing fewer than 50 colonies surviving postselection, most likely due to the poly(A)- design of this construct. After several days of growth, clones in individual wells are split to two plates, with one plate grown to confluence for the isolation of genomic D N A while the second plate is prepared for freezing. The plate to be frozen is prepared by treating each well with 30/zl of trypsin solution followed by the addition of an equal volume of 2x freezing medium, and then placing the plate in a Styrofoam container to cool slowly to - 7 0 °. Analysis of Targeted Embryonic Stem Cells The initial identification of homologous recombination between the targeting DNA and the endogenous D i d gene is performed by Southern 19M. Hooper, K. Hardy, A. Handyside, S. Hunter, and M. Monk, Nature (London) 326,

292 (1987). 20 R. Ramirez-Solis,A.

C. Davis, and A. Bradley,Methods Enzymol. 252, 855 (1995).

[421

TARG~TIN~E3 or PDC/KDC/BCKDC

471

blot analysis. The genomic DNA is isolated from colonies grown in the 96well plates. Ceils are washed twice with PBS and then 50/xl of genomic DNA extraction buffer is added per well. The plates are then sealed and incubated at 60° overnight with gentle rocking. Genomic DNA is recovered by adding 100/zl of DNA precipitation solution per well and then spinning the pilate at 2500 rpm for 10 min. After careful decanting of the solution, the wells are washed twice with 70% (v/v) ethanol and allowed to air dry. The samples are resuspended in 35/zl of restriction digestion buffer (supplied by the manufacturer) containing the SacI isoschizomer Ecl136II (New England BioLabs, Beverly, MA), and incubated at 37° overnight. To identify which G418-resistant clones have undergone homologous recombination events, the digested DNA from each clone is fractionated on a 0.8% (w/v) agarose gel in TAE buffer, transferred to Zeta-Probe membranes (Bio-Rad), and probed with a HindlII-SalI fragment from DNA upstream of the region contained within the targeting vector (Fig. 1, probe A). After hybridization in Church buffer, the membranes are washed with 1 x SSC and 0.1% (w/v) SDS at room temperature for three 20-min periods and then with 0.25x SSC and 0.1% (w/v) SDS at 65° for 20 min. Genomic Southern blot analysis with Ec1136II digestion yields an 11kb wild-type allele and a 9-kb targeted allele (Fig. 1). For further confirmation of the site-specific integration of the targeting DNA at the Dld locus, genomic DNA from putative targeted ES ceils is digested by several additional restriction enzymes (ApaI plus KpnI, EcoNI, or BstXI) and analyzed by Southern blot, using a 1-kb HindlII fragment that hybridizes to sequences immediately upstream of the EcoRI site (Fig. 1, probe B). 21 The presence of predicted wild-type and targeted alleles identifies heterozygous ES cell lines. Also, none of these lines demonstrates additional, unanticipated hybridizing bands, indicating that there are no additional insertions of the targeting construct into the genome. In addition to Southern blot analysis, the targeted ES cell lines are also analyzed for E3 enzymatic activity.22 Cells are washed in PBS twice, resuspended in 20 mM potassium phosphate (pH 7.5) containing 0.5% (v/v) Triton X-100, and then subjected to three freeze-thaw cycles. After centrifugation, aliquots of the supernatants are added to the reaction cocktail [1100mM potassium phosphate (pH 8.0), 1.5 mM EDTA, 3 mM dihydrolipoamide, 3 mM NAD+], to a final volume of 1 ml. The production of NADH is measured spectrophotometrically at a wavelength of 340 nm over 21 M. T. Johnson, H.-S. Yang, and M. S, Patel, Proc. NatL Aca& Sci. U.~A. 94, 14512 (1997). M. S. Patel, N. N. Vettakkorumakankav, and T. C. Liu, Methods EnzymoL 252, 186 (1995).

472

DETECTION AND CONSEQUENCES OF GENETIC DEFECTS

[42]

several minutes at 37 °. An approximately 50% reduction in E3 activity is observed in targeted ES cells. ES cells that are confirmed as heterozygous by molecular and biochemical analysis are karyotyped by preparing standard chromosome spreads and staining with Giemsa. Establishment of Chimeric Mice The targeted ( D i d +/-) ES cell lines with good karyotypes (>85% diploid) are microinjected into C57BL/6 recipient blastocysts and transferred to the uteri of CH3B16 pseudopregnant females (performed by J. Duffy at the University of Cincinnati, Cincinnati, OH). The offspring are judged to be chimeric on the basis of coat color. The recipient blastocysts are derived from a strain that is homozygous for the alleles, aCP, which results in a black coat color whereas the ES cells are derived from the 129/Ola strain that is homozygous AWcChp,producing a cream-colored coat. Contribution of ES cells in the chimeras can be visually assessed by the presence of areas of agouti fur, indicating contribution from the ES cells to the epithelium, with areas of a cream color indicating a region of higher ES cell contribution. These overtly chimeric animals are initially bred to the outbred Black Swiss line so that ES cell-derived progeny are identified by the presence of agouti offspring as a result of transmission of the dominant agouti allele. Identification of Genotype of Disrupted D i d Gene The offspring with coat color are genotyped by Southern blot or polymerase chain reaction (PCR) of their genomic DNA shortly before weaning. The genomic D N A from mouse tail tissue is isolated with a genomic extraction kit (GeneMate, ISC BioExpress, Kaysville, UT) according to the manufacturer instructions. The genomic D N A is analyzed by Southern blot as previously described or by PCR, using a trio of primers that simultaneously amplify regions of the wild-type and targeted alleles. The amplification conditions consist of 35 cycles of 95 ° for 1 min, 55 ° for 1 rain, and 72 ° for 1 min. The wild-type allele is amplified as a 0.9-kb fragment from the wildtype exon 9 to exon 10 by 5' common primer (5'-GGTGGAATFGGAATTGACAT-3') and 3' wild-type specific primer (5'-TTATFGACTGGAATrCTACCTITGGGATCT-3'). Under these conditions, the targeted allele does not amplify because the insertion of n e o disrupts the annealing of 3' wild-type specific primer. The targeted allele is amplified as a 1.1-kb fragment from the wild-type exon 9 to n e o by 5' common primer and 3' targeted allele specific primer (5'-ACCCCCTCTCCCCTCCTITrG-3'). More recently, the founder chimeras have been bred with the 129/J and 129/Oia inbred strains to maintain the mutant allele in a homogenous genetic background.

[42]

TARGETINGE3 OF PDC/KDC/BCKDC

473

Analysis of D/d +/- Animals In analyzing the agouti progeny, the presence of a 1:1 ratio of there are no dominant effects on viability from the introduced mutation. The D i d ÷I- progeny from the chimera/Black Swiss mating are examined for any obvious phenotypic effect of carrying a mutant E3 allele by performing gross anatomic surveys and following growth curves for D i d ÷/- and D i d ÷I÷ littermates. As may be expected from the human data, the D i d ÷I- littermates are phenotypically normal. Liver samples from the D i d ÷I- and D i d ÷j÷ littermates are assayed for E3 activity as described above. To assess the effect of the heterozygous state on four of the E3requiring multienzyme complexes, liver samples from both D i d ÷/- and D i d ÷/+ littermates are assayed for pyruvate dehydrogenase complex (PDC), ot-ketoglutarate dehydrogenase complex (KDC), BCKDC, and glycine synthase (GS). 23-z6 These assays follow the decarboxylation of specific substrates labeled at the C-1 position. The CO: produced is trapped in hyamine hydroxide and counted by a liquid scintillator. For determination of PDC, KDC, and BCKDC activities, crude liver homogenates are used. To measure total PDC and BCKDC activities, the samples are activated by treatment with phospho-E1 phosphatase and A protein phosphatase (New England BioLabs), respectively, in the presence of dichloroacetate and otchloroisocaproic acid, respectively, for inhibition of kinase activities. For measurement of GS activity, instead of liver homogenates, liver mitochondrial extracts are used. Liver samples are minced and homogenized in a buffer (70 mM sucrose, 220 mM mannitol, 2 mM HEPES and EDTA, pH 7.4). After six or seven passes by a motor-driven Teflon homogenizer, the unbroken cells are removed by centrifugation at 650g for 10 rain at 4°. The mitochondria are pelleted by centrifugation at 12,000g, washed twice with PBS, and resuspended in 20 mM Tris-HCl (pH 8.0), followed by three freeze-thaw cycles. The measurement of GS activity is determined according to a previously described protocol. 26 As a mitochondrial marker, citrate synthase activity is measured by the production of free CoASH, which can be measured spectrophotometrically by assaying the production of free CoASH after treatment with dithionitrobenzoic acid.z7 The D i d ÷/- animals have approximately 50% of wild-type E3 activity as well as four E3-requiring complex activities (Table I), suggesting that the disruption results in a loss of function. Northern blot analysis showed D l d ÷I÷ : D i d ÷I- indicates that

23D. S. Kerr, S. A. Berry, M. M. Lusk, L. L. Ho, and M. S. Patel, Pediatr. Res. 24, 95 (1988). 24D. T. Chuang,C. W. Hu, and M, S. Patel, Biochem. J. 214, 177 (1983). 25G. W. Goodwim,B. Zhang,R. Paxton,and R, A. Harris,Methods Enzymol. 166,189 (1988). 26H. Kochi,K.Hayasaka,K. Hirraga,and G. Kikuchi,Arch.Biochem. Biophys. 198, 589 (1979). 27p. A. Srere, Methods Enzymol. 7, 3 (1969).

474

[42]

DETECTION AND CONSEQUENCES OF GENETIC DEFECTS TABLE I COMPARISON OF" E3 AND E3-REQUIRING COMPLEXES IN D l d +/+ AND D l d +/- MICEa Enzymatic activity (mean _+ SD) b

Animals

E3

PDC

KDC

BCKDC

GS

CS c

D i d +/+ D i d +/%d

21.0 ___ 2.9 12.4 _+ 1.1 59

3.1 --- 0.3 1.5 + 0.1 48

5.1 --+ 0.4 2.7 + 0.4 53

0.79 -+ 0.05 0.43 _+ 0.04 54

0.87 _+ 0.06 0.42 + 0.10 48

0.17 _+ 0.01 0.17 -+ 0.01 100

" n = 4 except for GS, where n = 3. b Milliunits per milligram of total protein, except for GS, where entries represent milliunits per milligram of total mitochondrial protein. c CS, Citrate synthase, a mitochondrial protein as control. a Percentage (%) of enzymatic activity of control ( D i d +/+) animals.

that the Did +1- animals have approximately 50% of wild-type levels of Did mRNA in liver and kidney samples. 21 Identification of the Did-~- Embryos The litters from intercrosses between heterozygotes are found to be approximately 25% smaller. Genotypic analysis of progeny reveals the presence of Dld +/+ and Did +/- animals in the expected Mendelian ratio (66 : 130), but there are no Did -/- animals. The absence of Did -/- mutants indicates that a recessive prenatal lethal allele has been created. To determine the time of death, embryos from various stages of postimplantation development are dissected from the decidua and genotyped by PCR. Embryos are incubated with an embryo lysis buffer for at least 2 hr at 50°. For larger embryos, 7.5 days postcoitum (dpc) and later, embryos are genotyped with a trio of primers that allow for simultaneous amplification of both the targeted and wild-type alleles, using a common primer annealing to a sequence within intron 9 ( 5 ' - C A C T A A G C T C C A T C T - F C A G C C A T G A G 3'), a wild-type allele-specific primer annealing to a sequence within intron 10 ( 5 ' - G G T C T G T F I T F A T C T T T A G A G A G A G C C A A A A A - 3 ' ) , and targeted allele-specific primer annealing to a sequence within the/3-actin promoter of the neomycin marker (5'-CCTCCGCCCTrGTGGACA CT3'). The thermocycling parameters consist of 35 cycles of 95° for 1 min, 55° for 1 rain, and 72° for 1 min. In the case of smaller preimplantation embryos, additional primers are synthesized that are internal to the original set of primers to allow for nested PCR. The second set of primers includes a primer from exon 9 (5'-GGTGGAATTGGAATTGACATGGAGAT-3'), a wild-type allele primer ( 5 ' - G G T C T G T I T I T A T C T T r A G A G A G A G C CAAAAA-3'), and a mutant allele primer (5'-ACCCCCTCTCCCCT-

[42]

TARGETINGE3 OF PDC/KDC/BCKDC

475

CCTITrG-3'). The final amplification product for the targeted allele is 400 bp, whereas the wild-type allele product is 455 bp. For nested PCR, a 5-/zl sample from the first 50-/zl reaction is added to the second amplification cocktail and the sample is cycled with cycling parameters as described above. At the blastocyst stage, corresponding to 3.5 dpc, all embryos are similar in gross appearance, with the expected Mendelian distribution of genotypic classes: 16 : 35 : 14 for D l d +/÷ : D l d ÷/- : D i d -/-. At 7.5 dpc, which corresponds to several days after uterine implantation, two types of embryos are observed. Approximately 75% of the embryos (56 of 81) appear to be normal in size and morphology at pre- to early primitive streak stages and genotype as D l d ÷/÷ or D l d ÷/- class. 28The remaining 25% of embryos are much smaller and resemble normal 6.5 dpc egg cylinders, indicating that they are delayed in development. The genotypes of these embryos indicate that the majority of these embryos are D l d -/- class, suggesting that D l d -/- embryos survive at least this stage. One day later, at 8.5 dpc, the normally developing embryos have reached the head-fold stage and are all in the D l d ÷/÷ or D l d ÷/- genotypic class (11:18), whereas all the embryos genotyped as D i d -/- are abnormal in morphology, still appearing in size to be 6.5 dpc embryos. One day later, at 9.5 dpc, 42 decidua are collected, of which 8 contain only resorption sites with insufficient embryonic material for genotypic analysis. The rest of the embryos are genotyped as D l d ÷/÷ or D l d ÷/-, with a ratio of 11 : 23. Analysis of D/d-/- Embryos I7o assess the D i d -/- embryonic phenotype in more detail, histologic analysis is preformed. Decidua are collected at several time points around the established time of embryonic death and fixed overnight in 4% (w/v) paraformaldehyde in PBS. The following day, the decidua are dehydrated through increasing concentrations of ethanol and embedded in paraffin. A series of 7-/zm sections of the decidua are stained with hematoxylin-eosin. At 6.5 dpc, all embryos appear to be normal egg cylinders, suggesting that the D l d -/- class is not significantly delayed in growth or differentiation at this stage. One day later, at 7.5 dpc, the normal embryos have initiated gastrulation with the formation of the three germ layers: mesoderm, endoderm, and ectoderm. Approximately one-quarter of the embryos are presumed to be developmentally delayed by approximately 1 day. Further analysis of these abnormal embryos at 7.5 and 8.5 dpc reveals the presence of mesoderm in most embryos, suggesting that the homozygous mutant embryos initiate gastrulation and then cannot progress further in development. 28j. R. Cloughand D. G. Whittingham,J. Embryol. Exp. Morphol. 74, 133 (1983).

476

DETECTION AND CONSEQUENCES OF GENETIC DEFECTS

[42]

The D i d null phenotype provides direct evidence of the importance of oxidative metabolism during the early postimplantation period. Because of the lack of PDC and KDC enzymatic activities, the D i d -/- embryos would be unable to metabolize glucose oxidatively. Instead, embryos must rely exclusively on glycolysis, which converts pyruvate to lactate. An in vitro study using embryos in the presence of radiolabeled glucose demonstrated that virtually all glucose is converted to lactate at the egg cylinder period. 29 At 6.5 dpc, both D i d +/+ and D i d -/- embryos are similar in size, mophology, and phenotype, suggesting that anaerobic glucose oxidation may supply sufficient energy for this stage. However, at 7.5 dpc gastrulation begins, a period of rapid growth and differentiation that imposes a significant energy demand on the embryo. One plausible explanation for the cessation of development of the homozygous mutants at this stage is that glycolysis does not generate significant energy for D i d -/- embryos at this period. The generation of these animals has provided some of the first in vivo data indicating the importance of oxidative metabolism in early postimplantation embryogenesis. Prospect This chapter has described the techniques involved in generating and analyzing mice that harbor a null mutation in the D i d gene. The existence of these mice now opens the possibility for further studies to determine if heterozygotes manifest with metabolic differences compared with their wild-type littermates. It may be possible to subject these animals to extreme environmental conditions such as endurance exercise or diets enriched in branched-chain amino acids to bring out metabolic differences and in turn gain some insight into the roles of these metabolic pathways. Also, with the existence of an E3 null allele, it would be possible to introduce less severe mutations into this background to mimic more closely the phenotypes seen in humans to serve as a model for testing various therapeutic interventions. Acknowledgments We thank Dr. Clemencia Colmenares of the Cleveland Clinic for the El4.1 ES cell line, Dr. John Duffy of the University of Cincinnati for performing the blastocyst injections, and Dr. Terry Magnuson of Case Western Reserve University for helpful discussions and guidence. Mark Johnson was the recipient of a predoctoral fellowship supported by Metabolism Training Grant AM07319. This work was supported by U.S. Public Health Service Grant DK 42885. The first two authors (MTJ and HSY) made equal contributions to this article. 29 K. M. Downs and T. Davis, Development I l K 1255 (1993).