Human Molecular Genetics

Production of targeting constructs

We made a replacement vector using a 7.5 kb HindIII subclone (HLMG7.5) of the previously described 129 strain mouse HL genomic clone (7). The subclone contains mouse HL exons 2-5 and ends 1130 bp 5[prime] of exon 2 (Fig. 1).

We replaced the 1130 bp 5[prime]-region by a 797 bp amplified fragment, introducing a NotI site at the 5[prime]-extremity and XhoI and BamHI sites in exon 2. Amplification was performed using oligonucleotides HLM18 (5[prime]-AAGCGGCCGCGTCTGGAGAAATACAAA-3[prime], corresponding to residues 780-797 upstream of the exon 2 acceptor splice junction) and HLM23 (5[prime]-AACCATGGATCCTCTAGACTCGAGAATTCGTGCTCACCTTTGGTTT-3[prime], complementary to nine intronic and eight exonic nucleotides at the intron 1 acceptor splice junction plus cloning XhoI, XbaI, BamHI and NcoI sites). We cloned this fragment into naturally occuring NotI and NcoI sites of the vector, then inserted a bacterial neomycin resistance (Neo) gene oriented in the direction of HL transcription and driven by the herpes simplex thymidine kinase promoter (pMC1neo poly A; Stratagene) into exon 2 at the XhoI and BamHI sites. In the resulting transcript, HL sequences terminate at codon 23 and the neomycin resistance stop codon occurs ~1000 nt downstream. The Neo cassette is flanked by ~0.8 (5[prime]) and 6 kb (3[prime]) of HL gene sequence.

Targeting and culture of embryonal stem cells (ES cells)

Electroporation was performed using 10 µg linearized vector and 10⁷ cGR8 ES cells (13) at 240 V and 500 µF in a 0.4 cm Gene Pulse cuvette (Bio-Rad). Cells were cultured for 24 h in Glasgow minimum essential medium (MEM; Gibco BRL) with 0.1 mM MEM non-essential amino acids, 2 mM l-glutamine, 1 mM sodium pyruvate, 0.2 mM [beta]-mercaptoethanol, 10% fetal calf serum, 50 U/ml penicillin, 50 µg/ml streptomycin and 1000 U/ml leukemia inhibiting factor (Gibco BRL) in gelatinized (0.1%) tissue culture dishes (Nunclon). Selection was performed for 10-12 days in the same medium with 200 µg/ml G418 (Gibco BRL).

HL genotyping by PCR and Southern blotting

Pools of four G418-resistant colonies were screened by PCR. ES cell lysates were digested with 2 µg proteinase K in a volume of 50 µl at 50°C for 90 min. Amplification was performed in 100 µl containing 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl₂, 0.001% gelatin, 0.5% DMSO, 0.2 mM each dNTP and 2.5 U Taq DNA polymerase. The sense primer HLM28 (5[prime]-CCGTTCGCTTCAAATCAGGC-3[prime]), which is outside the targeting construct, corresponds to residues 1034-1054 upstream of exon 2; the antisense primer Neo-1 (5[prime]-ATCCATCTTGTTCAATGGCCGATCC-3[prime]) is complementary to nt 735-759 of the neomycin resistance cDNA in the pMC1neo poly A vector. Forty cycles of amplification were performed as follows: 94°C, 45 s; 60°C, 45 s; 72°C, 90 s. Amplification of a 1.4 kb fragment was diagnostic for the presence of a targeted allele.

For Southern blots, mouse genomic DNA (5 µg) was digested with HindIII or HindIII and XbaI. Following electrophoresis in a 0.8% agarose gel in 1× TAE buffer (14), the samples were transferred to a nylon membrane (Hybond-N; Amersham). Probes used were either an amplified fragment spanning intron 1 residues 799-1130 upstream of the acceptor splice site or the Neo cDNA. Hybridizations were carried out in 4× SSC, 5× Denhardt's solution (1× Denhardt's is Ficoll, bovine serum albumin and polyvinyl pyrrolidone, each at 0.2 mg/ml), 10% (w/v) dextran sulfate, 1% (w/v) SDS and 100 mg/ml sheared salmon sperm DNA. The membranes were washed three times at 65°C: once in 2× SSC, 0.5% SDS for 30 min; once in 1× SSC, 0.5% SDS for 30 min; once in 0.5× SSC, 0.5% SDS for 20 min.

Production and analysis of HL-deficient mice

Targeted ES cell clones were injected into C57BL/6J blastocysts and transferred to pseudopregnant recipient ICR strain females. The targeted HL allele was bred onto C57BL/6J, 129 and CD1 backgrounds. HL+/- heterozygotes were crossed and their offspring were genotyped. All animals studied were anesthetized with methoxyflurane (Janssen Pharmaceutica). Dissections were performed rapidly after decapitation.

Isolation of genomic DNA

Mouse tails (~0.5 cm) or yolk sacs were incubated overnight at 65°C in 400 µl 4 M urea, 100 mM Tris-HCl, pH 8.8, 0.5% N-lauroylsarcosine and 10 mM EDTA. Then 20 µl proteinase K, (2 µg/µl) were added and incubation was continued for 5 h at 65°C. DNA was extracted with phenol, precipited with ethanol and suspended in 10 mM Tris-HCl, pH 8.6, 1 mM EDTA.

Fibroblast lines from HL-deficient mouse embryos

Mouse fibroblast lines were generated as described (15). Briefly, 10.5 d.p.c. embryos were washed in phosphate-buffered saline solution and forced from a 3 ml syringe through an 18 gauge needle. The homogenate was placed in a gelatinized (0.1%) 25 cm² flask and cultured in Dulbecco's modified Eagle's medium (DMEM) containing 15% fetal calf serum (Hyclone). After reaching confluence (`passage 0'), cells were cultured in DMEM containing 0.37% sodium bicarbonate plus 10% fetal calf serum, 50 U/ml penicillin and 50 µg/ml streptomycin.

RT-PCR analysis of HL-deficient mRNA

Transcripts from the endogenous and targeted alleles were detected as follows. Reverse transcription of 10 µg total fibroblast RNA was initiated with HLM20 (5[prime]-ACTTTAGCCGGGGAGAC-3[prime]), complementary to residues 539-556 of the mouse HL cDNA. The reaction was carried out in a 30 µl volume containing 140 ng primer, 500 µM each dNTP, 17 µM dithiothreitol, 200 U Superscript RT RNase H^--free reverse transcriptase (Gibco BRL), 40 U RNasin (Promega) and 1× First Strand Buffer (Gibco BRL) at 42°C for 30 min and then at 95°C for 7 min. Amplification was carried out using the same buffer as described above and 50 µM dNTP with two pairs of primers: 5[prime]-AGTGAAGATGGCGTCAGTGA-3[prime] (nt 7-13, primer 1, Fig. 2a) and 5[prime]-GACTTCCACGATCTTCACCTGCTTT-3[prime] (nt 90-105, primer 5 or primers 1 and 2; Neo-1 above). One microliter of the above cDNA mixture was amplified for 30 cycles as follows: 94°C, 30 s; 62°C, 30 s; 72°C, 90 s.

For detection of transcripts containing 3[prime] sequences from both neomycin and HL, we performed reverse transcription as above except that we used a poly(T) primer. Then we amplified as above using primer 3 (5[prime]-GATTCGCAGCGCATCGCCTT-3[prime], ending 25 nt upstream from the neomycin resistance stop codon; Fig. 2a) and primer 6 (5[prime]-ACCAGGTCCTCAGTAGCCAA-3[prime], HL cDNA residues 845-826) or primer 4 (5[prime]-TGCCAAAGCAGGTGAAGATC-3[prime], HL cDNA residues 86-106) and primer 6.

HL assay

HL was assayed spectrophotometrically using HMG-CoA as substrate (16). One unit of enzyme activity is the amount which produces 1 µmol acetoacetate/min. Protein was assayed as described (17) with bovine serum albumin (Sigma) as the standard.

Microscopy

Conventional techniques were used for sample preparation. For histology, paraffin-embedded sections fixed in formalin (pH 7.0) were stained with hematoxylin/phloxin/saffron. Frozen sections were stained with hematoxylin/eosin or Oil Red-O. For electron microscopy, epon-embedded sections were stained with osmium tetraoxide.