Reevaluation of the Role of Extracellular Signal-Regulated Kinase 3 in Perinatal Survival and Postnatal Growth Using New Genetically Engineered Mouse Models

Abstract

The physiological functions of the atypical mitogen-activated protein kinase extracellular signal-regulated kinase 3 (ERK3) remain poorly characterized. Previous analysis of mice with a targeted insertion of the lacZ reporter in the Mapk6 locus (Mapk6^lacZ ) showed that inactivation of ERK3 in Mapk6^lacZ mice leads to perinatal lethality associated with intrauterine growth restriction, defective lung maturation, and neuromuscular anomalies. To further explore the role of ERK3 in physiology and disease, we generated novel mouse models expressing a catalytically inactive (Mapk6^KD ) or conditional (Mapk6^Δ ) allele of ERK3. Surprisingly, we found that mice devoid of ERK3 kinase activity or expression survive the perinatal period without any observable lung or neuromuscular phenotype. ERK3 mutant mice reached adulthood, were fertile, and showed no apparent health problem. However, analysis of growth curves revealed that ERK3 kinase activity is necessary for optimal postnatal growth. To gain insight into the genetic basis underlying the discrepancy in phenotypes of different Mapk6 mutant mouse models, we analyzed the regulation of genes flanking the Mapk6 locus by quantitative PCR. We found that the expression of several Mapk6 neighboring genes is deregulated in Mapk6^lacZ mice but not in Mapk6^KD or Mapk6^Δ mutant mice. Our genetic analysis suggests that off-target effects of the targeting construct on local gene expression are responsible for the perinatal lethality phenotype of Mapk6^lacZ mutant mice.

Keywords: ERK3; MAP kinases; mouse models; protein kinases; signal transduction.

FIG 1

Generation of mice expressing kinase-dead ERK3. (A) Schematic representation of the gene targeting strategy used to generate a catalytically inactive knock-in allele of Mapk6. The targeting vector harbors a neomycin resistance gene (neo^r) and DNA mutations that replace lysines 49 and 50 with alanine residues in exon 2 of the Mapk6 coding sequence. Exons 2 to 6 and restriction sites are shown (E, EcoRI; K, KpnI; S, ScaI). Open boxes correspond to untranslated regions (UTRs), and closed boxes indicate coding regions. Black triangles correspond to loxP sites. (B) Nucleotide and amino acid sequences of codons 49 and 50 in the kinase domain of wild-type and kinase-dead ERK3 protein. (C, upper) Southern blot analysis of EcoRI-digested genomic DNA with probe A showing two correctly targeted ES clones with wild-type (9.1 kb) and mutant (6.8 kb) alleles. (Lower) Southern blot analysis of ScaI-digested genomic DNA with probe B showing the expected 9.5-kb recombinant band. (D) PCR analysis of a litter from a Mapk6^KD/+ heterozygous cross at weaning. (E) Immunoblot analysis of ERK3 protein expression in lysates prepared from brain and skeletal muscle of wild-type and ERK3^KD/KD adult mice. (F) Chromatograms of genomic DNA sequences from Mapk6^+/+ and Mapk6^KD/KD mice confirming the mutation of lysines 49 and 50 to alanines in exon 2 of the Mapk6 gene.

FIG 2

Generation of mice expressing a conditional Mapk6 allele. (A) Schematic representation of the Mapk6 locus, targeting vector, recombinant conditional allele, and conditional deleted allele. The targeting vector carries a neomycin resistance cassette (neo^r) flanked by two FRT sites, and exon 3 is flanked by two loxP sites. Exons 2 to 5 and restriction sites are shown (N, NheI; N, NotI; S, SfiI). Open boxes correspond to UTRs, and closed boxes indicate coding regions. White and black triangles correspond to loxP and FRT sites, respectively. (B) Southern blot analysis of SfiI-digested genomic DNA showing two correctly targeted ES clones. Wild-type (16 kb) and targeted mutant (12 kb) alleles are indicated. (C) Immunoblot analysis of ERK3 protein expression in lysates prepared from wild-type (+/+), heterozygous Mapk6^Δ/+ (Δ/+), and Mapk6^Δ/Δ (Δ/Δ) E12.5 embryos. (D) Immunoblot analysis of ERK3 protein expression in lysates prepared from different tissues of Mapk6^Δ/+ and Mapk6^Δ/Δ adult mice.

FIG 3

Phenotypic analysis of Mapk6^KD/KD and Mapk6^Δ/Δ E18.5 embryos. (A to C) Body weight distribution of E18.5 embryos from Mapk6^lacZ/+ (A), Mapk6 ^KD/+ (B), and Mapk6^Δ/+ intercrosses. *, P < 0.05; ***, P < 0.001 (unpaired Student's t test). (D, left) Representative photographs of H/E staining of lung sections from E18.5 embryos of the indicated wild-type or mutant Mapk6 genotype. Scale bar, 250 µm. (Right) Saccular space quantification of E18.5 lungs from Mapk6 ^KD/+ (n ≥ 6) or Mapk6^Δ/+ (n = 4) intercrosses. Results are expressed as means ± standard errors of the means (SEM). AU, arbitrary units. (E) Representative images of E18.5 embryos of the indicated wild-type or mutant Mapk6 genotype. Arrows point to the wrists and show carpoptosis, which is present exclusively in Mapk6^lacZ/lacZ embryos.

FIG 4

Targeted insertion of the Mapk6^lacZ construct influences the transcription of neighboring genes. (A) Schematic representation of mouse chromosome 9 and neighboring genes of Mapk6 within a 1-Mb interval. (B to D) Relative mRNA expression of 15 genes within a 1-Mb interval around Mapk6 in E12.5 embryos of wild-type or homozygous Mapk6^lacZ/lacZ (B), Mapk6 ^KD/KD (C), and Mapk6^Δ/Δ (D) genotypes. Gene expression was measured by quantitative RT-PCR. Results are expressed as means ± standard deviations (SD). *, P < 0.05 (unpaired Student's t test). (E to G) Relative mRNA expression of 14 developmental genes on mouse chromosome 9 outside 1 Mb from the Mapk6 locus in E12.5 embryos of wild-type or homozygous Mapk6^lacZ/lacZ (E), Mapk6 ^KD/KD (F), and Mapk6^Δ/Δ (G) genotypes. Results are expressed as means ± SD. *, P < 0.05 (unpaired Student's t test). (H) Functional enrichment analysis of the 89 genes located within 5 Mb upstream or downstream of the Mapk6 locus on mouse chromosome 9 using Ingenuity Pathway Analysis software. Enrichment of biological functions is expressed as log₁₀ P value.

FIG 5

Impact of the loss of ERK3 activity or expression on postnatal growth in mice. (A) Growth curves of Mapk6^flox/+, Mapk6^Δ/flox, Mapk6^Δ/+, and Mapk6^Δ/Δ mice (pooled animals, ≥11) over 15 postnatal days. (B) Analysis of Mapk6^+/+, Mapk6^Δ/+, and Mapk6^Δ/Δ mouse (males, n ≥ 8; females, n ≥ 18) body weight at weaning (day 21). (C) Growth curves of Mapk6^+/+, Mapk6^KD/+, and Mapk6^KD/KD mice (pooled animals, n ≥ 12) over 15 postnatal days. (D) Analysis of Mapk6^+/+, Mapk6^KD/+, and Mapk6^KD/KD mouse (males, n ≥ 8; females, n ≥ 9) body weight at weaning. Results are expressed as means ± SEM. *, P < 0.05; **, P < 0.01 (unpaired Student's t test).