Comparative genomics suggests loss of keratin K24 in three evolutionary lineages of mammals

Abstract

Keratins are the main cytoskeletal proteins of epithelial cells and changes in the expression of keratins have contributed to the evolutionary adaptation of epithelia to different environments. Keratin K24 was proposed to be a differentiation marker of epidermal keratinocytes but the significance of K24 expression in the epidermis versus other tissues has remained elusive. Here, we show by RT-PCR, western blot, and immunofluorescence analyses that K24 is highly expressed in the epithelium of the cornea whereas its expression levels are significantly lower in other stratified epithelia including in the epidermis. To investigate the evolutionary history of K24, we screened the genome sequences of vertebrates for orthologs of the human KRT24 gene. The results of this comparative genomics study suggested that KRT24 originated in a common ancestor of amniotes and that it was lost independently in three clades of mammals, i.e. camels, cetaceans, and a subclade of pinnipeds comprising eared seals and the walrus. Together, the results of this study identify K24 as component of the cytoskeleton in the human corneal epithelium and reveal previously unknown differences of keratin gene content among mammalian species.

Figure 1

Loci and phylogenetic relationships human type I keratin genes. The schematic shows the loci of KRT24 and other human type I keratin genes and the phylogenetic relationships of the genes. Branches without significant bootstrap support are collapsed. Arrows indicate the genes with arrow tips pointing in the direction of gene transcription. The expression pattern of each keratin in human tissues is indicated above the genes. With the exception of KRT18 (chromosome 12q13), all type I keratin genes are clustered on chromosome 17q21. KRTAP genes, located between KRT40 and KRT33A, are not shown. Note that KRT222 (indicated by an arrow drawn with a broken line) encodes a keratin-like protein of unknown function. IRS, inner root sheath; ORS, outer root sheath.

Figure 2

KRT24 mRNA levels are high in the cornea. (a) KRT24 mRNA levels in human tissues and cell types were extracted from the Genevestigator database. Expression values in arbitrary units (a.u.) according to Genevestigator are shown. Bars indicate means and error bars show standard deviations. The number (n) of samples per tissue or cell type is indicated below the bars. P-values were calculated with the two-sided t-test. (b–e) Quantitative RT-PCR analysis of mRNAs of selected keratin genes in human cornea (n = 8) and epidermis (n = 5). Expression levels of KRT24 (b), KRT3 (c), KRT12 (d), and KRT10 (e) were calculated relative to that of GAPDH and normalized to the highest value (defined as 10 arbitrary units, a.u.) among the samples investigated. Bars indicate means and error bars show standard deviations. P-values (Mann-Whitney U Test) are shown above the graphs.

Figure 3

Immunodetection of K24 protein in the human cornea. (a) Western blot analysis of K24 and K14 in human conjunctiva, cornea and epidermis. The total protein blotted onto the membrane was visualized by Ponceau staining prior to incubation with blocking buffer and probing with antibodies. Results are representative for n ≥ 3 samples of each tissue. Positions of size markers are shown on the right. kDa, kilo-Dalton. (b) Immunofluorescence labeling of K24, K12, K10 (red) and K14 (green) in human cornea and skin. DNA was labeled with Hoechst 33258 dye (blue). The junctions between epithelia and mesenchymal tissues are indicated by broken lines. Results are representative for at least 3 samples of each tissue. Scale bars: 100 µm.

Figure 4

Evolution of the KRT24 gene locus. A model for the evolutionary history of KRT24 and neighboring keratin genes was developed from the comparison of corresponding gene loci in tetrapods. Genes are indicated by arrows that point from the 5′-end to the 3′-end of the coding sequence. A timed phylogenetic tree is depicted besides the gene loci. Gene gain and loss events were inferred by the principle of maximum parsimony from the species distribution of the genes. The presence or absence of KRT24, KRT223, and KRT224 genes in ancestors of modern species is indicated by lines of the same color as the corresponding gene in the schematic on the right. The scale bar below the phylogenetic tree indicates a period of 100 million years (MY). Note that KRT223 and KRT224 are pseudogenes in the human genome and that orthology relationships could not be faithfully resolved by molecular phylogenetics for Krt24-like genes of several species (*). Two copies of the Krt10 gene are present in marsupials but only one copy is shown in this schematic (**). IRS Krts, inner root sheath keratin genes (KRT25-28).

Figure 5

Loss of Krt24 in different clades of mammals. The presence (+) or absence (−) of functional Krt24 genes was mapped onto a phylogenetic tree of mammals. Only Krt24-deficient species and close relatives are shown.

Figure 6

Evolution of corneal epithelial keratins in mammals. The presence (+) or absence (−) of functional orthologs of human corneal epithelial keratins K5, K14, K3, K12, and K24 was mapped onto a phylogenetic tree of mammals. Gene loss events were inferred from the distribution of functional genes and known relationships of species. Putative keratin heterodimers in the cornea are listed for each species. The scale bar beneath the timed phylogenetic tree indicates a period of 20 million years (MY).