An absence of both lamin B1 and lamin B2 in keratinocytes has no effect on cell proliferation or the development of skin and hair

Abstract

Nuclear lamins are usually classified as A-type (lamins A and C) or B-type (lamins B1 and B2). A-type lamins have been implicated in multiple genetic diseases but are not required for cell growth or development. In contrast, B-type lamins have been considered essential in eukaryotic cells, with crucial roles in DNA replication and in the formation of the mitotic spindle. Knocking down the genes for B-type lamins (LMNB1, LMNB2) in HeLa cells has been reported to cause apoptosis. In the current study, we created conditional knockout alleles for mouse Lmnb1 and Lmnb2, with the goal of testing the hypothesis that B-type lamins are crucial for the growth and viability of mammalian cells in vivo. Using the keratin 14-Cre transgene, we bred mice lacking the expression of both Lmnb1 and Lmnb2 in skin keratinocytes (Lmnb1^Δ/ΔLmnb2^Δ/Δ). Lmnb1 and Lmnb2 transcripts were absent in keratinocytes of Lmnb1^Δ/ΔLmnb2^Δ/Δ mice, and lamin B1 and lamin B2 proteins were undetectable. But despite an absence of B-type lamins in keratinocytes, the skin and hair of Lmnb1^Δ/ΔLmnb2^Δ/Δ mice developed normally and were free of histological abnormalities, even in 2-year-old mice. After an intraperitoneal injection of bromodeoxyuridine (BrdU), similar numbers of BrdU-positive keratinocytes were observed in the skin of wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice. Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes did not exhibit aneuploidy, and their growth rate was normal in culture. These studies challenge the concept that B-type lamins are essential for proliferation and vitality of eukaryotic cells.

INTRODUCTION

The nuclear lamina, an intermediate filament meshwork lying beneath the inner nuclear membrane, is composed mainly of four proteins—lamins A, C, B1 and B2 (1,2). Lamins A and C are alternatively spliced products of a single gene (LMNA) (3,4), whereas lamin B1 and lamin B2 are products of distinct genes, LMNB1 and LMNB2, respectively (5,6). The lamins were initially divided into two groups, A-type (lamins A and C) and B-type (lamins B1 and B2), on the basis of different isoelectric points, but other considerations have strengthened the rationale for this categorization (2,7). For example, the sequences of lamins B1 and B2 are more similar to each other than to lamins A and C (5). Both lamin B1 and lamin B2 contain a farnesyl lipid anchor, whereas lamin C and mature lamin A do not (8). Also, lamins A and C are expressed late in embryonic development, primarily in differentiated cells, whereas lamins B1 and B2 are expressed in all somatic cell types, including undifferentiated cells at the earliest stages of embryonic development (9,10). In addition, hundreds of missense, non-sense and frameshift mutations in LMNA have been linked to a panoply of genetic diseases (including muscular dystrophy, cardiomyopathy, lipodystrophy, neuropathy and progeria), but these classes of mutations have not been reported for LMNB1 or LMNB2 (8).

The A-and B-type lamins have been reported to differ in terms of their importance for the growth and vitality of mammalian cells (11–13). A knockdown of LMNA in HeLa cells had no effects on cell viability, whereas a knockdown of LMNB1 or LMNB2 arrested growth and led to apoptosis (11). These findings seemed consistent with reports that B-type lamins have unique and important functions in the cell nucleus. For example, lamin B1, but not A-type lamins, was reported to localize to areas of DNA replication (14), and depleting a B-type lamin in Xenopus blocked DNA replication (15). In addition, Tsai et al. (16) reported that lamin B1 has a crucial role in the assembly of the mitotic spindle, and that the spindle complex was disrupted by a dominant-negative lamin B1 mutant. In addition, B-type lamins are thought to be important for chromatin organization and gene expression at the nuclear periphery (7,17).

To assess the functional relevance of the B-type lamins, our laboratory began by creating Lmnb1 and Lmnb2 knockout mice (12,13). Both knockout mice died shortly after birth. Lmnb1- and Lmnb2-deficient fibroblasts were able to grow in culture, although the absence of lamin B1 was associated with misshapen cell nuclei. The single-knockout models did not, of course, permit firm conclusions about the importance of the B-type lamins, given that each knockout model retained the other B-type lamin.

In the current study, we took the next step and examined the consequences of inactivating both Lmnb1 and Lmnb2 in the same cell. We generated conditional knockout alleles for Lmnb1 and Lmnb2 and then bred mice lacking both lamin B1 and lamin B2 in skin keratinocytes. We chose skin keratinocytes, rather than some other cell type, for several reasons. First, the skin and hair express large amounts of lamin B1 and lamin B2 (18). Second, on a practical level, the keratin 14-Cre transgene (K14Cre) is extremely effective for inactivating genes in keratinocytes (19,20). Third, the turnover of skin keratinocytes is high (21), and we reasoned that any effect of the B-type lamins on cell proliferation would be readily detectable in those cells. Fourth, the formation of the skin epidermis and hair involves complex developmental programs, and it seemed reasonable that these programs would be disrupted by disturbances in cell growth or gene expression.

RESULTS

Inactivating Lmnb1 and Lmnb2 in keratinocytes using conditional alleles

The targeting vectors for Lmnb1 and Lmnb2 (Figs 1 and 2) were designed to introduce an FRT-flanked βgeo cassette into intron 1, along with loxP sites flanking exon 2 (Figs 1 and 2). After identifying targeted cell lines, the βgeo cassette was excised with Flp recombinase, generating conditional knockout or ‘floxed’ alleles (Lmnb1^fl and Lmnb2^fl) (Figs 1 and 2).

Figure 1.

Generation of keratinocyte-specific Lmnb1 knockout mice. (A) Schematic of the Lmnb1 targeting vector containing a βgeo gene-trapping cassette flanked by FRT sites in intron 1 and a pair of loxP sites flanking exon 2. The gene-trapping cassette in the targeted allele (Lmnb1^βgeo) was removed with Flp recombinase, creating a floxed allele (Lmnb1^fl). With expression of Cre recombinase, the floxed allele is converted into a null allele (Lmnb1^Δ). Open squares represent exons; loxP and FRT sites are depicted as black triangles and black rectangles, respectively. PCR primers (F1, R1, LoxpR2) are indicated with arrows. (B) PCR of genomic DNA from Lmnb1^+/+, Lmnb1^+/βgeo and Lmnb1^fl/+ ES cells with primers F1 and R1. WT (wild-type), 540 bp; Lmnb1^βgeo (KO), 6.8 kb; Lmnb1^fl, 1.06 kb. (C) PCR of genomic DNA from Lmnb1^fl/fl, Lmnb1^+/+ and Lmnb1^fl/+K14Cre mice with primers F1 and LoxpR2. WT, 1.2 kb; Lmnb1^fl (fl), 1.8 kb; Lmnb1^Δ (Δ), 720 bp.

Figure 2.

Generation of keratinocyte-specific Lmnb2 knockout mice. (A) Schematic of the Lmnb2 targeting vector containing a βgeo gene-trapping cassette flanked by FRT sites in intron 1 and a pair of loxP sites flanking exon 2. The gene-trapping cassette in the Lmnb2^βgeo allele was removed with Flp recombinase, creating a floxed allele (Lmnb2^fl). With the expression of Cre recombinase, the floxed allele is converted into a null allele (Lmnb2^Δ). Open squares represent exons; loxP and FRT sites are depicted as black triangles and black rectangles, respectively. PCR primers (F1, R1 and R2) are indicated by arrows. (B) PCR of genomic DNA from Lmnb2^+/+ and Lmnb2^fl/+ ES cells with primers F1 and R1. WT (wild-type), 244 bp; Lmnb2^fl (fl), 692 bp. (C) PCR of genomic DNA from Lmnb2^fl/fl, Lmnb1^fl/+K14Cre and Lmnb2^fl/flK14Cre (Lmnb2^Δ/Δ) mice with primers F1 and R2. WT, 1.1 kb; Lmnb2^fl (fl), 1.53 kb; Lmnb2^Δ (Δ), 826 bp.

We used the K14Cre transgene to delete exon 2 and thereby inactivate Lmnb1 and Lmnb2 in skin keratinocytes. Keratinocyte-specific Lmnb1 (Lmnb1^fl/flK14Cre or Lmnb1^Δ/Δ), Lmnb2 (Lmnb2^fl/flK14Cre or Lmnb2^Δ/Δ) and double-knockout mice (Lmnb1^fl/flLmnb2^fl/flK14Cre or Lmnb1^Δ/ΔLmnb2^Δ/Δ) were born at the expected Mendelian frequency. The mice survived for >2 years and were indistinguishable from wild-type littermates.

Skin keratinocytes from newborn Lmnb1^Δ/ΔLmnb2^Δ/Δ mice contained only trace amounts of Lmnb1 and Lmnb2 transcripts, as judged by qRT-PCR (Fig. 3A and B). There was no effect of the keratinocytes knockouts on gene expression in other tissues (Supplementary Material, Fig. S1). Lamin B1 and lamin B2 proteins were undetectable in extracts from primary Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes (Fig. 3C), and immunofluorescence microscopy revealed a complete absence of lamins B1 and B2 in freshly isolated Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes (Fig. 3D; Supplementary Material, Fig. S2A). Microscopy also showed an absence of lamins B1 and B2 in keratinocytes of the ear skin (Fig. 3E) and tongue (Supplementary Material, Fig. S2B and C) from Lmnb1^Δ/ΔLmnb2^Δ/Δ mice.

Figure 3.

Inactivation of Lmnb1 and Lmnb2 in skin keratinocytes. (A and B) qRT-PCR studies of Lmnb1 (A) and Lmnb2 (B) expression in keratinocytes from Lmnb1^+/+ and Lmnb1^fl/flLmnb2^fl/flK14Cre (Lmnb1^Δ/ΔLmnb2^Δ/Δ) mice (n = 4/genotype). Data were normalized to β2-microglobulin and gene expression was compared with that in the Lmnb1^+/+ mice (which was set at 1.0). (C) Western blots of extracts from Lmnb1^+/+, Lmnb1^Δ/Δ, Lmnb2^Δ/Δ and Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes with antibodies against lamin B1, lamin B2 and lamin A/C. Actin was used as a loading control. (D and E) Immunofluorescence microscopy of primary keratinocytes (D) and ear skin (E) from wild-type (WT) and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice (four mice/group were examined) with antibodies against lamin B1 (green), lamin B2 (red) and keratin 14 (magenta). DNA was visualized with DAPI (blue). Scale bar, 20 µm.

Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes proliferate normally

Bromodeoxyuridine (BrdU) labeling studies revealed that mitotically active skin keratinocytes (i.e. BrdU-positive) were equally abundant in wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice (Fig. 4A and B). Similarly, there were similar numbers of keratinocytes expressing the phosphorylated form of histone H3 (a marker of mitosis) in wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice (Fig. 4C and D). Primary Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes exhibited normal cell growth curves (Supplementary Material, Fig. S3A) and had normal numbers of chromosomes (Supplementary Material, Fig. S3B). Lamin A/C expression was not upregulated, either in Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes (Fig. 3C, Supplementary Material, Fig. S4) or in immortalized Lmnb1^Δ/ΔLmnb2^Δ/Δ fibroblasts (Supplementary Material, Fig. S5).

Figure 4.

Normal proliferation of keratinocytes from Lmnb1^Δ/ΔLmnb2^Δ/Δ mice. (A) BrdU labeling studies in 40-day-old wild-type (WT) and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice. Mice were given an intraperitoneal injection of BrdU and sacrificed 1 h later. Ear skin was examined by immunofluorescence microscopy with antibodies against BrdU (green), lamin B1 (red) and keratin 14 (Krt14; magenta). DNA was visualized with DAPI (blue). epi, epidermis; derm, dermis. Scale bar, 50 µm. (B) Percentage of BrdU-positive keratinocytes in ear skin from wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice. Numbers of BrdU-positive cells (numerator) and total cells (denominator) examined are recorded within each bar (P = 0.72, t-test). (C) Immunofluorescence microscopy studies of the skin of wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice at P1, P4 and P7 with antibodies against phosphohistone H3 (ph-H3; green) and keratin 14 (Krt14; red). DNA was visualized with DAPI (blue). Scale bar, 50 µm. (D) Number of ph-H3-positive keratinocytes/mm of skin from wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice at P1, P4 and P7. A total of 10 mm of skin was analyzed for each genotype. No differences were observed.

Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes have an increased frequency of misshapen cell nuclei

To determine whether the absence of B-type lamins affects nuclear shape, we assessed the frequency of misshapen cell nuclei in primary keratinocytes from wild-type, Lmnb1^Δ/Δ, Lmnb2^Δ/Δ and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice (n = 3 cell lines/genotype; 300–900 cells/genotype). Keratincytes from wild-type mice exhibited a relatively high frequency of misshapen cell nuclei (blebs and folds) when compared with our earlier studies with cultured fibroblasts (22,23). When we examined the mutant keratinocytes, we found more misshapen nuclei (mainly nuclear blebs) in Lmnb1^Δ/ΔLmnb2^Δ/Δ than in wild-type keratinocytes (P < 0.0005), and there was a trend toward more misshapen nuclei in Lmnb1^Δ/Δ and Lmnb2^Δ/Δ keratinocytes (Fig. 5A and B). These differences were consistent in keratinocytes prepared from different mice (Fig. 5B). While nuclear blebs were relatively common in cultured Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes (Fig. 5A and B), we never observed blebs in the skin keratinocytes of Lmnb1^Δ/ΔLmnb2^Δ/Δ mice by immunofluorescence microscopy (with a lamin A/C antibody) or by routine histology (Supplementary Material, Figs S2, S6).

Figure 5.

Increased frequency of nuclear blebbing in Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes. (A) Immunohistochemical staining of keratinocytes from wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice with antibodies against Lap2 (green) and keratin 14 (Krt14; magenta). DNA was visualized with DAPI (blue). (B) Frequency of misshapen nuclei in wild-type, Lmnb1^Δ/Δ, Lmnb2^Δ/Δ and Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes (n = 4/genotype tested). Bars indicate the frequency of misshapen nuclei; the number of cells with nuclear blebs and the total number of cells examined are provided within each bar. Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes had more nuclear blebs, as judged by a χ² test. Significant differences between wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes were also observed by a t-test, by comparing the percentages of misshapen nuclei in keratinocytes prepared from different mice (P < 0.001). Slides were scored by two reviewers blinded to genotype. Circles represent the frequency of misshapen nuclei of each individual mouse.

Normal skin and hair in Lmnb1^Δ/ΔLmnb2^Δ/Δ mice

The skin and hair appeared entirely normal in Lmnb1^Δ/ΔLmnb2^Δ/Δ mice up to 24 months of age (through all stages ofthe hair cycle). Also, we observed no histological abnormalities in the skin or hair of Lmnb1^Δ/Δ, Lmnb2^Δ/Δ and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice, and skin thickness was similar (Supplementary Material, Fig. S6). Finally, electron microscopy of skin keratinocytes from wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice (n > 60 micrographs/genotype) uncovered no abnormalities (Fig. 6). Specifically, the nuclear membrane, nuclear pores and heterochromatin were indistinguishable in wild-type and Lmnb1^Δ/ΔLmnb2^Δ/Δ keratinocytes (Fig. 6).

Figure 6.

Electron micrographs of keratinocyte nuclei in the skin from Lmnb1^+/+, Lmnb1^Δ/Δ, Lmnb2^Δ/Δ and Lmnb1^Δ/ΔLmnb2^Δ/Δ mice, showing heterochromatin (HC), euchromatin (EC), nuclear pores (NP) and nuclear envelope (NE). Magnification in (A)–(D): ×22, 700. Scale bar, 1.0 µm. Magnification in (E)–(H): ×74, 200. Scale bar, 0.5 µm. (I) Percentage of heterochromatin in the nuclei of Lmnb1^+/+, Lmnb1^Δ/Δ, Lmnb2^Δ/Δ and Lmnb1^Δ/ΔLmnb2^Δ/Δ skin keratinocytes. We examined, by electron microscopy, a total of 26 WT nuclei, 28 Lmnb1^Δ/Δ nuclei, 31 Lmnb2^Δ/Δ nuclei and 35 Lmnb1^Δ/ΔLmnb2^Δ/Δ nuclei.

DISCUSSION

The B-type lamins, lamins B1 and B2, are fundamental building blocks of the nuclear lamina and are thought, based on cell biology studies, to have unique and important roles in cell growth and mitosis, DNA replication, chromatin organization and gene expression (2,7,24). Genetic studies support this view. When Lmnb1 and Lmnb2 were knocked down in cultured cells with siRNAs (11), the cells stopped proliferating and died. Also, Lmnb1 and Lmnb2 knockout mice die shortly after birth (12,13). In addition, the absence—at least thus far—of loss-of-function mutations in human LMNB1 and LMNB2 could be viewed as being consistent with an essential function for the B-type lamins in eukaryotic cells. In the current study, we explored the functional importance of lamins B1 and B2 in adult mouse tissues. We generated conditional knockout alleles for Lmnb1 and Lmnb2 and went on to create keratinocyte-specific knockout mice, reasoning that any unique and vital function for B-type lamins in the nucleus would likely manifest itself in that rapidly proliferating cell type. The knockout of Lmnb1 and Lmnb2 in keratinocytes was efficient; transcript levels for those genes were virtually abolished, and lamin B1 and lamin B2 proteins were undetectable in skin keratinocytes. Remarkably, there were no apparent consequences from the loss of B-type lamins. The skin and hair were normal, even at 2 years of age. The chromosome number in freshly isolated keratinocytes was normal, and the proliferation of keratinocytes in Lmnb1^Δ/ΔLmnb2^Δ/Δ mice was no different than in wild-type mice. By electron microscopy, there were no abnormalities in the nuclear envelope or in the distribution of heterochromatin within the cell nucleus. These observations suggest that the B-type lamins are dispensable in keratinocytes and argue against an essential role for the B-type lamins in DNA replication or cell proliferation.

Cultured Lmnb1^Δ/ΔLmnb2^Δ/Δ skin keratinocytes had significantly more misshapen cell nuclei than wild-type keratinocytes, but misshapen nuclei were never observed in the skin keratinocytes from Lmnb1^Δ/ΔLmnb2^Δ/Δ mice (either by routine histology or immunohistochemistry). In the case of some mutations—although perhaps not all (25)—we suspect that cultured cells are particularly susceptible to the development of nuclear blebs, perhaps because the pancake-like shape of cells grown on plastic renders nuclei more susceptible to shape abnormalities (26). Previously, we examined nuclear shape abnormalities in a Lmna knock-in model (Lmna^LAO/LAO) that synthesizes exclusively mature lamin A (bypassing prelamin A synthesis and processing) (26). By immunofluorescence microscopy, there were many misshapen nuclei in cultured Lmna^LAO/LAO fibroblasts but none was found in tissues of Lmna^LAO/LAO mice.

Our findings do not mean that the B-type lamins are devoid of purpose. In fact, we know that this is not the case—simply because Lmnb1 and Lmnb2 knockout mice cannot survive after birth. However, we can conclude that lamins B1 and B2 are dispensable for skin development and proliferation (and that A-type lamins appear to be sufficient for the integrity of the cell nucleus in vivo). Ultimately, we suspect that the B-type lamins will turn out to be dispensable in other tissues as well; indeed, we already know that mice lacking both Lmnb1 and Lmnb2 in hepatocytes (Lmnb1^fl/flLmnb2^fl/fl mice harboring an albumin–Cre transgene) are free of hepatocellular disease (S.H.Y., L.G.F., S.G.Y., unpublished data).

If B-type lamins are not uniquely important in proliferating cells such as keratinocytes, then what is their purpose and why have they been highly conserved through evolution? A likely clue, we believe, has emerged from studies of Lmnb2-deficient mice (13). Lamin B2 deficiency causes defective neuronal migration in the developing brain, leading to disordered layering of neurons in the cerebral cortex. Migration of cortical neurons during development is a saltatory process that involves repeated cycles of pulling the cell nucleus forward into the leading edge of the cell (13). The ‘tugging’ on the cell nucleus, applied by cytoplasmic dynein motors, exerts deformational forces on the cell nucleus and likely places a premium on the scaffolding provided by the nuclear lamina. The B-type lamins may be well suited for holding the nuclear envelope together, given that these lamins are anchored to the inner nuclear membrane by a farnesyl lipid. In any case, B-type lamins are clearly essential for neuronal migration in the brain but not for the development of skin or the proliferation of keratinocytes.

The human genetics of A- and B-type lamins are markedly different. There are hundreds of missense, frameshift, non-sense and splicing mutations in LMNA that elicit a panoply of human diseases (8). But for lamins B1 and B2, there are no diseases resulting from these classes of mutations, and the only bona fide disease association is autosomal-dominant leukodystrophy with a LMNB2 gene duplication (27). This discrepancy in disease associations remains somewhat enigmatic, but mouse models are starting to provide a framework of understanding. In light of the current studies, it would be difficult to claim that the absence of human mutations is due to a requirement for the B-type lamins in the proliferation and survival of eukaryotic cells. The developmental biology studies by Coffinier et al. (13) also provide important insights; those studies suggested that mutations in LMNB1 and LMNB2 might ultimately be found in association with neurodevelopmental diseases rather than with the usual spectrum of LMNA diseases (e.g. lipodystrophy, progeria, cardiomyopathy, muscular dystrophy).

MATERIALS AND METHODS

Generation of keratinocyte-specific Lmnb1 and Lmnb2 knockout mice

To create conditional knockout alleles for Lmnb1 and Lmnb2, we used ‘targeted trapping/conditional-ready’ gene-targeting vectors (28). Bacterial artificial chromosome (BAC) clones for Lmnb1 and Lmnb2 were isolated from a mouse BAC library (strain 129OlaHsd). Second, a βgeo gene-trapping cassette was inserted into the targeted BAC vector clones with the λ-Red recombineering system and Gateway cloning (Invitrogen). Both Lmnb1 and Lmnb2 targeting vectors were designed to insert one loxP site upstream of the βgeo cassette in intron 1, and a second loxP site in intron 2. The targeting vectors were linearized with AsiSI and electroporated into strain 129/OlaHsd ES cells; targeted ES cell clones were identified by Southern blotting with a flanking probe. Next, the targeted ES cells were transfected with a supercoiled Flp recombinase plasmid to excise the βgeo cassette, creating ES cells harboring a conditional knockout allele (exon 2 is flanked by loxP sites). Those ES cells were used to produce germline-transmitting chimeric mice. The ‘floxed’ alleles were not hypomorphic alleles; Lmnb1 and Lmnb2 expression levels in Lmnb1^fl/fl and Lmnb2^fl/fl mice were equivalent to those in wild-type mice (Supplementary Material, Fig. S7). A more detailed description of screening, genotyping and breeding is provided in the Supplementary Materials.

Studies with epidermal keratinocytes

Primary keratinocytes were isolated from mouse skin on postnatal Day 1 or 4 (P1 or P4) and seeded into 6-well tissue culture plates with mitotically inactivated 3T3 fibroblasts. The cells were cultured in Williams E medium with 0.07 mm CaCl₂ in 5% CO₂ at 37°C. For growth assays, equal numbers of cells were seeded into each well of a 6-well tissue culture plate along with mitotically inactivated 3T3 fibroblasts. Numbers of keratinocytes were counted with a Coulter counter at Days 1, 5, 8 and 11. To assess chromosome number, primary keratinocytes were incubated for 4 h with 10 μg/ml colcemid (Sigma). The cells were then centrifuged and treated with 0.075 m KCl and fixed in a mixture of glacial acetic acid and methanol (1:2.5). After spreading on slides, DNA was stained with Giemsa (Gibco). Images were obtained on a Nikon Eclipse E600 microscope with a 40×/0.75 objective and processed with SPOT software (Diagnostic Instruments).

Histological studies and immunofluorescence microscopy

Mouse skin was fixed for 4 h in 10% paraformaldehyde (PFA), dehydrated for 24 h in 70% ethanol and embedded in paraffin. Sections (5 µm) were stained with hematoxylin and eosin. For immunofluorescence microscopy studies, skin and tongue were embedded in Sakura Tissue-Tec OCT compound, frozen on dry ice, and 8 µm thick sections were cut with a cryostat. Slides were fixed in 4% formalin for 10 min, permeabilized with PBS containing 0.1% Tween-20 and incubated in blocking buffer (PBS containing 10% donkey or goat serum, 0.2% BSA, 1 mm CaCl₂ and 1 mm MgCl₂) for 1 h. Slides were then incubated overnight at 4°C with primary antibodies diluted in blocking buffer containing 0.1% Triton X-100. Secondary antibodies were incubated in a similar fashion for 0.5–1 h. After washing the slides, they were fixed with 4% PFA for 10 min, incubated in PBS/DAPI for 10 min and mounted with Prolong-Gold Antifade reagent (Invitrogen). The following antibodies were used for the immunochemical studies: lamin B1 (goat, 1:200, Santa Cruz), lamin B2 (mouse, 1:200, Invitrogen), lamin A (mouse, 1:200, Millipore), keratin 14 (rabbit, 1:800, Covance), phosphohistone H3 (rat, 1:1000, Sigma), Lap2 (mouse, 1:400, BD Laboratories) and BrdU (rat, 1:200, Abcam). Alexa Fluor 488- and Alexa Fluor 568-conjugated secondary antibodies were used at a 1:200 dilution. Lamin B2- and lamin A-specific monoclonal antibodies were labeled with Alexa Fluor 555 and Alexa Fluor 488 Antibody Labeling Kits (Invitrogen), respectively. Images were recorded with an Axiovert 200M microscope equipped with an AxioCam MRm and an ApoTome (all from Zeiss).

BrdU labeling

Forty-day-old mice were given an intraperitoneal injection of BrdU (40 mg/kg of body weight; Sigma) and euthanized 1 h later. Ear tissue biopsies were pretreated with 1 N HCl for 10 min on ice, 2 N HCl for 10 min at room temperature, followed by 10 min at 37°C and 0.1 m sodium borate (pH 8.5) for 12 min. Sections (8 µm) were cut and processed for immunofluorescence microscopy as described earlier.

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG online.

FUNDING

This work was supported by the National Institutes of Health Grants (HL76839, CA099506–07, HL086683, HL089781); the Ellison Medical Foundation Senior Scholar Program; March of Dimes Grant (6-FY2007-1012); a Beginning Grant-in-aid from the American Heart Association, Western States Affiliate (0865262F) and a fellowship from the Vascular Biology Program at UCLA (2 T32 HL069766:06).