• Open Access

Stem cell depletion in Hutchinson–Gilford progeria syndrome

Authors

  • Ylva Rosengardten,

    1. Department of Biosciences and Nutrition, Center for Biosciences, Karolinska Institutet, Karolinska University Hospital, Huddinge, Novum, SE-14186 Stockholm, Sweden
    Search for more papers by this author
  • Tomás McKenna,

    1. Department of Biosciences and Nutrition, Center for Biosciences, Karolinska Institutet, Karolinska University Hospital, Huddinge, Novum, SE-14186 Stockholm, Sweden
    Search for more papers by this author
  • Diana Grochová,

    1. Department of Biosciences and Nutrition, Center for Biosciences, Karolinska Institutet, Karolinska University Hospital, Huddinge, Novum, SE-14186 Stockholm, Sweden
    Search for more papers by this author
  • Maria Eriksson

    1. Department of Biosciences and Nutrition, Center for Biosciences, Karolinska Institutet, Karolinska University Hospital, Huddinge, Novum, SE-14186 Stockholm, Sweden
    Search for more papers by this author

Dr Maria Eriksson, Department of Biosciences and Nutrition, Center for Biosciences, Karolinska Institutet, Karolinska University Hospital, Huddinge, Novum, SE-14183 Huddinge, Sweden.Tel.: +46 8 524 81066; fax: +46 8 524 81170; e-mail:Maria.Eriksson.2@ki.se

Summary

Hutchinson–Gilford progeria syndrome (HGPS or progeria) is a very rare genetic disorder with clinical features suggestive of premature aging. Here, we show that induced expression of the most common HGPS mutation (LMNA c.1824C>T, p.G608G) results in a decreased epidermal population of adult stem cells and impaired wound healing in mice. Isolation and growth of primary keratinocytes from these mice demonstrated a reduced proliferative potential and ability to form colonies. Downregulation of the epidermal stem cell maintenance protein p63 with accompanying activation of DNA repair and premature senescence was the probable cause of this loss of adult stem cells. Additionally, upregulation of multiple genes in major inflammatory pathways indicated an activated inflammatory response. This response has also been associated with normal aging, emphasizing the importance of studying progeria to increase the understanding of the normal aging process.

Introduction

Hutchinson–Gilford progeria syndrome (HGPS or progeria) is a segmental premature aging disease that affects one in 4–8 million live births (DeBusk, 1972; Brown, 1992). Children born with HGPS appear normal at birth but develop symptoms within their first year of life. Clinical symptoms of progeria include alopecia, lack of subcutaneous fat, scleroderma-like skin changes, growth retardation, bone abnormalities, and joint stiffness. The average life span of patients with HGPS is 13 years, and most children die from heart-disease and other atherosclerotic complications (DeBusk, 1972; Merideth et al., 2008). A de novo single nucleotide mutation in exon 11 in the LMNA gene (c.1824C>T, p.G608G) is the most common mutation and causes 90% of HGPS cases (De Sandre-Giovannoli et al., 2003; Eriksson et al., 2003). The LMNA gene encodes the A-type lamins, which are the major proteins of the nuclear lamina situated underneath the inner nuclear membrane. The nuclear lamina is important in determining the shape and size of the nucleus and is involved in fundamental processes such as DNA replication, transcription, and repair (Dechat et al., 2008). The most common HGPS mutation partly activates a cryptic splice site, resulting in the production of a truncated mRNA transcript encoding a prelamin A protein with an internal deletion of 50 amino acids, known as progerin (De Sandre-Giovannoli et al., 2003; Eriksson et al., 2003). This internal deletion alters the posttranslational processing of prelamin A, which remains farnesylated and carboxymethylated, with an abnormal association with the inner nuclear membrane and the nuclear lamina, acting as dominant negative with serious effects on multiple cellular functions (Goldman et al., 2004; Capell & Collins, 2006; Dechat et al., 2007). The HGPS cellular phenotype, observed in fibroblasts from HGPS cases, includes nuclear blebbing, thickened nuclear lamina, abnormal distribution of nuclear pore complexes, and loss of peripheral heterochromatin (Goldman et al., 2004). Several reports have also showed an altered gene expression in cells from HGPS patients (Ly et al., 2000; Csoka et al., 2004). Interestingly, progerin has been found in normal cells and evidence points toward a similar molecular mechanism in progeria as in normal aging (Scaffidi & Misteli, 2006; McClintock et al., 2007; Rodriguez et al., 2009).

A proposed model to explain the disease mechanism causing HGPS suggests that premature exhaustion of stem cells inhibits the capacity for tissues to regenerate (Halaschek-Wiener & Brooks-Wilson, 2007). This model suggests that tissues characterized by a high degree of cell turnover because of continuous mechanical stress, such as the skin or tissues that undergo continuous growth, exhaust their progenitor cells resulting in an early depletion of stem cell pools. This model is supported by in vitro studies where progerin expression initiates differentiation in human mesenchymal stem cells (Scaffidi & Misteli, 2008). In addition, recent studies with induced pluripotent stem cells (iPSC) from HGPS patients have suggested that the proposed exhaustion of mesenchymal stem cells could be the cause of increased hypoxia sensitivity (Zhang et al., 2011).

To test the hypothesis of premature exhaustion of adults stem cells in vivo, we used our previously described mouse model, which displays several features of the HGPS skin phenotype (Sagelius et al., 2008a). Our previous studies using this model have shown that postnatal expression of the HGPS mutation results in a progressive phenotype with early stages characterized by epidermal hyperplasia and a hyperproliferative epidermis, and an end stage characterized by epidermal hypoplasia, hypoplastic sebaceous glands, loss of hypodermis and a fibrotic dermis (Sagelius et al., 2008a). Other mouse models of accelerated aging, and mouse models with a similar progressive skin phenotype, have been shown to have a reduction in stem cell number and an impaired stem cell function (Kuro-o et al., 1997; Waikel et al., 2001; Liu et al., 2007; Su et al., 2009). However, there has been no example of the opposite, where a human premature aging syndrome has been proven to be the result of stem cell depletion.

Results

Progeria mice display a reduced population of keratinocytes with stem and progenitor characteristics

To assay the number of stem cells in the skin of our progeria mice, we used a BrdU label-retaining technique. Four-day-old wild-type and progeria animals (tetop-LAG608G+; K5tTA+) received multiple BrdU injections at 24-h intervals to label mitotic cells. The BrdU label became diluted following cell division, and after a chase period of 70 days, only the cells that rarely divided were labeled. Label retention is consequently a marker of the proliferation history of a cell. Because stem cells are slow cycling by definition, these label-retaining cells were considered stem cells (Braun & Watt, 2004). Analysis of dorsal skin following the 70-day chase showed a significant reduction in label-retaining cells in progeria animals compared to wild-type (Fig. 1A–C). This reduction was evident in both transgenic animals with a milder skin phenotype and in the animals with a more progressed skin phenotype (Fig. 1C). Stem cells in the skin are believed to localize to specific regions (Trempus et al., 2003; Braun & Watt, 2004; Tumbar et al., 2004; Silva-Vargas et al., 2005). Therefore, we analyzed the label-retaining cells in different skin compartments (Fig. 1D,E). In the bulge, the sebaceous gland, and the interfollicular epidermis regions, where stem cells have been shown to localize, there was a significant difference in both progeria animal groups when compared to wild-type animals (Fig. 1D,E).

Figure 1.

 Expression of the HGPS mutation results in a reduced population of label-retaining cells in dorsal skin following a 70-day chase. Mice expressing the HGPS mutation from the date of birth (Day 0) exhibit a more severe and more rapid progression of the phenotype than animals expressing the HGPS mutation from postnatal day 21 (Day 21; Sagelius et al., 2008a). (A,B) BrdU immunofluorescent representative slides of dorsal skin from (A) wild-type and (B) progeria mice, BrdU (red) and keratin 5 (green). The total number of label-retaining cells per follicle region (C) and separated per follicle compartment (D,E). Mice were intercrossed on doxycycline, which was removed at the date of birth (C,D) or on postnatal day 21 (A–C,E). All values represent the mean ± SD (*< 0.05, **< 0.01). Scale bars are 50 μm.

To further study the population of stem cells in the skin, progeria and wild-type primary keratinocytes were isolated from animals of varying ages and analyzed for α6-integrin and CD34 expression using fluorescence-activated cell sorting (FACS). CD34 expression in mouse skin overlaps with the label-retaining cells in the bulge region of the hair follicle, and α6-integrin is expressed in the basal layers of the epidermis. Together, these two markers have been shown to mark cells with progenitor or stem cell characteristics (Trempus et al., 2003; Tumbar et al., 2004; Silva-Vargas et al., 2005). Keratinocytes isolated from progeria animals showed a decreased number of α6-integrinhighCD34high cells compared to wild-type keratinocytes, indicating a reduced population of cells with stem or progenitor cell characteristics in progeria animals (Fig. 2A). The decrease in this cell population was already evident within 13 weeks of transgenic expression (Fig. 2A; data not shown).

Figure 2.

 Progeria mice display a reduced population of keratinocytes with stem and progenitor characteristics. (A) Representative fluorescence-activated cell sorting (FACS) analysis of CD34- and α6-integrin-positive cells in keratinocytes extracted from wild-type and progeria mice. These keratinocytes were extracted from animals with transgene expression from postnatal day 21, followed by 22–27 weeks of transgenic expression (B–F) colony-forming assay. Keratinocytes from progeria mice (D,E) showed a reduced ability to form colonies when compared to keratinocytes extracted from wild-type mice (B,C). Colonies within the boxed areas in figures B and D, shown magnified in figures C and E, respectively, exhibit less dense cell growth in colonies from progeria (E) keratinocytes compared to wild-type (C). These less dense colonies were probably largely made up of remaining feeder layer cells. (F) Both the number of colonies and colony size were reduced in keratinocytes extracted from progeria mice compared to wild-type mice. Keratinocytes were extracted following 37 weeks of transgenic expression. Mice were intercrossed while being treated with doxycycline, which was removed at postnatal day 21. (G) Downregulation of epidermal stem cell markers Lrig1, Cd34, and Krt15 in progeria mice. Quantitative RT-PCR relative expression analysis of stem cell markers in keratinocytes isolated from 5-week-old animals (n = 3–4 wild-type and progeria mice). Mice were intercrossed while being treated with doxycycline, which was removed at date of birth. All values represent the mean ± SD (*< 0.05, **< 0.01, ***< 0.001). (H) Expression of the progeria mutation results in irreversible effects on the maintenance of cells with stem and progenitor cell characteristics. Representative FACS analysis of cells positive for α6-integrin and CD34. Mice were intercrossed while being treated with doxycycline, which was removed at the date of birth. Transgenic expression was suppressed at postnatal week 7. Keratinocytes were isolated from dorsal skin of wild-type and progeria mice at postnatal week 37, after 30 weeks of sustained transgenic suppression.

A complementary approach to study epidermal cells with stem cell properties was to analyze the ability of primary keratinocytes to form colonies in vitro. Stem cells form large self-renewing colonies, whereas their nonstem cell daughters divide only a small number of times before undergoing terminal differentiation (Barrandon & Green, 1987). The colony-forming assay showed that keratinocytes isolated from wild-type animals were able to form colonies that remained after 16 days of cultivation, whereas cells from progeria animals showed an impaired ability to produce colonies (Fig. 2B–F). Colonies from progeria animals were fewer in number, smaller in size, and exhibited a less dense cell growth pattern compared to wild-type (Fig. 2F,E compared to Fig. 2C). Downregulation of stem cell markers Lrig1, CD34, and Krt15 was also seen by quantitative RT-PCR in progeria mice compared to wild-type mice (Fig. 2G).

Impaired stem cell function caused by the expression of the progeria mutation is reversible to some extent

Previously, we have shown that the progeria skin phenotype is reversible following transgenic suppression (Sagelius et al., 2008b). This reversible phenotype suggested that stem cells were still present and able to regenerate the epidermis. In our previous study, the progeria mutation was expressed for 7 weeks before the transgene was suppressed. The phenotype including hair thinning, skin crusting, and hyperplasia of the interfollicular epidermis was already significant after 5 weeks of transgenic expression (Sagelius et al., 2008b). To analyze stem cells in animals with a reversed phenotype, we sustained transgenic expression in progeria animals for 7 or 10 weeks to develop a phenotype. After this period, the transgenic expression was suppressed, resulting in phenotype reversal (data not shown). Following 21 or 30 weeks of transgenic suppression, keratinocytes were isolated and analyzed for the expression of CD34 and α6 integrin using FACS analysis. The keratinocyte population isolated from progeria animals showed a decrease in α6 integrinhighCD34high cells, indicating a reduced population of cells with stem and progenitor cell characteristics after phenotype reversal (Fig. 2H). These results suggested that even though the progeria skin phenotype was fully reversed (Sagelius et al., 2008b), expression of the progeria mutation had effects on the adult stem cells in the skin, but not enough to cause any phenotype. This will most likely render the skin more sensitive in stress situations. However, animals with prolonged transgenic expression (13 weeks of transgene expression) were not able to reverse the external phenotype after transgene suppression (data not shown). This suggested that stem cells that were still present in animals after 7 or 10 weeks of transgene expression were capable of regenerating the epidermis, whereas after 13 weeks of transgene expression, the stem cells are either too few in numbers or have lost their ability to repopulate and rescue the diseased skin.

Impaired epidermal wound healing in progeria mice

An important function of epidermal stem cells is to actively contribute to wound repair (Ito et al., 2005). Following wounding of the skin, healing can be divided into three separate stages: inflammation, new tissue formation, and remodeling (Gurtner et al., 2008). The stem cells of the skin are mostly active in tissue formation, in which cellular proliferation and migration close the open wound. In addition to impaired stem cells, ulcerated lesions have previously been described in our mice (Sagelius et al., 2008a), indicating defective wound healing. To study wound healing, we created 3-mm wounds on the backs of wild-type and progeria animals. The wound area was excised 4 and 7 days postwounding and analyzed histologically. Four-day-old wild-type wound edges exhibited well-organized proliferation and migration (Fig. 3A,C,I), whereas wound edges in progeria animals were uneven and disorganized (Fig. 3B,D,J). After 7 days, the wounds from wild-type animals were completely re-epithelized (Fig. 3E,G), whereas progeria animals still had un-epithelized wounds and a delay in collagen regeneration (Fig. 3F,H). Additionally, wounds of progeria animals showed evidence of increased inflammation in the connective tissue and impaired wound cleansing, with an increased amount of foreign materials in the wound as a result (Fig. 3B,F; data not shown). To analyze the proliferative potential of cells involved in wound healing, a short-term chase was performed by injecting the animals with a dose of BrdU 1 h before the tissue was collected. BrdU was incorporated into dividing cells and marked proliferating cells. Both wild-type and progeria animals displayed proliferating cells at the wound edge (Fig. 3I,J). Seven days postwounding, progeria animals showed a significantly higher number of BrdU-labeled cells, indicating increased proliferation (Fig. 3K). However, impaired migration of progeria keratinocytes delayed the healing of the wounds in these animals compared to wild-type animals. This impaired migration of keratinocytes could be the result of reduced expression of integrins (O’Toole, 2001). To test this, we analyzed transcripts from progeria and wild-type keratinocytes, and the results showed reduced expression of β1- and α6-integrins in cells from progeria compared to wild-type mice (Fig. 3L).

Figure 3.

 Impaired epidermal wound healing in progeria mice. (A–J) 3-mm wound cross-sections from wild-type mice 4 days (A,C,I) and 7 days (E,G) postwounding and from progeria mice 4 days (B,D,J) and 7 days (F,H) postwounding. Masson’s trichrome (A,B,E,F) shows increased inflammation in connective tissue (marked with asterisks) and a delay in collagen regeneration under the wounds of the progeria mice (indicated with arrowheads; B,F) compared to wild-type mice (A,E). (C,D,G–J) Immunofluorescence results from a short-term chase with BrdU indicating actively dividing cells (red, highlighted with an arrow) and keratin 5 (green). Uneven and disorganized wound edges, disrupted re-epithelization (D,H,J), and increased proliferation of BrdU-positive cells at the wound edges of progeria mice (J) compared to wild-type (C,G,I). Enlargement of wound edges, I of C and J of D. (K) Quantification of dividing cells at the wound edge in 4-day- (wild-type, n = 4; progeria, n = 4) and 7-day- (wild-type, n = 4; progeria, n = 4) old wounds. (L) Reduced expression of β1-and α6-integrins in progeria mice as analyzed by quantitative RT-PCR of keratinocytes isolated from 5-week-old animals (n = 4 wild-type and progeria mice). Vertical dashed lines in (A–H) indicated wound centers. Scale bars in (A–H) = 400 μm, in (I,J) = 100 μm. Values represent the mean ± SD (*< 0.05, ***< 0.001). Mice were intercrossed while being treated with doxycycline, which was removed at date of birth.

DNA damage, senescence, and a senescence-associated secretory phenotype in progeria mice

In cells with DNA double-strand breaks, phosphorylated H2AX (γH2AX) localizes to form characteristic foci (Rogakou et al., 1999). Analysis of keratinocytes from progeria mice showed an increased frequency of cells with numerous γH2AX foci, suggesting an increased number of cells with severe DNA damage or with an impaired DNA damage repair mechanism (Fig. 4A,B). Severe or irreparable DNA damage, especially DNA double-strand breaks, is one trigger of cellular senescence with enhanced expression of the senescence-associated β galactosidase (SA-β-gal; Dimri et al., 1995; Rodier et al., 2009). Staining for SA-β-gal in skin from wild-type and progeria mice revealed a tendency of increased senescence in the interfollicular epidermis of progeria mice skin (Fig. 4C,D). Staining of SA-β-Gal in the hair follicle and sebaceous gland has previously been described as background staining (Dimri et al., 1995) and was not considered in the analysis.

Figure 4.

 Increased DNA damage, senescence, and downregulation of stem cell maintenance pathways in progeria mice. (A,B) Keratinocytes isolated from progeria mice have a higher frequency of cells with DNA double-strand breaks. (A) Representative staining of cytospun cells isolated from a progeria mouse, where green is keratin 5, red is γH2AX, and blue is DAPI. (B) The frequency of cells with more than five γH2AX loci was counted from a minimum of 180 keratinocytes from six (wild-type) and five (progeria) mice. (C) Skin section from progeria mice displays activation of endogenous β-gal (blue), indicating a senescent phenotype. The picture shows a representative negatively stained region in the wild-type sample and a positively stained region from a progeria mouse. Staining of SA-β-Gal in the hair follicle and sebaceous gland, as is seen in the wild-type sample, is seen in all samples and considered background (Dimri et al., 1995). (D) The frequency of positively stained microscope fields (20× magnification) was calculated from a minimum of 52 microscope fields per animal (wild-type, n = 3; progeria, n = 3). (E–F) Keratinocytes isolated from 5-week-old animals display activated inflammatory response (E) and downregulation of Wnt and Notch pathways (F), as analyzed by quantitative RT-PCR of target genes. The break in the y axis designates a change in scale (n = 3–4 wild-type and progeria mice). Values represent the mean ± SD (*< 0.05, **< 0.01, ***< 0.001). Scale bars are 20 μm.

There is growing evidence that senescent cells secrete inflammatory factors that alter the behavior of surrounding cells to a pro-inflammatory phenotype, and this set of characteristics is termed the “senescence-associated secretory phenotype” (SASP) (Coppéet al., 2008). Senescence-associated secretory phenotype proteins are generally induced at the mRNA level (Coppéet al., 2008); therefore, we performed quantitative RT-PCR analysis of selected inflammatory markers in keratinocytes from 5-week-old wild-type and progeria animals. All SASP markers showed a significant increase in the keratinocytes from progeria mice (Fig. 4E). In agreement with the increased proliferation we have previously reported in these mice (Sagelius et al., 2008a), the SASP proteins promote proliferation (Freund et al., 2010). Chronic inflammation is also associated with normal aging and has been shown to impair stem cell function (Freund et al., 2010).

Wnt and Notch pathways are downregulated in progeria mice

Based on our results revealing impaired adult stem cell function and previous observations of fibroblasts from progeria children (Scaffidi & Misteli, 2008), we analyzed genes known to be involved in stem cell maintenance. Analyzing multiple genes within the Wnt and the Notch pathways in keratinocytes isolated from animals following 5 weeks of transgenic expression showed a decrease in the expression of genes linked to both pathways compared to wild-type mice (Fig. 4F). This is in agreement with a disease mechanism that results in impaired and depleted stem cells, and it is most likely a secondary effect from a reduced stem cell population.

p63 is downregulated in progeria mice

To address the molecular mechanism underlying the impaired and altered stem cell function in progeria animals, we conducted mRNA expression analysis in keratinocytes isolated from progeria and wild-type animals. The transcription factor p63, a member of the p53 family, is required for normal epidermal development and differentiation. The gene encoding p63, the Trp63 gene, contains two promoters generating two distinct classes of protein. The two classes differ in the N-terminal end of the proteins; TAp63 contains the transactivation domain, while the truncated form ΔNp63 lacks this domain. Both TAp63 and ΔNp63 can be alternatively spliced at the C- terminal end, resulting in a total of six different isoforms (Candi et al., 2008). ΔNp63 is the predominantly expressed isoform of the p63 protein in epidermis. Knockdown experiments have shown that decreased expression of ΔNp63 in the epidermis leads to increased proliferation in suprabasal cells of the epidermis as well as impaired wound healing, similar to what we see in our progeria mice (Koster et al., 2007; Sagelius et al., 2008a). p63 has also been shown to be important for maintaining the proliferative state of transit amplifying cells and preventing premature onset of their terminal differentiation (Candi et al., 2008). It has also been shown that p63 deficiency in vivo results in a shortened lifespan and many features of accelerated aging (Keyes et al., 2005; Su et al., 2009). These studies have been performed on p63 deficient mice with all isoforms of p63 ablated in an inducible and tissue-specific model (Keyes et al., 2005) and in one model with ablated TAp63 expression in the epidermis (Su et al., 2009). Quantitative RT-PCR performed on keratinocytes isolated from the skin of progeria mice showed a downregulation of ΔNp63 already after 5 weeks of transgene expression compared to wild-type mice (Fig. 5A). Downregulation of p63 expression was further confirmed at the protein level (Fig. 5B,C). To further examine the p63 downregulation in our progeria mice, we performed quantitative RT-PCR on downstream targets of p63 (Fig. 5D). While there was no significant difference between keratinocytes from wild-type and progeria mice for most of the genes tested several of the genes did show a trend toward being downregulated already after 5 weeks of transgenic expression (Igfbp3, Itga3, Emp1, and Tcf3; Fig. 5D). Phlda1 and Krt1 was significantly upregulated (Fig. 5D). Phlda1 has previously been shown to be a p63 target and upregulated in keratinocytes upon p63 knockdown (Truong et al., 2006; Barton et al., 2010). Krt1 has previously been shown to be downregulated upon p63 knockdown (Truong et al., 2006), the discrepancies with our results are most likely caused by the epidermal hyperplasia seen in our mice, with increased expression of Krt5, Krt1, Krt10 (Sagelius et al., 2008a).

Figure 5.

 Downregulation of p63 in progeria mice and HGPS patients. (A) Quantitative RT-PCR relative expression analysis of Trp63ΔN in keratinocytes isolated from 5-week-old animals (n = 3–4 wild-type and progeria mice). (B) Quantification of p63 stained cells in the dorsal skin of 5-week-old wild-type and progeria mice in p63 immunohistochemistry staining of skin sections from a wild-type and progeria mouse. Cells with strong p63 staining were counted in four to ten hair follicle regions per animal, and values represent the mean percentage of strongly stained cells among the total number of cells in the hair follicle region from three (wild-type) to four (progeria) biological replicates. (C) Representative pictures of immunohistochemistry staining on skin sections from wild-type and progeria mice. (D) Quantitative RT-PCR of p63 downstream target genes. The break in the y axis designates a change in scale. Values represent the mean ± SD (*< 0.05, **< 0.01, ***< 0.001). (E) Quantitative RT-PCR shows a decreased expression of TAp63 relative to age in lymphoblasts from HGPS patients, but not in HGPS sibling lymphoblasts or adult control lymfoblasts. Values represent the mean ± SD (*< 0.05, **< 0.01, ***< 0.001). Scale bars are 20 μm.

p63 is downregulated in HGPS patients

To examine whether the downregulation of p63 seen in keratinocytes from progeria mice also applies to HGPS patients, we performed quantitative RT-PCR on samples from HGPS patients and controls. We could not detect any p63 isoform in primary fibroblasts form HGPS patients (data not shown). While the mRNA for ΔNp63 was not detectable in lymphoblasts by quantitative RT-PCR, the mRNA levels of TAp63, the major isoform in lymphoblasts, were quantified in cells from HGPS patients and controls (Fig. 5E). There was no difference in the TAp63 expression levels between progeria and their unaffected siblings (P = 0.716) or the adult control group (P = 0.157). However, for the TAp63 expression levels and age of sampling, there was a striking degree of correlation, R= 0.845, P = 0.001, for the progeria samples (Fig. 5E). This inverse correlation with a decrease in TAp63 expression levels with increased age of progeria patients was not seen in any of the other control groups (Fig. 5E).

Discussion

The genetic basis for HGPS was discovered 8 years ago, and since then, there has been enormous progress not only in progeria research but also in the field of lamins and the nuclear envelope and its implications in normal aging. In this study, we have expanded the knowledge of the disease mechanism in HGPS using our previously characterized inducible mice model that recapitulates multiple features of the HGPS skin phenotype (Sagelius et al., 2008a). Our findings clearly indicate that the expression of the most common HGPS mutation in postnatal epidermis of mice results in a reduced number of epidermal cells with stem and progenitor properties. These results contrast with another report that proposes that there is an accumulation of epidermal stem cells in a mouse model with lamin A premature aging (Espada et al., 2008). A possible explanation for this discrepancy could be that this model did not express the HGPS mutation, instead it was a mouse model that was null for the prelamin A-processing enzyme, Zmpste24, which might also affect not only prelamin processing but also other proteins in the cell. In addition and in contrast with our mice model, these mice did not have a progressive phenotype suggestive of stem cell depletion (Pendás et al., 2002).

Our progeria mice exhibit impaired wound healing because of impaired stem cell function. Delayed wound healing and loss of elasticity in the skin are phenomenon of normal aging, and it has been proposed that these effects are caused by impaired tissue-specific stem cell regeneration because of the accumulation of DNA damage and telomere shortening (Sharpless & DePinho, 2007). The extent to which the stem cells are active in maintaining normal tissue homeostasis and repair correlates with the severity of these aging effects (Rando, 2006).

Expression of the progeria mutation (LMNA c.1824C>T, p,G608G) has previously been shown to have an effect on stem cells with aberrant differentiation of human mesenchymal stem cells to the osteogenic and adipogenic lineage in vitro (Scaffidi & Misteli, 2008). This was preceded by an activation of downstream targets of the Notch signaling pathways. In this study, we have seen downregulation of both the Notch and Wnt signaling pathways in the epidermis of mice as a result of the expression of the progeria mutation. Notably, reduced Wnt signaling has been recently reported in Δ9Lmna mice that develop a progeroid phenotype (Hernandez et al., 2010), which together with our data shows that impairment of the Wnt pathway might be a general defect in premature aging diseases and that reactivation of Wnt signaling could be one potential therapeutic approach for HGPS. Wnt signaling reduction is also in agreement with a proposed disease mechanism resulting in impaired and depleted stem cells, and it is most likely a secondary effect from a reduced stem cell population. Future studies, especially from mouse models, will tell whether a HGPS disease mechanism that results in early depletion of adult stem cells in the epidermis could be applied to other tissues that are affected in HGPS.

In a previous study from our laboratory using the same mouse model, we showed that the progeria phenotype is reversible (Sagelius et al., 2008b). In that study the transgenic expression was sustained until postnatal week 7. We here report that while 7 or 10 weeks of sustained expression of the HGPS mutation is reversible, after 13 weeks of sustained expression it is no longer possible to reverse the skin phenotype. Our results from these studies emphasize the importance of introducing therapy as early as possible in children with progeria.

Previously, it has been shown that progerin expression induces cell senescence associated with DNA damage. Here, we showed that this senescence might be a consequence of p63 protein impairment. This is supported by other observations that ablation of p63 in proliferating epithelia leads to several features of accelerating aging (Keyes et al., 2005). p63-induced senescence leads to the upregulation of several pro-inflammatory cytokine genes and to a SASP. Senescence-associated secretory phenotype in turn may have a deleterious effect on epithelial cells in early stages of our progeria mice – leading to their hyper-proliferation. Chronic inflammation associated with SASP directly or indirectly disrupts the function of epidermal stem cells. Expression of SASP factors is dependent on a function of the transcription factor NF-kB, and it has been shown that cells lacking lamin A/C proteins fail to activate efficient transcription by NF-kB (Lammerding et al., 2004). It is, therefore, possible that the dominant negative form of the lamin A protein, progerin, could inappropriately activate NF-kB pathway with consequent activation of SASP proteins. This hypothesis remains to be tested. Chronic inflammation is associated with physiological aging and is a cause of many age-related diseases such as cancer or atherosclerosis; therefore, SASP might be one of the possible links between premature and normal aging.

In this study, we have shown that the major isoform of p63 in the epidermis, ΔNp63, is downregulated on both mRNA and protein level. In addition, we have shown that expression levels of several p63 targets are affected. We have also shown that TAp63, the isoform of p63 in lymfoblasts, is downregulated in HGPS patients. However, additional experiments addressing the direct effect from progerin expression on p63 and the cause of its downregulation are warranted. In conclusion, in this study, we demonstrated that the expression of the HGPS mutation in the skin results in impaired function and depletion of epidermal stem cells. We show that progerin expression in the skin reduces the expression of p63, and induces DNA damage or possibly impairs DNA repair mechanisms. We also show that there is an increased expression of cytokines and increased inflammation. Whether this “SASP” response is the direct cause of the stem cell depletion in the skin, and if this is also the same mechanism for other affected tissues in progeria, remains to be investigated.

Experimental procedures

Transgenic mice

Mice were housed within the animal facilities at the Karolinska Institutet, Huddinge, Sweden, and animal breeding and genotyping were in accordance with previously described procedures (Sagelius et al., 2008a). Heterozygous tetop-LAG608G animals from the F1 line VF1-07 (Sagelius et al., 2008a) and K5tTA (Diamond et al., 2000) were intercrossed while being treated with 100 μg ml−1 doxycycline (D9891; Sigma-Aldrich, St. Louis, MO, USA) and 2.5% sucrose, which was replaced with normal water on postnatal day 0 or 21. The conditions for housing and feeding were in agreement with previously described procedures (Sagelius et al., 2008a,b). Animal studies were approved by the Stockholm South Ethical review board, Dnr S148-03, S141-06, S107-09.

BrdU labeling and label-retaining cell analysis

For label-retaining cell analysis, 4-day-old progeria (n = 6) and wild-type (n = 6) littermates received five BrdU (5-Bromo-2′-deoxy-uridine, Sigma B9285-50MG) injections (I.P.) at a 24-h intervals (50 mg kg−1). Animals were sacrificed 70 days after the last injection, and dorsal skin samples were collected and incubated at 4°C overnight in 4% paraformaldehyde (pH 7.4). After fixation, the samples were dehydrated in ethanol and xylene and embedded in paraffin. To analyze BrdU label-retaining cells, 4-mm sections were used for immunofluorescent staining. The sections were rehydrated, followed by antigen retrieval in 10 mm citrate buffer in a microwave and 2 m HCl at 37°C. Before primary antibody incubation, the sections were blocked in normal goat serum and mouse-to-mouse blocking reagent (Scytek, Logon, UT, USA). The following primary antibodies were used: anti-BrdU (1:25, 7580; Becton Dickinson, Franklin Lakes, NJ, USA) and anti-Keratin5 (1:1,000, PRB-160P; BioSite), and the corresponding secondary antibodies were: Alexa 555-conjugated goat anti-mouse (1:100, A-21422; Invitrogen Probes, Invitrogen, Carlsbad, CA, USA) and FITC-conjugated goat anti-rabbit (1:200, ab6717 abcam). BrdU-labeled cells were counted per hair follicle region where each follicle was divided into five areas: the bulb, the sebaceous gland, and the surrounding area including the bulge, the infundibulum, the interfollicular epidermis, and the dermis, stretching to the next hair follicle. In each group, skin sections from six animals were analyzed, and labeled cells from three anatomical regions from each mouse were counted.

Microscopy

Immunofluorescent imaging was performed on the Zeiss Axioplan 2 (Carl Zeiss AG, Oberkochen, Germany) compound fluorescence microscope, equipped with an HBO 100 watt AttoArc mercury lamp. Zeiss Plan-NEOFLUAR 40x/0.75 and 20x/0.5 dry lens were used. Imaging was performed at room temperature, and images were taken with the Zeiss Axiocam MRm camera and Axiovision software. Samples from the wound healing study were imaged with a Zeiss LSM510 laser scanning confocal microscope using sequential scanning in frame mode. A Plan-Neofluar 40x/1.3 Oil DIC immersion objective and LSM 5 Software were used (Carl Zeiss, Göttingen, Germany). Immunohistochemical imaging was performed on a Nikon Eclipse E1000, using Nikon plan apo 40x/0.95 and 20x/0.75 lenses. Images were taken with a Nikon DXM1200 camera and the Nikon ACT-1 imaging software. Quantification was performed with Fiji (http://pacific.mpi-cbg.de), a scientific image processing application based on ImageJ (http://rsb.info.nih.gov/ij).

Keratinocyte isolation and staining

Primary keratinocytes were isolated from adult progeria animals and wild-type littermates as previously described (Jaks et al., 2008). Briefly, animals were sacrificed by cervical dislocation, and the fur was trimmed before the skin was taken and placed into Ca2+ free PBS (d-PBS; Invitrogen). Subcutaneous tissue was removed, and the skin was floated on trypsin (T4424; Sigma-Aldrich) for 2 h at 32°C. After incubation, the epidermis was scraped into S-minimal essential medium (S-MEM; Invitrogen) supplemented with 0.01% soybean trypsin inhibitor (T9128; Sigma-Aldrich) and 0.5% BSA (bovine albumin fraction V solution (7.5%); Invitrogen) and gently mixed with a magnetic stirrer at room temperature for 20 min. The cell suspension was filtered through a 70-mm cell strainer (BD Falcon, Becton Dickinson, Franklin Lakes, NJ, USA), and viable cells were counted after staining with tryphan blue.

For cytospin preparations, 100 000 isolated cells per slide were centrifuged for 15 min at 28 g in a Shandon Cytospin 4 (Thermo Electron Corporation, Waltham, MA, USA). After drying, cytospin preparations were stored at −80°C until immunofluorescence staining. After fixation with 4% paraformaldehyde (pH 7.4), cells were permeabilized in PBS containing 1% NP-40 (Surfact-Amps NP-40; Pierce Biotechnology, Rockford, IL, USA) and blocked in 5% normal goat serum and 0.1% Birj (Surfact-Amps 58; Brij 58; Pierce Biotechnology) in PBS before incubation with the primary antibodies anti-Keratin5 (1:1,000, PRB-160P; BioSite) or anti-Keratin5 (1:1000, PRB-160P; BioSite) co-stained with anti-phospho-Histone H2A.X (γH2AX; 1:500, JBW301; Millipore, Billerica, MA, USA) at 4°C overnight. Samples were incubated with the secondary antibodies FITC-conjugated goat anti-rabbit (1:200, ab6717 abcam) and Alexa Fluor594-conjugated goat anti-mouse (1:200, A21125; Molecular Probes) in the dark for 30 min at room temperature before mounting in vectashield mounting media containing DAPI (Vector laboratories, Burlingame, CA, USA). The fraction of keratin 5-positive cells was calculated from a minimum of 400 isolated keratinocytes from wild-type (n = 3) and progeria (n = 3) animals. Results from the keratinocyte isolation showed that > 90% of the cell populations was keratinocytes. The frequency of cells with more than five γH2AX foci was calculated from the isolated keratinocytes of wild-type (n = 6) and progeria (n = 5) animals. A minimum of 180 keratin 5-positive cells were counted for each animal.

Colony-forming assay

Keratinocytes isolated from animals with transgene expression from their date of birth (wild-type, n = 3; progeria, n = 3) were isolated after 13 weeks of transgene expression, and animals with transgene expression from their postnatal day 21 (wild-type, n = 2; progeria, n = 2) were isolated after 37 weeks of transgene expression (postnatal week 40). Thirty thousand viable keratinocytes per well were seeded onto collagen IV precoated six-well plates (BD bioscience). The cells were grown in epidermal keratinocyte medium (CnT-02 CELLnTEC) on γ-irradiated (30 Gy) NIH T3T feeder layer cells (100 000 cells per well). After 16 days of cultivation, the cells were rinsed in PBS and fixed in 4% paraformaldehyde (pH 7.4). To visualize the keratinocyte colonies, the cells were stained with 1% rhodamine B (R6626; Sigma-Aldrich). Colony size and number were analyzed. Two six-well plates per experiment were used.

Fluorescence-activated cell sorting

Single-cell suspensions of isolated keratinocytes were blocked with mouse-to-mouse blocking reagent (Scytec) before incubation with PE-Cy5-conjugated anti-a6 integrin (1:5, CD49f; BD biosciences) and PE-conjugated anti-CD34 (1:40; RAM34 eBiosciences, San Diego, CA, USA) antibodies. The antibodies were diluted in 2% chelex-treated (Chelex 100 Resin; Bio-Rad, Hercules, CA, USA) FBS and incubated for 2 h at 4°C. From each sample, a minimum of 32 000 events were analyzed using a FACScalibur flow cytometer (BD bioscience). Keratinocytes from animals with transgene expression from postnatal day 21 (wild-type n = 4, progeria n = 6) of ages between 25 weeks and 40 weeks old and keratinocytes from animals with transgene expression from their date of birth (wild-type, n = 3; progeria, n = 3) were analyzed at 13 weeks of age. Reversible effects on the stem cell pool were investigated using isolated keratinocytes from animals bred on doxycycline treatment that was removed at the date of birth. To allow phenotypes to develop, the animals (wild-type, n = 3; progeria, n = 4) were kept on normal water for 7 weeks after which doxycycline was reintroduced. Transgenic expression was suppressed for 30 weeks before animals were sacrificed, and keratinocytes were isolated and analyzed by FACS.

Quantitative RT-PCR

RNA was isolated from keratinocytes from 5-, 9- and 12-week-old wild-type (n = 3–4 per individual age group) and progeria (n = 3–4 per individual age group) animals (with transgenic expression induced at the day of birth) using TriZol® Reagent (Invitrogen). Random hexamers and SuperScript II Reverse Transcriptase (Invitrogen) were used for cDNA synthesis from 800 ng RNA. Primer sequences and conditions will be provided upon request. All reactions were run in triplicate, and data were only accepted when the variation among the triplicates was < 0.3 units for CT < 30 and > 0.5 units for CT  ≥ 30. To calculate relative changes in gene expression, we used the comparative CT method, the 2−ΔΔCT method (Schmittgen & Livak, 2008). Data were interpreted as the expression of the gene of interest relative to the reference gene (Actb) in progeria animals compared to wild-type animals.

Wound healing assay

Using a 3-mm biopsy puncher (Medicarrier, Stockholm, Sweden), wounds were introduced on the lower backs of 10-week-old progeria (n = 4) and wild-type (n = 4) animals and analyzed after 4 and 7 days. Animals were bred on doxycycline, which was removed at the date of birth. One-hour prior to sacrificing, the animals were injected (I.P.) with 250 mg kg−1 BrdU (5-Bromo-2′-deoxy-uridine, Sigma-Aldrich, St. Louis, MO, USA) in PBS. Dorsal skin and wounds were fixed at 4°C overnight in 4% paraformaldehyde (pH 7.4), dehydrated in ethanol, and embedded in paraffin. Four-micron sections were stained with hematoxylin and eosin (H&E) or Masson’s trichrome stain according to standard procedure. BrdU integration was detected by immunofluorescence staining on 4-μm sections using anti-BrdU (1:25; 7580 Becton Dickinson) counterstained with anti-keratin 5 (1:1000, PRB-160P; BioSite) using the same procedure as for BrdU labeling analysis and label-retaining cell analysis.

β-galactosidase staining

Dorsal skin from wild-type (n = 3) and progeria (n = 3) animals was mounted in OCT (Tissue-Tek) and rapidly frozen in liquid nitrogen before being sectioned into 10-mm sections. The sections were stained according to the manufacturer’s recommendations using the Senescence Cell Histochemical Staining Kit (Sigma-Aldrich). After staining, the sections were washed in PBS and mounted in fluoromount-G (SouthernBiotech, Birmingham, AL, USA) before being analyzed on a bright field microscope. Sections were analyzed by grading microscope fields as strongly stained or not stained. The frequency of strongly stained fields was calculated. A minimum of 52 fields were analyzed per animal.

p63 Immunohistochemistry

Immunohistochemistry for p63 (1:500, 4A4 sc-8431; Santa Cruz Biotechnology, Santa Cruz, CA, USA) was performed on 4-mm paraffin sections of dorsal skin from 5-week-old animals (n = 4 for wild-type and progeria). A pressure cooker and 10 mm sodium citrate (pH 6.5) were used for antigen retrieval, and slides were blocked in peroxidase and mouse-to-mouse blocking reagent (Scytek Laboratories, Logan, Utah, US) following a protocol that was in accordance with previously published procedures (Hanif et al., 2009).

Lymphoblast cultures and cultivation conditions

Epstein-Barr virus (EBV)-transformed lymphoblastoid cell lines from HGPS patients (AG03259, AG10801, AG03344, AG03506, AG10579, AG10587, HGALBV057 and HGALBV009) their unaffected siblings (AG03262, AG03507, AG03263, AG03342, and AG03508) and unaffected adults (HGMLBV066, HGMLBV010, HGMLBV023, HGFLBV021, HGFLBV031, AG03346, AG03505, AG03345, AG10583, HGFLBV067, GM11036, GM13056, GM11037, GM10858, GM10859, GM11870, GM11871, GM13113, GM13114, GM13118, and GM07023) were obtained from Coriell Cell Repositories (Camden, NJ, USA) and the Progeria Research Foundation Cell and Tissue Bank.

The cells were grown in suspension in T75 flasks (Sarstedt). Growth medium was RPMI 1640 (Gibco, Invitrogen) supplemented with 15% heat inactivated (56°C for 30 min) fetal bovine serum, 2 mm L-Glutamine and 1× penicillin- streptomycin (Gibco). Total RNA was isolated from lymphoblasts using TriZol® Reagent (Invitrogen), and the RNA was analyzed by quantitative RT-PCR the same way as isolated keratinocytes.

Statistical analysis

Data were analyzed using unpaired Student’s t-test, and a two-tailed P-value of 0.05 to 0.01 was considered significant (*), a P-value of 0.01 to 0.001 was indicated as **, and a P-value smaller than 0.001 was indicated as ***. For calculating correlation coefficients, linear regression analysis was used.

Acknowledgments

We would like to acknowledge Björn Rozell for pathological competence. We would also like to thank Viljar Jaks and Inderpreet Kaur Sur for technical suggestions regarding keratinocyte isolation and culturing, Konstantin Yakimchuk for technical assistance and advice concerning FACS analysis, Åsa Bergström for technical consultation and Hanna Sagelius for technical assistance. This work was supported by grants to ME from the Swedish Medical Research Council, the Center for Biosciences, Karolinska Institutet, the Marcus Borgström foundation, the Lars Hierta foundation, the Åke Wiberg foundation, and Svenska läkare sällskapet. YR is supported by the Karolinska Institutet KID funding and the Royal Swedish Academy of Sciences. The authors declare no conflict of interest.

Ancillary