Mice with reduced expression of the telomere‐associated protein Ft1 develop p53‐sensitive progeroid traits

Summary Human AKTIP and mouse Ft1 are orthologous ubiquitin E2 variant proteins involved in telomere maintenance and DNA replication. AKTIP also interacts with A‐ and B‐type lamins. These features suggest that Ft1 may be implicated in aging regulatory pathways. Here, we show that cells derived from hypomorph Ft1 mutant (Ft1 kof/kof) mice exhibit telomeric defects and that Ft1 kof/kof animals develop progeroid traits, including impaired growth, skeletal and skin defects, abnormal heart tissue, and sterility. We also demonstrate a genetic interaction between Ft1 and p53. The analysis of mice carrying mutations in both Ft1 and p53 (Ft1 kof/kof ; p53 ko/ko and Ft1 kof/kof ; p53 +/ko) showed that reduction in p53 rescues the progeroid traits of Ft1 mutants, suggesting that they are at least in part caused by a p53‐dependent DNA damage response. Conversely, Ft1 reduction alters lymphomagenesis in p53 mutant mice. These results identify Ft1 as a new player in the aging process and open the way to the analysis of its interactions with other progeria genes using the mouse model.

in human fibroblasts results in senescent phenotypes, including the activation of the p53 pathway, nuclear deformity, heterochromatin alterations, and senescence. In addition, AKTIP reduction affects lamin A expression in human cells (Burla, Carcuro, et al., 2016). Altogether, the properties of AKTIP place this protein at the crossroad of multiple pathways that have been associated with progeroid phenotypes.
The Hutchinson-Gilford progeria syndrome (HGPS) is the bestcharacterized example of progeria, caused by a mutation in exon 11 of the LMNA gene leading to the production of a truncated form of lamin A (De Sandre-Giovannoli et al., 2003). Patients with HGPS develop multiorgan abnormalities, including skeletal defects and absence of subcutaneous fat. They show a limited growth and die in the teenage years, prevalently due to cardiovascular problems leading to infarction or stroke (Hennekam, 2006). Mouse models reflect several aspects of the human disease; the LAKI model, for example, carrying the G608G mutation in the LMNA gene, is characterized by reduced lifespan and body weight, and skeletal and skin defects (Osorio, Navarro, et al., 2011).
The idea that lamins play a pivotal role in determining premature aging is also supported by the discovery of progeroid disorders different from HGPS. For example, restrictive dermopathy patients carry recessive mutations in the ZMPSTE24 gene, which encodes the proteolytic enzyme involved in lamin A maturation (Barrowman, Wiley, Hudon-Miller, Hrycyna & Michaelis, 2012). Also, in this case, a mouse model replicates the progeroid phenotype of the disease (Osorio, Ugalde, et al., 2011). A partial recovery of the ZMPSTE24 À/À phenotype is obtained by depletion of the tumor suppressor protein p53, pointing to a role of DNA damage in the pathophysiology of this progeria (Varela et al., 2005).
In addition to the LMNA gene, several genes involved in DNA metabolism have been implicated in progeroid syndromes.
They include the WRN and BLM genes, which encode members of the RecQ DNA helicase family and are responsible for the Werner and Bloom syndromes, respectively (Bachrati & Hickson, 2003).
Telomere dysfunctions have also been linked to progerias.
The involvement of AKTIP in telomere maintenance and regulation of lamin A (Burla et al., 2015;Burla, Carcuro, et al., 2016) prompted us to investigate whether this protein contributes to preventing premature aging. We thus generated mice bearing a mutation in the Ft1 gene. We report here that Ft1 mutant mice exhibit multiple progeroid traits, including impaired growth, skeletal and skin defects, and sterility. We also demonstrate an interplay between Ft1 and p53. Ft1 mutant mice carrying mutations in p53 (Ft1 kof/kof ;p53 ko/ko and Ft1 kof/kof ;p53 +/ko ) showed a partial rescue of the progeroid traits observed in Ft1 single mutants, suggesting that these traits are at least in part caused by the p53-mediated response to the DNA damage elicited by mutation in Ft1.

| Generation of Ft1 kof mice and characterization of derived MEFs
Given that AKTIP is required for DNA replication and cell proliferation (Burla et al., 2015), we reasoned that full knockout (ko) of Ft1 would cause physiological damage incompatible with mouse survival. Thus, we produced animals with reduced Ft1 levels using the knockout first (kof) strategy, based on the insertion into the target gene (referred as kof allele) of the bgeok cassette (Testa et al., 2004) (Figure 1a), which traps and truncates Ft1 nascent transcript reducing the expression of the gene (Figure 1a). Transgenic founders and subsequent generations were screened by PCR (Figure 1b), and two independent kof lines (lines 107 and 588) were selected and analyzed for mRNA reduction. q-PCR on tail biopsies from Ft1 kof/kof and Ft1 +/kof animals showed that Ft1 expression was significantly reduced compared to wild-type (wt) mice ( Figure 1c).
In Ft1 kof/kof mice from lines 107 and 588, Ft1 expression was reduced to 18% and 12%, respectively; in Ft1 +/kof animals from the same lines, Ft1 expression was reduced to 64% and 52% (Figure 1c). The analysis of 573 F1 progeny from crosses between Ft1 +/kof conformed to Mendelian ratios, although we observed a slight nonsignificant trend of embryonic lethality of Ft1 kof/kof animals ( Figure 1d). We next investigated whether MEFs from Ft1 kof/kof mice exhibit the same phenotypes as those previously observed in RNAi cells depleted of AKTIP or Ft1 (Burla et al., 2015;Burla, Carcuro, et al., 2016). We first checked the Ft1 subcellular localization by immunostaining MEFs with an anti-Ft1 antibody. In human cells, AKTIP is enriched at the nuclear rim where it partially co-localizes with lamins (Burla, Carcuro, et al., 2016). Consistent with these results, Ft1 +/+ MEFs displayed a Ft1 signal at the nuclear periphery (Figure 1e), while the signal was undetectable in the Ft1 kof/kof MEFs (Figure 1e,f).
In line with these results, Western blotting showed a strong reduction of Ft1 in Ft1 kof/kof MEF extracts ( Figure 1g).
We then asked whether MEFs from Ft1 kof/kof mice activate the DNA damage response (DDR) and exhibit telomere defects. Compared to wt MEFs, Ft1 kof/kof MEFs displayed substantial increases in 53BP1 and cH2AX foci, indicating that Ft1 is required for the maintenance of genome integrity (Figure 2a-f). In addition, double immunofluorescence staining of cH2AX and TRF1 showed that Ft1 kof/kof MEFs exhibit a significant increase in cH2AX/TRF1 colabeled foci (Telomere Dysfunction induced Foci, TIFs) compared to control MEFs, suggesting that the DDR of mutant cells was at least in part linked to telomere dysfunction (Figure S1c,d). To determine the nature of telomere defects in Ft1 kof/kof MEFs, we performed in situ hybridization with a TTAGGG probe. The analysis of metaphase spreads showed that Ft1 kof/kof MEFs exhibit multiple telomeric signals (MTS, also known as fragile telomeres) and sister telomere associations (STA) (Figure 2g-i and Figure S1e). These types of telomere aberrations are considered hallmarks of defective telomere replication (Sfeir et al., 2009). Ft1 kof/kof MEFs showed a small but not statistically significant increase in telomere fusions compared to matched control MEFs. In addition, Ft1 kof/kof MEFs displayed a frequency of telomeres with a TTAGGG signal comparable to that of control, confirming (Burla et al., 2015) that an impairment of the Ft1 function does not result in telomere loss ( Figure S1a,b).
Finally, we evaluated the status of lamin A in MEFs. Consistent with our previous results on human AKTIP (Burla, Carcuro, et al., 2016), wt cells displayed partial co-localization of lamin A with Ft1 ( Figure S1f). Altogether, our results indicate that Ft1 kof mutation cause DDR, telomere defects and abnormal lamin distribution, which are well-known hallmarks of aging. We thus asked whether Ft1 kof mutant mice exhibit signs of premature aging. In mouse models of progeroid disorders, premature aging alterations mostly affect body growth, fertility, bones, skin and heart. We therefore focused on these phenotypic traits in our analyses on Ft1 kof mutant mice.

| Ft1 kof mice display growth defects, reduced lifespan and sterility
Macroscopic observation of Ft1 kof/kof mice revealed that mutant animals (n = 170) display a significant reduction in body weight compared to controls (Figure 3a-c); 21% of the animals showed a 30% reduction in body weight compared to controls; henceforth, we will F I G U R E 1 Generation of Ft1 kof/kof mice. (a) wt allele (+) of Ft1 and kof cassette inserted into the gene to generate animals with reduced Ft1 expression. F, forward; R, reverse; Frt, target site for FLP recombinase; loxP, target site for Cre recombinase; SA, splicing acceptor site element from engrailed 2; IRES, internal ribosomal entry site from the Encephalomyocarditis virus; lacZ, b-galactosidase gene; neo, neomycin phosphotransferase, selectable marker. (b) PCR of gDNA from Ft1 +/+ , Ft1 +/kof , and Ft1 kof/kof animals. (c) q-PCR on tail cDNA of animals from two independent mouse lines. Results present the ratios between Ft1 exons 6 and 3 normalized to GAPDH. Ft1 kof/kof mice n = 53, Ft1 +/kof animals n = 15, and wt n = 30. ***p < .001 in Student's t test. (d) Mendelian distribution; no significant difference between observed and expected Ft1 kof/kof genotypes (p = .056 v 2 test). (e) MEFs stained with anti-Ft1 antibody (red in merges). Scale bar 5 lm.  refer to these mice as severely affected Ft1 kof/kof mice, abbreviated with SA Ft1 kof/kof or SA mutant mice. By selecting a cohort of animals with a mild (non-SA) phenotype, we monitored body weight over a 100-week period and subdivided lifespan in three major intervals: young 3 < weeks < 20; juvenile 21 < weeks < 60; adult 61 < weeks < 100. We observed that the difference in body weight between wt and Ft1 kof/kof animals significantly increases as mice age (Figure 3b,c). Ft1 kof/kof mice had a reduced lifespan compared to wt, SA Ft1 kof/kof animals died at day 12-14, while the remaining population displayed a median survival of 113 weeks ( Figure 3d). We also observed reduced fertility in inbreeding; when Ft1 kof/kof non-SA males were crossed with Ft1 +/kof females, we did not observe any pregnancies ( Figure 3e). Altogether, these observations show that Ft1 expression is critical for mouse growth, survival and fertility.

| Ft1 kof mice display skeletal alterations
Bone is altered in progeroid patients and mouse models for progeria syndromes (Bergo et al., 2002;Mounkes, Kozlov, Hernandez, Sullivan & Stewart, 2003;Osorio, Navarro, et al., 2011). Radiographic analyses of whole skeletons were collected at day 12 from eight SA Ft1 kof/kof mice and three wt animals. X-ray images showed reduced skeleton  Table S2 F I G U R E 2 DNA damage, telomere aberrations and lamin A alterations in Ft1 kof/kof MEFs. (a-c) Staining for anti-53BP1 (red in merges) in MEFs (a) and quantification (b, c). (d-f) Staining with anti-cH2AX (red in merges) in MEFs (d) and quantification (e, f). Graphs (b, c, e, f) show mean AE SEM; **p < .01; ***p < .001 in Student's t test. Scale bars 5 lm. (g) Partial DAPI-stained (red) metaphases from MEFs showing telomeric FISH signals (black and white; green in merges) including enlargement of single chromosomes with multiple telomeric signals (MTS). MTS are indicated by arrows. (h, i) MTS (h) and STA (i) frequencies in MEFs. Graphs show mean AE SEM; *p < .05; ***p < .001 in v 2 test. (j) Immunostaining for lamin A in MEF nuclei. Scale bar 5 lm. (k) Quantification of lamin A signal at nuclear rim in MEFs. Graphs show mean AE SEM; *p < .05 in Student's t test from two independent experiments on two MEF cultures (n = 100 cells/culture). (l) Z stack projections and quantification showing the altered distribution of lamin A in Ft1 kof/kof MEF nuclei. See also Figure S1 LA TORRE ET AL.
| 5 of 13 regular columnar and conjugational cartilage, although slightly shorter than controls (Figure 4g). In mutant mice, newly formed bone trabeculae were also shorter, with a poorly defined osteoblastic rim, as compared to wt (Figure 4h). TRAP cytochemistry did not reveal significant differences in osteoclast numbers relative to bone surfaces between Ft1 kof/kof samples and controls, suggesting that the F I G U R E 4 Bone alterations in Ft1 kof/kof mice. (a) X-ray on total body of Ft1 kof/kof and wt mice at day 12. Arrowheads indicate spine defects and craniofacial dysmorphisms in Ft1 kof/kof animals. (b) Quantification of the angle formed by the cervical and thoracic vertebrae; ***p < .001 in Student's t test. (c) X-ray images of femurs and tibias from Ft1 kof/kof and wt at day 12. (d) X-ray of tail and magnification of caudal vertebrae (dotted white box) of Ft1 kof/kof and wt at day 12. (e, f) Pseudocolored femur images (e) and relative quantification (f) showing that X-ray absorption is lower in Ft1 kof/kof as compared to wt. Student's t test, *p < .05. (g) H&E-stained sections of caudal vertebrae (top) and femurs (bottom) of Ft1 kof/kof and wt at day 12. (h, i) TRAP (h) and relative quantification (i) on femur sections shows no significant differences in TRAP-positive cells between Ft1 kof/kof and matched wt (p = .67 Student's t test). See also Table S2 osteopenic defects cannot be ascribed to increased osteoclastogenesis ( Figure 4i).
Altogether, these results show that mutations in Ft1 cause bone defects that partially phenocopy those observed in progeroid models caused by mutations in lamin coding genes or in genes involved in DNA metabolism (Bergo et al., 2002;Chen et al., 2012;Saeed et al., 2011).

| Ft1 kof animals display skin and heart alterations
Several studies have shown that skin and heart are typically altered in premature aging disorders associated with impaired DNA metabolism, lamin, or telomere defects (Bergo et al., 2002;Cao & Hegele, 2003;Mounkes et al., 2003;Watson et al., 2013). We found that in SA Ft1 kof/kof mice adipose tissue deposits are strongly reduced compared to age-matched controls (Figure 5a). The analysis on H&E-stained skin sections clearly showed the absence of subcutaneous fat layer in SA mutant animals, a defect similar to the skin defects described in mice carrying mutations in the lamin A coding gene (Mounkes et al., 2003) (Figure 5b).
The heart of SA Ft1 kof/kof mice was smaller than in controls, with a reduction in size proportional to the overall body reduction (Figure 5c). In addition, analysis of H&E-stained hearts of Ft1 kof/kof animals and wt mice showed a difference in tissue architecture. In Error bars indicate SEM; ***p < .001 in Student's t test. (g) q-PCR quantification of the p21 senescence marker expression in wt and matched Ft1 kof/kof mice at day 12 and at 6 months after birth. *p < .05 and ***p < .001 in Student's t test. See also Table S2 hearts of SA Ft1 kof/kof animals, there was no apparent fibrotic tissue and the number of nuclei per area was higher than in wt hearts, suggesting an increase in the nuclear/cytoplasmic ratio (Figure 5c,d).
To gain additional insight into the origin of the defects detected in SA Ft1 kof/kof hearts, we immunostained heart sections for cH2AX to reveal DNA damage foci. In mutant hearts, the frequency of cells with more than 5 foci was significantly higher than in controls, indicating DDR activation (Figure 5e,f). We also investigated whether mutant hearts express the p21 DDR and senescence marker to a higher extent than to control hearts. q-PCR analysis revealed that p21 is upregulated in heart extracts from both 12 days SA mutants and 6-month non-SA Ft1 kof/kof mice compared to age-matched controls ( Figure 5g).
Collectively, these results indicate that mutations in Ft1 affect the skin and heart structural organization, and activate DDR and senescence pathways.

| p53 and Ft1 genetically interact
p21 is a potent inhibitor of cyclin-dependent kinase (CDK) that mediates p53-dependent cell cycle arrest in response to DNA damage; it has been shown that p21 is activated by p53. We thus asked whether p53 contributes to the phenotypic traits observed in Ft1 kof/kof mice. To test this possibility, we generated Ft1;p53 double mutant using p53 ko mice (Jacks et al., 1994). Double-mutant mice were examined for several phenotypic traits, particularly those affected in Ft1 kof/kof single mutants.
We first analyzed fertility of mutant animals. In contrast with Ft1 kof/kof male mice that were sterile, Ft1 kof/kof ; p53 +/ko and Ft1 kof/kof ; p53 ko/ko male mice gave progeny when crossed to wt females, indicating that mutation in one or both p53 alleles rescues sterility (Figure 6a). We next examined body weight and survival; the body weight deficiency observed in Ft1 kof/kof mice was rescued in Ft1 kof/kof ;p53 +/ko mutants at least until the 24th postnatal week.

Analysis of death causes revealed a further interplay between
Ft1 and p53 (Figure 6d,e and Figure S3). Homozygosity for Ft1 kof did not result in malignant tumors, and Ft1 kof mutations were modestly cancer-protective in a p53 ko background (Table S1 and Figure S3).  Table S1). In addition Mice with reduced levels of Ft1, both in the presence or absence of p53, appeared to be sensitive to other pathologies, including hepatitis, bone marrow aplasia, peritonitis, nephritis, and pneumonia (Figure 6e, and Table S1 and Figure S3h,i). Thus p53 deficiency in Ft1 kof/kof mutant mice rescues the sterility and the reduced body weight phenotypes, but a concomitant deficiency of p53 and Ft1 affects lymphomagenesis.

| Ft1 mutant cells are sensitive to DNA damaging agents
The DDR foci observed in the MEFs and heart of Ft1 kof/kof mice, and the telomeric aberrations found in Ft1 kof/kof MEFs suggest that Ft1 mutant cells might be defective in DNA repair. To address this issue, we determined the sensitivity of Ft1 kof/kof MEFs to DNA damaging agents. We exposed Ft1 kof/kof MEFs to nonlethal doses of the radiomimetic compound bleomycin, which creates DNA double-strand breaks (DSBs). Cell density assessment at 10 days after treatment showed that Ft1 kof/kof MEFs are significantly more sensitive to the drug compared to passage-matched wt MEFs ( Figure S4a). Increased sensitivity of Ft1 kof/kof MEFs to DNA damage was also observed after treatment with hydroxyurea, which depletes the cells of dNTPs, generating stalled replication forks that can collapse into DSBs ( MEFs compared to wt MEFs. We also observed an excessive doublings of Ft1 kof/kof ;p53 +/ko MEFs with respect to Ft1 +/+ ;p53 +/ko cells ( Figure 6i). An increase in the proliferation rate of MEFs bearing mutation in p53 has been reported previously (Lang et al., 2004;Ma, Choudhury, Hua, Dai & Li, 2013).
Collectively, these results suggest that Ft1 deficiency renders cells more sensitive to DNA damaging agents, resulting in proliferation defects that are (over) rescued by the presence of a single p53 ko mutant allele.

| DISCUSSION
Human progeroid syndromes and their related animal models have been instrumental to identify factors involved in normal human aging. The cellular defects found in progeroid diseases that also characterize normal human aging include DNA damage and genome instability, telomere attrition, epigenetic alterations of histones, aberrations in the nuclear lamina, and cell senescence (de Boer et al., 2002;Liu et al., 2005;Osorio, Ugalde, et al., 2011;Varela et al., 2005).
Here we analyzed the cellular, developmental, and physiological phenotypes of Ft1 mutant mice, focusing on traits related to the aging process. Importantly, our analysis of MEFs from Ft1 kof/kof mice confirmed and extended our previous results obtained on the mouse and human cells depleted of Ft1 or AKTIP (Burla et al., 2015;Burla, Carcuro, et al., 2016). Specifically, we showed that Ft1 kof/kof mutant MEFs exhibit fragile telomeres and sister telomere associations, TIFs, DNA repair foci, increased sensitivity to bleomycin and hydroxyurea, and reduced cell proliferation. In addition, we confirmed that in Ft1 mutant MEFs there is an alteration in lamin A, resulting in a strong F I G U R E 6 Ft1 kof/kof mouse phenotype is p53 sensitive. (a) Pups generated by animals bearing mutations in Ft1 and/or p53; ***p < .001 in Student's t test. Whiskers represent the minimum and the maximum values observed for each mating and the boxes the 25th to the 75th percentile. (b) Body weight in Ft1 kof/kof animals in the presence or absence of a null mutation in p53; note that loss of a single p53 allele dominantly rescues the Ft1-dependent body weight reduction; *p < .05; ***p < .001 in Student's t test. (c) Survival of Ft1 kof/kof ;p53 +/ko mice is decreased compared to that of Ft1 +/kof ;p53 +/ko and Ft1 +/+ ;p53 +/ko animals (p < .001-log-rank-Mantel-Cox test). (d, e) Case analysis on wt mice and mice bearing mutations in p53 (p53 ko/ko ) and Ft1, showing that Ft1 mutation impacts on lymphomagenesis and inflammatory conditions. See also Figure S3 and Table S1. (f, g) Cell survival response in MEFs from Ft1 +/+ ;p53 +/+ , Ft1 kof/kof ;p53 +/+ , and Ft1 kof/kof ;p53 +/ko mice upon increasing doses of bleomycin (f) or hydroxyurea (HU) (g) showing that cells homozygous for mutations in Ft1 and bearing a null mutation in p53 (p53 +/ko ) are less sensitive to DNA damage than Ft1 mutant cells bearing two wt copies of p53. Graphs show mean AE SEM; **p < .01 in Student's t test. (h) Western blotting analysis of p21 and p53 expression in Ft1 +/+ ; p53 +/+ , Ft1 kof/kof ; p53 +/+ , and Ft1 kof/kof ;p53 +/À . (i) Population doubling (pd), showing that Ft1 kof/kof MEFs have a reduced pd compared to Ft1 +/+ cells; this phenotype is rescued by a p53 ko mutation. Each dot represents the mean AE SEM of the cumulative pd at the indicated day; *p < .05, ***p < .001 in Student's t test. See also Figures S4 and S5 LA TORRE ET AL. | 9 of 13 reduction in the lamin nuclear rim. Thus, Ft1 mutant MEFs display many traits that have been previously observed in progeroid syndromes and progeroid animal models, as well as in normal human aging.

Consistent with the results on mutant MEFs, our analysis of
Ft1 kof/kof mutant animals detected progeroid phenotypes. Ft1 kof/kof mice displayed multiple traits that have been previously observed in several progeroid models (Table S2). We found that Ft1 mutant mice have reduced body weight, fertility defects, and reduced lifespan, as previously observed in models of laminopathies (Bergo et al., 2002;Osorio, Navarro, et al., 2011) and telomeropathies (Mart ınez et al., 2009). In addition, the growth defects of Ft1 kof/kof mice were exacerbated with aging, suggesting that the effects of Ft1 mutations intercept the normal aging-induced degeneration pathways. Ft1 kof/kof mice also displayed skin and bone defects, which were previously observed in lamin mutant mice (Bergo et al., 2002;Mounkes et al., 2003;Osorio, Navarro, et al., 2011), in Tert ko animals (Rudolph et al., 1999), and in mice with reduced Trf1 expression (Mart ınez et al., 2009).

Skeletal alterations and lipodystrophy have been imputed to failures
in the proliferation of mesenchymal stem cell progenitors, which are sensitive to lamin mutations and senescence (Scaffidi & Misteli, 2008). Mutant hearts were smaller than those of wt animals and showed a higher nuclear density compared to wt, with an increase in the nuclear/cytoplasmic ratio. In addition, we found that mutant hearts display DNA damage and activation of the DDR, and up-regulation of p21 expression. The relationships between increased DNA damage and a change in nuclear density in the mutant hearts are unclear. A possible explanation is that DNA damage and the related inflammation process induce local cell reprogramming. This explanation is consistent with the observation that cellular reprogramming in vivo occurs following tissue injury (Yanger et al., 2013).
The fact that the organismal phenotypes observed in Ft1 mutant animals have also been found in models specifically defective in lamin structure and/or expression, or bearing mutations in genes required for telomere maintenance or DNA repair, poses an interesting question. Which of the cellular phenotypes observed in Ft1 mutant MEFs (defective lamin behavior, telomere dysfunction, DNA damage) is responsible for the organismal progeroid phenotypes?
Answering this question is difficult because the traits that characterize Ft1 mutants at the cellular level are deeply interconnected. For example, alterations in lamin function affect DNA replication and repair, epigenetic modification of chromatin and transcription (Gonzalo & Kreienkamp, 2015). Moreover, multiple interactions link telomeres to the lamin network, including the association of telomeres with the nuclear envelope .
Finally, telomeres recruit and interact with many DNA repair factors, which play crucial functions in telomere maintenance (Doksani & de Lange, 2014). Thus, current information does not allow identification of the specific cellular phenotype that leads to progeroid traits observed in Ft1 mutant mice. The most likely hypothesis is that all cellular defects observed in Ft1 mutant MEFs contribute to the organismal phenotype of mutant animals. It is indeed quite possible that these defects lead to senescence in most if not all tissues, causing developmental defects and infertility.

| Relationships between Ft1, p53, and cancer
We have shown that p53 deficiency in Ft1 mutant MEFs induces cell over proliferation and rescues the sensitivity to both bleomycin and hydroxyurea. Consistent with these findings, in Ft1 kof/kof mutant mice, mutations in p53 rescue the body weight and sterility phenotypes but do not improve survival. Impairment of the p53 function also ameliorates the progeroid phenotypes in BRCA1-deficient mice (Cao, Li, Kim, Brodie & Deng, 2003) and in HGPS mouse models (Varela et al., 2005). However, p53 deficiency worsens the progeroid phenotype in telomere dysfunctional mice (Begus-Nahrmann et al., 2009). An explanation for this discrepancy is that p53 deficiency allows beneficial propagation of damaged cells rescuing certain progeroid traits. However, when cellular damage is extensive and the regenerative capacity of tissues is severely limited, p53 deficiency would become deleterious and accelerate aging (Lopez-Otin, Blasco, Partridge, Serrano & Kroemer, 2013). Our results are consistent with this model; they indicate that mutations in Ft1 result in a relatively mild genomic damage that triggers DDR-related checkpoints, which are abolished by mutations in p53 allowing resumption of cell proliferation.
The relationships between mutations in progeria-related genes and cancer are also complex. Progeroid models have been used to study the interplay between aging and cancer, given that age is a major risk factor for cancer developing. It has been shown that some forms of progeria can exert a protective role against tumor development (de la Rosa et al., 2013). On the other hand, mutations in the WRN helicase causing a segmental progeroid syndrome have been associated with an elevated cancer risk (Blander et al., 2000). We found that Ft1 kof mutation does not induce cancer and that p53 Ft1 double-mutant mice do not exhibit an increase in the overall frequency of malignancies. However, p53 ko combined with Ft1 deficiency induced an increase in the diffusion of lymphomas as compared to the restricted localization of this type of tumor in p53 ko mice. It has been reported that T-cell lymphomas in p53 ko mice are oligoclonal and generated by a characterized sequence of mutational events (Dudgeon et al., 2014). We therefore hypothesize that in p53 Ft1 double mutants, this sequence is altered causing multiclonality and/or histotype change of lymphomas.
In conclusion, we have shown that mutations in Ft1 affect lamin, telomeres, DNA repair, and cell senescence. At the organismal level, these mutations result in a number of phenotypes that have been previously observed in several progeria mouse models. Thus, we believe that Ft1 is a new player in both the normal and accelerated aging processes, and that Ft1 mutant mice will be instrumental to analyze the interactions between Ft1 and other mouse progeria genes.

| Mice
ES (HEPD0589_6_H06) from C57Bl/6 animals carrying the knockout first mutations in the Ft1 gene (referred as Ft1 kof) were generated by the International Knock-out mouse consortium. Injections into C57Bl/6 blastocyst were performed in EMBL (Monterotondo, Italy).
Offspring were weaned at 3 weeks, and tail biopsies were genotyped and transgene expression analyzed. When needed, mice were anesthetized by intramuscular Zoletil 20 (Virbac S.A., France), or euthanized by asphyxiation with carbon dioxide or cervical dislocation.

| Cells
MEFs were isolated and cultured as described in Rinaldo et al. (2012). Population doubling (pd) was calculated with the formula Log (n t /n 0 ) 9 3.33. For Bleomycin and Hydroxyurea sensitivity assay, cells plated 24 hr in advance were treated with Bleomycin (Sanofi Aventis) or Hydroxyurea (Sigma) for 7 hr and replaced with medium w/o drugs. Cell density was calculated 10 days after treatment by staining with crystal violet (5% in methanol, Sigma) for 10 min and analyzed by IMAGEJ.

| Genotyping
Tail biopsies were digested overnight at 56°C with a proteinase K/ SDS solution; genomic DNA (gDNA) was extracted using the NucleoSpin â Tissue columns kit (Macherey-Nagel, Duren, Germany) following manufacturer's instructions. Mice were PCR genotyped using the following primers:

| q-PCR
RNA was extracted using the TRIzol reagent (Invitrogen) according to manufacturer, after DNaseI treatment (Invitrogen) was reverse transcribed into cDNA with oligo d(T) primer and OMNISCRIPT RT KIT (Qiagen). q-PCRs were performed as described (Burla et al., 2015) using following primers:

| Western blotting
Western blotting was carried out as described in Burla et al. (2015).

| Histology, immunohistochemistry, and TRAP
Skin, bone, and heart were fixed in 4% formaldehyde. Tissues were cleared with ascendant alcohol concentration, embedded in paraffin, and sectioned at 3.5 lm. Sections were hydrated with descendant alcohol concentration, stained with Hematoxylin (Carlo Erba) and Eosin (Sigma), cleared, and mounted with DPX Mountant for Histology (Sigma). For cH2AX analysis on paraffin, embedded heart sections were treated as previously described (Martinez, Ferrara-Romeo, Flores & Blasco, 2014). Tissues were incubated overnight with an anti-cH2AX (Abcam 2893) diluted in BSA 3%, Triton X-100 0.1%, and the day after incubated for 1 hr at room temperature with the LA TORRE ET AL. | 11 of 13 pertinent secondary antibody (anti-rabbit-ALEXA 555, Invitrogen A21430). Slides were counterstained with Mayer hematoxilin (Carlo Erba) and mounted with DPX mounting solution for microscopic evaluation (Sigma). Pictures were taken with ZEISS-Axio Phot (Zeiss) microscope connected to Progress-C5 JENO-PTIK camera with the software PROGRESS MAC (Capture PRO). TRAP staining was performed according to manufacturer's instructions (Sigma 387A).

| X-ray and bone density analysis
Total body X-ray images were taken using Faxitron MX-20 (Faxitron X-ray Corp.) at 24 kV for 6 s; images captured with Medical Imaging Film HM Plus (Ferrania). Cervical-thoracic vertebrae angle quantification was measured with Photoshop CS6 plugin. Femurs were imaged using a Faxitron MX20 operating at 24 kV for 4 s. Image density was determined as previously described (Bassett, van der Spek, Gogakos & Williams, 2012).

| Statistics
Kaplan-Meier curves were analyzed using the log-rank (Mantel-Cox) test. Inheritance of kof allele was analyzed using the Mendelian ratio for heterozygous mating, and v 2 test was applied. The Kolmogorov-Smirnov test was used to compare gray-level cumulative frequency distributions in X-ray image quantification. Independent data sets were compared with the Student's t test (unpaired, two-tailed).

ACKNOWLEDGMENTS
This work has been supported by Grants EU FP7 Brainvectors (no. 286071), Telethon GEP15033 and PRF 2016-67 to IS; by AIRC IG 2014 to MG. This work is in the memory of P. Bianco.

CONF LICT OF I NTEREST
We have no conflict of interest.