Patients signed approved consent forms to participate in the studies of the International Registry of Werner Syndrome (Seattle, WA, USA) prior to the initiation of the study. This study was approved by the Human Subjects Division, Institutional Review Board of the University of Washington, Seattle, WA, USA.

Human diploid fibroblast cell lines were established from biopsies of skin samples. Cell lines 82-6 and 88-1 (normal) were obtained from newborn foreskin and the abdominal skin of a 62 year old female, respectively. Fibroblasts were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Thermo Scientific Hyclone, OH, USA) and penicillin and streptomycin (Invitrogen), as previously described (Chen et al., 2003). The media was changed every 3–4 days. All cells were cultured at 37°C in 5% CO₂. Primary fibroblasts were immortalized with a construct encoding hTERT (Rubio et al., 2002). MG-132 (10 μM, Calbiochem, La Jolla, CA, USA) and cycloheximide (100 μg/ml, Sigma, St. Louis, MO, USA) was used when treating fibroblasts.

To determine the fraction of proliferating cells, a 5-bromo-2′-deoxyuridine (BrdU) incorporation assay was carried out using a BrdU labeling and detection kit (11 296 736 001, Roche Applied Science, Indianapolis, IN, USA) according to the manufacturer’s instructions. Briefly, primary fibroblasts from control, LMNA mutant and WRN mutant cell lines were grown for 24 h on cover slips. Cells were then labeled with BrdU (10 μM) for 24 or 48 h, fixed in 70% ethanol with 50 mM glycine for 20 min at -20°C and processed for immunofluorescence. Cells were then incubated with the primary antibody, anti-BrdU (mouse monoclonal antibody, clone BMG 6H8 IgG1, 1:10 dilution) for 30 min at 37°C followed by Alexa Fluor 594-conjugated goat anti-mouse antibody (1:200 dilution; A-11005, Molecular Probes, Eugene, OR, USA) for 30 min at 37°C. Cells were mounted onto slides using Vectashield with DAPI (Vectashield, Burlingame, CA, USA) and observed and photographed with Nikon Upright (Nikon Eclipse E600) microscope at the Keck Center for Imaging, University of Washington, Seattle, WA, USA. The percentages of cells with BrdU-labeled nuclei were determined by counting 300 cells per data point.

For western blot analyses, cells were lysed in 2× Laemmli buffer [0.5 M Tris pH 6.8, 10% sodium dodecyl sulfate (SDS) glycerol]. Fifteen micrograms of total protein isolated from primary fibroblasts or immortalized fibroblast cell lines were separated on NuPAGE 4–12% gradient Bis–Tris gel (Invitrogen) and transferred to polyvinylidene fluoride (PVDF) membranes. The membranes were incubated with anti-TRF2 (IMG-124A, clone 4a794.15, Imgenex, San Diego, CA, USA and sc-32106, Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), anti-TRF1 (sc-5596, Santa Cruz Biotechnology Inc), anti-POT1 (sc-27952, Santa Cruz Biotechnology Inc.), anti-TIN2 (IMG-282, clone 59B388.1, Imgenex), anti-hRap1 (IMG-289, Imgenex), anti-TPP1 (ab54685, Abcam Inc., Cambridge, MA, USA) at a dilution of 1:2000 or anti-β-actin (clone AC-15, Sigma, St Louis, MO, USA) at a dilution of 1:40,000 in a milk blocking solution, then with corresponding biotinylated secondary antibodies against mouse, rabbit or goat IgG (Vector Laboratories, Burlingame, CA, USA). The reactions were visualized with western lightening chemiluminescence reagent (NEL100, Perkin Elmer, Waltham, MA, USA) according to the manufacturer’s instruction. The bands corresponding to TRF2, TRF1, POT1, TIN2, hRap1, and TPP1 were quantified using TotalLab software (Newcastle, UK) and normalized against β-actin levels.

For the measurement of the TRF2 protein half-life, immortalized fibroblast cell lines from the control, LMNA mutant and WRN mutant were treated with cycloheximide (100 μg/ml) for 8–12 h. Total protein extracts were subjected to the western analysis of TRF2 and β-actin as described above. After normalizing to β-actin levels, TRF2 levels were calculated in order to estimate the protein half-lives.

Quantitative reverse transcription polymerase chain reaction (RT-PCR) for TRF2 mRNA was performed using Taqman gene expression assay system (Hs00194619_m1, Applied Biosystem, Foster City, CA, USA) and normalized against mRNA level of GAPDH (Hs02758991_g1, Applied Biosystem). Total RNA was isolated from the cultured cell pellet using Qiagen RNA isolation kit (Qiagen, Valencia, CA, USA) following the manufacturer’s instructions. Reverse transcription was carried out with 2 μg isolated total RNA. Quantitative PCR was performed with a Rotor-Gene 3000 (Corbett Research, Sydney, Australia). The data analysis was performed as previously described (Huang et al., 2008).

Mean telomere lengths were determined by qPCR as previously described (Huang et al., 2008). Genomic DNA was isolated via a salting out method and qPCR was performed with a Rotor-Gene 3000 (Corbett Research, Sydney, Australia). The data were analyzed as previously described (Huang et al., 2008). The intra-assay and inter-assay variability (CV) for qPCR was 3 and 5%, respectively.

Telomerase activity in hTERT-immortalized cell lines was measured using the TRAPEZE XL Telomerase detection kit (S7707, Millepore, Billerica, MA, USA). Cells were lysed in ice-cold 3-[(3-cholamidopropyl)dimethylammonio]-1-propane-sulfonate (CHAPS) lysis buffer for 30 min. 2 μl of the cell extract was analyzed for the telomerase repeat amplification protocol (TRAP) assay. In the first step telomeric repeats were added to a substrate oligonucleotide by telomerase; then, the extended products were amplified by Taq polymerase using PCR with TS and RP (reverse) primers, generating a ladder of products in 6 base increments. The PCR products were separated on a 10% non-denaturing polyacrylamide gel (PAGE). The gel was stained with SYBR Safe (S33102, Invitrogen) to visualize the TRAP ladder using FluorChem Q imaging system (Alpha Innotech, San Leandro, CA, USA).

Immuno-fluorescent in situ hybridization was performed as described previously (O’Sullivan et al., 2004) with modifications. Cells were grown on coverslips overnight and then fixed the next day with 3.7% paraformaldehyde in phosphate buffered saline (PBS; pH 7.4) plus 0.2% Triton X100 for 10 min at room temperature. Cells were blocked for 30 min at 37°C and incubated with primary antibody, mouse anti-TRF2 (IMG-124A, clone 4a794.15, Imgenex) or rabbit anti-γ-H2AX (2577, Cell Signaling Technologies, Beverly, MA, USA) for 30 min at 37°C. They were then incubated with secondary antibody, donkey anti-mouse Alexa Fluor or donkey anti-rabbit Alexa Fluor (Invitrogen Molecular Probes, Eugene, OR, USA) for 1 h at room temperature, and then post-fixed with 3.7% paraformaldehyde for 5 min. After washing in PBS, cells were processed for FISH. Cells were treated with 100 μg/ml of RNase A (Sigma-Aldrich, St. Louise, MO, USA) in PBS for 45 min. at 37°C, then dehydrated in an ethanol series and air dried. Peptide nucleic acid (PNA) probe (200 nM, Cent-FAM, TelC-FITC, or TelC-Cy3; Panagene Inc., Daejeon, Korea) in hybridization buffer was placed on top of the coverslips, heated to 75–78°C for 15 min, then removed, wrapped in foil and allowed to sit at room temperature overnight. The next day the coverslips were washed with 70% formamide in Tris buffer, washed with preheated (55–60°C) Tween 20 solution, dehydrated in an ethanol series and then air dried. Coverslips were mounted on slides with DAPI (Vectashield) and sealed with clear nail polish.

The images were obtained in Z-series with Deltavision deconvolution microscope (Applied Precision, Issaquah, WA, USA) at the Keck Center (University of Washington, Seattle, WA, USA). The images were deconvolved using SoftWorx software (Applied Precision, Issaquah, WA, USA). They were analyzed for average telomere intensity (which is proportional to telomere length), and average TRF2 intensity using a software program, 4xColocalization, version 138, provided by the University of Washington Nathan Shock Center, as described previously (O’Sullivan et al., 2004). The images were also analyzed for co-localization (telomere and TRF2; telomere and γ-H2AX; TRF2 and γ-H2AX). Statistical significance was determined by the Student t-test (Chen et al., 2003; Huang et al., 2008).

To assess the degree of nuclear irregularity, nuclear contour ratios (NCRs) were determined in randomly selected nuclei of LMNA mutant fibroblasts using MetaMorph software as described previously (Saha et al., 2010). The nuclei were also analyzed for average TRF2 intensities. The correlation between NCR and TRF2 intensities was analyzed using Pearson’s correlation coefficient, r. For all experiments, a two-tailed P value of <0.05 was considered significant.

We first analyzed the steady state protein expression levels of the shelterin subunits TRF2, TRF1, POT1, TIN2, TPP1, and Rap1 in primary fibroblasts carrying a heterozygous R133L mutation derived from Registry# PORTU8010 and in primary fibroblasts carrying a heterozygous L140R mutation derived from Registry# NORWAY1010 (Chen et al., 2003). The controls were two normal primary fibroblast cultures, 82-6 and 88-1 (Oshima et al., 1995; Huang et al., 2008) as well as cultures of WRN mutant fibroblasts, Registry# MCI7885, derived from a patient with classical Werner syndrome (Oshima et al., 1995; Yu et al., 1996). The latter was included as a “control” since these fibroblasts undergo accelerated telomere shortening by different mechanisms, namely defects in lagging strand synthesis of telomeres (Crabbe et al., 2004). We examined fibroblasts in early passages (BrdU labeling index higher than 60%) in order to detect proximal changes rather than changes secondary to telomere shortening and replicative senescence.

Western analysis showed that most subunits of the shelterin complex were reduced in both types of LMNA mutant fibroblasts (Figure 1A). Steady state levels of TRF2 were 46 and 45%, POT1 was 57 and 56%, TIN2 was 36 and 35%, and Rap1 was 46 and 40%, in R133L and L140R mutants, respectively, compared to the control, 82-6 fibroblasts (Figure 1B). The other two components, TRF1 and TPP1, were also slightly reduced to 73 and 62% in R133L mutants and 74 and 92% in L140R mutants (Table 1). Interestingly, WRN mutant fibroblasts also showed reduced protein expression of all of the shelterin components.

Since the length of the telomere determines the number of available binding sites for the shelterin complex, cells with short telomeres (i.e., late passage or senescent cells) would require and presumably express lower amounts of the shelterin components. To rule out this possibility, relative telomere lengths were measured by quantitation of fluorescence in situ hybridization images (Q-FISH), as established for normal fibroblasts undergoing replicative senescence (O’Sullivan et al., 2004). Using this method, we found that the telomere lengths of young LMNA mutant fibroblast cultures were comparable to those of controls after normalizing to centromere signal intensities (Figure 1C). Mean telomere lengths were also measured by quantitative PCR, showing no significant differences among LMNA mutant fibroblasts and controls (Figure 1C). The mean telomere lengths of the two controls were similar despite the difference in the age of donors, 82-6 (newborn) and 88-1 (adult; Figure 1C). This may be due to the fact that there is a wide range of mean telomere lengths and replicative life spans among normal individuals. The labeling index determined by the 48 h BrdU incorporation and population doubling (PD) of the culture were 94% in 82-6 (PD 9), 89% in 88-1 (PD 7), 71% in R133L (PD 5), 96% in L140R (PD 6), and 91% in MCI7885 (PD 6), confirming that the LMNA mutant fibroblast cultures used in these experiments were indeed from young passages. This indicates that the reduction of shelterin proteins observed in LMNA mutants was not a direct result of shortened telomeres.

Next, we focused on the role of TRF2, a major component of the shelterin complex, in preventing telomere erosion. To determine whether the reduction of TRF2 proteins in LMNA mutant fibroblasts occurs via a transcriptional mechanism or post-transcriptional degradation, we analyzed TRF2 expression at the mRNA level. The qRT-PCR results showed that the steady state levels of TRF2 mRNA in LMNA mutant fibroblasts were not significantly different when compared to the controls (Figure 1D). These results suggest that the TRF2 protein may undergo degradation in LMNA mutant cells, resulting in low TRF2 protein levels while the cells still retain long telomeres.

We previously reported that primary LMNA mutant fibroblasts, but not control fibroblasts or immortalized LMNA mutant cells, exhibit abnormal nuclear morphology, as assessed by NCR (Chen et al., 2003). In control fibroblasts, 82-6, the contour ratio was 0.842 ± 0.036, whereas in R133L the contour ratio was 0.749 ± 0.118 (P < 0.005); for L140R, the ratio was 0.815 ± 0.074 (P < 0.05). We noted that within the LMNA mutant cultures, the cells with more distorted nuclei tended to have lower levels of TRF2 (Figure 2A). Quantitative analysis revealed a significant positive correlation between NCR and TRF2 protein level per nucleus in both R133L and L140R mutant cultures: Pearson’s correlation coefficient, r, was 0.365 for R133L (P < 0.005) and was 0.379 for L140R (P < 0.05; Table 2). The presence of a correlation between and NCR and TRF2 levels supports the idea that abnormal lamin A/C function, as reflected by abnormal nuclear morphology, may be causally related to the TRF2 degradation. As expected, control fibroblast cultures did not show significantly abnormal NCRs.

Next, we assessed the intranuclear localization of TRF2 in relation to telomeres using immuno-FISH. Co-localization analysis of TRF2 and telomeres revealed that the fraction of signals of telomeric foci that overlapped with signals of TRF2 foci was slightly less in LMNA mutant fibroblasts 66% in R133L and -68% in L140R as compared to 78% in the control, 82-6 (Figures 2B,C). Approximately 34% (P = 0.06) of telomeric foci in the R133L mutant and 32% (P = 0.174) in the L140R mutant were free of TRF2. These results agree with our western results, which showed a reduction of TRF2 protein in mass cultures. The TRF2 proteins in LMNA mutants, albeit reduced in amount, were co-localized to the telomeres (Figure 2B). These findings are consistent with the hypothesis that LMNA mutant fibroblasts in young passages might have lost optimum levels of shelterin complexes prior to telomere shortening, and that the reduced levels of TRF2 protein observed in LMNA mutants could be a causal step rather than a consequence of accelerated telomere attrition.

It has been reported that progerin induces a DNA damage response at the telomeres (Benson et al., 2010). We asked whether non-progerin producing LMNA mutations, R133L and L140R, also cause the accumulation of DNA damage and if so, where. To answer these questions, we assayed a marker of DNA damage, γ-H2AX foci. In control fibroblasts, up to approximately 4% of cells were positive for >5 γ-H2AX foci (Figures 3A,B). The fraction of cells with >5 γ-H2AX foci was increased to 22% (P = 0.06) in R133L and 36% (P = 0.01) in L140R mutants (Figure 3B).

To assess the degree to which the DNA damage foci in young passages of LMNA mutant cells were localized to the telomeres, we analyzed the co-localization of γ-H2AX and telomeres using immuno-FISH (Figures 3A,C). In control cells, γ-H2AX signals were hardly visible as foci at standard exposures whereas in LMNA mutant cells γ-H2AX foci tended to be larger than those seen in the controls (Figure 3A). The majority of the γ-H2AX foci present in LMNA mutant cells were not co-localized to telomeric DNA (89% in RI33L and 87% in L140R; Figure 3C). Telomeric γ-H2AX represented relatively minor fractions of the total foci. These data indicate that, during early passages, R133L and L140R LMNA mutant fibroblasts may sustain non-telomeric genomic damage prior to the accumulation of damage at the telomeres.

To further confirm that the reduction of shelterin complex proteins and the accumulation of DNA damage seen in primary LMNA mutant fibroblasts was not due to shortened telomeres, we analyzed TRF2 and other shelterin protein expressions in LMNA mutant and control fibroblasts immortalized with hTERT. Telomerase activity in all hTERT-immortalized cell lines was confirmed with a TRAP assay (data not shown). Western analysis showed an overall reduction of most of the shelterin subunit proteins in mutant hTERT lines in a pattern similar to what was observed in primary fibroblasts; TRF2, TRF1, POT1, TIN2, and Rap1 were reduced to 54, 43, 49, 79, and 59%, respectively, in R133L mutant and 37, 36, 47, 59, and 31%, respectively, in L140R mutant hTERT fibroblasts cell lines, as compared to control hTERT lines (Figure 4A; Table 1). Both Q-FISH analysis and quantitative PCR of the telomeres showed that the telomeres in LMNA mutant hTERT cells were not shorter than those of control hTERT lines (Figure 4B). These data indicate that the reduction of the shelterin complex subunits occurs without shortened telomeres in LMNA mutant hTERT lines, consistent with our findings in primary fibroblasts described above.

As observed in primary fibroblasts, TRF2 mRNA levels in the LMNA mutant hTERT lines were comparable to those of control hTERT lines, indicating that the reduction of TRF2 occurs post-transcriptionally (Figure 4C). In normal human diploid fibroblasts TRF2 undergoes rapid proteasome-mediated degradation after dissociation from the telomeres during replicative senescence (Zhou et al., 2009; Fujita et al., 2010). When proteasome degradation was inhibited by MG-132, TRF2 protein levels in R133L and L140R mutants were partially restored from 49 to 67% and 33 to 63%, respectively (Figure 4D). This suggests that a part of the post-translational degradation of TRF2 in LMNA mutants is through proteasomal degradation. These data indicate that the stabilization of telomeres by hTERT does not prevent the degradation of TRF2 and that there may be other mechanisms that initiate TRF2 reduction in LMNA mutant fibroblasts.

To examine whether DNA damage is involved in the reduction of TRF2 in LMNA mutant cells with stable telomeres, we analyzed DNA damage foci in hTERT-immortalized LMNA mutant and control fibroblasts. The results showed that the fraction of cells positive for >5 γ-H2AX foci in LMNA mutant hTERT fibroblasts was decreased compared to the LMNA mutant primary fibroblasts. However, the number remained higher (9 and 14% in R133L and L140R mutants, respectively) when compared to the control hTERT cell lines, which were devoid of cells with >5 γ-H2AX foci (Figure 5A). Co-localization analysis revealed that only 12 and 5% of total γ-H2AX foci co-localized with TRF2 (a marker of telomeres) in R133L and L140R mutant cells, respectively, whereas the rest of the DNA damage signals were distributed to non-telomeric DNA (Figure 5B). These results indicate that mutant lamin A/C causes the accumulation of non-telomeric DNA damage and the reduction of TRF2 in hTERT-immortalized fibroblasts without substantial telomeric damage, as observed in young primary fibroblasts.