Differential stem cell aging kinetics in Hutchinson-Gilford progeria syndrome and Werner syndrome

Hutchinson-Gilford progeria syndrome (HGPS) and Werner syndrome (WS) are two of the best characterized human progeroid syndromes. HGPS is caused by a point mutation in lamin A (LMNA) gene, resulting in the production of a truncated protein product—progerin. WS is caused by mutations in WRN gene, encoding a loss-of-function RecQ DNA helicase. Here, by gene editing we created isogenic human embryonic stem cells (ESCs) with heterozygous (G608G/+) or homozygous (G608G/G608G) LMNA mutation and biallelic WRN knockout, for modeling HGPS and WS pathogenesis, respectively. While ESCs and endothelial cells (ECs) did not present any features of premature senescence, HGPS- and WS-mesenchymal stem cells (MSCs) showed aging-associated phenotypes with different kinetics. WS-MSCs had early-onset mild premature aging phenotypes while HGPS-MSCs exhibited late-onset acute premature aging characterisitcs. Taken together, our study compares and contrasts the distinct pathologies underpinning the two premature aging disorders, and provides reliable stem-cell based models to identify new therapeutic strategies for pathological and physiological aging.


INTRODUCTION
Progeroid syndromes are heritable human disorders characterized by progeroid features that recapitulate typical features of normal aging. Among all the progeroid syndromes, Hutchinson-Gilford progeria syndrome (HGPS) and Werner syndrome (WS) are best characterized (Kudlow et al., 2007). HGPS is a sporadic autosomal dominant syndrome, and most HGPS patients were heterozygous for LMNA mutation (p.G608G/+). LMNA encodes A-type lamins that belongs to the family of nuclear lamina proteins, and a point mutation (p.G608G) in LMNA creates an aberrant splicing site in exon 11, resulting in the production of a truncated protein, progerin (Chojnowski et al., 2015;DeBoy et al., 2017;Luo et al., 2014). Another commonly seen progeroid syndrome is WS, caused by mutations in WRN gene that encodes a RecQ DNA helicase (Yu et al., 1996) important to DNA replication and DNA damage repair. Lossof-function WRN leads to genomic instability, heterochromatin alterations, and cell growth defects, which contribute to WS pathogenesis Murfuni et al., 2012;Ren et al., 2017a;Ren et al., 2011;Seki et al., 2008;Shamanna et al., 2017;Zhang et al., 2015).
Both HGPS and WS patients present a wide range of aging-associated syndromes such as alopecia, lipodystrophy, osteoporosis and atherosclerosis. Studies on fibroblasts from HGPS and WS patients reveal features of accelerated cellular senescence and decreased proliferation potential (Brunauer and Kennedy, 2015;Chen et al., 2017;Cheung et al., 2014;Cheung et al., 2015;Kudlow et al., 2007;Liu et al., 2011a). Despite these common features, differences exist between HGPS and WS in the scope, intensity and duration of symptoms. For example, most patients with HGPS show symptoms resembling aspects of aging at a very early age and die at a median age from 11 to 13. By comparison, WS patients usually develop normally in the childhood and can live up to their fifties (Cox and Faragher, 2007;Ding and Shen, 2008;Hennekam, 2006;Kudlow et al., 2007;Mazereeuw-Hautier et al., 2007;Muftuoglu et al., 2008;Oshima et al., 2017).
In recent years, technologies based on stem cells and gene editing have been widely used to model various human diseases (Atchison et al., 2017;Duan et al., 2015;Fu et al., 2016;Liu et al., 2011a;Liu et al., 2012;Liu et al., 2014;Liu et al., 2011b;Lo Cicero and Nissan, 2015;Miller et al., 2013;Pan et al., 2016;Ren et al., 2017b;Wang et al., 2017;Yang et al., 2017;Zhang et al., 2015). Of note, HGPS-specific induced pluripotent stem cells (iPSCs) and WS-specific iPSCs and embryonic stem cells (ESCs) have been separately generated. Based on the findings by us and other groups, although the iPSCs and ESCs do not have any premature aging defects, mesenchymal stem cells (MSCs) and vascular smooth muscle cells (VSMCs) derived from these pluripotent stem cells display premature aging, consistent with the observations in fibroblasts from HGPS and WS patients (Chen et al., 2017;Cheung et al., 2014;Liu et al., 2011a;Miller et al., 2013;Zhang et al., 2011). Both being typical cases of progeroid syndromes, comparative analysis on HGPS and WS is very limited. More information about the similarities and differences in the pathological processes and molecular mechanisms of HGPS and WS remains to be uncovered via comparative studies.
Here, we successfully developed a reliable and isogenic platform for side-by-side investigation of HGPS and WS. Taking advantage of gene editing, we generated human ESCs harboring heterozygous LMNA p.G608G mutation and WRN deficiency, mimicking HGPS and WS, respectively. Notably, a genetically enhanced HGPS-specific ESCs bearing biallelic LMNA p.G608G mutation were also created. We found that HGPS-and WS-MSCs, but not ESCs or ECs, exhibited typical aging-associated characteristics. Interestingly, distinct aging kinetics were detected between HGPSand WS-MSCs. For the first time, we achieved a contemporaneous comparison between HGPS and WS under the same genetic background to unravel the molecular and cellular differences, opening a window into the understanding of the pathology of human aging and providing a platform for screening for therapeutic strategies against aging-associated disorders.

Generation of LMNA-mutated and WRN-deficient human ESCs
Using a genome-editing technique with a helper-dependent adenoviral vector (HDAdV), we generated heterozygous and homozygous LMNA-mutated human ESC lines (Fig. 1A). Combined with our previously reported WRN-deficient human ESCs , we obtained ESCs with heterozygous (LMNA G608G/+ ), homozygous (LMNA G608G/ G608G ) LMNA mutation, and homozygous WRN deficiency (WRN −/− ) under the same genetic background ( Fig. 1B-D). All the three ESC lines displayed normal karyotypes and morphologies indistinguishable from those of WT-ESCs ( Fig. 1B and 2A). All clones expressed the pluripotency markers OCT4, SOX2, NANOG, and were hypomethylated at the OCT4 promoter region ( Fig. 1B and 2B). Each cell line was maintained for more than 30 passages without detectable growth abnormalities (data not shown) and was assessed for pluoripotency by differentiation into the three embryonic germ layers in vivo, by teratoma formation (Fig. 2C). Ki67 staining and cell cycle analysis also confirmed comparable proliferation potential of HGPS-ESCs and WS-ESCs with that of WT-ESCs ( Fig. 2D and 2E). As expected, progerin was suppressed in both HGPS-ESCs and WS-ESCs (Fig. 1D). In addition, the levels of nuclear lamina component LAP2β, and heterochromatin markers H3K9me3 and HP1α were each normal in HGPS-ESCs and WS-ESCs compared to WT-ESCs ( Fig. 2F and 2G). These data indicate that despite the progeroid-associated mutations, premature senescence phenotypes and chromosomal Zeming Wu et al. instability are well concealed in HGPS-ESCs and WS-ESCs at the pluripotent stage.

HGPS-MSCs and WS-MSCs exhibit aging-associated phenotypes with different kinetics
Clinical observations in HGPS and WS patients indicate that premature aging disorders are often accompanied with defects in mesenchymal lineages, such as lipodystrophy, osteoporosis and atherosclerosis (Cox and Faragher, 2007). MSCs are adult stem cells originated from mesoderm and can be differentiated into osteocytes, chondrocytes, adipocytes and many other cell types (Lepperdinger, 2011;Marofi et al., 2017;Uccelli et al., 2008). We postulated that MSC exhaustion may play an important role in premature aging disorders. Here, HGPS-ESCs and WS-ESCs were differentiated into HGPS-MSCs and WS-MSCs. Both MSC lines expressed MSC-specific markers including CD90, CD73 and

RESEARCH ARTICLE
Zeming Wu et al. CD105 (Fig. 3A) and exhibited multiple-lineage differentiation potentials including adipogenesis, osteogenesis and chondrogenesis, though the differentiation ability of WS-MSCs towards adipocytes and osteoblasts was partly compromised ( Fig. 3B-D).
Having shown the distinct senescence-associated kinetics in HGPS-MSCs and WS-MSCs, we continued to evaluate other aging-related phenotypes. Increased DNA damage response is an important feature of aging (Brunauer and Kennedy, 2015;Burtner and Kennedy, 2010;Liu et al., 2005;Lopez-Otin et al., 2013;Mostoslavsky et al., 2006;Musich and Zou, 2011;Saha et al., 2014;Wang et al., 2009;Zhang et al., 2015). Here, increase in γ-H2AX and 53BP1 double-positive cells, indicative of increased DNA damage response, was observed only in WS-MSCs at passage 3 ( Fig. 6A). At passage 9, both WS-MSCs and HGPS-MSCs exhibited increased DNA damage response, with the most observed in homozygous HGPS-MSCs (Fig. 6A). Increased size and decreased number of nucleoli can also serve as aging biomarkers (Buchwalter and Hetzer, 2017;Tiku et al., 2016). We observed that only WS-MSCs had fewer but larger nucleoli at early passages, and both WS-MSCs and HGPS-MSCs exhibited increased size and decreased numbers of nucleoli at late passages (Fig. 6B).
Taken together, these results suggest that HGPS-MSCs and WS-MSCs exhibit aging-associated phenotypes with different kinetics, and progerin exerts a dose-dependent effect on cellular senescence of HGPS-MSCs.

HGPS-ECs and WS-ECs do not exhibit phenotypes of accelerated senescence
Arterosclerosis have been observed in HGPS and WS patients, and progerin is widely present in the vascular cells including endothelial cells (Lo et al., 2014;McClintock et al., 2006;Miyamoto et al., 2014;Olive et al., 2010). As the inner layer of blood vessels, endothelial cells have unique functions in vascular biology, including barrier effect, vascular tone control, blood clotting regulation and inflammatory response (Bochenek et al., 2016;Hansen et al., 2017;Sturtzel, 2017). To explore whether LMNA mutation or WRN deficiency may cause aging-associated defects in endothelial cells (ECs), HGPS-ESCs and WS-ESCs were differentiated into HGPS-ECs and WS-ECs, respectively. CD31 and CD144 double-positive cells were sorted (Fig. 7A). All EC

RESEARCH ARTICLE
Zeming Wu et al.
To be noted, both HGPS-ECs and WS-ECs were more apoptotic compared to WT-ECs at baseline, indicating

RESEARCH ARTICLE
Zeming Wu et al.

DISCUSSION
Human progeroid syndromes are characterized by typical features resembling normal aging, and therefore studies on progeroid disorders have provided important clues to understanding the molecular mechanisms underlying premature and normal aging (Burtner and Kennedy, 2010;Cao et al., 2011;Ding and Shen, 2008;Dreesen and Stewart, 2011;Kudlow et al., 2007;Miyamoto et al., 2014;Scaffidi and Misteli, 2006). As two of the best characterized progeroid syndromes, HGPS and WS have attracted a lot of attention during the last decade; related studies have been greatly conducive to our understanding of the pathology of these two disorders (Atchison et al., 2017;Chen et al., 2017;Cheung et al., 2014;De Sandre-Giovannoli et al., 2003;Ding and Shen, 2008;Egesipe et al., 2016;Kubben et al., 2016;Kudlow et al., 2007;Li et al., 2016;Liu et al., 2011a;Liu et al., 2011b;Lo Cicero and Nissan, 2015;Scaffidi and Misteli, 2006;Zhang et al., 2011;Zhang et al., 2015). However, there are no effective treatments so far and more information about the molecular pathology of these two premature aging syndromes are to be unveiled.
In this study, we generated LMNA-mutated and WRNdeficient human ESC lines with the same genetic background, making it possible to compare and contrast the cellular consequences of the genetic defects underlying HGPS and WS side-by-side. Similar to the iPSCs derived from the fibroblasts of HGPS and WS patients, HGPS-and WS-ESCs did not show any premature aging defects, indicating that pluripotent stem cells are able to conceal aging defects caused by LMNA mutation or WRN deficiency (Liu et al., 2012;Zhang et al., 2013). Upon mesenchymal differentiation, however, HGPS-and WS-MSCs exhibited agingassociated phenotypes that recapitulate those reported in fibroblasts and iPSC-derived MSCs from HGPS and WS patients Cheung et al., 2015;Compagnucci and Bertini, 2017;Zhang et al., 2011), with different kinetics. By measuring proliferation potential, SA-β-gal positivity, cell cycle, DNA damage response, and nuclear architecture, we showed that WS-MSCs had early-onset mild premature aging phenotypes while HGPS-MSCs exhibited late-onset acute premature aging characterisitics. To some extent, these dynamic features may mimic the patterns of disease progression of these two premature aging disorders (Fig. 10). To our knowledge, this is the first study evaluating the similarities and differences of HGPSand WS-stem cells side by side. Our platform provides powerful tools to study aging by mimicking human genetic diseases in a petridish, facilitating the understanding of the pathology of different types of progeroid disorders and more importantly, making it possible for targeted high-throughput drug screening in human genetic background.
In addition, we observed that the homozygous HGPS-MSCs exhibited more severe aging phenotypes with a higher level of progerin than the heterozygous MSCs. Thus, the MSCs with homozygous or heterozygous LMNA mutation generated in our study also provide opportunities to investigate the role of progerin in a dose-dependent manner. Given the propriety of higher homogenicity in MSCs bearing homozygous LMNA mutation (e.g., expression of progerin), these cells may be particularly amenable to mechanistic studies using multi-omics techniques.
Different from HGPS-MSCs, HGPS-ECs did not display any premature senescence phenotypes, consistent with previous observations in HGPS-iPSC-derived ECs . Similarly, WS-ECs did not show aging defects, either. These results indicate that the senescence-associated defects caused by LMNA mutation or WRN deficiency are cell-type-specific (Fig. 10). However, further analyses show that these cells were not otherwise normal; HGPS-ECs and WS-ECs were more apoptotic at baseline than WT-ECs. Moreover, WS-ECs, but not HGPS-ECs, exhibited a more pronounced response to inflammatory factor TNF-α, again indicating different molecular pathologies between the two progeroid syndromes.
Therefore, we have generated in vitro models to compare and contrast the pathogenesis of HGPS and WS for the first time, providing high-throughput platforms to efficiently screen for effective treatments for both progeria syndromes and normal aging. In the future, it would be interesting to employ multi-omics technologies, including genomics, epigenomics, transcriptomics, proteomics and metabonomics, to unravel the molecular patterns of HGPS and WS under the same human genetic background, shedding light on the complex mechanisms underlying premature and normal aging and providing new evidence for the prevention and treatment of age-associated disorders.

RESEARCH ARTICLE
Zeming Wu et al.
Bisulfite sequencing of the OCT4 promoter Bisulfite treatment of DNA was carried out by using EZ DNA Methylation Kit (Zymo Research) following the manufacturer's instructions. About 1 μg of genomic DNA was used. A genomic fragment of the OCT4 promoter was amplified using LA Taq Hot Start Version (TAKARA) as previously described. In brief, PCR products were purified by using gel extraction kit (Qiagen), and subsequently cloned into the pMD20 T vector (Transgen). 7 clones from each sample were sequenced with the universal primer M13. Primers used for PCR: meF-OCT4, 5′-ATTTGTTTTTTGGG-TAGTTAAAGGT-3′; meR-OCT4, 5′-CCAACTATCTTCATCTTAA-TAACATCC-3′.

Teratoma analysis
Teratoma assay was performed as described (Duan et al., 2015;Fu et al., 2016;Zhang et al., 2015). Briefly, 5 × 10 6 ESCs were administrated subcutaneously into NOD/SCID mice (male, 6-8 weeks). 6-12 weeks after injection, mice were killed and teratomas were analyzed by immunostaining. All animal experiments were conducted with approval of the Institute of Biophysics, Chinese Academy of Science.

Fluorescence-activated cell sorting (FACS)
MSCs or ECs were collected by using TrypLE Express (Gibco) and washed by PBS twice. Cells were incubated with primary antibody diluted with 10% FBS in PBS for 1 h at room temperature and then sorted by using a flow cytometer (BD FACSAria IIIu).

RESEARCH ARTICLE
Zeming Wu et al.

Analysis of cell cycle distribution
For cell cycle analysis, about 1 × 10 6 cells were processed with the Click-iT EdU Flow Cytometry Assay Kits (Invitrogen) according to the manufacturer's instructions. In brief, the cells were harvested after 2 h incubation with EdU and stained with Alexa Fluor 647 dye azide and propidium iodide. Cells were examined by fluorescence-activated cell sorting (FACS) using a flow cytometer (BD LSRFortesa).

Clonal expansion assay
The single-cell clonal expansion assay was carried out as described (Duan et al., 2015). Briefly, 2,000 MSCs were seeded in a gelatincoated 12-well plate. The relative cell density was then determined by Image J after crystal violet staining.

SA-β-Gal staining
SA-β-Gal staining was performed as described previously (Duan et al., 2015;Zhang et al., 2015). Briefly, cultured cells were washed in PBS and fixed at room temperature for 5 min in 2% formaldehyde and 0.2% glutaraldehyde. Fixed cells were stained with SA-β-Gal staining solution at 37°C overnight, percentage of SA-β-Gal positive cells were then calculated.

Measurement of cell apoptosis and nitric oxide (NO)
For cellular apoptosis analysis, cells were collected freshly and stained with Annexin V-EGFP Apoptosis Detection Kit (Vigorous Biotechnology), and then apoptotic cells were quantified by FACS. For NO detection, cells were treated with DAF-FM (Molecular Probes) for 30 min and quantified by FACS.

Dil-Ac-LDL uptake assay
In brief, ECs were collected after 6 h incubation with Dil-Ac-LDL (Molecular Probes) in EC culture medium. For FACS analysis, cells were collected by using TrypLE Express (Gibco) and measured by a flow cytometer (BD LSRFortesa). For immunofluorescence detection, cells were processed following the immunofluorescence microscopy protocol.

In vitro tube formation assay
Briefly, 5 × 10 4 cells were suspended in 500 μL EC medium and then seeded on Matrigel coated plate. After 6-8 h, lattice-like vessel structures formed and then cells were stained with Calcein-AM (Invitrogen) and examined by using fluorescence microscope (Olympus).

RT-PCR
Total RNA was extracted by using TRIzol reagent (Invitrogen). 2 μg of RNA was converted to cDNA by using GoScript Reverse Transcription System (Promega), and 1/50 volume of the cDNA reaction was applied to PCR using primers for human GAPDH, Progerin. RT-qPCR was performed by using iTaq Universal SYBR Green Supermix (Bio-Rad).

Immunofluorescence microscopy
Cells seeded on microscope coverslips were fixed with 4% formaldehyde in PBS for 30 min, permeabilized with 0.4% Triton X-100 in PBS for 25 min, and blocked with 10% donkey serum in PBS for 1 h. The coverslips were incubated with primary antibody (diluted with 1% donkey serum in PBS) overnight at 4°C and then incubated with fluorescence-labeled secondary antibody (diluted with 1% donkey serum in PBS at 1:500) at room temperature for 1 h. Hoechst 33342 (Invitrogen) was used to stain nuclear DNA.

Statistical analysis
Student's t-test was used to analyse differences between different cell lines. Results were presented as mean ± SEM. P values < 0.05, P values < 0.01 and P values < 0.001 were considered statistically significant (*, **, ***).