Compromised Chondrocyte Differentiation Capacity in TERC Knockout Mouse Embryonic Stem Cells Derived by Somatic Cell Nuclear Transfer

Mammalian telomere lengths are primarily regulated by telomerase, consisting of a reverse transcriptase protein (TERT) and an RNA subunit (TERC). We previously reported the generation of mouse Terc+/− and Terc−/− embryonic stem cells (ntESCs) by somatic cell nuclear transfer. In the present work, we investigated the germ layer development competence of Terc−/−, Terc+/− and wild-type (Terc+/+) ntESCs. The telomere lengths are longest in wild-type but shortest in Terc−/− ntESCs, and correlate reversely with the population doubling time. Interestingly, while in vitro embryoid body (EB) differentiation assay reveals EB size difference among ntESCs of different genotypes, the more stringent in vivo teratoma assay demonstrates that Terc−/− ntESCs are severely defective in differentiating into the mesodermal lineage cartilage. Consistently, in a directed in vitro chondrocyte differentiation assay, the Terc−/− cells failed in forming Collagen II expressing cells. These findings underscore the significance in maintaining proper telomere lengths in stem cells and their derivatives for regenerative medicine.


Introduction
The ends of eukaryotic chromosomes are capped with telomeres, copies of a hexamer repeat sequence, and associated proteins, which play central roles in stabilizing the ends of chromosomes during replication. Telomeres are primarily regulated by telomerase, a ribonucleoprotein (RNP) consisting of the protein subunit TERT and the RNA subunit TERC. A short segment within TERC is used as a template for a TERT-catalyzed reverse transcription reaction to elongate telomeres. At the organism level, telomerase is inactivated in somatic cells, whose telomeres shorten each time they proliferate [1,2]. In embryonic tissues and adult stem cells (e.g., hematopoietic stem cells), telomerase is expressed as a mechanism to maintain proper telomere lengths [3][4][5]. In humans, loss of function mutation in genes that encode telomerase components, such as TERT or TERC, cause premature aging and age-related diseases, including dyskeratosis congenital (DC), aplastic anemia, and idiopathic pulmonary fibrosis (IPF), which are collectively referred to as "telomere syndromes" to reflect the short and dysfunctional telomeres commonly found in these patients' cells [6][7][8]. Because mice are characterized with excessive long telomere lengths (~100 kb vs.~10-15 kb in humans) [9,10], early generations of telomerase deficient mice that are homozygous Tert or Terc knockout are viable and  Figure 1D), and immunofluorescence staining ( Figure 1E) to evaluate the expression of conventional pluripotent marker genes. Pluripotent gene expression profiles were indistinguishable among Terc −/− , Terc +/− and wild-type Terc +/+ ntESCs. However, the cell growth rate was Terc genotype dependent, with the slowest in Terc −/− ntESCs and the fastest in wild-type cells ( Figure 1F). The telomere lengths, as measured by the Southern blot ( Figure 1G) and T/S ratio ( Figure 1H), were also Terc dependent; the longest in the wild-type, followed by heterozygous knockout, and the shortest in the homozygous knockout. These results show that the expression of conventional pluripotency markers was not sensitive to telomerase insufficiency, while telomere lengths and the cell grow rate were. (D) Western blot analysis of pluripotency by detecting NANOG, octamer-binding transcription factor 4 (OCT4), Sal-like protein 4 (SALL4), and SRY (sex determining region Y)-box 2 (SOX2) in ntESC lines. TUBULIN is used as the internal control. Mouse embryonic fibroblast (MEF) is used as the negative control for pluripotent markers. (E) Immunofluorescent staining shows the positive signal of SOX2 and SALL4 in ntESC lines. 4',6-diamidino-2-phenylindole (DAPI) is used for nuclei stain. (F) Growth curve of ntESCs in six-days of culture in embryonic stem cell (ESC) medium (n = 3). *** indicates significant difference between groups (p < 0.001). ns: no significant differences. (G) Telomere length analysis by telomere restriction fragment (TRF) in three genotypes of ntESCs. Red line indicates the medium length of genomic DNA. (H) Comparison of telomere length by telomere to single-copy gene ratio (T/S ratio). * indicates significant difference between groups (p < 0.05) analyzed by unpaired student t-test. . *** indicates significant difference between groups (p < 0.001). ns: no significant differences. (G) Telomere length analysis by telomere restriction fragment (TRF) in three genotypes of ntESCs. Red line indicates the medium length of genomic DNA. (H) Comparison of telomere length by telomere to single-copy gene ratio (T/S ratio). * indicates significant difference between groups (p < 0.05) analyzed by unpaired student t-test.

Compromised Spontaneous In Vitro and In Vivo Differentiation Capacity in Terc
The germ layer differentiation potency of ntESCs of different Terc genotypes were first evaluated by spontaneous in vitro differentiation of embryoid bodies (EBs). Although EBs can be derived from all genotypes of ntESCs with similar EB formation efficiency, the size of EBs from Terc −/− ntESCs was significantly smaller than those derived from Terc +/+ and Terc +/− ntESCs (Figure 2A,B). RT-PCR assay reveals that these EBs, regardless of genotypes, all expressed signature germ layer-specific genes, including endoderm, SRY (Sex-Determining Region Y)-Box 17 (Sox17) and GATA binding protein 4 (Gata4); mesoderm, Brychury (Bry) and heart and neural crest derivatives expressed 1 (Hand1); and ectoderm, paired box 6 (Pax6), at similar levels with the exception of ectodermal marker gene SRY (sex determining region Y)-box 1 (Sox1), whose expression level was significantly lower in the Terc −/− group than in those in the Terc +/− groups ( Figure 2C,D).
We next conducted a more stringent pluripotency test, the teratoma assay. ntESCs of 1 × 10 6 of each genotype were transplanted to the hind legs of the BALC/C Nu mice. The average weight of the teratoma was smaller in the Terc −/− group (0.41 ± 0.12 g), as compared to the Terc +/− (1.13 ± 0.71 g) and wild-type groups (0.79 ± 0.20 g) ( Figure 2E). Furthermore, hematoxylin and eosin (H&E) staining of teratoma in the Terc −/− group failed to reveal cartilage, a mesoderm derived cell type that was observed in Terc +/− and wild-type groups; whereas cells for endoderm (e.g., ciliated epithelium) and ectoderm (e.g., neuron like) were observed in all groups ( Figure 2F). These results indicate that Terc −/− ntESCs possess compromised germ layer differentiation capacity.  We next conducted a more stringent pluripotency test, the teratoma assay. ntESCs of 1 × 10 6 of each genotype were transplanted to the hind legs of the BALC/C Nu mice. The average weight of the teratoma was smaller in the Terc −/− group (0.41 ± 0.12 g), as compared to the Terc +/− (1.13 ± 0.71 g) and wild-type groups (0.79 ± 0.20 g) ( Figure 2E). Furthermore, hematoxylin and eosin (H&E) staining of teratoma in the Terc −/− group failed to reveal cartilage, a mesoderm derived cell type that was observed in Terc +/− and wild-type groups; whereas cells for endoderm (e.g., ciliated epithelium) and ectoderm (e.g., neuron like) were observed in all groups ( Figure 2F). These results indicate that Terc −/− ntESCs possess compromised germ layer differentiation capacity.

Compromised In Vitro Differentiation Capacity to Chondrocytes in Terc −/− ntESCs
To confirm the observation in the teratoma assay that mesodermal cartilage differentiation capacity in Terc −/− ntESCs was compromised, we subjected ntESCs of different Terc genotypes to a 30-day long directed differentiation protocol for derivation of chondrocytes in vitro ( Figure 3A). It has been reported that this chondrogenic differentiation protocol leads to the formation of cartilage with its typical extracellular matrix. The expression of collagen type II (Col2a1) is used as an indicator for chondrocytes. In addition, the presence of a key molecule within the cartilage matrix, Aggrecan, is used as an indicator for cartilage formation, which is stained dark blue using Alcian Blue.
Along the time course of differentiation, massive cell deaths were observed in the Terc −/− group, but not in the wild-type and Terc +/− groups ( Figure 3A). Furthermore, cells in the Terc −/− group did not express either the early cartilage marker SRY (Sex-Determining Region Y)-Box 9 (Sox9), or the late cartilage marker Col2a1 at day 20 or day 30, as determined by semi-quantitative RT-PCR ( Figure 3C). In contrast, in the wild-type and Terc +/− groups, Sox9 was detected on both day 20 and day 30, while Col2a1 was detectable on day 30 but not day 20 following differentiation ( Figure 3C,D). Real-time PCR analysis also confirmed the significantly decrease of Sox9 gene in the Terc −/− group compared with the wild-type and Terc +/− groups ( Figure 3E). Consistently, Alcian blue staining for Aggrecan was positive in cells of wild-type and Terc +/− but not Terc −/− , on day 30 post differentiation ( Figure 3B).
These results show that Terc −/− , but not the wild-type and Terc +/− ntESCs, failed to form Collagen II expressing chrondrocytes and Aggrecan-positive cartilages after directed in vitro differentiation. These results show that Terc −/− , but not the wild-type and Terc +/− ntESCs, failed to form Collagen II expressing chrondrocytes and Aggrecan-positive cartilages after directed in vitro differentiation.

Discussion
We discovered that the differentiation capacity of ntESCs to mesodermal chondrocyte is vulnerable to short telomeres associated damages. In an early study, Liu et al. reported similar differentiation defects in mouse Terc −/− mesenchymal stem cells [23]. These are consistent with clinical findings that telomere syndrome patients suffer primarily in mesoderm lineage cell types, such as muscle, connective tissue, cartilage, bone, and blood cells [24][25][26]. On the other hand, Aguado et al. reported that iPSCs of long telomeres, as indicated by the high expression of the shelterin-complex protein TRF1, differentiate more efficiently into cardiomyocytes (mesodermal linage) than those with relatively short telomeres [6]. These findings indicate that the differentiation capacity of stem cells into mesodermal linage cell types are sensitive to telomere lengths.
The transcripts of Sox1, an ectodermal marker gene, were significantly decreased in Terc −/− ntESCs compared to other genotypes ( Figure 2D); however, the teratoma assay did not reveal ectodermal differentiation defects of Terc −/− ntESCs ( Figure 2E). It is possible that the decrease of Sox1 could be compensated by other Sox family proteins [27,28]. It remains to be further evaluated if and to what extent the ectodermal differentiation is affected in ntESCs of telomerase deficiency.
More generally, our work underscores the importance for quality control of proper telomere maintenance in regenerative medicine. On one hand, telomere length maintenance is critical for the unlimited self-renewal, pluripotency, and chromosomal stability of PSCs [17]. On the other hand, critically short telomeres lead to slowed self-renewal, chromosomal instability, and compromised pluripotency [17]. Therefore, ensuring healthy telomere lengths should be considered an important quality control parameter toward clinical applications of PSCs and their derivatives. We show, in the present study, that conventional PCR and immunostaining-based examinations of pluripotency markers is unable to identify the germ layer differentiation defects in PSCs that are of short telomeres (e.g., Terc −/− ntESCs). Our work suggests that stringent tests, such as a teratoma assay, should be conducted to evaluate the PSCs' quality especially, when they are at high risk to possess short telomeres, for example haploinsufficiency of a telomerase component gene (Terc or Tert). This is also valid for trans-differentiation-based regenerative medicine, in which a somatic cell type (e.g., fibroblast) is converted to another somatic cell type (e.g., cardiomyocyte), because during trans-differentiation the somatic cells do not go through a "reprogramming to pluripotency" stage, as in the iPSC derivation process, in which telomeres can be reportedly elongated [29][30][31]. Hence it is very possible that the short telomeres in the original somatic cell type are inherited in the target cell type. Without proper telomere lengths, which are often not evaluated, trans-differentiation derived cells may be suboptimal for the intended therapies.
In summary, the present work demonstrates that Terc −/− ntESCs suffer from compromised mesodermal chondrocyte differentiation in vitro and in vivo. It provides new evidence to include telomere length as a parameter in the quality control process in regenerative medicine.

In Vitro Differentiation
For spontaneous differentiation of ntESCs, the confluent cells were trypsinized, and aggregated embryoid body (EB) was placed onto the top of a petri-dish by suspension culture, with the concentration of 1000 cells per drop in the differentiation medium, which was composed of GlutaMax, NEAA, BME, P/S in Dulbecco's Modified Eagle Medium (DMEM, 11995-065, Thermo), and 20% fetal bovine serum (FBS, 10437028, Thermo) in DMEM medium. After 2 days of culture, EBs was re-plated into a new petri dish for further culture for another 8 days, and medium was changed every 2 days. For chondrogenic cells induction, the ntESCs were trypsinized and transferred to the differentiation medium, separated to drops that were 20 µL per drop, and were dripped onto the top of the petri dish. After 3 days, EBs were collected to petri dishes and cultured with the same medium with additional supplementation of transforming growth factor beta 1 (TGF-β 1, 10 ng/mL, , 100-21, PeproTech, Inc., Rocky Hill, NJ, USA), and bone morphogenetic protein-2 (BMP-2, 10 ng/mL, 355-BM, R&D Systems, Minneapolis, MN, USA). EBs were transferred to the tissue culture dishes on gelatin in the differentiation medium containing BMP-2 (10 ng/mL), insulin-Transferrin-Selenium (1 µg/mL, 41400-045, Thermo), and L-Ascorbic acid (50 mg/mL, A4403, Sigma) on the fifth day in chondrocyte differentiation. The medium was changed every 2 days until the thirtieth day in chondrocyte differentiation [34]. Differentiated chondrogenic cells and teratoma section were stained with Alcian blue (A5268, Sigma) for 30 min.

Teratoma Assay
BALB/c Nu mice were purchased from BioLASCO Taiwan Company (Taipei, Taiwan) and maintained in individually ventilated cage (IVC) following the approved protocol reviewed by the Institutional Animal Care and Use Committee of NTU according to protocol number NTU-105-EL-164 (Valid from 2017/08/01 to 2020/07/31). The ntESCs were trypsinized and intramuscular injected into the hind leg 1 × 10 6 per site. After 6 weeks, mice were sacrificed and the tumors were dissociated and fixed in 10% formalin. Fixed samples were embedded and analyzed by hematoxylin and eosin (H&E) stain.

Telomerase Activity Measurement
Telomerase activity was measured by an enzyme-linked immunosorbent assay (ELISA) using a commercial kit (S7700, Millipore). About 1 × 10 5 -10 6 cells from each sample were lysed by 1×3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) lysis buffer. Three hundred ng of extract protein from each sample was used for PCR following manufacturer's instructions. Each PCR experiment included positive control (293T cells) and negative control (heated 293T cells). A serial dilution of TSR8 template provided by the manufacture was used to establish the standard curve. The yield of the PCR reaction was determined by measuring the fluorescence in a spectrofluorometer. Each reaction was performed in duplicate.

Quantitative Real-Time PCR for Telomere Assay
Quantitative PCR (real-time PCR or qPCR) was used to measure relative telomere lengths (RTL) of ntESCs, as previously described [35]. Briefly, genomic DNA was extracted from cells using the DNA Isolation Kit (High Pure PCR Template Preparation Kit, 11796828001, Roche, Basel, Switzerland). For each sample, 20 ng of DNA was used in each reaction. PCR reactions were performed on the SYBR Green detection system (KK4603, Kapa Biosystems, Inc., Woburn, MA, USA), using telomeric primers (5 -CGGTTTGTTTGGGTTTGGGTTTGGGTTTGGGTTTGGGTT-3 and 5 -GGCTTGCCTTACCCTTACCCTTACCCTTACCCTTACCCT-3 ). For each PCR reaction, a standard curve was made by serial dilutions of known amounts of mouse genomic DNA. The telomere signal was normalized to the signal from the single-copy gene (36B4: 5 -ACTGGTCTAGGACCCGAGAAG-3 and 5 -TCAATGGTGCCTCTGGAGATT-3 ) to generate a telomere to single-copy gene ratio (T/S ratio) indicative of relative telomere length of the given sample. Each reaction was performed in triplicate.

Statistical Analysis
The signal intensity of agarose gel electrophoresis was analyzed with Image J software (version 1.47, National Institutes of Health, Bethesda, MD, USA) [36] and was normalized to internal control gene Gapdh. The data were statistically analyzed by GraphPad Prism software by one-way ANOVA following Tukey's test. Standard deviation of the mean (SEM) was shown in all figures unless indicated. Significance was defined as follows: * = p < 0.05, ** = p < 0.005, *** = p < 0.001.

Conclusions
In this study, we evaluated the differentiation capacities by in vitro embryoid body (EB), in vivo teratoma, as well as an in vitro directed differentiation assay in Terc −/− , Terc +/− , and wild-type (Terc +/+ ) ntESCs. Our work revealed Terc −/− ntESCs suffered from compromised mesodermal chondrocyte differentiation, which elucidated for the significance in maintaining proper telomere lengths in stem cells and their derivatives for regenerative medicine.