SIRT6 enhances telomerase activity to protect against DNA damage and senescence in hypertrophic ligamentum flavum cells from lumbar spinal stenosis patients

Lumbar spinal stenosis (LSS) is a condition wherein patients exhibit age-related fibrosis, elastin-to-collagen ratio reductions, and ligamentum flavum hypertrophy. This study was designed to assess the relationship between SIRT6 and telomerase activity in hypertrophic ligamentum flavum (LFH) cells from LSS patients. We observed significant reductions in SIRT6, TPP1, and POT1 protein levels as well as increases in telomerase reverse transcriptase (TERT) levels and telomerase activity in LFH tissues relative to non- hypertrophic ligamentum flavum (LFN) tissues. When SIRT6 was overexpressed in these LFH cells, this was associated with significant increases in telomerase activity and a significant reduction in fibrosis-related protein expression. These effects were reversed, however, when telomerase activity was inactivated by hTERT knockdown in these same cells. SIRT6 overexpression was further found to reduce the frequency of senescence-associated β-galactosidase (SA-β-Gal)-positive LFH cells and to decrease p16, MMP3, and L1 mRNA levels and telomere dysfunction-induced foci (TIFs) in LFH cells. In contrast, hTERT knockdown-induced telomerase inactivation eliminated these SIRT6-dependent effects. Overall, our results indicate that SIRT6 functions as a key protective factor that prevents cellular senescence and telomere dysfunction in ligamentum flavum cells, with this effect being at least partially attributable to SIRT6-dependent telomerase activation.


INTRODUCTION
Lumbar spinal stenosis (LSS) is a relatively common cause of lower back pain in older adults, and it is characterized by age-associated fibrosis, reduced elastin-to-collagen ratios, and ligamentum flavum (LF) hypertrophy [1,2]. Under normal conditions, the LF is highly elastic and is composed of approximately 80% elastin fibers and 20% collagen fibers, with the relative frequency of collagen fibers in the LF rising as a function of age, degeneration, and tissue hypertrophy [3,4]. The onset of LSS is associated with a range of different pathological changes including increases in the expression of inflammatory cytokines, matrix metalloproteinases, and pro-fibrotic growth factors [5]. Cellular senescence is thought to be an important driver of intervertebral disc degeneration [6][7][8]. The specific molecular mechanisms governing LF cell senescence in the context of LSS development, however, remain to be defined.
Sirtuins are NAD+ dependent histone deacetylases that control genomic stabilization, inhibit inflammation, and slow the aging process [9]. Seven total sirtuins (SIRT1-7) have been described in mammals, with each member of this family having different localization patterns and target proteins within cells [10]. SIRT6 is a sirtuin with a high degree of substrate-specificity that plays important roles in maintaining chromosomal integrity and functionality through influencing DNA repair, gene AGING expression, and genomic stabilization [11]. SIRT6 deficiencies have been linked to premature aging phenotypes in mice [12], whereas SIRT6 overexpression is associated with an increased lifespan in these same animals [13], thus suggesting that SIRT6 is a key regulator of both physiological aging and agingassociated disease.
Cellular senescence is closely linked to the structure and function of telomeres, which are located on chromosomal ends and which grow shorter with each round of cell division. When these telomeres reach a critical length, cells cease to divide and are considered to be senescent [14]. Telomeres can be lengthened by telomerase, which is a ribonucleoprotein complex composed of RNA and the human telomerase reverse transcriptase (hTERT) enzyme, and which functions by adding repeating telomere sequences to the 3' ends of DNA, thereby forestalling senescence and extending the cellular lifespan [15]. Nagai et al. [16] demonstrated that when SIRT6 was knocked down in human chondrocytes, this was associated with premature senescence, telomere dysfunction, and DNA damage. Whether SIRT6 impacts telomerase activity in hypertrophic LF (LFH) cells from LSS patients, however, remains to be determined. The present study was therefore designed to assess how SIRT6 impacts DNA damage, telomere dysfunction, and cellular senescence in these cells, with our hypothesis being that SIRT6 can promote telomerase activation and thereby protect LF cells from aging-related damage.

Ligamentum flavum hypertrophy is associated with the downregulation of SIRT6 and the activation of telomerase activity
When we compared gene expression levels in hypertrophic ligamentum flavum (LFH) and nonhypertrophic ligamentum flavum (LFN) tissue samples from LSS patients, we found that the expression of SIRT6 and of the telomere-related genes POT1 and TPP1 was significantly reduced in LFH samples relative to LFN samples through both transcriptomic highthroughput sequencing and qRT-PCR analyses ( Figure  1 and Table 1). In total, 3,935 differentially expressed genes (DEGs) were identified through hierarchical clustering analyses (Figure 1A, 1B and Supplementary  Table 1), with 1,952 and 1,983 of these genes being upand down-regulated, respectively, in LFH tissues. DEGs associated with telomere function are shown in Table 1. Six of these DEGs were selected for qRT-PCR-based validation, revealing TPP1, SIRT6, and POT1 to have been significantly downregulated in LFH samples relative to LFN samples ( Figure 1C). This was further supported by immunohistochemical staining analyses, which demonstrated that TERT was upregulated and POT1, TPP1, and SIRT6 was downregulated in LFH tissues at the protein level relative to LFN tissues (Figure 2A-2E). We similarly detected significantly increased telomerase activity in LFH samples relative to LFN samples using a telomerase PCR ELISA kit ( Figure 2F). Together, these findings suggested that alterations in the expression of SIRT6 and telomerase activity were associated with LF hypertrophic pathology.

SIRT6 overexpression enhances telomerase activity and inhibits fibrosis in ligamentum flavum cells
We next assessed the impact of SIRT6 on LF cell functionality by transducing LFH cells for 48 h with a lentivirus encoding human SIRT6, resulting in significant protein-level SIRT6 upregulation relative to untransfected LFH control cells ( Figure 3A, 3B and Supplementary Figure 1). This SIRT6 overexpression was associated with significantly enhanced telomerase activity, although this activity remained at a low level in LFN cells ( Figure 3C). SIRT6 overexpression was also associated with significant reductions in TGF-β1, α-SMA, and collagen I protein levels in LFH cells ( Figure  3D, 3E and Supplementary Figure 2). Together, these findings suggest that SIRT6 can suppress the myofibroblastic differentiation of LFH cells by enhancing telomerase activation.

SIRT6 overexpression enhances telomerase activity and fibrosis-related protein expression through a mechanism dependent upon hTERT upregulation
To examine the role of telomerase activation in the SIRT6-mediated inhibition of fibrosis in LFH cells, we next utilized an shRNA-encoding lentivirus to knock down hTERT in these same cells. Such knockdown did not adversely impact SIRT6 expression ( Figure 4A, 4B and Supplementary Figure 3), but it did reverse SIRT6 overexpression-dependent enhancement of telomerase activity ( Figure 4C) and downregulation of TGF-β1, α-SMA, and collagen I ( Figure 4D, 4E and Supplementary Figure 4). This, therefore, suggests that SIRT6 can inhibit fibrotic differentiation in LFH cells at least in part via the activation of telomerase activity.

Knockdown of hTERT reverses the beneficial impact of SIRT6 overexpression on premature cellular senescence in LFH cells
To assess the functional importance of SIRT6 and telomerase activity in the context of LF cellular senescence, we next conducted SA-β-gal staining in these cells at 48 h after SIRT6 overexpression and/or AGING hTERT knockdown ( Figure 5A). SA-β-gal staining was increased at baseline in LFH cells relative to LFN cells (18.6% vs. 4.2%). SIRT6 overexpression significantly reduced the frequency of SA-β-gal-positive LFH cells to 11.8%, whereas hTERT knockdown increased this frequency to 28.4%. Importantly, hTERT knockdown in SIRT6-overexpressing cells reversed the beneficial impact of SIRT6 overexpression and increased the   AGING frequency of SA-β-Gal positive LFH cells to 26.2% ( Figure 5B). Trends in the expression of senescencerelated genes including p16 INK4a , MMP3, and long interspersed nuclear element 1 (L1) were consistent with frequencies of SA-β-gal-positive cells, with the exception of L1, which was inhibited in pSIRT6 and sh-hTERT co-transduced cells, while the same was not true for p16 or MMP3 ( Figure 5C-5E). Similarly, hTERT knockdown inhibits telomerase activity regardless of SIRT6 overexpression ( Figure 5F). Together, these findings thus suggest that telomerase activity is an essential mediator of the SIRT6-driven inhibition of premature senescence.

SIRT6 induces telomerase activation and thereby protects LFH cells from DNA damage and telomeric dysfunction
To understand the mechanisms whereby SIRT6 overexpression inhibits premature LFH cellular senescence, we next assessed whether such overexpression was associated with any significant changes in γH2AX foci formation or TIF prevalence within these cells, given that these are reliable markers of DNA damage and telomeric dysfunction. We observed significantly higher numbers of γH2AX foci and telomere dysfunction-induced foci (TIFs) in LFH AGING cells, whereas numbers of TRF-1 foci were decreased relative to LFN cells. Importantly, SIRT6 overexpression significantly decreased the numbers of TIFs and γH2AX foci in LFH cells. In contrast, hTERT knockdown was sufficient to significantly increase the number of γH2AX foci and TIFs within LFH cells, thereby reversing the impact of SIRT6 overexpression on the formation of these foci ( Figure 6A-6D).
Together, these findings suggest that SIRT6 can protect LFH cells against premature senescence via activating telomerase and thereby preventing DNA damage and telomeric dysfunction.

DISCUSSION
While disc degeneration occurs naturally over the course of human aging, it can be associated with several debilitating conditions including lumbar disc herniation, lumbar canal stenosis, segmental lumbar instability, and degenerative lumbar scoliosis. The hypertrophy of LF cells is a key driver of LSS, and is best characterized by reductions in the relative abundance and organization of elastin fibers within the LF coinciding with increases in the abundance of collagen fibers [3,4]. Liu et al. [17] previously demonstrated that inhibiting cellular AGING senescence is a potentially viable approach to preventing aging-associated intervertebral disc degeneration and related conditions.
Sirtuins are closely linked to aging responses in mammals, and many studies have demonstrated that SIRT1, SIRT3, and SIRT6 serve as inhibitors of agingrelated processes [18][19][20]. Moreover, SIRT6 acts as a suppressor of aging-related genes such as p16, MMP3, and L1 [21,22]. In this study, we confirmed that SIRT6 localizes to the nuclei in human LF tissue samples, and we found that it was expressed at much lower levels in LFH cells relative to LFN cells. Importantly, we demonstrated that SIRT6 can prevent both DNA damage and telomeric dysfunction in LF cells, thereby protecting them against premature aging.
The structural and functional integrity of telomeres is dependent upon telomerase-mediated elongation and on the shelterin complex (TRF1, TRF2, TIN2, hRAP1, AGING TPP1, and POT1), which shields these structures from degradation [23]. Telomerase activity is primarily mediated by TERT, which functions as a component of a ribonucleoprotein complex in order to synthesize repeating telomere sequences that are attached to the ends of each chromosome [24]. In a study utilizing TERTknockout mice, Cheng et al. [25] demonstrated that reduced telomerase activity was associated with Representative images of LF cells stained for γH2AX (red) and TRF-1 (green) and counterstained with DAPI (blue). Scale bar = 5 μm. The mean number of γH2AX foci (B), TRF-1 (C) and telomere dysfunction-induced foci (TIFs, D) in cells were quantified using ImageJ. Data are means ± SD of three replicates. *p<0.05, **p<0.01 vs. LFN. #p<0.05 vs. LFH control.
AGING increased cellular senescence, reduced autophagy in renal tubular epithelial cells, and impaired renal functional recovery in response to ischemia-reperfusion injury. In a separate analysis, Liu et al. [26] overexpressed hTERT in hepatic tissues from elderly rats prior to liver transplantation and thereby found that hTERT was able to protect these cells from apoptotic death.
Telomerase is present in the nucleus wherein it repairs chromosomal DNA when cells are divided or DNA is damaged. As such, telomerase activity is suppressed at baseline in healthy tissues, whereas it is activated in tumors and proliferating cells [27]. Our results clearly demonstrate that LFH tissues exhibited increased TERT and telomerase activity as well as decreased POT1 and TPP1 expression relative to LFN tissues, consistent with a role for enhanced telomerase activity in hyperplasic tissues [27,28]. Dechsupa et al. [29] demonstrated that LFH cell samples from LSS patients exhibited increased oxidative DNA damage and shorter average telomere lengths relative to LFN cells from these same patients. However, this prior study did not examine telomerase activity or the underlying molecular mechanisms governing this phenotype.
In this study, we found that overexpressing SIRT6 was sufficient to bolster telomerase activity in LFH cells while simultaneously reducing the expression of the fibrosis-related proteins TGF-β1, α-SMA, and collagen I in these same cells. We further confirmed that such SIRT6 overexpression was able to protect LFH cells from telomeric dysfunction and DNA damage. SIRT6 knockdown has previously been demonstrated to drive significant increases in DNA damage and telomeric dysfunction in human chondrocytes [16], whereas SIRT6 overexpression protects against myocardial damage via altering telomeres [30] and prevents myofibroblast differentiation via the consequent inactivation of the TGF-β1/Smad2 pathway [31]. We determined that telomerase inactivation mediated by hTERT knockdown in LFH cells was sufficient to enhance DNA damage, senescence, and myofibroblastic differentiation while also reversing the protective impact of SIRT6 overexpression on these same phenotypes. Razdan et al. [32] previously demonstrated that telomeric dysfunction can drive human fibroblast transdifferentiation into myofibroblasts. Our results are thus consistent with findings from many previous reports, suggesting that SIRT6 overexpression can prevent the myofibroblastic differentiation of LF cells through a mechanism dependent upon changes in the expression and activity of telomerase.
In summary, in the present study we found that SIRT6 overexpression in human LF cells was sufficient to protect them against DNA damage, premature senescence, and myofibroblastic differentiation through a mechanism dependent upon telomerase activation. Future studies regarding the relationship between SIRT6 and telomerase may offer novel insights into the development and progression of LSS, and have the potential to identify important new therapeutic targets for the treatment of this disease.

Patient specimens
LF tissue samples were collected from 33 patients (20 males, 13 females) with LSS at L4/L5 that had undergone decompressive surgery at Renji Hospital (Shanghai, China). Patients were between the ages of 50 and 70 (mean: 58.1±7.4 years). LFH samples were collected from those patients exhibiting evidence of LF hypertrophy at the L4/L5 level, while non-hypertrophic samples from the L3/L4 level in these same patients were collected as LFN controls. Following collection, these LF tissue samples were either snap-frozen and stored at −80° C, or were immediately used to isolate LF cells. The ethics committee of Renji Hospital approved this study, which was consistent with the Declaration of Helsinki. All patients provided written informed consent prior to study participation.

Sequencing analysis
High-throughput transcriptomic sequencing was used to compare patterns of gene expression in LFH and LFN samples. Briefly, tissue RNA was extracted with TRIzol (Invitrogen, CA, USA), after which chloroform and isopropyl alcohol were used to precipitate this RNA. Sequencing and subsequent bioinformatics analyses were conducted by Cloud-Seq Biotech (Shanghai, China). Briefly, total RNA was treated with a Ribo-Zero rRNA Removal Kits (Illumina) to remove rRNA based on the manufacturer's instructions, after which RNA libraries were constructed using rRNA-depleted RNAs with a TruSeq Stranded Total RNA Library Prep Kit (Illumina) according to the manufacturer's instructions. Libraries were then subjected to quality control and quantification using the BioAnalyzer 2100 system (Agilent Technologies, USA), and 10 pM libraries were denatured to yield single-stranded DNA molecules that were captured on Illumina flow cells, amplified in situ as clusters and finally sequenced for 150 cycles on an Illumina HiSeq Sequencer as per the manufacturer's instructions.

qRT-PCR
TRIzol (Invitrogen) or an RNeasy mini kit (Qiagen, CA, USA) were used to extract RNA from samples, with AGING RNAlater (Qiagen) having been used for RNA preservation. A RetroScript kit (Ambion, TX, USA) was used to reverse transcribe RNA (2 μg/sample) based on provided directions. SYBR Green Rox Master Mix (Qiagen) was used for qRT-PCR reactions, which were performed using an ABI 7300 instrument (Applied Biosystems, CA, USA). GAPDH was used as a normalization control. All primer sequences are compiled in Table 2.

LF cell culture
The isolation of LF cells was conducted as in prior studies [33]. Briefly, collected LF tissue was minced into ~0.5-1 mm 3 cm pieces and was then successively digested with 0.25% trypsin and 250 U/mL type I collagenase (Sigma, USA) in T25 flasks. Cells were then cultured in DMEM containing 10% FBS and 1% penicillin/streptomycin (Gibco, USA) at 37° C in a humidified 5% CO2 incubator until confluent. The media was replaced every third day. Trypsin was used to passage cells when confluent, and experiments were conducted using LF cells from the third passage.

Telomerase activity analysis
Extracts from LF tissues or cells were used to conduct the telomeric repeat amplification protocol (TRAP reaction). Telomerase activity was quantified using the Telo TAGGG Telomerase PCR ELISA kit (Roche, Mannheim, Germany) according to the manufacturer's instructions.

Statistical analysis
Data are means ± standard deviations (SDs) from three independent experiments. GraphPad Prism 5 (GraphPad, CA, USA) was used to compare data via unpaired Student's t-tests or one-way ANOVAs with Tukey's post hoc test as appropriate. A two-tailed P < 0.05 was considered to be statistically significant. Supplementary Figure 3. LFH cells were infected using lentiviral vectors encoding SIRT6 (pSIRT6), an hTERT-specific shRNA (sh-hTERT), or a control shRNA, after which Western blotting was used to assess the expression of SIRT6 in these cells, with α-tubulin being used for normalization. 1, non-infection control; 2, pSIRT6 + shRNA control; 3, sh-hTERT; 4, pSIRT6 + sh-hTERT.