Telomere length and hTERT genetic variants as potential prognostic markers in multiple myeloma

Telomere dysfunction is a notable event observed in many cancers contributing to their genomic instability. A major factor controlling telomere stability is the human telomerase reverse transcriptase catalytic subunit (hTERT). Telomere shortening has been observed in multiple myeloma (MM), a plasma cell malignancy with a complex and heterogeneous genetic background. In the present study, we aimed to analyse telomere length and hTERT genetic variants as potential markers of risk and survival in 251 MM patients. We found that telomere length was significantly shorter in MM patients than in healthy individuals, and patients with more advanced disease (stage III according to the International Staging System) had shorter telomeres than patients with less advanced disease. MM patients with hTERT allele rs2736100 T were characterized with significantly shorter progression-free survival (PFS). Moreover, allele rs2736100 T was also found to be less common in patients with disease progression in response to treatment. hTERT rs2853690 T was associated with higher haemoglobin blood levels and lower C-reactive protein. In conclusion, our results suggest that telomere length and hTERT genetic variability may affect MM development and can be potential prognostic markers in this disease.

concept of protection of critically short telomeric DNA by telomerase 11 .Nevertheless, it seems that the telomere shortening in haematological malignancies might still be associated with occurrence of aberrant karyotype 12,13 .In addition, inhibition of telomerase activity by GRN163L (Imetelstat), a lipid-conjugated thio-phosphoramidate oligonucleotide, was shown to be effective in the treatment of MM patients in both in vitro and in vivo studies 14 .Short telomeres may be responsible for chemotherapy resistance, as telomere shortening can result in adaptation of malignant cells, allowing their increased proliferation to overcome the cytotoxicity of treatment 15 .
The occurrence of telomere shortening also depends on the expression level, promoter mutations and genetic variability within the gene coding for hTERT 16,17 .The hTERT gene is located on the shorter arm of chromosome 5 (5p15.33)and consists of 15 introns, 16 exons and a 260 bp promoter core 18 .Aberrant gene variants may play a key role in transforming somatic cells into malignant cells by activating telomerase and other related signalling pathways, e.g.Wnt/β-catenin pathway 19,20 .
In the present study, we aimed to assess telomere length and hTERT genetic variants as potential markers associated with risk, survival, response to treatment and clinical course of the disease in MM.For this purpose, telomere length as well as six SNPs (rs2853690, rs2736100, rs33954691, rs35033501, rs2735940, rs10069690) located within the hTERT gene were analysed.

Telomere length in multiple myeloma patients and healthy individuals
We observed that telomere length was significantly shorter in MM patients than in healthy individuals (median length 2.35 vs 4.96, p < 0.001, Fig. 1).This association was further confirmed in a multivariate generalized linear model analysis which included, alongside telomere length, also age and sex of patients and controls.Telomere length was proved to be an independent factor of MM risk (p = 0.010, Table 1).
Telomere length was also shorter in patients with more advanced disease (stage III according to the International Staging System; ISS) than in patients with less advanced disease (median length 2.27 vs 2.81, p = 0.031, Fig. 2).Nevertheless, patients with less advanced disease (ISS I-II) still had significantly shorter telomeres than healthy individuals (median length 2.81 vs 4.96, p < 0.001, Fig. 2).Telomere length also weakly correlated with blood albumin concentration (R = 0.293, p = 0.004).No associations with overall or progression-free survival or other clinical parameters was observed.

hTERT genetic polymorphism and MM survival
Genotype frequencies for all the six analysed hTERT SNPs (rs2853690, rs2736100, rs33954691, rs35033501, rs2735940, rs10069690) are presented in Table 2.All the SNPs were in Hardy-Weinberg equilibrium in both patients and healthy controls.Linkage disequilibrium (LD) analysis showed that rs2736100, rs2735940, rs10069690 were in relatively high LD (Fig. 3).No difference was observed in the allele and genotype frequencies between MM patients and healthy individuals.We performed an analysis of both overall (OS) and progression-free survival (PFS) to test if any of the hTERT SNPs might affect it.No association with OS was observed.We noticed that alleles rs2736100 T and rs2735940 C corresponded with shorter PFS during the first 3 years, but longer PFS thereafter (Supplementary Fig. S1).When analysing only early survival (first 3 years), patients with allele rs2736100 T were characterized with significantly shorter PFS than patients without this allele (p = 0.043, Fig. 4a).A similar association, although not statistically significant, was also observed for allele rs2735940 C (p = 0.057, Fig. 4b) and for the GTC (rs10069690 G, rs2736100 T, rs2735940 C, based on the LD data as shown in Fig. 3) haplotype (p = 0.069, Fig. 4c).These  observations were confirmed in analyses in a Cox proportional hazards regression model including age and later ISS stages in addition to each of the risk alleles/haplotypes.Allele rs2736100 T (p = 0.035), but not allele rs2735940 C (p = 0.102) or haplotype GTC (p = 0.075), was found to be an independent factor of shorter PFS.When stratified according to the ISS criteria, patients with less advanced disease (ISS I) on diagnosis more commonly carried the rs2853690 T allele than those diagnosed with more advanced (ISS II-III) disease (22/61 vs 35/167, p = 0.025, Fig. 5a).Similarly, patients with less advanced disease (ISS I) also carried allele rs33954691 T more commonly than other patients (13/61 vs 17/167, p = 0.044, Fig. 5b).Detailed information on distribution of hTERT genotypes in patients with different ISS stages is shown in Supplementary Table S1.Allele rs2853690 T was also associated with higher haemoglobin blood levels (median value 11.7 vs 10.7 g/dL; p = 0.006) and lower  www.nature.com/scientificreports/C-reactive protein (1.8 vs 4.6 mg/dL; p = 0.034) compared to patients without this allele.Interestingly, we also observed that alleles rs2736100 T, rs2735940 C, and haplotype GTC were less common in patients with progressive disease in response to treatment (p = 0.024, p = 0.020, p = 0.031 for the two alleles and the GTC haplotype, respectively).We also analysed differences in telomere length between carriers of various hTERT alleles.MM patients with alleles rs10069690 G had longer telomeres than patients without this allele (2.33 vs 3.99 kb, p = 0.047).No associations with other SNPs included in this study were observed.

Discussion
The current knowledge of human cancer development shows that telomere dysfunction may be a key event causing genomic instability and disease progression in many types of solid tumours e.g.renal cell carcinoma 22 , glioma 23,24 , esophageal squamous cell carcinoma 25 , gastric cancer 26 , ovarian and breast cancer [27][28][29][30] .Additionally, telomere shortening has also been observed in haematological malignancies, both acute and chronic leukaemias [31][32][33][34] , some lymphomas [35][36][37] , as well as bone marrow failure syndromes (myelodysplastic syndrome (MDS)) 38,39 and aplastic anaemia [40][41][42] .It is not surprising that diseases related to telomere length manifest so prominently in the bone marrow, given that it is a place where turnover of cells is remarkably high, with app. 10 9 cells being produced every hour 43 .Haematopoiesis is based on the ability of stem cells to self-renewal and differentiation.In humans, there is clear evidence that short telomeres cause quantitative and qualitative defects in HSCs that manifest as stem cell depletion [44][45][46][47] .
MM is a very complex haematologic malignancy with a heterogenous genomic landscape, involving accumulation of mutations, as well as structural and copy number changes.These genetic variations occur in combination with point mutations and affect various cellular pathways, including genome maintenance 3 .Telomere dysfunction is one of the mechanisms that may lead to the genetic and clinical heterogeneity observed in MM, therefore analysis of telomere length may have prognostic significance.Our present study was conducted on a group of newly diagnosed untreated MM patients.We observed that significant shortening of telomere length in MM patients compared to healthy individuals.This is contrast to a study by Campa et al., as they found a link between longer telomeres and an increased risk of MM 48 .However, most other studies seem to confirm our observation.Similar results were reported by Wu et al. in a study analysing CD138 positive cells isolated from the bone marrow of patients with newly diagnosed or relapsed MM and healthy donors.These results showed significantly reduced telomere length in MM patients compared to telomere length in plasma cells from healthy donors 11 .Cottliar et al. studied bone marrow cells from patients with MM and with MGUS.Similarly to our results, they observed short telomeres in MM patients at diagnosis and during relapse.Additionally, they noted that telomere length in bone marrow cells was restored after disease remission 49 .In a three-dimensional (3D) telomere analysis, Klewes et al. showed changes in 3D nuclear architecture during disease progression from MGUS to MM, resulting in increased telomere attrition and consequent shorter telomeres in MM, as well as in relapsed MM, compared to MGUS 50 .
Hyatt et al. stratified their group of MM patients using a length threshold for telomere dysfunction (telomere length below 3.81 kb) previously defined by Lin et al. 51,52 .They showed that MM patients with short telomeres (< 3.81 kb) had significantly shorter overall survival.In their study, short telomeres were also associated with significantly worse survival in high-risk ISS patients.Moreover, in a multivariate modelling analysis, telomere length as well as patient's age and ISS were identified as predictors of MM risk.Based on these data, Hyatt et al. proposed that each ISS prognostic subset could be further stratified for risk according to telomere length, supporting the inclusion of this parameter as a refinement of the ISS system 51 .
Telomere length can be influenced by genetic variability, including mutations in the telomerase components hTERT and TERC, and various single-nucleotide polymorphisms (SNPs), some of which are located near genes with known roles in telomere maintenance 53 .In addition, SNPs can affect telomere length, as their variants can modulate the expression of a broad spectrum of genes e.g.Myc, TP53 and NFKB 30,54,55 .For example, − 1327C>T (rs2735940) is a SNP located in the promoter region of hTERT gene (hTERTp) and is a T/C transition 1327-bp upstream of the transcription start site and is able to induce expression of hTERT by upregulating its transcriptional activity in vitro and hTERT mRNA expression in vivo 53 .This SNP has been extensively analysed in various types of solid cancers.However, it has not been studied much in haematological diseases, except for one study on patients with childhood ALL 56 .In this study, we observed that MM patients carrying C allele hTERTp rs2735940 had shorter PFS during the first 3 years.Sheng et al. reported that the TT genotype and T allele were associated with ALL in Chinese children 56 .However, conflicting results were obtained by Eskandari et al., who concluded that rs2735940 does not increase the risk of ALL 57 .In patients with renal cell carcinoma, the CC genotype hTERTp rs2735940 was associated with shorter time to disease progression and shorter overall survival 58 .In our present study, we did not observe any significant associations between rs2735940 and telomere length.However, there are reports suggesting that hTERT rs2735940 polymorphism affects telomere length and that longer telomeres were associated with an increased breast cancer (BC) and lung cancer risk 59,60 .Our previous study showed that women with BC and the CC genotype had longer telomeres than those with TC and TT genotypes.Additionally, BC patients with the rs2735940 C allele were characterized by more invasive tumours than patients with the TT genotype 30 .
The next SNP, rs2736100, is located in intron 2 of the hTERT gene and based on the result of the Evolutionary and Sequence Pattern Extraction through Reduced Representation score, it is located within a putative regulatory region 61,62 .It is the most frequently analysed polymorphic variant of hTERT.The C allele of rs2736100 was found to be associated with longer telomeres, which is consistent with the direct regulatory effect of the rs2736100 genotype on hTERT gene expression 29 .Interestingly, the effect strength of the rs2736100 polymorphism may vary between populations, as demonstrated, e.g. in Swedish and Chinese males with myeloproliferative neoplasms (MPNs) 63 .In addition, rs2736100 C is also associated with increased blood cell count, a hallmark of MPN 64,65 , and has been identified as a risk variant for MPN in the Icelandic population 64 67 .In the present study, we observed that the T allele rs2736100 was associated with shorter early PFS in MM patients.We confirmed this observation (for T allele rs2736100) by a Cox proportional hazards regression model analysis, including age and later ISS stages in addition to each of the risk alleles.In our previous study, patients with CLL carrying C allele were characterized by longer telomeres with less advanced disease (Rai 0-I or Binet A) compared to patients with the C allele, but exhibiting more advanced stage of CLL 32 .In women with BC we showed that patients with T allele rs2736100 had more invasive tumours than BC patients with other genotypes 30 .
Another SNP, rs10069690, was much more extensively studied.It is known to act as a risk factor in many cancers 68 and was found to be a marker of decreased risk for MM in an earlier study 48 .However, we were not able to confirm this association in our present study.Three other SNPs included in our study were predicted to affect either miRNA binding (rs2853690) or splicing (rs33954691, rs35033501) based on an analysis by the National Institute of Environmental Health Sciences SNP Function Prediction tool 69 .Of these, we found rs2853690 T and rs33954691 T to be more common in patients with less advanced disease (lower ISS stage).Additionally, rs2853690 T was also associated with higher haemoglobin and lower C-reactive protein levels, low haemoglobin and high CRP being common features of MM progression 70,71 .All three of those SNPs have been relatively poorly studied.The rs2853690 is known to affect circulating hTERT mRNA levels and response to treatment in patients with rectal cancer, as well as spontaneous preterm labour in pregnant women 67,72 .rs33954691 was shown to affect longevity and leucocyte telomere length 73,74 .It was also associated with increased risk of radioiodine-refractory papillary thyroid carcinoma 75 .All three SNPs were analysed in our previous study on chronic lymphocytic leukaemia (CLL), although only rs35033501 was associated with CLL risk 32 .
The regulation of the hTERT gene is a very complex process and hTERTp can be activated by multiple mechanisms in haematological malignancies.Moreover, genetic variation in hTERT may modulate telomere length and thus such genetic variants may be risk factors for cancer development.So, it appears that the optimal telomere length is a balance of cell proliferation, senescence and control.Shortening of telomeres to critical length results in a loss of telomere protection, leading to chromosomal instability, which can contribute to the abnormalities in the hematopoietic process.

Patients and controls
The study included 251 newly diagnosed Polish MM patients and 226 healthy blood donors serving as the control group.Both groups were nearly equally divided into men and women (the ratio of females was 0.502 and 0.420, respectively).Blood samples were collected at diagnosis after obtaining informed consent from patients.All methods were used according to the Declaration of Helsinki.The study was approved by the Wroclaw Medical University Bioethical Committee (ethical approval code: 369/2019).According to International Staging System (ISS) stratification, 61 (24.3%) patients were in stage I at diagnosis, 76 (30.3%) were in stage II, 91 (36.3%) were in stage III, and 23 lacked data on ISS.Most patients were administered either cyclophosphamide, thalidomide, dexamethasone (CTD)-38.8%,bortezomib, thalidomide, dexamethasone (VTD)-13.2%,or bortezomib, melphalan, prednisone (VMP)-11.0%as first line therapy.Response to treatment was as follows: complete response (n = 43), very good partial response (n = 32), partial response (n = 81), minor response (n = 7), stable disease (n = 20), progressive disease (n = 19).

Figure 1 .
Figure 1.Telomere length in multiple myeloma (MM) patients and healthy individuals.

Figure 2 .
Figure 2. Telomere length in MM patients with less advanced (stages I-II according to the ISS) and more advanced (stage III) disease.Healthy individuals are also included for comparison.Patients with more advanced disease are characterized by shorter telomeres than those with less advanced disease, although both groups of patients have significantly shorter telomeres than healthy individuals.

Figure 3 .
Figure 3. Linkage disequilibrium between hTERT SNPs included in the study.Darker colours represent higher D′ values.Results were obtained from the Haploview 4.2 software 21 .

Figure 5 .
Figure 5. Frequency of alleles rs2853690 T (a) and rs33954691 T (b) among patients with more and less advanced disease at diagnosis according to the ISS criteria.Both alleles are more common in patients with ISS I.

Table 1 .
Results of a multivariate logistic regression analysis of multiple myeloma (MM) risk.Significant values are in bold.

Table 2 .
hTERT genotype frequencies in multiple myeloma (MM) patients and healthy individuals.
MM patients (N =

251) Healthy individuals (N = 226)
. Tong et al. observed a higher frequency of the CC genotype and C allele in Chinese AML patients 66 .Furthermore, Rampazzo et al. observed longer telomeres at diagnosis and greater telomere erosion during neoadjuvant chemoradiotherapy in patients with rectal cancer with CC genotype compared to patients with the other genotypes