Telomere length and its association with systemic lupus erythematosus in an Asian population: A Mendelian randomization study

Key Messages

(1) Longer telomere length has been linked to a higher risk of SLE in Asian population.

Introduction

Telomere is a highly regulated dynamic complex at the end of chromosomes, consisting of thousands of TTAGGG tandem repeats of short DNA sequences and shelterin protein. It protects the genetic information of chromosomes and can be shortened due to cell division and oxidation, resulting in the loss of structure and function. ¹ Loss of telomeres can lead to cell senescence, apoptosis, and other cellular activities. Inflammation, oxidative stress, and increased cell replication are major environmental factors that can accelerate telomere shortening. Studies have shown that shorter telomeres are causally associated with aging-related outcomes, such as coronary heart disease and Alzheimer's disease among others.

Systemic lupus erythematosus (SLE) is a highly inherited autoimmune disease of unknown etiology and could damage all major organ systems. It mainly occurs in women of childbearing age and can lead to an increased risk of premature death. The incidence and prevalence of SLE vary slightly among countries. The incidence of SLE is 2-7.6/100 000 person-years in the United States, with a prevalence of 19-159/100 000, while that of SLE is about 5/100,000, with an estimated prevalence of 68/100,000 in Sweden. ²

Accelerated aging of the immune system represented by telomere shortening or immune aging is associated with various rheumatic diseases and telomere shortening is considered to be one of the causes of disease onset and progression. ³ Premature atherosclerotic cardiovascular disease is one of the main reasons for the high incidence rate and mortality of SLE patients. While oxidative stress increases in SLE, it also promotes the progression.^4–7 These activities can lead to the shortening of telomere length. Previous studies have shown that the telomere length of mononuclear and polymorphonuclear cells in SLE patients was shortened, especially in those under 45 years old. ⁸ A study involving 154 SLE patients showed that shorter telomere length was observed in SLE patients. ⁹ Similar results were reported in other observational studies.^10–13 A meta-analysis of 472 cases from eight different studies showed that telomere length was significantly shortened in SLE patients regardless of race or assessment method. ⁹ However, because of small sample sizes and residual confounding, it is difficult to draw firm conclusions regarding the causes or consequences between telomere length and SLE. In order to minimize residual confounding bias, we used a two-sample Mendelian randomization (MR) approach to examine the potential association between telomere length ¹⁴ and SLE ¹⁵ from genome-wide association study (GWAS) summary statistics.

Methods

Results

Supplementary Table 1 describes the data and sample size for each cohort. Of the ten SNPs of genome-wide significance, one (rs28365964 and the proxy SNPs) cannot be found in the SLE GWAS and was removed, leaving nine SNPs (the first nine SNPs in Supplementary Table 2) selected as instrumental variables in the primary MR analysis. The F-statistics of the nine SNPs was 861.9, implying strong IVs for the MR analysis.

The scatter plot for nine SNPs effect on TL and SLE was shown in Figure 1 (panel A). Using the IVW method, longer genetically predicted TL was associated with higher odds of SLE (OR = 1.72, 95%CI: 1.21, 2.46, p=0.01). MR-Egger regression (OR = 2.01, 95%CI: 0.88, 4.48, p = 0.14), weighted median (OR = 1.74, 95%CI: 1.26, 2.40, p = 0.01), and weighted mode method (OR = 1.69, 95%CI: 1.03, 2.78, p = 0.05) obtained similar results. These estimates were also plotted in Figure 1 (panel A).

OPEN IN VIEWER

Figure 1.

Panel A. Scatter Plot for the Effects of SNPs on Leukocyte Telomere Length and SLE. IVW: inverse-variance weighted, MR: Mendelian randomization. SLE: systemic lupus erythematosus, SNP: single-nucleotide polymorphisms. The horizontal axis represents the effects of each genetic variant on leukocyte telomere length and the vertical axis denotes the effects of each genetic variant on SLE. The orange lines around the solid black points are the corresponding confidence intervals for the effects. The slopes of solid lines represent the estimates from IVW, weighted median, weighted mode, and MR-Egger regression analyses. Panel B. Single SNP Plot; Panel C. Leave-One-Out Analysis; Panel D. Funnel Plot for the single-nucleotide polymorphisms (SNPS).

We additionally conducted a leave-one-out analysis, which produced comparable results and did not find any significant effect where a single SNP might dominate the results (Figure 1, Panel B and C). Funnel plots do not imply the presence of heterogeneous SNPs (Figure 1, Panel D). By MR-Egger regression, there is no evidence for directional pleiotropy (intercept: 0.017, p = 0.69).

We also used 15 SNPs as IVs for TL that were identified from the trans-ethnic meta-analysis of the TL GWAS as a secondary MR analysis. The results as shown in Supplementary Table 3 were largely comparable to those in Table 1 where nine SNPs were used as the primary IV.

OPEN IN VIEWER

Table 1.

Association between telomere length and systemic lupus erythematosus (SLE) using Mendelian randomization (MR) analysis using the first nine SNPs as IV.

Methods	OR	95% CI	p-value
Inverse-variance weighted	1.72	1.21, 2.46	0.01
MR-Egger regression	2.01	0.88, 4.58	0.14
Weighted median	1.74	1.26, 2.40	0.01
Weighted mode	1.69	1.03, 2.78	0.05

Discussion

Our MR analysis, based on TL and SLE GWAS summary statistics, examined the potential causal relationship between TL and SLE in the Asian population. We compared several MR estimates and concluded that longer TL was associated with a higher risk of SLE. Contrary to previous observational studies, our study provides evidence supporting the potential causal role of TL in SLE risk.

Several studies had examined TL in patients with SLE. The SOLVABLE cohort involving 154 patients found that SLE women had a shorter TL compared with healthy controls. ³ Similar results were observed in a UK cohort with 63 SLE patients ² and SLEIGH study with 59 SLE patients. ¹⁰ These studies, notably, used case-control designs and may be prone to reverse causation and unmeasured confounding. To further examine the relationship between TL and SLE, large-scale prospective cohort studies with sufficient confounders collection are necessary. In our MR study, which could control for unmeasured confounding and reverse causation, we found that the observed association of TL and SLE was in the opposite direction to those reported in the case-control studies. This implies that the shorter TL reported in SLE patients may be a consequence of SLE progression rather than the cause of SLE development. Although the biological mechanisms of longer TL and higher risks of SLE remain unknown, a few potential explanations might help. The higher telomerase activity in CD19 (+) cells may also be concerned with the pathogenesis of SLE. ¹² However, there is a lack of studies concerning the biological functions of TL in SLE. Further research is warranted to clarify how longer TL confers SLE development and progression using cell or animal models.

A few limitations should be acknowledged when interpreting the results. First, the validity of the MR results relies on the assumptions of instrumental variables selected in the analysis. Although the MR approach could minimize bias due to confounding, it cannot capture all confounding factors such as parental confounding. Second, the no-pleiotropy assumption may not hold true. To alleviate such concerns, we performed MR-Egger regression and found no obvious evidence to support pleiotropic effects. Third, because of genetic heterogeneity among different ethnic populations, our analysis is limited to the Asian population and could not be generalized to other ethnic groups. Taken together, our results should be interpreted with caution and may be subject to biases that the MR approach cannot fully control.

Conclusions

In summary, our present study findings support a potential causal relationship between longer TL and higher risks of SLE by considering unmeasured confounding using the MR approach in an Asian population.

Supplemental Material