Abstract
Objectives:
Systemic sclerosis (SSc) is an autoimmune fibrotic disease affecting multiple tissues including the lung. A subset of SSc patients with lung disease exhibit short telomeres in circulating lymphocytes, but the mechanisms underlying this observation are unclear.
Methods:
Sera from the Johns Hopkins and UCSF Scleroderma Centers were screened for autoantibodies targeting telomerase and the shelterin proteins using immunoprecipitation and ELISA. We determined the relationship between autoantibodies targeting the shelterin protein TERF1 and telomere length in peripheral leukocytes measured by qPCR and flow cytometry and fluorescent in-situ hybridization (Flow-FISH). We also explored clinical associations of these autoantibodies.
Results:
In a subset of SSc patients we identified autoantibodies targeting telomerase and the shelterin proteins that were rarely present in rheumatoid arthritis, myositis and healthy controls. TERF1 autoantibodies were present in 40/442 (9.0%) SSc patients and were associated with severe lung disease (OR 2.4, p=0.04, Fisher’s exact test) and short lymphocyte telomere length. 6/6 (100%) patients with TERF1 autoantibodies in the Hopkins Cohort and 14/18 (78%) patients in the UCSF Cohort had a shorter telomere length in lymphocytes or leukocytes, respectively, relative to the expected age-adjusted telomere length. TERF1 autoantibodies were present in 11/152 (7.2%) patients with idiopathic pulmonary fibrosis (IPF), a fibrotic lung disease believed to be mediated by telomere dysfunction.
Conclusions:
Autoantibodies targeting telomere-associated proteins in a subset of SSc patients are associated with short lymphocyte telomere length and lung disease. The specificity of these autoantibodies for SSc and IPF suggests that telomere dysfunction may have a distinct role in the pathogenesis of SSc and pulmonary fibrosis.
Keywords: telomere, systemic sclerosis, interstitial lung disease, idiopathic pulmonary fibrosis
Introduction
Systemic sclerosis (SSc) is an autoimmune chronic fibrosing disease of unknown etiology that results in vasculopathy and multi-organ fibrosis. The disease is heterogeneous with a wide range of possible clinical manifestations that include skin thickening, interstitial lung disease (ILD), and Raynaud’s phenomenon1. The majority of SSc patients develop ILD2, which has some clinical similarities with the progressive lung scarring seen in idiopathic pulmonary fibrosis (IPF)3–5. Telomere dysregulation has been observed in both SSc and IPF6–8, although it remains unclear if there are common mechanistic pathways underlying telomere dysfunction in these diseases.
Telomeres are repetitive nucleotide sequences that protect the ends of chromosomes from deterioration and fusions with neighboring chromosomes. Telomeres shorten with each cell division, serving as a “molecular clock” for cellular aging9. Telomeres are elongated by telomerase containing a telomere-specific reverse transcriptase (hTERT) that adds telomere repeat sequences to the end of telomeres. hTERT is one component of the human telomerase ribonucleoprotein (RNP), which is composed of the telomerase RNA component (hTR), hTERT, and the accessory proteins DKC1, NOP10, NHP2, and GAR110. Other proteins associate with the telomerase complex and act as regulators of telomerase function, including the six shelterin proteins TERF1, TERF2, POT1, TPP1, TIN2L, and RAP111.
Telomere dysregulation is implicated in lung disease associated with IPF and autoimmune disease including SSc12. Germline mutations in hTERT or hTR are present in familial clusters of IPF, and patients with such mutations have markedly shortened telomeres6,13,14. The literature on telomere dysregulation in SSc is conflicting and heterogeneous, in part due to variability in assays used to measure telomere length. Several studies have identified a subgroup of SSc patients with markedly short telomeres in lymphocytes8,15–17 who seem to be at increased risk of ILD15,18. The association between germline mutations in telomere-associated genes and IPF, together with the short telomeres observed in some patients with systemic sclerosis-ILD, raises the possibility that the fibrotic lung disease observed in these two patient subgroups might be phenocopies, potentially representing the consequence of inherited and acquired defects in telomere function.
Distinct SSc clinical phenotypes have been defined by the presence of specific autoantibodies. These autoantibodies often target intracellular nuclear proteins that maintain chromosome structure and function, including proteins involved in mitosis, DNA replication, and DNA repair19,20. Subgrouping SSc by autoantibodies has utility in predicting clinical manifestations, and can provide insights into the biological mechanisms underlying this disease21. Since telomere lengths are relatively short in a subset of patients with systemic sclerosis, we hypothesized that this subgroup may be defined by an immune response with autoantibodies targeting the telomerase complex that is associated with a specific clinical phenotype. In this study, we identify autoantibodies targeting multiple telomere-associated proteins in a subset of SSc patients and demonstrate an association with shortened peripheral leukocyte telomere length and fibrotic lung disease.
Methods
Patient Cohorts
Sera were obtained from consecutive patients who met classification criteria for SSc at the Johns Hopkins (JH) and the University of California, San Francisco (UCSF) Scleroderma Centers. These two independent cohorts have similar databases and collect identical demographic and longitudinal clinical information including pulmonary function test data and organ-specific disease severity assessed by the Medsger Disease Severity Scale (supplemental methods 1). We also assayed sera from healthy controls and patients with myositis, rheumatoid arthritis (RA), and IPF (supplemental methods 2). The Institutional Review Boards at Johns Hopkins and UCSF approved the different components of this study.
Patient and Public Involvement
Patients were recruited to participate in the longitudinal cohorts during routine clinical visits and all patients signed informed consent. Results will be disseminated through conference presentations.
Immunoprecipitation Assays for Autoantibody Detection
Cell lysate immunoprecipitation:
To determine if patient sera contain autoantibodies targeting hTERT, we developed an immunoprecipitation (IP) assay using a cell lysate overexpressing telomerase. A cell line overexpressing the telomerase RNA component (hTR) and FLAG-tagged human telomerase (hTERT) was generated using a Flp-In T-Rex 293 cell line per the manufacturer’s instructions (Thermo Fisher). 50 ug of cell lysate was pre-cleared with protein A beads in NP40 Lysis Buffer (supplemental methods 3) and immunoprecipitated with 1 ul patient serum. Immunoprecipitates were electrophoresed on SDS-PAGE gels blotted with anti-FLAG antibody (Millipore, Sigma) and visualized using enhanced chemiluminescence (ThermoFisher) in a FluorChem M chemi-luminescence imager (ProteinSimple). The data were quantitated by densitometric scanning of the blots and analyzed using ImageJ22. Each sample set was calibrated with the same positive reference IP that was run on each blot. The cut-off for a positive autoantibody was defined as the mean + 4 SD of the healthy controls.
IP using 35S-methionine-labeled proteins:
Complementary DNAs for human POT1, TPP1, TIN2L, TERF1, TERF2, RAP1, NHP2 and DKC1 (GenScript) were used to generate 35S-methioinine-labeled proteins by in vitro transcription and translation (IVTT) per the manufacturer’s protocol (Promega). The radiolabeled proteins were immunoprecipitated with patient sera in Lysis Buffer, and the products were electrophoresed on SDS-PAGE gels and visualized by fluorography23.
TERF1 ELISA
The detailed ELISA protocol is in supplemental methods 3. ELISA plates were coated overnight at 4°C with 200 ng/well of recombinant full-length TERF1 protein (Sino Biological). Patient sera were used at 1:200 dilution and secondary antibodies were horseradish peroxidase-labeled. The color was developed using SureBlue peroxidase reagent (Seracare Life Sciences) and the absorbance was read at 450 nm. The same positive reference serum (with an optical density (OD) in the linear range) was included on every plate as a calibrator. The cutoff for autoantibody positivity was set as the mean plus 4 standard deviations of 50 healthy controls. All positive sera were re-tested by ELISA alongside an uncoated well; ODs of the uncoated wells were subtracted from those obtained with TERF1 coated wells.
Immunoblotting Assays
Recombinant TERF1 protein (200 ng/lane) was electrophoresed on SDS-PAGE gels and transferred to nitrocellulose membranes for the immunoblotting assays. Sera from patients and healthy controls were used at 1:2000 dilution for these immunoblots (see supplemental methods 3 for details).
Other Autoantibody Assays
The JH SSc sera were screened for autoantibodies targeting SSc-associated autoantibodies using the line immunoblot platform (EuroImmun: Systemic Sclerosis [Nucleoli] profile). U1RNP autoantibodies were assayed using a commercially available ELISA (Inova Diagnostics, CA). Euroimmun results were considered positive per the manufacturer’s guidelines (supplemental methods 1). Autoantibodies in the UCSF cohort were derived from clinically indicated commercial testing.
Telomere Length Measurements
Two assays were used to measure telomere length: (i) A PCR-based assay measured telomere length in peripheral leukocytes from the UCSF SSc Cohort as previously described24,25 (supplemental methods 4); and (ii) Flow-FISH was used to measure telomere length in banked peripheral blood mononuclear cells (PBMC’s) prospectively collected from a subset of the JH SSc Cohort with and without TERF1 autoantibodies. Flow-FISH was done on all samples in batch as previously described26,27. Telomere lengths were compared with a validated nomogram of telomere length among healthy controls27.
Statistics
Fisher’s exact test was used to evaluate differences in the frequency of TERF1 autoantibodies between different patient cohorts. The various demographic, clinical, and serologic features of systemic sclerosis, as well as differences in telomere length, were compared between the TERF1 autoantibody-negative and -positive patients using the Wilcoxon rank-sum test or student’s t-test for continuous variables and the Fisher’s exact test for dichotomous variables. All statistical analyses were 2-sided and were conducted using JMP Version 9 (SAS Institute Inc). p<0.05 was considered significant.
Results
Discovery Cohort: Sera from a subset of SSc patients IP telomerase and the shelterin proteins
To test whether SSc patients have autoantibodies targeting telomerase (hTERT), we screened 200 sera from the JH SSc Cohort for these autoantibodies by IP using lysate made from HEK 293 cells overexpressing FLAG-tagged hTERT and telomerase RNA (hTR)28. The IPs were electrophoresed and hTERT was visualized by immunoblotting with anti-FLAG. Of the 200 JH SSc sera screened with this assay, 6 (3.0%) immunoprecipitated hTERT. We did not identify hTERT autoantibodies in 30 healthy control sera (Figures 1A,B).
Figure 1:
Autoantibodies targeting the telomerase/shelterin complex in scleroderma. A&B: Immunoprecipitations (IPs) were performed with patient sera (JH scleroderma cohort, n=200) using lysate made from HEK293 cells overexpressing hTERT-FLAG as input. IPs were detected by blotting with an anti-FLAG antibody. hTERT autoantibodies were found in 6/200 of the scleroderma patients (Scl 1–6) and 0/30 healthy controls. C: IPs were performed with patient sera using the six 35S-methionine-labeled shelterin proteins generated by in vitro transcription and translation. IPs with anti-FLAG was used for positive controls. At least one shelterin autoantibody was detected in 6/200 scleroderma patient sera. In total, 7/200 (4%) scleroderma patients either immunoprecipitated hTERT or had an autoantibody targeting at least one shelterin protein.
The same 200 sera from the JH SSc Cohort were screened for autoantibodies targeting the 6 shelterin proteins (POT1, TPP1, TIN2L, TERF1, TERF2, RAP1) by IP using 35S-methionine-labeled protein generated by IVTT as input. 7/200 (3.5%) SSc sera immunoprecipitated either hTERT or one of the shelterin proteins, and 6 of these patients had multiple telomere-associated autoantibodies (Figure 1C). In contrast, 0/30 healthy controls had a shelterin autoantibody. Representative images of negative controls are shown in supplemental Figure S1. None of the 7 patients with telomere-associated autoantibodies had autoantibodies targeting DKC1 or NHP2.
Validation Cohort: TERF1 Autoantibodies Detected by ELISA in the JH and UCSF SSc Cohorts
As TERF1 was the most common of the shelterin autoantibodies and overlapped with multiple other telomere-associated autoantibodies, we developed an ELISA to screen for TERF1 autoantibodies. 5/6 patients with TERF1 autoantibodies identified by IVTT IP were positive by ELISA (all except Scl 3, Figure 1). In total, the ELISA detected TERF1 autoantibodies in 22/200 (11.0%) of the JH Cohort. As a validation cohort, we screened 242 sera from the UCSF SSc Cohort by ELISA and identified TERF1 autoantibodies in 18/242 (7.4%) patients. Table 1 includes demographic and clinical features of both cohorts. Of the 40 patients total with TERF1 autoantibodies identified by ELISA, 7/40 (18%) were positive by TERF1 IVTT IP. While the ELISA likely detected more positive sera compared to IVTT IP because of differences in TERF1 protein conformation used in the assays, we set up a third assay (immunoblotting of recombinant TERF1 protein) to better address the issue of the discrepant TERF1 autoantibody readouts. Using this, we confirmed the presence of TERF1 autoantibodies in 25/32 (78%) sera that were ELISA-positive but IP-negative using patient sera to immunoblot TERF1 protein (Supplemental Figure S2). We expanded the number of healthy controls screened to 78 and found that the prevalence of TERF1 autoantibodies among SSc patients in both cohorts (40/442 [9.0%]) was significantly higher compared to healthy controls (1/78 [1.3%]), p=0.01).
Table 1.
Demographics and disease characteristics of patients in the Johns Hopkins (JH) and University of California, San Francisco (UCSF) Systemic sclerosis (SSc) Cohorts. 200 patients were in the JH cohort and 242 in the UCSF Cohort. Results are depicted as median with interquartile range (IQR) and mean with standard deviation (SD). Ku autoantibody data was not available from the UCSF cohort. No SSc-specific autoantibody is defined as seronegativity for centromere, Scl70, and RNA polymerase III. MRSS=modified Rodnan Skin Score. Pulmonary function test and echocardiogram data were reported as the maximum (max) right ventricular systolic pressure (RVSP) and the minimum (min) diffusion capacity (DLCO) and forced vital capacity (FVC) recorded in the longitudinal database. Wilcoxon rank-sum test or student’s t-test were used for continuous variables and Fisher’s exact test for dichotomous variables.
| JH SSc Cohort (n=200) | UCSF SSc Cohort (n=242) | P value | |
|---|---|---|---|
| Age (years), mean (SD) | 57.9 (13.4) | 54.6 (13.2) | p=0.009 ** |
| Sex | |||
| Female, n [%] | 171 [85%] | 207 [86%] | P=1.0 |
| Male, n [%] | 29 [15%] | 35 [14%] | |
| Race | |||
| Caucasian, n [%] | 153/198 [77%] | 142/241 [59%] | P<0.0001 *** |
| African American, n [%] | 36/198 [18%] | 23/241 [10%] | |
| Asian or Indian, n [%] | 9/198 [5%] | 76/241 [32%] | |
| SSc Type | |||
| Limited, n [%] | 127 [64%] | 160 [66%] | P=0.62 |
| Diffuse, n [%] | 73 [36%] | 82 [34%] | |
| Disease Duration at time of bleed | |||
| From onset of RP, median (IQR) | 12.5 (6.6–21.4) | 11.1 (5.4–20.0) | P=0.09 |
| From onset of non-RP symptom, median (IQR) | 12.1 (6.1–18.2) | 9.4 (4.3–16.3) | P=0.001 ** |
| Autoantibody status | |||
| Centromere, n [%] | 62/199 [31%] | 60/241 [25%] | P=0.16 |
| U1RNP, n [%] | 17/200 [9%] | 19/237 [8%] | P=0.86 |
| Scl70, n [%] | 44/199 [22%] | 64/241 [27%] | P=0.32 |
| RNA polymerase III, n [%] | 39/199 [20%] | 43/237 [18%] | P=0.81 |
| Ku, n [%] | 11/199 [6%] | ||
| No SSc-specific Ab, n [%] | 64/199 [32%] | 82/241 [34%] | P=0.69 |
| Clinical Features | |||
| History of cancer (ever), n [%] | 38/200 [19%] | 39/242 [16%] | P=0.45 |
| Mortality, n [%] | 16/242 [7%] | 6/200 [3%] | P=0.12 |
| Inflammatory arthritis (ever), n [%] | 40/200 [20%] | 55/242 [23%] | P=0.56 |
| Digital ulceration or gangrene (ever), n [%] | 47/200 [24%] | 100/242 [41%] | P<0.0001 *** |
| SSc renal crisis (ever), n [%] | 5/200 [3%] | 11/242 [5%] | P=0.31 |
| Myopathy (ever), n [%] | 43/200 [22%] | 23/242 [10%] | P=0.0005 *** |
| Max MRSS, mean (SD) | 10.9 (10.4) | 6.6 (7.2) | P<0.0001 *** |
| Severe muscle disease (ever), n [%] | 51/200 [26%] | 18/242 [7%] | P<0.0001 *** |
| Severe heart disease (ever), n [%] | 53/194 [27%] | 52/242 [22%] | P=0.18 |
| Severe lung disease (ever), n [%] | 59/196 [30%] | 92/242 [38%] | P=0.09 |
| Max RVSP (mmHg), mean (SD) | 36.2 (11.8) | 38.4 (21.3) | P=0.17 |
| Min DLCO % predicted, mean (SD) | 64.8 (20.5) | 52.3 (21.3) | P<0.0001 *** |
| Min FVC % predicted, mean (SD) | 74.4 (19.2) | 70.7 (22.6) | p=0.07 |
p< 0.05,
p<0.01,
p<0.001
TERF1 Autoantibodies in Other Rheumatic Diseases
To determine the specificity of TERF1 autoantibodies for systemic sclerosis, we screened 60 RA and 60 myositis sera for TERF1 autoantibodies by ELISA. The mean age of the systemic sclerosis, RA, and myositis cohorts were similar, and detailed demographic and clinical information for these cohorts is in supplemental Tables S1 and S2. In each of the RA and myositis cohorts, 1/60 (1.7%) patients had a positive TERF1 autoantibody, which was similar to healthy controls (Figure 2). TERF1 autoantibodies were significantly more frequent among SSc patients (JH and UCSF combined) compared to RA or myositis (40/442 [9.0%] vs 1/60 [1.7%], p=0.05 in each case).
Figure 2:
TERF1 autoantibodies detected by ELISA in healthy controls (n=78), the combined JH and UCSF scleroderma cohorts (Scl, n=442), rheumatoid arthritis (n=60), and myositis (n=60). Fisher’s exact test was used to compare the frequency of TERF1 autoantibodies between different cohorts. * p< 0.05
TERF1 Autoantibodies and Telomere Length in Leukocytes
We next sought to determine if autoantibodies targeting the telomerase/shelterin complex are associated with abnormalities in telomere length. Telomere length was measured by qPCR in peripheral leukocytes from all UCSF SSc patients using the same banked blood draw from which the TERF1 autoantibodies were assayed. Figure 3A shows telomere length plotted by age for all patients with and without TERF1 autoantibodies. Given that telomeres shorten with a constant linear rate in middle-age29, we calculated each patient’s expected telomere length using a linear regression model based on the relationship between age and telomere length among the TERF1 autoantibody-negative patients (expected telomere length (bp)=7028 – 12.62 *[years of age]). The difference between the patient’s telomere length and the expected telomere length was then calculated for each patient. Compared to patients without TERF1 autoantibodies, significantly more patients with TERF1 autoantibodies had a shorter telomere length than the expected age-adjusted telomere length (14/18 [78%] vs 96/224 [43%]), p=0.006). Furthermore, the difference between the patient telomere length and the expected age-adjusted telomere length was significantly more negative for patients with TERF1 autoantibodies compared to patients without TERF1 autoantibodies (median −230 [IQR −572 to −18] vs 53 [−272 to 304] bp, p=0.01, Wilcoxon rank-sum) (Figure 3B).
Figure 3:
Peripheral blood leukocyte telomere length measured by qPCR in 242 scleroderma patients from the University of California, San Francisco (UCSF) Scleroderma Center. A: Relationship between leukocyte telomere length and age for TERF1 autoantibody positive (n=18) and negative (n=224) patients. B: Patients with TERF1 autoantibodies have a significantly shorter telomere length relative to the expected age-adjusted telomere length compared to patients without TERF1 autoantibodies. Statistics were performed using Wilcoxon rank-sum test, * p< 0.05.
We next sought to confirm the association between TERF1 autoantibodies and short telomeres in the JH SSc Cohort using the Flow-FISH assay, which is known to be more accurate, reproducible, sensitive and specific compared to qPCR16 and can simultaneously differentiate telomere length in lymphocytes and granulocytes. We identified 6 patients with TERF1 autoantibodies and 10 patients without TERF1 autoantibodies who presented for routine clinical visits and agreed to donate PBMC’s. ELISAs performed on serum collected concurrently were used to determine TERF1 autoantibody status. Telomere length was measured on PBMC’s using Flow-FISH. The delta TL (telomere length), which is the difference between the patient’s telomere length and the median telomere length for a healthy person of the same age, was significantly more negative for the TERF1 autoantibody-positive patients compared to the TERF1 autoantibody-negative patients in lymphocytes (median −1132 [IQR −1552 to −996] vs −254 [−950 to 464] bp, p=0.03, Wilcoxon rank-sum). This difference was not observed in granulocytes (median −706 [IQR −1686 to 22] vs −829 [−1122 to −446] bp, p=0.8, Wilcoxon rank-sum) (Figure 4). The two patients with the highest titer hTERT autoantibodies both had telomere lengths below the 10th percentile in lymphocytes and granulocytes.
Figure 4:
Telomere length measured by Flow-FISH (flow cytometry and fluorescent in-situ hybridization) in lymphocytes and granulocytes of 6 TERF1 autoantibody-positive scleroderma patients and 10 TERF1 autoantibody-negative scleroderma patients. A & B: Nomogram of telomere length relative to age in lymphocytes and granulocytes relative to a healthy control population depicted by percentiles. Patients with TERF1 autoantibodies are depicted in blue, and those without TERF1 autoantibodies are in red. C & D: Patients with TERF1 autoantibodies have shorter telomere lengths in lymphocytes (C) but not granulocytes (D) compared to TERF1 autoantibody-negative patients. Delta TL is the difference between the patient telomere length and the median telomere length of a healthy control population. kb=kilobases, bp=base pairs. Statistics were performed with Wilcoxon rank-sum test, * p < 0.05.
Clinical and Serologic Associations with TERF1 Autoantibodies in Systemic sclerosis
After identifying the existence of TERF1 autoantibodies in SSc and demonstrating an association of these autoantibodies with short telomeres in lymphocytes, we next explored associated clinical and serological features (Table 2). The JH and UCSF SSc Cohorts use standardized clinical definitions with harmonization in clinical data acquisition, enabling the evaluation of clinical associations for all 40 TERF1 autoantibody-positive and 402 TERF1 autoantibody-negative patients. The length of clinical follow-up was similar between patients with and without TERF1 autoantibodies (6.8 ± 6.2 vs 6.0 ± 6.0 years, p=0.46). Patients with TERF1 autoantibodies tended to be slightly younger (52.6 ± 13.7 vs 56.4 ± 13.3 years, p=0.10). The presence of TERF1 autoantibodies was significantly associated with a history of severe lung disease (OR 2.4 [CI 1.2–4.8], p= 0.04) and a lower percent predicted diffusion capacity (DLCO) within one year of serum collection (58.0 vs 67.9, p=0.02, student’s t-test). There was also an association with a history of severe muscle disease (OR 3.0 [CI 1.4–6.1], p=0.005) and inflammatory arthritis (OR 2.1 [CI 1.1–4.3], p=0.04). Non-white race was strongly associated with severe lung disease (OR 2.3 [CI 1.5–3.5], p<0.0001) and was also associated with the presence of TERF1 autoantibodies (OR 2.5 [CI 1.3–4.8], p=0.005). The association between TERF1 autoantibodies and severe lung disease was not statistically significant after adjusting for race (OR 1.73 [CI 0.88–3.4] p=0.11).
Table 2.
Clinical and serologic characteristics among TERF1 autoantibody positive (n=40) and negative (n=402) systemic sclerosis (SSc) patients from the Johns Hopkins (JH) and University of California, San Francisco (UCSF) SSc Cohorts. Results are depicted as median with interquartile range (IQR) and mean with standard deviation (SD). No SSc-specific autoantibody is defined as seronegativity for centromere, Scl70, and RNA polymerase III. MRSS=modified Rodnan Skin Score. Pulmonary function test and echocardiogram data were reported as the maximum (max) right ventricular systolic pressure (RVSP) and the minimum (min) diffusion capacity (DLCO) and forced vital capacity (FVC) recorded in the longitudinal database. Wilcoxon rank-sum test or student’s t-test were used for continuous variables and Fisher’s exact test for dichotomous variables.
| TERF1 Ab positive (n= 40) | TERF1 Ab negative (n= 402) | p-value | |
|---|---|---|---|
| Age (years), mean (SD) | 52.6 (13.7) | 56.4 (13.3) | P= 0.10 |
| Sex, female, n [%] | 34 [85%] | 344 [86%] | P= 1.0 |
| Race, Caucasian, n [%] | 19/40 [48%] | 276/399 [69%] | P=0.008 ** |
| African American, n [%] | 10/40 [25%] | 49/399 [12%] | |
| Asian, n [%] | 11/40 [28%] | 74/399 [19%] | |
| SSc Type, limited, n [%] | 24 [60%] | 263 [65%] | P=0.49 |
| Disease duration | |||
| From onset of RP, median (IQR) | 13.9 (7.2–22.5) | 11.6 (5.9–20.6) | P=0.08 |
| From onset of non-RP symptom, median (IQR) | 12.2 (8.4–17.5) | 10.4 (4.5–17.1) | P=0.08 |
| Autoantibody status | |||
| Centromere, n [%] | 10/40 [25%] | 112/400 [28%] | P=0.85 |
| U1RNP, n [%] | 10/40 [25%] | 26/397 [7%] | P=0.0006 *** |
| Scl70, n [%] | 11/40 [28%] | 97/40 [24%] | P=0.70 |
| RNA polymerase III, n [%] | 5/39 [13%] | 77/397 [19%] | P=0.39 |
| Ku, n [%] | 4/22 [18%] | 7/177 [4%] | P=0.02 * |
| No SSc-specific Ab, n [%] | 17/40 [42%] | 129/400 [32%] | P=0.22 |
| Clinical features (ever, max/min) | |||
| History of cancer (ever), n [%] | 7/40 [18%] | 70/402 [17%] | P=1.0 |
| Mortality, n [%] | 1/40 [3%] | 21/402 [5%] | P=0.71 |
| Inflammatory arthritis (ever), n [%] | 14/40 [35%] | 81/402 [2O%] | P=0.04 * |
| Digital ulceration or gangrene (ever), n [%] | 12/40 [30%] | 140/402 [35%] | p=0.60 |
| SSc renal crisis (ever), n [%] | 1/40 [3%] | 15/402 [4%] | P=1.0 |
| Myopathy (ever), n [%] | 8/40 [20%] | 58/402 [14%] | P=0.35 |
| Max MRSS, mean (SD) | 8.0 (9.3) | 8.6 (9.0) | P=0.44 |
| Severe muscle disease (ever), n [%] | 13/40 [33%] | 56/402 [14%] | P=0.005** |
| Severe heart disease (ever), n [%] | 14/39 [36%] | 91/397 [23%] | P=0.08 |
| Severe lung disease (ever), n [%] | 20/40 [50%] | 131/398 [33%] | P=0.04 * |
| Max RVSP (mmHg), mean (SD) | 39.9 (20.1) | 37.3 (17.7) | P=0.46 |
| Min DLCO % predicted, mean (SD) | 53.0 (20.6) | 58.4 (21.9) | P=0.13 |
| Min FVC % predicted, mean (SD) | 66.5 (20.4) | 72.9 (21.2) | P=0.07 |
| DLCO % predicted, mean (SD) | 58.0 (22.5) | 67.9 (23.4) | P=0.02 * |
| FVC % predicted, mean (SD) | 75.0 (21.3) | 80.2 (20.5) | P=0.18 |
p< 0.05,
p<0.01,
p<0.001
TERF1 autoantibodies were associated with U1RNP autoantibodies in the combined cohorts (OR 4.8 [CI 2.1–10.8], p=0.0006) and Ku autoantibodies in the JH cohort (OR 5.4 [CI 1.4–20.2], p=0.02) (Table 2, Figure 5). Ku autoantibody status was not available for the UCSF cohort. There was no association with the frequent SSc-specific autoantibodies anti-Scl-70, anti-centromere, or anti-RNA polymerase III. Absence of these SSc-specific autoantibodies was observed slightly more frequently in patients with TERF1 autoantibodies compared to patients without TERF1 autoantibodies (17/40 [42%] vs 129/400 [32%], p=0.22), although this difference was not significant.
Figure 5:
Overlap among TERF1 (n=22), U1RNP (n=17), and Ku (n=11) autoantibodies in the Hopkins (JH) scleroderma cohort. 40/200 (20%) of patients in the JH cohort had at least one of these autoantibodies.
TERF1 Autoantibodies in Idiopathic Pulmonary Fibrosis
To address whether TERF1 autoantibodies might be present in other syndromes in which telomere dysfunction and lung fibrosis are prominent, we screened 152 patients with IPF and identified TERF1 autoantibodies in 11/152 (7.2%) patients, compared to only 1/78 (1.3%) positives among healthy controls (p=0.06) (Figure 6). Further details on the IPF cohort are in supplemental Table S3. The patient in the IPF cohort with the highest TERF1 autoantibody titer had a positive ANA (1:160, speckled) and subsequently developed symptoms of SSc approximately 2 years later. It was determined that this patient most likely had systemic sclerosis-ILD rather than IPF, although the TERF1 autoantibody had preceded the other clinical features of systemic sclerosis. The other IPF patients with TERF1 autoantibodies did not have a positive ANA and have not, to our knowledge, developed any features of a systemic autoimmune disease.
Figure 6:
TERF1 autoantibodies detected by ELISA in healthy controls (n=78) and the University of California, San Francisco (UCSF) Idiopathic Pulmonary Fibrosis (IPF) cohort (n=152).
Discussion
We describe a subgroup of SSc patients with autoantibodies targeting the telomerase and shelterin complex, characterized by short telomeres in lymphocytes and the presence of lung disease. These autoantibodies are also present in a subset of patients with IPF, but are rarely detected in healthy controls, RA, or myositis. While prior studies have demonstrated telomere dysregulation in SSc30, to our knowledge this is the first description of highly specific autoantibodies targeting the telomerase/shelterin complex in a rheumatic disease. The association of these autoantibodies with short telomeres in systemic sclerosis, and the lower prevalence of these autoantibodies in other chronic inflammatory diseases such as RA that are also known to have telomere dysfunction31, suggests that the mechanism of telomere dysregulation in SSc may be distinct from other inflammatory diseases. The presence of germline mutations in the essential telomerase genes in familial and sporadic IPF supports a causal role for telomere dysfunction in pulmonary fibrosis6,32. However, germline mutations account for only a fraction of IPF cases with short telomeres, suggesting there are other mechanisms of telomere dysregulation33. Our findings raise the possibility that an immune response directed against telomere-associated proteins may also be implicated in telomere shortening in a distinct subgroup of IPF patients.
Although the presence of autoantibodies targeting the shelterin protein TERF1 were enriched among SSc patients with short telomeres, the majority of patients with short telomeres did not have these autoantibodies. Therefore there may be multiple distinct mechanisms leading to short telomeres in systemic sclerosis, and the presence of these autoantibodies may be indicative of one such mechanism. We speculate that the subset of patients with TERF1 autoantibodies have abnormal processing and presentation of telomere-associated proteins, leading to an immune response against the multimolecular telomere complex. In support of this hypothesis, patients developed autoantibodies targeting multiple components of the telomerase and shelterin complexes, suggesting loss of tolerance and epitope spreading across multiple related proteins34.
The association between autoantibodies directed against telomere-associated proteins and short telomeres could also indicate that these autoantibodies exert a directly pathogenic effect on telomeres. We only observed telomere shortening in lymphocytes from SSc patients with TERF1 autoantibodies. It is possible that telomeres in granulocytes might be spared from telomere shortening in the setting of a pathogenic autoantibody because of the short life-span of a granulocyte (only a few days), while lymphocytes survive and circulate for weeks to months. Additional research is needed to understand the biology underlying this highly specific association between telomere-associated autoantibodies and shortened telomeres in SSc and the significance of these autoantibodies in IPF.
The association between lung disease and TERF1 autoantibodies in SSc is consistent with previous studies which have found more severe lung disease among SSc patients with short telomeres in lymphocytes,18,35 and further supports a role for telomere dysfunction in the pathogenesis of SSc lung disease. The patient in the IPF cohort with the highest titer TERF1 autoantibody developed systemic symptoms consistent with SSc several years after enrollment, suggesting that some patients who meet diagnostic criteria for IPF actually have SSc lung disease. TERF1 autoantibodies may serve as an early biomarker to predict subsequent progression to a more definitive diagnosis of systemic sclerosis. This patient was notable for having a high-titer ANA unlike the other IPF patients with TERF1 autoantibodies, suggesting that in most cases TERF1 autoantibodies may still be an indicator of telomere dysfunction in classic IPF.
TERF1 autoantibodies were associated with anti-Ku and anti-U1RNP, specificities which are predictive of SSc overlap syndromes. However, most patients with TERF1 autoantibodies did not have Ku or U1RNP autoantibodies, indicating that TERF1 autoantibodies may provide non-redundant clinical information and could serve as another biomarker for an overlap phenotype. The protein Ku is involved in telomere capping36,37 and it is therefore possible that Ku and TERF1 autoantibodies may both reflect underlying telomere dysfunction.
Strengths of this study are the use of two diverse and well-characterized longitudinal SSc cohorts and the use of two different telomere length assays. There was some discordance in the sensitivities of the 3 assays used to detect TERF1 autoantibodies, and future validation studies are needed to determine the prevalence of these autoantibodies in independent cohorts. Limitations of this study include the small sample size for the Flow-FISH assay, lack of age-matching between the cohorts, and the fact that ILD was not radiographically confirmed on all patients in the SSc cohort. Future studies using RA and myositis cohorts systematically screened for ILD will enable the presence of TERF1 autoantibodies in RA- or myositis-ILD to be evaluated. We had limited power for subgroup analysis, and further studies are needed to characterize the relationship between race, TERF1 autoantibodies, and lung disease severity. While we did not identify an association between TERF1 autoantibodies and FVC in IPF, studies performed on larger cohorts are needed to make any definitive conclusions regarding the relationship between TERF1 autoantibodies and disease severity in IPF.
In summary, we describe a novel subgroup of SSc and IPF patients with autoantibodies targeting the telomerase/shelterin complex that in SSc is associated with short telomeres in peripheral lymphocytes and the presence of lung disease. Telomere-associated autoantibodies may be pathogenically important in the fibrotic lung diseases with telomere dysfunction.
Supplementary Material
Key Messages:
What is already known about this subject?
A subset of systemic sclerosis patients have markedly short telomeres in lymphocytes and a higher prevalence of interstitial lung disease.
What does this study add?
The presence of autoantibodies targeting telomere-associated proteins in systemic sclerosis and their association with short telomeres provides important insights into telomere dysfunction in systemic sclerosis, and raises the possibility that some forms of telomere dysfunction can be acquired through an aberrant immune response.
The association of telomere-associated autoantibodies with interstitial lung disease in systemic sclerosis and the presence of these autoantibodies in idiopathic pulmonary fibrosis supports a role of telomere dysregulation in pulmonary fibrosis.
How might this impact clinical practice?
These autoantibodies could serve as novel biomarkers for systemic sclerosis and specifically for systemic sclerosis lung disease.
Acknowledgments
The cell line overexpressing hTERT and hTR was generated by Alexandra Pike. We thank the Johns Hopkins Myositis Center and the Johns Hopkins Arthritis Center for providing patient samples and Dr Erika Darrah for providing healthy control serum. We also thank Dr. Mary Armanios for helping with Flow-FISH through the Pathology Flow-FISH Core.
Funding:
BA was supported by the Jerome L. Greene Foundation, Scleroderma Research Foundation, and T32 AR048522. LCR and AAS were supported in part by the Donald B. and Dorothy L. Stabler Foundation. The UCSF Scleroderma Cohort was supported by the Lennox Foundation and the Scleroderma Research Foundation. The UCSF ILD Cohort was funded by the Nina Ireland Program for Lung Health. The Johns Hopkins Scleroderma Center Research Registry receives support from the Johns Hopkins inHealth Precision Medicine Initiative, the Scleroderma Research Foundation, and the Chresanthe Staurulakis Memorial Discovery Fund. Samples from the Rheumatoid Arthritis Cohort were obtained through a pilot project supported by the Patient Centered Outcomes Research Institute (PCORI), IP2-PI000737 and additionally supported by the Camille Julia Morgan Arthritis Research and Educational Fund. This study was additionally supported by P30-AR053503 and P30-AR070254. Funding of the Myositis Cohort is from the Huayi and Siuling Zhang Discovery Fund and the Peter Buck Foundation.
Footnotes
There are no competing interests for any author.
This study was approved by the Hopkins IRB (IRB00239503)
The data are stored as deidentified participant data which is available upon request to Brittany Adler (Brittany.adler@jhmi.edu)
Publisher's Disclaimer: Disclaimer: The findings of this study including its conclusions represent those of the authors and do not reflect those of the National Institutes of Health (NIH), the National Institute of Arthritis Musculoskeletal and Skin Diseases (NIAMS), the Patient Centered Outcomes Research Institute (PCORI), its Board of Governors or its Methodology Committee.
References
- 1.Jaeger VK, Wirz EG, Allanore Y, et al. Incidences and Risk Factors of Organ Manifestations in the Early Course of Systemic Sclerosis: A Longitudinal EUSTAR Study. Assassi S, ed. PLoS One. 2016;11(10):e0163894. doi: 10.1371/journal.pone.0163894 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bussone G, Mouthon L. Interstitial lung disease in systemic sclerosis. Autoimmun Rev. 2011;10(5):248–255. doi: 10.1016/j.autrev.2010.09.012 [DOI] [PubMed] [Google Scholar]
- 3.Desai SR, Veeraraghavan S, Hansell DM, et al. CT features of lung disease in patients with systemic scerosis: Comparison with idiopathic pulmonary fibrosis and nonspecific interstitial pneumonia. Radiology. 2004;232(2):560–567. doi: 10.1148/radiol.2322031223 [DOI] [PubMed] [Google Scholar]
- 4.Huang J, Beyer C, Palumbo-Zerr K, et al. Nintedanib inhibits fibroblast activation and ameliorates fibrosis in preclinical models of systemic sclerosis. Ann Rheum Dis. 2016;75(5):883–890. doi: 10.1136/annrheumdis-2014-207109 [DOI] [PubMed] [Google Scholar]
- 5.Herzog EL, Mathur A, Tager AM, Feghali-Bostwick C, Schneider F, Varga J. Review: Interstitial lung disease associated with systemic sclerosis and idiopathic pulmonary fibrosis: How similar and distinct? Arthritis Rheumatol. 2014;66(8):1967–1978. doi: 10.1002/art.38702 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Armanios MY, Chen JJ-L, Cogan JD, et al. Telomerase Mutations in Families with Idiopathic Pulmonary Fibrosis. N Engl J Med. 2007;356(13):1317–1326. doi: 10.1056/NEJMoa066157 [DOI] [PubMed] [Google Scholar]
- 7.Bilgili H, Białas AJ, Górski P, Piotrowski WJ. Telomere Abnormalities in the Pathobiology of Idiopathic Pulmonary Fibrosis. J Clin Med. 2019;8(8):1232. doi: 10.3390/jcm8081232 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Artlett CM, Black CM, Briggs DC, Stevens CO, Welsh KI. Telomere reduction in scleroderma patients: A possible cause for chromosomal instability. Br J Rheumatol. 1996;35(8):732–737. doi: 10.1093/rheumatology/35.8.732 [DOI] [PubMed] [Google Scholar]
- 9.Levy MZ, Allsopp RC, Futcher AB, Greider CW, Harley CB. Telomere end-replication problem and cell aging. J Mol Biol. 1992;225(4):951–960. doi: 10.1016/0022-2836(92)90096-3 [DOI] [PubMed] [Google Scholar]
- 10.Schmidt JC, Cech TR. Human telomerase: biogenesis, trafficking, recruitment, and activation. Genes Dev. 2015;29(11):1095–1105. doi: 10.1101/gad.263863.115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Lange T Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev. 2005;19(18):2100–2110. doi: 10.1101/gad.1346005 [DOI] [PubMed] [Google Scholar]
- 12.Newton CA, Oldham JM, Ley B, et al. Telomere length and genetic variant associations with interstitial lung disease progression and survival. Eur Respir J. 2019;53(4). doi: 10.1183/13993003.01641-2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tsakiri KD, Cronkhite JT, Kuan PJ, et al. Adult-onset pulmonary fibrosis caused by mutations in telomerase. Proc Natl Acad Sci U S A. 2007;104(18):7552–7557. doi: 10.1073/pnas.0701009104 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Garcia CK. Insights from human genetic studies of lung and organ fibrosis. J Clin Invest. 2018;128(1):36–44. doi: 10.1172/JCI93556 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lakota K, Hanumanthu VS, Agrawal R, Carns M, Armanios M, Varga J. Short lymphocyte, but not granulocyte, telomere length in a subset of patients with systemic sclerosis. Ann Rheum Dis. Published online January 24, 2019:annrheumdis-2018–214499. doi: 10.1136/ANNRHEUMDIS-2018-214499 [DOI] [PubMed] [Google Scholar]
- 16.Gutierrez-Rodrigues F, Santana-Lemos BA, Scheucher PS, Alves-Paiva RM, Calado RT. Direct Comparison of Flow-FISH and qPCR as Diagnostic Tests for Telomere Length Measurement in Humans. PLoS One. 2014;9(11):113747. doi: 10.1371/journal.pone.0113747 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Katayama Y, Kohriyama K. Telomerase activity in peripheral blood mononuclear cells of systemic connective tissue diseases. J Rheumatol. 2001;28(2):288–291. [PubMed] [Google Scholar]
- 18.Adamali HI, Delgado CM, Stock C, et al. Telomere (TL) shortening is associated with disease severity in scleroderma (SSC) associated interstitial lung disease (ILD). Eur Respir J. 2012;40(Suppl 56). [Google Scholar]
- 19.Moroi Y, Peebles C, Fritzler MJ, Steigerwald J, Tan EM. Autoantibody to centromere (kinetochore) in scleroderma sera. Proc Natl Acad Sci U S A. 1980;77(3 I):1627–1631. doi: 10.1073/pnas.77.3.1627 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schild-Poulter C, Su A, Shih A, et al. Association of autoantibodies with Ku and DNA repair proteins in connective tissue diseases. Rheumatology. 2007;47(2):165–171. doi: 10.1093/rheumatology/kem338 [DOI] [PubMed] [Google Scholar]
- 21.Ho KT, Reveille JD. The clinical relevance of autoantibodies in scleroderma. Arthritis Res Ther. 2003;5(2):80–93. doi: 10.1186/ar628 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012;9(7):671–675. doi: 10.1038/nmeth.2089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fiorentino D, Chung L, Zwerner J, Rosen A, Casciola-Rosen L. The mucocutaneous and systemic phenotype of dermatomyositis patients with antibodies to MDA5 (CADM-140): A retrospective study. J Am Acad Dermatol. 2011;65(1):25–34. doi: 10.1016/j.jaad.2010.09.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Vasilishina A, Kropotov A, Spivak I, Bernadotte A. Relative Human Telomere Length Quantification by Real-Time PCR. In: ; 2019:39–44. doi: 10.1007/978-1-4939-8931-7_5 [DOI] [PubMed] [Google Scholar]
- 25.Liu S, Wang C, Green G, et al. Peripheral blood leukocyte telomere length is associated with survival of sepsis patients. Eur Respir J. 2020;55(1). doi: 10.1183/13993003.01044-2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Baerlocher GM, Vulto I, de Jong G, Lansdorp PM. Flow cytometry and FISH to measure the average length of telomeres (flow FISH). Nat Protoc. 2006;1(5):2365–2376. doi: 10.1038/nprot.2006.263 [DOI] [PubMed] [Google Scholar]
- 27.Alder JK, Hanumanthu VS, Strong MA, et al. Diagnostic utility of telomere length testing in a hospital-based setting. Proc Natl Acad Sci U S A. 2018;115(10):E2358–E2365. doi: 10.1073/pnas.1720427115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Pike A, Strong M, Ouyang JP, Connelly C, Greider C. TIN2 functions with TPP1/POT1 to stimulate telomerase processivity. bioRxiv. Published online October 5, 2018:435958. doi: 10.1101/435958 [DOI] [Google Scholar]
- 29.Whittemore K, Vera E, Martínez-Nevado E, Sanpera C, Blasco MA. Telomere shortening rate predicts species life span. Proc Natl Acad Sci U S A. 2019;116(30):15122–15127. doi: 10.1073/pnas.1902452116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hohensinner PJ, Goronzy JJ, Weyand CM. Telomere dysfunction, autoimmunity and aging. Aging Dis. 2011;2(6):524–537. [PMC free article] [PubMed] [Google Scholar]
- 31.Schönland SO, Lopez C, Widmann T, et al. Premature telomeric loss in rheumatoid arthritis is genetically determined and involves both myeloid and lymphoid cell lineages. Proc Natl Acad Sci U S A. 2003;100(23):13471–13476. doi: 10.1073/pnas.2233561100 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Alder JK, Chen JJL, Lancaster L, et al. Short telomeres are a risk factor for idiopathic pulmonary fibrosis. Proc Natl Acad Sci U S A. 2008;105(35):13051–13056. doi: 10.1073/pnas.0804280105 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cronkhite JT, Xing C, Raghu G, et al. Telomere shortening in familial and sporadic pulmonary fibrosis. Am J Respir Crit Care Med. 2008;178(7):729–737. doi: 10.1164/rccm.200804-550OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Vanderlugt CJ, Miller SD. Epitope spreading. Curr Opin Immunol. 1996;8(6):831–836. doi: 10.1016/S0952-7915(96)80012-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Lakota K, Hanumanthu VS, Agrawal R, Carns M, Armanios M, Varga J. Short lymphocyte, but not granulocyte, telomere length in a subset of patients with systemic sclerosis. Ann Rheum Dis. Published online January 24, 2019:annrheumdis-2018–214499. doi: 10.1136/annrheumdis-2018-214499 [DOI] [PubMed] [Google Scholar]
- 36.Ghosh G, Li G, Myung K, Hendrickson EA. The lethality of Ku86 (XRCC5) loss-of-function mutations in human cells is independent of p53 (TP53). Radiat Res. 2007;167(1):66–79. doi: 10.1667/RR0692.1 [DOI] [PubMed] [Google Scholar]
- 37.Hsu HL, Gilley D, Galande SA, et al. Ku acts in a unique way at the mammalian telomere to prevent end joining. Genes Dev. 2000;14(22):2807–2812. doi: 10.1101/gad.844000 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Greidinger EL, Flaherty KT, White B, Rosen A, Wigley FM, Wise RA. African-American race and antibodies to topoisomerase I are associated with increased severity of scleroderma lung disease. Chest. 1998;114(3):801–807. doi: 10.1378/chest.114.3.801 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.