SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To investigate whether sera or purified IgG from patients with polymyositis (PM) and patients with dermatomyositis (DM), with or without interstitial lung disease (ILD), can activate endothelial cells (ECs).

Methods

Patients' sera were selected based on the presence or absence of anti–Jo-1, anti-SSA, or anti–U1 small nuclear RNP autoantibodies. The presence of autoantibodies was determined by line blot assays. Cultured human microvascular ECs derived from lung tissue (HMVEC-L) were incubated with sera or purified IgG from 22 patients with PM, 7 patients with DM, and 10 healthy individuals as controls. Assessment of intercellular adhesion molecule 1 (ICAM-1) expression was conducted by immunofluorescence (n = 22) and by cell-based enzyme-linked immunosorbent assay (ELISA) (n = 20). Serum levels of soluble ICAM-1 (sICAM-1) were determined by ELISA.

Results

Sera from PM patients with ILD who were positive for anti–Jo-1 autoantibodies had a significantly stronger effect on the expression of ICAM-1 by HMVEC-L in comparison with sera from healthy controls and patients with other autoantibodies. Purified IgG did not induce ICAM-1 expression. Higher serum levels of sICAM-1 were found in patients with myositis compared with healthy controls.

Conclusion

EC activation with ICAM-1 expression could contribute to the multiorgan involvement, including the development of myositis and ILD, in patients carrying anti–Jo-1 autoantibodies. The EC-activating factors are not the autoantibodies themselves, but might be systemic factors associated with these autoantibodies.

Polymyositis (PM) and dermatomyositis (DM) are inflammatory myopathies characterized by muscle weakness and inflammatory infiltrates in skeletal muscle tissue. Interstitial lung disease (ILD) is a common extramuscular manifestation. Autoantibodies, including anti–histidyl–transfer RNA synthetase (anti–Jo-1), anti-SSA (anti–Ro 52/anti–Ro 60), anti–U1 small nuclear RNPs (anti-RNPs), and anti–signal recognition particle (anti-SRP), are often present (1).

Multiorgan involvement in myositis could possibly be explained by the contribution of microvessels. Endothelial cells (ECs) of the microvessels in muscle tissue both in patients with PM and in patients with DM express activation markers, such as intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule, and the proinflammatory cytokine interleukin-1α (2, 3). It is not known whether these ECs are activated via systemic factors or via local factors. The aim of our study was to investigate whether sera or purified IgG from patients with myositis, with or without ILD and different autoantibody specificities, could activate human lung ECs.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Sample collection.

Sera from 22 patients with PM and 7 patients with DM (median age 50 years [range 20–79 years], median disease duration 0.07 years [range 0–22 years]) (4) were selected based on the presence or absence of anti–Jo-1, anti–Ro 52/anti–Ro 60, and anti-RNP autoantibodies, identified from review of the patients' records (Table 1). Definitions of ILD were applied as described previously (5). Levels of creatine phosphokinase (CPK) and the erythrocyte sedimentation rate (ESR) were retrieved from the patients' records. Due to limited availability of the sera, some samples were available for analysis by only 1 of the assessment methods; for stimulation experiments assessed by immunofluorescence, 18 PM samples and 4 DM samples were used, while for cell-based enzyme-linked immunosorbent assay (ELISA), 16 PM samples and 4 DM samples were used (Table 1). Soluble ICAM-1 (sICAM-1) levels in the sera and cell supernatants were measured by ELISA (n = 22 patients). The control group consisted of 10 healthy individuals (median age 50.5 years [range 27–73 years], 6 women and 4 men).

Table 1. Demographic characteristics and clinical, laboratory, and serologic data in patients with PM or DM at the time of serum sampling*
Patient/age/sexDiagnosisDisease duration, yearsCPK, μcat/literESR, mm/hourPositive autoantibody (levels)MedicationCumulative clinical manifestations/overlap syndromes
  • *

    All serum samples were subjected to stimulation and assessed by both immunofluorescence and cell-based enzyme-linked immunosorbent assay (ELISA) (patients 1, 2, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, and 20), by cell-based ELISA only (patients 3, 4, 5, 6, 13, 18, and 19), or by immunofluorescence only (patients 21, 22, 23, 24, 25, 26, 27, 28, and 29). CPK = creatine phosphokinase (normal <2.5 μcat/liter in men and <2.0 μcat/liter in women [for units/liter, factor by 60]); ESR = erythrocyte sedimentation rate (normal <30 mm/hour); PM = polymyositis; ILD = interstitial lung disease; Pred. = prednisolone (dosage in mg/day given in parentheses); CSA = cyclosporin A; RP = Raynaud's phenomenon; RA = rheumatoid arthritis; NA = not available; AZA = azathioprine; SSc = systemic sclerosis; SS = Sjögren's syndrome; MCTD = mixed connective tissue disease; DM = dermatomyositis; SR = skin rash; IVIG = intravenous immunoglobulin; SLE = systemic lupus erythematosus; PA = polyarthritis; JA = juvenile arthritis.

  • Positivity for autoantibodies to Jo-1 (histidyl–transfer RNA synthetase), histones, Ro 42, signal recognition particle (SRP), Ku, SmB, Ro 60, La, and the U1 RNPs was determined by grading for the densitometry levels, as follows: 0–5 = negative (Neg.), 6–10 = (+), 11–25 = +, 26–50 = ++, and >50 = +++.

1/43/FDefinite PM0>76.865Jo-1 (+++)No therapyILD/–
2/23/FDefinite PM0>76.89Jo-1 (+++)No therapyILD/–
3/46/FDefinite PM51.08Jo-1 (+++)Pred. (7.5), CSAILD, arthritis, RP/RA
4/79/FProbable PM4.50.7NAJo-1 (+++)Pred. (5), AZAILD, dysphagea
5/70/MDefinite PM220.925Jo-1 (++)Pred. (7.5)ILD
6/61/MProbable PM<11.011Jo-1 (++)Pred.(40), AZAILD
7/49/MProbable PM0>76.85Jo-1 (+++), histonesNo therapyILD/–
8/34/FProbable PM<181.423Jo-1 (++), Ro 52 (+++)No therapyILD, RP/SSc, SS
9/64/FDefinite PM05321Jo-1 (+++), Ro 52 (+++)No therapyILD/–
10/64/FDefinite PM011.1100Jo-1 (+++), Ro 52 (+++)No therapyILD, RP/MCTD
11/69/FDefinite PM06.78Neg.No therapy–/–
12/54/MDefinite DM<10.44Neg.Pred. (40)–/–
13/69/FProbable DM3.81.26Neg.Pred. (7.5)SR
14/33/FDefinite PM<13.56SRP (++)No therapyRP/–
15/41/FProbable PM09.2<28SRP (+)No therapyRP/–
16/48/MDefinite PM040.110SRP (+++)Pred. (15)–/–
17/66/FDefinite PM017105KuNo therapyRP/–
18/53/FDefinite DM150.816KuNo therapySR
19/74/FProbable DM3.892.07KuPred. (10), IVIGSR
20/44/FDefinite PM0717SmBNo therapyRP/–
21/55/MDefinite PM0>76.840Ro 52 (+++)No therapy–/–
22/51/MProbable PM03.190Ro 52 (+++), Ro 60, LaNo therapyRP/–
23/30/FDefinite DM05.443Ro 52 (++), Ro 60, LaNo therapyILD/–
24/30/FProbable DM0<0.3258Ro 52 (+), Ro 60, LaPred. (10)–/RA, SLE
25/21/FProbable PM09.272RNP 70K (+++), RNP-ANo therapyRP/MCTD
26/50/FDefinite DM013.514RNP 70K (++), RNP-A, SmBNo therapyRP/MCTD, SS
27/47/FDefinite PM09.8104RNP 70K (+++), RNP-A, SmB, Ro 52 (+)No therapyRP/MCTD, PA, SS
28/20FProbable PM04.85RNP 70K (+++), RNP-A, RNP-C, SmBNo therapyRP/MCTD
29/56/FDefinite PM022.460RNP 70K (+++), RNP-A, RNP-C, SmBNo therapy–/MCTD, JA

Serum levels of circulating immune complex (IC) were measured by C1q binding assay (n = 22 patients) (Table 2), which included a control group consisting of 20 healthy individuals (median age 42 years [range 24–53 years], 8 women and 12 men). All patients and control subjects gave their informed consent, and the local ethics committee at the Karolinska Hospital, Stockholm, approved the study.

Table 2. Antibodies, mitogens, and products used in this study*
Antibody, mitogen, or productClone/catalog no.IsotypeDilution or concentrationSupplier
  • *

    MWCO = molecular weight cutoff; ICAM-1 = intercellular adhesion molecule 1; IF = immunofluorescence; ELISA = enzyme-linked immunosorbent assay; TNFα = tumor necrosis factor α; vWF = von Willebrand factor; HRP = horseradish peroxidase; PBS = phosphate buffered saline; TMB = tetramethylbenzidine; sICAM-1 = soluble ICAM-1; ICs = immune complexes; HMVEC-L = human microvascular endothelial cells derived from lung tissue; EGM = endothelial growth medium; FBS = fetal bovine serum; hFGF-B = human fibroblast growth factor B; R3-IGF-1 = long R insulin-like growth factor 1; hEGF = human epidermal growth factor; VEGF = vascular endothelial growth factor; GA-1000 = gentamicin sulfate/amphotericin B; EBSS = Earle's balanced salt solution.

NAb Protein G Spin kit89949Binding capacity 2.2–3.0 mg IgGPierce, Rockford, IL
Slide-A-Lyzer dialysis cassettes66203MWCO 10 kdPierce, Rockford, IL
Nanosep centrifugal devicesOD010C34MWCO 10 kdPall Life Sciences, Ann Arbor, MI
Anti-CD54 (ICAM-1)84H10 (IF)/MCA1615 (cell-based ELISA)Mouse IgG11:25/4 μg/mlSerotec, Oxford, UK
TNFα (IF)/cell-based ELISA40-300-01 AB/210-TARecombinant10 ng/mlSigma Aldrich, Täby, Sweden/R&D Systems, Abingdon, UK
Anti-vWFF8/86Mouse IgG11:40Dako, Glostrup, Denmark
Anti-CD31EN4Mouse IgG11:100Sanbio, Uden, The Netherlands
Secondary antibody (IF)Biotin goat anti-mouse IgG11:500Caltag, Burlingame, CA
Secondary antibody (cell-based ELISA)00006983/P0260HRP-conjugated rabbit anti-mouse1:1,500Dako, Glostrup, Denmark
Irrelevant antibody (IF)X0931Mouse IgG15 μg/mlDakopatts, Copenhagen, Denmark
Avidin and biotinBlocking kitVector, Burlingame, CA
Oregon Green 488Green fluorescent fluorochrome (with avidin)2 μg/mlMolecular Probes, Eugene, OR
PBS–glycerinMounting medium9:1Merck, Darmstadt, Germany
TMB peroxidase substrate (cell-based ELISA)080371/50-76-00KPL, Gaithersburg, MD
ELISA kit for sICAM-1/sE-selectinR&D Systems, Europe, Abingdon, UK
High-sensitivity ELISA kit for TNFαR&D Systems, Europe, Abingdon, UK
Bindazyme C1q binding kit for circulating ICsThe Binding Site, Birmingham, UK
HMVEC-L6358-1 (IF)/6F3754 (cell-based ELISA); cryopreserved at passage 4/passage 3Clonetics, BioWhittaker, Verviers, Belgium/Lonza, Walkersville, MD
Complete EGM2-MV Bulletkit (CC-3202)Basal EGM supplemented with 5% FBS, 0.1% hFGF-B, 0.1% R3-IGF-1, 0.1% hEGF, 0.1% VEGF, 0.04% hydrocortisone, 0.1% ascorbic acid, and 0.1% GA-1000Clonetics, BioWhittaker, Verviers, Belgium
Factor-free EGM2-MV Bulletkit (CC-3202)Basal EGM supplemented with 0.1% human serum albumin, 0.1% ascorbic acid, and 0.1% GA-1000Clonetics, BioWhittaker, Verviers, Belgium
EBSS–saponin solutionEBSS supplemented with 0.1% saponinRiedle de Haen, Seelze, Germany

Autoantibody determination.

Patients' sera were analyzed for autoantibodies to SmB, SmD, U1-70K RNP, U1-A RNP, U1-C RNP, Ro 52/SSA, Ro 60/SSA, La/SSB, centromere B, topoisomerase I/Scl-70, Jo-1/histidyl–transfer RNA synthetase, ribosomal P antigen, and histones. The autoantibodies were detected using an Inno-Lia Update antinuclear antibody line blot assay (Innogenetics, Ghent, Belgium). In addition, autoantibodies to Jo-1, Ro 52, Mi-2, PM-Scl, Pl-7, Pl-12, Ku, and SRP were analyzed using a Euroline Myositis Associated Antigens line blot assay, with a newly developed line blot for anti-SRP. Results were quantified by densitometry (Euroimmune, Lübeck, Germany). The findings yielded by both methods were congruent in all but 1 serum sample, in which anti–Ro 52 was detected only in the Euroline myositis line blot assay. A sample that tested positive against an autoantigen by either method was considered positive.

IgG purification.

IgG fractions were affinity purified on Nab protein G spin columns, and then dialyzed and concentrated using Nanosep centrifugal devices (Table 2). The final protein concentrations of the IgG fractions were evaluated by nephelometry (Clinical Chemistry, Karolinska University Hospital Solna, Stockholm, Sweden).

EC culture.

Human microvascular ECs derived from lung tissue (HMVEC-L) were grown to 70–90% confluence, at 37°C in 5% CO2, with complete endothelial growth medium (EGM) (Table 2). At passage 6, the HMVEC-L were seeded onto glass coverslips placed in 24-well chambers (2,000 cells/well) for immunofluorescence testing, and in 96-well microtiter plates (4,638 and 6,933 cells/well) for cell-based ELISAs. After 2 hours of adherence, complete EGM was added to the coverslips. The cells were grown to confluence between 40 hours and 70 hours. Thereafter, a minimum of 30 minutes of incubation with growth factor–free EGM (Table 2) was performed. The cells were incubated (4 hours for immunofluorescence, and 24 hours in triplicate for cell-based ELISA), at 37°C in 5% CO2, with EGM alone (i.e., unstimulated cells), with tumor necrosis factor α (TNFα; 10 ng/ml) as a positive control, or with 9.1–10% heat-inactivated serum or purified IgG (50 μg/ml, for the cell-based ELISA only) from patients or control subjects. Induction of ICAM-1 was assessed in 2 independent experiments with all sera, with the exception of 6 serum samples (2 from controls and 4 from patients with other autoantibodies) that were excluded due to technical reasons.

For stimulation with purified IgG, initial experiments with the cell-based ELISA were performed (for 4 hours) to select a suitable concentration of IgG from among the 5 tested (200, 50, 12.5, 3.125, or 0.78 μg/ml) using the sera from 3 control subjects, 3 anti–Jo-1 autoantibody–positive patients, and 3 patients with other autoantibody specificities. No significant differences between the 5 different concentrations of IgG were obtained (results not shown). The purified IgG concentration of 50 μg/ml was therefore chosen for a further ELISA experiment (performed for 24 hours) using sera from all patients and controls.

The endotoxin content in the cell-free supernatants was found to be negative in all but 2 samples (1 control and 1 patient serum) as measured by the Limulus amebocyte lysate test (Clinical Microbiology, Karolinska University Hospital Solna). Neither sample induced ICAM-1 expression.

Immunofluorescence analysis of intracellular ICAM-1 expression.

Cells were fixed with 2% formaldehyde and then rinsed in Earle's balanced salt solution–saponin buffer (Table 2) in between the different incubations. Cells were blocked for 5 minutes with 2% fetal calf serum, pH 7.4, at 37°C, and thereafter incubated with 0.1% saponin complemented with avidin and biotin for 15 minutes at room temperature. Incubations for 30 minutes, first with primary antibody, subsequently with biotinylated secondary antibody, and finally with Oregon Green 488, were performed (Table 2). Nuclear staining was conducted with 4′,6-diamidino-2-phenylindole. Irrelevant isotype-matched control antibody and unstimulated cells were used as negative controls. In addition, as controls for the preserved phenotype of the HMVEC-L, antibodies to von Willebrand factor and CD31 were used.

Indirect immunofluorescence images were coded and evaluated (at 400× magnification) using a fluorescein isothiocyanate–filtered Nikon Eclipse E800 microscope (Amsterdam, Holland). The percentage of cells staining positive for ICAM-1 was estimated. Only slides with a minimum of 100 cells were used for statistical analysis purposes.

Cell-based ELISA analysis of cell surface ICAM-1 expression.

After washing with growth factor–free EGM, the cells were fixed with 0.025% glutaraldehyde solution in phosphate buffered saline (PBS)–Mg2+Ca2+ (100 μl/well) for 10 minutes at room temperature (RT). Between incubations, washing was performed with PBS–Mg2+Ca2+. Cells were blocked with assay buffer (0.5% bovine serum albumin in PBS–Mg2+Ca2+) at 100 μl/well for 15 minutes at RT. One hour of incubation at 37°C with monoclonal ICAM-1 and secondary antibodies (each 50 μl/well) was performed, followed by development using tetramethylbenzidine substrate (100 μl/well) for 5 minutes in the dark at RT, before the reaction was stopped with 1M H2SO4 (50 μl/well). Optical density (OD) values were measured at 450 nm/562 nm. Only OD values reflecting the induction of ICAM-1 (i.e., the mean OD value of the unstimulated wells subtracted from the mean OD value of the stimulated wells, in triplicate experiments) were used for statistical analyses.

Statistical analysis.

Data were analyzed using GraphPad Prism statistical software, version 4.0 (GraphPad, San Diego CA). The Kruskal-Wallis test (analysis of variance for multiple comparisons) with Dunn's post hoc test and Mann-Whitney U test were used to compare the patient groups. Spearman's rank correlation coefficient was used to test for correlations. Results are expressed as the median and range. P values less than or equal to 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Serum characteristics.

Clinical and laboratory data for all of the patients with myositis are presented in Table 1. All anti–Jo-1–positive patients had PM and ILD. All patients had significantly higher sICAM-1 serum levels than did control subjects (median 387 ng/ml, range 198–1,012 ng/ml versus 233 ng/ml, range 97–257 ng/ml; P = 0.0001). There was no difference in sICAM-1 serum levels between anti–Jo-1–positive patients and anti–Jo-1–negative patients (median 461 ng/ml, range 287–603 ng/ml [n = 5] versus 385 ng/ml, range 198–1,012 ng/ml [n = 16]). No significant differences in the CPK levels or the ESR were observed between anti-Jo-1 autoantibody–positive patients and anti-Jo-1 autoantibody–negative patients (Table 1). The serum levels of sICAM-1 in patients did not correlate with the CPK levels or the ESR. The supernatant sICAM-1 levels in all patients (median 31 ng/ml, range 11–54 ng/ml) correlated with the serum levels of sICAM-1 (r = 0.865, P < 0.0001), which can be explained by the 9.1–10% serum content.

Effects of sera on ICAM-1 expression in HMVEC-L as assessed by immunofluorescence.

There was no significant difference in the percentage of ICAM-1–positive ECs between all myositis patients and controls (median 7.9%, range 3.0–17.4% versus 6.0%, range 2.7–8.9%). When patients were subclassified according to autoantibody profiles, anti–Jo-1–positive sera induced significantly more ICAM-1–positive cells (median 11.6%, range 6.8–17.4% [n = 6]) than did sera from patients with other autoantibody specificities (median 6.9%, range 3.0–11.5%; P < 0.05 [n = 16]) or controls (median 6.0%, range 2.7–8.9%; P < 0.05 [n = 9]) (Figures 1A and B). No correlation was found between the percentage of ICAM-1–positive ECs and the levels of anti–Jo-1 autoantibodies.

thumbnail image

Figure 1. Expression of intercellular adhesion molecule 1 (ICAM-1) after stimulation of human endothelial cells (ECs) derived from lung tissue (HMVEC-L) with sera from healthy controls or from patients with polymyositis or dermatomyositis positive for anti–Jo-1 autoantibodies or other autoantibody profiles (as described in Table 1). A, Percentage of ICAM-1–positive cells among the total number of ECs. Circles are the mean results of 2 independent stimulation experiments with an individual serum, as assessed by immunofluorescence. B, A representative immunofluorescence staining of perinuclear ICAM-1 expression (bright yellow; indicated by arrows) in HMVEC-L at 4 hours after stimulation with anti–Jo-1–positive serum from an individual patient. C, Optical density (OD) values of ICAM-1 induction. Circles are the mean results of 2 independent stimulation experiments with an individual serum, as determined by cell-based enzyme-linked immunosorbent assay (ELISA). In 1 of the 2 experiments, only 4 serum samples from patients positive for other autoantibodies and 2 control serum samples were studied. D, A representative immunofluorescence staining of perinuclear (solid arrows) and cytoplasmic/cell membrane (broken arrows) expression of ICAM-1 (bright green) in HMVEC-L at 24 hours after stimulation with anti–Jo-1–positive serum from an individual patient. The blue color is nuclear counterstaining. Inset, Unstimulated cells (incubated with medium alone). In A and C, the horizontal bars represent the median for each group. ∗ = P < 0.05. (Original magnification × 400 in B and D.)

Download figure to PowerPoint

Effects of sera and purified IgG on ICAM-1 expression in HMVEC-L as assessed by cell-based ELISA.

There was no significant difference between ICAM-1 induction in ECs stimulated by patient sera and that stimulated by control sera (median OD [× 10−3] 97, range 1.0–239 versus 80, range 4.0–152). Anti–Jo-1–positive sera induced significantly more ICAM-1 (median OD [× 10−3] 144, range 1.0–239) than did sera positive for other autoantibody specificities (median 79, range 26–142; P = 0.036) or control sera (P = 0.05 [n = 10 per group]) (Figure 1C). The corresponding immunofluorescence staining of ICAM-1 expression at 24 hours is displayed in Figure 1D.

Furthermore, there was no significant difference in ICAM-1 induction between ECs stimulated by purified IgG from patient sera (n = 16) and that stimulated by purified IgG from control sera (n = 10) (median OD [× 10−3] 12, range −123 to 167 versus −8.1, range −114 to 197), and ICAM-1 was induced by purified IgG irrespective of autoantibody profiles in the patients (median OD [× 10−3] −1.4, range −123 to 167 in anti–Jo-1–positive sera [n = 10] versus 28, range −56 to 34 in sera positive for other autoantibodies [n = 6]) compared with healthy controls. No correlation was found between the OD values of ICAM-1 expression after stimulation with either sera or IgG and the levels of anti–Jo-1 autoantibodies. Moreover, there was no difference in ICAM-1 induction in ECs between sera from patients with PM and sera from patients with DM, as determined with either method.

Levels of C1q-binding ICs.

The levels of C1q-binding ICs did not differ between patients (n = 22) and controls (n = 20) (median 1.2 μg/ml, range 0.13–35.8 μg/ml versus 1.5 μg/ml, range 0.1–9.5 μg/ml). The levels of C1q-binding ICs in anti–Jo-1–positive patients (n = 6) did not differ from those in anti–Jo-1–negative patients (n = 16). Patients with anti-SSA or anti-SSB autoantibodies alone (n = 4) had significantly higher levels of C1q-binding ICs (median 10.0 μg/ml, range 1.6–35.8 μg/ml) than did controls (median 1.5 μg/ml, range 0.1–9.5 μg/ml; P = 0.03) or anti–Jo-1–positive patients (median 0.96 μg/ml, range 0.14–6.48 μg/ml; P = 0.02).

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Sera from PM patients with ILD and positivity for anti–Jo-1 autoantibodies induced more ICAM-1–positive HMVEC-L in comparison with sera from healthy controls or patients with other autoantibodies, as assessed by 2 methods, immunofluorescence and cell-based ELISA. This difference between groups could not be demonstrated using stimulations with purified IgG from sera.

The mechanisms responsible for the activated ECs in muscle tissue of PM and DM patients are not known. We utilized commercially available HMVEC-L, which is derived from lung tissue, for stimulation experiments, because lung involvement is a frequent finding in PM and DM (1). As a measure of EC activation, we assessed the expression of ICAM-1, which has previously been shown to be overexpressed in ECs from patients with myositis (2, 3). The increased ICAM-1–inducing capacity stimulated with anti–Jo-1–positive sera, which was expressed as a higher percentage of cells positive for ICAM-1 by immunoflurescence, was confirmed using the quantitative cell-based ELISA. A longer incubation time was applied to determine the cytoplasmic ICAM-1 expression by the cell-based ELISA as compared with that with the perinuclear staining detected by immunofluorescence.

The induction of ICAM-1 expression in ECs was not achieved by purified IgG, which rules out the anti–Jo-1 antibodies themselves as a stimulator of ECs. Instead, it is likely that serum factors associated with the presence of anti–Jo-1 autoantibodies can bind to ECs and cause an activation that is similar to that seen in vivo. The potential to activate ECs was not restricted to anti–Jo-1–positive sera, since increased serum levels of sICAM-1 were also confirmed in patients with other autoantibody specificities (6). Our experiments, however, are the first to show that myositis sera may induce ICAM-1 expression in ECs from a relevant target organ. Notably, all of the patients with anti–Jo-1 antibodies, whose HMVEC-L exhibited the strongest ICAM-inducing capacity in our study, had ILD, which is known to be strongly associated with these autoantibodies.

A limitation of this study is that we were not able to determine which serum factors induced the endothelial expression of ICAM-1. The effect on EC activation was seen both in patients receiving immunosuppressive treatment and in those who had not received immunosuppressive treatment, which suggests that there is no role for general inflammation–associated molecules; this was also suggested by the lack of relationship of EC activation with the ESR or serum CK levels. This activation is not likely to be explained by the levels of circulating TNFα, since these levels were not higher in anti–Jo-1–positive patients compared with the other patients (data not shown). In DM patients, complement-mediated microvessel damage has been suggested as a mechanism of EC activation (7), an explanation that is less possible in our experiments since we used complement-inactivated sera. Moreover, the anti–Jo-1–positive patients had normal levels of C1q-binding ICs, whereas elevated levels were associated with the presence of anti-SSA/anti-SSB autoantibodies, confirming the results from another study (8).

Instead, an interesting mechanism for cell activation, associated with the presence of anti–Jo-1 antibodies, has been suggested from the observation that sera from myositis patients with these particular antibodies have a distinct type I interferon (IFN)–inducing capacity (5). This capacity has also been seen in patients with cutaneous lupus erythematosus, in whom increased expression of type I IFN correlated with increased expression of ICAM-1 (9), indicating a role of IFNα in the expression of ICAM-1. Furthermore, increased type I IFN–inducible gene expression in peripheral blood mononuclear cells from PM and DM patients also correlated with disease activity (10). Finally, type I IFN has been indicated as a risk factor for endothelial progenitor cell depletion and endothelial dysfunction in systemic lupus erythematosus (11). Thus, one possible mechanism behind the EC activation in patients with myositis might be via the induction of IFNα.

Thus, taken together, the findings of this study indicate that the association between ILD and anti–Jo-1 autoantibodies in sera from patients with PM and the in vitro activation of ECs suggest a new way to investigate and understand the pathogenesis of myositis-associated ILD.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Barbasso Helmers had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Barbasso Helmers, Englund, Engström, Åhlin, Heimbürger, Lundberg.

Acquisition of data. Barbasso Helmers, Englund, Engström, Åhlin, Heimbürger, Rönnelid, Lundberg.

Analysis and interpretation of data. Barbasso Helmers, Englund, Engström, Åhlin, Fathi, Janciauskiene, Heimbürger, Rönnelid, Lundberg.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We thank Inger Wedin, Helena Storfors, and Camilla Orbjörn for technical assistance, Heidi Wähämaa and Prof. Ulf Andersson for providing the TNFα, and Prof. Lars Klareskog for critical reading of the manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES