Detection of psoriasin/S100A7 in the sera of patients with psoriasis
K.S.Anderson*†, J.Wong*, K.Polyak†, D.Aronzon* and C.Enerbäck‡
*Cancer Vaccine Center and †Department of Medical Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02115, U.S.A.
‡Department of Clinical Genetics and Department of Dermatology, Sahlgrenska University Hospital, S-41345 Göteborg, Sweden
Correspondence to Karen S. Anderson.
E-mail:
Conflicts of interest
None declared.
Copyright Journal Compilation © 2009 British Association of Dermatologists
KEYWORDS
autoantibodies • biomarker • psoriasis • S100A7
ABSTRACT
/ / / / /BackgroundPsoriasis is a disease of dysregulated inflammation and epithelial hyperproliferation in the skin, involving both the innate and adaptive immune system. Psoriatic keratinocytes express high levels of psoriasin (S100A7), a small calcium-binding protein.
ObjectivesTo determine if patients with active psoriasis have elevated serum levels of psoriasin and psoriasin-specific autoantibodies.
MethodsBlood was collected from 14 patients with psoriasis vulgaris at the start of narrowband ultraviolet (UV) B therapy and from 11 of these patients every 2weeks during the course of the UVB treatment. Patient and control sera were tested for psoriasin antigen levels by sandwich enzyme-linked immunosorbent assay, and for psoriasin autoantibody titres using recombinant purified psoriasin and overlapping peptides.
ResultsWe confirmed strong and specific expression of psoriasin in psoriatic epidermis by immunohistochemistry. Systemic psoriasin antigen levels tended to be lower in patients (mean 213ngmL−1) than in controls (mean 331ngmL−1, P=0·308) and decreased with increasing disease severity. Psoriasin-specific autoantibodies were detected in a subset of patients with psoriasis and healthy normal donors (mean 0·347 vs. 0·255units, P=0·246). The epitopes recognized by the autoantibodies were mapped to an external loop domain of the molecule but did not show corresponding T-cell immunogenicity.
ConclusionsAlthough psoriasin is overexpressed in psoriatic skin lesions, systemic levels of psoriasin tended to be lower with increasing disease severity, which may be due to the presence of psoriasin-specific autoantibodies. Neither psoriasin nor psoriasin-specific autoantibodies appear to be promising serum biomarkers for clinical psoriasis.
Accepted for publication 11 August 2008
DIGITAL OBJECT IDENTIFIER (DOI)
10.1111/j.1365-2133.2008.08904.xAbout DOI
Article Text
Psoriasis is considered to be a genetically determined disease of dysregulated epidermal hyperproliferation and inflammation, which is initiated and maintained by multiple mechanisms of the immune system. Recent findings suggest a pathological collaboration between innate immunity (mediated by antigen-presenting cells and natural killer T lymphocytes) and acquired immunity (mediated by T lymphocytes) that results in the inflammatory infiltrate seen in psoriatic plaques.1 Hyperproliferation of the keratinocytes may be induced by the chemokines and cytokines secreted by inflammatory cells.
Calcium-binding S100 proteins are involved in many biological processes.2 Several S100 proteins, specifically S100A7 (psoriasin), S100A8 (calgranulin A) and S100A9 (calgranulin B) have been found to be markedly upregulated in psoriatic skin lesions.3,4 The expression of all three S100 proteins was induced by the proinflammatory cytokine interleukin (IL)-22 in psoriatic skin, and the serum level of IL-22 correlated with psoriasis disease severity.5 We and others have previously demonstrated a strong correlation between S100A7 and S100A9 expression and keratinocyte differentiation in normal and psoriatic skin.4,6 Psoriasin expression is also regulated by interferon (IFN)-γ,7 it has been proposed to be a chemoattractant of CD4+ lymphocytes in the skin,8 and it upregulates vascular endothelial growth factor.9 These findings suggest a potential role for psoriasin in the dysregulated differentiation, increased angiogenesis and intense inflammatory reaction characteristic for psoriatic lesions.
Elevated antigen levels of S100A8/S100A9 molecules have been detected in the sera of patients with psoriasis and were found to correlate with disease activity.10 As S100A7, S100A8 and S100A9 are coexpressed in psoriatic skin keratinocytes and are implicated in the pathogenesis of psoriasis,11,12 we hypothesized that psoriasin/S100A7 may also be detected in the sera of patients with psoriasis and may be a potential biomarker of disease severity. However, we found that the level of serum psoriasin was decreased in patients with psoriasis and decreased with increasing disease severity. Because psoriasin is partially secreted, we hypothesized that psoriasin may induce psoriasin-specific autoantibodies that could be used as surrogate biomarkers in patients with psoriasis. Secreted antigens, including the extracellular domain of Her2 in breast cancer, can induce autoantibodies that are highly specific and sensitive biomarkers for breast cancer.13 To test this hypothesis, we developed a highly sensitive enzyme-linked immunosorbent assay (ELISA) for the detection of IgG antipsoriasin autoantibodies, confirmed the specificity of these antibodies by Western blotting, and mapped the interactive epitopes using overlapping peptides. A subset of patients with psoriasis and healthy donors had detectable autoantibodies to psoriasin, but this was independent of disease severity and serum psoriasin levels. Although the number of patients analysed was fairly small, based on our results neither psoriasin antigen nor antipsoriasin autoantibodies appear to be promising serum biomarkers for psoriasis.
Material and methods
/ / / / /Patients and serum samples
Patients with psoriasis vulgaris treated at the psoriasis dermatological clinics at the Department of Dermatology, Sahlgrenska University Hospital, Göteborg, Sweden, who had not received systemic (e.g. methotrexate or ciclosporin) or topical (e.g. local glucocorticoids) treatment for at least 4weeks prior to sample collection, were selected for the study. The study was performed in the autumn and the patients had not received strong summer sun exposure just before the study. Patients with other types of psoriasis (guttate psoriasis and psoriatic arthritis) were not included. Overall 14 patients were analysed, four women and 10 men. The mean age of the patients was 47years (range 21–64). Patients received narrowband ultraviolet (UV) B treatment 3days a week for 8–10weeks. Patients were treated by whole-body exposure using Philips TL-01 lamps. Increments were according to standard procedures. Peripheral blood was collected at the start of and every 2weeks during the course of UVB treatment. Serum was immediately frozen on liquid nitrogen and stored at −80°C. Peripheral blood mononuclear cells (PBMC) were isolated using ACCUSPIN™ (Sigma-Aldrich, St Louis, MO, U.S.A.) and cellular viability was confirmed by trypan blue exclusion assay. The study was approved by the ethics committee of Göteborg University and all patients gave written informed consent. Control serum was obtained with informed consent from healthy blood donors at the Dana-Farber Cancer Institute.Immunohistochemistry
Tissue samples for immunohistochemical studies were selected from the files of the Departments of Dermatology and Pathology at Sahlgrenska University Hospital. Informed consent was requested from the patients under approval by the ethics committee of the University of Göteborg. Tissue samples were de-identified prior to the study. Immunohistochemical analysis was performed essentially as described.14S100A7 protein purification and peptides
cDNA encoding for full-length human S100A714 was subcloned as an N-terminal His-tagged fusion molecule in pQE30 (Qiagen, Valencia, CA, U.S.A.) and expressed in bacteria. Protein was purified under denaturing conditions in 6molL−1 guanidine-HCl with 100mmolL−1 NaCl and 10mmolL−1β-mercaptoethanol. Protein was purified using a nickel NTA column (Qiagen) using a slow urea step gradient with 0·5 molL−1 NaCl and 20% glycerol. Protein purity was confirmed by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE), and concentration determined by bicinchoninic assay (Pierce Biotechnology, Rockford, IL, U.S.A.). Overlapping peptides (15-mer, overlapping by 10) spanning the full-length molecule were obtained from Invitrogen (Carlsbad, CA, U.S.A.). S100A7–glutathione-S-transferase (GST) fusion proteins and GST control protein were expressed from pGEX4T-1 (Amersham, Piscataway, NJ, U.S.A.) and purified as described.15Sandwich enzyme-linked immunosorbent assay for the detection of S100A7 in serum
Monoclonal antibodies 817, 1068 and 396 were generated by standard immunization with recombinant psoriasin.14 A mixture of two capture antibodies, clones 817 and 1068, each at 2μgmL−1 in carbonate binding buffer pH 9·6, were coated overnight at 4°C on to 96-well Nunc-Immuno plates (Nalge Nunc, Rochester, NY, U.S.A.). The wells were blocked with 1% bovine serum albumin, then 50μL of human serum was added in duplicate to each well at room temperature for 1h. The detection antibody 396 was biotinylated using an EZ-Link biotinylation kit (Pierce Biotechnology), used at 2μgmL−1, and detected with streptavidin–horseradish peroxidase (HRP) and SuperSignal ELISA Substrate (Pierce Biotechnology) using a Victor3 plate reader (PerkinElmer, Waltham, MA, U.S.A.).Enzyme-linked immunosorbent assay and Western blot for the detection of S100A7-specific autoantibodies
His-tagged psoriasin antigen and peptides were diluted to 5μgmL−1 in binding buffer and coated overnight at 4°C on to 96-well Nunc-Immuno plates (Nalge Nunc) as described.15 Plates were blocked with 2% nonfat milk and 50μL of sera (1:300 dilution) was added in duplicates followed by incubation with secondary antibody (goat antihuman IgG-HRP, 1:5000; Jackson ImmunoResearch, West Grove, PA, U.S.A.) and detection with TMB+ Substrate (Dako, Carpinteria, CA, U.S.A.). For Western blotting, 10μg of S100A7-GST or GST antigen was separated by SDS-PAGE, transferred to nitrocellulose membrane, immunoblotted with patient sera at 1:500 and goat antihuman IgG-HRP (Jackson ImmunoResearch) and detected with ECL reagent per manufacturer's instructions (NEN, Boston, MA, U.S.A.).In vitro stimulation of peripheral blood mononuclear cells and interferon-γ enzyme-linked immunospot assay
PBMC from patients with psoriasis were stimulated in cell culture for 1week in the presence of 10ngmL−1 IL-7 (Endogen Inc., Woburn, MA, U.S.A.) from day 0 and 20IUmL−1 IL-2 (Chiron Corp., Emeryville, CA, U.S.A.) added on days 1 and 4. Peptides were added on day 0 at 10μgmL−1. On day 7, 50000 PBMC per well were coincubated with peptide pools corresponding to final concentrations of 4μgmL−1 of each individual peptide in triplicate in enzyme-linked immunospot (ELISPOT) plates (Millipore, Bedford, MA, U.S.A.) for 24h. IFN-γ secretion was detected following the manufacturer's instructions (Mabtech AB, Nacka Strand, Sweden) and imaged using an ImmunoSpot Series Analyzer (Cellular Technology Ltd, Cleveland, OH, U.S.A.).Statistical analysis
Psoriasis Area and Severity Index (PASI) score was confirmed with a self-reported patient opinion index of disease severity (a scale of 0–10 with 10 at start of treatment). Patients were considered responders if their PASI score improved by at least 75% compared with baseline or their self-opinion score improved to below 4. Three patients who withdrew before at least three samples were collected were not included in the analysis. Kappa statistic was used to analyse associations among the outcome of responder status using the PASI and patient opinion at the last available time point. The statistical significance of the differences between patients and normals in ELISA was analysed using two-sided Student's t-test.Results
/ / / / /Clinical response to ultraviolet B treatment
In order to determine if psoriasin may be used as a serological biomarker of psoriasis, 14 patients with active psoriasis were enrolled in a clinical trial through the dermatology clinic in Göteborg, Sweden. Individual PASI scores were recorded over time (Fig.1). Baseline PASI scores ranged from 2·0 to 25·3 (mean 8·5). Patients received standard UVB therapy and serum samples were collected at weeks 0, 2, 4, 6 and 10weeks follow-up. Patients were deemed eligible for the analysis if assessed at 4weeks or later after therapy was initiated (n=11) and the last time point was used for evaluation. Nine of 11 eligible patients (82%) had at least a 75% improvement in PASI score and all patients had a decrease in PASI score with UVB therapy (P0·003). This is consistent with published data on the rate of response to UVB therapy.16 Two patients (patients 8 and 14) had a transient increase in PASI score at week 6 after initiation of treatment, but they subsequently improved. PASI score was confirmed with a self-reported patient opinion index of disease severity shown in Table1. Ten of 11 (91%) patients had a severity index of improvement <5. Kappa statistic of 0·62 (n=11) showed significant agreement between the two scoring systems./ Fig1.Decrease in Psoriasis Area and Severity Index (PASI) score with ultraviolet (UV) B treatment of psoriasis. Fourteen patients with moderate to severe psoriasis vulgaris at baseline underwent UVB treatment and evaluation every 2weeks until study completion. Individual PASI scores at 2-week intervals are shown, and decreased in all patients with treatment.
[Normal View ]
Table1Decrease in clinical disease severity with time (A–E) on ultraviolet (UV) B treatment
Patient / PASI score / Patient opinion
A / B / C / D / E / A / B / C / D / E
1 / 8·4 / 8·4 / 2 / 0·5 / 1·5 / 10 / 10 / 4 / 2 / 4
2 / 8·3 / 5·4 / 2·7 / 1·6 / 0·9 / 10 / 7 / 5 / 3 / 2
3 / 3·9 / 1·8 / 1·3 / 0·8 / 10 / 6 / 4 / 2
4 / 25·3 / 15·7 / 12·9 / 6·4 / 10 / 6 / 5 / 3
5 / 3·3 / 3·3 / 0·7 / 0 / 10 / 10 / 5 / 0
6 / 2 / 10
7 / 15·6 / 10
8 / 9 / 7 / 5 / 6·3 / 2·2 / 10 / 7 / 6 / 8 / 3
9 / 4·2 / 3·4 / 1·9 / 0·4 / 0 / 10 / 9 / 7 / 3 / 0
10 / 5·4 / 4·5 / 2·7 / 1·35 / 0 / 10 / 8 / 5·5 / 4·25 / 3
11 / 4·2 / 2·5 / 10 / 6·5
12 / 8·2 / 5·7 / 0·4 / 0·2 / 0 / 10 / 6 / 0 / 0 / 0
13 / 3·6 / 2·8 / 1·6 / 0·4 / 0 / 10 / 8·5 / 6 / 3 / 0
14 / 17·2 / 14·4 / 11·2 / 12·8 / 4·8 / 10 / 8 / 7 / 8 / 5
Fourteen patients with moderate to severe psoriasis vulgaris at baseline (A) underwent UVB treatment and evaluation every 2weeks (B, C and D) until study completion (E). Clinical improvement after UVB treatment is shown by Psoriasis Area and Severity Index (PASI) score and self-reported patient opinion of severity (range 1–10, with 10 at baseline).
Psoriasin is specifically overexpressed in psoriatic skin lesions
Elevated antigen levels of S100A8/S100A9 molecules have been detected in the sera of patients with psoriasis, and they correlated with disease activity.10 As psoriasin/S100A7 expression is also upregulated in psoriatic skin6 and it is a partially secreted protein,14 we hypothesized that psoriasin would also be elevated in the serum. To develop a highly sensitive sandwich ELISA for psoriasin, three distinct monoclonal antibodies were raised in mice by S100A7 injection.14 Immunohistochemical analysis of psoriatic skin lesions using the monoclonal antibody clone 1068 demonstrated high psoriasin expression in psoriatic keratinocytes and in the epithelial cells of normal hair follicles (Fig.2). Psoriasin expression was often stronger in keratinocytes in the superficial parts of the hyperplastic epithelium and absent in the basal layers. In the keratinocytes, strong staining was often observed in the peripheral parts of the individual cells and faint staining could sometimes be observed in apparent intercellular bridges (desmosomes). Low expression of psoriasin was observed in keratinocytes of normal skin. In some cases, the epithelium was completely negative, but it often showed faint staining in superficial parts, which represent the more differentiated cells (data not shown)./ Fig2.S100A7 (psoriasin) is strongly overexpressed in psoriatic skin. Immunostaining using monoclonal antibody 1068 directed against psoriasin at low (a) and high (b) power. There is preferential staining in differentiated spinous layer cells of the central and superficial parts of the hyperplastic and elongated rete ridges.
[Normal View ]
Serum levels of S100A7 antigen are decreased in the sera of patients with psoriasis
We developed a sandwich ELISA for the detection of S100A7 protein in the sera of patients with psoriasis. Detection limits were calibrated using recombinant purified His-tagged S100A7 and S100A7-GST (Fig.3 and data not shown). Both His-S100A7 and S100A7-GST had comparable limits of detection. GST alone was not detected and the data depicted in Figure3 are representative of more than five replicate assays. The ELISA had a detection limit of 20ngmL−1, which is significantly below the 365ngmL−1 reported mean serum levels of S100A8/S100A9 in normal individuals.10 The range of detection was 20–2000ngmL−1 and titration curves of recombinant S100A7 protein were included with every assay. ELISA analysis of serum protein concentration of psoriasin was performed in duplicate on all 14 patients with psoriasis at baseline and run concurrently with sera from 14 healthy normal volunteers twice. In contrast to S100A8/S100A9, the mean serum levels of S100A7 antigen were decreased in patients with psoriasis (Fig.4a, mean 213ngmL−1) compared with normals (mean 331ngmL−1) but this was not statistically significant (P=0·308). S100A7 serum protein levels were comparable with the reported mean serum levels of S100A8/S100A9 in normals (365ngmL−1).10/ Fig3.Development of an enzyme-linked immunosorbent assay (ELISA) for the detection of psoriasin antigen. Recombinant His-tagged psoriasin protein was purified and titrated in a sandwich ELISA to determine the level of sensitivity of detection of psoriasin protein. The assay consisted of two capture monoclonal antibodies (mAbs) (including mAb 1068 shown in Figure2) and an independent biotinylated mAb detector antibody. The assay shown here was representative of over five different quantitative assays. Both His-tagged and glutathione-S-transferase (GST)-tagged psoriasin protein showed similar limits of detection to 20ngmL−1 psoriasin, and GST protein alone was not detected (not shown).
[Normal View ]
/ Fig4.Serum levels of psoriasin/S100A7 are decreased in patients with psoriasis. (a) Serum levels of psoriasin were assayed using the psoriasin enzyme-linked immunosorbent assay in sera from all 14 patients with psoriasis at baseline, and in sera from 14 healthy volunteers. Levels of psoriasin detected in the sera were lower in patients (mean 213ngmL−1) vs. healthy donors (mean 331ngmL−1) but this difference was not significant (P=0·308). (b) Psoriasin levels do not increase with increased disease severity as measured by Psoriasis Area and Severity Index (PASI) score. The patients with psoriasis shown in (a) were separated into two groups, those with lower PASI score (n=7, mean psoriasin level 239ngmL−1) and those with higher PASI score (n=7, mean psoriasin level 188ngmL−1): differences in psoriasin level between the groups were not significant (P=0·25).
[Normal View ]
The patients assayed in Figure4(a) were divided into two groups by disease severity as measured by PASI score. The mean levels of psoriasin detected in the sera of patients with low PASI score (mean 239ngmL−1) decreased with increased severity of the disease (188ngmL−1, P=0·504; Fig.4b). This result is in contrast to the correlation of S100A8/S100A9 level with disease severity, where mean levels of 1315 ngmL−1 were observed in patients with high PASI score.10 UVB therapy did not significantly affect S100A7 antigen levels in the 11 patients included in the study (data not shown, P=0·56).
Autoantibodies to S100A7 are not elevated in the serum of patients with psoriasis
We hypothesized that the lack of increased psoriasin protein levels in the serum of patients with psoriasis could be due to autoantibodies to S100A7 that may result in the clearance of antigen from the serum and could themselves be potential biomarkers of disease. To detect these autoantibodies we developed a sandwich ELISA using purified recombinant His-tagged S100A7 antigen refolded in urea following denaturation with guanidium chloride. Sera from the 14 patients with active psoriasis and from 14 healthy controls were assayed in duplicate, repeated twice (Fig.5a). We determined that patients with active psoriasis had slightly increased autoantibodies to S100A7 compared with healthy controls; however, these differences were not statistically significant (mean 0·347 vs. 0·255units, P=0·246). Antipsoriasin autoantibody levels also did not correlate with disease severity and serum S100A7 protein levels (data not shown). The half-life of IgG in serum is 23days;17 thus, IgG autoantibody levels might decrease following 4–6weeks of therapy. Correlating with this, autoantibodies to p53 antigen, a biomarker of breast cancer, decrease rapidly and specifically after induction chemotherapy.15 In the 11 patients analysed S100A7 autoantibody levels did not significantly change after UVB treatment (P=0·83). The specificity of the autoantibody identified by ELISA for psoriasin was confirmed by Western blotting sera from a patient with breast cancer (a disease also associated with S100A7 overexpression14) that had elevated S100A7 autoantibodies detected by the ELISA (Fig.5b)./ Fig5.Detection of S100A7-specific antibodies in sera. (a) Detection of S100A7-specific antibodies by enzyme-linked immunosorbent assay (ELISA). Recombinant His-tagged S100A7 protein was adhered to 96-well plates, and detected with sera from healthy donors (left) and patients with psoriasis (right). No detectable difference in antibody levels was observed (mean 0·347 vs. 0·255units, P=0·246). O.D., optical density. (b) S100A7 autoantibodies detected by ELISA are specific to S100A7 protein. Recombinant glutathione-S-transferase (GST) and S100A7-GST fusion proteins were purified and separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis and detected by Coomassie staining (left). The antigens were immunoblotted using 1:500 dilution of sera from a patient with breast cancer that had elevated levels of S100A7 autoantibodies by the ELISA shown in (a). The S100A7-GST fusion protein is specifically detected (arrow), while the control GST protein is not (right).
[Normal View ]
Epitope mapping of anti-S100A7 antibodies