- Email: [email protected]
Keratin 8 is a novel autoantigen of rheumatoid arthritis
Biochemical and Biophysical Research Communications 465 (2015) 665e669
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Keratin 8 is a novel autoantigen of rheumatoid arthritis Xiaoxu Wang 1, Peng Chen 1, Jiawen Cui, Chunhe Yang, Hongwu Du* 112 Lab, School of Chemistry and Biological Engineering, University of Science and Technology Beijing, Beijing 100083, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 July 2015 Accepted 31 July 2015 Available online 12 August 2015
Objective: This research aims to verify Keratin 8 (K8) as a specific autoantigen in rheumatoid arthritis (RA). Methods: First, total RNA was extracted from HaCaT cell to obtain cDNA by inverse transcription. Then, PCR was performed to amplify corresponding gene by K8 primers. Next, cloning, expression, and purification technology were used to obtain the recombinant human K8 (rhK8). At last, the purified rhK8, after identified by mass spectrometer, was used to perform further disease-related Western blotting and ELISA test with real clinical samples. Results: Purified rhK8 was successfully obtained and then Western blotting confirmed antigenicity of K8 in rheumatoid arthritis. The reactivity of serum IgG against rhK8 was further detected in 34 of 50 RA patients (68%). The reactivity of RA serum IgG antibodies against K8 was significantly higher than healthy controls and systemic lupus erythematosus (SLE) patients. Conclusion: This research confirmed Keratin 8 as a novel autoantigen of RA. © 2015 Elsevier Inc. All rights reserved.
Keywords: Rheumatoid arthritis Autoantigen Keratin 8
1. Introduction Rheumatoid arthritis (RA) is a chronic autoimmune disease without a certain etiology. It typically manifests synovitis, multijoint involvement and may lead to multisystem diseases such as atherosclerosis and lung diseases [1]. Sometimes, RA patients showed definitive chronic liver diseases, primary biliary cirrhosis [2,3], inflammatory bowel disease [4] and steatocystoma multiplex [5]. Since all these diseases were remarkably related to keratins [6], we considered keratins also had some uncertain connections with RA. Actually, there were already some antigens identified in RA, including heterogeneous nuclear ribonucleoproteins [7], enzyme peptidylarginine deiminase 4 [8], citrullinated peptide/protein antibodies, such as cyclic citrullinated peptide, citrullinated filaggrin and citrullinated vimentin, etc [9]. Keratin is the largest family of intermediate filament. Keratin had caused widespread concentration because of its wide distribution [10], multiple functions [11] and disease association. However, only a few reports talked about the roles of keratin in autoimmune diseases [6]. In 2013, keratin, type II cuticular Hb4
* Corresponding author. No. 30 XueYuan Road, Haidian District, School of Chemistry and Biotechnology Engineering, University of Science and Technology Beijing, Beijing 100083, China. E-mail address: [email protected] (H. Du). 1 These authors Xiaoxu Wang and Peng Chen contributed equally to this work. http://dx.doi.org/10.1016/j.bbrc.2015.07.161 0006-291X/© 2015 Elsevier Inc. All rights reserved.
(KRT84/K84) was suggested as a specific antigen of RA with proteomic approaches [12]. Considering that K8 and K84 fall into similar sequence homology groups, type II [13] and K8 is also remarkably disease related [6], we assume that K8 may also be a autoantigen in RA. Actually, one of the useful auxiliary diagnoses was to detect the so called anti-keratin antibody (AKA) in clinical application, which was a useful marker for RA diagnosis as early as 1979 [14], but with the technology progress, the later so-called AKA's antigen was considered as filaggrin [15]. Afterwards, there were also antibodies, whose antigen might also be filaggrin, used in medical diagnosis called antibodies to human epidermis filaggrin (AFA) and anti-perinuclear factor (APF) [16,17]. However, AFA and APF's positivity of individuals was not identical to AKA's exactly, so there might be some special antigens exist in each antibody profile [18,19]. We also found protein sequences of K8 and filaggrin were not similar after comparison. We assumed that keratin 8 was another potential self-antigen of so called AKA. The total prevalence rate of RA in China is 0.2e0.4% [20,21]. Although it is considering lower than worldwide prevalence rate (0.5e1%) [22], with China's large population, the number of RA patients is still very large, so more basic researches related to RA are needed. In this study, gene recombination technology was used to construct an expression system of K8 in Escherichia coli. Mass spectrometry was used to confirm purified rhK8 protein. The connection between K8 and patients was reassured by Western blotting and ELISA.
666
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
2. Materials and methods 2.1. Sera collection In this research, 167 Han Chinese subjects were enrolled and used to assess the serological criteria. Experimental group: 60 confirmed RA patients with an average of 52 years old (Range from 22 to 77, 43 female and 17 male patients). Disease controls: 50 SLE patients with an average of 33 years old (Range from 16 to 62, 41 female and 9 male patients). Healthy controls: 57 healthy people with an average of 25 years old (Range from 21 to 32, 47 female and 10 male). All 60 diagnosed RA patients met the criteria of the American College of Rheumatology [23]. Sera were stored at 80 C immediately after aliquoted. All simples were provided by Chinese PLA General Hospital with ratification and all patients gave informed consents. 2.2. Cell line The HaCaT cell line was provided by Cell Lines Service (Eppelheim, Germany). The condition of cultivation was set, according to our former research [24]. DMEM (HyClone, UT), containing 10% fetal bovine serum (HyClone, UT) was used to cultivate HaCaT. Then the cell was used to extract total RNA for next experiment.
dried target gel pieces were then digested by trypsin (Sigma, MO), mixed with 20 mL of 0.05 M NH4HCO3, overnight at 37 C. The target protein band was identified, using a 4700 Proteomics Analyzer MALDI-TOF/TOF mass spectrometer (Applied Biosystems, Foster City, CA). Mass spectrometry data were analyzed with Mascot bioinformatics database search engine (Matrix Sciences, London, UK). 2.5. Western blotting We used methods which were described in detail previously [24,27] with a few modifications. The purified rhK8 was loaded into 10% polyacrylamide gel and separated. Then, the target proteins were transfer to polyvinylidene fluoride (PVDF; Merck Millipore, MA) membranes. The PVDF membranes were blocked with 5% nonfat milk in PBS at 4 C overnight. After blocking, we cut up PVDF membranes into strips and incubated them with sera of RA patients, disease controls or healthy controls (1:1000, diluted in 1% nonfat milk) overnight at 4 C. Then, the PVDF membranes were washed by 0.2% PBST for three times to remove uncombined antibodies and each wash lasted for 10 min. Later, we Incubated the PVDF membranes with horseradish-peroxidase conjugated goat anti-human IgG (1:10,000, diluted in 1% nonfat milk) (ImmunoHunt, Beijing, China) for 1 h at 37 C. The potential target bands were detected by enhanced chemiluminescence kit (Applygen, China) and exposed later.
2.3. Protein expression and purification 2.6. Dot-ELISA Protocols of gene recombination, protein expression and purification, certified by our previous study, were used [25]. In brief, total mRNA was extracted from aforementioned cell line by TRIzol reagent (Invitrogen, CA). The designed primer: Forward: (EcoR I) 50 CCGGAATTCGTGACCCAGAAGTCCTACAAG-30 , Reserve: (SaI I) 50 CGCGTCGACTTATTAATCAGAGGACTCAGACACCAG-30 , were used to amplify K8 gene by RT-PCR, carried out with the kit's instruction (Fermentas, MD). Then, the K8 gene was linked with pET-28(a) vectors, which were transformed into E. coli DH5a. Transformational E. coli DH5a strain was recovered in 400 mL Super Optimal Broth (SOC) cultivation for 1.5 h. Later, we transferred the bacterium culture solution to Luria Broth (LB) cultivation, containing kanamycin (50 mL/mL). Extracting recombinant plasmids from E. coli DH5a and transforming these plasmids into high expression engineering strain E. coli BL21, using the same method that mentioned in this section. Under the optimal condition: 37 C, 200 rpm shaking speed, overexpression of K8 was achieved by adding IPTG (1 mM) when optical density (OD) reaches 0.4e0.6. Finally, the recombinant K8 with N-terminal hexahistidine-tag was purified by Ni-NTA resin (Qiagen, Hilden, Germany), according to instruction supplied by manufacturer. After determining the concentration of protein by a BCA assay kit (Biosynthesis Biotechnology, Beijing, China), the purified protein was stored at 80 C for further experiments. 2.4. Mass spectrometry The method in our previous study was applied to this experiment [24,26]. Briefly, the target protein band was cut off from SDSPAGE gel. After destaining the target gel pieces with 25 mM NH4HCO3 and 50% acetonitrile, the mixture was dried by a vacuum centrifuge. Then, 25 mM NH4HCO3 and 10 mM dithiothreitol were added to cover the target gel pieces and reduced for 2 h at 37 C. The DTT solution was replaced by roughly same volume of 55 mM iodoacetamide in 25 mM NH4HCO3 and incubated for 45 min at room temperature in the dark. Next, 50% acetonitrile in 25 mM NH4HCO3 were used to wash the target gel pieces for 10 min. The liquid mixture was completely dried by vacuum concentrator. The
To perform this experiment, we referred to our previous research [28]. We spotted about 0.07 mg keratin 8 on the dull side of a 0.22-mm nitrocellulose membrane. After drying the membrane, 10% goat sera (diluted in PBS) was used to block the antigen discs for 1 h at 37 C. Then the sera of RA patients (1:500, diluted in PBS) or healthy controls (1:500, diluted in PBS) were added to the discs and incubated for 1 h at 37 C. PBST (phosphate buffered saline with 0.1% Tween-20, pH 7.4) was used to wash the discs for 3 times later. After removing the liquid phase and dry, the discs were incubated with goat anti-human IgG/HRP (1:1000, diluted in PBS) (ImmunoHunt, Beijing, China) for 1 h at 37 C. Next, the washing process was repeated as aforementioned and 3, 30 -diaminobenzidine solution (ZSGB-BIO, Beijing, China) was added and colorated for 5 min. At last, adding deionized water to stop the reaction. 2.7. ELISA The detection of potential autoantigen with ELISA was performed, according to previous studies [25,27]. Coating the 96-well microtiter plates with rhK8 (100 ng/mL, dissolved in 0.05 M carbonic buffer, pH ¼ 9.6) overnight at 4 C. After removing the liquid phase, each well was blocked with 100 mL goat sera (1:10, diluted in PBS) for 2 h at 37 C. Then the liquid phase was removed again and the sera of RA, SLE patients and healthy controls (1:100, diluted in PBS) were used to incubate the 96-well microtiter plates for 2 h at 37 C. Five-time wash was performed here to remove unbind autoantibodies. Next, 100 mL goat anti-human IgG/HRP (1:20,000, diluted in PBS) (ImmunoHunt, Beijing, China) was added to each well and incubated for 1 h at 37 C. Again, wash each well with 0.1% PBST for five times to elute the second antibodies. Orderly, adding 50 mL tetramethylbenzidine (TMB) A (0.1 M citric acid, 0.2 M NA2HPO4, 0.6 g hydroperite/L) and 50 mL TMB B (5 mM citric acid, 0.4 mM EDTA-NA2, 0.2 g TMB/L) to each well. After keeping the plates at a dark, room-temperatured place for 15 min, the reaction was stopped by adding 50 mL stop buffer (2 M H2SO4). Finally, OD value of each well was measured by ELISA reader (Tecan, Switzerland) at 450/620 nm.
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
667
2.8. Statistical analysis Receiver operating characteristic (ROC) analysis was carried out with MedCalc (Version 9.2.0.1, Mariakerke, Belgium). Fisher's exact test was performed to compare differences between cohorts. The scatterplot of ELISA result was evaluated by GraphPad Prism (Version 5.0, GraphPad Software Inc., La Jolla, CA). P values were considered to indicate statistical significance only when they were less than 0.05. 2.9. Antigenic determinant prediction As aforementioned, since K84 was a probable antigen of RA [12], K8 and K84 were selected to perform sequence alignment using BLAST to evaluate their homology [29]. Then Bepipred Linear Epitope Prediction [30] was used to predict epitopes of two proteins. Potential common epitopes of two proteins were selected with standards that amino acid lengths were no less than 8 and cooccurred in K8 and K84. 3. Results 3.1. Expression and purification of recombinant K8 protein Human K8 gene was amplified by Polymerase Chain Reaction technique. The production was verified by 0.8% Agarose Gel Electrophoresis. A target band was obtained at 1500 bp and no other specific amplification appeared (Fig. 1)A. After cultivating under the optimum condition mentioned before, a rough 53-KDa protein band appeared. And the band was the putative K8 protein when compared with recombinant E. coli BL21 (DE3) strain without induction of IPTG. Then his-tag contained target protein was obtained by Ni-NTA resin (Fig. 1B). 3.2. Confirmation of keratin 8 The protein obtained by Ni-NTA resin was reassured by mass spectrometry. The result shows that individual ions scores are more than 41, indicating the putative protein is highly homologous identity (P < 0.05). And protein scores are greater than 70, which is significant (P < 0.05) to rank a protein hit (Fig. 1C).
Fig. 2. Demonstration of K8 antigenicity in RA. (A) Western blotting result. Three positive results among 10 RA patients were named RA1, RA2 and RA3. A 53-kDa band was observed. Two negative results were selected randomly from 7 negative results, named RA4, RA5. All healthy controls showed negative results. (B) Dot-ELISA result. One positive resultdRA1dwas then used to perform Dot-ELISA to assure the antigenicity of K8. Compared with blank control (PBS; using PBS instead of patient sera) and healthy control (HC1), only RA1 showed positive result (the dotted line circle).
3.3. Western blotting and Dot-ELISA K8 associated Western blotting was performed with sera of 10 RA patients and 5 healthy people, who were selected randomly from our sera bank. Among these 10 RA patients, 3 patients showed positive antigeneantibody reaction and all healthy controls showed negative results, indicating K8 was a potential autoantigen in RA patients (Fig. 2A). And the Dot-ELISA result of RA1 was consistent with the result of Western blotting (Fig. 2B). 3.4. Significance of K8 in RA ELISA of K8 with 50 RA patients was carried out in order to confirm the antigen-specificity of K8 in RA patients. At same time, disease controlsd50 SLE patients' serad were involved in ELISA with the same operating method. 56 healthy people's sera were used as healthy controls. The cut-off value of positivity of K8 in RA disease was obtained by ROC curve (Fig. 3A). The cut-off value was 0.205, at which the sensitivity and specificity were 68% and 80.36%, respectively. All ELISA results were showed in scatterplot (Fig. 3B). The positive rates of K8 in HC, RA and SLE patients were 20% (11 of 56 healthy controls), 68% (34 of 50 RA patients. P < 0.001 compared with healthy controls) and 24% (12 of 50 SLE patients).
Fig. 1. Gene amplification, expression, purification and identification of Keratin 8. (A) PCR production of K8 gene. (B) Expression and purification of K8 protein. Symbol “M” and Bands 2e4 represented DNA Marker, purified K8 protein (His-tag contained) and protein expression of BL21 strains which contained the recombinant plasmid K8-pET-28a (þ) with or without 1 mM IPTG induction, respectively. (C) Mass spectrogram. The target protein was identified as K8. The molecular weight of target protein was 53.670-kDa, conformed to expected size: 53-KDa. Symbol “*” noted the matched peptides displayed in the top right corner.
668
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
Fig. 3. Significance of K8 in RA. (A) Receiver operating characteristic (ROC) curve was drawn to indicate prevalence of K8 antibodies in RA. (B) Scatterplot of all ELISA results demonstrated the difference between RA and HC. The marker “***” manifests P < 0.001.
Fig. 4. Antigenic epitopes prediction and comparison between K8 and K84. (A) Prediction result of K8. (B) Prediction result of K84. The three similar epitopes were emphasized with red underlines. Epitope sequences were displayed beyond the underlines. (C) Amino acid properties of three similar epitopes were further analyzed. For each site, amino acids, which had similar properties (in green) and those were identical in both keratins (in red) were showed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
4. Discussion Historically, the term Keratins widely aroused people's attention to RA in 1979, with the discovery of so-called AKA [14], which had been widely used as a clinical diagnosis marker of RA. Later, experiment demonstrated that one probable target protein of socalled AKA was filaggrin [15], though, no other detailed evidences could illustrate flilaggrin was the only antigen of AKAs. Usually, AKA was detected by indirect immunofluorescence and no confirmed ELISA test method had been established [18]. Since K8 was a keratin that must exist in epidermal keratinocytes [13], we assumed the antibody of K8 was a supplement of so called AKA and may be detected by ELISA in the future. Actually, some keratins had been demonstrated that played significant roles in human diseases. Among all keratins, K8 was considering highly disease-related [6]. It had been reported that many K8 related biological processes such as K8 genetic mutation and hypersumoylation were connected with liver injury and liver diseases [31,32]. K8 was also useful for tumor classificatory diagnosis [33,34] and could be used as a prognostic marker [35]. In 2013, an experiment had identified that K84 was a screening citrullinated protein in tissues of synovial in RA. Even though they found K84 was a putative antigen [12], except K84, other keratins were not confirmed related to RA so far.
In this study, we evaluated K8 as a potential antigen, which was demonstrated by ELISA and Western blotting. Since K8 and K84 belong to a same sub-family: type II family in keratin and they are all specific antigens in RA, we make a further investigation to predict the role K8 and K84 play in RA. In the result of Bepipred Linear Epitope Prediction, twenty-six and thirty-six epitopes were predicted for K8 and K84, respectively. According to the standards mentioned before, three significant homologous epitope pairs were selected (Fig. 4), indicating the similar epitopes engaged in two keratins and might cause antigen-activation in RA. In conclusion, we identified anti-K8 autoantibody in some of the patients with rheumatoid arthritis and K8 was probably one of the real antigens of traditional so-called AKA. Conflict of interest All co-authors have no competing interests. Acknowledgments This work was supported by Beijing Nova Program of Science and Technology (No. 2007B024), Program for New Century Excellent Talents in University (No. NECT-12-0773), and the National Natural Science Foundation of China (No. 31371203).
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
Transparency document Transparency document related to this article can be found online at http://dx.doi.org/10.1016/j.bbrc.2015.07.161.
[17]
[18]
References [1] S. Gertel, H. Amital, Putative approaches to bypass the citrulline-specific autoimmune response in rheumatoid arthritis, Isr. Med. Assoc. J. 16 (2014) 587e590. [2] P.R. Mills, R.N. MacSween, W.C. Dick, et al., Liver disease in rheumatoid arthritis, Scott. Med. J. 25 (1980) 18e22. [3] C. Caramella, J. Avouac, P. Sogni, et al., Association between rheumatoid arthritis and primary biliary cirrhosis, Jt. Bone Spine 74 (2007) 279e281. [4] O. Palm, B. Moum, J. Jahnsen, et al., The prevalence and incidence of peripheral arthritis in patients with inflammatory bowel disease, a prospective population-based study (the IBSEN study), Rheumatol. Oxf. 40 (2001) 1256e1261. [5] D. Sohn, T.C. Chin, M.J. Fellner, Multiple keratoacanthomas associated with steatocystoma multiplex and rheumatoid arthritis. A case report, Arch. Dermatol. 116 (1980) 913e915. [6] D.M. Toivola, P. Boor, C. Alam, et al., Keratins in health and disease, Curr. Opin. Cell Biol. 32 (2015) 73e81. [7] G. Steiner, K. Hartmuth, K. Skriner, et al., Purification and partial sequencing of the nuclear autoantigen RA33 shows that it is indistinguishable from the A2 protein of the heterogeneous nuclear ribonucleoprotein complex, J. Clin. Invest. 90 (1992) 1061e1066. [8] E.H. Halvorsen, S. Pollmann, I.M. Gilboe, et al., Serum IgG antibodies to peptidylarginine deiminase 4 in rheumatoid arthritis and associations with disease severity, Ann. Rheum. Dis. 67 (2008) 414e417. [9] K. Conrad, D. Roggenbuck, D. Reinhold, et al., Profiling of rheumatoid arthritis associated autoantibodies, Autoimmun. Rev. 9 (2010) 431e435. [10] I. Kurokawa, K. Takahashi, I. Moll, et al., Expression of keratins in cutaneous epithelial tumors and related disordersedistribution and clinical significance, Exp. Dermatol. 20 (2011) 217e228. [11] Y.H. Chou, O. Skalli, R.D. Goldman, Intermediate filaments and cytoplasmic networking: new connections and more functions, Curr. Opin. Cell Biol. 9 (1997) 49e53. [12] X. Chang, Y. Zhao, Y. Wang, et al., Screening citrullinated proteins in synovial tissues of rheumatoid arthritis using 2-dimensional western blotting, J. Rheumatol. 40 (2013) 219e227. [13] J. Schweizer, P.E. Bowden, P.A. Coulombe, et al., New consensus nomenclature for mammalian keratins, J. Cell Biol. 174 (2006) 169e174. [14] B.J. Young, R.K. Mallya, R.D. Leslie, et al., Anti-keratin antibodies in rheumatoid arthritis, Br. Med. J. 2 (1979) 97e99. [15] M. Simon, E. Girbal, M. Sebbag, et al., The cytokeratin filament-aggregating protein filaggrin is the target of the so-called “antikeratin antibodies”, autoantibodies specific for rheumatoid arthritis, J. Clin. Invest. 92 (1993) 1387e1393. [16] C. Vincent, M. Simon, M. Sebbag, et al., Immunoblotting detection of
[19] [20] [21]
[22] [23]
[24] [25]
[26] [27]
[28]
[29] [30] [31]
[32] [33] [34]
[35]
669
autoantibodies to human epidermis filaggrin: a new diagnostic test for rheumatoid arthritis, J. Rheumatol. 25 (1998) 838e846. M. Sebbag, M. Simon, C. Vincent, et al., The antiperinuclear factor and the socalled antikeratin antibodies are the same rheumatoid arthritis-specific autoantibodies, J. Clin. Invest. 95 (1995) 2672e2679. C. Vincent, F. de Keyser, C. Masson-Bessiere, et al., Anti-perinuclear factor compared with the so called “antikeratin” antibodies and antibodies to human epidermis filaggrin, in the diagnosis of arthritides, Ann. Rheum. Dis. 58 (1999) 42e48. S.L. Slack, M. Mannik, B.A. Dale, Diagnostic value of antibodies to filaggrin in rheumatoid arthritis, J. Rheumatol. 25 (1998) 847e851. Y.J. Xiang, S.M. Dai, Prevalence of rheumatic diseases and disability in China, Rheumatol. Int. 29 (2009) 481e490. S.M. Dai, X.H. Han, D.B. Zhao, et al., Prevalence of rheumatic symptoms, rheumatoid arthritis, ankylosing spondylitis, and gout in Shanghai, China: a COPCORD study, J. Rheumatol. 30 (2003) 2245e2251. S.E. Gabriel, The epidemiology of rheumatoid arthritis, Rheum. Dis. Clin. North Am. 27 (2001) 269e281. F.C. Arnett, S.M. Edworthy, D.A. Bloch, et al., The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis, Arthritis Rheum. 31 (1988) 315e324. P. Chen, H. Yan, Y. Tian, et al., Annexin A2 as a target endothelial cell membrane autoantigen in Behcet's disease, Sci. Rep. 5 (2015) 8162. H. Du, G. Chen, S. Wang, et al., Immunological screening and characterization of highly specific monoclonal antibodies against 20 kDa hGH, Bioanalysis 4 (2012) 2161e2168. Y. Xun, P. Chen, H. Yan, et al., Identification of prohibitin as an antigen in Behcet's disease, Biochem. Biophys. Res. Commun. 451 (2014) 389e393. H. Du, C. Li, H. Jin, et al., Generation and evaluation of antibodies against human MGF E-peptide by reverse phase protein microarray and reverse competitive ELISA, Bioanalysis 5 (2013) 2269e2275. H. Du, J. Liu, Y. Xun, et al., Determination of deoxynivalenol, zearalenone, aflatoxin B1, and ochratoxin by an enzyme-linked immunosorbent assay, Anal. Lett. 47 (2014) 1912e1920. S.F. Altschul, W. Gish, W. Miller, et al., Basic local alignment search tool, J. Mol. Biol. 215 (1990) 403e410. J.E. Larsen, O. Lund, M. Nielsen, Improved method for predicting linear B-cell epitopes, Immunome Res. 2 (2006) 2. N.T. Snider, S.V. Weerasinghe, J.A. Iniguez-Lluhi, et al., Keratin hypersumoylation alters filament dynamics and is a marker for human liver disease and keratin mutation, J. Biol. Chem. 286 (2011) 2273e2284. P. Strnad, S. Paschke, K.H. Jang, et al., Keratins: markers and modulators of liver disease, Curr. Opin. Gastroenterol. 28 (2012) 209e216. R. Moll, M. Divo, L. Langbein, The human keratins: biology and pathology, Histochem. Cell Biol. 129 (2008) 705e733. H. Tsubochi, T. Suzuki, S. Suzuki, et al., Immunohistochemical study of basaloid squamous cell carcinoma, adenoid cystic and mucoepidermoid carcinoma in the upper aerodigestive tract, Anticancer Res. 20 (2000) 1205e1211. T. Knosel, V. Emde, K. Schluns, et al., Cytokeratin profiles identify diagnostic signatures in colorectal cancer using multiplex analysis of tissue microarrays, Cell. Oncol. 28 (2006) 167e175.
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Keratin 8 is a novel autoantigen of rheumatoid arthritis Xiaoxu Wang 1, Peng Chen 1, Jiawen Cui, Chunhe Yang, Hongwu Du* 112 Lab, School of Chemistry and Biological Engineering, University of Science and Technology Beijing, Beijing 100083, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 July 2015 Accepted 31 July 2015 Available online 12 August 2015
Objective: This research aims to verify Keratin 8 (K8) as a specific autoantigen in rheumatoid arthritis (RA). Methods: First, total RNA was extracted from HaCaT cell to obtain cDNA by inverse transcription. Then, PCR was performed to amplify corresponding gene by K8 primers. Next, cloning, expression, and purification technology were used to obtain the recombinant human K8 (rhK8). At last, the purified rhK8, after identified by mass spectrometer, was used to perform further disease-related Western blotting and ELISA test with real clinical samples. Results: Purified rhK8 was successfully obtained and then Western blotting confirmed antigenicity of K8 in rheumatoid arthritis. The reactivity of serum IgG against rhK8 was further detected in 34 of 50 RA patients (68%). The reactivity of RA serum IgG antibodies against K8 was significantly higher than healthy controls and systemic lupus erythematosus (SLE) patients. Conclusion: This research confirmed Keratin 8 as a novel autoantigen of RA. © 2015 Elsevier Inc. All rights reserved.
Keywords: Rheumatoid arthritis Autoantigen Keratin 8
1. Introduction Rheumatoid arthritis (RA) is a chronic autoimmune disease without a certain etiology. It typically manifests synovitis, multijoint involvement and may lead to multisystem diseases such as atherosclerosis and lung diseases [1]. Sometimes, RA patients showed definitive chronic liver diseases, primary biliary cirrhosis [2,3], inflammatory bowel disease [4] and steatocystoma multiplex [5]. Since all these diseases were remarkably related to keratins [6], we considered keratins also had some uncertain connections with RA. Actually, there were already some antigens identified in RA, including heterogeneous nuclear ribonucleoproteins [7], enzyme peptidylarginine deiminase 4 [8], citrullinated peptide/protein antibodies, such as cyclic citrullinated peptide, citrullinated filaggrin and citrullinated vimentin, etc [9]. Keratin is the largest family of intermediate filament. Keratin had caused widespread concentration because of its wide distribution [10], multiple functions [11] and disease association. However, only a few reports talked about the roles of keratin in autoimmune diseases [6]. In 2013, keratin, type II cuticular Hb4
* Corresponding author. No. 30 XueYuan Road, Haidian District, School of Chemistry and Biotechnology Engineering, University of Science and Technology Beijing, Beijing 100083, China. E-mail address: [email protected] (H. Du). 1 These authors Xiaoxu Wang and Peng Chen contributed equally to this work. http://dx.doi.org/10.1016/j.bbrc.2015.07.161 0006-291X/© 2015 Elsevier Inc. All rights reserved.
(KRT84/K84) was suggested as a specific antigen of RA with proteomic approaches [12]. Considering that K8 and K84 fall into similar sequence homology groups, type II [13] and K8 is also remarkably disease related [6], we assume that K8 may also be a autoantigen in RA. Actually, one of the useful auxiliary diagnoses was to detect the so called anti-keratin antibody (AKA) in clinical application, which was a useful marker for RA diagnosis as early as 1979 [14], but with the technology progress, the later so-called AKA's antigen was considered as filaggrin [15]. Afterwards, there were also antibodies, whose antigen might also be filaggrin, used in medical diagnosis called antibodies to human epidermis filaggrin (AFA) and anti-perinuclear factor (APF) [16,17]. However, AFA and APF's positivity of individuals was not identical to AKA's exactly, so there might be some special antigens exist in each antibody profile [18,19]. We also found protein sequences of K8 and filaggrin were not similar after comparison. We assumed that keratin 8 was another potential self-antigen of so called AKA. The total prevalence rate of RA in China is 0.2e0.4% [20,21]. Although it is considering lower than worldwide prevalence rate (0.5e1%) [22], with China's large population, the number of RA patients is still very large, so more basic researches related to RA are needed. In this study, gene recombination technology was used to construct an expression system of K8 in Escherichia coli. Mass spectrometry was used to confirm purified rhK8 protein. The connection between K8 and patients was reassured by Western blotting and ELISA.
666
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
2. Materials and methods 2.1. Sera collection In this research, 167 Han Chinese subjects were enrolled and used to assess the serological criteria. Experimental group: 60 confirmed RA patients with an average of 52 years old (Range from 22 to 77, 43 female and 17 male patients). Disease controls: 50 SLE patients with an average of 33 years old (Range from 16 to 62, 41 female and 9 male patients). Healthy controls: 57 healthy people with an average of 25 years old (Range from 21 to 32, 47 female and 10 male). All 60 diagnosed RA patients met the criteria of the American College of Rheumatology [23]. Sera were stored at 80 C immediately after aliquoted. All simples were provided by Chinese PLA General Hospital with ratification and all patients gave informed consents. 2.2. Cell line The HaCaT cell line was provided by Cell Lines Service (Eppelheim, Germany). The condition of cultivation was set, according to our former research [24]. DMEM (HyClone, UT), containing 10% fetal bovine serum (HyClone, UT) was used to cultivate HaCaT. Then the cell was used to extract total RNA for next experiment.
dried target gel pieces were then digested by trypsin (Sigma, MO), mixed with 20 mL of 0.05 M NH4HCO3, overnight at 37 C. The target protein band was identified, using a 4700 Proteomics Analyzer MALDI-TOF/TOF mass spectrometer (Applied Biosystems, Foster City, CA). Mass spectrometry data were analyzed with Mascot bioinformatics database search engine (Matrix Sciences, London, UK). 2.5. Western blotting We used methods which were described in detail previously [24,27] with a few modifications. The purified rhK8 was loaded into 10% polyacrylamide gel and separated. Then, the target proteins were transfer to polyvinylidene fluoride (PVDF; Merck Millipore, MA) membranes. The PVDF membranes were blocked with 5% nonfat milk in PBS at 4 C overnight. After blocking, we cut up PVDF membranes into strips and incubated them with sera of RA patients, disease controls or healthy controls (1:1000, diluted in 1% nonfat milk) overnight at 4 C. Then, the PVDF membranes were washed by 0.2% PBST for three times to remove uncombined antibodies and each wash lasted for 10 min. Later, we Incubated the PVDF membranes with horseradish-peroxidase conjugated goat anti-human IgG (1:10,000, diluted in 1% nonfat milk) (ImmunoHunt, Beijing, China) for 1 h at 37 C. The potential target bands were detected by enhanced chemiluminescence kit (Applygen, China) and exposed later.
2.3. Protein expression and purification 2.6. Dot-ELISA Protocols of gene recombination, protein expression and purification, certified by our previous study, were used [25]. In brief, total mRNA was extracted from aforementioned cell line by TRIzol reagent (Invitrogen, CA). The designed primer: Forward: (EcoR I) 50 CCGGAATTCGTGACCCAGAAGTCCTACAAG-30 , Reserve: (SaI I) 50 CGCGTCGACTTATTAATCAGAGGACTCAGACACCAG-30 , were used to amplify K8 gene by RT-PCR, carried out with the kit's instruction (Fermentas, MD). Then, the K8 gene was linked with pET-28(a) vectors, which were transformed into E. coli DH5a. Transformational E. coli DH5a strain was recovered in 400 mL Super Optimal Broth (SOC) cultivation for 1.5 h. Later, we transferred the bacterium culture solution to Luria Broth (LB) cultivation, containing kanamycin (50 mL/mL). Extracting recombinant plasmids from E. coli DH5a and transforming these plasmids into high expression engineering strain E. coli BL21, using the same method that mentioned in this section. Under the optimal condition: 37 C, 200 rpm shaking speed, overexpression of K8 was achieved by adding IPTG (1 mM) when optical density (OD) reaches 0.4e0.6. Finally, the recombinant K8 with N-terminal hexahistidine-tag was purified by Ni-NTA resin (Qiagen, Hilden, Germany), according to instruction supplied by manufacturer. After determining the concentration of protein by a BCA assay kit (Biosynthesis Biotechnology, Beijing, China), the purified protein was stored at 80 C for further experiments. 2.4. Mass spectrometry The method in our previous study was applied to this experiment [24,26]. Briefly, the target protein band was cut off from SDSPAGE gel. After destaining the target gel pieces with 25 mM NH4HCO3 and 50% acetonitrile, the mixture was dried by a vacuum centrifuge. Then, 25 mM NH4HCO3 and 10 mM dithiothreitol were added to cover the target gel pieces and reduced for 2 h at 37 C. The DTT solution was replaced by roughly same volume of 55 mM iodoacetamide in 25 mM NH4HCO3 and incubated for 45 min at room temperature in the dark. Next, 50% acetonitrile in 25 mM NH4HCO3 were used to wash the target gel pieces for 10 min. The liquid mixture was completely dried by vacuum concentrator. The
To perform this experiment, we referred to our previous research [28]. We spotted about 0.07 mg keratin 8 on the dull side of a 0.22-mm nitrocellulose membrane. After drying the membrane, 10% goat sera (diluted in PBS) was used to block the antigen discs for 1 h at 37 C. Then the sera of RA patients (1:500, diluted in PBS) or healthy controls (1:500, diluted in PBS) were added to the discs and incubated for 1 h at 37 C. PBST (phosphate buffered saline with 0.1% Tween-20, pH 7.4) was used to wash the discs for 3 times later. After removing the liquid phase and dry, the discs were incubated with goat anti-human IgG/HRP (1:1000, diluted in PBS) (ImmunoHunt, Beijing, China) for 1 h at 37 C. Next, the washing process was repeated as aforementioned and 3, 30 -diaminobenzidine solution (ZSGB-BIO, Beijing, China) was added and colorated for 5 min. At last, adding deionized water to stop the reaction. 2.7. ELISA The detection of potential autoantigen with ELISA was performed, according to previous studies [25,27]. Coating the 96-well microtiter plates with rhK8 (100 ng/mL, dissolved in 0.05 M carbonic buffer, pH ¼ 9.6) overnight at 4 C. After removing the liquid phase, each well was blocked with 100 mL goat sera (1:10, diluted in PBS) for 2 h at 37 C. Then the liquid phase was removed again and the sera of RA, SLE patients and healthy controls (1:100, diluted in PBS) were used to incubate the 96-well microtiter plates for 2 h at 37 C. Five-time wash was performed here to remove unbind autoantibodies. Next, 100 mL goat anti-human IgG/HRP (1:20,000, diluted in PBS) (ImmunoHunt, Beijing, China) was added to each well and incubated for 1 h at 37 C. Again, wash each well with 0.1% PBST for five times to elute the second antibodies. Orderly, adding 50 mL tetramethylbenzidine (TMB) A (0.1 M citric acid, 0.2 M NA2HPO4, 0.6 g hydroperite/L) and 50 mL TMB B (5 mM citric acid, 0.4 mM EDTA-NA2, 0.2 g TMB/L) to each well. After keeping the plates at a dark, room-temperatured place for 15 min, the reaction was stopped by adding 50 mL stop buffer (2 M H2SO4). Finally, OD value of each well was measured by ELISA reader (Tecan, Switzerland) at 450/620 nm.
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
667
2.8. Statistical analysis Receiver operating characteristic (ROC) analysis was carried out with MedCalc (Version 9.2.0.1, Mariakerke, Belgium). Fisher's exact test was performed to compare differences between cohorts. The scatterplot of ELISA result was evaluated by GraphPad Prism (Version 5.0, GraphPad Software Inc., La Jolla, CA). P values were considered to indicate statistical significance only when they were less than 0.05. 2.9. Antigenic determinant prediction As aforementioned, since K84 was a probable antigen of RA [12], K8 and K84 were selected to perform sequence alignment using BLAST to evaluate their homology [29]. Then Bepipred Linear Epitope Prediction [30] was used to predict epitopes of two proteins. Potential common epitopes of two proteins were selected with standards that amino acid lengths were no less than 8 and cooccurred in K8 and K84. 3. Results 3.1. Expression and purification of recombinant K8 protein Human K8 gene was amplified by Polymerase Chain Reaction technique. The production was verified by 0.8% Agarose Gel Electrophoresis. A target band was obtained at 1500 bp and no other specific amplification appeared (Fig. 1)A. After cultivating under the optimum condition mentioned before, a rough 53-KDa protein band appeared. And the band was the putative K8 protein when compared with recombinant E. coli BL21 (DE3) strain without induction of IPTG. Then his-tag contained target protein was obtained by Ni-NTA resin (Fig. 1B). 3.2. Confirmation of keratin 8 The protein obtained by Ni-NTA resin was reassured by mass spectrometry. The result shows that individual ions scores are more than 41, indicating the putative protein is highly homologous identity (P < 0.05). And protein scores are greater than 70, which is significant (P < 0.05) to rank a protein hit (Fig. 1C).
Fig. 2. Demonstration of K8 antigenicity in RA. (A) Western blotting result. Three positive results among 10 RA patients were named RA1, RA2 and RA3. A 53-kDa band was observed. Two negative results were selected randomly from 7 negative results, named RA4, RA5. All healthy controls showed negative results. (B) Dot-ELISA result. One positive resultdRA1dwas then used to perform Dot-ELISA to assure the antigenicity of K8. Compared with blank control (PBS; using PBS instead of patient sera) and healthy control (HC1), only RA1 showed positive result (the dotted line circle).
3.3. Western blotting and Dot-ELISA K8 associated Western blotting was performed with sera of 10 RA patients and 5 healthy people, who were selected randomly from our sera bank. Among these 10 RA patients, 3 patients showed positive antigeneantibody reaction and all healthy controls showed negative results, indicating K8 was a potential autoantigen in RA patients (Fig. 2A). And the Dot-ELISA result of RA1 was consistent with the result of Western blotting (Fig. 2B). 3.4. Significance of K8 in RA ELISA of K8 with 50 RA patients was carried out in order to confirm the antigen-specificity of K8 in RA patients. At same time, disease controlsd50 SLE patients' serad were involved in ELISA with the same operating method. 56 healthy people's sera were used as healthy controls. The cut-off value of positivity of K8 in RA disease was obtained by ROC curve (Fig. 3A). The cut-off value was 0.205, at which the sensitivity and specificity were 68% and 80.36%, respectively. All ELISA results were showed in scatterplot (Fig. 3B). The positive rates of K8 in HC, RA and SLE patients were 20% (11 of 56 healthy controls), 68% (34 of 50 RA patients. P < 0.001 compared with healthy controls) and 24% (12 of 50 SLE patients).
Fig. 1. Gene amplification, expression, purification and identification of Keratin 8. (A) PCR production of K8 gene. (B) Expression and purification of K8 protein. Symbol “M” and Bands 2e4 represented DNA Marker, purified K8 protein (His-tag contained) and protein expression of BL21 strains which contained the recombinant plasmid K8-pET-28a (þ) with or without 1 mM IPTG induction, respectively. (C) Mass spectrogram. The target protein was identified as K8. The molecular weight of target protein was 53.670-kDa, conformed to expected size: 53-KDa. Symbol “*” noted the matched peptides displayed in the top right corner.
668
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
Fig. 3. Significance of K8 in RA. (A) Receiver operating characteristic (ROC) curve was drawn to indicate prevalence of K8 antibodies in RA. (B) Scatterplot of all ELISA results demonstrated the difference between RA and HC. The marker “***” manifests P < 0.001.
Fig. 4. Antigenic epitopes prediction and comparison between K8 and K84. (A) Prediction result of K8. (B) Prediction result of K84. The three similar epitopes were emphasized with red underlines. Epitope sequences were displayed beyond the underlines. (C) Amino acid properties of three similar epitopes were further analyzed. For each site, amino acids, which had similar properties (in green) and those were identical in both keratins (in red) were showed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
4. Discussion Historically, the term Keratins widely aroused people's attention to RA in 1979, with the discovery of so-called AKA [14], which had been widely used as a clinical diagnosis marker of RA. Later, experiment demonstrated that one probable target protein of socalled AKA was filaggrin [15], though, no other detailed evidences could illustrate flilaggrin was the only antigen of AKAs. Usually, AKA was detected by indirect immunofluorescence and no confirmed ELISA test method had been established [18]. Since K8 was a keratin that must exist in epidermal keratinocytes [13], we assumed the antibody of K8 was a supplement of so called AKA and may be detected by ELISA in the future. Actually, some keratins had been demonstrated that played significant roles in human diseases. Among all keratins, K8 was considering highly disease-related [6]. It had been reported that many K8 related biological processes such as K8 genetic mutation and hypersumoylation were connected with liver injury and liver diseases [31,32]. K8 was also useful for tumor classificatory diagnosis [33,34] and could be used as a prognostic marker [35]. In 2013, an experiment had identified that K84 was a screening citrullinated protein in tissues of synovial in RA. Even though they found K84 was a putative antigen [12], except K84, other keratins were not confirmed related to RA so far.
In this study, we evaluated K8 as a potential antigen, which was demonstrated by ELISA and Western blotting. Since K8 and K84 belong to a same sub-family: type II family in keratin and they are all specific antigens in RA, we make a further investigation to predict the role K8 and K84 play in RA. In the result of Bepipred Linear Epitope Prediction, twenty-six and thirty-six epitopes were predicted for K8 and K84, respectively. According to the standards mentioned before, three significant homologous epitope pairs were selected (Fig. 4), indicating the similar epitopes engaged in two keratins and might cause antigen-activation in RA. In conclusion, we identified anti-K8 autoantibody in some of the patients with rheumatoid arthritis and K8 was probably one of the real antigens of traditional so-called AKA. Conflict of interest All co-authors have no competing interests. Acknowledgments This work was supported by Beijing Nova Program of Science and Technology (No. 2007B024), Program for New Century Excellent Talents in University (No. NECT-12-0773), and the National Natural Science Foundation of China (No. 31371203).
X. Wang et al. / Biochemical and Biophysical Research Communications 465 (2015) 665e669
Transparency document Transparency document related to this article can be found online at http://dx.doi.org/10.1016/j.bbrc.2015.07.161.
[17]
[18]
References [1] S. Gertel, H. Amital, Putative approaches to bypass the citrulline-specific autoimmune response in rheumatoid arthritis, Isr. Med. Assoc. J. 16 (2014) 587e590. [2] P.R. Mills, R.N. MacSween, W.C. Dick, et al., Liver disease in rheumatoid arthritis, Scott. Med. J. 25 (1980) 18e22. [3] C. Caramella, J. Avouac, P. Sogni, et al., Association between rheumatoid arthritis and primary biliary cirrhosis, Jt. Bone Spine 74 (2007) 279e281. [4] O. Palm, B. Moum, J. Jahnsen, et al., The prevalence and incidence of peripheral arthritis in patients with inflammatory bowel disease, a prospective population-based study (the IBSEN study), Rheumatol. Oxf. 40 (2001) 1256e1261. [5] D. Sohn, T.C. Chin, M.J. Fellner, Multiple keratoacanthomas associated with steatocystoma multiplex and rheumatoid arthritis. A case report, Arch. Dermatol. 116 (1980) 913e915. [6] D.M. Toivola, P. Boor, C. Alam, et al., Keratins in health and disease, Curr. Opin. Cell Biol. 32 (2015) 73e81. [7] G. Steiner, K. Hartmuth, K. Skriner, et al., Purification and partial sequencing of the nuclear autoantigen RA33 shows that it is indistinguishable from the A2 protein of the heterogeneous nuclear ribonucleoprotein complex, J. Clin. Invest. 90 (1992) 1061e1066. [8] E.H. Halvorsen, S. Pollmann, I.M. Gilboe, et al., Serum IgG antibodies to peptidylarginine deiminase 4 in rheumatoid arthritis and associations with disease severity, Ann. Rheum. Dis. 67 (2008) 414e417. [9] K. Conrad, D. Roggenbuck, D. Reinhold, et al., Profiling of rheumatoid arthritis associated autoantibodies, Autoimmun. Rev. 9 (2010) 431e435. [10] I. Kurokawa, K. Takahashi, I. Moll, et al., Expression of keratins in cutaneous epithelial tumors and related disordersedistribution and clinical significance, Exp. Dermatol. 20 (2011) 217e228. [11] Y.H. Chou, O. Skalli, R.D. Goldman, Intermediate filaments and cytoplasmic networking: new connections and more functions, Curr. Opin. Cell Biol. 9 (1997) 49e53. [12] X. Chang, Y. Zhao, Y. Wang, et al., Screening citrullinated proteins in synovial tissues of rheumatoid arthritis using 2-dimensional western blotting, J. Rheumatol. 40 (2013) 219e227. [13] J. Schweizer, P.E. Bowden, P.A. Coulombe, et al., New consensus nomenclature for mammalian keratins, J. Cell Biol. 174 (2006) 169e174. [14] B.J. Young, R.K. Mallya, R.D. Leslie, et al., Anti-keratin antibodies in rheumatoid arthritis, Br. Med. J. 2 (1979) 97e99. [15] M. Simon, E. Girbal, M. Sebbag, et al., The cytokeratin filament-aggregating protein filaggrin is the target of the so-called “antikeratin antibodies”, autoantibodies specific for rheumatoid arthritis, J. Clin. Invest. 92 (1993) 1387e1393. [16] C. Vincent, M. Simon, M. Sebbag, et al., Immunoblotting detection of
[19] [20] [21]
[22] [23]
[24] [25]
[26] [27]
[28]
[29] [30] [31]
[32] [33] [34]
[35]
669
autoantibodies to human epidermis filaggrin: a new diagnostic test for rheumatoid arthritis, J. Rheumatol. 25 (1998) 838e846. M. Sebbag, M. Simon, C. Vincent, et al., The antiperinuclear factor and the socalled antikeratin antibodies are the same rheumatoid arthritis-specific autoantibodies, J. Clin. Invest. 95 (1995) 2672e2679. C. Vincent, F. de Keyser, C. Masson-Bessiere, et al., Anti-perinuclear factor compared with the so called “antikeratin” antibodies and antibodies to human epidermis filaggrin, in the diagnosis of arthritides, Ann. Rheum. Dis. 58 (1999) 42e48. S.L. Slack, M. Mannik, B.A. Dale, Diagnostic value of antibodies to filaggrin in rheumatoid arthritis, J. Rheumatol. 25 (1998) 847e851. Y.J. Xiang, S.M. Dai, Prevalence of rheumatic diseases and disability in China, Rheumatol. Int. 29 (2009) 481e490. S.M. Dai, X.H. Han, D.B. Zhao, et al., Prevalence of rheumatic symptoms, rheumatoid arthritis, ankylosing spondylitis, and gout in Shanghai, China: a COPCORD study, J. Rheumatol. 30 (2003) 2245e2251. S.E. Gabriel, The epidemiology of rheumatoid arthritis, Rheum. Dis. Clin. North Am. 27 (2001) 269e281. F.C. Arnett, S.M. Edworthy, D.A. Bloch, et al., The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis, Arthritis Rheum. 31 (1988) 315e324. P. Chen, H. Yan, Y. Tian, et al., Annexin A2 as a target endothelial cell membrane autoantigen in Behcet's disease, Sci. Rep. 5 (2015) 8162. H. Du, G. Chen, S. Wang, et al., Immunological screening and characterization of highly specific monoclonal antibodies against 20 kDa hGH, Bioanalysis 4 (2012) 2161e2168. Y. Xun, P. Chen, H. Yan, et al., Identification of prohibitin as an antigen in Behcet's disease, Biochem. Biophys. Res. Commun. 451 (2014) 389e393. H. Du, C. Li, H. Jin, et al., Generation and evaluation of antibodies against human MGF E-peptide by reverse phase protein microarray and reverse competitive ELISA, Bioanalysis 5 (2013) 2269e2275. H. Du, J. Liu, Y. Xun, et al., Determination of deoxynivalenol, zearalenone, aflatoxin B1, and ochratoxin by an enzyme-linked immunosorbent assay, Anal. Lett. 47 (2014) 1912e1920. S.F. Altschul, W. Gish, W. Miller, et al., Basic local alignment search tool, J. Mol. Biol. 215 (1990) 403e410. J.E. Larsen, O. Lund, M. Nielsen, Improved method for predicting linear B-cell epitopes, Immunome Res. 2 (2006) 2. N.T. Snider, S.V. Weerasinghe, J.A. Iniguez-Lluhi, et al., Keratin hypersumoylation alters filament dynamics and is a marker for human liver disease and keratin mutation, J. Biol. Chem. 286 (2011) 2273e2284. P. Strnad, S. Paschke, K.H. Jang, et al., Keratins: markers and modulators of liver disease, Curr. Opin. Gastroenterol. 28 (2012) 209e216. R. Moll, M. Divo, L. Langbein, The human keratins: biology and pathology, Histochem. Cell Biol. 129 (2008) 705e733. H. Tsubochi, T. Suzuki, S. Suzuki, et al., Immunohistochemical study of basaloid squamous cell carcinoma, adenoid cystic and mucoepidermoid carcinoma in the upper aerodigestive tract, Anticancer Res. 20 (2000) 1205e1211. T. Knosel, V. Emde, K. Schluns, et al., Cytokeratin profiles identify diagnostic signatures in colorectal cancer using multiplex analysis of tissue microarrays, Cell. Oncol. 28 (2006) 167e175.