High-throughput Methods Integrated Analyses For Atopic Dermatitis: OMICS Studies For Constructing Hub Proteins And For Probing Biomarkers

BackgroundAtopic dermatitis(AD) patient populations have been prevalent along with modernization and industrialization.AD is not only a critical skin disease,but also is the hot issue of the public health.AD is a chronic relapsing inflammatory skin disease that's characterized by the distribution of eczematous skin lesions.AD typically reveals lichenification,pruritic excoriations and dry skin with a variety of pathophysiologic manifestations.Since AD is a systemic disorder,it simultaneously triggers asthma,allergic rhinitis and food allergy,where the serum IgE and peripheral eosinophils are elevated.Several cells such as eosinophils,T lymphocytes, Langerhans cells and keratinocytes have been regarded as important for the pathogenesis of AD since these cells are directly related with several blood factors such as cytokines,chemokines and IgE,and they are also related with microbial infection.Although most AD patients have high serum IgE levels,one subgroup of AD patients lacks the sensitization to aero-allergens or food allergens.Two types of AD have been defined,and they are the extrinsic and intrinsic types.The extrinsic type of AD(ADe) has a Th2 cell mediated high serum IgE level and this condition is associated with IgE-mediated sensitization;this is seen in 70-80%of AD patients.As a counterpart to ADe,the intrinsic type(ADi) is not involved with IgE-mediated sensitization and allergies,and ADi is seen in about 20-30%of AD patients.High-throughput screening(HTS) methods such as DNA microarray and 2D-PAGE/MALDI-TOF approaches are currently being used to determine the candidate genes associated with AD.With HTS,although we can obtain a great deal of genes/proteins that are overexpressed or down-regulated in AD disease,it is very difficult to choose the most important genes.Furthermore,it is hard to explain the meaning of those proteins.In this regard,interactomic tools such as performing PPI mapping may help to find the hub proteins among the detected candidate genes.Since the next step of functional studies is a time-and cost-consuming process,the number of target proteins must be limited;hence,to make the right choice,computational prediction on the basis of HTS results could be critical.The candidate genes listed in the PPI maps may provide important information to piece together the complex nature of AD disease.Regardless of the various approaches that are currently being used to treat AD,the number of AD patients has greatly increased and more effective treatments that offer quick relief of severe symptoms still remain to be developed. Thus,more insights regarding the complex gene/protein regulation are needed into the pathogenesis of AD,in addition to elucidating the immunologic aspects of this disease.Besides AD studies,we also conducted the molecular level studies of environmental toxicology.Once the environmental chemicals absorbed into the body, they can interact with the most important biological macromolecules such as proteins. The chemicals not only can form the covalent protein adducts,but also may affect the protein conformation as well as lead to protein misfolding and thus,result in environmental health problems.In this study,we used CK-BB as a model material for trying to explore the effects of small molecules on the enzyme structure and function. Based on the understanding of CK-BB unfolding and refolding,we further studied the mechanisms of CK-BB binding with sodium dodecyl sulfate(SDS) that is abundantly presented in modern life and also,studied the effects of two environmental chemicals such as acrylamide and Zn²⁺ which may be related to degenerative diseases. Method1.All AD samples were divided into the two types defined as the extrinsic(ADe) and intrinsic(ADi) types.AD skin tissues were collected from non-asthmatic atopic patients.Non-atopic control subjects who had neither a personal/family history nor any signs of atopic diseases were selected for the study.Punch biopsies of 3-5 mm diameter were individually obtained from the lesional skin of both the ADe and ADi patients.We recruited 20 ADe patients,20 ADi patients and 22 normal specimen donors.The mean values for each group's total IgE level,eosinophilia count,age and SCORAD score were 1739 U/mL,625/μL,27.4 year old and 50.09 for the ADe group and 103 U/mL,248/μL,28.2 year old and 46.3 for the ADi group,respectively. The gender of all samples was male.Serum was collected according to previously reported methods and stored at-80℃after proper aliquots.DNA microarray was performed following by the total RNA preparation, labeling and hybridization:the individually prepared normal,ADe,and ADi RNA samples were correspondingly prepared,and the further cDNA template synthesis and labeling for each of samples were conducted.The two kinds of commercial DNA chips(Affymetrix Human Genome U133 Plus 2.0 Array chip;and GenePlorer Twinchip Human-8K) were individually applied for different samples.These chips were hybridized with a mixture of fluorescently labeled cDNAs from the control and AD samples at 42℃for 16 h in the presence of hybridization solution.Image analysis to obtain the gene expression ratios was performed properly and the logged gene expression ratios were normalized by a LOWESS regression procedure.By using DEG and SAM,the data sets were analyzed.The genes showing significant expression changes(＞2-fold) were selected with a cutoff of 5%of the q value.An Affymetrix microarray was used to analyze ADe and ADi skin tissues,and the 8K cDNA microarray was used to analyze atopic keratinocytes.These tools were probed the dysregulated genes at the transcriptional levels.The free-flow electrophoresis (FFE) was used to analyze atopic fibroblasts and two-dimensional electrophoresis (2-DE) was used to analyze serums with proper treatments.Based on the above results,the interactomic tool was also adapted to predict putative hub genes or proteins.2-DE analyses of AD patient-derived serums were conducted after treating with multiple affinity removal columns(MARC).In addition,fractionation method by using Phenyl Sepharose resin was conducted to analyze AD serum proteins.Serums were properly fractionation with a column following by TCA precipitation.In the next step,PPI of AD related proteins were predicted by bioinformatics algorithms.The prediction of PPI uses most of the major types of PPI algorithms. They are:1) Protein Structural Interactome MAP(PSIMAP),a method that uses the structural domain of the SCOP(Structural Classification of Proteins) database;2) Protein Experimental Interactome MAP(PEIMAP),a common method that uses public resources of experimental protein interaction information such as HPRD, BIND,DIP,MINT,IntAct and BioGrid.2.Recombinant human CK-BB was expressed and purified according to the previous reports.CK-BB activity was measured following proton generation during the reaction of ATP and creatine with the indicator thymol blue at 597 nm.The fluorescence emission spectra and the ANS labeling fluorescence were studied for monitoring the tertiary structural changes.The far-UV circular dichroism(CD) spectra were recorded in the range of 190-250 nm for measuring secondary structural changes.Isothermal titration microcalorimetric measurements were performed with a 3101 TAM III Thermostat in accordance with a previous report.We retrieved the known homologous structures of CK-BB from the Protein Data Bank(PDB) and found that bovine retinal creatine kinase chain A(PDB entry:1g0w) was a very good structural template(97%sequence identity) as a close homologue for CK-BB.The docking procedures were categorized to 1)calculate receptor and ligand 3D coordinates,2)characterize docking site,3)compute grid score,and 4)determine the docking for complementarity with receptors and ligand.Result1.From the Affymetrix microarray results,406 genes were selected:130 genes were obtained from normal vs.AD(ADe and ADi);327 genes were obtained from normal vs.ADe;146 genes were obtained from normal vs.ADi.The 8K cDNA microarray results showed that 157 genes were significantly dysregulated in primary atopic keratinocytes compared to the control subjects.In the next step,we obtained 93 candidate proteins by using FFE in AD fibroblasts.In AD serums,30 proteins were detected as up-regulated and 19 proteins were down-regulated in AD serums after using MARC treatment.With adapting fractionation method,9 proteins were detected as dysregulated.In Affymetrix microarray analysis,five hub genes such as the PIK3R1,IGF1R, MAPK13,SFN and YWHAE were predicted by using the PSIMAP algorithm and eight_genes such as CQRIN,COL14A1,CLCA4,PTPRC,IL10RA,C1S,PHLPP and SLITRK6 were predicted by using the PEIMAP algorithm.In 8K cDNA microarray, 25 genes were predicted for AD keratinocytes hub proteins.In FFE result, overlapping of the PEIMAP and PSIMAP algorithms detected 4 hub proteins:I55423, ARR3,YWHAE,and EEF1A1.In 2-DE results of AD serums,26 proteins were predicted for AD serum samples.SFN and PTPRC were also further validated by real-time PCR in AD tissues. The results showed that the transcriptional level of SFN is up-regulated 4.5-fold in ADi,and PTPRC is down-regulated 60%in ADe and 46%in ADi.In the ELISA results,MMP1 and MMP10 were consistently detected as up-regulated in the serum of ADi patients.The results of ELISA studies of the serum of ADi patients showed that both MMP1 and MMP10 were significantly up-regulated(almost 2-fold) compared to normal and ADe patients.Statistical significance was confirmed by p-values.Randomized pairings and the Wilcoxon-rank sum test were applied.2.We have also investigated the inactivation by using SDS to study CK-BB's folding behaviors.SDS strongly inhibited the CK-BB activity in a noncompetitive inhibition manner(K_i=1.22 mM).The tertiary structural change was induced by SDS binding with the exposure of hydrophobic surface.The changes of thermodynamic parameters for the SDS ligand binding such as enthalpy,Gibbs free-energy,and entropy were obtained as-13±7.0 MJ/mol,8.39 kJ/mol and-42.754 kJ/(K·mol),respectively.The predicted structure had a Root Mean Square Deviation of 0.51 A.The docking between CK-BB and SDS was successful with significant scores(-4.67 kcal/mol, AutoDock4 and-48.32 kcal/mol,DOCK6).We studied the inhibitory effects of acrylamide on CK-BB.We found that CK-BB was kinetically inactivated by acrylamide accompanied by the disruption of the hydrophobic surface.Acrylamide mainly interacted with the thiol(-SH) residue of CK-BB and resulted in alkylation.A computational docking simulation supported that acrylamide directly bound to the active site of CK-BB where CYS74 and GLY73 residues interacted mainly.Zn²⁺ inactivated the activity of CK-BB in a dose dependent manner(IC₅₀=0.06 mM).The spectroflurorimetry results showed that Zn²⁺ conspicuously induced the tertiary structural change of CK-BB with exposure of its hydrophobic surfaces.We also found that CK-BB aggregation was induced by Zn²⁺.This aggregation was dependent on the temperature and the enzyme and Zn²⁺ concentrations.Conclusion1.With respect to explaining all the events that occur in the pathogenesis of AD or for probing how many genes are involved in the development of AD,our large-scale microarray data can suggest new information.In the initial part of results,the data were mainly focused on profiling the AD associated genes in a large scale DNAchip array and this provides useful information for the overall insight into the AD disease at the level of skin tissues,but our results do not include any detailed functional study. Then,primary cultured keratinocytes and fibroblasts were studies.The profiling of differentially expressed proteins in AD serums was also conducted.In respect to the complexity of the AD pathogenesis where many factors are associated with AD and various triggers are known to exist among different patient,our data suggest that human epidermal keratinocytes and dermal fibroblasts actively participate in AD disease along with the other types of cells.However,some of the genes/proteins observed in the primary cultured keratinocytes and fibroblasts may not be representative of the in vivo changes,and the differences between the mRNA levels and the expressed protein levels limit our understanding with regard to their exact roles in AD. Taken together,the large scale microarray provided information to further understand the complex factors associated with pathogenesis of AD.The approach of combining the expression data analysis and predicted protein interaction partners performed in large scales can bring valuable and more reliable gene targets for AD. The predicted protein interaction partners and validated hub proteins can offer valuable and reliable targets for further functional studies of AD treatment.The number of the AD patients is increasing rapidly and being able to make a specific diagnosis and administering case sensitive treatment are needed.Therefore,OMICS approaches that are based on high-throughput technologies may help researchers gain additional insight or they may reveal pieces of the puzzle into the pathogenesis of AD.According to the results,several points are suggested in this study:i)the bioinformatic tools combined with the HTS data identified the new candidate hub genes/proteins,which can be useful in the further studies,especially for the functional studies or clinical trials of drug development;ii)several important genes such as SFN, PTPRC,MMP1 and MMP10 have been newly suggested in this study,which they are significantly up-regulated or down-regulated in AD patients-derived samples and they can be biomarkers for AD;iii)most of the predicted hub proteins and real-time PCR probing candidate genes/proteins detected in this study have not been previously reported to be involved in AD pathogenesis.This may be due to the unknown role of these factors in AD pathogenesis;iv)in addition to the fact that the combinations of several research areas such as genomics,proteomics,and interactomics can help guide research into the complex pathogenesis of AD and assist in designing precise therapeutic strategies to find biomarkers or to determine the right clinical target molecules.2.In this section,we used CK-BB as a model enzyme to explore changes of structures and functions induced by small chemical molecules and elucidated the toxicological pathways of CK-BB unfolding and inactivation,as well as aggregation. The aggregation phenomenon has been generally recognized as an important factor of degenerative disease.Furthermore,it is worthy to study further for relationships between industrial chemicals and chronic neurotoxicity of human health.