It is of critical importance to identify new biomarkers and hints of mechanisms of disease states and disease progression in illnesses affecting large populations such as chronic inflammatory diseases in order to personalize treatment and accurately monitor disease progression. Methods based on genomic approaches and amongst them global gene expression profiles play an increasing role in such efforts.
The pathomechanism of these chronic inflammatory diseases has been examined at different levels including cell cultures, animal models and clinical samples. Importantly, as the molecular signature of disease across tissues is overall more prominent than the signature of tissue expression across diseases , there is good reason to believe one can detect disease-specific gene expression signatures in PBMC samples which are easy-to-access.
There have been studies analyzing the gene expression patterns of these diseases in PBMC samples which led to the identification of genes, which we also found to be significantly different between disease and control patients. Examples include PTGS2 in psoriasis ; PTPN22 in RA  and GZMK in IBD . Other groups compared different chronic disorders to each others such as juvenile arthritis and spondyloarthropathy  or psoriatic arthritis, RA and spondyloarthropathy , but such an extended number of genes have not yet been found in peripheral blood. Importantly the gene expression profiles of RA, IBD and psoriasis have not been compared previously. We tried to fill this gap with our study and identify markers of chronic inflammation that overlap among the 3 conditions and also ones that discriminate them from each other.
It should be noted though gene expression analyses of these diseases provide little insight into the pathomechanisms of the disease simply because it is impossible to sort out what is cause and what is consequence. However it can still provide hints or clues about the potential pathways affected. Therefore such datasets, although indirectly, can form the basis of more mechanistic studies. This is why we have carried out and presented the gene interaction analyses of the genes identified and linked to the different diseases. Significantly the existence of overlapping gene sets suggests that there are common pathways impacted among the studied diseases. A common characteristic of all diseases is an altered inflammatory response.
In addition, genetic association studies have already identified SNPs linked to these diseases. A comparison of our dataset to these can provide further support to those findings [Table 1].
Several of the now identified differentially expressed genes have been linked to the studied diseases although either at a different level or different model. Regarding the IBD gene panel, OLR1/LOX1  and CCR1  both have immune response function and are expressed in human intestinal cells. The homophilic ligation of CEACAM1, which is a cell-surface molecule, on T cells leads to the inhibition of a range of T-cell functions therefore it might be a new potential therapeutic target in the treatment of IBD . MMP9, a gelatinase, that showed a 7-fold up-regulation in diseased samples in our study is an important mediator of tissue injury in colitis and is up-regulated in sputum samples of CD patients . Garg et al. suggested that developing strategies to block MMP-9 activity in the gut might be of benefit to IBD . TIMP1 is a natural inhibitor of matrix metalloproteinases (MMPs) and might be considered a potential drug target. The DNA methylation of IFNG that regulates immune response was analyzed within the mucosal compartment in both normal and IBD populations . PTGS2/COX2 is related to IBD expression level in colon biopsy . ABCA1, BASP1, GZMK, HBB, SEC14L1, THEM2 and TNFAIP6 have never been associated with IBD; therefore further analyses should be carried out.
In the psoriasis panel, CCR1 that plays a role in inflammatory response is associated with psoriasis via the analysis of skin lesion samples . Koczan et al. identified pairs of genes such as PTGS2/COX2 and NR1D2 which allowed an accuracy of disease stage prediction of 86%, based on gene expression patterns . ADORA2B, IFRD1, ITGA2B, MAP4K1, PADI4, PDE4A, SEC14L1 and TREM1 have no documented association with psoriasis yet. Therefore these are candidates for further SNP/genetic association or mechanistic studies.
Regarding the significantly changing genes in RA samples, IL8 is related to RA based on expression . Van Der Voort et al. demonstrated that the expression of ADAM10 is strongly enhanced in RA synovia . CCL4 was expressed in PBMC  and CCL5 showed high serum levels . Kehlen et al. found up-regulation of TNFAIP6 in fibroblast-like synoviocytes of patients with RA . HMGB1  and PLCB1  only showed difference in protein level in RA patients. PTGS2/COX2 mRNA levels in PBMC samples from RA patients were within the normal range or below normal . ADAM12, ADAM33, ADORA2B, AKT3, ANXA3, CDKN1C, CTSL1, CUGBP1, CUGBP2, CYP51A1, FGL2, HBB, HSP90AA1, KLRF1, NR1D2, SLC33A1 and THEM2 have no evidence of association with RA and should further be studied.
Gene expression profiling may allow early diagnosis, aid in identifying prognosis or subtypes. We found genes that differentiate between CD and UC in IBD of which some have already been associated with the disease but at a different level or in a different tissue. Alexander et al. reported that the expression level of FOS was two- to threefold higher in involved than in uninvolved areas of the colons of two UC patients . van der Pouw Kraan at el. found that ETS2 may be an important transcription factor driving inflammation in acute as well as chronic inflammatory diseases such as IBD . As oppose to UC, IL-18 may serve as indicators of acute phase reactivity in CD, according to Haas et al. , that correlates with our findings. Mitsuyama et al. suggested that IL-6 trans-signaling may play role in the development of IBD and imply the possibility of a selective therapeutic strategy to target this signaling .
A gene panel containing 8 genes shows significant differences between patients with and without arthritis in psoriasis. Hitchon et al. described that MMPs, especially MMP2 and MMP9, have been implicated in several features of inflammatory arthritis including angiogenesis and bone erosions . Partsch et al. measured the protein level of IL4 in synovial fluid of patients with psoriatic arthritis (PsA) .
Another gene set including HBB, GZMK and PADI4 separates different states of prognosis in RA regarding MR-confirmed bone erosion. In RA, PADI4 is a target of autoantibodies; and its increased expression in synovia of RA patients , also its relation to the disease at SNP level have already been confirmed . Osteoclasts are shown to be mainly involved in the bone-destruction of RA which may indicate further studies regarding the role of PADI4 in bone erosion.
We consider a key aspect of our work is the identification of five genes including ADM, AQP9, CXCL2, IL10 and NAMPT that differentiate between samples from patients with chronic inflammation and healthy controls. ADM that plays role in response to wounding is found to be distributed on the surface of the human colonic mucosa  and the plasma level of ADM in RA patients was significantly higher compared to healthy controls in synovial tissue . CXCL2 that has function in immune response is over-expressed in osteoarthritis fibroblasts rather than rheumatoid fibroblasts , up-regulated in psoriatic epidermis , and also in epithelial tissue . IL10 is expressed in PBMCs of psoriasis , RA  and IBD patients . The expression of NAMPT which is a pre-B cell-enhancing factor is increased in colonic biopsy specimens of IBD patients compared to healthy controls , it is also up-regulated in plasma and synovial fluid of RA patients ; and in PBMC of psoriasis in the diseased stage . AQP9 that has an activity in immune response has never been associated with any of these medical conditions.
These genes might serve as universal markers of chronic inflammation. As expected, IL10 seems to be the key gene in this network regarding biological processes (Figure 3b). As the number of biological processes related to ADM is relatively high and it has no direct interactions with the genes we examined, there might be other potential markers that play role in the pathomechanism of the conditions we analyzed that also have similar pathways or targets as ADM. Chronic inflammatory and autoimmune diseases share a number of phenotypic and genetic characteristics suggesting common etiological pathways or pathomechanisms. Becker et al. used meta-analyses of whole-genome scans and found non-random clustering of disease susceptibility loci for autoimmune diseases . Our results may suggest that all of the chronic inflammatory conditions we analyzed share similar pathogenetic background as reflected by peripheral gene expression.
Personalized medicine is becoming an integral part of healthcare and the key challenge is to establish a strategic focus on biomarker-based clinical tests. Non-invasive or minimally invasive diagnostic tests present less danger to a patient than invasive tests such as biopsies; and can help clinicians by providing them with valuable information in decision-making. A biomarker assay based on the gene panels described above might save time by reducing a list of preliminary disease impressions to a definitive diagnosis. The evaluation of these results and the identification of genes with altered expression or marker genes could also be potential targets for novel and more effective therapies and may lead to important insights into the pathogenesis of chronic inflammatory diseases. It might also have benefits to look at the identified gene panels in terms of SNPs in order to identify potential genetic changes associated with these medical conditions. Such novel and known biomarkers or a panel of such markers can play a major role in the development of personalized medicine.