Implications of LncRNAs and CircRNAs in psoriasis: a review

ABSTRACT Psoriasis is a chronic inflammatory skin disease characterized by skin infiltration of immune cells and abnormal epidermal thickening. The initial pathogenesis has not been fully elucidated. Non-coding RNAs (ncRNAs), which include long ncRNAs (lncRNAs) and circular RNAs (circRNAs), comprise the majority of genome transcripts and are important influencers of gene transcription and post-transcription modulations. Emerging roles of ncRNAs in psoriasis were identified recently. This review summarizes the existing studies of psoriasis-related lncRNAs and circRNAs. A considerable proportion of the studied lncRNAs and circRNAs regulate keratinocyte mobility, such as keratinocyte proliferation and differentiation. Some lncRNAs and circRNAs are tightly related to keratinocyte inflammation reactions. Other reports demonstrated that they are also implicated in modulating immune cell differentiation, proliferation, and activation. This review might illuminate future psoriasis research and highlight that lncRNAs and circRNAs might act as therapeutic targets.


Introduction
Current research progress confirmed that psoriasis is a systematic and complex disease and presently cannot be cured. Although the application of biological agents that target IL−17A, IL−23, or TNF-α significantly improved the treatment of psoriasis, they cannot completely eliminate the disease [1]. Therefore, there remains an urgent need to explore the initial pathogenesis of psoriasis. Genetic susceptibility is tightly related to psoriasis risks, particularly in the presence of the HLA-C × 06:02 risk allele [2,3]. Comprehensive assessment of psoriasis from a genetic perspective (including coding and non-coding genes) is necessary. Given that the genetic coding basis of psoriasis was reviewed elsewhere [4,5], the present review focused on advances in non-coding RNAs (long ncRNAs [lncRNAs] and circular RNAs [circRNAs]).
Non-coding RNAs (ncRNAs) are a class of RNAs without the capability for protein translation. However, they have diversified regulatory roles in many biological processes by affecting gene expression and cellular protein location and function. It was assumed that > 98% of the human genome is transcribed into ncRNAs, while only ~ 2% of RNAs encode proteins [6,7]. The ncRNAs provide an additional regulatory layer to balance the intracellular activity. The ncRNAs can be divided according to their length and structure: linear lncRNAs that are > 200-nucleotides long, microRNAs (miRNAs) that are ~ 22-nucleotides long, and circRNAs, a newly defined ncRNA characterized by a stable and covalently closed loop with no 5' and 3' polar ends. The ncRNAs constitute an important and complex epigenetic regulation [8], where recognizing ncRNA function gradually opens a new door to the study of complex diseases, including psoriasis, which is believed to be tightly related to epigenetic regulation [9,10]. Indeed, previous studies demonstrated that most genetic susceptibility loci identified for psoriasis fall into non-coding regions [11,12].
LncRNA expression is tissue-specific [13], which renders differentially expressed lncRNAs more likely to become disease biomarkers and potential therapeutic targets. For example, Lam et al. performed RNA sequencing (RNA-seq) analysis of biopsied skin and detected 4022 lncRNAs, among which more than one-third were novel and skin-specific. Notably, >40% of these novel lncRNAs were differentially expressed in psoriasis skin as compared with healthy skin [14].
The circRNAs are a class of evolutionarily conserved and endogenously expressed ncRNAs with tissue-specific expression patterns. Emerging research suggested that circRNA dysregulation is closely related to complex pathologies, such as cancer, neuronal disease, and autoimmune disease [15][16][17].
Although there are few reports on circRNAs in psoriasis research, current studies demonstrated a distinguished circRNA expression landscape of psoriasis lesions from both non-lesional and healthy skins [18,19].
Exploring lncRNA and circRNA regulation and function in psoriasis will facilitate in-depth understanding and the search for new biomarkers of this complex disease. The present review summarizes the current research progress on lncRNAs and circRNAs in psoriasis.

LncRnas and psoriasis
LncRNAs are > 200-nucleotides long and are incapable of coding for proteins [20]. Extensive transcriptome analysis demonstrated that the majority of the mammalian transcriptome is composed of lncRNAs, which were once considered genomic transcription noise. Nevertheless, lncRNAs are associated with diverse biological functions [21]. For example, lncRNAs function as cis or trans transcription regulators [22][23][24][25], mRNA processing modulators and mRNA activity regulators by sponging miRNAs [26,27], and protein locational and functional regulators [28,29]. Despite the elucidation of potential regulatory mechanisms, the biological relevance of the vast majority of lncRNAs and their specific effects in different diseases remain uncertain.
The pivotal role of lncRNAs in skin homoeostasis and pathogenesis has attracted considerable attention. For example, the methylation modification of the lncRNA Pvt1 was critically involved in maintaining epidermal progenitor cell stemness and contributed to epidermis self-renewal and wound healing [30]. Similarly, Li et al. identified a new lncRNA, WAKMAR1 (wound and keratinocyte migration-associated lncRNA 1), which regulated a network of protein-coding genes important for cell migration by interfering with E2F1 promoter methylation [31]. Recent studies identified numerous lncRNAs that are important in the pathology of the skin inflammatory response, for example, H19, MALAT1 (metastasis-associated lung adenocarcinoma transcript 1), NEAT1, and GAS5 [32].

LncRNA profiles in psoriasis
Gupta et al. determined that human skin contains > 6000 lncRNAs, of which approximately 1000 are differentially expressed in psoriasis [33]. Our lab also conducted RNA-seq analysis of psoriasis, where we identified a total read of 105,136 lncRNAs, among which 1940 lncRNAs were differentially expressed between psoriasis lesions and non-psoriasis skin, and 1559 lncRNAs were differentially expressed between psoriasis lesions and healthy skin. Notably, the non-psoriasis skins also demonstrated an significant signature of 1201 differentially expressed lncRNAs when compared with healthy skin. Besides, all three comparisons shared a considerable proportion of differentially expressed lncRNAs. This proportion of lncRNAs might consistently contribute to psoriasis development both in the pre-lesion formation stage and the subsequent lesion formation stage [34]. A study using weighted gene co-expression network analysis obtained 76 lncRNA -mRNA co-expression pairs in psoriasis. Among these pairs, lncRNA AL162231.4 and CCL27 mRNA are on the same chromosome with overlapping regions. According to the theory that a synonymous lncRNA regulates the corresponding mRNA, the author speculated that AL162231.4 might regulate CCL27 expression in psoriasis. That study proposed that identifying synonymous lncRNAs and mRNAs might be an effective approach to elucidate psoriasis pathogenesis [35].
Using an IL−22-induced HaCaT cell model, Qiao et al. determined that 1438 lncRNAs were differentially expressed, which indicated that lncRNAs are susceptible to inflammation signals in keratinocytes [36]. Furthermore, Luo et al. determined that lncRNA SPRR2C (small proline-rich protein 2C) acts as a hub gene with a critical effect on psoriasis pathogenesis by responding to IL−22 treatment. SPRR2C modulated the IL−22stimulated HaCaT cell phenotype through the miR−330-STAT1-S100A7 axis [37]. Ahn et al. conducted WGCNA on the RNA-seq results of skin samples from psoriasis patients (preand post-treatment with the TNF-α inhibitor adalimumab) and healthy controls and determined that > 50% of co-expressed genes in most of the psoriasis-and treatment-associated network modules were lncRNAs; therefore, the authors speculated that lncRNAs are critical in regulation pathways involved in psoriasis pathogenesis [38]. In Table 1, we summarize several important lncRNAs that are associated with psoriasis.

Psoriasis susceptibility lncRnas
In 2005, Sonkoly et al. identified a psoriasis susceptibility-related lncRNA termed PRINS (psoriasis susceptibility-related RNA gene induced by stress), whose expression was higher in the uninvolved epidermis of psoriatic patients compared with that in both psoriatic lesions and healthy epidermis [53]. Subsequent investigations revealed that the stress state induced PRINS expression in HaCaT cells cultivated in vitro, which is essential for keratinocyte survival under stress conditions [53]. Genomewide association studies identified TRAF3IP2 as a common susceptibility locus for psoriasis [54]. Interestingly, the antisense of TRAF3IP2, which is an lncRNA termed TRAF3IP2-AS1, is also vital in regulating psoriasis, especially in inhibiting the IL −17A-induced activation of NF-κB signalling and MAPK signalling by binding to SRSF10 and thereby blocking the subsequent recruitment of ACT1 [55]. A psoriasis-prone variant of TRAF3IP2-AS1, A4165G (rs13210247), was identified as a gainof-function mutant with enhanced binding affinity to SRSF10. Interestingly, lncRNA E130307A14-Rik was functionally identical to TRAF3IP2-AS1. E130307A14-Rik was overexpressed by delivering E130307A14-Rik and SRSF10 lentivirus into a mouse model of psoriasis, and significantly inhibited IL−17A-and TNF-α-induced pro-inflammatory effects. Therefore, TRAF3IP2-AS1 and SRSF10 May be therapeutic targets for psoriasis [55]. Rakhshan et al. reported that a single-nucleotide polymorphism (SNP) gene, rs12826786, which is a variant of HOX transcriptional antisense RNA (HOTAIR), was associated with psoriasis risk [56]. An important lncRNA, HOTAIR is extensively involved in affecting tumour cell proliferation and apoptosis, and metastasis [57], but its involvement in psoriasis has not been well documented. Another study genotyped three HOTAIR SNPs (rs12826786, rs1899663, rs4759314) by screening the loci of key candidate SNPs in HOTAIR. Two variant loci, rs12826786 and rs4759314, were eventually demonstrated as being associated with psoriasis risk [56]. Similarly, the antisense ncRNA in the INK4 locus, termed ANRIL, which is influenced by STAT1 signalling [58], is involved in autoimmune diseases, such as multiple sclerosis and inflammatory reactions [59]. Three ANRIL variants (rs1333048, rs4977574, rs10757278) were associated with psoriasis risk. For example, the rs10757278 G allele and the rs1333048 C allele were more prevalent in psoriasis cases. However, the rs4977574 A allele was protective against psoriasis [60]. Table 2 summarizes the variant psoriasis risk-associated lncRNAs.

LncRnas contribute to keratinocyte dysfunction in psoriasis
Keratinocytes are the main cell types that comprise the outmost layer of skin. As intrinsic immunity members, keratinocytes process antigens and present them to immune cells rapidly and non-specifically [64]. Keratinocyte dysregulation seriously contributes to psoriasis induction, development, and phenotype formation. Figure 1 depicts the lncRNAs that contribute to psoriatic keratinocyte dysfunction and the related mechanism.
After PRINS was associated with susceptibility to psoriasis, a subsequent study suggested a mechanism by which PRINS induces the onset of psoriatic lesions but does not maintain the disease. This led to some researchers wondering whether PRINS regulates coding proteins and thereby affects disease maintenance [53]. A subsequent study demonstrated that PRINS controls the expression of G1P3, an anti-apoptotic protein that contributes to the maintenance of keratinocyte hyperproliferation in psoriatic lesions [40].
A host gene for miR−31, lncRNA MIR31HG was first reported to promote cancer cell proliferation and invasion [65]. A study on psoriasis reported that MIR31HG was significantly upregulated in psoriasis lesions as compared with healthy skin. Knocking down MIR31HG in HaCaT cells Inhibits lps-induced maturation of DCs; suppresses T cell proliferation and promotes production of treg cells; and inhibits inflammatory responses by interacting with intranuclear NF-κB in lpsstimulated macrophages. [32,44,45] FABPSP3 Up Skin/Keratinocyte It maintains KMT2C overexpression through recruitment of human antigen R (HuR) thereby enhancing its downstream response to proliferation and inflammation. [46] AGAP2-AS1 Up Skin/keratinocyte Upregulation of AKT 3 by sponge miR−424-5 p activates the AKT/mTOR pathway, thereby promoting proliferation of keratinocytes. [47]
significantly decreased KRT6 and KRT16 expression and greatly reduced the percentage of S-phase cells while increasing the percentage of G2/M-phase cells. These results suggested the role of MIR31HG in psoriasis by promoting keratinocyte proliferation. Furthermore, the authors demonstrated that MIR31HG expression appeared to be NF-κB-dependent [41]. KLHDC7B-DT was positively regulated by interleukin enhancer binding factor 2 (ILF2), and both were overexpressed in psoriatic tissues and an M5 (IL−17A, IL−22, IL −1α, oncostatin M, TNF-α) induced keratinocyte model. The investigators determined that knocking down KLHDC7B-DT in keratinocytes prohibited the M5-induced proliferation and inflammatory cytokine secretion. By contrast, overexpressing KLHDC7B-DT in HaCaT cells led to cell hyperproliferation and apoptosis prohibition. Via RNA pull-down assay and mass spectrometry detection, the authors identified a mechanism through which KLHDC7B-DT could interact with ILF2 and activate the STAT3-JNK pathway [42].
LncRNA involvement in the epigenetic regulation of psoriatic keratinocytes has also received much attention. Recently, Huang et al. reported that lncRNA FABPSP3 contributed to the upregulation of lysine methyltransferase 2C (KMT2C) in psoriatic lesions by recruiting human antigen R (HuR) to maintain KMT2C mRNA stability. PIK3R3 overexpression promoted cell proliferation and cyclin D1 expression in psoriatic keratinocytes. The author reported that FABPSP3 promoted PIK3R3 transcription via KMT2C by regulating histone H3 lysine 4 trimethylation enrichment at the PIK3R3 promoter and histone 3 lysine 4 monomethylation at the PIK3R3\enhancer, thereby accelerating keratinocyte proliferation [46]. Furthermore, Xian et al. reported that downregulating methyltransferase-like 3 (METTL3) in psoriasis stabilized lncRNA AGAP2-AS1 by reducing its m 6 A modification; AGAP2-AS1 then functioned as a competitive endogenous RNA (ceRNA) by sponging miR −424-5p to upregulate AKT3 and activate the AKT -mTOR pathway, hence promoting keratinocyte proliferation [47].
A CRISPR interference (CRISPRi) screen of lncRNAs in human keratinocytes identified progenitor renewal-associated ncRNA (PRANCR) as an important regulator of epidermal homoeostasis. Depleting PRANCR significantly inhibited keratinocyte proliferation and resulted in the intrinsic loss of clone formation capability. In contrast, there was no difference in apoptosis induction after PRANCR had been depleted [49].
According to the rules of ceRNA regulation, the downregulated lncRNAs in psoriasis could not protect the target mRNAs from miRNA-mediated degradation. Therefore, most of them were related to the coding genes that were also downregulated in psoriasis. For example, lncRNA H19 protected S1PR3 from miR−766-3p-mediated degradation. In psoriatic skins, miR−766-3p overexpression was correlated with the downregulation of both lncRNA H19 and S1PR3 [63]. Using a keratinocyte model, He et al. reported that miR−766-3p activated the cell survival-related AKT -mTOR signalling; however, the signalling was inactivated by H19. The import of exogenous miR−766-3p counteracted the inhibitory effect of H19 on keratinocyte proliferation in a S1PR3dependent manner [51].
Encoded on chromosome 12, LINC00941 is a key regulator of human epidermal homoeostasis. LINC00941 is most highly expressed in undifferentiated progenitor keratinocytes and its abundance decreases significantly during cell terminal differentiation. The finding suggested an important role for LINC00941 in suppressing premature keratinocyte differentiation. Interestingly, LINC00941 and SPRR5 are inversely regulated in epidermal differentiation. Moreover, more than half of the LINC00941-induced genes were repressed by SPRR5. That study reported an interesting pair of lncRNAs with opposite roles in keratinocyte differentiation [50].

LncRnas regulate keratinocyte inflammation in psoriasis
Excessive keratinocyte proliferation is closely related to inflammation conditions, where a substantial proportion of lncRNAs that have been identified to regulate keratinocyte proliferation is also involved in regulating inflammation. Figure 2 summarizes their regulation mechanisms. Apart from affecting keratinocyte proliferation and differentiation, PRINS reduced IL−6, IL−8, and CCL−5 secretion from keratinocytes, but did not affect the expression of other inflammatory factors, such as IL−1α, IL−1β, and TNF-α [39]. Furthermore, PRINS demonstrated very high binding affinity to IL6 mRNA, and the specific interaction disrupted IL−6 translation and secretion in normal human epidermal keratinocytes (NHEKs) [39]. As IL−6 and IL−8 are actively involved in the inflammatory response in psoriasis [67], PRINS has a potential inhibitory effect on psoriasis inflammation.
Introduced in the previous section, ANRIL is induced by inflammatory stimuli such as TNF-α in an NF-κB-dependent manner. In contrast to PRINS, ANRIL activates IL−6 and IL −8 transcription through interaction with the transcription factor Yin Yang 1 (YY1), which functionally binds the IL−6 and IL−8 promoters upon activation [61].
As previously described, KLHDC7B-DT was significantly upregulated in psoriatic tissues, where it acts as an inflammation promoter. For example, knocking down KLHDC7B-DT greatly weakened the M5-induced inflammatory response in keratinocytes. Moreover, ablating KLHDC7B-DT significantly reversed the ILF2-mediated induction of IL−6 and IL−8 in keratinocytes [42].
The lncRNA FABPSP3 is an underlying cause of KMT2C overexpression in psoriasis by maintaining KMT2C stability. In addition to promoting keratinocyte proliferation, KMT2C is a well-documented inflammatory responder via the induction of PIK3R3, which subsequently mediates Akt -NF-κB pathway activation and results in the production of inflammatory mediators, such as IL−6, IL−8, CCL20, and S100A9 [46].
Conversely, H19, which is significantly downregulated in psoriasis epidermis, reduced the expression of inflammatory cytokines in IL−17A-, IL−22-and M5-induced keratinocyte models, while the addition of miR−766-3p to these cell models attenuated H19 regulation of inflammation [51].
Tang et al. reported that lncRNA MEG3 was downregulated in psoriatic skin and acted as a target molecule of TNFα. Similar downregulation of MEG3 was detected in both a TNF-α induced keratinocyte model and TNF-α induced mouse model. The author demonstrated that recovering MEG3 expression blocked TNF-α regulation on keratinocytes, which included enhancing inflammation and suppressing autophagy, likely via PI3K -AKT-mTOR signalling [68].

LncRNA regulates immune cell functions in psoriasis
Psoriasis was considered an immune-mediated inflammatory skin disease, where diverse immune cell types that infiltrate the skin tissue contribute to the inflammation cascade and subsequent phenotype formation. Currently, there are relatively few reports on the regulating roles of lncRNA in immune cell functions in psoriasis. However, when referring to another disease such as cancer, numerous studies indicated that lncRNAs were associated with the development and functions of various immune cell types. We believe that more efforts should be made to explore lncRNA regulation of psoriasis-associated immune cells.
A recent report using integrated analysis identified several immune-related lncRNAs as diagnostic biomarkers of psoriasis. For example, LINC01137, CCDC18-AS1, and CARMN were positively correlated with activated memory CD4 + T cells, myeloid dendritic cells (DC), neutrophils, M1 macrophages, and T follicular helper (Tfh) cells and were negatively correlated with T regulatory cells (Treg). LINC01215, MAPKAPK5-AS1, TPT1-AS1, EPB41L4A-AS, and LINC01214 were negatively correlated with activated memory CD4 + T cells, activated myeloid DC, neutrophils, M1 macrophages, and Tfh cells and were positively correlated with Treg [69]. In this section, we summarize several important immune-related lncRNAs, although some of them were not reported in relation to psoriasis, which might suggest implications for future psoriasis studies.
HOTAIR is an extensively explored lncRNA with a notable role in regulating immune cell functions. In lipopolysaccharide (LPS)-induced macrophages, HOTAIR critically mediated NF-κB signalling activation and the subsequent secretion of inflammatory cytokines such as TNF-α, IL−6, macrophage inflammatory protein−1B (MIP−1B), and inducible nitric oxide synthase (iNOS) [70]. Crucially, SNPs in the HOTAIR locus are associated with psoriasis risk [56]. These SNPs may influence susceptibility to psoriasis by regulating inflammatory factor levels in immune cells. However, there remains a lack of sufficient evidence for the association between HOTAIR and psoriasis.
On the contrary, MALAT1 was negatively associated with LPS-induced innate immune responses in human macrophages. Cells with small interfering RNA (siRNA)-mediated knockdown of MALAT1 had significantly increased TNF-α and IL−6 production in conjunction with LPS stimulation. Mechanistically, MALAT1 exerted an inhibitory effect on inflammation by interacting with NF-κB in the nucleus, thereby inhibiting its DNA binding activity [44]. Notably, MALAT1 mediated the inflammation progress in a highglucose environment by upregulating IL−6 and TNF-α expression in endothelial cells [71]. This implied that the anti-or pro-inflammatory effects of MALAT1 differ between cell lines and stimulations. Wu et al. used an LPS-stimulated DC model and reported that MALAT1 overexpression prevented IL−6, IL−12, and interferon (IFN)-γ induction, whereas it promoted the secretion of the anti-inflammatory cytokine IL−10. Furthermore, MALAT1-suppressed DC had markedly decreased expression of co-stimulatory molecules CD80, CD86, and MHCII, indicating that MALAT1 inhibited the maturation of LPS-induced DC. Moreover, T cells cocultured with LPS-treated MALAT1-overexpressing DC had significantly reduced proliferation capacity and increased percentages of Treg. Therefore, MALAT1 promoted the transition of LPS-stimulated DC to a tolerant phenotype, which led to effector T cell inactivation [45]. The DC-specific intercellular adhesion molecule 3 grasping non-integrin (DC-SIGN) is an intrinsic immune receptor expressed mainly by DC and macrophages and is involved in immunosuppressive maintenance and immune tolerance. MALAT1 overexpression in DC promoted DC-SIGN expression by functioning as a miR−155 sponge [72]. That study highlighted the role of MALAT1 in mediating DC immune suppression. Psoriasis tissues had increased MALAT1 levels [32]; nevertheless, the exact functions require clarification.
Analysis of data from the Immunological Genome Project (ImmGen) compendium of immunocyte gene expression determined that lncRNA FLICR (Foxp3 long intergenic ncRNA) was specifically expressed in Treg and co-occurred with Foxp3 during cell differentiation [73]. FLICR negatively regulated Foxp3, and this effect was particularly marked in IL −2 deficiency conditions. Analysis of transposase accessible chromatin (ATAC) high-throughput sequencing inferred that FLICR affected chromatin accessibility in the Foxp3 conserved non-coding sequence 3 (CNS3, which is important for Treg differentiation) accessible region 5 (AR5) through a local cis-inhibitory manner. Therefore, FLICR in cis targets the Foxp3 CNS3/AR5 region, thereby inhibiting Treg activity, and this effect is more pronounced in the absence of IL−2 [73]. Considering that Foxp3 stability is essential for maintaining Treg dynamic homoeostasis to prevent autoimmunity, the inhibitory effect of FLICR on Foxp3 May exacerbate psoriasis, which requires further elucidation.
The IL−17-derived T helper cells (Th17 cell) are important in the pathogenesis of autoimmune diseases, including psoriasis.
The peripheral Th cell-expressed lncRNA NONHSAT079547.2 was positively correlated with IL−17 expression in Hashimoto's thyroiditis. Further investigation suggested a mechanism by which NONHSAT079547.2 acts as a ceRNA to IL17 mRNA, where it competitively binds with miR−4716-5p, which could directly target the IL17 mRNA sequence [74].

LncRnas link psoriasis with cardiometabolic diseases
Psoriasis is considered a systemic disorder rather than a skinexclusive disease, where a particularly convincing amount of evidence indicated that psoriasis is associated with cardiovascular diseases [75,76]. Psoriasis-related systemic inflammation has been identified as an independent risk factor for cardiometabolic disease [77]. The genetic and molecular pathways shared by psoriasis and cardiometabolic diseases were well reviewed elsewhere [78,79]. However, existing reports on the shared lncRNAs between the two disease types remain highly limited. Herein, we summarize the shared lncRNAs studied in both psoriasis and cardiometabolic diseases.
The NF-κB-dependent ANRIL, which may serve as a psoriasis biomarker, is located in the INK4 locus, a welldefined genetic risk locus associated with coronary artery disease. Furthermore, it was suggested that the NF-κB-dependent ANRIL regulates inflammatory responses in coronary artery disease by activating IL−6/8 transcription. H19, which regulates IL−17A expression in psoriasis, is also dysregulated in cardiometabolic diseases. Specifically, H19 was enriched in cardiac and vascular tissue, and a multifunctional role of H19 in cardiometabolic diseases was suggested. H19 polymorphisms were associated with ischaemic stroke, and ischaemic stroke patients have higher levels of circulating H19 [80,81]. Furthermore, H19 regulated cardiomyocyte function and was affected in the progression of cardiac remodelling [82,83]. Moreover, H19 is widely involved in other cardiometabolic diseases, such as pulmonary hypertension, aneurysmal disease, haemorrhagic stroke, hypoxicischaemic encephalopathy, myocardial infarction, and coronary artery disease, which was well-reviewed by Busscher [84]. In addition to mediating pro-inflammation reactions of macrophages in psoriasis [70], HOTAIR also affects adipocyte differentiation and function, where it is specifically expressed in gluteal but not abdominal subcutaneous adipose tissue, and may be involved in determining the metabolic properties of gluteal compared with abdominal adipocytes. Therefore, HOTAIR might be associated with cardiometabolic risk by regulating peripheral body fat development [85]. In psoriasis, MEG3 might act as an antagonist to inflammatory signals [68]. However, MEG3 methylation modification might induce disorders, for example, the higher blood levels of MEG3 methylation in children due to prenatal lead exposure are more likely to generate rapid adiposity gain, which is a risk factor for childhood obesity and cardiometabolic diseases in adulthood [86]. Furthermore, MALAT1 negatively regulates the innate immune responses in psoriasis [44,45,72]. However, MALAT1 aggravated cardiomyocyte scorch death in diabetic cardiomyopathy by targeting NLRP3, a wellcharacterized inflammasome [87].
We believe that more efforts should be made to explore the dysregulated lncRNAs common to both psoriasis and cardiometabolic diseases, which may aid the clarification of the mechanistic association between the two disease types.

CircRnas and psoriasis
Unlike most linear mRNAs and lncRNAs with 5' N7methylguanosine (m 7 G) caps and 3' polyadenylated tails, circRNAs are covalently closed single-stranded RNAs (ssRNAs) also present in eukaryotes [88,89]. Although circRNAs were known of for many years, their roles were largely underestimated until the development of RNA-seq technologies and the further study of their molecular aspects [90]. In skin diseases, circRNAs were associated with melanoma occurrence, progression, and tumour malignant behaviours, which included metastasis and invasion [91,92]. RNAseq identified many differentially expressed circRNAs in psoriasis, and some circRNAs may be candidate biomarkers or involved in psoriasis pathogenesis [19,93]; several circRNAs may contribute to the development and pathophysiology of hypertrophic scars [94,95]. In this section, we review the current understanding of circRNA biology and circRNA functions in relation to psoriasis.

Characterization of circRnas
CircRNAs are generated by the back-splicing of precursor mRNA (pre-mRNA), which competes with canonical mRNA splicing to produce a linear RNA. CircRNAs can be divided into three main types based on biogenesis from different genomic regions: exonic circRNAs that consist of a single exon or multiple exons and account for most of the known circRNAs [96], intronic circRNAs that only contain introns and may rely on a consensus motif containing a specific base distribution [97], and exon-intron circRNAs, which contain sequences derived from both exons and introns ( Figure 1) [98].
CircRNAs are primarily derived from canonical splice sites, and their expression is determined by pre-mRNA transcription levels [99,100]. CircRNA biogenesis can be modulated by transcription factor activity and epigenetic modifications, such as methylation [101][102][103]. Additionally, circRNA processing events are determined by the regulation of circularization, which involves three main hypotheses: lariat-driven circularization, in which the looping structure of the intron sequences flanking the downstream splice donor site and upstream splice acceptor site brings these sites into close proximity; intron pairing-driven circularization, in which looping can be mediated by base pairing between inverted complementary repeat sequences (e.g. Alu elements); and RNA-binding protein (RBP)-mediated circularization, in which RBPs are dimerized and bind to specific motifs in the flanking introns to facilitate back-splicing ( Figure 3) [104,105].
Increasing evidence suggested that circRNAs modulate gene expression at the transcriptional level and/or posttranscriptional processing, regulate splicing, sponge miRNAs or proteins, and code peptides or proteins [88,89,106] ( Figure 4). CircRNAs binding to miRNAs can act as ceRNAs, thereby modulating the ability of miRNAs to target mRNAs [107]. Additionally, circRNAs can also bind to other RNA molecules, such as mRNAs, to regulate the stability and functions of these RNAs [108]. CircRNAs can bind to specific RBPs and thereby influence the functions of these RBPs [109]. Although most circRNAs are present in the cytoplasm [89,90,105], some can be found in the nucleus of human cells via generation from processed intron lariats or backsplicing with retained introns [89,90,104,105]. CircRNAs abundantly expressed in the nucleus are more likely to modulate transcription and splicing [110]. Many circRNAs have been validated to carry open reading frames (ORFs) and may encode peptides or proteins in an internal ribosome entry site (IRES) or N6-methyladenosine (m 6 A)-dependent manner [110][111][112][113]. In particular, circRNAs are abundant in exosomes, which means they are signal transmitters [114].

CircRNA roles in psoriasis
CircRNAs are key in regulating cell proliferation, the immune system, and the inflammatory response [89,115,116]. An increasing number of studies have proven that many circRNAs are abnormally expressed between psoriatic lesions and normal healthy skin tissues [18,19,[117][118][119], and might be involved in psoriasis occurrence and development through a variety of molecular mechanisms, which provide more evidence that circRNAs are potential novel therapeutic targets and biomarkers for psoriasis diagnosis, prognosis, and therapy.
Many attempts have been made to discover psoriasis biomarkers; however, no specific marker can accurately screen disease severity and treatment efficacy. CircRNAs can exist stably in cells and body fluids, such as serum and plasma, which indicates their potential as diagnostic biomarkers [120]. Several studies profiled circRNA expression in serum, peripheral blood mononuclear cells (PBMC), Treg, or lesional skin from psoriatic patients (Table 3).
These studies suggested that some circRNAs might be promising psoriasis screening biomarkers, while others suggested circRNAs could be used for psoriasis diagnosis. In the future, combining circRNA expression detection with clinical psoriasis diagnosis processing might improve psoriasis diagnostic accuracy as compared with traditional diagnosis approaches.
Treg are a subgroup of immunosuppressive CD4 + T cells that dominantly mediate immune tolerance and maintain immune homoeostasis [127]. In psoriasis, impaired and dysfunctional Treg result in the abnormal proliferation of effector T cells, which disturbs immune homoeostasis and was associated with psoriasis development. hsa_circ_0003738 was obviously stimulated in psoriatic Treg [122]. Importantly, knocking down hsa_circ_0003738 in psoriatic Treg restored their suppressive functions by inhibiting the secretion of the pro-inflammatory cytokines IL−17A and IFN-γ [122]. Moreover, hsa_circ_0003738 can harbour miR−562 to relieve the suppression of the miR−562 target gene IL−17RA, thereby promoting IL−17A signalling activation in psoriatic Treg. Moreover, hsa_circ_0003738 sponged miR−490-5p and released the inhibition of the target gene IFNGR2, and promoted IFN-γ signalling pathway activation in psoriatic Treg.
Yang et al. verified that hsa_circ_0004287 was enriched in PBMCs from psoriasis patients [128] and protected against psoriasis inflammation. In macrophages, hsa_circ_0004287 alleviated the stability of its host gene MALAT1 by competitively binding to growth factor 2 mRNA-binding protein 2 (IGF2BP2) in a m 6 A-mediated RNA modification-dependent manner [128]. IGF2BP3 regulated mRNA stability by serving as a m 6 A reader [129], and m 6 A sites were found within fulllength MALAT1 [130]. This finding suggested that circRNAs might modulate their host gene expression via RNA modification, which warrants further studies.
Finally, we hypothesize that certain circRNAs in various psoriasis-related functional cells may participate in psoriasis occurrence and progression by sponging numerous miRNAs and/or modulating host gene expression, which may present a new theoretical framework for additional research on circRNA functions in psoriasis. Furthermore, it was suggested that circRNAs might become the clinical biomarkers of psoriasis. However, this hypothesis is far from convincing. Therefore, more extensive and intensive research is needed to reveal the association between circRNAs and psoriasis, and the underlying molecular mechanism.

Conclusion
As a chronic recurrent skin disease, the complex mechanism of psoriasis is unclear, where it is a serious skin disease with multiple participating and influencing cells and factors. Genetic alteration may be a key factor among the initiating, maintaining, and relapsing elements of psoriasis [4,12]. Research on the non-coding genes, which comprise the majority of the human genome, will aid in-depth understanding of the background mechanism of various complex and incurable diseases, including psoriasis.
According to the lncRNA profiles of psoriasis reported by different laboratories, approximately one-fifth of lncRNAs are differentially expressed [14,34], which indicates that lncRNAs are profoundly involved in psoriasis pathogenesis. Current research of lncRNAs in psoriasis mainly focuses on their regulation of keratinocyte proliferation, differentiation, and inflammation, and the functions of immune cells, including DC, macrophages, Treg, and Th17 cells. Several lncRNAs might act as psoriasis biomarkers and/or therapy targets, such as the IL−22-responsive SPRR2C [37], psoriasis susceptibility gene PRINS [39,40], NF-κB-dependent ANRIL [61], IL −17A-promoting H19 [51], and the innate immune responserelated HOTAIR [70] and MALAT1 [44,45,71]. However, considering that the functions of most differentially expressed lncRNAs in psoriasis remain unknown, substantial efforts should be made to systematically screen the role of these differential lncRNAs in psoriasis in conjunction with the diversity of lncRNA regulation modes.
Recent studies also indicated a considerable number of reports on the existence of stable, functional small peptides (micropeptides) translated from ncRNAs [132,133]. These findings added another layer of lncRNA and circRNA regulating functions. We look forward to more psoriasis-related research reports that can reveal the innovative functions and mechanisms of lncRNAs and circRNAs, which may enable a more comprehensive and in-depth understanding of this complex skin disease.