Background

The human genome predominantly comprises of non-coding DNA [1]. Non-coding DNA, although not translated into proteins, plays a significant role in gene expression regulation during the transcription phase. The non-coding regions harbor untranslated regions (UTRs) that include 5’ UTRs and 3’ UTRs. 5’ UTRs are situated upstream of the protein-coding region, while 3’ UTRs are present downstream of protein-coding regions. UTRs, as we know, are not just passive bystanders in the process of gene expression. They are vital components for gene expression regulation and pre-mRNA processing, which affect protein outcomes [2, 3]. The 5’ UTRs control translation initiation by serving as an entry point for ribosome and interaction with TFs [4]. However, 3’UTRs are involved in the mRNA subcellular location, stability, and polyadenylation [5, 6].

Various studies in the literature have highlighted SNPs in UTR regions cause disease development, including cancer [7, 8]. UTR functions are commonly disrupted in oncological disorders [9, 10]. SNPs in the 3’ UTR region are reported to modify the cellular homeostasis balance toward cancer [11]. 5’ UTR mutations in BCL2 CXCL14 and TAOK2 genes caused abnormal gene translation [12]. Furthermore, the 5’ UTR mutation in the APC gene added a false start codon, which hindered the translation initiation process, leading to colon cancer development [13]. 5’ UTR (g.−130G > A) variant in the MATE2-K gene associated with poor metformin response [14]. PRKCZ is a serine-threonine kinase that belongs to the atypical group of protein kinase C [15]. PRKCZ is present on chromosome 1(1p36.33–p36.2) on the forward strand [16]. PRKCZ enzyme is polyfunctional, regulating cellular homeostasis and behavioral phenotype [17, 18]. PRKCZ is a regulator of tumorigenesis due to the downstream target of the P13K, and it also activates ERK/MAPK and NF-kb, JAK/STAT pathways [19,20,21]. PRKCZ expression dysregulations lead to poor prognosis and a lower survival rate, as reported in different cancers such as prostate cancer, breast cancer, glioblastoma, colon cancer, and ovarian cancer [21,22,23,24,25]. A splice variant in the 3’ region of PRKCZ was highly expressed in malignant prostate cancer cells and tissues [26]. rs436045 of PRKCZ was significantly associated with type 2 diabetes (T2DM ) pathogenesis [27]. A previous study found that PRKCZ promoter methylation was involved in T2DM development [28]. PRKCZ hypermethylation induced EMT in Head and neck squamous cell carcinoma [29].

The present study was mainly focused on untranslated region variants, i.e., 5’ UTR and 3’ UTR, ofPRKCZ. The current research aimed to explore the impact of PRKCZ 5’UTR and 3’UTR variants on transcription factor binding sites and the impact of TFs on gene regulation and miRNA-mediated interactions. The study also focused on studying the impact of 5’UTR and 3’UTR variants on the stability of mRNA secondary structure. eQTLs analysis was targeted to investigate the effect of UTR variants on the expression of PRKCZ in different tissues. Variation tolerance analysis was targeted to analyze the probability of variation in the gene with disease onset. This study provides preliminary in-silico based analyses that can be investigated further to understand the functional and clinical implications of PRKCZ UTR variants.

Methods

Data collection

The data relating to PRKCZ gene variants was downloaded from the Ensembl genome browser [30], COSMIC [31] and gnomAD [32]. The data contained variant IDs, allele variations and genomic coordinates. Ensembl genome browser includes a wide range of SNPs, both common and rare variants from various sources [30, 33], gnomAD focuses on genomic variants from whole-genome sequencing data [32], and COSMIC primarily contains somatic mutations from cancer genomes. It has lower UTR variants because it mainly focuses on coding region mutations [34]. The variant data obtained from the databases comprised two categories: coding regions and non-coding regions. As per the research objective, non-coding region variants encompassing 5’ UTR and 3’ UTR were included in the study and coding region variants were excluded. In order to process high-quality data, all the UTRs were screened out from the databases and thoroughly investigated to remove redundant data. The UTR SNPs were mapped on the PRKCZ genome sequence (GRCh38/hg38 genome assembly) according to the information retrieved from the databases.

Regulatory function analysis

The RegulomeDB database [35] was accessed to determine the functional impact of 5’ UTRs and 3’UTRs variants. This tool classifies the variants into six classes ranging from 1 to 7, and Classes are further subdivided into subcategories, namely 1a, 1b, 2a, 2b, 3a, and 3b and it also provides scores to UTR variants: with variants having a score of 1 or nearly equal to 1 have functional dysregulations and score less than 1 indicate lower functional dysregulation (Supplementary Material 1, Table S1).

Identification of transcription factor binding sites

The Alibaba tool was utilized to analyze the impact of UTR variants on transcription binding sites (TFBS) [36]. The tool combines TRANSFAC database data and generates pairwise alignment to find the TFBS. Alibaba tool takes the FASTA format as an input and provides TFBS as an output. It facilitates the identification of the number and type of transcription factors that bind to both the wild-type and mutant sequences. To elucidate the upstream and downstream proteins of PRKCZ that regulate diverse transcription factors in the nucleus, a thorough analysis was conducted through pathway mapping and gene annotations obtained from the Kyoto Encyclopaedia of Genes and Genomes (KEGG) database [37]. The pathway was visually represented using Microsoft PowerPoint.

microRNA annotation and structure prediction

To predict the effect of UTR variants on microRNA binding sites, PolymiRTS [38] was utilized for 3’ UTR variants, and the SNPinfo tool [39] was used for 5’ UTR variants. The PolymiRTS tool takes variant ID as an input and provides miRNA ID, functional class and context + score as an output. The database classified the effects of these variants into four functional categories: ‘D,’ ‘N,’ ‘C,’ and ‘O. The ‘D’ class indicates that the derived allele affects a conserved miRNA site, the ‘N’ class disrupts a non-conserved miRNA site, the ‘C’ class creates a new miRNA site, and the ‘O’ class provides no information on the ancestral allele. Additionally, the PolymiRTS Database provided insight into the Context + score change, where higher negative scores suggest a greater likelihood of disease development due to miRNA binding site disruption. Similarly, the SNPinfo tool also takes a variant ID as input and provides a transcription binding site if present, its genomic location and allele alteration as an output.

RNA secondary structure prediction

The study investigated the effect of both 5’ UTR, located upstream of the initiation site, and 3’ UTR, located downstream of the termination site of mRNA transcript, on the pre-processing, stability, and translation of mRNA structures. The RNAFold [40] web server was utilized to predict the secondary structure of mRNA and the minimum free energy value as a measure of the stability of the mRNA structure. RNA secondary is an important component in transcription and translation [6, 41]. RNAFold utilizes dynamic programming and McCaskill PF algorithms to predict mRNA secondary structure. It takes DNA sequence as an input to assess RNA secondary structure.

Expression quantitative trait loci (eQTLs) prediction

The GTEx portal [42] was employed to discover the single tissue eQTLs. eQTLs are genetic variants linked to various gene expression variations [43]. The GTEx portal takes variant ID as input and provides information on gene expression variations in different tissues, exclusively human tissues. The tool is an invaluable source for understanding the impact of non-coding variation correlated with other diseases, such as cancer, diabetes, stroke, etc.

Variation tolerance analysis

The Human Gene Database [44] was utilized to determine the genetic variation tolerance in the PRKCZ gene. The data provides a residual variation tolerance score and gene damage index (GDI). The residual variation score shows the gene’s susceptibility to genetic variation, while GDI characterizes the relationship between gene damage and the likelihood of disease.

Results

Data collection and processing

Data for both the 3’ and 5’ UTR variants was obtained from the Ensembl genome browser. The total number of 3’ and 5’ UTR variants obtained from the Ensembl genome browser were 576, including 139 5’ UTR variants (Supplementary Material S2, Table S2) and 223 3’ UTR variants (Supplementary Material S3, Table S3). Furthermore, 192 variants were downloaded from genomAD, consisting of 124 5’ UTR variants and 68 3’ UTR variants. The COSMIC database only contained 22 variants, all of which were 3’ UTR variants (Fig. 1A). The present study exclusively incorporated UTR variants that exhibited nucleotide changes, while those variants that lacked information related to nucleotide changes were excluded from the analysis (Fig. 1B).

Fig. 1
figure 1

PRKCZ corresponding UTR variants data retrieved from multiple sources, i.e., Ensembl genome browser, genomeAD, and Cosmic Panel (A) shows the number of 3’ and 5’ UTR variants obtained from databases, while (B) exhibits the UTR variants chosen for further analysis

Full size image

Regulatory functional analysis

The present study conducted a detailed regulatory analysis of both the 3’ and 5’ UTR variants through utilizing the RegulomeDB tool. The RegulomeDB tool provided information on 100 5’ UTR variants (Fig. 2A) (Supplementary material 4, Table S4) and 154 3’ UTR variants (Supplementary material S5, Table S5). Furthermore, variants with a probability score ranging from 0.60 to 1 were included for functional analysis, whereas those with a score lower than 0.60 were excluded. After applying the pre-determined inclusion and exclusion criteria, a total of 25 5’ UTR variants and 24 3’ UTR variants were identified to have a score above the 0.60 threshold. This information is depicted in Fig. 3A and B. Figure 3C provides a visual representation of the regulomeDB class rank to which the identified variants belong. Additionally, Fig. 3D displays the percentage distribution of the variants across the different rank classes.

Fig. 2
figure 2

Functional analysis of UTR variants by RegulomeDB. The data is segmented by score and rank for both 3’ UTR and 5’ UTR. (A) displays the score-wise category for 3’ UTR variants, while (B) shows the score-wise classification for 5’ UTR variants. A score of 1 or near 1 indicates a higher probability of affecting miRNA function. (C) demonstrates the rank-wise categorization for different classes for both 3’ UTRs and 5’ UTRs

Full size image
Fig. 3
figure 3

UTR variants included for further analysis based on score and rank. A) selected 5’ UTRs along with their score probability, (B) selected 3’ UTRs along with their score probability, (C) the identified variant’s class ranking, and (D) the percentage distribution of both 5’ and 3’ UTR variants across the rank classes

Full size image

Transcription factor binding site analysis

A comparative analysis of the 5’ UTR wild type and mutant sequences was conducted, which revealed that of the 25 variants analyzed, 11 variants caused the deletion of transcription factor binding sites, 8 variants led to the creation of a new site, whereas 6 variants caused no effect on TFBS ( Supplementary Material S6, Table S6) (Fig. 4A). However, 5’ UTR variants (rs1389053287, rs1282623960, rs192386882, and, rs57743955) led to the addition of entirely new YY1, repressor, Oct1, and YY1 mutant TFBS respectively (Fig. 4B). Based on the comparative analysis of 3’ UTR, it was found that 4 variants resulted in the deletion of TFBS, while 14 variants led to the addition of TFBS. Additionally, 6 variants did not affect the TFBS (Supplementary Material S77, Table S7) (Fig. 4C). 3’ UTR variants (rs1017752535, rs755735985, rs375170466, rs1006027187, and rs371058786) led to the addition of AP-2alpha, AhR, Da, GR, and USF mutant TFBS respectively (Figs. 4D and 5). TFs identified through the Alibaba tool were also analyzed to investigate their impact on PRKCZ expression. Moreover, KEGG pathway analysis revealed that different signaling pathways wereinvolved in PRKCZ gene expression. UTR genetic variants could lead to overexpression of certain TFs c-myc, c-Jun and fos, PDK and NF-KB complex, which ultimately could initiate positive feedback that promoted PRKCZ overexpression. These pathways encompass MAPK, WNT canonical and planner polarity pathway, NF-KB, and PI3K pathway. MAPK pathway was found to activate c-myc, c-Jun and fos, while NF-KB pathway promoted the transcription of NF-KB complex transcription. Similarly, the WNT canonical and planner polarity pathway also activated c-Jun, c-myc and PDK. All these TFs mentioned triggered PRKCZ translation and expression. Moreover, PRKCZ underwent activation within the cytoplasm through the AKT and PDK pathway via PI3K activation. Additionally, PRKCZ exerted an influence on the activity of ERK and RAC1 proteins. Consequently, PRKCZ influenced ERK and RAC1 to enhance the activity of C-myc, C-Jun, and Fos.

Fig. 4
figure 4

Comparative analysis of TFBS by Alibaba web server for wild type and variant. (A) illustrates the comparative analysis of 5’ UTR for wildtype and variant, (B) showcases the mutant TFBS induced by 5’ UTR variants (rs1389053287, rs1282623960, rs192386882, and, rs57743955) C) displays the comparative study of the 3’ UTRs for wildtype and variant and (D) illustrates the mutant TFBs induced by 3’ UTR variants (rs1017752535, rs755735985, rs375170466, rs1006027187, and rs371058786)

Full size image
Fig. 5
figure 5

PRKCZ signaling schematic representation constructed using the KEGG database. P13K, MAPK, NFKB WNT canonical and planner polarity pathway regulate PRKCZ transcription. Furthermore, PRKCZ interacts with ERK and CDC42/RAC1 to maintain positive feedback

Full size image

Analyzing the impact of UTR variants on miRNA binding sites

The PolymiRTS database contained valuable data regarding variants that affected miRNA binding. Seven 3’ UTR were observed to alter the miRNA binding sites, including its genomic location, ancestral allele, functional class, and Context + score (Table 1). Moreover, it was found that one SNP affected multiple miRNAs. SNPS rs80119748 altered the binding site of hsa-miR-3186-3p at conserved site, rs11553733 altered the binding site of various miRNAs; it created a new binding site of has-miR-4746-3p and hsa-miR-876-3p, and it altered the binding site of hsa-miR-140-5p hsa-miR-6499-3p at the conserved site, rs113534097 impacted the binding site of two miRNA such as hsa-miR-188-5p, hsa-miR-6866-3p by creating new binding sites. rs143864233 affected the binding sites of hsa-miR-1307-3p, hsa-miR-1469, hsa-miR-4638-5p at conserved sites. rs202100130 altered the binding site of hsa-miR-3943 at the conserved site, whereas it created new binding sites for following miRNA such as hsa-miR-4313, hsa-miR-3127-3p, hsa-miR-6769a-3p, hsa-miR-6756, 3p, rs1050601, hsa-miR-132-5p, hsa-miR-3148 d, hsa-miR-6124 d, hsa-miR-4497, hsa-miR-4664-3p, hsa-miR-4668-5p, hsa-miR-6784-5p. The SNPinfo database indicated that rs57743955 5’ UTR variant created TFBS. Among 5’UTR variants, it was found that solely one variant had a discernible impact on the TFBS (Table 2).

Table 1 List of PRKCZ 3’ UTR variants impacting 3’ UTR region interaction with miRNA binding sites along with functional class and context + score analyzed through PolymiRTS tool
Full size table
Table 2 PRKCZ 5’ UTR variants modulating the miRNA TFBS analyzed through SNPinfo
Full size table

Analyzing the impact of UTR variants on RNA secondary structure

Out of 25 5’ UTR, eight variants were found to alter the RNA secondary structure. rs1450239046 removed one stem-loop. Conversely, rs12822623960 removed two stem loops, rs1471967977 removed two stem loops and increased the positional entropy, rs1319211422 removed one stem-loop and created a larger stem-loop, and rs1292130789 removed three stem-loops. Variant rs192386882 also caused the removal of three stem loops and caused an increase in the positional entropy, rs145641717175 rearranged nucleotides in the stem-loop, enlarged it, and increased the positional, and rs1444012826 added one stem loop and decreased the positional entropy (Fig. 6A). Out of 24 3’ UTR, 6 variants either removed or added stem loops. The rs1038816956 removed a stem-loop leading to a decrease in positional entropy; the rs754595816 added one stem-loop, causing a larger stem-loop; the rs753475181 added two stem-loops, leading to high positional energy.the rs1006004882 and rs779822936 added one stem-loop, resulting in a decrease in positional entropy and the rs755735985 added a stem-loop, causingan increase in positional entropy (Fig. 6B).

Fig. 6
figure 6

PRKCZ RNA secondary prediction and positional entropy by RNAFold. (A) Shows RNA secondary structure for both wild type (W) and mutant (M) for 3’ UTR variants, while (B) Shows RNA secondary structure for both wild type (W) and mutant (M) for 5’ UTR variants. Positional entropy entropy is denoted with different colors; red depicts the highest, and blue represents the lowest positional entropy

Full size image

The non-coding variants exerted varied effects on the mRNA structural stability by either enhancing or reducing it or causingno effect. The results showed that among 5’ UTR, ten variants caused an increase, six caused a decrease, and nine variants had no effect on the mRNA stability (Fig. 7A) (Supplementary Material S8, Table S8). Moreover, 3’ UTR analysis showed that among all3’ UTR variants, seven variants led to an increase, six variants caused a decrease, and ten variants had no effect on the mRNA structure stability (Fig. 7B) (Supplementary Material S9, Table S9). Furthermore, a collective analysis of 5’ UTR and 3’ UTR variants effect on mRNA stability is depicted in Fig. 7C.

Fig. 7
figure 7

Minimum Free Energy calculation by RNAFold due to UTR variants (A) variation in MFE in 5’ UTR mutations, (B) variation in MFE in 3’ UTR, and (c) shows a collective analysis of the decrease, increase, and no change in mRNA stability upon 5’ UTR and 3’ UTR variations

Full size image

Analyzing the impact of UTR variants on expression by eQTLs

The GTEx portal revealed that only one variant, rs17777295, was found to be expressed in the lungs and thyroid with a p-value of 1.9e-116 and 1.5e-19, respectively (Table 3).

Table 3 PRKCZ UTR variant single tissue eQTL analysis by GTEx portal with the p-value
Full size table

Analyzing variation tolerance

The Human Gene Database revealed that PRKCZ had a high probability of intolerance, with scores of 8.25% and 1.63%, respectively (Table 4). In practical terms, this means that genetic variants could impactPRKCZ functionally that could increase the likelihood of disease onset. Specifically, the database indicated that as gene damage increases, the probability of the gene-to cause disease decreases.

Table 4 Genetic variation tolerance of PRKCZ along with residual variation intolerance score, Gene Damage Index Score and its impacts analyzed by the human Gene Database
Full size table

Discussion

Gene variations in UTR regions have been linked to various diseases. Literature has highlighted the involvement of UTR variants in several neurological and oncological disorders [45,46,47,48]. These regions are of extreme significance due to their essential function in regulating gene expression through the addition, deletion, or modification of transcription factors and transcription binding sites. However, the literature does not contain data concerning the role of PRKCZ UTR variants in disease onset. Experimental procedures can be inefficient in terms of both time and cost. Therefore, the present study utilized in-silico approaches to determine the effect of UTR variations in PRKCZ. The present study has examined the impact of PRKCZ UTR variants on interactions with transcription factors and miRNA. RNA secondary structure, eQTLs, gene network, and variation tolerance analysis were also done.

The study presented here filtered out PRKCZ UTR variants with a high probability of functional dysregulations. The results are consistent with the past studies that identified UTR variants in PRKCE, PKCI, and XRCC1 with functional implications [49,50,51]. Another study identified 97 non-coding SNPs involved in coronary artery disease development [52]. TFBS analysis revealed that 5’ UTR variants added YY1, repressor, and Oct1, whereas 3’ UTR variants added AP-2alpha, AhR, Da, GR, and USF binding sites. YY1 is a zinc finger transcription factor that regulates diverse functions such as proliferation, differentiation, and embryogenesis [53]. A previous study found that YY1 is closer to the translation start and had activator and repressor roles in gene regulation [54]. Repressor on the other hand, is a transcription factor that represses transcription by blocking the binding of different transcription factors to the promoter. The Oct1 transcription factor is pro-oncogenic and is involved in immune modulation, cytotoxic, and oxidative resistance, as well as metabolic reprogramming [55]. The role of AP-2alpha is significant in maintaining equilibrium between growth and differentiation. Its association with epidermal growth factor is direct, and studies showed that it causes hyperproliferation in epidermal cancers [56]. AhR transcription factors play a role both in immunity toxicity and T cell development [57]. Da has an attachment site downstream of the transcription site where polymerase binds [58]. Da binding to the site where polymerase normally attaches affects gene expression regulation [59]. GR TF binding to glucocorticoid response elements (GREs) controls many physiological processes, such as developmental, metabolic, and immune [60, 61]. The USF transcription factor family is involved in various cellular processes, for example, iron homeostasis, embryonic development, and carcinogenesis [62]. Our findings support the conclusions of earlier studies that explored the functional implication of varying TFBS due to non-coding variants [49, 50, 63].

The current study identified 5’ and 3’ UTR variants affecting miRNA binding. 5’ and 3’ UTR variants disturb miRNA binding sites, affecting the translation process and leading to various pathologies [64]. A previous study found that 3’ UTR SNP rs2229295 introduced binding sites for miR214-5p and miR550-5p, causing abnormal glucose metabolism [65]. MicroRNAs are essential components that play a crucial role in regulating the stability of mRNA [6]. The present study found that 3’ UTR caused variation in miRNA binding, which could affect the translation process. 3’ UTR SNP impact miRNA. In the past study, FOXC2 3’ UTR SNPs (rs201118690, rs6413505, and rs201914560) caused binding site variation in miRNA (impact on three miRNA, hsa-miR-6886-5p, hsa-miRS-6886-5p, and hsa-miR-6720-3p) [66]. 3’ UTR (rs12516, rs3092995 and rs8176318) in BRCA1 impacted interaction with miR-103 seed sequence [65]. Another study showed that UTR variants impacted the binding of miRNA with PRKCE mRNA, causing PRKCE overexpression [49]. These findings align with prior studies that analyzed the impact of the 3’ UTR variant on miRNA binding site and gene expression [45,46,47,48]. A previous study identified 110 3’ UTR SNPs in 67 genes that impacted miRNA binding sites [67]. Both 3’ and 5’ UTR variants impacted mRNA secondary structure and stability. It had been emphasized in the literature that mRNA stability is a crucial factor for mRNA regulation and processing [68]. An in-silico study reported that the 3’ UTR variant mutation was responsible for mRNA destabilization in the LIF gene, correlated to schizophrenia development [69]. The present study found that one variant was associated with tissue eQTLs. GTEx analysis had largely been employed in different studies to find the link between tissue eQTLs with coronary arteries, diabetes, and schizophrenia [70, 71]. According to the variation tolerance analysis conducted by the human gene database, it had been determined that the PRKCZ gene was intolerant, which put it at a higher risk of genetic variation. The GDI and residual intolerance score metrics are invaluable sources for determining the risk of genetic variation of a gene in the disease pathogenesis [72].

Conclusion

Our scientific investigation has employed in-silico approaches to determine the functional impact of UTR variants in PRKCZ. The present study has examined the involvement of TFs in regulating the expression of the PRKCZ gene. The PRKCZ UTR variants affected miRNA binding sites and mRNA stability. eQTLs analysis indicated that only one UTR variant showed expression in specific human tissues. Gene tolerance analysis revealed that PRKCZ is an intolerant gene that is prone to a high degree of genetic variation.

Limitations and Future Perspectives

It is imperative to validate the findings of the current study at the population level through genotyping following high-throughput sequencing techniques. Additionally, these discoveries can also be investigated in in-vitro and in-vivo models to determine the impact of variations in complex biological systems. Future studies may explore the potential correlation between identified TFs and PRKCZ by analyzing their co-expression.