Investigation of Microbial Translocation, TLR and VDR Gene Polymorphisms, and Recurrence Risk in Stage III Colorectal Cancer Patients

Simple Summary Gut microbial dysbiosis and microbial passage into the peripheral blood leads to colorectal cancer (CRC) and disease progression. Toll-like (TLR) and vitamin D (VDR) receptors play important role in the immune modulation and polymorphisms that may increase CRC risk. The aim of the current study was to demonstrate the prognostic value of microbial DNA fragments in the blood of stage III CRC patients and correlate such microbial detection to TLR/VDR polymorphisms. TLR/VDR polymorphisms and presence of microbial DNA in CRC patients highlight their role in cancer development and progression. Abstract Gut microbial dysbiosis and microbial passage into the peripheral blood leads to colorectal cancer (CRC) and disease progression. Toll-like (TLR) and vitamin D (VDR) receptors play important role in the immune modulation and polymorphisms that may increase CRC risk and death rates. The aim of the current study was to demonstrate the prognostic value of microbial DNA fragments in the blood of stage III CRC patients and correlate such microbial detection to TLR/VDR polymorphisms. Peripheral blood was collected from 132 patients for the detection of microbial DNA fragments, and TLR/VDR gene polymorphisms. In the detection of various microbial DNA fragments, TLR and VDR polymorphisms was significantly higher compared to healthy group. Homozygous individuals of either TLR or VDR polymorphisms had significantly higher detection rates of microbial DNA fragments. Mutational and MSI status were significantly correlated with TLR9 and VDR polymorphisms. Significantly shorter disease-free survival was associated with patients with BRAF mutated tumors and ApaI polymorphisms, whereas shorter overall survival was associated with the detection of C. albicans. The detection of B. fragilis, as demonstrated by the multivariate analysis, is an independent poor prognostic factor for shorter disease-free survival. TLR/VDR genetic variants were significantly correlated with the detection of microbial fragments in the blood, and this in turn is significantly associated with tumorigenesis and disease progression.


Introduction
Although screening and new treatment strategies have been established for colorectal cancer (CRC), it remains a major health issue [1]. CRC is the third and second most common Since August 2018 to June 2022, a total of 132 consecutive patients from the Department of Medical Oncology, University Hospital of Heraklion, aged >18 years old, with newly diagnosed and with histologically documented stage III CRC were enrolled in the study. None of the patients had history of other malignancy. Moreover, a total of 100 healthy individuals, aged >18 years old were also enrolled in the study.

Blood and Primary Tissue Samples
Peripheral blood (5 mL in EDTA) was collected from all patients and control subjects enrolled in the study, and the QIAamp DNA Blood Mini Kit (QIAGEN, Hilden, Germany) was used for DNA extraction. Concerning the primary tissues, for microdissection of representative formalin fixed paraffin embedded (FFPE) specimens, a piezoelectric microdissector (Eppendorf, Hamburg, Germany) was used to collect cancer cells [22], following evaluation of the appropriate area by an experienced pathologist. DNA extraction was performed using both the MasterPure ™ Complete DNA and RNA Purification Kit (Epicenter, Madison, WI, USA) following the manufacturer's instructions. NanoDrop ND-1000 v3.3 (Thermo Fisher Scientific, Wilmington, DE, USA) was used for DNA quantification.

Microbial DNA Amplification
For the microbial DNA amplification, each genes' target reagents and PCR conditions used have already been covered in detail by our team [6,23]. In brief, a set four primer pairs were used for the detection of bacterial genomic DNA encoding 16S rRNA, glutamine synthase of Bacteroides fragilis, β-galactosidase gene of Escherichia coli, and 5.8S rRNA found in Candida albicans. The samples' DNA integrity was checked using the reference gene human glyceraldehyde phosphor-dehydrogenase (GAPDH). In order to identify bacterial DNA in blood samples, 16S rRNA was employed as a reference.

Toll-Like Receptor (TLR) and Vitamin D Receptor (VDR) Genotyping
For genotyping of TLR genetic variants, polymerase chain reaction and restriction fragment length polymorphism methodologies (PCR-RFLP) were performed. In brief, TLR2 196-to-174 Ins/Del genetic variants were determined by PCR, whereas TLR4 (Asp299Gly and Thr399Ile) and TLR9 (T1237 and T1486C) genetic variants were determined by PCR-RFLP. All materials and conditions for each gene target involved in the current study have already been described previously by others and our group [20,21,24].
For the genotyping of VDR genetic variants, as in the case of TLRs, PCR-RFLP methodology was performed. Each genes' target reagents and PCR conditions used have already been covered in detail by our team [20,21].
In brief, for all single nucleotide polymorphisms (SNPs) of the TLR gene or VDR gene at the TaqI, ApaI, FokI, and BsmI positions, the patients were characterized as wild type, heterozygous, or homozygous in the absence of the restriction site in both alleles, the presence of the restriction site in one allele, and presence of the restriction site in both alleles, respectively.

Mutational Analysis
KRAS, NRAS, and BRAF mutational analyses were performed by Sanger sequencing, following amplification of KRAS exon 2, 3, and 4; NRAS exon 2, 3, and 4; and BRAF exon 15. Moreover, microsatellite instability (MSI) status was also evaluated. All materials and conditions for each gene target involved in the current study have already been described previously by our group [25][26][27][28].

Patients and Healthy Donors Characteristics
In total, 132 stage III CRC patients and 100 healthy donors have been enrolled in this single-centered study. The median age of the patients and healthy subjects enrolled was 62 years (range: 36-83 years) and 66 years (range: 35-89 years), respectively. Most of them were males (patients: 59.1%; healthy donors: 54%) and <70 years old (patients: 68.9%; healthy donors: 70%). Most of the patients had a good performance status (PS) (99.2%) and had left sided tumors (75.8%), mainly on sigmoid (50.7%). All patients were diagnosed with adenocarcinoma, and mucinous features were observed in 18 (13.9%) patients, who received CAPOX (capecitabine + oxaliplatin) as an adjuvant treatment (55.7%). The patients' characteristics and demographics are demonstrated in Figure 1 and Tables 1 and S1. were males (patients: 59.1%; healthy donors: 54%) and <70 years old (patients: 68.9%; healthy donors: 70%). Most of the patients had a good performance status (PS) (99.2%) and had left sided tumors (75.8%), mainly on sigmoid (50.7%). All patients were diagnosed with adenocarcinoma, and mucinous features were observed in 18 (13.9%) patients, who received CAPOX (capecitabine + oxaliplatin) as an adjuvant treatment (55.7%). The patients' characteristics and demographics are demonstrated in Figure 1 and Tables 1 and S1.

TLR and VDR Genetic Variants Analysis and Clinical Outcoume
All CRC patients and healthy subjects were also evaluated for the presence of genetic variants in both VDR and TLR genes. For the case of VDR genetic variants, 20 (15.2%), 49 (37.1%), 20 (15.2%), and 17 (12.9%) patients presented the homozygous mutant genotype for TaqI, ApaI, FokI, and BsmI polymorphisms, respectively (Tables 3 and S1). A significant difference was demonstrated for all VDR gene polymorphisms between CRC patients and healthy donors (p < 0.001) (Tables 3 and S1).

Univariate and Multivariate Analysis for Cox Regression Analysis
According to the Cox regression univariate analysis, BRAF V600E mutations, histology (adenocarcinoma vs. mucinous), the detection of microbial DNA encoding for glutamine synthase of B. fragilis, and the detection of ApaI mutant alleles of the VDR gene are significantly associated with a shorter disease-free survival (DFS); and the detection of microbial DNA encoding for 5.8S rRNA is significantly associated with shorter overall survival (OS) ( Table 6). Based on the Cox regression multivariate analysis, adjusting for above mentioned factors, B. fragilis is a significant independent factor linked to shorter OS (HR: 33.85, 95% CI: 1.8-622.4, p = 0.018) ( Table 6).

Discussion
Different pathways have been under research to understand tumorigenesis and develop new treatments against CRC. TLRs, microbiota, and VDRs are some of the areas that are under evaluation, involved in pathways that modulate immunity against cancer. In the current research, we aimed to evaluate the existence of TLRs and VDRs polymorphisms and of microbial translocation and their correlation with prognosis, in stage III CRC patients.
The passage of intestinal microorganisms into the bloodstream (also known as microbial translocation) is a phenomenon mainly met because of microbial dysbiosis [23]. This disturbance in microbial composition has been proven as one of the ways of tumorigenesis and tumor growth in CRC [6,29]. The microbiota that has escaped to the blood stream is detected mainly by its fragments or its products [6]. As our group has previously demonstrated, 16SrRNA, E. coli, B. fragilis, and 5.8S rRNA microbial fragments detected are correlated with tumorigenesis and progression and may have a prognostic role in CRC patients [6]. This study confirms the previous results, focusing on stage III CRC patients. Numerically, all DNA fragments were more frequently detected in CRC patients than in the healthy donors and this detection was statistically significant, except in the case of the detection of β-galactosidase gene of E. coli.
TLRs are a part of innate immunity, contributing mainly to the recognition of external factors that could be pathogens [10]. MyD88-dependent pathway plays a crucial role in the immune reaction and CRC related to inflammation [7,10]. Previous studies have shown that different polymorphisms have been detected in CRC patients. TLR 3/4/7/8/9 and their polymorphisms seem to have a prognostic role for CRC [20,30]. Specifically, high expression of TLR4-mediated MyD88 signaling has been correlated with poor prognosis, even in stroma and CAFs related to CRC [31,32]. Also, TLR4 along with TLR2 and TLR3 may have a prognostic role for CRC through regulation of NFκB pathway, leading to tumorigenesis [33]. TLR7 and TLR8 expressed by CD133 + cells have been linked to worse prognosis in CRC patients [34]. On the other hand, there are some controversial results concerning TLRs and their prognosis in various tumor types [35,36]. Specifically for CRC, high expression of TLR5 in tissue seems to be linked with better prognosis [37]. On a previous study of our group, it was demonstrated that higher frequencies of TLR2, TLR4, and TLR9 polymorphisms in CRC patients, in comparison to healthy individuals, are correlated with worst prognosis [20]. TLR2-196 to-174 del/del genotype, TLR4 Asp299Gly, TLR4 Thr399Ile, TLR9 T1237C, and TLR9 T1486C homozygous genotypes were all detected in statistically significant higher levels in the disease setting and were also correlated with worst prognosis [20]. All the above results are in accordance with the results of the current study; that is, higher rates of all TLR genetic variants detected in CRC patients compared to health individuals . Particularly for the homozygous mutant genotype for TLR4-D299G,  TLR4-T399I, TLR9-T1237C, TLR9-T1486C, and TLR2-196 to-174 bp the detection was up to 56.8% in CRC patients, while none of healthy donors were detected with any of the variations. Nevertheless, no statistically significant prognostic value was demonstrated in stage III CRC patients.
The multiple roles vitamin D plays in carcinogenesis, protection, or therapy on CRC have been being researched over the last handful of decades [19]. Vitamin D fulfills its role by binding to its receptors, so VDR polymorphisms could affect the signaling pathway vitamin D activates [38]. More than 60 single nucleotide polymorphisms (SNPs) of VDR gene have been referred to in previous studies as related to carcinogenesis and prognosis at different tumor types [39]. The SNPs are found mainly in the promoter region in exons 2-9 and in the 3 -UTR (3 -untranslated regions) of the gene [40]. However, only some of them are directly related and functionally important in CRC. These include TaqI (rs731236; Thermus aquaticus I), ApaI (rs7975232; Acetobacter pasteurianus sub. pasteurianus I), BsmI (rs1544410, Bacillus stearothermophilus I), and FokI (rs2228570; Flavobacterium okeanokoites I), located in exon 9, in the intron between exons 8 and 9 and in exon 2, respectively [41][42][43][44][45]. Previous studies have highlighted the role of VDRs in immune modulation mainly through regulation of gut microbiota and microbial translocation. VDR conditional knockout (VDR ∆IEC ) in the epithelium of colon or low intestinal VDR protein levels may lead to microbial translocation [46]. Reduction of JAK/STAT (Janus kinases/signal transducer and activator of transcription proteins) signaling is another form of interference with gut microbiota and inflammatory responses [47]. Also, vitamin D has a promoting role for TLRs through binding to VDRs, leading to activation of innate immunity and modulation of gut microorganisms [48,49]. Different studies have tried to evaluate the role of VDR polymorphisms in CRC, with many controversial results, possibly because of the investigation of different populations [50][51][52]. Nevertheless, meta-analyses and reviews have shown a significantly higher level of detection of the homozygous mutant genotypes of TaqI and ApaI genotypes in CRC patients [53], whereas CRC tumorigenesis was correlated strongly with BsmI, FokI, and TaqI polymorphisms [54]. We have previously showed that the homozygous mutant genotypes of all TaqI, ApaI, BsmI, and FokI are significantly more frequent in CRC patients of all stages comparing to healthy donors [21]. However, the detection was in lower levels in early stages [21]. Herein, we validated the results above. Although the homozygosity of the polymorphisms were not so frequent, even in CRC patients, all evaluated polymorphisms were significantly more frequently detected in stage III CRC patients compared to the healthy population. This strongly suggests that VDR polymorphisms can contribute to CRC tumorigenesis. Regarding their correlation with TLR polymorphisms, almost all VDR and TLR polymorphisms were significantly correlated, promoting the hypothesis of their synergy. Correlating the detection of microbial fragments with VDR and TLR polymorphisms in stage III CRC patients, not all detected fragments were significantly correlated almost every polymorphism, and this is in contrast to our previous study. However, the correlation remains high, mainly for VDR polymorphisms and TLR4-D299G. All these results strengthen the hypothesis that VDR and TLR polymorphisms act together for the remodeling of gut microbiota and dysbiosis, disturbing the homeostasis and leading to immune modulation.
Finally, taking for granted the prognostic significance of tumor mutational status, KRAS and BRAF mutations were correlated with TLR9-T1237C, TLR9-T1486C polymorphisms, and TLR9-T1486C, respectively. MSI high status was correlated mainly with TLR9-T1237C and all VDR polymorphisms. Having in mind that TLR9 participates in CRC tumorigenesis through inflammation, the results can be biologically reasonable [55][56][57].
To our knowledge, our research remains the only effort to investigate the detection and possible significance for CRC tumorigenesis and progression of microbial translocation, TLR and VDR polymorphisms, and their correlation in stage III CRC patients. This research is prospective, includes a homogeneous population, is well-distributed, and took into consideration important pathological features. However, limitations do exist. The sample is relatively small and the total time of follow up needs to be extended for safer prognostic results. Despite the limitations, TLR and VDR polymorphisms, as well as microbial translocation, seem to maintain an important role to CRC tumorigenesis and progression in stage III CRC patients and could modulate immune reaction to CRC. On that basis, more research remains to be done to validate these results and drive to new, multi-targeted preventive and therapeutic options.

Conclusions
In conclusion, our study remains the first attempt to evaluate the detection and possible prognostic significance of TLR and VDR polymorphisms and microbial translocation in the Greek population in stage III CRC patients. All polymorphisms were significantly more frequently detected in CRC patients than in healthy donors, and the same was observed for microbial fragments with the exception of the β-galactosidase gene of E. coli. Also, these parameters seem to correlate each other, empowering the hypothesis that immune modulation against CRC is a complex axis with possible multiple preventive and therapeutical targets.

Institutional Review Board Statement:
The study has been approved by the Ethics Committee/Institutional Review Board of the University Hospital of Heraklion (Number 7302 /19-8-2009). All the procedures performed were in accordance with the ethical standards of the institutional and/or national research committee and the 1964 Helsinki declaration, and its later amendments or comparable ethical standards.

Informed Consent Statement:
All patients signed a written informed consent form for their participation.
Data Availability Statement: All relevant data are within the paper and its Supporting Information files.

Conflicts of Interest:
The authors declare no conflict of interest.