Effect of Quercetin on ABCC6 Transporter: Implication in HepG2 Migration

Quercetin is a member of the flavonoid group of compounds, which is abundantly present in various dietary sources. It has excellent antioxidant properties and anti-inflammatory activity and is very effective as an anti-cancer agent against various types of tumors, both in vivo and in vitro. Quercetin has been also reported to modulate the activity of some members of the multidrug-resistance transporters family, such as P-gp, ABCC1, ABCC2, and ABCG2, and the activity of ecto-5′-nucleotidase (NT5E/CD73), a key regulator in some tumor processes such as invasion, migration, and metastasis. In this study, we investigated the effect of Quercetin on ABCC6 expression in HepG2 cells. ABCC6 is a member of the superfamily of ATP-binding cassette (ABC) transporters, poorly involved in drug resistance, whose mutations cause pseudoxanthoma elasticum, an inherited disease characterized by ectopic calcification of soft connective tissues. Recently, it has been reported that ABCC6 contributes to cytoskeleton rearrangements and HepG2 cell motility through purinergic signaling. Gene and protein expression were evaluated by quantitative Reverse-Transcription PCR (RT-qPCR) and western blot, respectively. Actin cytoskeleton dynamics was evaluated by laser confocal microscopy using fluorophore-conjugated phalloidin. Cell motility was analyzed by an in vitro wound-healing migration assay. We propose that ABCC6 expression may be controlled by the AKT pathway as part of an adaptative response to oxidative stress, which can be mitigated by the use of Quercetin-like flavonoids.


Introduction
Colorectal cancer (CRC) is the third most common cause of cancer-related death worldwide [1]. This high mortality may be explained by the relatively high incidence and high rates of local regional and metastatic disease in CRC. The incidence is increasing in the context of organised screening programmes and an ageing population. The preponderance of advanced and metastatic disease is still a reality despite research advances in treatment. Quality of life is a vital aspect of our care for patients with CRC, and individualised treatments may help to limit side effects. Since prognostic biomarkers may allow for more tailored treatments for our patients, circulating tumour cells (CTCs) have potential to be helpful in this context.
The aim of this systematic review is to examine how CTCs in patients with CRC can be used as prognostic biomarkers. We will summarise the progress reported in the literature so far, and will discuss future research.

Circulating Tumour Cells (CTCs): Isolation Enrichment and Detection
CTCs are epithelial cancer cells from the primary tumour or metastases that gained access to the circulatory system, and are detectable in sampled peripheral blood. They are believed to be directly involved in the biology of the metastatic process [2]. From this perspective, CTCs may then be defined as surrogate tumour material. Different techniques have been developed to isolate CTCs from the bloodstream, based on biophysical methods (deformability, size, density, and surface charge), and/or immunoaffinity status.
Once CTCs are isolated, they are available for multiple analyses such as genetic, epigenetic, transcriptomic, proteomic, and identification of surface cell markers or living cell properties. All these analyses have the potential to be used in the clinical setting for either diagnosis, prognosis, prediction of recurrences, or to adapt therapeutics for different types of solid cancers [4].

Review Methodology
We searched for published studies that reported the use of CTCs as a prognostic marker for CRC patients. Our search was performed with PubMed from 1975 to 1 February 2021. As search terms, we used "colorectal cancer (CRC)", "circulating tumour cells (CTC)", and "prognosis biomarker". All articles in English or French reporting information about CTCs, CRC and prognosis were included in our initial search. We excluded all articles where the study population was not patients with CRC. Case reports were excluded. Any articles about CTCs that did not integrate prognostic factors were also excluded.
Articles were eligible for inclusion if they reported either quantitative data (presence/absence or number of CTCs) or qualitative data (specific CTC elements).
For the definition of "prognosis", we included articles that discussed overall survival regardless of the treatment, and those that discussed the sensitivity or resistance to a specific treatment. We were also interested in discussions regarding treatment modulation, such as adding or cancelling treatment, drug intensification or de-escalation, possibility of targeted therapy, alternative treatment including early change or interruption of active care (Figure 1). These were of interest because they facilitate precision medicine that enables us to give the best care with the least side effect as possible.

Results
We found 399 articles in our initial search. After excluding reports (Table 1) not meeting the selection criteria (Table 1), there were 65 relevant studies ( Figure 2). We classified studies according to type of marker used (quantitative, qualitative or both), techniques

Results
We found 399 articles in our initial search. After excluding reports (Table 1) not meeting the selection criteria (Table 1), there were 65 relevant studies ( Figure 2). We classified studies according to type of marker used (quantitative, qualitative or both), techniques (Table 2), and their CTC detection time points (Figure 3). They are summarised in Tables 3 and 4 and ??.

Types of Markers
The majority of articles use CTCs as a quantitative prognostic biomarker (Table 2). This is likely to be due to the only FDA-approved method (CellSearch) being a CTC count method. CellSearch, an EpCAM based technique, was used in 14 of the 77 included studies.
EpCAM is part of the cell adhesion molecule (CAM) family. It was first identified in colon cancer in 1979 [70]. Epithelial Cell Adhesion Molecule (EpCAM) is expressed by normal epithelial cells and is highly expressed in a majority (70%) of primary tumour.
Concerning colon adenocarcinomas, EpCAM is overexpressed in 81% [71] and it is known to promote tumour expansion. EpCAM is involved in numerous independent pathways [72]. It plays a part in the activation of the Wnt signalling pathway that is known for its role in carcinogenesis. On one hand, EpCAM activates this pathway through its interaction and stabilization of Lrp6. On the other hand, it downregulates the expression of some this pathway inhibitors (fgf3). EpCAM is also known to promote oncogenesis through its interaction with β -catenin, by activating cell proliferation via proto-oncogenes, such as c-Myc or Cyclin A [73].
Thus, EpCAM is often use to characterise CTCs. However, during metastatic spread, epithelial-mesenchymal transition (EMT) occurs and tumour cells lose their epithelial marker, one of them being EpCAM. This cancer expansion modality highlights a clear limit of anti-EpCAM based detection system [74].
In addition to this loss of specific markers, CTCs are also difficult to separate because of their low concentration in the blood. To tackle these difficulties, newer strategies have been proposed, such as the using blood samples close to the tumour, such as mesenteric and portal vein blood samples, and detection methods such as specific gene reverse transcriptase quantitative polymerase chain reaction (RT-qPCR). This quantitative RNAbased method for CTC detection is more sensitive but has limitations. Target genes need to be carefully selected as they can be weakly expressed in normal blood cells leading to false-positives. Moreover if chosen target genes are expressed differently from one CTC to another, this may lead to an inaccurate tumour cell count [75].
After total mononuclear cell RNA extraction, RT-PCR is conducted on mRNA that are supposed to be specific to CTCs. More and more genes are being explored, focusing on different CTCs characteristics, such as epithelial (CEA, CK20, and CK19) and stem cells-like potential.

Types of Markers
The majority of articles use CTCs as a quantitative prognostic biomarker (Table 2). This is likely to be due to the only FDA-approved method (CellSearch) being a CTC count method. CellSearch, an EpCAM based technique, was used in 14 of the 77 included studies.
EpCAM is part of the cell adhesion molecule (CAM) family. It was first identified in colon cancer in 1979 [70]. Epithelial Cell Adhesion Molecule (EpCAM) is expressed by normal epithelial cells and is highly expressed in a majority (70%) of primary tumour.
Concerning colon adenocarcinomas, EpCAM is overexpressed in 81% [71] and it is known to promote tumour expansion. EpCAM is involved in numerous independent pathways [72]. It plays a part in the activation of the Wnt signalling pathway that is known for its role in carcinogenesis. On one hand, EpCAM activates this pathway through its interaction and stabilization of Lrp6. On the other hand, it downregulates the expression of some this pathway inhibitors (fgf3). EpCAM is also known to promote oncogenesis through its interaction with -catenin, by activating cell proliferation via proto-oncogenes, such as c-Myc or Cyclin A [73].
Thus, EpCAM is often use to characterise CTCs. However, during metastatic spread, epithelial-mesenchymal transition (EMT) occurs and tumour cells lose their epithelial marker, one of them being EpCAM. This cancer expansion modality highlights a clear limit of anti-EpCAM based detection system [74].
In addition to this loss of specific markers, CTCs are also difficult to separate because of their low concentration in the blood. To tackle these difficulties, newer strategies have been proposed, such as the using blood samples close to the tumour, such as mesenteric and portal vein blood samples, and detection methods such as specific gene reverse transcriptase quantitative polymerase chain reaction (RT-qPCR). This quantitative RNAbased method for CTC detection is more sensitive but has limitations. Target genes need to be carefully selected as they can be weakly expressed in normal blood cells leading to false-positives. Moreover if chosen target genes are expressed differently from one CTC to another, this may lead to an inaccurate tumour cell count [75].

Prognostic Thresholds
For studies that utilised the CellSearch system, there were threshold changes according to CRC stage. For metastatic CRC, CTC count above or equal to 3 CTC per 7.5 mL of blood was used by Matsusaka et al., 2011 [5], Aggarwal et al., 2013 [7], and Sastre et al., 2013 [9] as a positive marker to determine "high CTC" patients. These studies all reported a significant correlation between baseline "high CTC" status and reduced survival. Camera et al., 2020 [17], similarly reported a prognostic difference in the presence of disease refractory to standard treatment or unresectable CRC. Coumans et al., 2012 [6], reported that survival for patients with metastatic CRC patients with CTCs is reduced by 6.6 months for each 10-fold CTC increase. Krebs et al., 2015 [11] and Aranda et al., 2020 [16] showed benefits of escalating therapy from a three-drug regimen to a four-drug chemotherapy regimen for patients above their CTC prognostic threshold.
For patients undergoing surgery, CTC count above or equal to 1 CTC per 7.5 mL of blood was used by Bork et al., 2015 [10] and van Dalum et al., 2015 [13] to determine survival. These finding suggest a reduced surgical utility for patients above the prognostic threshold.

Techniques
There were multiple quantitative techniques other than CellSearch used in the included studies. The epithelial markers CK20 and CEA were the most commonly used among included studies. Taniguchi et al., 2000 [76] used CEA whereas Katsumata et al., 2006 [24], and Hinz et al., 2017 [44], used CK2O, and reported a significant association between CTC at time of surgery and survival. Allen-Mersh et al., 2007 [25] and Ito et al., 2002 [21] used either one or both of these markers to investigate the correlation between CTC detection after surgery and survival. For rectal cancer patients, CK20 was chosen by Hinz et al., 2015 [39], to predict non-response among rectal cancer undergoing neo-adjuvant chemoradiation.
Other newer techniques are being developed such as combined markers detection [31] or new molecule presence, like survivin [29]. Patients with ≥1 CTC per 7.5 mL had a significantly worse OS in the non-metastatic group, as well as in the complete cohort. This strong prognostic factor for both groups was confirm by multivariate analysis.  There was a significant correlation between dichotomous (CK20, CEA), qPCR covariate and DFS or OS. Patients with CEA mRNA-positive portal blood had a lower 4-year recurrence rate than patients with CEA mRNA-negative portal blood. There was no significative correlation between DFS and CEA mRNA positivity.  There was a significant correlation between OS and the different CRC patient groups.  They classified patients with at least four of the six-gene panel markers below cut-offs, as low-CTC. Moreover, those with three or more markers above cut-offs in the high-CTC group. Patients with high B1 CTC had a significant lower median PFS and OS than patients with low CTC markers.  Patients where in the high/low CTCs groups, when at least four markers were above/below the individual cut-offs. There was a significant correlation between all CTC markers that presented an expression above the cut-off (high CTC) and shorter OS and PFS rates, both when analysed at B1. Patients with increased expression of CTC markers during treatment had a shorter PFS and OS times than patients presenting decreased expression.     There was no significant correlation between positive rate of the mutated gene in the blood and tumour stage. Patient with p53 and/or K-RAS gene mutation-positive findings had significantly shorter OS than patients testing negative. There was no significant correlation between CTC detection alone and progression-free or OS. However, there was a significant correlation between nuclear PD-L1 (nPD-L1) expression and short survival.  Prospective CTC detection before neoadjuvant chemoradiation (S1) and before (S2) surgery.

No
Locally advanced rectal cancer undergoing neoadjuvant chemoradiation followed by surgery There was a significant correlation between CTC counts decreased between S1 and S2 in patients and pCR or partial response. Patients exhibiting pCR had negative TYMS and RAD23B CTCs. There was an association between, RAD23B expression and non-responder status at S1 and S2.

Qualitative Data from CTCs
There was some overlap in the reporting of quantitative and qualitative markers. When CTC are detected without being destroyed, many complementary qualitative analyses are available. Study investigators have focused on markers of the aggressiveness of disease, expressed as either proteins (surface cell marker) or genes.
New marker development is focusing on discovering specific CTC subpopulations that may have particular effects on prognosis. Zhao et al., 2017 [67] combined CTC count and EMT state profile to predict disease aggressiveness. Tseng et al., 2015 [58] correlated the presence of CD133+ CD44+ CD54+ CTC subpopulation with lower survival in metastatic patients that did not undergo metastasis resection. Abdallah et al., 2016 [59] used MRP1 positive CTC as a prognosis marker in patient that were going to initiate a new chemotherapy agent.
Since markers are not always easy to determine [64], some clinicians try to associate their detection with other techniques. One of them is to collect blood closer from the tumour. Tseng et al., 2015 [58], used CTC analysis from mesenteric venous blood collected while the patient was on the operating table.

Main Findings
The main finding from our search of the literature regarding the prognostic utility of CTCs for patients with CRC is that there are multiple quantitative and qualitative tests that have been used, with some success, in predicting overall survival and success of treatment, such as surgery and chemotherapy.
The presence of CTCs is a key risk factor for disease progression and severity. In breast cancer, they are considered as efficient prognostic biomarkers and included in the 7th AJCC Cancer Staging Manual (new category M0(i) introduced, and defined "by the presence of circulating or disseminated tumour cells not exceeding 0.2 mm detectable in bone marrow, circulating blood or other non-regional tissues of non-metastatic patients". For patients with CRC, this type of staging is not yet undertaken. Indeed, there is ongoing debate regarding the utility of CTC detection due to the variations in techniques and conflicting results [77,78]. Our review confirmed that there are a range of heterogeneous techniques and results.
Most of our selected studies integrate metastatic patients (46/77). Additional prospective studies that investigate early CRC are needed in order to determine whether CTCs have any prognostic utility in this context. Rectal cancer was the subject for only 3 of the 77 included studies, and requires further evaluation in prognostic studies.
Isolation and examination of CTCs allows different analyses focusing on a variety of cell elements. The studies included in our review show that whatever quantitative or qualitative techniques (or both) are used, CTCs have potential as biomarkers to predict differences in outcomes such as survival and disease prognosis. To improve CTCs detection, new biomarkers are being studied and selected in order to help distinguish CTCs from other cells such as hematopoietic cells.

Detecting Circulating Tumour Cells
Since CTCs are epithelial cells that are found in the circulation, one option to detect their presence would be to use epithelial markers, such as CEA and cytokeratins. CEA is part of the immunoglobulin superfamily and is the product of the CEACAM5 gene [79], and is involved in cellular adhesion. It is commonly found on the surface of small and large bowels, rectum, pancreas, lung, and kidney cells. Being at a low rate for healthy people, it is already being used in routine clinical practice as a biomarker for CRC. However, its specificity is not the best as it can be increased in heavy smokers, inflammatory bowel disease, chronic obstructive pulmonary disease, pancreatitis or other adenocarcinomas than colorectal (such as ovaries, lung, kidney). On the RNA side, CEA mRNA can be detected in almost all epithelial cells, including CTCs, and is not found in non-epithelial cells [80]. Cytokeratins are part of the intermediate filaments of epithelial cells cytoskeleton. They are expressed in a tissue-specific manner. CK20 expression is limited to the gastric and intestinal epithelium, urothelium, and Merkel cells, and from malignancies that originate from these sites [30,81,82]. CK20 is not expressed in hematopoietic cells [83]. The main limitation of these biomarkers is that cancer cells undergo epithelial-mesenchymal transition (EMT) in the circulation, which can cause down regulation of epithelial markers. Therefore, these markers will only partially detect CTCs. Yokobori et al., 2013 [37] chose to detect CTCs by using a marker that is not lost during the EMT process: plastin3 (PLS3). It codes for a ubiquitous protein that inhibits depolymerization of actin fibres. PLS3 is actually an EMT inducer and is therefore overexpressed by CTCs [84]. While working on EMT, Armstrong et al., 2011 [85] found that CTCs expressed both epithelial, and stem cell markers. Some of the included studies utilised stemness markers such as survivin, CD44 variants 6 and 9, and CD133. A good strategy seems to combine both categories of markers (epithelial and stemness) [42]. Survivin is an inhibitor of apoptosis and is highly conserved in CTCs cells. By limiting this programmed cell death, tumour cells develop aggressiveness [86]. Survivin is found in many cancer tissues, including CRC [29] and is not expressed in normal ones [87].
CD44 is a gene involved in adhesion cells and growth and invasion in tumour cells. CD44 is expressed by metastasis initiating cells [88] also known as cancer stem cells in CRC, but its exact function in these particular cells is still unknown, especially the exon that plays the central role [89]. CD133 (also known as prominin-1), is a transmembrane protein, initially described on the surface of hematopoietic stem cells; it is now known as a stemness marker for normal and cancer cells. It is not specific for CRC [90], but is considered as a key marker of tumoural stem cells in CRC [31]. Metastatic colon tissues are formed of CD133 (+) and (−) cells that can both initiate tumour cells. Since CD133 is also known to be expressed by endothelial cells, its expression by RT-PCR might be caused by bone marrow-derived circulating endothelial cells [35].
Other biomarkers with a key role in cancer biology have also been investigated in order to identify subpopulations of CTCs with a higher potential for aggressiveness [35]. Epithelial growth factor receptor (EGFR) is expressed in different cell types except hematopoietic cells and is commonly used as a therapeutic target in CRC. As more and more anti-EGFR resistance is identified, it may become important to detect CTC modification in terms of its expression of EGFR during anti-EGFR treatment [41]. Similarly, as TYMS polymorphisms seem predict response to 5FU-based chemotherapy [91], which is the main chemotherapy used as a neoadjuvant setting before rectal surgery, and also the ultraviolet excision repair protein, RAD23 homolog B (RAD23B), which potentially be "induced by the genetic damage introduced by radiotherapy" [92] before rectal surgery, both monitoring of thymidylate synthase (TYMS) and excision repair protein, RAD23 homolog B (RAD23B), can be used to predict resistance to chemotherapy/radiotherapy used in rectal cancers [69].

The Future
Our updated review confirms the potential of CTC biomarker detection as a prognostic tool, in keeping with the findings of a previous meta-analysis [93]. Molecular characterization at both cellular and genomic levels may be of utility in the prognosis for patients with CRC, whatever the initial stage (metastatic or not). Furthermore, where CTCs remain present despite chemotherapy treatment, this may be used to indicate failure of therapy and predict survival. CTCs remain of great interest to clinicians who look after patients with CRC due to their potential for utilization in clinical practice. They might give a "snapshot" of the disease within time and space, and enable the individualised therapy for patients. This may be possible by identifying patients at risk, facilitate more accurate cancer surveillance and potentially the adaptation of treatments. One focus for the future application of such a test would be the search for the most sensitive and specific (and cost-effective) method for use in routine daily clinical practice.
There were a number of articles excluded because they focused on other liquid biopsy techniques. However, they demonstrate the breadth of interest in the search for applicable circulating prognostic biomarkers for CRC. Some include the "OMICS" (circulating tumoural DNA, miRNA, circulating protein), the microenvironment (tumourderived exosomes, circulating immune cells with tumour associated macrophage) and neovascularisation (circulating endothelial cell, VEGF [Vascular Endothelial Growth Factor], EPC [endothelial progenitor cells], CEC [Circulating endothelial cells]) [94]. Interestingly, some of these other biomarkers made CTCs less prevalent. For example, DNA mutation analysis is now often assessed on circulating tumour DNA rather than on CTC [95].
Several interesting avenues require further evaluation concerning the detection and analysis of CTCs. There may be some role in early diagnosis through screening [96]. CTCs may also be helpful in identifying cancer primary site [97] in cases where there is diagnostic uncertainty. Further investigations regarding tumour biology and the metastatic process are in development, focusing on tumour microenvironment, the epithelial to mesenchymal transition, circulating cancer cell stemness, and others [98,99]. Furthermore, when CTC are isolated without being destroyed, CTC-derived xenografts could lead to patient specific drug screening [100][101][102].

Conclusions
We identified 77 studies that reported data regarding the prognostic utility of CTCs for patients with CRC. CTCs seem to have a high potential for use as a prognostic biomarker in CRC in both quantitative and qualitative terms. Ongoing investigations are required to evaluate the role of CTC analysis in routine clinical practice.