Liquiritin represses proliferation , migration and invasion of colorectal cancer cells through inhibition of the miR-671/HOXB3 signaling pathway

Background: Colorectal cancer is a common malignant tumor and ranks third in cancer-related deaths. Considering that the side-effects of current therapies limit the clinical effectiveness, exploring new anti-colorectal cancer drugs from natural products is critical for the treatment of colorectal cancer. Among these drugs, liquiritin is an active component of the traditional Chinese herb Glycyrrhiza Radix and has been found to possess powerful anti-inammatory and anti-tumor abilities. However, the direct molecular target of liquiritin remains unknown. Therefore, the aim of the present study was to identify potential molecular target of liquiritin to mediate its anti-colorectal cancer effect. Methods: The function of liquiritin in cell proliferation, apoptosis, migration and invasion was estimated by CCK-8 assay, colony formation, EdU assay, �ow cytometry analysis, TUNEL assay, wound healing assay and transwell assay, respectively. Animal experiment was carried out to further con�rm the role of liquiritin in vivo. H&E staining analysis, TUNEL staining and immunohistochemistry (IHC) assay were adopted for histological analysis. The mechanism research was conducted with RT-qPCR, western blot and luciferase report assay. Results: In this study, we found that liquiritin signi�cantly inhibited the proliferation, migration, invasion and EMT processes of SW480 and HT-29 cells in a dose-dependent manner. MiRNAs have been extensively identi�ed as drug targets in various studies. However, the miRNAs functioning as the direct targets of liquiritin remain unknown. In our study, we tested 8 potential pathogenic miRNAs screened from colorectal cancer patients, and we found that only the expression level of miR-671 was diminished due to liquiritin treatment. More importantly, overexpression of miR-671 partially abrogated the anti-tumor effects of liquiritin on colorectal cancer. Conclusions: Collectively, our �ndings demonstrated that liquiritin exhibits great potential in the treatment of colorectal cancer through regulation of miR-671.


Background
Colorectal cancer is a common malignant tumor and is the third leading and second leading cause of cancer-related death among males and females, respectively [1].Recently, the incidence of colorectal cancer worldwide increased to approximately 655, 000 people annually [2].Although the 5-year survival rate of early colorectal cancer is approximately 90%, the survival rate of metastatic colorectal cancer has decreased to 15% [3,4].Hence, invasive metastasis, especially extensive metastases to liver, lung, bone and abdominal cavity, are the leading causes of death.Colorectal cancer is a multistep, multistage and multigene cytogenetic disease, and it includes interactions between many factors, such as activation or inactivation of oncogenes and tumor suppressor genes, mismatch mutation of repair genes, gene overexpression and epigenetic modi cation of genome [5][6][7].Current therapies, especially chemotherapy drugs, have achieved remarkable advances, but their side-effects restrict the clinical effectiveness.In this regard, many studies have focused on the development of drugs from natural products, which are considered safe.Therefore, exploring new anti-colorectal cancer drugs from natural products is critical for the treatment of colorectal cancer.
Glycyrrhiza Radix, as a traditional Chinese medicine, is one of the most commonly used drugs in clinical practice, and it belongs to leguminous plants [19,20].Currently, Glycyrrhiza Radix is widely used for anticancer effects, liver protection, liver detoxi cation, anti-in ammation effects, anti-allergy effects, immune regulation and other aspects [21,22].Liquiritin (Fig. 1A), an active component of Glycyrrhiza Radix, possesses multiple pharmacological activities including anti-in ammation, neuroprotection, anti-cancer activities [23,24].For example, liquiritin induces apoptosis and autophagy in cisplatin (DDP)-resistant gastric cancer [25].In addition, liquiritin exhibits signi cant inhibitory effects on the growth of cervical cancer cells via activation of caspase-3 [26].Furthermore, in combination with isoliquiritin and isoliquirigenin, liquiritin signi cantly aggravates apoptosis in non-small cell lung cancer cells via regulation of p53 and p21 [27].However, the speci c role and the direct molecular target of liquiritin remain unknown, which restricts its clinic application and potential structure modi cation to improve its bene cial functions.Thus, this study was performed to explore the precise effects of liquiritin on colorectal cancer in vitro and in vivo, as well as its underlying mechanism.
MicroRNAs (miRNAs) are a type of noncoding RNAs present in various organs and implicated in the regulation of multiple biological processes.Mature miRNAs are typically 19 to 25 nucleotides in length.It should be noted that abnormal expression patterns of miRNAs exist in human tumors, including the ampli cation, deletion, aberrant post-transcriptional control, epigenetic alterations or defects of miRNAs [8][9][10].Hence, under certain conditions, these dysregulated miRNAs activate the proliferation of cancer cells, resist apoptosis, trigger invasion, promote metastasis signals, and induce angiogenesis [11,12].For the colorectal cancers, various miRNAs are also involved in the regulation of cell growth and metastasis of colorectal cancer.For instance, miR-141-3p is decreased in colorectal cancer, while overexpression of miR-141-3p signi cantly inhibits colorectal cancer cell proliferation, migration and invasion by targeting TRAF5 [13].MiRNA-766 aggravates cell apoptosis through regulation of p53/Bax signaling via targeting MDM4 in human colorectal cancer [14].Furthermore, miR-214 enhances the chemo-sensitivity of 5-FU by targeting HSP27 in colorectal cancer [15], indicating that miRNAs also function as drug targets.However, the miRNAs that function as the direct targets of liquiritin remain unknown.In our study, we tested 8 potential pathogenic miRNAs screened from colorectal cancer patients.Of these miRNAs, only miR-671 was sensitive to liquiritin treatment.MiR-671, as a member of the miR-671 family, is closely associated with the development of malignant tumors.For example, overexpression of miR-671 induces growth and metastasis of glioblastoma multiforme via targeting CDR1-AS, CDR1, and VSNL1 [16].In addition, miR-671 is induced in clinical prostate cancer tissues, and repression of miR-671 remarkably inhibits prostate cancer cell proliferation by targeting SOX6 [17].Moreover, miR-671 suppresses gastric cancer cell proliferation and enhances cell apoptosis by targeting URGCP [18].Nevertheless, the speci c effects and mechanisms of miR-671 on colorectal cancer have not been elucidated.Herein, we found that liquiritin inhibited cell proliferation, promoted cell apoptosis, reduced cell migration, reduced invasion and suppressed EMT in a dose-dependent manner in colorectal cancer.Mechanistically, liquiritin decreased the expression of oncogenic miR-671 but increased HOXB3 levels.More importantly, overexpression of miR-671 partially abrogated the bene cial effects of liquiritin on colorectal cancer progression, indicating the regulatory role of miR-671 in relaying liquiritin signals.In conclusion, our study indicated that liquiritin impedes the progression of colorectal cancer via regulation of miR-671/HOXB3.

Cell viability assay
Cell viability of colorectal cancer cells treated with liquiritin was determined by the Cell Counting Kit 8 assay (CCK-8, Dojindo Molecular Technology, Rockville, MD, USA).In brief, HT-29 and SW480 cells (1 ⋅ 10 4 cells/well) were cultured in 96-well plates.The medium was replaced with medium containing indicated concentrations of liquiritin (10, 30 and 90 µM) and incubated in a 5% CO 2 atmosphere at 37℃ for 24, 48 and 72 h.Subsequently, 10 µL of CCK-8 solution was added to each well for an additional 3 h.Finally, a microplate reader was used to measure the absorbance at 450 nm.

Cell proliferation assay
Colony formation and EdU analyses were performed to determine the effects of liquiritin on the proliferation of HT-29 and SW480 cells.For the colony formation assay, HT-29 and SW480 cells (2 × 10 4 cells/well) were seeded into 24-well plates, and the medium was replaced with medium containing indicated concentrations of liquiritin (10, 30 and 90 µM) every 2-3 days.After incubation for 12 days, these cells were xed with paraformaldehyde for 30 min and stained with 10% crystal violet for 30 min.For the EdU assay, cells were incubated with medium containing liquiritin (10, 30 and 90 µM) for 48 h and 100 µL of EdU (50 µM) was added into each well and cultured for an additional 8 h.SW480 and HT-29 cells were xed and stained according to the manufacturer's protocol (Life Technologies, NY, USA).

Flow cytometry analysis
The role of liquiritin in apoptosis of HT-29 and SW480 cells was assessed by ow cytometry analysis.In brief, HT-29 and SW480 cells (1 ⋅ 10 4 cells per well) were seeded in 24-well plates, and treated with indicated concentrations of liquiritin (10, 30 and 90 µM) for 48 h.Subsequently, 300 µL of 1 × binding buffer was added to each well, and the treated cells were washed twice with PBS.Then, 5 µL of Annexin V-PE was added for 15 min, and 5 µL of 7-AAD solution was added for 5 min.Finally, 200 µL of 1 × Binding Buffer was added for ow cytometry analysis.

TUNEL assay
For in vitro experiments, the TUNEL assay was performed to explore the effects of liquiritin on the apoptosis of HT-29 and SW480 cells according to the manufacturer's protocol (Roche, USA).In brief, HT-29 and SW480 cells were treated with indicated doses of liquiritin (10, 30 and 90 µM) for 48 h.Cells were then xed with methanol-free formaldehyde solution for 30 min, and cells were permeabilized with permeabilization solution for an additional 2 min.The TUNEL reaction mixture was subsequently added to slides and incubated for 1 h at 37℃ in dark.Representative images were acquired with a microscope.
For in vivo experiments, the apoptosis of para n-embedded tumor sections was detected by the TUNEL assay.Brie y, sections were dewaxed, and permeabilized with proteinase K for 15 min at room temperature.Sections were then treated with 3% H 2 O 2 to block endogenous peroxidases, and incubated with equilibration buffer and terminal deoxynucleotidyl transferase enzyme.Finally, sections were incubated with antidigoxigenin-peroxidase conjugate.Finally, DAB was used for visualization.Sections were examined under a light microscope.

Wound-healing assay
The migration abilities of HT-29 and SW480 cells treated with liquiritin were evaluated by a woundhealing assay.Brie y, HT-29 and SW480 cells (5 ⋅ 10 6 cells/well) were seeded into 6-well plates and grew to con uence.A standard wound (< 3 mm) was created on the cell monolayer (time set as 0 h), and then cells were subjected to liquiritin treatment at indicated doses in serum-free medium for 48 h.The width of the scratch was imaged and recorded under a microscope (⋅ 200).

Transwell migration and invasion assays
The migration and invasion abilities of SW480 and HT-29 cells in response to liquiritin treatment were determined using a transwell chamber assay.For the migration assay, 5 ⋅ 10 4 SW480 or HT-29 cells were seeded into the transwell chambers.For the invasion assay, Matrigel was diluted with medium (1:3), and 35 to 45 µL of the solution was added to transwell chambers.For transwell migration and invasion assays, the lower wells of the chamber were lled with 600 µL of medium containing 20% FBS.SW480 or HT-29 cells (5 ⋅ 10 4 cells) were treated with indicated doses of liquiritin (0, 10, 30 and 90 µM) and allowed to migrate or invade for 24 h.Subsequently, non-migrated cells were removed from the upper side of the membrane with cotton swabs.Cells on the lower side of the membrane were stained with 350 µL of Wright-Gimsa dye for 5 min.Finally, four elds of view were imaged under a light microscope (× 400).

Luciferase reporter assay
The putative binding site in the 3'-untranslated regions (3'-UTR) of HOXB3 was mutated using mutagenesis kit.Wild type and mutant sequences were ampli ed and inserted into the vector to construct luciferase reporter plasmids according to the manufacturer's recommendations (Promega, USA).The luciferase activities were detected with the Dual Luciferase Reporter Kit (Promega, USA).The 3' UTR region of human HOXB3 cDNA containing one putative target site for miR-671-3p was ampli ed and inserted into the vector to construct luciferase reporter plasmids according to the manufacturer's recommendations (Promega, Madison, WI, USA).A mutant version with a mutation of 7 bp (GAACCGG mutated to GACAAGC) from the sites of perfect complementarity was also generated.Wild-type (WT) and mutant (Mut) inserts were con rmed by sequencing.For transfection, 200 ng of WT or Mut reporter vector was transfected into either HT-29 or SW480 cells together with miR-671-3p mimics.
Relative luciferase activities (ratios of Renilla luciferase signal normalized to re y luciferase) were determined 48 h after transfection.Equal amounts of NC mimic were used for all the transfections.

Xenograft tumors in nude mice
To detect the effect of tumor growth in vivo, SW480 cells (1 × 10 6 cells per mouse) were injected subcutaneously into the right axilla of the nude mice (Model Animal Research Center of Nanjing University, Nanjing, Jiangsu, China).Tumors were allowed to grow until their average diameter reached over10 mm.Then, the experimental nude mice were treated with indicated doses of liquiritin (10, 20, 40 mg/kg i.p.) every 3 days for 30 days.Following a 30-day period, nude mice were euthanized by a high dose of anesthesia (150 mg/kg pentobarbital, intraperitoneal injection) as per protocols set by the Ethical Committee of Jiangsu Cancer Hospital.After the nude mice were sacri ced, neoplasms were isolated for further analyses.Tumor volumes were recorded by slide calipers every week and calculated with the formula: volume = 0.5 × length × width × depth.

Hematoxylin and eosin (H&E) staining
Tumor slices were stained with hematoxylin for 5 min, rinsed for 1 min, and incubated with 1% ammonia (30 s).Slices were then rinsed with running water (1 min).Slices were stained with by 0.5% H&E (for 1 min), rinsed (for 30 s), made transparent, and mounted with neutral gum.

Statistical analysis
Graph Pad Prism 5.0 statistical software was used to perform all statistical analyses.Measurement data were represented as the mean ± SD.Statistical differences among multiple groups were analyzed by oneway ANOVA followed by a Bonferroni post hoc analysis.A value of P < 0.05 was considered as statistically signi cant.

Liquiritin represses the proliferation of colorectal cancer cells
As shown in Fig. 1B, the CCK-8 assay revealed that liquiritin inhibited HT-29 and SW480 cell viability in a dose-dependent manner.To further assess the effects of liquiritin on the proliferation of colorectal cancer cells, we performed the EdU assay and found that liquiritin dose-dependently repressed the proliferation of HT-29 (inhibition by 61%) and SW480 cells (inhibition by 68%) in a dose-dependent manner (Fig. 1C).These results were con rmed by the colony formation analysis (Fig. 1D).These data suggested that liquiritin executes inhibitory effects on colorectal cancer cell proliferation.

Liquiritin aggravates apoptosis of colorectal cancer cells
To determine the effect of liquiritin on the cell apoptosis of HT-29 and SW480 cells, ow cytometry analysis was performed.As expected, liquiritin (10, 30 and 90 µM) signi cantly induced SW480 and HT-29 cell apoptosis compared to the untreated group (Fig. 2A).In addition, the TUNEL assay revealed that the apoptosis of HT-29 and SW480 cells was gradually augmented with increasing concentrations of liquiritin (Fig. 2B).In HT-29 and SW480 cells treated with liquiritin at the indicated doses, the mRNA and protein expression levels of cleaved-caspase-3, cleaved-caspase-9, and BAX were decreased, whereas the mRNA and protein expression levels BCL2 were increased (Fig. 2C and 2D).All of the effects of liquiritin occurred in a dose-dependent manner.

Liquiritin inhibits migration and invasion of colorectal cancer cells
To identify the potential role of liquiritin in the regulation of colorectal cancer cell migration and invasion, wound-healing, transwell migration and transwell invasion analyse were performed.As shown in Fig. 3A, the wounding healing rate in liquiritin (10, 30 and 90 µM) groups was signi cantly decreased compared to that of untreated group.Moreover, the transwell migration and invasion assays (Fig. 3B and 3C) showed that the migration and invasion of HT-29 and SW480 cells were notably repressed by liquiritin in a concentration-dependent manner (10, 30 and 90 µM).At the molecular level, the expression levels of migration-and invasion-associated proteins, including MMP-2 and MMP-9, were decreased in response to liquiritin treatment at the indicated doses (Fig. 3D).In addition, epithelial-mesenchymal transformation (EMT) plays an important role in the occurrence, development and metastasis of tumors [28,29].Key transcription factors, including E-cadherin, N-cadherin and vimentin, are extensively involved in the regulation of the EMT process [30].Therefore, RT-qPCR and Western blot assays revealed that liquiritin signi cantly enhanced E-cadherin level and downregulated the expressions of N-cadherin and vimentin (Fig. 3E and 3F).

Liquiritin suppresses the growth of colorectal cancer xenografts in vivo
We next used SW480 tumor xenograft models to evaluate the suppressive effect of CLQ in vivo.As shown in Fig. 4A, the average tumor volume in the control group reached 2.5 mm 3 , whereas 10 mg/kg, 30 mg/kg and 90 mg/kg liquiritin treatments remarkably decreased tumor volume by 81.8%, 61.1% and 40.8%, respectively, compared to the control group.Consistently, the xenograft tumor weights were reduced by 73.7%, 56.6% and 31.6% in the 10 mg/kg, 30 mg/kg and 90 mg/kg liquiritin treatment groups, respectively (Fig. 4B and 4C).Histologically, H&E staining analysis showed that liquiritin treatment signi cantly decreased tumor cell density and blurred tumor cell borders but increased severe necrosis (Fig. 4D).TUNEL staining revealed that liquiritin increased tumor cell apoptosis in vivo in a dosedependent manner (Fig. 4E).Immunohistochemistry (IHC) analysis indicated that the number of Ki-67positive cells gradually reduced in response to liquiritin treatment (Fig. 4F), con rming the inhibitory effects of liquiritin on tumor growth.Additionally, liquiritin treatment led to enhanced expression of Ecadherin and reduced N-cadherin levels (Figure S1A and 1B).These data indicated that liquiritin exerts anti-colorectal cancer properties in vivo.

Liquiritin decreases the expression of miR-671 and promotes the expression of HOXB3 in colorectal cancer in vitro and in vivo
MiRNAs have been extensively identi ed as drug targets in various studies [31].However, the miRNAs that function as direct targets of liquiritin remain unknown.Based on a previously performed microarray, eight potential miRNAs were screened from colorectal cancer patients.We hypothesized that these miRNAs may be the target of liquiritin.We tested these miRNAs in our settings and found that only the expression level of miR-671 was diminished due to liquiritin treatment (Fig. 5A).In contrast to normal colonic epithelial cells NCM460, miR-671 level was elevated, and the HOXB3 level was downregulated in colorectal cancer cells (Figure S1C and 1D).On the other hand, according to previous studies, miR-671 and HOXB3 levels are abnormally expressed and exhibit important effects on the growth and metastasis in the pathogenesis of cancer [16][17][18][32][33][34].Therefore, we further examined the dose-dependent effect of liquiritin on the expression levels of miR-671 and HOXB3 in vitro and in vivo.As expected, liquiritin reduced the expression of miR-671 in HT-29 and SW480 cells (Fig. 5B).Similar results were demonstrated in the in vivo system (Fig. 5C).Furthermore, RT-qPCR and Western blot analyses showed that liquiritin increased the expression of HOXB3 in HT-29 and SW480 cells in a dose-dependent manner (Fig. 5D and  5E).In agreement with these results, the mRNA and protein expression levels of HOXB3 in xenografts were enhanced correspondingly with liquiritin treatment (Fig. 5F).These data suggested that liquiritin may exhibit inhibitory effects on the growth of colorectal cancer via regulation of miR-671 and HOXB3.

HOXB3 is a direct target gene of miR-671
To validate if HOXB3 is a target gene of miR-671 that mediates its function in the development and progression of colorectal cancer, TargetScan analysis was employed, which indicated that HOXB3 may be a candidate target of miR-671 (Fig. 6A).To further dissect the relationship between HOXB3 and miR-671, HT-29 and SW480 cells were transfected with NC mimics or miR-671 mimics.The overexpression e ciency was presented in Fig. 6B.The luciferase reporter assay revealed that miR-671 inhibited the activity of the human HOXB3 3'UTR region.More importantly, the inhibitory effects of miR-671 were abolished when the miR-671-3p-binding sites in the HOXB3 3'UTR region were mutated in HT-29 and SW480 cells (Fig. 6C).Moreover, overexpression of miR-671 signi cantly decreased the HOXB3 expression in HT-29 and SW480 cells at transcriptional and translational levels (Fig. 6D and 6E).Therefore, HOXB3 may be a direct target gene of miR-671.

Liquiritin affects the proliferation of colorectal cancer cells through regulation of miR-671/HOXB3
To further demonstrate the role of miR-671 in mediating the inhibitory effect of liquiritin on cell proliferation, we transfected either miR-671 or NC mimics into HT-29 and SW480 cells for 48 h, followed by liquiritin treatment.The overexpression e ciency was presented in Fig. 6B.)CCK-8 and EdU analyses indicated that overexpression of miR-671 accelerated the proliferation of HT-29 and SW480 cells by 1.63fold and 1.59-fold, respectively.More importantly, this overexpression abrogated the liquiritin-inhibited proliferation of HT-29 and SW480 cells (Fig. 7A and 7B), indicating that miR-671 mediates the bene cial effects of liquiritin on the proliferation of colorectal cancer cells.

Liquiritin affects the migration and invasion of colorectal cancer cells through regulation of miR-671/HOXB3
To further con rm if miR-671 is involved in the liquiritin-induced inhibition of colorectal cancer cell migration and invasion, we transfected HT-29 and SW480 cells with miR-671 mimics for 48 h followed by liquiritin treatment.As shown in Fig. 8A to 8C, overexpression of miR-671 alone promoted the migration and invasion of HT-29 and SW480 cells, while this overexpression partially antagonized the repression effect of liquiritin on cell migration and invasion.In addition, RT-qPCR and western blot assays indicated that overexpression of miR-671 repressed E-cadherin expression but promoted N-cadherin and vimentin expression at both mRNA and protein levels (Fig. 8D and 8E).Importantly, overexpression of miR-671 partially abrogated the bene cial effects of liquiritin on the EMT process.These results indicated that miR-671 partially mediates the function of liquiritin in the migration and invasion of colorectal cancer cells.

Discussion
Colorectal cancer is a common malignancy that has been ranked the third fatal cancer in recent years.In China, the morbidity and mortality of colorectal cancer have been increasing year by year [1,2].Surgical resection is the most effective treatment for colorectal cancer, but these patients have poor prognosis and are prone to recurrence after surgery [35].Therefore, colorectal cancer treatment is largely dependent on medical treatment.However, although 5-uorouracil is the most basic chemotherapy drug for colorectal cancer, its single drug e ciency is not high [36].Hence, conventional chemotherapy is not functional for the treatment of colorectal cancer.In recent years, new chemotherapeutic drugs, such as capecitabine and irinotecan, as well as the molecular targeting drug cetuximab, have been developed and achieved certain curative effects.However, shortcomings, including high price, large toxicity and side effects, are inevitable in such treatments [37][38][39].As a result, developing highly effective, low cost anti-colorectal cancer drugs is still a top priority.In this study, we demonstrated liquiritin as a potential anti-colorectal cancer drug and identi ed the direct molecular target, namely, miR-671, which relays the liquiritin signals.
Traditional Chinese medicine has the advantages of multitarget, high e ciency, low toxicity and cost in the treatment of cancer.These herbs have unique superiorities in stabilizing the condition and alleviating pain, thus further improving life quality and extending the survival time of cancer patients.Recently, the anti-tumor effects of nature plant extracts have become one of the hotspots in anti-cancer drug development, and they have attracted increased attention from oncologists [38][39][40][41][42]. Glycyrrhiza Radix is a traditional Chinese medicine with the positive functions in clearing heat, clearing toxic material, relieving spasm, relieving pain, reinforcing spleen function, reconciling the stomach, expelling phlegm and arresting coughing [19,20].Liquiritin is one of active ingredients extracted from the root of Glycyrrhiza uralensis.Emerging studies have shown that liquiritin has multiple bene cial effects, such as antioxidation, anti-allergy, immunity-enhancing, anti-thrombosis and neuroprotection effects [23,24].More importantly, liquiritin has powerful inhibitory effects on the development of various cancers, such as gastric cancer, cervical cancer and lung cancer [25][26][27].However, the effects of liquiritin on colorectal cancer remain unknown.In the present study, liquiritin repressed cell proliferation but remarkably induced apoptosis of HT-29 and SW480 cells.These inhibitory effects of liquiritin on cancer growth were further con rmed by an in vivo xenograft experiment.Hence, liquiritin may be a promising anti-colorectal cancer drug.
To enter the blood or lymphatic circulation, stationary epithelial-derived tumor cells migrate and invade, causing tumor metastasis, which is the main cause of death for cancer patients [43].In the present study, we also found that liquiritin substantially diminished the migration and invasion of HT-29 and SW480 cells as evidenced by decreased migrated distances and reduced protein levels of MMP-2 and MMP-9.Moreover, the EMT process promotes the migration and invasion of stationary tumor cells [44].The concept of EMT was rst proposed by Hay and Greenburg in 1982, and it was mainly used to describe the physiological process of embryonic stage [45].Further studies have broadened the EMT concept and found that this process is also involved in the chronic degeneration of mature organs, brosis of mature organs and tissue remodeling and that it is closely associated with the metastasis of tumors.As EMT is an early process of tumor metastasis, we examined the effect of liquiritin on the cancer cell EMT process.
We found that liquiritin repressed the protein expression levels of EMT hallmarks, such as N-cadherin and vimentin, but increased E-cadherin levels.Given the importance of EMT in versatile biological processes, the pharmacological effects of liquiritin on diseases, such as chronic degeneration, remain to be further investigated.
Recent studies have shown that chemotherapeutic drugs affect the expression of miRNAs in cancer cells, suggesting that miRNAs are extensively involved in the anticancer effects of chemotherapeutic drugs.For example, Kaempferol inhibits growth and metastasis of hepatocellular carcinoma HepG2 cells via decreasing miR-21 [46].Ginkgetin induces G2-phase arrest in colorectal cancer cells by regulating b-Myb and miRNA34a [47].Dioscin exhibits protective effects against pancreatic cancer through suppression of Akt1 mediated by miR-149-3p [48].These previous ndings promoted us to hypothesize that targeting miRNAs may provide new strategies and ideas for the development of new anticancer drugs.From clinical sample sequences, eight miRNAs were identi ed to be potentially involved in the development of colorectal cancer.However, only miR-671 responded to liquiritin treatment, indicating its regulatory role in drug treatment.Indeed, miR-671 has been reported to be highly expressed in many tumors, such as breast cancer, glioblastoma multiforme, and gastric cancer, and down-regulation of miR-671 suppresses biological functions of tumor cells [16][17][18].The role of miR-671 in cancer progression is still debated.For example, miR-671 inhibits the invasion and cell adhesion of breast cancer by binding to the 3'UTR of FOXM1 [49], whereas miR-671 promotes glioblastoma multiforme proliferation and migration by targeting CDR1-AS, CDR1, and VSNL1.Nevertheless, the speci c effects and underlying mechanisms of miR-671 on colorectal cancer and liquiritin-mediated colorectal cancer have not been elucidated.In the present study, liquiritin decreased miR-671 expression, while overexpression of miR-671 promoted proliferation, suppressed apoptosis, enhanced migration, enhanced invasion and enhanced EMT in HT-29 and SW480 cells.These data suggested that miR-671 was an onco-miRNA in the colorectal cancers.Of note, overexpression of miR-671 partially abrogated liquiritin-induced inhibition of proliferation, migration, invasion and EMT as well as liquiritin-induced apoptosis in HT-29 and SW480 cells, suggesting that miR-671 is a potential liquiritin target.Moreover, it has been reported that miR-671 affects the expression of various genes to regulate cancer development [16][17][18].Therefore, TargetScan was used to predict the target candidates of miR-671, and HOXB3 was found to be a potential candidate of miR-671.HOXB3 is a member of the HOX family, and it has been reported to be highly expressed in prostate cancer, pancreatic cancer and breast cancer, playing a crucial effect on the occurrence and progression of cancer [50][51][52][53].
Notably, HOXB3 is described as an oncogene in lung tumors due to its epigenetic silencing effects on the RASSF1A tumor suppressor.Consistently, overexpression of HOXB3 promotes cell growth, migration and invasion in endometrial cancer.HOXB3 also accelerates the progression of prostate cancer via transactivation of CDCA3.All of these ndings indicate that HOXB3 functions as an oncogene in cancer progression.However, the role of HOXB3 in the colorectal cancer has not been identi ed.In our study, we demonstrated that HOXB3 was increased in response to the liquiritin treatment.In addition, the luciferase reporter assay indicated that HOXB3 was a target gene of miR-671 in HT-29 and SW480 cells.With the reduction of miR-671 induced by liquiritin, the expression level of HOXB3 was unexpectedly increased in vitro and in vivo.As HOXB3 is a molecular target of miR-671, this unexpected increase may be caused by the repressive effects of liquiritin on the miR-671.Hence, further studies are needed to investigate the direct role of HOXB3 by manipulating the expression of HOXB3 in the presence of liquiritin.Moreover, the bene cial effects of either liquiritin or suppressed miR-671 on colorectal tumor may not be simply due to one molecule.Although we found an undesirable increase of HOXB3, the net effects of liquiritin and the suppression of miR-671 caused by liquiritin impeded tumor progress.Broad bioinformatics and molecular biological analyses need to be performed to identify the direct molecular target of liquiritin/miR-671 in future studies.

Conclusion
To sum up, liquiritin, as a type of Glycyrrhiza Radix, retarded cell growth, migration and invasion of colorectal cancer in vitro and in vivo through regulation of the miR-671/HOXB3 signaling pathway.Therefore, liquiritin may be a powerful candidate compound for anti-colorectal cancer drug development.

Supplementary Files
This is a list of supplementary les associated with this preprint.Click to download. FigS1.tif