Leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5) is a 7-transmembrane receptor reportedly expressed in stem cells of the intestinal crypts and hair follicles of mice. Overexpression of LGR5 is observed in some types of cancer; however, there has been no specific assessment in colorectal tumorigenesis. We performed quantitative RT-PCR for LGR5 expression in 37 representative cancer cell lines, and showed that LGR5 mRNA was frequently overexpressed in colon cancer cell lines. Moreover, LGR5 expression was higher in colon cancer cell lines derived from metastatic tumors compared with those from primary tumors. In clinical specimens, there was significant overexpression of LGR5 in 35 of 50 colorectal cancers (CRCs), and in seven of seven sporadic colonic adenomas, compared with matched normal mucosa. This suggests up-regulation of LGR5 from the early stage of colorectal tumorigenesis. LGR5 expression showed marked variation among CRC cases and correlated significantly with lymphatic invasion, vascular invasion, tumor depth, lymph node metastasis, and tumor stage (IIIC vs. IIIB). In addition to cancer cells, crypt base columnar cells of the small intestine and colon were shown by in situ hybridization to express LGR5. This is the first report suggesting the involvement of LGR5, not only in early events but also in late events in colorectal tumorigenesis. (Cancer Sci 2010)
Colorectal cancer (CRC) is one of the most common causes of cancer death in the world.(1) Despite improved surgical techniques, adjuvant/neoadjuvant chemotherapy, and molecular target therapy, patients with progressive disease still have poor clinical outcomes. The 5-year survival rate for patients with localized CRC is about 90%, whereas the rate for patients with regional and distal metastases is only 60%, and 10%, respectively.(1,2) New markers for cancer prognosis and therapeutic options targeting invasion and metastasis are needed to improve treatment of CRC.
Mutations in both alleles of the adenomatous polyposis coli (APC) gene are associated with >70% of human intestinal carcinomas.(3) This mutation results in the accumulation of β-catenin in the nucleus and aberrant activation of a Wnt target gene program that initiates the transformation of intestinal epithelial cells.(4–7) Patients with familial adenomatous polyposis (FAP) carry the APC mutation, which develops multiple colonic adenomas, indicating that loss of the APC “gatekeeper” function is associated with the earliest stages of intestinal tumorigenesis.(3)
Recent studies suggest that leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5), also known as GPR49, is a target of Wnt signaling.(7–9) LGR5 is a member of the G-protein-coupled receptor (GPCR) family comprising proteins with seven transmembrane domains. GPCRs function as receptors for various classes of ligand, including peptide hormones and chemokines.(10) To date, the ligand for, and function of, LGR5-related signaling remain unclear. A recent study showed that LGR5-null mice exhibited neonatal lethality, characterized by ankyloglossia and gastrointestinal distension.(11) It was also reported that LGR5 is a marker for stem cells in the small intestine and colon,(12) and in the hair follicles of mice.(13) Therefore, LGR5 may play a crucial role in the biological function of stem cells. As we reported previously, LGR5 is overexpressed in basal cell cancers (BCCs) and hepatocellular carcinomas (HCCs).(9,14) LGR5 expression is down-regulated in colon cancer cell lines as a result of Wnt signaling suppression, and is up-regulated in intestinal adenomas from humans with FAP and mice with the germline APC mutation.(7,8) Therefore, LGR5 may be involved in colorectal carcinogenesis as a Wnt target gene. We specifically analyzed LGR5 gene expression using a panel of cancer cell lines and colorectal surgical specimens.
Materials and Methods
Patients and samples. Surgical resections were performed on 395 patients with CRC at Keio University Hospital, Japan, between December 2005 and April 2008. Curative operations were performed on 89 patients with pT3/T4 and stage II/III CRC. From the 89 patients, we selected 50 patients: 25 cases each positive and negative for lymph node metastases. Surgically resected specimens of cancerous tissue and matched normal mucosal tissue were obtained from the 50 patients for quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis. All samples were taken from surgically resected material, immediately frozen in liquid nitrogen and stored at −80°C. Adequate lymph node dissection was performed in all 50 cases and a pathological diagnosis was made according to the TMN classification. These 50 patients comprised 27 males and 23 females; average age 66.0 years (range, 48–87 years). Clinicopathologically, there were no significant differences between the two groups except for lymphatic invasion and venous invasion (Table 1). None of the 50 patients had synchronous cancers, ulcerative colitis, or FAP and none had received pre-operative chemotherapy or radiation therapy. These 50 cases were evaluated for the association between LGR5 expression and clinicopathological findings. Tissue specimens were also obtained from seven patients with sporadic colorectal adenomas. Specimens were collected, stored, and analyzed, as described previously. The study was approved by the ethics committee of the School of Medicine, Keio University.
Table 1. Patient characteristics of 50 cases of pT3/T4 colorectal cancer
Lymph node metastasis
*Fisher’s extract test; †tumor invasion, TNM classification by International Union Against Cancer (UICC) was adopted. ‡Right side, from cecum to transverse colon; §left side, from descending colon to rectum; ¶χ2-test. There was no significant difference between lymph node metastasis and clinicopathological findings except for lymphatic and vascular invasion.
Cell culture. Thirty-seven cell lines derived from colon cancers (COLO 201, COLO 205, HT-29, HCT-15, SW1116, SW480, SW620, LoVo, WiDr, Caco-2, and HCT 116), HCCs (Hep G2, PLC/PRF/5, KIM-1, KYN-2, and Li7), ovarian cancers (ES-2, TOV-21G, RMG I, RMG II, RMG V, TOV-112D, RMUG-S, RMUG-L, HTOA, and KF 11), and lung cancers (SK-LU-1, Calu-3, A549, PC9, H1650, H1975, SK-MES-1, SW 900, H69, H146, and H889), were used. Details of the colon cancer cell lines are shown in Table 2. All cell lines were obtained from ATCC (Manassas, VA, USA) except for PC9 (Immuno-Biological Laboratories, Takasaki, Japan), and KIM-1, KYN-2, Li7, RMG I, RMG II, RMG V, RMUG-S, RMUG-L, HTOA, and KF.(9,15–17) All cell lines, except SK-LU-1, Calu-3, A549, PC9, H1650, H1975, SK-MES-1, and SW 900, were routinely maintained at 37°C with 5% CO2 in RPMI-1640 containing 10% FBS. SK-LU-1, Calu-3, A549, PC9, H1650, H1975, SK-MES-1, and SW 900 were routinely maintained at 37°C with 5% CO2 in DMEM containing 20% FBS.
Table 2. Characteristics of colon cancer cell lines
†,‡Indicate origin from same patient. NA, not available.
Colon, distant lymph node
Colon, metastatic lymph nodule
Quantitative RT-PCR. Total RNA was isolated from cultured cells and CRC tissues using RNeasy Mini Kits (Qiagen, Tokyo, Japan), and RNAiso (Takara Bio, Shiga, Japan), with a slight modification of the DNase treatment. RNA quality was assessed in the RNA 6000 Nano LabChip using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). cDNA was synthesized using the PrimeScript 1st strand cDNA Synthesis Kit (Takara Bio). Quantitative RT-PCR analysis was performed on a Thermal Cycler Dice Real Time System using SYBR Premix Ex Taq (Perfect Real Time) (Takara Bio). The primer sequences for GAPDH were 5′-ATCATCCCTGCCTCTACTGG-3′ and 5′-TTTCTAGACGGCAGGTCAGGT-3′, and those for LGR5 were 5′-TCTCAGCCATGGTGAACAA-3′ and 5′-TAGCGAATCACCAGGAAGGT-3′. Fold-induction values were calculated using the 2−ΔΔCt method and LGR5 expression was normalized to GAPDH. All experiments were performed in triplicate and repeated in at least three separate experiments; representative data from these experiments are presented.
Immunohistochemical analyses. For immunohistochemistry, normal intestinal mucosal tissues were analyzed. Formalin-fixed and paraffin-embedded serial sections were incubated with Ki-67 mouse monoclonal antibody (1 : 200; Dako, Glostrup, Denmark) and chromogranin A rabbit polyclonal antibody (1 : 4000; Dako) for 1 h at room temperature for detection of proliferating cells and enteroendocrine cells, respectively, in intestinal crypts. Primary antibody binding was visualized using an ImmPRESS Anti-Mouse Ig Kit (Vector Laboratories, Burlingame, CA, USA) for Ki-67, and swine antirabbit secondary antibody (Dako) for chromogranin A, with diaminobenzidine as the chromogen. The sections were counterstained with hematoxylin and mounted.
In situ hybridization. Tissues were fixed in 10% formalin and embedded in paraffin, and 5-μm sections were cut and placed on poly-l-lysine-coated slides. Digoxigenin-labeled LGR5 sense and antisense probes were generated from a 359-bp fragment of LGR5 (corresponding to nucleotides 1730–2088, GenBank ID NM_003667) using the DIG-RNA labeling Kit (SP6/T7) (Roche Diagnostics, Mannheim, Germany). In situ hybridization (ISH) was performed according to the following protocol. De-paraffinized sections were treated with proteinase K (10 μg/mL) for 9 min at 37°C, and acetylated and prehybridized in hybridization solution (50% formamide, 5× SSC, 1% SDS, 0.05 mg/mL yeast RNA, 0.05 mg/mL heparin) for 1 h at 65°C. Hybridization was performed for 16 h at 70°C with hybridization solution containing the probe (1 ng/μL). Post-hybridization washes were performed as follows: 3× washes with 50% formamide and 4× SSC for 15 min at 70°C, 3× washes with 50% formamide and 2× SSC for 15 min at 65°C and 3× washes with 0.5 mg/mL levamisol in Tris-buffered saline-Tween-20 (TBST) for 10 min at room temperature. Following the washes, blocking was performed with 1% sheep serum in TBST containing 0.5 mg/mL levamisol for 30 min. Sections were incubated with anti-digoxigenin-alkaline phosphatase (Roche Diagnostics), diluted 1 : 3000 in TBST for 2 h at room temperature, followed by blocking. Sections were visualized using BM Purple AP substrate (Roche Diagnostics), followed by equilibration in NTMT (0.1 M Tris-HCl [pH 9.5], 50 mM MgCl2, 0.1 M NaCl, and 0.1% Tween-20).
Statistical analysis. Paired t-tests were used to analyze differences in LGR5 expression between CRCs and the corresponding colonic mucosal tissues. Unpaired t-tests were used to analyze the association between LGR5 expression and clinicopathological parameters. StatView (version 5.0) software (Abacus Concepts, Berkeley, CA, USA) was used and P-values <0.05 were considered significant.
Overexpression of LGR5 in colon cancer cell lines. To evaluate the expression of LGR5 in CRC compared with other types of cancers, qRT-PCR analysis was performed on a panel of representative cancer cell lines derived from different organs: 11 colon cancer cell lines, five HCC cell lines, 10 ovarian cancer cell lines, and 11 lung cancer cell lines (Fig. 1). LGR5 expression in each of the cell lines was normalized to the average LGR5 expression in the 37 cell lines. LGR5 mRNA was overexpressed in colon cancer cell lines (5/11), whereas LGR5 overexpression was rare in the cell lines derived from other cancers, only occurring in Hep G2 among the HCCs (1/5) and in H889 among the lung cancer cell lines (1/11). Colon cancer cell lines more frequently displayed elevated mRNA levels for LGR5 compared with other cell lines. Elevated expression of LGR5 was observed in all four colon cancer cell lines derived from metastatic sites (LoVo, SW620, COLO 201, and COLO 205) but in only one of seven derived from primary sites (P = 0.0288) (Fig. 1). The mean expression of LGR5 was three times higher in cell lines derived from metastatic sites compared with those derived from primary sites.
Overexpression of LGR5 in colorectal cancers and adenomas. LGR5 expression was assessed by qRT-PCR in surgical specimens from stage II/III CRCs. As shown in Figure 2(a), LGR5 expression was markedly elevated in most cancer tissues compared with normal mucosal tissues. Of the 50 cases, 35 (70%) showed LGR5 expression levels more than three times higher than those of the matched normal mucosal tissues.
LGR5 expression in sporadic adenomas was also investigated. As shown in Figure 2(b), all seven colorectal adenomas showed marked up-regulation of LGR5 compared with normal colonic mucosa. The mean expression of LGR5 in CRCs and adenomas was significantly higher than that in normal mucosa (0.039 ± 0.069 vs. 0.003 ± 0.002, P = 0.0005 and 0.016 ± 0.005 vs. 0.002 ± 0.001, P = 0.0003, respectively) (Fig. 2c). Variation in LGR5 expression was observed among cancer tissues, whereas adenoma tissues and normal mucosal tissues showed comparatively stable LGR5 expression.
Localization of LGR5-expressing cells in colorectal tumors and normal intestinal mucosa. Since a specific antibody to LGR5 for immunohistochemical analysis was not available, ISH was performed to confirm LGR5 expression in CRC cases (Fig. 3a–f). LGR5 mRNA was detected in the tumor cells in the colon cancer cases. However, LGR5 was not expressed in the stromal tissue surrounding the tumors. LGR5 mRNA was diffusely expressed in the entire tumor, including the tumor edge and invasive front (Fig. 3a–c). Similar results were obtained in adenomas (Fig. 3g–i).
In contrast, no clear positive signals were observed in normal mucosal tissue, except for the crypt base cells of the small intestine (Fig. 4c–f) and colon (data not shown). Careful observation showed that crypt base columnar (CBC) cells between Paneth cells expressed LGR5 mRNA (arrowheads, Fig. 4d). Immunohistochemistry on serial sections showed that Ki-67-positive cells were mainly the transamplifying cells and small numbers of CBC cells in the intestinal crypt (Fig. 4g,h). Figure 4(h) shows Ki-67 staining of nuclei in LGR5-positive CBC cells. LGR5-positive cells may have low proliferative activity compared with transamplifying cells. Chromogranin A-positive enteroendocrine cells (arrow, Fig. 4j) were occasionally detected in the crypt base. However, LGR5 positive cells were negative for chromogranin A.
Correlation of LGR5 overexpression with malignant potential. The clinicopathological significance of variable expression of LGR5 in CRC was evaluated. When LGR5 expression normalized to that of matched normal mucosa, LGR5 expression was significantly greater in cases with lymph node metastasis, lymphatic invasion, vascular invasion and pT4 stage tumors (P = 0.0004, P = 0.01, P = 0.0206 and P = 0.0024, respectively) (Fig. 5). Moreover, LGR5 overexpression correlated significantly with the number of lymph node metastases (P < 0.0001), and LGR5 expression in stage IIIC tumors was approximately two-fold higher than in stage IIIB tumors (26.6 ± 24.5 vs. 14.0 ± 13.0, P = 0.0245) (Fig. 6a,b). However, LGR5 expression did not correlate with the age or gender of the patient, or with tumor differentiation, location, or size (Fig. 5).
It is thought that the adenoma–carcinoma sequence underlies the development of CRC, and that the progression of adenoma to carcinoma is a slow process driven by a linear and stepwise series of genetic alterations involving key genes such as APC, KRAS, and TP53.(18) Biallelic mutation of APC, resulting in aberrant activation of Wnt signaling, was observed frequently in CRC.(3) Previous studies showed that LGR5 was overexpressed in BCCs(14) and HCCs, with activation of Wnt signaling by a β-catenin mutation.(9) Moreover, from recent findings, it was suggested that LGR5 may be involved in colorectal carcinogenesis as a Wnt target gene.(7,8) Barker et al.(12) performed lineage-tracking experiments in mice and showed that Lgr5-expressing CBC cells at crypt positions 1–4 fulfilled all the criteria for putative intestinal stem cells. They also showed that selective transformation of Lgr5-enhanced green fluorescent protein-positive stem cells after knockout of Apc, and efficiently induced neoplasia in both the small intestine and colon of mice.(19) The present study has shown that LGR5 was markedly overexpressed in the majority of advanced CRCs (70%) compared with normal mucosal tissue, suggesting that overexpression of LGR5 is a common feature of CRC. Moreover, the ISH signal for LGR5 was detected in the CBC cells, which are thought to be intestinal stem cells, in both small intestine and colon. LGR5 was also detected in adenoma and CRC cells. These findings suggest that LGR5 may be a marker for intestinal stem cells, as well as for colorectal adenoma and carcinoma in humans. Taken together with recent findings,(12,19) the present data suggests that the origin of human CRC may be intestinal stem cells expressing LGR5.
Interestingly, LGR5 expression was higher in colon cancer cell lines derived from metastatic tumors compared with those derived from primary tumors. Therefore, lymph node metastasis positive and negative cases of CRC were compared. Marked variation in LGR5 expression was observed among CRC samples, compared with the samples of normal mucosa and adenoma samples. Increased expression of LGR5 in CRCs was significantly correlated with lymph node metastases of advanced CRCs. These results suggest that LGR5 expression may be associated with the malignant potential of CRC. Since the mechanism of LGR5 involved in colorectal tumorigenesis remains unclear, further studies are needed to elucidate the functions of LGR5. However, we show LGR5 as a promising prognostic marker in management of cancer patients, because patients with lymph node metastases have poor clinical outcomes after curative surgery.(1,2) Thus, the intensity of LGR5 expression could be an indicator of the need for postsurgical adjuvant therapy in CRC patients.
LGR5 was originally isolated as a leucine-rich, orphan G-protein-coupled, seven-transmembrane receptor. LGR5 is a member of the glycoprotein hormone receptor subfamily that includes thyroid-stimulating hormone (TSH), follicle-stimulating hormone (FSH), and luteinizing hormone (LH) receptors.(20,21) The GPCR families are major targets for pharmaceutical development.(22) Thyroid-stimulating hormone (TSH) suppression following thyroidectomy has been reported to reduce the rate of recurrence of thyroid cancer.(23,24) Luteinizing hormone-releasing hormone (LHRH) agonists, whose receptors are also a member of the GPCR family, provide an additional class of agents for the treatment of premenopausal women with hormone receptor-positive breast cancer.(25,26) Luteinizing hormone-releasing hormone (LHRH) agonist therapy has also been the mainstay of treatment for advanced prostate cancer.(27) Orphan GPCRs are potential new drug targets and are currently being investigated for their potential usefulness in the treatment of debilitating diseases, including obesity, cardiovascular disease, inflammation, and cancer.(28–31) The present study suggests that LGR5, a member of the GPCR family, plays an important role in colorectal tumorigenesis, and these findings are supported by recent studies.(7,8,19) Therefore, LGR5 may be a promising therapeutic target for the management of patients with CRC.
In conclusion, LGR5, a putative intestinal stem cell marker, was frequently overexpressed in adenomas and CRCs. These results suggest that LGR5 plays important roles, not only in early events, but also in late events (invasion and metastasis) in colorectal tumorigenesis. Therefore, LGR5 may be a novel prognostic marker and a candidate therapeutic target in CRC.
The authors would like to thank M. Fujiwara, M. Iwata, and C. Fukushima for providing technical assistance, and Dr Miwa Komuta for helpful discussions. This work was supported by a Grant-in-Aid for Scientific Research (B) from the Ministry of Education, Culture, Sports, Science and Technology of Japan; and The Third Term Comprehensive 10-Year Strategy for Cancer Control from the Ministry of Health, Labor and Welfare of Japan.