This article is mentioned in:

Introduction

The circadian gene network sets and regulates the circadian rhythms in organisms as oscillations with an approximate 24 h cycle (1). At the molecular level, circadian rhythms are controlled by the interaction between positive and negative feedback loops consisting of several key clock regulators (2,3). A model encompassing a feedback system involving heterodimer transcriptional factors (Clock and Bmal1), two cryptochromes (Cry1 and Cry2), and three Period (Per1, Per2, and Per3) regulators has been widely described. Among these clock genes, hClock is at the core of the circadian gene family.

Previous studies have revealed that the disruption of circadian rhythms is one of the endogenous factors that contributes to the initiation and development of colorectal carcinoma (CRC). Krugluger et al reported that the expression of hPer1 mRNA in CRCs was lower than that found in normal cells (4). Wang et al detected that the expression of the hPer2 protein in CRC cells was lower than that in non-cancerous cells (5). Correlations between the decreased hPer2 level and the patient’s age, histological grade and TNM stage were also detected. The human Clock (hClock) gene functions upstream of hPer2 to regulate its expression (5). Therefore, the circadian hClock gene may play a vital role in the development of CRC.

Hoffman et al observed that the hCLOCK protein expression level was approximately two times higher in breast tumor cells compared with that observed in the adjacent normal tissues and three times higher than that in normal breast cells (6). Alhopuro et al reported that ~53% of hClock genes in CRC cells had point mutations in the DNA sequence (7). Oshima et al demonstrated that the expression level of hClock mRNA in CRC cells was significantly higher compared with that found in normal tissues (8).

CRC is one of the most commonly observed malignancies and is the leading cause of cancer-associated mortality worldwide (9). In 2011, ~141,210 new CRC cases and 49,380 mortalities were predicted in the USA (10). In China, CRC is the fourth leading cause of cancer mortality in large cities and the fifth in more rural areas. However, in Shanghai, the incidence of CRC and its mortality rate ranked second and third, respectively, for females (11). As surgical approaches and conventional therapeutics have been incapable of fully controlling the outcomes of CRC, there is an urgent need to develop more effective treatments.

The circadian rhythm is interconnected with many aspects of cellular function, such as cell proliferation, migration and differentiation, and it also helps regulate the digestive system (12,13). The molecular mechanism governing the development of CRC is a multiple factor and multistep process, yet the molecular changes underlying it remain unexplored. The core circadian gene, hClock, may participate in the development of CRC, but the correlation between them has not been reported.

In this study, immunofluorescent histochemistry and quantitative polymerase chain reaction (qPCR) were used to characterize the role of hClock in the development of human CRC. The correlation between the hClock gene and tumor-related genes Bak, Bax, Bid, HIF-1α, ARNT, tumor necrosis factor receptor I (TNFR I), TNFR II and VEGF were also examined. Based on our data, the expression of hClock in human CRCs and its possible effects were discussed. Our investigation may provide a new angle for understanding tumorigenesis and a new target for CRC treatment development.

Materials and methods

Tissue samples

In total, 30 specimens were obtained from the Tumor Tissue Bank of the Center of Pathological Research in the Institutes of Biomedical Sciences at Fudan University (Shanghai, China). The specimens were obtained from patients who had undergone a radical CRC resection between January 2007 and September 2008 from the General Surgery Department of Huashan Hospital Affiliated to Fudan University (Shanghai, China). The specimens were all obtained between 10:00 am–14:00 pm. The cancerous tissue was dissected from the cancerous mucosa while the paired non-cancerous tissue was obtained from the normal section of the colorectal mucosa of the same patient. All tumor tissues were pathologically confirmed as CRC and the patients did not receive any presurgical treatment nor exhibited any metastasis. Following surgical removal, samples were immediately preserved in liquid nitrogen and stored at −80°C. Complete medical records including the patient’s age, tumor site, tumor type, histological grade and Dukes’ stage were also obtained. This study was approved by the Medical Ethics Committee of Huashan Hospital, Fudan University, Shanghai, China. Written informed consent was obtained from the patients.

Immunofluorescent histochemistry

Frozen tissues were cut into 6 μm sections and stored at −80°C. Sections were fixed at 4°C with 70% ethanol solution and blocked by 10% normal goat serum solution at 37°C for 60 min. Subsequently, sections were incubated with goat anti-human hClock antibody (1:50, Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) as the primary antibody at 4°C overnight. Following a brief wash with PBS, secondary donkey anti-goat IgG fluorescent antibody (1:200; Invitrogen Life Technologies, Carlsbad, CA, USA) was added for 2 h at room temperature. Following several washes with PBS, the slides were mounted. Sections which were only incubated with 10% normal goat serum solution, served as negative controls. The image analysis software Image-Pro Plus (Media Cybernetics, Rockville, MD, USA) was used to measure the Integrated Optical Density (IOD), which is the cumulative OD of the positive immunofluorescent expression of a selected area and represents the total protein expression of that area.

RNA extraction and first-strand cDNA synthesis

Preparation of total RNA from tissue samples was performed using the TRIzol reagent (Invitrogen Life Technologies). The quantity of total RNA was determined using UV spectrophotometry and RNA integrity was assessed using agarose gel electrophoresis. First-strand cDNA was prepared with oligo-dT primers using a commercial cDNA synthesis kit (ReverTra Ace qPCR RT kit, Toyobo, Osaka, Japan). The cDNA was subsequently amplified for 33 cycles with specific primers for Bak, Bax, Bid, HIF-1α, ARNT, TNFR I, TNFR II and VEGF. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) served as an internal reference (Table I). Sequence data were analyzed using the basic local alignment search tool (BLAST) on the National Centre for Biotechnology Information (NCBI) website (http://www.ncbi.nlm.nih.gov).

Table I Primer pairs used for qPCR.

Table I

Primer pairs used for qPCR.

Gene	Primer (5′-3′)	Product length (bp)
hClock	Forward: GGCTGAAAGACGACGAGAAC Reverse: GGTGTTGAGGAAGGGTCTGA	210
ARNT	Forward: CAAGCCCCTTGAGAAGTCAG Reverse: GGGGTAGGAGGGAATGTGTT	165
Bak	Forward: GGGTCTATGTTCCCCAGGAT Reverse: AATGGGAGCAGGGGTAGAGT	165
Bax	Forward: TCTGACGGCAACTTCAACTG Reverse: TTGAGGAGTCTCACCCAACC	188
Bid	Forward: GAGGATTGTGGCCTTCTTTG Reverse: CTCAGCCCAGACTCACATCA	185
HIF-1α	Forward: GAAAGCGCAAGTCCTCAAAG Reverse: TGGGTAGGAGATGGAGATGC	167
TNFR I	Forward: GTGCCTACCCCAGATTGAGA Reverse: TGTCGATTTCCCACAAACAA	175
TNFR II	Forward: GAGTTGTGGAAAGCCTCTGC Reverse: CAGGGGGAGAAACAAACAAA	193
VEGF	Forward: AAGGAGGAGGGCAGAATCAT Reverse: ATCTGCATGGTGATGTTGGA	226
GAPDH	Forward: ACAGCCGCATCTTCTTGTGCAGTA Reverse: GGCCTTGACTGTGCCGTGAATTTT	226

[i] qPCR, quantitative polymerase chain reaction; TNFR, tumor necrosis factor receptor.

PCR reactions containing SYBR-green were amplified on a Corbett Real Time PCR machine (Bio-Rad, Hercules, CA, USA). Following reverse transcription, the cDNA was diluted to 10 ng/μl. The PCR reaction mixture contained 4 μl of diluted cDNA, 1 μl of 10 μM primers (0.5 μl of upstream primers, 0.5 μl of downstream primers), 10 μl SYBR Green Realtime PCR Master mix (Bio-Rad) and 5 μl of H2O, into a final volume of 20 μl. The samples were subjected to 40 cycles of amplification (denaturation at 95°C for 15 sec, primer annealing for 30 sec and extension at 72°C for 30 sec).

Following each qPCR, melting profiles were performed in order to rule out the possibilities of nonspecific PCR products and primer dimers. Data from the qPCR was represented by fluorescent intensity, ΔRn, and number of cycles. ΔRn=Rn+ − Rn−, Rn+ refers to the fluorescent intensity of products at any time, Rn− refers to the systematic basic fluorescent intensity (the fluorescent intensity at the 6–15th cycles). The threshold was set to the intermediate value of logΔRn vs. cycle function. The CT value stands for the number of cycles at the cross point of ΔRn vs. threshold. The higher the CT value, the greater the number of steady cycles, thus the lower the gene expression concentration. As the amplification efficiencies of the target genes and the internal control were equal, the relative changes to the target gene expression in the tumor cells compared with that in normal colorectal mucosa (ΔCT calibrator value) were calculated using the equation 2−ΔΔCT, where ΔΔCT = ΔCT(non-cancer) − ΔCT(cancer). The ΔCT values were determined by subtracting the average GAPDH CT value from the average target gene CT value.

Statistical analysis

Data are reported as the means ± SD. Comparison between the groups was performed using the paired Student’s t-test. Relative analysis was conducted using Pearson’s χ2 test. IBM SPSS Statistics 19.0 was used for all data analysis. P<0.01 was considered to indicate a statistically significant difference.

Results

Immunofluorescent histochemical analysis of hCLOCK protein expression

To analyze expression of the hClock gene in CRC, we first examined the expression of hCLOCK protein in 30 paired colorectal cancerous and non-cancerous tissues using immunofluorescent histochemistry. For these 30 cases, hCLOCK protein expression in colorectal tissues exhibited three typical patterns. Type I (11 cases): the immunofluorescence was markedly stronger in the cancerous cells compared with that in the paired non-cancerous cells, particularly in the nucleolus and cytoplasm. Type II (17 cases): there were no significant differences in the immunofluorescence between the cancerous and non-cancerous cells. Type III (2 cases): the immunofluorescence was stronger in the non-cancerous cells compared with that in the paired cancerous cells (Fig. 1).

Figure 1 Immunofluorescent histochemistry of hCLOCK protein in human colorectal cancers. (A) Cancer tissue, (magnification, ×40). (B) Paired normal mucosa tissue (magnification, ×40).

In these 30 specimens, the average cumulative OD of the hCLOCK protein in cancerous tissues was 5389.59±2385.20 and that in paired non-cancerous tissues was 4558.34±1385.87, P<0.05 (P=0.034). These results revealed that the hCLOCK protein was expressed in all specimens and a higher expression level of hCLOCK protein was found in human CRCs compared with that in the paired non-cancerous tissues.

Correlation between hCLOCK protein expression and clinical-pathological features in CRC patients

In order to analyze the clinical and pathological relevance of high hCLOCK protein expression in human CRC, we assessed the correlation between hCLOCK expression and the respective clinical-pathological parameters in the 30 cases of CRC. As shown in Table II, an increased expression of hCLOCK protein was found more in male than female patients (P=0.039). We observed a higher expression of hCLOCK in poorly differentiated (P=0.042), or late-stage, Dukes’ grade (P=0.032) tumors. Furthermore, 64.3% of tumors with lymph node metastasis expressed higher hCLOCK levels. However, no correlations were found between high levels of hCLOCK expression and age, tumor site or pathology type.

Table II Correlation between increased hCLOCK protein expression levels in CRC and clinical-pathological features.

Table II

Correlation between increased hCLOCK protein expression levels in CRC and clinical-pathological features.

Clinical-pathological features	Increased hCLOCK expression in tumor	Non-increased hCLOCK expression in tumor	P-value
Gender			0.039
Male	8	11
Female	3	8
Tumor site			>0.05
Colon	7	11
Rectum	4	8
Pathology type			>0.05
Ulcer	6	11
Protrude	5	6
Infiltrating	0	2
Histological grade			0.042
I		3
I–II, II	7	13
II–III, III	4	3
Dukes’ stage			0.032
0	0	1
A	0	4
B	4	7
C	7	7

[i] hCLOCK, human CLOCK, CRC, colorectal carcinoma.

Determination of hCLOCK mRNA level by qPCR

We further analyzed hClock mRNA levels in the 30 paired cancerous and non-cancerous tissues by qPCR. The quantity of hClock mRNA was normalized to the endogenous reference gene, GAPDH. The normalized hClock mRNA expression (ΔCt) of tumorous tissues was subsequently compared with the ΔCt of the paired non-tumorous tissues from the same CRC patient in order to determine their relative expression levels (ΔΔCt) (Fig. 2). The ΔCt of hClock in tumorous tissues (5.91±1.36) was significantly lower than that found in paired non-tumorous tissues (6.26±1.47; P=0.002), indicating a significantly higher level of hClock expression in human CRCs than in paired non-cancerous tissues.

Figure 2 Distribution of quantitative polymerase chain reaction ΔΔCt of hClock mRNA in tumor and paired non-tumor tissues.

Amongst the 30 specimens, 12 tumor tissues demonstrated a high expression of the hClock gene. Furthermore, in these 12 tumor tissues, 10 specimens also demonstrated a higher expression of the hCLOCK protein. The correlation coefficient (r) between hClock mRNA and the hCLOCK protein was 0.474, P<0.01 (P=0.008). The expression of the hClock gene and the hCLOCK protein revealed a strong positive linear correlation in human CRCs.

Correlation between the hClock gene and tumor-related genes

To further investigate the potential role of hClock at the molecular level, the correlation between the hClock gene and a number of tumor-related genes were determined by qPCR. As shown in Table III, there was a strong positive linear correlation between the hClock gene and the ARNT, HIF-1α and VEGF genes in human CRCs. There was no significant correlation between hClock with Bak, Bax, Bid, TNFR I and TNFR II. The higher expression of ARNT, HIF-1α and VEGF in poorly differentiated, or late-stage, Dukes’ grade tumors was observed. Furthermore, high VEGF expression was found in 14 tumor cases with lymph node metastasis.

Table III Correlation between the hClock gene and other tumor-related genes.

Table III

Correlation between the hClock gene and other tumor-related genes.

Statistical significance	Bak	Bax	Bid	HIF-1α	ARNT	TNFR I	TNFR II	VEGF
r	0.114	0.005	−0.272	0.543	0.608	0.370	−0.280	0.714
P-value	0.548	0.980	0.415	0.046	0.027	0.068	0.083	0.008

Discussion

It has been demonstrated that the circadian clock has important roles in orchestrating temporal integrations of physiology and behavior with the environment. This master clock, which resides in the hypothalamic suprachiasmatic nucleus in mammals, synchronizes multiple peripheral oscillators to ensure temporally coordinated physiology (1,14,15). Clocks located outside the brain, peripheral clocks, are important in regulating physiological functions and pathological situations.

Previous studies have suggested that the disruption of circadian rhythms, which are regulated by circadian genes, was one of the endogenous factors that contribute to tumor initiation and progression. Studies have reported evidence regarding the beneficial effects of chronotherapy, which refers to the delivery of chemotherapy according to the circadian rhythms (16,17). The hClock gene is at the core of the circadian gene family and directly, or indirectly, regulates certain clock-controlled genes (18,19). The role of hClock in human CRCs is unclear. In this study, we aimed to investigate this question directly. We first examined the protein expression of hCLOCK in CRC patients using immunofluorescent histochemistry. The results revealed that hCLOCK was expressed in all CRC specimens and there was a higher expression of hCLOCK in human CRC tissues compared with that found in the paired non-cancerous tissues (P=0.034).

We further analyzed hCLOCK expression patterns with the respective clinical-pathological features of these colorectal tumor tissues. We found that cancerous tissues with poorly differentiated (grade II or above), or late-stage, Dukes’ grade tumors (stage B or C) demonstrated increased expression of hCLOCK protein. The greater expression of hCLOCK in the CRC tissues of male patients compared with that in paired non-cancerous tissues was observed, but not in female patients. This may be further validated by using a larger sample set in the future.

qPCR analyses revealed that hClock was expressed in all of the specimens. A significantly higher level of hClock expression was found in human CRC compared with that in the paired non-cancerous tissue (P=0.002). hClock had a positive linear correlation with the hCLOCK protein in human CRCs (r=0.474, P=0.008). A higher expression of hClock was prevalent in tumor malignancies of a higher degree. These results suggested that the hClock is closely related to the development and prognosis of human CRCs.

We hypothesized that the circadian gene hClock may be a proto-oncogene. A proto-oncogene is usually a normal gene with rhythm, which is overexpressed and subsequently loses its rhythmic character in tumor cells. Hoffman et al observed deregulation of tumor-related genes in the hClock-silent cell strain, and that the protein hCLOCK expression in breast tumor cells was significantly higher compared with that in the adjacent normal tissues (6). Fu et al reported that c-myc in the tumorous cells of mice was regulated by circadian genes (20). The hClock gene correlated with cell cycles and had the characteristics of rhythmicity and specificity (21). This hypothesis requires further validation by examination with a larger sample set.

Circadian genes are important in regulating certain downstream clock-controlled genes (CCGs), including a variety of tumor-related genes (22). In our study, there was a significant positive linear correlation between the hClock gene and the ARNT, HIF-1α and VEGF genes in human CRCs. There was no significant correlation between hClock and Bak, Bax, Bid, TNFR I or TNFR II. Malignant tumors promote the secretion of various proangiogenic factors, of which the most important is VEGF. VEGF has a marked proangiogenic function and is closely related to tumor differentiation, Dukes’ stage tumors and tumors with lymph node metastasis.

When tumors were under environmental hypoxic conditions, the heterodimer HIF-1α/ARNT would be overexpressed, resulting in a higher VEGF expression. hClock may interact with HIF-1α/ARNT and activate VEGF to stimulate tumor angiogenesis and metastasis (23,24).

In conclusion, our data suggest that aberrant expression of hClock may correlate with the development and progression of CRCs. However, the mechanisms leading to a high expression of hClock in CRCs and how this promotes tumorigenesis remains unclear. Further studies using a larger sample set are required in order to elucidate the role of hClock in human CRCs.

Acknowledgements

This study was supported by the Natural Science Foundation of China (81070234 and 81000355).

References