Background
Cancer metastasis is the process whereby cancer cells spread from the primary tumor to one or more other places in the body. More than 90 % of cancer-associated deaths are directly related to cancer distant metastasis [
1]. In fact, all cancers can form metastatic tumors, including cancers of the blood and the lymphatic system (leukemia, multiple myeloma, and lymphoma). Metastatic tumor cells spread through two major highways—lymphatic vessels and blood vessels—to form secondary foci at common sites such as the lungs, liver, brain, and bones [
2]. The current understanding of cancer metastasis development is mostly derived from mouse models. It is now known that the formation of a metastasis involves a complex molecular cascade through which cancer cells leave the site of the primary tumor (intravasation), enter the blood and/or lymphatic vessels (circulation), and are disseminated to distant anatomical sites (arrest and extravasation), where they can growth of secondary tumors at the target organ site (colonization) [
1]. During this process, a group of enzymes called matrix metalloproteases (MMPs) acts as “molecular scissors,” which are secreted by cancer cells to cut through the proteins that inhibit the movement of migrating cancer cells.
The success rate of metastatic cancer cells forming secondary foci in distant organs is very low, with an estimated rate of 0.01% from primary tumor cells [
3]. The reasons for this low success rate may include the following: (1) cancer cells normally live tightly connected to their neighbors and the meshwork of proteins surrounding them, and any detachment from other cells can lead to cancer cell death (anoikis); (2) cancer cells are often quite large compared with other blood cells, and they are easily damaged or get stuck when traveling through the vessels, which leads to cell death; and (3) highly heterogeneous cancer cells may be recognized and destroyed by cells in the immune system. Although some types of metastatic cancer can be cured with current therapies, most cannot. Therefore, the first priority of these therapies is to shrink the cancer or slow its growth to help relieve cancer-related symptoms.
Circulating tumor cells (CTCs) are cancer cells that have been shed from the vasculature of a primary tumor and circulate in the bloodstream [
4]. Some of these CTCs acquire the capability to extravasate and colonize secondary sites, spreading tumors to distant vital organs and causing the majority of cancer-related deaths. In recent decades, numerous groups have tried to develop new diagnostic assays to detect CTCs in the peripheral blood of tumor patients. By applying these cutting-edge technologies, anti-metastasis therapies for blocking cancer metastasis in patients is now possible. If metastatic cancer cells can be kept dormant, this will transform cancer into a chronic but manageable disease.
With imminent breakthroughs in the recent study of metastasis, three classes of genes have been distinguished—metastasis initiation genes, metastasis progression genes, and metastasis virulence genes—whose gain or loss of function specifically enables tumor cells to circulate, target, penetrate, and colonize distant organs [
5]. Metastasis initiation genes provide an opportunity for primary tumor cells to enter circulation. These genes have cell-motility-, invasion-, and angiogenesis-related abilities that enable tumor cells to target the vasculature in the microenvironment, enter circulation, and be disseminated to distant organs [
6,
7]. Metastasis progression genes, which contribute to primary tumorigenesis, fulfill additional functions that are more advantageous to the metastasis site [
8]. This process acts as a rate-limiting function in primary tumor growth during metastatic colonization. Metastasis virulence genes provide a selective advantage and aggressiveness to secondary colonization sites [
9,
10]. These genes rarely present “poor-prognosis” gene-expression signatures in primary tumors. In addition to these metastasis genes, nearly 30 metastasis suppressors have been identified so far [
11]. The first metastasis suppressor, nm23 protein, was identified in the mid-1980s. Other metastasis suppressors are well known due to their important functions in cell and molecular biology, such as the cadherin family (E-cadherin), caspase-8, stress-activated MAPK signaling (p38), and tissue inhibitors of metalloproteinases (TIMPs). These genes are responsible for blocking tumor metastasis in the metastasis initiation state, and in the future, they may provide a way to develop novel therapeutic agents for cancer metastasis by targeting metastasis suppressors.
So far, more than 200 clinical trials have incorporated CTC counts as a biomarker in patients during metastasis screening of various types of solid tumors [
12], including breast, gastric, and hepatocellular cancers, using density gradient centrifugation, immunomagnetic separation, side population, cell sorting, and further analysis via flow cytometry, reverse transcription polymerase chain reaction (RT-PCR), gene chips, and quantitative PCR (QPCR) [
13‐
15]. QPCR analysis involves the modification of the PCR principle, which preferentially binds to double-stranded DNA, to measure gene expression. Using TaqMan- or SyBGreen-based QPCR platforms has facilitated the understanding of the clinical relevance of the gene expressions of CTCs in both cancer patients and healthy subjects [
16,
17]. For instance, higher cytokeratin-7 (CK7) and epidermal growth factor receptor (EGFR) expressions (4- to 8-fold) in cancer cells have been found in lung and breast cancer patients, whereas normal leukocytes are present in a very low level of expressions [
18]. Thus, quantification of these metastatic-expressing messenger RNAs (mRNAs) is essential in distinguishing normal expression in blood from that with the presence of CTCs.
In 2005, de Kok and colleagues used GUS genes to normalize the variability between clinical tissue samples using QPCR measurements [
19]. In that study, 13 housekeeping gene expressions were measured among 80 epithelial tissue samples, including normal tissue and tissue from colorectal, breast, prostate, skin, and bladder tumors, with different cancer staging, from noninvasive to metastatic carcinomas. The results demonstrated that the expression patterns of hypoxanthine-guanine phosphoribosyl-transferase (HPRT) and the GUS genes were the two most accurate in reflecting the mean expression pattern compared with the other 13 selected genes. The QPCR results showed a very precise accuracy, with ±1.3 and ±1.4 PCR cycles (Ct) in HPRT and GUS normalization, respectively, which indicates that the bias from all clinical tissue samples was less than two times the standard deviation (2-SD).
In this study, we aimed to design a quick and reliable method to monitor human CTCs in a mouse model using both bioluminescent imaging in vivo and QPCR-based analysis. Well-designed primer sets of ß-glucuronidase (GUS) genes for both human and mouse sensitivity were used to detect CTC numbers during cancer metastasis development in orthotropic-developed breast-tumor-bearing mice. In addition, we used the IVIS Spectrum CT System with 3D–imaging (hereafter, IVIS) to clearly illustrate the strong correlation between lung metastasis development and CTC numbers in peripheral mouse blood. The results from this study revealed the molecular basis of CTCs in cancer development, which can be applied as biomarkers to accelerate translational medicine in clinical investigations.
Methods
Cell culture
Human mammary gland epithelial adenocarcinoma cell lines MDA-MB-231 was purchased from the American Tissue Culture Collection (ATCC, Manassas, VA) and maintains in DMEM/F12 medium. The cells were incubated with 10% (v/v) foetal bovine serum (FBS, Biological Industries, Israel), 100 units/ml penicillin and 100 mg/ml streptomycin in a 37 °C incubator with 5.0% CO2.
Transfection and cell line selection
MDA-MB-231 cells were transfected with pcDNA3 plasmids expressing the firefly luciferase gene (the gene sequences were originally from
luc4.1; Chris Contag, Stanford University, Stanford, CA, USA) by electroporation, as described previously [
20]. Briefly, 5 × 10
6 cells were washed twice with PBS and mixed with 10 μg of plasmid. Two pulses were applied for 20 milliseconds under 1.2 kV on the pipette-type MicroPorator MP-100 (Digital Bio, Seoul, Korea). The stable cells were selected 48 h later with G418 (6 mg/mL). The bioluminescent derivatives of MDA-MB- 231 cells were used for further in vivo studies.
Animal experiments
Four-week-old severe combined immunodeficient (SCID) female mice were purchased from the National Science Council Animal Center (Taipei, Taiwan) and housed in micro-isolator cages at the Laboratory Animal Center in National Defense Medical Center (Taipei, Taiwan). This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee (IACUC) at National Defense Medical Center (Permit Number: IACUC-15-240). All surgeries were performed under isoflurane anesthesia and all efforts were made to minimize suffering. During the experiment, no stress or abnormal behaviors due to tumor bearing were observed in the mice. The health status of the animals was monitored once daily by a qualified veterinarian. Food and water were replaced every two days.
Bioluminescent (IVIS) and tumor multimodality (CT/DLIT/FLIT) imaging
Bioluminescent imaging was performed with a highly sensitive, cooled CCD camera mounted in a light-tight specimen box (In Vivo Imaging System - IVIS; Xenogen). For in vivo imaging, animals were given a serial numbers of luciferase stable expressed MDA-MB 231 breast cancer cells (1 X 105, 104, 103, 102 cells and control-PBS only) by tail vein injection. After 15 min, the mice were i.p. injected with D-luciferin (200 mg/kg) for fifteen minutes. Animals were placed onto the warmed stage inside the camera box and received continuous exposure to 2.5% isoflurane to sustain sedation during imaging. Every group of mice was imaged for 30 s. The light emitted from the mice were detected by the IVIS camera system, integrated, digitized, and displayed. Regions of interest from displayed images were identified and were quantified as total photon counts or photons/s using Living Image® software 4.0 (Caliper, Alameda, CA.).
The orthotropic tumor model was used to mimic the cancer in humans through use of immune competent and severe combined immune deficiency (SCID) mice (6–8 week old). Five mice were anesthetized with 2% isoflurane and each implanted with 5 × 106 luciferase expressing MDA-MB-231cells into the mammary fat pad. Five mice injected PBS were presented as control. All the mice were scarified for blood collection after 10-week of cancer cell injection. Throughout the study, all mice were kept in an environmentally controlled room with temperature and relative humidity maintained between 69 and 75 F (21–24 C) and 43–65%, respectively.
Blood samples
100–150 ul of blood was obtained by cardiac puncture from mouse and processed according to standard separation protocols. Total DNA was isolated from human cell lines and mouse leukocyte using AxyPrep blood genomic DNA miniprep kit by following the manufacturer protocol. NanoDrop quantification were used for DNA quantity (260/280) measurement. All DNA samples contained at least 10 ng/ul DNA.
Real-time quantitative PCR
Human GUS primers (forward: AGTGTTCCCTGCTAGAATAGATG and reverse: AAACAGCCTGTTTACTTGAG) and mouse GUS primers (forward: GCAGGCTTTCAAGAGTTCA and reverse: TATGAGCTGGTCCTCCATTTC) were synthesised by Genomics BioSci and Tech (Taipei, Taiwan). A LightCycler thermocycler (Roche Molecular Biochemicals, Mannheim, Germany) was used for QPCR analysis. One microliter of sample and master-mix were first denatured for 10 min at 95 °C and then incubated during 40 cycles: denaturation at 95 °C for 5 s; annealing at 60 °C for 5 s; elongation at 72 °C for 10 s and detected for fluorescent intensity. The PCR samples were all performed melting curve analysis for non-specific PCR product detection. The human GUS fluorescence intensity was measured and normalised to the mouse GUS expression by using the built-in Roche LightCycler Software, Version 4.
Absolute quantitative QPCR
For generate the absolute quantitative standard curve for QPCR analysis. We used PCR product of mouse GUS gene and cloned into TA cloning vector (pTA® Easy Cloning Kit) which purchased from Genomics BioSci and Tech (Taipei, Taiwan). After following the steps of gene sequence, E.coli amplification, plasmid purification and determination of molecular weight, the copy number of GUS gene were calculated and diluted into 108 to 102 per μl. Each copy number of GUS gene were measured with its accuracy and the liner correlation.
Statistical methods
All data were expressed as mean ± SD and performed student t-test analysis for the pairwise samples. All statistical comparisons were performed using the SigmaPlot graphing software (San Jose, CA, USA) and the Statistical Package for the Social Sciences v.13 (SPSS, Chicago, IL, USA). A P-value <0.05 was considered statistically significant and all statistical tests were two-sided.
Discussion
QPCR has been widely used for the detection of CTCs in the peripheral blood of various types of cancers [
21‐
25]. However, the sensitivity and specificity of CTC detection can be extremely variable based on the experimental design, the markers chosen, and the method of cell enrichment. Among these factors, a reliable and detectable CTC marker is one of the most important in determining whether or not the experiment will be a success. To date, reference control genes are the most frequently used method of normalizing the mRNA concentration in samples. These reference control genes are often referred to as “housekeeping genes,” which remain unchanged in the tissues or cells under investigation. In our experimental design, the GUS gene was selected to identify the cells of humans and mice. GUS degrades glycosaminoglycans, including heparan sulfate, dermatan sulfate, and chondroitin-4,6-sulfate. In molecular biology, the GUS reporter gene system is a powerful tool that is often used for the assessment of gene activity in mammalian and plant cells. Therefore, monitoring β-glucuronidase activity through the use of a GUS assay can determine the spatial and temporal expression of the gene in question [
26]. GUS genes are essential for cell biology and constitutive action, and an increasing number of studies have used them for mRNA normalization.
The results from the two experimental mRNA and DNA detections in the target genes using QPCR analysis showed varying levels of CTCs in the blood samples, and each had advantages and limitations according to their expression and stability. Taking mRNA as an example, due to its instability, the reverse transcription process was required to transform mRNA into complementary DNA (cDNA) under PCR experimentation, specifically, the denaturation step of PCR. When a constitutive-expression housekeeping gene was selected, the cDNA copy number included an abundance of target genes. However, trying to detect a high similarity of genes in two samples (e.g., the human and mouse GUS genes in this study) was a difficult task when designing primers. By comparison, the gene sequences of GUS in the human and mouse mRNA in the coding regions showed 77.8% of shared identity, which resulted in difficulty in finding primer sets that would precisely measure one gene expression without disturbing the other. On the other hand, high stability is one of the best-known advantages of using DNA as a template because CTCs can be measured using purified DNA or enriched leukocyte during QPCR analysis.
The long sequence of DNA (including exon and intron) also provides more possibilities and easier conditions for designing suitable PCR primers for target genes in different species. Except for gene duplication (also referred to as gene amplification), a significant drawback of using DNA as a template for cell quantification is its lower sensitivity during QPCR measurement, compared with gene expression detection. This results in researchers using more blood or leukocyte taken from the mouse, or adding more amplification cycles during PCR analysis to overcome this natural defect. Thus, using mRNA reverse transcription cDNA as the source template and having well-designed target primer sets is a better strategy when evaluating the cell numbers of human CTCs in peripheral mouse blood using QPCR analysis.
Human xenograft tumors implanted in immunocompromised mice provide an important approach for the assessment of tumor growth, invasion, metastasis, angiogenesis, and the effects of the tumor’s microenvironment [
27]. However, it is difficult to investigate CTCs numbers in xenograft tumor models during cancer development. In recent decades, using bioluminescence and fluorescence has made it easier to address these issues, especially in xenograft models. Moreover, using an IVIS provides a method that facilitates and enhances the quantification of tumor progression and treatment efficacy by measuring the progression in the bioluminescence or fluorescence associated with tumor growth. In this study, we found that the luciferase measurement from the IVIS showed a linear and trustworthy correlation with the number of CTCs. In addition, the metastatic cancer that developed in the secondary tissue infiltration was also detected by the IVIS through its 3D–images of blood-enriched organs.
Finally, using QPCR-based primer sets to quantify the cell numbers of the CTCs in the xenograft mouse model with human breast cancer cells, we compared both bioluminescent imaging and QPCR measurements in the mice’s bloodstream to examine the replicability and reliability of the whole system. The data from the orthotropic breast tumor animal model illustrated the most direct and trustworthy evidence in this preclinical study, which showed a positive correlation between the metastasis event and the CTC numbers. This technique lays a strong foundation for future studies to examine therapeutic responses and answer biological questions involving CTC-associated molecules and signaling pathways during cancer metastasis. However, the limitation for this study is that only one type of human breast cancer cell was tested in animals. Ideally, CTC detection using realtime PCR in preclinical animal should involve different types of cancer cells to ensure human GUS expression can be precisely measured in mouse blood. Furthermore, the amount of blood collected from mice could be little, from volume 50–500 ul in submandibular blood collection or tail-vein blood collection. This would cause less sensitivity of CTC detection in animal model. In clinic, to overcome this limitation, all CTC related clinical trials have used cell enrichment as an essential and necessary step to increase the sensitivity of CTC detection.
The results from the current investigation should facilitate the discovery of novel therapeutic targets and the development of specific inhibitors and drugs for clinical practice. We expect that this preclinical system involving the IVIS and QPCR analysis will accelerate the biological profiling of CTCs, which will largely improve the diagnostic capabilities used in clinical oncology. In this study, we used breast cancer as a target for CTCs; however, this technique should also be applicable to mouse models for other human cancers, with important implications for cancer metastasis therapy.
Acknowledgements
We would like to acknowledge both the Instrument Center of Taipei Medical University and the National Defense Medical Center for the use of the real-time PCR and the IVIS SpectrumCT In Vivo Imaging System, especially Miss Po-Li Chen and Pei-Yeh Lee for their technical support.