Recent evidences show that nervous system acts as a crucial part of cancer microenvironment. Infiltration of nerve fibers into cancer microenvironment has an important active role in cancer progression. The stimulations of both cancer growth and metastasis by members of nervous system such as neurons and glial cells have been demonstrated. However, how the nervous system is built in cancer is largely unknown. Here we show that a fraction of cancer stem cells (CSCs) derived from patients with gastric carcinoma and colorectal carcinoma are capable of producing neurons that are involved in tumor neurogenesis and tumor growth. Cancer stem cell monoclone derived from a single cancer stem cell was able to generate neurons including sympathetic and parasympathetic neurons to take part in the nervous system in cancer tissues. Knocking down the neural cell generating capability of the human CSCs inhibited the growth of xenograft tumors in mouse model. Our data demonstrate that human CSCs are able to produce one of most important components in the cancer microenvironment that are required for cancer development and progression.
The observations on the association between cancer and nervous system can be traced back to early years of ninteenth century.1 Nerves have an important role in tumor growth, cancer invasion and even metastasis and are considered to be components of cancer microenvironment.2 A process termed perineural invasion that cancer cells can grow around and eventually invade existing nerves has been observed in many kinds of cancers and is generally associated with poor survival and prognosis.3,
Materials and methods
Cancer stem cell isolation and culture
Tumor surgical specimens were collected in accordance with a protocol approved by the West China Hospital of Sichuan University Institutional Ethics Committee. Informed consent was obtained from all patients. Colorectal cancer stem cell and gastric cancer stem cell were derived from colorectal and gastric adenocarcinoma tumors and functionally validated as described previously.12,13 In in vitro differentiation assays, cells were seeded on coverclips pretreated with Matrigel Matrix Growth factor reduced (Corning, Bedford, MA, USA) and induced to differentiate in Dulbecco's modified Eagle's medium medium containing 2% fetal bovine serum and B27 (Thermo) with vitamin A. Following shRNAs were used and the corresponding lentiviruses were from Genepharma (Shanghai, China): Microtubule Associated Protein 2 (MAP2) shRNA1 ( 5′-GCGCCAATGGATTCCCATACA-3′), MAP2 shRNA2 (5′- GCACCTGACCTTCCTGAAATG-3′) and control shRNA ( 5′-TTCTCCGAACGTGTCACGT-3′).
MAP2 promoter-driven expression of ZsGreen
Human MAP2 promoter (1487 bp)14 was cloned by PCR and confirmed by sequencing. The promoter was inserted into pLVX-IRES-ZsGreen1-EF-puro lentiviral vector to replace the original CMV promoter. Lentiviruses were produced and tittered as described elsewhere.15
Coverclips and frozen sections were fixed with 4% paraformaldehyde or methanol/acetone. In experiments that paraformaldehyde was used for fixation permeablization was performed with 0.5 to 1% Trion X-100. After blocked with 5% bovine serum albumin in PBS-Tween for 1 h, fixed cells or frozen sections were incubated with primary antibodies overnight at 4 °C in PBS-Tween with 3% bovine serum albumin. The primary antibodies used were: Beta-3-tublin (Chicken, Novus, Littleton, CO, USA nb100-1612), NuMA (Rabbit, Abcam, Cambridge, MA, USA ab84680), NuMA (Goat, Santa-Cruz, Dallas, TX, USA sc-18557), MAP2 (Rabbit, Santa-Cruz sc-20172), CDX2 (Mouse, Origene, Beijing, China TA500251), CK20 (Rabbit, Abcam ab-76126), TH (Chicken, Abnova, Taipei City, China PAB29094), Vacht (Rabbit, Sigma, St Louis, MO, USA SAB4200559), SV2 (Goat, Santa-Cruz sc-11936), Synapsin I (Rabbit, Abcam ab-64581). Secondary antibodies specific to the appropriate species were used (1:500; Jackson ImmunoResearch Laboratories, West Grove, PA, USA & Thermo-Fisher, Waltham, MA, USA). All immunofluorescent staining results of in vitro cultured cell shown in this article were replicated for more than five times. All of the immunofluorescent staining results of frozen sections of xenografts were replicated more than 3 times.
Animal experiments were performed as described previously.12,13 In detail, male or female nude mice (BALB/C strain), 4–6 weeks old, were purchased from the Beijing Experimental Animal Center of the Chinese Academy of Sciences (Beijing, China). Mice in this study were housed under pathogen-free conditions, and all animal studies were carried out according to the animal protocol approved by the Sichuan University Institutional Animal Care and Use Committee. In all experiments, a small aliquot of cells was set aside to confirm cell counts and viability using conventional techniques (that is, Trypan blue exclusion) or 7-AAD staining. Once cell counts and viability were confirmed, cells were diluted to appropriate injection doses for intraperitoneal or subcutaneous injection. In intraperitoneal injection, cells were suspended in PBS and injected into mice intraperitonealy. In subcutaneous injection, cells were mixed with Matrigel (Corning) at a 1:1 ratio, and injected subcutaneously in nude mice on the ventral wall. No randomization or blinding techniques were applied in this study. Injected mice were killed when the established criteria for end-stage diseases were reached.
RNA extraction and real-time quantitative PCR
Total RNA of cells was extracted with a Trizol reagent kit (Takara Biotechnology Co., Ltd. Dalian, China) according to the manufacturer’s protocol. Subsequently, reverse-transcription of RNA by PCR was performed using a Takara SYBR real-time PCR kit for target gene.
Data were statistically analyzed. For two experimental comparisons, a two-tailed unpaired Student’s t-test was used. The difference between experimental groups was assessed by one-way analysis of variance testing. When cells were used for experiments, three replicates per treatment were chosen as an initial samplesize. All n values defined in the legends refer to biological replicates. If technical failures such as inadequate subcutaneous or intraperitoneal injection occurred before collection, those samples were excluded from the final analysis.
The authors declare that the data supporting the findings of this study are available within the paper.
Neural cells with human-specific marker appear in tumor xenografts in mice
In order to determine the involvement of neural system in the tumors that were generated from human cancer stem cells (CSC) in vivo, we transplanted the CSCs that were previously isolated from the patients with gastric and colorectal carcinoma12,13 into nude mice via subcutaneous and intraperitoneal injections to produce human cancer xenografts. Tumor innervations were observed. The neural staining showed that many ganglia generated from mouse nerve system infiltrated into the interstitial tissues in the tumor masses (Figure 1a). The IF staining of frozen sections of intraperitoneal tumor xenografts showed a significant fraction of neural cells carried the human-cell-specific nuclear antigen nuclear mitotic apparatus protein (NuMA).16 The neural cells with human origin were distributed in the ganglia closed to tumor masses and in tumor tissues (Figures 1b–d). In addition, the sections of subcutaneous xenografts were detected fewer neural cells with human markers in the ganglia, in comparison with intraperitoneal tumor xenografts. The results suggest that gastric CSCs and colorectal CSCs may have the capacity to differentiate into neural cells to generate nervous system in tumor tissues in tumor xenografts.
CSC monoclones derived from single human cancer stem cells can generate neural cells in vitro
To determine the neural differentiation potential of gastric and colorectal CSCs, Cultured gastric and colorectal CSCs were induced to differentiate for more than 7 days. Neural cell foci among the epithelial cancer cells were detected in IF staining of differentiated CSCs (Figure 2A). To verify the capacity of CSCs to differentiate into neurons, gastric and colorectal CSC monoclones were generated from single cancer stem cells. The CSC monoclones were then differentiated with serum induction. IF staining demonstrated that the CSC monoclones did generate neural cells. We also examined the undifferentiated CSC spheres and did not observe neural cells (Figure 2B and data not shown). Then, the gastric and colorectal CSC monoclones were induced to differentiate by medium optimized for neural stem cell to produce neuron for further examination of their differentiation potential. The capabilities to produce neural cells of different CSC monoclones are different (Figure 2C), suggesting that CSCs derived from the patient samples are heterogeneous and only a fraction of CSCs are able to give rise to neural cells. The capacity of the CSC monoclones to form tumor xenografts in nude mice was verified by subcutaneous and intraperitoneal implantation (data not shown). These data ruled out the possibility that our results were caused by cross neural cell contamination when we isolated the CSCs from the patient samples and confirmed that gastric and colorectal CSCs are able to produce neural cells.
To provide further evidence validating neural differentiation capacity of the gastric and colorectal CSCs, we cloned the human MAP2 promoter and generated lentiviral construct for MAP2 promoter-driven zsGreen expression (Figure 3A) and determined the zsGreen expression of transduced CSCs in vitro and in tumor xenografts in mice. IF staining of the cultured differentiated CSC showed the cells expressing zsGreen were positive for MAP2 (Figure 3B). In addition, zsGreen-positive cells in frozen section of mouse celiac xenograft were detected (Figure 3C). These results provided evidence supporting our discovery that the gastric and colorectal CSCs carry neural differentiation potential.
Human cancer stem cells can produce functional neurons
We examined the potential to form synapses, one of the most important structures of functional neurons,17 of the neural cells derived from the gastric and colorectal CSCs. A fraction of differentiated CSCs were stained positive for synapse markers (Figures 4a and b), suggesting that the human CSCs are able to produce functional neurons. Previous evidence has demonstrated that the autonomic neural system contributes to cancer progression.9,10,18,
Knocking down the neural generating ability of human cancer stem cells decreases the tumor growth in vivo
Recent data have revealed that the crosstalks between cancer cells and nervous system are critical for tumorigenesis, tumor growth and tumor metastasis.2,22,23 After knowing that the human CSCs were able to produce neurons, we speculated that CSC derived neurons may have the ability to promote tumor progression. Accordingly, we ablated the neural generating capabilities of the human CSCs by MAP2 RNAi (5′ end RNAi: shRNA1 and 3′ end RNAi: shRNA2; Figure 6A–C) to explore the involvement of neurons derived from the human CSCs in tumor progression. The detection of MAP2 mRNA showed that both 5′ end and 3′ end shRNAs markedly decreased the level of MAP2 mRNA. The ethynyl deoxyuridine incorporation of the CSCs were not affected by MAP2 knock-down, suggesting that the CSC neural differentiation potential are not related to the proliferation in vitro (data not shown). We also did not detect increased apoptosis or autophagy in in vitro differentiated cells after knocking down MAP2. In contrast, the capacity of colorectal cancer stem cells to give rise to neural cells and the TH-positive cells was remarkably reduced by MAP2 knock-down (Figure 6B and C). The shRNA1 was more effective in reducing the neural differentiation capacity of CSCs in culture. Consistently, the tumorigenesis capacity of the human CSCs and the tumor growth in nude mice were markedly reduced by MAP2 shRNA1 and slightly knocked down by MAP2 shRNA2 (Figure 7). The results demonstrated that the human CSC derived neural cells have an important role in tumorigenesis and promote the tumor growth.
The supporting tissues in tumor microenvironment, including blood vessels, connection tissues and fibroblast, are indispensable for all the steps, such as tumorigenesis, growth, progression and metastasis, of cancer development in human.2,17,24,25 Recent studies have led to the conclusion that nerves are an important component of cancer microenvironment.2,23,26 Autonomic nerves especially sympathetic nerves have a significant role in the progression of cancers such as prostate cancer and gastric cancer.9,11 Like prostate cancer, colorectal cancer has an environment, which is rich in autonomic nerve fibers, and the presence of nerve fibers is associated with poor prognosis.27 Here we show that xenograft tumors derived from human colorectal cancer stem cell were infiltrated by host nude mice nerve fibers and this demonstrate that colorectal cancer cells can interact with nerve fibers just like prostate cancer cells. Our result suggests that crosstalk between neural system and cancer cells are common among solid malignant tumors. Nerve infiltration is observed in human cancer xenograft tumors formed subcutaneously and intraperitonealy, suggesting that the crosstalk between cancer cells and nerve fibers can happen in different tumor initiation environment.
As shown here, both human gastric and colorectal cancer stem cells are able to give rise to neurons when transplanted intraperitonealy in nude mice. This capacity has also been verified in vitro and has been confirmed both in vivo and in vitro by monoclonal cancer stem cells. Cancer stem cells are known to be able to give rise to component of tumor microenvironment such as endothelial cells and pericytes.28,
We also identified that when the isolated CSCs are subcultured for >30 passages, they will gradually lose the neural differentiation capability (data not shown). Furthermore, we initiated to test the cancer stem cells derived from the patients with lung adenocarcinoma. The preliminary results showed that lung cancer stem cells also generated neural cells (data not shown).
Previous data suggested that β-blockers may help to improve survival and prognosis of cancer patients.31,
The institutional Ethics Committees of West China hospital approved this study. We thank the participants and their families for their kind cooperation, generosity, and patience. We appreciate Professor Dr Anming Meng for suggestion and advice and Dr Tie Chen for the help on the culture of gastric cancer stem cell. This work was supported by the National Basic Research Program of China (to XM 2015CB942800), the Nature Science Foundation of China (to XM 81361120381; to CF 81402446).
XM designed and supervised the whole work. SZ and YS initiated the experiments. RL and CF, helped by WS, YL, YL and DY performed the cell culture and animal work. QH, XL and WM performed the FACS assays and sorting. ZZ designed and supervised the clinic program of colorectal cancer. JH designed and supervised the clinic program of gastric cancer. WL designed and supervised the clinic program of lung cancer. LL designed and supervised the surgery program of lung cancer. ZZ, JH, WL and LL performed the clinic observation, clinic data collection, patient sample collection. XM helped by HX analyzed the data. XM and RL, wrote the manuscript
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/