Beyond Neurons: Long Distance Communication in Development and Cancer

Cancer cells communicate with many cell types in their surrounding parenchyma as a means of promoting tumor growth and metastasis, including endothelial cells, fibroblasts, adipocytes, neurites, and immune cells. The crosstalk that occurs between cancer and immune cells, particularly macrophages, is a striking example of long-distance communication between two different cell types. Broadly, this crosstalk is involved in remodeling the microenvironment toward a pro-tumorigenic state and the recruitment of cells to distant sites in the body (i.e., recruitment of macrophages to the tumor, or tumor cells to a site of metastasis). Understanding these mechanisms in the context of cancer may lead to insight into the long-distance crosstalk in other systems such as embryogenesis.

It has long been known that chronic infiltration of a tissue by macrophages and inflammation can promote neoplastic transformation, although the exact mechanism remains unclear. Early on in tumor formation, tumor cells secrete factors, cytokines, and ECM components into their environment in response to stimuli such as hypoxia to attract macrophages, causing their activation and polarization (Kim et al., 2009; Park et al., 2019). Tumor-associated macrophages (TAMs) are converted to the M2 phenotype (anti-inflammatory, proangiogenic) through tumor cell cytokine secretion and activation of the Wnt pathway (Ojalvo et al., 1950). Cancer cells also recruit macrophages to the tumor by secreting exosomes containing IL-8, VEGF, HGF, and CD44 mRNA that re-educate nearby monocytes, which differentiate into TAMs once they enter the tumor stroma (Baj-Krzyworzeka et al., 2006; Richards et al., 2013). It has been suggested that TNT formation may be linked to tumor–macrophage interactions (Hanna et al., 2019); TNT formation dependent on macrophage-secreted M-sec and EGF has been shown to increase 3D breast cancer invasion in vitro (Carter et al., 2019). Once TAMs are recruited to the tumor site, they in turn secrete factors that act on tumor cells to support their growth and metastasis (Coussens et al., 2000) and suppress the immune response against cancer cells (Solinas et al., 1950; Pathria et al., 2019). There is also evidence that TAMs can reduce the chemosensitivity of cancer cells, with one study demonstrating that miRNA contained in exosomes secreted by M2 macrophages conferred cisplatin resistance to gastric cancer cells in vivo (Zheng et al., 2017). Interestingly, TAMs also promote pro-metastatic changes in the tumor microenvironment in response to chemotherapy treatment (Sanchez et al., 2019).

One of the more striking examples of long-distance communication between TAMs and cancer cells is the priming of the pre-metastatic niche. For cancer cells to disseminate from the primary tumor and colonize a secondary location, they must be able to exit the circulation, migrate, and survive in the new site. Cancer cells can induce the early priming of these secondary sites, forming what is known as the premetastatic niche, before leaving the primary tumor. For example, pancreatic ductal carcinoma cells secrete exosomes from the primary tumor site which circulate to the liver where they fuse with Kupffer cells (macrophages of the liver) and cause activation of fibrotic and pro-inflammatory pathways that support metastases (Costa-Silva et al., 2015). TAMs also participate in priming of the premetastatic niche, secreting inflammatory cytokines which increase the success of lung metastasis (Nguyen et al., 2009). These observations suggest that cancer cells can play an instructor cell-like role, signaling to stromal cells to model microenvironments into favorable conditions in advance of the cancer cell arrival. Moreover, TAMs, aside from secreting factors that enter the circulation, are also directly recruited to the premetastatic niche and their presence is necessary for cancer cell survival at the secondary metastatic site (Hiratsuka et al., 2008; Gil-Bernabé et al., 2012; Kim et al., 2019).

Better understanding of how macrophages mediate long-distance communication during development may provide key insights for deciphering their role in cancer. The observation that breakdown in this process leads to invasion of melanocytes in zebrafish provides tantalizing evidence that this process may function in metastasis of highly aggressive cancers such as melanoma. Just as TAMs mediate communication between tumors and pre-metastatic niches, during stripe formation macrophages mediate communication between two distant and distinct cell types. The mechanisms by which macrophages identify and distinguish these two cell types may point to more general mechanisms by which TAMs preferentially target certain cell and tissue types. Similarly, the importance of airinemes and other projections (Eom et al., 2015) in this process may lead to better understanding of the pathological roles of these structures in metastasis.

It is important to understand methods of long-distance communication in non-neural networks. Herein we discussed recent insights into non-neural networks for communication at multiple scales and compare the two contexts of development and neoplasia to suggest what each field may learn from the other, summarized in Figure 1 and Table 1. Bioelectric signals can electrically couple cells during development, providing high level instructions for patterning and morphogenesis, and in cancer initiate signaling cascades to increase malignancy. TNTs offer the ability to control cell behavior beyond the local environment, sharing bioelectric, and morphogenic signals between cells during development and a variety of cargo in cancer cells to control cellular identity and shape the microenvironment. Lastly, we discussed the ability of macrophages to regulate long distance networks between cells in development and cancer, acting as relay cells to facilitate crosstalk between different cell types during development and establishing metastatic niches for cancer cells in distant organs. Although in most of the examples we provided these non-neural networks participate in different functions in development versus cancer, comparing the mechanisms of these networks is a valuable tool for both fields. Clearly development is a coordinated effort among cells to achieve large-scale patterning and tissue formation, but cancer, while not a coordinated effort of the body, can be considered coordinated within the cancer “organism”; there is simultaneously a breakdown of normal signaling but also the establishment of new communication routes with the recruitment of macrophages and formation of TMPs. Understanding how communication in these non-neural networks occurs will allow us to control it in disease states of both embryonic and post-natal organisms.

The long-distance phenomena described here may have profound implications for many different fields of biology. Developmental biologists would be well advised to look for novel players in well-studied processes at much greater distances than previously examined, as underscored by the recent discover of bioelectric injury mirroring (Busse et al., 2018). Disease researchers may find pathological sources in surprising accessory cell types located far from the affected cells. Clinically, the long-distance phenomena we review here may enable diagnosis or treatment of inoperable disease at readily manipulable surrogate sites. And, finally, the prevalence of long-distance communication may force us to re-evaluate our notions of cell autonomy as we uncover more evidence that a cell’s behavior is not a local phenomenon, but rather an emergent integration of information from throughout the body.

PM and SP conceived of the manuscript topic and drafted the manuscript. MO and ML critically revised it. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer MZ declared a shared consortium with one of the author ML to the handling editor.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.