Axon-like protrusions promote small cell lung cancer migration and metastasis

Abstract

14Metastasis is the main cause of death in cancer patients but remains a poorly understood 15 process. Small cell lung cancer (SCLC) is one of the most lethal and most metastatic types of 16 human cancer. SCLC cells normally express neuroendocrine and neuronal gene programs but 17 accumulating evidence indicates that these cancer cells become relatively more neuronal and less 18 neuroendocrine as they gain the ability to metastasize. Here we show that mouse and human 19 SCLC cells in culture and in vivo can grow cellular protrusions that resemble axons. The 20 formation of these protrusions is controlled by multiple neuronal factors implicated in 21 axonogenesis, axon guidance, and neuroblast migration. Disruption of these axon-like 22 protrusions impairs cell migration in culture and inhibits metastatic ability in vivo. The co-option 23 of developmental neuronal programs is a novel molecular and cellular mechanism that 24 contributes to the high metastatic ability of SCLC. 25 65 cells that have transitioned to a more neuronal cell state. 66 5 68 To investigate SCLC migration, we developed an assay in which SCLC cells, which 69 classically grow in culture as floating spheres or aggregates, are grown as a monolayer under 70 Matrigel ((Denny et al., 2016) and Methods). Unexpectedly, we noticed that cells from some 71 SCLC cell lines (N2N1G, 16T, 6PF) derived from the Rb f/f ;p53 f/f (DKO) and Rb f/f ;p53 f/f ;p130 f/f 72 (TKO) genetically engineered mouse models form long cellular protrusions into cell-free spaces 73 ( Figure 1A-B). To determine whether these structures specifically project into cell-free areas or 74 they also exist within monolayers, we cultured a minor fraction of fluorescently-labeled, RESULTS 67 SCLC cells can form long cellular protrusions in culture and in vivo 75GFP positive SCLC cells with control SCLC cells. We found that SCLC cells also form protrusions 76 when they are in close contact with surrounding cancer cells (Figure S1A). Similar mixing 77 experiments performed in subcutaneous allografts also documented the growth of protrusions by 78 SCLC cells in vivo (Figure 1C-D). Finally, similar structures also extend from SCLC micro-79 metastases in the liver in the autochthonous TKO mouse model and after intravenous 80 transplantations of SCLC cells (Figure S1B-C). 81Human SCLC patient-derived xenografts (PDXs) recapitulate many important features of the 82 human disease (e.g. (Gardner et al., 2017, Saunders et al., 2015). To label rare cancer cells 83 within human SCLC PDXs and identify whether they had protrusions in unperturbed tumors, we 84 used DiI tracing. DiI is a lipophilic dye that diffuses within cell membranes and has been widely 85 employed to label projections from individual neurons (Heilingoetter and Jensen, 2016, Mufson 86 et al., 1990). Protrusions from SCLC cells were easily identifiable in two out of three PDX 87 models ( Figure 1E-F). In the 2D monolayer culture system, not all human SCLC cell lines 88 formed protrusions, but NCI-H446 cells formed long protru...

Single cell profiling reveals novel tumor and myeloid subpopulations in small cell lung cancer

“…In particular, we found that variant (non-SCLC-A) SCLC subtypes are enriched in local/distant SCLC metastases in humans, with SCLC-N upregulating pathways in EMT, axonogenesis, and metastasis 20 . These findings are consistent with 1) a previously reported mouse model with invasive and metastatic SCLC tumors overexpressing NEUROD1 17 and 2) cell lines with an axonogenic phenotype that correlates with NEUROD1 expression and promotes metastasis 33 . In contrast to SCLC-A, we found an enrichment of ligand-receptor interactions in SCLC-N. One biphenotypic tumor identified ligands secreted from the SCLC-A compartment that activate Notch, Wnt, TGFβ/Activin and pro-axonogenic signaling in SCLC-N.…”

“…Interestingly, we found that SCLC-N was enriched in two main axonogenic signaling pathways: ephrin (EFNB1 and EPHB2, among others) 30 and semaphorin (SEMA6A and NRP2, among others) 31 . Prior studies have shown that the axonogenesis program coordinates cell polarity with neuronal migration 32 ; it has been implicated in SCLC metastasis 33 , but never before in SCLC-N specifically. We have shown that LUAD hijacks endodermal developmental preprint (which was not certified by peer review) is the author/funder.…”

Neuronal signatures in cancer

“…Several novel therapeutic paradigms arise that exploit the neuronal microenvironment and neurodevelopmental characteristics: (a) Prevention of the outgrowth of neurite‐like protrusions to inhibit invasion and interconnection; (b) Inhibition of tumor‐tumor cell communication; (c) Inhibition of neuronal input and neurotransmitter signaling; (d) Depletion of neuronal intratumoral subpopulations and tumor‐subtype specific treatments. Pharmacological or genetic (eg, by antisense oligonucleotides) targeting of neurodevelopmental pathways and molecules, such as Gap43, Fez1 or Ttyh1, could inhibit the outgrowth of neurite like‐protrusions. It remains to be elucidated if these protrusions also play a role for therapy resistance outside the brain, but first evidence exists that demonstrates that the inhibition of neurite‐like protrusions can be used to reduce the metastatic potential 30,33 . Many molecules involved during neurodevelopment seem to be of limited importance in the mature brain and might hence be suitable targets for novel tumor‐specific therapies 3 .…”

“…Neurite‐ and TM‐like protrusions, which are also driven by neurodevelopmental pathways including Gap43, have recently been found in small cell lung cancer (SCLC), 30 where they promoted invasion and metastasis. It needs to be seen whether those protrusions are also used for the formation of multicellular networks and for interconnection with host cells, for example, those of the brain, an organ where these tumors frequently metastasize to.…”

Multi-omics Data Analyses Construct TME and Identify the Immune-Related Prognosis Signatures in Human LUAD

“…Several studies had reported similar findings in various cancer types. [53][54][55][56] Finally, according to the detected genetic or epigenetic alterations, a prognosis model was constructed and was found to have a superior performance with a higher C-index or AUC value compared to the previous models. [14][15][16] To our surprise, the trained models showed higher accuracy of 5-year survival prediction on the training sets compared with the result from the 1-year survival prediction, whereas the superiority was lost when the prediction was done using the test set.…”