Research ArticleOncology Open Access | 10.1172/JCI170953

Saptarshi Sinha,1 Brennan W. Callow,2 Alex P. Farfel,2 Suchismita Roy,1 Siyi Chen,2 Maria Masotti,3 Shrila Rajendran,2 Johanna M. Buschhaus,2,4 Celia R. Espinoza,1 Kathryn E. Luker,2,5 Pradipta Ghosh,1,6,7,8 and Gary D. Luker2,4,5

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Sinha, S. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Callow, B. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Farfel, A. in: JCI | PubMed | Google Scholar |

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Roy, S. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Chen, S. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Masotti, M. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Rajendran, S. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Buschhaus, J. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Espinoza, C. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Luker, K. in: JCI | PubMed | Google Scholar |

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Ghosh, P. in: JCI | PubMed | Google Scholar

1Department of Cellular and Molecular Medicine, School of Medicine, UCSD, La Jolla, California, USA.

2Center for Molecular Imaging, Department of Radiology,

3Biostatistics, School of Public Health,

4Department of Biomedical Engineering, and

5Biointerfaces Institute, University of Michigan, Ann Arbor, Michigan, USA.

6Moores Comprehensive Cancer Center,

7Department of Medicine,

8School of Medicine, and Veterans Affairs Medical Center, UCSD, La Jolla, California, USA.

Address correspondence to: Pradipta Ghosh, Departments of Medicine and Cellular and Molecular Medicine, University of California San Diego, 9500 Gilman Drive (MC 0651), George E. Palade Bldg, Rm 232, 239, La Jolla, California 92093, USA. Phone: 858.822.7633; Email: prghosh@ucsd.edu. Or to: Gary D. Luker, Departments of Radiology and Biomedical Engineering, A524 BSRB, 109 Zina Pitcher Place, Ann Arbor, Michigan 48109-2200, USA. Phone: 734.764.2890; Email: gluker@umich.edu.

Find articles by Luker, G. in: JCI | PubMed | Google Scholar |

Published October 31, 2024 - More info

Published in Volume 134, Issue 24 on December 16, 2024
J Clin Invest. 2024;134(24):e170953. https://doi.org/10.1172/JCI170953.
© 2024 Sinha et al. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Published October 31, 2024 - Version history
Received: March 29, 2023; Accepted: October 24, 2024 View PDF Abstract

Estrogen receptor-positive (ER+) breast cancer commonly disseminates to bone marrow, where interactions with mesenchymal stromal cells (MSCs) shape disease trajectory. We modeled these interactions with tumor-MSC co-cultures and used an integrated transcriptome-proteome-network-analyses workflow to identify a comprehensive catalog of contact-induced changes. Conditioned media from MSCs failed to recapitulate genes and proteins, some borrowed and others tumor-intrinsic, induced in cancer cells by direct contact. Protein-protein interaction networks revealed the rich connectome between “borrowed” and “intrinsic” components. Bioinformatics prioritized one of the borrowed components, CCDC88A/GIV, a multi-modular metastasis-related protein that has recently been implicated in driving a hallmark of cancer, growth signaling autonomy. MSCs transferred GIV protein to ER+ breast cancer cells (that lack GIV) through tunnelling nanotubes via connexin (Cx)43-facilitated intercellular transport. Reinstating GIV alone in GIV-negative breast cancer cells reproduced approximately 20% of both the borrowed and the intrinsic gene induction patterns from contact co-cultures; conferred resistance to anti-estrogen drugs; and enhanced tumor dissemination. Findings provide a multiomic insight into MSC→tumor cell intercellular transport and validate how transport of one such candidate, GIV, from the haves (MSCs) to have-nots (ER+ breast cancer) orchestrates aggressive disease states.

Graphical Abstract Introduction

Estrogen receptor-positive (ER+) breast cancer, the most common subtype, preferentially disseminates to bone and bone marrow (1, 2). Bone metastases remain incurable and cause debilitating symptoms, e.g., pain, fractures, and life-threatening hypercalcemia (3). Bone marrow harbors disseminated tumor cells (DTCs) in breast cancer. DTCs may persist in a clinically occult state for decades before proliferating and potentially traveling to other organs to produce delayed recurrences in approximately 40% of patients with ER+ breast cancer (4). Recurrent disease is more aggressive, relatively resistant to treatment, and currently incurable. Little progress has been made in treating breast cancer that has already disseminated to bone (5), largely because we lack a comprehensive understanding of molecular mechanisms that make ER+ breast cancer cells more aggressive in the bone marrow niche (6).

Interactions with stromal cells in the bone marrow environment are proposed to shape key functions of breast cancer cells that determine disease progression and outcomes (7). Mesenchymal stem cells (MSCs), a multipotent cell type that contributes to formation of bone, fat, and cartilage, are a major stromal cell type driving aggressive phenotypes of disseminated ER+ breast cancer cells in bone marrow. MSCs regulate ER+ breast cancer cells through mechanisms such as secreted cytokines and direct intercellular interactions (8). Work by us and others has established that direct interactions with MSCs, rather than soluble molecules, drive changes in gene expression and metabolism that promote cancer stem-like cell states, resistance to anti-estrogen drugs, and metastasis in ER+ breast cancer (9–11). Breast cancer cells and MSCs can establish channels of intercellular communication that transport molecules and organelles. Two major structures for intercellular communication include gap junctions and tunneling nanotubes (TNTs), both of which share connexin 43 (Cx43) as an essential molecular player. Gap junctions are intercellular channels comprised of various subsets of connexin proteins (12). Lymph node and systemic metastases in breast cancer and other malignancies upregulate gap junctions and Cx43, suggesting these structures contribute to essential steps in the metastatic cascade (13). TNTs are actin-rich long, thin tubes that may form between cells of the same or different types, thereby allowing long-range communication via the exchange of materials (14, 15). TNTs serve as “exchange freeways” for a broad range of intracellular contents, including microRNAs, proteins, and organelles. Such exchanges, between tumor cells or between tumor and stromal or immune cells (8, 16, 17) shape drug resistance of cancer cells; regulate proliferation; and promote metastasis.

Despite these insights, only a limited number of molecules transferred from MSCs via gap junctions or TNTs have been implicated in aggressive phenotypes of breast cancer cells, whereas the identity of most remains unknown. We tackle this knowledge gap by performing a comprehensive multiomic network analysis to identify molecules and biologic processes regulated by direct contact between bone marrow MSCs and ER+ breast cancer cells. We identified genes and proteins that cancer cells acquire from MSCs through direct transfer (“borrowed” components) and additional molecules induced because of such transfer (“intrinsic” components). We used bioinformatic approaches to explore the implications of these gene expression changes on patient outcome and subsequently prioritized one borrowed gene/protein (CCDC88A/GIV) to explore further and validate through a series of in vitro and in vivo studies. Our findings establish a comprehensive resource for changes induced in ER+ breast cancer cells by contact with bone marrow MSCs and pinpoint GIV as a promising target to overcome aggressive phenotypes acquired by breast cancer cells in bone marrow.

Results

Multiomic analysis of close contact between ER+ breast cancer cells and MSCs. DTCs in bone marrow displace hematopoietic stem cells from supportive niches (18), where they establish direct contact with various subsets of MSCs and eventually gain aggressiveness. To identify how MSCs may impact ER+ breast cancer cells through direct contact, we used a 2D co-culture model combining MCF7 or T47D human ER+ breast cancer cells with human HS5 or HS27a MSCs in a 1:9 ratio (Figure 1A). After 3 days of co-culture in medium with low serum and physiologic concentrations of glucose, we recovered cancer cells using EpCAM immunomagnetic beads, capitalizing on expression of EpCAM exclusively on tumor cells of epithelial origin. We previously showed that such EpCAM-based immunoisolation has negligible MSC contamination (19) and confirmed through rigorous characterization studies showing that tumor cells subjected to such contact culture gain advantageous features such as metabolic plasticity (20), resistance to estrogen-targeted therapies, and enhanced survival of disseminated tumor cells early in the metastatic process (19, 21). To control for effects of soluble molecules produced by MSCs, we cultured MCF7 or T47D cells for 3 days in conditioned medium from MSCs (Figure 1A). We used monocultures of MSCs and tumor cells as additional comparator groups. We analyzed all samples by bulk RNA sequencing; we also analyzed mono- and contact cultured MCF7 cells by TMT proteomics.

Figure 1

Multiomic analysis reveals RNA/protein acquired by ER+ breast cancer cells from MSCs in contact co-culture. (A) Experimental set up for recapitulating disseminated ER+ breast cancer cells in bone marrow amidst mesenchymal stem cells (MSCs). Two different types of co-culture models were used, one with conditioned media (CMed; top) and one with direct contact co-culture (CC; bottom). MC, monoculture. (B) Violin plots display the composite score of a 49-gene tumor dormancy score (DS) specific for ER+ breast cancers and previously validated in 4 independent cohorts to predict recurrence. (C) Top: Venn diagram depicts multiple DEG analyses between different indicated groups that catalog suppressed genes or proteins in contact co-culture (CC) or conditioned media (CMed) and overall differential gene expression between MCF7 versus HS5 bone marrow MSCs; bottom. Bottom: Genes/proteins induced in MCF7 cells in co-culture with HS5 MSCs are binned into 3 categories (connected by arrows) based on likely mechanisms for induction. Red = 39 uniquely upregulated transcripts in cancer cells in contact co-culture also identified by proteomics. (D) Gene ontology analysis (GO Biological processes) on transcripts and/or proteins acquired by MCF7 cells from MSCs during contact co-culture. See Supplemental Figure 1 for Reactome pathway and GO enrichment analyses on genes and/or proteins within each category in C. (E) Reactome analysis on transcripts and/or proteins reduced in MCF7 cells during contact co-culture with MSCs. See Supplemental Figure 2 for analyses on genes and proteins suppressed in cancer cells.

MCF7 cells exposed to contact culture, but not conditioned media, correlated with substantial lowering of a previously defined 49-gene signature for tumor cell dormancy (Figure 1B); low dormancy scores confer an approximately 2.1-fold increase in risk of recurrence in 4 independent cohorts of patients with ER+ breast cancers (P < 0.000005) (22). These findings suggest that our co-culture model captures key transcriptomic changes in ER+ breast cancer cells of translational relevance to more aggressive disease in patients.

We carried out combinatorial multiomics analyses to determine what proteins or transcripts expressed highly in MSCs are borrowed by MCF7 cells. To this end, we first created a catalog of genes differentially expressed (DEGs; upregulated) in MSCs, but not in MCF7 cells (MCF7 vs. HS5; n = 1471; Figure 1C). This catalog served as a denominator for all likely candidate genes that a tumor cell may lack originally but acquire from MSCs via intercellular communication. Next, we performed pairwise DEG analysis on the other MCF7 samples subjected to different co-culture conditions. Direct contact co-culture induced many genes (n = 482) and proteins (n = 295) (Figure 1C). Conditioned media induced few genes in cancer cells (n = 13; Figure 1C). An overlay showed the following: (a) Exposure to conditioned media alone induced only a single unique protein/transcript in MCF7 cells, implying soluble mediators or exosomes as potential modes of communication; (b) By contrast, contact culture accounted for 487 unique proteins/transcripts, implying direct contact as the major mode of tumor-MSC communication and the largest share of uniquely induced genes; (c) Contact culture increased 242 unique proteins/transcripts (39 transcripts and 203 proteins with no proteome-transcriptome overlap) in MCF7 cells. These transcripts/proteins were not highly expressed in HS5 cells. We presumed this last category to be tumor cell–intrinsic response to contact culture, unlikely to be transferred directly from MSCs to breast cancer cells. We note that neither the experimental design nor the analytical steps can conclude definitively whether transport of a given target occurred as a protein or as mRNA because unknown factors (transcript/protein half-life) likely confound such conclusions. Similarly, while we infer that 487 unique RNAs/proteins could be transferred during contact culture, i.e., borrowed directly by cancer cells from the MSCs, we cannot exclude some cancer cell–intrinsic contributions to the overall levels. We did observe, however, enrichment of the borrowed RNA/proteins for biological processes that govern tube morphogenesis, organization of extracellular encapsulated structures, and movement of anatomic structures in the context of multicellular systems (Figure 1D). Enrichment of these processes further suggests the potential for direct transfer of RNA/protein from MSCs to cancer cells through structures such as TNTs. Reactome and gene ontology analyses of the genes induced by conditioned media, contact culture and intrinsic response present a comprehensive picture of distinct modes by which MSCs shape behavior and function of cancer cells (Supplemental Figure 1; supplemental material available online with this article; https://doi.org/10.1172/JCI170953DS1). A similar analysis for transcripts and proteins downregulated during contact culture and when exposed to conditioned media revealed genes/proteins that reduce the dependence on ER signaling (Supplemental Figure 2). We provide a full catalog of all upregulated (Supplemental Data Set 1) and downregulated (Supplemental Data Set 2) genes/proteins during contact culture.

A gene signature of contact culture is prognostic and predictive in ER+ breast cancers. We noted 39 unique, upregulated genes identified by both RNA seq and TMT proteomics in MCF7 cells during contact co-cultures (Figure 1C). Direct contact culture uniquely upregulated almost all genes (n = 38; Figure 1C) except for one (KYNU) upregulated in conditioned media. Hierarchical clustering, an unsupervised learning technique used to group similar objects into clusters, showed the 39 genes grouped the samples into 2 distinct groups within a dendrogram: MCF7 cells subjected to culture in direct contact with HS5 cells (Figure 2A) clustered with the HS5 monoculture samples. By contrast, MCF7 cells cultured in conditioned media clustered with the MCF7 monoculture samples (compare MC and CMed; Figure 2A). A composite score of the levels of these 39 genes confirmed their statistically significant increase in contact cultures but not conditioned media (Figure 2B). This pattern of gene induction repeated when we swapped HS5 cells with another MSC line, HS27a, in co-cultures (Figure 2C) or swapped MCF7 cells with another ER+ breast cancer line, T47D (Figure 2D). These findings demonstrate that the 39-gene signature represents a conserved feature of how MSCs impact the properties or behavior of ER+ breast cancer.

Figure 2

Genes uniquely upregulated in MCF7 tumor cells in contact co-culture with HS5 MSCs are identified reproducibly in other tumor-MSC co-culture models. (A) Heatmap displays hierarchical unsupervised clustering of samples used in Figure 1A using the 39 uniquely upregulated genes in cancer cells in contact culture. MC, monoculture. (B–D) Violin plots display the composite score of the 39 genes in A in various tumor cell-MSC co-culture models, e.g., MCF7↔HS5 (B), MCF7↔HS27a (C) and T47D↔HS5 (D). Statistical significance for B–D was assessed by 1-way ANOVA and the P values are corrected for multiple comparisons using Tukey’s method.

The 39 genes showed enrichment of processes related to growth factor, PI3K/Akt, and IL4/IL13 and IL10-centric tolerogenic cytokine signaling (Figure 3A), suggesting that acquisition of these 39 genes from MSCs may impact these aspects of cancer cell biology and/or behavior, including evasion of the immune system.

Figure 3

A contact culture signature derived from proteome-transcriptome overlap carries prognostic and predictive value in ER+ breast cancer. (A) Reactome pathways enriched in the 39 genes identified in Figure 1C. The PI3K/Akt signaling pathway (red), growth factor signaling pathways (teal blue), and the tolerogenic cytokine pathways (navy blue) are highlighted. (B–G) Kaplan-Meier survival plots on HER2-negative breast cancer patients from 2 independent cohorts with known outcomes over time (relapse/metastasis-free survival), stratified as high versus low expression of the 39-gene signature (see Methods). Statistical significance was assessed by log-rank analyses. RD, residual disease; PCR, pathologic complete response; Rx, treatment.

To test translational relevance of the 39-gene signature, we applied it to a publicly available dataset that prospectively tracked outcomes of ER+, Her2-negative breast cancers following neoadjuvant taxane-anthracycline chemotherapy (23) (Figure 3, B–E). Kaplan-Meier curves revealed that high levels of the contact culture signature correlated with significantly worse distant relapse-free survival in patients with ER+ (Figure 3B), but not ER–, breast cancers (Figure 3C). High levels of the close contact signature also predicted worse outcomes in patients with treatment resistant, but not treatment sensitive, disease (Figure 3, D and E). We reproduced these findings in an independent Her2-negative ER+ cohort from a large dataset, i.e., METABRIC (Molecular Taxonomy of Breast Cancer International Consortium; ref. 24), a landmark genomic and transcriptomic study of 2000 individual breast tumors (Figure 3, F and G). These findings establish that the 39 genes presumed to be borrowed by ER+ breast cancer cells via direct contact with MSCs carry translationally relevant information; they identify patients at greater risk for death from recurrent breast cancer.

Borrowed genes integrate with a cancer cell–intrinsic response during contact culture. Although TMT based proteomics can reliably detect as small as approximately one-tenth of the changes compared with label-free proteomics (25), it suffers from interference and distortion, which disproportionately impact low abundance proteins that often are the ones borrowed (26, 27). To circumvent this limitation and improve detection of relevant hits, we utilized a protein-protein interaction (PPI) network approach to identify relevant proteins, based on the assumption that the proteins function through meaningful interactions with each other. Briefly, we used the 39 proteins as seeds to build a PPI network consisting of 2743 proteins (Figure 4A). We overlaid the network-derived list with borrowed (n = 487) and intrinsic (n = 242) gene or protein sets (see Figure 1C) acquired by tumor cells via contact culture (n = 487) and genes or protein sets upregulated in tumor cells in contact culture but not highly expressed in MSCs. Generating the proteome network using the STRING database (see Methods) and then intersecting the proteome with the differential transcriptome increased signals (i.e., hits) by leveraging the sensitivity of transcriptomics and specificity of proteomics. This process produced a more focused network using 159 genes that ER+ breast cancer cells borrow from MSCs and 76 genes in the cancer cell intrinsic response group. Figure 4C displays the multilayer network connecting borrowed (pink) and intrinsic response (green) genes. Within the PPI network, 1482 nodes/proteins interacted within the borrowed layer and 835 nodes/proteins interacted within the intrinsic layer. The borrowed response interactome included Akt, growth factors and/or their receptors (EGFR, VEGFA, WNT5A), and integrins (Figure 4C), as noted earlier (Figure 3A). The borrowed network also revealed meaningful enrichment of other candidates, including the gap junction protein, GJA1/CX43, which has been implicated in establishing TNTs that facilitate exchange of molecules (28). The 2 layers shared 567 common nodes/proteins, indicating meaningful connectivity between 2 layers (P value of 0.00096 by hypergeometric test). The shared nodes were enriched for 3 prominent themes: membrane trafficking, immune signaling, and growth factor signaling (Figure 4D).

Figure 4

A multilayer network analysis to explore connectivity between the borrowed and intrinsic components of upregulated proteins in cancer cells after co-culture with MSCs. (A) Workflow for protein-protein interaction (PPI) network analysis using the 39 gene signature from Figure 1E as seeds for the STRING database. (B) Overlaps between the PPI network and genes/proteins from Figure 1C as likely borrowed from MSCs or upregulated in the cancer cell–intrinsic response during contact co-cultures. (C) Multilayer PPI network shows connectivity between borrowed (n = 159) and intrinsic (n = 76) proteins, with key nodes labeled. (D) Reactome pathways (left) enriched in 567 proteins shared between borrowed and intrinsic layers of the PPI network in C. Venn diagram (right) shows the total nodes in each layer and overlap.

Findings suggest functional integration between the 2 components; the induction of one (the intrinsic cancer genes) may either enable functions of and/or reflect consequences of the other (the borrowed genes). Supplemental Data set 3 presents the edge and node list of the entire PPI network and various subnetworks.

CCDC88A/GIV as a key gene borrowed by cancer cells. To validate and/or characterize key components of the borrowed transcriptome/proteome and yet reduce the risk of noise (of STRING database) and artifacts (of 2D culture), we further refined this list (n = 159) generated using 2D cultures by leveraging a published dataset using the same cell lines as us except in 3D cultures (GSE152312) (Figure 5A) (29). We carried out this refinement with a 2-fold intention: (a) to enhance the physiological context and translational relevance because TNTs can form in 3D culture systems (30); and (b) to provide continuity between 2D and 3D TNT biology, which has been difficult to achieve in a field still in its infancy (31). Refinement gave a handful of genes (n = 19) also induced during 3D close contact culture in both MCF7 and T47D cells (Figure 5B). We noted that 2 of the 19 genes (CCDC88A and EGFR) are key components of a recently described phenomena, growth signaling autonomy, which endows breast tumor cells with plasticity and stemness (among other features) especially during hematogenous dissemination in metastasis (32, 33). Gene ontology (GO) analysis of these 19 genes revealed that CCDC88A (which encodes the protein Gα-interacting, vesicle-associated [GIV or Girdin]) is commonly shared among key processes upregulated in ER+ breast cells during reprogramming of signaling to adopt a mesenchymal state (34–38) (Figure 5C). Compared with MCF7 monocultures, 2D direct contact cultures, but not conditioned medium, markedly upregulated CCDC88A (Figure 5D). Focused analysis of the CCDC88A/GIV-subnetwork from the multi-layer interactome (Figure 4C) revealed that the interactome of GIV could support numerous pathways and processes involved in cell projections and membrane domains (Figure 5E), suggesting that CCDC88A associates with direct intercellular contacts.

Figure 5

Integration of 2D and 3D co-culture–derived omics pinpoints GIV as a central orchestrator of the co-culture borrowed gene signature. (A) The 158 borrowed genes from our 2D co-cultures were further filtered using a public dataset from 3D co-cultures of MCF7 and T47D cells with HS5 MSCs. Threshold ROC AUC > 0.85 identified 19 genes. Violin plot shows the composite score for these 19 genes in 3D monoculture (MC) versus contact (CC) cultures. (B) Heatmap of z score–normalized expression of the 19 genes. (C) Gene ontology (GO) analyses identify CCDC88A as an invariant player across processes enriched within the 19 genes. (D) Shows induction of CCDC88A in MCF7 cells in contact coculture (CC) but not conditioned media (CMed). Only significant P values are displayed (Welch’s t test). (E) Graph displays GO cellular components enriched in the CCDC88A/GIV-subcluster in C. Blue highlights denote processes required for transport via tunneling nanotubes. (F) MCF7 cells (red nuclei; “T”) connected with a nanotube to HS5 MSCs in contact coculture (left). Boxed area is magnified on the right. Arrows mark the nanotube. Supplemental Figure 3 shows additional images. Bar plots (right) show average number of TNTs in each condition per field. (G and H) Violin plots show induction of GJA1 (G) and TNFAIP2 (H) in MCF7 cells in contact co-culture (CC) but not in conditioned media (CMed). P values for A, F, G, and H were derived by 1-way ANOVA and corrected for multiple comparisons by Tukey’s method.

TNTs constitute a major pathway to transfer materials, including RNA and proteins, between cells of the same and different types (39) through direct intercellular contacts. Using interferometry microscopy, we identified numerous TNTs connecting MCF7 and HS5 cells (Figure 5F and Supplemental Figure 3). Images revealed significantly more heterotypic TNTs connecting MCF7 cells with HS5 or HS27a stromal cells than homotypic TNTs connecting MCF7 cells in monocultures (Figure 5F). TNT formation in contact cultures also associated with elevated expression of 2 proteins previously implicated as central factors essential for their formation: (a) the gap junction protein GJA1 (28) (connexin 43, CX43; Figure 5G) and the TNT marker, TNFAIP2 (40) (M-Sec; Figure 5H). To investigate whether gap-junction or TNT-facilitated intercellular communication is functional, we conducted the well-established calcein-AM transfer assay (41, 42). Co-culture of MCF7 cells stably expressing mCherry (recipients) with calcein-labeled HS5 MSCs (donors) or calcein-labeled MCF7 cells showed successful transfer of calcein into mCherry-labeled MCF7 cells (Supplemental Figure 4A). Both donor cells (HS5 and MCF7 cells) transferred calcein to recipient MCF7 cells. Although MCF7 cells formed both heterotypic and homotypic TNTs, transfer is markedly higher from MSCs (Supplemental Figure 4A). We also detected transfer from MCF7 cells to HSCs, albeit at lower efficiency (Supplemental Figure 4B). These results confirm that our 2D contact cultures allowed intercellular transport.

Together these findings suggest that TNTs in contact cultures could provide a route for breast cancer cells to borrow f