CoChat

An in vitro vascularized micro-tumor model of human colorectal cancer recapitulates in vivo responses to standard-of-care therapy

Open access

Published version of a preprint

The DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index.

Abstract

Around 95% of anti-cancer drugs that show promise during preclinical study fail to gain FDA-approval for clinical use. This failure of the preclinical pipeline highlights the need for improved, physiologically-relevant in vitro models that can better serve as reliable drug-screening and disease modeling tools. The vascularized micro-tumor (VMT) is a novel three-dimensional model system (tumor-on-a-chip) that recapitulates the complex human tumor microenvironment, including perfused vasculature, within a transparent microfluidic device, allowing real-time study of drug responses and tumor-stromal interactions. Here we have validated this microphysiological system (MPS) platform for the study of colorectal cancer (CRC), the second leading cause of cancer-related deaths, by showing that gene expression, tumor heterogeneity, and treatment responses in the VMT more closely model CRC tumor clinicopathology than current standard drug screening modalities, including 2-dimensional monolayer culture and 3-dimensional spheroids.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1039/d0lc01216e

Cites

50 of 50 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Vascularized micro-tumors (VMTs) are supported by gravity-driven flow through living human microvessels within a microfluidic platform arrayed for high throughput experiments. (A) The microfluidic chip is bonded to a bottom-less 96-well plate via chemical glue and oxygen plasma. Zoomed view shows a single device unit with 3 tissue chambers (T1-3) fed through microfluidic channels, 2 loading ports (L1-2), medium inlet and outlet (M1-2), and a pressure regulator (PR) to prevent gel bursting during loading. (B) Fluorescent image shows a vascularized micro-tumor with EC (mCherry, red) forming the vascular network, fibroblasts (Az, blue) supporting the tissue, and HCT116 cancer cells (GFP, green). Tissue chamber is ∼2 × 1 × 0.1 mm. Scale bar is 200 μm. (C) COMSOL simulation on a perfused vascular network shows the 2D surface velocity of medium flowing through the vessels.
Fig. 1 Vascularized micro-tumors (VMTs) are supported by gravity-driven flow through living human microvessels within a microfluidic platform arrayed for high throughput experiments. (A) The microfluidic chip is bonded to a bottom-less 96-well plate via chemical glue and oxygen plasma. Zoomed view shows a single device unit with 3 tissue chambers (T1-3) fed through microfluidic channels, 2 loading ports (L1-2), medium inlet and outlet (M1-2), and a pressure regulator (PR) to prevent gel bursting during loading. (B) Fluorescent image shows a vascularized micro-tumor with EC (mCherry, red) forming the vascular network, fibroblasts (Az, blue) supporting the tissue, and HCT116 cancer cells (GFP, green). Tissue chamber is ∼2 × 1 × 0.1 mm. Scale bar is 200 μm. (C) COMSOL simulation on a perfused vascular network shows the 2D surface velocity of medium flowing through the vessels.CC BY-NC · PubMed Central
Fig. 2 The VMT models key features of in vivo tumor vasculature. (A) Comparison of VMO and VMT on day 5 and (B) day 8 shows irregular vasculature in the VMT that becomes more prominent as the tumor grows. Arrowheads show skinny, non-perfused vessels as areas of vascular disruption in the VMT. EC (mCherry, red), HCT116 cancer cells (GFP, green). (C) Quantification of total vessel length, (D) vessel diameter, (E) number of junctions and (F) number of endpoints, all are mean and SD, n = 12. Statistical significance was calculated by unpaired t-test. (G) Fluorescent image at T = 0 min shows angiogenic sprouts near an HCT116 tumor cluster (arrowhead). EC (GFP, green), HCT116 cancer cells (azurite, blue). Perfusion with 70 kD rhodamine dextran at T = 1 min and T = 2 min reveals a significant leak from the tumor-associated vessel. (H) Quantification of the permeability coefficient shows that the VMT is twice as leaky as the VMO. Mean and SD, n = 12. Statistical significance was calculated by unpaired t-test. ***p < 0.0001, ***p < 0.00001.
Fig. 2 The VMT models key features of in vivo tumor vasculature. (A) Comparison of VMO and VMT on day 5 and (B) day 8 shows irregular vasculature in the VMT that becomes more prominent as the tumor grows. Arrowheads show skinny, non-perfused vessels as areas of vascular disruption in the VMT. EC (mCherry, red), HCT116 cancer cells (GFP, green). (C) Quantification of total vessel length, (D) vessel diameter, (E) number of junctions and (F) number of endpoints, all are mean and SD, n = 12. Statistical significance was calculated by unpaired t-test. (G) Fluorescent image at T = 0 min shows angiogenic sprouts near an HCT116 tumor cluster (arrowhead). EC (GFP, green), HCT116 cancer cells (azurite, blue). Perfusion with 70 kD rhodamine dextran at T = 1 min and T = 2 min reveals a significant leak from the tumor-associated vessel. (H) Quantification of the permeability coefficient shows that the VMT is twice as leaky as the VMO. Mean and SD, n = 12. Statistical significance was calculated by unpaired t-test. ***p < 0.0001, ***p < 0.00001.CC BY-NC · PubMed Central
Fig. 3 Comparison of gene expression, tumor growth and response to drug treatment between VMT, 2D and 3D monocultures and xenograft tumors. (A) Nanostring cancer panel of 770 cancer-related genes in HCT116 CRC cell line-derived monolayers, VMT and xenograft tumors. (B) Pathways enriched with differentially expressed genes comparing VMT and xenograft tumor. (C) Growth curves of HCT116 cells grown in monolayer, in mice or in the VMT and treated with standard of care, FOLFOX. Mean and SD shown, n = 16. Statistical significance was calculated by unpaired t-test. See Methods for a description of how tumor size was calculated in each geometry. (D) Representative images of HCT116-VMT (vessels red, tumor cells green). (E) Growth curves of SW480 cells in monolayer, spheroids, VMT and mice treated with FOLFOX. Mean and SD shown, n = 16. Statistical significance was calculated by unpaired t-test. (F) Representative images of SW480-VMT. **p < 0.001, ***p < 0.0001.
Fig. 3 Comparison of gene expression, tumor growth and response to drug treatment between VMT, 2D and 3D monocultures and xenograft tumors. (A) Nanostring cancer panel of 770 cancer-related genes in HCT116 CRC cell line-derived monolayers, VMT and xenograft tumors. (B) Pathways enriched with differentially expressed genes comparing VMT and xenograft tumor. (C) Growth curves of HCT116 cells grown in monolayer, in mice or in the VMT and treated with standard of care, FOLFOX. Mean and SD shown, n = 16. Statistical significance was calculated by unpaired t-test. See Methods for a description of how tumor size was calculated in each geometry. (D) Representative images of HCT116-VMT (vessels red, tumor cells green). (E) Growth curves of SW480 cells in monolayer, spheroids, VMT and mice treated with FOLFOX. Mean and SD shown, n = 16. Statistical significance was calculated by unpaired t-test. (F) Representative images of SW480-VMT. **p < 0.001, ***p < 0.0001.CC BY-NC · PubMed Central
Fig. 4 scRNAseq reveals emerging HCT116 tumor heterogeneity in the VMT platform. (A) Experimental design. HCT116 CRC cells, EC and fibroblasts were grown individually as 2D monocultures and then pooled or co-cultured within the VMT, harvested on day 7 of culture and processed for droplet-based single cell RNA sequencing as separate samples. (B) Heatmap of top 10 differentially-expressed genes shows that clustering resolution retains unique marker gene expression between cell types within the aggregated dataset. The number of clusters emerges from the resolution used in Seurat and was optimized to balance the number of clusters versus the number of unique differentially-expressed genes in each cluster. (C) Distribution of each cluster from the aggregated data between monolayer and VMT culture. (D) Proportion of tumor cell types found in the aggregated dataset that are in VMT and monolayer based on scRNAseq clustering. Tumor C and tumor D populations are found exclusively in the VMT, whereas tumor B is found predominantly in monolayer. (E) Proportion of stromal cell types found in the aggregated dataset that are in VMT and monolayer. VMT is enriched for fibroblast B population and lacks EC B. (F) Immunofluorescent staining of EPCAM and KRT5 in monolayer and the VMT showed heterogeneous and higher expression of both EPCAM and KRT5 in the VMT. Scale bar is 50 μm.
Fig. 4 scRNAseq reveals emerging HCT116 tumor heterogeneity in the VMT platform. (A) Experimental design. HCT116 CRC cells, EC and fibroblasts were grown individually as 2D monocultures and then pooled or co-cultured within the VMT, harvested on day 7 of culture and processed for droplet-based single cell RNA sequencing as separate samples. (B) Heatmap of top 10 differentially-expressed genes shows that clustering resolution retains unique marker gene expression between cell types within the aggregated dataset. The number of clusters emerges from the resolution used in Seurat and was optimized to balance the number of clusters versus the number of unique differentially-expressed genes in each cluster. (C) Distribution of each cluster from the aggregated data between monolayer and VMT culture. (D) Proportion of tumor cell types found in the aggregated dataset that are in VMT and monolayer based on scRNAseq clustering. Tumor C and tumor D populations are found exclusively in the VMT, whereas tumor B is found predominantly in monolayer. (E) Proportion of stromal cell types found in the aggregated dataset that are in VMT and monolayer. VMT is enriched for fibroblast B population and lacks EC B. (F) Immunofluorescent staining of EPCAM and KRT5 in monolayer and the VMT showed heterogeneous and higher expression of both EPCAM and KRT5 in the VMT. Scale bar is 50 μm.CC BY-NC · PubMed Central
Fig. 5 Lineage hierarchy reconstruction reveals differential tumor states within HCT116-derived VMT. (A) Monocle analyses for HCT116-derived VMT tumor subset showing monocle trajectory plot and pseudotime plot (inset). (B) Pseudotime heatmap for HCT116-derived VMT transition subset showing 3 differentially-expressed clusters across the pseudotime trajectory. (C) Time-lapse confocal imaging showing an EMT event within the HCT116-derived VMT, characterized by collective and coordinated migration of tumor cells. Arrow shows movement of a cluster of tumor cells. Scale bar is 50 μm.
Fig. 5 Lineage hierarchy reconstruction reveals differential tumor states within HCT116-derived VMT. (A) Monocle analyses for HCT116-derived VMT tumor subset showing monocle trajectory plot and pseudotime plot (inset). (B) Pseudotime heatmap for HCT116-derived VMT transition subset showing 3 differentially-expressed clusters across the pseudotime trajectory. (C) Time-lapse confocal imaging showing an EMT event within the HCT116-derived VMT, characterized by collective and coordinated migration of tumor cells. Arrow shows movement of a cluster of tumor cells. Scale bar is 50 μm.CC BY-NC · PubMed Central
Fig. 6 scRNAseq reveals emerging SW480 tumor heterogeneity in the VMT platform. (A) Experimental design. SW480 CRC cells were grown as 2D or 3D (spheroid) monocultures, EC and fibroblasts were grown as 2D monocultures and then pooled or all 3 cell types were co-cultured within the VMT, harvested on day 7 of culture and processed for droplet-based single cell RNA sequencing as separate samples. (B) Heatmap of top 10 differentially expressed genes with select cluster-discriminative genes noted. (C) Distribution of each cluster from the aggregated data between monolayer, spheroid and VMT. The number of clusters emerges from the resolution used in Seurat and was optimized to balance the number of clusters versus the number of unique differentially-expressed genes in each cluster. (D) Proportion of tumor cell types in VMT, monolayer and spheroid based on clustering. Monolayer consists predominantly of tumor C, spheroid of tumor A, and tumor B is found mostly in the VMT. (E) Proportion of stromal cell types in VMT and monolayer. VMT is enriched for fibroblast B population and lacks fibroblast A and EC B. (F) Immunofluorescent staining of EPCAM and KRT5 in monolayer and the VMT reveals higher expression in the VMT. Scale bar is 50 μm.
Fig. 6 scRNAseq reveals emerging SW480 tumor heterogeneity in the VMT platform. (A) Experimental design. SW480 CRC cells were grown as 2D or 3D (spheroid) monocultures, EC and fibroblasts were grown as 2D monocultures and then pooled or all 3 cell types were co-cultured within the VMT, harvested on day 7 of culture and processed for droplet-based single cell RNA sequencing as separate samples. (B) Heatmap of top 10 differentially expressed genes with select cluster-discriminative genes noted. (C) Distribution of each cluster from the aggregated data between monolayer, spheroid and VMT. The number of clusters emerges from the resolution used in Seurat and was optimized to balance the number of clusters versus the number of unique differentially-expressed genes in each cluster. (D) Proportion of tumor cell types in VMT, monolayer and spheroid based on clustering. Monolayer consists predominantly of tumor C, spheroid of tumor A, and tumor B is found mostly in the VMT. (E) Proportion of stromal cell types in VMT and monolayer. VMT is enriched for fibroblast B population and lacks fibroblast A and EC B. (F) Immunofluorescent staining of EPCAM and KRT5 in monolayer and the VMT reveals higher expression in the VMT. Scale bar is 50 μm.CC BY-NC · PubMed Central
Fig. 7 Pathway enrichment for VMT-derived cells. (A) VMT-derived tumor pathway enrichment for HCT116 and SW480 compared to monolayer cultures. (B) Violin plots for Wnt-related target genes and signaling modulators for HCT116 and SW480. Wnt high cells correlate with a stem-like signature. (C) Pathway enrichment for stromal populations in HCT116 and SW480 VMTs. All pathways shown were below a cut-off of p-value < 0.001 and combined score > 100.
Fig. 7 Pathway enrichment for VMT-derived cells. (A) VMT-derived tumor pathway enrichment for HCT116 and SW480 compared to monolayer cultures. (B) Violin plots for Wnt-related target genes and signaling modulators for HCT116 and SW480. Wnt high cells correlate with a stem-like signature. (C) Pathway enrichment for stromal populations in HCT116 and SW480 VMTs. All pathways shown were below a cut-off of p-value < 0.001 and combined score > 100.CC BY-NC · PubMed Central
Fig. 8 Targeting TGF-β signaling with galunisertib blocks tumor growth in the VMT but not in monolayer or spheroids. (A) Violin plots showing upregulation of genes involved in TGF-β signaling in the SW480-VMT derived fibroblast B population. (B) The TGF-βR1 inhibitor galunisertib dose-dependently inhibits SW480 growth in the VMT. Tumor size in both the VMT and the spheroid was normalized to 1 on day 0, the first day of drug treatment. Data are mean and SD from n = 16. Statistical significance was calculated by unpaired t-test. (C) Galunisertib has no effect on SW480 spheroid growth. (D) XTT assay for SW480 cells grown in monolayer shows no significant difference between control and galunisertib-treated conditions at 48 hours. (E) As for (D) but at 96 hours (48 hours post-treatment).
Fig. 8 Targeting TGF-β signaling with galunisertib blocks tumor growth in the VMT but not in monolayer or spheroids. (A) Violin plots showing upregulation of genes involved in TGF-β signaling in the SW480-VMT derived fibroblast B population. (B) The TGF-βR1 inhibitor galunisertib dose-dependently inhibits SW480 growth in the VMT. Tumor size in both the VMT and the spheroid was normalized to 1 on day 0, the first day of drug treatment. Data are mean and SD from n = 16. Statistical significance was calculated by unpaired t-test. (C) Galunisertib has no effect on SW480 spheroid growth. (D) XTT assay for SW480 cells grown in monolayer shows no significant difference between control and galunisertib-treated conditions at 48 hours. (E) As for (D) but at 96 hours (48 hours post-treatment).CC BY-NC · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

1 more claim about this paper

Create a free account to see every claim other papers make about this one.

See all claims — free account

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 4

  1. RStudio Team, RStudio: Integrated Development Environment for R, RStudio, Inc., Boston, MA, 2016
  2. 10XGenomics, Vartrix single-cell genotyping tool
  3. Tickle T., Tirosh I., Georgescu C., Brown M. and Haas B., inferCNV of the Trinity CTAT Project, Klarman Cell Observatory, Broad Institute of MIT and Harvard, Cambridge, MA, USA, 2019
  4. Sharma G. and Stein-O'Brien G., projectR: Functions for the projection of weights from PCA, CoGAPS, NMF, correlation, and clustering, 2016

Cited by

See all 127

Metadata

Authors
Hachey, S. J., Movsesyan, S., Nguyen, Q. H., Burton-Sojo, G., Tankazyan, A., Wu, J., Hoang, T., Zhao, D., Wang, S., Hatch, M. M., Celaya, E., Gomez, S., Chen, G. T., Davis, R. T., Nee, K., Pervolarakis, N., Lawson, D. A., Kessenbrock, K., Lee, A. P., Lowengrub, J., Waterman, M. L., & Hughes, C. C. W.
Affiliations
  • University of California, Irvine (US)
  • Department of Molecular Biology and Biochemistry
Corresponding author
Hughes, Christopher C. W.
Venue
Lab on a Chip
Published
2021
Volume
21
Issue
7
Pages
1333-1351
ISSN
  • 1473-0189 (ISSN-L)
  • 1473-0197
OpenAlex
W3130617653
Cited by
124
References
50
Citations by year
2026: 19 · 2025: 21 · 2024: 30 · 2023: 24 · 2022: 18 · 2021: 11
FWCI
6.53
Type
journal article
Language
en
Open access
Locations
  • Lab on a Chip (publishedVersion, cc-by-nc)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
Funders
  • Edwards Lifesciences
  • Chao Family Comprehensive Cancer Center (P30A062203)
  • National Institutes of Health (CA217378, CA180122, TR-000481)
Topics
  • Physical Sciences › Engineering › Biomedical Engineering › 3D Printing in Biomedical Research
  • Health Sciences › Medicine › Oncology › Cancer Cells and Metastasis
  • Physical Sciences › Engineering › Biomedical Engineering › Microfluidic and Capillary Electrophoresis Applications
Keywords
colorectal cancer, tumor-on-a-chip, microphysiological systems, microfluidics, tumor microenvironment, drug screening, tumor heterogeneity, tumor-stroma interactions, preclinical model validation, multicellular spheroids, gene expression
MeSH terms
Antineoplastic Agents, Drug Evaluation, Preclinical, Humans, Colorectal Neoplasms, Lab-On-A-Chip Devices, Tumor Microenvironment

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.