Integrated genome and tissue engineering enables screening of cancer vulnerabilities in physiologically relevant perfusable ex vivo cultures

Abstract

Genetic screens are powerful tools for both resolving biological function and identifying potential therapeutic targets, but require physiologically accurate systems to glean biologically useful information. Here, we enable genetic screens in physiologically relevant ex vivo cancer tissue models by integrating CRISPR-Cas-based genome engineering and biofabrication technologies. We first present a novel method for generating perfusable tissue constructs, and validate its functionality by using it to generate three-dimensional perfusable dense cultures of cancer cell lines and sustain otherwise ex vivo unculturable patient-derived xenografts. Using this system we enable large-scale CRISPR screens in perfused tissue cultures, as well as emulate a novel point-of-care diagnostics scenario of a clinically actionable CRISPR knockout (CRISPRko) screen of genes with FDA-approved drug treatments in ex vivo PDX cell cultures. Our results reveal differences across in vitro and in vivo cancer model systems, and highlight the utility of programmable tissue engineered models for screening therapeutically relevant cancer vulnerabilities.

Keywords: Bioprinting; Breast cancer model; CRISPR screens; Medulloblastoma model; Point-of-care functional screens.

Fig. 1.

3D-Printing of Perfused Tissue Constructs and Transcriptomic Profiling of MDA-MB-231 Tumor Models. (A) Schematic representation of the perfusable construct manufacturing procedure. A PVA scaffold of desired geometry is printed and inserted into a silicone holder. The PVA scaffold is then encapsulated within a matrix formulation of desired composition. The matrix formulation is allowed to gelate and simultaneously, the PVA scaffold slowly dissolves. The scaffold is then evacuated using warm media. At this stage, the construct can be perfused, though nutrients will be distributed only by diffusion. The construct is then removed from the holder, and placed into the chamber of a larger silicone construct that allows for open perfusion on all sides. At this stage, interstitial flow may be introduced, and perfusion is possible around as well as within the construct. (B) Fluorescent images (left) showing the growth of GFP-labeled MDA-MB-231 cells in matrices at low flow rates without interstitial flow (15 μL/min) and high flow rates with interstitial flow (500μmL/min). Scale bars: 2 mm. Cell density counts (right) for MDA-MB-231 cells in matrices at low flow rates (n = 6) and high flow rates (n = 10). Error bars show standard deviation. (C) Schematic representation of MDA-MB-231 culture conditions. MDA-MB-231 cells are distributed across four conditions: 2D-cell culture, static matrices, perfused printed constructs, and tumors in mice injected orthotopically in the mammary fat pad. Conditions are sustained over 2 weeks, after which RNA is extracted, amplified, and sequenced. (D) Hierarchical clustering diagram comparing relative transcriptomic profiles of all expressed genes in MDA-MB-231 cells grown in various conditions (n = 2 for each condition). (E) Top-20 Metascape-generated enriched (left) and depleted (right) Gene Ontology Biological Process Pathways that are shared by perfused print and tumor conditions in comparison to the cell culture condition.

Fig. 2.

Kinome-wide CRISPRko screens in MDA-MB-231 cells cultured in in vitro 2D conditions, in the ex vivo perfused 3D model, and in in vivo orthotopic locations in mice. (A) Schematic representation of the kinome-wide CRISPR knockout screen. MDA-MB-231 cells are lentivirally transduced with a 3152-element CRISPR knockout library. Cells are collected and distributed across three conditions: 2D-cell culture, perfused printed constructs, and tumors in mice injected orthotopically in the mammary fat pad. Conditions are sustained over 4 weeks, after which gDNA is extracted, amplified, sequenced, and processed. (B) Results of hierarchical clustering performed for top hit genes of all individual replicates for each treatment condition. (C) Venn diagram comparing top depletion and enrichment hits for cell culture, mammary tumor, and perfused print growth conditions, as determined by the MAGeCK algorithm. Criteria consisted of p-value < 0.0027, as well as a LFC value reflecting status as an enriched or depleted hit.

Fig. 3.

Transcriptomic profiling of PDX medulloblastoma cells cultured ex vivo in various model conditions. (A) Schematic representation of PDX medulloblastoma culture conditions. PDX medulloblastoma cells are isolated from tumors, then distributed across several conditions: suspension cell culture, static matrices, perfused printed constructs, and tumors in mice injected orthotopically in the cerebellum. Conditions are sustained over 10 days, after which RNA is extracted, amplified, and sequenced. (B) Hierarchical clustering diagram comparing relative transcriptomic profiles of all expressed genes in PDX medulloblastoma cells grown in suspension cell culture, static matrix, perfused print, and orthotopic tumor conditions (n = 2 for each condition). (C) Top-20 Metascape-generated enriched (left) and depleted (right) Gene Ontology Biological Process Pathways that are shared by perfused print and tumor conditions in comparison to the suspension cell culture condition. (D) Survival of PDX medulloblastoma cells over time when grown in suspension cell culture (left), and comparison of PDX medulloblastoma cell survival in suspension cell culture vs. perfused prints (right).

Fig. 4.

Targeted CRISPRko screen in PDX medulloblastoma cells cultured in the ex vivo perfused 3D model, and in in vivo orthotopic locations in mice. (A) Overview of CRISPR knockout library design. Target genes of single-guide RNAs were assembled from a list of genes with FDA-approved drugs targeting their products. (B) Schematic representation of the targeted CRISPR knockout screen. PDX-medulloblastoma cells are isolated from tumors, and lentivirally transduced with a 74-element CRISPR knockout library. Cells are collected and distributed in perfused printed constructs and tumors in mice injected orthotopically in the cerebellum. Conditions are sustained over 6 weeks, after which gDNA is extracted, amplified, sequenced, and processed. (C) Results of a principal component analysis performed across all genes of all individual replicates for each treatment condition. (D) Comparison of effect sizes (Z-Score) for all 74 sgRNA targets in perfused prints and orthotopic tumors. AAVS1 and NTC respectively represent sgRNAs targeting the adeno-associated virus integration site 1 and non-targeting controls, both serving as controls. (E) Metabolic activity measurements of PDX medulloblastoma cells grown in perfused prints when treated with various drugs inhibiting top depletion targets of the CRISPRko screen (n = 3 with P values *P < 0.05 and **P < 0.01). (F) Timeline for a potential point-of-care functional oncology setting outlining the use of focused CRISPR screens in a perfused printed model for patient-specific cancer vulnerability screening.