Mesenchymal stem cells support growth and organization of host-liver colorectal-tumor organoids and possibly resistance to chemotherapy

Abstract

Despite having yielded extensive breakthroughs in cancer research, traditional 2D cell cultures have limitations in studying cancer progression and metastasis and screening therapeutic candidates. 3D systems can allow cells to grow, migrate, and interact with each other and the surrounding matrix, resulting in more realistic constructs. Furthermore, interactions between host tissue and developing tumors influence the susceptibility of tumors to drug treatments. Host-liver colorectal-tumor spheroids composed of primary human hepatocytes, mesenchymal stem cells (MSC) and colon carcinoma HCT116 cells were created in simulated microgravity rotating wall vessel (RWV) bioreactors. The cells were seeded on hyaluronic acid (HA)-based microcarriers, loaded with liver-specific growth factors and ECM components. Only in the presence of MSC, large tumor foci rapidly formed inside the spheroids and increased in size steadily over time, while not greatly impacting albumin secretion from hepatocytes. The presence of MSC appeared to drive self-organization and formation of a stroma-like tissue surrounding the tumor foci and hepatocytes. Exposure to a commonly used chemotherapeutic 5-FU showed a dose-dependent cytotoxicity. However, if tumor organoids were allowed to mature in the RWV, they were less sensitive to the drug treatment. These data demonstrate the potential utility of liver tumor organoids for cancer progression and drug response modeling.

1. Introduction

Despite treatment advances, metastatic tumor progression is still not well understood, in particular the cause of initiation of mechanisms behind activation of tumor cell growth and metastasis, and how metastatic tumor sites respond to therapies [1, 2]. Metastasis from colon to liver is common in patients with colon carcinoma, leading to the 2^nd highest number of cancer-related deaths in the U.S [3]. Patients with hepatic metastatic disease from colorectal cancer (CRC) have a poor prognosis, with 5-year survival rates below 5% [4]. Unfortunately, existing model systems in which scientists study cancer often focus on the primary site, and not necessarily the metastatic sites, essentially rendering many in vitro metastasis studies incomplete. Additionally, cancer research has been limited due to the inability to accurately model tumor progression and signaling mechanisms in a controlled environment. Animal models allow limited manipulation and study of these mechanisms, but are not necessarily predictive of results in humans [5]. Traditional in vitro 2D cultures fail to recapitulate the 3D microenvironment of in vivo tissues, drug diffusion kinetics, or patient drug scaling [6–8]. Furthermore, the differences cells experience on tissue culture dishes can place a selective pressure on cells that could substantially alter their molecular and phenotypic properties, rendering them ineffective for in vitro studies. We recently demonstrated that on 2D tissue culture dishes, metastatic CRC cells appeared epithelial, but when transitioned to 3D organoid environments they “switched” to a mesenchymal and metastatic phenotype [9, 10].

Bioengineered 3D tissue and tumor models can be biofabricated using a variety of technological approaches. Examples include bioprinting technologies [11–16], 3D photopatterning [17], cellularized soft lithography [18, 19], and rotating wall vessel (RWV) bioreactor culture [20–25]. The latter, RWV culture, is an established methodology that employs a rotating bioreactor to generate low fluid shear stress rotational forces, which simulates microgravity conditions. Simulated microgravity allows cells to self-aggregate into spheroids or to nucleate around microcarrier beads for adherent suspension cultures. To date, a wide variety of tissue types have been modeled as 3D organoids using RWV technology, including lung, colon, intestine, liver, vaginal epithelium, breast, and others [20–22, 24–28]. This platform is built on the combination of a modular hydrogel platform [15, 29, 30], which has been demonstrated extensively in the application of biofabricating tissue and tumor organoids [9, 25]. Our team recently published a study using this platform to create 3D organoids containing a hepatic cells line, Hep-G2, and metastatic colorectal cancer cells, HCT116 [9]. The organoids supported HCT116 tumor cell growth over time, induced expression of in vivo-like mesenchymal and metastatic markers, including active signaling pathways, and responded to chemotherapeutical drugs. However, Hep-G2 cells are transformed cancer cells and have lost much of their ability to metabolize drugs notably having lost expression of CYP3A4 – the most common isozyme of P450 in the liver.[31] This loss of function means that while Hep-G2 cells, are robust and easy to propagate, they are a sub-optimal model cell to study liver toxicity and potential metabolic interactions.

Liver tissue has an organized anatomy to support hepatic function, in which specific hepatic cells are localized. The hepatocytes are the major cell type that provides the liver function, along with liver sinusoidal endothelial cells and biliary epithelial cells. The hepatic stellate cells, of a mesenchymal origin, are imperative to liver regeneration, but are difficult to culture in vitro. Importantly, these cells have shown to support the growth of liver metastases [32]. In the current study we describe new 3D liver tumor organoids comprised of primary human hepatocytes, HCT116 colorectal cancer cells, and mesenchymal stem cells (MSC) as a surrogate for the stellate cells. We are using this tumor organoid model to observe tumor cell growth and tumor tissue maturation, and perform anti-cancer drug studies.

2. Materials and Methods

2.1. Liver-specific hydrogel microcarrier preparation

HA hydrogel-coated microcarrier beads were prepared as previously described [25], but together with a liver ECM supplement that has been employed to create cell-supportive liver ECM hydrogels [11, 33]. For microcarrier preparation, cross-linked dextran Sephadex G-50 beads (GE Healthcare Biosciences, Uppsala, Sweden) were coated with a formulation of Hystem-HP hydrogel (ESI-BIO, Alameda, CA) without the PEGDA cross-linker. Specifically, the thiol-functionalized hyaluronic acid and thiol-functionalized gelatin were allowed to oxidatively crosslink by disulfide bond formation over time. The hydrogel solution was prepared by dissolving Heprasil and Gelin-S (thiolated and heparinized HA and thiolated gelatin components, respectively) in water to make 2% w/v solutions, and then mixed in equal volumes. Sephadex beads (0.5 g) were added to a round bottom flask and the pressure was reduced in the flask using a vacuum. A valve was opened to allow 10 mL of the ungelled Extracel solution to flow into the flask. The reduced pressure environment allowed the aqueous hydrogel solution to be taken up into the pores of the beads. Excess liquid was removed, and the hydrogel solution was allowed to crosslink overnight in air by disulfide bond formation. The coated beads were then transferred to 50 mL centrifuge tubes, frozen, and lyophilized to yield the HA-coated microcarriers (HAMs).

2.2. Cell culture

Human colon carcinoma cells (HCT116, expressing red fluorescent protein [RFP]) were expanded in 2D on tissue culture plastic using 15 cm tissue-treated dishes until 90% confluence with Dulbecco’s Minimum Essential Medium (DMEM, Sigma, St. Louis, MO), containing 10% fetal bovine serum (FBS, Hyclone, Logan, UT). Cells were detached from the substrate with Trypsin/EDTA (Hyclone) and resuspended in media before use in further studies. Mesenchymal stem cells (MSC) were sourced from bone marrow-derived mesenchymal stem cells (Lonza, Walkersville, MD) and expanded on tissue-treated dishes 90% confluence with α-Minimum Essential Medium (α-MEM, Lonza), containing 10% FBS. Cells were detached as described above. Primary human hepatocytes were purchased from Triangle Research Laboratories (RTP, NC). Hepatocytes were thawed according to manufacturer instructions using Hepatocyte Thawing Medium (Triangle Research Labs) and used immediately after thawing.

2.3. Rotating wall vessel bioreactor culture

Prior to cell culture, HAMs were sterilized in Ca²⁺ and Mg²⁺-free PBS (HyClone, Logan, UT) by autoclaving at 115 °C for 15 min. The sterilized beads were then stored at 4 °C. Cell culture medium comprised of the DMEM described above with the addition of amphotericin B was used for all bioreactor culture studies. In preparation for seeding on HAMs, cells were initially cultured as described above. Trypsin–EDTA was used to free the cells from their substrate. Cells and HAMs in medium were added to a 50 mL RWV bioreactor (Synthecon, Houston, TX) to reach a final density of 80,000 cells/5 mg beads/mL medium with a ratio of 10:1:1 Hepatocytes:HCT116:MSC. Rotation of the RWV bioreactors was started immediately and continued for 14 days. Medium was first changed on day 5 of culture in order to allow the cells to bind to the beads, after which medium was changed every 2^nd day. The RWV bioreactors were set to initially rotate at 18 rpm; the rate of rotation was manually increased throughout culture to keep the clusters in “freefall” as they increased in size. Aliquots containing organoids were removed on days 5, 8, 11, and 14 days post-seeding and used to assess growth through size quantification of light microscopy images (n of 10 or higher) and MTS assays (n of 4) quantified at 490 nm on a plate reader. In addition, composite images were taken in which the organoids were also imaged with epifluorescence at 594 nm to analyze the progression of the RFP-labeled HCT116 cells within the organoids. Proportion of RFP-labeled HCT116 regions in the organoids was then determined using a custom MatLab scrypt to analyze the composite images, as described previously [34].

2.4. Liver function characterization

Urea and albumin production were measured by collecting supernatant from the RWV cultures. Urea production was measured using a colorimetric assay, Quantichrom Urea Assay Kit, (BioAssay Systems) following manufacturer’s instructions. Samples were measured in a 96-well clear assay plate (Corning) using plate reader set to 430nm (SpectraMax M5, Molecular Devices). Data were analyzed using two-sample unequal variance t-test. Albumin production was measured using Human Albumin ELISA kit (Alpha Diagnostic International) according to manufacturer’s instructions. Samples were measured using plate reader set to 450nm (SpectraMax M5, Molecular Devices) and data were analyzed using two-sample unequal variance t-test.

2.5. Histology and immunohistochemistry (IHC)

Aliquots of organoids removed from the RWV on day 5, day 8, and day 11 and organoids were fixed with 4% paraformaldehyde for 1 hour, dehydrated with graded ethanol washes followed by xylene, embedded in paraffin, and sectioned at 5 μm. All incubations were carried out at room temperature unless otherwise stated. Slides were warmed at 60° C for 1 hr to increase bonding to the slides. Antigen retrieval was performed on all slides and achieved with incubation in Proteinase K (Dako, Carpinteria, CA) for 5 min. Sections were permeabilized by incubation in 0.05% Triton-X for 5 min. Non-specific antibody binding was blocked by incubation in Protein Block Solution (Abcam, cat. #ab156024) for 15 min. Sections were incubated for 60 min in a humidified chamber with the primary CD-44 (raised in rabbit, cat. #ab157107, Abcam) and matrix metalloproteinase 9 (MMP9, raised in rat, cat. # ab38898, Abcam) antibodies each at 1:200 dilution in antibody diluent (Abcam, cat. #ab64211).

Following primary incubation, slides were washed 3 times in PBS for 5 min. Samples were then incubated for 1 hr with anti-rabbit, anti-mouse, or anti-rat Alexa Fluor 488 secondary antibodies (Invitrogen) as appropriate in antibody diluent (1:200 dilution). Cells were counterstained with DAPI for 5 minutes, and washed 3 times with 1X PBS prior to fluorescent imaging. Negative controls were performed in parallel with the primary antibody incubations and included incubation with blocking solution in place of the primary antibody. No immunoreactivity was observed in the negative control sections. Samples were imaged with fluorescence at 488 nm, 594 nm, and 380 nm with a Leica DM 4000B upright microscope.

2.6. Drug treatments

5-FU (5-FU, Sigma-Aldrich, cat. #F6627) was solubilized in DMSO then added to DMEM-containing culture media at a concentration of 100mM and sterilized by filtrations. Organoids were removed from the bioreactor at days 6 or day 10 and incubated with 0, 1, 10, or 100 mM 5-FU for 24 hours as previously described [9].

MTS assays on drug treated organoids were performed as described above for untreated organoids. Live-dead assay was performed by subjecting treated organoids to Live/Dead kit solution as instructed in the manufacturer’s protocol (ThermoFisher, cat. #L3224). The organoids were incubated in live-dead solution for 1 hr at 37C then imaged immediately.

2.7. Macro-confocal microscopy analysis

Confocal microscropy was performed on a Leica TCS-LSI macroconfocal microscope. Aliquots of organoids were removed from the RWV, fixed, stained, and then carefully placed on a microscope slide for imaging. Excess moisture was removed from the sample to decrease noise. Laser power was set to 30%. Gain was set to maximize signal while minimizing over-exposure. Organoids were imaged using full-length z-stacks which were then compressed into a single image with maximum projection processing.

2.8. Statistical Analysis

The data are generally presented as the means of number of replicates ± the standard deviation. All experiments were performed with n = 3 or higher. Values were compared using Student’s t-test (2-tailed) with two sample unequal variance, and p < 0.05 or less was considered statistically significant.

3. Results and discussion

3.1. Tissue-tumor organoid formation and progression

We followed a protocol as described previously to produce primary hepatocyte and tumor organoids. To create 3D liver tumor organoids, combinations of primary hepatocytes, MSCs, and HCT116 cells at a 10:1:1 ratio, or primary hepatocytes and HCT116 cells alone at a 10:1 ratio, and microcarriers coated in a liver specific hydrogel, based on hyaluronic acid and gelatin, as previously described [9]. HCT116 cells were genetically labeled with RFP to facilitate visualization in a multiple cell-type organoid and measureing their growth. The cells and microcarriers were added to RWV bioreactor vessels and cells generally adhered to and proliferated around the hydrogel-coated microcarriers over time, resulting in 3D organoid formation (Figure 1A). Importantly, within days, we observed that the presence of MSCs was essential to organoids’ formation. In the absence of MSCs, only a few cells adhered to the microcarriers or to each other and failed to form cellular organoids.

Figure 1. Colon carcinoma cell growth inside liver tumor organoids.

(A) Schematic drawing of liver tumor organoid formation in the RWV bioreactor. (B) MSCs are required for organoid formation. (C) Bright light microscopy of liver tumor organoids that include MSCs. (D) HCT116 cell growth inside the liver tumor organoids is monitored by overlaying fluorescent with bright light images, showing progressive increase in tumor regions over time. (E-I) Measurements of organoids’ size, proportion of red fluorescent protein-labeled HCT116 cells within the organoids, metabolic activity, albumin and urea production over time. The presence of MSCs drove an increase in organoids’ size metabolic activity and urea secretion but had little effect on albumin secretion Statistical significance between compared groups: * p < 0.05.

Organoid size and growth was measured on days 5, 8, 11, and 14, by light and fluorescent microscopy of small aliquots containing organoids, removed from the RWV. Figure 1C shows representative images of organoids composed of a mixture of primary hepatocytes, MSCs, and HCT116 cells. Overlaying a fluorescent image of the RFP-labeled HCT116 cells onto bright light images (Fig 1D) allowed us to estimate tumor cell proportion without disturbing the overall structure of the organoid. At day 5, tumor cells are lightly dispersed through the organoid volumes. On days 8 and 11, tumor cells rapidly expanded to assume a large portion of the overall organoid volume, as evidenced by both the area and intensity of the RFP-labeled cells. Organoids containing MSCs formed into roughly spherical aggregates and in contrast, organoids produced without MSCs formed significantly smaller (p < 0.05) and unstable spheroids through day 8 and were unable to sustain aggregate growth by days 11 and 14 (Figure 1E). Overall metabolic activity increased through day 14 in organoids containing MSCs (Figure 1F). Conversely, metabolic activity was significantly lower in organoids without MSCs (p < 0.05). Next, we employed an in-house developed MatLab script [34] to quantify the proportion of red fluorescent tumor cells within the organoids in images such as those in Figure 1D. Figure 1G shows that the proportion of tumor in the organoids consistently increased over time, reaching up to approximately 70% of the organoid at day 14.

Albumin and urea production are 2 of the primary functions of hepatocytes, and we used their quantities as a measure of hepatocyte function in the liver tumor organoids. We assessed the effects of MSC interaction by assaying the concentration of albumin and urea in media samples from the RWVs containing organoids produced with MSCs and those without, as shown in. Albumin concentrations were similar in media samples from either organoid type, at both time points (Figure 1H). Urea levels are higher in MSC-containing organoids, but drop in both organoid combinations from day 5 to 8, significantly more in the absence of MSCs (Figure 1I). These results suggest that while the presence of MSCs do seem to be required to support aggregation of cells into organoid form, they may not have a significant effect on hepatocyte albumin and urea secretion.

3.2. Stroma-driven organoid self-organization

MSCs have an important role in tissue regeneration and repair by remodeling the ECM and secreting growth-promoting factors [35, 36]. In the liver, stellate cells move to damaged areas and aid in tissue repair [37]. Given that organoids were stable only in the presence of MSCs, we hypothesized that the MSCs were remodeling or inducing self-organization of the organoids. Hematoxylin and eosin (H&E) staining showed hydrogel present within the organoids at days 5 and 8, possibly providing a scaffold for cells to attach to and grow from. A “shell” of stromal cells began to form by day 8, and by day 11 the shell had completely enveloped the organoid, inside of which cells with larger cytoplasm, likely hepatocytes, resided. These results were confirmed by immunohistochemical staining for MMP-9, a member of the matrix metalloproteinase family and is used by cells to cleave ECM components to facilitate mobility and remodeling (Figure 2D–F). MMP-9 was initially located throughout the organoid, and by day 8 and 11, MMP-9 is mostly co-localized with RFP-positive cells at the central portion of the organoid, suggesting that the HCT116 cells are continuing to cleave and clear the hydrogel. IHC staining for CD44 was used to detect MSC, showing sparse distribution in the organoids on days 5 and 8, likely due to the low initial numbers of MSCs during the organoid formation stage. However, by day 11, CD44-positive cells were clearly observed as an organized shell at the organoids’ periphery, with interspersed RFP-positive cells which are likely HCT-116 cells that have yet to move to the spheroid core. Alternatively, presence of RFP-positive cells may be indicative of intermingling between stromal and cancer cells evidenced by the colocalization of the red and green signal. Overall, these results suggest that the MSCs had a profound effect on the formation and maturation of the liver tumor organoids similar to the supportive role stellate cells play in native liver.

Figure 2. Histological examination of liver tumor organoids containing MSCs.

(A-C) H&E staining of organoids shows steady growth in organoid size as well as the formation of a cellularized shell enveloping the organoid. Hydrogel embedded in the organoids is visible as dark purple. (D-F) MMP-9 expression changes over time from wide spread to the center of the organoids, co-localized with red fluorescent HCT116 cells and is absent in the outer shell. (G-I) CD44 expression shows a reciprocal image compared with MMP-9 positive HCT116 cells. Initially, CD44 staining is low and is gradually move to the outer shell and mark the location of MSCs. By day 11 HCT116 cells remain primarily in the core of the organoid, but are also found interspersed in the MSC-rich stromal shell of the organoid. Immunofluorescence: Green – MMP-9 (D-F) and CD44 (G-I); Blue – DAPI; Red – Red fluorescent protein (HCT116 cells).

3.3. Chemotherapeutic agent screening assessment

In previous studies, we showed that 2D models of colorectal cancer are difficult to use for chemotherapeutic studies because the window for sensitivity is very narrow [38]. Using high concentrations will kill the bulk of cells while low concentrations have little effect. One of the benefits of utilizing 3D models for drug studies is a wider range of sensitivity that allows identification of optimal drug concentrations. In addition, 3D models often exhibit sensitivity more closely matching in vivo conditions making them useful for in vitro to in vivo screening. Also, as described above, 3D systems recapitulate aspects of in vivo architecture facilitating other important aspects such as more accurate diffusion and mass transport kinetics. To test the sensitivity of the liver tumor organoids to chemotherapeutic drugs we used the nucleoside analogue 5-fluorouracil (5-FU) [39]. As per previous experiments [9], aliquots containing organoids were removed after 6 and 10 days and incubated for 24 hours with 1mM, 10mM, or 100mM of 5-FU. Metabolic activity, measured by MTS, was used to assay the cellular response to the drug treatment. Organoids removed at day 6, showed a steadily decreasing metabolic activity at each increasing concentration of 5-FU (Figure 3A). In organoids removed at day 10, the dose dependency remains, but there is smaller difference between the 10mM and 100mM 5-FU response (Figure 3B), compared with organoids removed at day 6. When normalized to the no drug treatment control, it appears that the day 10 reorganized organoids are less sensitive to the drug (Figure 3C). It is possible that the lower sensitivity of day 10 organoids to 5-FU treatment is due to the large size and the stromal reorganization of the organoids. These results are supported by H&E and live-dead staining of day 10 organoids following 5-FU treatment (Figure 3D and 3E, respectively). Although there is a noticeable decrease in areas containing viable cells with increasing 5-FU concentration, even at the higher 5-FU concentration (100mM) there is a rim of viable cells around the organoid. Based on the results shown in Figure 2, we expect that the more resistant cells are the MSCs. As such, it may stand to reason that both hepatocytes and HCT116 cells have a higher metabolism than MSCs, potentially explaining MSC resistance to the drug treatment.

Figure 3. Response to 5-FU treatments.

Organoids were removed in aliquots from the bioreactor after 6 and 10 days and incubated for 24 hours with 0, 1mM, 10mM, and 100mM of 5-FU. (A-B) Metabolic activity, measured by MTS reagent, was used to assay the cellular response to the 5-FU treatment showing a dose response curve on day 6 (A) or 10 (B) of RWV culture and then assayed for metabolic activity. At both time points, organoids responded to 5-FU in a dose dependent manner. (C) However, when the data in A and B from these two drug screening experiments were normalized to the non-treated organoids’ control condition and overlaid (C), it appears that organoids that were cultured through day 10 of the culture were less sensitive to 10 and 100 mM 5-FU than organoids from day 6. Statistical significance: * p < 0.05. (D-E) H&E and LIVE/DEAD staining of day 10 organoids after 5-FU treatment Drug treated organoids were fixed, sectioned, stained with H&E and imaged (D), or stained with LIVE/DEAD reagents and imaged whole on a macro-confocal microscope (E). Under drug treatments, organoids show a dose dependent live to dead ratio. These organoids from the day 10 5-FU study appear to showing a partially viable internal cellular content at 10mM 5-FU, whereas at 100mM 5-FU live cells are mostly only visible as a semblance of the outer stromal layer composed of MSCs, suggesting they may protect the liver-tumor core at lower concentrations, but fail to do so at higher concentrations despite maintaining viability in the layer itself. Scale bars: 100μm.

Discussion

For many years cancer research experiments were aimed at the cancer cell alone, devoid of its surroundings and at non-physiological conditions such as tissue culture plastic dishes. Notably, tissue culture plastic is several orders of magnitude greater in stiffness than any of the soft tissues in the body [11, 40–42], and as such is an inadequate resource for trying to recapitulate the tumor microenvironment. In recent years, researchers have made substantial findings with regards to cancer-stroma and cancer-ECM effects, largely referred to as the cancer microenvironment. The cancer microenvironment is a complex space that contains stromal cells, a multitude of ECM components and proteins, as well as a plethora of signaling, paracrine, and growth factors. Together, these components of the microenvironment push and pull cancer cells between phenotypes and have a significant effect on the long-term progression of a tumor and response to therapy [1, 2]. To better capture the interactions between the cancer cell and its microenvironment, we have developed a host-liver colorectal-tumor organoid replete with primary hepatocytes and mesenchymal stromal cells as a substitute for hepatic stellate cells, as tumor stromal elements for metastatic colon carcinoma cells. Inclusion of hepatocytes allows assay of liver toxicity stemming from chemotherapeutics, adds the interference of hepatocyte metabolism to drug response study, as well as mimics the cellular environment of the native liver as perceived by a metastatic colorectal cancer cell. Under pathological conditions, hepatic stellate cells have been shown to activate and transform towards a myofibroblast phenotype that supports the growth of cancers in the liver by rapidly producing ECM [43]. Indeed, liver tumor organoids failed to grow in the absence of MSCs, indicating that the MSCs provide essential support for tumor growth, similar to hepatic stellate cell in liver cancer. By tracking fluorescently labeled HCT116 cells, we demonstrated the active proliferation of the tumor cells when the MSCs were present. Besides supporting tumor cell growth in the liver organoids, inclusion of MSCs in these organoids resulted in tumor-like tissue organization and maturation. The MSCs migrate to the periphery of the organoid and created an organized shell-like tissue encapsulating the tumor cells and hepatocytes at the core of the organoid. Interestingly, our past study utilizing Hep-G2 cells did not require MSC inclusion, indicating Hep-G2 cells may possess stromal remodeling capability that hepatocytes do not.[38] These results further demonstrate the capacity of the RWV bioreactor conditions to create a more physiologically relevant tumor models compared with tumor cells in cell culture dishes.

One of our main motivations for this system is its integration into anti-cancer drug studies. Drug development costs sky-rocket after the initial screening on 2D cell cultures, when moved into human trials, mainly because the cell culture models fail to recapitulate the complex behavior of cells inside the tumor. At the same time, many candidate drugs drop during the cell culture screening and are missing from the discovery pipeline. This drawback of conventional drug development studies necessitates an easy to build and test model that integrates human cells to more accurately mimic in vivo physiology. We hypothesized that the liver tumor organoid system will better recapitulate the response of both the tumor and stromal cells to chemotherapeutic drug treatments. Our results demonstrated not only a dose dependent response of the tumor liver organoid to a range of 5-FU concentrations but also that more organized organoids were less sensitive to the treatment. With our RWV-based culture, we are also able to generate a large number of organoids in a single batch, facilitating high-throughput drug studies.

There is a growing need for more complex and predictive in vitro models as drug development costs continue to grow. The host-liver colorectal-tumor organoid described here shows potential for seamlessly integrating into the drug development pathway and streamlining drug discovery. Many chemotherapeutics are prolific killers of cancer, but are also highly toxic to healthy tissue; one of the major attractions of a model like this is the option to study the effects of a drug on the host tissue in addition to the cancer cell. Further iterations on this model design include several additional features. First, we wish to assess the role of the MSC component in more detail, comparing them to primary human stellate cells. While specific to liver, primary stellate cells are less widely available as a commercial cell source, while MSCs sources are plentiful. Additionally, we are actively exploring will more completely integrate measures of hepatocyte response to drug compounds, including viability, metabolic activity, and functional output such as albumin and urea secretion, and drug metabolism capabilities. This platform with these expanded output metrics will facilitate the development of chemotherapeutics, which specifically target cancer cells and spare host tissue. Our system also provides a framework for adaptation to other tissue types as well as integration of patient specific samples. These avenues position our organoid-based approach as a transformative platform for use in drug screening, toxicology studies, and personalized medicine.