We describe a protocol for the production, culture, and visualization of human cancers, which have metastasized to the peritoneal surfaces. Resected tumor specimens are cut using a vibratome and cultured on permeable inserts for increased oxygenation and viability, followed by imaging and downstream analyses using confocal microscopy and flow cytometry.
Pseudomyxoma peritonei (PMP) is a rare condition that results from the dissemination of a mucinous primary tumor and the resultant accumulation of mucin-secreting tumor cells in the peritoneal cavity. PMP can arise from various types of cancers, including appendiceal, ovarian, and colorectal, though appendiceal neoplasms are by far the most common etiology. PMP is challenging to study due to its (1) rarity, (2) limited murine models, and (3) mucinous, acellular histology. The method presented here allows real-time visualization and interrogation of these tumor types using patient-derived ex vivo organotypic slices in a preparation where the tumor microenvironment (TME) remains intact. In this protocol, we first describe the preparation of tumor slices using a vibratome and subsequent long-term culture. Second, we describe confocal imaging of tumor slices and how to monitor functional readouts of viability, calcium imaging, and local proliferation. In short, slices are loaded with imaging dyes and are placed in an imaging chamber that can be mounted onto a confocal microscope. Time-lapse videos and confocal images are used to assess the initial viability and cellular functionality. This procedure also explores translational cellular movement, and paracrine signaling interactions in the TME. Lastly, we describe a dissociation protocol for tumor slices to be used for flow cytometry analysis. Quantitative flow cytometry analysis can be used for bench-to-bedside therapeutic testing to determine changes occurring within the immune landscape and epithelial cell content.
Pseudomyxoma peritonei (PMP) is rare syndrome with an incidence rate of 1 per million people per year1. Most PMP cases are caused by metastases from appendiceal neoplasms. Given that mice do not have a human-like appendix, modeling this type of cancer remains extremely challenging. While the primary disease is often curable by surgical resection, treatment options for metastatic disease are limited. Therefore, the rationale for developing this novel organotypic slice model is to study the pathobiology of PMP. To date, there are no appendiceal organoid models that can be perpetually cultured; however, a recent model was shown to be useful for the pharmacological testing of therapeutic agents and immunotherapy2. As such, we have adapted an organotypic slice culture system, which has been used in other types of human cancers, such as brain, breast, pancreas, lung, ovarian, and others3,4,5,6.
In addition to appendiceal neoplasms, PMP occasionally results from other tumor types, including ovarian cancers7, and in rare circumstances, intraductal papillary mucinous neoplasms8 and colon cancer9. Additionally, these tumors tend to grow slowly, with poor engraft rates in patient-derived xenograft (PDX) models10,11. Given these challenges, there is an unmet need to develop models to study this disease to begin to understand the pathobiology of PMP, and how these cancer cells: are recruited to the peritoneal surfaces, proliferate, and escape immune surveillance.
While cut from the systemic vascular circulation, tumor slices do contain cellular and acellular components, including the extracellular matrix, stromal cells, immune cells, cancer cells, endothelial cells, and nerves. This semi-intact microenvironment allows for the functional investigation of these cell types, which is uniquely advantageous compared to 3D organoid cultures, which consist only of cancer cells12. While organotypic slice cultures are advantageous in some respects, they are also inherently a low-throughput-based approach, compared to 3D organoids, which can be expanded, and are suitable for multiplexed investigational therapeutic drug screening13,14,15. In the case of PMP, there have been no reports documenting reliable establishment and perpetual passaging of PMP-derived organoids16. This is likely due to the slow growing nature of PMP-derived tumor cells, as well as the low number of malignant epithelial cells found within these mucinous tumors. Given the need to develop models to study PMP, organotypic slices are uniquely suited to study this disease. We present a protocol for preparing, imaging, and analyzing PMP from human specimens.
The deidentification and acquisition of all tissues were performed under an IRB-approved protocol at the University of California, San Diego.
1. Preparation of human PMP tissues for tissue processing and culture
2. Confocal imaging of living human tissue slices
NOTE: Once the organotypic tumor slices have been prepared, it is essential to determine the tissue viability to perform an efficient and optimized downstream analysis. Given that tumor specimens are 150-250 µm thick, confocal, or two-photon microscopy is highly recommended over wide-field microscopes for determining studies in situ. Flow cytometry can also be used for determining the viability and cellular populations (methods are below). However, the spatial resolution is lost during flow cytometry analysis, and viability is likely underestimated given the need to disassociate tumor slices by mechanical and chemical methods.
3. Disassociation of living slices for flow cytometry
NOTE: Disassociation of living tissue slices can be used for several downstream applications, including immunotyping, assessment of the viability, and interrogation of the changes to cell populations after pharmacological intervention. Steps should be taken to ensure that tissue quality and cell viability are maintained during the disassociation process.
4. Pharmacological intervention using living tissue slices for viability and proliferation analysis
NOTE: Once the organotypic tumor slices have been prepared, interventional approaches can be performed by using several methods, including drug testing, siRNA, as well as viral infection of living tissue slices. Here we will discuss drug testing, as well as downstream functional readouts, which include viability analysis using local proliferation.
In short, human tumor specimens from PMP are obtained under an IRB-approved protocol. The tissue is prepared, micro-dissected, and solidified in an agarose mold to be cut using a vibratome (Figure 1A; Video 1). Once cut, tissue slices are placed and cultured on permeable insert membranes (Figure 1B), which can be utilized for imaging assays in situ, as well as for cellular and functional interrogation using flow cytometry, confocal imaging analysis, and cytotoxicity assays. A representative workflow of the culture and processing of human PMP tissue slices is shown in Figure 1C. Initial imaging analysis can be performed by light microscopy (Figure 2A–B). The staining of mucin in tissue slices were visualized using Periodic acid-Schiff (PAS) staining (Figure 2C) as well as mucin-specific antibodies (Figure 2D). To assess the viability, slices cut from different regions are incubated in calcein AM and propidium iodide to determine their health and viability (Figure 2E–F). Most cells (roughly 85%) are viable (calcein AM+; green), while some cells are dead (propidium iodide; red) when visualized by confocal microscopy after initial cutting. The extracellular matrix can be visualized using an incidence white light at an angular position of 180°. Tumors treated with chemotherapy prior to surgery often show more PI signal. The proliferation of tissue slices can be tracked using EdU added in the slice culture media. For detection, a commercial cell proliferation assay kit was used, and the assay was performed on fixed slices after 72 h of culture. Actively proliferating cells in culture can be measured in the FITC channel (green). The cell nuclei (DAPI) will overlap with cells that have proliferated (EdU+) during culture (Figure 2G). These results are quantified to determine the number of proliferating cells (DAPI+ nuclei) within the tumor slice (Figure 2H). The results shown here indicate that cultured slices are viable and proliferative during slice culture.
To identify immune cell functionality, Ca2+ imaging studies were performed on living tumor slices (Figure 3; Video 2). Using this approach direct and paracrine signaling mechanisms can be identified. To label local immune cells, in situ cytolabeling was performed by incubating slices with a CD11b-PE conjugated antibody along with Fluo-4-AM (Figure 3A). Pseudocolor scaled images showing intracellular Ca2+ levels better allow for visualizing differential levels of intracellular Ca2+ (Figure 3B). Spontaneous activity was tracked and time lapse images are shown for before (Figure 3C), during (Figure 3D), and after (Figure 3E) an intracellular Ca2+ response within the highlighted (white dotted circle box; Figure 3A,B) CD11b+ immune cell. The raw traces of Ca2+ responses in the CD11b immune cell (red) along with other responsive (black) and non-responsive cells (blue) are shown (Figure 3F). Quantification of the area under the curve of the raw traces shown from the responding immune cell (CD11b+; red Figure 3F) compared to the quantification of the non-responding cell (blue; Figure 3F) were quantified (Figure 3G). These results indicate that live cell functional tracking can be performed utilizing this living tissue slice platform.
Downstream analysis using flow cytometry can be used to immunotype as well as investigate changes in cellular profiles in response to therapeutic agents. Results may be due to direct signaling or indirect (paracrine) signaling. Representative results of immunotyping of a patient tumor sample with PMP are shown in Figure 4A. The tumor immuno-profile was quantified and the donor specimen 337 was found to have high levels of M2-macrophages (Figures 4B). These results indicate that utilizing PMP-derived tissue slices allow for researchers to interrogate the unique donor cellular landscape of the patient tumor samples. Further analysis and interrogation of the molecular, cellular, and genomic landscape is warranted to strategize a bench-to-bedside therapeutic approach that may be beneficial for patient outcomes.
Pharmacotyping and downstream analysis of tissue slices from PMP human tissues are a direct method to measure chemoresistance and drug sensitivity (Figure 5). The results show cell killing of tumor slices as confirmed by cytotoxicity analysis as well as TUNEL confocal imaging in response to the cytotoxic drug bortezomib (2 μM) and 5-fluorouracil (5-FU; 2 μM) (Figure 5A–C). As tumor killing responses are highly heterogeneous, it is important to stratify data on the basis of tumor grade and prior HIPEC, as repeated chemotherapy has been shown to be an ineffective method to ablate tumor cells in vivo and ex vivo17,18. In addition to TUNEL analysis, luminescent viability analysis was performed using sequentially matched slices (Figure 5D). It is essential to use sequentially matched tumor slices when performing cytotoxicity assays, as unmatched slices will lead to unreliable results.
Figure 1: Schematic illustration of pseudomyxoma peritonei ex vivo organotypic slice culture platform with representative tissue slices. A representative workflow for preparing organotypic slice cultures for immunotyping and to investigate functional properties of pseudomyxoma tumors of appendiceal origin. (A) The image of a live mounted tumor nodule from a pseudomyxoma peritonei (PMP) neoplasm. The tumor nodule is embedded in agarose, which is fixed to a vibratome using super glue. (B) The representative distribution of cutting tumor slices and arrangement in a trans-well culture dish. (C) The scheme representing workflow of slice culture and interrogation of tumor slices. Please click here to view a larger version of this figure.
Figure 2: Confocal imaging and functional analysis of living tissue slices. (A) Brightfield image 2x magnification of a representative tissue slice cut from a patient with pseudomyxoma peritonei of appendiceal origin. (B) A 10x magnification of the representative tissue slice shown in A. (C) PAS staining of a tissue from a patient with pseudomyxoma peritonei. Scale bar 1 mm. (D) Immunofluorescence imaging using MUC2 antibody (green), EPCAM (red), and DAPI (blue). Scale bar 50 µm. (E) Immunofluorescence imaging of using live (calcein AM; green) and dead (propidium iodide; red) dyes from donor 39. Scale bar 50 µm. (F) Quantification of viability analysis using three separate slices from donor 39 during initial tissue cutting. Error bars are the standard error of the mean. (G) Immunofluorescence imaging of proliferation using EdU (green). Cell nuclei are labeled with DAPI (blue). (H) Quantification of the number of EdU positive cells per 1,000 cells. N = 3 representative slices from donor 39. Error bars are the standard error of the mean. Please click here to view a larger version of this figure.
Figure 3: Live cell tracking using calcium imaging of living tissue slices. (A) In situ cytolabeling of a living tissue slice from donor 292. CD11b labeled macrophages shown in red. Calcium imaging dye Fluo-4-AM (green). Macrophage co-labeled with CD11b and Fluo-4-AM highlighted in white. (B) Pseudocolor scale of a living tissue slice (shown in A). Macrophage highlighted in white as in A. Scale bar showing 100 µm applies to A. Pseudocolor scale bar shown as mean gray values. (C–E) High magnification image of Fluo-4-AM before (C), during (D), and after (E) calcium activation. (F) Quantification of calcium levels shown by raw traces. The CD11b+ macrophage highlighted in A-E is highlighted in red. A non-responding cell is highlighted in blue. The other responding cell traces are shown in black. (G) Quantification of the normalized area under the curve shown between CD11b+ macrophage and a non-responding cell. Please click here to view a larger version of this figure.
Figure 4: Immunotyping of living human tumor slices from pseudomyxoma peritonei. (A) Representative scheme for performing flow cytometry characterization of immune cell profiles from pseudomyxoma tumors of appendiceal origin. (B) The percentage of immune cell subsets are color coded as well as the donor average bar plotted. The color markers above are provided for immune cell subsets within pie charts. Please click here to view a larger version of this figure.
Figure 5: Representative analysis for pharmacological intervention using organotypic slice culture. (A) Immunofluorescence of representative images from TUNEL analysis on untreated slices (left column), 5-FU (2 μM; middle column), or Bortezomib (2 μM; right column) treated slices. (B) Quantification of TUNEL immunofluorescence as a percentage of TUNEL positive DAPI+ cells from donor 339. (C) Quantification of TUNEL immunofluorescence as a percentage of TUNEL positive PanCK+ cells from donor 339. (D) Quantification of drug-treated slices using luminescent viability analysis from donor 339 pooled from 2-3 sequentially matched biological replicates. CTL denotes control. Concentration for control, 5-FU (2 μM), and Bortezomib (2 μM). Please click here to view a larger version of this figure.
Video 1: Cutting living human tissue slices from metastatic mucinous tumor nodules. Please click here to download this Video.
Video 2: Calcium imaging of living organotypic tumor slices from pseudomyxoma peritonei. Pseudocolor images of maximum projection from confocal images of a living tissue slice loaded with Fluo-4-AM at time point 00:00. Scale bar is 50 μm (bottom right). Please click here to download this Video.
This manuscript describes a technique that can be used to culture, interrogate, and analyze human pseudomyxoma peritonei (PMP) tumor specimens. We have utilized numerous downstream functional assays to interrogate the tumor immune microenvironment and a platform for bench-to-bedside testing.
While the method is highly efficient in our hands, it will require some practice to cut tumor specimens using a vibratome. Namely, we encountered problems that were due to highly mucinous samples, as well as samples that were improperly microdissected to remove uncuttable scaffolds (abdominal wall, ascites, various ECM proteins). As such, microdissection of the PMP tissue to remove excess mucin is a critical step of this protocol.We report an 80% success rate in patients with solid nodules as previously reported. High grade samples were technically less challenging to cut18. Moreover, the preparation of agarose (without FBS) and solidification of the tissue in agarose is also important for this protocol in order to produce tissue slices. Additionally, it will be essential for the researcher to be aware of information regarding tissue conditions and surgical parameters of the specimen (location of the metastasis, pre-treated samples), as these will help novel insights, as well as help to solve potential troubleshooting issues. Another critical step in this protocol is in the tumor cell functional analysis. Here, it is essential to use an epithelial marker (such as cytokeratin staining) in combination with a functional readout (apoptosis, proliferation) to assess efficacy. This is necessary in this particular type of tumor, because many PMP tumors have less than 10% of the cellular composition as tumor cells18. As such, flow cytometry, or confocal imaging is the preferred method of analysis. However, high throughput assays using luminescence-based imaging may be appropriate in tumors with higher tumor cell content, or if assessing the role of therapeutics on the TME.
Despite being a powerful tool, the organotypic slice culture has certain limitations. For example, a challenge of this technique is the cellular heterogeneity within tumors. To account for tumor heterogeneity, sequentially cut slices from similar tumor regions should be matched between groups for treatment and cell viability analysis. Cutting and preparing slices requires practice to microdissect uncuttable tumor scaffolds. These skills are required for successful cutting and optimization of tissue slices to limit experimental biases. Another limitation is tissue viability during slice culture. While it is possible to maintain brain or cardiac slices for months in cultures19,20, slice cultures from PMP have currently only been tested for roughly 1 week in culture18, which prevents long-term experimental investigation, including chronic drug treatment and genetic manipulation. Improvements in culture methods such as adding supplements and growth factors may be necessary to prolong the lifespan of PMP slice cultures.
Organotypic slice culture is a widely used technique to study various organs, including the brain, kidney, and liver. We have adopted this technique for use during PMP, providing us with an alternative model system that, to the best of our knowledge, is the first model to examine the PMP tumor microenvironment in a semi-intact preparation. In conjunction with 3D patient-derived organoids and explant cultures, this technique offers a robust and flexible platform to rapidly assess treatment efficacy and allow immediate translation of novel therapies from bench to bedside.
The authors have nothing to disclose.
The authors would like to thank Kersi Pestonjamasp from the Moores Cancer Center imaging core facility for help with the microscopes UCSD Specialized Cancer Support Center P30 grant 2P30CA023100. This work was additionally supported by a JoVE publication grant (JRW), as well as generous gifts from the estate of Elisabeth and Ad Creemers, the Euske Family Foundation, the Gastrointestinal Cancer Research Fund, and the Peritoneal Metastasis Research Fund (AML).
1 M CaCl2 solution | Sigma | 21115 | |
1 M HEPES solution | Sigma | H0887 | |
1 M MgCl2 solution | Sigma | M1028 | |
100 micron filter | ThermoFisher | 22-363-549 | |
22 x 40 glass coverslips | Daiggerbrand | G15972H | |
3 M KCl solution | Sigma | 60135 | |
5 M NaCl solution | Sigma | S5150 | |
ATPγS | Tocris | 4080 | |
Bovine Serum Albumin | Sigma | A2153 | |
Calcein-AM | Invitrogen | L3224 | |
CD11b | Biolegend | 101228 | |
CD206 | Biolegend | 321140 | |
CD3 | Biolegend | 555333 | |
CD4 | Biolegend | 357410 | |
CD45 | Biolegend | 304006 | |
CD8 | Biolegend | 344721 | |
CellTiter-Glo | Promega | G9681 | |
DMEM | Thermo Fisher | 11965084 | |
DPBS | Sigma Aldrich | D8537 | |
FBS, heat inactivated | ThermoFisher | 16140071 | |
Fc-block | BD Biosciences | 564220 | |
Fluo-4 | Thermo Fisher | F14201 | |
Gentle Collagenase/Hyaluronidase | Stem Cell | 7912 | |
Imaging Chamber | Warner Instruments | RC-26 | |
Imaging Chamber Platform | Warner Instruments | PH-1 | |
LD-Blue | Biolegend | L23105 | |
L-Glutamine 200 mM | ThermoFisher | 25030081 | |
LIVE/DEAD imaging dyes | Thermofisher | R37601 | |
Nikon Ti microscope | Nikon | Includes: A1R hybrid confocal scanner including a high-resolution (4096×4096) scanner, LU4 four-laser AOTF unit with 405, 488, 561, and 647 lasers, Plan Apo 10 (NA 0.8), 20X (NA 0.9) dry objectives. | |
Peristaltic pump | Isamtec | ISM832C | |
Propidium Iodide | Invitrogen | L3224 | |
Vacuum silicone grease | Sigma | Z273554-1EA |