Elsevier

Biomaterials

Volume 34, Issue 20, July 2013, Pages 4777-4785
Biomaterials

A bioengineered microenvironment to quantitatively measure the tumorigenic properties of cancer-associated fibroblasts in human prostate cancer

https://doi.org/10.1016/j.biomaterials.2013.03.005Get rights and content

Abstract

Stromal–epithelial cell interactions play an important role in cancer and the tumor stroma is regarded as a therapeutic target. In vivo xenografting is commonly used to study cellular interactions not mimicked or quantified in conventional 2D culture systems. To interrogate the effects of tumor stroma (cancer-associated fibroblasts or CAFs) on epithelia, we created a bioengineered microenvironment using human prostatic tissues. Patient-matched CAFs and non-malignant prostatic fibroblasts (NPFs) from men with moderate (Gleason 7) and aggressive (Gleason 8–9 or castrate-resistant) prostate cancer were cultured with non-tumorigenic BPH-1 epithelial cells. Changes in the morphology, motility and phenotype of BPH-1 cells in response to CAFs and NPFs were analyzed using immunofluorescence and quantitative cell morphometric analyses. The matrix protein gene expression of CAFs, with proven tumorigenicity in vivo, had a significantly different gene expression profile of matrix proteins compared to patient matched NPFs. In co-culture with CAFs (but not NPFs), BPH-1 cells had a more invasive, elongated phenotype with increased motility and a more directed pattern of cell migration. CAFs from more aggressive tumors (Gleason 8–9 or CRPC) were not quantitatively different to moderate grade CAFs. Overall, our bioengineered microenvironment provides a novel 3D in vitro platform to systematically investigate the effects of tumor stroma on prostate cancer progression.

Introduction

The cell microenvironment through multiple controlled signals directs fundamental cellular processes such as migration, proliferation, survival, and differentiation. Aberrant cues can result in diseases, such as cancer [1], [2]. Specifically, in prostate cancer, the stromal microenvironment is critical in determining epithelial cell differentiation and function [3]. In tumor stroma, carcinoma-associated fibroblasts (CAFs) have a distinct phenotype relative to normal prostate fibroblasts (NPFs) and confer tumorgenicity on non-tumorigenic prostatic epithelial (BPH-1) cells when xenografted in vivo [4], [5], [6], [7]. Based on these observations, CAFs are a therapeutic target.

At present, the only proven method of investigating CAF tumorigenicity is tissue recombination, where stroma and epithelia are co-grafted into immune-deficient mice and the outcome is whether or not a tumor is formed. We and others have widely used this bioassay, yet the techniques are lengthy and technically challenging; only recently we developed a method for unbiased semi-quantitative analyses [4]. To advance the field, there is a need to develop improved in vitro models to mimic the cancer cell interactions with the microenvironment and allow quantitative analyses of the outcome, in terms of their effect on the epithelium. In turn, this will enable the comparison of relative differences between different patient CAFs. Ultimately, this data is fundamental to identify mechanisms that underpin therapeutic targeting of tumor stroma.

In the past, in vitro models of cancer cells have largely depended on the use of two-dimensional (2D) tissue culture on plastic or glass surfaces. Although much has been learned from these studies about how paracrine mechanisms instruct or direct cell behavior, these approaches are limited by the fact that the cells on surfaces often experience artificial polar environments, atypical elastic properties of their environment, and non-physiological cell contacts and densities. Moreover, the flat surface of the tissue culture plate represents a poor topological approximation of the more complex three-dimensional (3D) architecture of the extracellular matrix (ECM) [8]. Therefore, more physiologically mimetic model systems to study both normal and abnormal functions of cells and tissues are desirable [9]. There is strong evidence exists that the highly porous nano- and microtopography that results from the 3D fibrillar associations of ECM proteins is essential for cell adherence, cytoskeletal organization, migration, signal transduction, morphogenesis and differentiation in cell culture [10], [11].

The hypothesis of this study was that production of ECM components in a cellularized co-culture method would allow in vitro validation of prostatic CAFs. Herein, a bioengineered approach was developed to compare the effects of human primary stromal fibroblasts on epithelial cell morphology and motility and test if there is a change with prostate cancer progression.

Section snippets

Isolation of primary prostatic stromal cells (CAFs and NPFs)

Human prostate specimens were obtained by the Australian Prostate Cancer BioResource with approval from the Cabrini Institute (03-14-04-08), Epworth Hospital (34306) and Monash University Human Research Ethics Committees (2004/145). Tissue was collected from radical prostatectomy specimens for CAFs and NPFs as previously described [4]. Transurethral resection of the prostate (TURP) tissue was used for benign prostate fibroblasts (BAFs) and CAFs from castrate resistant prostate cancer (CRPC).

Morphological and molecular characterization of prostatic stroma

For this study, primary fibroblasts were isolated from human radical prostatectomy specimens. Tissue pieces were selected from confirmed malignant areas and adjacent non-malignant areas, as assessed by a pathologist. Carcinoma-associated fibroblasts (CAFs) and normal prostatic fibroblasts (NPFs) were isolated from digested tissue and cultured in medium that selects for fibroblasts instead of prostate epithelial cells. As an additional control, benign-associated fibroblasts (BAFs) were isolated

Discussion

Xenograft models where epithelia are recombined with stroma to study tumorigenicity are based on complex cellular interactions that cannot be replicated in standard 2D cell cultures because cues from the cell matrix play a critical role in directing and maintaining cell fate in vivo. Misregulation within the extracellular space can cause cell death or other aberrant behaviors associated with developmental defects and diseases such as fibrosis and cancer. Thus, as one thinks about cell culture

Conclusions

This co-culture technique is a reliable in vitro model to quantify stromal–epithelial interactions in the tumor. The cellularized matrix, which includes production of ECM, more accurately mimics the microenvironment found in vivo. Using this method, quantitative comparisons were made between fibroblasts from individual men at different stages of prostate cancer progression. Our data demonstrate the ability of CAFs to induce phenotypic changes in BPH-1 cells, regardless of the tumor stage.

Acknowledgments

This research was funded by the Prostate Cancer Foundation of Australia (Project Grant NCG4712, Movember Young Investigator Grants to R.A.T., S.J.E. and M.G.L.), Australian Prostate Cancer Research Centre-Queensland, Peter and Lyndy White Foundation, Australian Research Council (Discovery Project – DP110103890), Cancer Australia (1044458), National Health and Medical Research Council (Fellowships to G.P.R. and M.G.L.; 1035721) and Victorian Prostate Cancer Research Consortium (Scholarship to

References (25)

  • M. Ao et al.

    Cross-talk between paracrine-acting cytokine and chemokine pathways promotes malignancy in benign human prostatic epithelium

    Cancer Res

    (2007)
  • M.S. Joesting et al.

    Identification of SFRP1 as a candidate mediator of stromal-to-epithelial signaling in prostate cancer

    Cancer Res

    (2005)
  • Cited by (53)

    • The functional cross talk between cancer cells and cancer associated fibroblasts from a cancer mechanics perspective

      2021, Biochimica et Biophysica Acta - Molecular Cell Research
      Citation Excerpt :

      CAFs provide nutrients for cancer cells by secretion of exosomes [152], and coordinate some cancer cell behaviors such as proliferation, angiogenesis and invasion to other tissues [153–155]. In an in vitro study using 3D matrix, the key role of CAFs in increased motility and the subsequent invasive behavior of prostate cancer cells were confirmed [156]. It has been observed that in some occasions during invasion to other tissues, CAFs accompany cancer cells [157].

    • Cancer-associated fibroblasts of the prostate promote a compliant and more invasive phenotype in benign prostate epithelial cells

      2020, Materials Today Bio
      Citation Excerpt :

      In this study, we compared the mechanical properties of primary CAFs and non-malignant prostate tissue fibroblasts (NPFs) isolated from malignant or benign regions of prostate tissue from the same patient [36]. Previous works have identified distinct proteomic and epigenetic profiles in patient-matched pairs of stromal cells, altered interactions with epithelial and immune cells, as well as aberrant ECM deposition, compared with normal fibroblasts from the prostate [17,37–41]. Here, we compare the mechanical characteristics of patient-matched CAFs and NPFs as well as their co-cultures with benign prostatic epithelial cells using AFM and RT-FDC.

    • In vitro three-dimensional modeling for prostate cancer

      2020, Biomaterials for 3D Tumor Modeling
    • Tissue engineered human prostate microtissues reveal key role of mast cell-derived tryptase in potentiating cancer-associated fibroblast (CAF)-induced morphometric transition in vitro

      2019, Biomaterials
      Citation Excerpt :

      Prostatic cancer-associated fibroblasts (CAFs) and their non-malignant counterparts (non-malignant prostatic fibroblasts; NPFs) exhibit biochemical, proteomic and epigenetic differences [3–6]. Functionally, these differences enable CAFs to initiate and potentiate tumourigenicity in adjacent epithelia [3,4,7,8]. In particular, CAFs alter the physical environment of the epithelia by depositing an aberrant extracellular matrix (ECM) [9].

    View all citing articles on Scopus
    1

    Authors contributed equally.

    View full text