|Home | About | Journals | Submit | Contact Us | Français|
The successful isolation and cultivation of prostate stem cells will allow us to study their unique biological properties and their application in therapeutic approaches. Here we provide step-by-step procedures on the basis of previous work in our laboratory for: the harvesting of primary prostate cells from adolescent male mice by a modified enzymatic procedure; the isolation of an enriched population of prostate stem cells through cell sorting; the cultivation of prostate stem cells in vitro; and characterization of these cells and their stem-like activity, including in vivo tubule regeneration. Normally it will take approximately 8 hours to harvest prostate cells, isolate the stem cell enriched population, and set up the in vitro sphere assay. It will take up to 8 weeks to analyze the unique properties of the stem cells, including their regenerative capacity in vivo.
Epithelial stem cells are of interest due to their capacity for organ replenishment and for their potential role in cancer-initiation. During the normal life span of an adult organ, stem cells operate to replace lost or damaged tissue to ensure proper organ function1. More recently, epithelial stem cells have also been demonstrated as a target population for cancer2. Due to their long-lived nature and inherent self-renewal capacity, adult stem cells are a likely cell-of-origin for many cancers3. The isolation of these cells and investigation into their properties will be useful for understanding their basic biological processes and for designing new therapies.
The prostate represents an ideal model system to investigate the properties of adult stem cells due to the seemingly unlimited ability of the rodent prostate gland to undergo cycles of involution after androgen-depletion and subsequent organ regeneration in response to androgen add-back4. Perhaps the most important reason to study prostate stem cells (PrSCs) is that they share the property of androgen-independence (or castration-resistance) with the subset of prostate cancer cells that survive in an androgen-deprived environment and can drive the lethal phase of the disease, termed hormone-refractory or castration-resistant prostate cancer (CRPC)5,6. Identifying critical self-renewal pathways in PrSCs may provide new therapeutic targets for the treatment of CRPC.
Several methods have been described for the isolation of primitive cells from the mouse prostate7,8,9,10,11,12,13. While genetically engineered mouse models can be useful for stem cell isolation, they limit the widespread use of such an approach. We have previously reported the isolation of PrSCs from wild-type mice capable of self-renewal and tri-lineage differentiation in vivo9. PrSCs can be reproducibly isolated by fluorescence activated cell sorting (FACS) using the antigenic profile Lin- Sca-1+ CD49f+ (LSC)10 or Lin- Sca-1+ CD49fhi Trop2hi (LSCT)9. These cells possess a basal phenotype and primarily reside in the region of the gland proximal to the urethra. Primitive cells with basal characteristics and an antigenic profile similar to PrSCs (LSC/LSCT) can be reproducibly isolated from un-fractionated prostate when primary cells are cultured in the prostate sphere assay10,14. Here we will describe our methods to isolate PrSCs from primary prostate tissue, culture PrSCs in vitro, and measure PrSC activity using quantitative in vitro and in vivo assays.
Epithelial cells quickly lose their self-renewal potential when they are cultured in two-dimensions15. We have developed a three-dimensional culture system to maintain and expand primitive prostate cells that retain the capacity for both self-renewal and differentiation14. Cells are suspended in a semi-solid matrix comprised of prostate epithelial growth medium (PrEGM) and Matrigel. Matrigel is comprised primarily of laminin, collagen, fibronectin and other extracellular matrix (ECM) components. This in vitro environment closely mimics the ECM-rich basement membrane where basal cells reside in the gland. More importantly, basal cells express high levels of ECM-binding integrins that promote proper cell signaling and likely keep them in an undifferentiated state16. The culture system is supplemented with selected growth factors and cytokines including EGF and FGF in the PrEGM media.
Three assays (the colony-forming assay, the sphere-forming assay and the in vivo prostate-regeneration assay) have been developed to measure primitive prostate cell activity10. Similar to other epithelial stem cell assays, the colony-forming assay is performed in a two-dimensional culture system and measures both proliferative colony-forming activity and differentiation. Colonies are clonal in origin, express basal and luminal keratins, and can be induced to undergo differentiation upon addition of androgen10,14,15. The sphere-forming assay is performed under three-dimensional conditions, as described above. Spheres are also clonal in origin, are comprised of several hundreds of cells, and can be dissociated and re-plated to measure self-renewal activity10,14,15. Sphere cells undergo spontaneous differentiation with the most primitive cells residing around the outside and the more mature cells oriented towards the lumen, similar to the architecture of the native gland14. Finally, the in vivo prostate-regeneration assay17,18 measures the ability of PrSCs to form prostatic tubules when combined with inductive stroma and implanted under the kidney capsule or skin (subcutaneous) of immunodeficient mice. Regenerated tubules are indistinguishable from primary prostate tubules, with an outer basal layer, an inner luminal layer and rare neuroendocrine cells9. The subrenal regeneration assay can be used to study self-renewal in vivo by implanting prostate cells from transgenic mice harboring a probasin promoter driven luciferase, and performing androgen cycling15. While the in vitro colony-forming and sphere-forming assays take 7–10 days, the in vivo prostate-regeneration assay takes considerably longer (6–10 weeks). We have functionally defined PrSCs based on their ability to generate colonies, spheres and prostatic tubules in these assays.
Steps 1–6, isolation of prostate cells: ~3–4 hours
Step 7A or 7B, enrichment of PrSCs: 2–3 hours
Steps 8–13, culturing PrSCs: 1 hour set-up, followed by a 7–10day incubation
Step 14, counting prostate spheres: ~ 30 minutes – 1 hour
Steps 15–20, dissociating prostate spheres: 2 hours
Step 20A, Immunostaining: up to 1 week
Step 20B, Sphere passaging: 2 hours, followed by a 7–10day incubation
Step20C, Colony assay: 15 minute set-up, followed by a 10 day incubation
Step20D, Subcutaneous and subrenal regeneration assay: 2 hour set-up followed by 8-week incubation
Complete characterization of PrSCs: 8 weeks for performing all of the experiments
In our hands we can get ~1.0 – 1.5×106 prostate cells from an 8 to 10 week old C57/Bl6 mouse. We get around 20–23% LSC staining and 8–10% LSCT staining (when gating on the Lin- fraction). From one 8- to 10-week old prostate we can obtain approximately 8–10×104 LSC cells by FACS count, which correlates to approximately 3–5×104 cells by hand count. We get 2–4×104 LSCT cells by FACS count, which correlates to approximately 1–2×104 cells by hand count.
One in 35 LSC cells and 1 in 11 LSCT cells can form spheres in the first generation, as reported in our most recent publications9,10. It is important to remember that basal cells express higher levels of matrix binding integrins, which may give these cells a growth advantage in the extracellular matrix-rich Matrigel. The additional growth factors in the Matrigel and PrEGM may also drive more progenitor-like cells to give rise to spheres in addition to the stem cells.
The majority of primary prostate spheres are solid in appearance (Figure 5a), and a subset have secretions in the center (Figure 5b). The spheres contain mostly basal cells that express high levels of cytokeratin 5 and low levels of cytokeratin 8 (Figure 6a and b). The cells located on the outer rim are the most primitive basal cells that express p63 (Figure 6c). Full length AR protein cannot be detected in the majority of the spheres (Figure 6d). Western blot analysis has revealed that the sphere cells do express AR protein, but in the absence of testosterone, the protein is quickly degraded. Only small bands representing the degraded protein are seen on Western blots, and not the full length receptor14.
In vitro passaging: One in 20 dissociated first generation sphere cells will form secondary spheres. In later generation, the sphere forming activity becomes closer to 1 in 10.
In vitro colony assay and differentiation: Approximately 1 in 25 primary sphere cells will form colonies in the colony assay (Figure 5c). The colonies appear epithelial, with a cobblestone morphology (Figure 5d). In the absence of DHT, 95% of the colonies are CK5/CK8 double positive, which is indicative of an intermediate/transit-amplifying phenotype (Figure 6e–g). None of these colonies express full length AR (Figure 6h). There are occasional CK8 single positive colonies, which are probably derived from a luminal-restricted colony-forming cell. Another small percentage of the colonies is CK5 single positive. With the addition of DHT, the majority of the colonies become CK8 high and CK5 low to negative(Figure 6i–k). Most of the colonies also express AR in the presence of DHT(Figure 6l).
In vivo prostate regeneration: Primary isolated LSC and LSCT cells will grow in vivo in both the subrenal and subcutaneous regeneration assays. Cells tend to have more robust growth under the kidney capsule, most likely due to the rich blood supply and inductive microenvironment existing in that area. Stem cells cultured in vitro in the prostate sphere assay will grow better in the subcutaneous regeneration assay. Subcutaneous grafts will likely have a largely mesenchymal or fibroblastic appearance due to UGSM outgrowth, with scattered tubules (Figure 5e–g). One in a few hundred cells from an enriched preparation of stem cells (LSCT) will be expected to generate prostatic tubules. Regenerated tubules should contain a typical double-layered epithelial appearance (Figure 5h). Immunostaining will show an outer layer of cells expressing the basal keratins (K5, K14) and the transcription factor p63 (Figure 6m and n). The inner epithelial layer is larger and columnar in shape, expressing high levels of the luminal keratins (K8, K18) and AR(Figure 6m and o). Occasional cells within the basal layer or between layers will express the neuroendocrine marker synaptophysin(Figure 6p). In the subrenal regeneration experiment, the tubule forming activity of the LSC and LSCT cells are similar to the subcutaneous assay, but the tubules appear larger (Supplementary Figure 3a and 3b). As in the subcutaneous regeneration assay, cells in the outer layer express the basal markers CK5 and p63 (Supplementary Figures 3c and 3d), while the inner layer cells express the luminal markers CK8 and AR (Supplementary Figures 3eand 3f). The subrenal regeneration assay is also preferred for carcinogenesis studies. Use of SCID mice for the regeneration assays allows testing of the tubule regenerating capabilities of both primary mouse and human prostate cells in a parallel manner. Studies to purify and characterize stem and progenitor cells from the human prostate using the in vivo model are underway in our laboratory.
We thank Houjian Cai and Yang Zong for their helpful comments on the protocols and for demonstrating the procedure in Supplementary Figure 4. R.U.L. is supported by the California Institute for Regenerative Medicine training grant (T1-00005 and TG2-01169). A.S.G. is supported by Ruth L. Kirschstein National Research Service Award GM07185. This work has also been supported by the Prostate Cancer Foundation Challenge Award for Defining Targets and Biomarkers in Prostate Cancer Stem Cells. O.N.W. is an Investigator of the Howard Hughes Medical Institute.