Biomedical Engineering Reference
In-Depth Information
The prostate gland forms during gestation as a derivative of the urogenital
sinus, in which the epithelial buds protrude and elongate into the surrounding
mesenchyme (Risbridger et al. 2005; Thomson andMarker 2006). The buds can
bifurcate to form branches with terminal tips, and epithelial cell proliferation
primarily occurs at the leading edge. This branching morphogenesis gives rise to
the network of secretory ducts, and, in humans, prostate morphogenesis is
complete by birth. In rats, the buds are solid epithelial cords that are primarily
undifferentiated cells at day 1 that are positive for basal cell markers cytoker-
atins (K) K5 and K15, and negative for luminal markers K8 and K18 and are
low in AR expression (Prins and Birch 1995). As development proceeds basal
cells differentiate into luminal cells positive for K8 and AR. The development of
the prostate is androgen dependent, with involution occurring upon the with-
drawal of androgen and regeneration with restoration of androgen, suggesting
that there are androgen-independent prostatic stem cells that survive androgen
withdrawal and can regenerate the prostate.
1.3 The Cancer Stem Cell
A malignant tumor is composed of a heterogeneous population of cells with
varying degrees of tumorigenic potential, and only a subset of cancer cells can
initiate and propagate a tumor. This was first demonstrated in 1961 when
researchers' harvested recurrent cancer cells from patients and then autotrans-
planted the cells. Tumors only formed when injected with up to 1 million cells
(Southam and Brunschwig 1961). Other studies showed that only a small
fraction of cancer cells could form colonies in vitro (Hamburger and Salmon
1977; Bruce and Van Der Gaag 1963). These observations led investigators to
hypothesize that the clonogenic cells arose from cancer stem cells (CSCs) that
maintain the rest of the population. However, the stochastic model of tumor-
igenesis was equally plausible.
The stochastic model states that all cancer cells can proliferate extensively,
form colonies in vitro, and initiate new tumors, but only a small fraction has the
probability of finding a permissive environment, or niche, for tumor growth
(Huntly and Gilliland 2005; Perryman and Sylvester 2006). In contrast, the
cancer stem cell model posits that most cancer cells are unable to proliferate
extensively, cannot form colonies in vitro, and are unable to initiate new tumors
(reviewed in Wicha et al. 2006). Only a rare, biologically unique subpopulation
can exhibit these behaviors. The heterogeneous tumor, like normal tissue, is
governed by a cellular hierarchy and at the top is the stem cell (Huntly and
Gilliland 2005). To prove the latter hypothesis, one would have to isolate these
CSCs and show that they exhibit clonogenic capacity whereas their non-cancer
stem counterparts do not. Indeed, Lapidot et al. showed that CD34
+
CD38
cells could be identified as AML stem cells (Lapidot et al. 1994; Bonnet and
Dick 1997). NOD/SCID mice injected with CD34
+
CD38
leukemic cells
