mikos: “9026_c001” — 2007/4/9 — 15:50 — page3—#3
Fundamentals of Stem Cell Tissue Engineering 1-3
differentiate into neural cells [32,33], cardiac myocytes [34,35], vascular support cells (pericytes, smooth
muscle cells) [36,37], and perhaps other tissues [38,39]. The multipotential of culture expanded MSCs
provides the stimulus to consider them as candidates for various Tissue Engineering strategies. Such
strategies, by necessity, require both scaffolds and various growth factors/cytokines to manage the pro-
liferation and differentiation of MSCs to form specific tissues in vivo. In some cases, the tissue defect,
itself, and its microenvironment provide instructional support; in other cases, pretreatment of the cells or
placing the cells within unique scaffolds provides the instructional cues [40]. The details of these experi-
ments provide the experimental basis for improving these early Tissue Engineering logics in preparation
for their clinical use.
1.3 Fundamental Principles
When in doubt of how to manage the multiple parameters of tissue repair/regeneration, Mother Nature
should be asked to reveal her secrets as a guide. Philosophically, we believe that many of the funda-
mental principles of Tissue Engineering involve the recapitulation of specific aspects of embryonic tissue
formation [41,42]. For example, the embryonic mesenchyme that will form the cartilage anlagen of long
bones has a high ratio of undifferentiated progenitor cells to extracellular matrix (ECM). This embryonic
mesenchymal ECM is composed of type I collagen, hyaluronan, fibronectin, and water. In experiments
with chick limb bud mesenchymal progenitor cells in culture, we showed that the molecular weight of
hyaluronan is instructive to these progenitor cells, in that high molecular weight chondro-inductive or
chondro-permissive [43,44]. Others have shown that high molecular weight hyaluronan is antiangiogenic.
Indeed, the exclusion of blood vessels is also chondrogenic. Thus, a scaffold of hyaluronan coated with type
I collagen or fibronectin to bind the MSCs would be expected to be chondrogenic; indeed, we have shown
it to be just that [45]. The scaffold must be quite porous to allow the newly differentiated chondrocytes to
fabricate their unique and voluminous ECM that controls the cushioning properties of the cartilaginous
tissue. By mimicking the cell density and ECM of embryonic progenitor mesenchyme, cartilaginous tissue
forms in large, full thickness defects in adult rabbit knees [41].
In the absence of hyaluronan, the key physical characteristic of prechondrocytes is their close proximity
to their neighbors and maintenance of the cells in a rounded shape [20,21]. One way to achieve this
condition is to place adult MSCs in a type I collagen lattice. The cells will bind to the lattice fibrils and
rapidly contract the lattice to bring the cells into a high density configuration [46]. In culture, in the
presence of TGF-β [20,21] or in vivo by the contracted network excluding blood vessels, the MSCs will
form cartilage tissue. Thus, mimicking the embryonic microenvironment both chemically and physically
can result in the specific differentiation of MSCs.
In contrast, the rules for bone formation are quite different from those governing cartilage formation
[13,47,48]. In this case, we again studied embryonic chick limb development and observed that vascu-
lature is the driver for bone formation. In the context of Tissue Engineering scaffolds, bone formation
requires rapid invasion of blood vessels into the pores of the scaffold. For example, porous calcium
phosphate ceramics coated with fibronectin to bind MSCs provide an inductive microenvironment for
bone formation in subcutaneous or orthotopic sites [52]; it may be that the calcium phosphate, itself,
is informational, but more likely it binds osteogenic growth factors that stimulate the MSCs. The MSCs
bind to the walls of the pores in the ceramic where they divide and, as vasculature invades from the host
tissue at the implantation site, the cells differentiate into sheets of osteoblasts and fabricate the lamellae
of bone [49]. The vasculature does not go to the walls of dead-ends of the ceramic and in this location,
the MSCs divide and pile-up on one another and form compact areas of cartilage. Thus, in the two dif-
ferent microenvironments (vascular and avascular), the MSCs form two very different tissues (bone and
cartilage). This bone/cartilage forming capacity has been quantified and has become our gold standard
for judging the quality of MSC preparations [19,50].
Again, for emphasis, mimicking Mother Nature, especially her very efficient embryological events, is,
for us, a fundamental rule of engineering tissue repair or regeneration in adults.