Engineered Tissues as Customized Organ Germs

“How closely should the biological construct or substitute mimic the function of the natural tissue in order to be considered efficacious?”–With this question, Bob Nerem and Athanassios Sambanis introduced in the very first article of the newborn journal Tissue Engineering ¹ what turned out to be a central and still unresolved issue. In the subsequent years, the classical paradigm of the field, naturally attracted by the vision of “body parts on demand,” tended to be dominated by the attempt to generate preformed constructs replicating as faithfully as possible the structural architecture and biological/mechanical functionality of the target tissue. This approach has led to a large segment of scientific literature, which includes the contributions of the author, focused on the identification of conditions to reach advanced tissue maturation ex vivo, even in the awareness that the need for such functionality is largely not substantiated by experimental data.

Along with this main stream, it has been increasingly recognized that an ideal graft should not necessarily replicate the properties of the damaged tissue, provided that it contains the necessary and sufficient instructive elements for its regeneration. The relevant signals, possibly of structural, biochemical, or biophysical nature, can be substantially different from those controlling tissue homeostasis, and rather aim at activating pathways of organogenesis. The concept has been implemented, for example, in the context of bone formation using mesenchymal stromal cells. As opposed to directly inducing an osteoblastic phenotype, cells can be differentiated to generate hypertrophic cartilage tissue, which represents the primordial structure for the majority of the ossification processes. Upon implantation, such a template is in fact efficiently remodeled following the molecular/cellular events typical of bone development and repair, ultimately replicating the architecture and full functionality of a native bone organ.^2,3 The approach has been proposed also for tooth or salivary gland regeneration, whereby epithelial and mesenchymal cells from the corresponding embryonic structures, appropriately primed in vitro and subsequently implanted, developed into new functional organs in adult mice.^4,5 In these cases, the engineered grafts were attractively described as “organ germs.” Other remarkable examples of self-organogenesis, including those based on embryonic stem cell systems, are the subject of a recent outstanding review. ⁶

The paradigm of an organ germ clearly evokes the vision to recapitulate processes occurring during development, and therefore merges with the modern concept of “developmental engineering.” ⁷ According to this strategy, the engineering of a graft should rely on features typical of embryogenesis, including the temporal progression driven by sequential molecular switches (stage and path dependence), the relative uncoupling of interfaces between different developing entities (modularity), and the stability and reproducibility of events (robustness). The principle implies the in vitro priming of a process up to a defined stage, from which the organ germ will be capable to autonomously progress to engraftment, remodeling, and tissue regeneration, thanks to the in vivo bioreactor represented by the body itself. The vision is obviously exposed to a large degree of complexity, not only due to the intrinsic difficulty to identify and consistently activate upstream master molecular pathways, but also since the graft should interact with an environment that is not any longer embryonic. This implies substantial differences, for example, in the potency of resident stem/progenitor cells, in the structure and physical properties of the surrounding tissues, and in the strong regulatory role of postnatal inflammatory and immunological processes. Therefore, more recently, the term “developmental re-engineering” was proposed. ⁸ Despite the intrinsic challenges and complexity, the target of engineering a process rather than merely engineering a tissue represents an important milestone in the debate introduced by Nerem and Sambanis.

In the same first Volume of Tissue Engineering, Eugene Bell highlighted the strong instructive role of the extracellular matrix (ECM), “enabling cells to carry out their programs of cell division, morphogenesis, differentiation, and tissue building,” thanks to “its high molecular diversity and the supramolecular structure into which it is organized”. ⁹ The article provocatively compared the “low information content of currently known man-made biomaterials…with extracellular matrices that have evolved over the course of hundreds of millions of years.” Whereas it is clear that concepts and paradigms for the design of materials have seen a staggering progress in the past 20 years, including the possibility to mimic and engineer features typical of ECMs, ¹⁰ we must admit that some of the best established products in regenerative medicine still rely on the performance of a native, decellularized ECM. ¹¹ As compared with living cell-based grafts, the decellularization of an ECM offers the additional opportunity of off-the-shelf storage, with clear logistic and commercial advantages.

If we make an effort to combine the use of ECM-derived materials with the principles of developmental engineering, we may come to the conceptual realization that an ECM ideally suited to activate neo-organogenesis may need to be different from that derived from adult and fully developed tissues. In fact, the ECM should carry the information to generate a tissue as opposed to maintaining its homeostatic properties. Moreover, the delivered instructions should most likely be tissue specific as opposed to generic, all-purpose regenerative cues. One could thus conceive a tissue engineering approach to generate a specialized ECM, whereby ex vivo generated tissues would be blueprinted to reach a specific composition/structure/stiffness and subsequently decellularized. ¹² The resulting biomaterial, generated by the activity of living cells as opposed to synthetic chemistry, would not only represent a physical scaffold for cell motility and mechanosensing, but also a drug delivery vehicle capable to establish morphogen gradients. ¹³ Such ECMs, designed to initiate well-defined regenerative processes, could be laid down by suitable cell lines, engineered to express the relevant backbone proteins and morphogenic factors and additionally allowing to achieve standardization and repeatability of manufacturing. ¹⁴

Taking the liberty to propose a visionary perspective that could apply 20 years from today without bothering to discuss the associated challenges and bottlenecks, the approach could be stretched to a further degree of customization. Considering that paradigms for regenerative medicine will likely overlap with the current trends of the pharmaceutical industry, the engineered tissue or ECM could be personalized not only to target specific pathologies, but also based on the age, stage of disease progression, lifestyle, and even the genetic background of the individual. The extreme product diversification could still be combined with the definition of standard categories, and thus remain compatible with industrialized manufacturing systems.

Going back to the initially cited question, the author's personal view is that the engineered grafts used to treat our grandchildren would not have to mimic the properties of the target native tissue, but rather activate its developmental/regenerative programs in a way that matches individualized characteristics and requirements. Time will reveal if off-the-shelf, engineered organ germs will ever be available with a similar modality as personalized pills.