Opinion

The Thermodynamics of Development in

Bioartificial Tissue Design
Petros Lenas1,*

The fabrication of bioartificial tissues with authentic structures that could Highlights
assure their clinical efficacy remains a challenging problem. A new paradigm Bioartificial tissue implants designed
as developmental modules can be
has emerged that designs bioartificial tissues as intermediate in development
clinically efficacious by installing in situ
tissue forms, which can inherently progress autonomously on developmental natural developmental processes of
pathways, self-organizing their cells into tissue structures as in their in vivo tissue formation.

development. Biological processes involved in energy exchange between co- Developmental modules are thermo-
developing tissues are responsible for cell organization into the thermodynam- dynamic units operating as dissipative
ically robust cellular patterns of tissue structures. Bioartificial tissue design structures that stabilize their patterns
(tissue structures) by employing biolo-
rules that aim towards in vitro recapitulation of these processes can ensure the gical processes involved in energy dis-
thermodynamic operation of developing tissues, leading to formation of the sipation (energy import and entropy
export).
cellular patterns of tissue structures.
Thermodynamic analysis of develop-
mental modules casts light on the key
Overview: Restoring the Thermodynamic Operation of In Vitro Developing biological parameters involved in energy
Tissues Can Help to Fabricate Bioartificial Tissues with Authentic Tissue dissipation for pattern formation that
should be controlled in vitro for the for-
Structures mation of authentic tissue structures.
Tissue engineering was introduced three decades ago with the goal of replacing diseased or
damaged tissues with bioartificial tissues fabricated in the laboratory using cells growing inside 3D The fabrication of complex bioartificial
tissues and organs requires the in vitro
biomaterials, named ‘scaffolds’. Despite intense research efforts, clinical translation still faces
fabrication of thermodynamically inter-
difficulties, especially for tissues with complex structures, because their multiscale anatomical acting bioartificial developmental mod-
features cannot be easily reproduced in vitro. The exploitation of in vivo developmental mecha- ules that cooperate in energy
nisms, which invariably lead to authentic tissue structures, has been therefore proposed as a new dissipation.
paradigm for the design of bioartificial tissues [1,2]. Despite the complexity of the in vivo develop-
In vitro studies on systems of intercon-
mental mechanisms, their use in tissue engineering is feasible because they are self-established by nected modules can decipher general
the developing tissues themselves through the operation of evolutionarily designed genetic principles of the spatiotemporal orga-
programs [3]. Therefore, tissue engineering should aim at the fabrication of intermediate in nization of development that optimizes
energy dissipation.
development tissue forms that have an inherent capacity to initiate and orchestrate by themselves
developmental mechanisms that lead to the organization of their cells into tissue structures, Principles of development may allow
instead of aiming at the direct fabrication of bioartificial tissues that resemble the final tissue thermodynamically permissible clinical
structure [3]. Indeed, recent studies in tissue engineering have shown that implants designed as optimization of bioartificial develop-
mental modules by modification of
intermediate in development tissue forms can install in situ after implantation natural develop- genetic programs of biological pro-
mental processes of tissue formation, thus improving their clinical efficacy. cesses involved in energy dissipation.

However, it is unclear which of the biological processes taking place during in vivo tissue 1
College of Science, Harbin Institute
development are crucial for the formation of such autonomously developing tissue forms such of Technology (Shenzhen), Shenzhen
that they can be recapitulated in vitro, permitting the fabrication of bioartificial analogs. The study of 518055, China
the developing tissues as thermodynamic systems can provide a methodological framework to
address this question. Cells and organisms are open thermodynamic systems that maintain their
*Correspondence:
robust organization by importing energy from their environment, for example nutrients, to build their plenas@hit.edu.cn,
structures through which entropy generated by cellular processes (e.g., metabolic byproducts) is lenaspetros@yahoo.com (P. Lenas).

1116 Trends in Biotechnology, November 2018, Vol. 36, No. 11 https://doi.org/10.1016/j.tibtech.2018.06.006

© 2018 Elsevier Ltd. All rights reserved.
exported to the environment. Similarly, developing tissues operate as multicellular open thermo-
dynamic systems that use the imported energy to set in place biological processes that lead to
organization of their cells into robust spatial patterns of tissue structures through which entropy is
exported to their surrounding co-developing tissues and finally, through the circulation, to the
environment. Analysis of the thermodynamic role of the biological processes taking place in
developing tissues can indicate which of these processes are crucial for their continuous thermo-
dynamic operation which is necessary for cell organization into tissue structures. Bioartificial tissue
design rules compatible with the thermodynamic laws can be then identified that can ensure the in
vitro recapitulation and control of the biological processes that sustain the thermodynamic
operation of developing tissues. Furthermore, the thermodynamic interactions among co-devel-
oping tissues may provide general principles for the spatiotemporal organization of development.

Developmental Modules: The Autonomous Building Blocks of Developing

Tissues
Intermediate in development tissue forms, called ‘developmental modules’, are quasi-autono-
mous, multicellular building blocks of developing tissues with an inherent capacity to progress
through developmental pathways [4]. For example, as has been understood for a century, half
of the limb area is capable of developing into a morphologically normal limb when it is excised
and transplanted to ectopic sites in the embryo [5]. The autonomy of developmental modules
relies on interactions among the cells inside the module, which activate the necessary,
evolutionarily designed, signaling pathways [6]. Such a developmental module is formed,
for instance, during the development of long bones, which takes place through a cartilage
template that is gradually replaced by bone in the developmental process of ‘endochondral
ossification’ [7]. The part of a long bone where new bone growth takes place is called ‘growth
plate’, and this is a developmental module consisting of differentiating chondrocytes arranged
in a pattern of parallel columns (Figure 1A). The columnar pattern is robust to environmental
perturbations because its formation is not controlled by external factors but by internal to
growth plate interactions between chondrocytes [8]. These interactions establish a global,
spatially extended signaling loop that controls the rate of chondrocyte differentiation [9]
(Figure 1B). The growth plate induces and controls bone formation at its end via signals
secreted by the chondrocytes that terminate their differentiation and die there by apoptosis
(Figure 1A).

Harnessing the Autonomy of Developmental Modules: In Situ Installation of

Developmental Processes by Implants
Fabricating bioartificial bone by differentiation of stem cells into bone cells (osteoblasts) inside
scaffolds requires signaling molecules to be incorporated into the scaffolds that will orchestrate
the processes of cell differentiation and spatial organization into bone tissue structures which is
a challenging problem because most of these signals are not known. Alternatively, the
developmental biology-inspired paradigm suggests the use of bioartificial cartilage, instead
of bioartificial bone, as an implant, paralleling indirect bone formation through a cartilage
template in endochondral ossification [3]. The stem cells should thus be differentiated in vitro
into cartilage cells (chondrocytes) instead of bone cells. The cartilage template could then
autonomously induce and control bone formation after its implantation, operating as a devel-
opmental module with its chondrocytes providing the necessary signaling molecules for bone
formation. Chondrocytes produce factors that induce the differentiation of stem cells to bone
cells [10] such that, when implanted, they could transform the stem cells of the recipient tissue
found in their vicinity into bone-forming cells. In addition, they produce factors that induce the
growth of new blood vessels inside the developing bone, assuring cell viability [11]. Several

Trends in Biotechnology, November 2018, Vol. 36, No. 11 1117

 Blood supply
Metaphysis
Bone marrow
Bone developmental

Osteoblasts
module

Calcified
carlage
Ihh
Apoptoc

Hypertrophic

-
developmental module

Pre-
Growth plate

hypertrophic

Proliferang

Resng
chondrocytes
+ PTHrP
Oxygen
Epiphysis gradient

Blood supply

(A) (B)

Figure 1. The Growth Plate. (A) Chondrocytes from the proliferating and resting zone enter their differentiation program
and move along the longitudinal axis of the bone aligned in parallel columns. Only the proliferative zone is adequately supplied
with blood, thereby establishing an oxygen gradient that falls towards the hypertrophic zone. Terminating their differentiation,
chondrocytes have consumed their energy, and because of the lack of oxygen they die from apoptosis and are replaced by
osteoblasts that produce bone matrix. Bone grows in length as new chondrocytes from the proliferating and resting zone
enter the columns. (B) Pre-hypertrophic and hypertrophic chondrocytes secrete Indian hedgehog (Ihh) which induces the
expression of parathyroid hormone-related protein (PTHrP) by periarticular (resting) chondrocytes. PTHrP diffuses back and
reaches the proliferating chondrocytes, retarding their entry into the differentiation program. As new chondrocytes start their
differentiation and continuously enter the columns extending them, the limits on the differentiation rate set by this signaling
loop allow other processes, such as matrix secretion and organization, to complete the construction of a dense cartilage
matrix between the chondrocyte columns that ensures the structural integrity of the columnar pattern.

recent experimental studies in tissue engineering have validated these advantages of the
indirect developmental route for bone formation through bioartificial cartilage that operates
as developmental module [12–14]. For instance, bioartificial cartilage templates made in vitro by
differentiating adult human stem cells inside collagen-based scaffolds into chondrocytes was
found to autonomously trigger, in the absence of any instructive external signals, the whole
phenomenon of endochondral ossification after implantation [12]. The bioartificial cartilage
templates were transformed in vivo to ‘bone organ’, with a size, structure, and functionality
comparable to native bones, by inducing all the processes involved in bone formation –
including the ingrowth of blood vessels and the establishment of a functional bone marrow
containing hematopoietic stem cells [12].

The use of developmental modularity in bioartificial tissue design is an important milestone in

tissue engineering because it is directly related to the clinical efficacy of implants since it
encompasses their capacity to induce natural developmental processes for tissue formation in
situ after implantation [15]. The examples mentioned above establish a new paradigm for tissue
engineering with the design of bioartificial tissues as intermediate in development tissue forms
that can be autonomously transformed into final tissue forms either in vivo after implantation or
in vitro with appropriate in vivo-like methods for 3D cell growth and differentiation. The
fabrication of such tissue forms therefore necessitates the development of new methods in
tissue engineering that can use information from developmental biology in terms of bioartificial

1118 Trends in Biotechnology, November 2018, Vol. 36, No. 11

tissue design, as proposed herein through thermodynamic considerations that indicate the
crucial in vivo biological processes which must be recapitulated in vitro for the formation of
authentic tissue structures.

The Thermodynamics of Development: Developmental Modules as

Dissipative Structures
Dissipative structures are robust structures that are formed outside thermodynamic equilibrium
and are encountered at various scales in inanimate and living systems, for example in fluid
convection, chemical autocatalysis, and in the organization of cells, organisms, and ecosys-
tems (Box 1). They are ordered structures that differ from equilibrium structures (e.g., crystals) in
that their patterns are stabilized by continuous degradation of energy taken from their envi-
ronment into exported entropy, a process known as ‘energy dissipation’ (Box 1). All living
systems are dissipative structures because they dissipate (i.e., degrade by their metabolic
activities) the energy of environmental sources into metabolic wastes and heat that export into
the environment (Figure 2A). The analogies of the thermodynamic interpretations of pattern
formation observed between the Rayleigh–Bénard (R-B) convection system, an extensively
studied inanimate dissipative structure of globally organized convective fluid motion [16]
(Figure 2B), and the developmental module of the growth plate (Figures 1A and 2C) (Box 2),
can clarify the thermodynamic role of the biological processes involved in the formation of the
patterns of developmental modules. Both systems use imported energy to organize their com-
ponents into macroscopic spatial patterns and maintain their patterns by continuously exporting
entropy into their surroundings: heat at the low-temperature plate in R-B convection (Figure 2B),
and chondrocytes dying by apoptosis [17] (or ‘chondroptosis’ [18] – cell self-disintegration – which
is more reminiscent of entropy production in physicochemical systems) in the growth plate
(Figures 1A and 2C).

The thermodynamic significance of the biological processes taking place in the formation of the
growth plate pattern (Box 2) may indicate a deeper relationship between developmental

Box 1. The Thermodynamics of Life

The second law of thermodynamics states that isolated systems tend towards thermodynamic equilibrium that is
characterized by maximum entropy, namely maximum molecular disorder or randomness. Living systems, however,
appear to defy the second law by remaining in a highly organized state. Erwin Schrödinger (Nobel Laureate in Physics in
1933) resolved this apparent contradiction by noting that living systems are not isolated but open: they import energy
from their environment and export to it the entropy generated by their internal processes, such that the entropy of the
whole system, organism plus environment, increases in agreement with the second law [35]. Ilya Prigogine (Nobel
Laureate in Chemistry in 1977) provided the crucial link between entropy export and self-organization. According to
Prigogine, open systems use the energy of the environment to self-organize into robust spatial patterns that increase the
rate of entropy export [36], as do the convection rolls in the R-B convection system (see Figure 2B in main text) and the
apoptotic death in the chondrocyte columns of the growth plate (see Figures 1A and 2C in main text). Prigogine named
such patterns ‘dissipative structures’ because they dissipate energy: they degrade imported energy into energy with
high entropy content. For example, kinetic energy is degraded through friction into thermal energy that is distributed to
random motions of the molecules of the environment, and is therefore of high entropic content, unable to do work.
According to recent studies in non-equilibrium thermodynamics, inanimate physicochemical systems have a tendency
to evolve towards dissipative structures. They continuously absorb energy from their environment to overcome
activation barriers and adopt new structures by re-arranging their components. When it happens that the new randomly
adopted structure dissipates the absorbed energy as heat, the system has no energy to jump back the barrier which
makes the new structure more robust and irreversible than the previous one. The longer time-window of existence of the
new structure allows it to be further modified randomly to a more irreversible one that absorbs and dissipates energy
more efficiently until finally very robust dissipative structures appear and persist (‘dissipative adaptation’ [37]). The
thermodynamic stability of such inanimate dissipative structures makes highly probable the involvement of their
physicochemical mechanisms in the biological processes of living systems, as has been proposed for ancestral
developmental mechanisms [19], and may explain the origin of the robustness of developmental modules.

Trends in Biotechnology, November 2018, Vol. 36, No. 11 1119

 Environment Environment Environment
Buyoancy
Entropy Entropy
(metabolic byproducts, (apoptosis)
heat) Heat of high entropy

Cold plate Viscosity

Anabolism

Carlage matrix
c c c
ADP ATP

Catabolism Hot plate Temperature

gradient Oxygen
Heat of low entropy gradient
Energy
(carbohydrates) Energy
(blood supply)

(A) (B) (C)

Figure 2. Dissipative Structures. (A) Cells dissipate/degrade via catabolic reactions low-entropy energy taken from environmental sources (such as carbohydrates)
into high-entropy wastes (such as metabolic byproducts and heat) that they secrete into the environment, thereby keeping their entropy low while increasing the entropy
of the environment. Part of the imported energy is used by the cell for self-organization into spatial patterns through anabolic reactions that build the complex structures
of macromolecules. (B) In Rayleigh–Bénard (R-B) convection, a temperature gradient is created when a horizontal fluid layer contained between two parallel plates is
heated from below. For small temperature gradients, heat is transferred across the fluid by conduction via microscopic molecule-to-molecule collisions. Above a
temperature gradient, the system uses part of the imported thermal energy to build up the kinetic energy required to maintain macroscopic fluid movements that are self-
organized in a pattern of counter-rotating convection rolls. The energy imported by heat causes the thermal expansion of fluid elements, making them less dense, which
creates an upward buoyancy force that overcomes the viscous forces that prevent collective fluid movements. Fluid reaching the upper cold plate is cooled exporting
entropy and, because it is denser, it sinks. (C) In the growth plate, energy provided by the blood supply in the chondrocytes of the proliferating zone is used by the cells
for the construction of the pattern of parallel chondrocyte columns through cell multiplication and differentiation, as well as through the construction of cartilage matrix
around the chondrocyte columns. Chondrocytes reaching the other side of the columnar pattern export entropy by dying through apoptosis.

mechanisms and the physical mechanisms of pattern formation in inanimate dissipative

structures. Interestingly, it has been argued that developmental mechanisms have originated
in the early history of multicellular life through physical mechanisms that were in place long
before the elaboration of complex genetic programs for pattern formation by evolution [19].
These mechanisms, which were progenitors of the contemporary tissue shape transformations
in embryonic processes, involved the action of physical forces, such as gravity, phase
separation, adhesion, and interfacial tension, on multicellular clusters that exhibit viscoelastic
behavior susceptible to such forces. The ancestral physical morphogenetic mechanisms were
later stabilized by evolution, which installed complex genetic programs that fine-tuned and
consolidated them [19]. For example, it has been proposed that a primitive reflex response to
mechanical deformation, such as the phagocytosis of particles in response to physical contact,
evolved into a mechanically induced genetic program for the generation of a permanent gut in
Drosophila [20]. The involvement of physical forces similar to those operating in inanimate
dissipative structures is notable in the growth plate. Gravity, for instance, is involved in pattern
formation in both systems, R-B convection and growth plate. In R-B convection, gravity creates
buoyancy forces that lead to the upward motion of hot fluid elements in rotating convection rolls
[16] (Figure 2B). In the growth plate, gravity generates mechanical loading that exerts tension on
the cytoskeleton of chondrocytes [21]. In turn, the tension of the cytoskeleton determines its
interaction with the extracellular matrix that induces and controls the chondrocyte movements
such that they could be placed on the top of each other to form columns [21] so as to avoid the
stiff matrix of the intercolumnar space [22].

1120 Trends in Biotechnology, November 2018, Vol. 36, No. 11

 Box 2. The Thermodynamics of the Growth Plate
The extended resemblance of the growth plate (see Figures 1A and 2C in main text) to the R-B convection system
(Figure 2B) in the thermodynamics of pattern formation demonstrates the feasibility of describing developmental
modules as distinct thermodynamic units of a developing organism and renders the growth plate a model-systems for
thermodynamic studies in development. Numerous analogies can be detected between the two systems.
(i) Parallel and continuously renewed structures of macroscopic spatial patterns (columns of differentiating chon-
drocyte in the growth plate [7], rotating convection rolls of fluid in R-B convection [16]).
(ii) Gradients along which energy is transferred and patterns are formed (oxygen [25], temperature [16]).
(iii) Unidirectional energy input at one side of the gradient (epiphyseal blood vessels [24], heating at the bottom plate
[16]) and entropy export at the opposite side of the gradient (apoptosis of hypertrophic chondrocytes [17], heat
export from the upper low-temperature plate [16]).
(iv) Robustness of regular patterns inside a parameter range, and irregularity of patterns at high energy input
(disorganization of the columnar pattern with a high chondrocyte differentiation rate induced by genetic mod-
ifications of the Ihh/PTHrP signaling loop [30], deformation of the regular closed periodic trajectories of the fluid
elements of the rotating convection rolls into irregular, turbulent trajectories for high temperature gradients, as with
boiling water [16]).
(v) Reformation of the patterns after spatial disturbance (removal of proliferative and hypertrophic zones [38],
temporal stirring).
(vi) Lack of patterns when energy import is low (local apoptosis dispersed throughout the growth plate when
epiphyseal vascularization is impaired [39], heat transfer by conduction, i.e., local molecule-to-molecule inter-
actions throughout the fluid layer, in low heating [16]).
(vii) Competing forces for the creation of patterns (matrix stiffness that pushes chondrocytes along the columns [22]
and cell-to-cell adhesion that keeps them close to each other [40], buoyancy that pushes the fluid elements up
and viscosity that keeps them bound to the bulk fluid [16].
Such extended and detailed thermodynamic correspondence between an inanimate and a living dissipative structure
may indicate that the genetic programs of the biological processes in developmental modules were built evolutionary to
support energy dissipation that was initially performed in their ancestral form by physical mechanisms which have been
proposed for the origin of developmental mechanisms [19].

Thermodynamically Relevant Bioartificial Tissue Design Rules:

Developmental Modules as Mutually Sustained Dissipative Structures
Thermodynamic Scaffold Design
The above considerations justify the study of developmental modules as dissipative structures
for determining new rules for the fabrication of authentic tissue structures, in direct relation with
the physical laws. How, for instance, the structure of developmental modules has evolved to
optimize their biological processes that contribute to entropy export by cell death, or by any
other relevant mechanism such as the degradation of extracellular matrix by enzymes secreted
by chondrocytes [23], is a question of paramount interest in evolutionary biology but also with
important practical implications in the fabrication of bioartificial tissues. To pursue questions
related to the thermodynamic significance of tissue structures, parameters related to the spatial
distribution of cell differentiation and metabolic functions in developmental modules, which
when lumped together appropriately provide global thermodynamic indices related to energy
dissipation (energy import and entropy export), should be first determined. These parameters
will then permit the design of thermodynamically optimal scaffold geometries that could
facilitate the energy dissipation and consequently the in vitro formation of multicellular patterns
similar to authentic tissue structures. A scaffold design methodology, based on thermody-
namically relevant indices that incorporate key design parameters, can circumvent the vast
combinatorial combination of parameters that are presently tested with trial-and-error
approaches, which create difficulties in deciphering their relative importance [2].

In Vitro Control Parameters for Energy Dissipation

Both systems, inanimate dissipative structures and developmental modules, engage their
constituent elements in massive, global transfer of imported energy along gradients.

Trends in Biotechnology, November 2018, Vol. 36, No. 11 1121

Growth plate chondrocytes carry the energy provided to them by the epiphyseal veins [24]
from the proliferating to the hypertrophic zone along an oxygen gradient [25] (Figures 1A
and 2C), as the convection rolls of the R-B convection system do with the imported heat
from the hot to the cold plate [16] (Figure 2B). The low oxygen concentration in the
hypertrophic zone is responsible for the entropy export by apoptotic cell death. The lack
of oxygen compromises metabolically the chondrocytes and predisposes them to apoptotic
death that is triggered by apoptogens produced by hydrolysis of the cartilage matrix by cells
carried by the capillaries that invade the hypertrophic zone [17] (Figure 1A). However, in
virtue of genetic programs added by evolution to support the putative physical mechanisms
of energy dissipation in ancestral developmental mechanisms [19], developmental mod-
ules, in contrast to inanimate dissipative structures, create and maintain their gradients by
endogenous processes. In this way they are able to self-control energy transfer such that
their patterns are established and persist without external interference. The oxygen gradient
along the growth plate is maintained by the chondrocytes themselves because they inhibit
angiogenesis in the whole growth plate, except in the hypertrophic zone, by secreting
various anti-angiogenic factors [26]. Furthermore, through calcium secretion, chondrocytes
calcify the extracellular matrix in the last part of hypertrophic zone, thereby increasing its
density [27] and restraining the diffusion of oxygen and nutrients [28] such that oxygen
availability decreases drastically [25]. On the contrary, in inanimate dissipative structures,
the gradients are imposed and maintained by external constraints (temperatures in the
plates in R-B convection, Figure 2B) and the gradients and patterns disappear when these
constraints are removed.

The maintenance of gradients that induce and direct energy transfer for pattern formation,
instead of the commonly used high uniform concentrations of metabolites and oxygen to
promote cell viability, therefore constitutes a crucial control objective for the in vitro fabrication
of bioartificial tissues with authentic tissue structures. For the same reason, special care should
be taken in vitro to preserve apoptosis, which, although detrimental at the cell level, represents
the entropy that should be exported for the spatial organization of cells into tissue structures.
Indeed, counterintuitively, but in agreement with the perception of apoptosis as an entropy
export mechanism, decreased apoptosis in the growth plate of mutant mice leads to loss of
chondrocyte organization into parallel columns [29]. Similarly, preventing entropy export that
takes place through the degradation of extracellular matrix by chondrocytes also leads to
disorganization of the growth plate [23].

In Vitro Compliance with the Thermodynamic Limits of Biological Processes

Although dissipative structure patterns are generally robust inside a given parameter range,
they are unstable and tend to disappear outside this range (Box 2). For instance, the convection
rolls of the R-B convection system are disorganized, and the motion of fluid becomes turbulent
in large temperature gradients [16] (Figure 2B). Similarly, for the columnar pattern of the growth
plate to be maintained, the chondrocyte differentiation rate must be within a given range which
is determined by the signaling loop of Indian hedgehog (Ihh)/parathyroid hormone-related
protein (PTHrP) (Figure 1B). The column elongation by new chondrocytes that start their
differentiation program and enter the columns can be then balanced with the construction
of a dense cartilage matrix in the intercolumnar space. The high density of the intercolumnar
matrix stabilizes the columnar pattern because it does not allow the chondrocytes to move
outside the columns [22]. However, when genetic modifications in the signaling loop increase
the chondrocyte differentiation rate above a given limit, many differentiating chondrocytes enter
the columns, extending them rapidly such that construction of the dense intercolumnar matrix
cannot be completed, and the chondrocytes can then move to other directions outside the

1122 Trends in Biotechnology, November 2018, Vol. 36, No. 11

columns, disorganizing the columnar pattern [30]. Therefore, the differentiation rate becomes a
pivotal parameter that should be controlled appropriately in vitro instead of simply achieving cell
differentiation or aiming at rapid differentiation.

Complex Tissues Fabricated In Vitro by the Assembly of Interacting Developmental

Modules
Interactions between developmental modules also highlight their differences from inanimate
dissipative structures. Hypertrophic chondrocytes of the growth plate module induce
angiogenesis, and thus energy import, in the adjoining module of the developing bone
[11]. In turn, the newly formed blood vessels in the bone module carry cells from the
underlying bone marrow that degrade the cartilage matrix, producing apoptogens that
induce entropy export by apoptosis in the module of the growth plate [17]. Dying chon-
drocytes leave behind a calcified cartilage matrix between the columns that becomes the
scaffolding for bone deposition [27], in this way pre-patterning cell organization in the bone
module, which will therefore continue supporting entropy export in the growth plate module
inducing chondrocyte apoptosis. Restoring such interactions in vitro between bioartificial
developmental modules will allow the modular fabrication of complex bioartificial tissues
and organs from a thermodynamically permissible assembly of mutually sustained co-
developing developmental modules, as takes place in vivo (Figure 3). For instance, a
thermodynamically and developmentally meaningful in vitro process for the fabrication of
bioartificial osteochondral tissue for articular cartilage repair requires that bone formation
takes place on a bioartificial growth plate instead of joining bioartificial cartilage and
bioartificial bone components that are fabricated separately on different scaffolds. In this
way, bone can penetrate the growth plate, resulting in the integration of bone and cartilage
phases, which is necessary for the mechanical strength of the osteochondral tissue.
However, when two separately fabricated bioartificial cartilage and bone components
are joined, the mechanical strength is low because of the lack of integration, requiring
the insertion of an artificial compact layer between the two tissues [31] that complicates the
fabrication process and is difficult to accurately match the mechanical strength of the
natural integrated osteochondral tissue.

Preliminary Practical Considerations for Fabricating Authentic Tissue Structures

The examples presented above show that difficult problems in tissue engineering, such as
optimizing scaffold geometry, determining in vitro parameters that must be controlled to
assure the formation of robust bioartificial tissue structures, and fabricating complex tissues
and organs with integrated tissue components, can be better approached through thermo-
dynamic considerations that indicate crucial biological processes involved in energy dissipa-
tion in developing tissues. The major advantage of the thermodynamic approach in
comparison to traditional approaches in tissue engineering is that it can assure the structural
robustness of the fabricated bioartificial tissues and consequently the reproducibility of
bioartificial tissue properties which is necessary for manufacturing and regulatory procedures.
Even though such an approach requires tissue engineers to be familiar with concepts of
developmental biology, its application should be attempted because it is a constructive
approach that leads to bioartificial tissues with progressively more complex structures
overcoming the need for separate designs. The limited analysis presented above can provide
some preliminary practical considerations for the application of the thermodynamic approach
in the fabrication of bioartificial tissues.

The fabrication of authentic tissue structures should start with the in vitro restoration of the
thermodynamic operation of the simplest developmental module, in terms of the dependence

Trends in Biotechnology, November 2018, Vol. 36, No. 11 1123

 Developmental
module 2

on e
Dissipave

c ul
s
ra od processes
in term
In
te

Dissipave
structure
Dissipave
processes

Dissipave Complex
structure
Developmental developing
module 1 ssue

Figure 3. Thermodynamic Design of Bioartificial Tissues. Thermodynamic design rules for the fabrication of a
bioartificial developmental module (developmental module 1, e.g., growth plate), aim towards the in vitro recapitulation of
biological processes that are involved in the energy dissipation (dissipative processes; e.g., energy input by the provision of
nutrients and oxygen, cell apoptosis, formation of an oxygen gradient). Such dissipative processes lead to cellular
organization into tissue structures/patterns (dissipative structures; e.g., chondrocyte columnar pattern). Conversely,
dissipative structures sustain the dissipative processes because energy is transferred through their patterns for export
into the environment (e.g., differentiating chondrocytes carry energy as they move along the columns, and form an oxygen
gradient that causes their apoptotic death through which entropy is exported). The mutually reinforcing interaction
between dissipative processes/dissipative structures (blue arrows) makes feasible the in vitro fabrication of bioartificial
developmental modules. If conditions are chosen that allow the operation of biological processes involved in energy
dissipation, this will lead to the initial formation of the pattern of a dissipative structure, and this structure subsequently by
itself will sustain the dissipative processes that are needed for its maintenance. More complex structures (e.g., bioartificial
osteochondral tissue) can be fabricated with the interconnection of bioartificial developmental modules (e.g., develop-
mental module 1, growth plate; and developmental module 2, developing bone/bone marrow) that restores cooperative
interactions (red arrows) that facilitate dissipative processes in both modules (e.g., induction of hypertrophic chondrocyte
apoptosis by the cells of bone/bone marrow module, vascularization of the bone/bone marrow module by signals secreted
by hypertrophic chondrocytes).

on other co-developing tissues according to developmental biology, for a developing tissue

system (Figure 3). For instance, to fabricate a bioartificial osteochondral tissue, the growth plate
should be chosen as the first bioartificial tissue to be made because of its independence on
external factors for the differentiation of mesenchymal stem cells into chondrocytes [32], for cell
organization into a columnar pattern [8], and for the deposition of calcium and calcium
phosphate formation (mineralization) in cartilage matrix [11]. Biological processes involved
in energy dissipation should be selected for appropriate control in an in vitro system. The design
of an in vitro system for 3D cell differentiation should aim to avoid artificial constraints that
compromise the biological processes necessary for thermodynamic operation instead of
applying specific conditions. This is because the cells of a developmental module are
evolutionarily programmed to form and preserve the necessary gradients that allow energy
dissipation, so that the thermodynamic operation necessary for the formation and maintenance
of cellular patterns is self-sustained (Figure 3).

1124 Trends in Biotechnology, November 2018, Vol. 36, No. 11

However, some specific externally imposed conditions are necessary because no develop- Outstanding Questions
mental module is fully autonomous but depends on other co-developing developmental How thermodynamic principles can
modules with which cooperates in energy dissipation (Figure 3) to a degree that has to be apply in the spatiotemporal organiza-
tion of development representing the
determined from relative information in developmental biology and ultimately tested experi- operation of developmental modules
mentally. For instance, apoptogens, which originate from matrix degradation by cells carried to as dissipative structures and their
the hypertrophic zone from the bone marrow [17], should be externally added in a bioartificial interactions for mutually sustainable
energy dissipation and pattern
growth plate to continually remove entropy by cell apoptosis. A bioartificial growth plate
formation?
developmental module could be then connected to an in vitro system for bone marrow culture
(e.g., co-culture in double-layered hydrogel construct) to provide apoptogens to the growth How can thermodynamic consider-
plate but also form a thermodynamically more independent integrated system of two interacting ations of development be integrated
developmental modules as it is needed for the development of the more complex osteochon- into a new methodology that provides
a complete set of design rules for bio-
dral tissue (Figure 3). This arrangement would provide the necessary cells, such as osteoclasts,
artificial tissue fabrication from a
osteoblasts, and epithelial cells, to restore thermodynamically meaningful interactions between thermodynamically optimal assembly
the two modules, leading to degradation of the cartilage matrix (entropy export), deposition of of bioartificial developmental
bone matrix (pattern formation), and the development of blood vessels inside the developing modules?

bone (energy import), respectively. Such interconnection of developmental modules can

Which thermodynamic constraints
provide bioartificial osteochondral tissue with integrated cartilage and bone phases, and
determine the structure and function
can also control the relative thickness of cartilage and bone adjusting cartilage matrix deposi- of developmental modules? Is there
tion and resorption by the relative number of osteoblasts and osteoclasts. room for improvement of their opera-
tion for specific clinical applications?
How can engineered tissues be clini-
The Future of Tissue Engineering: Towards Deciphering General Principles cally optimized by genetically modify-
ing the biological processes of energy
of Development
dissipation in bioartificial developmen-
The interaction of tissue engineering with developmental biology within the framework of the tal modules?
thermodynamics of developmental modules can advance our fundamental knowledge of devel-
opment which will subsequently allow the fabrication of clinically efficacious bioartificial tissues. Can thermodynamic criteria for the
Elucidating the way that developmental modules cooperate to support each other in energy evolution of developmental modules
be extracted and tested in bioartificial
dissipation which is needed for the formation of the cellular patterns of tissue structures could lead
developmental modules in vitro via
to general principles that operate through genetic programs to control the spatiotemporal genetic modifications of their biological
organization of development (see Outstanding Questions). Such principles could describe processes involved in energy dissipa-
how whole-body development is thermodynamically organized by co-developing, interacting tion shedding light in the past of evo-
lution but also permitting predictions?
entropy-exporting units so as the development of the different parts of an organism to be
spatiotemporally coordinated and the entropy produced by the processes of cell organization
into tissue and organ structures throughout a developing body to be optimally exported to the
environment. Tissue engineering can assist developmental biology by experimentally pursuing
such studies in bioartificial developmental modules made in vitro from stem cells, and assembled
in systems of interconnected modules of gradually increasing complexity, instead of dealing
directly with the overwhelming complexity of multiple interacting modules in a developing organ-
ism. Such studies, conducted with human stem cells, will constitute a strong scientific basis for
clinical translation because they could determine if the clinical performance of bioartificial tissues is
bounded by natural developmental constraints of the current evolutionary time, or, as has been
recently proposed by tissue engineers, whether it is possible to redesign developmental modules
to have optimal behavior in the microenvironment of an adult [33] or damaged tissue [34], which
could be achieved by thermodynamically permissible genetic modification of the biological
processes that participate in energy dissipation in developmental modules.

References
1. Kaplan, D.L. et al. (2005) It takes a village to grow a tissue. Nat.
Biotechnol. 23, 1237–1239
2. Ingber, D.E. et al. (2006) Tissue engineering and developmental
biology: going biomimetic. Tissue Eng. 12, 3265–3283
3. Lenas, P. et al. (2009) Developmental engineering: a new para-
digm for the design and manufacturing of cell-based products.
Part I. From three-dimensional cell growth to biomimetics of in
vivo development. Tissue Eng. Part B Rev. 15, 381–394

Trends in Biotechnology, November 2018, Vol. 36, No. 11 1125

4. Raff, R.A. (1996) The Shape of Life: Genes, Development, and the
Evolution of Animal Form, University of Chicago Press
5. Harrison, R.G. (1918) Experiments on the development of the
forelimb of Amblystoma, a self-differentiating equipotential sys-
tem. J. Exp. Zool. 25, 413–461
6. Von Dassow, G. et al. (2000) The segment polarity network is a
robust developmental module. Nature 406, 188–192
24. Wirth, T. et al. (2002) The blood supply of the growth plate and the
epiphysis: a comparative scanning electron microscopy and his-
tological experimental study in growing sheep. Calcif. Tissue Int.
70, 312–319
25. Schipani, E. et al. (2001) Hypoxia in cartilage: HIF-1alpha is
essential for chondrocyte growth arrest and survival. Genes
Dev. 15, 2865–2876

7. Long, F. and Ornitz, D.M. (2013) Development of the endochon- 26. Shukunami, C. et al. (1999) Spatiotemporal pattern of the mouse
dral skeleton. Cold Spring Harb. Perspect. Biol. 5, a008334 chondromodulin-I gene expression and its regulatory role in vas-
cular invasion into cartilage during endochondral bone formation.
8. Abad, V. et al. (1999) Determinants of spatial polarity in the growth
Int. J. Dev. Biol. 43, 39–49
plate. Endocrinology 140, 958–962
27. Amizuka, N. et al. (2012) Histology of epiphyseal cartilage calcifi-
9. Vortkamp, A. et al. (1996) Regulation of rate of cartilage differentiation
cation and endochondral ossification. Front. Biosci. 4, 2085–
by Indian hedgehog and PTH-related protein. Science 273, 613–622
2100
10. Gerstenfeld, L.C. et al. (2002) Chondrocytes provide morpho-
28. Williams, R.M. et al. (2007) Solute transport in growth plate
genic signals that selectively induce osteogenic differentiation of
cartilage: in vitro and in vivo. Biophys. J. 93, 1039–1050
mesenchymal stem cells. J. Bone Miner. Res. 17, 221–230
29. Donohue, M.M. and Demay, M.B. (2002) Rickets in VDR null mice
11. Gerber, H.P. et al. (1999) VEGF couples hypertrophic cartilage
is secondary to decreased apoptosis of hypertrophic chondro-
remodeling, ossification and angiogenesis during endochondral
cytes. Endocrinology 143, 3691–3694
bone formation. Nat. Med. 5, 623–628
30. Lanske, B. et al. (1996) PTH/PTHrP receptor in early development
12. Scotti, C. et al. (2013) Engineering of a functional bone organ
and Indian hedgehog-regulated bone growth. Science 273, 663–
through endochondral ossification. Proc. Natl. Acad. Sci. U. S. A.
666
110, 3997–4002
31. Da, H. et al. (2013) The impact of compact layer in biphasic
13. Kuhn, L.T. et al. (2014) Developmental-like bone regeneration by
scaffold on osteochondral tissue engineering. PLoS One 8,
human embryonic stem cell-derived mesenchymal cells. Tissue
e54838
Eng. Part A 20, 365–377
32. Aubin, J.E. (1998) Bone stem cells. J. Cell Biochem. (Suppl. 30/
14. Bernhard, J. et al. (2017) Tissue-engineered hypertrophic chondro-
31), 73–82
cyte grafts enhanced long bone repair. Biomaterials 139, 202–212
33. Tonnarelli, B. et al. (2014) Re-engineering development to instruct
15. Martin, I. (2014) Engineered tissues as customized organ germs.
tissue regeneration. Curr. Top. Dev. Biol. 108, 319–338
Tissue Eng. Part A 20, 1132–1133
34. Dennis, S.C. et al. (2015) Endochondral ossification for enhancing
16. Manneville, P. (2006) Rayleigh–Bénard convection: thirty years of
bone regeneration: converging native extracellular matrix bioma-
experimental, theoretical, and modelling work. In Dynamics of
terials and developmental engineering in vivo. Tissue Eng. Part B
Spatio-Temporal Cellular Structures: Henri Bénard Centenary
Rev. 21, 247–266
Review. Springer Tracts in Modern Physics (Vol. 207) (Mutabazi,
I. et al., eds), In pp. 41–65, Springer 35. Schrödinger, E. (1944) What Is Life? The Physical Aspect of the
Living Cell. Cambridge University Press
17. Shapiro, I.M. et al. (2005) Fate of the hypertrophic chondrocyte: micro-
environmental perspectives on apoptosis and survival in the epiphyseal 36. Nicolis, G. and Prigogine, I. (1977) Self-Organization in Non-
growth plate. Birth Defects Res. C Embryo Today 75, 330–339 Equilibrium Systems: From Dissipative Structures to Order
Through Fluctuations, Wiley
18. Roach, H.I. et al. (2004) Chondroptosis: a variant of apoptotic cell
death in chondrocytes? Apoptosis 9, 265–277 37. England, J.L. (2015) Dissipative adaptation in driven self-assem-
bly. Nat. Nanotechnol. 10, 919–923
19. Newman, S.A. (2012) Physico-genetic determinants in the evo-
lution of development. Science 338, 217–219 38. Abad, V. et al. (2002) The role of the resting zone in growth plate
chondrogenesis. Endocrinology 143, 1851–1857
20. Farge, E. (2003) Mechanical induction of Twist in the Drosophila
foregut/stomodeal primordium. Curr. Biol. 13, 1365–1377 39. Maes, C. et al. (2004) Soluble VEGF isoforms are essential for
establishing epiphyseal vascularization and regulating chondro-
21. Aszodi, A. et al. (2003) Beta1 integrins regulate chondrocyte rota-
cyte development and survival. J. Clin. Invest. 113, 188–199
tion, G1 progression, and cytokinesis. Genes Dev. 17, 2465–7249
40. Romereim, S.M. et al. (2014) A dynamic cell adhesion surface
22. Prein, C. et al. (2016) Structural and mechanical properties of the
regulates tissue architecture in growth plate cartilage. Develop-
proliferative zone of the developing murine growth plate cartilage
ment 141, 2085–2095
assessed by atomic force microscopy. Matrix Biol. 50, 1–15
23. Inada, M. et al. (2004) Critical roles for collagenase-3 (Mmp13) in
development of growth plate cartilage and in endochondral ossi-
fication. Proc. Natl. Acad. Sci. U. S. A. 101, 17192–17197

1126 Trends in Biotechnology, November 2018, Vol. 36, No. 11