Review

In search of principles for a Theory of Organisms

GIUSEPPE LONGO1,2 , MAËL MONTÉVIL2,3 , CARLOS SONNENSCHEIN2,4 and ANA M SOTO1,2,*

1
Centre Cavaillès, République des Savoirs, CNRS USR3608, Collège de France et Ecole
Normale Supérieure, Paris, France
2
Department of Integrative Physiology and Pathobiology, Tufts University School of Medicine,
Boston, MA, USA
3
CNRS, UMR8590 IHPST, Paris, France
4
Centre Cavaillès, École Normale Supérieure, and Institut d’Etudes Avancees de Nantes, Nantes, France

*Corresponding author (Fax, 617-636-3971; Email, [email protected])

Lacking an operational theory to explain the organization and behaviour of matter in unicellular and multicellular
organisms hinders progress in biology. Such a theory should address life cycles from ontogenesis to death. This theory
would complement the theory of evolution that addresses phylogenesis, and would posit theoretical extensions to
accepted physical principles and default states in order to grasp the living state of matter and define proper biological
observables. Thus, we favour adopting the default state implicit in Darwin’s theory, namely, cell proliferation with
variation plus motility, and a framing principle, namely, life phenomena manifest themselves as non-identical
iterations of morphogenetic processes. From this perspective, organisms become a consequence of the inherent
variability generated by proliferation, motility and self-organization. Morphogenesis would then be the result of the
default state plus physical constraints, like gravity, and those present in living organisms, like muscular tension.

[Longo G, Montévil M, Sonnenschein C and Soto AM 2015 In search of principles for a Theory of Organisms. J. Biosci.] DOI 10.1007/s12038-
015-9574-9

Whether you can or cannot observe a thing depends on much either. This is due in part to the preponderance of ‘prag-
the theory which you use. It is the theory which decides matic systems biology,’ a practice that emphasizes large-scale
what can be observed. molecular interactions which is technology-driven and does not
claim explicit theory commitments (O’Malley and Dupre 2005).
– Einstein, oral remark quoted in Salam (1990) Fifty years ago the notions of program, information and
signal borrowed from mathematical theories seemed to resolve
the differences between physics and biology made explicit by
Kant: teleology, now under the guise of a program, became
1. Motivation acceptable to reductionists who claimed that biology could be
reduced to physics and chemistry (Jacob 1974). However, the
Biologists acknowledge a crisis in their midst: On the one hand, notions of program, information and signal borrowed from the
reams of data acquired from a reductionist perspective (for mathematical theories of information have been hindering
example, ‘transcriptomics’) do not provide the anticipated un- progress in biology (Longo et al. 2012a). These notions re-
derstanding of the subject matter of their interest; on the other quire searching for biological specificity at the molecular level
hand, the application of mathematical modeling has not helped which is not coherent with that of biomechanics, another

Keywords. Autopoiesis; biological organization; criticality; default state; development; emergence; mathematical symmetries

http://www.ias.ac.in/jbiosci J. Biosci. * Indian Academy of Sciences

 Giuseppe Longo et al.

important contributor to phenotypes. Yet biomechanics be- Boyle and Mariotte), allowed for the development of a
came prominent in several biological fields including cancer theory at that level of observability. Further work at the
research. Indeed, tumours are palpable, a fact obviously related macroscopic level originated modern thermodynamics
to their being more rigid than the surrounding normal tissue. (TD), another major achievement of the 19th century; this
Thus, technical knowhow is being introduced together with was well before Boltzmann’s unification with an atomistic
new operational definitions which may not be coherent with perspective. Notwithstanding, atoms and molecules are
the ones already in use.1 proper observables in another physical theory, quantum me-
The above-referred practice contributes to the feeling that chanics (QM).
biological complexity is unfathomable and that generaliza- When living entities die, they decompose into particles of
tions and global concepts are unhelpful.2 Thus, the theoret- inert matter, and in turn, living organisms assemble the same
ical bases underlying the experimental programs being inert matter in novel ways. The emergence of these novelties
pursued remain only implicit. Meanwhile, the explosion of requires suitable theoretical constructs (Longo and Montévil
data continues unabated having neither sound theoretical 2012; Longo and Montévil 2014; Saçlioglu et al. 2014). That
bases nor an adequate language to make sense of them. is, besides the already acknowledged physical principles and
While acknowledging the immense complexity of organ- default states, additional principles and theoretical require-
isms, we dare to think that what Darwin achieved for evolu- ments are needed to describe proper biological observables,
tion could eventually materialize for ontogenesis: this will be such as phenotypes. These extensions of physical laws into
achieved through the elaboration of an appropriate theoreti- biology must be compatible with physical theory about inert
cal framework. matter, i.e. organisms should not violate the laws of thermo-
dynamics, gravity or the quantum properties of their compo-
nent particles. Yet, these principles may not suffice to make
2. The role of theories
the biological dynamics intelligible at the phenotypic level.
In biology, other than Darwin’s theory of evolution
At the beginning of the Scientific Revolution, scientists (1859), the creation of global theories has not been as suc-
thought that they had direct access to their outside world: cessful as in physics (Soto and Sonnenschein 2012). Among
God’s will was to make nature intelligible to creatures like other subjects, Darwin’s theory on the origin of species
us. The separation of science and religion was a long and addressed common descent, encompassed a long time-
complex process: Kant’s philosophy and Darwin’s theory frame and provided an adequate explanation of phylogeny.
were major contributors to this separation. Ever since, sci- The fundamental principles in Darwin’s theory are (a) de-
entists acknowledge that they are inside the world they wish
scent with modification and (b) natural selection. However,
to observe and study. As a result of this realization, objec-
biology has yet to produce a theory of organisms that would
tivity had to be constructed through scientific theories that
encompass ontogeny and life cycles, i.e. phenomena occur-
would provide intelligibility principles to frame observa-
ring on a time-scale going from conception to death (Elsasser
tions, experiments and explanations. In the 16th and 17th
centuries, physicists developed theories that provided an 1987). Recently, several worthy contributions have been
accurate description of phenomena of the inert. In this made in this area (Kupiec 2010; Deacon 2012; Newman
theory-rich context, scientists were aware that theories de- 2012; Davies 2013). Approaches based exclusively on
termine which are the proper observables while, conversely, stochasticity and natural selection like the ontophylogenesis
the choice of observables was a major theoretical commit- theory (Kupiec 2010) provide new perspectives; however,
ment. For example, the decision to investigate the relation- they are insufficient to frame a theory of organisms because
ship between pressure, volume and temperature well before the molecular events at the core of the ontophylogenesis
considering the atomic structure of gases (17th century, theory are causally dependent on cell, tissue and organismal
contexts and these contexts are not addressed by this
approach.
1
In operationalism, scientific terms are defined by the experimental Since the 1970s, the thermodynamics of dissipative sys-
operations which determine their applicability. For example, epidermal tems provided an opportunity to examine the relevance of
growth factor is a misnomer because it does not induce proliferation.
However, it induces cell spreading in one assay: when cells spread, the
self-organizing physical systems to the understanding of the
circumference of the colony is longer, and as only the cells in the emergence of life, as exemplified by the pioneering work of
periphery proliferate, EGF in this assay seems to increase cell number. Prigogine and his school (Nicolis and Prigogine 1977),
When given in vivo, it induces early eye opening, a process character- Kauffman (Kauffman 1993) and others. For instance,
ized by cell death. So, in one operational definition it increases cell Cottrell elegantly highlighted the role of thermodynamics
number, in the other, it increases cell death.
2
For example, some biologists believe that there are exceptions to
while stressing the need for a biological perspective that
every rule, and that there are no valid ‘principles’, ‘rules’ or overarching would bring to this analysis distinctive biological character-
concepts. istics such as historicity and purposiveness (Cottrell 1979).

J. Biosci.
 Theory of Organisms

Unfortunately, despite these promising beginnings, the de- 3.2 Default states and symmetries in physics
velopment of a theory of organisms has been hindered by the
misuse of metaphors borrowed from theories of information Since Galileo and Descartes, physicists made the default
(Longo et al. 2012a). To remedy this situation, we are state of inert matter explicit: this is inertia, an instance of
proposing instead explicit principles for the elaboration of fundamental conservation laws.4 Briefly, if no cause (force)
a theory of organisms that would make such a theory robust, modifies the properties of an object, the object conserves its
and open to change when challenged by empirical evidence. properties.5 In short, a default state is what happens when
Our analysis purposefully refers to physics, not only nothing is done to the object in question. Three centuries
because biological theory should not violate physical laws, later, E Noether’s theorems provided a deeper understanding
but also because there is a tradition in organismal biology to for this default state by mathematically justifying conserva-
use both the similarities and differences between these dis- tion properties of energy and momenta in terms of symme-
ciplines to advance biological knowledge. For example, tries in the state equations (van Fraassen 1989). Ever since,
Helmholtz improved the understanding of both the physiol- symmetries (and their breaking) acquired an even more
ogy of nerve conduction and muscle metabolism while de- fundamental role in physics (table 1). We are referring to
veloping the principle of conservation of energy in physics the notion of ‘symmetries’ in the broad sense given to it by
(Lenoir 1982). An additional example is the current wave of modern physics, as transformations preserving the key in-
physicists and mathematicians entering the field of Systems variants observed and proposed by the intended theory. For
Biology while carrying with them the theoretical framework example, Noether’s theorem enables us to understand the
of physics.3 Most importantly, biology may make it possible conservation of energy and momentum on the basis of time
to identify a new physical principle (Moore 2012). There- and space symmetries of fundamental equations, respective-
fore, before stating the proposed principles for a theory of ly. In short, the conservation of these quantities is grounded
organisms, we will elaborate on some relevant relationships on the idea that the ‘laws’ of physics are the same at different
between physics and biology (table 1). positions and times. However, the notion of symmetries is as
old as Archimedes’ principles as exemplified by the lack of
3. From physics to biology movement of a scale with equal and symmetric weights on
each arm. Moreover, physicists have proposed the existence
3.1 The impact of physics on biology of a particle for symmetry reasons; the existence of anti-
matter was proposed this way. In biology, instead, the types
of symmetries usually referred to are a subset within the
First, we will go over the fundamental role that symmetries
broad category of symmetries in mathematics; for example,
and conservation principles play in physical theories, in
symmetry with respect to an axis on a plane. They represent
particular by defining the default state. Then, we will discuss
specific and simple cases of transformation (a space rotation,
the appearance of new observables, which is related to the
for example) preserving the properties of the geometric
concept of emergence, and how these concepts relate to
structure under examination.
symmetries and conservation principles. We will then posit
that biology is characterized by a relentless breaking of
symmetries, and propose a theoretical bridge between phys-
ics and biology which we are calling extended criticality. 3.3 Emergence in physics and in biology
Extended criticality provides the theoretical frame for a non-
conservative default state. The latter entails an intrinsic The construction of theories in physics relies on mathemat-
source of variation, which is a necessary concept in evolu- ical symmetries. There is no obvious continuity between
tionary biology together with a principle of non-identical theories when they are based on different symmetries. For
iteration of morphogenetic processes. example, in classical mechanics there is no time arrow (phe-
nomena are reversible) whereas in thermodynamics, like in
3
However, applying existing physical theories directly to biology biology, time is oriented (phenomena are irreversible)6
might be misleading. For example, in numerous studies the use of the (table 1).
concepts of energy and temperature stems from a fruitful mathematical
analogy with statistical mechanics. However, the analog of temperature
4
and energy in models of interacting birds in a flock is not and cannot be Inertia was first introduced by Kepler, but with a different meaning:
expressed in the proper physical units (Joules for energy and Kelvin for the default state was rest (quiescence), and thus, it could not explain
temperature); as a result, these applications lack a theoretical justifica- why a planet keeps orbit without being pushed by some agency.
5
tion for this energy to be conservative which is a crucial invariant In the context of Galilean relativity, speeds are arbitrary or relative,
property for statistical physics. A specific example of why any scientific and entail a change in position; therefore, position cannot be conserved.
theory developed within a specific domain cannot be directly applied to The speed of an isolated object is conserved.
6
another domain is given in Longo et al. (2012a) about the application of For example, irreversibility means that we can tell when a movie is
the concept of information in biology. being played forward or backward.

J. Biosci.
J. Biosci.
Table 1. From physics to biology: A comparison of fundamental principles for theory construction

Physical world

Far from-equilibrium/self-organization
Linear/equilibrium physics Classical thermodynamics physics Biological world

Time No arrow of time Arrow of time Arrow of time Arrow of time (adds a
biological level of
irreversibility)

Conservation Conservation principles Conservation and introduction of a Conservation and a basic non-conservation Non-conservation principle,
(energy, momentum, etc) non-conservation principle principle associated to randomness in the new possibilities
(production of entropy) self-organization process

Description space Stable Microscopic: stable. Microscopic: stable. Not stable over time
Macroscopic: shrinks over time. Macroscopic: increases over time (emergence) *
(emergence, yet causally reducible).

Mathematical symmetries Stable symmetries. The system gets more symmetric over Simple symmetry breaking (the system Ubiquitous symmetry changes.
Giuseppe Longo et al.

time (measured by entropy increase). becomes less symmetric on the basis of

former symmetries).

Framing principle Conservation of energy. Increase of entropy. Identical iterations (at the statistical level). Non identical iteration.
Historicity No (past and future are No (peculiar features from the past No (a few features are akin to historicity but Fully historical systems
equivalent). are destroyed by the dynamics). the framework is ahistorical and the objects (objects are historical and
are spontaneous). not spontaneous).

Default state Uniform rectilinear movement. Stationary state with maximal entropy Stationary state under constant flows Proliferation with variation
(equilibrium). (non-equilibrium). and motility.

*A new circumstance enables new outcomes, which are not predicted (anticipated) by the description of the system (Longo and Montévill 2013)
 Theory of Organisms

Emergence is the appearance of a new observable that concept (Bak and Sneppen 1993; Kauffman 1993). More-
cannot be derived from the root theory. The transition from over, along critical transitions, the ‘local structure’ is corre-
water molecules to a liquid is a relevant example of emergence lated to the ‘global’ one, a phenomenon which echoes the
of new properties, such as fluidity and incompressibility. coherence of organisms. Typically, at critical transitions,
These novelties force a radical change of mathematical sym- correlation lengths9 tend to infinity and the very object of
metries and thus of theories from QM to hydrodynamics (HD). investigation changes, say, from vapour to a snowflake. This
Similarly, at its inception, TD disregarded gas particles and transition happens at a point which marks the passage from
their classical trajectories, and focused instead on macroscopic one object to another and from one symmetry to another.
observables, such as temperature, volume and pressure. Later Within a given theoretical framework, renormalization
on, Boltzmann unified TD to the Newton–Laplace theory of methods allow the representation of the critical transition to
trajectories of gas particles by adding the hypothesis of ‘mo- a ‘new object’ by a class of models parameterized by the
lecular chaos’, a strong, limit property, based on a fundamental scale 10 (table 1). In biology, various phenomena are
symmetry.7 As a result, the theoretical symmetries were totally analysed explicitly as critical transitions in the physical
changed. For example, time becomes irreversible (table 1). In sense, such as the formation of a coherent depolarization of
both cases a new unifying theory, including theoretical sym- thousands of mitochondria inside cardiomyocytes (Aon et al.
metry changes, is needed. Recapitulating, in physics a theory is 2004), the critical dynamics of the transcriptome (Nykter
ruled by its mathematical symmetries, which in turn determine et al. 2008) and the coherent activities of neurons (Werner
its proper observables (table 1). Conversely, the appearance of 2007) (see Longo and Montévil (2014) and Mora and Bialek
new observables forces a change of theory. (2011) for additional examples).
A biological example of emergent phenomena is the From the view point of extended critical transitions an
advent of novelties such as ears and hearing which evolved organism is understood as being in a permanent transition
from the double jaw bones of some vertebrates. These evo- with all the main signatures of criticality, such as changes of
lutionary processes imply a change of observables and, thus, symmetries, the constant reconstruction of correlation
of theoretical symmetries (Longo and Montévil 2014). In the lengths and the formation of a new global structure (Bailly
previous examples from physics, changes of observables and and Longo 2011; Longo and Montévil 2014) (table 1). Thus,
symmetries required a change of theory. In contrast, Darwin phylogenetic and ontogenetic trajectories involve ‘cascades
proposed a global theory of evolutionary dynamics, one with of symmetry changes’ which contribute to the historicity of
little mathematics and with no explicit theoretical symme- phylogeny and ontogeny by generating biological variability
tries, yet based on principles that allow us to understand and anatomical and functional diversity (Longo et al. 2012b;
changes like the generation of new organs and functions8 Longo and Montévil 2014). This is the first of the points we
and of new phenotypes. Implicitly, Darwin’s is a theory wish to make regarding the unity of these two theories
about symmetry changes; that is, a theory of the changes of addressing life at different time-scales.
biological organization along phylogenesis. Symmetry The theory of critical phase transitions is relevant to our
changes are also relevant for ontogenesis, as exemplified proposal because this theory is concerned with the formation
from the generation of a metazoa from a single cell. We thus of new objects and symmetries at and beyond the critical
propose principles for a theory of organisms in which sym- transition point (figure 1). Here emergence of a new object is
metry changes are incessant, and occur in an ‘extended mathematically treated as a point-wise transition at the crit-
critical interval’ (Bailly and Longo 2011; Longo and ical point (Longo and Montévil 2014). Extended critical
Montévil 2014) (see below). Unification between ontogene- transitions, instead, span a non-trivial interval such as an
sis, and phylogenesis will require a closer examination of organism’s lifetime. In this context, an organism continually
criticality. undergoes critical transitions, whereby both the objects and
the symmetries change. The organism and its components
are permanently reconstructed with variations.11 This repre-
3.4 From criticality to extended criticality
sents a sharp departure from physics where a radical change
of symmetry implies a change of theory. So far, only a single
In the 1980s, theoretical biologists became interested in the
physics of criticality where symmetry breaking is a central
9
Correlation length: distance at which different parts still influence
each other.
7 10
More precisely, this unification uses the notion of ergodicity. The The renormalization methods provide refined mathematical tech-
ergodic hypothesis states that a particle spends time in regions of the niques to deal with the situation at a symmetry change (Lesne 1998).
11
phase space of same energy proportionally to their volume, which is a This situation is very different from the physical notion which in-
new symmetry of the phase-space. volves a single point and a single symmetry change. At the mathemat-
8
We use a broad definition of function: the physiologic activity of an ical limit, this may be viewed as a dense set of critical transitions in the
organ or part. intended interval (Longo and Montévil 2014)

J. Biosci.
 Giuseppe Longo et al.

Figure 1. This figure represents a simulation of a phase transition, which is a simple model of magnets: the Ising model. The system is a
grid where each location can be either 1 (white) or −1 (black). At high temperatures, the system is disordered, and is a uniform random mix
of black and white; as such its macroscopic description does not change if we swap black and white: it is symmetric and the global
magnetization is 0. At low temperatures, the system is magnetized, and is either dominated by black or white; the magnetization has become
a new relevant global observable and the symmetry is broken. At the critical point, the point of transition between these two states, the
system has a global behaviour with patterns at all scales.

transition like the one at the critical point can be accommo- of stationary systems than those equations of equilibrium
dated into a single theory. In summary, the passage from (Cottrell 1979; Nicolis and Prigogine 1977). In this instance,
theories of the inert to a theory of organisms must accom- physics approximates biology because there is no such thing
modate continual symmetry changes within one theoretical as an organism at thermodynamic equilibrium. Organisms
frame. Extended criticality is an attempt in this direction. use flows but are not a consequence of flows (table 1). Next,
we further elaborate on this distinction.
Far-from-equilibrium physical systems are understood by the
3.5 Physical systems at equilibrium, far-from-equilibrium analysis of their instantaneous flows. Indeed, the shape of a
and organisms flame can be calculated from the flows of energy and matter that
go through it, whereas the shape of an organism cannot. Far-
Physical systems at equilibrium are fully described by con- from-equilibrium systems are ahistorical13 because they appear
servation properties and the associated equational symme- spontaneously and can be analysed independently. In contrast,
tries. In contrast, far-from-equilibrium dissipative physical organisms are not spontaneous but historical, meaning that they
systems like flames, micelles and Bénard cells12 are an are a consequence of the reproductive activity of a pre-existing
organization of flows of energy and/or matter. In a mathe- organism. Organization cannot be deduced from flows operating
matical sense, flows and boundary conditions fully deter- within and upon organisms; instead, understanding biological
mine and thus causally trigger and maintain the structure of organization requires a historical analysis. This perspective rein-
any such system (Nicolis and Prigogine 1977). Within this forces Dobzhansky’s assertion that ‘nothing in biology makes
theory, the quantity which allows the mathematical analysis sense except in the light of evolution’ given that evolution is
of the system is entropy production, that is, the energy precisely historical. Historicity is thus the second point we stress
dispersal rate instead of energy conservation. This analysis in this quest for unification between the theories of evolution and
is based on a non-conservative quantity, i.e. entropy increase of organisms.
associated to time irreversibility. When the flows of energy, In an organism, each cell division changes local symme-
matter and entropy are constant, these systems are called tries because each of those divisions forces new local and
stationary. However, not all the mathematical symmetries potentially global correlations. These changes yield variabil-
of equilibrium are still applicable. Instead, the balance equa- ity and adaptability to organisms. In the context of evolution,
tions of the flows provide more suitable tools for the analysis the advent of new functions and organs are additional

12 13
Bénard cells are one of the simplest types of self-organization of Because of their ahistoricity, self-organized far-from-equilibrium
matter generated by natural convection: regular patterns of liquid move- physical systems are used as a paradigm for the origin of life, a subject
ment occur in a layer of fluid heated from below. that is not addressed herein.

J. Biosci.
 Theory of Organisms

examples of symmetry changes. In our approach, reproduc- Newton’s laws for understanding mechanics. It should be
tion with modification involves ‘symmetry changes’ and remembered that Newton posed his laws ‘axiomatically,’ as
may be viewed as multiple critical transitions which span core principles of his theory.14
the irreversible time of both phylogenesis and ontogenesis.
This is the third common point towards the unification of the 4.1.1 Proliferation with variation: Darwin explicitly stated
theories addressing these two different time-scales. ‘…There is no exception to the rule that every organic being
naturally increases at so high a rate, that, if not destroyed, the
4. Foundations for a Theory of Organisms earth would soon be covered by the progeny of a single pair’
(Darwin 1859). Reproduction obligatorily involves
We propose two founding principles: (1) the default ‘modification’ (descent with modification, in Darwin’s
state of cells, meaning proliferation with variation and words). Reproduction with variation is intrinsic to
motility, and (2) the framing principle of non-identical organisms regardless of whether they are unicellular or
iterations of a morphogenetic process. These principles multicellular (Sonnenschein and Soto 1999; Soto and
take place in the context of extended critical transitions. Sonnenschein 2011). Darwin’s narrative implies that
reproduction with variation is a default state and he
4.1 The default state in biology: A nexus between the describes it as a limit case. This is the third, and probably
theories of evolution and of organisms most important, point in common between the theories of
evolution and organisms.
The default state of proliferation applies to the first com-
In order to provide a theoretical transition between physics
mon ancestor, i.e. the cell from which all organisms arose
and biology, we will define a default state which is a limit
(figure 2). In fact, microbiologists consider axiomatic that
case. Inertial movement as uniform rectilinear movement is a
proliferation is the default state of prokaryotes and unicellu-
limit state and physicists made all physical movements in-
lar eukaryotes (Luria 1975). On the contrary, despite lacking
telligible as departures from it. By describing this default
evidence, there had been a consensus among biologists that
state, Galileo could focus on the analysis of the forces
consider quiescence as the default state of cells in multicel-
constraining it, such as gravity and friction.
lular organisms (Alberts et al. 2008; Alberts 2010). We
The ‘unconstrained’ condition proper to the biological
posited, instead, that proliferation was retained as the default
default state requires adequate physical conditions, such as
state with the advent of metaphyta and metazoa. This con-
specific intervals of temperature, pressure and pH. Sufficient
clusion is supported by the conservation of the cell cycle
nutrients provide a flow of energy and matter canalized
components throughout eukaryotes (Sonnenschein and Soto
through metabolic processes. In these unrestrained, limit
1999) and by experimental evidence (Soto and
conditions, cells constitutively exert their default state, i.e.
Sonnenschein 1985; Sonnenschein et al. 1996;
proliferation with variation plus motility (table 1). The de-
Sonnenschein and Soto 1999; Ying et al. 2008; Leitch
fault state should not be conflated with conditions necessary
et al. 2010). Additionally, the default state of proliferation
for life. For example, metabolism is necessary for cells to be
has been adopted advantageously as a fundamental principle
alive given that metabolism happens regardless of whether
in theories of carcinogenesis and of development
cells are proliferating or quiescent, moving or immobile
(Sonnenschein and Soto 1999; Soto and Sonnenschein
(‘metabolism happens, whatever happens’). In contrast, a
2010; Minelli 2011).
default state is what happens when nothing is done to the
Variation (as in proliferation with variation) should be
object or system (‘a default state happens when nothing
understood as symmetry changes; each cell division gener-
happens to prevent it’). Consequently, the nature of what
ates variations that correspond to symmetry changes associ-
can be done to the system is defined by the theory, like the
ated to critical transitions. How does intrinsic variation
concept of force in classical mechanics; according to New-
manifest itself? One obvious example is the unequal distri-
ton’s first law force modifies (accelerates) inertial move-
bution of macromolecules and organelles during cell divi-
ment. Thus, the analysis of the constraints to proliferation
sion (Huh and Paulsson 2011); another is stochastic gene
with variation and motility is fundamental to the intelligibil-
expression (Kupiec 1983; Taniguchi et al. 2010; Tyagi 2010;
ity of organismal biology. This is equivalent to the role of
Marinov et al. 2014). Additional variation is due to somatic
14
The term constraint has been used in various contexts. The concept mutations and aneuploidy, events described among cells of
used herein is (i) constraint is something that remains invariant with normal mammalian kidneys and brains (Martin et al. 1996;
respect to the duration of the process being constrained, (ii) a constraint Rehen et al. 2001).
changes the process being constrained (Montévil and Mossio 2015).
Additionally, like for the mechanical default state (inertia) a constraint
acts on the biological default state. In biological systems, constraints 4.1.2 Motility: Motility encompasses intracellular, cellular,
enable the emergence of new processes and phenomena. tissue and organismic movements (Stebbings 2001). The

J. Biosci.
 Giuseppe Longo et al.

Figure 2. The emergence of living organisms and their further evolution, as we know it, implies that the default state of the first cell (i.e.
‘THE CELL’) and those of their daughters must have been proliferation-with-variation and motility. The question mark indicates that the
process that generated THE CELL from a pre-biotic soup is unknown.

evolution of the cytoskeleton protein actin, which is believed open and close in response to light (van Doorn and van
to have been present in all early organisms, supports the Meeteren 2003), and like animal cells, can move organelles
concept that motility, like proliferation with variation, is using actin and myosin (Ueda et al. 2010).
the default state of all cells (Buss 1987; Sonnenschein and
Soto 1999). From chemotaxis to swimming, motility
immediately suggests the idea of agency.15 Agency has 4.2 Default state and constraints
dominated the debate on the differences between inert
and living matter for most of the 19th century. Motility During early development, increases in the size of an organ-
suggests a comparison with classical physics because both ism occur by generation of new cells and production of
involve trajectories in space. In physics, an external force extracellular matrix. Cell proliferation is the result of the
is required to obtain a change of inertial movement, default state and is constrained. Organ shape is a result of
interpreted as a conservation property grounded on the cells’ default state plus physical constraints, like gravity,
theoretical symmetries. In contrast, a cell or an organism and those ones created by the living organism, like muscular
will spontaneously move by using forces and flows of tension.
energy and matter. Cells do not require external stimuli Inert matter requires causes to change states or properties.
to move. Movement along a direction represents a The causal structure of a physical process is determined
symmetry change, and corresponds to a non-conservation mathematically by a set of equations justified by symmetries,
property as it is not inertial. Moreover, these actions are such as equilibria. For example, Newton’s equations relate
accompanied by critical transitions ( Werner 2007; forces to their effects, i.e. as causes of acceleration at equi-
Cardamone et al. 2011). librium. In physics, causes and constraints on the default
Motility should not exclude movements other than loco- state are synonymous. Organisms, instead, are agentive and
motion. Plants may be attached to the ground by their roots, thus capable of initiating activity by themselves. The default
but they can generate movement of their parts, for example state is a cause in biology; by contrast, anything that affects
by growing towards a source of light. Flowers and leaves the default state is a constraint. For example, gravity be-
comes a constraint for evolution and embryogenesis, and not
a cause of biological dynamics. Gravity influences numerous
15
Agency is the ability to act, that is, to initiate an action. if not all biological processes and it is a main determinant of

J. Biosci.
 Theory of Organisms

morphogenesis as exemplified in the formation of the antero- scale unicellular ancestors of cells of multicellular organisms
posterior axis in chick embryos. However, the formation of are assumed to have had proliferation as their default state.
this axis is not a consequence of gravity, but of the activity of Thus, it is the shift from proliferation with variation and
cells under the constraint of gravity: in our theoretical frame motility to quiescence, which should be explained, instead of
and language, gravity is a constraint and not a cause. In sum, being assumed. Cell proliferation is achieved by the execu-
biological dynamics is grounded on the default state, that is, tion of a cell cycle process, which does not stop until two
proliferation with variation and motility. quite similar but non-identical cells are formed. The cell
Constraints narrow down the range of ‘possibles’. For cycle is a representation of enzymatic reactions and physical
example, fibroblasts removed from subcutaneous tissues of processes leading to the duplication of the cell components
an animal have the same size when dissociated and including DNA, and the faithful separation of two daughter
explanted into a culture dish. Soon after, their size varies cells. Textbooks use the metaphor of the cycle operating like
over a wide range ( Rubin and Hatie 1968; Rubin 1988). a dishwasher performing a series of stereotyped tasks
Constraints also enable other ‘possibles.’ The bottom of a (Alberts et al. 1994). If quiescence were its default state, it
cell culture dish prevents displacement in that direction would be difficult to activate the ‘cell cycle machine’ by
reducing the possible movements, but at the same time this organizing the cell cycle components which entail a very
constraint enables cells to crawl along a plastic surface. In a complex network of constraints. Instead, when proliferation
multicellular organism, other constraints to the default state is the default state, it becomes easier to prevent complex
are relevant. Mechanical constraints limit proliferation, var- machinery from functioning: a simple mechanical or chem-
iation and motility, like when cells compress each other; they ical constraint may suffice. For example, a switch will do it
adhere to other cells, and stretch each other. Fibers in the for an engine and an inhibitor will do it for a cell. To sum the
extracellular matrix apply tensile forces on structures situation up, the difference between proliferation or quies-
allowing movement in certain directions and constraining it cence as a default state corresponds to the difference be-
in others (Barnes et al. 2014). Sequential smooth muscle tween preventing something complex from happening and
differentiation exerts compressive stress on the endoderm causing it to happen. The latter is much harder to conceive,
and mesenchyme of the small intestine, causing buckling explain and realize.
and folding which leads to the formation of villi (Shyer
et al. 2013). Muscle contraction shapes the bones to which 4.3 The framing principle of biological morphogenesis
they are attached by inducing tissue accrual on the side of
tension (Muller 2003; Rot-Nikcevic et al. 2007), and chem-
Generating phenotypes from a single cell, be it LUCA (Last
ical interactions may also constrain the default state, as when
Common Universal Ancestor) (Steele and Penny 2010) or a
serum albumin restricts proliferation of estrogen sensitive
zygote, is an essential component of phylogenesis and onto-
cells, and estrogen cancels the action of this constraint
genesis. Organisms, be they unicellular or multicellular, are a
(Sonnenschein et al. 1996).
consequence of the inherent variability generated by prolif-
We posit that organisms do not have stable symmetries that
eration, motility (Sonnenschein and Soto 1999) and self-
would allow us to spell out the actual phylogenetic and onto- organization (Mossio and Moreno 2010), all of which oper-
genetic trajectories that they follow because such trajectories ate within the framing principle we propose: life phenomena
depend on the history and random changes of symmetries of are non-identical iterations of a morphogenetic process by
the objects considered, organisms in this case (Longo and which organization is iterated and maintained (table 1). For
Montévil 2014). The default state of proliferation with varia- example, branching morphogenesis is a ubiquitous process
tion corresponds to Darwin’s key idea of evolution being that generates a repetitive, yet non-identical pattern whereby
‘descent with modification’, on which selection operates. length of branches and branching angles vary (figure 3).
In contrast to physics where conservation principles frame The framing principle cannot be derived from the default
the theories, in biology, the default state of proliferation with state alone. It may be instantiated as autopoiesis16 or more
variation is a non-conservation principle (Longo et al. generally, closure to which variation is added, given that
2012b; Longo and Montévil 2013). This rationale does not living autopoietic processes require permanent changes in
conflict with physical principles as it concerns new observ- their constructing and reconstructing activities (Montévil and
ables, i.e. phenotypes. In fact, proper principles and observ- Mossio 2015). Far-from-equilibrium physical processes have
ables are being added at different and interacting levels of autopoietic characteristics by iterating shapes and physical
biological organization and determination. Thus, the theory
16
of organisms that we envisage becomes a compatible exten- Autopoietic organization is defined as a unity by a network of
productions of components which (i) participate recursively in the same
sion of physical theories.
network of productions of components which produced these compo-
Finally, the recent phylogenetic history of every cell is the nents, and (ii) realize the network of productions as a unity in the space
proliferation of its parental cell. On a far longer temporal in which the components exist (Varela et al. 1974).

J. Biosci.
 Giuseppe Longo et al.

Figure 3. Ductal development in whole mounts of mammary glands from C57Bl6 mice illustrates the principle of never identical
iterations on the branching pattern of a ductal system. In Panel A, the arrows point to two mammary gland epithelial structures (buds) at day
15 of embryonic life (E15); the ductal tree emerges from the growth of these buds. Panels B to D show the ductal tree at E19. Arrows
indicate the origin of the ductal tree. Scale bars indicate 100 μm. Panels E and F show mammary trees at post-natal day 21. Scale bars
indicate 500 μm.

structures along optimal trajectories. Flames, micelles and years while evolution generated diverse living forms from
Bénard cells remained unchanged over the last 4 billion LUCA up to the reader of this page. In our approach, we

J. Biosci.
 Theory of Organisms

expand the notion of autopoiesis by including in it the perform the functions of the cells within the normal tissue of
concept of variability. The latter is expressed as theoretical origin. This combination of higher complexity and lower
symmetry changes. By these means we go beyond properties organization represents a true hallmark of cancer.
that could be purely physical since autopoiesis would be
enriched by biological variability and historicity, both asso-
5. Implications for biological research
ciated with symmetry changes.
What are the benefits of adopting the principles for a
4.4 Complexity versus organization theory of organisms delineated above? First, the pro-
posed principles would help to move away from opera-
Oftentimes, organization17 and complexity are considered tional definitions. From the notion of gene to that of
synonyms. However, while complexity is mostly meant to growth factor, the use of operational definitions has
span the material world, including inert and living matter, resulted in contradictions and ambiguities that hinder
organization should be viewed as an exclusive attribute of the establishment of general and stable concepts (Moss
life and of machines invented by humans. Complexity and 2003; Sonnenschein and Soto 1999). Additionally, the
organization ought to be explicitly defined and distin- proposed principles enable alternative explanations to the
guished. We posit that phenotypic complexity is a quantifi- mechanistic ones inherent to the molecular biology rev-
able characteristic of static structures as exemplified by the olution. The latter do not represent an explicit theoretical
anatomy of an organism. Complexity can be quantified by frame, but mostly refer metaphorically to common sense
enumerating its components such as the number of cell notions, such as ‘information’ and ‘program’. The use of
types, tissues, organs, organ-connected components, connec- these metaphors forces explanations to be molecular and
tions and nodes within networks, fractal dimensions of cells to follow predictable causal chains (Longo et al. 2012a).
and organs, etc. (Bailly and Longo 2009; Bizzarri et al. Instead, by insisting on the search of constraints to the
2011; Longo and Montévil 2014). From the perspective of default state, multilevel biomechanical and bioelectrical
the framing principle, instead, organization refers to a dy- explanations become as legitimate as molecular ones.
namic state of interdependence of levels that includes both Second, our principles radically change both observ-
structures and functions as well as integration and regulation. ables and determination vis-à-vis the theoretical frames
Organization cannot happen without complexity, but orga- proposed by physical theories. Such a change enables us
nization is not reducible to complexity. to anchor reasoning and modelization on robust biolog-
Carcinogenesis illustrates how complexity and organiza- ical principles. Indeed, as implied by Turing, there is an
tion are not equivalent. For instance, in the mammary gland, epistemological gap between modelization and imitation
precancerous lesions like ductal carcinoma in situ are more (Turing 1950, 1952). While the former is based on a
complex than normal ducts. This lesion appears as a partial theory concerning the object of knowledge, the latter is
occlusion of ducts which results in more than a single lumen. not. For example, individuation becomes the result of
Multiple lumena indicate higher measurable topological non-identical morphogenetic iterations. This principle ex-
complexity (figure 4). Moreover, the epithelium-stroma in- cludes a Platonic conception of ideal or perfect organs
terface has a higher fractal dimension than that of their or structures which would be determined as an optimal
normal counterparts (Bizzarri et al. 2011).18 However, a solution of an equation.
cancerous tissue is less organized in the sense that it does Third, the principles we propose enable the construction
not adequately perform the function of the normal tissue of and discussion of mathematical models on the bases of
origin. For example, a blocked duct interferes with excretory biologically relevant assumptions. For example, in ecology
function. Additionally, cells within cancer tissues may not the commonly used equations addressing population size are
not theory-based. Taking into consideration the default state
and the notion of constraint it is feasible to obtain theoreti-
17
The term organization appeared in the early 15th century, in both cally meaningful equations in which the food term is not a
English and French; it represents a conflation of ‘to construct’ and cause per se: this constraint becomes relevant only when
‘organ,’ the musical instrument. Organization became associated with
the meaning of organized living beings in the 18th century. In our there is scarcity of food. In an ecological context there are
narrative, organization is compatible with the notion of ‘organizational numerous parameters that enable the population to grow in
closure’. The latter is a ‘distinct level of causation, operating in addition number, however, those parameters actually play a role on
to physical laws, generated by the action of material structures acting as growth (i.e. the growth rate) only when they are limiting the
constraints’ (Mossio and Moreno 2010). increase of the population. Food superabundance will not
18
‘Cells within a cancer’ is not synonymous with ‘cancer cells’. When
considering that cancer is a tissue-based disease the phenotype of make the population grow faster.
individual cells within the cancer tissue is determined by the tissue. Fourth, we are proposing an alternative to the use of the
For a detailed explanation, see (Soto and Sonnenschein 2011) notions of program, information, and signal specificity in

J. Biosci.
 Giuseppe Longo et al.

Figure 4. Complexity versus organization. Panel A shows cross-sections of normal mammary gland ducts. Panel B shows a cross-section
of a ductal carcinoma in situ. Note that while normal ducts have a single lumen, carcinoma in situ has multiple ones, thus showing higher
topological complexity than normal ducts. Panel C shows various cross-sections of normal ducts, and Panel D shows a section of a large
carcinoma tumour. The arrow points to the area that is magnified in the inset. The tumour contains many structures similar to the carcinomas
in situ.

biology. In particular, biological variation is not noise as in theory of organisms while providing founding principles
those information-based theories (Huang 2009). Rather than adequate for framing experiments and mathematical
as a priori determination as presupposed by the notion of modelling.
developmental program, the two proposed principles lead to
a conception of biological specificity that is defined directly
Acknowledgements
with respect to the trajectory of organisms and their cells in
time and space as a cascade of symmetry changes. This work was conducted as part of the research project
‘Addressing biological organization in the post-genomic
6. Conclusions era’, which is supported by the International Blaise Pascal
Chairs, Region Ile de France. AMS is the incumbent Blaise
Pascal Chair of Biology at the École Normale Supérieure,
Research on organismal biology is being conducted in the
and GL is the Chair’s host. Additional support was provided
absence of a global theory. Instead, its conceptual frame-
by grants from the Avon Foundation, the Italian Space
work is loosely based on the mathematical theories of infor- Agency and the National Institute of Environmental Health
mation and on operational definitions. This combination of Sciences of the NIH (ES08314 and U01-ES020888). The
notions has generated contradictions and hindered progress. content does not necessarily represent the official views of
In spite of a few successful outcomes, the application of the funding agencies including the National Institute of
physical principles without proper analysis of the differences Environmental Health Sciences or the National Institutes of
between biological and physical situations has also contrib- Health. MM was supported by a grant from the Région Île-
uted to the current crisis. Our analysis of the differences de-France, DIM ‘Problématiques transversales aux Systèmes
between the physics of inanimate and living matter provides Complexes’. CS is a Fellow of the Institut d’ Études
a sound perspective for the construction of a much needed Avancées de Nantes. GL’s work is part of the project ‘Les

J. Biosci.
 Theory of Organisms

lois des Dieux, des hommes et de la nature’ in that Institute. Kupiec JJ 2010 On the lack of specificity of proteins and its
The authors are grateful to Paul-Antoine Miquel, Matteo consequences for a theory of biological organization. Prog.
Mossio, Andrew Moore, Cheryl Schaeberle and Michael Biophys. Mol. Biol. 102 45–52
Sweeney for their critical input. The authors have no com- Leitch HG, Blair K, Mansfield W, Ayetey H, Humphreys P,
Nichols J, Surani MA and Smith A 2010 Embryonic germ cells
peting financial interests to declare.
from mice and rats exhibit properties consistent with a generic
pluripotent ground state. Development 137 2279–2287
Lenoir T 1982 The strategy of life: teleology and mechanics in
References nineteenth-century biology (Dordrecht, Holland: D. Reidel
Publishing)
Lesne A 1998 Renormalization methods: critical phenomena, cha-
Alberts B 2010 Model organisms and human health. Science 330 1724
os, fractal structures (New York: Wiley)
Alberts B, Bray D, Lewis JG, Raff M, Roberts K and Watson JD
Longo G and Montévil M 2012 The inert vs. the living state of
1994 Molecular biology of the cell (New York, NY: Garland
matter: extended criticality, time geometry, anti-entropy - an
Publishing Inc.)
overview. Front. Physiol. 3 39
Alberts B, Johnson A, Lewis J, Raff M, Roberts K and Walter
Longo G and Montévil M 2013 Extended criticality, phase spaces
P 2008 Molecular biology of the cell (London: Garland
and enablement in biology. Chaos, Solitons Fractals 55 64–79
Science)
Longo G and Montévil M 2014 Perspectives on organisms: bio-
Aon MA, Cortassa S and O’Rourke B 2004 Percolation and criti-
logical time, symmetries and singularities (Berlin: Springer)
cality in a mitochondrial network. Proc. Natl. Acad. Sci. USA
Longo G, Miquel P-A, Sonnenschein C and Soto AM 2012a Is
101 4447–4452
information a proper observable for biological organization?
Bailly F and Longo G 2009 Biological organization and anti-entropy.
Prog. Biophys. Mol. Biol. 109 108–114
J. Biol. Syst. 17 63–96
Longo G, Montévil M, and Kauffman S 2012b No entailing laws,
Bailly F and Longo G 2011 Mathematics and natural sciences: the
physical singularity of life (London: Imperial College Press) but enablement in the evolution of the biosphere; in Genetic and
Bak P and Sneppen K 1993 Punctuated equilibrium and criticality evolutionary computation conference (Philadelphia: ACM)
in a simple model of evolution. Phys. Rev. Lett. 71 4083–4086 Luria SE 1975 36 lectures in biology (Cambridge: MIT Press)
Barnes C, Speroni L, Quinn K, Montevil M, Saetzler K, Bode- Marinov GK, Williams BA, McCue K, Schroth GP, Gertz J, Myers
Animashaun G, McKerr G, Georgakoudi I, et al. 2014 From RM and Wold BJ 2014 From single-cell to cell-pool
single cells to tissues: interactions between the matrix and hu- transcriptomes: stochasticity in gene expression and RNA splic-
man breast cells in real time. PLoS One 9 e93325 ing. Genome Res. 24 496–510
Bizzarri M, Giuliani A, Cucina A, D’Anselmi F, Soto AM and Martin GM, Ogburn CE, Colgin LM, Gown AM, Edland SD and
Sonnenschein C 2011 Fractal analysis in a systems biology Monnat RJ Jr 1996 Somatic mutations are frequent and increase
approach to cancer. Semin. Cancer Biol. 21 175–182 with age in human kidney epithelial cells. Hum. Mol. Genet. 5
Buss LW 1987 The evolution of individuality (Princeton, NJ: 215–221
Princeton University Press) Minelli A 2011 Development, an open-ended segment of life. Biol.
Cardamone L, Laio A, Torre V, Shahapure R and DeSimone A Theor. 6 4–15
2011 Cytoskeletal actin networks in motile cells are critically Montévil M and Mossio M 2015 Biological organisation as closure
self-organized systems synchronized by mechanical interactions. of constraints. J. Theor. Biol. 372 179–191
Proc. Natl. Acad. Sci. USA 108 13978–13983 Moore A 2012 Life defined. BioEssays 34 253–254
Cottrell A 1979 The natural philosophy of engines. Contemp. Phys. Mora T and Bialek W 2011 Are biological systems poised at
20 1–10 criticality? J. Stat. Phys. 144 268–302
Darwin C 1859 On the origin of species (London: Clowes and Moss L 2003 What genes can’t do (Cambridge, MA: MIT Press)
Sons) Mossio M and Moreno A 2010 Organisational closure in biological
Davies J A 2013 Mechanisms of morphogenesis (Elsevier Academic Press) organisms. Hist. Philos. Life Sci. 32 269–288
Deacon T 2012 Incomplete nature: how mind emerged from matter Muller GB 2003 Embryonic motility: environmental influences and
(New York: Norton) evolutionary innovation. Evol. Dev. 5 56–60
Elsasser WM 1987 Reflections on a theory of organisms (Quebec: Newman SA 2012 Physico-genetic determinants in the evolution of
Orbis Publishing) development. Science 338 217–219
Huang S 2009 Non-genetic heterogeneity of cells in development: Nicolis G and Prigogine I 1977 Self-organization in non-
more than just noise. Development 136 3853–3862 equilibrium systems (New York: Wiley)
Huh D and Paulsson J 2011 Non-genetic heterogeneity from sto- Nykter M, Price ND, Aldana M, Ramsey SA, Kauffman SA, Hood
chastic partitioning at cell division. Nat. Genet. 43 95–100 LE, Yli-Harja O and Scmulevich I 2008 Gene expression dy-
Jacob F 1974 The logic of life (New York, NY: Pantheon Books) namics in the macrophage exhibit criticality. Proc. Natl. Acad.
Kauffman SA 1993 The origins of order (Oxford: Oxford Univer- Sci. USA 105 1897–1900
sity Press) O’Malley MA and Dupre J 2005 Fundamental issues in systems
Kupiec JJ 1983 A probabilistic theory for cell differentiation, em- biology. BioEssays 27 1270–1276
bryonic mortality and DNA C-value paradox. Specul. Sci. Rehen SK, McConnell MJ, Kaushal D, Kingsbury MA, Yang AH
Technol. 6 471–478 and Chun J 2001 Chromosomal variation in neurons of the

J. Biosci.
 Giuseppe Longo et al.

developing and adult mammalian nervous system. Proc. Natl. Soto AM and Sonnenschein C 2012 Is systems biology a promising
Acad. Sci. USA 98 13361–13366 approach to resolve controversies in cancer research? Cancer
Rot-Nikcevic I, Downing KJ, Hall BK and Kablar B 2007 Devel- Cell Int. 12 12
opment of the mouse mandibles and clavicles in the absence of Stebbings H 2001 Cell motility. Encyclopedia of life sciences (Na-
skeletal myogenesis. Histol. Histopathol. 22 51–60 ture Publishing Group) (www.els.net)
Rubin H 1988 Can we speak of a mechanism for malignant trans- Steele M and Penny D 2010 Origins of life: common ancestry put to
formation?’; in Theories of carcinogenesis: facts, fashion or the test. Nature 465 168–169
fiction? (Washington, DC: Taylor and Francis) pp 31–43 Taniguchi Y, Choi PJ, Li GW, Chen H, Babu M, Hearn J, Emili A
Rubin H and Hatie C 1968 Increase in the size of chick embryo and Xie XS 2010 Quantifying E. coli proteome and tran-
cells upon cultivation in serum-containing medium. Dev. Biol. scriptome with single-molecule sensitivity in single cells. Science
17 603–616 329 533–538
Saçlioglu C, Pekcan Ö and Nanjundiah V 2014 Group behaviour in Turing AM 1950 Computing machinery and intelligence. Mind 59
physical, chemical and biological systems. J. Biosci. 39 177–189 433–460
Salam A 1990 Unification of fundamental forces (Cambridge: Turing AM 1952 The chemical basis of morphogenesis. Philos.
Cambridge University Press) Trans. R. Soc. Lond. [Biol]. 237 37–72
Shyer AE, Tallinen T, Nerukar NL, Wei Z, Gil ES, Kaplan DL, Tyagi S 2010 Genomics. E. coli, what a noisy bug. Science 329
Tabin CJ and Mahadevan L 2013 Villification: how the gut gets 518–519
its villi. Science 342 212–218 Ueda H, Yokota E, Kutsuna N, Shimada T, Tamura K, Shimmen T,
Sonnenschein C and Soto AM 1999 The society of cells: cancer Hasezawa S, Dolja W, et al. 2010 Myosin-dependent endoplas-
and control of cell proliferation (New York: Springer Verlag) mic reticulum motility and F-actin organization in plant cells.
Sonnenschein C, Soto AM and Michaelson CL 1996 Human serum Proc. Natl. Acad. Sci. USA 107 6894–6899
albumin shares the properties of estrocolyone-I, the inhibitor of van Doorn WG and van Meeteren U 2003 Flower opening and
the proliferation of estrogen-target cells. J. Steroid Biochem. closure: a review. J. Exp. Bot. 54 1801–1812
Mol. Biol. 59 147–154 van Fraassen BC 1989 Laws and symmetry (Oxford: Oxford Uni-
Soto AM and Sonnenschein C 1985 The role of estrogens on the versity Press)
proliferation of human breast tumor cells (MCF-7). J. Steroid Varela FG, Maturana HR and Uribe R 1974 Autopoiesis: the
Biochem. 23 87–94 organization of living systems, its characterization and a model.
Soto AM and Sonnenschein C 2010 Environmental causes of Biosystems 5 187–196
cancer: endocrine disruptors as carcinogens. Nat. Rev. Werner G 2007 Metastability, criticality and phase transitions in
Endocrinol. 6 363–370 brain and its models. Biosystems 90 496–508
Soto AM and Sonnenschein C 2011 The tissue organization field Ying QL, Wray J, Nichols J, Batlle-Morera L, Doble B, Woodgett
theory of cancer: a testable replacement for the somatic mutation J, Cohen P and Smith A 2008 The ground state of embryonic
theory. BioEssays 33 332–340 stem cell self-renewal. Nature 453 519–523

MS accepted 20 April 2015; accepted 17 August 2015

Corresponding editor: SCOTT F GILBERT

J. Biosci.