The Laws of Complexity and
Self-organization:
A Framework for Understanding
Neoplasia
Nat Pernick, M.D.
PathologyOutlines.com, Inc.
Bingham Farms, Michigan (USA)
Posted online: 6 December 2011, revised 7 December 2011
Email: NatPernick@hotmail.com
Abstract
Background: Current biologic research is based
on a reductionist approach. Complex
systems are broken down into combinations of simpler systems or parts, which
can then be studied more readily. Although
this approach is rational, it has failed to bring about the understanding
necessary to substantially reduce cancer-related deaths. Complexity theory suggests that emergent
properties, based on interactions between the parts, are important in
understanding fundamental features of living systems. Applying complexity theory to neoplasia may
yield a greater understanding of physiologic systems that have gone awry.
Methods and Findings: The laws of complexity and
self-organization are reviewed and summarized, and applied to neoplasia:
1. In living systems, the whole is greater than the sum of
the parts.
2. There is an inherent inability to predict the future,
particularly for living systems.
3. Life emerges when the molecular diversity of a closed
system exceeds a threshold of complexity.
4. Much of the order in organisms is due to generic
properties that emerge from a network of gene products.
5. Numerous biologic pressures push cells towards disorder.
6. Organisms resist these biologic pressures towards disorder
through multiple layers of redundant controls.
7. Neoplasia occurs as these multiple controls are breached.
The resulting neoplastic process reflects
the nature of the breaches, the individual’s germline configuration and the
network state of the cell of origin.
Conclusions: In the framework of the laws of
complexity and self-organization, cells maintain order by redundant control
features that resist the inherent biologic pressure in the cell towards
disorder. Neoplasia can be understood as
the accumulation of changes that undermine these control features, which may
lead to dysregulated growth and differentiation. Studying the neoplastic
process within this context may generate new approaches to treatment.
Introduction
On 23 December 1971, President Nixon signed the National
Cancer Act of 1971, generally viewed as beginning the “war on cancer” in the
United States (note 1).
Since then, the U.S. National Cancer Institute has spent over $100
billion on this effort (New York Times, 2009 Jun 27).
Death rates due to cancer have decreased for both sexes (J Natl Cancer Inst 2011;103:714), and the 5 year relative survival
rate for all cancers has increased from 50% in 1975-1977 to 68% in 1999-2006 (Cancer Facts and Figures 2011, p. 2). However, in
2011, there will still be an estimated 571,950 cancer deaths just in the United
States, many in otherwise healthy people (Cancer Facts and Figures 2011, p. 1).
Our current research efforts are based on a reductionist
approach to understanding biology. Under
this approach, analysis of complex systems can be reduced to studying
combinations of simpler systems, which can themselves be reduced to simpler
parts (Reductionism:
Interdisciplinary Encyclopedia of Religion and Science).
A cell can be thought of as a complex machine whose actions can be
understood by analyzing each of the components.
We study disease by finding and analyzing the defective components. For example, 80-90% of follicular lymphoma
cases have the t(14;18)(q32;q21) chromosomal translocation, which brings the
BCL2 proto-oncogene under the transcriptional influence of the immunoglobulin
heavy chain gene. This leads to
overexpression of a functionally normal bcl2 protein, which inhibits these
cells from their usual condition of undergoing apoptosis. This allows additional genetic mutations to accumulate,
which leads to neoplasia of follicular center cells in some patients (Haematologica
2008;93:1773).
The reductionist approach is logical and predictable, and
has led to improvements in survival for some malignancies. However, many neoplastic processes cannot be
described in so logical a manner, and this approach has failed to improve
survival for the major causes of cancer related death. This may be because the reductionist approach
ignores fundamental properties of living systems that are not necessarily
logical and predictable. Understanding
these properties may yield a greater understanding of physiologic systems that
have gone awry during neoplasia. To this
end, this paper summarizes the laws of complexity and self-organization, and
applies them to neoplasia (note 2).
Table 1 - The Laws of Complexity and
Self-Organization Relevant to Neoplasia
1. In living systems, the whole is greater than the
sum of the parts.
2. There is an inherent inability to predict the
future, particularly for living systems.
3. Life emerges when the molecular diversity of a
closed system exceeds a threshold of complexity.
4. Much of the order in organisms is due to generic
properties that emerge from a network of gene products.
5. Numerous biologic pressures push cells towards
disorder.
6. Organisms resist these biologic pressures towards
disorder through multiple layers of redundant controls.
7. Neoplasia occurs as these multiple controls are
breached. The resulting neoplastic
process reflects the nature of the breaches, the individual’s germline configuration
and the network state of the cell of origin.
1. In living systems, the whole is
greater than the sum of the parts.
The reductionist approach is based on the premise that the
whole is equal to the sum of the parts.
Using this approach, a large system can be analyzed by breaking it down
and studying the parts and the connections between the parts, which is assumed
to provide an understanding of the entire system (EMBO
Rep 2004;5:1016). For living organisms, we begin
by examining large structures. To gain
further understanding, we look deeper into organs, cells, biochemistry,
chemistry and physics (Reinventing
the Sacred, p.
10). As Francis Crick explained, “the
ultimate aim of the modern movement in biology is to explain all biology in
terms of physics and chemistry” (note 3).
This approach makes sense for studying complex mechanical
structures created by man, such as automobiles, which have thousands of parts
and utilize numerous engineering principles.
By using the reductionist approach, even a non-engineer can begin to
understand how an automobile works.
However, an automobile lacks features of even the simplest living
organism, such as the ability to respond to the environment, to reproduce or to
evolve.
In living systems, as discussed below, it is the
interactions between molecules that create life. Individually, the molecules can be considered
as “dead.” Collectively, these molecules
develop emergent properties, the missing features that make the whole greater
than its parts (Complexity 2002;7:18).
For example, the properties of a protein are not equivalent
to the sum of the properties of each amino acid. Proteins can catalyze a chemical reaction or
recognize an antigen not only because their amino acids are arranged in a
specific order, but because their tertiary structure and function are
additionally determined by external factors (EMBO
Rep 2008;9:10). Yet the tertiary structure and
function cannot be deduced from analysis of individual amino acids.
Emergent properties also arise in cell reproduction. Various molecules engage in linked processes that
culminate in the reproduction of a cell as an entire entity. As Kauffman notes, “it is a closure of work
tasks that propagates its own organization of processes” (Reinventing
the Sacred, p.
94). Although no laws of physics are
broken, it is not reducible to physics.
How it works is something science “hardly knows how to talk about” (Reinventing
the Sacred, p.
47). Thus, this first law of complexity
and self-organization is really a recognition that these emergent properties
exist and are not explained by the traditional reductionist approach.
2. There is an inherent inability to
predict the future, particularly for living systems.
The deterministic nature of Newton’s three laws of motion
suggested that given initial and boundary conditions, one could exactly
determine a system’s evolution and predict its future (note 4).
However, there are several reasons why we cannot predict the future,
particularly for living systems.
First, many properties of living systems are emergent, and
although they do not violate the laws of physics, they cannot be predicted by
them. As indicated above, given the
amino acid sequence of a protein, we cannot predict its tertiary
structure. One explanation for this
inability is that “this is a formidable problem given the very large number of
degrees of freedom and hence very large conformational space of a polypeptide
chain” (J
Mol Biol 2009;393:249). However, a more
basic explanation of our inability to predict is that the overall folding
mechanism of proteins is an emergent property, based on properties between the
parts (amino acids) not described by any known law of nature. Similarly, we cannot predict emergent
properties at higher levels. Even with a
complete map of its genome, we cannot determine basic features of an organism,
such as what it will look like, or the impact of a genetic change. All we can do is to try to explain, after the
fact, why things are the way they are.
Second, quantum physics dictates that there is inherent
uncertainty in all matter at the subatomic level. Even the location of an electron can only be
determined as a probability (Relational Quantum Mechanics, Stanford Encyclopedia of Philosophy).
This is reflected in the Heisenberg
Uncertainty Principle, which states that for an electron or any other
particle, one cannot simultaneously measure,
with an arbitrary degree of accuracy and certainty, both its present
position and future momentum (The Uncertainty Principle, Stanford Encyclopedia of
Philosophy). This is not a criticism of our ability to
measure accurately, but a statement about the laws of physics. Thus, living and nonliving systems have features
that cannot be predicted with precision because they are composed of subatomic
particles whose features can only be determined as probabilities.
Third, many aspects of living systems, whether cells or the
weather, are best described by nonlinear equations, which have mathematical
features known as chaotic properties (note 5). Although
we can accurately model the weather by a series of equations, these equations
are exquisitely sensitive to initial conditions (Journal of Atmospheric Sciences 1963;20:130).
Lorenz, in his pioneering work on chaos theory, found that his computer
model of the weather experienced exponential divergence when he reran it
substituting the figure 0.506 for 0.506127 (Technology
Review (MIT), March/April 2011). This is
often called “the butterfly effect,” because a butterfly flapping its wings in
one city can affect factors that eventually change the weather across the world
(Predictability: Does the Flap of a
Butterfly's Wings in Brazil set off a Tornado in Texas, 1972 (AAAS)).
This inability to predict the future of well-understood systems is
disturbing to many. It is not due to our
lack of intelligence, and will not be altered by more powerful computers. It is an inherent property of the non-linear
world in which we live.
The fourth limitation is that living systems may be based on
incompressible algorithms. This means
that there is no shorter means to predict what a living system will do than to
simply observe what happens (At Home in the Universe, p. 23).
We would like a model that provides a shorter, compressed description,
such as an equation. However, if each
part of the living system contributes to the whole, then a model of only part
of the system may be markedly inaccurate.
Finally, we cannot predict the future of living systems
because the function of some molecules is dependent on evolutionary pressures,
which themselves cannot be predicted (Cold Spring Harb Symp Quant
Biol 1964;29:137). For example, selection may favor individuals
with the human sickle cell mutation, but only in geographic areas where
falciparum malaria is endemic, because this mutation protects erythrocytes from
infection (Nature Education-Natural Selection).
However, we cannot predict the impact of this particular mutation on
survival in the local environment without knowing the evolutionary pressures of
all other human molecules, and how they reinforce or counteract each
other. In addition, we must know how the
physical environment will change over time, how the genomes of competing
organisms will change over time, and how all these factors will interact. Each of these factors cannot be predicted
because of the first four limitations above, and because we have no natural
laws to predict specific evolutionary changes.
3. Life emerges when the molecular diversity of a closed
system exceeds a threshold of complexity.
According to Kauffman, life is the emergent collective
property of a modestly complex mix of molecules which catalyze each other’s
formation (The Origins
of Order,
Chapter 7). Individually each of the molecules is
inert. However, with a large enough
collection of molecules of sufficient complexity, confined to a small space so
they are more likely to interact with each other, a self-sustaining network or
web of reactions may be formed that can reproduce and evolve. Unlike some models of the origin of life, the
molecules do not reproduce themselves.
Rather, the network has the property that the last step in the formation
of each molecule is catalyzed by some other molecule in the network.
Under the right circumstances, there is a high probability
that a subset of a large set of complex molecules will form a network of
molecules that catalyze the formation of each other. A complex protein has numerous clefts or
grooves on its surfaces due to the tertiary structure of its polypeptide chain,
which may bind the transition states of reactions, a requirement for a catalyst
(Proc Natl Acad Sci USA 1994;91:4103). The likelihood
that a specific molecule will catalyze a specific reaction is very low, but
with a large number of molecules and possible reactions, the probability that some molecule will catalyze some reaction is high (At Home in the Universe, Chapter 3).
This is analogous to the “birthday problem.” In a group of 50 people, the probability that
any of them will have the same birthday as the first is only 49/365 (13%), but
the probability that any two people in the group will have a common birthday is
97% (Wikipedia).
As reactions are catalyzed, additional molecules are
created, which further increases the diversity of molecules. The number of possible catalyzed reactions
will increase faster than the number of molecules. Eventually, a threshold is crossed in which a
network of molecules exists that constitutes a collectively autocatalytic set,
in which each molecule’s formation is catalyzed by other molecules. This reaction network will also couple
exothermic and endothermic reactions (The Origins
of Order,
Chapter 8). Many alternative pathways are possible to
produce this autocatalytic network, but only one of them is required.
This model of the origin of life suggests an answer to the
question of why free living cells have an apparent minimal complexity. Mycoplasma
genitalium, the smallest known genome that constitutes a cell,
contains approximately 470 genes, a large number for the simplest organism (note 6). Based on the above model, we can hypothesize
that a lesser number of genes would lack the complexity to create a
self-staining network. Other theories of
the origin of life, such as the replicating RNA theory, provide no explanation for
this minimal complexity.
4. Much of the order in organisms is
due to generic properties that emerge from a network of gene products.
Each cell coordinates the activities of approximately 20,000
genes and their products (Nature 2004;431:931).
Not only does the cell coordinate complex activities such as mitosis,
but these activities occur without any oversight. The molecules interact spontaneously with
each other, and lead to the creation of new cells and new living organisms with
the same properties. If we are to get a
deeper understanding of diseases whose secrets have defied decades of determined
research, we may need to understand the general principles underlying these
processes.
The traditional view is that the sole source of order in
organisms is natural selection as described by Darwin (On the Origin of Species, 1859); the patterns of a pine cone or
the artistry of mitosis are due to “chance caught on the wing,” a description
of natural selection by Jacques Monod (Le Hasard et la nécessité (Chance and Necessity, 1970). An alternative view, advanced by Kauffman and
others, is that order is an expected emergent property of molecular networks,
based on structural properties of networks that are not dependent on details on
the particular molecules (The
Origins of Order: Self-Organization and Selection in Evolution, 1993). Genes, RNA
and proteins form a complex parallel processing network in which each molecule
is connected to some other molecule, and switches each other off and on. During ontogeny, gene products turn each
other on and off to generate precursor cells.
Once cells become differentiated, some of this switching continues, but
the cells are relatively stable.
Consider a hypothetical gene product A, which could be
considered active or not, based on its state of phosphorylation. Based on inputs from kinases or other gene
products, protein A may change its properties in the next instant of time. Any change to gene product A may affect other
gene products it is connected to in the network, and for which it acts as an
input. If we consider all the gene
products in a cell, we can consider all of their properties at a particular
moment in time as the state of the cell.
In the next instant, each gene product may change its properties based
on relevant inputs, leading to a new state.
The path that a network traverses over time, based on changes in the
state of each gene product at each moment in time, is called its state space or
state cycle.
Theoretically, a cell with 20,000 types of gene products,
and numerous copies of each, could have an almost infinite length to its state
space, and never return to its original state.
For example, a network of N gene products, with only one copy and two
possible properties for each gene product, would theoretically have a state
cycle of length 2N. For N =
20,000 gene products, this is approximately 106000. However, a state cycle this large does not
happen, due to three network properties that induce order.
The first network property that induces order is the
surprising finding that if each gene product is regulated by at most 2 inputs,
which is relatively common, the median length of the state cycle is only
the square root of the number of gene products, or 141 if N is 20,000 (J Theor Biol 1969;22:437, The
Origins of Order, p. 479). This network
property creates inherent stability, even in networks with large numbers of gene
products, as the cell network is localized to a very small percentage of its
possible state space. In contrast,
networks with 5 or more inputs per gene product have an exponential rate of
growth and chaotic behavior (The
Origins of Order,
p. 194).
The second network property that maintains order is that
most genes are regulated by “canalyzing” Boolean functions (J Theor Biol 2011
Aug 25 [Epub ahead of print]), which means that one input can completely determine the
property of the gene. For example, the
Boolean OR function is canalyzing. In
the OR function, if one input is active, then the result is active, regardless
of the value of the other input. A
series of canalyzing functions, such as a network of genes regulated by the OR
function, is considered a “forcing structure,” because changing one input to
active propagates the “active” status throughout the network. If there is a loop, then the network is
frozen in this active state. Thus,
genetic networks with canalyzing functions are stable, which further limits the
length of the state cycle (Proc Natl Acad Sci USA
2004;101:17102).
The third network property that maintains order is the P
parameter, which may provide order even when molecules are regulated by a large
number of inputs, as long as the molecule has only two states, such as active
and inactive. Assume that a Boolean
function (such as AND or OR) exists to determine the state of a particular
molecule. If there are K inputs for the
molecule, there are 2K possible combinations of input values. The P parameter measures the proportion of
these combinations that create an active or inactive state, whichever is
larger. Thus, it has a value of 0.5 to
1.0 (19). If the molecule exists 50% of
the time in each state, the P parameter is 0.5.
If the molecule exists 100% in one of the states, the P parameter is
1.0. As P moves from 0.5 to 1.0, the
network is tuned to becoming more orderly. The P parameter helps explain the
order induced by gene products with at most 2 inputs. In these networks, a large number of Boolean
functions have P values of .75, which promotes stability. In contrast, as the number of inputs
increases, functions with such high values of P become rare (At Home in the Universe, p. 84).
Kauffman terms the results of these three network properties
“order for free.” As a result of these
properties, cellular networks actually have only a limited number of
states. Mutations and other
perturbations can change the functional connections between some of the
molecules in the network, but this usually does not change the stability of the
network very much due to these order inducing properties.
These generic network properties have profound implications.
Cellular networks, once they arise, have an inherent order independent of the
particular molecules or reactions involved.
This order is part of the interactions between the parts described above
that makes the reductionist approach inadequate for understanding biologic
systems.
5. Numerous biologic pressures push
cells towards disorder.
How do these laws of complexity and self-organization relate
to neoplasia? Tension exists in living
systems between order and disorder, an example of the tradeoffs inherent in
living systems to achieve compromise between conflicting interests. A certain amount of order is required within
cells, tissues and organs for living systems to function. However, numerous pressures push cells
towards disorder.
First, all systems, living and nonliving, tend towards
increasing disorder or entropy, based on the second law of thermodynamics (Wikipedia). Without external
energy, entropy tends to increase, which means cells have to spend their
limited resources on countering this pressure towards disorder. Cells can use entropy to drive useful
tasks. For example, entropy promotes DNA
self-assembly by favoring the packing of many small molecules, which outweighs
the cost of forming loops with thousands of base pairs (Biophys J 2006;90:3712).
However, using
entropy in this way still
requires significant cellular resources.
Second, the process of creating an autocatalytic network,
discussed above, promotes disorder. As a
closed network of molecules begins reacting with itself, it produces an
increasing number of new molecules, which catalyze further reactions. This exponential expansion of the network
creates continuing disorder, and itself has no natural stopping point until
limiting factors are reached.
Third, Kauffman suggests that the earth’s environment is
supracritical, which puts pressure on living systems to breach their orderly
constraints (At Home in the Universe, Chapter 6).
In an uncontrolled nuclear reaction, a uranium 235 atom absorbs one
neutron, leading to nuclear fission, which releases more neutrons, which then
act on other uranium atoms. Similarly,
it is believed that life on Earth is expanding at an exponential rate. Life may have begun with a small collection
of amino acids (Science 1953;117:528, Astrobiology 2003;3:291), but today, over 10 million
different organic molecular structures have been cataloged (Wikipedia). The presence of
these molecules throughout our biosphere puts constant pressure on living
systems to interact with them in novel ways and abandon their ordered ways.
Fourth, natural
selection disfavors rigid order in living systems. Organisms do have the ability to maintain genetic
information unchanged over long time periods.
For example, the 16S and 23S ribosomal RNA genes are the most conserved DNA sequence segments across phylogeny, with
minimal changes over billions of years of evolution (Orig Life Evol
Biosph 2008;38:517). They apparently
are under strong evolutionary pressure and positive selection, as changes would be deleterious to
the organism.
However, although conservation of DNA is possible, cells
must also be flexible to respond to a complex environment, to facilitate
development, and to exhibit minor modifications when mutated. In addition, cells must have the ability to
undergo change sufficient to form new species. From the Cambrian explosion onwards, the fossil record shows a number
of large extinction events which wiped out a significant fraction of Earth’s
species, possibly caused by loss of habitat and environmental changes (Extinctions
and Biodiversity in the Fossil Record 2002). Rigid order would doom species amidst these
changes.
Finally,
the ability of living systems to maintain viability after mutational changes demonstrates an inherit flexibility
not present in a completely ordered regime.
Many systems are incapable of sustaining small changes without
substantial loss of function. For
example, some scientists speculate that life began with an “RNA world”
dominated by RNA polymerase (Nature 1986;319:618). However, a simple network with
this type of polymerase is subject to “error catastrophe” (J Virol 2006;80:20) because a small change in the
polymerase will most likely lead to erroneous replication if no error
correction mechanisms are in place.
Similarly, computer programs are also exquisitely sensitive to
changes. Almost any change to valid
computer code will lead to a nonfunctioning program. Thus, the ability of complex systems to
survive small changes and actually improve in function should not be taken for
granted.
Kauffman believes that organisms maintain a position between
order and disorder that he terms the “edge of chaos,” an evolution derived
compromise between order and surprise that may be optimal to coordinate complex
activities and to evolve further (At Home in the Universe, p. 86; Proc Natl Acad Sci USA
2005;102:13439). This “edge of chaos” is a type
of phase transition, similar to that of water and ice at the freezing point.
6. Organisms resist these biologic
pressures towards disorder through multiple layers of redundant controls.
It appears that organisms have multiple layers of redundant
controls that resist the above pressures towards disorder.
First, the structure of biologic networks inherently resists
these pressures towards disorder. Based
on interactions between the components, a large “frozen” component forms, whose
state does not change over time even as the states of other molecules change (The Origins
of Order, p.
206; Physica
D: Nonlinear Phenomena 1990;42:135).
In addition, transient changes to the state of a gene product tend not
to propagate to the remainder of the network.
This not only maintains network stability against minor changes, but
also provides the basis for cellular homeostasis. Although cells can theoretically exhibit an
almost infinite number of states, there are actually only approximately 300
cell types, which maintain a consistent phenotype due to the relatively stable
expression of a large percentage of their gene products (The
Origins of Order,
p. 467). These homeostatic states are called
attractors, which limit the behavior of the network to a small portion of its
state space. Thus, a cardiac myocyte
cannot transform to a glial cell, and only rarely do mature cells change their
state of differentiation at all.
Second, cells have external membranes with transporter
proteins that act as “border controls” (Science
2008;322:709). This limits the entry of novel
molecules into the cell that might create new reactions or alter existing
ones. In addition, molecules are
compartmentalized to limit their possible reactions. For example, hematopoietic stem cells
physiologically produce the c-Abl tyrosine kinase, which is sequestered
in the cytoplasm by binding to 14-3-3 proteins (Nat Cell Biol 2005;7:278). In
response to DNA damage, c-Abl is targeted to the nucleus, but only under
tightly controlled conditions (Cancer Control 2009;16:100).
Third, cells have robust processes to limit errors. This prevents altered protein structure and function,
which might disturb existing network interactions or create new ones. Transcription is a complex process, with an
estimated error rate of 1 per 100,000 nucleotides (Nature Education 1(1)-DNA Replication and Causes of
Mutation). DNA repair processes,
including proofreading and mismatch repair, correct most of these errors. Cells also limit known causes of DNA damage,
such as reactive oxygen species, which would otherwise increase the error rate
of transcription and translation (J Biol Chem
1969;244:6049, J Biol Chem 2001;276:38084).
Fourth, cells have several mechanisms to respond to injury
or DNA damage, which might eventually alter proteins. Through an underdetermined intracellular
audit, some injured cells undergo apoptosis through activation of caspases and
bcl2 family members (Mol Neurobiol 2010;42:4). In addition, transcription
factors and signaling molecules can recognize potentially lethal stimuli and
initiate cycle arrest, autophagy or protein synthesis shutoff (Carcinogenesis 2011;32:955). The importance of the apoptotic pathway is
suggested by the high frequency of a defective p53 response in most cancers,
either by p53 gene mutations or deletions, or by other alterations in the p53
pathway (Science 1991;253:49).
Fifth, key cellular processes have numerous controls that
tightly regulate their activity. For
example, various mitotic checkpoints ensure genomic integrity by delaying cell
cycle progression in the presence of DNA or spindle damage (Mol Cell Biol 2010;30:22, DNA Repair (Amst) 2009;8:1047). These multiple levels of control prevent inappropriate initiation of
mitosis or other key processes, which reduces the propagation of transformed
cells.
Finally, humans have an extensive and varied
immune surveillance system of B and T cells, NK cells and macrophages. These immune cells have both innate and
adaptive properties, and maintain order by destroying cells with disordered
properties (Cell
2010;140:883). Their importance is suggested
by the impact of immunosuppression, which is usually associated with a markedly
elevated risk of malignancy (Nat Rev Nephrol 2010;6:511).
7. Neoplasia occurs as these
multiple controls are breached. The
resulting neoplastic process reflects the nature of the breaches, the
individual’s germline configuration and the network state of the cell of
origin.
The laws of complexity and self organization give us a
framework to better understand neoplasia.
Cells are not just the sum of 20,000 gene products interacting with each
other in a predictable way (law 1), but networks of interactions that emerge
between these molecules, which cannot be predicted (laws 2-3). These networks possess a great deal of
stability, independent of the specific nature of the molecules (law 4). On the other hand, these cells are also under
tremendous pressure to breach the control mechanisms that maintain order by
resisting the inherent biologic pressures in the cell towards disorder (law
5). In fact, a certain amount of
disorder is promoted by natural selection.
Thus, it is not surprising that cells will escape some of these
controls. Only the presence of multiple
redundant controls at various levels (law 6)
maintains sufficient order in cells and organisms to allow them to
function, consistent with the “multiple-hit” theories that more than one
mutation is necessary for neoplasia (Br J Cancer 1953;7:68, Proc Natl Acad Sci USA 1971;68:820).
To illustrate the key idea of the inherent biologic forces
promoting disorder, consider a plan to put a large number of teenagers in a
small area, and get them to create some kind of ordered community. A large number of adults would be required to
get the teenagers to obey the rules. Yet
teenagers inherently don’t like being told what to do, and even with multiple
layers of enforcement, there will be frequent attempts by the teenagers to work
together to break the rules, with occasional success, and with varied responses
from the adults. This type of tension
exists in our cells on a regular basis.
Of course, molecules cannot think, but they have an inherent tendency to
want to react with other molecules in novel ways. Another difference is that the control
mechanisms are also part of the molecules being controlled, and not a separate
class. When order starts to break down, it
can create chaos with the teenagers, and neoplasia within the cell.
Nature of breaches
It may be possible to classify neoplasia based not just on
morphologic changes, but on the type of controls that have been breached. For example, low grade gastric MALT lymphoma
is strongly associated with chronic Helicobacter
pylori infection, and H. pylori
eradication provides an excellent long-term outcome (Gut 2011 Sep 2
[Epub ahead of print]). It is possible that over long
periods of time, these inflammatory changes induce changes to the cellular
network that is otherwise “frozen,” which leads to neoplasia.
Some genetic changes alter the distribution of molecules
within the cell, allowing them to interact with other molecules previously
restricted. The t(9;22) translocation
leads to a bcr-abl fusion protein in hematopoietic stem cells, which leads to constitutive tyrosine kinase
activation and translocation of the fusion protein from the nucleus to the
cytoplasm. As a result, the fusion
protein now has numerous additional targets for catalysis based on its new
location, which may disrupt existing networks (Cancer Control 2009;16:100).
Similarly, alterations in networks associated with error
control (Mech
Ageing Dev 2008;129:391), apoptosis (Cancer Epidemiol Biomarkers Prev 2009;18:1680), other key pathways (Science 2008;321:1801) and immune surveillance (Int J Biol Sci 2011;7:651) can be viewed as breaches in the
physiologic mechanisms that resist pressures towards disorder. Tumors with similar types of breaches may
have common clinical or morphologic features, or respond to similar types of
treatment, an area of future investigation.
Germline configuration
The nature of the neoplasia is
also affected by the germline configuration of the organism. A large percentage of genes have more than
one allele, and many individuals are heterozygous for a large fraction of these
alleles (RC Lewontin, The Genetic Basis of Evolutionary Change, 1974). Individual variation in these alleles means
individuals have inherent differences in the genetic networks in their
cells. These differences may involve
genes known to be important control mechanisms.
To date, at least 54 familiar cancer syndromes have been identified,
with mutations in p53 or other known oncogenes (J
Natl Cancer Inst Monogr 2008;(38):1).
Some allelic variations may affect
important control mechanisms only under specific circumstances. For example, the CCR5 gene encodes a
chemokine receptor whose role in normal immune function is unclear, but which is
the principal receptor for macrophage-tropic viruses, including HIV1. Approximately 80% of Caucasians have this
allele, rendering them susceptible to HIV1 infection, which if untreated,
impairs immune surveillance and leads to increased rates of various malignancies. However, 20% of Caucasians are heterozygous
for the CCR5 delta 32 mutation, which produces a shorter nonfunctional gene
that does not migrate to the cell surface.
As a result, these patients are relatively resistant to HIV1 infection (Cell 1996;86:367,
PLoS One 2011;6:e22215). Thus, the CCR5 delta 32 mutation has an
impact on neoplasia, but only in the presence of an HIV epidemic.
Network state of cell of origin
The phenotype and state of differentiation of the cell of
origin, a reflection of its network state, constitute a somewhat permanent
variation to the germline configuration that may also influence the nature of
the neoplastic process. Cellular
differentiation is due to the capacity of genes to modify the activity of other
genes. The network properties that
provide “order for free,” discussed above, also create variability in the
effect of breaches of control mechanisms.
In many cell types, a change in a molecule’s function will have minimal
impact on the existing network in that cell, because the molecules they would
typically affect are “frozen” in a particular state by other members of the
network. For example, t(14;18) is
important only in lymphocytes that produce IgH; in other cells, the altered
molecules cannot propagate within the network.
The control mechanisms may also change due to hormones,
environmental influences, or unknown maturational factors. Thus, mutations may have a different impact
in neonates, children and adults.
Although cells may look similar morphologically at different points in
development, there may be varying degrees of homeostasis based on different
control mechanisms. Perhaps part of the
reason Wilm’s tumor or neuroblastoma occur primarily in children/young adults
is that the cells in older adults have developed additional control mechanisms
that make these cells more resistant to the changes required to develop these
neoplastic conditions.
Summary
In the framework of the laws of complexity and
self-organization, cells maintain order via redundant control features that
resist the inherent biologic pressure in the cell towards disorder. Neoplasia can be understood as the accumulation
of changes that allow cells to override these control features, leading to
altered networks, which may lead to dysregulated growth and differentiation.
The nature of the neoplasia may be affected by the nature of the breaches, the
germline configuration and the network state of the cell of origin. A better understanding of how these factors
interact with each other may generate new approaches to treatment.
Notes
Note: all links were retrieved on 30 November 2011 unless
otherwise indicated.
Note: links in green are to free full text references.
(1) The National Cancer Act followed
President Nixon’s pledge in his State of the Union address on 22 January 1971:
I will also ask for an appropriation of an extra $100
million to launch an intensive campaign to find a cure for cancer, and I will
ask later for whatever additional funds can effectively be used. The time has
come in America when the same kind of concentrated effort that split the atom
and took man to the moon should be turned toward conquering this dread disease.
Let us make a total national commitment to achieve this goal.
See National
Cancer Institute pages #1, #2, video of President Nixon
signing the bill into law
(2) This paper is based primarily on these
works of Dr. Stuart A. Kauffman. The
Origins of Order contains detailed discussions supporting the underlying
theories:
Kauffman, SA. The
Origins of Order: Self-Organization and Selection in Evolution,
Oxford University Press, 1993
Kauffman, S. At Home in the Universe: The Search for the
Laws of Self-Organization and Complexity, Oxford University Press, 1996
Kauffman, SA. Investigations,
Oxford University Press, 2002
Kauffman, SA. Reinventing
the Sacred: A New View of Science, Reason, and Religion, Basic Books,
2010
(3) Crick FHC. Of Molecules and Men (1966),
reprinted by Prometheus Books (2004), http://www.amazon.com/exec/obidos/ASIN/1591021855/pathologyoutl-20
See also Steinbach, Peter J: Classical and Quantum Mechanics
- in a Nutshell, http://cmm.cit.nih.gov/intro_simulation/node1.html
(5) For examples of chaos in biologic
systems, see Phys
Rev E Stat Nonlin Soft Matter Phys 2010;82:011924 (CNS neurons), Phys Rev
E Stat Nonlin Soft Matter Phys 2009;80:016213 (blood flow), Med Biol Eng Comput 2008;46:433 (coronary circulation).
(6) Mycoplasma
genitalium is
reported to have from 475-82 protein encoding genes (see Science 1995;270:397 and Wikipedia). Smaller genomes
have been documented in bacterial endosymbiots, but they are not capable of
independent life, and are considered a transition to becoming an
organelle.
-- End --