The Evolution of Multicellularity
The evolution from single cells to biofilms ("training wheels" for multicellularity)
and then to full-blown multicellular life required at least the beginnings of four key
organizing principles. Multicellular computing is following a similar path.
The transition from single-cells to multicellular
life did not happen in one evolutionary leap.
We do not know with much precision when and how the strategies arose
that support today's multicellular organisms
nor what alternatives were tried and failed.
We know that what we see today survived the test of time.
It appears that multicellular life evolved from single cells
in two stages. First, single cell organisms evolved the ability to form
loose cooperative communities, called biofilms,
that can perhaps be thought of as “training wheels” for
multicellular life. Perhaps the earliest colony bacteria were
the cyanobacteria that evolved more than three billion
Biofilms remain common today. Present-day examples of biofilms include
slime mold, dental plaque, films on rocks in streams and many more. They are
complex ecologies
of single-cell organisms that typically include algae, bacteria, protozoa,
cyanobacteria, fungi, and viruses.
Perhaps one billion years ago true multicellular organisms
formed – plants, animals, and fungi – known generically as Metazoans.
Unlike cells in biofilms, all cells in a Metazoan organism share the same DNA.
As the organism develops, the cells' genetic programs direct them to sequester
and permanently silence much of their DNA. They thereby become specialized.
Some organisms have multiple stages of stable forms, e.g., insects that
exhibit larva, pupae, and adult forms. But these developmental stages all
involve programmed cell differentiation. For most cells, stem cells being
the exception, differentiation and the resulting cell specialization is
dramatic and irreversible. The full complement of genes and DNA control sequences in a multi-cellular
genome is far more complex than that of most single cell organisms
[1].
Yet any given type of
cell – and there are about 250 different types in humans – is functionally
much simpler than a typical single cell organism.
Each differentiated cell type uses just a subset of the 25,000 total human genes.
For example, all cells in the body have the gene for hemoglobin, but only
red blood cells make that protein. Along with this
specialization, the cells
must coordinate their activities by sending messages
to each other. They also work cooperatively to develop their "body," which is a
stigmergy structure. And they need
apoptosis mechanisms to remove cells that have
outlived their usefulness or become dangerous.
Without all four of those organizing principles operating together
in a coordinated manner,
true multicellularity would not have been possible.
Might a different set of basic multicellular principles have worked just as well?
Possibly. But, if so, we cannot know what they might have been.
It is difficult to argue that any one factor is primarily responsible for the
evolution of multicellularity. Conventional wisdom once asserted that the primary
benefit of multicellularity, hence presumably what drove its evolution, was the
division of labor, or specialization, provided by differentiated cells.
(see Maynard Smith, J. & Szathmáry, E. The Major Transitions in Evolution, 1995).
But preexisting biofilms already used all four principals, not just specialization.
They used stigmergy to structure their colonies and polymorphic messaging for
quorum sensing and apoptosis.
It was the cooperation of unlike species of single-cell organisms that had
already given biofilms an advantage over single independent cells.
From that perspective, it would seem that the messaging, stigmergy, and apoptosis that support
cooperation are at least as responsible for the evolution of multicellularity.
Multicellular organisms also benefit from the advantages of scale, which is an organism-level
property rather than a property of individual cells. Larger organisms can be
more mobile,
they can pool sensory information, e.g. about vibration, light, and their chemical
environment over a wider area, and they can be more stable because they are not as subject
to the random effects of Brownian motion. So they have a new range of competitive stratagies
available for foraging, hunting, and defense.
In the world of computing, the evolution toward multicellularity began
when PCs were used as terminals to mainframes in place of dedicated
terminals. As more “smarts” or software function
migrated from the mainframe to the PC terminals, the interaction
between client and server became richer and more varied. Then, with the emergence of the
Internet, the central role of the mainframe was eclipsed. Today
we see loosely organized general-purpose computers in web communities, P2P networks,
and ad hoc grids such as SETI at home. These loosely organized communities are
comparable to biofilms. Some Grid architectures are more formal and specialized
and therefore are more analogous to small Metazoa such as the hydra or perhaps
small jellyfish. In any case, we now see at least some aspects of all
four principles already at play.
Specialization becomes more
common. Polymorphic messaging,
especially in Service Oriented Architectures (SOA), Web Services,
and Web 2.0 mashups grows rapidly. New and interesting
stigmergy structures
abound. And the beginnings of apoptosis mechanisms
are showing up.
The advantages of scale also become evident in the Internet. Many of the more novel
apps in iPhones/iPads benefit from pooling information across many individual
machines at disparate locations. Some devices are mobile and contribute local
information such as GPS position, acceleration and orientation, vision (from their
embedded cameras), and sound.
Will computing require additional, or different basic principles?
We cannot yet know. But the fact that the same four principles seem to be already
emerging in the rather different world of multicellular computing
suggests that these are at leat a valuable initial set of organizing principles.
[1] Some single cell organisms, e.g. certain species of Amoeba, can have genomes more than 100 times larger than the human genome. See for example
Sizing up genomes: Amoeba is king, Edward R. Winstead, Feb. 2001. Note however that the number of base pairs in a genome
does not necessarily determine its complexity.
Contact: sburbeck at mindspring.com
years ago. Their fossil remains are visible today because these colonies
secreted a thick gel as protection from strong solar radiation. This gel, in turn,
trapped sand and debris from the surf which, together with lime secreted by the
bacteria, formed the beautiful patterns of the Stromatolite fossil reefs
visible in Australia (see image at left). These structures vary in size from
twig-size to semi-truck size.
The need for all four key principles of multicellularity
The benefits of multicellularity
Parallels with the evolution of multicellular computing
Last revised 5/14/2010