Ganti's Chemoton theory from the eyes of a chemist
The idea of understanding a living cell as a fluid chemical automaton goes back to Tibor Ganti (1.1.1), who, in the mid seventies saw the earliest cell on Earth as entities in which an autocatalytic metabolism was coupled to the self-reproduction of a membrane and a genetic subsystem.[1]
Figure 1
The metabolic core of Ganti's chemoton model (Figure 1) was inspired
by the formose reaction (1.1.2).
As in the formose reaction Ganti assumed the uptake of a single
resource molecule XA (the equivalent of formaldehyde). The
metabolism consisted of five reactions connecting five intermediates to
form an autocatalytic cycle where the doubling (A5 ← → 2 A1) was the
equivalent of the split of aldotetroses into 2 molecules of
glycolaldehyde.
Ganti saw the two other subsystems having evolved at the periphery
of the metabolic core, the first being the lipid generator. Lipid
precursor T' splitting from the metabolic engine is converted into T*
and then combines with a reaction product R from the genetic subsystem
to yield Lipid T. The re-translation into the implementation landscape
of chemistry could mean either a head to tail combination, or, as
lipids usually consists of two hydrophobic "tails" a combination of a
second tail with a head-tail construct. In any case it means that the
number of lipids reflects the number of internucleotide bonds formed in
the process of genetic reproduction as well as the number of
metabolites from which the precursors were generated. This is now
called "stoichiometric coupling". The only assumption needed is that
the whole reaction network could run as a finite state engine
reproducing earlier states in later division and growth cycles.
Oszillations however like in the Brusselator did NOT show up by
themselves. They had to be "induced" in the dynamic simulations.
Inspite of this and other conceptual problems (1.1.3),
Ganti's chemoton theory holds importants conceptual insights (1.1.4)
for protocell implementation.
One may wonder why Ganti in the seventies did not increase the
plausibility of the chemical part of his model, especially if one
considers that Ganti was trained in chemistry. The simple answer is
found in Moore's law. Assuming a doubling period of 18 months, 30 years
ago computers were 230/1.5, viz. ca. 106 times
slower than today. A simulation of Ganti's chemoton for 1000 time
points takes a second or so today (SimFit) but took almost two weeks in
Ganti's times. Meaningful changes of the model itself as well as some
of his parameters and initial concentrations were thus out of his
access to computing power behind the iron curtain. We therefore have to
take the Chemoton idea as a rough sketch giving us the basic idea how a
chemical automaton could operate but leaving it to ourselfes to start
thinking about possible implementation schemes.
A more realistic model explicitly considering elementary events in
template replication, coupled to micelle reproduction and energy uptake
from a photochemical engine was introduced by Rasmussen et al. [5]
In a sense, the Los Alamos bug (as it is coined today) adds
compartimentization to Kauffman's "autonomous agent", a reaction system
based on a parabolic replicator coupled to a photochemical engine.[6]
In contrast to Kauffman's model, the photochemically driven metabolism
is much better worked out here pointing to a reaction scheme that may
be within possible implementation scenarios in chemistry.
Whatever protocell model will finally be at the level of chemical implementability one has to see that any effort to integrate a protocell has to be a stepwise approach. As there are three subsystems there are three formal couplings to be explored: A metabolic - genetic coupling, a metabolic - membrane coupling, and a membrane - genetic coupling. Systems arising from such couplings were recently coined infrabiological systems.[7]