from Thursday, October 16, 2003 9:00-10:00 PM Eastern
© by International Society for Complexity, Information, and Design
Our guest speaker today is Lynn Caporale. Lynn received her Ph.D. in
Molecular Biology from the University of California at Berkeley.
After teaching and doing research at New York University, Memorial/Sloan-Kettering
Cancer Center, Rockefeller University, and Georgetown University
Medical School, she moved to Merck Research Laboratories, where she
spent over a decade focused on the discovery of new medicines.
Dr. Caporale has held research and senior executive positions with
various biotechnology companies and in the pharmaceutical industry
and currently is an independent consultant in drug discovery and
functional genomics. Her most recent book is titled Darwin in the
I'm now going to turn the chat over to Dr. Caporale.
Thank you Micah. I'll begin with a little background about my book
and other publications.
Darwin and Wallace proposed that evolution takes place through variation
followed by selection. However, they lacked the molecular tools to
suggest what the source of that variation might be. Biologists
later incorporated genes and mutation to say that evolution takes place
through random mutation followed by selection. That mutation
is random was not Darwin’s idea.
In the 1980s, I suggested that because there is more than one DNA triplet
that encodes most amino acids, additional information could be transmitted
through a protein coding sequence [in addition we now know the large
extent of non-protein coding sequences in the genome]. Such information could affect the rate and type of variation along
More recently, I have discussed how Natural Selection, by acting on
intrinsic variations in the probability of distinct types of mutation,
acts on the mechanisms that generate genome variation, thus incorporating
information about the world into the mechanisms that generate genome
variation between generations.
This implies that intelligence about the world can evolve within a
genome under the pressure of Natural Selection and affect the future
evolution of that genome.
I look forward to a good discussion this evening.
Dr. Caporale, Many thanks for being here. In order to 'anticipate'
the future and respond adaptively to that 'anticipation,' a system
must: 1. Store (abstract) durable representations of circumstances
as they are encountered; 2. Store (abstract) durable representations
of responses & outcomes in association with those circumstances;
3. Recognize that current circumstances are similar to the initial
stages of one of the stored representations of past circumstances;
and 4. Respond based on some sort of evaluation of prior responses & outcomes
in those past circumstances. A system that can do those four things
will give the appearance of "anticipating" the future,
but since there is no explicit or implicit model of that future in
the system, that is an appearance only. "Anticipation" then
is recognizing current circumstances that are similar to past circumstances.
Given that, is there any suggestion in your work that "teleology" (in
the sense of goals imposed by an internal or external intelligent
agency) necessary to give a full account of biological phenomena?
There are several points made in your statement and question, so let
me begin by giving a specific example of what I mean by "information
about the world" that affects variation, and then we can discuss
your broader issues in those terms. But first, bottom line, my observations
do not require an external intelligenct agency.
One specific example: it is to the advantage of pathogenic bacteria
to vary their coat proteins, for example to stay one jump ahead of
our immune system. It turns out that the DNA that encodes the coat
protein varies at a higher rate than DNA needed for " housekeeping" chores
What does your theoretical stance have to say about communication between
genomes of the same species, different species, etc?
do you mean molecular communication-- growth signals for example, or
do you mean the transfer of DNA between different individuals?
Molecular communication, mainly.
There definitely is communication amoung genomes. FOr example,
biofilms are multi-cellular bacterial structures that depend upon communication
between different bacteria to be built. But I am a little uncertain
re where you are taking this wrt evolution. [end]
Are you saying that somehow an animal's genome can somehow communicate
the necessary changes needed to its sex cells? I thought sex cells
are unchangeable in that respect?
Perhaps it's better if I explain the overall mechanism-- selection
changes the genome. In other words, suppose there are two individuals,
with an intrinsically different probability of mutation at certain
locations in their genomes. The genome that tends to make more destructive
changes will tend to have fewer progeny over the course of many generations.
So, the majority of progeny [looking at the percentage of the progeny
of each of the two types of genomes] would be from the genome that
tended to make less destructive changes. In other words, the species
genome has "learned" through selection not to mutate in
Chromosomal rearrangements, especially duplications, are often theorized
to be essential for creating signalling pathways (e.g. the various
Ras/MAPK kinase cascades). However, rates and causes of chromosomal
rearrangements are less well understood than those of point mutations.
Furthermore, one may anticipate that chromosomal rearrangements would
have a greater impact and require a longer time to become functionally
essential, than an increase in point mutations. Is there good direct
evidence for, or a good theoretical indication that, evolvability
on this scale has evolved? For example, if such mechanisms existed,
the Cambrian explosion wouldn't seem particularly surprising. But
it's difficult for me to envision how such broad-scale mechanisms
could have evolved and then been maintained.
for some reason my connection got dropped. I'm back now
the one re chromosomes? It's a very important subject. Unfortunately
the question is no longer on my screen. there was a good review on
that subject in Ann. Rev. Genetics 2002 by Heinz Saedler.
Recently it was found that the sloppy Pol IV gene activated in E. coli
under late stationary phase conditions. The outcome of this gene
activation will be an increased mutation rate per generation. Is
this at all relevant to your thesis?
This is part of it. I wrote a review in Ann. REv. Microbiology 2003
which discusses the "mutation phenotype" of a cell. A genome
must have a balance between fidelity and exploration. Under certain
circumstances, it is good to increase the
rate of variation. However, if this were done randomly through the
it would be less adaptive than if certain types of mutations [eg
for the bacteria, in the coat proteins] were more likely than others.
SO, the role of "mutator polymerases" and other mechanisms
of decreased fidelity is an interesting area to explore.
Interestingly, such "Mutator polymerases" may also to be
involved in the generation of diversity in our own immune response.[end]
aba phil I noticed in reading your Review article that the many new
insights you described were not related to Darwinian Concepts in
the sense that the same program would have been followed if one had
apprached with the conclusion that Darwin's Theory was substantially
wrong. The Theory seemed non-relevant to the modern enterprise. Would
you please comment on this.
it would help me to understand what you are saying if you point
out where the same program would have been followed even if Darwin
were wrong. Darwin suggested that evolution takes place by variation
followed by selection. What my work is doing is deepening our understanding
of the power and scope of selection [ie it works on the mechanisms
that generate variation as much as it works on beaks and wings]
Are you suggesting that somehow environmental changes may be inherited
sans involvement of germinal cells.?
I am asking because logically there is no reason why environment shouldn't
add on its imprint on the species for the benefit of future generations
facing the changed environs.
The germ line is very much involved . In fact, if you are interested
in evolution in multicellular eukaryotes [as I suspect we all are]
then the place to look closely is at meiosis and mitosis in the germ
line. For example, the histocompatibility locus [responsible for
protection against viruses, but discovered because it makes tissue
transplantation challenging] is a site of high variation among individuals
and it also is a site of high recombination in the germ line.
In solid mechanics complex material behavior can be phenomenologically
modeled using internal variables which may be analogous to your additional
information. The choice of types of internal variables (scalar, 2nd
order tensor, ...) place limits on the types of behavior which can
be described. Does your identification of process variables limit
the types of mutations which may take place? Are there any inherit
If I understand your question, the only limits are that we must wind
up with As, Ts, Gs, and Cs. But there can be point mutations [a single
A changes to a G-- and of course that will mean a T changes to a
C on the other strand], insertions [eg ATA becomes ATTA], deletions,
and recombination, where some information is moved between chromosomes.
Is that what you meant?
I guess that I was wondering if your modeling of the process rate or
location would limit the types of mutuations
On reflection you were asking how the information might constrain the
mutations-- so for example, insertions and deletions might become
more probable in a sequence such as TTTTTT or GATCGATCGATC as the
polymerase slips and loses register for example, and just such sequences
are found in the area of the changeable coat proteins.
Also, there is "site directed" recombination, where sequences
at the border of certain DNA regions make them likely to be cut and
pasted and moved to other sites. Much of our antibody diversity is
generated that way and, yes, some bacteria change their coat proteins
by cutting and pasting [eg the bacteria responsible for Lyme disease]
Amazon.com's Synopsis of your recent book says "Most importantly,
by exploring the genome and its evolutionary strategies in wonderful
detail, Caporale disperses the nagging doubt that natural selection
could have produced human life unassisted." Is that a fair statement
of your view?
Yes, I think that understanding that natural selection can act on the
mechanisms that generate genome variation makes evolution by natural
selection more conceiveable than is a mechanism of mutation by completely
random mutation. [end]
Evolutionary pressures are quite different on, say, bacteria and humans.
What classes of pro-evolvability features do you think are maintained
in higher (i.e. long-generation-time, low-brood-size) vertebrates,
and which would be limited to bacteria and single-celled eukaryotes?
I think we have inherited many of the same mechanisms from the distant
past, but you are right that there are differences in the rate of
selection that is available to us and therefore the value of certain
types of variation. One thing we don't do [although some feel some
evidence disagrees with what I am about to say] is take up DNA from
the environment and incorporate it into our genome. In contrast,
bacteria do that all the time.
On the other hand, the "slippery" DNA that is used in bacteria
may also be used in our own genomes, as a variety of proteins, including
the androgen receptor for example, have these slippery sequences and
do vary between individuals [end]
Would you say that mutations caused by environmental stressors (e.g.
uv, chemicals, ionizing radiation) that ultimately affect a gene
or genes leading to increased survival, growth etc. could result
in stable propagation of the mutation and the mutation could indeed
be in the sequences recognized by a regulatory transcription factor?
yes, if they lead to increased survival as you say.[end]
Earlier you used the phrase “information about the world.” This
reminds me of Chris Adami’s work in Artificial Intelligence.
Shannon information cannot address meaning or “aboutness.” It
is merely combinatorial and probabilistic. How did stochastic ensembles
in a prebiotic environment acquire semantic message-meaning “about
Pre-biotic is harder for me to envision than DNA-based life. However,
unless I am misunderstanding you, selection "teaches" what "works" in
the environment in which the organism [or pre-organism?] finds itself.
Constantly changing environments also "teach" the importance
of being able to change. But expand on your question, if you feel
that I am missing the point. Parenthetically, I am glad that you
mentioned Shannon as it was my reading Campbell's "Grammatical
Man" that sort of incubated in my subconscious and led me to
propose that additional information could be "transmitted" "under" a
protein coding sequence.
SO in a sense I do see changing probabilities of mutation as being
the way a population of learns.
I wonder how the same historical environmental changes could have taken
place that results in two different kinds of genomic survivability,
viz. prokaryotes and eukaryotes. Given that, I wonder if the same
historical environmetal changes would not have better produce 'cancer'
type of eukaryotic genome in terms of survivability as they are similar
to prokaryotes and low-level eukaryotes rather than the fragile complex
multicellular organ specialised organism.
There are many types of niches in the world-- some of which can be
filled by single celled organisms, and others by multicellular organisms--
it is not either/ or.
Dr. d' has sent in three messages, which I'll let you extract a question
I don't think Dr. Lynn answered my question about how can acquired
info. can be incorporated into genome for transmittal sans the involvement
of germinal cell. Way back in 1960 when I was working at SKI (Biophysics
Dept.) on the transformation of fibroblasts into Rous Sarcoma cancer
cells I found how an acquired Viral RNA could be incorporated into
Dna and transmitted by mitosis to daughter cells. Later on Rubin
and Temin won novel prizes on my data. :-)
Somehow Miss Lynn is avoiding the question as to the mechanism of somatic
mutations finding their way into inheritable DNA without using the
I am not trying to resuscitate Lysenko or Lamarck but Miss Lynn has
not explained how somatic mutations can find their way into inheritable
DNA avoiding the germinal cell classical route.
Dr. D-- RNA is an interesting and exciting area of work now-- I didn't
realize that you were asking about that. Yes, a retrovirus can get
into the DNA> Also, there is a LOT of speculation, now that RNAi
and other small RNAs have been discovered that perhaps they can affect
the germ line. In fact, it is known that they cause, at least, epigenetic
changes that can affect gene expression for more than one generation.
So in this line of speculation, assuming the RNA is made in response
to some property of the environment, it then brings some information
into the DNA. If that is what you were asking about, the whole area
of RNAi is a fascinating one, as is the question of retroviruses
[and the subject of an interesting science fiction book, Darwin's
Radio], but I do not know of clear examples where this actually has
changed an inherited property of the genome in response to a specific
environmental event. I have been focused on presenting specific,
strong examples, to encourage people to think about the power
of natural selection
Hm. Grammatical Man didn't suggest that to me! :) However, there is
increasing interest in applying linguistic analyses - grammatical
analyses - to genes. What do you think of that effort?
And a great question to look to the future. Yes, I am very interested
in that. It would help to identify new types of information, including
the information that I have suggested. But I would look not only
at T, G, A, C as we write them, but at the physical properties, such
as hydrogen-bonding patterns, in space, which is what the proteins
and other molecules see.
Again, in response to RBH's question, there is a great deal of information
that we have not yet thought is there--- rather than simply ask about
what we know, trying to understand the "language" of DNA
is certain to be a very rewarding line of investigation. [end]
ISCID would like to thank Lynn Caporale for tonight's chat. It was
quite fun and insightful.
Thank you all for your questions, and to the moderator for hosting
© by International Society for Complexity, Information, and Design