6. Behavioral Genetics I

[MUSIC PLAYING] Stanford University. [SIDE CONVERSATIONS] OK, let’s get going. Let’s see. First off, apologies for Friday. Sometimes, what seems
like a flawless set of connecting flights on paper
turn out not to be in reality. So hopefully, people made
good use of the time. OK. We are ready for our next
bucket, our next bucket, our third one in this
series, just in time to convince you that we
just keep jumping these. As soon as you get accustomed
to one approach, here we are, yet another
one, where we are going to trash everything
that’s come before us. OK. What have been the two
broad approaches so far? The first one, the
sociobiological/evolutionary psychology one–
behavior evolves exactly under the same sort
of wind tunnel of selection as does the heart of
a giraffe, blah, blah, that whole song and dance. Following certain
rules, you could generate fairly
structured predictions about social behavior. And then, the we-win
version of using that is, here’s what we predict. Here is the way we explain
this complex system of social behavior,
using these rules which assume certain degrees of
heritability of behavior, following certain
rules of evolution. And until you come up
with a better explanation for how this goes on, we win. This is how behavior works. Then we shifted over
to the molecular end. And what we saw was, on a
certain sort of trendy level, molecular biology is the
answer to the people who would sit around to
the sociobiologists and say, show me the genes. Show me the genes
for what you keep talking about inferentially. And what we saw was how
evolutionary change works out in the DNA level. Very importantly,
the common mechanisms for microevolutionary
change and how that supports the picture of gradualism. And then seeing all of these
relatively new, unexpected ways in which you are getting
big time changes in DNA, all these amplifying
effects of macroevolution, and suddenly support for
punctuated equilibrium in two ways. The first one being,
as soon as you get one, little, tiny mutation
affecting transcription factors, affecting
splicing enzymes, affecting entire
networks, it is going to be real rare
that you don’t get a mutation that’s a disaster. Stabilizing. Stasis, that’s the
long-term feature of the punctuated equilibrium. The equilibrium part
and the punctuated seeing how, under circumstances
of extreme selection, something that is come up with
is wildly adaptive, wildly advantageous, quickly
fixates in the society. And thus, we have
step functions. OK. So what we do today–
and Wednesday is now shift to another field–
thinking about behavior in a genetic context,
in a heritable context, this whole field of
behavior genetics. And what you’ll see is,
depending on your stripe, either these are a bunch of
really powerful approaches for being able to infer some
sort of genetic components to behavior, often getting
a whole lot of insight into what’s going on,
and in some cases, a more negative view,
more critical one, the entire field is gibberish. And a way to summarize the
view of it as being gibberish was this great cartoon
I saw a while back. OK, two scientists are
standing around the lab. And they’ve got
their lab coats on, and their test tubes, and going
about doing science stuff. And one of them is
saying to the other one, you know how sometimes you’re
on the phone with someone, and you’ve been talking
for a long time, and it seems like they
decide they want to get off, but they don’t want to
say they want to get off, so they say, I probably
shouldn’t keep you any longer, even though you’re not the
one wanting to get off, because they’re the one
who wants to get off? Well, I think I found
the gene for that. [LAUGHTER] And that winds up being one of
the criticisms of the behavior genetics. Once again, this world of
inferring genetic bases to behavior in often the most
deterministic possible way and using techniques which
are often complete nonsense. And what we’ll see are the
tools of behavior geneticists, and all the criticisms,
and why this winds up being a very contentious field. OK. So this being another version
of getting at, how do you know when a behavior
has a genetic component. And we’re immediately
allowing ourselves to be a little more subtle
here, not determined by genes, not determined purely by nature. All of that merely to
have a genetic component, a genetic influence, how
do you begin to do that? And this is a world where
people, the scientists, don’t sit around coming up
with evolutionary models. They don’t sit around trying
to find mutations in genes. They have a very different
sort of strategy. They look for patterns
of shared traits among the individuals who
have differing degrees of shared genes and
infer, related to this, and infer genetic
influences from that. OK, what’s the simplest most
mindless version of doing this? The basic initial
rules, centuries ago, when behavior genetics
started , was the notion, if you see a trait that
is universal in a species, obviously, it’s genetic. Obviously, it’s hard-wired. Obviously, it’s instinctual. Obviously, that’s an
extremely limited approach. Yes, indeed. You will see some
species, like flies, in which certain behavioral
traits are determined by relatively small numbers
of genes and are universal. This is obviously
going to fall apart when you get to something
more interesting than a fly. So what’s a much more
the starting point for the whole field
is to say, ooh, look. Here’s some behavioral
traits that run in families. Genes run in families. Therefore, we may have
just found evidence for a genetic influence
on these traits. And of course, it
takes about an eighth of a second to come up with the
retort to that, which is, yes, genes run in families, but
environment runs in families as well. So the standard
response is to now put a little bit more
strictures on it, to say, OK, it’s not just
that genes run in families, but as we know, from a few weeks
ago, the more closely related you are, the more genes
you share in common. Our Mendelian rule
of half the genes with full sibling, et cetera. Following thus, the logic that,
if you have a behavioral trait that becomes more common
the more closely related two individuals are, now
you’re inferring something about genetics. And of course, the
problem there is not only, as you become more closely
related to somebody, do you share a greater
percentage of your genes, you share environment more. Obviously, you are sharing
environment on the average, in most cases, much
more with a sibling than with a first cousin, than
with an eighth cousin. The trouble is shared genes
and shared environment tend to co-variant families. So that greatly weakens what
was the initial classic approach to the field 70, 80 years ago. OK. So you’ve got to come up
with something fancier, something more informative,
something more subtle. And what you wind
up doing then is, let’s control for environment. Yes, obviously,
the eighth cousin is living in a different
world than your full sibling. Let’s control for environment
under circumstances where you examine
relatives where they have the same
environment, and they differ in terms of the
amount of genes they share. And what is this
classic approach? Look at identical twins
versus fraternal twins, monozygotic twins, from one
zygote, monozygotic, identical twins versus dizygotic. And the general notion there
is, OK, identical twins share 100% of their genes. Fraternal twins share
50% of their genes. So if identical twins are raised
in the exact same environment, and fraternal twins are raised
in the exact same environment, they all have
environments shared. What’s the only
difference then to explain any behavioral
differences, if you see a trait that is
shared to a greater extent in a pair of identical
twins, than in fraternal twins? What’s the only difference? They all grow up with
identical environments. It’s because the identical
twins have twice the amount of genes in common, 100%. You could thus infer a genetic
influence on that trait, because it’s the
identical environment. Twins get raised
absolutely the same. And the only difference
is the degree of relatedness, the
degree of shared genes between monozygotic
and dizygotic twins. That one should take about
2/8 of a second to demolish, for the simple
starting point that, oh, yeah, sometimes fraternal
twins, dizygotic twins, are different sexes. OK. So that complicates things. So you come back and you’re a
little bit more rigorous now. And you restrict your
comparisons of monozygotic to dizygotic, to
same-sex dizygotic pairs. And then you
institute this rule, OK, same-sex twins, whether
identical or fraternal, are raised essentially
in the same environment. The same environment. So if you see a greater
sharing of traits among the monozygotic
twins than the dizygotic, the only place that greater
sharing could be attributed to is the fact that they
have more genes in common. Ah. We have just identified
a behavior that has a strong genetic component. So the big problem
with that approach is– anyone– who’s
a twin in here? Whoa. That’s a lot of hands. Identical twins? Dizygotic twins? Any triplets? OK. Just to extend that,
armadillos always give birth to four
identical offspring at once. [LAUGHTER] OK. [LAUGHTER] So that having been prompted,
what we then move on to is the obvious problem
with this entire approach. Hurray. Monozygotic twins get raised in
virtually the same environment. Dizygotic twins
of the same gender get raised in virtually
the same environment. That’s not true
in the slightest. There is far more
differentiation of environment for dizygotic twins than
for monozygotic twins. They are treated
differently, as pairs. Thus, if behavioral
traits are more in common between monozygotic
twins than between dizygotic, it could be because these guys
share more genes in common, and/or because they share
more environment, a completely flawed approach in that regard. The next complication. It turns out, monozygotic
twins don’t always have the exact same environment. Well, yes. Obviously, one of them gets
sneezed on at a higher rate by a kid sister, or something. And that’s going to produce just
decades worth of consequences. But even in other
domains, monozygotic twins don’t get treated exactly the
same, in some circumstances. And here is what it looks
like, if I had drawn it. OK. So you’ve got your twins. And this is during fetal life. And there are fetuses in there. And OK, there’s one with
a little thing of hair. And there’s the other. OK. And what you have
is a placenta forms. And here’s the deal. With monozygotic twins, if they
split during the first five days after conception,
they wind up, each one, with
their own placenta. On the other hand, if the
split occurs between days five and 10, there’s already
commitment to one placenta, which they then share. So monochorionic
pregnancy, or bichorionic. And what does that
wind up meaning? In the monochorionic,
the two fetuses share a blood stream,
to a greater extent, than in the bichorionic
circumstance. In these cases, it’s
separate blood flow from mom. OK, it’s still the same
mom, and it’s still ultimately the same
blood, but it’s going to be subtle differences
in the levels of stuff in the bloodstream. With the monochorionic, the
environment for these fetuses are much more similar,
in terms of whatever is carried in the bloodstream. OK, great. That’s a great factoid. And that’s like interesting,
things about identical twins. But like different
levels of nutrients, whether it’s coming off
of a blood vessel here, versus there, OK, maybe
that makes a difference. But these are going to
be minuscule differences. Nothing interesting. As one example of it,
monochorionic identical twins have more similar IQs than do
bichorionic identical twins. Just one demonstration that
this could make a difference. This is relevant. The IQs are more similar. And this has been one reason
why marmoset monkeys– marmoset monkeys, pair
bonding, so you know what they’re all about–
among other things, marmoset monkey mothers
always give birth to twins, because there is
two parents around to take care of them. And it’s got a completely
screwy circulatory system for the twins that are
unlike either of these. And there’s all sorts
of people who’ve been interested in
marmosets over the years– the ones who couldn’t afford
to by themselves armadillos– interested in terms of looking
at the differences in blood flow during fetal
life there of twins. OK. So we start off here that,
just because you see something more in common in
monozygotic twins than dizygotic, that
doesn’t tell you the first thing about
whether it’s genes involved. They share our
environment much more. Then even trying to
make sense of what’s going on in monozygotic. What you’ve got there
are fundamental potential differences,
environment, starting at a very important
early state of life. So monochorionic
versus bichorionic. What else then? How about another one where you
can look at a behavioral trait, looking at traits and
traits differences, according to some
genetic trait and where environment is the same? How about gender differences? Whoa. OK. So genes have something
to do with producing what your gender is. And environment being
the same, if you see differences between
females and males, that’s attributable to
the different genetics. OK. And you should be able to nuke
that one within seconds, as well, which is the
notion of shared environment, the notion
of identical environmental experience. To give you a sense
of how subtle this is, at one hour of life, on the
average, the level of activity, the rate of movement, the
amount of movement of limbs, on the average, is
greater among newborn boys than newborn girls. Whoa. That’s there like within
an hour of getting born. That’s not a whole lot
of time for environment having gone on there. Maybe you’re seeing a
strong genetic effect. What other studies have shown
is, from the first moments of post-natal life,
mothers are already interacting differently with
baby girls than with baby boys. From the first interaction, from
the very first holding of them, there’s differences in
how long they are held. There are differences in
proximity to the body, to the face. So Whoa. Sex differences in
behavior at one hour. Sex differences in environment
within mere moments after being born. So that weakens that one. And this came through
in another realm. This was a study that was
done in the ’80s, that was enormously influential, by
a pair of scientists at Johns Hopkins, Benbow and Stanley. And it had to do with a program
that I’d bet a lot of you guys had something or
other to do with back when, which is the Johns
Hopkins Gifted Youth Program thing, which I bet
all sorts of you guys qualified for
at various points and got the Johns Hopkins
blue ribbons pasted to your forehead. And that was part of
this massive study that’s been going on for decades
and decades of kids who are very gifted,
academically, and in a number of different realms. And Benbow and Stanley were
some of the senior researchers on this. And this was a study
they did when they had 40,000 kids in their database. And they asked a
very simple question, which was, well, what
does IQ distribution look like in this population as
a function of what sex you are? And back came a very
interesting finding, which was a highly, highly,
highly significant difference in the average IQ
between girls and boys, in the direction of
boys having a higher IQ. Moreover, when you looked at
the tail of the distribution, the highest IQ
range, what you saw there was approximately a 13:1
ratio, way out at this extreme. They then followed up and did
the exact same thing– no, they didn’t do IQ. What am I talking about? They did math. Ignore that. Erase all of that. OK, let’s start over. [LAUGHTER] So lots of you probably got the
Benbow and Stanley sweatshirts when you were part
of that program. So they were looking at math
skills, yes, junior high school kids who were taking the
math part of the SATs. That’s part of getting
into this Hopkins program. So they took the math SATs and
they looked at those 40,000 scores and saw that
there was a gender difference in the average
score on the math SATs, with boys scoring higher. Not only that, but
when they looked out at the tail here– OK, I’m back
on track– when they looked out at the tail here at the
highest math achievement, there was a ratio of
13:1, boys to girls OK, so that was pretty striking. And in their paper, which
was published in Science, was a very, very
critical phrase. Their rationale for
doing this study on kids that age was the
fact that– at least, at the time in most
schools, middle school, junior high school– kids have
not differentiated classes yet. They’re not yet
at the point where you could choose to take
extra math classes, to choose to take the absolute minimum. This was still the stage where
most kids in this country are getting the exact
same math classes. Everyone is still
getting the same. And thus, at that point, they
could include in their paper a phrase along the lines of,
since all the children had essentially identical
educational environments, any gender differences
seen in it are reflecting– and they used the
word, “biological”– are reflecting
biological differences. This was one hell
of a famous study. This was front page
all over the place. This wound up being shown in
a very large article in Time, and in Newsweek, the
Reader’s Digest– which I know, sort of,
everyone is supposed to make fun of– but the
Reader’s Digest, at least– does it still exist? Yes. Yes, it does. OK. Throughout the 1980s,
the Reader’s Digest was the most widely
read magazine in this country,
which is astonishing, showing what percentage of
people read in the bathroom, so having to go with the
Reader’s Digest there. And the Reader’s Digest covered
this and used the phrase, “The math gene,”
and discussing how this was the definitive study
showing that more boys have the “math gene.” Like, you already know
that’s like nonsense on so many different levels. But this was all over the place. This was the study
that definitively showed genetic differences
in math skills by sex, and definitively showed that
this was manifest at a stage before there were different
educational environments, in terms of math. And what, of course,
completely rips apart that study– and it was shameful
that thing was ever published, let alone got as
much attention as it did– is the fact that the
environment was not exactly the same. Endless number of studies
have shown, beginning by first grade, if
it is a simple math problem at that stage for
the same hands put up, a boy is more likely to
be called on than a girl. Studies showing that, for
the same correct answer, boys in elementary
schools are more likely to be praised for the
correct answer than are girls. By the time junior high
school is coming around, guidance counselors are
already differentially, by sex, advising, once
you get into high school, to take more elective math. Tremendous, massive
differences in environment. And this study, using
this whole argument of, if there’s an
identical environment, and you see differences,
it’s due to genes. And predicated on
that, if that’s not true in our afterlife,
it was ludicrous that they were making this
argument about 13-year-olds. So this was a paper
with enormous influence in every major newspaper
in the country. Scientists have discovered
the “math gene.” And boys have them
more than girls. We will come back to
that study later on, because something much, much
more interesting was going on. OK. So those were the limits in
the sex difference end of it. So what can you do next? There’s a flip side,
another approach, that the behavior
geneticists use, which is basically the converse. Now, you look at
individuals getting raised in the same environment
who don’t share genes. Same environment,
different genes. Rather than just now
the different genes, same environment nonsense
about gender differences, even down to monozygotic
twins, all of that, but now the flip
side, same environment and different genes. What was the paradigm for this? The one that is used over,
and over, and over again, the standard approach in
this part of the field is adoption studies. You take someone who
is adopted as a child, and they are now raised
in a household of people who they are not related to,
their adoptive, non-biological parents. And what you now
begin to look at are patterns of shared traits. Specifically, what
is looked at is, when you see a trait in
someone who was adopted, who are they more likely
to share that trait with? Their biological
parents, or parent? Or with an adoptive parent? Now, the logic of this is
completely straightforward. And this has been sort of the
standard paradigm in animal studies of the genetics
of behavior for centuries, something called
cross-fostering. You take a newborn litter
and another newborn litter, and you switch them
between moms so that they were raised by different moms. And they’re raised with, thus,
someone they’re not related to, but someone who they now have
environmental shared with them. Or in the litter
cross-fostering studies, you will take half of
a litter and switch it to another mother, half– so
you see how the iterations there go. So that was essentially
the rationale for how this would
be done in humans. And this was the
basis of one study in the late ’60s,
early ’70s, which was the landmark study
in genetic psychiatry, in behavioral genetics as a
whole, a study carried out by a guy at Harvard,
named Seymour Kety. And this was a phenomenal study
and phenomenally important. Here’s what Kety did. Kety was dealing with the
notion at the time of making sense of schizophrenia. And as we will see when we get
to the schizophrenia lecture, the number of nutty
ideas out there as to what the cause of this
disease is is just staggering. But what he was
interested in was getting at the notion that
was kind of floating around in some corners of the
field at the time, which is schizophrenia has a
biological component, a genetic component. And what that was was
viewed as very unlikely. But what Kety did was
try to go and test this. So what’s he going to do? He’s going to look at
adopted individuals who are schizophrenic
and see are they more likely to share that
trait with a biological parent or adoptive parent. You see the logic already. OK, how many
schizophrenic adoptees are you going to find
out there where you also were able to figure out who
the biological parents were. This was not an easy task. And Kety’s insight,
his intuition, was to go to one of the places
on Earth where that was most easily done, which
is Scandinavia, where all the Scandinavian
countries keep records like you cannot believe about
everything on Earth. People understanding,
for example, how the age of
puberty onset in girls have been decreasing
for centuries. They’re always using
Scandinavian data, because every single thing
that could possibly be recorded has been written down and
stored someplace there in moth-proof vaults. And what he was able to do in
ways that no human subjects committee on earth would
approve these days, is go through the entire
database of adoptees in Denmark and identify all of the cases
where somebody adopted had wound up with
schizophrenia, and then able to go back and see
who the biological parents and the adoptive parents were. Was a staggeringly large study. He and a team of psychiatrists
spent years in Denmark doing this, because one
of the things they did was they then did follow-ups. And they themselves
interviewed all of the potentially
schizophrenic individuals on there, so that
they would have the same diagnostic standards
all across the board. As we’ll see in the
schizophrenia lecture, it’s a very squishy diagnosis. So having the same
psychiatrists interviewing every possible person,
hugely important control. This is part of what took
them years and years. And then they finally
were able to say what the patterns
of similarity were. And here’s what they
wound up saying. You take any random
person on the street who is schizophrenic, and you
take any other random person off the street, and you
ask what are the odds that the second person
shares this trait with the first person? About 1% likelihood. What’s that another
way of stating? In the population
as a whole, there’s about a 1% incidence
of schizophrenia. So you start with
the circumstance where the biological
parents, neither of them had schizophrenia, and neither
of the adoptive parents do. And what’s the incidence in
this population among adoptees? A 1% schizophrenia rate. That’s just average
people off the street. That’s the usual rate across
Western European population. So 1% rate. Now, let’s have the
person growing up in a different
sort of household. This is a person who
has no biological legacy of schizophrenia,
but was brought up in an adoptive household
with a schizophrenic parent. And what you saw
then was a 3% chance, which, with this
enormous sample size, was a highly significant,
reliable number. So in a rough kind of way,
being raised in a household with an adoptive parent
who is schizophrenic approximately triples your risk
of a schizophrenia diagnosis. But then, the really
critical one, which is you were raised in a household
where neither parent, adoptive parent,
is schizophrenic, but you have a biological legacy
among your biological parents of schizophrenia. What do you see? A 9% incidence. Approximately a three-fold
increase above that, almost a 10-fold difference
now over what you see in the general population. This one number was what
roared through the field. This was viewed as
the clearest evidence to date for a genetic component
to a psychiatric disorder. Regular old person off
the street, 1% rate. Have a biological parent
with schizophrenia and share no environment with
them, because you got adopted away, and almost 10-fold higher
chance of getting the disease. Then, final thing, looking at
the incredibly rare people who got screwed on more different
fronts than you can imagine, who had a biological parent
with schizophrenia, and phew, got out of there, and landed
in an adoptive household with a schizophrenic parent. [LAUGHTER] So you get the
double whammy there. And what you saw
was a 17% incidence. What’s interesting
about that number? OK, what this cell does,
what this number does, is reflect the
increased risk by having a biological legacy
of schizophrenia, plus the increased
risk of having an adoptive environmental
legacy of schizophrenia. So let’s see. What’s the difference
between 1 and 3? That’s 2. So the difference between
1 and 9, that’s 8. So thus, it should be
about 10 percentage points. It should be about a 10% rate. What are you seeing here? A synergism. Get yourself a biological
legacy and get yourself a schizophrenia
household to grow up in, and it is not just adding
the two degrees of risk. There was a synergism, a
non-additive synergism. That is an important hint
for us of stuff to come. So this was this landmark study. This was phenomenally
difficult to have pulled off. It got Kety a number of
Nobel Prize nominations. This was the study that showed
the first definitive modern science evidence for
a heritable basis to a psychiatric disorder. And this became
the gold standard for how to do behavioral
genetic studies. And in the aftermath
of that, people began to do adoptive
studies on heritability between biological and
adoptive households, heritability of depression,
heritability of alcoholism, heritability of criminality. And you can see heading
off in all sorts of interesting directions
from there all sorts of interesting ones, and
them always producing a number that would be higher
in this cell of the matrix than that cell, and always
producing the notion that one has just shown a
strong genetic component to whatever that trait is. That trait– IQ. That trait– criminal behavior. That trait– alcoholism. These were some
very loaded studies, in terms of the implications. So what is the problem
with that approach? A number of problems. The first one is that, under
the best of circumstances– best of circumstances, best
of circumstances for people trying to publish
papers out of this– under the cleanest
of circumstances, the individuals in the
study were adopted away, were taken from the biological
parents, a quarter of a second after being born. No postnatal shared
environment whatsoever. What one knows, of
course, is that’s not the case with adoption. And there’s varying amounts
of lag time before it occurs. And they were never able to
factor that into the analyses. OK, so environment– so
you’ve got two and a half days worth of environment with
your biological parents at the beginning there. OK, so that’s a confound
“give me a break, though.” Just a couple of days, that’s
going to explain differences like these? What, of course, is
also linking around in there is the
huge, whopping topic of prenatal affects, prenatal
environment shared with mom. And we’re about to
see in a little while some absolutely
astonishing realms in which prenatal environment
has very long-term effects. OK, so that’s a huge confound. So how could you
control for that? You see a trait that an
individual shares in common with a biological parent,
despite being adopted away a second after
birth, all of that. And thus, you can infer
there’s a biological, there’s a genetic
component to this trait. Uh-oh. Wait a second. Shared environment
with the mother, that may explain some
of the shared traits. How do you get by that then? The difference in the likelihood
of sharing a biological trait, a trait with a
biological mother, versus a biological
father, is the measure then of the prenatal effect. If a trait is shared 10% with
the biological father and 17% with the biological
mother, the 7% is attributable to
prenatal effects. That was the general
conclusion that the field made at that point for dealing
with this irksome little problem, this pesky little thing
of prenatal environment, which is going to come back big
time in a few minutes. That was how they
were distinguished. So some more problems
with that approach. One is one that absolutely
tortures behavior geneticists the world over, one
which makes them just want to have people
being inbred strains where they could keep them in
cages and keep track of them. The problem being that, often,
the guy saying he’s the father ain’t actually the father. Oh, issues of
paternity uncertainty. That sure screws up
your genetic studies, if you’re trying to attribute
stuff to someone who turns out not to be related. Very much higher rates than
the people at Reader’s Digest would have you believe, the rate
at which the person claiming to be the father is not
actually the biological father. OK, that’s a bummer. That makes things
more complicated. One other big confound
in the adoption approach. And this was something that
was emphasized by a guy at Princeton, Leon Kamin,
a psychologist there, doing a very good job of
showing that adoptive family placements were non-random. When a child is adopted,
you don’t sit there and close your eyes
and spin the globe and put your finger
down in some place, and two minutes later this kid
born in, like, Shaker Heights, is running around with some
camel herders in Rajasthan. This is not done. It is not random placement. Instead, what is a policy in
virtually every adoptive agency in this country
is to try to match the kids, as much as
possible, along a number of different domains. In other words, you are also
sharing a lot of biology with the adoptive parents. And that completely
screws up the analyses. Adoption is non-random,
how it is done. One does not just
spin the globe. And instead, there are
very intentional attempts to try to match for certain
traits, traits which have genetic influences on them. OK. So that’s a big problem. So the adoptive approach
had tons and tons of interesting findings,
enormously influential. But over the years, people
have realized, more and more, prenatal effects,
paternity uncertainty. And from day one, being pointed
out that adoptive parents have higher than random rates of
shared genes with the adoptees, in most cases, in this country. OK. So what becomes
the next approach? And this one wound up
being the gold standard, the high watermark, of how
to do behavior genetics. Behavior geneticists who are
able to do this sort of study, the rest of the
behavior geneticists hate them, because they’ve
got the best toys out there to play with. And they’ve got the
coolest things going. And they’re always
snotty, because they’ve got the best possible
circumstance, the most perfect thing you can imagine, which
is identical twins separated at birth. Whoa. That must be one hell of an
experiment to have pulled off, identical twins
separated at birth. It turns out, every now and
then, a pair of identical twins are adopted very
soon after birth where each one is adopted
into a different household. Perfect. Perfect. Different environments,
different households, identical genes, you could
not possibly ask for something better than that. People, like, wet their pants
when the identical twins separated at birth paradigm
burst on the scene. This was so wonderful. People wrote poems
about identical twins separated at birth, sonnets. It was the best. It was so informative,
because look at the power of this approach. This is an incredibly
powerful approach. The two individuals are raised
in different environments, but they’ve got the
exact same genes. So the question
becomes, where are you going to find identical
twins separated at birth? And there’s one obsessive
geneticist, behavior geneticist, University of
Minnesota, Tom Bouchard, has done an entire career not
only studying these folks, but God knows how he
finds these people. But he first started
publishing when he had 40 pairs of identical
twins separated at birth re-united in adulthood. And that made for some
incredible, bizarre, heartwarming stories
of discovering your long-lost identical sibling
and produced a mountain’s worth of data. It was a bizarre
literature from day one. First one being, that you would
get the perfect case there. And what these guys would be
reporting in the literature was totally nutty stuff. OK, so you’ve got one of these
pairs of identical twins. And they’re born. And Wolfie winds up being
raised in Uruguay by neo-Nazis. And Shmuel gets raised in
Israel by his highly orthodox whatevers. [LAUGHTER] And then, as a result of
a game show quirk of fate, they’re suddenly
brought back together. And there is Wolfie and Shmuel,
who are identical twins. And what do they report? The most amazing thing
they have in common, they both flush the toilet
both before and after they go to the toilet. [LAUGHTER] You think I’m being facetious. Go back to that literature
when that first came out and the coverage in the press. And it would be things like,
Wolfie and Shmuel, they both have, like, a
poodle named Fluffy. And the flushing the toilet
before and after going to the bathroom, that was one
of the landmark early findings of these studies. They would find twins
that would do that. They would find twins who
were both married to somebody named Congolia, or something. [LAUGHTER] And they’d, oh, my god. This is totally amazing. That was what hit
the pages, initially, these obscure, little,
bizarre similarities within a backdrop
of, well, what’s the data actually showing? And what’s by now a twin
registry of probably a couple of hundreds sets
of these twins– and this has become the cottage
industry of the best behavior genetics around– what has
come out of that literature is the most solid, reliable
findings is about 50% heritability of IQ. About 50% heritability
of where you are on the introversion,
extroversion scale, and about 50% heritability for
degree of aggression. That’s kind of interesting. And what we will see is
there’s all sorts of problems with this approach as well. First one being,
starting right off, back to that Kamin
guy from Princeton, his critique– turns
out there were not a whole lot of
Wolfies and Shmuels. Even though they got adopted
into different families, there again was the non-random
placement in families, more similar environments
than one would have anticipated purely by chance. So that is a confound. OK, so what’s the
solution for that one? I know. Let’s look at monozygotic
twins separated at birth and reunited on Oprah
at age 50, and then look at dizygotic twins
separated at birth and reunited after the
commercial break on Oprah, and see what similarities are. And if you see more things in
common with the monozygotics, rather than the
dizygotics, you’ve just controlled for the
non-random placement in the adoptive homes. The extent to which
the monozygotics have traits more in common
than the dizygotics, that reflects the identical genes. That was the interpretation. That was a very powerful
sort of analysis, one that, nonetheless,
winds up being very limited. Because in this case,
because of tiny sample size, it’s really hard to
have done those studies. What’s another feature of
the whole behavior genetics approach? Here’s another one. If you see traits that
occur, behavioral traits, in the absence of any
learning, in the absence of any environmental experience,
in the absence of anything that can count as being
non-genetic, if you see that, you’re looking at a
genetic influence. And what are the examples
that are always given? The fact that all babies
all over the universe start smiling, and they use
the exact same set of muscles. And they always start smiling
socially around, roughly, the same age. And what you also
see– I don’t want to know how these were done–
but these classic photographs that you can get filming
of kids in utero– I don’t know what fiber
optic something or other was doing that– but
the demonstration that fetuses smile. Fetuses smile during
the third trimester. This is a motoric pattern that
is shared among all humans. OK. Well, social smiling,
you see the flaw there, which is that is very subject
to shaping of behavior. You are three months
old, and you’re watching all this social smiling
going on around you, and how the mannequins they have
at home are nowhere near as interesting as the
animated faces there. And you sort of try
it out yourself. What you see it’s the exact
same developmental time course for smiling among
congenitally blind babies, babies who never
see anybody smiling. So you see the
motor pattern here being something that is
arguably fairly universal and occurring in the absence
of any sort of training, how you go about smiling. The other example
that’s always used is with congenitally
deaf babies. And what you get there is
another universal, which is beginning to babble
at the exact same age that hearing kids
begin to babble, and the same argument
being made there. Of course, you have a very
uphill task there of ruling out any environmental similarities. Because once again,
once again, an area that has been utterly
under-appreciated in this whole field
is the whole world of prenatal
environmental effects. And the theme that’s going to
come out of that is environment does not begin at birth. And some environmental
effects prenatally are enormously
influential forever after. And if you showed up
on the scene one second after that individual
was born, all of the tools of modern
behavior genetics would tell you that
you were looking at a genetic trait,
where it is one instead that was brought about by
the prenatal environment. OK, let’s take a
five-minute break, and then we will
pick up on that. –good ones. First one being going back
to that gender difference business, the assumption
running through the field that, OK, boys and girls are
raised in the same environment by their parents. And the only thing that
differs is the genes. Like even those people,
behavior geneticists, believe that could be the case. No, they recognized this
was a very limited approach, and thus would limit
themselves to circumstances like the first hour of
post-natal life, which we already saw is a
flawed assumption, or under circumstances
where everybody’s had the same number
of math problems to take in their first
12 years of life. And we see the
problem with that. Nonetheless, there
was the recognition that that was a very
limited set of tools for getting at these issues. The other useful thing
that was brought up was somebody pointing out
that, in the extended notes for behavior genetics, in the
second paragraph, somewhere in there, about the
fourth or fifth line, is the very clearly
typed out word, “unable.” And pointing out that,
actually, the word was supposed to be “able.” So you might want to
take a look at that and kind of keep
that one in mind. OK, perhaps I should
take a look at it also. Moving on. Moving on now to this
business, all of the approaches we’ve been seeing
about comparing monozygotic with dizygotic. By the way, with the
identical twins, 2/3 of them have monochorionic. 1/3 are split. And how many of you who are
identical twins absolutely know in your heart of
heart right now whether you were a monochorionic
or bichorionic twin? OK. Well, that didn’t
work very well. OK. So pushing on. What we see here is, with
all of these approaches, the adoption, the twins
separated at birth, the twins, mono versus
dizygotic, et cetera, et cetera, all of
those were predicated on one simple
assumption, which is, environment begins at birth. And that has been
completely destroyed in some incredibly interesting
ways in recent years. We have very vibrant
literature at this point. First way that it
can go down, which is your prenatal environment. What are you having as
a prenatal environment? Who are you sharing
environment with? Obviously, with your mother. You are sharing blood. You are sharing blood,
and thus the things that she is experiencing in
the world around her that make for a different environment
than the person standing next to her gets translated
into effects on the fetus. First domain, hormonal ones. Here’s one example of something
that you will wind up seeing. This was worked on by a
guy named, Fred Vom Saal, at University of Missouri. And what he did was
look at the fact that rats give birth to
litters of about a dozen kids at a time. And there’s some
circulatory system thing. They look like a whole
necklace of fetuses there. And the circulatory
system is such that everybody’s getting
blood, but you’re getting preferentially
the blood that is from the siblings
right around you. There’s some sort
of looping thing that occurs with the blood
system that looks just like that. And what you wind up getting is
you have a particularly shared blood environment with the
siblings on either side of you. And what he asked was
something very simple. You are a female rat fetus. And in one case,
you’re sitting there with brothers on each side. In another case,
one brother and one sister, or in the final case,
obviously, with two sisters on either side. And what you wind up getting
is a different hormonal environment. How does that
translate out later? The more male siblings you
have around you as a fetus, the later you’re going
to reach puberty. That’s interesting, suggesting
that very local endocrine effects here play out
in something like that. Also, it predicts how
estrogen levels are going to drop in you later in life. So this winds up being
one very interesting prenatal environment. Here’s another one. Here we have, in humans,
the age of one’s mother when she gave birth– and
extrapolating a little bit here at both ends, but just assume
this is kind of the age range. And what we see here is
the age of puberty onset in the offspring. And what is seen is very young
mothers and very old mothers have offspring who reach
puberty later than women in a more intermediate age. What does that
appear to be due to? Differing estrogen levels. Higher levels of estrogen
at this point in life– it’s actually not symmetrical. It’s skewed a little
bit this way– and that seems to be
the driving force on it. Whoa. The age at which
you reach puberty has to do with how old the
fetal sack was that you hung out for nine months? That has an influence? Absolutely. So prenatal effects. More. Another version of it. Suppose now the
hormone you’re getting inundated with through
the bloodstream is a stress hormone. A stress hormone–
glucocorticoids, we will learn all about
those down the line– a stress hormone,
because mom is stressed. What are some of
the consequences? For the same prenatal
stress, as an adult, you will have a smaller
brain– if you’re a rat. You will have a thinner cortex. You will have less
learning abilities. You will be more
prone towards anxiety. You will have fewer of those
benzodiazepine receptors that we heard about
the other day. You will have more of
a cognitive decline when you are a
doddering old rat. All sorts of stuff
will go differently throughout your entire life. But get this. OK, look at this mechanism. So you are a rat. And your mother
was stressed when you were a fetus back when. And you were marinated
in those glucocorticoids when you were a fetus. Your brain, overall,
will be smaller. There’s one particular
brain region, which I won’t mention
right now because it’s not really critical,
there’s one brain region that’s particularly hard hit. What does that brain region do? Among other things, it helps to
turn off the stress response. So if that part of
the brain is smaller, you’re not as good at blocking
glucocorticoid secretion at the end of stress. And somebody with a normal-sized
whatever, something stressful occurs, and they recover. And you do this instead. Because this mysterious part
of the brain is smaller, is not giving as much of a
negative feedback signal. And for people new to
endocrinology, that’s something you’ll be getting
in a couple weeks. The net result is, if this
part of the brain is smaller, you will have more lifetime
exposure to glucocorticoids. So what happens next? What happens next, in
addition, baseline is also elevated in these individuals. So the net result is a lot
more cumulative exposure. So you are a female rat. And you were in a mother
who was stressed prenatally when you were a fetus. And as a result, in addition
to all the other problems that you’ve got lifelong,
you secrete higher than expected
glucocorticoid levels. And eventually,
you get pregnant. And thus, your fetus
is going to be exposed to elevated
glucocorticoid levels and will be born with a
somewhat smaller brain, thinner cortex, et cetera, et cetera. What have we just shown? An environmental manipulation
on a pregnant female manifesting itself
two generations later in the grandchildren. And when this was first
described in the early ’60s, this was called the
grandmother effect. And eventually, it
was shown to go out about four or five generations. The magnitude the
effect would get smaller with each generation,
before it disappeared. But look at what this is about. This is inheriting a
trait that is not genetic. And this wound up being the
first example of what is now called, non-Mendelian
inheritance of traits, non-genetic
inheritance of traits. And all you’ve got going
here is prenatal environment. Extremely powerful observation. And what you also then have
is, your some researcher, and again, you come
along one second after the animal is born. And you wind up studying,
saying oh, look at this. This rat tends to have elevated
glucocorticoid levels, just like mom. And this rat tends to have a
thinner cortex, just like mom. And this rat– and
if you’ve never heard of prenatal
environmental effects, what’s the only conclusion
you could make? There are genetic
influences on these traits. So these non-Mendelian,
non-genetic transmission of traits are really,
really important. Next thing. What else is floating around
in mom’s bloodstream that gets shared, besides hormones? Nutrients. Nutrients. And this winds up producing
something extremely interesting as well. OK, it’s all a blur by now. Have we already talked
about Dutch Hunger Winter? It was in the video we
watched while you were gone. Oh. [BLOWS RASPBERRY] OK. [LAUGHTER] Well, now that you all know
about– what did I say? That was only like
100 people, though. OK. OK, for those of you who
didn’t– Dutch Hunger Winter, here’s the deal. If you were a fetus in Holland
during the winter of 1944, something very interesting
happened with you. 1944, Holland is still
occupied by the Nazis. Nazis are being
pressured on all fronts. In the process of losing,
they’re falling back. They’re losing a lot of
the land they’ve occupied. And what they decide
to do that winter is, because they need some
food and because they wanted to punish the Dutch for
beginning to be more openly in the resistance, what
they did was, that winter, they diverted all of the
food in Holland to Germany. And historically known as
the Dutch Hunger Winter. Essentially, people went
from a perfectly fine, healthy– amid the
context of a war– Western-European diet,
down to something like that from out of nowhere. 40,000 people starved to death
during the Dutch Hunger Winter. Something very
interesting occurred, if you were a third-trimester
fetus, during the Dutch Hunger Winter, because it only
lasted for the winter. The Allies came in,
liberated Holland after that. And it went to something
like this, something resembling a step
function of starvation for about three months. If you were third-trimester
fetus during the Dutch Hunger Winter, your body learned
something important, which is here is not a whole
lot of calories out there. During third trimester,
fetuses are in some way– and this is metaphorical–
deciding, learning how much, in the way of
nutrients out there in the world, how readily
do calories come in? How is the fetus finding out? By way of mom’s circulation. Mom is starving. And thus, much lower levels of
nutrients in the bloodstream. And the fetus, at that
point in development, is saying, metaphorically,
well, what’s it like out there in
this place I’m going to be heading to rather soon? What’s the nutritional
profile like? There’s not much in the
way of food out there. And as a result, the fetus
has– and the term used now is “metabolic programming.” There is metabolic programming,
or metabolic imprinting– notice the word
“imprinting” here being used in a completely
different sense than we’ve heard about
already– metabolic programming to produce what is called
a thrifty phenotype. The fetus, realizing
there is nothing out there in the way of plentiful
food, what it does is it programs its
pancreas to function in a certain way forever after. What does the pancreas do? The smallest smidgen of
food hits the bloodstream, and the pancreas is
pumping out insulin, which helps store all that
stuff and quickly scarf up every bit of nutrient
in the bloodstream and store it away,
because you’ve got to be as efficient as possible. And you’ve metabolically
programmed your kidneys, so that your kidneys are
incredibly good at retaining salt. Because along
with starvation, there’s going to
be salt shortages. So really hold on to salt.
And you are born with a body with a thrifty metabolism,
very good at retaining salt, and spectacular at storing
away every bit of nutrients that hits the bloodstream. So at that point, you go
back to this sort of diet. And you have that for
the rest of your life. And what has now been shown
with the Dutch Hunger Winter individuals, the ones who were
third-trimester fetuses then, as adults, they have a
19-fold increased incidence of obesity, hypertension,
diabetes, and what’s called metabolic syndrome. What’s that about? Their body has programmed
to be extremely thrifty with metabolism. And as such, it is
forever after storing away every of the
grotesque Westernized diets that we all wallow in. And what you’ve got then
is setting individuals up for a much, much higher risk
of these metabolic disorders. If you were a newborn at
the time, it didn’t happen. A newborn during
here, the metabolic programming is
already over by then. If you were a first-trimester
fetus, didn’t happen. The metabolic programming
hasn’t started yet. Second, the later part or
second to third trimester is when the programming goes on. And this was a
landmark observation. This was really important. Among other things,
you get the people who do not like the notion
of these subtle effects. And they’re saying, OK, sure. This can happen,
but this is subtle. 19-fold is not subtle. And this was the
landmark study that ushered in what is now called
a whole field of fetal origins of adult disease. Fetal origins of
adult disease, which a lot of realms
of medicine think is outrageous and couldn’t
possibly work this way. And it is popping up in
more and more domains. Elevated levels of stress
hormones during fetal life, increased likelihood
of anxiety disorders as an adult, independent
of post-natal environment. Other examples like that, all of
these being ones of programming around that time. And the Dutch Hunger Winter
one is the iconic example. You know, what we’re all
accustomed to is yeah, you study something
in, like, a planaria. And then you study it in a rat. And you study it in a monkey. And then you study it
in a college freshman. And finally, when it’s– now,
you can conclude something maybe, maybe about humans. This was first discovered in
this population of humans. This was no, is this
relevant to humans. Interestingly, there
was another population at the time that went
through something or other like that, which was people
in the city of Stalingrad who were under
siege by the Nazis and had essentially years
worth of severe starvation. They didn’t get a Dutch
Hunger Winter phenomenon, because the food
coasted off like this. And afterward, it
took years for it to reach a
Western-European average. You don’t get it under
those circumstances. It’s a step function like this. OK. So think about this now. So you were a Dutch
Hunger Winter fetus. And as a result, you have
a very thrifty metabolism. And 30 years later,
you’ve gotten pregnant. You’re having a
perfectly normal diet, normal intake of calories. But you’ve got this
thrifty metabolism. And as a result,
your body is really good at pulling nutrients
out of the bloodstream, because you secrete more
insulin than most people would. With your thrifty
metabolism, you are pulling a disproportionate
share of the calories out of the bloodstream. And thus, your fetus is getting
a disproportionately smaller amount of calories. And thus, your fetus is
born with a milder version of the Dutch Hunger
Winter phenomenon. And this has now been
shown in the grandchildren of Dutch Hunger Winter fetuses. This is that exact same deal. This is non-Mendelian,
non-genetic transmission of traits, multigenerationally. Absolutely astonishing
that this could work. And the biology is
all in place for it. What we will see in a little
while is what the mechanism is, and it’s been identified down
to the molecular level of what went on in these Dutch Hunger
Winter– OK, I’ll give it away. Remember that epigenetic
business the other day? There are epigenetic
changes in the gene’s coding for things related to insulin in
the Dutch Hunger Winter babies. So enormous effect. Another example of it,
another dietary one, which is, if you are
female– and female human, among other species, this
has been shown in– there’s an issue of how much estrogen
does your mother consume when you were a fetus. Where’s estrogen coming
from in the diet? From a lot of different types
of plants, phytoestrogens. And what the studies show
is that increased exposure to estrogen derived
from phytoestrogens during fetal life, and there
is a small but consistent increased risk of
estrogen-dependent breast cancer, 20 years
later, 90 years later. Again, very subtle prenatal
environmental effects playing out forever, even unto
the generations after you. Another realm of prenatal
effects, learning. OK, this sounds ludicrous
right off the bat, that you can have
learning prenatally. You have learning prenatally. You can show this first in rats. Here’s what you do. You take a rat fetus, and
you can inject into it a particular flavor of
water along with sucrose. And the fetus absorbs it. The fetus actually drinks
amniotic fluid, which I find to be deeply creepy. But nonetheless, you
inject this stuff in there. And you are doing this a
number of days running. And this fetus is now
drinking this flavor that has a lot of sucrose in it. It tastes good. It tastes good? You’re a fetus. What do you mean,
it tastes good? The fetus learns about it. Because after birth,
given a choice between two neutral
flavors, it will prefer the flavor it was
exposed to that it was drinking when it was a fetus. How weird is that? More ones– more ones? OK, that was good grammar. More ones. Here’s one from humans. And this was a study,
which is as strange as you can get,
looking at the fact that the diaphragm is
a very good resonating membrane, something or other. Mother’s voices are heard in
the fetal sack quite readily. And if I don’t know how they
got the fiber-optic camera in there, I have no idea
how they got the microphones in there for recording that. But it is a very resonant
chamber, the diaphragm and the amniotic fluid. Fetuses hear a lot of
what’s going on out there, most notably mom’s voice. So this was a study, a
totally inspired one. And what you had
was, in this one, pregnant women spent
their last trimester loudly reading either
The Cat in the Hat over and over, like
four times a day or something, until they
went mad, absolutely mad– and that’s before the
kid was even born– or reading some random
collection of sentences that controlled for word length,
that controlled for rhythmicity, all of that. Then, you get the
newborn some time later, and you give them a test of
which they prefer to listen to. How do you do that
with a newborn? Something you can do is,
when newborns like something, they make more sucking
motions with their mouth. So you’ve got a sucko-meter
thing in there measuring it. And the newborns prefer
hearing The Cat in the Hat. They learned it. Not a huge effect,
but nonetheless, what is up with that? Of course, a follow-up question
that maybe half the people in here would be wondering is,
well, what about the fathers? And they have the study with the
fathers reading Cat In The Hat and, you know, like a megaphone
on mom’s belly and reading. And it doesn’t work. It doesn’t resonate
enough that side. Sounds from the mother. OK, so what we see here is this
whole world of stuff going on before behavior geneticists
show up and say, environment is just started. All these prenatal
effects of hormones, nutrition, sensory stimulation,
amazing, in some cases, multigenerational. Yeah? Did they catch up with
those children later, The Cat in the Hat
children, and see if they had a tendency to rhyme more? [LAUGHTER] Where are they now? Remarkably, all of them are
heads of states of countries. [LAUGHTER] People are still trying
to understand that one. But very good
follow-ups on that. So we’ve got this punch
line here, over and over, showing the power
of prenatal effects. One final study. And this was one carried out by
a scientist at Berkeley named, Darlene Francis, which took
some amazing surgical skills. So there are different
strains of rats that have been bred for
different levels of anxiety. And we’ve already heard about
one possibility for that. It turns out some
genetic differences in the promoter to the gene for
the benzodiazepine receptor. There’s all sorts
of strains that have been bred for high,
low levels of anxiety that people have studied. They’ve been bred. These are transmissible traits. These are heritable,
these are genetic traits. Here’s what Darlene
Francis did, which was she did an adoption study. What’s the adoption paradigm
we’ve heard already? Right after birth, you
cross-foster the rats, or the kids get adopted. That’s not the
adoption study she did. She transferred fetuses. She figured out how
to do the surgery to remove a fetus early on
in development from one rat mom to another rat mom where
they developed perfectly normally, once she
had this surgery down. And you know what turns
out to be the case? It’s not a genetic trait. It was not a genetic trait. You grew up with
the anxiety levels matching the strain
of your mother, even if the strain of the
individual whose body you fetally developed in– you take
a mom from the high anxiety strain, and you take a mom
from the low anxiety strain. And you take fetuses from
the high anxiety mom, and they go through gestation
in the low anxiety mom. And as adults, they
are low anxiety. It was not a genetic trait,
it was a prenatal one having to do with another one
of those multigenerational begatting by having early
experience influencing the nature of the pregnancy
you would eventually have, and thus, influencing
your fetus, producing a different pregnancy
for them down the line. This demonstration
of cross-fostering, of adopting, as early
as you possibly could, enormously important study. One really difficult one. Yet another wave pounding
on this point, environment does not begin at birth. And some of the most
important environment is not occurring
starting at birth. And everything about
behavior genetics, classically, was
predicated on there’s no environment before that. OK, so what do they
come back with? What’s the response? We’ve already seen
a possible way of controlling for that, which
is, if you see traits shared in common with the
biological father, and then you see
more traits shared in common with the biological
mother, the increased degree of sharedness with the
mom reflects, not the genes, because you’re getting
the same amount of genes from each parent. It reflects the
prenatal environment. That would be the control that
would be used in these studies. The extent to which
a trait is more shared with a biological mother
than with a biological father is a reflection of
prenatal effects, because you get the same amount
of genes from each parent. Naturally, this turns out to
be vastly messier than this. Because there’s a
whole world in which you’re getting more genetic
influences from your mother than from your father. OK. First one is, once
again, if it turns out the person who’s claiming to
be the father isn’t actually the father, that kind of
changes the whole map. And once again,
that is a problem running through all of human
behavior genetics literature. But here, here’s the next one. Here, we have– just make sure
we’ve got our cliche in place here– what is this? This is the– OK, which is the
what of the cell? Powerhouse, The powerhouse. OK, here we have the powerhouse
of the cell, mitochondria. And in this rather odd cell,
there’s one mitochondrion, but it is standing in for
all of the powerhouses. And we’ve got mitochondria. Something that, when you
first learn about this, is just flabbergasting,
which is, here you’ve got a cell with
its DNA, and its nucleus, and its double helix
thingy happening there. And it turns out
that mitochondria have their own DNA. And this is part
of explaining one of the, like, truly amazing,
adventurous, nutty ideas that have turned out to be true. Scientist, named Lynn Margulis,
University of Massachusetts, 30 years ago or so, she
noted that this business of mitochondria
having their own DNA, and came up with this hypothesis
that mitochondria used to be independent organisms. That in some symbiotic
whatever, billions of years ago, got into cells that had
no mitochrondia at the time, and that there’s been
a symbiosis ever since. Mitochondria have
DNA, which are related to mitochondrial function. Not a ton of the
stuff, nonetheless, you have every gene that is critical
for mitochondrial function. No, that’s not true. A few of them have
wound up in here. But all of the genes
in here are pretty important for
mitochondrial powerhousing, all that sort of thing. And these are derived from
a completely different world of DNA than these. So now consider this. This cell is an egg. This cell is not an egg. This cell is just merely
carrying genetic information that looks like that. What you’ve got is sperm,
all they are carrying are the DNAs, the
genes, the DNA. And they don’t have cytoplasm,
the fluid-y environment in a cell. This is one major
dense packing job. Here’s the critical implication. Eggs have mitochondria,
sperm don’t. So right at the point
of fertilization, you have gotten all of your
mitochondria from your mother. And thus, all of
the genes related to mitochondrial
function that are contained in the mitochondria
you don’t get from fathers. It is exclusively
inherited from mothers. Mitochondrial DNA solely
comes from the mother. So what we’ve just
seen as a first pass is it’s not an even 50/50 split. You get a disproportionate
share of your DNA coming from the mother. Really important. And one that’s been
used by all sorts of evolutionary geneticists
to trace legacies. If these mitochondrial DNA is
only passed along female lines, that allows you to figure
out all sorts of stuff about evolution. It has given rise to the Eve
hypothesis, that somewhere back then was a woman, probably
some sort of early hominid, who is ultimately the ancestor,
the great, great, great, great, great, grandmother of every
single human on Earth. And it can be traced through
the mitochondrial DNA. So that’s an asymmetry. Next one. Next source of asymmetry
is back to that business about imprinted genes. From a couple of weeks
ago, you remember that one, genes that
are working differently, depending on which parent
you are getting them from. So that one’s a violation
as well of this rule that you get all of your DNA in
equal amounts from each parent. This is another thing that
possibly skews the ratio. Now, here’s another
very interesting thing. So you’ve got the egg here. And not only does it have this
cytoplasm with mitochondria floating around,
but in addition, there’s other stuff
floating around in there, like transcription factors. Sperm don’t have
transcription factors. Sperm, all they’re doing is on
this, like, suicide swimming mission there. And they’re not
making any new genes. All they’re doing is
this one, long spurt of racing for the end. And in the case of
the eggs, though, you have transcription factors. You have all sorts of
proteins in the cytoplasm. You’ve got a fully functioning
cell, instead of this, sort of, much more streamlined version. All the transcription factors
that come in a fertilized egg are coming from the mother. So what does that
wind up meaning? Transcription factors,
those are proteins. We’re not talking
about genes here. We just saw how
you get more genes from your mother
than your father, these mitochondrial genes. OK, but transcription
factors, the father has genes for
transcription factors. The mother has genes for
transcription factors. What’s the significance of
getting your transcription factors from your mother
in the fertilized egg? OK. Consider here two genes. The first one codes for
transcription factor A. We know that already. Transcription factors
are typically proteins, so they have their own
genes all the way down. And this codes for gene
X, whatever that is. So here’s what you’ve got. There are promoters responsive
to transcription factor A. Transcription factor A turns
on the synthesis of the protein coded for by gene
X. And in addition, transcription factor A turns
on transcription factor A gene. It’s a positive feedback
loop where it makes more and more of the stuff. That’s the way this particular
transcription factor works. So suppose this is
the only thing that can activate transcription
of gene X. So suppose you’ve got some
environmental event which, as a result of it, knocks out
the activity of transcription factor X in an egg. The egg is fertilized. And as a result of transcription
factor A not being expressed, it doesn’t express more. This is the only thing that
drives more expression. And you never make gene X. Now, somewhere along
the line in your body, you are soon making eggs
which contain these genes, of course, but
where you have never expressed transcription factor
A, because this was knocked out in the egg. So because of that, you
never express gene X. And that egg gets fertilized. And you pass on that
trait to your offspring. You pass that on, this
acquired trait of transcription factor A not working. If this is the
circuitry that you have, it doesn’t matter which
version of gene X you get, you are never going to express
that gene for generations and generations, forever,
if this is the loop. What’s going on here, what a
lot of people think is relevant, are some environmental
toxins that are known to
disrupt the activity of certain
transcription factors. And what that does is
induce heritability in a non-genetic way of
non-expressing of a gene. The gene’s being inherited,
but it will never ever be expressed. What have you just acquired? A Lamarckian trait. You remember Lamarck. Everybody learns about Lamarck,
in order to mock him viciously. And Lamarck had the notion that
the way evolutionary change works, the way
inheritance works, is you experience
something, and it causes a change in your body. And as a result, you pass
on that acquired trait to your offspring. Ludicrous. Lamarckians have been mocked
and pilloried for centuries, except in the Soviet Union in
the 1930s, where it gave rise to Lysenkoism, a very horrific
piece of genetic history. But what you’ve got is
this complete trashing of the notion of you
acquire some trait from the environment, and you
pass it on to your offspring. This is Lamarckian inheritance. This is an environmental factor
that knocks this transcription factor out of action. And if this is the
wiring that you’ve got, this gene will
never be expressed. And as a result, it will never
be expressed in your offspring, in your grandkids, et
cetera, all the way down. This is Lamarckian
inheritance of a trait. And again, where the best
evidence for this has been is with environmental
toxins that knock out, that have some of these
mutating effects in eggs. They are not mutations
in a classical DNA sense. But nonetheless, they
are now heritable. So that pops up also. So have we got here? We have the simple
assumption that, if you see more sharing
of a trait with the mother than with the father,
that’s reflecting prenatal environment. And what we’ve seen
here is totally messing this up is the
fact that you do not get equal genetic
influences from each parent. You are getting more
genetic material, you are getting more
genes from your mother, because the mitochondrial DNA. Even if you are getting
equal amounts of DNA, expression of them will
have different consequences because of imprinted genes. Finally, in this
world, having nothing to do with the amount of
genes or the actual DNA, you can have this
Lamarckian inheritance of traits due to
environmental perturbations. What we see here
are ways ranging from extremely subtle and rare
to some rather substantial ones with the mitochondria where
you are not getting equivalent inheritance from both parents. So that confuses things a lot. OK. So after all of that,
you do, nonetheless, get circumstances where
behavior is influenced by genes. And by every rule,
it could be shown. And by the most
contemporary of techniques where people find
the gene and the DNA, and they’ve traced out the steps
in showing that there really are genetic influences
on behavior, and ones that withstand every
single one of these criticisms. OK. So sometimes, you’ve
got genes regulating or genes influencing behavior. But now we bring in a
whole other possibility. And this is something
that was emphasized by a psychologist named, Judith
Rich Harris, a number of years ago, in a very important book
of hers, called The Nurture Assumption, which has a
lot to do with arguing the relative importance
of influences of peer versus parents. Nonetheless, she
focused in one section of the book on the
genetics of behavior and focused on what she calls
indirect genetic effects. What would indirect
genetic effects be? OK. You’ve got that
trait that I referred to before, one of
the most reliable of traits in the
identical twins separated at birth business, 50%
heritability of where you are on the introversion,
extroversion continuum. OK. So that one has
held up pretty well. Amid all the possible complaints
about these various approaches, that one appears
to be quite solid. And you’re immediately
off and running with, OK, genes for
extroversion, for sociality, for all of that. What she shows instead is
something else is happening. There’s a very, very heritable
trait from parent to offspring, one of the most heritable
physical traits out there, which is your height and
your appearance in general, that those are highly
heritable traits. And suddenly, you
have a phenomenon that is well-documented,
which is people who are taller are treated
better and considered more attractive, comma,
he says bitterly. [LAUGHTER] What you’ve got is people
are treated differently along those lines. And what is known
also is, people who are treated more
positively during development, during childhood,
become more extroverted. What we have here is not
heritability of the trait where you are in
the introversion, extroversion continuum. What you have is heritability
of a physical trait, which causes you to be
treated differently in the world, which then brings
about changes in personality. And studies have
since shown that most of the heritability of the
introversion, extroversion is mediated by physical
traits in between. So that’s a completely
indirect way in which you could have gotten to this. More cases. More cases of this. Let’s see. OK, you can show, in various
bird, turkey, hen species, that there are chicks, that
there is heritability of rank. You could be born
to a low-ranking mom in the pecking
order, high-ranking, all the perfect
studies, and controls, and cross-fostering,
all of that, and there is
heritability of rank. But another indirect
genetic effect that was subsequently
demonstrated, which is it’s not the rank that’s
being inherited, it’s a particular
version of genes related to melanism of
your feathers, the color and iridescence
of your feathers. And it turns out, if you have
a certain color pattern, all the other like
roosters, and chicks, and hens, and poultry,
peck at you more often. And you’re reduced
to subordination. This is not inheritance
of a social dominant trait or a social subordination trait. This is inheritance
of a gene having to do with the color and
iridescence of your feathers, which wind up producing
your social rank. Another example. Here’s another one. OK, back to chicks
again, which is chicks appear to be intuitively
able to peck at grubs shortly after birth. That they’re able to
peck down and get grubs. And by all the rules
of behavior genetics, with all the constraints
and criticisms answered, this appears to be
a heritable trait. But it turns out that this
is not what is heritable. What is heritable, bizarrely, is
the tendency of newborn chicks to find their toes to
be very interesting and to peck at their toes. And they quickly learn that this
doesn’t feel all that great. But if you do it in somewhat
more sloppy of a way, you get one of
these things that’s squirmy that tastes good. They start off, what is
genetic is the tendency to peck at your feet. God knows why. If you are a newborn
chick, because of extremely elegant
high technology studies in which newborn
chicks are put in galoshes or something and
they don’t pack, and they don’t show the
seemingly innate ability to peck for grubs. So here, we have
a behavioral trait which, in fact, is indirectly
mediated by something else. More examples. There is, by now, a literature
showing approximately 70% heritability– and I
keep using this word. We are going to dissect the
word, “heritability” big time in a short while–
There’s about 70% heritability of political party
affiliation in this country. Sharing that behavioral
trait with your parents. Whoa. What is that about? That’s sure disturbing. And that sure makes
you want to procreate in the name of your political
stances, or whatever. And this appears to
hold up pretty well to some of the standard
criticisms in the literature. And it’s got nothing
to do with this. What’s the mediating variable? A large really
interesting literature showing, when you compare
political or social progressives with political
or social conservatives, one of the most reliable
personality differences is how they feel
about ambiguity. Conservatives, on the average
do not like ambiguity. They are much more
ambiguity-averse. And you can start
it with showing ambiguous sensory
stimuli in kids and looking at heart
rate responses to it. And that tends to be a
stable personality difference at political extremes. You will see, some
time later on, that there’s a whole world
of moral development in kids where there’s various
scales measuring. One’s Kohlberg’s Stage
of Moral Development. There was an old
literature suggesting political differences as to
how fancy of a Kohlberg stage you got to. We will see that,
despite what struck me as the intuitive sort of
logic of what was found. That one hasn’t held up. But one that does
hold up is difference in ambiguity tolerance. And that’s probably
the mediating trait. It is not inheriting a tendency
to like elephants, or donkeys, or whatever. It is instead having
this intermediate trait. Final example of what would
be an intermediate trait. There’s a whole bunch
of rat and mouse strains that have been
developed that have differing levels of aggression,
high aggression strains. And you do all the
proper controls, and you can show that
this is a genetic trait. Whoa, heritability
of aggression. We’re suddenly back to twin
adoption studies and Kety with heritability of
criminality, all of that, heritability of aggression,
what’s actually going on in all of the strains identified
to date by spontaneous traits coming up and then
breeding for it. What you see instead
is the strain that is so aggressive, and
so pissy, and so impossible, and so constrained by the law
and order of rodent society, and all of that. They’ve got a lower threshold
for pain sensitivity. Things hurt them more readily. And they’re more likely
to become aggressive at that point. It turns out it’s genetic
differences in the neurobiology of pain sensitivity. So what we’re seeing
here, over and over is, amid the gazillion
of criticisms we’ve had about when does
environment actually start, and when do
environmental assumptions and being treated
the same go down the tubes, and differential
inheritance of genes, yeah some traits do appear
to have some fairly strong genetic components. But even once you get
that far, they very often are through some
very indirect routes. OK more things here. More things? We just covered that. OK, a little bit
more on epigenetics, which is that
whole business, you remember, from the
other day, which is after you’ve got
the DNA, and then you have that whole world of
protein coatings, which I’ve been very careful not
to give the jargony name for, because it’s not important. But what you’ve got
is this whole world where regulation is not so
much at the level of genes– and in 95% of the DNA,
that’s the on/off switches– but whether the transcription
factors can even get in there, and this world of
epigenetic changes that will cause lifelong
differences in how readily transcription factors
get to something or other. What have you got at that point? The exact same possibilities
as this one here. If instead of, due to
some environmental toxin, you knock a transcription factor
out of business in a fertilized egg where there’s the set up
of genes, if for some reason, whatever, you do not
have access to it because of an epigenetic
change, it’s going to be the exact
same consequence, multigenerational inheritance
of non-genetic traits due to epigenetic, rather
than genetic differences. That is turning out to be what
went on in the Dutch Hunger Winter people and
the animal models of epigenetic differences and
access of transcription factors to genes related to
insulin metabolism. That turns out to
be a critical one. Here is one of the coolest
examples of this to date. And this is work done by a guy
at McGill University, named Michael Meaney. And what he has focused
on is what started off as a very artificial
literature, which is, take yourself a newborn rat,
and for the first two weeks or so of its life,
every day, you pick it up for three
minutes and you pet it. And now, you put it back. And all else being
equal, it will have a bigger brain in
adulthood, better learning abilities, more resistance
to a whole bunch of neurological insults,
lower glucocorticoid levels, et cetera, that
whole world of what came to be known as
neonatal handling. On the other hand,
pick up the rat, take it away from mom
for, instead of 3 minutes, an hour and a half. Then each day, put him back. And as an adult, it’s going
to have a smaller brain and a shorter life expectancy. Three minutes away
from mom does wonders. An hour and a half of
being petted does not. That, in and of itself,
is interesting in terms of what counts as stimulation,
what counts as stress. OK, so hurray. What we’ve just
learned is just how generations of rat-petting
graduate students can influence the lineages of
rat brains and all of that. And what Meanie started
looking at with this phenomenon being one that was around
forever– first identified around 1960 by a
guy named Seymour Levine in the Psychiatry
Department here, and no longer alive– but
that started this whole world of neonatal handling. What Meanie did
was say, well, rats did not evolve, whatever
is going on here, for the purpose of
doctoral theses, what’s the natural equivalent
in the world of a rodent? And it turns out that what
happens when you pick up a rat for three minutes and
do this and put it back, mom is all excited and
goes and checks out the pup and nestles it, and licks it,
and whatever other stuff there. And it has all this attention. Whereas, if you take the pup
out for an hour and a half, when you put him back
with mom, mom basically ignores the pup for
long periods of time. You’re changing the
mother’s behavior. OK, so that’s an
indirect effect. And what he
proceeded to show was the critical thing about
the handling was not what you’re doing to the
rat during that time, it’s the fact that
you’re causing dramatic changes in maternal
behavior based on that. So that’s interesting. But that still doesn’t
solve the problem of why did the system
evolve for grad students manipulating maternal behavior. And what he then
proceeded to look at was normal variation in
rat mothering styles, because some rat
mothers are– OK, I know this is a
value judgement– but some rat mothers are better
mothers than other mothers. Some rat mothers, they
simply are better. They’re better. They’re nicer they
have better souls. [LAUGHTER] And in these rat mothers,
how do you determine that by these sorts of measures? Licking and grooming. How much time do you
spend licking your baby? And how much time do you
spend grooming your baby? And what Meanie
proceeded to show is that’s what the neonatal
handling phenomenon was about. When you have moms who lick and
groom their kids an awful lot, what you do is produce the
same sort of better outcome. From the three minutes
of petting deal there, you get the kid who is
bigger, and healthier, and smarter, that sort of thing. Moms who hardly ever lick
and groom their pups, they produce pups
that, as adults, are like the ones
that were separated for an hour and a half a day. It is a reflection
of mothering style in the rats and the
variability there. Next thing he showed was that
this was multigenerational. If you lick and groom your
baby rat daughter a whole lot, as an adult, she will be
more of a licker and groomer. And he’s already shown what some
of the neurological mechanisms are for that. For development,
what have we got? Yet again, one of these
non-Mendelian inheritance of traits deals going on. In this case, not even prenatal. Your early experience is going
to cause lifelong changes in your brain, which
will make you more likely to reproduce the
same early experience for your offspring. Off you go. The final thing he
did, which stands as a landmark in the field
of behavioral neurobiology, is he figured out what
the epigenetic change is. One of them, or
rather two of them is identified by now,
what gets changed by how mom often or
un-often licks you, grooms you, all of that? You change the access
of transcription factors relevant to activating
genes for making receptors for stress hormones, making
receptors for estrogen, making receptors for a whole
bunch of different hormones. Showing the epigenetic
changes there, that’s how you go from moms
differing maternal style to lifelong differences
in expression of all sorts of genes. How’s this? What you wind up seeing there
as this permanent mechanism, it is also reversible,
what he has since shown, which is you have a
baby rat who spends the first half of its
infancy with some totally terrible, negligent, distracted
mom who pays no attention and doesn’t do any licking. Now cross-foster the pup
to a more attentive mother, and you can change the
epigenetic pattern. So all of this has
two themes going on. Early experience, causing
really persistent differences in how this stuff
works long after, and experience later on having
the potential to reverse some of this. All of this stuff, once again,
would be mistaken for genetic. What we have here
is what appears to be a genetic style of what
sort of mother rat you are. And it’s not genes,
it’s the mothering style setting up the offspring for
being a similar type of mother. Incredibly important
studies demonstrating this. What remains unclear
is how you get from mom licking you to
something epigenetic happening here. His crew is pounding
away at that. OK, so what have we
got at this point? We have gone through
over and over here– where have we gotten
to at this point? We’ve gone over and over here at
all of the classical techniques in the field of
behavior genetics, does it run in families,
adoptive studies, identical versus
monozygotic twins, twins separated at birth. We saw all of the
problems with it. And most dramatically,
most excitingly these days, prenatal environmental effects. We saw that trying to separate
maternal prenatal effects from paternal
genetic effects hits a wall, when you get all
of these weird-o hereditary things, including potentially
non-genetic Lamarckian inheritance of a trait. And what we’ve seen is how this
stuff playing out early in life has multigenerational
consequences. What we’re going to pick
up with on Wednesday is now looking at how
people in this business find the actual genes, and
ultimately, gene environment interactions that make the last
two hours basically irrelevant. OK. For more, please visit
us at stanford.edu.