WEBVTT

00:00:00.000 --> 00:00:07.000
So just trying to remind you that
the replication fork looks something

00:00:07.000 --> 00:00:15.000
like this where 5 prime to 3 prime
and 5 prime to 3 prime.

00:00:15.000 --> 00:00:22.000
This is what's known as the leading
strand because DNA,

00:00:22.000 --> 00:00:30.000
the synthesis of the new strand can
go --

00:00:30.000 --> 00:00:37.000
Which is going 5 prime to 3 prime is

00:00:37.000 --> 00:00:43.000
going in the same direction as the
movement of the replication fork.

00:00:43.000 --> 00:00:49.000
The other strand, which is known as
the lagging strand,

00:00:49.000 --> 00:00:55.000
the DNA synthesis is actually going
backwards to the movement of the

00:00:55.000 --> 00:01:01.000
replication fork,
which means it has to go and then

00:01:01.000 --> 00:01:06.000
start up here and go again.
And it's continually jumping.

00:01:06.000 --> 00:01:10.000
And I told you that the little RNA
primer is used to start each strand.

00:01:10.000 --> 00:01:15.000
And then the DNA polymerase is able
to elongate that.

00:01:15.000 --> 00:01:20.000
And then at the end these little
nicks in here,

00:01:20.000 --> 00:01:24.000
the RNA has to be removed,
fill in the gap and then it's sealed

00:01:24.000 --> 00:01:29.000
up by the enzyme DNA ligase,
which we'll talk about when we talk

00:01:29.000 --> 00:01:34.000
about recombinant DNA.
Someone asked,

00:01:34.000 --> 00:01:38.000
I had mentioned why this strategy of
using RNA was beneficial,

00:01:38.000 --> 00:01:42.000
and that has to do with the fact
that the fidelity,

00:01:42.000 --> 00:01:46.000
which is going to be the next thing
I'm going to focus on of DNA

00:01:46.000 --> 00:01:50.000
replication as not,
you can get a much higher accuracy

00:01:50.000 --> 00:01:54.000
if you have the end of a primer
already there and then carry out the

00:01:54.000 --> 00:01:58.000
chemistry in there.
No enzyme has ever achieved the

00:01:58.000 --> 00:02:02.000
accuracy that you see in DNA
replication if it's

00:02:02.000 --> 00:02:06.000
starting a strand.
So RNA polymerase,

00:02:06.000 --> 00:02:10.000
which constantly starts strands to
make RNA copies,

00:02:10.000 --> 00:02:14.000
as we'll talk about,
is not as accurate as DNA

00:02:14.000 --> 00:02:18.000
replication.  And by putting a
little bit of RNA,

00:02:18.000 --> 00:02:22.000
because the cell has to start a new
strand.  Before it gets here there's

00:02:22.000 --> 00:02:26.000
no strand at all on this lagging
strand so it needs to make this

00:02:26.000 --> 00:02:30.000
little RNA primer.
It needs to make a little primer.

00:02:30.000 --> 00:02:34.000
And by making it out of RNA then it
can tell what doesn't belong there.

00:02:34.000 --> 00:02:39.000
It doesn't matter if it's not quite
as accurate as the rest of DNA

00:02:39.000 --> 00:02:43.000
replications because it's going to
take it out anyway and fill it in

00:02:43.000 --> 00:02:48.000
using the DNA polymerase.
And if you think about that maybe

00:02:48.000 --> 00:02:52.000
you can see one of the reasons that
the cell has chosen or nature has

00:02:52.000 --> 00:02:57.000
chosen through evolution to use
little RNAs to begin the strands.

00:02:57.000 --> 00:03:01.000
OK.  Well, in any case, the
fidelity of DNA replication is

00:03:01.000 --> 00:03:05.000
really pretty amazing.
Incidentally, just speaking of DNA,

00:03:05.000 --> 00:03:09.000
many of you wrote some very
thoughtful things about Vernon

00:03:09.000 --> 00:03:12.000
Ingram's visit.
I didn't give him a whole lot of

00:03:12.000 --> 00:03:15.000
warning and he had to go and change
his schedule and move meetings

00:03:15.000 --> 00:03:19.000
around in order to come to talk to
you.  And it was very nice of you.

00:03:19.000 --> 00:03:22.000
Many of you wrote some very
thoughtful things,

00:03:22.000 --> 00:03:25.000
which I'm going to pass onto him.
I want him to know that many of you

00:03:25.000 --> 00:03:29.000
appreciated his visit.
I also saw a lot of you reacted to

00:03:29.000 --> 00:03:33.000
his advice about crowded labs.
That has been my experience,

00:03:33.000 --> 00:03:37.000
too.  And one thing about the
scientific process is it's not just

00:03:37.000 --> 00:03:41.000
one person.  You're in with a group
of people, just as Vernon described,

00:03:41.000 --> 00:03:45.000
and that group of people becomes the
creative engine that drives all the

00:03:45.000 --> 00:03:49.000
science within that lab.
And so you're not only picking your

00:03:49.000 --> 00:03:53.000
project, you're looking for a group
of people to work with.

00:03:53.000 --> 00:03:57.000
And, as Vernon said, if the lab is
really doing hot stuff they tend to

00:03:57.000 --> 00:04:01.000
attract a lot of people.
So a crowded lab can sometimes be a

00:04:01.000 --> 00:04:04.000
really good indicator.
No absolutes, and there's an

00:04:04.000 --> 00:04:08.000
exception to everything,
but that was a good piece of advice

00:04:08.000 --> 00:04:12.000
he gave you if you're looking for
UROPs sometimes.

00:04:12.000 --> 00:04:16.000
OK.  So, anyway,
DNA fidelity.  Remember I said we've

00:04:16.000 --> 00:04:20.000
gone from, our bodies have somewhere
from like 10 to 20 billion miles of

00:04:20.000 --> 00:04:24.000
DNA in them if we could take all the
human DNA and stretch it out?

00:04:24.000 --> 00:04:28.000
But that fidelity is done at an
error rate of about one mistake to

00:04:28.000 --> 00:04:32.000
every ten to the minus tenth
nucleotides replicated.

00:04:32.000 --> 00:04:37.000
Which I said if you were typing all
the time it would be like sort of

00:04:37.000 --> 00:04:43.000
making one mistake every 38 years.
So it's an astonishing degree of

00:04:43.000 --> 00:04:49.000
fidelity.  Something that's beyond
anything within our experience.

00:04:49.000 --> 00:04:55.000
And there are three principles that
go.  One is polymerase is really

00:04:55.000 --> 00:05:01.000
good at the base pair recognition
telling that an A is paired with a T

00:05:01.000 --> 00:05:06.000
or a G is paired with a C.
And discriminating against

00:05:06.000 --> 00:05:12.000
everything else there's a phenomenon
known as proofreading,

00:05:12.000 --> 00:05:18.000
and I'll tell you how that works.
And then there's a third system

00:05:18.000 --> 00:05:24.000
called mismatch repair.
And all three of these contribute

00:05:24.000 --> 00:05:30.000
to this very, very low-frequency of
errors, one mistake for

00:05:30.000 --> 00:05:36.000
approximately every ten to the tenth
nucleotides replicated.

00:05:36.000 --> 00:05:40.000
So the first thing is I've pointed
out to you several times that if you

00:05:40.000 --> 00:05:45.000
draw the hydrogen bonds between an A
and a T base pair,

00:05:45.000 --> 00:05:49.000
the two hydrogen bonds or the three
hydrogen bonds between a G and C

00:05:49.000 --> 00:05:54.000
base pair, that the shapes of this
pair and that pair are virtually

00:05:54.000 --> 00:05:59.000
identical.  You can pick them up and
lay it right down on top.

00:05:59.000 --> 00:06:02.000
Now, if you actually look at it
you'll see you could draw some base

00:06:02.000 --> 00:06:05.000
pairs between,
for example, a G and a T.

00:06:05.000 --> 00:06:09.000
In fact, you can draw two hydrogen
bonds, which is the same as between

00:06:09.000 --> 00:06:12.000
an A and a T.  But the one thing I
hope you can see,

00:06:12.000 --> 00:06:16.000
just from the shapes even without
being able to see the individual

00:06:16.000 --> 00:06:19.000
atoms, is that a GT base pair
doesn't have the same shape as the

00:06:19.000 --> 00:06:23.000
correct base pairs.
So when I showed you that little

00:06:23.000 --> 00:06:26.000
movie the other day where this is
the template nucleotide,

00:06:26.000 --> 00:06:30.000
this is the incoming nucleotide and
there's this alpha helix

00:06:30.000 --> 00:06:34.000
that's swinging up.
What's happening in there is that

00:06:34.000 --> 00:06:40.000
the enzyme is checking the way that
the incoming nucleotide is the

00:06:40.000 --> 00:06:45.000
correct shape to go with the base
pair.  And you can sort of see it's

00:06:45.000 --> 00:06:50.000
flipping it right into a very narrow
little slot in the enzyme.

00:06:50.000 --> 00:06:56.000
So it's not only asking for sort of
hydrogen bonds, it's asking

00:06:56.000 --> 00:07:01.000
for the exact shape.
If you just did it by thermodynamic

00:07:01.000 --> 00:07:07.000
grounds you'd make about one mistake
in a hundred because that's about

00:07:07.000 --> 00:07:13.000
the discrimination between the
correct base pairs and some of these

00:07:13.000 --> 00:07:19.000
other ones.  This works so well.
You get more like one mistake in

00:07:19.000 --> 00:07:25.000
ten to the fourth or ten to the
fifth.  We're still quite a distance

00:07:25.000 --> 00:07:31.000
away from the ten to the tenth,
but this is one of the things.  It's

00:07:31.000 --> 00:07:37.000
looking for the correct shape
of the base pair.

00:07:37.000 --> 00:07:41.000
Now, the second thing that helps
with fidelity is a phenomena known

00:07:41.000 --> 00:07:54.000
as proofreading --

00:07:54.000 --> 00:08:00.000
-- exonuclease.  Things
called a nuclease.

00:08:00.000 --> 00:08:12.000
That means it can degrade DNA.
And the exo works at an end.

00:08:12.000 --> 00:08:19.000
And, furthermore,

00:08:19.000 --> 00:08:24.000
the directionality of this
proofreading was something that

00:08:24.000 --> 00:08:30.000
puzzled people initially because
it's going 3 prime to 5 prime.

00:08:30.000 --> 00:08:35.000
And when people started to purify
DNA polymerases or complexes of DNA

00:08:35.000 --> 00:08:40.000
polymerases involved in replication
there seemed to be a puzzle because

00:08:40.000 --> 00:08:46.000
the polymerase,
as I've told you,

00:08:46.000 --> 00:08:51.000
goes 5 prime to 3 prime,
but the same enzyme complex had an

00:08:51.000 --> 00:08:57.000
exonuclease that went in
the opposite direction.

00:08:57.000 --> 00:09:01.000
So this seemed very peculiar at
first in the sense if you were

00:09:01.000 --> 00:09:06.000
trying to polymerase DNA in this way
why in that same enzyme would you

00:09:06.000 --> 00:09:10.000
have something that wanted to
degrade DNA in the other way?

00:09:10.000 --> 00:09:15.000
And the answer turned out that this
was known as a proofreading

00:09:15.000 --> 00:09:20.000
exonuclease, as I've put up here.
And here's the principle of how it

00:09:20.000 --> 00:09:24.000
works.  Suppose you were replicating
the DNA and there was a G.

00:09:24.000 --> 00:09:29.000
And if you put a C in there it very
quickly goes on and continues

00:09:29.000 --> 00:09:34.000
the replication.
If it puts in a T,

00:09:34.000 --> 00:09:38.000
let's say, this is not a very good
base pair.  It wouldn't have the

00:09:38.000 --> 00:09:42.000
right shape.  So when the enzyme
came up looking for that 3 prime

00:09:42.000 --> 00:09:47.000
hydroxyl, which would be right at
the end of that T,

00:09:47.000 --> 00:09:51.000
things are not in the right place.
And so the polymerase activity

00:09:51.000 --> 00:09:56.000
slows down.  And as that primer
terminus, if it sits there for a

00:09:56.000 --> 00:10:00.000
little bit, it's able to just peel
off the DNA, flip up,

00:10:00.000 --> 00:10:04.000
and there's this function that does
just what you'd do if you were

00:10:04.000 --> 00:10:09.000
typing and you made a mistake.
You'd just hit the delete key and

00:10:09.000 --> 00:10:13.000
take off the last nucleotide that
you did.  And I have a little movie

00:10:13.000 --> 00:10:17.000
showing you that.
This is a crystal structure.

00:10:17.000 --> 00:10:21.000
This is the DNA template.  And the
polymerase catalytic activity site

00:10:21.000 --> 00:10:25.000
is right here.
And in this little movie it's just

00:10:25.000 --> 00:10:30.000
added an incorrect base pair and the
polymerase is sort of stalled.

00:10:30.000 --> 00:10:35.000
And the actual nuclease function is
physically separate on the protein

00:10:35.000 --> 00:10:41.000
structure.  But what you'll see in
the movie is that if the polymerase

00:10:41.000 --> 00:10:47.000
cannot go very well eventually this
thing will come up and it will chop

00:10:47.000 --> 00:10:53.000
off one nucleotide,
come back and try it again.

00:10:53.000 --> 00:10:59.000
Let's see.  I think if we do this,
oopsy-daisey.  Let me see if I can

00:10:59.000 --> 00:11:04.000
get this to work here.
Nope, it's not working.

00:11:04.000 --> 00:11:09.000
OK.  Well, anyway, I'm going to
skip it for right now.

00:11:09.000 --> 00:11:13.000
I don't want to waste time.
But, in any case, the end would go

00:11:13.000 --> 00:11:18.000
up here and it would take off one
nucleotide.  So there at least are

00:11:18.000 --> 00:11:23.000
two of the ways that the polymerase
is able to work with such fidelity.

00:11:23.000 --> 00:11:28.000
It selects for the correct base
pair shape.

00:11:28.000 --> 00:11:32.000
And then after it's done in addition
it sort of looks back,

00:11:32.000 --> 00:11:36.000
just as if you were a very slow
typist, and every time you typed a

00:11:36.000 --> 00:11:40.000
letter you looked back and said did
I make a mistake?

00:11:40.000 --> 00:11:44.000
And if you made a mistake then
you'd delete and then just try again.

00:11:44.000 --> 00:11:48.000
And that gets the cell another
maybe two orders of magnitude of

00:11:48.000 --> 00:11:52.000
accuracy.  So we're up to about one
mistake in ten to the seventh base

00:11:52.000 --> 00:11:56.000
pairs replicated.
The third system,

00:11:56.000 --> 00:12:00.000
which is called mismatched repair,
turns out to be very important for a

00:12:00.000 --> 00:12:04.000
whole variety of reasons.
And before I tell you about it,

00:12:04.000 --> 00:12:08.000
I want to first introduce the idea
of DNA repair in general.

00:12:08.000 --> 00:12:12.000
One of the things that's wonderful
about DNA --

00:12:12.000 --> 00:12:19.000
-- as you've learned,

00:12:19.000 --> 00:12:22.000
is it's got the information in two
copies.  It's in a complimentary

00:12:22.000 --> 00:12:25.000
form but it's like having the
photograph and the negative.

00:12:25.000 --> 00:12:29.000
And if your kid sister pokes a hole
with a pair of scissors through the

00:12:29.000 --> 00:12:32.000
picture of your boyfriend or your
girlfriend, you're not really in

00:12:32.000 --> 00:12:35.000
trouble as long as you've got the
negative because you can get the

00:12:35.000 --> 00:12:39.000
information back again.
And that same principle applies in

00:12:39.000 --> 00:12:43.000
DNA repair.  So if you have some
kind of lesion in DNA,

00:12:43.000 --> 00:12:48.000
and this might have come from going
outside in the sunlight,

00:12:48.000 --> 00:12:53.000
your DNA absorbs in the UV and it
undergoes photoreactions,

00:12:53.000 --> 00:12:57.000
they tend, for the most part,
to just effect one of the two

00:12:57.000 --> 00:13:03.000
strands of DNA.
Or if you smoke,

00:13:03.000 --> 00:13:09.000
which I hope none of you do,
there are many chemicals in smoke

00:13:09.000 --> 00:13:15.000
that will react with DNA,
and they'll modify one strand.

00:13:15.000 --> 00:13:21.000
And so what the cell has is a
system that has many kinds of repair

00:13:21.000 --> 00:13:27.000
systems, but it has a special type
of repair system known as nucleotide

00:13:27.000 --> 00:13:33.000
excision repair.
And you could think of this as a

00:13:33.000 --> 00:13:39.000
protein machine that constantly
scans the DNA looking for little

00:13:39.000 --> 00:13:44.000
distortions.  And if it finds it
then what it needs to do is it needs

00:13:44.000 --> 00:13:50.000
to make cuts, remove the DNA and
make a little gap.

00:13:50.000 --> 00:13:55.000
And now you can see what it can do
now.  Once it's got a little gap the

00:13:55.000 --> 00:14:01.000
information over here is
a complimentary form.

00:14:01.000 --> 00:14:05.000
So if a DNA polymerase were to come
along it could fill in that gap and

00:14:05.000 --> 00:14:10.000
seal it up and then you'd be back to
ordinary DNA, the lesion would be

00:14:10.000 --> 00:14:15.000
gone.  And I made a silly little
PowerPoint thing here to show it.

00:14:15.000 --> 00:14:20.000
So if you were to, say, damage the
guanine with something,

00:14:20.000 --> 00:14:25.000
say one of the carcinogens you find
in cigarette smoke,

00:14:25.000 --> 00:14:30.000
you could think of this protein
machine as being a sort of pair of

00:14:30.000 --> 00:14:35.000
scissors that have a conditionality
in them.

00:14:35.000 --> 00:14:38.000
As this protein machine scans along
the DNA the scissors aren't

00:14:38.000 --> 00:14:42.000
activated until it recognizes
there's a distortion here,

00:14:42.000 --> 00:14:46.000
at which point then it senses that
there's some bump in the DNA.

00:14:46.000 --> 00:14:50.000
And it's very cleaver the way it
does it because the nuclease

00:14:50.000 --> 00:14:54.000
activities, the things that are
going to cut the DNA are actually

00:14:54.000 --> 00:14:58.000
some distance away,
a few nucleotides away from the

00:14:58.000 --> 00:15:02.000
lesion.
So even if this is distorting the

00:15:02.000 --> 00:15:06.000
DNA, the scissors are able to work
out here and out here.

00:15:06.000 --> 00:15:10.000
It makes two cuts.  That was a huge
surprise.  Nobody expected that when

00:15:10.000 --> 00:15:14.000
they started to do the biochemistry.
And then in principle once you cut

00:15:14.000 --> 00:15:18.000
it now you can remove this little
nucleotide and then a DNA polymerase

00:15:18.000 --> 00:15:22.000
can just come in,
and following those A pairs with T,

00:15:22.000 --> 00:15:26.000
G pairs with C, copy it along and
then would seal it up to get to the

00:15:26.000 --> 00:15:30.000
end.  And I've actually shown you a
picture of what happens if a human

00:15:30.000 --> 00:15:35.000
is missing that system.
When I was showing you how profound

00:15:35.000 --> 00:15:39.000
an effect you could get from just
losing one single gene or a mutation

00:15:39.000 --> 00:15:44.000
affecting one single gene,
this disease called xeroderma

00:15:44.000 --> 00:15:48.000
pigmentosum.  They're a variety of
different groups.

00:15:48.000 --> 00:15:53.000
And the one on the left is an
example.  That's someone who is

00:15:53.000 --> 00:15:57.000
missing one of the genes that
encodes one of the proteins involved

00:15:57.000 --> 00:16:01.000
in nucleotide excision repair.
And this is really,

00:16:01.000 --> 00:16:05.000
really important for fixing up the
damage we get all the time in

00:16:05.000 --> 00:16:08.000
sunlight.  So if you miss that
repair system and you got out in the

00:16:08.000 --> 00:16:12.000
sun then you get all kinds of
lesions and people are very

00:16:12.000 --> 00:16:16.000
susceptible to skin cancer.
And I told you fortunately now you

00:16:16.000 --> 00:16:19.000
don't find people with this disease
looking like that because at least

00:16:19.000 --> 00:16:23.000
in developed countries we recognize
it.  They're kept out of the sun.

00:16:23.000 --> 00:16:26.000
And these were the kids who I said
are called ìchildren of the moonî

00:16:26.000 --> 00:16:30.000
because they, for example,
go to summer camps where they do

00:16:30.000 --> 00:16:34.000
everything at night so they won't
get exposed to sunlight.

00:16:34.000 --> 00:16:39.000
But that's what happens to us if we
miss that excision repair.

00:16:39.000 --> 00:16:45.000
And, again, what makes that
possible is that the information is

00:16:45.000 --> 00:16:50.000
there twice in a double-stranded DNA.
I also showed you a little movie

00:16:50.000 --> 00:16:56.000
early on when I was showing you,
I'm going to actually run this in

00:16:56.000 --> 00:17:02.000
QuickTime because it works a little
more smoothly, I think.

00:17:02.000 --> 00:17:07.000
So I showed you this when we were

00:17:07.000 --> 00:17:11.000
talking about DNA because I wanted
you to sort of get that sense of

00:17:11.000 --> 00:17:14.000
what it was like to kind of fly down
the groove of a DNA.

00:17:14.000 --> 00:17:18.000
But what I didn't emphasize was
this protein that was bound to the

00:17:18.000 --> 00:17:22.000
DNA.  That's a protein that's a DNA
repair protein.

00:17:22.000 --> 00:17:25.000
And it's one of these things that
looks for lesions in the DNA.

00:17:25.000 --> 00:17:29.000
And as we fly along the major
groove this little green thing is

00:17:29.000 --> 00:17:33.000
actually the lesion that that
protein is looking for.

00:17:33.000 --> 00:17:38.000
And it sort of puts fingers down
into the groove and it's able to

00:17:38.000 --> 00:17:43.000
sense that.  And you can sort of see
how this protein is bound to DNA.

00:17:43.000 --> 00:17:49.000
This is a lesion that we get all
the time from oxidative damage.

00:17:49.000 --> 00:17:54.000
And remember I said oxygen is bad
for DNA?  So our bodies have to have

00:17:54.000 --> 00:18:00.000
systems that are able to do that.
So DNA repair is very important for

00:18:00.000 --> 00:18:05.000
life.
We'll just finish flying down the

00:18:05.000 --> 00:18:09.000
major groove one more time here.
OK.  I'm going to go back to

00:18:09.000 --> 00:18:20.000
PowerPoint.

00:18:20.000 --> 00:18:30.000
OK.  So mismatched repair is a form
of repair that's got

00:18:30.000 --> 00:18:37.000
that same idea.
Let's think about it if we had a

00:18:37.000 --> 00:18:42.000
replication fork here,
and let's say there was a G here and

00:18:42.000 --> 00:18:46.000
the T got misincorporated,
but in this case it wasn't removed

00:18:46.000 --> 00:18:51.000
by the proofreading which happens
about one in ten to the seventh

00:18:51.000 --> 00:18:56.000
times.  Now if that strand is fixed
up, excuse me,

00:18:56.000 --> 00:19:01.000
is continued then you'd end up with
a GT base pair.

00:19:01.000 --> 00:19:05.000
And the next time you copied it this
strand would give rise to a GC but

00:19:05.000 --> 00:19:09.000
this one would give rise to an AT.
And then you'd have a mutation that

00:19:09.000 --> 00:19:13.000
now would have changed.
And if it affected an important

00:19:13.000 --> 00:19:17.000
gene that could be bad for you.
So the cell has what's known as a

00:19:17.000 --> 00:19:25.000
mismatch repair --

00:19:25.000 --> 00:19:29.000
-- that works in exactly the same
logic as here.  That it

00:19:29.000 --> 00:19:34.000
basically comes along.
It scans the DNA.

00:19:34.000 --> 00:19:41.000
It finds the bump because this is
not a proper base pair.

00:19:41.000 --> 00:19:48.000
And then it fills it in and you're
back to ordinary DNA with a GC base

00:19:48.000 --> 00:19:55.000
pair.  There's one little wrinkle.
For this system to work it has to

00:19:55.000 --> 00:20:03.000
do one other thing that's different
from that kind of DNA repair.

00:20:03.000 --> 00:20:08.000
Can anybody see what it is?
Why don't you talk to the person

00:20:08.000 --> 00:20:14.000
next to you and see if you can
figure it out.

00:20:14.000 --> 00:20:19.000
This system must be doing something
else in order for this to work.

00:20:19.000 --> 00:20:25.000
OK, you can ask somebody.  What do
you think?

00:20:25.000 --> 00:20:35.000
What if I removed the gene?

00:20:35.000 --> 00:20:43.000
Would that work?

00:20:43.000 --> 00:20:47.000
What would happen if I took the gene
instead?  Say I made the little gap

00:20:47.000 --> 00:20:52.000
over on this strand instead,
cut it here?

00:20:52.000 --> 00:20:59.000
Yeah.  So which one is the one

00:20:59.000 --> 00:21:03.000
that's right, the old strand
or the new strand?

00:21:03.000 --> 00:21:07.000
The old strand,
yeah.  See, this is the old and this

00:21:07.000 --> 00:21:11.000
is the new.  And the term that's
usually used, it's known as the

00:21:11.000 --> 00:21:16.000
daughter strand,
the new strand.  So the other thing

00:21:16.000 --> 00:21:20.000
this system has to do is it not only
has to be able to detect that

00:21:20.000 --> 00:21:25.000
there's an incorrect little base
pair in there,

00:21:25.000 --> 00:21:29.000
but it also has to know which is the
parental strand,

00:21:29.000 --> 00:21:33.000
the template strand,
and which is the daughter strand,

00:21:33.000 --> 00:21:37.000
the newly synthesized strand.
And this system makes the assumption

00:21:37.000 --> 00:21:41.000
that the strand that's old is the
one that's correct and the mistake

00:21:41.000 --> 00:21:44.000
is on the new one.
You guys see that?

00:21:44.000 --> 00:21:48.000
OK.  So that gets another two or
three orders of magnitude in

00:21:48.000 --> 00:21:52.000
accuracy and that's what
brings it up.

00:21:52.000 --> 00:21:55.000
Now, the people who made this,
who formulated this model for

00:21:55.000 --> 00:21:59.000
mismatch repair,
complete with the feature that it

00:21:59.000 --> 00:22:03.000
needed to recognize the old and new
strand, that's a bit of a trick,

00:22:03.000 --> 00:22:07.000
if you think about it because it's
DNA on both sides.

00:22:07.000 --> 00:22:10.000
And there are several different
ways used in nature,

00:22:10.000 --> 00:22:14.000
so I'm not going to go into it,
but there's at least a couple of

00:22:14.000 --> 00:22:18.000
different ways of doing that trick.
You could sort of see if you were

00:22:18.000 --> 00:22:22.000
the replication fork and you talked
to that you could certainly,

00:22:22.000 --> 00:22:26.000
just from the geometry of that,
if you wanted, you could probably

00:22:26.000 --> 00:22:29.000
keep track of who's old and new.
E. coli has a very cute trick,

00:22:29.000 --> 00:22:33.000
but it's not universal so I won't go
into it, but the people who did the

00:22:33.000 --> 00:22:36.000
seminal stuff,
I had to just quickly show you a

00:22:36.000 --> 00:22:39.000
couple of pictures.
When I showed you that picture of

00:22:39.000 --> 00:22:43.000
the DNA 50th, the guy sitting in the
front row was Miroslav Radman who

00:22:43.000 --> 00:22:46.000
was one of the two people.
He's a European scientist

00:22:46.000 --> 00:22:49.000
originally from Croatia.
And he collaborated with someone

00:22:49.000 --> 00:22:53.000
you've heard about before,
Matt Meselson, who was up at Harvard.

00:22:53.000 --> 00:22:56.000
And it was with the Meselson-Stahl
experiment that showed the

00:22:56.000 --> 00:23:00.000
semi-conservative mechanism
of DNA repair.

00:23:00.000 --> 00:23:03.000
This was a little reception.
And Matt was talking to Alex Rich

00:23:03.000 --> 00:23:07.000
who's in the MIT Biology Department.
And I was amused because remember

00:23:07.000 --> 00:23:11.000
how Vernon told you how Francis
Crick would run up and down the

00:23:11.000 --> 00:23:15.000
stairs in the Cambridge lab and he
was talking all the time?

00:23:15.000 --> 00:23:19.000
And I've heard Vernon say you could
never really tell whether an idea

00:23:19.000 --> 00:23:23.000
came from Watson or Crick because
they'd just talk,

00:23:23.000 --> 00:23:27.000
talk, talk all the time.
So this was at sort of nice

00:23:27.000 --> 00:23:31.000
reception at the DNA 50th.
And within a couple of minutes,

00:23:31.000 --> 00:23:35.000
I looked over and there were
Miroslav Radman and Matt Meselson

00:23:35.000 --> 00:23:39.000
talk, talk, talk.
They were in the corner drawing

00:23:39.000 --> 00:23:43.000
pictures on a board.
I also showed you actually a

00:23:43.000 --> 00:23:47.000
picture of one of the genes that's
involved in recognizing this

00:23:47.000 --> 00:23:51.000
mismatch, because there's a protein
that recognizes that mismatch and

00:23:51.000 --> 00:23:55.000
it's given the name of mute S.
And when I was showing you some

00:23:55.000 --> 00:23:59.000
proteins it had one that had a lot
of alpha helices.

00:23:59.000 --> 00:24:04.000
This is actually a picture of mute S.
It's a dimer.

00:24:04.000 --> 00:24:09.000
That's why some of it's green and
some of it's blue.

00:24:09.000 --> 00:24:14.000
And this is DNA viewed end on and
it's recognizing a GT mismatch in

00:24:14.000 --> 00:24:19.000
DNA in that picture.
Now, this may sound very esoteric,

00:24:19.000 --> 00:24:24.000
you know, and obviously important
for life and an important part of

00:24:24.000 --> 00:24:29.000
sort of understanding how life works
if you're interesting in studying

00:24:29.000 --> 00:24:34.000
molecular biology.
It may not seem to have very much

00:24:34.000 --> 00:24:39.000
connection to your real life.
But, in fact, in this case mismatch

00:24:39.000 --> 00:24:44.000
repair does because it affects the
frequency with which,

00:24:44.000 --> 00:24:49.000
if you lose it, then when you
replicate your DNA you're going to

00:24:49.000 --> 00:24:54.000
make more mistakes.
And I need to just give you a very

00:24:54.000 --> 00:24:59.000
quick introduction to cancer so you
can see why this is important.

00:24:59.000 --> 00:25:03.000
Cancer comes from the fact that
remember a human cell or a

00:25:03.000 --> 00:25:07.000
multi-cell like us that has many
kinds of different cells starts out

00:25:07.000 --> 00:25:11.000
from one cell.
And I talked about first you get

00:25:11.000 --> 00:25:15.000
the embryonic stem cells that can
become anything.

00:25:15.000 --> 00:25:19.000
And the cells become successively
more and more and more specialized

00:25:19.000 --> 00:25:23.000
as they go along.
So ultimately a cell that's in your

00:25:23.000 --> 00:25:27.000
retina or in, say,
the lining of your colon needs to

00:25:27.000 --> 00:25:32.000
know that's where it belongs.
And it also needs to know that it

00:25:32.000 --> 00:25:37.000
cannot just keep replicating.
So if this is actually showing a

00:25:37.000 --> 00:25:42.000
little picture of the lining of your
intestine.  And there's a single

00:25:42.000 --> 00:25:46.000
layer of cells right along the
inside edge of your intestines.

00:25:46.000 --> 00:25:51.000
This is the cells through which all
the nutrient exchange happens and

00:25:51.000 --> 00:25:56.000
everything else when your body
extracts nutrients as food stuff

00:25:56.000 --> 00:26:01.000
passes through your intestine.
And so what happens with cancer is a

00:26:01.000 --> 00:26:05.000
cell that's normally a part of your
body has to obey a whole set of

00:26:05.000 --> 00:26:10.000
rules.  And what you can think of
when someone starts to develop

00:26:10.000 --> 00:26:15.000
cancer is that what started out as
an ordinary cell undergoes some kind

00:26:15.000 --> 00:26:19.000
of successive changes in its DNA
that gradually causes it to forget

00:26:19.000 --> 00:26:24.000
the rules that make it be part of an
organized body system.

00:26:24.000 --> 00:26:29.000
So if we take a look here at all
these different cells.

00:26:29.000 --> 00:26:33.000
But let's imagine just one of the
gets a change that makes it forget

00:26:33.000 --> 00:26:38.000
to stop, or it should know to stop
replicating when it touches its

00:26:38.000 --> 00:26:42.000
neighbors, but if a cell were to
lose that control what would happen?

00:26:42.000 --> 00:26:47.000
Well, it would then begin to
proliferate.  And then what happens

00:26:47.000 --> 00:26:52.000
in cancer is the cell will,
now there are more of them, and one

00:26:52.000 --> 00:26:56.000
cell with acquire an additional
mutation that will lead to a further

00:26:56.000 --> 00:27:01.000
loss of growth control.
You can see now the cells are

00:27:01.000 --> 00:27:05.000
starting to become sort of funny
shapes.  And then one of the cells

00:27:05.000 --> 00:27:09.000
in here will undergo yet another
change.  And right at this point,

00:27:09.000 --> 00:27:13.000
up until now, the cancer has, even
though the cells are dividing and

00:27:13.000 --> 00:27:18.000
have lost some of their growth
control they're still staying in the

00:27:18.000 --> 00:27:22.000
same place.  So that would be sort
of, you know, like a wart or

00:27:22.000 --> 00:27:26.000
something like that,
or what you would hear as a benign

00:27:26.000 --> 00:27:30.000
tumor.
You can go in surgically and take it

00:27:30.000 --> 00:27:34.000
away.  But then the other thing that
can happen is cells can forget where

00:27:34.000 --> 00:27:38.000
they're supposed to be in the body.
And when that happens they say the

00:27:38.000 --> 00:27:42.000
cells metastasize and become
metastatic or a malignant tumor.

00:27:42.000 --> 00:27:46.000
And what that means is the cell is
beginning to, it's acquired yet

00:27:46.000 --> 00:27:50.000
another change that's made it forget
which part of the body it's supposed

00:27:50.000 --> 00:27:54.000
to be in.  And they've signified it
here as being a change in this cell

00:27:54.000 --> 00:27:58.000
that then leads to,
you can see here right now it's

00:27:58.000 --> 00:28:02.000
starting to invade into
the whole intestine.

00:28:02.000 --> 00:28:06.000
Or if one of those cells comes off
lose in your bloodstream it can land

00:28:06.000 --> 00:28:10.000
somewhere else in your body and then
start to grow there.

00:28:10.000 --> 00:28:15.000
And that's what happens when
somebody has metastatic cancer.

00:28:15.000 --> 00:28:19.000
You cannot really cure it because
now there are cancer cells all over

00:28:19.000 --> 00:28:24.000
the body.  And that usually is a
very difficult situation to get any

00:28:24.000 --> 00:28:28.000
kind of cure on.
So to put this in perspective,

00:28:28.000 --> 00:28:32.000
you needed to have a number of
changes to go from an ordinary cell

00:28:32.000 --> 00:28:37.000
to a metastatic cancer cell.
So each one of these changes there

00:28:37.000 --> 00:28:43.000
was some kind of change in the DNA.
Either there was a mutation or

00:28:43.000 --> 00:28:48.000
maybe a chromosome was lost or
something like this so that you need

00:28:48.000 --> 00:28:53.000
a series of successive genetic
alterations.  So there was a very

00:28:53.000 --> 00:28:59.000
key insight that a number of people
had after we understood the

00:28:59.000 --> 00:29:04.000
mechanism of mismatch repair.
Because some people realized that if

00:29:04.000 --> 00:29:09.000
a human cell had lost mismatch
repair then the frequency of each

00:29:09.000 --> 00:29:14.000
one of these changes would go up.
It wouldn't affect what the change

00:29:14.000 --> 00:29:19.000
was.  It wouldn't actually have
anything to do,

00:29:19.000 --> 00:29:24.000
if you lost mismatch repair it
wouldn't affect directly the ability

00:29:24.000 --> 00:29:29.000
of this cell to stop dividing when
it touches its neighbors.

00:29:29.000 --> 00:29:35.000
But it would increase the chances
that a mutation somewhere would have

00:29:35.000 --> 00:29:41.000
that effect.  And if every one of
these steps goes now a hundred or a

00:29:41.000 --> 00:29:47.000
thousand times faster,
you can see that if somebody loses

00:29:47.000 --> 00:29:54.000
mismatch repair in a cell then the
chances of that cell coming into a

00:29:54.000 --> 00:30:00.000
cancer are very high.
So there was a kind of human cancer,

00:30:00.000 --> 00:30:06.000
it's a susceptibility to colon
cancer called hereditary

00:30:06.000 --> 00:30:12.000
nonpolyposis colon cancer.
You don't need to remember the name.

00:30:12.000 --> 00:30:18.000
It's often abbreviated HNPCC for
people who cannot remember the name.

00:30:18.000 --> 00:30:23.000
But it was a kind of susceptibility
to cancer that ran in families.

00:30:23.000 --> 00:30:29.000
So it was thought to be genetically
determined in some way.

00:30:29.000 --> 00:30:33.000
And one of the interesting things
was a number of the people who had

00:30:33.000 --> 00:30:38.000
this disease would show a kind of
instability of the genome if they

00:30:38.000 --> 00:30:42.000
looked in the tumors.
They just looked at the DNA.

00:30:42.000 --> 00:30:47.000
It seemed to be undergoing changes
at a much faster rate.

00:30:47.000 --> 00:30:51.000
And the insight that came out was
that the people who had this disease

00:30:51.000 --> 00:30:56.000
had, for example,
a mutation affecting what we can

00:30:56.000 --> 00:31:01.000
think of as a human
homolog of mute S.

00:31:01.000 --> 00:31:06.000
And we'll talk about genetics of
humans in a small number of weeks,

00:31:06.000 --> 00:31:11.000
but I think most of you know that
for most genes,

00:31:11.000 --> 00:31:16.000
except for the genes associated with
the sex chromosomes,

00:31:16.000 --> 00:31:21.000
you get one copy of a gene from mom
and another copy of a gene from dad.

00:31:21.000 --> 00:31:26.000
So under most circumstances we
would have two good copies of this

00:31:26.000 --> 00:31:32.000
gene encoding a human
homolog of mute S.

00:31:32.000 --> 00:31:36.000
What does that human homolog of mute
S do?  The same thing as the

00:31:36.000 --> 00:31:41.000
bacteria.  It recognizes a mismatch
in DNA and fixes it up.

00:31:41.000 --> 00:31:45.000
So it turned out that what the
people with this disease have is

00:31:45.000 --> 00:31:50.000
they have one of the genes.
The gene they got from mom or the

00:31:50.000 --> 00:31:55.000
gene they got from dad is broken.
So they're still OK. They have one

00:31:55.000 --> 00:32:00.000
copy of mismatch repair
in every cell.

00:32:00.000 --> 00:32:06.000
But if a cell ever had lost that
copy of the good version now that

00:32:06.000 --> 00:32:13.000
cell and all of its descendents
would mutate at something like a

00:32:13.000 --> 00:32:20.000
hundred or a thousand times the
normal probability.

00:32:20.000 --> 00:32:27.000
And so they would progress down
this pathway.

00:32:27.000 --> 00:32:33.000
And so the polyposis means that if
they look in the colons of people

00:32:33.000 --> 00:32:39.000
who have this disease they find lots
and lots of little growths or polyps

00:32:39.000 --> 00:32:45.000
that are on their way to progressing
down this disease.

00:32:45.000 --> 00:32:51.000
Even in these people it takes quite
a while.  And so once they knew that

00:32:51.000 --> 00:32:57.000
they were able to go in and through
colonoscopies find these cancers

00:32:57.000 --> 00:33:02.000
and remove them.
And most of you will not have that

00:33:02.000 --> 00:33:06.000
disease, but this is now a kind of
cancer that's pretty much

00:33:06.000 --> 00:33:11.000
preventable as long as it gets
detected.  It can take in a normal

00:33:11.000 --> 00:33:15.000
person as long as 20 years or
something for an initial cell that

00:33:15.000 --> 00:33:19.000
underwent this initial change to go
all the way down to becoming

00:33:19.000 --> 00:33:24.000
metastatic.  So when you get older,
and this certainly applies to most

00:33:24.000 --> 00:33:28.000
of your parents or in this age group,
you should have ask them if they've

00:33:28.000 --> 00:33:32.000
had a colonoscopy.
It's not the world's most fun

00:33:32.000 --> 00:33:35.000
procedure because,
you know, they stick a probe and

00:33:35.000 --> 00:33:39.000
look inside your intestine,
but it isn't that bad.  And what

00:33:39.000 --> 00:33:42.000
they do is if they see one of these
little polyps they can catch it

00:33:42.000 --> 00:33:45.000
before it's progressed far enough to
be metastatic.

00:33:45.000 --> 00:33:49.000
And then there's no problem.
I had my first one done about,

00:33:49.000 --> 00:33:52.000
I don't know, three or four years
ago and they found one.

00:33:52.000 --> 00:33:55.000
And they took it out and I'm fine.
But if it had been left there and

00:33:55.000 --> 00:33:59.000
allowed to progress then some years
down the line I would have

00:33:59.000 --> 00:34:02.000
gotten colon cancer.
And I'm going to have to go back and

00:34:02.000 --> 00:34:06.000
get checked again in another year or
two.  But it is something that you

00:34:06.000 --> 00:34:10.000
should check with your parents
because everybody should have a

00:34:10.000 --> 00:34:13.000
colonoscopy.  My hope is by the time
you guys reach an age when this

00:34:13.000 --> 00:34:17.000
comes they'll probably have some
kind of little blood test or

00:34:17.000 --> 00:34:20.000
something where you won't have to go
through this indignity.

00:34:20.000 --> 00:34:24.000
But right at the moment it's
something everyone should do,

00:34:24.000 --> 00:34:28.000
I think.  I just wanted to make one
other comment about basic research

00:34:28.000 --> 00:34:32.000
because there's another thing here.
Actually, my lab was the first lab

00:34:32.000 --> 00:34:36.000
to clone the mute S gene.
We cloned it, we sequenced it,

00:34:36.000 --> 00:34:40.000
and we looked in the databases.  And
at that time in the late eighties

00:34:40.000 --> 00:34:44.000
there was nothing else that looked
like it.  I thought it would be like,

00:34:44.000 --> 00:34:48.000
there were some sort of similar
mutants, and here's what it looked

00:34:48.000 --> 00:34:52.000
like.  This is a culture of E.
coli.  And there are about ten to

00:34:52.000 --> 00:34:56.000
the ninth cells per mil.
And we plated about ten to the

00:34:56.000 --> 00:35:00.000
ninth or ten to the eighth on a
plate with a drug on it.

00:35:00.000 --> 00:35:03.000
And you can see they almost all died,
but there were maybe three or four

00:35:03.000 --> 00:35:07.000
that survived.
And then their descendents were

00:35:07.000 --> 00:35:11.000
able to grow up and form a colony.
This is how we recognized something

00:35:11.000 --> 00:35:15.000
was defective and what we now know
as mismatch repair.

00:35:15.000 --> 00:35:18.000
If you took this mutant of E.
coli and plated it out, you'd see

00:35:18.000 --> 00:35:22.000
you got a lot more drug-resistant
colonies.  That's the difference

00:35:22.000 --> 00:35:26.000
that I was describing,
the importance of mismatch repair.

00:35:26.000 --> 00:35:30.000
If you don't have mismatch repair
you can see, you get a lot more

00:35:30.000 --> 00:35:33.000
mistakes that show up as mutants.
So I was studying that.

00:35:33.000 --> 00:35:37.000
And we cloned the mute S and mute L
genes which are another gene that's

00:35:37.000 --> 00:35:41.000
involved in this.
Didn't see anything in the database,

00:35:41.000 --> 00:35:44.000
but there were very similar mutants
in streptococcus pneumonia that

00:35:44.000 --> 00:35:48.000
people had isolated.
Remember streptococcus pneumonia in

00:35:48.000 --> 00:35:51.000
the transformation experiments?
So I thought, well, maybe these are

00:35:51.000 --> 00:35:55.000
the same genes on an evolutionary
basis.  So I phoned some labs,

00:35:55.000 --> 00:35:59.000
and I found one that was sequencing
what turned out to be

00:35:59.000 --> 00:36:02.000
a homolog of mute S.
We tried to publish our papers in a

00:36:02.000 --> 00:36:06.000
medium fancy journal because I
thought this was a pretty cool

00:36:06.000 --> 00:36:10.000
result that two bacteria that were
evolutionarily very diverged had

00:36:10.000 --> 00:36:14.000
this conserve mechanism for mismatch
repair, but the reviewer said,

00:36:14.000 --> 00:36:18.000
you know, this is a pretty
specialized topic,

00:36:18.000 --> 00:36:22.000
it's not of general interest,
it should go in, the phrase they use

00:36:22.000 --> 00:36:26.000
is ìa more specialized journalî.
So it was published in the Journal

00:36:26.000 --> 00:36:30.000
of Bacteriology which is a really
wonderful journal,

00:36:30.000 --> 00:36:33.000
but it basically deals with bacteria.
And about a week after that paper

00:36:33.000 --> 00:36:36.000
came out my phone rang and it was a
guy from Emory.

00:36:36.000 --> 00:36:39.000
And he said, ìI work on mouse.
We were sequencing a gene,î it

00:36:39.000 --> 00:36:42.000
doesn't matter what,
ìand we sequenced in the wrong

00:36:42.000 --> 00:36:45.000
direction.  And we seem to have
something called mute S.

00:36:45.000 --> 00:36:48.000
Do you know anything about mute S?
And a couple of days after that I

00:36:48.000 --> 00:36:51.000
got a phone call from somebody at
NIH.  And they said the same thing,

00:36:51.000 --> 00:36:54.000
ìWe were trying to sequence this
gene in humans.

00:36:54.000 --> 00:36:57.000
We kind of sequenced in the wrong
direction and found mute S.

00:36:57.000 --> 00:37:00.000
So within a week of the paper
coming out I knew there were mouse

00:37:00.000 --> 00:37:04.000
and human homologs.
And that led from these sorts of

00:37:04.000 --> 00:37:09.000
studies, which my first graduate
student worked on,

00:37:09.000 --> 00:37:14.000
to the identification of the human
homologs.  And then not me but

00:37:14.000 --> 00:37:19.000
others made the connection between
mismatch repair and cancer.

00:37:19.000 --> 00:37:24.000
But this is the way a lot of things
happen with basic research.

00:37:24.000 --> 00:37:29.000
This doesn't look like anything
that's very important.

00:37:29.000 --> 00:37:32.000
And it sure doesn't look like it's
going to lead to an insight into

00:37:32.000 --> 00:37:36.000
cancer, but this is very much the
way it goes.  I've had this happen

00:37:36.000 --> 00:37:40.000
twice with another set of genes in
my life that turned out to be

00:37:40.000 --> 00:37:44.000
important for cancer as well.
And, as I said, what happens,

00:37:44.000 --> 00:37:47.000
if you lose mismatch repair, then
all these alterations happen much

00:37:47.000 --> 00:37:51.000
more quickly and the cells can
become cancerous.

00:37:51.000 --> 00:37:55.000
I've included a couple of outtakes
because I actually made this slide

00:37:55.000 --> 00:37:59.000
with my son's pillowcase on our
dining room counter.

00:37:59.000 --> 00:38:04.000
And our cats, who you saw at some
point earlier in the year,

00:38:04.000 --> 00:38:09.000
thought this was the weirdest thing
they had ever seen,

00:38:09.000 --> 00:38:14.000
when I brought these plates home.
So, OK, anyway.  All right.  So one

00:38:14.000 --> 00:38:19.000
other thing to tell you about DNA
replication before I move

00:38:19.000 --> 00:38:28.000
on, and that is --

00:38:28.000 --> 00:38:33.000
-- the initiation of DNA replication.
In E. coli there's one great big

00:38:33.000 --> 00:38:39.000
piece of DNA.  And it's all one
giant circular chromosome.

00:38:39.000 --> 00:38:45.000
And if you realize what I've told
you about DNA replication,

00:38:45.000 --> 00:38:50.000
I've talked to you only about once
you have a replication fork

00:38:50.000 --> 00:38:56.000
established how you keep it going.
But, as you might guess, a really

00:38:56.000 --> 00:39:02.000
important point of biological
control is the initiation

00:39:02.000 --> 00:39:07.000
of DNA replication.
And so the way cells do that is they

00:39:07.000 --> 00:39:12.000
have a special sequence in their DNA.
It's written just with Gs and Cs

00:39:12.000 --> 00:39:17.000
and As and Ts,
but it's a word sort of written in a

00:39:17.000 --> 00:39:23.000
different language than the kind of
genetic code we're going to be

00:39:23.000 --> 00:39:28.000
talking about in the next couple of
lectures.  And what it means is

00:39:28.000 --> 00:39:34.000
ìstart replication hereî.
And so in E. coli these terms are

00:39:34.000 --> 00:39:41.000
called origin DNA replication.
And, for example, in E. coli it's a

00:39:41.000 --> 00:39:48.000
stretch of DNA that's about 250 base
pairs long.  And it's got a sequence

00:39:48.000 --> 00:39:55.000
that lets proteins bind and they
kind of are able to make a little

00:39:55.000 --> 00:40:01.000
bubble like this.
And it's at the edges of this little

00:40:01.000 --> 00:40:05.000
bubble where it's able to start a
replication fork.

00:40:05.000 --> 00:40:09.000
And one of the secrets to control
of cell division is that cells are

00:40:09.000 --> 00:40:13.000
able then to control whether the
protein that sees the origin is

00:40:13.000 --> 00:40:17.000
there or not.  And it won't start a
new round of replication unless

00:40:17.000 --> 00:40:21.000
everything is right.
Then it can make the things that

00:40:21.000 --> 00:40:25.000
initiate a new round.
And after that it finishes.

00:40:25.000 --> 00:40:30.000
Our eukaryotic cells with a lot
more DNA use the same thing.

00:40:30.000 --> 00:40:34.000
The same idea,
but there tend to be multiple

00:40:34.000 --> 00:40:39.000
origins.  And you get a little
bubble and another little one down

00:40:39.000 --> 00:40:44.000
here.  And once you get the
replication forks established then

00:40:44.000 --> 00:40:49.000
these kind of merge.
And then eventually we end up with

00:40:49.000 --> 00:40:53.000
the two strands of DNA.
But I just mention that in passing

00:40:53.000 --> 00:40:58.000
because it's an example of how even
though the DNA is nothing but Gs and

00:40:58.000 --> 00:41:03.000
Cs and As and Ts,
you can kind of write words in there

00:41:03.000 --> 00:41:08.000
that mean different things.
Some of them on the genetic code

00:41:08.000 --> 00:41:13.000
tell you what the order of amino
acids in the cell are,

00:41:13.000 --> 00:41:17.000
but everything else has to be
encoded in the DNA,

00:41:17.000 --> 00:41:22.000
too.  And here's a really nice
example of how that works.

00:41:22.000 --> 00:41:27.000
Now, we're going to switch at this
point from worrying about how DNA is

00:41:27.000 --> 00:41:32.000
replicated to how information is
stored and interpreted.

00:41:32.000 --> 00:41:39.000
And there's a figure that most of

00:41:39.000 --> 00:41:44.000
you have probably seen,
DNA goes to RNA goes to protein.

00:41:44.000 --> 00:41:48.000
This is the usual direction of
information flow.

00:41:48.000 --> 00:41:53.000
The information for making proteins
is encoded in the DNA,

00:41:53.000 --> 00:41:58.000
as we'll talk about in more detail,
and an RNA copy of some piece of

00:41:58.000 --> 00:42:03.000
that, one gene's worth usually,
gets made in RNA.

00:42:03.000 --> 00:42:12.000
And then that information in the RNA
is used to direct the sequences of

00:42:12.000 --> 00:42:22.000
amino acids that appear in a protein.
And this is a four letter alphabet,

00:42:22.000 --> 00:42:31.000
if you want, A, G, T and C.
This is a four letter alphabet,

00:42:31.000 --> 00:42:39.000
A, G, U and C, where the uracil and
the thiamine have the same base

00:42:39.000 --> 00:42:47.000
pairing capacity.
And this is a 20 letter alphabet.

00:42:47.000 --> 00:42:54.000
All those 20 amino acids that you

00:42:54.000 --> 00:42:58.000
were looking at,
at the chart over at the back of the

00:42:58.000 --> 00:43:02.000
exam.
So from the point of view of

00:43:02.000 --> 00:43:06.000
information storage and information
flow there are some interesting

00:43:06.000 --> 00:43:11.000
things that had to come up in order
for the information to flow in that

00:43:11.000 --> 00:43:15.000
way.  But before I do that I want to
just get you to think about DNA as

00:43:15.000 --> 00:43:20.000
an information storage device.
This is MIT.  I'm almost sure in

00:43:20.000 --> 00:43:24.000
this room there are some people that
are experts in high density

00:43:24.000 --> 00:43:29.000
information storage.
And even if you're not most of us

00:43:29.000 --> 00:43:35.000
have now a lot of experience with it.
Your computer can do gigabytes of

00:43:35.000 --> 00:43:40.000
information.  Your iPod probably has
a 40 megabyte hard drive in it or

00:43:40.000 --> 00:43:46.000
something like that.
So you have some experience with

00:43:46.000 --> 00:43:51.000
high density information storage.
So here's the question.  How much

00:43:51.000 --> 00:43:57.000
DNA would it take to encode
everybody who's alive on earth today,

00:43:57.000 --> 00:44:02.000
6 billion and a bit people?
And let's argue that all we need is

00:44:02.000 --> 00:44:08.000
a single cell's worth of DNA because
everybody started out a single

00:44:08.000 --> 00:44:14.000
fertilized egg and went on.
Yeah?  OK.  Enough DNA to fill one

00:44:14.000 --> 00:44:19.000
human being.  Anybody else got any
sense?  All right.

00:44:19.000 --> 00:44:25.000
This is, I think, the most amazing
demo.  I did this when I was

00:44:25.000 --> 00:44:30.000
teaching for the first time.
The amount of DNA it would take to

00:44:30.000 --> 00:44:34.000
encode everybody who's alive on
earth, one cell of everybody who's

00:44:34.000 --> 00:44:39.000
alive on earth today is this little
thing in here,

00:44:39.000 --> 00:44:43.000
which you probably cannot see even,
but I took a picture of it.  There

00:44:43.000 --> 00:44:48.000
are about six times ten to the minus
twelfth grams of DNA in a human cell.

00:44:48.000 --> 00:44:53.000
And if you multiple that out by 6
billion people it comes out to 36

00:44:53.000 --> 00:44:57.000
milligrams of DNA.
And I weighed out 40 something

00:44:57.000 --> 00:45:02.000
milligrams of DNA.
So there's actually more DNA there

00:45:02.000 --> 00:45:06.000
than you need to encode everybody
who's alive on earth today.

00:45:06.000 --> 00:45:11.000
And I don't know how this hits you,
but I've been working on DNA my

00:45:11.000 --> 00:45:16.000
entire life.  And every time I do
this, you know,

00:45:16.000 --> 00:45:20.000
I think I understand this molecule,
but I don't really think I do at

00:45:20.000 --> 00:45:25.000
some more fundamental level.
It's absolutely amazing how much

00:45:25.000 --> 00:45:30.000
information is stored
in that molecule.

00:45:30.000 --> 00:45:34.000
So the one point I will,
actually, I think it's close enough.

00:45:34.000 --> 00:45:37.000
Why don't we just call it a day,
and I'll pick this stuff --