\~w

My title is "what is a gene ?” but what I really want to
talk about (8) some recent work on the fine-structure of the gene; work
which may lead us to understand more about the TE ee
molecular basis of genetics and perhaps something about that
gZenes act.
Let me first sketch for you the classical picture of the
gene. You can find this in any elementary textbook of genetics.
It starts by describing some simple breeding experiments -- usually
the historic experiments of Mendel on peas - and shews how
they can be explained by a straightforward theoretical scheme.
Then the complications due to "recombination" - as it is called -
are introduced, and finally the reader learnd how all these
ideas can te understood in terms of the chromosomes, the small
thread-like bodies within the nuclei of cells. The chromosomes
carry the hereditary determinants, and the genes, the units of
heredity, are strung out along the chromosome, in a linear
order, like beads on a string.
The gene, then, is the unit of heredity. It is the building
brick of the mee ohn eee to a child from its parents,
half from each of them, , which mainiy determines tue sort of
child it is. It is not known how many different geneé there are,
but each human being probably has more than 10;000 of-_them, di fterus ones.
That is the standard picture - but I want to explain to
you that. the gene - the gene in the very old-fashioned sense
of the indivipie unit of heredity <- has been split, and split
decisively. To do this I shall discuss a simpler system so that

nr
you can grasp the eseential poits more easily .
The system I shall describe 13 tnat of a virus that attacks
bacteria. You may feel that tnis is rather remote from human
genetics, but it is one of the astonishing discoveries of modern
biology how similar the basic biochemical and genetic processes
are in all living material. This virus is quite a small object
by normal standards - it is too small to be seen even with a
very poverful light microscope - but as viruses go it is quite
large. Its molesular weight - if we may use Such a term - is
about two hundred million (the molecular weight of water is 18)
and it is twenty or thirty times bigger than a small virus such
as the poliomyelitus virus. It is made up of only two types of
chemical - protein and deoxyribonucleic acid ( DNA for short).
Most if not atl of the protein component corresponds to the body
of the virus ‘aed the genetic part appears to consist mainly if
not entirely of DNA.

Let me say first, without giving the evidence, that the
genetic material appears to be all in one yikers long piece - in
a higher organism we would say that there was one chromosome -
and that there is only one copy of it in the virus. To use the
genetic jargon it is haploid, whereas human beings, for example,
are diploid; that is, we have two copies of each chromosome, one
from each parent. It is the fact that the virus is haploid which
simplifies the explanation. The exact number of genes in this
virus is not known, but a conservative estimate would be twenty,
arranged in a linear oraer along the chromosome.

When the virus comes into contact with a suitable bacterial
-eell it attaches itself by its tail (it has a polygonal head
and a long thin rigid cylindrical tail). All the DNA then engters
Vw

the bacterial cell and most of the protein stays outside. |
Twenty minutes later the cell bursts open and a hundred or so
new complete virus particles are released. Thus in this short
time the system has made many copies of the infecting virus.
Moreover the copies are exact copies. Various strains of the
virus exist and the viruses that come out of the cell are exactly
like the particular virus particle which went in. In other words
the virus breeds trué,

Very occasionally, however, the replicating mechanism
makes a mistake, and a slightly different virus emerges. If
we now infect a secong time with this changed virus its descendents
are all like itself;tiat is, of the altered form. In a higher
organism such a change would be called a mutation, and there
Seems no reason why we Should not use the word here, Tne virus,
then, can experience a mutation, and the matetton will breed
true. |

Now what will happen if one infects with two different
mutants of tne virus simultaneously ? It can be shown that
both penetrate into the cell, both multiply inside the cell,
and copies of both emerge when the cell is burst open. But the
interesting fact is that, in addition, new virus strains appear
whose genes come partly from one parent virus and partly from
the other. The picture evoked to explain this is that during
the copying process inside the cell a new copy may occasionally
start to be formed on one parent and then, at a random point,
skip§ over and finishes the process of copying on the other
parent. This process is called "yecombination" - again because

of its similarity to a similar process in higher organisms.
\+-

It is this recombination which allows geneticists to map
the genes on the chromosome. If two genes are very close together
the chance of the skipping-over process taking place between them
is rather small. If, on the otuer hand, they are far apart it
will happen more often. By finding how frequently recombination
occurs for any two genes one can get Some idea of their distance
apart, and it is by tnis method that geneticists have shown
that genes are arranged in a linear array along the lengta of
the chromosome.

Let us forget recombination for a moment, and think about
what happens if we have a virus one of whose genes is defective.
If this virus is used by itself it may be unable to reproduce.
However of another strain of virus, with that particular gene
intact, is used simultaneously, it may be able to do that part
of the work for two, and both viruses may be able to reproduce,
In fact if the second virus has a defect in a different gene,
of which the first has a good copy, they may be able to help
each tw other. A pretty picture !

Fortunately it has some technical use. For suppose both

these two viruses, instead of having their defects in different
genes, had a mistake in the Same gene. Then neither of them would
have a good copy, and reproduction might not be possible. Thus
this test ( of seeing whether two defective viruses can assist
each other or not) shows whether the defects are in the same
eene or not. This is one of the definitions of the gene, in fact

- the unit of physiological action, and it is the meaning

I shall be using when I use the word gene.
It is not the only definition, however. We could. define the
gene as the “unit of mutation". That is, the smallest part of
a chromosome which can change, or mutate, independent of the
other parts. Cr again, we could define the gene - and indeed
many geneticists do - as the unit of recombination - the
smallest part that can te moved about by the recombination
process: the smallest step in the skipping-over process if
you like.

So there we have three defintions of the gene. The unit of
action, the unit of mutation and the unit of recombination,
and for many years it was felt that, in spite of some exceptions,
these three definitions refered to the same thing - the gene,
as it was called.

It is now becoming clear that this picture was x oversimplified.
Let me give you the new picture straig it away, to help you
follow the argument. In the new picture the unit of action is
a large thine, roughly zumparakke corresponding to what used to
be thought of as the gene, but the other two units now appear
to be very much smaller. We can put this another way. Let us for
be iousas wee the word gene for the unit of action. Then we
can summarise the new discoveries by saying that not only
can geneticists map the order of genes on a chromosome, but they
can map the order of the different changes within a gene. In
other words, the gene - the unit of action - has-been split
by the genetic mavping technique.

Now I would not want you to think that this is a totally
new idea, because evidence along these lines has been available

for years. What has transformed the situation is the weight of
iN

the new evidence. This has mainly come from two sources. Drs.
Pontecorvo and Roper, working in Glasgow on the mould Aspergillus,
and Br. Benzer working at Purdue University, in the States, on

the virus system I have described. Br. Benzer's system has a
technical advamtage over that of the Glasgow workers because it
can be pushed more quickly to finer limits, but the work on

the mould makes us think that the new data on the virus is not

a freak, but likely to be of general occurence. Moreover cases are
known in flys, in corn and in other organisms.

Let me briefly Summarise what Dr. Benzer finds. For
technical reasons he has studied one particular gene, known
as rMII. dis first finding is that r II is not one gene, but
two, next to one another, and with closely} related functions.

He has picked up mutants of this pair of genes on about a

thousand different occasions. Sometimes he finds that a newly-
ariaing mutant appears by his tests to be identical with one

he has found before, but often it is quite different from

any previous one. In all he has so far found about 250 different
mutants in his two genes combined. In other words each of his
genes has been marked (or split, if you like) in over 100
different places.

His next finding is that he can arrange all these different
mutants in a linear order, as if they were a lot of different
marks on a piece of string. There are no real exceptions to this.
There are a few apparent exceptions, but these appear to be due
to "deletions" as they are called - meaning the actual loss of

S

a small piece of the string. In every Such cae he never finds
A
\

a reverse mutation -'Vene lost piece of the string never

reappears again{, 4S one wight expect}

As far as he can tell, allowinz for tue limitations of his
technique, tnere is no space between his two genes. AS soon as
one ends, according to his mapping, the other begins, without
any appreciable gap.

Another interesting finding is that there are mutants which
he can tell are different, but which he cannot seperate on the
map; they appear to occur at tie same point. You might wonder
how he knows they are different. The character which distinguishes
them is their reverse mutation rate - the rate at waich they
mutate back to the original version. For example one particular
mutation site is very common. He has picked it up about 130
times out of his set of 1,000 . About half or ete. have one
back-mutation rate, whereas the remahdder (with one exception)
have another one. In fact he appears to be finding that the
unit of mutation is not quite the same as the unit of recombination
though both appear much smaller than the unit of physiclogical
action.

It is not unreasohable to ask how small his smallest unit
is. How near 1s he to the molecular limit ~- the molecular
"graininess" if you like - of the genetic material ? It is
possible to make a very rough estimate of this if certain g
simplifying assumptions are made. The gentic material consists
mainly of DNA. Let us assume that it is wholely DNA. Now DNA,

as you know, is a polymer, made up of small molecules called

nucleotides joined end to end. The total number of nucleotides
\>

in the whole virus is about 490,090 though there are reasons
for veleiving that only part of this corresponds to the
genetic map. Knowing roughly the total length of the gentic
BAD, an e-scale
rapping Bemzer has calculated the number of nucleotides which

 

correspond to his smallest map-distance. The answer, which is
only approximate, comes to around half a dozen - and it might
he fewer. In other words by genetic means he is dividing the
gene down to, or almost down to, the molecular level. You can
understand why those of us who are interested in the molecular
basis of genetics find his work tremendously interesting.
Finally we may ask - what next ? It is a very reasonable
speculation that each gene controls, directly or indirectly, the
production of a particular protein molecule. liore precisely,
that it controls the, order ore ene potas in one polypeptide
chain of a protein molecule, What we suspect is that the linear
order inside a gene corresponds to the linear order of amino
acids along the polypeptide chain of the relevent protein, wich
the-aeli-te-pyedueing. If we could find the protein commeep =
£5 Dr. Benzer's gene we might be able to discover, with modern
techniques, the change in the order of the amino acids produced
by any particular mutant. If we could do this we could soon see
if the two orders - the linear mapping of the gene and
the linear arrangement of the amino acids ~- were related.
Dr. Benzer is coming to work with us at Cambridge next year, and

this is exactly what we shall try to do.