91av

THE human genome is about a metre of DNA containing three billion pairs
of chemical building blocks known as bases. The American government has
committed itself to discovering the sequence of bases in this DNA. Sequencing
DNA now costs between $3 (Pounds sterling 1.75) and $5 per base pair, so
the project will probably cost more than $10 billion.

And what about Britain’s contribution to the project? With classic British
understatement, the government has agreed to support a British project to
map the human genome by supplementing the Medical Research Council’s grant-in-aid
by Pounds sterling 11 million over three years.

Can anything be done with Pounds sterling 11 million in the face of
this colossal American effort and the possibility of a corresponding effort
by the Japanese? Is it worth doing anything at all? Involvement with the
genome project is important for Britain. Whoever gets the human genome data
first will decide what will happen to them, and will be in an unassailable
position to dictate terms over its commercial, including its medical, exploitation.
Britain has to buy itself a seat at the international bargaining table,
and we will probably have less than five years to establish our credentials.
Bidding will not wait for the project to be completed – it will start as
soon as there is anything worth selling.

Given the competitiveness of the human genome game, is the Pounds sterling
11 million enough to ensure that Britain is taken seriously? In all probability,
yes, provided that the money is spent imaginatively. A subtle, finely tuned
approach may still mean that British research is world-class, even if it
is relatively poorly funded.

The strength of the British approach lies in its potential to put the
powerful techniques of molecular genetics into the hands of people who have
an important biological problem to solve. So the best strategy, at least
for Britain’s small-scale studies, is to find a way of ensuring that research
groups with interesting biological, commercial or medical problems have
help with the methodology and technology of molecular genetics. Rather than
having a methodology looking for applications, we should ensure that anyone
with a good problem has access to this methodology.

In Britain, this strategy is about to be realised. On the eighth floor
of the Clinical Research Centre’s building at Northwick Park Hospital at
Harrow in Middlesex, a suite of laboratories is almost ready – the Human
Genome Mapping Project (HGMP) Resource Centre. British scientists are being
formally invited to join the project and use its resources. But what will
the centre actually do? To answer this, we need to explore what the project
can, and cannot do.

The ultimate goal of the American human genome project is the complete
sequence of all three billion base pairs in the human genome – to produce
a complete ‘physical map’ of the genome. The alternative which is being
pursued in the British project is to produce a gene map – the sequence of
base pairs making up individual genes and their positions within the complete
genome.

These two maps sound identical, but they are not. The most unexpected
feature of the human genome (a feature it shares with the genomes of other
higher animals) is that most of it does not consist of genes – which account
for only about 3 per cent of the total DNA. The purpose of this great quantity
of non-coding DNA is not known, if indeed it has any. Whether it is truly
deserving of the description ‘junk DNA’ remains to be seen. One possibility
is that it may act as the source of new genes. The ‘higher’ an animal is
in evolutionary terms the more genes it has, and these genes have to come
from somewhere. A mutation in a gene may create a new gene, but it also
destroys one. Mutations in the spare DNA may create a new gene without losing
anything useful.

But even if this spare DNA does have some purpose over evolutionary
time, in everyday medical terms that purpose is secondary to that of the
genes. So there seems little point in attaching the same priority to obtaining
detailed information about the ‘junk’ DNA as about the genes themselves.
And this reasoning underpins the operation of the resource centre – its
strategy is to go after the human genes that appear medically or commercially
important, sequence them and map their position within the genome.

Even this much more limited aim, concentrating on about 3 per cent of
the total DNA, presents formidable problems. The 100 million base pairs
are contained in 30 000 to 100 000 different genes. The total number of
human genes that have been completely sequenced so far worldwide is probably
less than 2000 – though many more have been partially sequenced. Although
the problem remains formidable, it does start to look much more practical
on limited British resources.

It looks even more promising if one suggestion is taken up – to limit
the sequencing for many genes to a few hundred bases at the end of each
gene: Sydney Brenner, director of the MRC’s Molecular Genetics Unit at Cambridge,
argues that this information will almost certainly be sufficient to tell
us whether the gene has already been sequenced, and so stop sequencers from
wasting their time doing it again. But there is a much more important reason
for adopting this approach: a few hundred bases may be sufficient to identify
what sort of a protein the gene codes for and tell scientists whether the
gene is of any interest to their research.

Many computerised databases now available link DNA sequences to three-dimensional
structural motifs in proteins – the bits of a protein that form hormone
receptor sites, for instance, or regions that span membranes or bind DNA.
The computer algorithms that link sequence and structure are improving and
this gives added incentives to molecular biologists to obtain new sequences.
Plainly, a strategy of looking at only parts of genes should be cost-effective,
maximising the useful information that can be obtained for a rather limited
investment.

But in the American project, cost, although very important, is not the
rate-limiting factor at the moment – while technical difficulties are. Ironically,
the cost-cutting British approach may well be at an advantage in purely
technical terms.

Gearing up the technology and organisation for a full-scale assault
on the physical map may take the Americans some years, providing a breathing
space for the British effort to make a decent impact on the gene map.

The technical difficulties arise from the sheer amount of DNA to be
sequenced. To tackle the job, the DNA has to be broken up into manageable
bits. The largest individual pieces of DNA that can be sequenced are somewhere
in the region of 300 to 500 base pairs long. The average gene is at least
two or three times as long as this, while large proteins may have genes
that run into thousands of base pairs. Long pieces of DNA have to be broken
into smaller pieces and sequenced. The breaking up must be done in more
than one way so that overlapping sequences of base pairs can be obtained
which allow different pieces to be matched up. This procedure can be done
randomly – the ‘shot gun’ approach which obtains information quickly but
leaves gaps which are difficult to fill – or in a semi-ordered or fully
ordered way which are useful in filling the gaps, although not so useful
in the early stages when speed matters. As the length of the DNA to be sequenced
becomes longer, the complexity of overlapping and matching the small pieces
becomes very great.

Using these methods the largest single DNA that has been fully sequenced
so far is the genome of the cytomegalovirus with about 250 000 base pairs,
less than one ten-thousandth the size of the human genome. A number of laboratories
around the world are believed to be attempting or about to be attempting
genomes an order of magnitude larger – the genome of the bacterium Escherichia
coli, for example, which contains about five million base pairs (of which
nearly a million have been sequenced already though not as a single continuous
length of DNA).

The British strategy of limiting sequencing to simple genes or parts
of genes avoids the problems of exponentially increasing complexity. Still,
there will be serious problems to tackle even with the ‘genes only’ approach.
For instance, the dystrophin gene involved in muscular dystrophy is contained
within a piece of DNA about 2.3 million base pairs long which must be close
to the limit of what can be sequenced with existing methodology and technology.

The British gene mappers will locate the position of individual genes
using a variety of techniques. Again, the challenge is to find ways of chopping
up the genome into manageable, identifiable chunks that can be stored and
retrieved when needed. The genome is spread over 24 chromosomes, but even
the smallest individual chromosome is too large to be handled successfully.
The genome must be broken up into pieces of a convenient size – small enough
to be handled, but large enough for the entire ‘library’ of the pieces making
up the total genome to be stored easily.

The crucial feature of the storage, however, is that a large number
of faithful copies of each piece can be made very readily. To do this, the
library is ‘cloned’ by inserting the fragments into the DNA of potentially
rapidly reproducing cells – as they copy their own DNA, they will also copy
the extra piece. The two cells that are traditionally chosen are those of
E. coli and of yeast. In E. coli the inserted pieces are called cosmids,
which contain typically about 40 000 base pairs. In yeast, the inserted
pieces are known as YACs (yeast artificial chromosomes) and are roughly
an order of magnitude larger, from about 250 000 up to a million base pairs.

Of course, however carefully the initial human DNA is broken up, the
pieces from different cells will vary in size to some extent. And when the
fragmented DNA is then purified, some will inevitably be lost. This presents
something of a problem since you cannot then be certain that any particular
gene will be stored in your cosmid or YAC library. By increasing the number
of cosmids or YACs, however, the chances can be increased to any predecided
level. The HGMP Resource Centre at Northwick Park has decided to use YAC
libraries rather than cosmids, and is negotiating the transfer of a number
of existing libraries to the centre.

The centre aims to transform what is now a rather ad hoc activity into
a large-scale, systematic professional service. It intends to bring the
methods of molecular genetics within reach of many research groups which
at present do not have it. From human cells the centre will collect messenger
RNAs, the chemical intermediary between genes and the protein-manufacturing
system of the cell. It is then relatively easy to make complementary DNA
from them. After this cDNA is sequenced, researchers can use it as a molecular
probe which will show the gene’s position on the YAC. The centre will generate
some of its own cDNAs from human tissues collected at Northwick Park Hospital
– it is in close contact with the Kennedy-Galton Cytogenetics Unit at the
hospital. It will also act as a catchment centre for cDNAs from research
groups outside. The availability of the cDNAs will be advertised to the
scientific community and they will be distributed to interested laboratories.
When the centre is fully operational, it is expected to be able to process
up to 100 cDNAs a week. Over five years, an appreciable fraction of the
genes in the genome could be identified and located.

The resource centre will also keep in touch with genome programmes in
other countries and exchange data with them. The need for networking to
provide online access to databases led to the realisation that the British
genetics community is not as computer literate as it might be – many reputable
laboratories are not connected to networks. To improve awareness, a task
force from the centre will visit laboratories to help and advise, offering
as a carrot a user-friendly system giving access to standard molecular biology
packages.

But it’s not just human genes that are important to the biological aim
of understanding what genes do. A decent understanding usually comes by
comparing one thing with another. In parallel with the human genome, the
mouse genome will be surveyed, mapped and sequenced. And not only the mouse.
The Laboratory of Molecular Biology, for instance, is at the centre of an
international effort to sequence the eight million base pairs in the genome
of the nematode Caenorhabditis elegans. There should be plenty of scope
for making interesting comparisons between one species and another (see
Box above and on previous pages).

The idea of a resource centre is something of a novelty to British research.
Yet it provides a solution to a problem that has traditionally faced agencies
that distribute public money for research. The motives of individual scientists
only occasionally coincide with the motives of governments. Indeed, for
the most part, the motives of scientists are probably not even comprehensible
to government. Yet the Human Genome Mapping Project Resource Centre may
well help to bridge the gap. On the one hand it should help individual scientists
to achieve their own ambitions, however theoretical they seem to outsiders,
and at the same time provide a respectable base for exploiting anything
that looks commercially promising.

Traditionally, research in Britain is supported ‘reactively’. Scientists
are trusted to know what the best problems are and how best to solve them.
The notion of dragooning biologists into working in a direction decided
by someone else has not generally found much favour. Yet this is the first
time in the life sciences that a large, well-defined and testing problem
has existed alongside a maturing technology that could, with concerted effort,
crack it.

So it makes sense to try a radically new approach. With the resource
centre, the Medical Research Council is trying to create a sort of sheepdog
to round up and point the community of gene mappers and sequencers in the
same direction. In doing so, the MRC is putting its reputation on the line:
there will presumably come a day of reckoning from the government paymasters.
For the sake of British science, the hope is that the government of the
day likes what it is given. But to succeed, the resource centre will need
the cooperation of Britain’s molecular scientists, who, seeing on which
side their bread is buttered today, will, it is hoped, work together in
the hope of jam on their bread tomorrow.

Dr John Galloway is head of public relations at the Cancer Research
Campaign in London.

* * *

WHOSE IDEA WAS IT ANYWAY – AND WHO’S PICKING UP THE BILL?

THE idea of mapping and sequencing the human genome in a detailed and
systematic way seems to have arisen in several places at once. The touchpaper
was the realisation that the technology existed, or could certainly be developed,
to enable scientists to tackle the ultimate problem in molecular genetics.
In Britain, the prime mover was Sydney Brenner, for some years director
of the Medical Research Council’s Laboratory of Molecular Biology in Cambridge
and now director of its Molecular Genetics Unit. Brenner’s idea was in time
transformed into a formal bid to the government by a board jointly representing
the MRC and the Imperial Cancer Research Fund, which shrewdly claimed and
bought a stake in the project. The Cabinet Office showed interest in a stand-alone
British project, and finally the board drew up a detailed proposal which
the Advisory Council on Science and Technology then approved.

Margaret Thatcher gave this project the ultimate seal of approval, ensuring
that new money was made available: Pounds sterling 2.3 million in the financial
year 1989/90, rising to Pounds sterling 4.6 million in 1991/92. After that,
the MRC will have to find the money from its grant-in-aid; extra money will
not be earmarked by the government.

Although the money is coming to the MRC, which will act as paymaster,
the genome project is not an exclusive MRC project, but one into which other
research organisations will tap.