Many factors influence the rate of progress of scientific research: the development of new technical procedures; the availability of more refined equipment; recognition by funding agencies of new areas of research ripe for further investigation; and effective collaboration between individuals with different skills and expertise. But in our cloning research, the greatest catalyst of all was a conversation in a Dublin bar.

As we all know, informal chats between research workers are extremely important for the exchange of ideas. And it was just such a conversation in that Dublin bar that transformed my view of the potential importance of methods for the cloning of mammals. As a result, a new line of research led ultimately to the birth of Dolly the sheep, the first clone of an adult animal.

The process of cloning is technically very demanding because it involves moving the genetic information from an egg and substituting the nucleus from a donor cell. It is the genetic information in the new nucleus that determines almost all of the offspring's characteristics. The eggs of amphibians are relatively large when compared with those of mammals, so methods for nuclear transfer in these species were developed first.

In about 1960, the aim of cloning experiments in amphibians was to be able to understand the mechanisms that regulate the development to adulthood of an embryo that is just a single cell. At that time, an important hypothesis was that cells discard the genetic information that is no longer required at later stages of development. This led to the prediction that an embryo produced by nuclear transfer would not be capable of normal development if there had been a permanent change in the genetic information by the stage of development from which the nucleus was transferred.

ÖйúAƬ

It was readily established that nuclei from very early stages of development retain the ability to direct development to an adult frog. However, as nuclei were transferred from progressively later stages, the proportion of embryos that developed to adulthood decreased significantly. If nuclei were transferred from the tadpole stage of development, only a very small proportion of cloned embryos were able to form adult frogs. But when nuclei were transferred from adult tissue, none of the resulting embryos was able to develop further than the tadpole stage. Specifically, no adult amphibians have ever developed following the transfer of nuclei from adult tissue, even today.

These results had a profound effect on biological thinking until the birth of Dolly. They showed that there is something profoundly different about the organisation of genetic information in adult cells, but the nature of the difference was not clear.

ÖйúAƬ

One possibility was that specific pieces of genetic information are lost as differentiation takes place as originally suggested. An alternative hypothesis suggested that all the genetic information might be retained in the cell, but that the complex regulatory mechanisms that direct expression of the genes required for the formation of each tissue could not be reversed by cloning procedures at that time.

Meanwhile, other research in mammals led to the development of methods for the recovery and transfer of embryos between females. In addition, sophisticated equipment was developed that made it possible to perform microsurgery on mammalian embryos, which are so small that they cannot even be seen without the aid of a microscope. In this way, embryos were divided into two or even four fragments in order to produce a larger number of offspring from particularly valuable parents. All this made it possible to begin to think of carrying out nuclear transfer in livestock species and laboratory mice.

The experience in amphibians led people to believe that cloning had only limited potential and that it would not be possible to clone adult mammals. Nevertheless, there was considerable commercial interest in being able to produce a number of identical offspring by multiplying embryos from valuable livestock.

The laboratory mouse was the first mammal in which nuclear transfer was attempted. With hindsight, it seems that research progress in this species was diverted by a spectacular report by Karl Illmensee in 1979, which subsequently proved impossible to repeat. In order to obtain some clarity on this matter, a distinguished group of independent researchers was asked to examine the results. In addition, one member of this group, Davor Solter, conducted his own research to discover what he could achieve by nuclear transfer in the mouse.

In 1984, Solter and James McGrath developed the first reproducible methods for nuclear transfer in the laboratory mouse and published several key papers describing their work. It was in one of these, published in the eminent journal Science, that Solter said: "The cloning of mammals, by simple nuclear transfer, is biologically impossible." This had an enormous influence on scientific and public expectations.

Some of the key experiments in livestock were carried out by a Danish veterinarian, Steen Willadsen, while working at the University of Cambridge. It was he who earlier had developed the methods of embryo-splitting referred to above. As a result, he had the equipment and skills required to perform the first cloning experiments. In his first report, published in the journal Nature in 1986, he described successful nuclear transfer from sheep embryos that had eight or 16 cells.

Seen from the post-Dolly era, it is easy to underestimate the importance of this experiment. At the time, it represented a distinct breakthrough because it established the first procedure for routine nuclear transfer in mammals and created commercial interest among livestock breeders.

As research continued, people began to develop hypotheses that could account for the apparent limitations of nuclear transfer. Considered in very general terms, it seemed that the procedure in the mouse was successful only to the two-cell stage, whereas in sheep and cattle, development could be obtained following transfer from the eight- or 16-cell stage. This contrasted with success in amphibians from stages with many thousands of cells. One factor that correlated at least approximately with this difference was the species-specific stage of development at which the genes in the embryo assume control of development.

ÖйúAƬ

The very earliest stages of development are brought about by factors in the egg produced while it is still within the ovary of the mother. At a specific stage of development, the genetic information of the embryo begins to function and assume control. This occurs when there are two cells in mouse embryos, eight cells in sheep and cattle, and more than 5,000 cells in amphibians. So for a time, those of us who were involved in the field thought that it might not be possible to obtain clones following nuclear transfer from stages of development far beyond this critical point. The interpretation that we offered was that beyond this stage, there were changes in the organisation within the nucleus that existing cloning procedures could not reverse.

The meeting of the International Embryo Transfer Society was the most important scientific gathering for those of us carrying out nuclear transfer. It is held annually in January; as most members are based in North America, it meets for two consecutive years there, but every third year convenes elsewhere. In 1987, the meeting was in Dublin.

ÖйúAƬ

It was during that meeting, and at this stage in the development of the cloning technique, that the particularly important conversation took place in the bar. It was with a friend, Geoff Mahon, who I knew from when we were both at Cambridge. He played a key role at Granada, the company in Texas where Steen Willadsen had worked briefly.

Geoff told me that Steen had been able to obtain calves following the transfer of nuclei from embryos at the blastocyst stage of development. This was very important information because this embryo stage has many more than eight cells. The result meant that development was possible following the transfer of nuclei from cells that were several days past the critical point when the embryo assumes control of its own development. This in itself was of profound biological importance.

Moreover, it is from this stage of development that embryo stem cells are derived. Steen's success suggested to me in that revelatory moment that if we could derive embryo stem cells from livestock species, nuclear transfer was feasible. Taken together, these two techniques would make it possible to introduce precise genetic modifications into livestock species for the first time. It was this eureka moment that led me to change the emphasis of my own research completely, and ultimately led to the birth of Dolly.

As it happened, I was planning a visit to Australia later that year and was able to see Steen at his new position in Calgary on my way back to Edinburgh. I had known him when he worked in Cambridge. He quickly confirmed that he had indeed produced cattle following nuclear transfer from the blastocyst stage and was extremely open and generous when describing the techniques that he had used.

I landed back in Edinburgh on 6 June 1987 and sought commercial funding to support our new research project. I remember specifically that on 10 October that year, we held our first meeting with a group of partners. It was their funding that enabled us to recruit Keith Campbell to the Roslin Institute.

The work was a team effort. Without Keith and other valued colleagues, it would not have reached a successful conclusion.

Whereas my previous research experience had been with the reproductive biology of livestock species, the main objective of Keith's research was to understand the mechanisms that regulate the growth and division of cells. He quickly recognised that coordination of these mechanisms in donor cells and recipient eggs would be an absolute must for the development of cloned embryos. As a result, we were able to improve the efficiency of procedures for transferring nuclei from early embryos.

In a series of critical experiments, we found that nuclei that had been induced to hibernate had greater potential to support normal development. In the first experiment, two sheep, Megan and Morag, were born, following the transfer of nuclei from differentiated cells derived from embryos by our collaborator, Jim McWhir. This in itself was very exciting, because the donor cells were by far the most differentiated cells that had been used successfully up to that point.

This result prompted us to use cells taken from adult tissue, and Dolly was born. Since then, many laboratories have confirmed our fundamental observation that it is possible to clone adult mammals from most species.

The results of the research project were entirely unexpected. We failed to obtain embryo stem cells, which meant that our initial approach to the multiplication and genetic modification of livestock was not open to us. However, we had far exceeded our expectations in that we had been able to clone adult animals. Since then, precise modifications to livestock species have been introduced by making genetic changes in cells from adult animals that were then used for nuclear transfer.

The most exciting of these projects enabled a Japanese-owned company based in the US to generate cattle that produce human rather than bovine antibodies. There is every reason to expect that in the near future, these antibodies will offer the first treatments for human diseases that we cannot treat at present.

These events illustrate two influences on scientific research. First, informal communication is often very important for the development of ideas. It is vital that the organisation of research facilities and scientific meetings should facilitate this type of interaction. Second, the outcome of research cannot be predicted.

ÖйúAƬ

Of course, this helps to make a research career very exciting. It also shows that a significant part of academic research funding should be directed to basic discovery. It is in this way that we are most likely to come across important surprises.