Below are excerpts of an informal, wide-ranging discussion.

On other attempts to clone mice from zygotes

Davor Solter

The way we did zygotes before, we removed two pronuclei, and whatever was in the nuclei was removed. Now, if I understand Eggan's technique correctly, the nucleus is now open again and so whatever is inside the nucleus can hypothetically spill into the cytoplasm.

So the difference between oocytes and zygotes could be whether the nuclear content could spill into the cytoplasm or not. A paper by Polish researchers in Reproduction4 would support this. The researchers enucleated zygotes differently; they broke the nucleus and sucked out chromatin.

It might partially recover the reputation of this scientist from Switzerland5. [Karl Illmensee, who claimed to have cloned mice from zygotes but whose work was discredited.] He couldn't figure out how to put a large pipet into cytoplasm and used a very thin pipet, so it was impossible to remove the rather large pronuclei intact, so as he was removing them, he always broke them. At the time we thought he left some chromosomes there, but it's possible that he let whatever-it-was spill, so he was in a way right. He always claimed that this should be re-examined.

It is significant in that it shows we never know what is right. We should never be sure about anything. At least for me this is the most interesting lesson.

Also, one should not be too clean in an experiment. I always thought that our major advantage was that we did it so cleanly. We always removed the entire pronuclei and their contents, and now it looks like trying to be too clean is not a good idea.

On what can be learned from the success of chromosome transfer

We didn't know how to find the right factors when they were in the cytoplasm, so I don't know how we will find them when we think they are in the pronuclei. The problem is that none of it will get us that much closer to understanding what exactly is in the oocyte or the zygote that makes reprogramming possible.

And why does it work so badly [for cloning genetic material from donor cells]? One would expect that biology should work with a little bit more reliability. Nobody in any other system would work with experiments that work only once in a hundred times.

The issue was always 'who is the right recipient for the clone?' 'What cytoplasm can reprogram?' Based on our work and the work of others that the zygote didn't work and the oocyte did work [for cloning] we always hypothesized that during the maturation something crucial disappeared from the cytoplasm, and therefore that some factor that was essential for reprogramming was lost or used up or disappeared.

And the logic followed that oocyte cytoplasm is the one that reprograms the sperm nucleus and the oocyte nucleus, so it [the reprogramming ability] is the property of an oocyte and not the property of a zygote. Now both of these papers3,4 would argue that this is not the case, that the difference between zygotes and oocytes is that in an oocyte what was in the nucleus is now spilled into the cytoplasm. It would be interesting to see if the germinal vesicle [the nucleus of the pre-ovulation oocyte] is also capable of reprogramming. Now one should start looking for the relevant factors that are not in the cytoplasm but in the nucleus.

One should not be too clean in an experiment. We always removed the entire pronuclei and their contents, and now it looks like trying to be too clean is not a good idea.

My impression of the paper in Reproduction is that the donor nucleus was in interphase. So then, the idea that both [the donor and the recipient cells] need to be synchronized cannot be right.

In practical terms, none will be of great use because people will still use oocytes as the recipient because they are easier to get than zygotes. Plus if there is the slightest possibility that you have to synchronize the donor and recipient, you would have to do additional work using the donor cells to synchronize. So I think as far as cloning is concerned that we are probably going to stay using oocytes and donor cells that are not in mitosis.

On what experiments should be tried now

I know people use cytoplasm from the two-cell stage [for cloning from blastomeres] but I don't know if it would go much beyond that6. It might work with the two or four cell stage, no reason not to try. It would be difficult to try in mice because the embryonic genome is already activated, and that activation could cause problems.

If it only works when mitosis is going on [when the pronuclear membranes have broken down], it's impossible to distinguish whether this is the perfect stage of the cell cycle [or due to spilled contents of the nuclei]. You'd have to get into this phase in the cell cycle without getting the pronculei membranes to dissolve, and I have no idea how to do that, but you can just try to break the pronuclear membranes before mitosis starts and see whether such zygotes are acceptable [as recipient cells for cloning].

At the time we were doing nuclear transfer, we were trying to think of a way to break the pronuclear membrane, but I couldn't think of any way to make absolutely sure that the chromosomes were all removed. At that time, we didn't really know that the chromosomes were all attached to the nuclear membrane and you could pull them all out, but even then that wouldn't be sufficient, because you'd always worry that something would break and you'd leave chromosomes behind.

Before, we assumed that the cytoplasm of the oocyte was capable of reprogramming and the cytoplasm of the zygote was not. We thought if you could inject oocyte cytoplasm into the zygote and make it suitable, you would have evidence that somewhere in the oocyte is a reprogramming factor. And you could compare oocyte and zygote cytoplasm. Now we are guessing that the factors are in the nucleus. Does it mean that they are present in every nucleus? Can one take the zygote, remove the pronuclei and then inject nuclear extract from any kind of cells and then do nuclear transfer? It sounds bizarre because you should have transferred that with the donor nucleus, so it might be that there is something truly specific in the nucleus of the oocyte and of the zygote.

Is it present only in the female pronucleus or also the male pronucleus? The next experiment would be to remove each pronuclei intact. One could season the donor nucleus in the oocyte and then transfer it to the zygote. So there is a whole series of micromanipulation experiments that could be done.

On lab work in the 1970s and 1980s

In developmental biology, everything was going so slowly. The mouse blastocyst has two parts, and people used to do experiments separating those to study cell lineages. That was extremely tedious. It required micromanipulation, but maybe only one or three people could do this. I was doing expression of antigens during embryogenesis.

And suddenly one day I thought, “I know the trophoblast is impermeable; there's a tight cell junction between the cells and if I expose the blastocyst to antibody and then wash it and put it in complement, the complement will kill all outside cells and the inner cell mass will be left intact.” And I tried, and in one hour I had 30 or 40 separated, clean inner cell masses, which at that time anybody would take days to make 3 or 4. It really worked. It was probably the only moment in my life where you could go from an idea to a solution in such a short time7.

We spent a lot of frustrating days discovering imprinting but not believing it. We were transferring male and female pronuclei [to make cells containing only male or female contributions to the zygote] We were fully expecting them to develop normally and they wouldn't and they wouldn't, and we just couldn't believe that they wouldn't. And it took us six months to a year before we decided that we should start believing it8.