Eggs and scrambled chromosomes

It's not quite meiosis, but it's pretty cool

Making eggs is pretty hard. Even in mice, the only species for which scientists can currently grow eggs in vitro, the process is difficult and inefficient. Research on human in vitro oogenesis has made considerable progress recently, but we’re still not at the point where a 50 year old woman (or man) can have biological children using lab-grown eggs.

But what if growing eggs from stem cells wasn’t actually required?

A new paper (Mikhalchenko et al.), recently published by the labs of Paula Amato and Shoukhrat Mitalipov, reported research on an alternative method, in which a nucleus from an adult mouse cell is injected into a donor egg. Subsequently, the egg becomes haploid through an unusual process of cell division, and is able to be fertilized by sperm.¹ Although the results of the paper show that it will be very difficult to make this process work for humans, it could eventually provide an alternative to in vitro oogenesis for people who want to obtain oocytes containing their own genetic material.

So how does this work?

Prior to fertilization, eggs maintain their chromosomes in metaphase of meiosis II, the last stage of meiosis. The machinery to separate chromosomes is present, but it’s not activated until the sperm enters.

The authors followed up on a previous paper in which they noticed that injecting an adult (somatic) cell nucleus into a metaphase II stage egg can cause the chromosomes to segregate, sort of like what happens normally during metaphase II. Interestingly, this method of nuclear transfer is very similar to what’s used for cloning, except that the egg cell was also injected with a sperm.

The nuclear transfer method from Lee et al. 2022 (CC-BY). PN refers to pronucleus, and PPB refers to pseudo-polar body.

In this paper, the authors refined their method and performed experiments using different mouse strains. Over the years, researchers have developed strains of mice which are so inbred that all the mice in each strain have genetically identical chromosomes.² By crossing two inbred strains (B6 and FVB), the authors generated hybrid mice with a consistent, yet distinguishable, set of chromosomes.

Looking at where the chromosomes went after injecting a nucleus into a metaphase II egg, the researchers found that the distribution was basically random. Out of a total of 63 oocytes injected, none ended up with all 20 mouse chromosomes properly segregated. Also, there was no recombination happening between the different chromosomes. But the authors’ previous paper reported better chromosome segregation, and even some live mouse offspring! So why was this suddenly worse?

The answer turned out to be that the method works better for inbred mice. By some unclear mechanism, chromosomes which are genetically identical are better able to segregate properly during meiosis II.

This is quite surprising to see. Until now, it was thought that proper chromosome segregation is set up during prophase of meiosis I, based on recombination between homologous chromosomes. Metaphase of meiosis II is a much later developmental stage, and the proteins necessary for recombination are no longer present.

Recombination-related proteins during prophase of meiosis I (in vitro!)

And because recombination is required for proper chromosome pairing during meiosis, the nuclear transfer method is very sloppy about assigning which chromosomes to go where.

So, there must be some other mechanism that promotes proper segregation during meiosis II, although apparently it only works if the chromosomes are extremely similar. It will be quite interesting to see the followup research on this!

What does this mean for humans?

Although this paper describes some very cool biology, it actually makes me a bit less optimistic that this method could be applied to humans. First, even the most inbred humans are not nearly as inbred as mouse strains, and it looks like inbreeding is key to efficiency here.

Yeah, even the worst of the Habsburgs only had an inbreeding coefficient of 0.254. Those are rookie numbers!

Second, using this method will require large numbers of donor oocytes, and most of the embryos produced will be aneuploid. Even for inbred strains, the rate of successful segregation was only 14 – 20%.

However, if the mechanism that causes proper segregation for inbred chromosomes is better understood, perhaps it could be adapted to work with arbitrary chromosomes. And if the efficiency of this nuclear transfer method is improved, it might one day be a complement to in vitro oogenesis. Large batches of eggs could be grown using a single stem cell line, and then a nucleus from a patient could be swapped in just prior to fertilization. It will be interesting to see how this research progresses!

1. I actually first learned about the lab’s work when they were applying for an Astral Codex Ten grant and I was a reviewer. I thought it was super cool, but I couldn’t blog about it until now, in order to maintain confidentiality. Although ACX Grants didn’t have enough money that year to fund this project, they ended up getting an Open Philanthropy grant and I’m happy to see their research published.

2. Aside from de novo mutations.