Breeding strategies for mice

The text on this page is taken from an informal compilation of opinions of contributors to the online VOLE List. As such, they are not peer reviewed and may contain differences of opinion. Those wishing to contact the list may contact Adrian Smith.

A question about the optimal strategy for breeding lines of mice that breed "true", such as homozygous x homozygous.

What would you consider the optimal way to maintain these lines, as some sources seem to suggest using sibling matings and others disagree, so your collective advice on the subject would be really appreciated, please!

You could contact the Breeding and Colony Management Advice Service of the NC3Rs.

I'll start my comment by saying that I am absolutely NOT a genetics or breeding expert. I am certain others are much better equipped to answer than me, but I'll start the ball rolling  .... Short answer:

We don't brother-sister mate in homozygous (hom) colonies and we try to regularly backcross to Wildtype (WT) to maintain the background genetics as close to commercially available C57B6 colonies as possible.

Long answer:

I suspect the answer is "it depends". There are a number of issues that need to be considered. For example, if they have a closed colony for hom x hom, what are they using as their control WT (assuming they need one)? Continuing to brother-sister mate will continue to inbred the line - how does this compare with any control that is being used? Why is this necessary to continue to inbreed when you have fixed the alteration for the gene of interest and the animals are already inbred?

If you brother-sister mate (as I understand it) you may "fix" additional unwanted mutations quicker than by having a wider breeding strategy - JAX have a webinar on this (can't remember the relevant size of colony exactly, but JAX say that this runs at potentially 3 a year in a medium sized colony if I remember correctly). Effectively when JAX doing brother-sister matings they do full genome analysis periodically and if they identify something unexpected/ they don't want, they just eliminate that line - so they have multiple brother-sister lines going at the one time. If you don't do the genetic analysis (not just genotyping for the alteration of interest - not that I expect that this is being done at all in a homozygous colony) how do you know what you've got? Random mutations becoming "fixed" will be slower with a non-brother-sister strategy - but you will still get some; but it's less risky from that perspective.

The other issue is what you are doing about genetic drift. JAX go back every 10 generations and use frozen genetic material to "reset" the background to slow down genetic drift. If you are breeding only within the colony (whether brother-sister or wider strategy) how are you dealing with the fact that these mice will no longer reflect the background strain as bred at commercial breeders/ elsewhere. You will have a unique colony. This is absolutely fine if it doesn't matter and/ or if it is reported correctly in the papers using the animals.

We have a few Hom bred colonies. We view these as a genetic alteration that we wish to maintain (as far as possible) on a C57B6J background. We therefore don't brother-sister mate. General experience over the years means that the technicians feel it's generally a bad strategy as they find this causes issues with the breeding performance and with pup size. As it's our policy to avoid it, it's not done without specific permission from me as the NVS (as my usual experience is that people do it when they aren't managing the colony properly, i.e. have too few/ too old pairs and they "end up" needing to do it, so we need a wider discussion). We also backcross periodically (generally every couple of years) to the currently available C57B6J WT commercially available at the time. This obviously means we have a generation of heterozygotics and then we re-cross these to go back to the homozygotes. With complex genetic alterations (our worst has 8 different gene alterations) we obviously can't use this strategy and so periodically freeze eggs down and these can then be used as a "time capsule" to go back to the previous background as required. I asked at a webinar on genetics as if there were better approaches then this one for complex genetics, but no-one appears to have one.

Final comment - and this is very much just my own humble opinion - I have always found that there are far more instances of people using too few animals in their breeding strategy rather than too many - driven by funding more often than not (only one or two pairs to maintain a colony of valuable GA mice is a pretty risky strategy in my opinion - you only need one pair to stop producing and the other mum to get stuck littering and you risk losing it, before we consider the background genetics issues). I absolutely hate wastage - but I have seen too many issues caused by colonies squeezed to a ridiculously small size to feel that the current pressure for reduction in numbers beyond everything is providing an overall reduction in animal harm - and possibly sometimes the reverse.

Just in case I didn’t understand the question, please allow me to state that we could have an outbred colony with all mice being HO for a particular locus, and those would not be bred B-S (brother-sister). The athymic nude mouse, for example, is an outbred mouse with a recessive trait for the particular allele, and mice can be either hairless (HO-recessive) or have fur (HE or HO-dominant).

Inbred mice should be bred Brother-Sister to reduce the chances of genetic drift (which still happens after 20 generations). If we intend to use the “same” inbred mouse over the years, we should resuscitate original embryos as we approach the 20th generation. The nucleus colony (smaller group of few pairs that produce future breeders) would ideally always be paired B-S, whereas this may not be as critical for the production colony (larger group of breeding pairs that produce animals for use). Because litters will not offer 50% of each gender, you could use the spares from separate litters for the production colony. Depending on the production requirements, there could be a middle size colony called expansion colony, also bred B-S.

I agree entirely that inbred mice should be B-S mated to reduce the chances of genetic drift. And if you want to maintain your "inbredness" then B-S for your nucleus colony is what you should do (although bearing in mind the pitfalls of fixing an unwanted gene, checking for it and dealing with it if/ when it occurs - which is something that is not always done well in academia). However - that's not what we (my specific units) are aiming to do. We are aiming to keep a fixed gene for one (or more) loci on a C57BJ background - i.e. we don’t want to further inbred our colony (and therefore move it further away from a "true" C57B6J). We therefore don’t B-S mate and refresh with commercial WT genetics more regularly.

I am not arguing for a particular approach - I'm arguing that the scientist who is using the colony should be aware of what they are wishing to achieve scientifically and discuss it with the vet so the correct local strategy is used for the desired outcome and that the strategy (and strain nomenclature) is correctly included in relevant papers so others know what is done and therefore can make decisions based on it if they wish to try and keep the line or reproduce the studies.

I am just wondering what others have as guidelines for breeding mice (primarily GA lines on C57B6J background) or if you have any thoughts/ comments on our strategy laid out below. I am told that we are causing a loss of breeding time by a new PI who thinks our guidelines are too draconian. These are set out below, but I should add that we are always prepared to allow animals smaller/ younger/ older, etc. on a case by case basis. The PI however wants the guidance changed. I am reluctant to do so, as I would prefer a structure which we stay within UNLESS there is a positive action to make a different decision:

- matings not set up before 8 weeks for females (we will consider earlier if the females are big enough), 6 weeks for males (to reduce the weight differential

- matings not set up with females <18g (and we prefer bigger) - based on mean weight of females (for general C57B6J) being 18.7g at 8 weeks, with a S.D. of 1.2g

- a pair will not be set up if the male more than around 20% heavier than the female, especially if the female is <18g

- we generally breed as monogamous pairs for colony maintenance, and trios (with males removed once the females are obviously pregnant) for expansion breeding for specific studies

- males removed from pairings once the female is 6 months old (she is kept for a couple of weeks to see if she is pregnant and allowed to have that litter is she is). If the pair is breeding well at that 6 months we will extend based on breeding performance (if the number weaned is reducing significantly or they lose a litter having previously been successful, we would call it a day)

- we recommend taking down a pairing if the female is not obviously pregnant within 6-8 weeks of the pair being set up OR if the pair has two consecutive failed litters (no pups weaned) OR have failed to wean pups in 2 out of 3 attempts

- recommended minimum colony size is 3 pairs, with the parity preferably being spread over a range so that pairs are replaced sequentially and not together to try and avoid the higher loss rate of first litters causing a hiatus in breeding

The only thing we do differently is pairing the animals at the time of weaning and keeping the trios like the breeding pairs, i.e., the male is kept within the cage when the females litter down, and always brother-sister matings.

Most of our units would have a similar approach to the one that you suggest and they don't pair at weaning but wait until the mice are 6-8 weeks of age. (Sometimes because the genotyping results aren't back from the PI in time for weaning).

We wouldn't recommend brother-sister mating for a GA line, as the idea is to have whatever the gene is on as near a "normal C57BL/6 xxx") background as possible.The lines should be outcrossed back onto "commercial" WT periodically to maintain their integrity (or revived from cryo if it's a complicated one).

Regarding the number of pairs - this is another ongoing 'items for discussion' as many researchers ask to have a tickover colony with as few as 2 breeding pairs, usually for financial reasons. We strongly advise against this, as it's highly likely to fix unwanted alleles in the population (genetic drift) but also very likely that they'll lose the line entirely if the mice fail to breed well. We advise 4 pairs minimum, assuming that the line breeds reasonably well. NC3Rs /MLC have some excellent information about maintaining GA colonies to which we direct researchers but in practice.

One thing that we want to do sometime is to cost up the intermittent breeding system vs the 'tickover' system in one of the facilities and just see whether the cost differential for colony maintenance is more perceived than real.

Brother-sister mating is the only mating system that guarantees the inbred genetic background. Doing any other will increase the genetic drift of the GA breeding colony. Of course, I presume you have the gene of interest in an inbred background and not in a mixed one. You backcross every ten generations or when you discover a significant departure from your inbred background, but most people don’t bother to do a genetic control even though they are really cheap today. So sticking to the theoretical ten generations works.