Healthy clones: Dolly the sheep's heirs reach ripe old age | Hacker News Reader

Healthy clones: Dolly the sheep's heirs reach ripe old age

reuters.com

111 pointsbenologist25 comments

This outlines one of the problems with genetic engineering. The debug/test cycle is a few lifetimes.

That's also the problem with antibiotic resistance due to natural selection - the test/debug cycle is only a few lifetimes.

>This outlines one of the problems with genetic engineering. The debug/test cycle is a few lifetimes. Well, with CRISPR ([1]), we at least can hot patch.

1. https://en.wikipedia.org/wiki/CRISPR

And wait the rest of the lifetime to see if any additional defects were added.

Unless models are accurate enough.

Well, that's the point: we can now (mostly, theoretically) fix the issues when they appear, instead of hoping they would never show up, and the costs of fixing will only go down over time.

Not even close, not to mention you have no clue about the impact of genetic engineering on future generations.

That literally describes every major human endeavor from power production to resource extraction to biotechnology to agriculture to government to experimental physics research.

With most of these, if you notice something going wrong, you have the option of not continuing. Biotech is significantly different in that respect.

I know this is tongue in cheek, but I think it's instructive to look at different "development" disciplines and see both why they're different and see what we can learn from them.

For example, in genetic engineering, given the constraint that the debug/test cycle is lifetimes, they use different approaches:

1. Plan, model and simulate

2. Use living models that have fast lifetimes (bacteria, fruit flies, etc.)

3. Use results from #2 to improve #1 and continue to tighten the feedback loop.

In traditional engineering, there's groups that do nothing but test the stress points of different materials, subtract out a safety margin and put that into design databases so when you build a bridge, you know exactly how much steel to buy to make sure it doesn't fall down.

and so on.

It's also instructive for other engineering disciplines to learn from software (and why it's very different). Lots of traditional engineering disciplines look down on software engineering because, from the outside it looks a lot like beta-testing in the wild. But part of building software is doing so and providing working software!

I have worked places that put very close to zero bugs in the wold, and buggy messes. The problem is customers don't care or are already locked in. How often do the people buying corporate software actually use that software?

I am now imagining a world where people can code a genome in a REPL or IDE. When they are happy with the results, the print it out in a large 3D printer. If they don't like the organism they created, then they kill it and throw it in a recycling bin so that the materials can be reused.

I've read a fantasy book where this was solved by moving genetic experiments to a faster-time universe and checking out on them every few generations.

Makes for a good headline but genetically identical (clones) of mice and other animals have been routine test subjects in pretty much any study for a very long time now.

Curious, but how does a researcher acquire cloned mice ? The JAX Mice, which are the most-published mouse models in the world[1] does not offer clones for sale on their website. Anyone know the source and cost ?

[1]https://www.jax.org/jax-mice-and-services

You'd probably have to make them, or get a colleague to.

Reading over this study, http://www.cell.com/cell-stem-cell/fulltext/S1934-5909(13)00... and some of its citations it's within the means of a 3 person team.

I'm guessing GP is referring to the fact that most popular mouse strains for research are, in theory, isogenic (genetically identical). This is achieved by inbreeding the mice for multiple generations until they are statistically certain to be homozygous at every gene locus. The standard is 20 generations. Thus making all the mice in a strain as close genetically as clones or twins

There are of course issues with random mutation. So the mice end up not being _exactly_ identical. JAX actually puts a lot of effort into trying maintain the genetic stability of their strains through methods such as freezing founder sperm and rebreeding it into later generations.

Naive question: do the clones have to have all genes for recessive disorders eliminated so they can interbreed without risk? If so, how did they get that?

They probably simply killed the ones with [noticeable] problems.

I've never done it, but I think those are just naturally bred out by breeding healthy mice from each successive generation.

Many of the strains they sell aren't healthy. If you enter a condition or disease like 'lung cancer' or 'paralysis' in https://www.jax.org/mouse-search, you'll find they sell a mouse that's susceptible to a form of it.

For example, https://www.jax.org/strain/004863:

"Homozygous Snap25 knock-out mice exhibit neonatal lethality. Late embryos lack spontaneous movement or sensorimotor reflexes, and abnormal innervation and muscle fiber development is observed."

Breeding mice that die at birth is an unsolved problem I'm sure they are thinking of, so they sell the heterozygous variant (carriers of the gene)

> Breeding mice that die at birth is an unsolved problem I'm sure they are thinking of

That sentence is both horrible, absurde, macabre and logical at the same time.

For anyone interested the museum in Edinburgh has put Dolly back on display in one of its new galleries

Is it definitely the original? I've heard there are a lot of copies on the market.

Other than being kept outside rather than in a barn, what are the reasons for these clones having longer life expediencies than Dolly the Sheep? Improvements in this somatic cell nuclear transfer process?

The paper in Nature Communications: http://www.nature.com/ncomms/2016/160726/ncomms12359/full/nc...

Reply on news.ycombinator.com