Hi Maria,

I did a many plasmid extractions from yeast (regular plasmid extraction and a couple of screens) and I never got any big unexpected size changes. I would suggest 1) sequencing the insert ends (over the multiple cloning site) 2) re-transformation of the plasmid in your mutant yeast cells for a functional assay. In addition, the expected and visible size of DNA might be a little bit different with Agarose gels.

I hope this helps.

Hi Maria,

I agree with what Thomas wrote, and would add that I'd check several E. coli transformants for their plasmis (I imagine that after you isolated the plasmid from yeast, you then transformed E. coli). Perhaps the E. coli transformant was an odd-ball. It is possible that it was altered in E. coli - if so, you could use a STBL E. coli strain to prevent rearrangements.

It is always possible that the plasmid was altered in Saccharomyces if that conferred some advantage (I haven't seen that, but it is possible). You'd want to figure out what Saccharomyces didn't like about the construct and then how to avoid this.

Good luck.

Hi Maria

I am involved in yeast genetics and had extracted number of times yeast plasmid.

I never faced such trouble of alteration but yet some times plasmid get integrated some how into the genome. However, u could opt for a new transformation rom new preparation of plasmid from source and then again extracting total DNA from yeast . Re-transform it in E.coli. I hope it would work as per my experience.

Recombination and integration may be reason of alteration in yeast but should conform your insert through PCR using gene specific primers. and untransform strain used as negative control.

Such problems are frequently encountered while trying to recover plasmids from S. pombe in E. coli. A solution I found was same as suggested by Ronda. Take 4-6 independent E. coli transformants and recheck the different isolates for complementation of the yeast mutation. Hope this helps.

I have encountered near the same problem as Maria described. I have been working with the large plasmids containing beta-galactosidase reporter constructs about 13 kb. My colleagues and I have found that if such a big plasmid is transformed into yeasts by Li-acetate method they did not show any enzymes's activity. Plasmids purified from yeasts have altered restriction map and fragments of unexpected size are excised from them. The same picture is observed for old yeast, let say more than month old. I suspect that some plasmid rearrangements are possible and there is a literature on this theme. The rate of plasmid rearrangements is very low and could not be detected under normal conditions. I think that inserts of prokaryotic origin is prone to be subjected for rearrangements because they could not form normal chromatin structure that protects DNA from such alterations.

The solution of the problem is very simple - use DMSO in transformation mixture. For unknown reasons yeast transformation with DMSO protects plasmid DNA from rearrangements. However, you can still obtain the "wrong" colonies, but this time they are only comprise a very minor fraction: 2-3 colonies are sufficient to find the right colony.

Thank you everybody for your suggestions, they are very helpful.

I will indeed have to sequence into the gene from both sides to see whether I have at least the right sequences flanking it.

Unfortunately, complementing the yeast mutation with the re-isolated plasmid will require to first make a second vector with a different auxotrophic marker containing my suspicious insert. As I said, the gene of interest is an essential gene, i.e. plasmid shuffle is necessary to always retain a copy in the cells.

I suspect that there might have been homologous recombination, since it is an essential gene. However, is it known whether recombination is more likely for essential genes than for non-essential genes, when you transform plasmid copies into deletion backgrounds?

I can rule out an individual "odd clone", since I already had two different preps from two independent yeast colonies (clones #1 and 2), each of which were transformed into E.coli. From each of these transformations, I made minipreps of 3 E.coli-colonies each (named #1.1-1.3 and 2.1-2.3). And strangely, all of them have the unexpected pattern on my agarose gel.

Also, the uncut plasmids run at a different size than the single digest, so I assume the vector was properly linearized. The double digest I then performed to cut out my expected insert also gave me 2 bands, one of which had the (expected) vector size, the other one (insert of gene?) being smaller than expected.

Also, regarding the EtBr supercoiling effect: We use SyBr in my lab, no idea whether this could have the same effect. But it's also intercalating.

Anyways, thank you all for your help.

Dmitry, that is very interesting!

My plasmid is supposedly ~10 kb large, so rather big. My problem is that we obtained only the yeast strain from another group, neither the plasmid nor the E. colis. The yeast was fresh, definitely not kept on plates for one month.

I will keep in mind to use DMSO in the LiAc transformation of large plasmids. How much do you normally use? And is it also possible to include DMSO in electroporation?

yes, this type of problem is encounted sometimes. This could be due to recombination, deletions , etc happening . Check this by doing a PCR. To avoid this use different host vector systems.

I have had the same using a pRS-based library for complementation, never being able to recover the complementing clone for a mutant. Using a library based on a different vector (good old YCp50) we were able to recover intact clones, for a gene we later also found to be essential. Indeed, the insert had integrated in the pRS-case.

So besides the transformation procedure, the vector backbone may also be a factor.

Yuriy, I'm assuming Maria amplified the yeast plasmid in E. coli, and I don't expect the natural 2mk plasmid to have an ampicillin resistance cassette. Therefore she should not be having this in het restriction analysis.

Yuriy, I passed through E.coli. Also, the size of the plasmid is not close to the one of 2µ (6.3kb).

Maria, I think DMSO can be used in yeast electroporation. I saw electroporation protocols for mammalian cells, for example Melkonyan, H., et al., "Electroporation efficiency in mammalian cells is increased by dimethyl sulfoxide (DMSO)," Nucleic Acids Research 24 (21), 4356-4357 (1996). That means that DMSO in principle can be used in electroporation, so you just need to find optimal concentration. DMSO also affects membrane so may be it should be added in decreased quantity compared to the traditional DMSO-LI-acetate method in order to avoid cells death.

So, I had the plasmids sequenced from the 5' and 3' end of the MCS. Apparently, there is a whopping 2.5 kb (the first half) of the gene missing, including the ATG.

I suspect that the original clone was already wrong, since both independent preps show the exact same altered sequence, and also it corresponds exactly to an internal restriction site (the one I think was used for cloning). I have no idea whether the whole problem occured only after transformation and maybe recombination in the original deletion strain. Strange though that the strain is still viable in my hands, since without an ATG I don't think there is any protein expressed....