Yeast alpha tubulin western

[mkc3]

Hi,

I have a nagging problem with the detection of S. cerevisiae alpha tubulin (Tub1) that I hope to get some advice on. When running SDS-PAGE gels on yeast lysates followed by a western using either YOL1/34 or YL1/2 (rat monoclonal antibodies specific for alpha-tubulin), I can detect a single ~110 kDa band but nothing at ~50 kDa. When I include human cell lysate (from U2OS) as a control, I get a ~50 kDa band. The yeast band appears to be a dimer, and is of about 2-3x the intensity of the human band.

I have also used an anti-beta tubulin mouse monoclonal (KMX-1), in which case I get a very weak band for yeast running at ~50 kDa, and a much stronger one for human cells (5-6x the intensity of the yeast band) at the same molecular weight.

I have tried preparing lysates three ways now, grinding in liquid nitrogen, glass bead disruption (using a modified RIPA buffer that I also used to lyse the human cells) and Novagen YeastBuster (using the manufacturer's buffer and reducing agent). All give the same result. I have also tried lysing 6 different wild-type strains just to make sure the problem I face is not specific to our lab's wt strain.

Btw, I boil my lysates for 3-5 min after adding hot sample buffer with BME.

I am wondering if somehow the yeast tubulin heterodimer survives boiling in sample buffer Does anyone have any ideas/suggestions?

Thank you.

[mkc3]

Just an update - I've determined that with YOL1/34, I consistently get a single band of roughly equal intensities in human and yeast lysates, just that the human alpha tubulin is ~50 kDa and the yeast one seems to run at ~110 kDa. With the KMX-1 though, it is clear that both human and yeast lysates give a ~50 kDa band (the yeast band runs a little bit higher on the gel) but that the human beta tubulin band is about 50x as intense. So I'm increasingly leaning in favour of the idea that the yeast tubulins are running as a heterodimer in my SDS-PAGE gels, with a small amt that dissociates, whereas the human tubulins dissociate and run as monomers.

[mkc3]

So the problem turned out to be Calbiochem, which I first purchased YOL1/34 and YL1/2 from. After buying YOL1/34 from Abcam, this problem was resolved. I started getting comparable signals between YOL1/34 and KMX-1 both in yeast and in human cells.

Millipore, which has acquired Calbiochem, refused to entertain any queries as to the identify of their anti-tubulin antibody clones but I will no longer purchase these from them and I would advise others to do the same.

[pigfarmer]

Thanks, I will heed your advice. I usually stick with Pierce (under Fisher), AbCam, or SCBT.

[mkc3]

Originally Posted by pigfarmer

Thanks, I will heed your advice. I usually stick with Pierce (under Fisher), AbCam, or SCBT.

SCBT's goat polyclonals raised against N- or C-terminal epitopes of yeast proteins also tend to perform poorly, in my experience. However, when raised against whole protein, they can be very potent

[rjpalumbo24]

Originally Posted by mkc3

So the problem turned out to be Calbiochem, which I first purchased YOL1/34 and YL1/2 from. After buying YOL1/34 from Abcam, this problem was resolved. I started getting comparable signals between YOL1/34 and KMX-1 both in yeast and in human cells.

Millipore, which has acquired Calbiochem, refused to entertain any queries as to the identify of their anti-tubulin antibody clones but I will no longer purchase these from them and I would advise others to do the same.

I too have been having this issue for about 2 years now myself with the YOL 1/34 clone (from Chemicon International). My advisor didn't care that Tub1p wasn't being detected at the right molecular weight because for whatever reason, it was mostly constant. Our post-doc was convinced it was cross-reacting with something else. I was convinced it was dimerization, as discussed below.

I have noticed that sometimes, independent samples will have a bit less of the 110 kDa band (which is blindingly intense) and a faint, faint band around 55 kDa. I made a fascinating discovery when trying to transfer small molecular weight proteins in the absence of SDS (semi-dry, Bio-Rad). Under these transfer conditions, bands were present at 55 kDa, in addition to the very intense bands at 110 kDa; however, sometimes, there are no bands at 55 kDa, and no noticeable corresponding increase in the amount of 100 kDa. So these two observations had me convinced that my lysis conditions were causing the majority of my Tub1p to run at 110 kDa—presumably due to dimerization. We boil at ≥100ºC in SDS+DTT for 3-10'. and I know that that can cause some proteins to aggregate (I am also having this issue with a new target protein), so I was going to considering denaturing at 60ºC.

However, this is good news, that an Abcam anti-Tub1p antibody gives bands at 55 kDa. Are they as intense as the bands you were detecting at 110 kDa? Or as faint as the ones you could sometimes detect at 55 kDa? As I say, sometimes I see less 110 kDa and more 55 kDa (and via SDS- transfer, even more 55 kDa, or none at all), which sounds like a dimerization issue—wherein the majority of Tub1p would be dimerized. However, were your Abcam antibody to detect the 55 kDa as well as the Calbiochem at 110 kDa, then that would suggest that either the Calbiochem/Chemicon antibodies preferentially recognizes dimerized Tub1p, whereas the Abcam preferentially detects independent Tub1p—both of which could arise from how the antibodies were raised. This still presents a problem because, why then do I sometimes detect less 110 kDa and more 55 kDa in single lanes only, and why is some proportion of Tub1p dimerized (again, likely due to lysis conditions, perhaps even as specific as individual tubes being prepared simultaneously)? If the degree of dimerization is not consistent between samples, regardless of which antibody is used, this could lead to inaccurate relation of a target to the Tub1p control.

Any thoughts on this would be greatly appreciated. It would be nice to put this Tub1p issue to bed.