Hi Michael,

the most important aspect of dealing with low abundance samples is sample prep. Once the sample is in the MS, there is little you can do to improve a poor quality sample. Any of the mass specs you list should be fine for the analysis. There may well be variations in how they are setup and it may well depend on what you want to do. Are you just after identification, or are you quantitating in some way?

Back to sample prep. Have the pieces you are using been silver stained? Was an MS compatible stain used? At the very least you will need some sort of high sensitivity clean-up. We use the DIY Stage Tip setup, but there are many commercial offerings out there. What PTMs are you investigating? If it's phosphorylation, then you are obviously looking at TiO2 or IMAC for clean-up. We prefer TiO2. The PTM you are examining may well affect the MS choice, as some PTMs are better identified using HCD, ETD, etc. Given the choice of mass spectrometers, your lab should have experienced staff that can assist with the best MS setup for the samples you'll be preparing.

Finally, if you haven't prepared similar samples in the past and the samples are precious, I'd suggest you run some sort of test with standards (at a similar quantity level) to make sure you are confident in what you are doing before committing to the real samples.

Good luck.

The final paragraph in Peter Hains' comment is important. Before you try to analyze results from an experiment you should make sure you can analyze data from a known sample. Don't do actual experiments at the same time that you're learning how to do them.

:)

Michael, all of the three MS will do a good job in principle for what you are looking to do. I would recommend to look at your experimental setup more closely though:

- you are stating that you are analyzing a pulldown. Do you have controls to analyze, to rule out non-specific binding as the source of your proteins/PTMs? Without a propoerly designed control, the result from a pulldown is more or less meaningless

- if you are using enrichment then the concept of using control samples becomes somewhat stretched due to limited reproducibility of e.g. TiO2 cleanup. If this sort of result is crucial for you, you might want to consider SILAC labeling (heavy sample, light control) for your experiment to correct for differential loss

- tackling PTM analysis from faint silver-stained bands has a high chance of failure - at this level of abundance, it is easy to lose/overlook sub-stoichiometric PTMs

- for literature, look for Nesvizhskii AI, Gingras AC, Aebersold A

Thanks a lot for the answers!

No, I`m not trying to do quantitation with these samples. The staining was done in an MS compatible, but I didn`t do any stage tip cleanup afterwards, I just eluted from the gel band, dried and resuspended. Why is it helpful to use stage tips afterwards?

I`ll try to come up with a sample that`s similar enough to do the optimization on it. Won`t be easy though.

The design of the pull-down should be OK, with appropriate controls. I`m pulling down on the protein level, then separate on the gel (mainly to get rid of the abundant bait protein), cut in slices and digest. There`s no enrichment on peptide level, because the bait seems to work better/only with proteins.

Thanks again.

Why clean-up the sample? Generally it's a good idea. Even though you are going to separate your samples on a RP system, it is still a good idea to clean-up (and as such desalt and concentrate) those samples prior to loading onto the LC system. This will keep excess salt out of the MS system (this may depend on your setup) and ensure your sample is a high quality before loading onto the LC-MS.

Even though it doesn't seem to make sense to do a small scale "RP" clean-up prior to loading your samples onto an RP system, I think you'll find the vast majority of people take this approach. Do a comparison and see for yourself why it's a good idea.

Dear Michael,

One important thing to take care of when you go down to the silver stain level is to avoid plastic as much as possible. Once the peptides are released from your gel slice they will slowly get (irreversibly) adsorbed to the plastic walls of your tube where you keep them before MS analysis. So use the "low bind" tube from Eppendorf and try to run the LC-MS asap after digestion.

Good luck

Hi, thanks again for the answers.

We normally do clean up after in-solution digestion, but consider them clean enough after the in-gel digest in which we do quite a lot of ACN washing. Probably it`s still a good idea to do it.

Thanks about the tip with the low bind tubes. I think that`s very important.

I`ve been trying to come up with a good standard for optimizing the conditions. I`d like to spike some known modified peptides into a background of about the same complexity as my sample. However BSA as background seems of too low complexity and a (bacterial) crude lysate too complex. Does anyone have an idea what I could use, i.e. something that represents about 50 proteins or so?

Thanks

Dear Michael,

If you could put your hands on an E Coli periplasmic extract this should contain around 300 proteins. Be careful that the dynamic range is also important to take into account, try to stay below 10e4.

Enjoy.

Regards

Sounds good. Have you ever used that and if so, how do you prepare it?

Thanks

Yes I did used it but I didn't prepared myself. There are protocols around in the literature and I think you play with an osmotic shock.

Could try to find you a detailed protocol if you want me to.

Best regards.

Dear Michael,

Silver staining is not a good staining even if it is compatible with MS. Dynamic range is very poor, each protein is stained differently and is a very strong oxidant that modified proteins. I recommend you use Sypro Ruby, is a very good choice!

On the other hand, you are interested in any of the bands you detect? If not, you better run the sample in a stacking gel to concentrate the whole "pull down" in the stacking/ressolving gel interface. Then you cut the band and digest; you must desalt always if you want to obtain a good MS analysis. Be careful with low bind tubes, in our hands we get a lot of interference in the LC-MS analysis.

Best regards,

Anabel

Mol. Cel. Proteomics 2011, 10: 1-14

Thanks.

The main reason to run it out on the gel rather than doing in-solution digestion or, as you say, just run it a bit into the gel was so that I could remove the bait. Not sure if that was a good idea. I was thinking recently that maybe I should have tried proteominer beads, however that might have lead to quite huge losses.

When we use Eppendorf low bind tubes, we normally don`t have problems with contaminants, but recently we tried to switch supplier and encountered the same problem. Which tubes do you use?

Do you by chance have a reference for the Sypro Ruby vs Silver or is it in the MCP paper?

@Didier: Thanks, I`ll try if I can dig something up. I`ll come back to you if I can`t if that`s OK

Michael, in our lab we use Coomassie staining by default, however we do use it to cut bands. Instead, the whole lane is cut into 11 or 23 equidistant slices, which are digested in-gel and analysed by nanoLC/MS/MS, so basically we analyze the whole lane. MS as a detector is nowadays significantly more sensitive than gel staining, so it will detect proteins even where there is no discernible stained band. The Coomassie serves rather as a QC and to monitor the running of the gel; we choose Coomassie over Silver mainly because it is more readily compatible with MS, and cheaper.

...sorry for the mistake - we do NOT use the Coomassie staining to cut bands.

Do you really need to run a gel? If repetitive performance of the protocol you used resulted in similar pictures, maybe you can try to perform in solution desalting and digestion for a higher yield.

As stated above I would not run it on a gel and instead just perform a in solution digestion. In-GeL digestion is good and all when dealing with highly complex samples but you will also get sample loss (depending on starting material), especially at the peptide level. Since your samples are pull downs the complexity and the yield I would imagine is already pretty low, no need to make it worse. Try just a normal in solution digestion followed by stage-tip or some type of C18 material clean up. Run on a long LC gradient of 2-4 hours.

Thanks, everyone. Yes, my tendency now is also to avoid the gel and rather go for in-solution digestion.