Hi Charissa,

the best option is to perform an sds page and stop the electrophoresis right after the migration front has entered the separation gel. This results in a single protein band right at the border between stacking and separating gel. It is sufficient to load about 5-10 µg of protein per lane. Staining should be performed by Coomassie.

If run on a decent MS, the in-gel digest will result in about 600-1000 identified proteins.

Hi Charissa,

the best option is to perform an sds page and stop the electrophoresis right after the migration front has entered the separation gel. This results in a single protein band right at the border between stacking and separating gel. It is sufficient to load about 5-10 µg of protein per lane. Staining should be performed by Coomassie.

If run on a decent MS, the in-gel digest will result in about 600-1000 identified proteins.

Good suggestion, Sascha.

To digest the whole lane is possible but it takes more reagents and the sample may be diluted, so concentration would be required after the digestion.

If you are going to digest the whole sample, digestion in solution may work too.

Depending on the complexity of the sample you are running you could also let the gel run completely to the bottom and cut up the entire lane into different fractions. That way you can also make sure about the size of the protein you are looking for and because of the various fractions you have a better chance to identify your protein of interest. We did this to identify a protein expressed in plants and ran the total protein extract on the gel and identified our protein at the correct size. If your sample is however already halfway purified or if you expect to have high relative concentrations you might not need to do this.

I would start with Sascha's suggestion above and if you find nothing try fractionating. In solution digestion or on column digestion would also work and would give similar results as the single band digestion.

Also the MS detection procedure you are using would make a big difference. We used LC MS with a C18 column and acetonitrile gradient coupled to an orbitrap.

Normally, protein samples are run on a gel so that a particular band / bands can be chosen for subsequent identification, or as Jihui suggested - to make digestion and identification easier. If you wanted to go down the path of cutting out the whole lane and performing digestion on the whole lot, then you would be undoing the process of gel separation.

Catherine's objections are obviously true; however, you remove a lot of unwanted sample matrix, e.g. dust, particles, salt, detergents, lipids etc., and in addition you efficiently denature the proteins by the in gel procedure. You might loose very small proteins >10 kDa during staining.

For the protein ID it is not a hug gain to know the actual mass of the protein, as it is done by MS/MS using the amino acid sequence. It is; however true that you lose the information on the structural integrity of your target protein

Hi Charissa,

The suggestion of Sascha is absolutely a good idea that i have performed. Nervertheless, as Cobus has mentionned, the mass spec is crucial and an orbitrap is the best way to analyse your sample. I think you should test, if possible (dependently of the nature of your sample and the need of solubilisation/denaturation of some protein with SDS), an direct analyse of the "crude" sample with an orbitrap mass spectrometry without an electrophoresis step.

I would recommend separating the proteins by SDS-gel electrophoresis, it will be purified (more or less), you will obtain the MW (+/- 5%) and in addition protein ID by mass spec. If other tests are done, for example immunoblotting, that will add info as well for the particular band.

Thank you all very much for your very nice answers. Very helpful for someone like me who is trying to figure out the best approach.

To give you a little more background, the proteins I want to try to identify are secreted in medium (the medium doesn't give a background) after which I concentrate them and wash in ammonium bicarbonate containing DTT to get rid of salts etc. Afterwards I checked my samples making use of SDS-PAGE, which makes me think I have figured out my set up for protein secretion, collection and concentration. One of the abundant bands I digested and analysed with the Orbitrap gave nice results.

Now, I eventually want to try to identify all proteins that are secreted using TMT. And from your answers I conclude there are two ways to go about this: 1) run the samples into a gel, excise the band for trypsin digestion (such as Sacha suggested), and 2) do an in solution digestion (what would be the best starting conditions for that?) I think it's safe to say the proteins I look at are all soluble.

The following is more an opinion than an answer: I would stick to the in gel approach, as you already have experience with these digests. They are very robust, in respect to chemical composition of the sample, and sample amount.

If you are interested in the secretome, you have a number of concentration steps. Your proteins of interest usually end up in the pellet - together with a bunch of dust, hairs cells debris etc. It is good to get rid of those components, as they might mess up your analysis.

The in solution digests have a number of pitfalls - all can be overcome; however, this takes time. For the SDS page approach you have already a set of methods established in a your lab. TMT is based on the peptides after the digest, and is therefore strongly dependent in the reproduciblity of your digest. Establishing solution based digests from literature, if you do not have somebody close by with experience, will take about 3 month to yield reproducible results.

I completely agree with sascha. Your own experience with gel-based methods is defining the first way to examine your sample. If you have somebody close by with experience in in-solution digestion and direct mass spec, so do it but first try with gel based approach!! Good luck!!

YES. It is possible, but be aware that even a single stained band has more than one protein in a SDS-PAGE resolution. So you should expect to have different levels of complexities of protein mixture in the whole line, therefore you should consider to use a strategy to reduce this complexity. One possible strategy is fractionating proteins in classes based on chromatography (either affinity, size exclusion etc) and divides the lines in zones based on protein separation in the SDS-PAGE. By using this strategy I could, safely, identify about 80 different proteins. I have seen articles showing that you can identify more than 200 proteins in a complex mixture of proteins from gels. You can also improve your resolution on the identification of the protein by using suitable strategy of analysis of your peptides sequences on your MS/MS spectrum.

Good luck,

Hello Charissa,

Use the gel approach if you are limited by the protein amount in the starting material. If your proteins secreted in medium are giving you more than 50ug of starting material, then I would use a modified version of Filter Aided Sample Preparation (FASP). Simply use a Amicon 3 KD MWCO filters and concentrate your sample. This way you concentrate your sample and get rid of unwanted salts and other small material that you don't want in your sample IN ONE STEP. We have been very successful with this method for concentrating proteins in urine prior to Mass Spec. I can provide you with a more detailed protocol if you wish. Best of luck

This is just an opinion based upon my experience alone.

1, In my experience in-gel digestion has always yielded less peptides after LC-MS/MS compared to solution digest. These samples were relatively clean and concentrated, they contained only salts and detergents, but not dust etc. Peyman's approach should help you concentrate your sample and with the rest of the issues.

2. Do you want to quantitate or identify the proteins? Again, in my experience TMT, or iTRAQ for that matter, labeling leads to a decrease in the number of peptides identified in an experiment. Therefore, if you want to only identify the proteins and can run multiple LC-MS/MS experiments, you should consider not using the TMT labeling.

3. Solution protein digest will require less manual labor that in-gel digestion.

all the best

Thank you Peyman for your very nice offer of sharing your protocol. Please, if you can get around to it I would love to have a look at it.

For now you all convinced me to try the in gel digestion first to see how that goes. Good point mentioning the reproducibility. I am incubating new samples now and will harvest and prepare them towards MS Monday/Tuesday. Depending on how occupied the Orbitrap is I hope to run them asap after that and will get back to you with my results. Have a nice weekend!

Hello,

Why don't you use just in solution digestion in this case? If you don't want to lower sample complexity than there is no meaning of loading the sample on gel.

We would take a rather different approach to in-gel, which inevitably needs many more LC-MS/MS runs, and can give poor yields. We would take the secretome and concentrate it, followed by a global analysis of everything present. For us, the ability to see low abundance proteins outweighs any clean up that a gel might give. From a secretome, with no pretreatment other than concentration, we'd expect to see about 1000 proteins.

Incidentally, we have that FASP consistently underperforms against other methods. A secretome sample in particular is largely soluble proteins, and the gains from FASP may not be apparent.

As regards concentration, then precipitation can be effective, but we would always use a digestion enhancer (we are enthusiastic about Rapigest but 'other products are available ':-) ) to ensure consistency of digestion.

Since your study is comparative, why not start with a label-free analysis? You can then decide which labelling protocol to use after wards. I'd also consider metabolic labelling if you can.

If you use MCO filters, I suggest giving them a good wash first by passing buffer through repeatedly. They are full of plasticisers or humectants (glycerol, PEG, etc) that can play havoc with the MS analysis! Blank runs are your friend here and make sure you look at the BPI chromatograms and the spectra of very large peaks.

Rob Beynon

It is worth considering Robs answers as he is well regarded in the field and has published a number of papers in the field.

The question is what you are trying to achieve. If you are just trying to characterize the proteome as a whole then digesting everything will cause problems at the LC/MS unless you are doing a 2D LC-MS approach because there will simply be too many peptides for the MS to capture so you will lose one hell of a lot of data. We do SILAC and we have also characterized un-labelled proteins in the same way. We pour mini gels and use a five well comb and we load up to 50 ug of protein per lane (ideally triplicate biological replicates, but certainly technical replicates) and cut the lanes into 10 equal bands before running each band on an LC-MS run. This way we get quite good coverage of the proteome. We find we have no problems with in-gel digestion and get good peptide yields.

In terms of quantitation- Label free is good but generally requires far greater numbers of replicates to get good data and you probably need to invest in a good piece of label free software- Progenesis LC-MS is very easy to use but costs a lot, there are fre alternatives but these require a bit more skill. We do SILAC and this is easiest if you are generating orbi-trap data as this plugs straight into MaxQuant which is free. It is a little more tricky if you use other MS machines although there are rumors of maxquant supporting other data types rather than .RAW in the near future.

If you do digestion for the whole lane, you will destroy the purpose of separating all proteins in your sample into different cluster of bands according to their molecular weight. I would suggest you to do protein digestion for each band so that the number of proteins could be minimized and it would be easier to handle when you do the MS analysis subsequently.

generally speaking，In-gel digestion of proteins separated by single bands is better, since, single band contains the proteins which have the same MW, and it can help you to check the results of MS identification.

Hello,

I was googling and I found this discussion so I will try to ask you guys.

I'm not in the MS field, I never done MS and I probably don t understand everything.

I looked at the 2013 Science paper from Mann ("Direct Proteomic Quantification of the Secretome of Activated Immune Cells

"), where they analyze using MS the secretome. I would like to reproduce this experiment but I need first to understant how to prepare the sample that is my cell supernatant. I found a lot of protocol for the preparation of cell lysate sample, but not so much about supernatant. In the paper they have a description of the secretome preparation but I have no idea about the volumes that I need to use(they have concentration)..for example If I grow cells in 48 well plate and I obtain around 500 uL of supernatant how much volume do I need for each sample for the MS?

Any suggestion?

Thanks!

As others mentioned before, you cannot directly do in-gel digestion of whole lane. However, SILAC Kits, provided by Invitrogen, had been used by my colleague for protein identification. This pre-label approach could let you combine the lane for MS. You may ask for some trial samples if you are interested in. Here is the link about SILAC information: http://www.invitrogen.com/site/us/en/home/Products-and-Services/Applications/Protein-Expression-and-Analysis/Mass-Spectrometry/MassSpec-Misc/Protein-Characterization.html. Good luck~

You are actually following GeLC-MS approach, where LC-MS/MS runs are done subsequent to 1-dimensional PAGE and in-gel digestion. Performing in-gel digestion of all the bands in a lane is similar to carrying out in-solution digestion, instead there is no need to do PAGE at all; in-solution digestion would suffice. Rather than working on the entire lane of a gel, you can combine 2 or 3 or 4 bands on a lane and carry on digestion on those 2 or 3 or 4 bands together. The step of peptides' extraction after in-gel digestion is important. If the peptides are not properly extracted from the gel onto the extraction solution, then the number of protein- identifications will get decreased.

In gel Digestion is generally considered when you know the size of your protein/proteins in case you need to see all the proteome better to digest and perform SCX-followed by reverse phase HPLC (as the recent paper from Pandey Lab : the map of human proteome).