I would generally agree with the authors of this paper, that targeted proteomics is a far more quantitative and robust method than western blotting for absolute quantification of protein levels. The argument that they make seems very clear and concise. I would add that one of the authors is an old colleague of mine, so I have a bit of conflicted interest!

My view is that one should use the right method for your experiment. For a lot of experiments, western blotting is absolutely the right way to go. It is very good as a "reality check" for confirming other results before you go into more expensive, complicated methods; with well characterised antibodies, it can deliver excellent results; and generally, it is a cheap and relatively quick experiment that is within the budget of most molecular biology labs to do without too much hassle. The investment of about £2,500 in equipment nowadays will provide a system that gets you from sample to result in half a day, with moderate human interaction and running costs. However, my final year UG students regularly use westerns as an easy experiment to "critically analyse" in papers, as usually only one result is given, it's not quantitated, and there is no analysis of reproducibility or errors.

Mass spectrometry has come on such a long way in the last ten years that experiments previously undreamed of have become routine. MS is a very powerful and well validated method, and the great advantage is that the validation cuts across different samples. The downside is that the instruments are very expensive, there is higher running cost, and in most cases you want someone who really knows what they are doing to run (and possibly analyse) the data.

The question, as ever, is what do you want to get out from your experiment? If you want a qualitative and semi-quantitative answer, quickly, then westerns will do fine. If absolute quantitation, with associated errors and methodological validation, is essential, I'd go for MS.

HTH.

I would generally agree with the authors of this paper, that targeted proteomics is a far more quantitative and robust method than western blotting for absolute quantification of protein levels. The argument that they make seems very clear and concise. I would add that one of the authors is an old colleague of mine, so I have a bit of conflicted interest!

My view is that one should use the right method for your experiment. For a lot of experiments, western blotting is absolutely the right way to go. It is very good as a "reality check" for confirming other results before you go into more expensive, complicated methods; with well characterised antibodies, it can deliver excellent results; and generally, it is a cheap and relatively quick experiment that is within the budget of most molecular biology labs to do without too much hassle. The investment of about £2,500 in equipment nowadays will provide a system that gets you from sample to result in half a day, with moderate human interaction and running costs. However, my final year UG students regularly use westerns as an easy experiment to "critically analyse" in papers, as usually only one result is given, it's not quantitated, and there is no analysis of reproducibility or errors.

Mass spectrometry has come on such a long way in the last ten years that experiments previously undreamed of have become routine. MS is a very powerful and well validated method, and the great advantage is that the validation cuts across different samples. The downside is that the instruments are very expensive, there is higher running cost, and in most cases you want someone who really knows what they are doing to run (and possibly analyse) the data.

The question, as ever, is what do you want to get out from your experiment? If you want a qualitative and semi-quantitative answer, quickly, then westerns will do fine. If absolute quantitation, with associated errors and methodological validation, is essential, I'd go for MS.

HTH.

I agree Nicholas. What is your view on using westerns to validate MS though? This is the predicament I currently find myself in.

Thanks for your response!

Melanie, I guess that it depends on the position that you are in - in other words, whether it is your PI or a fussy referee that is asking for the data.

If it's your PI, then you need to have the conversation about what he or she really wants to see. It might be that the western is the right idea, but (s)he has not communicated this to you properly. A good conversation will probably help here.

If you have the fussy referee problem, then you should talk it through with your PI. Usually responding to referees is a trade-off about doing what you can, quickly, to allay most of the concerns, and arguing away a small part of the concerns as intractable or disproportionate to the value of the data. It may be that finding a friendly colleague who is good at westerns and has some of the necessary materials, and buying in a couple of reagents, is going to be less hassle than dealing with other concerns. Sometimes you just have to do what it takes to get the paper out. You have spent many months or even years of your and others' time on the paper, and $10,000s of resources, so investing a few hundred dollars and a few days to neuter a referee's comments might be a very small extra. Hopefully more experienced colleagues will be able to point you in the right direction as to what is necessary and what is just too much to do.

Hope this helps!

Out with the old, in with the new ;).

We generally start with MS, but also like to support interesting or potentially significant IDs with Westerns. I like Nicholas Harmer's mini-essay on this.

With all the principle technological arguments in this discussion - it is also a matter of accessibility. Every biochemically oriented lab can establish Western relatively quickly and at moderate cost. The PIs and experimenters have full control (plus/minus the antibody specificity) over the experimental setup and timelines. Targeted MS, however, is still the domain of research MS groups, and to a smaller degree core facilities. Biochemists have to be ready to give away the control over the experimental design to others in this case, and likely also have to wait a few weeks before getting instrument time. In my experience this is the biggest obstacle for a more widespread adoption of the "Mass Western".

Thank you all for your input- it is much appreciated and useful.

However, I am more so trying to gauge the current opinion on whether using westerns to validate MS is an outdated request.

I come from a background in psychology and behavioural neuropharmacology, but have now found myself in the biochemistry lab. I have used MS to take a proteomic approach to my latest study.

I am aware that the majority of journals ask for validation of MS results using westerns, and it has come to the attention of myself and my colleagues that this appears to be something that is just routinely performed, without actually questioning whether western validation actually has any added value to MS results these days. In this way, we have found this recent paper by Aebersold and colleagues very interesting. This is not to say that westerns are no longer useful, maybe just no longer necessary for the validation of MS. I'd like to put forward this argument in my next publication to justify my reasoning for not validating MS using westerns. We have all of the equipment and expertise to perform both MS and westerns.

As I said, this is not my principle field of research and I am very interested in hearing what others have to say about this.

My expectation is that this will strongly depend on the type of result you are trying to publish. It is mostly accepted to use just MS for global proteomics studies. If you are trying to publish on individual proteins and their mode of action in a discrete biological context, however, it is already good practice to show results from two technologically independent experimental approaches (if available), so a convincing approach might be Western AND MS for quite a while.

Thank you Christof- Very helpful.

Hi Melanie. Essentially agree with what was said earlier. In a nutshell western is technically simple, cheap and commonly shared methodology and this goes a long way in explaining its gold standard status. However, in my experience don't expect to reproducibly see variations of less than 50% unless you're ready to run 4 or 5 gels. In addition WB has a serious signal to noise issue with cross-reacting epitopes, especially with low abundance targets (arguments against). On the up-side it provides info on the complete protein whereas MS requires fragmentation, so one can never be sure that the protein of interest (rather than splice variants, fragments etc..) is analyzed. MS has its own, distinct drawbacks and is even less quantitative unless to start going into more sophisticated and pricey variants such as SILAC-based analyses. So both are complementary but for routine analysis of a single target (your case), western is still the way to go in my opinion. Unless the referee/critic is out to lunch (which has been seen before…), I would quote the famous godess Nike and just do it; if you’ve never done it, WB is a very useful technique to learn…Good luck!

I think one thing that has been left out of the previous answers is the difference in mass spec approaches. Absolute quantification as described by Abersold and colleagues is more quantitative and precise than western blotting thus there should be no need for western blot validation. On the other hand, shotgun techniques like bottom-up proteomics, which identify hundreds to thousands of proteins are less quantitative and it is common to see western blots or targeted studies of specific proteins of interest.

Just one example: http://www.ncbi.nlm.nih.gov/pubmed/20336678

D1 protein ("main chloroplast protein") was easily detected by Western blot but not found with a proteomic approach.

Another drawback of a proteomic assay - you don't know real size of a protein detected.

On my mind, proteomics is good for a wide first glance, and Western blot - for a precise work.

And don't forget: Western Blots are still the way to go if you want to obtain (semi-)quantitative results about posttranslational modifications such as phosphorylation.

Ralph, I would not fully agree. Provided your PTMs are in an MS-accessible region of the protein, Targeted MS can be used in the same way as a Western. Contrary to Western it is per se sequence-specific, and is sometimes able to address features for which it is not possible to raise specific antibodies. Of course it also has drawbacks of its own that Western does not suffer from. As stated above, my expectation is that it will be beneficial to have both complementary techniques available for quite a while.

I have seen more than one protein system generate conflicting Western data between different researchers/publications: the differences were (potentially) explained via temporal structural variation (very rugged proteins), despite using the same protein sequence etc.

I have also observed fundamental image densitometry performed very poorly, despite being a much more well-established and simpler technique compared to mass spectrometry. There is no doubt it can be done well and accurately, but it is a mistake to think that a Western is either trivial or assured to be robust.

I agree with Aebersold, Burlingame, and Bradshaw for the majority of cases, and the tables are continuing to tip further in their favor with every year. However, Westerns remain invaluable for orthogonal confirmation, edge cases, and when there is a structural component to the question.

I agree with the authors. Targeted proteomics is far more quantitative, sequence specific and can also be used to quantify PTMs. I agree with Christof Lenz here that targeted MS is still the domain of research MS groups so for groups that are not MS oriented its easier to stick with Westerns and experimental procedures that are familiar.

The technical pros and cons for each of the two technologies are clear. But the real costs associated with each is not. It would help for those working in these areas to provide some real numbers in dollar amount. The high costs of antibodies give the impression that western blotting may be more expensive. By comparison, the mass spectrometry instrument certainly costs more, but if it is installed in a centralized MS facility to provide the analysis to users who only pay the cost of running those samples, does this still cost more than western blotting?

Since WB is so widely used, the pros and cons and the alternative methodology for quantifying proteins should be discussed thoroughly to publicize the problem.

Which technique works best for you and your skill set.

The small thing to add to this discussion - If you have good Ab's WB required much less sample than average MS run in centralized core facilities.

Hello,

as the discussion here is more focusing around Western blotting vs. mass spectrometry I would like to come back to the paper cited above. I think the key message of Aebersold, Burlingame and Bradshaw is to “posit the request to that the request to validate quantitative MS data by Western blotting is no longer justified”.

I have submitted multiple articles to Journals, where reviewers asked for validation of the mass spectrometry data by Western blotting. This can sometimes lead to a rejection of the publication. Especially in journals, which do not frequently publish data applying mass spectrometric methods, this could be an issue.

I think the article from Aebersold, Burlingame and Bradshaw can really help here to facilitate publications, which are using MS based methods for protein quantitation.

Also it should be understood, that mass spectrometry methods advanced in recent years and numerous publications indicated that the quantitation of proteins via MS can be robust, accurate, and reproducible. As the list of publications cited in the article of Aebersold et al. is rather short I would like to add two more articles related:

Molecular Systems Biology 4 Article number: 222 doi:10.1038/msb.2008.61

Citation: Molecular Systems Biology 4:222

Selected reaction monitoring for quantitative proteomics: a tutorial

Vinzenz Lange, Paola Picotti, Bruno Domon & Ruedi Aebersold

or

Monitoring protein expression in whole-cell extracts by targeted label- and standard-free LC-MS/MS. Katharina Bluemlein and Markus Ralser.

Nature Protocols, 2011, 6:6, 859-869

Best regards

Matthias

I agree with the authors that reviewers will ask for validation of the mass spectrometry data by Western blotting. So the two technologies were used in our articles when we submitted articles.

I guess the question becomes whether or not we should be validating our results by orthogonal, complimentary techniques. Then it depends on whether a suitable validated antibody is available or whether you have access to an instrument capable of detecting the protein or PTM that you are interested in.