Interesting question.

What we usually do is first to determine that there is only one protein present and not a homogeneous mixture of proteins (e.g. when many cysteines and methionines are present, sometimes proteins are partially oxidized, some proteins are heavily modified with other PTMs, and sometimes we observe salt adducts).

If these potential problems are all eliminated, a 58 kDa protein may still pose a problem for FULL top-down sequence coverage. We often experiment with various fragmentation techniques and vary the parameters as well. If one is lucky, one can do this all with a single infusion, maybe half an hour acquisition. If one is unlucky, the protein may already precipitate before one even gets to the acquisition. Or the protein may just not fragment under the chosen conditions.

Our experience is that it is seldom as straightforward as it would be when one takes some apomyoglobin from Sigma and analyzes it in top-down mode. For that reason, we treat top-down always as a research project.

Is the goal to obtain the protein sequence, or to prove top down works (never easy). If you have pure protein, I might be more inclined to suggest that you perform bottom up sequencing, using proteases of different specificity, having resolved the peptides, run them in a high mass accuracy and resolution instrument, then assembling the sequence by overlaps. It is certainly non-trivial. 

If you think the protein has homologs of known sequence from other species, then it can be easier. Does, for example, a PMF search, or an MS/MS search, reveal related proteins. 

Can you illuminate your problem a little more?

While agreeing with John, I think we can give at least a range of protein concentrations.

I will write mine for Bottom up sequencing with a MALDI TofTof or a high res LC-MS/Ms instrument: an amount of 500 ug (8-9 nmol of your protein) would be enough for an experienced user to determine the sequence with a DB search and multiple analysis (tech replicates, different enzymes etc).

If you want a complete de novo sequencing experiment, I assume (I never did it on my own) that you might require mg ( high nmol to umol in your case) of purified protein.

There are two question that John asked, which are VERY important:

- is the team experienced with proteomic sample prep and proteomic mass spec analysis?

- can you confidently assess the purity of your protein? ( nr: purity assessed via SDS page cannot be trusted for mass spectrometry. Even HPLC purity assessment is not 100% reliable)

Hi!!! Thank you for all the answers and I'm sorry about the late. 

I work with an L-amino acid oxidase isolated through fast liquid protein chromatography from a snake venom. I have this doubts because I'm planing doing and internship abroad next semester in a lab who works with sequencing of proteins by MS. 

This enzyme is around 58 kDa and it has high homology with other L-amino acid oxidases isolated from snake venoms.

Gisele,

a quick database search shows that the L-amino acid oxidases from snake venoms can carry glycosylation sites, e.g.; http://www.ncbi.nlm.nih.gov/pubmed/11453999.

Trying to analyze such a protein in top-down mode is probably not your best choice. As I said before, beware of multiple PTMs! As stated here in this discussion, consider analyzing it in bottom-up mode using multiple enzymes and putting the pieces of the puzzle together afterwards. Bottom-up is considerably more sensitive than top-down analysis, and thus probably your best choice, particularly when you think you are sample limited, although you can probably collect more venom and isolate the oxidase if necessary.

I agree with Sonja.

Bottom up will allow you to repeat experiments.

Since you are going to perform the proteomic experiments elsewhere, prepare amounts up to few mg if you can. Most likely you will not use that much, but you never know. 

In our lab we have already around 70% of its sequence obtained from bottom up (another student did this). I was thinking about complement this sequence analysing LAAO through another method (top down) to complement this sequence. 

Sonja, Why did you mention the glycosilation sites? Does it disrupt analysis through top down or another reason?

What kind of structrual characterization of LAAO do you think I can do using MS?

Glycosylation sites (and other PTMs) add (micro)hetereogeneity to your protein, and make it often impossible to resolve the mass of the protein. So, while your chromatogram may show one peak, you have 100 versions of it that are not separable by conventional chromatographic methods. You think you have one component but you do not. In fact, peaks are often so hetereogeneous that the mass can not be calculated.

In addition, glycosylated peptides do not necessarily fragment very well, so sequence information is often lost. Which may be the case for the peptides that you are missing in the data that are already available to you.

There are lots of bottom-up sequencing strategies, using trypsin, ArgC, LysC, AspN, GluC, chymotrypsin, pepsin, in addition to limited digestions (shorter times than recommended resulting in somewhat larger peptides, aka middle-down approaches). All of these should help you to achieve your goal, complementing the information that you already have.

This is not a trivial task.