Mass Spectrometry and Proteomics
Which type of MS should I use?
Ion trap – We use this instrument almost exclusively for the identification of proteins from enzymatic digests at high sensitivity from e.g. 1D and 2D gel spots or unfractionated IP pull downs. It can also be used to analyze small molecules, protein up to approx. 35 kDa. and to fragment (MSMS) molecules below approx. 3000 Da. Samples are dissolved and infused via an HPLC system or sprayed in via nanospray needle or syringe.
MALDI-TOF - MALDI TOF is good for mass measurement of proteins up to 500 kDa., peptides, conjugates, nucleic acids up to 30 kDa, polymers and even small molecules. MALDI Mass Mapping, the enzymatic digestion of of two samples followed by differential analysis of their mass spec profiles, can be very informative. Accuracy is between 0.1 and 0.01%. This mass spectrometry depends on the creation of good crystals, and the data are proportional to the size and quality of the particular crystal hit with a laser. It is therefore not quantitative. The procedure is fast and can have accuracy of 5 ppm for peptides under 2500 daltons when using internal calibration. Larger molecules show less accuracy.
Q-TOF - Sample prep is similar to the Ion-trap. PPM accuracy using a model QSTAR instrument. Used primarily to analyze small molecules, protein up to approx. 80 kDa. and peptides. Mass spec sequencing/protein identification, post translational modification analysis and quantitative mass spec are typical uses. Please contact:
KI Proteomics Lab
Ioannis Papayannopoulos (Facility Director)
Office E17-354 617 324-2609
Lab E18-313 617 324-2246 lab
Tech Asst: Rick Schiavoni email@example.com
ESI Triplequadrupole - (This instrument has been placed in storage due to mechanical difficulties, however, almost any technique that could be accomplished on the triplequad can be accomplished at higher sensitivity and with better accuracy on our model QSTAR Q-TOF instrument. See above. )
Samples are dissolved and infused via a syringe pump or an HPLC system. Our triplequad has 0.01% accuracy on proteins greater than 5 kDa, as opposed to about 0.1% seen with MALDI, however, 30X more sample is required. DNA analysis is about 30X more accurate but requires about 30X more sample.
(back to top)
Ion Trap LCMS Protein Identification
Nature Reviews-Molecular Cell Biology, “The ABC’s (and XYZ’s) of Peptide Sequencing); Steen and Mann, Vol 5, Sept 2004, p699]
We use a Thermo Electron Model LTQ Ion Trap mass spectrometer connected to an Agilent Model 1100 Nanoflow HPLC system. This instrument is excellent for mass spectrometry sequencing (ms/ms) of peptides to identify proteins and has a sensitivity level in the attomole range. Basically, the protein or collection of proteins as in the case of unfractionated pull downs from an immunoprecipitation, are digested with an enzyme, usually trypsin. Different enzymes can be used to increase coverage of the protein. The peptides from the digest are loaded onto a reverse phase capillary separation column and run into the mass spectrometer where they are ionized and their molecular weight determined. The ions are then fragmented in the trap producing mostly type Y and type B ions. We use Sequest software (Anal Chem. 2002 Nov 1;74(21):5593-9. Probability-based validation of protein identifications using a modified SEQUEST algorithm. MacCoss MJ, Wu CC, Yates JR 3rd.) to match the fragmentation patterns with the theoretical fragment patterns of tryptic peptides in the non-redundant protein database from NCBI. Statistical significance of a match is high if one knows the parent molecular weight of a peptide and several contiguous amino acid residues.
We offer three types of services:
- Routine autosampler analysis using capillary column chromatography in which sample is loaded first onto an enrichment trap and then transferred to an analytical C18 column. This is adequate for CBB stained bands.
- Manual loading. This technique has a sensitivity of about 50 femtomoles sample loaded on column.
- Manual loading of individual samples directly onto capillary analytical columns. This technique has a sensitivity of about 5 femtomoles sample loaded on column.
(back to top)
Please call the lab or office to discuss your samples if you have not used this service before.
Consider newer fluorescent stains such as Sypro Ruby from Molecular Probes in lieu of silver stain. Sypro ruby appears to give better mass spectrometry results and is easier to use.
We prefer to do the enzymatic digestion ourselves to improve the chance of success. If you perform the digestion, we will insist on seeing your sample prep protocol to avoid damage to our mass spectrometer. No detergents are permissible. Colloidal CBB is very sensitive and works well. If using silver stain, please ask for our hand-out notes regarding keratin contamination elimination and reagent changes to make the stained band more compatible with trypsin digestion. Also see: Gharahdaghi, F., Weinberg, C., Meagher, D., Imai, B., and Mische, S. Mass spectrometric identification of proteins from silver-stained polyacrylamide gel: A method for the removal of silver ions to enhance sensitivity. Electrophoresis 20, 601-605 (1999).
(back to top)
See Pricing Chart.
(back to top)
We email to you the Sequest report in MS Word format. There are many ways to look at the data from a confidence point of view, and you are encouraged to contact the lab to discuss your results.
The following are suggested minimum criteria for data to be considered valid. Each matching peptide fragmentation pattern should be inspected for accuracy. Two or more different peptide matches inspire confidence, although it is possible to identify a protein from a single match (a.k.a. one-hit-wonders) after careful inspection of the fragment pattern match from a fairly long peptide and after making sure that pI, MW, species, function, etc. all make sense.
||>1.5 acceptable, >2.0 good, >1.9 rigorous
||2.0 acceptable, >2.2 good, >2.7 rigorous
||>2.5 acceptable, >2.9 good, >3.5 rigorous
(back to top)
How to Order
See Ordering Information.
(back to top)
Overview of the Technology
We are using the Applied Biosystems Model Voyager DE-STR instrument for Matrix Assisted Laser Desorption Instrument Time of Flight (MALDI-TOF) mass spectrometry. MALDI-TOF mass spectrometry works on the principle that ions of different molecular weights travel at different speeds, providing they are accelerated with the same potential and carry the same charge. Fortunately, most MALDI ions are singularly charged (MH+) i.e. carry one extra proton (+H) or have been singularly deprotonated. With larger molecules, you may see a small amount of the MH++ and perhaps MH+++ ions. The time of flight is proportional to mass--heavier ions take longer to reach the detector. For a more thorough explanation, please see chapter 1.3 of the manufacturer's operators manual entitled "Voyager Biospectrometry Workstation User's Guide" which can be downloaded from the Applied Biosystems web page or from our Biopolymers FTP server.
Basically, the sample is mixed with matrix solution to assist ionization and then dried on a sample plate. A laser beam is fired at the sample crystals simultaneously with the application of a high voltage pulse. The time it takes for sample ions to drift through the flight tube to the detector is proportional to their molecular weight.
The fact that the sample is initially in crystal form reduces the ability to get information on the quantity of sample loaded. The analysis of one well-formed crystal in a very dilute sample can give a greater signal than the data from the analysis of multiple, poor-quality crystals in a far more concentrated sample. If you need quantitative data, you are better off using an ESI mass spectrometer, where sample is dissolved and sprayed into the MS. The amount of sample is directly proportional to the signal for that sample.
The MALDI mass detection range for peptides/proteins is between 500 and 350,000 Da. Ions below 500 Da can be measured if the amount of sample (usually ca. 10+ pmole) provides a signal that sufficiently exceeds those of the numerous matrix artifact ions. In general, we prefer to analyze small molecules on our ESI instrument, which provides valid data in the 30 to 80,000 Da range.
(back to top)
General Protein and DNA Samples
We would ideally like to load 1-100 pmol of peptide, 10-100 pmol of protein, and 10-100 pmol of DNA. As for volume, we generally mix 1 ul of your sample with 1 ul of matrix. We will concentrate the sample if necessary.
The sample should be free of salts and detergents because these may inhibit ionization. We can desalt your sample using reverse phase or ion exchange resin. Removal of detergent is often sample-specific, so we prefer that you do this yourself. You should precipitate or dialyze your sample away from detergent or contact us about other alternatives (also see ESI Sample Requirements). Since MALDI analysis is inexpensive, you may want to rely on the dilution of your sample to negate the effects of detergent and salt, and if that fails, proceed with sample clean-up. Remember, 1 ul of a 1 pmol/ul solution is more than sufficient in most cases.
Tell us the MW of each portion, as this affects how we go about the analysis.
Mass Mapping for Protein Identification Samples
We would like 1 pmol of in-gel protein, but as little as 125 fmol is potentially feasible from CBB (e.g. G250) or silver stained gels (a special silver stain protocol is required--see below). Consider newer fluorescent stains such as Sypro Ruby from Molecular Probes in lieu of silver stain. Sypro ruby appears to be give better mass spectrometry results and is easier to use. We appreciate a control gel slice from a blank area of the gel of approximately equal size to the sample to enable subtraction of gel artifact ions prior to database searching.
Do the required de-stain procedure, excise the band of interest, and place in an eppendorf tube. Keep the protein/gel ratio as high as possible, i.e. trim off any non-protein containing gel. Keep frozen, if possible.
(back to top)
MALDI Mass Fingerprint Mapping for Protein Identification
Most of the samples we process for MALDI mass fingerprint mapping for protein identification come from SDS-PAGE bands that have been stained with either CBB or silver. We can also assist in HPLC of protein intended for digestion and mapping. Solution phase digestion provides better coverage. We perform a tryptic digest on the gel band or sample, isolate the resulting peptides, and measure their ion molecular weights at high accuracy. We generate a list of ions that are present in the sample but absent from the control band and do a database search to determine the identity of the sample. It is most helpful if the number of proteins in the band are minimized. When more than two or three proteins are present the search becomes difficult. Use a 2D gel if necessary to reduce the sample complexity. We would much rather have 200 fmol of a single protein than 1 pmol of a mixture. Use our Ion Trap protein identification service for identifying gel bands that are likely to contain more than one protein or to identify the proteins in an immunoprecipitation prior to SDS-PAGE separation.
We have found that CBB stained samples of 250 fmol (we are talking about sample in the gel band, not sample loaded onto the gel) or greater work virtually 100% of the time, while CBB samples of 125 fmol work about 50% of the time. One must be especially careful with silver stained samples to avoid keratin contamination, which can dwarf the signal from the actual sample.
If using silver stain, please ask for our hand-out notes regarding keratin contamination elimination and reagent changes to make the stained band more compatible with trypsin digestion. Also see: Gharahdaghi, F., Weinberg, C., Meagher, D., Imai, B., and Mische, S. Mass spectrometric identification of proteins from silver-stained polyacrylamide gel: A method for the removal of silver ions to enhance sensitivity. Electrophoresis 20, 601-605 (1999).
(back to top)
See Pricing Chart.
(back to top)
Data will be emailed in a Word document. The data file can be saved as a text file which can be opened in Excel and plotted in histogram format to give a reasonable graphic. Unfortunately, our instrument spectra files can only be opened using the manufacturer's software. You are welcome to use this software on our instrument computer to expand small segments of the spectra and examine them in detail.
General Protein or DNA Output
We usually provide at least two printouts of the spectra taken. One will show the entire range of the spectrum measured and some will detail the masses of interest. We often provide you with a spreadsheet file that details the intensity (peak height) and peak area for each mass measured.
Mass Mapping for Protein Identification Output
We provide you with enough of the output from a database search program, usually Protein Prospector from UCSF (MS-FIT), that we think is necessary to identify the protein(s) present in the sample. To see if more than one protein is present, we subtract the ions used to identify the first protein and do a second search. This is done repeatedly until no further matches are discerned.
The MOWSE score for the actual protein present is usually several orders of magnitude higher than the score of other proteins listed in the output report. The grid display is very useful in visualizing which protein identifications were returned from which ions.
Most importantly, we provide a manually generated peak list for all ions found that were not present in the control gel band. You can cut and paste this list into Protein Prospector or any other search program and vary the parameters to see if different results are obtained. You can use the peak list to re-search at a later date when the genome of interest is more complete. We "de-isotope" the peak list for you, i.e. we report only the C12 isotope for each ion.
We use internal calibration, unless otherwise noted, and this results in less than 10 ppm error, with the majority of ions having less than 5 ppm error. Good precision (a consistent error bias) is indicative of a high confidence match, e.g. most of the matching ions are all of negative 3 ppm error +/- 2 ppm. On the other hand, if a matching peptide had to be modified by two oxidized methionines, three phosphates, and a pyroglutamic acid conversion in order to align with an ion from the peak list, and the MW error is positive 25 ppm, while all the other ion errors had an error of negative 3 ppm +/- 2 ppm...it is unlikely that that particular peptide match is valid.
If you requested to do the database search portion of the process yourself, we provide you with a spreadsheet file that details the intensity (peak height) and peak area for each mass measured. The peak list and de-isotoping are generated automatically by the instrument software and are not verified. Any trypsin and gel blank artifact peaks are not subtracted. Below is part of the "Information with your MALDI data" form that we include with your data:
||Internal calibration used. Typical accuracy is +/- 5 ppm.
||External calibration used. Typical accuracy is +/- 25 ppm.
||High confidence match found.
||More than one protein appears to be present.
||Manual peak list included with report (each isotope cluster verified by eye)
||Computer generated peak list with intensities included. If you are doing the database search yourself, we have attached a de-isotoped and non-de-isotoped list.
||A gel blank was analyzed and the ions in the peak list are not present in the blank
||The peak list was rather short for a protein of this size, possible cause-too little sample.
||An accurate peak list was generated but no outstanding matches found. You should look at the matches to see if they make sense to you. You should search again later when the genome databases are more complete.
||Check to see if we searched with the correct Cys modification e.g. IAM, IAA, acrylic acid, etc.
||The match required many post translational modifications and may not be valid.
||We think the PPM error range is somewhat wide. You should look at the matches to see if they make sense to you.
||No reliable ions obtained. Possible cause-too little sample.
||We used the Protein Prospector MS-FIT database search program and pasted the results into your report. A MOWSE score (look under the MS-Digest column) one or two orders of magnitude higher than the subsequent protein's score is usually significant. Please call the lab if you have any questions whatsoever.
(back to top)
How to Order
See Ordering Information.
(back to top)
The Koch Institute Proteomics Core Facility provides mass spectrometry-based analyses of protein samples for which one or more of the following is desired: Protein identification, protein characterization (including modifications), intact protein MW determination, protein quantitation. Samples which are not proteins or peptides, for which structural or MW information is needed, may also be submitted, as long as they are amenable to analysis with the instrumentation and methods currently available.
- PROTEIN IDENTIFICATION
Proteins separated by gel electrophoresis
Protein identification from gel bands is generally routine and as long as there’s enough protein and the keratin contamination is kept under control it yields results. In general Coomassie-stained bands contain enough protein to identify it, unless the keratin contamination is very high. A weakly stained Coomassie band may contain a 200-300 fmol protein (10s of ng) whereas a strongly-stained Coomassie band may contain pmol amounts of protein (100s of ng). Although the more protein there is the more confident the protein identifications will be, consider that by overloading a gel you may end up obscuring some of the weaker bands, especially if these are close to the bands of the major proteins.
Silver-stained gel bands contain less protein, typically 100 fmol or less (low ng range) and the sensitivity of fluorescent stains is in-between that of Coomassie and silver.
There are several staining kits available commercially from the major suppliers of biochemicals, which are compatible with mass spectrometric analysis (not all gel staining methods are). If you (or others in your lab) have submitted samples before and obtained useful results then obviously continue doing what you (or your colleagues) did before. Otherwise please talk to us before submitting samples so that we can help you increase the chances of success and avoid wasting time, effort and money.
Issues to consider are keratin contamination and how to minimize it and the use of staining protocols that are compatible with mass spectrometric analysis. Keratin comes primarily from skin, wool clothing and dust. While handling samples and gels wear a lab coat and gloves (nitrile is said to be better than latex) and avoid touching samples, gels, etc. with bare hands. Rinse all containers, tubes, with HPLC-grade methanol and HPLC-grade water. If available, run the gel in a laminar flow hood, such as a tissue culture hood, if not available use a clean surface in an dust-free area (e.g. away from an air vent). After running and staining the gel place it in a rinsed Petri dish (avoid Ziplock bags) in 1-2% acetic acid, cover it and bring it over to our lab so we can cut the gel band(s) of interest in our laminar flow hood (if the timing doesn't work the gel can be stored overnight in the refrigerator in the covered Petri and this should not cause problems to its subsequent analysis). If you need to scan the gel ensure that the scanner surfaces have been rinsed with clean methanol and water and wiped clean. By the way, “old style” gel fixing (e.g. using glutaraldehyde or formaldehyde or similar protein cross-linking rteagents) must be avoided as it makes it impossible to digest proteins in gels and recover proteolytic peptides.
For staining you can use any of the "mass spectrometry compatible" stains sold by several vendors. For Commassie, Gel Code Blue (Pierce/ThermoFisher) or Commassie Brilliant Blue R-250 and G-250 (Pierce or Bio-Rad) are all fine, if you follow the instructions that come with the staining kits. For silver, SilverQuest from Invitrogen or Silver Snap from Pierce are fine. Fluorescent gel stains (e.g. SYPRO Ruby, SYPRO Tangerine) are also mass spectrometry compatible. If you plan to use a stain and you cannot determine (e.g. from the product description) whether it is compatible with mass spectrometry please let us know and we can make inquiries.
Although cutting gel bands is straight forward, it is important to do it in a clean area such as a laminar flow hood. If you have not done this before please talk to us before excising gel bands for analysis. We can arrange for you to bring your gels to our lab and we can cut gel bands you wish to identify.
We prefer to carry out the in-gel digestions, under conditions that reduce sample contamination and optimize digestion and recovery of peptides, using protocols that have been applied successfully to thousands of samples. However we will accept already-digested samples, but if you have not submitted digested sample before please talk to us before doing this for the first time as we may be able to offer some helpful suggestions. We use trypsin for digestion; if you need your protein digested with another enzyme please discuss this with us.
Proteins in solution
We routinely identify proteins in solutions, from a few to several hundreds. The same caveats apply as with in-gel digestion with regard to contamination (keratin is also a problem with solution samples, as proteolytic peptides from keratin often obscure or suppress signals from peptides from the less abundant proteins). Excessive amounts of salts can also be problematic, so using a method such as acetone precipitation to prepare protein samples may be necessary; please talk to us about this if you have not done this sort of thing before. In addition there are certain other materials and chemicals, such as detergents and polymers, which are much more problematic with solution samples compared to gel bands,. This is because excised gel bands are washed extensively prior to digestion so that many buffers and detergents can be removed without protein loss, but solution samples are not readily amenable to washing. Although methods such as MW filtration or dialysis can be used, they result in some protein loss and also some detergents and polymers stay with the protein anyway. Please talk to us regarding the use of detergents as we may be able to recommend alternative ways of preparing samples as well as some “mass spectrometry friendly” detergents.
For on-gel protein sample it would be helpful to run a silver stained gel using ~10-15% of the sample; if you see bands, even weak ones, this means that there’s probably enough material in the sample remaining to identify some proteins.
Identifying proteins in immunoprecipitation experiments can be done in two ways: Either by running the immunoprecipitated proteins on a gel and submitting one or more gel bands or by submitting the mixture of immunoprecipitated proteins without separation. The same issues outlined above apply. Please note that if you run a silver stained gel using 10-15% of the immunoprecipitated sample and the only bands you see are related to the antibody used in your experiment it may not be possible to identify any of the other (and more relevant) proteins in your sample.
- PROTEIN MOLECULAR WEIGHT MESUREMENT
We can measure protein molecular weights accurately, with errors in the range of 25-50 ppm (this is less than 3 Da for a 60 kDa protein) for single proteins. Protein mixtures also work but the data analysis (and the results generated) may get complicated because we do this by electrospray ionization, which generates many multiply charged ions for each protein and if there are several proteins in the sample the overlapping multiply charged ion envelopes may overwhelm the deconvoluation software. If you do not need the higher mass accuracy you may consider submitting your protein sample to the Biopolymers Core Facility for MALDI-TOF MS analysis, which would give you a MW with a mass error of 0.05-0.1% (30-60 Da for a 60 kDa protein), which is still much better than what you can estimate from SDS-PAGE.
Detergents should be avoided, as they suppress the ionization of protein samples and other contaminants such as polymers (PEG or PPG) have similar deleterious effects. However low-to-moderate concentration salt buffers (e.g. PBS, Tris) are fine, as protein samples are desalted on-line prior to MS analysis.
As with protein samples that are submitted for protein identification, please talk to us if what you need is something less straight-forward than simple MW determination, like the assessment of modifications such as phosphorylation or glyosylation or N- or C-terminus truncation. Depending on the heterogeneity of the sample and the MW the aforementioned modifications may or may not be discernible from the MW measurement of the intact protein and may require additional analysis of the proteolytic peptides obtained from digestion of the protein.
- PROTEIN CHARACTERIZATION
We can obtain detailed protein sequence information of relatively pure protein samples by digesting the protein and analyzing the resulting proteolytic peptides by LC-MS. The difference between protein characterization and protein identification is that in the former the identity of the protein is known and so is the amino acid sequence and the purpose of this analysis is to confirm the correctness of the sequence (e.g. of a point mutation, a splice variant or the presence of a tag) or identify specific known or suspected modifications to the amino acid sequence. For this it is important that as much of the protein sequence (e.g. 80% or more) is covered by identified proteolytic peptides, whereas a protein can be unambiguously identified from one or a few peptides that cover a mere 10-20% of the amino acid sequence. In general, for this type of work it is preferable that the protein sample be submitted in solution because in-gel digestion of proteins usually results in less than 50% sequence coverage (some parts of the protein embedded in the gel are not accessible to the enzyme and some of the peptides thus generated cannot be extracted from the gel – these problems generally do not arise when the protein is digested in solution). Typically larger amounts of protein (several pmol or 100s of ng) are needed for characterization (vs.10-50 times less for identification).
The aforementioned constraints notwithstanding, we will do our best to identify modifications in protein samples submitted for identification, whether in solution or as gel bands. We ask you to indicate on the sample submission form what type of modifications you are interested in and we will set up the database search software to look for the modifications indicated.
- PROTEIN QUANTITATION
We can provide estimates of protein amounts using label-free approaches as well as labeling methods such as SILAC and iTRAQ. For label-free quantitation the appropriate controls need to be run, for SILAC quantitation only certain isotopically labeled amino acid combinations are supported by the software we have available (you do the cell culture) and the iTRAQ chemistry, although relatively straight forward, requires following a specific protocol in detail. If you are interested in any of these methods please talk to us before generating samples for analysis in order to ensure that we are equipped to handle your request.
- ANALYSIS OF MOLECULES OTHER THAN PEPTIDES AND PROTEINS
We will analyze non-peptide samples as long as they behave like peptides. That is, they ionize in positive ion mass spectrometry, are water soluble and amenable to reversed phase HPLC analysis. Please note that the mass accuracy we can obtain for small molecule MW measurements (10-25 ppm) and the mass resolution (10,000-12,000) of our instrument may fall short of journal requirements for confirming the structure of a newly synthesized molecule. Also, since such samples are not regularly submitted for analysis, we may wait until enough small molecule samples have been submitted before setting up the instrument to analyze them. Therefore, wait times may be longer than usual for such analyses.
- INSTRUMENTS AND SOFTWARE
We have two mass spectrometers, an LTQ ion trap from Thermo and a QSTAR quadrupole-time-of-flight instrument from Applied Biosystems. Both are electrospray instruments and each is connected to a nanoflow HPLC. For the analysis of proteolytic peptides we use very small HPLC columns (50 or 75 mm internal diameter) at low flow rates (~300 nL/min), which increases sensitivity vis-à-vis analyses at higher flow rates. For identification of proteins we use either the SEQUEST database search software for the data generated with the LTQ instrument or the MASCOT database search software for the data generated with the QSTAR instrument (and sometimes with the LTQ data as well). We keep current with the various protein databases, with the help of the Bioinformatics Core Facility staff. For protein characterization and quantitation and MW determination we use software supplied with the instruments as well as third-party software. Please talk to us if you would like more information on our instrument and their capabilities of on the software we have available.
- CONTACT INFORMATION
The Koch Institute Proteomics Core Facility lab is in E18-313 (phone: 617-324-2246). The facility personnel are:
Ioannis A. Papayannopoulos, Ph.D. Richard P. Schiavoni
E17-354 (right across from E18-313) E17-415
Please feel free to contact us if you have questions or to discuss specific items related to your projects
How to Order
See Ordering Information.
(back to top)