MALDI-TOF Mass Spec

Table of Contents

Letter from the Editor
Index of Experts
Mass Spec Experts:
Catherine Fenselau
Sarah Hart
David Lubman
Peter O'Connor
Andy Pitt
Chris Sutton
Jonathan Sweedler
Tim Veenstra

Download the PDF version here

Letter from the Editor

Here at GT, we're bringing you something new this month: an instrument focus guide with recommendations for choosing a mass spec. We've collected an esteemed roster of proteomics experts who offer advice on what factors to consider when planning your next MALDI-TOF purchase.

Due to its speed, sensitivity, and high resolution, MALDI-TOF is one of the most robust — and highthroughput — methods of determining molecular weight. Everything from proteins, peptides, and oligonucleotides can be ferreted out of mixtures, and MALDI-TOF is being utilized in a wide array of 'omics fields, including metabolomics and glyco-proteomics.

Whether you're a veteran at mass spec or new to the technology, the number of flavors of instrument on the market can be daunting. That's why we've included a range of questions to keep in mind when figuring out what you need in a MALDI-TOF. What will you primarily be using it for? How does your existing setup come into play? How can you make sure your new machine performs optimally on your unique samples? Do you need a service contract? Hopefully, this guide will arm you with practical buying advice on your quest for that just-right MALDI-TOF.

— Jeanene Swanson

Index of Experts

Genome Technology would like to thank the following contributors for taking the time to respond to the questions in this guide.

Catherine Fenselau
University of Maryland

Sarah Hart
University of Manchester

David Lubman
University of Michigan

Andy Pitt
University of Glasgow

Chris Sutton
University of Bradford

Jonathan Sweedler
University of Illinois at Urbana-Champaign

Peter O'Connor
Boston University
School of Medicine

Tim Veenstra
National Cancer Institute-Frederick

Catherine Fenselau
University of Maryland


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

CF: We use our MALDI-TOF to analyze proteins and peptides — for example, to survey proteins in an HPLC fraction and to survey peptides produced in a proteolysis reaction. When we develop optimal conditions for our new chemical cleavage reaction we monitor the progress of the microwave-supported acid cleavage with rapid MALDITOF analysis. We take advantage of the speed, sensitivity, mass range, tolerance for salts, and reliable production of singly charged ions.

GT: What factors were most important to you in choosing a mass spec?

CF: Reliability. Engineering. Service resolution. Computer control.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

CF: Not at all. We have altered some of our workflows to take full advantage of the MALDI-TOF.

GT: What steps would you recommend to people looking to buy a MALDI-TOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

CF: Get a list of users from the manufacturer and talk to the users about performance, reliability, and service.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

CF: We accepted the manufacturer's tests done in our lab. We established reference spectra for sensitivity and calibration.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

CF: Among my highest priorities

Sarah Hart
University of Manchester


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

SH: The primary applications within the Michael Barber Centre for Mass Spectrometry are protein and peptide analyses, most of which is carried out within the general context of proteomics. We currently have two MALDITOF instruments within the lab, our workhorse AB Voyager-DE STR, which was installed around eight or nine years ago, and a more recent acquisition, a Bruker Ultraflex II TOF/TOF. The Voyager gets used quite a bit by outside labs for simple PMF and sample checking, analysis of synthetic peptides, etc., since it is such a simple platform to use. I shall primarily discuss the Ultraflex here since most labs performing MALDI-TOF would not consider buying an older type of instrument.

GT: What factors were most important to you when you were choosing which mass spec to purchase?

SH: The most obvious criteria for most people are sensitivity, flexibility, and robustness. The Ultraflex sensitivity is pretty good, with routine observation of femtomole levels of sample as applied to the target. But the real advantage is the sheer number of applications which are ready to run; at the moment we have standard MS/MS, LCMALDI and in-source decay set up, iTRAQ workflows for quantitative analysis are underway, and we are preparing to set up MALDI imaging. The ability to tune the laser beam diameter to the particular application of choice is highly appealing. Another aspect is ease of use. As a research lab we all need to be able to use all of the equipment within the Centre; this instrument is very easy to learn and run. Finally, it also helped that we got an attractive discount when we bought an ion trap concurrently.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

SH: Since MALDI-TOF per se is a relatively routine method, our existing workflows have been readily accommodated. On the whole, rather than altering our workflows, we have found that the new instrument has enabled us to do a number of experiments which we would otherwise have been unable to do using our Voyager, including in-source decay. Being able to rapidly generate product ion spectra from detected precursors also means that ready identification of unknowns has been speeded up as well.

GT: What steps would you recommend to people looking to buy a MALDI-TOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

SH: First off, before you even apply for money, speak with your local colleagues in surrounding labs about their existing instrumentation and things they wish they could do better.

Once funding is in place, get in touch with sales reps for the companies whose instruments you found of interest during your preliminary enquiries to organize demonstrations using real samples which you are interested in (any instrument will be able to identify a tryptic digest of BSA, but you need to test that it will help move your research ahead). Again, speak with colleagues to find out what they have done in demos and what they would recommend/would never do again.Apps chemists with the instrument vendors are often a useful source of information about what you can expect to achieve within a typical demonstration. I also find that the ABRF listings (www.abrf.org) are an invaluable resource for finding out what other labs worldwide think of existing instrumentation. Consider your service and support needs at this stage, too; some labs will be happy dismantling and cleaning instruments themselves, whilst many will want to consider a comprehensive service contract.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

SH: Initial quality assessments were performed with the manufacturer's test set of samples to ensure that the instrument exceeded the performance specifications (these are often deliberately modest; examine them in detail if you can). Since the installation of this instrument, we have developed an in-house standard for calibration and sensitivity checks of mass spectrometry equipment, in terms of mass resolution, sensitivity, and dynamic range, based upon the QconCAT technology developed by Rob Beynon and Simon Gaskell. We use this regularly to monitor instrument performance and perform external calibration.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

SH: Facile cleaning of the source region when performing large numbers of MS/MS experiments on a MALDI-TOF is an important factor, since the source can get dirty very rapidly when the laser is running all the time. Laser lifetime for cartridge-based lasers is an important factor in keeping an instrument running. I have been particularly impressed by the ability to conduct online tech support sessions with our new instrument; often this can get us straight back on our feet, and saves the time spent waiting for an engineer visit.

David Lubman
University of Michigan


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

DL: We use our MALDI-TOF for basic studies on proteomics and glycoproteomics. We have the Shimadzu QIT-TOF MS.

GT: What factors were most important to you when you were choosing which mass spec to purchase?

DL:We needed a MALDI mass spec which could perform multiple stages of tandem mass spectrometry (MSn) and do so with high mass accuracy.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

DL: No. We were using a Micromass TofSpec for proteomic studies. The use of the new instrument allowed MSn studies. Initially we could only use DHB matrix with this instrument, but we now can use CHCA also.

GT: What steps would you recommend to people looking to buy a MALDITOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

DL: There are several companies selling MALDI-TOF instruments. I would decide whether you need a MALDITOF or if you need MS/MS capabilities. Some of the instruments are designed for high throughput; others for high- or low-energy CID; some have better mass accuracy; others have MSn capability. I would look at your needs versus the specs of the instruments. Each instrument has its own advantages and disadvantages.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

DL: We are not a service lab, but we tested a number of standards to evaluate performance. Ultimately we needed the sensitivity to study glycopeptides and glycans from serum. The MSn capabilities were essential.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

DL: Shimadzu has provided excellent help in servicing and updating the instrument in terms of software development. This is essential.

Peter O'Connor
Boston University School of Medicine


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

PO: I do proteomics and glycomics. While our group does have a Bruker Reflex IV MALDI-TOF, we prefer our home-built MALDI-FTMS instrument. The MALDI-TOF is, however, useful for checking the quality of collaborators' samples and for simple in-gel digest protein ID experiments.

GT: What factors were most important to you when you were choosing which mass spec to purchase?

PO: Performance is critical in terms of resolution, mass accuracy, and sensitivity.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

PO: We don't run so many samples that the workflow changed much. For us, every sample is unique and requires application of a wide variety of methods to generate useful results. The MALDI-TOF was another tool in the toolbox.

GT: What steps would you recommend to people looking to buy a MALDI-TOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

PO: First, evaluate your needs. Do you need 10k resolving power and 10 ppm mass accuracy routinely? If so, add it to your specifications. Do you need MS/MS capabilities? Consider a TOF/TOF, MALDI-qTOF, or trap-TOF systems. If you need 50k RP and 1 ppm mass accuracy, consider the MALDI-FTMS. If your goal is routine in-gel digest experiments or the equivalent, a MALDI-TOF is what you need.

Second, get a set of standard samples from the pool of projects that you work on and visit all the manufacturers to try them out. Don't take the hardest samples in your projects; if they don't already work for you, you can't know what to expect from a new instrument. If you get no data, you don't know if the problem is with the sample or the instrument.

Third, evaluate the company in terms of their reputation with respect to quality and service. Call a number of people that you know who are knowledgeable about instruments and get their opinions. Again, you need to know your needs in this respect. Some instruments are a terrible choice for fiddly academics, but a great choice for routine MS facilities, and vice versa. Some places need reliable service contracts; others don't.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

PO: A series of standard protein digest experiments were done. The instrument was calibrated, tuned, and tested to meet not only the manufacturer's specs (which are always very "safe" targets for them), but also to meet the expected performance that one can determine by calling knowledgeable colleagues and by reading the literature with a skeptical eye. Particularly, off-the-record conversations at conferences or by telephone are the best way to get the opinions you need. Also, you need a lot of these opinions, as most of the members of the field have known biases.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

PO: We don't use service contracts, so I prefer instruments which are easy to maintain. This means simple, open, flexible design. It's critical to be able to take the instrument apart, clean it, reassemble it, get it working again, and calibrate it without needing to call anyone. Clearly, this is different than operation in a commercial MS facility or high-throughput operation.

Andy Pitt
University of Glasgow


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

AP: We currently have two MALDI-TOF mass spectrometers in general use, a 4700 Proteomic Analyzer from Applied Biosystems and a Bruker Ultraflex II (with SmartBeam), which are used for a range of activities focused mainly on proteomics in its broadest sense, as well as limited analysis of other biopolymers.

The 4700 is used mainly for routine identification of 2D-PAGE gel spots as a firstpass method of identification, especially in high-throughput applications; for quantitative analysis (iTRAQ or others) of protein samples fractionated by liquid chromatography; and for LC-MALDI for GeLC and bottom-up quantification workflows, as a complementary technology to our electrospray instruments. The 4700 has also been used for the analysis of lipids, short oligonucleotides, and synthetic polymer conjugates.

The Bruker Ultraflex II is dedicated mainly to mass spectrometric imaging of tissue samples. We do use it occasionally for standard analysis, using anchor-chips for the analysis of very small samples, such as peptides isolated from major histocompatibility complexes, and weak gel spots.

GT: What factors were most important to you when you were choosing which mass spec to purchase?

AP: Performance is always the main criterion coupled to good technical support, closely followed by flexibility of operation coupled to high-quality software. As we are an academic institution, cost is also an issue — especially the ongoing running costs. Service contracts can be painfully expensive, and some of the less helpful manufacturers do not include full software upgrade with these, which is a hidden cost that easily catches you out.

Some of the factors in good performance are sensitivity and resolution (mass accuracy). Our main criterion for sensitivity is to be able to use a low laser power and small laser sport size so that we can acquire as many spectra as possible from each individual sample spot, especially when performing MS/MS on the MALDITOF/ TOF instruments. Resolution and mass accuracy in automated mode should be better than 10 ppm to help limit the search space and improve statistics for protein identification.

High-quality service and support staff are a real bonus; it is great to be able to pick up the phone and talk to an experienced service engineer.

Software that forms part of an integrated package for protein identification is essential, and it would be a real bonus if the data was compatible with other instruments from the same manufacturer, such that there is not a significantly different user interface for each instrument and that data can be easily merged and analyzed together and coupled directly to higher-level statistical analysis. Unfortunately, this is still not the case for a few manufacturers. Ease of operation is less of an issue, but poor integration with databases, especially archived information, and unnecessarily complex file structures do not make data management in higher-throughput studies very easy.

Robustness of the hardware is another key operating criterion; for example, in the imaging studies we clean the source regularly (sometimes every few days) and this is only possible on some instrumentation.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

AP: Actually, very little. We are driven by the need to generate the best quality data possible and provide as broad-based analysis as possible, and are fortunate to have the in-house expertise. The flexibility of working in the academic environment to be able to modify our workflows as necessary to accommodate changes in instrumentation. Workflows often need to change to take best advantage of the instrumentation, although the basics remain the same.

GT: What steps would you recommend to people looking to buy a MALDI-TOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

AP: This is often a tough process, as the performance of most instruments is very similar on paper. The first step is to decide what the instrument will be mainly used for, and to be honest about this. There will always be a compromise to be made between all the things you would like an instrument to do, and what the capabilities of the different instruments are. It may seem very attractive to have an instrument able to perform some of the more esoteric tasks, but all too often this results in a compromise of the bread and- butter tasks. Also, although it is a cliché, don't believe everything that you read; check out the facts from another source. Contact some users of the instrumentation who are doing similar things that you want to do, and if possible don't just rely on those suggested by the vendor. Most people using the instruments will be willing to give you an honest evaluation. Whatever information you get will give you useful ammunition for further discussion with vendors.

To my mind nothing can make up for getting a chance to see a selection of your day-to-day samples run on a range of instruments by visiting the demonstration labs of the vendors. Often samples are required in advance to give an opportunity to see some data from the instrument, but I am always more pleased if I can turn up with some additional samples in my pocket and get these run to see that the systems are then robust enough to deal with my real samples. Choose as broad a range of realistic samples as is necessary to cover the work that is likely to be performed. The opportunity to talk at length with the applications or demonstration scientists during the demo gives a good indication of the likely support that you will get after the purchase. This may be time consuming and a little bit costly, but is likely to pay off in the future.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

AP: Obviously ensuring the instrument is operating to the manufacturer's specifications before signing off, and then using these tests occasionally to check performance. However, these tests are usually designed to work well, and we have found that sometimes they are not a good indicator of performance on non-optimal samples. We therefore have a test sample (a protein digest) that we run should we feel that the machine is not operating as it should, and we also test this using a recent sample for which data was obtained that was satisfactory. Also, as most of our samples for proteomics applications contain trypsin, this is always used as an internal control.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

AP: Ongoing maintenance is a key issue. These are high-tech instruments, and replacement parts are expensive, especially some items that have limited lifetimes such as high repetition rate lasers, turbomolecular pumps, and fast-switching high voltage supplies. Unfortunately, they will go wrong from time to time. To keep instruments operating at their maximum potential, servicing is critical, and when a breakdown occurs, it is much simpler to be able to switch out all the suspect components, rather than work about costs. Also, the "deep clean" that is performed at preventive maintenance visits ensures the machines keep operating at maximum potential. The frequency of these PM visits will depend on the amount of use the instrument receives. If running close to 24/7, then six-month intervals should be considered as matrix deposits and other non-volatile materials can build up relatively quickly. A sensibly priced comprehensive service contract, with good turn-around on problems, is important. Having local service engineers certainly helps with this.

Chris Sutton
University of Bradford


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

CS: I work in the Institute of Cancer Therapeutics at the University of Bradford. The institute is geared to the processes of drug synthesis all the way through to phase one clinical trials, including pre-clinical model testing and various biological systems of validating targets. I joined two and a half years ago, and at this stage there was no proteomics or MALDITOF mass spectrometers in the institute.

My background is one of having worked in the mass spec industry, and my preference coming into academia was to work with a type of mass spectrometer I was familiar with. In the end, we actually went for a Bruker Ultraflex II instrument.

GT: What factors were most important to you when you were choosing which mass spec to purchase?

CS: Again, having come from the mass spec industry, I had a very clear perspective of the commercial side of what was required in making instruments appropriate for the market. The highest priority was performance, but particularly in relation to the key application, which was separating proteins in the LC-MALDI strategy.

Additional priorities were their ability to meet standard specifications as defined by their literature and an indication that the vendor provided solid software that provided solutions to the kind of proteomics applications we were looking at. So, fully automated LC-MALDI acquisition software, plus integration with database search engines to identify proteins from complex mixtures.

The other major priority was the application support and the expertise within that company to back up their software and help a customer to be able to settle these applications in their own laboratories.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

CS: I don't have a straightforward answer to this because I didn't have an existing setup within the university. I had, if you like, a blank sheet of paper to work from at the end of the day and the responsibility to decide which was the approach that I wanted to take. It wasn't so much a matter of the existing setup as creating a setup that made the choice of instrument easier.

GT: What steps would you recommend to people looking to buy a MALDI-TOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

CS: Coming into academia, I can see these decisions are career-defining decisions, and therefore I think it's very important to make the time to find out about the equipment, to actually design and define the samples that you want to be evaluated, and to make sure you prepare them in a way that when they're provided to vendors that everyone's on a level playing field. I think it's very important to take control of the sample prep, decide what you're expecting to see from each sample. This helps you to get the results that you want, and it helps the application scientists and the vendor to make sure you get the data you want. And I think it's a good way of establishing a relationship with the vendor for the long term.

I think it's important to understand what the choice of instruments are, and to actually get opinions about those instruments independent of the salespeople who support those systems.

I think it's also important to take the time to go to the demos, not to just depend on remote sample analysis, and to try and do this for at least three vendors. Another aspect one has to take into account is what their post-sales support is like.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

CS: Some time was taken to actually carry out this process. The engineer obviously did the installation, performed the tests that they are required to do to have it signed off. I then also ran the same standards that I provided in the demo to make sure that we were seeing the same level of performance. For LC-MALDI, the very nature of the technique involving a number of pieces of hardware, a number of software systems, did require one or two months of evaluation and then further discussions with, in this case, Bruker, to optimize the performance before I considered it fully acceptable. So there was an ongoing process there for up to two months. Once everything was installed it was very much the case of me being left to my devices, having had the basic training, to really refine my understanding of how everything worked. And then for the vendor to support me when I'd had a go and then needed that boost of refined training.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

CS: An extended warranty was available — that is very important to establish in the beginning. The very nature of mass spectrometers, whether these are ESI-based or MALDI, is they are heavily used; there will be wear and tear. Parts will need changing, so regular servicing is required whether by myself or by an engineer. I think it's very important to establish that the vendor is going to have that ability to support you. And that it is covered by the service contract or the extended warranty.

Jonathan Sweedler
University of Illinois at Urbana-Champaign


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

JS: Our most recent MALDITOF we purchased is the Bruker Ultraflex II TOF/TOF, which we acquired just over a year ago. We use it for three major applications related to neuropeptide characterization, with these peptide characterization applications using a range of sample types. One important application includes peptidomics, which involves characterizing the endogenous peptide content of a tissue homogenate. For these samples, we use a variety of MS instruments, one of which uses multiple stages of capillary liquid chromatography followed by MALDI-TOF or TOF/TOF. We also perform direct peptide characterization from tissues, tissue extracts, and even samples as small as individual neurons. These manipulations benefit from the range of sample targets available from Bruker. While single-cell samples require extra care in terms of minimizing sample losses during cell isolation and matrix addition, the sample complexity is reduced for these samples compared to larger tissue homogenates. This reduction in complexity aids peptide characterization. Lastly, we perform a variety of mass spectrometric imaging experiments to determine the spatial distribution of peptides, and we also monitor peptide release from brain slices. For the imaging experiments, we apply the MALDI matrix using several approaches and find the ImagePrep an important addition that offers greater sample-to-sample repeatability than our previous airbrush- based matrix application methods. Some of our work involves technology development in terms of new protocols and approaches for sampling neuropeptide distribution and release from a range of animal models.

GT: What factors were most important to you when you were choosing which mass spec to purchase?

JS: We require high sensitivity, imaging capabilities, and tandem MS while directly working with brain tissues. We evaluated a number of available instruments in making our decision. The Bruker Ultraflex has the highest sensitivity for our neuronal tissue samples, the easiest and most flexible imaging software, produces high-quality tandem MS data directly from tissues, and is equipped with an excellent quality stage that provides precise movements producing high spatial resolution images.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

JS: We had a different manufacturer's older instrument in the laboratory and wanted to upgrade it, and ended up with a new company's instrument. Our existing protocols actually did not influence our decision. Because of the improved imaging software and the capabilities of tandem MS, we changed several protocols we had been using to confirm a peptide sequence when we received the new Bruker Ultraflex II MALDI TOF/TOF.

GT: What steps would you recommend to people looking to buy a MALDITOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

JS: The specifications that one needs for an instrument greatly depend on the way it will be used. A first question is whether a laboratory requires MS imaging rather than semi-automated proteomics experiments. To be practical, MS imaging applications require a high repetition rate laser, long laser lifetimes, unique software, and operating protocols. While our applications require as high as possible sensitivity for peptides as we occasionally work with samples in the picoliter to nanoliter volume range (such as single neurons), other applications require a different set of performance specifications. For example, there is not a single mass analyzer that delivers the ultimate performance for all applications and so perhaps this question should be broadened. While a specific sample type or application may require MALDI, you should ask whether you need multiple stages of fragmentation or the mass accuracy of MALDI-FTMS. Ask a series of broad questions related to your particular needs and it often becomes obvious which instrumental platform will be the best fit.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

JS: Because of our laboratory's experiences with MALDI MS using a range of neuronal samples, we tested several samples and sampling protocols that we knew produced robust results. When the new instrument was purchased, its operation was different enough that there was indeed a several month training period; we have since learned how to obtain better data from the same samples and now work with some samples that we could not characterize previously. The ability to use our old sample plates directly with the new instrument using a special stage inset made the transition even easier.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

JS: We currently have a service contract and the instrument has been very reliable. We selected our instrument more on the basis of performance specifications and less on QC and maintenance.

Tim Veenstra
National Cancer Institute-Frederick


GT: For what type of research do you use your MALDI-TOF, and which one do you have?

TV: We use our MALDI-TOF primarily for protein identification and characterizing post-translational modifications (PTM). Protein identification is done through separation using one-dimensional SDS-PAGE, followed by in-gel digestion of the protein bands that are extracted from the gel. For mass spectral identification, our laboratory uses the ABI 4700 TOF/TOF. We like this instrument because it has tandem mass spectrometry capabilities, which are important for validating protein identification and PTMs.

GT: What factors were most important to you when you were choosing which mass spec to purchase?

TV: Sensitivity and mass accuracy were critical in our selection. Sensitivity is an obvious criteria. Almost every aspect of protein chemistry migrates towards greater sensitivity in our measurements. Mass accuracy is critical in that it enhances our confidence in the identification of an unknown protein as well as in the characterization of any type of PTM. Probably the primary reason we purchased the 4700 over other MALDI-TOF instruments was its tandem MS capability. While high mass accuracy is great in determining the presence of a specific PTM, nothing is more reassuring than having MS/MS confirmation. Another important factor is ease of use. Many scientists will make use of a single piece of equipment. Therefore it is important that the ability to run a particular type of mass spectrometer is not unique to an individual.

GT: How much did your existing setup — from sample prep to data analysis — influence your choice of instrument? Did you have to change your workflow to accommodate the mass spec?

TV: I always think it is important not to let the mass spectrometer dictate your workflow or the type of research you are involved in. Your research focus always needs to dictate the type of instrument that is best suited for your laboratory. When purchasing the 4700, we were looking for an instrument that would fit into our existing workflow and data analysis capabilities. What we needed was high-throughput protein identification of proteins that had been separated by either one- or two-dimensional SDS-PAGE. The 4700 has tremendous speed for this application. What we have been pleasantly surprised with is this mass spectrometer's capabilities in detecting specific modifications in purified proteins.

GT: What steps would you recommend to people looking to buy a MALDI-TOF to ensure that they investigate options from a number of vendors and find an instrument that suits their needs?

TV: The first thing people need to determine is what is the primary need for this instrument. This determination will dictate the types of samples that they should prepare for a demonstration of the instrument. I cannot understand why laboratories rely solely on a series of standard solutions to evaluate an instrument. While controlled samples are necessary for standardizing the demo across different instruments, it is important to provide samples that reflect the workflow of the laboratory. For instance, if your major effort is the identification of gel-separated proteins, then prepare a series of these and send them to different vendors to get a sense of protein coverage and mass accuracy on the peaks. If you are interested in something more elaborate, such as tissue imaging, then it is important to determine if the vendor has a "package solution" that you can evaluate. The package solution should include necessary sample preparation devices (e.g. matrix deposition) as well as software analysis tools. The bottom line is that it is important that a component of the demo samples be reflective of actual samples that your laboratory would actually analyze on a day-to-day basis.

GT: When your instrument arrived, what did you do to test the machine and establish quality controls for your lab?

TV: Once the instrument has been installed and shown to pass specifications by the manufacturer's technicians, we will calibrate the instrument ourselves before running experimental samples. While we could put the instrument through another set of demo-type studies, we feel it is important to move the instrument quickly into operation mode to evaluate how it will behave in the normal lab workflow. The calibration standards are used routinely for quality control in the laboratory. The standards are purchased directly from the manufacturer of the instrument. While QC standards are run at regular intervals, it is also quite easy for most experienced scientists to determine if the instrument is acting "funny" during the analysis of experimental samples.

GT: How do ongoing maintenance and QC factors of your machine affect your choice of instrument?

TV: One of the biggest frustrations is equipment failure. It is important that routine maintenance be straightforward so that a scientist can quickly conduct any routine maintenance that is required. Another very important issue is service and support from the vendor.