Buying a mass spectrometer from Bruker Daltonics can be a bit like shopping for a new car. On a visit to the company's Billerica, Mass., headquarters you can ogle various models ranging in size, performance, and, of course, price all parked around a spacious showroom. There's the OmniFlex desktop instrument, the UltraFlex TOF-TOF, two kinds of LC/MSs, the MicroTOF LC, a BioTOF II and a BioTOF Q, an FT/MS, and the AutoFlex TOF for SNP analysis. (To see Bruker's latest model, the 12-Tesla Apex Q hybrid mass spec, you'll need to make a trip to the basement skunkworks.)
As a prospective buyer, you're welcome to bring your own sample along and take any of the $5 million worth of equipment for a test drive. And if you're anything like the academic research director who came in the other day to purchase an ion trap and was wowed into leasing a TOF-TOF as well, there's a good chance you'll leave here having spent more than you intended.
From his second-floor office that looks out over the demo floor, Bruker CEO Frank Laukien recalls the company's short, rapid climb to success in the life sciences instrumentation marketplace. The German-born son of a Bruker Company founder earned a degree in physics from MIT and a PhD in chemical physics from Harvard in 1988 before starting up his own substance detection instrument business. Fifteen years later, though current world events have spiked that market, life sciences sales account for 80 percent of Bruker Daltonics' revenues — half of that in proteomics.
After launching its Fourier transform hybrid mass spec at Pittcon, and announcing a merger with Bruker AXS in March, the 1,150-employee, soon-to-be renamed Bruker Biosciences is more determined than ever to become a major player in proteomics — a field that Laukien is betting the company's future on.
"[Proteomics] is not going to be like genetic sequencing was with that one-time extreme ramp-up driven by competition," he says. "It's going to be slower growth, perhaps, but much more sustainable."
To be sure, Bruker Daltonics has nearly tripled itself in the last four years, going from about $40 million in revenue in 1998 to $116 million in 2002. Its new other half, AXS, which specializes in x-ray crystallography and lab automation, doubled in the same period. This year, Laukien says he expects $250-$260 million in revenue and a compounded annual growth rate of greater than 15 percent for the combined company.
GT visited him and the Bruker showroom in April to find out how he intends to make that happen.
Genome Technology: With regard to your recent merger: How will combining high-throughput, protein-analysis instrument sales with protein-structure hardware pursuits give you an advantage? Did your market research show that these efforts inside pharmaceutical companies or academic facilities are related or moving closer together?
Frank Laukien: The trends in the market, and the proteomics market is no exception, is toward more information-rich technologies. People are moving more toward getting detailed information … on target validation or … beyond the very large throughput screening towards more detailed understanding in expression proteomics, which we do with mass spectrometry.
It was exciting two years ago that we could identify so many proteins from gels. Now people want to move way beyond that; they want to have very large significant sequence coverage, they may want to identify and localize post-translational modifications, and … understand three-dimensional structures, and that is, of course, done primarily by x-ray crystallography and NMR.
It's not typically the same scientist that does both, but it may very well be two professors or groups in the same department, or it may be two scientists in the pharmaceutical or biotech company that report to the same director or vice president of research in drug discovery or development. One shouldn't [think] that we're trying to come out with one black box instrument that does both — that's not going to happen. But we certainly see the complementarity of moving beyond the expression proteomics and clinical proteomics to offer, under the same brand from the same supplier, tools that have much less throughput but really incredible information content if you have a protein and you want to understand not only its three-dimensional structure but how it interacts with other proteins or pieces of DNA or a small-target potential drug molecule.
We think integrating that from a bioinformatics software point of view and also from a marketing point of view is a valuable strategy. We're not suggesting that one plus one all of a sudden equals three, but we do see some chance for incremental sales from a broader marketing approach with a broader technology base and more products into proteomics. One plus one perhaps is 2.2. But that's still an incremental improvement for both of us.
GT: Your new Compass software will make the data generated by any of your mass spec instruments interoperable. Will you also now provide a way to link that data to data generated by protein structure instruments?
FL: Yes, at an even higher level. I'm not necessarily suggesting having a similar operator's user interface for someone who does x-ray crystallography because those are not the same people. But they're in the same research process, the same drug discovery process. In Proteinscape, our software that does all the workflow management and all the information management of our Proteineer System, we've already begun to include structural protein databases. At that very high workflow and information management level, as opposed to the instrument operator's level, we hope to get some additional integration and provide some real technical synergies, not just marketing synergies.
By combining with Bruker AXS, the only public providers of structural proteomics tools, while also expanding into clinical proteomics via a partnership into interaction proteomics, we'll have a broader and, after this merger, a rather unique set of tools — perhaps the most comprehensive set of tools that people would want to have for proteomics. It's a new strategy; people have done it on the expression proteomics side — Amersham wants to combine gels and mass spectrometers — but as one looks further downstream in proteomics, that hasn't been done yet.
GT: You've said that the merger could enable some more strategic activities such as acquisitions and that those would be most likely in the areas of instrumentation, robotics, separation, and potentially consumables. Can you explain what customers' needs are in these areas?
FL: Bruker AXS actually did a couple of small-scale acquisitions in the last two years that were nice consolidating acquisitions within their x-ray crystallography … but some of the midsized acquisitions for each of our companies, being smaller players, were not accessible. We believe that as we become a more midsized life science tools provider we can look at acquisition opportunities that we could not access individually because of scale and size.
We on the Daltonics side would generally be looking at complementary technologies whether they are in sample preparation, robotics, or liquid handling. Also the back end, bioinformatics, we do quite a bit of that within the company but it's also one of those things we might add from the outside.
For the combined Bruker Biosciences we may also be exploring, particularly in proteomics, potential additional technologies that we presently do not have within Daltonics or AXS. It does not have to be proteomics, but also we are certainly looking at life and advanced material sciences.
I'd like to not have this over-interpreted; we're very selective. All of our growth has been strong, organic growth. I think that's the key to our success in almost tripling Daltonics in the last four years.
And our internal innovation engine, our product development capability, is still a key feature that distinguishes us a little bit, at least in style, from some other companies in our industry. We'd like to very much maintain that. We're not all of a sudden switching to an acquisition-driven model. We'll still be quite selective when it comes to that.
GT: Do you see more interest in end-to-end mass spec platforms or do scientists still prefer buying one instrument at a time?
FL: You see both. A proteomics expert like a GeneProt or the Roche Proteomics effort under Hanno Langen doesn't buy the complete solution; they buy a bunch of mass spectrometers or bits and pieces and some of them have helped us develop some of the software solutions.
But customers that get into proteomics for the first time and don't have all the expertise or the time to apply all the expertise [prefer] one solution. We have customers at medical schools or at companies that are not proteomics powerhouses yet but want to have a proteomics effort for a certain target instead of research topics.
You see a continuum of what people want. Now that mass spectrometry for the life sciences has become so much more accessible, non-mass spectroscopists tend to be more interested in buying a complete solution. The experts that have been doing this for years tend to rather design it themselves — they have many of the pieces already, they may just need a TOF-TOF to complement their previous capabilities.
GT: Your new hybrid FT/MS instrument, the Apex Q, seems to be serving a whole new purpose in proteomics.
FL: We're the market leader in FT/MS but it has been primarily a research-rich field and it has been applied by pharmaceutical companies primarily to small-molecule analysis, combichem analysis, and all sorts of small-molecule applications. With the advent of these hybrid FT/MS systems, particularly also with the high-field magnets, we believe there is a real additional market segment opening for FT/MS and for mass spec in general.
All of a sudden this is very applicable to some of the most exciting research in proteomics. [One is] shotgun proteomics, where you get away from 2D gels and basically do an imperfect separation, just a single LC run, starting with a proteome or a subset of a proteome. You accept that you don't get perfect separation, but you then use the enormous resolution and separation power of the FT mass spec itself to separate remaining mixtures. This is what has been pioneered by Dick Smith's group at PNNL.
In addition, something great that came from Fred McLafferty's group and various professors and junior professors out of Cornell is the top-down proteomics approach. You don't have to do an enzymatic digest of your protein and then look at the peptide fragments. You can start with an intact protein and use the mass spec and its top-down dissociation capabilities, which are unique to us. Those capabilities are very exciting and there are a lot of publications coming out on them. I would still characterize this as a methodology research, but it's gone sufficiently mainstream that we think a lot of groups that previously wouldn't have gotten into FT/MS because it couldn't deal with large molecule problems are suddenly very interested in getting a high-field hybrid FT/MS to pursue both of these approaches.
GT: Is this capability what justifies the price of your hybrid instrument? At $1.2-$2 million, it's twice as expensive as Thermo Electron's new hybrid. Who will purchase such a costly piece of equipment?
FL: Just like in NMR, the price point depends on the size of the magnet and the configuration. An NMR can cost from $200,000 to $5 million. We have a 7-Tesla system including the hybrid capability and the top-down capability that is in the $850,000 range. That's going to be the instrument that, at least on price point, would be competitive with the other hybrid FT/MS that was recently introduced, although we think, given that magnetic field strength, that's really not what people will mostly want to buy for proteomics applications. We think that's going to be a very good machine for small-molecule metabolomics applications. Those are still important in FT/MS as well.
Somewhat in analogy to NMR, where most of the structural proteomics and similar work is done by necessity at higher magnetic field strengths, the same is true in FT/MS, which is why we have these 9.4-Tesla and 12-Tesla systems available now. They can cost anywhere from $1.2-$2 million.
It's not going to be everybody's system, but we found it remarkable that when Maldi TOF-TOF came out at price points of $500,000 to $600,000 and everybody had budgeted for high-end MALDI TOFs at maybe $350,000, within a half a year, because it was such a compelling new capability, customers did find the funding, not only in companies, but in a lot of universities and medical schools.
The Mayo Clinic just bought a $2 million FT/MS system from us. We have a lot of negotiations for high-field 9.4- and 12-Tesla systems going on right now, so we believe there will be a sufficiently interesting market, not just one or two systems a year.
If [performance] was just a little bit better than before it would be hard to justify the budget. But they can do new experiments that people have been dreaming of for proteomics. It's remarkable — there's a compelling scientific reason and you really can do something you couldn't do before. People tend to find budgets, so we're pretty optimistic that budgets will be available for hybrid FT/MS.
GT: Can you talk about where the current orders are coming from?
Mayo's was a 12-Tesla. We have orders from a Dutch university and a French pharmaceutical company, and a US national lab has upgraded their existing FT/MS system to the hybrid capability. And we have a lot of people in the pipeline evaluating this, including a number of pharmaceutical companies. Many of them have been looking at the 7-Tesla, but it was the only one that was available until recently.
The majority will go to national labs and major research universities and medical schools. If you look at an extreme at the $5 million 900-megahertz NMR systems, there's not a single one in industry but 20 or so have been sold worldwide. Most of them go to university consortia, NIH, Scripps. All the universities in the Michigan life sciences corridor combined to buy one, and the universities and medical schools in Cleveland have acquired one together. But for a pharmaceutical company to have $800,000 to $1.2 million is pretty typical if they think it's a compelling reason and it's not just another toy or technology platform.
GT:What are the most important areas in which mass spec technology still has room for improvement? Is it accuracy, sensitivity, pricing, footprint, throughput, or something else?
FL: Yes. Five times yes! We brought out a small benchtop electrospray TOF system that has better resolution and accuracy than prior generation electrospray TOFs. For our qTOF system, which we call the BioTOF 3, we've just doubled the resolution and accuracy. For our Maldi TOFs we can now do up to 100,000 samples a day, so there we've pushed the throughput.
And then "something else" is probably, when can you get something qualitatively different? I think Maldi TOF-TOF was and is a great example of that. You can get with such ease and sensitivity so much sequence information and even use it for de novo sequencing — that went beyond the expectations that we had when this first came out a year and a half ago. So Maldi TOF-TOF was "something else" — something qualitatively different, which I think is why it has become such a success for both ABI and us.
The other "something else" this year will be the hybrid FT/MSs with that top-down capability. That's really something qualitatively different where you can do things not just better and more accurately with better sensitivity, which is important, but there are great examples where you can do something really qualitatively new.
GT: One analyst said after your merger that the challenge for you now would be to improve the overall company's profitability. He said generally companies in the instrument market show operating margins of 10 to 20 percent, but that last year Bruker Daltonics did 5.9 and Bruker AXS was at 3.4 percent. First, is this analyst's assessment correct, and if so, how long do you expect it will be before your margins are on par with other publicly traded instrument companies such as Waters, Dionex, ABI, Thermo Electron, and PerkinElmer?
FL: That's absolutely correct. We were not a life sciences instruments company 10 years ago. Our roots in mass spectrometry were in the substance-detection business. That's what funded our ability to then move into life science mass spectrometry.
With that new business model we became profitable in mid-1999. We have good gross margins. And we have been driving top-line growth as well as bottom-line growth ever since then. So we're at the level where we're steadily moving our profitability. Our earnings before interests and taxes, depreciation and amortization are at 12 percent — about $15 million. If I look at the merger, the combined company in terms of profitability is about nine percent. It has also more than doubled in the last two years. Our goal as part of the merger is to bring that from nine to 15 or 16 percent for the combined company.
Obviously, for Bruker Daltonics it would be higher already. That puts us well into the profitability range of a lot of companies. It doesn't make us the most profitable; those tend to be Waters and Dionex. But I don't want to turn on a dime and cut my R&D spending way down to five or six percent. Then we could be very profitable, but right now we want to maintain this growth momentum that we've had. I want to increase the bottom line to where we're in the mid-range of profitability by 2005 and then in subsequent years drive that even further towards the high range.
GT: Why is it so important to grow larger?
FL: We've been debating that for a long time. We cherish "small is beautiful" and [saying] "look how fast we can grow, look how focused we are." We want to maintain that part of what arguably has been a recipe for success.
But there are advantages to scale. In terms of selling to large industrial accounts, some prefer to buy from larger suppliers. With more scale we can get more complementary marketing opportunities that wouldn't be accessible for each of us individually.
There's a clear discrimination in the capital markets against small caps these days. Two years ago, a pure play in mass spec was just wonderful. But then again, years ago everything was wonderful. And in this very bearish environment the fund managers are reluctant to invest even if they think, 'Wow, there's a hundred-million-dollar gem, that's a great company." Shareholder value for existing investors can also be proved for the capital markets by having the larger size and scale while maintaining the operating companies underneath that are still mostly focused on their business while also hopefully being able to take advantage of some of the operational costs and marketing product synergies.
GT: When you say you want to reach mid to large size, what number designates midsize?
FL: In our industry there are the big players that have $800 million to $2 billion market cap: ABI, PerkinElmer, Thermo. There's almost no midfield. You could argue that Dionex is in the midfield, that we with about $250 million this year will be in the midfield. Affymetrix has grown to that midrange level too. I once saw a list of 43 public companies that probably all have market caps below $150 million. Many of them have cash burn and many went public in the last two or three years.
There are a few big companies, and many of them acquire smaller companies — Waters acquired Micromass, ABI acquired Perseptive. And then you have all the small companies. It seems to be difficult to grow toward the center, but that's what we want to do.