Part two in a two-part series.
(To read part one, see Cell-Based Assay News, 6/22/05)
Kurt Herrenknecht, Head of Cellular Applications, Evotec Technologies
Dietrich Ruehlmann, Product Manager, Imaging and Analysis, BD Biosciences
Michael Sjaastad, Director of Marketing, Imaging, Molecular Devices
Joe Zock, Senior Customer Support Scientist; Manager of HCS User Services, Cellomics
LONDON — At the Marcus Evans Practical Experiences in High-Content Screening conference held here two weeks ago, Cell-Based Assay News sat down with representatives from four well-known HCS vendors — Evotec Technologies, BD Biosciences, Molecular Devices, and Cellomics — to discuss current trends in their industry and future challenges. In part one of this series, published in last week's Cell-Based Assay News, the reps weighed in on such topics as the HCS growth curve, the adoption of HCS in primary drug screening, and the importance of image analysis and informatics. In this week's installment, the participants discuss new HCS assay formats, reagents that will drive HCS, selling to academia versus pharma, and the integration of HCS into systems biology.
Regarding HCS formats — a lot of this was designed originally for well-plate applications, but now it is starting to incorporate flow, microfluidics, cellular arrays, culture plates, etc. Do these have a place in pharma, and are there developments of this sort on the horizon?
Zock: Yes. We're releasing next month a slide holder. You could argue that something like that is a long time coming, but I pushed very hard for a long time that we would never put a slide inside of [our] machine. That is a screening machine, it's not a microscope in a box, and it's more complex than that. But I've had to eat my words because of microarrays — tissue microarrays — and because of various other technologies such as [Whitehead Institute's David] Sabatini's [cellular array] patents. All of these things are slide-based, and so we're coming out with a holder. But that's only the beginning of what I can conceive of a lot of different changes as people get more complex, as they want to do more complex biological studies.
Ruehlmann: Do you want our slide holder? (Group laughs) We've had it for the last four years, and none of our customers use slides. I think your question goes in the direction of 'Should we abandon the SBS footprint?' And that's different. Should we be able to take the instrument and put it inside a real incubator and image all the plates, or something like this? I think [Acumen] tried this a few years ago, with the tape — remember this? (Asks other participants, who nod in agreement) I think there is a very high technology threshold. All of the infrastructure that's not imaging — which is our expertise — so, pipette handlers and everything else, is based on something that fits into the 96-well SBS footprint. And any vendor who attacks that will have a really hard time to overcome this resistance, even if it's technologically better. I think something has to be so substantially better to give up the well footprint, and then we can adapt quickly, and turn a microscope around very fast.
Sjaastad: We do need to be sensitive to the formats that are used in academia, because we're all looking in that direction. We're looking to the core facilities; we're looking to the labs that have the money. So I think we should be sensitive to all the existing formats — that's where the ideas come from.
Zock: I was approached three or four times during this conference by people asking 'What are your thoughts on making masters degree programs for high-content screening?' or 'What do you think about teaching young people about this?', so that when they come out they are asking 'Where is my imaging system?', because that is how they learned cell biology and how they now approach this. I think that this is going to happen, and it's going to require being sensitive to the needs of people in academia.
Perfect segue — what are the challenges in selling these platforms to academia? Is it significantly different from pharma?
(Participants nod in agreement, except Ruehlmann)
Ruehlmann: I'm the one that disagrees. I think the segmentation — pharma industry, and in the US, government and academia — is not necessarily useful, and I would like to point to the NIH Roadmap by Dr. Zerhouni, who has arguably one of the most impressive drug programs I've ever seen, and it's governmental. It's not profit-oriented, and you cannot go in there expecting a small lab with two post-docs. You are seeing buildings being built for high-throughput screening. You're seeing things that are enormously complex on the bioinformatics side — much more complex than any pharma I've seen. And at the same time, you look at small biotech companies — for instance, BioImage and Norak [now Xsira Pharmaceuticals] — both of them have two [GE Healthcare] IN Cell 3000's each, and they've published on that. Now that is million-dollar instrumentation that this small, venture-capital-funded company has; whereas you struggle to get instrumentation like this into AstraZeneca, for instance. So I think from our perspective, on our system, we don't differentiate any more based on where the customer is; we differentiate based on what the customer is trying to do. And I think that really has dissolved in academia, especially in the US, though I can't speak for Europe.
Herrenknecht: We see that also in Asia — for example, Korea. It's not really governmental-types of institutions, but, for example, the Institut Pasteur in Korea has recently purchased an entire high-throughput screening system from us, including two imaging systems. This is not pharma, this is not biotech. So we see a different type of customer shape.
Zock: I can say that Cellomics continues to see a fundamental difference in their approach — throughput numbers versus targets — and so we try to make the system flexible, or approach people from their perspective so that they are comfortable. This becomes critical when you're trying to sell a second system to them, or a third, or a fourth, or a seventeenth. So there is a need and a driver to get adoption in these larger pharmas globally. We're pushing that very hard because we don't have a million people — and no one does — but if it's globally adopted, then they start to talk to each other more than we talk to them. And we see that now; everyone sees that those people have multiple systems, and they have every platform there in some way, shape, or form, and they use them for different purposes.
Sjaastad: I want to jump in here, because that was the point of some slides we showed today. We analyzed data from three systems that weren't ours …
Zock: That raised some eyebrows …
Sjaastad: Yeah, and I had a very specific intention for doing that. It comes from customer requests. They don't want an incompatibility across platforms. We all have instruments that have strengths and weaknesses, and they are all going to settle into particular applications. But we're generating information, and I think that the more we can be cross-platform compatible with our information, the more we're going to be accepted within industry. So I'm behind that 100 percent.
Zock: And we're behind that and we're making products. Having GE have our back-end [informatics], for instance, and it will continue to go that way.
Ruehlmann: At BD Biosciences, we've established a standard for flow cytometry data. This data has a standard, and you can open it cross-platform, so from our end, this eventually will happen anyway, and we're certainly open to work on a standard. In fact, the National Institutes of Standards and Technology has offered to be a neutral ground to develop this.
Zock: This is very exciting.
Ruehlmann: It is, and for us, there is no issue. We will collaborate with anyone who is accepted on this.
How important are supporting technologies to HCS, and what supporting technologies are needed? I'm talking about reagents, assay technology, transfection techniques, cell culture …
Zock: Well, you can make a great razor, but razor blades are what drive it. Applications drive this business. Nobody comes in and says 'I love this thing because it has this nice green LED on the side of it.' People come in and say 'I've got these neurons, and I can't look at them any other way', or 'I want to look at tube formation', or 'I want to count cells.' Whatever it happens to be, it's application-driven. So yes, we hear that all the time. At our user group meeting here, the message was very strong: 'We don't have enough reagents.'
Ruehlmann: From BD's end, we work with Pharmigen on the antibodies; we work with BD Labware on the plasticware; we work with Clontech on the fluorescent proteins; we work with Research Triangle Park on the R&D; and it is very clear, these are all very synergistic technologies. And we have a very lucky position in that we offer all of these technologies independently so far, and flow cytometry and automated imaging are sort of the focal points where these technologies converge, because we need plasticware, antibodies, and some kind of reader. The auxiliary technologies that actually help the scientists discover their drug mechanism are extremely important to the business model.
Herrenknecht: I would like to raise a question: From the users of our instruments, have we any idea of how many very new applications are actually going to be used compared to the individual assays they are developing on their own? Yes, of course, you sell your reagents (indicates Ruehlmann), and your reagents business is certainly very different. But I would like to see numbers. You (indicates Sjaastad) pointed out this morning that you do not sell any sort of reagents …
Sjaastad: Not imaging kits. Molecular Devices sells a lot of other reagents, but at this point, we haven't ventured down that road. We may. But it is an interesting thing … there's an awful lot of reverse engineering that goes on. It's been an early-adopter technology — they sort of have their assays, and are looking for a detector. I think we're going to see a change in that. I think they're going to be looking to adopting HCS and seeing what they can do with it.
Zock: It's really two things. First, something novel that no one else can do that is enabling. For example, a MatriGel, or something like that. But then there is also 'How do I get my job done quickly, because I have only three weeks in which to do it?' The people that are in that boat — and it's not necessarily the assay developers, because their job is to reverse-engineer things — but it's the toxicologists, or the lead optimization biologists, or the target identifiers that don't necessarily develop assays for a living. They just want to use a tool that is ready made, and which they can use to get their job done. That's where kits and reagents actually shine.
Ruehlmann: And this is actually the philosophy of BD. We supply antibodies and reagents for flow cytometry without necessarily packing them into kits. We have done kits, and we will do kits — there is no question about that — but it's not locking someone into having to buy this kit in order to use it on an HCS platform. That is not going to happen. However, we do offer advice on the antibodies, and we do offer advice on the other reagents we are developing, and the idea is that the customer has a smorgasbord of validated reagents that they can apply — on our system, or on anybody else's.
Sjaastad: I don't think it's only assay kits. In addition to being able to find hits in new ways with cellular informatics, the points that we're finding right off the bat are the trends in the pipetting, the quality of the plates, and we're going to apply pressure all the way from cell dispensing to plates to media exchanges — and so we'll push all those fronts with this.
How is HCS being integrated with the idea of systems biology — combining this data with that of proteomics and microarray data, etc.? Is it being integrated with this data? Should it be?
Herrenknecht: I think the answer to this is a clear 'yes.' At Evotec we are also proud of government-funded research projects, and we are part of one big systems biology project, where our imaging system has been really central. It's really about generating high-throughput images, and having a lot of images in order to provide the informaticists with the data they need.
Ruehlmann: I think that systems biology is a buzzword, as much as I like the concept of it. The ultimate systems biology institute is MedLine. The data is there. And this is where bioinformatics companies come in. They need to distill the data, and make information of it. And this becomes systems biology, and from our end, we are very happy to be a provider of data, and we are also very happy to be a provider of part of the information. I don't see us being the absolute necessary cornerstone of every systems biology [project]. I think the data we, as HCS companies, contribute will be fundamental to systems biology.
Zock: I agree. People talk a lot about it, but it remains to be seen exactly how systems biology is going to be implemented. I think we're all standing there saying 'We've developed the tools. You can generate whatever data you want. But how are you going to put it together?'