AT A GLANCE
NAME: Stephen Hammond
POSITION: CEO, Scottish Biomedical, since 1996
BACKGROUND: Director, Glaxo Research Center, Verona, Italy, 1990-1996; Head of biology, antibacterial chemotherapy, Astra Research Laboratories, Sodertalje, Sweden, 1984-1990; Lecturer, microbiology, University of Leeds, 1978-1983; Postdoc, Bristol University and Purdue University, 1974-1977; PhD, 1972
As CEO of drug-discovery service firm Scottish Biomedical, Stephen Hammond has a first-hand look at the trends shaping the pharmaceutical discovery market.
According to a recent article published by Outsourcing-Pharma.com, Hammond voiced his support of high-content screening technologies at a recent conference for pharmaceutical outsourcing in London, and said that the method was poised to overtake traditional techniques such as HTS for generating drug leads.
This week, Hammond expanded on his comments in an interview with Cell-Based Assay News.
Tell me about Scottish Biomedical.
Scottish Biomedical is a drug-discovery service company based in Glasgow. It’s been going for about 12 years now, and we employ about 60 scientists, half of whom have PhDs. We are a pre-clinical CRO, and we work on behalf of big pharma and biotech, and even small biotech, who want to outsource the drug-discovery process.
We’ve got everything in terms of proteomic and genomic target identification and validation; developing high-throughput and high-content screens; cell-based assays; our own libraries and small-animal facility. We do everything up to proof-of-principle animal models; we don’t do regulatory toxicology. We do everything on pretty much a fee-for-service basis. At any one time we’ll be working for eight or nine companies. Some of those projects will be three years in duration, and some will be three months. We’ve got a few that we call fully integrated programs where we’ll identify a target, devise a screen, run our libraries, and get up to proof of principle in animals. For others we may just clone a target that the customer gives to us, develop a screen and validate it, or maybe run our library against one of the customer’s screens. It just depends.
You were recently quoted at the Pabord Pharmaceutical and Biotech Outsourcing conference as saying that high-content screening was poised to overtake other methods as a the preferred way to generate hits in drug discovery …
I see HCS as a step on from HTS rather than something completely new. But certainly the perception in industry 10 or 15 years ago was that HTS was going to solve all your problems. All you needed to do was to get these huge libraries, huge robots, and run 4 million compounds in zero time, and you would get wonderful hits. And the problem is that it just hasn’t gone that way.
We’re driven very much by customer demand. We do what people ask us to do. And we’ve noticed a trend over the past five years where people are asking for more and more cell-based assays, rather than ‘get the enzyme into solution.’ We’ve also found anecdotally with the enzymes that we’ve worked with – for example, one client asked us to clone an enzyme, and we did this high-throughput screen and found some compounds. But the compounds didn’t work as well in cells as you would imagine. At first we thought it was a permeability problem, but it wasn’t. It was the enzyme actually in the cell was in a different physiological state than what it was in solution. It was held as part of a complex in a cellular environment.
So HTS is moving down a route where you’ve got more and more potent inhibitors of that enzyme in solution, but away from compounds that have an effect on that target in a cell. Certainly we’ve noticed, then, that more people are asking for cell-based assays, and that leads to a paradigm shift. The instrumentation is different. We’ve got more reporter-based assays, more image-capturing assays, and more multiplexed assays, which is the latest thing – people want to look at pathways and multiplex to look at two or three things in a pathway. Because you can’t get the throughput on these sorts of assays that you can on a straight kinase or phosphatase assay, but the amount of information that you generate goes up tremendously. So you’ve got to invest in new endpoint readers and data-handling systems to cope with that. If you make that investment, I think the quality of the information you get is a lot better and more useful to customers.
Along with that, we are cutting down the sizes of the libraries we screen. If you couple a more iterative HCS with a more focused, iterative library, our experience shows that you get better, higher-quality, and more interesting hits than if you zapped 4 million compounds through in HTS.
Many people in this industry have said that HCS will never replace HTS, but will at the best be a complementary technology.
I think [the two technologies are] probably bedfellows, to some extent. If you’ve got something really simple, and you’ve got a big compound library, then yes, do HTS. But I think that HCS gives you much more information and understanding about the target. I think that the coupling of a good high-content screen with a good, well-focused and well-designed library is much more relevant and useful – particularly if you’ve got a more sophisticated target, and are moving away from a simple phosphatase, kinases, phosphoesterase, or something like that, into something more pathway-oriented.
Where do you draw a distinction between high-content screening and high-throughput cell-based assays?
I would think the key differentiator is that you’ve got a viable cell that’s doing something. In some HTS where you are using cells, like in a receptor-binding assay, that’s not HCS. To me, it’s when you’ve got an intact, growing cell, and you’ve got some sort of event going on in it with some sort of kinetic to it. The other thing that we’ve found with HCS because of this is that you get bonuses. With some of our HCS clients, we’ve found that some of the false positives are actually interesting compounds in their own right. If you’ve got a green fluorescent protein expressing somewhere, you can sometimes pick up false positives that went by a different route to the same end point, and they come out as a bonus or surprise. Sometimes they’re as interesting as your primary mechanism. You wouldn’t get this at all in HTS. It would just be a miss.
So to me, it is a cell-based assay where there is a kinetic measurement of something. It gives you a temporal resolution of an event you’re measuring. That, of course, presents a problem with data capture, because you may have a curve or several curves to interpret. The data used to define a hit is much more complicated than in an HTS screen. Often, for example, if you’ve got a phosphatase, and you’re doing it in a plate reader, you can see on the plate yourself that it’s yellow instead of green. It’s a crude all-or-nothing response, but with HCS, you get a more interpretive response.
“Certainly the perception in industry 10 or 15 years ago was that HTS was going to solve all your problems. … And the problem is that it just hasn’t gone that way.”
What types of assays, instruments, or reagents is your company using for HCS?
When we first started 12 years ago, we were basically using colorimetric and isotopic assays. Over the years we have gotten a lot more into fluorescence measurements, like fluorescence lifetime, FRET, BRET, polarization, AlphaScreen. Then came the FLIPR technology, and we’re heavily into that. We’re now using the second generation of machines for that. And we’ve got several machines for multiplexing, as well. A problem with the FLIPR-type assays is that you’re now measuring kinetics in a single well with cells, so basically it’s only measuring one scan of the plate at a time. The residence time of the plate in the machine is actually quite long, and that really slaughters your throughput. But we’re always looking at new things. We’re looking at the Guava instruments. The move is to get away from single-point colorimetric or isotopic assays to more time-resolved types of techniques.
Throughput is cited as one of the bigger problems with HCS…
And another problem here is miniaturization, in the sense that you’re seeding plates with cells that you want to stay alive in the actual assays. The way that you produce the inoculums and grow the cells for the assay reproducibly so it behaves the same across the plate in every well is quite difficult. In addition, your well price is going to go up.
How are you dealing with some of these challenges as you begin to implement HCS more?
There is no real quick fix. It would be nice to see someone develop technologies that reproducibly grow cells and plate them out for you in a reproducible format. We have people who do that, and that’s their skill. But it seems to take a real live cell biologist. You just can’t automate that. If you’re working with bulk standard cells – Hoechst or CHO, or things like that – then that’s OK. But some of our clients want more esoteric cell lines, like neuroblastoma lines, nerve cells, or liver cells, and that gets a lot trickier. One of the hopes we have, but we’re not there yet, is working with stem cells – actually taking an adult or embryonic stem cell line and applying differentiation protocols. I think that will be the thing that will happen over the next four or five years to help change this industry.
Scotland is certainly at the forefront of stem cell research. Does your company have a greater opportunity to explore this area because of that?
We are talking to various research organizations here in two areas: First of all getting better technologies to differentiate and expand stem cell populations reproducibly, and then actually increasing the variety of lines available. We are probably in a slightly more favorable situation than parts of the US, but it’s still not perfect. There is still a lot of concern about the use of that sort of material. The regulatory framework is certainly better than in the US, but it’s still quite constraining regarding what we can do.