Dr. Djaballah came to MSKCC in June 2003, and it took him nearly a year to get things up and running. The HTS laboratory became fully functional in August 2004. Today the Center’s chemical library consists of nearly 380,000 compounds, and its proprietary small interfering RNA libraries consist of nearly 21,000 duplexes covering 6,000 genes. The collection also includes libraries licensed from other commercial sources. The facility has now developed and validated nearly 56 assays, and these screens have led to the identification of several “hits,” initiated exploratory chemistry around several scaffolds, and led to patent filings and progression of compounds into the clinic. One of the screens contributed to FDA approval of the repositioning of a known drug for the treatment of retinoblastoma
Dr. Hakim Djaballah, Director of the High-Throughput Screening (HTS) Core Facility at the Memorial Sloan-Kettering Cancer Center (MSKCC) in New York City, talks to Tanuja Koppal, contributing editor to Lab Manager Magazine, about his experiences setting up a screening lab in an academic environment. After spending many months designing and setting up his lab, Djaballah is now finding it gratifying to see his screens yield new drug candidates that have moved to the clinic.
Can you provide some details on how you went about setting up the HTS lab at MSKCC?
It really boiled down to how much capital investment MSKCC was willing to make, and we had to make some very tough decisions to build a facility that could support the ongoing research in a very diversified fashion. I put a budget together based on the previous two labs that I had helped set up in biotechnology companies. My strategy was to make the screening facility very versatile to support both chemical and RNAi screens. The initial budget, drafted before I got here, was modeled based on what was happening in other academic labs, but it was later decided that the screening lab at MSKCC would be set up along the lines of the labs seen in industry. So I went about creating a business plan that included a mission statement and a return on investment of 10 years. MSKCC was happy to make the initial capital investment, which is very rare, and we did not have to depend on getting any grants funded. The versatility of the assays, the size and diversity of the chemical libraries, and the creation of databases and customized informatics solutions were all key aspects that contributed to the business plan. You really need a sound business plan from the get-go.
Are there any resources that can help lab managers draft budgets?
Unfortunately there are no resources or benchmarks to help lab managers put together budgets. There have been some publications on how much it would cost to screen a well but nothing around what it would take to build a screening facility. One of the reasons is that how much you spend is often a very sensitive issue and has to remain confidential.
How did you know what you needed from the get-go?
I made some assumptions, and one was that the assays run would be 50 percent in vitro and 50 percent cell-based, although there was more emphasis on in vitro in those days. When we went live in August 2004, the assays turned out to be 70 percent in vitro and 30 percent cell-based, but today it is 99 percent cell-based and only 1 percent in vitro. That’s the danger in making assumptions: sometimes if you are wrong, you could be stuck with a completely different setup than what you need. I took a gamble on high-content screening, and that has now proved to be a workhorse for us. However, pushing siRNA screens in the early days was a disaster since we were led to believe that there were not many off-target effects, and that got us off to a wrong start.
Did you opt for customized solutions or did you go with the plug-and-play systems?
Mostly we opted for customized solutions. There were a lot of negotiations with the vendors. The robotic platform we currently have is the largest the vendor (Thermo Fisher Scientific, Inc.) had ever built. Having things customized was more expensive, but it provided us with the flexibility to do many things. If you go along with what is available, it will work but you will end up doing things differently from what you had set out to do. Especially on the informatics side, if the data output is not mirrored to the database you will have problems loading the data.
What other services do you offer besides the actual screening?
We provide consultation, which means a lot of people come to us early on with their ideas for screening and we spend time to guide them to get the best reagents and set up the best assay possible. We train high school and medical students for a few weeks each year to give them a feel for how drugs are discovered and how HTS plays an important role in that process. We also train graduate students and postdoctoral fellows on a regular basis.
How do you go about hiring people, and what type of training do they receive?
Recruiting the right people is very important, and this decision can sometimes make or break a screening facility, especially in academia. In academia, you almost never get people who are already trained, due to the salary discrepancies between academia and industry. So I often recruit people as research technicians or senior scientists and train them for six to nine months. Every year I also take in two or three postdocs who are very interested in going to industry but first want to join a lab like ours to see if they really like doing screening.
Right now I have 14 people in the lab, but only five or six of them are fully trained, and the rest, who are mostly new recruits, are being trained. This is not an ideal situation, but it is a consequence of our recent expansion into RNAi screening. When you are starting a new lab, you should start out with no more than two people, or else managing everything can be disastrous. It’s important to first get them trained and then recruit more people. My personnel get a combination of internal and external training. I meet with them regularly to go through their progress with the use of certain technologies. They also receive training from vendors on site.
Currently are you experiencing any bottlenecks in your screening?
One of our bottlenecks is the incubator space to hold materials for cell-based assays. We can buy more incubators, but integrating them onto the robots will be a challenge.
We will either have to take things off the robot or we will have to expand the robots to accommodate the additional incubators. And this will have to be done by keeping the downtime to a minimum, probably no more than four weeks, which is difficult.
Could you have avoided this?
This probably happened because our initial goal was to run more chemical screens, and now with the addition of cell-based RNAi screens, the incubator capacities have nearly doubled. It’s a tough situation, because if you buy more than what you need initially, it’s going to hurt the capital investment. But if I were to make an assumption today, I would probably go with 80 percent cellbased assays and not 50 percent.
What aspect of setting up the lab was the most challenging?
Setting up the informatics was the most challenging aspect, since everything was custom-built. We had to work with a number of contractors, which was stressful and very time-consuming. In fact, we completed the screen before the software programs were ready. In pharma it was different because the investments were larger, we could hire a large informatics firm, and payments were made by milestones. On the other hand, the part that was most stimulating was putting together my team and performing my first screen. In academia, it is more gratifying because you can see compounds move into the clinic. In industry, the more the compound progresses, the less you hear about it.
If you missed the Ask The Expert webinar “Setting Up a High-Throughput Screening Lab”, originally broadcast on Tuesday March 23, 2010, click here to watch the archived video.