Among the most compelling presentations at SLAS2013 will be those that represent the "best in breed of new technologies and science—particularly those that are making a difference at a practical level," says Paul Taylor, M.S., chair of the SLAS2013 High-Throughput Technologies track. The following examples offer a glimpse of what's in store for conference participants in this evolving arena.
Shortly after biomedical engineer Anand Ramasubramanian, Ph.D., joined the faculty at The University of Texas at San Antonio (UTSA) in 2008, he gave a talk describing the high-throughput work he had done during his post-doctoral training. "We developed a chip that contained small matrices filled with cancer cells, and we used that for screening compounds instead of the standard 96-well plate," he explains.
Attending the talk was microbiologist José López-Ribot, Pharm.D./Ph.D., a pioneer in the area of Candida biofilms. After the talk, López-Ribot approached Ramasubramanian to discuss whether a similar chip might be developed to test antifungal compounds—and an interdisciplinary collaboration was born.
Candida albicans, the causative agent of Candidiasis, has emerged as an important health threat over the past decade, particularly for immunocompromised individuals who are at risk for opportunistic infections. The condition is often associated with biofilm formation on implanted medical devices such as indwelling catheters, making the fungus a leading cause of nosocomial infection in hospitals. Since these biofilms tend to be resistant to most antifungal drugs, as well as to host defenses, finding ways to rapidly screen for and identify candidate compounds has taken on some urgency, according to Ramasubramanian.
Screening for antifungal compounds with conventional 96-well plates requires large amounts of liquid reagent, which tends to be "messy to handle and can disturb the biofilm," says Ramasubramanian. Chip technology does away with those problems by encapsulating the biofilms in a matrix, where they can grow unencumbered. In its present format, the chip consists of a microarray of 768 distinct biofilms, each with a volume of about 50 nanoliters, on a single glass microscope slide. The platform "cuts reagent use and analysis time, minimizes labor-intensive steps and dramatically reduces assay costs" compared with conventional screening methods, he notes.
In the real world, people often are infected by multiple strains of fungi and/or bacteria. Therefore, Ramasubramanian and his colleagues recently started working with multiple species biofilms, both fungal and bacterial—and this will be a major focus of a presentation by Ramasubramanian and López-Ribot at SLAS2013. The team also is looking at potential synergies among existing drugs to create new forms of combination therapy. "For example, since Candida is also a problem in cancer patients undergoing chemotherapy, we are looking at the efficacy of treating with both traditional cancer drugs and traditional antifungals, to see if we can identify any potential synergies." Because many infections are associated with combat-related trauma, the team also is collaborating with the U.S. Army Institute of Surgical Research in San Antonio to determine the best therapies for wound infections.
Looking ahead, Ramasubramanian is eager to collaborate with groups willing to test biofilm chip technology to help develop therapies for infectious/neglected diseases.
"One reason I'm looking forward to SLAS2013 is to see who might be open to working with us in these areas," Ramasubramanian says. "I am a trained biomedical engineer and López-Ribot is a trained mycologist. If we had never talked to each other, I would not have known about the problems with the microplate platform and he would not have seen a potential solution."
Is nothing sacred? When it comes to smart technology and science, apparently not. Now, even the venerable Petri dish, that shallow vessel used to culture bacteria and other microorganisms since the 1800s, has undergone a major transformation. Changhuei Yang, Ph.D., professor of electrical engineering and bioengineering at California Institute of Technology, and colleagues have developed a "smart" version of the cell-culture dishes, dubbed ePetri. Yang will provide details on the invention, its benefits and potential applications in the laboratory and in the clinic at SLAS2013 as part of the High-Throughput Technologies track.
"I'll be driving home the point that Petri dish experiments have not really changed over the years, and that now is the time to look at ways to do these experiments more intelligently, with fewer manpower resources and significant cost savings," Yang says. The time is right, he explains, because "cell phone cameras have become almost ubiquitous, which really drives down the cost of the 2D sensor chips that form the heart of those cameras. We capitalized on the fact that those chips are highly sophisticated, yet cheap and readily available—and we were able to go the extra mile with them by getting their resolution to be comparable to what you get from a conventional microscope."
The ePetri dish system has three main components: the image sensor, a smartphone and Legos. The system sits inside an incubator; cells are grown directly on the sensor and a smartphone sitting (supported by Legos) above the sensor continuously scans the culture with a spot of light that moves back and forth. The sensor sends images via a cable to a laptop outside the incubator. "Because the cell cultures are scanned continuously, the resulting images are like a movie, with light coming in at different angles," explains Yang. "We can process those images and distill them to a high resolution."
With the new system, scientists don't have to take dishes in and out of an incubator or use a microscope to assess culture results, Yang continues. They also get a complete view of the cell culture. "There are many situations in which cells run around all over a Petri dish, and if you're looking in a conventional microscope and cells run out of your field of view, you're out of luck. With our system, you can see everything, and also focus in on specific areas of interest."
Yang and his team recently reported a longitudinal study of Euglena gracilis cultured in an ePetri platform, which demonstrated the system's ability to continually monitor motile organisms over time and in high resolution. The study provided a proof of principle that the miniaturized, automated culture monitoring platform can "greatly improve" experiments with such organisms, Yang says.
The team also has started a spinoff company to test various applications over the coming year. "As we improve the system—for example, by adding fluorescence detection capability— we are actively looking for other possible uses," Yang says. "By the time I give my talk at SLAS2013, I expect that we will have expanded our pool of testers and be ready to welcome others who are interested in testing applications to contact us." Although they currently are focusing on bioscience applications, "we hope that eventually we can drive the cost down to the point where our system will serve as a direct replacement for the conventional Petri dish used in the diagnostic market," he adds.
Smaller is better—especially for Andrew deMello, Ph.D., professor of biochemical engineering at ETH Zurich and co-founder (with John deMello and Donal Bradley) of Molecular Vision, an Imperial College London spinoff company that is developing low-cost diagnostic devices for use in physicians' offices and in the home by combining microfluidics with semiconducting polymer technology.
"It was really a fluke that I got into the field of microfluidics," deMello says. "As a postdoctoral student, I wanted to work on Raman spectroscopy with Richard Mathies at the University of California, Berkeley. By chance, he had just started developing microfluidic systems for analyzing and amplifying DNA—and that's how I got into it." His microfluidics experience laid the groundwork for deMello's current interest—microdroplets, which he sees as "defining new experimental paradigms for high-throughput chemistry and biology because of their unique characteristics and benefits."
Microdroplets can be made "very easily and quickly," deMello observes. "We don't have to talk to Intel or IBM to implement relatively simple micromachining methods. We can make systems that create very small droplets—about 1,000 per second—with volumes between about 10 picoliters and 10 nanoliters. Each droplet is surrounded by oil, and we can dose each one with different reagents to create a different chemical or biological environment for high-throughput experimentation."
Working with droplets also helps solve a longstanding problem in microfluidics—namely, that when small volumes are moved through small channels, "everything touches the surface," deMello explains. "It's not that important when you're working on a large scale, because the surface is relatively insignificant compared with a macro-scale volume. But on a microscale, "if you were making a device to do a diagnostic assay, for example, every time you push fluids through, you contaminate the device because the reagents interact with the surface. With microdroplets, we can get around the contamination issue because they never touch the surface to begin with."
Of course, working with microdroplets presents challenges, as well. "The most difficult challenge is that if I am creating a thousand different experiments in a thousand different droplets, and I have 10 picoliters in each droplet, and they're moving with relatively high linear velocities, I need highly sensitive and very rapid detection methods so I can probe each of them and pull out the information," deMello says. The team has been using fluorescence for this purpose, but they are now moving towards label-free methods, including vibrational spectroscopy, surface-enhanced Raman and infrared probing of droplets."
In addition to providing details on the benefits and challenges of microdroplets for high-throughput screening, deMello will speak at SLAS2013 as part of the High-Throughput Technologies track about some of the recent work with which his group has been involved. These include developing diagnostic point-of-care devices that can be used with saliva, urine or blood samples, as well as more exotic applications involving the processing of live organisms. For example, working with nematodes as a model organism, they are developing tools that might help evolutionary and developmental biologists unravel some of the mysteries surrounding the aging process.
deMello is looking to SLAS2013 as an opportunity to find collaborators in other areas of investigation, as well. "One of the very best things that happens when you give talks is sparking the interest of someone from a completely different field, who then asks for your help," he says. "We got interested in the aging problem when an evolutionary biologist who works with the C.elegans explained that they use tiny platinum tools like tweezers to pick up the worms and move them from one agar plate to the next. He said, ‘maybe we can use microfluidics so we no longer have to do this.' That was the catalyst for the work we're doing now."
Jeffrey Romine, Ph.D., and his colleagues at Bristol-Myers Squibb were in pursuit of a hepatitis C virus treatment for the close to 170 million people worldwide who are chronically infected with the virus. Rather than taking the traditional HTS route of doing in vitro screening for activity against a known protein, the team used a phenotypic approach—that is, a whole cell screen using a sub-genomic replicon, which allowed them to screen against a broader range of potential targets. The result: "We found a single compound, BMS858, that was discovered to have inhibitory activity against a novel target, the NS5A protein, a result that generated a lot of excitement because we knew we were onto something new," Romine says.
But as Romine, a chemist, began to do the requisite work of purifying, retesting and profiling the compound, something strange was going on. "We observed BMS-858 had oxidized and rearranged into something else, and when we determined the structure of the new compound, we found out that it didn't have any activity against HCV," he says. The structural rearrangement happened under other conditions, too—including incubation in cell media—yet the ability to inhibit HCV was very good, even when there was no hint of the parent compound left in cell media. "This was puzzling in light of our structure-activity relationships data that appeared normal," Romine continues. "For example, we could see that only one amino acid of a particular stereochemical configuration had activity, and that if we took that away, we lost activity. And if we made variations in other parts of the molecule the potency either went up or down as expected.
"So the next question was, ‘what is going on in the cell media?' We turned to a natural products isolation scientist at BMS, who solved the problem by isolating the active component from media. It turned out that BMS-858, a thiazolidinone, had dimerized; the entire molecule was intact, but it had coupled with itself. The HCV inhibitory activity was traced to this dimer," Romine explains. That could have been the end of the line for the program, he says, except that the director and co-chair of chemistry asked the question, "What can we deduce from this?" That question led to an idea of a simplified molecule that incorporated the important structural characteristics of the dimer. The ensuing synthesis gave rise to a compound with potency comparable to the dimer and that was stable in cell media. A full-phase drug optimization program was launched, which ultimately yielded a first-in-class drug candidate, BMS-790052, or Daclatasvir, which is now in Phase III clinical trials against HCV.
"In the real world, we're all under pressure to reach objectives. We metricize so we can measure progress and there's always that overhanging burden of uncertainty as to when, at a certain point, management will say, ‘that's enough,'" Romine observes. "Fortunately, in this case, the team had the tenacity to say, ‘We can't walk away from this because the virology profile is too strong; we've got potency against a novel target.' So the message is, ‘don't give up when you have activity—sometimes it pays to pursue problems until you solve them.'" Romine will describe the discovery adventure in detail at his High-Throughput Technologies track presentation at SLAS2013.
It's no secret that HTS and related drug-discovery technologies are undergoing a revolution that affects both back ends and end products. Ideally, these advances could be shared across companies and organizations to stimulate further collaborations and innovation. Part of what's standing in the way, according to Katheryn Shea, president of ISBER [International Society for Biological and Environmental Repositories] and vice president, BioServices Operations at SeraCare Life Sciences, is "lack of standardization, particularly in sample preparation procedures." The lack "makes it difficult to automate processes in a cost-effective manner" she says. The result: difficulty duplicating new findings for sensitive biomarkers, which slows discovery, related research and clinical applications.
Shea takes the position that automating processes could promote standardization by eliminating site-to-site differences and reducing the "human element" of sample processing in each site. But there are many hurdles to overcome before such a solution is feasible, or justifiable from a financial perspective, she acknowledges.
"For example, measuring cytokine levels in plasma would typically require a collection tube with a citrate preservative of a certain volume. To measure viral load, a tube with an EDTA preservative that could have a different volume would be used. The different tube types often have different dimensions. Differences in the input materials can make it more difficult to automate," Shea explains.
"Also, the processing methods are often different," Shea continues. "Samples from multiple protocols are received each day and each protocol can require that different components of the blood be saved and different processing protocols be used. Centrifuge speed, time, temperature and reagent variation is not uncommon. Another complicating factor is that everyone's blood is different and so the volume of the blood component you are isolating may be different from one donor sample to the next."
In her High-Throughput Technologies track presentation at SLAS2013, Shea will put forth "more questions than solutions," she admits. SLAS has a strategic alliance with ISBER. ISBER provides an international forum to address technical, legal, ethical and managerial issues related to repositories. She notes, "collaborations between automation experts and those in the very manual world of repositories is exactly what is needed. By working together we will be able to come up with solutions that allow for high quality sample preparations that allow researchers to measure the biomarkers of interest and not artifacts of sample processing variability."
December 10, 2012