May 15, 2013 (Vol. 33, No. 10)

Proteomics has undergone the same existential choice, and self-examination, as other biological sciences with respect to automation. On the surface, automation helps laboratories cope with very high-throughput workflows.

Return on investment can be swift and convincing in an era where large testing laboratories seek to do more with less. Even medium-throughput labs can be convinced that the assay accuracy, consistency, and reduced human error (and its consequence—rework) that robotics bring make an even more compelling case for automation.

What about discovery-stage science labs operating at the lower echelons of throughput? Suparna Mundodi, Ph.D., global product manager at Rainin Instruments, offered a glimpse into what is in store for them at CHI’s “PepTalk” program on protein sciences. There, Dr. Mundodi presented the case for semi-automation, as exemplified by her company’s PureSpeed protein purification system.

PureSpeed uses a semi-automated pipetting protocol resident on the company’s E4 XLS electronic pipette, which processes up to 12 samples in parallel. But she said the stars of the show are Rainin’s PureSpeed tips, which purify microscale volumes of target protein, at very high concentration, in as little as 15 minutes. The pipette-fitting PureSpeed tips contain small volumes of standard chromatography resins concentrated in the tip region. Protein A, protein G, IMAC, and other formats are possible.

Rainin’s technologic competitors in ultra-small-scale protein purification are gravity chromatography and spin columns. Gravity entails very long sample prep times and is unsuitable for parallel processing. “You can’t do three or four samples simultaneously because you have to run between columns pouring buffer.”

The E4 XLS pipette draws and discharges a protein solution through the resin bed, cycling as much as needed to ensure adequate binding of protein to resin. After washing away impurities, the purified protein is eluted into a tiny volume suitable for such concentration-dependent analyses as surface plasmon resonance, electrophoresis gels, and ELISAs.

The official book on PureSpeed is it bridges the workflow gap between very low and medium-high throughput experiments. Actually, said Dr. Mundodi, the 12-channel system sits within the sweet spot for most early-stage research. “If you look at the protein purification market, 12 samples can be high throughput in many instances.” Her own market research, conducted before PureSpeed was developed, suggested that the average number of samples in PureSpeed’s target market was just under 10.

That tells only part of the story. The many benefits of automated liquid handling notwithstanding, it is often simpler and less costly to forgo automation, particularly in labs whose workflows are constantly changing. “Many of our customers, who may be doing expression screening on 40 to 50 clones, find the semi-automated system more straightforward, even faster, than fully automated systems.”

Since it is semi-automated (or semi-manual, depending on one’s perspective), PureSpeed requires no programming or extensive method development. With the inexorable drain of robotics and IT talent from many labs, semi-automation represents a welcome step forward.

“The best part is it’s a pipette, which makes people comfortable with using it,” said Dr. Mundodi.

Full Automation

The theme of accomplishing more with less recurs in most analytical work. One way to achieve this is through automation, but many managers cannot be convinced of the value proposition. From the laboratory’s perspective, ROI is directly related to freeing high-level scientific talent from repetitive tasks. This was the theme of a talk by PerkinElmer on the marriage of the Janus® BioTx Pro automated liquid-handling system with PerkinElmer’s LabChip® microfluidics-based analyzer. LabChip GxII takes classic SDS PAGE and automates it within a microfluidic platform.

Together, the two instruments provide automated, small-scale protein purification and characterization. The Janus liquid handler prepares samples based on filter plates, tip-based chromatography, or mini-columns. Users can then feed purified samples into the analysis instrument of their choice, for example the PerkinElmer LabChip GXII microfluidic-based analysis system.

Krystyna Hohenauer, director of biotherapeutics development, life science & technology at PerkinElmer, explained a clear trend toward biotherapeutics and the need for rapid analysis and characterization of putative products. “Related are biotech patent expirations and the development of biosimilars.”

There is much analytical work to be done, she said, “but companies are not hiring more analysts. The ones who remain need much more efficient ways to characterize and purify proteins. They want to analyze for key characteristics much earlier in the development process.”

The combination of sample prep and GXII provides scientists with the information to make better, more timely decisions, and free them from bench time to engage in more high-level data analysis.

The Janus can be integrated into the GXII to fully automate the workflow from sample prep through to analysis and characterization, or the systems can be set up to be completely separate. This allows the researcher to utilize the instruments to best fit their sample throughput and development needs.

According to PerkinElmer, its Janus liquid-handling workstation facilitates proteomics analysis.

Life Raft?

Membrane-bound biomolecules present numerous challenges for analysis and characterization. One family of membrane structures, the lipid raft microdomains, are attracting attention for their possible roles in disease and wellness.

Although their precise role is still controversial, current thinking holds that lipid rafts organize or compartmentalize glycolipid-protein interactions necessary for normal biological function. “One purpose is to concentrate cell-signaling molecules,” explained Susan Daniel, Ph.D., assistant professor of chemical and biological engineering at Cornell University.

If molecules required for a signaling event are held within the raft, they’re available to carry out that function. Likewise, if outside molecules are required, and the cell can control their partitioning into and out of the raft region, the cell has a means to control how signaling events occur.

Dr. Daniel developed a technique for purifying lipid raft resident molecules from nonraft species through a biological membrane-based platform that achieves 2D separations based on chemical affinity for a particular membrane composition. She likened the technique to traditional extraction except that instead of oil and water, materials partition into one or another flat membrane material in close contact. Ionic strength and pH are controlled as they would be in electrophoresis.

The membrane sheets (one raft, one nonraft) are positioned parallel, within the same plane and with a 2Dp (line) interface. “But instead of a 2D interface you have a line formed by two planar sheets butting up against one another.” When a mixture is loaded and forced through, components migrate into the region with which they enjoy the higher affinity.

The membrane sheets (one raft, one nonraft) are positioned parallel, within the same plane and with a 2Dp (line) interface. “But instead of a 2D interface you have a line formed by two planar sheets butting up against one another.” When a mixture is loaded and forced through, components migrate into the region with which they enjoy the higher affinity.

The cell membrane is chemically heterogeneous (left). Lipid rafts, enriched in cholesterol and exhibiting local lipid ordering (light gray), sequester certain kinds of membrane biomolecules (red); while other biomolecules (green) prefer a more disordered, phospholipid-rich lipid environment (dark gray). (Right) The natural partitioning preference of membrane biomolecules within a mixture (yellow) can be exploited to separate species that prefer raft-like regions from those that do not. This separation is carried out using a supported lipid bilayer platform that is patterned with two different lipid phases. [Cornell University]

Purifying PARP-Associated Proteins

Guy Poirier, Ph.D., Canada chair in targeted proteomics at Laval University, described a high-throughput affinity purification—actually a profiling of the poly(ADP-ribose)-related response to DNA damage signaling.

When DNA damage occurs inside cells, DNA-dependent poly(ADP-ribose) polymerases (PARPs) synthesize anionic poly(ADP-ribose) (pADPr) scaffolds that bind to several proteins, resulting in formation of pADPr-associated multiprotein complexes. These complexes were for many years believed to be covalent—similar to more common post-translational modifications such as phosphorylation and methylation. Subsequently, the noncovalent interactions have emerged as central in events downstream from DNA damage, such as cancer.

Professor Poirier uses a combination of affinity purification and mass spectrometry-based proteomics to isolate pADPr-associated complexes, and assess protein dynamics with respect to pADPr metabolism.

His approach involves a substrate trapping strategy that uses a mutant but inactive poly(ADP-ribose) glycohydrolase (PARG), as mutant can act as a physiologically selective “bait” for isolating specific pADPr-binding proteins. Along with more standard antibody-mediated affinity purification methods, Prof. Poirier employs a pADPr macrodomain affinity resin to recover pADPr-binding proteins and their complexes.

“Only one PARP was known before the human genome was sequenced,” Prof. Poirier explained. “Now we know of six major PARPs, and several homologs.” PARP activity can rise up to 100-fold during DNA damage, with concentrations in the micromolar range.

PARP activity has been implicated in several diseases, including cardiovascular, neurodegenerative diseases, and cancer.

PARP repairs single-strand breaks that occur with frightening regularity in dividing cells. If these breaks persist during cell replication (as they might during times of low PARP activity), the single-strand breaks become double-strand breaks and the cell dies. A PARP inhibitor, therefore, could sustain DNA damage in susceptible cells, causing tumor cells to die.

The key to designing a therapeutic PARP inhibitor, many believe, is to identify PARP-binding proteins. A decade ago around 60 such proteins were known. “Today we know more than 1,000,” Prof. Poirier said. Thanks to the sensitivity of new mass spectrometry instruments and the availability of peptide/protein databases, experimental timelines have shrunk exponentially. “We can now see more than 100 proteins during a one-hour run. That would have taken several years not too long ago.”

Protein Crystal Structures

Established in 2004, the private-public University of Toronto partnership, the Structural Genomics Consortium (SGC), has deposited more than 1,300 structures into its protein database. Susanne Gräslund, Ph.D., discussed various high-throughput methods to generate these proteins within a range of expression hosts. Since 2011, the organization has also been involved in a large-scale project to generate recombinant antibodies to epigenetics proteins.

With a U.K. charity as its base, SGC relays funds from public and private donors to member labs in Toronto and Oxford, U.K. The group also maintains satellite labs worldwide. Approximately half of SGC’s members are dues-paying industrial firms. “It’s important to maintain that public-private balance,” Dr. Gräslund commented.

SGC’s core mission is to solve crystal structures of proteins relevant to human health, which explains the strong pharma/biopharma connection.

As part of its mandate, SGC has begun investigating chemical probes, mostly small molecules, that bind to protein targets. “These are not necessarily drug candidates, as the leap from binding to drug is very large,” Dr. Gräslund explained. “Rather, researchers use these molecules as starting points for drug discovery.”

The source of this molecular largesse is chemical libraries, whose access is provided by member companies. SGC also maintains academic collaborations, for example with the University of Colorado. A potential drawback to commercial partners is the unique agreement by which SGC publishes results freely and openly.

The consortium’s third charge is making monoclonal antibodies using purified proteins as antigens, and subsequently selecting synthetic recombinant antibodies using phage-display technology.

Binding is determined through in vitro assays and cell-based assays. “We do a lot of co-crystallizations as well,” Dr. Gräslund said. Most of the proteins employed in this work have already been solved, in their native unbound state, by crystallography. “Small molecules that bind sit in a specific locus, altering the protein’s conformation relative to its unbound state. Conformational changes cause shifts in thermal melt temperature.”

SGC is open to additional collaborations, both academic and corporate, and has space to host visiting scientists with common interests in proteomics. The consortium also holds formal courses in protein purification and crystallization.

Previous articleRuiYi, CMC Collaborate to Speed mAb Development
Next articleFDA Approves Roche EGFR Mutation CDx for NSCLC