August 1, 2008 (Vol. 28, No. 14)

David Daniels, Ph.D.

Cancer Detection, Assay Development, Early Drug Development, and Diagnostics Top List

The latest techniques used to advance biomarker research were among the topics discussed at CHI’s “Biomarker World Congress” held in Philadelphia recently. Several of the presenters at the meeting spoke to GEN about their work using biomarkers for cancer detection, assay development, early drug development, clinical development, and diagnostics.

Keld Sorensen, Ph.D., director of R&D at Luminex, is a big proponent of quantitation in array-based systems. “There are three different methods for the study of biomarkers in use today,” said Dr. Sorensen, “each has its benefits. qPCR is great for multiplexed analysis of a few markers in the sample, planar arrays are good for looking at more markers but have a lower throughput, and bead-based arrays enable the study of even more markers and at a much higher throughput. The key is to provide standard curve controls within the experimental workflow to allow for quantitative assessment.”

The current focus right now at Luminex is miRNA, single-stranded RNA molecules of about 21–23 nucleotides in length that function to downregulate gene expression either by impacting RNA processing or inhibiting translation. As a biomarker, the profile of miRNAs in the cell can be used to differentiate between normal and cancer cells.

The key issue for monitoring miRNAs is robust sample processing and avoidance of miRNA amplification variation from sample to sample. Luminex has built standard curve controls into its kits to provide a reference for each reaction. Luminex also employs LNA-based probes (Exiqon) to provide for Tm-optimization of the reaction conditions. By virtue of this approach, all target amplifications can be run at 72ºC instead of amplification at the natural Tm of the target. The latter can vary over a broad range, from 40º to 70ºC for the 22-base sequences. By normalizing the reaction conditions, assay relevance is maintained and the end-user can measure multiple targets in the same assay, said Dr. Sorenson.

To put these tools in the hands of research scientists, Luminex has launched the FlexmiR™ Select kits. These kits allow the end-user to pick from more than 320 different miRNAs to develop their own assay for miRNA to differentiate normal from cancer cells in whatever system they are studying, according to Dr. Sorensen.

The FlexmiR kits also come with nine control miRNAs per well, which can be used for standard curve generation. “We recommend testing a large number of miRNA targets with a few samples to identify the best 20–30 targets that differentiate cancer from normal cells in the system being studied,” added Dr. Sorensen. “Then, screen the large number of samples with the reduced set of targets in the assay.”

The FlexmiR kits are currently for research use only, but the bead-based xMAP® Technology multiplexing platform has received 510K approval for use in HLA-typing, respiratory virus testing, CF-testin, and other applications.

Assay Development

At Xceed Molecular, David Englert, Ph.D., vp of technology development, and his team have developed a new flow-through chip system, the Ziplex Microarray Workstation, which allows end users to process biotin-labeled cRNA samples to assay their biomarkers of interest. Ziplex enables fast processing of the samples, including hybridization, washing, and detection, said Dr. Englert, who added that it has been designed to assay up to 20–120 biomarkers simultaneously in up to eight samples.

The utility of this approach is based on a large body of research that points to the association of gene-expression profiles with disease states and clinical outcomes. While discovery work is typically performed on global expression arrays, clinically informative profiles involve a relatively small number of target genes. The research effort at Xceed Molecular is focused on refining and validating gene-expression biomarker signatures to develop clinically informative, focused assays.

The Ziplex workstation holds two square-well 96-well microplates for processing eight samples at a time. Xceed provides TipChips containing up to about 400 features (120–140 probes with replicates and controls), as well as the hybridization and wash buffers and the chemilluminescent detection reagents. Once the biotin-labeled cRNA samples are loaded onto the plate, the time to final results is less than three hours, which translates to a throughput of 24 samples per day, Dr. Englert noted. Data analysis is automatic. The output from the experimental run is in the form of numeric tables with QC data.

Xceed Molecular has established a fee-for-service business that provides for the manufacture of custom chips for the biomarkers of interest and processing of samples on the Ziplex Workstation. “All we need is the transcript sequences of interest,” indicated Dr. Englert. “We then manufacture and spot the corresponding capture probes on the flow-through chips called TipChips and process the samples.”

Early Drug Development

Biotrin International is focused on the development and implementation of cell-specific biomarkers. These preformed cytosolic proteins with a defined distribution are rapidly released in response to cellular injury. Detection of these proteins enables injury to be localized to distinct cell types in distinct organs. In a case study presented at the meeting, Martin Shaw, senior product manager, provided data demonstrating that the assay of specific biomarkers in urine provided early indications of injury to defined cell types and also gave information as to pathological changes.

“Traditionally in the clinic and drug development, the most studied renal parameter is serum creatinine. But it is a late biomarker whose levels reflect a balance between injury and regeneration,” Martin said. “That is, 50 percent of renal function may be lost before serum creatinine changes significantly enough to be detected and the decision for treatment is made. That’s not a particularly responsive biomarker for early disease state or fluctuation in response to therapeutics.” In drug development, the use of such insensitive biomarkers can lead to delays in detecting and excluding potentially nephrotoxic compounds, leading to increased cost and delay.

The biomarkers developed at Biotrin to monitor acute renal injury or response to therapeutic intervention include a-GST (alpha-glutathione S-transferase) from the proximal convoluted tubule and p-GST from the distal convoluted tubule. These markers are readily detected in urine, an accessible, noninvasive sample source, and can provide a total picture of renal injury, Shaw said.

In the complementary field of drug development, Biotrin has similar biomarkers available for rat studies plus RPA-1 (renal papillary antigen 1), a urinary biomarker for collecting duct injury (renal papillary necrosis). a-GST and RPA-1 are being evaluated as part of the International Life Sciences Institute consortium on the qualification of biomarkers and will soon be submitted to the FDA for acceptance.

For the discovery of new biomarkers, Biotrin uses a technique known as Histomics®. Specifically, the company raises antibodies against rat and human urine samples from animals, or subjects, with defined forms of renal injury. These antibodies are then fluorescently tagged and used to screen kidney sections for new cell-specific markers. Interesting antibodies are then used to develop immunoassays.

Autoantibodies

At Protagen Jens Beator, Ph.D., protein biochips director, and his team have developed the UNIarray program to systematically look for autoantibodies in patient sera.

“Autoantibodies are constitutive and dynamic components of the immune system,” Dr. Beator explained. “And in good alignment with the definition of a biomarker, they can change specifically with the development of diseases as it is already established for several autoantibodies. Further, autoantibodies are stable in serum or plasma for years, which makes both retro- and prospective studies easily possible.”

The development goal at Protagen is to create indication-specific biochips for clinical screening projects containing panels of (e.g., 20–200) diagnostic biomarkers that differentiate patient populations. The case study presented at the “Biomarker World Congress” highlighted the profile of autoantibodies that were developed for multiple sclerosis (MS) to differentiate relapsing/remitting MS from primary progressive MS and secondary progressive MS. The company defined a panel that shows promise to differentiate subgroups of therapeutic responders and nonresponders.

For development of the MS biochip, Protagen started the project by screening a human recombinant protein-expression library of >10,000 different proteins. Candidate biomarkers were confirmed using an antigen-verification biochip, which is a protein biochip printed with greater than 330 affinity-purified biomarker candidates for the disease state. The company also printed 1,968 other affinity-purified recombinant human proteins, including autoantigens from previous studies. Human serum samples from some 100 diseased and 100 healthy patients were analyzed with these chips to verify disease-specific antigens. To enable the determination of autoantibody titers in the test samples, a serial dilution of an IgG calibrator is spotted on the chip for standardization.

The result of this work is the first biochip prototype for serum-based diagnosis and therapeutic monitoring of multiple sclerosis, according to Dr. Beator.

The focus of in vivo toxicology studies at Merck Research Labs is on toxicity biomarker endpoints as defined as necrosis and degeneration, according to Katerina Vlasakova, research associate in the systems toxicology and safety assessment laboratories. These differ from other ADME endpoints in that they directly correlate with and diagnose an adverse event. Specifically, the scientists at Merck have developed multiplexed assays that target tissue toxicity biomarkers for kidney, heart, skeletal muscle, liver, and acute phase proteins.

“We work with urine for determining kidney toxicity by measuring clusterin, albumin, and Kim-1 levels. We use plasma samples for measuring toxicity in the other organs, looking at cardiac Troponin I, skeletal Troponin I, A2M, AGP, GST-a and other biomarkers,” shared Vlasakova. “We work with accessible biomarkers from plasma and urine sources. If there was an organ injury, those biomarkers will be detectable in plasma or urine.”

With the use of multiplex analysis the Merck scientists can monitor for all target organ toxicities in case the source of toxicity is unknown or just focus on one panel (e.g., kidney) if kidney toxicity is expected.

Taking a multiplex assay approach is key so that multiple endpoints can be obtained from one assay in a given sample. It also maximizes information capture where samples are limiting. Proteins have a different half-life so having multiple markers ensures that toxicity will be detected even when timing of sampling was not optimal for some.

As with the biomarkers used by Biotrin, these Tox biomarkers are tissue-specific and not compound-related, so that they are applicable for general use in all compound screening.

Merck Research Labs uses a Meso Scale Discovery Sector Imager 6000. The Meso Scale technology is similar to ELISA in that it uses antibodies for analyte capture and detection. It differs in several ways, though. First, the capture antibodies are adsorbed to a carbon surface instead of plastic used for ELISA. This enables a “denser” coverage of capture Ab. Second, detection antibodies are conjugated to an electrochemiluminescent label that emits light when electrochemically stimulated. This method of detection is more sensitive and results in wider assay detection range than can be achieved with ELISA, according to the company. The plate contains electrodes, and a CCD camera that measures luminescence. Individual carbon spots within a well allow for multiplexing assays.

The general protocol for an MSD assay is similar to an ELISA protocol; samples are diluted and pipetted on a blocked plate. Typically there are two 1–2 hr incubations (first with sample and second with detection antibody) with a wash after each step. After the last wash, Read buffer is added, and the plate is read on the Sector Imager 6000. Throughput is limited at the first step where individual samples have to be handled and diluted. The plate read itself, only takes about 1–2 minutes.

The main advantage is multiplexing; Meso Scale has commercialized a kidney toxicity 6-plex. In a 96-well plate format a Merck scientist can run 40 samples in duplicate, in addition to a standard curve. At the end of the day, data output in the form of light units can be captured from 240 measurements in a single plate. The use of a standard curve enables the conversion from light units to concentration.

“We are always looking to optimize the current antibody assays and searching literature for new promising biomakers,” noted Vlasakova. “The biomarkers we are developing tend to be more sensitive and more specific than classical biomarkers of tissue toxicity.”

Clinical Development and Diagnostics

The conventional approach for biomarker discovery using protein arrays is to spot known capture antibodies onto the surface of the array and then add the biological test sample to the array to look for proteins in the sample that is bound. At Theranostics Health, they take an alternative approach, they use a reverse-phase protein microarray wherein the patient sample is arrayed on the platform and probed with known, fully validated antibodies for the binding reaction.

According to Gabriela Lavezzari, Ph.D., laboratory director, Theranostics Health has developed a number of solutions to address the challenges of obtaining good results on microarrays. There is extreme variability at the time of sample collection that can dramatically impact the outcome of the analysis. The two key problems with sample collection are tissue heterogeneity and sample preservation.”

To solve the problem of tissue heterogeneity, Theranostics Health employs laser capture microdissection to isolate pure cell populations. This is particularly important for the isolation of cancer cells from surrounding normal stromal material.

Theranostics Health has also developed a room-temperature fixative that can be used instead of formalin fixation or sample freezing. The preservative has been shown to prevent both false positive and false negative results, and importantly, it allows for the capture of charged phosphoproteins and maintains the cellular morphology for pathology and immunohistochemical analysis, Dr. Lavezzari said.

The platform was designed to enable the detection of biomarkers in patient samples without bias. While 100-120 samples can be loaded onto a slide for testing with a single detection antibody, up to 1000 samples can be printed with smaller pin arrayers. Specifically, the patient sample lysate is mixed with low and high controls and a calibrator protein and then plated in triplicate in a serial dilution.

One class of detection antibody and detection chemistry are added to the arrayed samples to look for specific binding. The detection antibodies are chosen based on customer request, from a portfolio of greater than 290 fully validated antibodies obtained from commercial sources. The reverse-phase microarray platform provides a nonsubjective quantitative protein immunoassay on the biopsied samples, Dr. Lavezzari noted.

Contract Research Organizations

Pronota’s mission is to develop and commercialize diagnostic tests based on protein biomarkers it discovers using its high-performance protein biomarker discovery and verification technologies, explained Huw Davies, Ph.D., director of business development. The company also provides collaborative proteomic R&D services.

“We are addressing the two biggest challenges in proteomics today,” indicated Dr. Davies. “First, we have developed an ability to consistently mine for novel low-abundance proteins in complex clinical samples including plasma. And second, we have developed a platform for rapidly screening biomarker candidates arising from discovery programs in extended sets of clinical samples to prove clinical utility and performance.” These two challenges of protein biomarker development are addressed by Pronota’s MASStermind® and MASSterclass™ platforms respectively.

The MASStermind discovery platform was designed to be an effective miner of the deep proteome. It is also capable of identifying, not only intact proteins, but also the proteins that are the result of in vivo digestion of the parent protein. The technology is capable of reproducible unbiased discovery and quantitation of protein biomarkers down to the single digit to sub ng/mL level in blood, Dr. Davies explained.

The MASSterclass verification platform takes biomarker candidate hits generated from MASStermind or other sources and verifies their performance in new sample sets. Assays can be developed at low cost and within about six weeks (compared to nine to twelve months for an ELISA). It has a sensitivity level comparable to MASStermind. It means that the company can take larger biomarker hit lists into the verification stage thereby increasing the chances of discovering true biomarkers in each project.

NextGen Sciences just launched its protein biomarker services, biomarkerexpress™, which was designed to cover a range of rapid, robust, and cost-effective biomarker services including discovery, assay development, and monitoring from as little as 2 µL of sample, said Michael Pisano, CEO.

Because the technology is based on multiple reaction monitoring (MRM) using mass spec, the method provides the fastest assay development service on the market today, he reported. The assays do not require an affinity reagent for detection. It takes about four weeks to develop a single 20 minute assay for monitoring up to 30 proteins.

NextGen Sciences can currently monitor protein biomarkers in a range of samples including biological fluids (e.g., plasma, CSF, and urine), tissues, FFPE tissue, and cell lines. It is also building a knowledgebase library from its internal development activity. Currently the library contains information on about 400 proteins in plasma, 850 proteins in CSF, and 900 proteins in urine.

Previous articleBristol-Myers Squibb Bids $4.5B for ImClone
Next articleCary Institute of Ecosystem Studies Is Recipient of $750K EPA Award