The common need in cancer research and pharmaceutical drugdevelopment is to reveal the configurations of active signaling pathways indiseased tissues, to support target validation, trial design, patientselection, response assessment and if trials are successful, the diagnosticcomponent of theranostics. Importantly, the predictive power of measurements ofsignaling protein expression depends on the precision and accuracy of tissueanalysis tools.
For example, many techniques deployed today, such as thosebased on microarray detection, or analysis of sample lysates, provide data thatare in fact averages from volumes of tissue, including many cells not ofinterest. These methods blur out key proteomic information that reside at thecellular level, and relate to the signaling states of individual cells.
Role of signal transduction pathways in cancer
During the course of tumor progression, cancer cells acquirea number of characteristic alterations. These include the capacity toproliferate independently of exogenous growth-promoting or growth-inhibitorysignals, the tendency to invade surrounding tissues and metastasize to distantsites, the penchant for eliciting an angiogenic response and the ability toevade mechanisms that limit cell proliferation, such as inflammatory response,apoptosis and replicative senescence. These properties reflect alterations inkey cellular signaling pathways that in normal cells control cell proliferation,motility, and survival.
Many of the proteins currently under investigation aspossible targets for cancer therapy are signaling proteins that are componentsof these pathways. The nature of these signaling pathways and their roles intumorigenesis are the subject of intense study by pharmaceutical companies,motivated by the hope that progress in understanding these signaling pathwayswill accelerate drug development. It is a complex research task to identifyrelevant pathways, understanding them and demonstrating correlation withoutcome. An additional level of complexity arises from the fact that it isoften the interrelationship between pathway proteins and their localizationthat help characterize the pathway, rather than the mere presence of a protein.
Example: Detecting phospho-epitopes of AKT, ERK and S6
These three pathway markers are widely studied and play avital role in cancer pathogenesis. In this particular example, the goal is todetect the activation of PI3K/AKT, RAS, and MEK signaling pathways.
AKT has recently been found to play a paradoxical role: onone hand, it increases cancer cells' survival capability, while on the otherhand, it blocks their motility and invasion abilities, thereby preventingcancer from spreading [1]. It had been presumed that one could promote cancercell death by inhibiting AKT that controls the synthesis of proteins involvedin proliferation. Yet now, with this added complexity, the role of AKT must beunderstood further, so as not to promote metastases by inhibiting AKTexpression.
Activation of the MEK pathway up-regulates ERK proteinlevels, promoting cell division. This pathway is often up-regulated in humantumors and is thought to fulfill multiple roles in the acquisition of a complexmalignant phenotype. Accordingly, a specific blockade of the MEK pathway isexpected to result in not only an anti-proliferative effect, but also inanti-metastatic and anti-angiogenic effects in tumor cells.
Recently, potent small-molecule inhibitors targetingcomponents of the MEK pathway have been developed. Among them, BAY 43-9006 (Rafinhibitor), and PD184352, PD0325901 and ARRY-142886 (MEK1/2 inhibitors) havereached the clinical trial stage. The combination of MEK pathway inhibitors andconventional anticancer drugs might provide an excellent basis for thedevelopment of new chemotherapeutic strategies against cancer.
Finally, s6 is a ribosomal protein involved in translationof mRNAs. It is thought to play an important role in controlling cell growthand proliferation.
Automated, multiplexed tissue cytometry
Detecting pathway markers using conventional histology orimmunofluorescence is a challenge, given the need to observe many markerssimultaneously (i.e., to multiplex) inorder to gain a full understanding of the pathways involved and the phenotypes.Conversely, conventional multiplexing techniques, such as microarrays or flowcytometry, fail to provide the contextual information needed to confirmintracellular localization; also a requirement in order to confirm pathwaystate. What is needed is simultaneous measurement of multiple proteins, on aper-cell basis, set in the context of the original anatomy.
New platform technologies now offer us the opportunity toaccess this level of information, by utilizing an effective, practical andreliable platform for cytometric analysis of intact tissue sections. This canbe conceptualized as "tissue cytometry." The platform supports preclinical andclinical studies through the integration of multiplexed immunohistochemical(IHC) or immunofluorescent (IF) labeling strategies, robotic slide handling,and automated multispectral image acquisition and analysis. Multispectralimaging systems and advanced image analysis software together provide the idealplatform for this application.
The ideal imaging platform integrates: a) easy-to-implementmultiplexed staining protocols; b) an automated slide analysis system that canisolate marker signals from one another and from autofluorescence; and c)pattern recognition-based image analysis software for automatically segmentingimages and extracting quantitative data from cells of interest.
Multispectral imaging and automated image analysisaccelerates preclinical and clinical studies
Quantitative, independent and specific multi-label protocolshave been developed that in conjunction with easy-to-use multispectral imagingsystems and advanced learn-by-example software, can greatly accelerate clinicaland preclinical studies [2].
For example, today, approximately one-third of small-moleculekinase inhibitors in development or trial target pathways are associated withEGFR activation. Analysis of EGFR activation in tumor xenographs is typicallydone by immunohistochemical staining of tissue sections for phosphor-epitopesof EGFR. Samples are analyzed by eye by pathologists, either under themicroscope or on the computer screen as digital slides.
Typically, pathologists can process slides at an averagerate of 100 samples per day. A study of hundreds of samples takes days or weeks.On the other hand, if samples are stained with multiple color protocols thathelp guide image analysis and provide internal controls, such as a stain fortotal EGFR, slides can be analyzed automatically with image analysis software.Such software can then present segmentation results and associated markerintensity scores to pathologists for review, modification if necessary, andfinal approval.
In benchmark studies, results have shown that an analysisprocess that takes many days, at 100 slides per day, can be reduced to hours,at a rate of 200 to 300 slides per hour. The pathologists remain central to theprocess by training the image analysis algorithms to identify important tissueareas, and as a final quality control gate on image analysis results.
In a recent study performed at one pharmaceutical company, atrained image analysis solution accurately segmented tissue into regions ofinterest for 98 percent of samples in a large, 3,000-sample study.
There is another benefit to this approach, in addition toincreased productivity and shorter study durations. Data is more consistent,since stain intensity scores are based on measured signal levels from a digitalcamera instead of human visual perception, which can vary over time based onchanging ambient environments and is not well suited to capture the non-linearsignal levels inherent in chromogenic absorption.
Pharmaceutical companies are motivated by the hope thatprogress in understanding signaling pathway activity will accelerate drug development.The tasks of revealing activated pathways, understanding theirinterrelationships and determining correlation with outcome are challenging.The complexity inherent in signaling pathway activity can only be elucidated byrevealing key marker localization and distribution within tumor cells, ratherthan the mere presence of a protein independent of morphological context.
Combining multispectral imagingwith advanced image analysis tools to perform tissue cytometry rapidly and on alarge scale and using many markers at once has proven to enable a betterunderstanding of the mechanism of disease and potentially better, more preciseavenues of treatment.
Clifford C. Hoyt,a founder of Cambridge Research & Instrumentation Inc. (CRi) in Woburn,Mass., joined the company in 1987 as a staff scientist. He has played a centralrole in the development of many of CRi's core technologies, including theliquid crystal tunable filter for multispectral and polarized light imaging andthe integration of these core technologies into analytical instruments forapplications such as in vitro fertilization, high-throughput drug screening,stem cell research, in vivo small-animal imaging, live-cell biology andtissue-based immunohistochemical analysis. He holds 12 patents, has numerouspatents pending and is the author or co-author of 20 publications in technicaljournals.
Darren Leeis vice president of marketing at CRi. He has more than 20 years of experiencein marketing management, business development and engineering in the lifesciences and clinical diagnostics industries. He has held senior-levelmanagement positions at Primera Biosystems, a molecular diagnostics company,and Decision Biomarkers, a life-science systems developer. He has also servedin senior management positions at PerkinElmer and Packard Bioscience.
References:
1. Yoeli-Lerner, Yiu, Rabinovitz, Erhardt, Jauliac andToker: "Akt blocks breast cancer cell motility and invasion through thetranscription factor NFAT." Molecular Cell,November 2005.
2. Levenson, Fornari and Loda: "Multispectral imaging andpathology: Seeing and doing more." Expert Opinion on Medical Diagnostics, 2008.
