Past Seminar Series

Cell motility is fundamental to vertebrate development, cellular immunity, maintenance of animal tissues, disease progression, the survival of most single cell organisms and metastasis. Understanding how cells locomote must include 3D descriptions of behavior and dynamic architecture. To accomplish this, we have developed the computer-assisted 3D dynamic image analysis system, 3D-DIAS, which reconstructs the surface, nucleus and pseudopods of living, crawling cells, and tracks vesicle behavior in 3D. Using differential interference contrast microscopy to visualize the cell surface, nuclear membrane, and pseudopodial regions, and a near-real time laser scanning confocal microscope to visualize stained vesicles in living cells, 30 optical sections are obtained in a one second period through the z-axis of a cell, and this procedure is repeated every second. The optical sections are digitized into the 3D-DIAS program, which then image-processes each section, outlines edges and reconstructs the living cell as a faceted cell surface that encases the nucleus, pseudopods and vesicles. Over 100 parameters of motility and dynamic morphology are then automatically computed every second. The audience will be supplied with 3D red and blue glasses, and provided with a computer-assisted demonstration of 3D reconstruction, and the 3D dynamics of the cell surface, nucleus, F-actin enriched pseudopods and vesicles of live, crawling cells searching for and responding to gradients of chemoattractant. Dynamic 3D reconstructions will also be presented of monstrous HIV-induced T cell syncytia burrowing through endothelium, and neoplastic cells with altered cell behavior. Finally, new technologies now under development at the Keck Facility will be reviewed.

Sequenom has developed the technology to scan an array of samples in a chip format by Matrix Assisted Laser Desorption Ionization (MALDI) time of flight (TOF) Mass Spectrometry (MS). Current instrumentation uses samples that are 200 microns square. About 6 fmol of DNA in 6 nl is required; proteins and other analytes require far less material. Mass spectra can be acquired automatically at 3.5 seconds per sample. Ten 384-sample chips can be processed simultaneously. DNA analysis by Sequenom's MassArray™ technology covers a gamut of applications from sequencing and allele detection to mutation finding and gene expression analysis. Allele determination is the most well developed application and the one likely to see immediate large-scale application in cancer biology. In MS DNA sequencing, the high resolution allows unequivocal detection of heterozygotes and in some cases even allows the phase of compound heterozygotes to be called. The overwhelming advantage of MS-based genotyping is its accuracy. Trace amounts of alleles can also be analyzed by MS. This will greatly facilitate the analysis of heterogeneous tumor samples. The tiny size of the samples needed for MALDI scanning of chips poses a challenge for current methods of biochemical sample preparation. At present there are no robust techniques for preparing most kinds of biological samples on the nanoliter scale. When typical microliter preparation methods are used, 99.9% of the sample is wasted. Progress in developing new nanoliter sample preparation methods will be described.

Most pathogenic processes are based at the molecular level on abnormalities in protein expression, level, location, structure or modification. Improvements in patient management and treatment rely on the discovery of more selective and relevant protein targets and markers - to allow both better drug design and better diagnosis and treatment of disease. Proteomics - the direct quantification and identification of expressed proteins - has now reached a stage in technology sufficient to allow the direct application of a proteomics approach to the discovery of new disease-specific targets, mechanisms and markers. This presentation will deal with the following:

• the rationale behind a proteomic approach to gaining new insights into molecular pathogenesis,
• the complementarity of proteomics and genomics,
• the current status of the proteomics technology, particularly as applied at OGS, and
• a review of the current status of the programs underway at OGS for the discovery of new targets and markers in breast cancer and liver cancer.

Tissue microarray technology facilitates high-throughput molecular analyses of very large numbers of tissue specimens or cells (Kononen et al., 1998). The arrays are constructed by acquiring cylindrical biopsies from up to 500-1000 individual tumor tissues into a tissue microarray block, which is then sliced to over 200 sections for probing of DNA, RNA or protein targets. A single staining or in situ hybridization reaction provides information on all of the specimens on the slide, while subsequent sections can be probed with other probes or antibodies. Over 100 different tissue microarray blocks with over 10 000 tumors have already been arrayed at the University of Basel in Switzerland and at CGB/NHGRI/NIH. Based on our experience, the technology is likely to find numerous applications, such as:

1. Cancer gene discovery,
2. Validation and prioritization of cDNA microarray data,
3. Rapid translation of data from cell line, xenograft and animal models to human cancer,
4. Molecular basis of tumor progression,
5. Molecular profiling of large series of tumors with hundreds of biomarkers,
6. Evaluation of the diagnostic, prognostic, and therapeutic potential of newly-discovered genes and molecules,
7. Testing and optimization of DNA probes and antibodies,
8. Improved utilization of pathology archives, tissue banks and collaborative tissue resources, including clinical trial materials,
9. Surveys of human or animal tissues during development, differentiation, and disease, including surveys of tissues from transgenic animals.

Kononen et al. Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med. 1998 Jul;4(7):844-7.

The ability to sculpt microscopic features into planar solid state substrates emerged as a discipline known as micromachining in the late 1970's. Micromachining is a spin-off from the microelectronics community and thus uses similar principles for construction. Fabrication capabilities have continued to grow over the past two decades producing many intriguing actuator assemblies including miniature motors, tweezers, accelerometers, and pressure sensors. There has also been increasing interest in microscale chemical preparation and analysis procedures over the same time span. Most of these efforts have employed manual procedures coupled to large-scale instruments that can address small samples. It now appears feasible to consider using micromachining technology to fabricate miniature chemical instruments and possibly miniature chemical laboratories. Realization of these micro-laboratory components could allow a paradigm shift for chemical and biochemical synthesis and analysis similar to that provided to electronics by the transistor and integrated circuit. Some of the components we have developed for the laboratory-on-a-chip will be described. The use of such integrated microdevices for acquiring chemical and biochemical information will be discussed.

A major scientific challenge is to understand the molecular events that drive the evolution of premalignant lesions in actual tissue. Laser capture microdissection (LCM) was originated to provide a reliable method to procure pure populations of cells from specific microscopic regions of tissue sections; in one step, under direct visualization. The cells of interest are transferred to a polymer film that is activated by laser pulses. The exact morphology of the procured cells (with intact DNA, RNA and proteins) is retained and held on the transfer film. LCM technology has been successfully applied to DNA, and RNA analysis from frozen and fixed embedded tissue (Emmert-Buck et al., Science 274,998-01,1996; Bonner et al., Science 278 1481-83,1997; Luo Lin et al., Nature Med. Jan 1999). Through an NIH CRADA with Arcturus (arctur.com) LCM has been fully commercialized and is now used in more than 200 labs worldwide. In the past it has not been possible to extract, quantify and characterize the functional state of specific proteins expressed by individual subpopulations of cells in actual tissue. Consequently, an important ongoing and future goal is to extend our microdissection technology to include the molecular profiling of cancer progression into the realm of quantitative proteomics: characterization of known proteins, as well as discovery of new proteins associated with progression. LCM was applied to high throughput screening of microdissected protein profiles using solid phase protein bait binding arrays. Proteins, which bind to the bait, can be detected using the Ciphergen SELDI time of flight mass mapping. A mass mapping spectrum of protein peaks was successfully produced with a sensitivity of less than 200 cells. Using SELDI, specific differences in protein patterns could be detected between microdissected normal and invasive carcinoma (Pawletz et al., submitted). We envision the ultimate protein assay chip, which reads out the functional state of several or all proteins in signal transduction pathway. This will be accomplished using a microarray of protein bait capture ligands which recognize different functional states of specific pathway members (e.g. phosphorylation, cleaved, activated, bound to partner proteins, or unoccupied binding sites for protein or DNA).

The pharmaceutical industry is increasingly turning to high-through-put genomic approaches to enhance and accelerate the discovery of new drug targets and drug leads. Gene expression analysis can provide critical information about the molecular events and pathways involved in pathophysiological processes. GENE LOGIC has developed core technologies that rapidly identify novel targets relevant to specific diseases. Gene expression profiles can be determined using either open or closed systems. READSä is an open system that generates comprehensive expression profiles for normal and diseased tissues as well as cells treated with various drugs. Microarrays are closed systems that require knowledge of a gene for expression analysis. GENE LOGIC is developing a new biochip technology, the 3-D Flow-thruä chip, with increased sensitivity and faster response times than competing flat chip formats. Genes identified using READSä or extracted from existing databases can be used to design disease-specific microarrays for extremely rapid expression profiling, expression screening and pharmacogenomic applications. We believe that the integration of open and closed systems for high-throughput gene expression analysis with sophisticated bioinformatics tools will facilitate the discovery and development of new therapeutic compounds.

Integration of genome-wide assays with rapid selection of lead compounds and consequent efficacy and toxicity will accelerate and optimize drug discovery. A novel high-throughput genome technology, GenomeScreen(tm) has been developed that integrates ultra-sensitive and efficient genome-wide functional analysis in human cells with high-throughput screening and lead discovery. GenomeScreenä enables the rapid identification of downstream genes that are either up or down-regulated in response to a stimulus such as a therapeutic or toxic compound. Responsive, living cell clones compatible with high-throughput screening systems are rapidly identified. Aurora is developing an ultra-high-throughput screening system (UHTSSä) that will allow the screening of at least 100,000 compounds per day. The system is based on miniaturized 1ml assays and robotics coupled to ultra-sensitive fluorescent sensors, and can be used to select and test compounds in cell-based assays. This novel integrated approach is applicable for large-scale rapid screening of any compound library.

Gene expression or physiological function is usually measured with invasive techniques that require either tumor removal and consequent loss of temporal dynamics, or non-invasive techniques that have poor spatial resolution (mm to cm). In collaboration with investigators at the MGH and elsewhere, we have developed genetically engineered mice to visualize gene expression, transparent windows to visualize tissues, and computer-assisted microscopy to quantify drug delivery and physiological function continuously and non-invasively at high resolution (1-10 mm) in normal and tumor tissues in small animals. This powerful technology has allowed us to show that the size of the holes (pores) in the walls of tumor blood vessels depends not only on the type of tumor but also in what organs these tumors grow1. Furthermore, these pores become smaller as tumors shrink during angiogenic or anti-hormonal therapy². This decrease in pore size is likely to limit the delivery of drugs and gene carriers during subsequent therapy.

For the growth and progression of tumors, the generation of new blood vessels (angiogenesis) is essential. One of the most important growth factors responsible for angiogenesis is VEGF (vascular endothelial growth factor) which is generally believed to be made by cancer cells. Using genetically engineered mice that emit green fluorescence wherever and whenever VEGF is turned on, we have discovered that tumor cells co-opt the host cells (fibroblasts) into making VEGF3. Others have shown that VEGF is made in response to hypoxia (low oxygen levels)4. Hypoxia is sensed by cells through an intracellular molecule known as HIF-1a (hypoxia-inducible factor 1a). By deleting the gene for HIF-1a from cancer cells, in collaboration with European scientists we have shown that VEGF expression, oxygen level and angiogenesis decreased in these tumors. Surprisingly, these tumors did not shrink, but grew larger ⁵.

These finding collectively suggest that (i) the blood vessel wall can become resistant to drug penetration during tumor regression; (ii) host cells should be considered as additional targets for anti-angiogenic therapy; and (iii) tumors can become resistant to anti-angiogenic therapy if they have or develop HIF-1a mutation during the course of therapy. These findings have important implications for the delivery of molecular medicine to solid tumors and anti-angiogenic approach to cancer treatment^6,7.

1. Hobbs,, S. et al., PNAS, 95:4607-4612 (1998)
2. Jain, R.K. et al., PNAS, 95:10820-10825 (1998)
3. Fukumura, D. et al., cell, 94:714-725 (1998)
4. Helmlinger, G. et al., Nature Medicine, 3:177-182 (1997)
5. Carmeliet, P. et al., Nature, 394:485-490 (1998)
6. Jain, R.K. et al., Nature Medicine, 3:1203-1208 (1997)
7. Jain, R.K. et al., Nature Medicine, 4:655-657 (1998)

In 1991, TIGR researchers, while at the NIH, developed Expressed Sequence Tags (EST) sequencing for the discovery and mapping of genes. This has become the principal method of gene discovery and has contributed to many gene disease discoveries, including the colon cancer genes found by TIGR in collaboration with Vogelstein's group. Today ESTs comprise over 70% of all GenBank accessions. The existence of 100s of thousands of EST sequences required new algorithm development by TIGR that further led to a fundamental change ingenomics. Using the new algorithms (TIGR Assembler), TIGR was the first to sequence a complete genome of a free-living species, H. influenzae. With the successful sequencing of this genome in its entirety, it became apparent that the DNA of entire complex organisms many megabases in size could be accurately and rapidly sequenced by using the 'shotgun' sequencing strategy. This landmark event has been followed with nine additional genomes, including Heliobacter pylori (causes stomach ulcers), Borrelia burgdorferi (Lyme disease), and the Tuberculosis genome. TIGR is also actively involved in eukaryotic genome sequencing.

In the fall of 1998, the first chromosome of P.falciparum, which causes malaria, was published. TIGR received a grant from the NSF to sequence chromosome 2 of Arabidopsis thaliana and to annotate the complete genome. Across species, nearly half of the candidate genes that have been identified cannot be assigned a biological role. In August 1998, Celera Genomics was formed to sequence the entire human genome over a three-year period using Perkin-Elmer's Applied Biosystems Division technology, and the TIGR whole genome shotgun method. As a prelude to the human genome, Celera will sequence the fruit fly genome. Using 230 PE AB 3700 DNA sequencers, Celera will sequence over 100 million basepairs/24 hours.