Comparative genomic hybridization (CGH), which involves the simultaneous hybridization of differentially labeled total genomic DNA from test cells and reference normal cells to metaphase chromosomes, has been used extensively to screen tumor genomes for regions of DNA sequence copy number variation. Analysis of these hybridizations requires quantitative analysis of the ratio of intensities of the fluorescent hybridization signals as a function of position along the chromosomes, which basically serve as a convenient genetic map. The ratios need to be measured very accurately since changes of about ± 20% from the average for the genome indicate important genetic events. Widespread use of CGH over the past several years has identified numerous regions of the genome that may contain currently unknown cancer genes. For example, regions of increased copy number may indicate sites of oncogenes, while regions of copy number decrease relative to average for the genome may signify the presence of a tumor suppressor gene.

In human leukemia and solid tumors cytogenetic analysis provides critical information of diagnostic and prognostic importance. Frequently however, the unambiguous characterization of all aberrant chromosomes is difficult using conventional chromosome banding techniques alone. We have developed a novel approach, termed spectral karyotyping (SKY), based on the hybridization of 24 differentially labeled human chromosome painting probes that allows the simultaneous color identification of all human chromosomes. This genome scanning method combines Fourier spectroscopy, CCD-imaging, and optical microscopy to visualize the hybridization of differentially labeled chromosomes painting probes. One of the most important analysis algorithms is the spectral-based classification algorithm that enables multiple different spectra in the image to be identified and highlighted in classification-colors. This allows assignment of a specific classification-color to all human chromosomes based on their spectra. This algorithm assumes that the (reference) spectrum of each chromosome has been measured and stored in a reference library in the computer

The gram-positive bacterium Listeria monocytogenes is one of several intracellular pathogens which can move about within its host's cytoplasm using a form of actin-based motility. This motility plays an important role in the virulence of the microbe, which can cause serious disease in humans. Actin is a host-cell protein whose polymerization is required for the locomotion of animal cells. L. monocytogenes exploits this normal cellular machinery for its own movement by creating a “comet” tail of cross-linked actin filaments behind it. Actin polymerization occurs at the bacterial surface and is required to drive the bacterium forward.

We are interested in how L. monocytogenes induces the polymerization of actin and how this polymerization can lead to the generation of force. To study this, we have used time-lapse videomicroscopy and a cell-free system derived from Xenopus egg extracts. We have observed rare instances of a saltatory movement of L. monocytogenes in these extracts.

We have developed a multi-mode digital imaging system (1-3) which acquires images with a 12 bit cooled CCD camera. A multiple band pass dichromatic mirror and robotically controlled excitation filter wheels provide rapid wavelength selection for epi-fluorescence with DAPI, fluorescein or GFP and X-rhodamine fluorophores while maintaining image registration on the cooled CCD detector. Shutters select illumination either by epi-fluorescence or by transmitted light for phase contrast or DIC. A robotically controlled emission filter wheel in front of the CCD camera inserts an analyzer in the light path for DIC imaging. To maximize fluorescence light intensity, the analyzer is removed and an optical flat of equivalent optical thickness is inserted for fluorescence imaging. A slider is inserted at the field diaphragm position of the fluorescence epi-illuminator to provide in-focus slit and spot targets for 360 nm photoactivation of “caged” fluorophores. The microscope system is robotically controlled and image acquisition and analysis is performed using MetaMorph™ digital imaging software.

Recent methodology, to be described, now makes possible specific localization and analysis of genetic loci within 3-Dimensional interphase nuclei in intact cells and tissues with minimal perturbation of the chromosome structure (Dernburg and Sedat, 1997). These techniques define genetic loci that specifically interact with the nuclear envelope and interior structures; we are able to map all loci to highly localized 3-dimensional positions within Drosophila embryonic nuclei (Marshall et al, 1996). S-Dimensions-as-a-function-of-time (4-D) studies of live nuclei, from Yeast and Drosophila, allow dynamic chromosome interactions to be probed and quantitated. Our results suggest a very dynamic but highly determined and organized nucleus. Using these approaches, we can now study specific mechanisms leading to homologue chromosome pairing and position-effect variegation (Dernburg et al., 1996).

Our laboratory is interested in understanding how 10 and 30 nm chromatin fibers fold to form interphase and mitotic chromosomes. Experimentally this has been a very difficult problem to investigate due to a number of technical difficulties. A common approach to this level of chromatin organization has been to use protein extraction conditions which experimentally “unravel” the native chromosome architecture. The difficulty with this approach is separating in vitro produced artifacts from remnants of in vivo structure.

As an alternative strategy we are focusing on changes in interphase chromosome structure during cell cycle progression and initiation of transcription or DNA replication, with the goal of identifying intermediates in the pathway of chromosome condensation or decondensation. A key element of this strategy is preserving chromosome structure as close as possible to its in vivo structure while using 3-dimensional light and electron microscopy reconstruction methods to “computationally” unravel native chromosome structure.

For daughter cells to receive equal copies of the genome during mitosis, the replicated chromosomes must attach to and move bi-directionally on the mitotic spindle. A chromosome becomes attached to the spindle via a pair specialized structures, known as kinetochores, that are positioned on opposite sides of its primary constriction (one on each of the two chromatids). In addition to being the spindle attachment site, kinetochores are also involved in producing and/or transmitting the forces for chromosome motion. In vertebrates the kinetochore closest to a spindle pole at the time of nuclear envelope breakdown usually is the first to attach to the spindle. As a result of this attachment the now “monooriented” chromosome moves toward the closest pole where its only attached kinetochore initiates oscillatory motions toward and away from that pole until the unattached sister kinetochore acquires microtubules (Mts) from the opposite spindle pole.

The dynein ATPases are a large family of motor enzymes that provide the driving force for flagellar motility and contribute to microtubule-based transport inside cells. The challenge for the field is to appreciate the functional significance of the multiple dynein motors, to determine how the cell assembles a motor complex and targets each motor to its appropriate location, and to understand how a cell regulates the activity of each motor to accomplish its specific task(s). In our laboratory, we have capitalized on the highly ordered structural organization of the flagellar axoneme and on the ease of genetic analysis in Chlamydomonas to ask how a single cell controls the assembly and activity of its multiple flagellar dyneins. In particular, we have focused our efforts on the inner dynein arms, which are both necessary and sufficient for normal flagellar motility. The significance of this work derives from the conservation of axoneme structure across species, as well as the numerous functional homologies between the flagellar and cytoplasmic dyneins.

We are developing techniques for obtaining the thinnest possible serial sections of biological specimens. Our goal is to produce not just an occasional very thin section but large numbers of 5-15 nm serial sections suitable for 3-D reconstruction. We have worked out conditions that allow us routinely to cut serial sections ˜15 nm thick, and have achieved ribbons of sections with an average thickness of only 11 nm.

We have worked with the Leica Ultracut S and Ultracut UCT microtomes, which have proven to be exceptionally stable instruments, capable of very regular advance between cutting strokes. The resulting low variability in section thickness is essential for serial sections thinner than 15 nm. However, these machines have an inherent forward drift that increases the thickness of sections cut soon after a block has been mounted and prepared for cutting. This drift is initially 50-100 nm/min, but decreases to only 10-20 nm/min if the microtome is allowed to cycle for 0.5—1.5 hours.

Centrosome structure: The centrosome is the major microtubule-organizing center of animal cells - it initiates spindle formation at the onset of mitosis and plays an active role in chromosome segregation. At the most basic level, microtubule (MT) formation is regulated at least in part, through the functional availability of nucleation sites on the centrosome. Functional centrosomes can be isolated from Drosophila embryos. When visualized with IVEM Tomography (IVEM-T) ring complexes are visualized throughout the peri-centriolar material (PCM) (1). γ-tubulin has been implicated in MT nucleation. Through immuno-IVEM-T we have shown that the ring structures within the PCM, contain y-tubulin and further that the only place γ-tubulin is found is at the minus (nucleating) end of MTs (2). This strongly implicates these γ-tubulin containing ring structures in MT polymerization, until our studies there was no direct proof that γ-tubulin was located at the site of MT nucleation. Furthermore Zheng et al have been able to isolate similar ring structures that are soluble from both Xenopus and Drosophila embryos (3).

We are using electron tomography to investigate the internal organization of mitochondria and the interactions of these organelles with other cellular components. Two-axis tilt-series are recorded from 0.2-1.0-μm sections using either an AEI EM7 high-voltage or JEOL 4000 intermediate-voltage electron microscope. Reconstructions are computed using programs in SPIDER and visualized with the Sterecon system.

Three-dimensional reconstructions of conventionally fixed and plastic-embedded rat-liver mitochondria (isolated and in situ) indicate that the cristae, the infoldings of the inner membrane, are pleiomorphic, defining compartments that are connected to each other and to the surface of the inner membrane by narrow (30-40-nm diameter) tubular regions (Fig. 1). This is contrary to the “baffle” model commonly shown in textbooks and has important implications for the diffusion of ions, metabolites, and macromolecules inside the organelle. The same basic organization is also observed in reconstructions of mitochondria in rat-liver sections prepared by the cryo-fixation method of Tokuyasu(Fig. 2).

Gap junctions are specialized cell-cell contact areas by which cells communication with each other. Within these contact areas are tens to thousands of membrane channels. A gap junction membrane channel (also referred to as an intercellular channel) is unique among membrane channels in that it is composed of two oligomers with each of two adjacent tissue cells contributing one oligomer (called a connexon or hemichannel). The pore of the intercellular channel controls the passage of small molecules and ions from one cell to another.

We are interested in how the structure and surface topology of the gap junction connexon at its extracellular surface influences the docking and formation of an intercellular communicating channel. It has been demonstrated that connexons made from some connexins will dock and form functional channels with some but not all connexons made from other isoforms. This selectivity is surprising considering that the primary sequences of the docking domains are highly conserved.

Over the past decade, there has been increased interest in quantifying cell populations and morphological structures in tissue sections using image analysis systems. Automated analysis is now being used in limited pathological applications, such as PAP smear evaluation, with the dual aim of increasing the accuracy of diagnosis and reducing the review time. Applications such as these primarily use gray scale images and deal with cells that are well separated. Quantification of routinely stained tissue represents a more difficult problem in that objects can not be separated in gray scale and the images are more complex. Many of the existing semi-automated algorithms are specific to a particular application and are computationally expensive. Hence, we investigated general adaptive automated color segmentation approaches which alleviate these problems.

The first stage of this research concentrated on separating tissue color clusters in a color space histogram of the image.

Early electron microscopic studies documented that significant changes in the membrane systems of cardiac cells occur in both ischemic and non-ischemic heart failure. These studies relied on analysis of two-dimensional sections and although quantitative changes were observed, the overall organization of the tranverse tubules (T-tubules) and the sarcoplasmic reticulum could not be assessed. In a 3-dimensional study using high voltage electron microscopy (EM) of the T-tubules in spontaneously hypertensive rats, Nakamura and Hama (1991) observed that concomitant with an increase in surface area, the T-tubule system becomes progressively more disorganized and exhibits structural irregularities such as increased numbers of longitudinal tubules, numerous short dead end branches and complex tubular aggregates. These authors suggested that this disorganization may interfere with synchronous contraction over the entire cell.

In the present study, we examined the 3-dimensional organization of T-tubules in the left ventricle of explanted human hearts using confocal microscopy and EM tomography.

A life threatening blood clot is the major cause of heart attacks. Thrombolytic therapy attempts to restore blood flow by activating the body's own fibrinolytic system at the site of the occlusive thrombus. However, for unknown reasons, therapy is unsuccessful in greater than 20% of patients. We have previously shown that the endothelial cells lining the wall of the vessel can play a substantial role in inhibiting clot lysis. This is due chiefly to the secretion of inhibitory molecules by endothelial cells. However, endothelial cells also have receptors for fibrin and little is known about how the direct interaction of fibrin with cells may influence lysis. To investigate this we have undertaken a series of microscopic studies to analyse the influence of endothelial cells on clot structure. We report here that endothelial cells can organize clot fibers into tight assemblies. We also show that, at least in culture, fibrin can act in concert with antithrombotic molecules to dramatically affect endothelial structure

Non-confocal Laser Scanning Microscopy (LSCM) was developed to allow applying the tenents of flow cytometery to specimens attached to a microscope slide (1). Areas of the slide are scanned, fluorescently labled cells are automatically segmented, and a list of features for each cell is calculated and stored in a list mode data file. Figure 1 shows a block diagram of the LSCM. A collimated laser beam scans in the y direction, and values from photomultiplier and photodiode detectors are digitized to .25 micron resolution. The automated stage advances in .5 micron steps in the x direction. When cascading memory banks are filled, the “image” is segmented. Figure 2 shows contours drawn as part of the segmentation process. The innermost contour is drawn around pixels that exceed a predetermined threshold, and is used to define events. Other contours are used to define integration areas, background calculation and correction, and integrating nuclear vs. peripheral fluorescence areas.

Oocyte meiotic maturation involves complex processes of nuclear breakdown, spindle formation, and organelle movement necessary for successful fertilization and early development. Yet these take place in the mouse without the aid of centrioles; instead the spindle is organized by a number of microtubule organizing centers (MTOC).Multiple MTOC are seen by fluorescence microscopy to migrate from the cortex to the nucleus (germinal vesicle; GV) during the first 30 min of maturation. However, EM has not revealed the expected ultrastructure of MTOC known from previous studies. New information from the present study sheds light on MTOC number, location and ultrastructure prior to and during very early maturation.

Maturation occurs in vitro when oocytes are released from ovarian follicles into culture, but is blocked by dibutyrl cAMP or IBMX (3-isobutyl-l-methyl xanthine; Sigma). MTOC were studied after 0-2 h in culture medium; IBMX medium (0.2 mM); an inhibitor of MT (nocodazole, lμM); or MF (cytochalasin, 5 μg/ml) by fluorescent confocal microscopy (CM; Biorad 600) or standard EM.

The flowing behavior of individual erythrocytes in blood vessels is usually determined by their deformability, which is controlled mainly by the nature of their interior constituents and the flexibility of their surface membrane. Moreover, the physical behavior of erythrocytes passing through capillaries has been examined in vivo by light microscopy. However, little has been known about ultrastructural changes of such erythrocyte shapes flowing in blood vessels in vivo. Recently, a new technique was developed for freezing cells and tissues in vivo without stopping the blood supply, which was referred to as “in vivo cryotechnique”.This method has been also suitable for obtaining informations about dynamic morphological changes.

Seven female Balb/c mice were anesthetized peritoneally with sodium pentobarbital (100μg/g body weight), and their abdomen was opened through a pararectus incision. For artificial cardiac arrest, some mice were anesthetized with an excessive dose of the anesthetic (500μg/g body weight), their respiration and heart-beat were completely stopped, and the following procedures were done within one minute. A liver was put on a plastic plate without disturbance of blood circulation, and the “in vivo cryotechnique” was performed. Briefly, a cryoknife was pushed into the liver as fast as possible and the tissue was immediately poured with liquid isopentane-propane mixture (-193°C) (Fig.la,b).

Brush border myosin I (BBMI) was first identified as a major constituent of the microvillus cytoskeleton, where it links the actin filaments of the microvillus core to the overlying plasma membrane. Recently, however, BBMI was also found on isolated Golgi membranes. This observation suggests that cytoplasmic BBMI might act as a molecular motor for apically targeted secretory vesicles.

In an earlier study from this laboratory a 105-110 kD protein was isolated from the chicken small intestine, which displayed several characteristics of the BBMI. Using an antibody against this BBMI- like protein (MLP), the protein was detected in the apical brush border of the enterocytes. The limited resolution of the light microscopic immunohistochemical techniques used in that investigation, however, did not provide information about the intracellular distribution of MLP in the enterocytes.

In this study the intracellular localization of MLP was characterized using post-embedding colloidal gold immunocytochemistry. Pieces of the jejunum from 6-week-old male White Leghorn chickens were fixed and embedded in LR Gold resin.

We developed a new type 3-Dimensional Internal Structure Microscope (3D-ISM) for the observation of internal structures of samples. The internal structure of a sample is obtained by cutting it into thin slices and observing the cutting side continuously while cutting. The position of the camera, as well as the sample position are fixed. Therefore there is no shift between the sections, and this system can obtain a true color image of the sample, which is a high resolution and a high-quality three-dimensional image compared with X-ray CT and MRI. After repeatedly slicing a sample, the digital image data from the sectional views is transferred to a computer, where 3-dimensional images of the internal structure of the sample are reconstructed. Using this system we analyzed many biological organisms. In this paper, a mouse specimen has been cut and the 3-dimensional images are shown.

This article presents the outline of the device and the principle of the observation. FIG.l shows a corresponding block diagram and Fig. 2 a schematic view of the 3D-ISM.