Imaging Equipment
The Amnis' Imagestream 100
The Amnis’ ImageStream 100 platform is designed to acquire six multispectral images of each cell at rates of 5,000 cells per minute. Figure 1 illustrates the layout and key components of this platform. As in a flow cytometer, cells are hydrodynamically focused into a single-file core stream approximately ten microns in diameter. Sheath and sample flow are controlled by precision stepper motor-driven pumps driving flow at a velocity of approximately 30 mm/s, with a total volumetric rate of less than 4 ml/hour. As shown in the figure, cells are illuminated from the side with a 200 milliwatt, 488 nm laser for scatter and fluorescence imaging, as well as from behind for brightfield illumination. Light is collected from the cells with an imaging objective lens which is ultimately projected onto a custom charge-coupled detector (CCD). An infrared laser, not shown in the figure, illuminates the cells from the side. Prior to hitting the detector, side-scattered infrared light is reflected out of the imaging path and transmitted to a system for the determination of cell velocity and auto focus. The velocity information is used to maintain proper synchronization of the CCD detector, which is operated in time-delay integration mode (TDI), as described below.
After the infrared light is reflected, the remaining visible light passes through a spectral decomposition element. The decomposition element is a fan arrangement of dichroic mirrors that direct different spectral bands at different angles. When the spectral decomposition element is placed before the detector in aperture space, it causes the different spectral bands to be focused to laterally-offset positions across the CCD detector. With this technique, a single cell image is optically decomposed into a set of sub-images, each corresponding to a different color component. The arrangement of cell sub-images across a five-channel detector as a result of spectral decomposition of the composite cell images is illustrated in Figure 1 to the left of the detector. Though only five spectral channels are shown in the figure, the ImageStream 100 utilizes six spectral bands in total. Spectral decomposition can facilitate the location, identification, and quantitation of signals within the cell by physically separating on the detector signals that may originate from overlapping regions of the cell, as depicted in Figure 1. This image processing occurs in real time during the image formation process, rather than via digital image processing of a conventional composite image. Spectral decomposition also allows multimode imaging, the simultaneous detection of brightfield, darkfield, and multiple colors of fluorescence, by choosing distinct spectral bands for the different illumination modes. For example, 488 nm laser light can be used to excite fluorescein and phycoerythrin, which are detected in the 500 - 550 nm and 550 - 600 nm channels, respectively; scattered laser light itself can be used for darkfield imaging in a 480 - 505 nm channel; 620 nm backlighting is used for brightfield imaging in a 600 - 630 nm channel. Other possible configurations are outlined in Table 1. Optical image processing and multimode imaging can greatly increase the information obtained from a cell. To complement the high information content of multimode imaging with increased sensitivity of detection, the ImageStream platform uses time-delay-integration (TDI). TDI is a specialized detector readout mode that is commonly used in machine vision applications such as semiconductor wafer inspection and assembly line quality assurance, where there is fast relative linear movement between the camera and the object being imaged. In TDI, the image on the detector is read out continuously, one row of pixels at a time from the bottom of the detector chip. As each row is read out, the signals in the remaining detector pixels are shifted down by one row, causing the latent image to translate down the detector during readout. If the readout rate of the detector is matched to the velocity of the object being imaged, the image will not blur.
The primary advantage of TDI operation is the greatly increased image integration period it affords. In comparison to standard CCD imaging, TDI increases the integrated signal proportional to the number or rows on the detector. The practical limit on the number of rows is determined by the accuracy of the cell velocity measurement, since velocity errors result in cumulative tracking errors. Past efforts to apply TDI detection to flow cytometry met with limited success due in part to the difficulty of accurately measuring cell velocity. The flow velocity detection system employed in the ImageStream platform is accurate to better than one part in one thousand, permitting the use of a TDI detector with more than 500 rows. The resulting increased signal levels allow the detection of even faint fluorescent probes within cell images acquired at high speed. Each pixel on the ImageStream’s CCD chip is digitized at ten bit resolution, which corresponds to approximately three logs of dynamic range. In practice, higher dynamic range is achieved because each image generally covers numerous pixels. In the ImageStream 100, a ten micron diameter cell is imaged over approximately 300 pixels, providing a linear dynamic range that generally exceeds four decades.
Table 1. ImageStream Channels and Fluorchrome Choices |
|||||
| Channel 1 | Channel 2 | Channel 3 | Channel 4 | Channel 5 | Channel 6 |
| 470-500nm | 400-470nm | 500-560nm | 560-595nm | 595-660nm | 660-735nm |
| 488 SSC | Brightfield | Brightfield | Brightfield | Brightfield | Brightfield |
| FITC | PE | 7AAD | PE-CY5 | ||
| Alexa 488 | Cy-3 | Alexa610/PE | Alexa680/PE | ||
| GFP | Alexa 546 | Alexa647/PE | |||
| Sytox | PerCP | ||||
| Spec. Green | Draq-5 | ||||
| QD-705 | |||||
| Fluorescence is excited with a 488 nm laser. As a result, Channel 1 is the laser side scatter channel. 488-excitable fluorochromes are placed intochannels 3-6. Brightfield imagery may be placed into any remaining channel. | |||||
Recently, the IS100 was upgraded with the addition of a 405 nm excitation laser and an extended depth of field element which effectively maintains focus over a 16 µm range which improves imaging of relatively small punctuate staining such as is observed with fluorescence in situ hybridization. The 405 nm laser adds the detection of DAPI, HOECHST, Pacific Blue, Marina Blue, Alexa Fluor 430, Alexa Fluor 405, and Dyecycle Violet to channel 2 and Pacific Orange to Channel 3.
The ImageStream platform includes the “Image Data Exploration and Analysis Software” (IDEAS) package. IDEAS allows the visualization and photometric/morphometric analysis of data files containing imagery from tens of thousands of cells, thereby combining quantitative image analysis with the statistical power of flow cytometry. The IDEAS user interface is shown in Figure 2, which includes imagery and quantitative data from an analysis of 20,000 human peripheral blood mononuclear cells (data courtesy of T. George, Amnis). In this assay, whole blood was treated with an erythrocyte lysing agent and the cells were labeled with an anti-CD45-PerCP mAb (red) and a DNA binding dye (green). Each cell was imaged in fluorescence using channel 3 and 6 spectral bands as well as darkfield (blue, channel 1) and brightfield (gray, channel 5). Referring to Figure 2, a thumbnail image gallery in the upper left of the interface allows the “list mode” inspection of any population of cells. Cell imagery can be pseudo-colored and superimposed for visualization in the image gallery or enlarged, as shown at the bottom for the four different cell types. The software also allows one- and two-dimensional plotting of parameters calculated from the imagery. Dots that represent cells in the two-dimensional plots can be “clicked” to view the associated imagery in the gallery. The reverse is true as well: cell imagery can be selected to highlight the corresponding dot in every plot in which that cell appears. In addition, gates can be drawn on the plots to define sub-populations which can then be inspected in the gallery. Any parameter calculated from the imagery or defined by the user can be plotted. The dot plot on the left of Figure 2 shows the clustering resulting from an analysis of CD45 expression (x-axis) versus a darkfield granularity metric (y-axis) that is similar to side-scatter intensity measured by a conventional flow cytometer. The plot reveals lymphocytes (green), monocytes (red), neutrophils (cyan), and eosinophils (orange). The dot plot on the right substitutes a calculated nuclear texture parameter, “nuclear frequency” for CD45 expression on the x-axis, revealing a basophilic cell population (purple). Back-displaying the basophilic population on the left dot plot reveals that this population has the same mean CD45 expression as the lymphocyte population (green).
The nuclear frequency parameter is one member of the morphologic and photometric feature set that was developed and incorporated into IDEAS. Currently 41 parameters are calculated per image including 3 user defined parameters, which amounts to over 200 features per cell in assays that employ all six images. These include area, aspect ratio, standard deviation, total intensity of an area or mask, intensity (min, max, sum and average) of pixels in or outside of a mask, etc. Each parameter is automatically calculated for all six images of each cell when an image data set is loaded into IDEAS. Each cell is also assigned a unique serial number and time stamp, allowing kinetic studies of cell populations over the course of hours. IDEAS further allows the user to combine parameters from any number of images of the same cell. For example, the nuclear image mask can be subtracted from the brightfield image mask (which covers the entire cell) as a means of generating a mask that includes only the cytoplasmic region. Once defined, the cytoplasmic mask can be used to calculate the cytoplasmic area, the N/C ratio, the relative fluorescence intensity of probes in the cytoplasm and nucleus, etc.
Finally, IDEAS includes a powerful cross-correlation algorithm that quantitatively compares the degree of similarity of any two images of the same cell. The “similarity score” can be used to measure nuclear translocation, the degree of antibody capping, co-capping of multiple markers, and the co-localization of labeled molecules to specific cellular compartments.
CONFOCAL MICROSCOPY
The facility has a Leica TCS SP2 Spectral confocal fluorescence imaging system that is equipped with 4 laser excitation sources (405 nm diode laser, an argon laser with lines at 457, 477,488 and 514 nm, and two HeNe lasers, one at 543 and one at 633 nm). It has three continuously adjustable spectrophotometer detectors. Software allows 3-dimensional reconstruction of the obtained images. Confocal microscopy is commonly performed on tissue sections and cells stained with various fluorescent markers. Because of the 4 available laser excitation sources, our system has a wide fluorochrome application range including capabilities for viewing Hoechst and DAPI nuclear stains (blue diode). A confocal microscopist is available by appointment to assist users with the confocal microscope.
ELECTRON MICROSCOPY
Electron microscopy services include transmission (Hitachi H-600 and scanning EM (Etec Autoscan)). In addition to routine tissue processing (including special processing of primary cultures, cell monolayers, and subcellular organelles), several specialized techniques are offered, including pre- and post-embedding immunocytochemistry, scanning immuno-EM, quantitative EM autoradiography, cryoultramicrotomy, negative staining, critical point drying, extreme angle rotary platinum shadowing, and Kleinschmidt technique. An electron microscopist, trained and experienced in all aspects of sample preparation, data acquisition and photographic documentation related to these techniques, is available to investigators by appointment.
LIVE CELL IMAGING
The facility has two state-of-the-art Live Cell Imaging Systems (Leica AS MDW and Leica AF6000LX) that were delivered in June 2006 and March 2007, respectively. The live cell imaging systems enable time kinetic studies of intracellular fluorescent targets in a temperature-, humidity- and CO2 -controlled environment. Time intervals, observation duration, number of positions observed per time point and number of Z-positions acquired per field of observation are fully operator controlled. Use of the live cell imaging systems is available by appointment and a light microscopist is available to the investigators to assist in experimental and instrument set-up and data analysis.
