DSU Spinning Disk Confocal and Slidebook 4.2 Quick Guide Light Microscopy Core Facility University of Chicago Vers 0.8, February 2007 By: Vytas Bindokas, Ph.D. Core Director
2 The DSU system is optimized for low light, high speed, and multi-dimension image capture from live preparations. It is arc-lamp and optical filter-based, so most fluorescent probes can be captured (UV to NIR range). DIC transmitted images are also available, up to 8 sequential channels per image capture cycle. Turn on the system by the 4 switches on the controllers on the shelf under the air table (left to right order). The computer should be on, so press cntl-alt-del and logon through BSDAD. Launch slidebook from the desktop icons. In vers 4.2, you can create your own personal profile to store your capture settings in a private directory (use Add user and then the user pulldown menu. If you forgot to turn on the camera or the microscope you will get a string of error or warning messages. It opens a new Slide (memory folder to hold images). Press the FOCUS and CAMERA icons CAMERA FOCUS
3 View after launching the Capture (middle) and Focus (Right) windows. Pull main window to span both monitors like this. There are new hot keys to do this without mouse fiddling: cntl-e (capture), cntl-f (focus), cntl-shift-ecamera plus focus. Find your sample using the FOCUS VISUAL buttons. You will need to open the fluorescence and/or brightfield shutters. Locate the object of interest, then close the shutter(s), and pick the filter setting for the camera preview. Then make sure the object is centered and in focus for the camera (adjust stage if need be). Camera tab shows image brightness graph If this focus is not active, then press the START button under the camera focus tab. (can be stopped after a camera capture) (You can stop the camera, too) Note the BOLD filter designation show the current choice is the visual path. Note the Neutral Density filter selections. These decrease intensity in positions 1-6 (100% to 1.5%). You can set a camera exposure time and/or intensification (EM gain) by sliders on this window. The intensification can carry over to actual image captures, but the exposure here has nothing to do with the final snapshots. The camera is
4 most linear with intensification set to 127 and most sensitive at 255. AVOID blasting bright light at the camera at full gain (it will shorten the life of the expensive camera). Offset will general add background, so there really is no reason to use anything other than zero. Note that Optical filters are setup in duplicates the D and non-d versions. D will use the spinning disk (i.e., be confocal), the plain designation is regular wide-field fluorescence mode (regular microscope). The Disk settings cut the intensity since the pinholes reject out-of-focus light. The disk spins over 3000RPM and flickers the pinhole light over the sample versus a constant illumination in regular mode. Confocal illumination will bleach much less and cause far less photodamage to live cells than regular illumination. You can collect up to 8 color channels per image capture, and you can mix confocal and non-confocal images though it takes the system some time to move the disk in/out between channels. Since the regular mode is about 10X brighter, you will likely need to use higher neutral density filters to avoid saturating the camera signal. (saturation is indicated by RED portions of the image and the intensity graph piling up on the right edge) Don t do this! You can view a single color at a time and later collect up to 8 color channels using the CAMERA window. Now it s time to collect images we can save. Go to the CAMERA window (press camera icon or cntl-e, if not already open). If you are looking at fluorescent proteins, filter names are located under the Proteins buttons set. Note: the DsRed filter is also ideal for Texas Red or Alexa 594. Use the Z tab on the focus window to set the limits for 3-D volume captures. You can focus the microscope directly and press the TOP or Bottom buttons at limits of interest, or you can use the motor buttons to move 10, 1, 0.1 micrometers per press up/down. The last box and arrows allow you to type in an arbitrary movement distance (like 3 um), if desired. The autofocus can work, but generally is not useful. Check the step size and enter the spacing you want between the optical slices.
5 Select up to 8 channels to capture by clicking the checkboxes in the filter lists. You CAN mix filters form any of the lists. You will need to set a camera exposure for each channel (defaults to 100msec when checked). You can press the TEST button to see how bright it would be, then the ONCE button to automatically guess what would be best based on the current exposure. You might want to press FIND BEST, to have the system figure this out without the test phase, though it can take a bit of time. This must be done for each channel you are using (even those on different pages of filter listings, the ones that you can t readily see). You can also specify the ND filter for each channels as well as the camera intensification value, but then you must fully specify the values for every channel (default to current setting, i.e., don t change it from what is currently in use) (leave as used in the FOCUS window). You can change the ORDER of channel capture within a list, if you want to, by the Move Up/Move Down buttons. The channels from other menu lists will follow according to the order of the lists (and channels within lists). If you need 3-D volume capture and/or time lapse series of images, then check the boxes on the left. For 3-D, you can tell it to use the TOP and BOTTOM you set from the FOCUS window, or the other options to range around the current point. The range and step will import form FOCUS, or you can enter values here. If doing time lapse, you need to tell it how many images to capture and how often. (you can time lapse 3D volume captures this way, too) It will indicate how much memory your data will take, and will complain that it hasn t enough, if need be. There is an indicator above the 3D portion. If you need more space, you can have it store the data directly to the hard drive and later import the spooled images back into Slidebook. (see ADVANCED tab on this window) You can hit ENTER or press the START button to begin image capture. It will name the data as capture 1, capture2, etc, or you can specify a base name. You can later change the name (or add comment) by using the cntl-i button to adjust/read the image information (it stores the filter and camera settings, 3D parameters, time lapse parameters, etc). At the UPPER LEFT, the pulldown allows you to store the capture setups for easy re-use from day to day. If you started Slidebook under your own username versus default user you will have the capture setups stored privately versus the public hodgepodge.
Your SLIDE will list the captured content, the times, type, channel content, size,etc. You can get the full details via the cntl-i as shown. Each capture has its own display window and you can control the appearance for each. Use the up/down arrows at the left to scroll though series data (or use the mouse scroll wheel!). The dotted square icon allows you to draw regions of interest (ROI) and ZOOM the image every time the left mouse button is pressed. HOLD SHIFT DOWN and then click to un-zoom. Moving the mouse over the image reads out the intensity in each channel next to the channel names. You can toggle the display on/off with the checkboxes here. The right icon gives access to the lookup tables that control the on-screen appearance of each color channel. 6
Our camera is 14-bits (16383 brightness values), and the computer displays are 8 bits (255 levels), so there is a very wide dynamic range available to show the important image information. The default is to map the display to the min and max values in the image. Drag the bars at the graph left and right edges to map to the important part of the range. You can type numbers in, alternatively. Gamma (bowing the mapping line) allows you to increase/decrease (gamma <1, gamma >1, respectively) the display of the least bright part of the range (this example suppresses the background display). While the data goes down to 523, we only show the part above 870 in this display example. If there are multiple channels, you can do this for each color. Be sure to press the THUMB icon to have Slidebook remember that you want this display as the default view for the data (vs autoscale everything). The THUMB allows you to also re-open the series images on a key image versus the very first image (default). If the image has a user-defined color (you can tell by the presence of a color square next to the channel name). Clicking on the color square allows you to change the display color of that channel. Pick a preset or any color from the palette (click color then add to custom colors ). The checkbox allows you to toggle the display of that color channel on/off to help judge whether probes colocalize, etc. If the filter setting was configured with a default color, you can make it user-defined via the VIEW main menu (or press cntl-shift-d). You can also use a pseudocolor scheme where color corresponds to brightness by cntl-shift-p. You may have to select the channel for this display mode if there where multiple channels (which one did you want color coded?). 7
8 The ADVANCED camera/capture tab has a few important features for series imaging. The first tab controls how capture and display affect the rate of image capture. You likely won t change anything here. You may need the 4D tab to collect a single DIc image (or other filter-s) during serial captures. You might want to speed volume captures by collecting each color channel separately versus the default of changing filters to collect each at each slice (this is slower since the filters are constantly being changed). This example is set to collect colors as separate stacks. The data are merged back into one stack automatically. If you expect the image to change, such as a live preparation where tags can move rapidly, then use the default setting of collecting, let s say, red and green together at each Z slice. You may need to have it automatically refocus the sample during longer time lapse experiments. You need to tell it which channel to use, and how often to test focus. You may want to limit the search range to a couple of um to speed things up and not chase other cell layers, etc.
If you have huge data that won t fit available memory, you can have it spool the data to the disk (SPOOL tab). The default is to spool to memory. Spooling to disk and memory allows you to immediately view data after stopping a capture and a copy is made (a spl file) to disk so that if the software crashes you can still have the data. Spooling to disk only is the only option for large experiments. You will need to IMPORT the spool file to display the images after the experiment (not a hard task). If you try an overnight time lapse, do spool to disk! Notice the spool file should be in your data folder NOT the default of the program folder. We will remove data from the program folder if we find it there. Major display-related tools are under the VIEW main menu. Volume data can be manipulated by the three view tool that gives orthogonal cuts through the volume (like slicing down an apple at right angles) or a montage of small tiles, one per optical slice, or the 3D volume 9
10 The default 3D view uses ray casting to cast highlights and shadows. You can control the angle of the light via the arrow hitting the sun in the upper right corner of the rendering. You can likewise control the display range via the histogram plots next to each color (limit is 3 channels at one time). There is a transparency slider to make the view glass-like and reveal other channels within structures. The MIP, or maximum intensity projection, method is more traditional. This plots the brightest parts of each slice vs emphasis on the surface information. The 3D SURFACE (or 4D surface) renders a more solid rendering of the objects. This is a 4D surface view This is a 4D volume view. You can see up to 3 channels animated inside the volume over time as a movie. You can use the mouse to rotate and zoom as the data are being played back. Note the playback speed slider. Movies can be exported for presentations.
11 The IMAGE main menu allows you to import data from other programs and the SPOOL DATA collected in big experiments, etc. IMPORTANTLY, you will want to EXPORT the data as raw 16-bit tiff files so that you can use the images in other programs, especially ImageJ freeware from NIH. Do this at the end of the experiment since you can export all the raw data from images collected the same way in one operation. (if you do this more than once for the same data, Slidebook will not overwrite the output, but rather add a number to the names as versions ; this can get confusing). You may wan to keep the export data separate in an output directory for grins later on. An important feature is the ability to apply the same display settings to multiple images per slide. This would make staining controls look appropriately dark, for example. This feature is hidden under 4D Operations as Renormalize All. This will make the default (thumbnail) view uniform. Any image that is currently opened will NOT take on the new display range, so you ll need to re-open them. Alternatively, you can use the VIEW main menu task of VIEW EXPORT. This allows you export the images AS DISPLAYED in a PowerPoint or Photoshop-friendly image that can be shown to whoever without any other tweaks. Remember,
12 Image/export preserves the wide dynamic range of the raw data; the VIEW/Export limits you to just the part of the data on the screen. Other useful items is the ability to flatten 3D data (or 4D) to maximum intensity projections. It s beast to do this at the end of an experiment, since you can have it do all the captures collected the same way at one time. If you do it at the end, the projections will be placed in a new slide (memory folder). You can save that separately (or export as need be). This routine will not overwrite data, but creates version numbers.
Appendix 1 Olympus Disk Scanning Unit (DSU) Filter specs 13 Excitation filter wheel Emission filter wheel 1 - open 1 - open 2 350/50 DAPI 2 457/50 DAPI 3 405/12 blue-violet 3 528/38 Fitc-Cy2 4 490/20 Fitc-Cy2 4 617/73 tritic- Cy3 5 555/28 tritic- Cy3 5 685/40 Cy5 6 635/20 Cy5 6 470/30 CFP 7 436/10 CFP 7 525/40 EGFP 8 500/20 YFP 8 535/30 YFP 9 480/25 EGFP 9 620/60 DsRed 10 565/25 DsRed 10 - closed Neutral Density Filters 1 100% 2 25% 3 12% 4 6% 5 3% 6 1.5% Internal DSU filter Cubes Scope (visual) cubes 1 mirror 1 mirror 2 dapi/fitc/red/cy5 QUAD 2 dapi/fitc/red/cy5 QUAD 3 cfp/yfp 3 cfp/yfp 4 gfp/dsred 4 gfp/dsred 5 open 5 cy3/cy5 6 dic 6 dic
Appendix 2 DSU Objective Properties 14 objective NA WD um/pixel 10water 0.4 500 1.6 20water 0.7 400 0.8 40water 1.15 260 0.4 60water 1.2 280 0.267 100tirf/OIL 1.45 200 0.16 10dry 0.3 1000 1.6 Using the 150X TIRFM objective generates 0.106 um/pixels Appendix 3: DSU pinhole disk spacings * Disk3 is default in the scanner. We have D1, D2, D3, and D4. Ask to have it changed. Appendix 4: Live prep hints Use the Harvard stage heater system and pipe in moist 5% CO 2 for longer experiments. Use the two glasstopped covers. Set camera intensification to max and ND filter to 6% lamp output. This will allow volume captures for 20 hours at a 5 min interval with little or no bleaching of GFP/YFP. Onset of GFP expression can be detected with 1300msec exposures. Opt for longer exposures and less light. Hoechst 33342 can be co-collected using 405EX (vs 350) to decrease effects of UV light on live cells.