Acquiring Images with Volocity

Acquiring images is time consuming—acquisition times can be long and there are a lot of images to be acquired—and fluorescence signals are perishable. Therefore the goals is to acquire the best images possible and get everything needed the first time around, reducing the time spent imaging while preserving the signal for anything that does need to be imaged a second time. 

Before Turning on the Microscope

• Sign up to reserve the system on the calendar in the hallway
• Decide beforehand what will be imaged: which series, sections, which labels and how many channels
• For fluorescence, take an empty slide tray and collect the sections to be imaged from the freezer
• Allow time for the slide to warm to room temperature
• After the imaging session, return all slides to their home tray

Start Up the Hardware and Software

• Turn on the hardware and computer.
• Turn on the hardware from left to right, in the order numbered
• Sign in on the sign in sheet
• Boot the computer under Windows (hold down the option key during startup and select the Windows startup HD)
• Make sure all hardware, including the camera, are on before launching Volocity
• Launch Volocity and log in
• Select either 'Volocity' or 'Acquisition Only' from the pull down menu
• Open the Image Acquisition Log and the Landmark DB
• From the Image Acquisition Log spreadsheet, select 'Go to Live Form' under the 'Form' menu
• In the Landmark DB note the date imaged on the sheet for that series, in the cell for that section 

Before Putting a Slide on the Stage

• Check that the camera is set to monochrome (Fluorescence) or RGB (Nissls)
• Configure the Acquisition Setup
• Select Acquisition Setup from the Video menu, or double click the small green 'Acquisition Protocol Feedback Display .

• Under the Channels tab:
• Name the file (see the Working with LightRoom entry for naming conventions). Do this for every image.
• Add all channels to be acquired. Select 'Change Channels with Light Paths', choose channel from the pulldown menu. Click the '+' to add another channel
• Manage shutter speeds. Use 'Balanced Sample Protection' 
• Unless acquiring a z-stack, no other options be selected
• Under the Stitch tab,
• Capture XY using Ludl Stage Controller 
• Set the overlap to 15% 
• Select Create Stitched Image and Auto Align Fields (only
• Select 'Save Raw Tile'
• Click 'Ok' to return to the Video Preview window

• Configure a Light Path for each channel.
• For now, verify that the filters match the channel, that "Fuor' is showing in the Channel Controls pull down, and Monochrome in the camera pulldown menu
• For the channel to be use to scan the overview, set the objective to 5X and the gain to around 5

• 

• Calibrate the Stage
• Switch to XY View
• Select 'Calibrate Stage' from the 'Stage' menu. Note that there is no need to move the stage to the load position if there is no slide

Create an Overview

• Looking through the ocular, find the slide and section to be imaged and preview all (visible) channels
• For each channel make a mental note of the brightest parts, which should be used for setting the correct exposure
• Close the shutter and switch the light path to the camera tube by pulling out the silver post on the side of the microscope
• From the Live Video view, at 5X w/gain of about 5X, AutoExpose and focus
• Return to XY View and drag the marquee tool over the approximate area to be imaged
• From the Stage menu, select 'Scan Selected Area'
• When the overview is generated, double click on the part to be used to set exposure for that channel

Capture the Image

• Return to Live Video. Switch to the desired objective open the shutter and click the AutoExpose button. Leave the the gain high for focusing. Find the best focus. 
• AutoExpose again after reducing the gain and determining the offset. See below for typical settings, and Exposure, Gain and Offset in Volocity for more detailed explanation, but in general, keep exposure times as short as possible)
• Save these settings by clicking the disk icon
• Switch to the next channel verify the focus and objective, set gain, offset and exposure. Save the settings. 
• Repeat for all channels, switching back to XY View as needed to move the stage to the best area for calculating exposure for each channel.
• When each 'Light Path' has been set up, return to the XY View and drag the Marquee tool over the area to be imaged
• Hit the red Record button to acquire the image
• Create snapshots (under the Image menu) of the composite image and each channel individually. 
• Export each snapshot as an OME Tiff.

Tips & Troubleshooting

General:
 Keep the shutter closed as much as possible. Note that Volocity will at times open the shutter  automatically when switching between 'Light Paths'

Stitching

10% is recommended, and 12% usually works well, but 15% should allow the software to stitch lower contrast sections and is more forgiving when the camera is not perfectly aligned

Light Paths

The objective, exposure, gain, offset, fluorescence filters and condenser can be changed for each channel—which means that each of these values must be set for each channel. All changes must be saved before switching to different channel or the changes will be lost.

Quick Scan

Generate a low magnification view to precisely control image content, and to navigate to a specific part of a section or slide.

The 'best' channel is the most abundant (helps recognize location) and most stable (slowest to bleach). If background is very low in all channels, it can be artificially raised by increasing the offset (but remember to change it back)

It is helpful to set the gain high to keep exposure times low. This reduces potential bleaching of the signal, but mainly it reduces the  video lag making it easier to focus

Exposure, Gain and Offset
The rule of thumb is keep the exposure as low as possible but long exposure is better than high gain. For acquisition, an Exposure between 50 and 300 ms is fine, and 300 is preferable to repeatedly tweaking values and re-calculating exposure. For any exposure less than 100 msec, Gain should be 1.0X. The problems with long exposures are 1) the signal may bleach and 2) it adds to the acquisition time. Neither become an issue until well past 500 msec. Gain: The rule: Feel free to use up to 2.5X gain to control exposure time. Use 2.5 to 5X  gain only when exposure are greater than 300 msec at 2.5X, and use  higher than 5X only in extreme cases. Increasing gain always adds noise, it is just that below a certain level it is undetectable. We always want to be well below that level. Offset determines black levels. Blacks should be kept as dark as possible but no darker.

Setting Exposure:
After the Quick scan, set the Gain to 1.0x,  the Offset to around 0 and AutoExpose. If the Exposure is greater than 100 ms, feel free to increase the Gain

Working with LightRoom

LightRoom (LR) is the program that keeps track of the image files for us—how they are stored, which files belong together (e.g., individual channels and the merge, tiles in a mosaic, slices in a stack, etc.) and all edits of the original. LightRoom is also going to play a big part in the analysis because it has powerful tools for selecting, grouping and comparing images. 

To get the most out of LR, here's what we need to do:

Exporting Images: Import only whole images (stitched mosaics or projections). If Volocity cannot stitch the image, assemble it using Adobe Bridge and Photoshop (PS) before importing it. Export the composite image with all channels and each channel individually. Export only one copy of each. Use an R-G-B color scheme when you can, but as long as we have the individual channels we can assign or change colors in PS as needed.

File Naming: Files should be named according to their content, or more to the point, the filename should not refer to anything that is not in the file. To denote the relationship between the composite image (an image of all components) and its component channels (multiple images of individual component), use the following scheme:


Composite:
2013_05_31_001a_DAPI, Cy2-MCH, Cy3-NOS_G04f, lvl 24,25

Channel 1:
2013_05_31_001b_DAPI_G04f, lvl 24,25

Channel 2:
2013_05_31_001c_Cy2-MCH_G04f, lvl 24,25

Channel 3:
2013_05_31_001d_Cy3-NOS_G04f, lvl 24,25


Keywords:
Keywords must come from a defined vocabulary to be useful. Most but not all keywords have already been created. Existing keywords can be found in the "Keyword List" in panel on the right. Note that that most have been organized into folders. Unfortunately the Keyword List can't be used for Keywording, but there is an autocomplete function when typing in keywords. If the keyword you are adding does not show up in the autocomplete list, check the Keyword List for the right keyword or correct spelling.

For each image, add these keywords:

• Peptide/Antibody
• Label (Cy 2, etc.)
• Counterstain (e.g., DAPI, Nissl)
• Tissue & Series (e.g., G1204a)
• Labeling method (IHC, etc.)
• Imaging method (epifluorescence, confocal)
• Image Type (Mosaic, Projection, etc.)
• Test or Data series
• Image content: Brain structures (just the obvious ones: LHA, AHN, VMH, PVH...)
• Atlas Level(s). Should know this before the image is taken (from the landmark database)

Notes:

Vocabulary: "Mosaic" and "Projection" are more accurate and we should use them instead of "Stitch" and "Stack". 

Tags:
In addition to keywords images can also be tagged with Stars, Flags, and Colors. Tags can be assigned from the menu or by right-clicking the image. 

Star Ratings:

1. Original Tile or Component
2. A complete raw image (Stitched, Projected and/or single frames). 
3. Manipulated w/out altering content (Stitched, color corrected, basic levels or curves, etc.) incl. Cropped and/rotated (part of image discarded but not altered)
4. Content Edited (for brightness/contrast, size/resolution, etc.)
5. Output version/figure component

Almost images should get a 2 star rating when first imported, or rather we should avoid adding more 1 star images. Most data series images will make a trip through PS and return with a 3 star rating. We will try to keep PS edits to a minimum but multi-channel composite images will usually get a 4 star rating.

Flags ("Rejected", "Flagged", "Unflagged)

Most images will be unflagged, which is the default. A few images will be flagged under special circumstance:

• Use the "Rejected" flag  to  hide unusable images without deleting them
• Use the "Flag" to mark images that need attention, like missing or non-standard keywords, the file format is not usable or if content is not clear, etc.

Color Labels:An image can be tagged with one of 5 different colors (R, G, B, Y, P) that we can explicitly defined. The obvious use for Color labels is to keep track of channel information (use Purple for Cy 5, Alexa 647, etc., and Yellow for Fluorogold, which isn't relevant to the LHA project).

Metadata:
Metadata opens up another realm of options. For now let's leave it at that: optional. Note though that it migh be convenient to copy the file name into the 'Title' field. Also that the 'Caption' field ccontains all of the microscope info exported with an OME Tiff, which may be useful at some point.

Anti-FITC Test Results

2013_05_26_001
NEI (rb; 1:10k) A488 rb anti FITC 1:750
16 hour incubation in A488

2013_05_26_003
NEI (rb; 1:10k) A488 rb anti FITC 1:750
39 hour incubation in A488

2013_05_26_004
NEI (rb; 1:10k) Cy2ms anti FITC 1:1k
16 hour incubation in Cy2

2013_05_26_005
NEI (rb; 1:10k) Cy2ms anti FITC 1:1k
39 hour incubation in Cy2

2013_05_26_006
NEI (rb; 1:10k) Cy2ms anti FITC 1:2k
16 hour incubation in Cy2

2013_05_26_007
NEI (rb; 1:10k) Cy2ms anti FITC 1:2k
39 hour incubation in Cy2

Imaging in Volocity: Setting Exposure, Gain and Offset

Three successive images of the same section. Image 1 is overexposed, with a longer Exposure and lower Offset. Exposure was corrected in Image 2 by using navigating to the brightest part of the image and using AutoExposure. Contrast was improved by increasing Offset from 138 to 230. The effect of Gain was examined in Image 3. Holding Offset constant, increasing the Gain from 1.7 to 5X decreases Exposure from 101 to 36 ms, with no noticeable increase in noise or loss of detail. Therefore, at least within a certain range, Gain can be used to optimize Exposure time to minimize bleaching and keep acquisition times reasonable. (All images: Bilateral composite image of MCH (Zamir's rabbit anti-MCH, 1:1k) in the LHA. 10X, 7 X 5 tiles)

Image 1:

Exposure: 157ms

Gain: 1.7x
Offset: 138 levels

Image 2:

Exposure: 101ms
Gain: 1.7x
Offset: 230 levels

Image 3:

Exposure: 36ms
Gain: 5.0x
Offset: 230 levels

Volocity Standard Imaging Procedure Take 2

1. Turn on the hardware, turn on the computer (windows), then launch Velocity
2. Save as new library (yyyy_mm_dd_Volocity).
3. Calibrate the stage in the XY stage view before putting slide on:
1. lower the stage away from the objective.
2. stage-> calibrate stage
4. Look through the eyepiece of the scope to find region of interest under Fluor light and appropriate filter. Focus.
1. control of fluorescent shutter is on the computer as well as on the box #3
5. Switch light path to camera (pull out post). Make sure camera is set to monochrome, Volocity is set to Mono and Fluor (2 different pulldown menus)
   Starts off in black and white until you choose a channel to pass the light through
6. click on a channel, choose light path appropriate to the fluorophore
7. double click on the lime-green rectangular box (to the right of the snowflake) to open “Acquisition set-up”
1. change name (yyyy_mm_dd_001_primary_fluor)
2. select the number of channels
3. change focus using (NONE) if not taking z step
4. stitching
1. change XY using Ludl XY stage
2. dont correct for shading
3. auto align field
4. correct for brightness
5. save raw tiles
6. 12% overlap
8. make sure you keep the shutter closed as much as possible to avoid bleaching
9. in Live View, do a quick focus if necessary, then click Autoexpose to start to find the appropriate exposure
   a. start with DAPI channel
   b. look for brightest area or most intense staining when using Autoexposure
   i. Try to keep exposure between 50 and 250 ms (not always possible)
   ii. If exposure is too long, increase gain to reduce time
   iii. Offset controls black level, generally positive (+) value, should show some tissue 
   features while leaving ventricle black
10. click the floppy disk on the top right to save the channel settings
11. switch to the XY stage view at 5x and select an area with marquee tool, 'scan selected area' to get a sense of where we are in the section
12. switch back to Live view to correct exposure
13. re-check the acquisition set-up
14. in XY stage view, select desired region and hit record (usually at 10x or 20x)

   15.  Fill out Image Acq Log for each image before taking the next
                a. most relevant info for image, like total acq time, summarized in the "experiment log" tab
   16. Multichannel: add channels in acq setup
                a. Use a different light path for each channel
                b. Set exposure for each separately
                     i. use brightest area for each
                     ii. save before moving to the next channel

LHA Cytoarchitecture

LHA parcellation (from Swanson '03)

According to the atlas, the LHA lies within the hypothalamic lateral zone, which can be divided into 'state related' (LZs) and 'motor related' (LZm) regions. The LZs has only one component cell group, the LHAd. The LZm can be further divided into the LPO and the LHA. The LHAmo contains all of the lateral zone except the LPO and LHAd, and all of the LHA except LHAd. The cell groups with the LHAmo are: the juxtaparaventricular (LHAjp), juxtaventromedial (LHAjv), anterior (LHAa), suprafornical (LHAs), subfornical (LHAsf), magnicellular (LHAm), parvicellular (LHApc), ventral (LHAv) and posterior (LHAp) regions. Four of these are further divided into several zones or parts. 


Hypothalamic Lateral Zone = LZs + LZm

LZs = LHAd

LZm = LPO + LHAmo

LHAmo = LHAjp + LHAjv + LHAa + LHAs + LHAsf + LHAm + LHApc + LHAv + LHAp

The atlas description and annotations are below. The atlas page where the info can be found is shown in red and the annotation number in italics. Citations are listed at the end but can also be found in the atlas' Reference section.

p. 173

     169: LHAd: distinct cell sparse region at the rostrocaudal level of the ventromedial hypothalamic nucleus (personal observations). Highest concentration of MCH & H/O

Note: the LHAd is listed under LZs: Hypothalamic lateral zone, state related. LZs has only one part

p. 175

263: LZm = Hypothalamic lateral zone, motor-related = LPO  and all LHA except LHAd; "often thought of as an interstitial nucleus of the medial forebrain bundle, and the rostral end of the reticular formation (see Nauta and Haymaker, 1969)

   265: LHAmo = Lateral hypothalamic area, motor related = everything except LPO.

     266: LHAjp - small, moderately-densely packed neurons lies rostral to, and is distinct from, the PHHjd

     267: LHAjd - this region is quite distinct due to a high density of small to medium-sized neurons

    268: LHAjv (juxtaventromedial region; jvd, jvv,), LHAa (ad, ai, av):  distinguished by a lower density of neurons than is found in all surrounding areas. A dorsal zone receives a circumscribed input from the posterior basolateral nucleus of the amygdala (Petrovich et al., 2001) and is slightly less dense than a ventral zone. Note: Despite being listed in the same LHA division, the LHAa does not receive an input from the BLAp

     269: RCH: Swanson and Kuypers 1980. nucleus supraopticus diffusus of Gurdjian ('27)

     270: TU: Canteras et al. 1994. TUte from Petrovich et al., 2001, derived from Paxinos & Watson 1986. The TUl is frankly parvicellular and obviously corresponds to the traditional lateral tuberal nucleus, including a small protrusion on the base of the hypothalamus (Nauta & Haymaker, 1969)

     271: LHAs - Cytoarchitecture similar to LHAjd, but there tend to be more neurons that are larger, and the cell density is somewhat greater

     272: LHAsfa, sfp - Goto et al. 2001, 2004. The vertical and horizontal limbs of this region have clear cytoarchitectural differences

     274: LHAm - Paxinos and Watson '86

References Cited:

Canteras, N. S., R. B. Simerly, et al. (1994). "Organization of projections from the ventromedial and tuberal nuclei of the hypothalamus: A PHAL study in the rat." J Comp Neurol 348: 41-79.

Goto, M., L. W. Swanson, et al. (2001). "Connections of the nucleus incertus." J Comp Neurol 438(1): 86-122.

Goto, M., N. S. Canteras, et al. (2005). "Projections from the subfornical region of the lateral hypothalamic area." J Comp Neurol 493(3): 412-438.

Gurdjian, E. S. (1927). "The diencephalon of the albino rat." J. Comp. Neurol. 43: 1-114.

Haymaker, W., E. Anderson, et al. (1969). The Hypothalamus. Springfield, Ill.,, Thomas.

Paxinos, G. and C. Watson (1986). The rat brain in stereotaxic coordinates. Sydney, Academic Press.

Petrovich, G. D., N. S. Canteras, et al. (2001). "Combinatorial amygdalar inputs to hippocampal domains and hypothalamic behavior systems." Brain Res Brain Res Rev 38(1-2): 247-289.

Swanson, L. W. and H. G. Kuypers (1980). "A direct projection from the ventromedial nucleus and retrochiasmatic area of the hypothalamus to the medulla and spinal cord of the rat." Neuroscience letters 17(3): 307-312.

Substance P Preview

SP 1:50k (rb; Cy2SAv). Image taken 1/27/13. 

General Procedure for Image Acquisition using Volocity

• clean the slide before use to avoid refracting light if necessary
  
  

  (gently wipe with chemwipe, do not use cleaning solution)

  
  
  

  
  

1. Turn on the scope, turn on the computer (windows)
2. Open Volocity once the scope AND camera are on.
3. Save as new library (2013_04_02 Volocity).
4. Calibrate the stage in the XY stage view:
1. lower the stage away from the objective.
2. stage-> calibrate stage
5. Look through the eyepiece of the scope to find the tissue under transmitted light and focus.
1. control of fluorescent shutter is on the computer as well as on the box #3
6. Starts off in black and white until you choose a channel to pass the light through
7. click on a channel
8. double click on the lime-green rectangular box (to the right of the snowflake) to open “Acquisition set-up”
1. change name (2013_04_02_001_primary_fluor)
2. select the number of channels
3. change focus using (NONE) if not taking z step
4. stitching
1. change XY using Ludl XY stage
2. dont correct for shading
3. auto align field
4. correct for brightness
5. save raw tiles
6. 12% overlap
9. make sure you keep the shutter closed as much as possible to avoid bleaching
10. click autoexpose to start to find the appropriate exposure (you can also change gain and offset to improve the signal)
11. click the floppy disk on the top right to save the channel settings
12. switch to the XY stage view and select an area, scan selected area to get a sense of where we are in the section
13. switch back to Live view to correct exposure
14. re-check the acquisition set-up
15. in XY stage view, select desired region and record

Thionin for Nissl Staining

From Jackson Bittencourt's protocol (Salk Inst., 1989)

A. Stock Solution

1. 1 M Acetic Acid (Add 60 ml glacial acetic acid (GAA) to 940 ml DI)
2. 1 N Sodium Hydroxide (40g NaOH/L)

B. Mixing the Stain

1. Add 36 ml 1 N NaOH to 200 ml acetic acid stock solution. Add DI to make a final volume of 1000 ml
2. Heat mixture to about 60º C. Add 2.5 g Thionin (for 0.25%) slowly while stirring
3. Bring Thionin solution to boil for 45 min
4. Let cool and store in a brown, well stoppered bottle at RT
5. Filter stain before use.

States that Thionin should be good for 4 months (but also says to store at 37º C...)

vGAT Preview

Another experiment in posting 'Original Size' images...

Single channel image of Cy 3-labeled rb anti-vGAT (1:10k) in the LHA. Original is 5 x 4 tile mosaic of 1200 x1200 images taken at 20X. The full-size (5443 x 4387) image was sharpened, cropped, converted to sRGB, and saved as a JPEG for upload.

Volocity: Exporting Channels

In a nutshell, the secret to exporting individual channels is to

1. Turn off all other channels (One of the buttons in the sidebar on the right, available when viewing the acquired image)
2. Capture a screen shot (Under the'Image' Menu select 'Capture a Screenshot...') using the 'Current View Size' option
3. Export the screenshot as an OME TIFF
4. Repeat for each channel

Here's how it looks using a 2 channel image of CCK in green and NeuroTrace Far Red in the red channel. These images were minimally processed. They were exported from Volocity and taken into PS 5 via Lightroom. In PS they were assigned and converted to a color profile and auto contrast was applied. The image was then re-sized to 1024x1190, converted to sRGB and saved as a JPEG. The images below are 1) a snapshot of the Cy 2-rabbit anti-CCK in the green channel, 2) a snapshot of NeuroTrace FRed in the red channel, and 3) a snapshot of the composite Red + Green channel. Note that the NeuroTrace was color changed in Volocity. Exporting works the same for mixed colors, but exporting the individual channels becomes more important. This is because channels in Volocity refer to data regardless of color, whereas in PS and everywhere else channels refer to color without regard for content. In other words, most colors use more than one channel  (e.g., Purple = Red + Blue, Yellow = Red + Green, etc.) making it more difficult to make selective changes to the component labels.

1: Green Channel ( Cy 2-rabbit anti-CCK)

2: Red Channel ( NeuroTrace FRed)

3: Red and Green Channel Composite

Nissl staining protocol

Thionin Staining Protocol as of 2/5/13
Protocol based on the 2003 procedure (on the wall in the wet lab)

Part 1: de-lipidation
1. 70% ethyl alcohol (EtOH) 2 changes, 2 min each
2. 95% EtOH 2 solutions, 2 min each
3. 100% EtOH 4 solutions, 2 min each
4. Xylene 3 solutions, 15 min each (under hood)

Part 2: rehydrating
1. 100% EtOH 4 solutions, 2 min each
2. 95% EtOH 2 solutions, 2 min each
3. 70% EtOH 2 solutions, 2 min each
4. 50% EtOH 1 solution, 2 min
5. deionized H2O (dH2O) 1 solution, 2 min

Part 3: staining and differentiation
1. dip in thionin dye, 30 seconds
2. dip in dH2O, 30 seconds
3. 50% EtOH 1 solution, 2 min
4. 70% EtOH 2 solutions, 2 min each
5. 95% EtOH 2 solutions, 2 min each
6. If overstained, differentiate in 95% EtOH with glacial Acetic Acid (1.5 ml gAA/250 ml EtOH)*
7. 100% EtOH 4 solutions, 2 min each
8. Xylene 3 solutions, 5 min each
9. Coverslip and dry flat

Dynorphin Works!!!!

(but Volocity still Sucks!!!)

Stitched image of  Cy 2-labeled rb anti-Dynorpnin A (1:10 K, 10 X, 11 x 6 x 2 x1). Image should contain NeuroTrace counterstain in the red channel. At this magnification, both intense labeling and high background can be seen clearly so the sections processed at 1:20k should be imaged and compared. Original file = 11800 x 6550 (or approx. 13½ x 7½ feet at screen resolution). Source G1111F S2s3.

With the idea of verifying the specificity of the label, a portion of the above section was also imaged as a z-stack at 20 X. Unfortunately the rendered projection which looked great in Volocity, exported as a blurry piece of crap. The densest label is still apparent, but the individual fibers are less distinct in the projection than at 10X. Note also slight differences in brightness of the component panels in the lower image. Both were stitch in Volocity without the 'shading correction' step. While this doesn't appear to be an issue at 10X, this shows we should use this option if it doesn't cause problems because it would be very difficult to correct in PS now that we are dealing with composite images, rather than component tiles...

Assembled image of Cy 2-labeled rb anti-Dynorpnin A (1:10 K, 20 X) stitched stack. Original image is 2 channel, 10 slice mosaic (7 x 4 x 10 x 2). In Volocity it looked fantastic but it will export TIFF as indexed color, not RGB, which is blurry & has lost the NeuroTrace FRed counterstain. This version was curved, sharpened heavily and shrunk, but that still doesn't completely hide the blur.