Tips For Experiments

[Andrew Hardaway]

Hi,

We are close to launching FED3 oriented experiments. I've been running some 24 hour tests with each device just to get an idea of whether they work or not. Do you have any tips for running experiments? I got our machine shop to build some dividers for us and they work well. Happy to share those plans for anyone who use "OneCage". Are you still using those pads in lieu of cob bedding? That has been a slight issue so I've been just trying to maintain a thin layer of cob bedding so they don't pile it up. What about nestlet? Do you guys do an acclimation period before running an experiment? Anything else I haven't thought of?

thanks,

Andrew

[Lex Kravitz]

Hey Andrew!

This is a great idea to start a "tips" thread.  Hopefully others can chime in but I'll share our experience and tips.  The tips depend on the experimental design.  For using FEDs as a simple operant box replacement we've been mounting them on the side of these $20 plastic boxes from Amazon.  We just drill holes in the side and mount the FEDs on them:

https://www.amazon.com/Clear-Plastic-Hanging-Organizer-Handles/dp/B002JLTMR0/

This is great if you're doing 1-2 hour tests, we experience few issues with jamming, etc in these short sessions.  For overnight/multiday studies it's more challenging to keep things running smoothly in a home-cage environment (although we've done >2 week experiments so not impossible, just harder!).  Here's an example of 10 FEDs going for >2 weeks of free feeding:

For multi-day homecage experiments the important things we found to do were:

1) Use "iso-pad" bedding for overnight sessions.  Regular corncob bedding pieces can get kicked into the pellet well making it think there's a pellet and not dispensing unless the mouse removes it:

https://www.braintreesci.com/prodinfo.asp?number=ISO

We have run experiments with a thinner layer of regular corncob bedding as well, it's usually fine but it's a gamble.

2) Check the FEDs daily and make sure the pellet path is clear and working.  We use compressed air and spray it into the pellet path from both sides.  We try to do this every day, sometimes skip a day but it's important to keep dust from accumulating.  We also check the function of each FED every day by dispensing a single pellet and making sure everything looks smooth.  If you hear pellets grinding take the pellet disk out and blow any pellet crumbs out with compressed air until it dispenses smoothly.

3) If you're interested in measuring calories you can also weigh the FED itself daily to get another measure of total food removed.  With free-feeding, we find the discrepancy between the weight removed from FED and the estimated food intake from pellets is very low, ~5%.  Here are ~50 measurements of the weight of FED going down every 1-2 days (x-axis), vs. the estimated weight of pellets removed (# of pellets * 20mg, y-axis).  This discrepancy should always be an undercount BTW, as FED can dispense 2 pellets and only count 1, but it never counts a pellet if nothing is dispensed. Depending on how much you care about this stuff you may want to do these calibrations for all of your FEDs, particularly if you're 3D printing and tweaking the parts yourself.

With FR1 sessions we get higher error in this, ~10%, which I think may be due to mice being eager to get the pellet in the FR1 trials and stealing the pellet before it's detected by the pellet well, causing FED to drop a second pellet.  We've tried to minimize this with modifications to the pellet well but the mice are tricky!  Anyway point is, if you care about calories (and not just active/inactive pokes) you should weigh the FEDs and get an idea of error rates.

4) Look out for "hoarding" behavior - ie: pellets left over on the floor the next day.  Hoarding happens with all feeding systems, but if you care about calories you'll want to count these up as well.  Interestingly, when we do overnight FR1 sessions the hoarding completely goes away, so this is a behavior they only seem to do when food is freely available.  We also see hoarding behavior when we weigh food intake of regular chow using little dishes. 

5) We acclimate mice to FEDs before starting any real experiments.  Usually one night of free feeding will be sufficient for them to learn to eat from FED, even with regular chow in the cage.  And then one night of FR1 is sufficient for them to get pretty good at the active/inactive discrimination.  We do acclimination with regular chow pellets still in the cage so there's no risk of starvation if they don't figure it out.  This also gives us a chance to make sure the FEDs are all working for a couple days before starting any real data collection.

6) We do use Nestlets. I don't know if they're good or bad but they don't seem to cause issues. Sometimes they make their nest right in front of FED and when they do this I try to move it to the back of the cage away from FED.

7) Using an in-cage "insert" is also helpful for keeping FED clean and improving reliability.  We are 3D printing these inserts and combining with a laser-cut aluminum front. 

We get the laser cuts done here: https://www.ponoko.com/.  It looks super nice and the aluminum is easier to clean than the 3D printed material, photos here:

https://hackaday.io/project/106885-feeding-experimentation-device-3-fed3/log/175230-cage-insert-to-protect-fed3

Hope this helps, and any other tips people have would be appreciated!

[zane andrews]

Hi Everyone,

I just wanted to add a couple of thing to Andrew and Lex - we are using FED3s in the animals home cage with custom dividers. It works really well but we have found that animals are very sensitive to different FED3s in their home cage. We only have 1 experiment at a time and the same mouse gets the same FED3 the whole entire experiment. We dont run two cohorts with the same FED3s at the same time. Its also the best way to go back and collect data from the SD Cards.

We also couldn't get pellets during the lockdown so we sourced some homeopathic tablets (they worked really well and were cheap). Has anyone looked into non-caloric pellets?

We have managed to use create different contexts in the home cage to have an ABA model (FR acquire in Context A, Extinction Context B, Reinstatement Context A (which is on extinction mode but 2 pellets dispenser at start). To create different contexts we have used different odours, bedding and housing, and varied the position of the FED3. You can see nice 4-5 fold increase in responding when moved back to context A.

Cheers

Zane  

[Matias Andina]

Hi,

I am not completely sure this is the best thread but I was thinking whether it would be a good idea for people with different printers to share their G-code or modifications. Now that the repository in GitHub contains a lot of centralized info, it looks like a good place to do so.

I was also wondering whether people have imported the .stl files into a body object that can be better used to use the FED in other renders (using SolidWorks, Autodesk, ...). I have the housing mesh as object in solidworks (happy to share) but the direct import keeps the mesh format (a gazillion triangles) and might not be the best to move it around or align it with other things in the render. An actual solid object would come super handy when incorporating the FED into another setup that has to be custom built (or building different types of divisors).

Best,

Matias

[Lex Kravitz]

Matias this is a great idea! Do you want me to add you to the Github as a contributor so you can post your files, hopefully others will follow?   We could also make a separate page in the Wiki for these files?

[Matias Andina]

Sure, matias...@gmail.com is what I use for GitHub.

--
You received this message because you are subscribed to the Google Groups "FEDforum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fedforum+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fedforum/39865dfb-7bf3-4e67-a5aa-1a526550076an%40googlegroups.com.

[Matthew Buczynski]

Really appreciate all the feedback list this from the FED community, and happy to help contribute in some small way. I'll try and put everything in the appropriate conversations (tips for printing, building, etc). Much of our approach supports what has been said here already.

Bedding: We are having much more success with Iso-Pads than other forms of bedding.

Dividers: VT doesn't have a machine shop, so we made dividers using a filament printer, including a design with embedded magnets. The idea was to make them with nylon filament (which has been unsuccessful to date) make a dishwasher-safe version. Printed dividers seem to work fine. In addition

Group Acclimation: Although we haven't rigorously analyzed the data, we've found that group housing the mice with a FED in free-feeding mode for 48 hrs seems to reduce the number of mice that fail to acquire self-administration behavior during the first 1-2 days.

Cleaning Pellet Path: As mentioned by Lex, we used compressed air to clean out debris from the pellet path every 1-2 days during continuous use. We also sift the pellets using a wire mesh to help clear the debris and dust that accumulated before putting pellets into the FED.

Battery Swapping: As Zane mentioned, we try give each mouse the same FED throughout an experiment. Instead of waiting for them to charge up, we just swap out drained batteries with newly charged ones every 3-5 days during the pellet cleaning. We found that repeated pulling on the connectors on the battery and power button cable (blue arrow) eventually pulled out some of the wires from the connector. This was eliminated by adding a small amount of epoxy to the wire-connector junction (green arrow) before use.