Tuesday, 27 May 2014

Squeaking out that somethings wrong

Sheridan GK, Dev KK. Targeting S1P receptors in experimental autoimmune encephalomyelitis in mice improves early deficits in locomotor activity and increases ultrasonic vocalisations. Sci Rep. 2014;4:5051. doi: 10.1038/srep05051.

Fingolimod (FTY720) is an oral therapy for relapsing remitting multiple sclerosis (MS) and targets sphingosine 1-phosphate receptors (S1PRs). FTY720 also rescues animals from experimental autoimmune encephalomyelitis (EAE), an animal model of MS. The protective effects of FTY720 in EAE are primarily scored manually by examining weight loss and limb paralysis that begins around 10-12 days after immunisation. To our knowledge, pre-clinical effects of FTY720 on animal behaviour early in EAE have not been explored. Here, we developed an automated behaviour monitoring system to examine the early effects of FTY720 on subtle pre-symptomatic behaviour of mice induced with EAE. Our automated home-cage monitoring system (AHC-MS) enabled non-contact detection of movement and ultrasonic vocalisations (USVs) of mice induced with EAE, thus allowing detection of subtle changes in mouse behaviour before paralysis occurs. Mice receiving FTY720 emit longer USVs and display higher levels of motor activity than vehicle-treated EAE mice before clinical signs become apparent. Importantly, this study promotes the 3Rs ethics (replacement, reduction and refinement) in the EAE animal model and may also improve pre-screening of potentially novel MS therapies. In addition, this is the first report showing the early effects of FTY720 in EAE which underscores its protective effects.

The use of EAE is increasingly coming into the spotlight and is considered a substantial (compared to mild and moderate) procedure. 

There is an ethical case to be made that you may not really need to allow animals to become paralysed, as a read out for experiments. For example studies on immune function using knockouts where you use paralysis as a readout for gene activity.  You could simply do your test tube experiments and say record time to onset of pre-clinical weight loss, development of the first sign, etc....rather than letting paralysis occur. In this current study you could use vocalisations. However, do this and you run the risk of the referee pool from regions who are not so bothered about the 3Rs (not reading riting and rithmatic...but refinement, reduction and replacement) of animal use, who may bin your work.

However, just because in the past it was done, does not mean that it should be done. The 3Rs is now the corner stone of Animal experimentation in Europe. Times are changing.

It is becoming increasingly difficult to do such experiments in the UK. 

In this study they designed a monitor to see what the mices were doing and in the first few days after induction with an adjuvant (stimulates the immune system) and whooping cough toxin, there was perhaps a sickness behaviour where they did not move as much and they were off their water. The toxin makes the number of white blood cells grow ten times to make them get EAE, we don't use it. This no doubt is associated with immune growth factors which may you under the weather.

They used a drug FTY720 which is the ingredient drug in Gilenya/fingolimod to stop EAE and they found that these animals gave longer squeaks. We can't hear it because it is ultrasonic.

The mice destined to get EAE did not squeak as much..

ProfG has had loads of whacky ideas about how to monitor movement of mices, but this system was all automatic and so unbiased and provides constant monitoring. The camera and Bat sound detector went onto a standard cage, which is what you want. There are systems that can monitor movement that can be implanted like a microchip for dogs and cats that can detect movement and temperature 24/7 (when a mouse gets EAE it moves around less and gets colder) but the system only worked for one mouse per cage and required a special cage and was very expensive, the manufacturers at the time were not interested for developing a system that would track multiple animals in a cage as mice are social animals and prefer to live in groups. This system also appears to measure collective responses, at present (at least in sound), where you would want individual readings as (a) we currently randomise within cages, rather than having all animals in one treatment in CageA and the other Cage B (b) Not all animals develop disease at the same rate.

The authors "do not see this system as a complete replacement to the current clinical scoring methods (which can check on individual animals), but rather an additional tool that can be used alongside gold standard routines to assess beneficial drug responses in EAE". 

We have shown that the gold standard clinical scoring routines are less responsive to looking at recovery and have shown that animals move less as disease progresses. The merit may be a way to look at progressive deterioration over time and because of its constant monitoring small changes in loss of movement may be amplified, meaning experiments don't have to be as long...saving money on (mouse) hotel fees and less time in procedure for the beasties.

This is the type of approach that CRACK-IT aims to investigate. This is a NC3Rs programme aimed at developing practical approaches to address important issues to Industry, that have 3Rs potential.
Wonder how much the system costs?


  1. The idea is so obvious, i thought everybody does this. This should be pretty cheap as video motion detection sw ether free or commercial is so cheap and available this days, though tracking multiple objects must be more complicated but still not a serious tech task I believe

    P.S. every time Prof Mouse posts a mice or rats photo, I get the desire to acquire a pair for myself hehe

    1. Start with a pair and three week later you have an extra six to twelve mouths to feed and after 6 weeks they bred.

    2. Prof Mouse, there are some open-source libraries/projects on motion detection in video stream out there. If you are interested in this I possible could build a very basic testbed in my spare time to see what could be done with them in relation with application you mentioned and estimate how much effort would be needed to make it produce any useful data.

    3. There are a number of commercial cameras that can sense multiple dots but the challenge is to have animals (5-10) in a home cage ( eg M2 or M5 cage) and record multiple cages and I guess working in a light dark cycle however working in alight cycle would be of value.

    4. Hi Prof Mouse
      If infra-red ilght is suitable for animals and will not influence the experiment outcome somehow, than it’s not a big issue as ir-backlighting camera can be used for 24h

      In commercially available cameras or controllers there is no readily available way to quantify amount of motion they recorded (but you need to quantify it somehow, that is the data in question). I’m looking at Axis.com solution right now and it does not seem to be readily suitable for that kind of application, but it can be pretty easily adapted (e.g. the simplest solution — it can take snapshots every time the amount of motion exceeds the threshold, and than you can do simple analysis of those huge amount of snapshots off-line, etc).

      I agree, main challenge is to be able to track and generate data set for each individual of multiple beasts sharing a single cage, as you randomise inside the cage. And you do need need an individual data set for each mouse. This would require markers and and some kind of image recognition atop of that. Currently I got no idea on how this could be managed but possible there is some cheap/quick solution on that already need to look into that.
      For a single animal inside the cage problem does not seem to be so complicated. Also, if the issue above could be sorted out in economically-effective way, multi-cage scenario should not be a problem, simple add a camera to any other cage and mark animals somehow.

    5. Infrared I don't know some animals sense this, but I believe vision is not major sense for mice and quiet a few strains of mice are naturally blind and seem to do OK.

      There are cameras that can sense and track individual colours, one could I think give some colour to the fur, it is done with sheep. There must be a safe hair dye

    6. Yes, thats was the idea.
      But this would not work with IR camera (so no data on dark cycle).
      Though I think i got an idea how this could be sorted out, will test on available equip if this feasible

  2. MD what are you going to do now that EAE is being killed off?

    1. Evolve and move forward or start to pack shelves at the local supermarlet(:-)


Please note that all comments are moderated and any personal or marketing-related submissions will not be shown.