I am not familiar with how LPS influences the performance of mice in the
Morris Water Maze test, ofcourse this could be a possible explanation. Did
you also look into the following:
- Temperature of the water: with higher temperatures mice tend to show more thigmotaxis, the temperature should be around 25 degrees Celsius.
- Maybe you could perform a tail suspension or sucrose preference test to investigate whether all your animals or a specific group display a form of anhedonic behavior that could be linked to the thigmotaxis. If this is the case, one can at least exclude that the behavior is due to the experimental setup.
Thigmotaxis can also be an indicator of stress. If you have controlled for conditions as suggested by Felicia, I recommend you also check the housing conditions of mice. Irregular lighting, temperature, isolation or cramped up space can induce stress in mice.
I would agree with the previously provided answers, however I would inquire about the following: 1st what is the interval between injection of LPS/Control and water maze? 2nd you stated that you have controls, did their injections schedule match? If both Controls and LPS injected mice both display equal amounts of thigmotaxis, it is possible that the injection itself could be the issue not so much the LPS. Mice inherently more stressed and water maze results vary greatly between laboratories. The injections may be the cause of stress can you do the following to reduce the stress of injections: 1) give LPS or Control injections after maze, if it is daily injections or 2) if you could give injections well before testing e.g. 24/48hrs before start of maze. As for exclusion, did you have an a piori parameter for exclusion? If not than no they should not be excluded. A easy test would be to have a single day of visible platform performance after hidden platform and probe testing. Once an a prior exclusion factor is established, such as failing the visible platform test, only those animals can be excluded. As in all animal behavioral testing, consistency is paramount to good data.