The Use of Filters in the Sump for Monitoring the Health of Laboratory Zebrafish (Danio rerio)
Early detection of pathogens is imperative for the health of laboratory zebrafish and to ensure reproducible scientific results. While most pathogens are present as subclinical or chronic infections, their presence can be a confounding factor in data collection, and some infections can affect zebrafish health and reproduction. Current methods to test for pathogens sample myriad sources, including sump swabs, detritus, water collection or filtration, and whole-fish PCR or histopathology. Sentinel mice have been used in the past for mouse health monitoring, but there has been a recent shift to replace sentinel animals with filters in rack exhausts. We wanted to determine whether such methods could be translated from mouse racks to zebrafish systems. First, we tested whether nitrocellulose filters would affect the health and behavior of adult and larval zebrafish and found no adversary effects. Next, we placed filters in the sumps of zebrafish racks to be collected and tested for pathogens at monthly intervals using PCR, and results were compared with those detected on filters in which water was actively vacuum pumped through or swabs of sump biofilm. Results suggest that the efficacy of filters may wane with prolonged use, with sensitivity being greatest at 60 d and then decreasing at 90 d. Results also showed the limitations of current testing methods for zebrafish health monitoring, with efficacy of detecting pathogens varying widely based on the method of collection. Our recommendation is to test the filters after 6 wk of exposure, and supplement with whole-fish testing of clinically ill fish.
Example of filter assemblies placed in fish rack sumps. Depicted is a nitrocellulose (Nitro) filter, stabilized with a black, polyethylene plastic holder apparatus contained within a mesh polyethylene biofilter bag retrieved from a sump after its submersion period.
PES vacuum filter apparatus.
Box-and-whisker plots of larva body length (A) and head length (B) measured during experiment 1. For each measurement, larva exposed to nitrocellulose (Nitro) filters are compared with controls, which were not exposed to any filter. Significant equivalence results (where present) are represented as follows: §, P ≤ 0.0001; +, P ≤ 0.005.
Box-and-whisker plots of larva performance on visual motor response assay for light cycle (A) and dark cycle (B) measured during experiment 1. For each measurement, larva exposed to nitrocellulose (Nitro) filters are compared with controls, which were not exposed to any filter. Significant equivalence results (where present) are represented as follows: §, P ≤ 0.0001.
Box-and-whisker plots of larva body length (A) and head length (B) measured during experiment 2. For each measurement, larva exposed to nitrocellulose (Nitro) filters are compared with controls, which were not exposed to any filter. Significant equivalence results (where present) are represented as follows: §, P ≤ 0.0001.
Box-and-whisker plots of larva body length (A) and head length (B). In this case, larva exposed to nitrocellulose (Nitro) filters on a chronic basis were compared with the G2 of control larva, which were collected at a comparable time since birth and not exposed to any filters. Significant equivalence results (where present) are represented as follows: §, P ≤ 0.0001; ×, P ≤ 0.0005.
Box-and-whisker plots of larva body length (A) and head length (B). In this case, G1 offspring of fish that were chronically exposed to nitrocellulose (Nitro) filters were compared with the G1 offspring of control fish, which were not exposed. Significant equivalence results (where present) are represented as follows: §, P ≤ 0.0001; ×, P ≤ 0.0005.
Box-and-whisker plots of embryo body length (A) and head length (B). In this case, G2 offspring of fish that were chronically exposed to nitrocellulose (Nitro) filters were compared with the G2 offspring of control fish, which were not exposed. Significant equivalence results (where present) are represented as follows: §, P ≤ 0.0001.
Contributor Notes