Fig 1.
A Standard IVC cage position equipped with DVC sensing technologies. The grey board contains the 12 electrodes, while the orange runners have embedded IR sensors for food and water bottle detection. Electrodes numbering is shown in the right panel; B shows example of raw signal measured from electrode 5 when a mouse stays for T1 seconds in a cage area close to electrode 7, then moves towards and across electrode 5 during T2 seconds and then moves on to electrode 3 staying there for T3 seconds.
Fig 2.
Day and night time activity at CNR, JAX and KI.
A, C, E. Heat maps showing average global activity (all 12 electrodes) recorded from 5 cages with 5 female C57BL/6J in each cage, during 4 consecutive weeks (day 1 through 28 on the ordinate) at CNR (A), JAX (C) and KI (E); activity coded in color according to scale on the right side of the panels A, C and D. Each panel is 1 week; with days as rows on the ordinate and time h on the abscissa. To the left is day break with lights-on and white vertical line indicates transition to night time with lights-off. Cages were subjected to daily check-up which are done without removing the cage from the rack. Two more intrusive procedures are conducted weekly with a cage-change on Thursday and a second handling for weighing with health assessment on Monday; both procedures require cage removal from the rack, transferring to a changing station, opening of cage and removing animals from the cage; i.e. manual handling of each animal. The general day and night rhythm of activity alterations as well as the impact of procedures replicates well across sites. A main difference between sites was a generally higher level of activity (p<0.001) after at CNR (A) compared with both JAX (C) and KI (D) and also a slower increase in activity at lights-off at KI (cf. A-C-E and see also Figs 5 and 9). Note also that the response to lights-on starts prior to the event and last for an hour following lights-on while the response to lights-off occurs after darkness. Use of 30 minutes of dawn and dusk at light-on/off, i.e. slow light change transition, did not significantly alter the responses (not shown). B, D, F Actigrams showing global (all 12 electrodes) average activity per hour of the cages shown as heat maps in A, C and E (day 15 to 22). Each bin indicated by dotted vertical lines display the night time (12 hrs. of lights off) activity superimposed on day time activity (12 hrs. of lights on). The grey-shaded bin is day of cage-change. Abscissa denotes age of animals (in days) and ordinate average activity. The blue trace is average per h (of activity per minute) daytime activity (each bin starts at lights-on until lights-off) and shaded blue areas denotes the standard deviation. The red trace is the corresponding night time activity (each bin starts at lights-off until lights-on) i.e. with a 12 hrs. phase shift to the red traces, with the standard deviation indicated as shaded red area. As evident from the heat map and plots there is a significant difference between activity during day and night time at There is also a difference in activity between sites day and night (p<0.001).
Table 1.
Table of animals used in the different recordings analyzed.
Table 2.
Cage content and vivarium parameters.
Table 3.
Vivarium parameters continued.
Fig 3.
Activity following lights-off at CNR, JAX and KI.
A-C Average response per cage across weeks to lights-off in cages (n = 5; individual traces in color) of female mice housed at the three sites. Traces start at light-off (abscissa, minutes) and ends at lights on (dark-cycle is 720 minutes). Each trace corresponds to one cage and is the average activity across multiple weeks normalized to peak activity (= 1.0). The plots deriving from the different sites show clear differences with a short time-to-peak latency at CNR (Table 4), while mice at JAX (B; Table 4) show a bi-phasic response with an early short latency time-to peak followed by a decrease in activity and a second peak at about 400 minutes. At KI (C, Table 4) there is more protracted onset of activity and the time to peak matches the second peak observed at JAX, i.e. 400 minutes after lights-off. The late-night pattern of activity, >400 minutes, show less difference across sites (see also Fig 8).
Fig 4.
Day and night time activity at CNR, JAX and KI.
Box plots (cages) showing global activity (ordinate) of the five cages during day time (A, D, G) and night-time (B, E, H) across 11–13 weeks for each day of the cage-change cycle (abscissa; 0 being day of cage-change). Weeks at each site are presented separately for each day of the cage-change cycle. Asterisk denote day of cage-change while double asterisk indicates day of weighing. At all three sites there were significant variation in day-time activity between weekdays (p<0.001) and also between weeks at each site (p<0.02–0.001), however, the week-to-week variation in activity did not co-vary with the variation of activity between weekday (weekday*week p>0.05).
Table 4.
Response to lights off.
Fig 5.
Fraction of total (24h) activity occurring during day time and night time, respectively, at CNR, JAX and KI.
A, B and C show fraction (%) of total day (24 h) activity occurring during lights-on and lights-off, respectively, at the three sites. The distribution of activity between day and night time is highly significant (p<0.001) and quite similar across sites.
Fig 6.
Day and night time activity of female and male mice at KI.
Box plots (cages) showing normalized (see Material and methods) global activity (ordinate) in cages with female (A, B; n = 5) and male mice (C, D; n = 4) housed at KI, over several weeks (9–11), for each week day across a cage-change cycle (abscissa; 0 day of cage-change) for day-time (A, C) and night-time (B, D) activity, respectively. Asterisk denote day of cage-change while double asterisk indicates day of weighing. For both sexes there is a significant variation with weekday of the cage-changing cycle (p<0.001) and for the female but not the male mice across weeks (p<0.001), respectively. Analysis of effect of sexes and weekday on activity show that female mice show an increased activity during the day of the procedure (cage-change) while male mice show a significant impact by the procedure lasting several days both in day and night time activity. The difference of day and gender was significant at p<0.001.
Fig 7.
Distribution of activity during day and night time across the cage floor area at CNR, JAX and KI.
Actigrams showing average activity in cages (A-D) during lights-on and lights-off in the front, middle-front, middle-rear and rear rows of electrodes at the three sites. Continuous activity (ordinate) data during one weeks (abscissa; age of the animals in days) from the cages with female mice at CNR (A), JAX (B), and KI (C) followed by the cages with male mice at KI (D). In A-D grey shaded area indicate day of cage-change. As in Fig 2 the blue trace (plotted as average per h based on activations per minute) is the average daytime activity (each bin starts at lights-on, and ends at lights-off) and shaded blue area denotes the standard deviation. The superimposed red trace is average activity during lights-off (each bin starts at lights-off and ends at lights-off; 12 hrs. phase shift from the daytime data) with the standard deviation indicated as red shaded area around the average trace. Note the overall lower level of activity recorded from rear area of the cage floor across sites and sexes. The level of activity day and night time is clearly lower in cages with male mice compared with cages holding female mice at KI. For further information, see text.
Fig 8.
Impact of cage-change at CNR, JAX and KI.
A, C, E and G Heat maps of single cages across multiple weeks (ordinate, each line represents a cage-change of sequential weeks; n = 9–15) showing the response to a cage-change (ordinate shows activity and abscissa time; the cage change start at time 0) in female mice (A, C, E) at the three sites and male mice at KI (G). Color coding according to scale to the right of each panel. B, D, F and H are plots of average response sampled across weeks for each cage (n = 5) with female mice at the three sites (B, D and F) and male mice housed at KI (H). Number of repetitive observations (weeks) are 15 at CNR, 3 at JAX, 12 for female mice at KI and 8 for male mice at KI. Peak- (I, L, O and R) and average activity (J, M, P and S) during the response duration (FWHM; see Material and methods) as well as the duration of the response (K, N, Q and T) were calculated from plots shown in B, D, F and H. Top row of panels show data collected from CNR, row two data from JAX, row three from KI and row four shows data from male mice housed at the KI site. Depending on the small number of observations at the JAX site it was excluded from the statistical calculations. The difference in average activity (p<0.01) and duration (p<0.001) of response (FWHM) to cage-change among female mice was significant between CNR and KI. Within sites there is a significant variation in peak-, average-activity and duration of response across weeks (p<0.01–0.001). Among male mice house at KI there was only a small difference in average activity across weeks in the response to cage-change (p<0.05). Peak- and average-activity, and response duration to cage-change were all significantly different between male and female mice housed at the KI site (p<0.01–0.001).
Fig 9.
Impact by lights-on on activity at CNR, JAX and KI.
A, E and I Plots of average response of each cage across 8–13 weeks during each day of a cage-change cycle (0 is day of cage-change) with female mice at CNR (A), JAX (B) and KI (C). Peak-(B, F and J) and average activity (C, G and K) during the response duration, and the duration (FWHM) of the response (D, H and L) were calculated from such plots. Among the female mice the peak and average activation response (site factor, p<0.001) but not the duration of the response (p>0.05) were significantly different between sites and across weeks of observations but not weekday of the cage-changing cycle. Within sites peak- activation varied across weeks (p<0.03–0.001) while only at the KI site there was a significant variation in both peak- and average activity response with day in the cage-changing cycle (p<0.01).
Fig 10.
Impact by lights-on on activity of male mice at KI.
(A) Plots of normalized (see Material and methods) average response of each day during a cage-change cycle (0 is day of cage-change) sampled across 8 weeks from cages with male mice at KI (see also Fig 9C for female mice at KI). Peak (B) and average (C) activity during the response duration and the duration (FWHM) of the response (D) of male mice (computed from plots shown in A) show that the peak- and average activity response varied with day in cage-change cycle (p<0.001) but not across weeks (p>0.10). While female mice (Fig 9J, 9K and 9L) show only small day-to-day variations following cage-change (p<0.01) and larger differences across weeks (p<0.001), the males show an increased response to lights-on the day after cage-change followed by a decreased response for 1–2 days until return to base-line response (day 5; p<0.001) which does not vary significantly across weeks (p>0.10). Statistical analyses revealed that both the peak- and average activity response is different between sexes (p<0.001) and that each sex has a different response across days post cage-change (p<0.001). In contrast, the duration does not show a distinct sex or weekday dependence.