Seburn1 project protocol

Comprehensive metabolic survey of 16 inbred strains of mice   (2001)

Seburn KL




Body weight, caloric monitoring, activity and motor function, food and water intake

General. Experiments described here used a Comprehensive Lab Animal Monitoring System (CLAMS) (Columbus Instruments, Columbus, OH) that consists of individual live-in cages for mice that allow automated, non-invasive data collection. Each cage is an indirect open circuit calorimeter that provides measures of oxygen consumption and carbon dioxide production. The system also provides concurrent measurement of ingestive behavior (food, water consumption) and activity. A total of sixteen mice (8 male, 8 female) from each of 16 strains were tested. Groups of 8 individually housed mice were tested in each experiment. In all cases mice of the same gender were tested over the same 3-day period for each of the 16 strains. Data for male and female mice of a given strain were collected consecutively within a 7 day period for 11 of the 16 strains and within a 2-3 week period for the remaining 5 strains. All mice were maintained in the testing facility for at least 48 hours prior to the experiment. All experiments started between 8:30-10:30 am and continued for 3 consecutive 24-hour periods with a 12:12 light:dark cycle. Body weights were determined just before and after testing. Immediately upon completion of the experiments the mice were prepared for shipment to the laboratory of Dr. Thomas Hampton (MouseSpecifics, Inc., Boston, MA.) for non-invasive measurement of the electrocardiogram.

CLAMS records date and time-stamped data on a fixed sampling schedule for each mouse for multiple variables. In the configuration used for these experiments data for each mouse were recorded every 41 minutes such that each mouse had 105 samples recorded over the 72-hour period of the experiment. In all cases data were recorded over three transitions from light to dark and three from dark to light. The daily averages over 3 days of testing were submitted and are available unless specified otherwise. Further relevant details related to sampling are described in association with the different measurements.

Calorimetric monitoring. The cages in which animals were housed are indirect open circuit calorimeters. The system compares oxygen and carbon dioxide gas concentrations by volume at the inlet and outlet ports of the cage chamber through which ambient air flows at a constant rate (0.60 l/min). The difference in concentration between the two ports and the flow information is used to calculate oxygen consumption, carbon dioxide production and the Respiratory Exchange Ratio (RER). Heat production is also estimated with standard formulas using oxygen consumption and the RER. Data reported here were collected with an 8-cage system that utilizes single sensors for oxygen and carbon dioxide. Consequently, gas measures must be taken sequentially from each of the 8 chambers on a fixed cycle. The cycle time between samples/measures for any given chamber is therefore a function of the time required for the sample to reach the sensors and be measured and the number of chambers in operation.

Activity. Two arrays of infrared beams (2.5 cm inter-beam distance) surround each cage. One array is situated 3.2 cm above the floor of the cage and a second array at 7 cm above the floor. This configuration provided three measures of activity: 1) Total activity - any movement producing a beam break in the horizontal plane, 2) Ambulatory activity - movement producing sequential horizontal beam breaks of different beams (i.e. subset of total), 3) Rearing - produced by vertical movement when the mouse stands on it's hind legs to break elevated beam array. For these experiments beam breaks were monitored continuously and summed and recorded each time a metabolic measure was taken (i.e. every 41 min.).1 Activity measurements are expressed in normalized units of beam/breaks per minute.

Food and water consumption. A computer continuously monitors a food container that rests on an electronic weigh scale.2 The cumulative consumption is monitored and stored each time a metabolic measure is recorded.3 Individual water columns containing pressure transducers supplied drinking water to each individual mouse. Drinking by the mouse reduced pressure in the column and was converted to volume.



Investigator's notes

  • Mice were not acclimated to the CLAMS cages prior to the 3-day experiment. Pilot work showed that the initial response of mice to being placed in the cages and the recorded values were not significantly different if mice were acclimatized for 24 hours and then placed in the CLAMS 24 hours after acclimatization.

  • The selection of a 3-day paradigm was based on pilot work with a CLAMS prototype. This work showed that behavior during the initial 24-hours in the individual cages was characterized by a period of exploratory behavior and was generally dissimilar and more erratic in overall profile (activity, food and water consumption) than the 2nd and 3rd days that showed more consistent patterns. Data related to the exploratory phase of Day 1 will be reported in subsequent updates.

  • Consideration and comparison of these data should take into account that mice were individually housed without bedding. One of the consequences of this is that most mice lose weight when placed in CLAMS. The extent of the weight change is strain dependent and averaged less than -5% across all strains (range –11.4 to 3.3%). Further work in C57BL6/J mice showed that the weight loss occurred within the first 24 hours and then remained stable for up to five days (unpublished observation). The pattern of weight loss of other strains reported was not examined. Weight loss was reduced but not eliminated by increasing the temperature in the testing room to between 22-25°C (72-77°F). Other interventions, such as providing a nest, more easily accessible food and/or mixing of water with food, had no effect. An important consequence of this weight loss is, of course, the introduction of error into metabolic measures normalized by body weight measured at the beginning of the experiment. Although impractical, it is expected based on other work (Gordon et al., 1998) that weight loss could be reduced or eliminated entirely by raising ambient temperature to between 24-30°C (75-86°F).

  • Pilot work revealed that certain mouse behaviors could result in inaccurate metabolic values if they occurred at the time the sample was drawn from a particular cage. In particular if mice were feeding or were otherwise impeding airflow through the passage leading to the food receptacle the concentration of gases sampled would not be representative. This was reflected in non-physiological RER values (>1.1 or <0.5). Although RER values less than 0.7 are generally considered theoretically impossible we included them because we reliably saw such values in some strains. Moreover, such low values have recently been reported in birds and suggest the theoretical lower limit is likely an oversimplification (Walsberg and Wolf, 1995). Non-physiological values were filtered from the data and not included in the calculation of averages for the different epochs (i.e. Day, Light, Dark). Based on the sampling schedule 35 samples were recorded in each 24 hour period for a total of 105 for an entire experiment (~54 during light, ~54 during dark). Rather than making an arbitrary decision as to what number of samples represents a valid representation of metabolic data for any given strain during a given epoch we have reported the number of samples used to calculate the metabolic values reported for each of the epochs. Sample sizes are not displayed as measurements but are available in the project data set. Improvements in the CLAMS have eliminated or reduced this problem and when possible strains with significantly reduced samples (e.g. NOD/ShiLtJ) will be re-tested to confirm findings. Note that this limitation does not affect food, drink or activity values.

1 Because of this the correlation of activity with metabolic measures is poor because the latter measures are "snapshots" taken at the end of the period for which activity is summed.

2 Feeder design is such that spilled food is captured to avoid overestimating consumption and to minimize the amount of food the animal can remove from the receptacle for "storage" in their living space.

3 The measurement of water consumption was the measure most prone to inaccuracies. Mice frequently "played" with sipper tubes resulting in water loss not related to consumption. When this behavior resulted in clearly inaccurate data the values have not been reported. Also, mice were not acclimatized to the individual CLAMS cages prior to the experiment and frequently it would take some time before they began drinking. It was assumed that this was due to being unfamiliar with the drinking apparatus that was different than that of their home cages.