High Fructose Diet inducing diabetes rapidly impacts olfactory epithelium and behavior in mice

Type 2 Diabetes (T2D), a major public health issue reaching worldwide epidemic, has been correlated with lower olfactory abilities in humans. As olfaction represents a major component of feeding behavior, its alteration may have drastic consequences on feeding behaviors that may in turn aggravates T2D. In order to decipher the impact of T2D on the olfactory epithelium, we fed mice with a high fructose diet (HFruD) inducing early diabetic state in 4 to 8 weeks. After only 4 weeks of this diet, mice exhibited a dramatic decrease in olfactory behavioral capacities. Consistently, this decline in olfactory behavior was correlated to decreased electrophysiological responses of olfactory neurons recorded as a population and individually. Our results demonstrate that, in rodents, olfaction is modified by HFruD-induced diabetes. Functional, anatomical and behavioral changes occurred in the olfactory system at a very early stage of the disease.

after two-ways ANOVA followed by Fisher's LSD post-hoc tests. (3)(4)(5)(6) EOG kinetics at 1:1000 dilution in mineral oil. Bar graphs represent mean values (±SEM). ***: p < 0.001 after Student's t test. (3) Area under curve. (4) Rise time (time needed between 10 and 90% of the maximum response). (5) Fast decay time (between 100 and 80% of the maximum response). (6) Slow decay time (between 40 and 20% of the maximum response). representing the number of action potentials elicited by 7pA current (b3), the latency between the onset of the stimulus and the first spike (b4) and the average interval interspike of the spike train elicited (b5). All recordings performed in perforated patch and at a membrane potential of -67mV.
Data represented as mean ± SEM for neurons from control (white bars) and HFruD (dark bars) animals. *: p < 0.05 after Student's t test.     Glucose tolerance test and insulin assay. Intolerance to glucose was assessed using intraperitoneal glucose tolerance test (ipGTT) instead of oral glucose tolerance test because HFruD modifies glucose absorption at the intestinal level 2 . For this experiment, mice were placed in a non-inverted cycle and food was removed 5 hours before the beginning of the test, which took place during the light phase of animals. Glucose was then injected intraperitoneally (2g / kg body weight). Blood glucose was then determined with an Accu-Chek Performa glucometer (Roche) from blood sample collected from the tail vein, before and 15, 30, 45, 60, 90 and 120 minutes after glucose injection. Blood samples were also collected just before glucose injection to perform quantitative analysis of insulin level. Insulin serum levels were determined using an alphaLISA Human insulin immunoassay kit (Perkin Elmer). For ipGTT, a two way ANOVA was performed followed by Fisher's LSD post-hoc test. Areas under curve were analyzed using a Student's t test and glycemia and insulinemia using a Mann-Whitney's test. All statistics were performed using Statistica software.

Histology.
Flat mount preparation. Septal olfactory epithelia were dissected in a Ringer solution as described earlier 3 . One part of them was mounted in a perfused chamber and visualized using an Olympus  For OMP staining, total area of OMP labelling was quantified at similar antero-posterior coordinates in the nasal cavity (as described above for PCNA). Area was expressed in relative values of total olfactory epithelium area. This OMP area is correlated with the number of OMP-positive cells, as previously reported 7 as well as in our conditions (r² = 0.91, p < 0.05, data not shown).
The olfactory epithelium thickness was measured perpendicular to the lamina propria on the septal epithelium. Four different measurements were performed on each picture, and 4 pictures were used for each animal. The average thickness was then calculated for each picture and each animal and expressed in µm.
All statistical analysis were Mann-Whitney's tests performed with Statistica software. All images were taken and all analyses were performed blindly of the experimental groups.
Electroolfactogram. EOG recordings were performed from the olfactory mucosa in an opened nasal cavity on mouse hemiheads as described earlier 8 20 is composed of a test area of 40 x 40 x 10 cm (length x width x height), with in its center a hole. A beaker containing a small piece of odorized paper covered by a metallic grid and 1 to 2cm of litter is placed under the hole. Mice were first train for 3 days prior testing to explore the apparatus while mineral oil alone was added on the paper. This training period is essential to reduce the exploration time of the animals placed in a novel environment 21 . On the test day, animals were submitted to 6 consecutive trials of 3 minutes. Animals were allowed to rest in a cage for 4 minutes between each trial (another mouse was performing the task in the meantime). Open field test. Mice were placed in the same apparatus as for the habituation/dishabituation test, except that no hole was present in the test area. Mice were tested two times for 5 minutes and allowed to rest for 6 minutes between the two trials. Animals were filmed during the test (conducted in red lights). Videos were processed using a custom Matlab routine that tracked mouse centroid positions during 5min. Open filed test scoring was completely automated. Total walking distance, average speed and thigmotaxis (expressed as the percentage of total time spent close to the walls) were measured.
Buried food test. The buried food test 22,23 , which relies on the animal's natural tendency to use olfactory cues for foraging, is usually used to confirm ability to smell volatile odors. Here we modified a protocol from Yang and Crowley 19 in which we chose Leerdammer® cheese as the food stimulus instead of a chocolate cookie (HFruD animals being used to a high-sugar flavor) . Animals were habituated to cheese on two consecutive days before the actual test by placing a little piece of cheese in their home cages and check that it has been eaten on the following morning. On test day, food access was removed 5 hours before the experiment. . Animals were placed in a conventional mouse cage (33 x 19 x 13 cm) filled with 6 cm of bedding. Food or control item (a pebble of equal size as the piece of cheese) was buried at one randomly chosen spot (out of 8) in the cage. Test ended once either the animals found the item or did not achieve to find it within 10 minutes. The criterion used to stop the experiment was when the animal held the item with forepaws or with teeth.
Animals were tested for 3 consecutive trials (food, control then food item). Only the second trial with food was used to analyze the results. To test for motivation of animals to search for food, we tested the ability of animals to retrieve a visible food item on the bedding in a conventional rat style cage (42 x 26 x 19 cm). The criterion to stop the experiment was when the animal started to eat the cheese.
During data analysis of all behavioral assays, experimenters were blind to the experimental groups.
All statistical analyses (using Statistica software) consisted in two way ANOVA followed by Fischer's LSD post hoc tests except for the visible cheese experiment which was analyzed using a Student's t test.