Fig 1.
Diet Switch (DS) feeding paradigm isolates ‘body weight’ from ‘diet and/or energy balance’ factors.
(A) Experimental design schematic. Mice consume either RC or HFD for 8–10 weeks (Pre-DS), then are randomized to either continue consuming their respective diet (RC NoDS and HFD NoDS) or switched to the alternate diet (RC → HFD and HFD → RC). All mice were sacrificed Day 7 Post-DS. (B) Absolute body weight across the study. (C) Absolute body weight at Day 7 Post-DS. (D) Percent change in body mass (ΔBody Weight) normalized to Pre-DS weights (Day 0 Post-DS). (E) ΔBody weight on Day 7 Post-DS (normalized to Pre-DS weight). (F) Caloric intake from Pre-DS (Day 0) to Day 7 Post-DS. (G) Average caloric intake on Day 7 Post-DS. *p<0.05, ****p<0.0001 comparing RC → HFD vs HFD → RC; a+: p<0.1, a: p<0.05, a’: p<0.01, a”: p<0.001, a”‘: p<0.0001 compared to RC NoDS; b+: p<0.1, b: p<0.05, b’: p<0.01, b”: p<0.001, b”‘: p<0.0001 compared to HFD NoDS. Sample sizes for B-E: [RC NoDS, n = 16; RC → HFD, n = 21; HFD → RC, n = 19; HFD NoDS, n = 14]. Sample sizes for F-G: [RC NoDS, n = 15; RC → HFD, n = 21; HFD → RC, n = 18; HFD NoDS, n = 14].
Fig 2.
Spontaneous locomotor activity is only reduced in HFD NoDS compared to all other conditions.
(A) Pre-DS locomotor activity shows HFD-fed mice are less active compared to RC-fed controls. (B) Locomotor activity counts in 2-hour time bins on Day 7 Post-DS. (C) Aggregate (24-hour) locomotor activity counts Day 7 Post-DS. All conditions are significantly more active than HFD NoDS. Note that body weight and food intake were measured daily between 9–10 am (ZT2-3), so locomotor data are omitted during the ZT2-4 time window. *p<0.05, ***p<0.001, ****p<0.0001 comparing RC v HFD at indicated time point; b: p<0.05, b”: p<0.001 compared to HFD NoDS. RC NoDS, n = 15; RC → HFD, n = 21; HFD → RC, n = 19; HFD NoDS, n = 14.
Table 1.
Linear modeling results comparing weight-matched DS conditions.
Table 2.
Differentially expressed genes (DEG) from RNA-sequencing analysis.
Fig 3.
RNA-seq results and RT-qPCR confirmation of VH-expressed genes is consistent between sequenced and non-sequenced tissue.
(Top row) Normalized read counts from RNA-seq analysis. (Middle row) RT-qPCR validation using tissue used for RNA-seq. (Bottom row) RT-qPCR validation using independent tissue samples not used for RNA-seq. (A-C) Agrp and (D-F) npy gene expression is lowest in RC → HFD and HFD NoDS compared to HFD → RC and RC NoDS conditions. (G-I) Cartpt and (J-L) trh gene expression is highest in HFD NoDS, lowest in RC NoDS, and low to intermediate in both DS conditions. agrp: agouti-related peptide; npy: neuropeptide y; cartpt: cocaine- and amphetamine regulated transcript; trh: thyrotropin-releasing hormone. *p<0.05, **p<0.01, ****p<0.0001 comparing RC → HFD vs HFD → RC; a+: p<0.1, a: p<0.05, a’: p<0.01, a”: p<0.001, a”‘: p<0.0001 compared to RC NoDS; b+: p<0.1, b: p<0.05, b’: p<0.01, b”: p<0.001, b”‘: p<0.0001 compared to HFD NoDS. Sample sizes for top row: [RC NoDS, n = 5; RC → HFD, n = 6; HFD → RC, n = 5; HFD NoDS, n = 5]. Sample sizes for middle row: [RC NoDS: n = 13; RC → HFD: n = 18; HFD → RC: n = 11; HFD NoDS: n = 7]. Samples sizes for bottom row: [RC NoDS: n = 5; RC → HFD: n = 8; HFD → RC: n = 6; HFD NoDS: n = 3].
Fig 4.
Glucose, insulin, and leptin levels from trunk blood plasma (Day 7 Post-DS).
(A) Glucose, (B) insulin, and (C) leptin levels measured from blood plasma, which was collected immediately after sacrifice (ZT6-7). ****p<0.0001 comparing RC → HFD vs HFD → RC; a: p<0.05, a’: p<0.01, a”: p<0.001, a”‘: p<0.0001 compared to RC NoDS; b: p<0.05, b’: p<0.01, b”: p<0.001, b”‘: p<0.0001 compared to HFD NoDS. Sample sizes for glucose: [RC NoDS: n = 16; RC → HFD: n = 21; HFD → RC: n = 19; HFD NoDS: n = 14]. Sample sizes for insulin: [RC NoDS: n = 13; RC → HFD: n = 17; HFD → RC: n = 12; HFD NoDS: n = 13]. Sample sizes for leptin: [RC NoDS: n = 13; RC → HFD: n = 17; HFD → RC: n = 13; HFD NoDS: n = 13].
Table 3.
Linear modeling results using body weight and Δbody weight as predictors from all animals.
Fig 5.
Plotting standardized correlation coefficients to visualize how blood biomarkers, VH gene expression, and behavioral data relate.
For each output metric, standardized β coefficients (Stdβxy) were calculated for both weight metrics (body weight and Δbody weight) and plotted together here. Any values to the right of the y-axis are positively associated with increased body weight, while any values to the left of the y-axis decrease as body mass increases. Any values above the x-axis increase as Δbody weight increases (positive energy balance), while points below the x-axis decrease as Δbody weight increases. Points in the shaded area represent non-significant associations. Exact Stdβxy and p-values are shown in Table 3. REM: Rapid eye movement sleep; NREM: Non-REM sleep; cartpt: cocaine- and amphetamine regulated transcript; serpina3n: serine (or cysteine) peptidase inhibitor clade A member 3N; sbno2: strawberry notch homolog 2; trh: thyrotropin-releasing hormone; agrp: agouti-related peptide; npy: neuropeptide y; T4: Thyroxine.