Loss of Brain Angiogenesis Inhibitor-3 (BAI3) G-Protein Coupled Receptor in Mice Regulates Adaptive Thermogenesis by Enhancing Energy Expenditure

Effective energy expenditure is critical for maintaining body weight (BW). However, underlying mechanisms contributing to increased BW remain unknown. We characterized the role of brain angiogenesis inhibitor-3 (BAI3/ADGRB3), an adhesion G-protein coupled receptor (aGPCR), in regulating BW. A CRISPR/Cas9 gene editing approach was utilized to generate a whole-body deletion of the BAI3 gene (BAI3−/−). In both BAI3−/− male and female mice, a significant reduction in BW was observed compared to BAI3+/+ control mice. Quantitative magnetic imaging analysis showed that lean and fat masses were reduced in male and female mice with BAI3 deficiency. Total activity, food intake, energy expenditure (EE), and respiratory exchange ratio (RER) were assessed in mice housed at room temperature using a Comprehensive Lab Animal Monitoring System (CLAMS). While no differences were observed in the activity between the two genotypes in male or female mice, energy expenditure was increased in both sexes with BAI3 deficiency. However, at thermoneutrality (30 °C), no differences in energy expenditure were observed between the two genotypes for either sex, suggesting a role for BAI3 in adaptive thermogenesis. Notably, in male BAI3−/− mice, food intake was reduced, and RER was increased, but these attributes remained unchanged in the female mice upon BAI3 loss. Gene expression analysis showed increased mRNA abundance of thermogenic genes Ucp1, Pgc1α, Prdm16, and Elov3 in brown adipose tissue (BAT). These outcomes suggest that adaptive thermogenesis due to enhanced BAT activity contributes to increased energy expenditure and reduced BW with BAI3 deficiency. Additionally, sex-dependent differences were observed in food intake and RER. These studies identify BAI3 as a novel regulator of BW that can be potentially targeted to improve whole-body energy expenditure.


Introduction
Obesity results from a chronic imbalance between energy intake and expenditure [1,2]. It is a major risk factor for developing metabolic disorders such as cardiovascular disease, liver disease, and Type 2 diabetes (T2D) [3,4]. Of critical importance, adipose tissues regulate whole-body energy homeostasis [5,6]. In particular, white adipose tissue (WAT) is an

Quantitative Magnetic Resonance (QMR)
The mouse body composition (fat and lean mass) was determined using an EchoMRI™ 3-in-1 QMR machine (Echo Medical Systems, Houston, TX, USA). A system test was performed using a known fat standard before the measurements. Mice were weighed and placed into a clear holding tube capped with a stopper that restricted vertical movement Metabolites 2023, 13, 711 4 of 20 but allowed constant airflow. The tube was inserted into the machine, and the mouse was scanned using a primary accumulation of 1 for measuring total body water in addition to fat and lean mass. The QMR measurements were done in the UAB Small Animal Phenotyping Core supported by the NIH Nutrition & Obesity Research Center P30DK056336, Diabetes Research Center P30DK079626, and the UAB Nathan Shock Center P30AG050886A according to their standard protocols.

Cold Exposure, Comprehensive Laboratory Animal Monitoring System, and Thermoneutrality
Eight-week-old wild-type mice (on C57BL6/J background) were housed at the Center for Comparative Medicine at the University of Alabama at Birmingham under controlled conditions (25 ± 1 • C and 12 h:12 h light-dark cycle enforced by environmental chambers) within a Comprehensive Laboratory Animal Monitoring System (CLAMS; Columbus Instruments Inc., Columbus, OH, USA). All mice were housed singly and allowed to acclimatize to the CLAMS for 9 days before initiation of the experimental protocol. During the acclimatization period, mice received a standard rodent chow diet and had access to water. Subsequently, in the indirect-calorimetry chambers, the oxygen consumption and carbon dioxide production were measured once every 9 min. The total energy expenditure represented the average hourly energy expenditure over 72 h. Resting energy expenditure was calculated from the average over three nonconsecutive 10 min intervals during which energy expenditure was minimal, with at least 1 h between each period. Infrared beams determined locomotor activity for horizontal (x, y) activity. The standard chow diet provided in all studies was obtained from the Harlan NIH-31 Irradiated Open Formula Mouse/Rat Diet (4.7% fat). Time-of-day-dependent analysis for food intake, RER, EE, and activity exhibiting significant oscillations was performed using Cosinor analysis (as described [45,46]). Briefly, Cosinor analyses were conducted using nonlinear regression with PRISM GraphPad (Version 9.4.1(458)) and by fitting the data on the cosine curve, assuming that the free-running period is fixed at 24 h. Data were considered rhythmic if the p-value of the R2 of the linearized Cosinor function of f(t) = M + Cos(2πt/24) + Sin(2πt/24) was less than 0.05. In cases where both comparison groups significantly fit the cosine function, individual parameter estimates for mesor, amplitude, and phase were compared using t-tests determined from the pooled variances. The null hypothesis of no model effects was rejected in all analyses at p < 0.05.

Tissue Histology and Analysis
Freshly extracted adipose tissues were fixed with 4% paraformaldehyde, dehydrated, embedded in paraffin, and sectioned. Sections were stained with hematoxylin (H) and eosin (E). Images were processed for cell size using the Olympus cell Sens Imaging Software version 1.18 (RRID: SCR_014551). The average area of the eWAT lipid droplet was assessed through a semi-automated method.

Quantitative Real-Time PCR
Total RNA was isolated from tissues using TRIzol™ reagent (Invitrogen, cat# 15596018). Following RNA extraction, cDNA was synthesized using a high-capacity cDNA reverse transcription Kit (Applied Biosystems, cat# 4368814). The mRNA abundance of genes was determined by qPCR (ABI3000, Applied Biosystems) with PCR using Fast Start SYBR Green (Roche, cat# 4673484001). The steady-state mRNA levels for each gene were calculated via the comparative ∆CT method and expressed relative to β-actin mRNA. Primer sequences used for qPCR are listed in Table 1.

Statistical Analysis
Data are represented as means ± SEM. Statistical significance was performed using Student's two-tailed unpaired t-test for independent data. The significance limit was set at p < 0.05. Statistical analysis was performed using two-way or repeated-measure ANOVA. PRISM GraphPad (Version 9.4.1(458)) was used to perform a two-way ANOVA to investigate the main effects due to genotype. Repeated-measure ANOVA was used to determine the impact of genotype on energy homeostasis. A full model, including second-order interactions, was conducted for each experiment. Significant differences were determined using Type III sums of squares. The null hypothesis of no model effects was rejected at p < 0.05. Rhythmic parameter comparisons (phase and amplitude) of CLAMs data were calculated using Cosinor analysis. For the following equation, the initial values and default parameter ranges were set manually: f(t) = Mesor + A × Cos [(2 πt/T) + Acrophase]; where mesor (an acronym for midline estimating statistic of rhythm) = the mean of the oscillation; A = the amplitude (peak-to-trough difference); T = the period (24 h); acrophase = the timing of the cosine maximum; t = a timepoint.

Generation and Validation of BAI3 −/− Mice
The BAI3-deletion (BAI3 −/− ) mice on a C57BL/6J background were generated using CRISPR/Cas9 genome editing technology [47]. The full-length BAI3 transcript (NCBI ID: NM_175642.4) contains 30 coding exons, from exon 2 to 31 ( Figure 1A), that encodes an approximately 171 kDa size protein. The NCBI database lists seven additional predicted isoforms of BAI3. XM_036163213.1, XM_006495817.5, and XM_036163215.1 isoforms lack exon 24, exon 30, and exon 3, respectively, from the coding region spanning from exon 2 to exon 31, plausibly due to splicing. XM_017319555.3 and XM_036163221.1 isoforms are predicted to encode identical proteins translated from exon 4 to exon 31. Isoforms XM_006495818.5 and XM_036163223.1 are predicted to encode the shortest BAI3 protein fragment from exon 10 to exon 31 ( Figure 1A). To achieve complete deletion of the BAI3 from mice, exon 2 and exon 18 were targeted by CRISPR/Cas9. The F1 mice produced from the breeding of the founder mouse carried a 4 bp deletion (CAGT) in exon 2 and a 1 bp deletion (T) in exon 18, generating a premature stop codon and stopping protein translation. Further, these nonsense mutations introduced BslI and TaqαI restriction enzyme digest sites in exon 2 and exon 18, respectively. Mice carrying a mutation in exon 2 and exon 18 on the same allele were selected for breeding. Genotyping was performed by polymerase chain reaction (PCR) followed by restriction enzyme digest ( Figure 1B) using tail DNA from the wild-type control and BAI3-mutant mice. The PCR-amplified regions of exon 2 and exon 18 were subjected to BslI and TaqαI restriction enzyme digest, respectively, followed by DNA gel electrophoresis. As predicted, BslI digestion of the PCR-amplified region of exon 2 generated 204 and 63 bp DNA fragments from the BAI3 +/+ allele, whereas 148, 63, and 52 bp DNA fragments were generated from the BAI3-targeted allele. Digestion by TaqαI of the PCR-amplified exon 18 generated a 326 bp DNA fragment from the BAI3 +/+ allele, whereas 201 and 124 bp DNA fragments were generated from the BAI3-mutated allele. The absence of 204 and 326 bp DNA fragments after BslI and TaqαI digestion confirmed homozygous mutation on the same allele in mice in exon 2 and exon 18 of the BAI3 gene ( Figure 1B). Western blot analysis demonstrated the presence of the full-length BAI3 protein (~171 kDa; theoretical molecular weight = 171.13 kDa) in brain and muscle lysates prepared from BAI3 +/+ mice. The complete absence of this band in the brain and muscle of mice containing the BAI3-mutated allele confirmed the presence of BAI3-deletion mice (BAI3 −/− ) ( Figure 1C). In the brain, we also observed a~73 kDa in control but not in BAI3 −/− mice. It is possible that this~73 kDa band observed in the brain corresponds to a protein fragment generated from BAI3 isoforms, and it has been reported previously [33].
R PEER REVIEW 7 of 21

Tissue Expression Profile of BAI3
We assessed the tissue expression profile of BAI3. PCR was performed using cDNA from tissues of BAI3 +/+ and BAI3 −/− mice. Primers were designed to differentiate between BAI3 +/+ and BAI3-CRISPR-targeted alleles ( Figure 1D). The forward (F1, containing CAGT sequence present in the exon 2 of BAI3 +/+ mice) and reverse (R) primers were designed to generate a 217 bp DNA fragment when using cDNA prepared from BAI3 +/+ mouse tissues, (C) Western blots show protein expression in the whole brain and muscle lysates isolated from BAI3 +/+ and BAI3 −/− mice (n = 4). Actin was used as a loading control. (D) The tissue expression profile of BAI3 was determined by PCR using primers designed to differentiate between the cDNA obtained from BAI3 +/+ and BAI3 −/− mice. Actin was used as a loading control to show the presence of DNA (n = 3).

Tissue Expression Profile of BAI3
We assessed the tissue expression profile of BAI3. PCR was performed using cDNA from tissues of BAI3 +/+ and BAI3 −/− mice. Primers were designed to differentiate between BAI3 +/+ and BAI3-CRISPR-targeted alleles ( Figure 1D). The forward (F1, containing CAGT sequence present in the exon 2 of BAI3 +/+ mice) and reverse (R) primers were designed to generate a 217 bp DNA fragment when using cDNA prepared from BAI3 +/+ mouse tissues, but no band was expected from BAI3 −/− mouse tissue cDNA. Alternatively, FP2 (lacking CRISPR-targeted CAGT bases) and RP were designed to generate a 213 bp DNA fragment from cDNA prepared from tissues obtained from BAI3 −/− but not from BAI3 +/+ mice. Tissues such as brain, muscle, WAT, BAT, and islets from the BAI3 +/+ mice expressed only the 217 bp band ( Figure 1D). In contrast, BAI3 −/− mouse tissues expressed only the 213 bp fragment. These findings confirm the deletion of BAI3 in the BAI3 −/− mice. No or very low expression of BAI3 was detected in the liver (as was reported previously [48]). A similar tissue expression profile of BAI3 was observed using quantitative PCR ( Figure S1).

Loss of BAI3 Decreases WAT Tissue Weight
The reductions in BW are often associated with decreases in WAT tissue weight and adipocyte size/number, which improves inflammatory characteristics and aids in the clearance of stored energy [14,49]. To determine if the loss of BAI3 regulates adipose tissue weight, we assessed the weight of tissues harvested from male and female BAI3 −/− and BAI3 +/+ mice. In male mice, significant reductions in kidney, heart, and liver weights ( Figure 3B,C,E), but not of the brain (Figure 3A), were noted in the BAI3 −/− mice compared to that in the BAI3 +/+ control mice. However, normalization to BW revealed no significant differences in tissue weights between genotypes. Among the fat compartments, BAT, inguinal white adipose tissue (iWAT), and epididymal white adipose tissue (eWAT) weights were decreased in the BAI3 −/− mice when compared to that in the BAI3 +/+ control mice (Figure 3D,F,G). Normalization with the BW revealed a significant lowering of eWAT weight (~30% decrease; p = 0.02), but not of iWAT and BAT, in the BAI3 −/− mice compared to that in the BAI3 +/+ control mice. Moreover, the size of adipocytes in the eWAT was significantly decreased in BAI3-deficient mice ( Figure 3H). In female mice, brain, kidney, heart, and liver weights were similar ( Figure 4A-C,E) in the BAI3 −/− mice compared to the control BAI3 +/+ mice. As in males, normalization to the BW did not reveal any significant differences in the tissue weights between the genotypes in females. The eWAT, iWAT, and BAT tissue weights were significantly reduced in the BAI3 −/− mice compared to the control BAI3 +/+ mice (Figure 4D,F,G). After normalization to the BW, only eWAT tissue weight was significantly lowered (42% decrease, p = 0.04) ( Figure 4G) along with the reduction in adipocyte size ( Figure 4H) in the BAI3 −/− mice compared to the control BAI3 +/+ mice. These outcomes suggest that the loss of BAI3 decreases eWAT tissue weight as a percent of total BW and size of the adipocytes in both male and female mice.

BAI3 −/− Mice Have Enhanced Energy Expenditure
To gain insights into factors that caused reductions in BW upon the loss of BAI3, whole-body energy balance parameters were assessed at eight weeks of age in male and female BAI3 −/− and BAI3 +/+ mice housed at room temperature (~25 °C), a temperature that is below thermoneutrality for mice. More specifically, mice were housed in CLAMS cages (at ~25 °C) to determine whether BAI3 deletion impacted either behavior (i.e., physical activity, food intake) or metabolism (i.e., EE, RER). We initially performed repeated measures, two-way ANOVA, to investigate the main effects of genotype and time-of-day. Independent of the time-of-day, male BAI3 −/− mice exhibited a significant decrease in food intake (20.9%, p = 0.01) ( Figure 5A) and concomitant increases in RER (p < 0.01) ( Figure 5B) and energy expenditure (23.3%, p = 0.01) ( Figure 5C) relative to BAI3 +/+ mice. In contrast, no difference in physical activity was observed due to genotypes in male mice ( Figure 5D). In female BAI3 −/− mice, a significant increase in energy expenditure (17.74%, p = 0.04) was observed (in the absence of genotype main effects for food intake, physical activity, and RER) ( Figure 6A-D). Given that all energy balance parameters exhibited significant timeof-day main effects in male and female mice, we next performed Cosinor analysis. In all instances, energy balance parameters significantly fit cosine curves with a periodicity of 24 h, allowing for the comparison of mesor, amplitude, and phase between BAI3 −/− and BAI3 +/+ mice (Tables 2 and 3). In male mice, mesor was significantly altered for all four energy balance parameters; RER (0.883 ± 0.002 vs. 0.876 ± 0.001; p < 0.01) and energy expenditure (0.0225 ± 0.00016 vs. 0.0183 ± 0.00015; p = 6.45 × 10 −19 ) were increased, while food

Mesor
Amplitude Phase We next investigated the potential mechanisms by which energy expenditure was increased in male and female BAI3 −/− mice. We hypothesized that increases in energy expenditure in BAI3 −/− mice were secondary to the augmentation of thermogenesis. To initially test this hypothesis, energy expenditure was investigated in BAI3 −/− and control mice acclimatized to thermoneutrality. More specifically, male and female BAI3 −/− and control mice were placed in CLAMS cages at (~30 • C) for 10 days; during the last 2 days, energy expenditure was assessed. Under thermoneutrality conditions, neither a twoway ANOVA nor Cosinor analysis revealed significant differences in energy expenditure between male and female BAI3 −/− mice (relative to controls) (Figures 5E and 6E). Altogether, these data suggest a role for BAI3 in thermogenesis.

Discussion
BAI3 is expressed in the central nervous system, hippocampal neurons, Purkinj cells, heart, islets, testis, muscles, and small intestine [35,[49][50][51][52][53]. BAI3 is involved in di verse cellular functions, such as dendrite morphogenesis and synaptogenesis in Purkinj cells [54,55]. Further, BAI3 mediates synaptic connections between the anterior olfactory nucleus (AON) and granule cells in the olfactory bulb (OB), which is essential for the socia transmission of food preference (STFP) [56]. In the muscle, BAI3 interaction with th plasma membrane receptor protein stablin-2 was reported to regulate muscle myoblas fusion, a critical process determining muscle size and regeneration in adults [34,35,57]. In pancreatic islets, BAI3 is implicated in regulating insulin secretion [52]. These studie show that BAI3 is expressed in key metabolic tissues. However, its role in whole-body energy homeostasis remains unknown. Figure 7. The mRNA and protein abundance in BAT. Total mRNA was prepared from BAT. The steady-state mRNA abundance of genes was expressed relative to the β-actin was determined using qPCR in males (A) and females (B). Western blots assessing the relative protein abundance normalized to actin (C). Data are reported as the mean ± SEMs, n = 4, * p < 0.05.

Discussion
BAI3 is expressed in the central nervous system, hippocampal neurons, Purkinje cells, heart, islets, testis, muscles, and small intestine [35,[49][50][51][52][53]. BAI3 is involved in diverse cellular functions, such as dendrite morphogenesis and synaptogenesis in Purkinje cells [54,55]. Further, BAI3 mediates synaptic connections between the anterior olfactory nucleus (AON) and granule cells in the olfactory bulb (OB), which is essential for the social transmission of food preference (STFP) [56]. In muscle, BAI3 interaction with the plasma membrane receptor protein stablin-2 was reported to regulate muscle myoblast fusion, a critical process determining muscle size and regeneration in adults [34,35,57]. In pancreatic islets, BAI3 is implicated in regulating insulin secretion [52]. These studies show that BAI3 is expressed in key metabolic tissues. However, its role in whole-body energy homeostasis remains unknown.
This study reports findings from the initial BAI3 −/− mice characterization. These mice were generated by targeting exon 2 and exon 18 of the BAI3 gene by CRISPR/Cas9 approach to get a complete loss of protein translation from BAI3 transcripts (Figure 1). The deletion of full-length BAI3 protein was confirmed by the custom-made rabbit anti-mouse BAI3 antibody, which was generated in response to the epitope located in the intracellular C-terminal region of the BAI3 protein, common to all putative transcripts.
Our studies reveal (a) that the loss of BAI3 from both male and female mice leads to reduced BW, which is attributed to decreased food intake and enhanced energy expenditure, (b) reduced eWAT tissue weight and increased expression of genes and proteins known to enhance BAT activity, and (c) no differences in energy expenditure between BAI3 +/+ and BAI3 −/− mice at thermoneutrality. These data suggest that BAI3 regulates food intake and maintains adaptive thermogenesis by enhancing energy expenditure to regulate BW in adult mice. cAMP signaling is critical to the functioning of WAT and BAT in regulating whole-body energy homeostasis. In both WAT and BAT, increases in cAMP signaling in response to the activation of Gα s -coupled GPCRs such as β3-adrenoreceptor (AR) enhance PKA-mediated lipolysis to break down stored triacylglycerols to free fatty acids, which are an energy source in peripheral tissues [37][38][39]. Further, the activation of cAMP/PKA signaling in BAT increases the expression of Ucp1, a key thermogenic gene that increases cold-induced thermogenesis (adaptive thermogenesis), enhancing whole-body energy expenditure and insulin sensitivity [5,[41][42][43]. Our data show that BAI3 is expressed in WAT and BAT. Furthermore, BAI3 −/− mice exhibited reduced WAT tissue weight (in males and females), increased expression of thermogenic genes in BAT of males and females, and enhanced energy expenditure in the chronic cold. Thus, it is possible that the loss of BAI3, which is known to be coupled to the Gα i [34], could lead to increases in cAMP signaling-mediated energy expenditure. In muscle, BAI3 was reported to be coupled with Gα i [34]. Similarly, the BAI3 N-terminal fragment that binds its ligand complement 1q-like 3 (C1ql3) secreted protein was reported to effectively block the inhibitory effects of C1ql3 on insulin secretion from mouse islets in response to cAMP stimulation [52]. The molecular mechanism by which BAI3 regulates cAMP signaling and the functioning of adipose tissues remains to be elucidated.
Our study revealed sex-dependent differences in mice after the deletion of BAI3. Overall greater differences in the BW due to BAI3 deletion were observed in the male mice compared to the female mice. After baseline correction by normalizing BW for both genotypes at 3 weeks of age to zero, the male but not the female BAI3 −/− mice exhibited a significant reduction in BW from 8 to 10 weeks of age, suggesting that the loss of BAI3 decreased BW only in the adult male mice. This reduction in the BW may remain sustained after 10 weeks of age. The sexual dimorphism in BW gain is well established and attributed to eating disorders, adipose tissue function, and fat distribution in distinct adipose depots [58][59][60][61]. We observed that after performing the time-of-day analysis by the Cosinor approach, food intake in the BAI3 −/− male mice was significantly reduced compared to that in the BAI3 +/+ mice during the active night-time feeding phase in mice ( Figure 5A, Table 2), potentially contributing to the reduced BW in BAI3 −/− male mice. The hypothalamus is a key regulator of hunger and satiety in response to insulin and leptin signaling. Inhibition of NPY/AgRP neurons and stimulation of POMC/CART neurons decrease food intake, reducing BW [62,63]. Thus, the loss of BAI3 directly or indirectly may modulate NPY/AgRP and POMC/CART functions to regulate BW in adult males.
We observed increased EE in response to the deletion of BAI3 in both male and female mice without changes in total activity and lean mass/BW ratio. Moreover, no difference in EE was achieved at thermoneutrality upon BAI3 deletion in mice, implicating the role of BAT tissue in mediating the effects of BAI3 gene deletion in adaptive thermogenesis. A potential mechanism involves increased in mitochondrial oxidation of FFAs in the BAT to generate heat rather than ATP synthesis [14,64]. We observed reduced WAT tissue weight in both BAI3 −/− male and female mice (Figures 3 and 4). Thus, FFAs generated from the WAT could be utilized as fuels in BAT to enhance EE. The BAT tissue from BAI3 −/− mice exhibited enhanced expression of thermogenic genes (Dio2, Elov3, Ucp1, Pgc1α), which are known to increase EE by promoting BAT activity [65,66]. Our characterization of BAI3 −/− mice reveals that the loss of BAI3 increases EE by altering BAT activity for adaptive thermogenesis. Additional studies will be required to elucidate the mechanism of BAI3 in regulating energy metabolism in the obesity cold challenge. The outcomes presented in this study implicate that BAI3 is a metabolic regulator whose function can be altered to regulate whole-body energy expenditure, potentially alleviating obesity and metabolic disorders. In conclusion, our findings reveal that BAI3 regulates BW in male and female mice, particularly by regulating energy expenditure via genes involved in adaptive thermogenesis.

Supplementary Materials:
The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/metabo13060711/s1, Figure S1. Determine the BAI3 mRNA abundance in tissues obtained from BAI3 +/+ and BAI3 −/− male mice. The total mRNA was prepared from tissues, and steady-state mRNA abundance of BAI3 expressed relative to the β-actin was determined using qPCR (A-G). Data are reported as the mean ± SEMs, n = 4, * p < 0.05. Figure S2. Correlation between lean or fat mass with BW in BAI3 +/+ and BAI3 −/− male and female mice. Correlation of lean mass with BW (A and D). Correlation of fat mass with BW (B and E). Lean and fat mass as a percent of BW (C and F).

Informed Consent Statement: Not applicable.
Data Availability Statement: The data supporting this study's findings are available in this article's methods and/or Supplementary Material. The raw data supporting the conclusions of this article will be made available by the authors without undue reservation.