Phenotypic Studies

Body weight and food intake were monitored weekly in all mice after surgery. Body weight and food were weighed weekly on an Ohaus NV511 precision balance (Parsippany, NJ). Mice were single-housed after surgery for food intake measurement, but mice were group housed for studies that only included either glucose clamps or indirect calorimetry. Glycemic measures (including intraperitoneal [IP] glucose tolerance test [GTT]; 2 g/kg body weight) and insulin tolerance test (0.6 Unit/kg body weight, Humulin (Eli Lilly), IP) were performed in 12- to 18-week-old animals. Blood glucose was measured with a One Touch Ultra 2 glucometer (Johnson and Johnson) by tail vein draw. For all glycemic measures, mice were fasted for 4 to 6 hours prior to testing. Body fat and lean mass were analyzed in a cohort of mice using Minispec LF9011 (Bruker Optics, Billerica, MA). Body composition data was collected with assistance from Indiana University School of Medicine Mouse Metabolic Phenotyping Center. TSE Phenomaster cages were used for indirect calorimetry measurement, with assistance from the Indiana University School of Medicine Mouse Metabolic Phenotyping Center. Indirect calorimetry data was used to calculate fat oxidation and glucose oxidation via the Weir equation (23). Data were analyzed by taking into account whole-body mass and lean mass separately. Glucagon, insulin, and leptin were all assayed in plasma by commercial enzyme-linked immunosorbent assays (Glucagon: Mercodia, Uppsala, Sweden, RRID: AB_2783839 (24); Leptin: Crystal Chem, Elk Grove, IL, RRID: AB_2722664 (25); Insulin: Crystal Chem, Elk Grove, IL; RRID: AB_2783626 (26)). Corticosterone was measured with radioimmunoassay (MP Biomedicals, Santa Ana, California, RRID: AB_2801269 (27)) with assistance from the Chemistry Core supported by the Michigan Diabetes Research Center (University of Michigan).

Reagents

All AAVs used include titer from stock. AAV-Cre-GFP (1.1 × 10¹³ genome copies [GC]/mL), from Addgene plasmid 68544, was generated in at the University of Michigan Viral Vector Core with the support of the Molecular Genetics Core of the Michigan Diabetes Research Center (University of Michigan, Ann Arbor, Michigan, USA) or directly acquired from Addgene (2.7 × 10¹³ GC/mL) (105540). AAV-GFP (5.4 × 10¹² GC/mL) was acquired from the University of North Carolina Vector Core (6765) or Addgene (3.4 × 10¹³ GC/mL) (105539). It is important to note that the 2 AAV-Cre used have different promoters. Addgene 68544 has a cytomegalovirus (CMV), whereas 105540 has a hSyn promoter. All of these AAVs were serotype 9.

Hyperinsulinemic-Euglycemic Clamp

In both cohorts, only male mice were used for hyperinsulinemic-euglycemic clamps. In the first cohort, mice were catheterized (arterial and venous) at 11 weeks poststereotaxic injection, while mice were catheterized 3 weeks postinjection in the second cohort. Mice were allowed to recover from surgery and only went through the clamping procedure if the animals had no sign of being sick or ill. After a 5-hour fast, conscious, unstressed, catheterized mice were infused with 10 mU/kg/min insulin, and glucose was clamped at ~150 mg/dL for 120 minutes. After clamp, mice were administered (venous catheter) with [¹⁴C]-labeled 2-deoxy-D-glucose. Organs were collected for radioactivity measurements as a marker for glucose uptake, as previously described (28).

Brown Adipose Tissue Temperature

Temperature probes were surgically implanted directly above the interscapular BAT (as previously described (29)) in PACAP-2a-cre mice (as previously described (30)) that received AAV-DIO-hM3dq into the VMN at the Beth Israel Deaconness Medical Center. Animals were allowed 1 week of recovery and acclimation was performed with saline injection for 5 consecutive days. Data were collected using a handheld reader system (DAS-7006/7r; Bio Medic Data System) at the time points indicated.

Stereotaxic Injections of Viral Constructs

Animals 8- to 14-weeks of age were anesthetized with 1.5% to 2% isoflurane, in preparation for craniotomy. In cohort 3, 15- to 18-week-old mice were used because it was a short-term study compared with cohorts 1 and 2. After exposing the skull, bregma and lambda were leveled, a hole was drilled, and the contents of a pulled pipette at the coordinates of our target for ~25 nL/min were released. For the VMN-directed injections, 100 nL of virus was injected at anteroposterior (AP) −1.05 and −1.15, mediolateral (ML) ± 0.3, dorsoventral (DV) −5.55. Four total injections (2 on each side within the same drilled hole for a total of 200 nL per side) were administered per animal, in order to spread the virus throughout the full extent of the VMN and get almost complete knockdown throughout the nucleus. We allowed at least 5 minutes for the virus to diffuse into the brain and the pipette was raised slowly out of the hole in the skull. The hole in the skull was filled with bone wax, and the skin was closed with surgical sutures (Henry Schein). Analgesics (Carprofen, 5 mg/kg) were administered prophylactically to all mice to prevent postsurgical pain. The animals were allowed 4 weeks to recover from surgery before any experimental manipulation. For food intake studies, the animals were individually housed after surgery in order to collect food intake data. Otherwise, animals continued to be housed with littermates. GFP fluorescent reporters in all studies were used to confirm proper targeting of the brain region. If bilateral fluorescence was not observed within the dorsomedial VMN, these cases were omitted from analyses.

Perfusion and Immunohistochemistry

Brains were collected and processed as previously described (28, 31). Mice were anesthetized with isoflurane prior to transcardial perfusion with phosphate buffered saline (PBS) followed by 10% neutral buffered formalin (NBF) (Sigma Aldrich). Brains were removed and placed into 10% NBF overnight, followed by 30% sucrose for at least 36 hours. Brains were cut into 30-µm sections on a freezing microtome in 4 series and stored in antifreeze solution (25% ethylene glycol, 25% glycerol). Sections were washed with PBS and then treated sequentially with 1% hydrogen peroxide/ 0.5% sodium hydroxide, 0.3% glycine, and 0.03% sodium dodecyl sulfate. Pretreatment was followed by an hour in blocking solution (PBS containing 0.1% Triton, 3% normal donkey serum), followed by overnight incubation in blocking solution containing either chicken anti-GFP (GFP-1020, RRID:AB_1000240 (32), Aves, 1:1000), rabbit anti-NeuN (ABN78, RRID:AB_10807945 (33), Millipore, 1:500), or rat anti-mCherry (Life Technologies M11217, RRID:AB_2307319, 1:1000 (34)). The next day, the sections were incubated with fluorescent secondary antibodies (Molecular Probes, 1:200). The sections were mounted onto slides coverslipped with Fluoromount-G containing DAPI (Southern Biotech).

In Situ Hybridization

In situ hybridization for Adcyap1 was performed in perfused brain tissue, using DEPC-treated PBS and 30% sucrose solution. Once the brains sunk in the 30% sucrose solution at 4 °C, brains were sectioned on a freezing stage microtome at 30 µm increments. Series of brain sections were mounted onto SuperFrost plus slides (Fisher Scientific) then stored in a slide box at −20 °C. Slides were fixed in 10% NBF (Sigma Aldrich) for 20 minutes and subsequently dehydrated in increasing concentrations of ethanol, cleared in xylenes, rehydrated in decreasing concentrations of ethanol, and finally placed in prewarmed sodium citrate buffer (pH 6.0). The slides were then microwaved for 10 minutes followed by dehydration in increasing concentration of ethanol. The slides were allowed to air-dry for at least 1 hour. The Adcyap1 cDNA (template) was produced from mouse hypothalamic RNA. The following primers were used to amplify a 509 base pair sequence in the encoding region of the Adcyap1 gene with polymerase promoters: F (T3) 5′-(CAG AGA TGC AAT TAA CCC TCA CTA AAG GGA GA) ACC ATG TGT AGC GGA GCA AG-3′; R (T7) 5′-(CCA AGC CCT CTA ATA CGA CTC ACT ATA GGG AGA) CGG CGT CCT TTG TTT TTA ACC-3′. Primers were acquired from Integrated DNA Technologies (IDT; Iowa). The Adcyap1 riboprobe was generated by in vitro transcription using 35S-UTP as the radioisotope. Labeled riboprobe was diluted in hybridization solution and brain sections were hybridized overnight at 58 °C. The following day, sections were incubated in 0.002% RNAse A followed by stringency washes. Sections were dehydrated in increasing concentrations of ethanol, air-dried, and placed in x-ray film cassettes with BMR-2 film (Kodak, Rochester, NY, USA) for 2 days.

Quantitative Real-Time PCR

RNA was extracted from BAT using RNeasy Lipid Tissue Mini kit (Qiagen). The cDNA was synthesized by reverse transcription from mRNA using the iScript cDNA Synthesis Kit (Bio-Rad). Gene expression was performed by quantitative real-time RT-PCR using Taqman gene expression assays using StepOnePlus detection system (Applied Biosystems) with a standard protocol. Relative abundance for each transcript was calculated by a standard curve of cycle thresholds and normalized to RL32.

Cell Counts

Images from 30-µm thick sections containing Cy3 tagged NeuN were collected in the mediobasal hypothalamus using a Keyence BZ-X810 microscope with a 20× objective. The VMN was identified by overlaying the corresponding bregma level from an anatomical reference atlas (Allen Mouse Brain Atlas) on microscope images in ImageJ. Two regions of interest (ROI) encompassing the whole VMN were drawn bilaterally for each microscope image and the ROI areas (pixel/inch²) were measured. Image thresholding (Otsu Method) followed by watershed segmentation was performed to determine the number of NeuN-positive cells within each ROI. Data were reported as NeuN-positive cells by ROI area to correct for bregma level differences in ROI size. Both right and left sides of the VMN were averaged for a final value.

AAV^Cre Injection Ablates VMN Adcyap1 Expression

We tested the hypothesis that Adcyap1 in the VMN directly regulates energy expenditure and glucose metabolism by ablating Adcyap1 expression in the VMN. We conditionally ablated Adcyap1 in the VMN via microinjection of AAV^Cre into the VMN of 8- to 14-week-old Adcyap1^flox mice (21). By using this approach, we avoided the drawback of potentially eliminating PACAP outside of the brain using transgenic Cre mouse models (such as SF1^CRE) that also target cells in peripheral tissues. The peptide PACAP is expressed throughout the periphery and the SF1^cre mouse model would induce Cre activity within some peripheral tissues (44).

Indeed, Adcyap1 mRNA was present within areas known to contain the neuropeptide (eg, cortex, preoptic area, medial amygdala, ventral premammillary nucleus) but was almost completely ablated from the dorsomedial VMN of all AAV^Cre-injected mice (Fig. 2A-2C). In support of this qualitative evidence, we also quantified expression level from the films to confirm reduction in Adcyap1 in the VMN (t[36] = 13.26; P < 0.0001) following injection of AAV^Cre, but not within nearby areas such as the medial amygdala, ventral premammillary nucleus, and preoptic area (Fig. 2A-2C). AAV^Cre injection contained a GFP tag, in order to observe the spread of the virus. All included cases contained bilateral GFP expression throughout the VMN. Injection was not completely confined in the VMN and did spread to some degree into nearby areas in order to induce desired near-complete ablation of Adcyap1 mRNA in the VMN. Of note, GFP cells were observed in the posterior hypothalamus, arcuate nucleus, anterior hypothalamus, suprachiasmatic nucleus, and dorsomedial hypothalamus, depending on the injection site (Supplemental Fig. 1) (22). Importantly, we did not observe PACAP mRNA expression above background in any of these regions, although it is known that small populations of PACAP-expressing cells are found in much of the hypothalamus across development (45, 46). Furthermore, unlike our VMN-directed ablations, none of the other regions had the same localization of viral spread, which we use as indirect readout of PACAP-ablation across multiple mice. Therefore, while PACAP may have been ablated from these nearby regions if expressed there, it is unlikely that the loss of function we see across our metabolic tests in adult mice can be attributed to the regions other than the VMN. While AAVs have been used for many years without cytotoxic consequences, we wanted to ensure that we did not lesion the hypothalamus with our injections. DAPI-labeled nuclei were observed in a similar distribution in all groups (data not shown). To quantify this expression, we used NeuN as a biomarker for neurons. In the VMN, there was no difference in the density of cells in the VMN (Fig. 2D-2F). Therefore, AAV^Cre sufficiently ablated PACAP from the VMN without lesioning the nucleus.

Loss of Adycap1 in MBH Induces Obesity in Both Male and Female Mice

Experiments were designed to assess the function for MBH PACAP in metabolic function (see experimental design in Supplemental Fig. 1) (22). Adcyap1^MBHKO mice were noticeably larger than littermate/cagemate controls beginning at 8 weeks postinjection. As such, body weight was closely monitored in these mice for the duration and in subsequent cohorts of mice. Body weight steadily increased to almost 40% more than control mice 8 weeks after surgery. This increase in body weight was observed in both male and in female mice (Fig. 3A, Supplemental Fig. 3A (22)) (F[2,17] = 10.2; P < 0.0001) (F[2,17] = 14.59; P < 0.0001) with only a delayed and modest increase in food intake (Fig. 3B, Supplemental Fig. 3B (22)) (F[2,17] = 0.5095; P = 0.0367) (F[2,17] = 0.3172; P = 0.1513). Because there was a mild increase in food intake starting at week 7, we measured the food intake in mice following a 24-hour fast at 10 weeks postinjection. However, fasting/re-feeding food intake responses were normal in these mice, indicating that feeding responses were only modestly affected after the onset of obesity (Fig. 3C, Supplemental Fig. 3C) (22). Since effects were similar across sexes, data was combined for subsequent analyses of indirect calorimetry and glycemic regulation.

Adcyap1 ^MBH KO Mice Display Dysfunctions in Energy Expenditure

In order to perform indirect calorimetry measurements, we included a separate cohort of mice that were exposed to the metabolic chambers early into the process of the mice becoming obese (4 weeks postinjection). These mice did exhibit increased fat mass (t[11] = 4.474; P = 0.0008), but lean mass was not different compared with controls. (Figs. 4A, ​,4B).4B). Oxygen consumption, VO2 (F[2,10] = 25.94; P = 0.0301), and carbon dioxide production, VCO2 (F[2,10] = 27.07; P = 0.0185), were both reduced when normalized with lean mass (Figs. 4E, ​,4F,4F, ​,4G,4G, ​,4H).4H). Respiratory exchange ratio was also reduced (F[2,10] = 26.99; P = 0.0076) during the dark cycle (Fig. 4M, ​,4N)4N) in Adcyap1^MBHKO mice. Finally, fat oxidation was elevated (F[2,10] = 4.979; P = 0.0051) and glucose oxidation reduced (F[2,10] = 5.06; P = 0.0029) during the dark cycle (Fig. 3P-3S) in Adcyap1^MBHKO mice. Certainly, Adcyap1^MBHKO mice exhibit dysfunctions in underlying metabolic activity, consistent with reduced energy expenditure. Therefore, the marked increases in body weight in the Adcyap1^MBHKO mice were due to the metabolic dysfunctions in energy expenditure, and the delayed increase in food intake is likely secondary to changes in energy utilization.

Obese Adcyap1^MBHKO Mice Have Increased Interscapular BAT Weight and Markers for Lipid Oxidation

Upon sacrificing the mice after indirect calorimetry experiments (5 weeks postinjection), we noticed that the Adcyap1^MBHKO mice had considerably more interscapular brown adipose tissue (iBAT) than the controls (Fig. 4C). Indeed, iBAT weights were more than 4 times larger than controls. We next measured expression of targets in BAT tissue associated with thermogenesis and lipid oxidation (Supplemental Fig. 4A-E) (22). Indeed, mitochondrial uncoupling protein 2 (UCP2) (t[12] = 2.319; P = 0.0267), iodothyronine deiodinase 2 (DIO2) (t[12] = 2.157; P = 0.0315), and peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1a) (t[12] = 2.832; P = 0.011) were all increased in the knockouts relative to the controls. However, we did not observe any differences in mitochondrial uncoupling protein 1 (UCP1) and PR domain containing 16 (PRDM16). These data, taken together with the indirect calorimetry, indicate that these obese mice are metabolically inflexible and utilize lipids rather than carbohydrates for energy, which may further contribute to their obese phenotype.

Obese Adcyap1^MBHKO Mice Are Hyperglycemic and Hyperinsulinemic, yet Remain Insulin Tolerant

In addition to influencing energy expenditure, PACAP plays a critical role in energy mobilization and glucose clearance into peripheral tissues. Thus, we tested whether loss of PACAP in the MBH leads to hyperglycemia and decreased glucose tolerance. In support of this hypothesis, we observed that Adcyap1^MBHKO mice had higher fed and fasted (24 hours) blood glucose levels compared with the controls (F[1,38] = 20.16; P < 0.0001) (Fig. 5A). In addition to hyperglycemia, Adcyap1^MBHKO mice were hyperinsulinemic (F[1,75] = 16.49; P = 0.0001), but neither hyperglucagonemic nor hypercortisolemic at either fasting or fed states (Fig. 5B, ​,5D).5D). The Adcyap1^MBHKO mice also had impaired glucose tolerance (F[1,39] = 6.504; P = 0.0148) (Fig. 5E). Despite increased body weight and adiposity in addition to impaired glucose tolerance, Adcyap1^MBHKO mice were not insulin resistant (Fig. 5F). Adcyap1^MBHKO mice had a comparable decrease in blood glucose levels as controls in response to exogenous insulin administration (0.6 Units/kg; Fig. 5F). Glucagon and corticosterone levels 30 minutes post–insulin injection were not different between the groups (Fig. 5G, ​,5H).5H). Therefore, loss of PACAP in the MBH resulted in glucose intolerance, despite having normal insulin tolerance and counterregulatory responses.

Obese Adcyap1^MBHKO Mice Have Impaired Glucose Clearance

In order to evaluate glucose clearance independent from changes in circulating insulin and glucose, we evaluated glucoregulation in Adcyap1^MBHKO in a hyperinsulinemic/euglycemic condition. Blood glucose (12 weeks postinjection) was clamped at 150 mg/dL (Fig. 6A). Considerably less glucose was infused in order to sustain equivalent circulating glucose levels in Adcyap1^MBHKO compared with control mice (F[1,16] = 24.28; P = 0.0002) (Fig. 6B). Indeed, hepatic glucose production (F[1,32] = 6.28; P = 0.0175) and insulin clearance (t[16] = 5.845; P < 0.0001) were also lower following the clamp, potentially contributing to the hyperglycemia in Adcyap1^MBHKO mice (Fig. 6C, ​,6D).6D). At the end of the clamp, [¹⁴C]-deoxy-D-glucose was infused and organs collected. Measured radioactivity was lower in the gastrocnemius muscle (t[16] = 2.722; P = 0.0075), heart (t[16] = 3.872; P = 0.0007), BAT (t[16] = 6.182; P < 0.0001), subcutaneous fat (t[15] = 3.913; P = 0.0007), but not visceral fat of Adcyap1^MBHKO mice (Fig. 6E-6I). Overall, these data suggest that, in addition to the control of adiposity and energy expenditure, loss of MBH PACAP results in hyperglycemia and diminished glucose clearance consistent with aberrant glucose specific regulation in the obese state.

We thank Martin G. Myers Jr., David P. Olson, Paula Goforth, Alex Banks, and Jon Resch for helpful discussions. We thank the Michigan Diabetes Center (NIH P30 DK020572, including the Molecular Genetics, Animal Studies, and Microscopy, Imaging, and Cellular Physiology Cores). We thank the University of Michigan Animal Phenotyping Core (1U2CDK110678-01). We would also like to thank the American Diabetes Association (17-INI-15 to J.N.F.) and the Michigan Diabetes Center Pilot and Feasibility Grant for the funding to support this project. Lastly, we would like to thank the NIH (P30 DK046200, P30 DK057521, R01 DK075632, R01 DK089044, R01 DK096010, R01 DK111401 to BL, 5T32DK108740 to N.B.K., K08 DK118201 and T32 HL007374 to R.R., and 5T32AA007462 to D.L.H.) for funding to support this project.

Financial Support: NIH (P30 DK046200, P30 DK057521, R01 DK075632, R01 DK089044, R01 DK096010, R01 DK111401 to BL, 5T32DK108740 and UL1TR002240 to N.B.K., K08 DK118201 and T32 HL007374 to R.R., and 5T32AA007462 to D.L.H.)

The American Diabetes Association (17-INI-15 to J.N.F.)

Author Contributions: J.N.F., R.R., and N.B.K. collected and analyzed the data. J.N.F., N.B.K., B.L., D.J., D.L.H., and R.R. designed experiments. B.L. and R.R. generated Adcyap1^flox mouse model and provided feedback on studies. All authors reviewed and edited the manuscript. J.N.F. is the guarantor of the manuscript.