The effects of acylation stimulating protein supplementation VS antibody neutralization on energy expenditure in wildtype mice

Polyclonal rabbit antibody against human ASP (Anti-ASP) was developed as previously described [18]. The IgG fraction was isolated from plasma by gamma sepharose chromatography (GE Healthcare, Chicago, IL, USA) and the polyclonal antibody cross-reacted with mouse ASP [17]. Non-immune IgG (NI-IgG) was purchased from Sigma (Sigma, St Louis, MO, USA). For the production of recombinant ASP (rASP), the portion of the human C3 gene representing ASP (C3adesArg) was cloned into the NcoI/EcoRI restriction sites of pET32a(+) (Novagen, Madison, WI, USA). An extra alanine (GCC) codon was incorporated to be able to use NcoI as a cloning site. The vector was digested with NdeI to excise out the thioredoxin tag (TRX.tag) and religated to obtain pET_his_ASP vector containing the His-tagged ASP sequence. The cell line used to produce recombinant ASP was Origami B (DE3) (Novagen), a BL21 derived cell line lacking thioredoxin and glutathione reductase genes. Frozen OriB-pET-his-ASP scrapes were inoculated and grown overnight at 37°C with shaking (220 rpm) in bacterial media (25 g/L LB Broth, EMD Biosciences, Madison, WI, USA) supplemented with 10 ug/mL kanamycin (Invitrogen, Burlington, ON, Canada) and 100 ug/mL carbenicillin (Invitrogen). Cultures were diluted (1.5 mL/L) in bacterial media, incubated 5-8 hours (37°C, 220 rpm) until an OD₆₀₀ of 0.6-0.8 was reached, then supplemented with IPTG (Isopropyl ß-D-1-thiogalactopyranoside, QIAGEN, Mississauga, ON, Canada, 1 mM final concentration) and incubated overnight (25°C, 220 rpm). Cell pellets were collected (centrifugation 3000 ×g, 20 min), lysed with 10 mL Bugbuster (EMD Biosciences), 10 mg lysozyme (Sigma) and 10 μL benzonase nuclease (20 U/mL, EMD Biosciences), vortexed well, shaken 30 min at RT, centrifuged (3000 g, 5 min, 4°C) and the supernatant decanted. The lysis step was repeated with an additional 10 mL, then 5 mL. All supernatants were pooled, acidifed with 10N HCl (final 1 N HCl), stirred (15 min RT), centrifuged (3000 g, 20 min, 4°C), pooled, adjusted to pH 7.4 with 10 N NaOH and centrifuged (3000 g, 20 min, 4°C). rHis-ASP was purified with binding to Ni²⁺-Sepharose column (GE Healthcare). Following loading of rHis-ASP, the column was washed with buffer (50 mM NaH₂PO₄, 300 mM NaCl, 10 mM imidazole, pH 74.) and eluted with acidified buffer (0.625 N HCl in buffer, pH ~2.0). For removal of rHis-tag, rHis-ASP solution was carefully adjusted to pH 8.0, enterokinase (New England Biolabs, Pickering, ON, Canada) was added (0.063 ug/20 mL), and incubated overnight at RT (18°C) with gentle shaking. rASP was purified by HPLC as previously described [19].

3T3-L1 preadipocytes cells were obtained from ATCC (Manassas, VA, USA). Cells were routinely grown in Dulbecco's modified Eagle's medium (DMEM) + 10% (v/v) fetal calf serum (Invitrogen) at 37°C, 5% CO₂. 3T3-L1 preadipocytes were differentiated into mature adipocytes as described previously [7]. Differentiated adipocytes in 48 well plates were preincubated (37°C, 5% CO₂) in serum free medium for 1 hour then incubated with PBS (baseline synthesis), insulin (100 nM) (Sigma), rASP (100 nM), or rASP (100 nM) + Anti-ASP (0.2, 0.3, 0.4 μ g/mL, added before the addition of ASP) for 1 hour. Uptake of fluorescently labeled fatty acids was measured over time in adipocytes as described [20, 21] using a QBT™ fatty acid uptake assay kit (Molecular Devices, Sunnyvale, CA, USA) as outlined by the manufacturer. Following the addition of BODIPY-fatty acid, fatty acid uptake is measured in real-time over 120 min in a bottom-reading fluorescent microplate reader.

Male C57BL/6 mice (aged 7-8 weeks) were purchased from Charles River Laboratories (Wilmington, PA, USA). The mice were maintained on a high-fat diet (45% kcal from fat, Research Diet, New Brunswick, NJ, USA) throughout the study (total of 6 weeks) and housed individually in a sterile barrier facility with a 12 h light: 12 h dark cycle. Two weeks following the start of the high-fat diet regimen, the mice were surgically implanted with an osmotic mini-pump (model 2004) (Durect Corp., Cupertino, CA, USA). Briefly, the mice were first sedated with ketopiofen (5 mg/kg) then anesthetized with isofluorane. Under sterile conditions, a small incision was made between the skin and scapulae and the osmotic pump was inserted subcutaneously into the mouse. The incision was then closed with sutures according to manufacturer's instructions. All protocols were conducted in accordance with the CACC guidelines and approved by the University Animal Care Committee. The mice were divided into four groups where they were continuously delivered; i) non-immune IgG n = 6 (2 μg/μL); ii) Anti-ASP n = 7 (4 μg/μL); iii) PBS (vehicle) n = 6; or iv) recombinant ASP (95 pmol/μL) in PBS n = 5 for 4 weeks at a flow rate of 0.25 μL/hr. Twenty-eight days following mini-pump insertion (6 weeks on high-fat diet), mice were fasted overnight, euthanized, and tissues, blood samples and the mini-osmotic pumps were collected. Average pump rate was calculated as the difference between the initial volume loaded and the residual volume divided by 28 days.

Overnight fasting plasma samples were taken 9 days after the surgery by submandibular bleed and during sacrifice by cardiac puncture (day 28). Plasma human rASP levels were measured by an in-house ELISA which does not cross-react with mouse ASP, background non-specific (PBS) values were subtracted [18]. Rabbit IgG was measured by Easy-titer IgG assay kit (Thermo Fisher Scientific Inc., Rockford, IL, USA), to track the efficacy of pump delivery. Plasma TG (Roche Diagnostics, Indianapolis, IN, USA), non-esterified fatty acids (NEFA) (Wako Chemicals, Richmond, VA, USA) and glucose (Sigma) were measured using colorimetric enzymatic kits. Insulin and adiponectin were measured using RIA kits (Linco, St.Charles, MO, USA); interleukin-6 (IL-6) (BD, Franklin Lakes, NJ, USA) and CRP (Immunology, Consultants Laboratory, Inc., Newberg, OR, USA) were measured by ELISA. C3 was measured by immunoturbidimetric assay (Kamiya Biomedical Company, Seattle, WA, USA).

Body weight and food intake were measured 3 times a week. Food intake results are reported as cumulative food intake in kcal. Food efficiency was calculated as the food intake in kcal over body weight gain in grams. Calorimetry measurements were taken 15-17 days after the surgery. Oxygen consumption (VO₂) and carbon dioxide production (VCO₂) were measured over a 24 hour period in an open circuit system after 48 hours equilibration as previously described [22]. VO₂ and VCO₂ were calculated as mL/kg/min and RQ (respiratory quotient) was taken as the quotient of VCO₂/VO₂. Energy expenditure (kcal/kg/min) was calculated as 3.91VO₂+1.1VCO₂[23].

Quadriceps muscles were excised and placed into room temperature Krebs-Ringer Buffer (KRB) containing 1% fatty acid free BSA and 5 mM glucose. For fatty acid oxidation, the muscle was cut into small pieces then incubated with 1 mM palmitic acid [³H palmitate, 5 μCi/μL] (Perkin-Elmer Life Sciences, Boston, MA, USA) in KRB for 2 hours. Oxidation was measured as the production of ³H₂O into the aqueous phase following a Folch extraction as previously described [16]. The organic extract was dried down in a speedvac, resuspended in 100 μL chloroform/methanol (2:1) and lipids were separated using silica thin layer chromatography plates. The extracted lipids were separated using heptane/isopropyl ether/acetic acid (60:40:4) as the separation solvent for 45 min. The plate was developed in iodine vapour and diglyceride and triglyceride bands were identified and scraped into scintillation vials for counting. Following overnight solubilization of the tissue remnants in NaOH, total proteins were measured by Bradford assay (BioRad, Hercules, CA, USA). Glucose oxidation was performed as previously described [12].

Liver glycogen was degraded into glycosyl units by incubating small liver pieces (~50 mg) in 1 M HCl at 100°C for 3 hours [24]. The extracts were neutralized with TRIS-KOH and the supernatant was assayed for glucose using a commercial colorimetric kit (Sigma). Results are expressed as μmoles/g glycosyl units per liver weight. For glucose-6-phosphate measurements, small liver pieces (~25 mg) were homogenized in 6% perchloric acid. The supernatant was clarified by centrifugation and neutralized with KOH in the presence of phenol red. The resulting potassium perchlorate precipitate was removed by centrifugation and the supernatant was enzymatically assayed for glucose 6-phosphate by following NADPH accumulation at 340 nm in presence of glucose 6-phosphate dehydrogenase. Results are expressed as nmoles/g liver sample. Neutral lipids were extracted from muscle and liver samples (30-50 mg) in a heptane/isopropanol (3:2) solution overnight at 4°C. The next day organic extracts were assessed for TG content using a commercial colorimetric kit (Roche Diagnostics) and total protein content was measured as described above. Results are expressed as μmoles of TG per gram of protein. Glucokinase (GK) enzyme activity was measured as previously described [12] however; solution volumes were modified for a 96-well plate and ODs were measured on a plate reader (BioTek Instruments, Inc. Winooski, VT, USA).

Results are expressed as means ± SEM. Results were analyzed by two-way ANOVA (two variables), one-way ANOVA followed by Dunnett's post-hoc test (one variable, more than two groups) or two-tail t-test (two groups). For indirect calorimetry measurements body weight and food intake, Group and Time P values are indicated in the graphs. Average area-under the curve (AUC) was calculated for light (12 hours) and dark (12 hour) energy expenditure. Significance was set as P value < 0.05, where NS represents non-significant values. All graphs and statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software, San Diego, CA, USA).

Recombinant ASP (rASP) activation and antibody neutralization were first tested in an in vitro system. Insulin (positive control, a known lipogenic hormone) and rASP significantly stimulated fatty acid uptake in murine adipocytes to a similar degree (Figure 1A), with maximal stimulation of 285 ± 49% (P < 0.01) for insulin and 243% ± 24% (P < 0.05) for rASP, vs PBS set at 100%. This is also presented as a 2.7-fold (P < 0.01) increase for insulin and 2.4-fold (P < 0.01) increase for rASP in the incremental area-under-the-curve (Figure 1B). The addition of increasing concentrations of Anti-ASP significantly reduced rASP stimulated fatty acid uptake to basal levels (maximal stimulation: 98% ± 4%, 120% ± 10%, and 144% ± 18%, vs PBS set at 100% for 0.2, 0.3 and 0.4 μg/mL Anti-ASP respectively, not significantly different from PBS) (Figure 1A and 1B).

Mice were treated by osmotic mini-pump for 4 weeks with neutralizing Anti-ASP or control non-immune IgG (NI-IgG). Plasma samples were taken 9 days after implantation of the osmotic pump and at the end of the study to track the amount of antibody delivered into circulation. Figure 2A shows that on day 9 rabbit NI-IgG and Anti-ASP antibodies were detected (~4% of starting material). However, by the end, the levels of both NI-IgG and Anti-ASP were greatly reduced (~0.1% of starting material). Since there was no difference in osmotic mini-pump rate for PBS, rabbit antibodies or rASP treatments (data not shown), reduced antibody levels may be attributed to enhanced antibody clearance. Anti-ASP blocking antibody treatment had no effect on mouse body weight (NI-IgG: 29.2 ± 0.4 g and Anti-ASP: 29.6 ± 0.3 g, NS, n = 6-7), average food intake (NI-IgG: 13.8 ± 0.5 kcal/day and Anti-ASP: 14.3 ± 0.5 kcal/day, NS, n = 6-7) or food efficiency (NI-IgG: 200.1 ± 21.2 kcal/g and Anti-ASP: 165.6 ± 25.8 kcal/g, NS, n = 6-7), consistent with our previous report during a 10 day study [17]. There was also no difference in adipose tissue depots, liver or spleen weight between the two groups (data not shown). IL-6 levels were slightly but significantly elevated in Anti-ASP mice (P < 0.05, Figure 2B). However, CRP levels, a marker of inflammation, were normal (Figure 2C). Anti-ASP did not have an effect on fasting glucose levels after nine days of treatment (NI-IgG: 6.22 ± 0.62 mmol/L and Anti-ASP: 6.58 ± 0.54 mmol/L, NS, n = 6-7). Furthermore, there was no difference in plasma adiponectin, insulin, C3, or lipid levels between the two groups at the end of the study (Table 1).

Twenty-four hour energy expenditure was significantly elevated in Anti-ASP treated mice (P < 0.0001, Figure 2D). Analysis of dark and light 12-h cycles separately indicates a significant increase with anti-ASP for both time periods (P < 0.0001, 2-way ANOVA, AUC Figure 2E). While RQ increased with feeding (indicating increased glucose utilization during this period), there was no change in RQ in either dark or light cycles between the two groups, suggesting a whole body increase in both glucose and fatty acid oxidation (Figure 2F). To investigate the mechanism of augmented whole body energy expenditure, we evaluated glucose and fatty acid consumption in skeletal muscle and liver. In the skeletal muscle, glucose oxidation was significantly increased more than two-fold in Anti-ASP treated mice (P < 0.01, Figure 3A). On the other hand, palmitate oxidation was reduced (P < 0.05, Figure 3B) with a shift of palmitate from oxidation towards incorporation into diglycerides (P = 0.05, Figure 3C) with no change in TG incorporation or total TG mass (data not shown). In the liver, glycogen content was decreased by 34% (P < 0.05, Figure 3D) associated with a trend in reduced glucose-6-phosphate content (P = 0.08, Figure 3E) and no change in glucokinase (GK) activity (NS, Figure 3F). Also, there was no difference in liver TG mass (NI-IgG: 40.6 ± 4.6 μmoles/g and Anti-ASP: 43.0 ± 5.3 μmoles/g, NS, n = 5-7).

Next, we treated WT mice with human rASP for 4 weeks to enhance ASP action. Immunoreactive circulating human rASP was detected at ~0.3 pmol/mL on day 9 and remained constant until the end of the study period (Figure 4A). Interestingly, NEFA levels were significantly lower in the rASP treated mice (P < 0.05, Figure 3B). Fasting glucose levels were similar between rASP and PBS treated mice (PBS: 6.77 ± 0.74 and rASP: 6.81 ± 0.58 mmol/L, NS, n = 5-6). In addition, no difference in plasma TG, insulin or adiponectin were detected between PBS and rASP treated mice, however C3 levels were significantly elevated in the rASP treated mice (Table 2).

Monitoring of body weight during the treatment period indicated a significant increase in the rASP treated group within the middle time period of treatment (days 9-24, Figure 4C, P < 0.0043). However, at the end, a similar final fasting body weight was measured (PBS: 30.0 ± 0.6 g and rASP: 30.3 ± 0.8 g, NS, n = 5-6). This increase in body weight was associated with a small but significant increase in cumulative food intake (P < 0.0001, Figure 4D). Food efficiency, calculated as food intake relative to body weight, was comparable between the groups (PBS: 127.5 ± 18.8 kcal/g and rASP: 151.0 ± 41.2 kcal/g, NS). During this time period, energy expenditure was significantly reduced in rASP treated mice (P < 0.0001, Figure 5A), in both the light and dark cycles (Figure 5B). While respiratory quotient changes with feeding, there was no difference between the groups (Figure 5C). Analyses of ex vivo skeletal muscle oxidation revealed no change in glucose or palmitate oxidation for rASP and control mice (Table 2). Meanwhile, liver glycogen content was significantly increased by 59% in rASP treated mice (P < 0.05) with no change in glucose-6-phophate or TG content (Table 2).