hNAG-1 increases lifespan by regulating energy metabolism and insulin/IGF-1/mTOR signaling

hNAG-1 female mice have increased lifespan

A total of 200 female hNAG-1 mice and their wild-type (WT) littermates from two transgenic lines (1377 and 1398) were placed on either 10% (LFD) or 60% fat diets (HFD) (n = 25 in each line-genotype-diet group). Log-rank tests showed lifespan was significantly longer for hNAG-1 mice than for WT mice in both lines with both diets (Figure 1a and 1b, Table 1). On LFD, the median lifespan of hNAG-1 mice was 18.8 weeks (19.5%) and 13.0 weeks (12.8%) longer than WT littermates, respectively for lines 1377 and 1398 (log rank χ² = 4.60, 1 d.f., p=0.032 for line 1377; log rank χ² = 8.03, 1 d.f., p=0.005 for line 1398, Table 1). On HFD, the median lifespan of hNAG-1 mice was 22.4 weeks (23.6%) and 33.6 weeks (43.6%) longer than WT littermates, respectively for lines 1377 and 1398 (log rank χ² = 13.0, 1 d.f., p<0.001 for line 1377; log rank χ² = 28.9, 1 d.f., p<0.001 for line 1398, Table 1). Consistent with previous findings that HFD could shorten lifespan[8, 9], survival of line 1398 WT mice in the HFD group was significantly lower than the LFD group (median was 24 weeks lower, log rank χ² = 11.2, 1 d.f., p < 0.001). However, there were no significant differences in lifespan between HFD and LFD in either WT or hNAG-1 transgenic mice in line 1377. Consistent with previous findings that young hNAG-1mice (20 to 30 weeks) are resistant to obesity on a 12 week-long HFD [14], both mean body and abdominal white adipose tissue (WAT) weights are significantly reduced in old hNAG-1 mice (> 95 weeks) after prolonged feeding with either diets (Figure 1c and 1d, Supplementary Figures 1 and 2). With aging and especially on HFD, mice (C57/BL6) develop spontaneous skin lesions and mass growths (tumors or non-tumors) in organs and experience lymph node and spleen enlargement [17-19]. Fewer hNAG-1 mice have skin lesions, spleen enlargement and gross liver lesions compared to WT littermates (Supplementary Tables 1 and 2). Percentages of mice with lymph node enlargements were similar between hNAG-1 and WT mice. Collectively, hNAG-1 overexpression increases lifespan in female mice. hNAG-1 mice are resistant to diet-induced obesity, as well as age- and diet-induced pathological lesions.

hNAG-1 female mice have increased circulating growth hormone level

In mice, disruption of growth hormone (GH) signaling that leads to major alterations in IIS has a well-documented positive impact on lifespan [4]. The long-lived GH-resistant GH receptor (GHR)-KO mice which have high GH level, as well as Ames dwarf and Snell dwarf mice lacking GHexhibit dwarfism, increased subcutaneous adiposity, change in tissue sizes, and increased insulin sensitivity[1, 20]. Although hNAG-1 mice are significantly shorter in body length (Figure 2a) there are no significant differences in tissue sizes, length of femurs or organ weights (data not shown). Interestingly, circulating GH level is substantially increased in terminal serum of hNAG-1 mice (Figure 2b). This difference is more dramatic in young 1398 female mice (Supplementary Figure 3) suggesting a possible GH-resistant phenotype or simply a feed-back regulation due to reduced IGF-1 level in hNAG-1 mice (Figure 2c). However, the expression of GHR in liver, WAT, or brain tissue at both mRNA and protein levels was the same for hNAG-1 and WT mice (data not shown). The downstream GH signaling cascade including phosphorylation of JAK2, ERK1/2, and SMAD were not different in hNAG-1 mice. Recombinant hNAG-1 did not inhibit GH-induced downstream signaling or IGF-1 secretion in cell culture studies using HepG2 cells (Data not shown). These data suggest that hNAG-1 mice are not GH resistant and the effects on longevity in hNAG-1 mice appear to be independent of GH. However, the cause of the reduction in IGF-1 levels and elevated GH levels in hNAG-1 mice is unclear.

hNAG-1 female mice have increased insulin sensitivity

The resistance against HFD-induced obesity suggests that the transgenic expression of hNAG-1 in mice might positively affect age-associated metabolic disorders, such as insulin resistance. Similar to young mice [16], insulin and leptin levels are substantially reduced in hNAG-1 mice on LFD or HFD in old mice (Figure 2d).

Insulin tolerance test (ITT) shows that 95week-old (line 1398) hNAG-1 mice have significantly improved insulin sensitivity compared to WT littermates (Figure 2e). Because the body weights between hNAG-1 mice and the WT mice are quite different, it was not possible to match same weight at the same age. We decided to use the mice at the same age, which we think is more appropriate than weight. We also calculated the results using % of basal glucose level and found similar results as using glucose concentration (data not shown). Basal glucose levels are also significantly lower in 95 week-old hNAG-1 transgenic mice. Thus, consistent with association between improved insulin sensitivity with better survival [4], older hNAG-1 mice sustain higher insulin sensitivity which may play a role, in part, for the increased lifespan as observed above. hNAG-1 mice mimic caloric restriction with reduced circulating IGF-1 levels, improved insulin sensitivity, and extended lifespan. There is no significant difference in food intake between the mice on a LFD or HFD in both lines (Supplementary Figure 4 and 5).

hNAG-1 have increased energy expenditure and metabolism

Increased energy expenditure can extend lifespan and alter insulin sensitivity and resistance to obesity [21, 22]. Previously, we have shown that hNAG-1 expression increases metabolic oxidation and thermogenesis in male mice which is responsible, in part, for the resistance of hNAG-1 mice to obesity [14]. Enhanced O₂ consumption and higher heat expenditure are consistent with increased oxidative metabolism and is associated with protection from HFD-induced obesity and enhanced longevity [1, 23]. In this study, heat production is slightly but statistically significant lower during the day in female line 1398 hNAG-1 mice (Figure 3a). During the night, hNAG-1 mice have significantly much higher heat production than WT littermates (Figure 3a) with no difference in physical activity between the mice (data not shown). We then examined the expression of representative thermo-genesis genes in brown adipose tissue (BAT) in both 30 wk and 95 wk old mice. However, unlike the young male mice where most of these genes (UCP-1, PGC1a, ECH-1, Cox8b, PGC1b, etc) are up regulated in BAT [14], only few genes were up regulated in BAT in female mice due to larger variation (Supplementary Figure 6). In addition, hNAG-1 mice utilize more oxygen than WT mice during the night (Figure 3b) and have reduced respiratory quotient (RQ) during the day (Figure 3d), implying greater reliance on fats, as opposed to carbohydrates, as an energy source.

hNAG-1 female mice have decreased mTOR signaling activity

To further understand the mechanism of extended lifespan in hNAG-1 mice, we used whole genome microarray analysis to examine differential gene expression in abdominal WAT in line 1398 hNAG-1 mice. Differential category expression analysis showed significant differences between hNAG-1 mice and WT mice in key pathways in the regulation of metabolism and mammalian lifespan, including down regulation of IGF-1, PI3K, AMPK, and RPS6K families (Supplementary Fig. 7, Table 2). We validated a set of genes (mTorc2, GHRH, IGF1R, RPS6k, FOXO1, Raptor, Mlst8, Mapkap1, Pik3c9) by qRT-PCR in young female mice (Figure 4a). The expression pattern of all of these genes confirmed the microarray data with a similar expression pattern in old mice (> 95 wk) except for FOXO1 (Figure 4b).

mTOR pathway is emerging as a key regulator of aging, and is linked to IIS pathway [7, 24]. Inhibition of mTOR activity or knockout the downstream effector of mTOR, S6K1, increases lifespan in mice [25, 26].

Phosphorylation of members of mTOR signaling pathway (IGF1R, mTOR (ser2448), AKT (ser308), p70S6k (ser 384), and 4EPB-1) is down regulated in the WAT of young 1398 hNAG-1 mice (Fig. 3f), and similar pattern is observed in old mice (> 95 wk, Figure 5). The expression of Sirt1 and Sirt6 [6, 27] between hNAG-1 and WT mice was not different in liver (data not shown) and abdominal WAT (Figure 5).

Animals

The hNAG-1 transgenic mice were generated as described previously [34]. hNAG-1 is expressed in most tissues, liver is very low but good expression is found in the skin, colon, kidney, brain and is highly expressed in the WAT and BAT (data not shown). Experiments were performed in accordance with the “NIH Guidelines for the Use and Care of Laboratory Animals”. A total of 200 WT and hNAG-1 female transgenic mice at 50 wk old from line 1377 and 1398 were randomized and fed LFD or HFD (Research Diet) and water ad libitum (n=25/group). A total of 200 male mice were initially included for longevity study. However, due to relative high percentage of male hNAG-1 mice that develop, for unknown reasons, urinary blockage which can cause early death the male mice were removed from study. Food intake was measured once a week for 40 weeks. Body weights were measured twice weekly over the course of the study. The mice were euthanized and total fat was removed and weighed. Physical activity, VO₂, VCO₂, and body heat production were measured at 30 wk of age in line 1398 female mice as described previously [35].

Insulin Tolerance Test

Female mice, 95 wk old (1398), were fasted for 6 h and insulin (Sigma, St. Louis, MO) was intraperitoneally injected into mice at dose of 0.5 U/kg BW. Blood was collected from the tail at 0, 15, 30, 60, and 90 min after injection and glucose concentration was analyzed using a glucose meter.

Microarray Analysis

RNA samples from abdominal WAT of line 1398 female NAG-1 mice and WT mice at 30 wk old were extracted as described [16]. Sample hybridization was performed as previously described [36]. Microarray analysis was performed using Affymetrix Mouse Genome 430 2.0 Array by Microarray core at NIEHS. Pathway analysis was carried out using Ingenuity Pathway Analysis (IPA, http://www.Ingenuity.com). Separate statistical analyses were conducted for each transgenic line.

Real-time PCR

RNA samples from abdominal WAT and BAT of line 1398 female NAG-1 mice and WT mice at 30 wk old were extracted as described[16]. One microgram of RNA was reverse transcribed using iScript cDNA synthesis kit from BioRad. Real-time PCR assays were performed using Taqman master mix and primers (Applied Biosystems, Foster City, CA) by MyiQ PCR detection system (BioRad) for semi-quatitative real-time PCR analysis. Relative fold changes were calculated using the delta delta Ct method, with β-actin or GAPDH serving as control genes.

ELISA Analysis

Terminal bleeds from mice were incubated at room temperature for 1 hour in serum separator tubes (Sarstedt, Nümbrecht, Germany) and then spun at 10,000 × rpm for 5 minutes. Serum was collected and stored at −80°C until analysis. Mouse leptin, mouse growth hormone, human NAG-1, mouse IGF-1 (R&D, Minneapolis, MN) and mouse insulin (Alpco, Salem, NH) ELISA kits were used according to manufacturer's instructions.

Western Blots

Western Blots were performed as described[16]. Total protein in abdominal WAT was isolated using RIPA buffer (Pierce, Rockford, IL) supplemented with sodium fluoride and sodium vanadate. A total of 45 μg protein was electrophoresed on a 4-15% SDS-polyacrylamide Tris-Hcl gel (Bio-Rad) at 175V for 1 h. After transfer, membranes were blocked with 5% non-fat dry milk in 1x TBST (50Mm Tris pH7.5, 150 mM NaCl, 0.1% Tween-20) at room temperature for 1 hour. The membrane was probed overnight at 4°C with appropriate antibodies, at the dilutions as recommended by their manufactures. The next day, blots were rinsed and probed with the appropriate horseradish-peroxidase secondary antibody for 1 hour at room temperature in 5% milk in TBST and illuminated with Western Lightening TM Plus-ECL Enhanced Chemiluminescence Substrate assay kit. The membrane was stripped using Restore Western Blot Stripping Buffer according to manufacturer's instruction. After stripping, the membrane was re-probed for GAPDH for loading control.

Statistical Analysis

Data were analyzed by Dr. Grace Kissling of the NIEHS Biostatistics Branch who determined the appropriate method to use. Survival was compared between genotypes and diets using log-rank tests; survival curves were plotted from Kaplan-Meier estimates. Body weight was compared between genotypes and diets using polynomial growth curve analysis. Food consumption was compared between genotypes and diets using Mann-Whitney tests with Bonferroni correction for multiple testing. All p-values are two-sided and considered significant at the 0.05 level. Body weight and food consumption plots show mean ± standard error of the mean.