The effect of a pharmaceutical ghrelin agonist on lifespan in C57BL/6J male mice: A controlled experiment

Abstract Interventions for animal lifespan extension like caloric restriction (CR) have identified physiologic and biochemical pathways related to hunger and energy‐sensing status as possible contributors, but mechanisms have not been fully elucidated. Prior studies using ghrelin agonists show greater food intake but no effect on lifespan in rodent models. This experiment in male C57BL/6J mice tested the influence of ghrelin agonism for perceived hunger, in the absence of CR, on longevity. Mice aged 4 weeks were allowed to acclimate for 2 weeks prior to being assigned (N = 60/group). Prior to lights off daily (12:12 cycle), animals were fed a ghrelin agonist pill (LY444711; Eli Lilly) or a placebo control (Ctrl) until death. Treatment (GhrAg) animals were pair‐fed daily based on the group mean food intake consumed by Ctrl (ad libitum feeding) the prior week. Results indicate an increased lifespan effect (log‐rank p = 0.0032) for GhrAg versus placebo Ctrl, which weighed significantly more than GhrAg (adjusted for baseline weight). Further studies are needed to determine the full scope of effects of this ghrelin agonist, either directly via increased ghrelin receptor signaling or indirectly via other hypothalamic, systemic, or tissue‐specific mechanisms.


| INTRODUC TION , RE SULTS AND DISCUSS ION
Of the well-studied effects on lifespan in mouse models, detailed mechanisms underlying the health and longevity benefit of caloric restriction (CR) are still being investigated despite some limitations on the practical applications to humans (Longo et al., 2015;Sharples et al., 2015). A major limitation of animal models is that they cannot self-report hunger or other physiological sensations that would inform mechanistic work. Since ghrelin was first described and noted as a growth hormone secretagogue receptor (GHS-R) agonist (Kojima et al., 1999), much study has focused on the effects on hunger and appetite regulation along with other aspects of energy balance (Bouillon-Minois et al., 2021;Lewiński et al., 2021;Ouerghi et al., 2021).

Investigators examining effects on cognition in mouse models
have posited that the mechanism may involve interoceptive cues or signaling, rather than reduced energy intake per se (Dhurandhar et al., 2013;Kunath et al., 2015). In those studies, oral administration of a ghrelin agonist (LY444711, an orally active compound that binds with high affinity to and is a potent activator of the growth hormone secretagogue receptor 1a [GHS-R1a] receptor; Lugar et al., 2004), reduces Alzheimer's disease pathology and improves cognition in the APP-SwDI mouse model (Dhurandhar et al., 2013). Treatment also reduced levels of amyloid beta (Aβ) and neuroinflammation (as measured by microglial activation) at 6 months of age compared to controls (Ctrls), like the effect seen in the 20% CR group (gp) but with no significant difference in body weight (BW) or % body fat (Dhurandhar et al., 2013).

LY444711 binds to the human ghrelin receptor (GHS-R1a) and
is a functional agonist. This agonist (Lugar et al., 2004) produced orexigenic behavior in rodents, including stimulated energy intake (food consumption; 40%> than Ctrl at 10 mg/kg, 50%> than Ctrl at 30 mg/kg dose), positive energy balance (23%> BW with 2 weeks treatment at 10 mg/kg), acute higher respiratory quotient (RQ) with increased dose (3, 10, and 30 mg/kg) and increased adiposity (greater fat mass but no significant difference in lean mass by DXA; Tschöp et al., 2000). The authors conclude this substance is orally active with an extended half-life relative to native ghrelin. Despite various studies on ghrelin effects on food intake and body composition, studies of ghrelin agonists on lifespan-extending effects in rodent models are lacking. We tested the hypothesis that a pathway related to perceived hunger, as induced by an oral, exogenously administered orexigenic agent (i.e., a synthetic ghrelin agonist), would differentially affect lifespan in mice when pair-fed to Ctrls.

| Food intake
Food intake before the start of the protocol was significantly greater in Ctrls (3.25 ± 0.35 g/day) compared to GhrAg animals (2.94 ± 0.61 g/ day, p = 0.001, 95% CI for the difference = 0.13-0.49 g). On protocol, per study design, the Ctrl gp mean food intake was pair-fed to the GhrAg gp -Ctrl intake in Figure 1a.

| Longevity
Overall, GhrAg animals lived significantly longer than Ctrls (logrank test p = 0.003, unadjusted Cox proportional hazards model p = 0.004, Figure 1c), which remained significant after adjusting for baseline BW (p = 0.012). However, the details over the lifespans are complex. The age when the GhrAg animals achieve both the 50th and 90th quantiles of survival is later than the Ctrl animals (Ctrl:

| Observed pathologies
We saw gp differences in those requiring euthanasia per veterinary protocols ( Table 1): χ 2 (1) = 4.48, p = 0.034. The largest difference in pathologies observed in gross necropsy was the higher rate of enlarged seminal vesicles in the GhrAg gp, and they also exhibited higher rates of being moribund, resulting in euthanasia. The most common health issue observed in both gps was ulcerative dermatitis, some of which was resolved after treatment. See Supplement for additional details of how animal health status was monitored.

| DISCUSS ION
The results of this experiment indicate the ghrelin agonist with weekly adjusted pair-feeding lengthened lifespan and decreased BW. Since the GhrAg gp were pair-fed to Ctrl gp's mean ad libitum (AL) food intake, it is not known whether GhrAg animals perceived hunger or could be considered to be under a state of negative energy balance as reflected in lower BW at times compared to Ctrls.
In another study, LY444711 treatment for 2 weeks increased adiposity by stimulating food consumption and limiting fat utilization in male rats, while increasing RQ and carbohydrate utilization during the dark phase (Lugar et al., 2004). This possible 'perception' of inadequate food resources, even at low levels, may stimulate adaptive mechanisms to change body composition, which may also affect rates of aging (Kaiser et al., 2012). All GhrAg-treated animals consumed all food delivered daily in the present study. This amount was, at times, higher than observed in many other studies of the same sex of this mouse strain but may be attributed to the singlehousing condition (Rowland et al., 2018;Schipper et al., 2018). It is unknown whether GhrAg animals would show a similar degree of hyperphagia over the lifespan as has been observed in acute studies.

Strengths of this study include a rigorous experimental design,
including the single chemical identity tested, dosing related to similar diurnal patterns in humans, diet provision and monitoring with energy intake matched gps, strain historic context and information for a normative longevity profile and large sample size. Limitations of interpretation include the use of an inbred strain, use of only male mice, AL feeding variability among Ctrls with the treatment animals all being pair-fed to the Ctrl gp mean, time of food access (24 h in AL), feeding behaviors (daily provision with agonist), no ability to make comparisons with the time of day feeding, and long duration of GhrAg exposure. We did not systematically compare health factors to quantify the health span between groups. Increased physical activity has been observed in short-term experiments with this same ghrelin agonist but was not measured over the entire protocol here. Future studies may wish to consider machine-assisted analysis of movement to enhance the prediction of longevity outcomes, as reported by Hession et al. (2022).

F I G U R E 1 (a)
Mean food intake (g/mouse/day) of the Ctrl gp; (b) BW by gp (g) over the duration of the protocol; (c) lifespan outcomes by gp. For the first 88 days, the GhrAg pill concentration was 1.66% (by weight), after which, it was increased to 2.4% to achieve the 30 mg/ kg dose. Diet was initially AIN-93G for ~6 months, then switched to AIN-93M for the remainder of the experiment. Note: Shaded vertical areas in (b) indicate periods where gp means were significantly different for greater than one measurement interval. Some individual measurements that were significantly different but noncontiguous are not shaded.

TA B L E 1
Counts of death conditions and rates of ulcerative dermatitis and gross necropsy observations by gp. Other c 11 13 a Between gps found dead versus euthanized χ 2 (1) = 4.48, p = 0.034. b Not tested due to fewer than 5 cases per cell.
c Other observations of the euthanized animals included blood pooling, tumors, and abdominal and bladder abnormalities.

| Other ghrelin-related GHS-R agonists show similar effects in rats
Similar to the Lilly compound (LY444711), high-dose treatment with single agonist compounds (BIM-28131 and BIM-28125, which have higher binding affinity to ghrelin receptor GHS-1a than natural ghrelin) or exogenous human ghrelin have been tested in rat models, resulting in increased BW gain by promoting fat mass, with only BIM-28131 showing a significantly increased lean mass (Strassburg et al., 2008). Food intake increased during treatment with BIM-28131 or ghrelin, but no effects on BW-adjusted energy expenditure were observed over the 4-week study. With the lower dose, only BIM-28131 had a significant effect on BW. These results characterize BIM-28131 as a promising GHS-R agonist with an attractive action profile for the GhrAg of catabolic disease states such as cachexia (Strassburg et al., 2008).

Reports of mechanistic studies show increased hypothalamic
SIRT1 activity and SIRT1 protein expression in heart tissue when mice models were exposed to ghrelin signaling potentiators such as rikkunshito and atractylodin (Fujitsuka et al., 2016). Recent evidence in Drosophila melanogaster shows interesting linkages between CR and tissue-specific aging (CLOCK gene regulation in eyes) that extend lifespan (Hodge et al., 2022). Work in Caenorhabditis elegans demonstrates the roles of serotonin and dopamine receptor antagonism in mimicking CR that result in increased longevity by way of the Fmo genes (Miller et al., 2022). Mammals and C. elegans share a common ancestral Fmo gene (Petalcorin et al., 2005).

| CON CLUS IONS
These observations suggest ghrelin as a growth hormone secretagogue may influence a network of factors related to aging or body composition. Greater focus on tissue-specific effects, neurotransmitters, gene expression (e.g., UCP2; Wang et al., 2010), and metabolic phenotypes (e.g., body temperature, physical activity, diet composition effects and potential sex differences) as well as use of composite frailty index measures (Kane et al., 2016) are warranted.

ACK N O WLE D G E M ENTS
The authors are grateful to Dr. Steve Austad for comments on an earlier draft of this manuscript.

FU N D I N G I N FO R M ATI O N
Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under awards R01AG043972, P30DK056336, P30 AG050886, and P30NS47466.
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

CO N FLI C T O F I NTER E S T S TATEM ENT
None of the authors have any relevant items to disclose related to this research. The opinions expressed are those of the authors and do not necessarily represent those of the NIH or any other organization. The ghrelin agonist used was donated by Eli Lilly and Company to UAB. Indiana University has received grants, contracts, and inkind donations from Eli Lilly and Company for additional scientific and educational projects.

O PEN R E S E A RCH BA D G E S
This article has earned an Open Data badge for making publicly available the digitally-shareable data necessary to reproduce the reported results. The data is available at [https://osf.io/34cr8/ ].

DATA AVA I L A B I L I T Y S TAT E M E N T
Data and analytic code for the project entitled 'Aim 5 -Ghrelin agonist longevity outcome' are available here: https://osf.io/34cr8/.
Data are available under the terms of CC BY-NC-SA 4.0 by the authors.