IGF, insulin-like growth factor
GH, growth hormone
GHRKO, growth hormone receptor knock out
MLS, LS, (maximum) lifespan
In this post I would like to expound on an idea that allows us to make sense of the studies on adult-onset GH/IGF1 deficiency. As a disclaimer, let me emphasize that I am not working in this field, but I do try to keep up with the literature.
The search for independent pathways: why study these animals at all?
One unresolved question is linked with the evolvability and mutability of lifespans. Given that LS is quite flexible within and between species, we would expect the existence of lifespan assurance mechanisms, and most likely they should include signalling pathways and transcription networks, because these can change quickly over reasonable time frames. This is the optimistic view that many biogerontologists agree with. In contrast, the pessimistic view holds that CR-related pathways are a curiosity and normally changes in lifespan require thousands of independent mutations in wildly different pathways, precluding significant human lifespan extension using drugs or other interventions.
So far we know that there exist partly redundant longevity-assurance pathways that are all loosely linked to CR, anabolism and perhaps cellular "quality control" and multistress resistance. It will be very important to define the degree of overlap between these pathways to clarify whether the optimistic or pessimistic view is closer to reality. We need to know if there are pathways that are truly distinct from CR or that produce additive benefits with CR even if they are redundant.
In the end, we need to know how to combine interventions to achieve the best results, i.e. which genes are epistatic and hence in the same signalling cascade. We have to answer questions such as: How much of the effect of CR is due to GH and IGF1? Is protein and methionine restriction operating through the same mechanisms as CR? It seems, both of these overlap with CR but are distinct. What about mTOR? It seems linked to GH, IGF1 and CR but distinct (long-lived GH dwarfs have diminished mTOR signalling for example but mTOR inihibtion produces a different phenotype from dwarfism). What about the two novel players, c-myc and H2S?
GH vs IGF1 and other downstream pathways: what is more important to LS?
The evidence now is clear-cut that IGF1 explains only some of the benefits of GH-deficiency and it might be especially linked to cancer (see e.g. Ashpole et al. 2017).
From Bartke 2017, highlighting some differences:
Many of the physiological actions of GH are mediated by IGF-1, which also plays a key role in the negative feedback control of GH expression. Circulating IGF-1 levels provide a very useful measure of GH secretion and activity...Evaluating the functional and pathological implications of the concomitant decrease in GH and IGF-1 levels is complicated by the fact that while some actions of GH are mediated by IGF-1, some are not...For example, GH is lipolytic while IGF-1 is not, GH promotes insulin resistance while IGF-1 reduces it and mimics various insulin effects...
Hepatic production of the diffusible and toxic gas H2S is quickly emerging as the other important pathway downstream of GH (Hine et al. 2017). Dwarf mice like Ames or GHRKO produce more H2S in the liver and this can be counter-acted by giving GH and induced by a somatostatin analog. In contrast, IGF1 has no influence. The responsible enyzme for H2S production is cystathionine gamma lyase which is upregulated in the liver of these animals. The authors also provide some evidence that ATF4 and increased substrate availability through autophagy can induce H2S production and that JAK/STAT signalling downstream of the GH receptor modulates H2S production (at least in vitro) .
Early life treatment of Ames dwarfs with GH can reduce H2S production in the long term. The authors used the same treatment as the controversial study by Panici et al, 2010, which first showed that this treatment reverses the lifespan benefits of dwarfism. This longitudinal correlation does point to a causal role.
However, mice with less robust LS extension also produce more H2S, e.g. IRS1 KO and FGF21 transgenic animals. Simple hypothyroidism also promotes H2S expression, certainly arguing that this pathway is important in nutrient sensing. Finally, the authors show that mice unable to produce H2S do not respond to fasting by downregulating IGF1 and T4, which puts H2S both upstream and downstream of GH/thyroid signalling and suggests an involvement in feedback regulation.
GH models vs CR: which one is more robust and why?
Does CR operate solely through GH and IGF1 reduction? Do these interventions engage the same downstream pathways and hence aren't additive? I've blogged on this issue before and I think we have yet to reach a consensus.
If CR operates through GH then it makes no sense why CR would be more robust than GH dwarfism in certain ways: Adult-onset CR works well if proper protocol is followed (Ingram and de Cabo 2017). In contrast, adult-onset GH/IGF1 interventions flounder (see e.g. Ashpole et al. 2017, Duran-Ortiz et al. 2016 etc.). Second, CR produces similar LS extension to pituitary dwarfism or GHRKO even though there is plenty of residual circulating GH and Igf1. Third, CR has been successfully studied in dozens or hundreds of mouse genotypes and other organisms and GH/IGF1 disruptions haven't. (These studies have shown that CR works in many but not all organisms at least when it comes to some of the benefits.)
Over these many studies proper protocol has emerged as an important concept in the CR literature but no one knows why exactly. Not only does the right protocol enable adult-onset CR but it is also speculated that the wrong protocol, i.e. too severe CR, is responsible for many of the null studies of CR (e.g. recombinant strain data). No one has experimented with gradual growth hormone deficiency but it is the logical next step.
Adult-onset controversy: do we need a proper way to gradually decrease GH/IGF1?
Adult-onset IGF1 deficiency, a reduction of circulating hormone by at least 80%, extends female mean LS and somewhat shortens the mean LS of males. This effect seems to be mediated by a reduction of cancer in males. No effect on MLS. (Ashpole et al. 2017, Sonntag lab)
GH-receptor knockout at 6 weeks of age increases female MLS and the change is approx. from 150 to 177 weeks of age. No effect on mean LS or on males! (Junnila et al. 2016, Kopchick lab)
Only females could be used in the next study. It was shown that PAPP-A knockout during adulthood extends lifespan. PAPP-A increases the local availability of Igf1 by cleaving binding proteins and knockout from birth on extends mouse lifespan. However, I must admit that I don't know much about this model nor the involved group. At a quick glance, their mice always tend to be on the short-lived side. (Bale et al. 2017)
"disruption of the [insulin receptor, IR] in peripheral tissue of 15-weeks-old mice" does not extend LS, but homozygous knockout is harmful to males. (Merry et al. 2017, Ristow/Zarse lab) However, I would not take this too seriously, because IR interventions have always produced mixed results.
In all of these studies female mice respond better than males and if any benefits are realized they are much smaller than in mice who are GH/Igf1 deficient from birth. This could argue that Igf1 is necessary to maintain health, as Sonntag argues, and deficiency "programs" the mice to live longer only when present during early life. However, I believe there is a better working hypothesis. While I do agree with Sonntag and many others that late in life the anabolic effects of Igf1 could be health-promoting, it seems to me like adult-onset GH deficiency
Why does CR, which operates partly through reduced Igf1, work so well in adult mice?
This may be the reason. Over 30 years ago Weindruch and Walford pioneered the gradual induction of CR (see also de Cabo and Ingram 2017). From the original paper:
"We decided that the influences on survival of dietary restriction initiated in adults warranted further study because in previous studies (i) the dietary restriction was imposed abruptly rather than gradually"
If we want to properly test the hypothesis that loss of GH only improves lifespan when it exist from birth, we will need to develop an improved protocol to slowly induce adult onset growth hormone deficiency mimicking the standard protocol employed in the CR literature. While I don't want to ignore the clear null results it would be also less than prudent to ignore the seminal work of Walford and Weindruch highlighting the success of gradual onset CR, given the extreme similarity between CR and GH-interventions.
H2S is emerging as one of the most important mediators of CR. It might be too early to tell, but we can speculate whether the adult-onset GH treated animals failed to upregulate H2S.
Adult onset interventions
Ashpole, N. M., Logan, S., Yabluchanskiy, A., Mitschelen, M. C., Yan, H., Farley, J. A., ... & Georgescu, C. (2017). IGF-1 has sexually dimorphic, pleiotropic, and time-dependent effects on healthspan, pathology, and lifespan. GeroScience, 39(2), 129.
Bale, L. K., West, S. A., & Conover, C. A. (2017). Inducible knockdown of pregnancy‐associated plasma protein‐A gene expression in adult female mice extends life span. Aging Cell.
Disruption of the GH Receptor Gene in Adult Mice Increases Maximal Lifespan in Females. Junnila RK, Duran-Ortiz S, Suer O, Sustarsic EG, Berryman DE, List EO, Kopchick JJ. Endocrinology. 2016 Dec;157(12):4502-4513. Epub 2016 Oct 12.
Merry, Troy L., et al. "Impairment of insulin signalling in peripheral tissue fails to extend murine lifespan." Aging Cell (2017).
Lifespan extension in the spontaneous dwarf rat and enhanced resistance to hyperoxia-induced mortality. Sasaki T, Tahara S, Shinkai T, Kuramoto K, Matsumoto S, Yanabe M, Takagi S, Kondo H, Kaneko T. Exp Gerontol. 2013 May;48(5):457-63. doi: 10.1016/j.exger.2013.02.015. Epub 2013 Feb 20.
GH and ageing: Pitfalls and new insights. Bartke A, Darcy J. Best Pract Res Clin Endocrinol Metab. 2017 Feb;31(1):113-125. doi: 10.1016/j.beem.2017.02.005. Epub 2017 Feb 24. Review.
Hine, Christopher, et al. "Hypothalamic-Pituitary Axis Regulates Hydrogen Sulfide Production." Cell Metabolism 25.6 (2017): 1320-1333.
Ingram, Donald K., and Rafael de Cabo. "Calorie Restriction in Rodents: Caveats to Consider." Ageing Research Reviews (2017). APA