Coffeehouse notes on GH/IGF-1, CR and life span

Both Ames and growth hormone receptor knock-out (GHRKO) mice have disruptions in the GH pathway (11).

Surprisingly, however, studies suggest that GHRKO mice show more robust life span extension than Ames dwarfs. For one, the GHRKO model holds the absolute longevity record for lab mice. Secondly, it is much less responsive to caloric restriction (CR), while reducing calories extends the lifespan of Ames dwarfs quite robustly. Third, Ames dwarfs lack TSH, which might be beneficial, yet do not outlive GHRKOs (alternatively, this means TSH is only a minor player).

However, this notion in itself is controversial, because these mice have defects in the same pathway. In Ames dwarfs growth hormone (GH) is absent and GHRKO mice simply lack the GH receptor.

The hypothesis of the highly robust GHRKO mouse has two important implications worth exploring. On the one hand, it suggests that if CR fails to increase lifespan in this model it may work exclusively through GH signaling, which I don't believe. In addition, it would mean that the GHR receptor might have functions independent of GH binding and that the GHRKO mouse is meaningfully different from the Ames dwarf. When I talked to researchers at the last conference, some were quite convinced that there is a "magic ingredient" to the GHRKO mouse responsible for its robustness. Let's call it magic, because no one really knows what it is.

So it is reasonable to hypothesize that GHRKO, with the magic pathway fully suppressed, imposed on top of Ames dwarfism, magic pathway still partly active, would further improve life span and healthspan. Before we discuss the paper by Gesing et al. 2016 (1) testing this proposition, I want to give an introduction to GH in aging and briefly review key studies and controversies.

A. The biology of GH signalling - what is the magic pathway?
What other functions of the GHR are conceivable in addition to binding pituitary GH?

Many cells* and cancers are known to produce autocrine GH which never leaves the cell (7), which could also promote a local increase in IGF-1. Interestingly, local IGF-1 levels are quite variable between different tissues in GH-deficient models (e.g. high brain IGF-1 in Ames), but, unfortunately, I am not aware of any head to head comparisons of GHRKO vs Ames.

*GH "has, for instance, been found postnatally in the central and peripheral nervous systems, in the immune system (in the thymus, spleen, tonsils, lymph nodes, Peyers patches and Bursa of Fabricius), in reproductive tissues (the ovary, oviduct, uterus, mammary glands, placenta, testis and prostate), in gastrointestinal tissues (in hepatic tissues, pancreatic tissues, salivary tissues and the alimentary tract), in skeletal and dental tissue, in muscular tissues, in cardiovascular tissues (the heart and blood vessels), in respiratory tissues (in lungs and gills) and in integumentary tissues (skin)"

Alternative ligands for the GH receptor exist such as prolactin, but pituitary dwarfs lack the cells which secrete this hormone. Could there be other ligands which activate the GHR in pituitary dwarfs? Perhaps minor ligands could have some effect if dwarfs are hypersenstive to GH as reported in some studies (13).

Perhaps the receptor is necessary for the function of other signalling molecules than GH and related molecules. Although, it is true that the receptor may signal independently of the canonical JAK2 and STAT5 way, instead involving Src, this function appears to be ligand dependent (6).

The receptor itself may play a function independent of its ligand, but it is unclear how this relates to aging. Ligand-indepdent dimerization has been described, but I don't think this leads to downstream signalling.

Brooks, A. J., & Waters, M. J. (2010) (6b)

Important concepts in GH signalling.

Coffeehouse notes would not be complete without a sketch on a napkin-style drawing.
It is clear from the diagram that mutations in GHR and upstream of it should have more robust effects on lifespan than those in the IGF-1 pathway. Hypothetical pathways converging on GHR are depicted as well.

B. CR, Dwarfism and GHRKO
Let's briefly review the studies of CR superimpositon. Sorry, for the mixed life span meassures and guesstimates, but I am working with the inconsistent reporting as seen in the literature. A more detailed breakdown of the studies can be found under "D. Data".

There is a wall at 1150-1200 days for mean and ~1400d for maximum life span, which is hard to break through. Different GH dwarfs - GHRKO (2, 3), GHRHKO, Ames (9), Snell - live to this age and so do CR mice in some studies. For control mice the limit is around 900 and 1150 days, respectively.
Taking these numbers into account, the studies by Bartke 2001, Gesing 2016 and Sun 2013 suffer from short-lived controls. The dwarfs in Bartke 2001 and Sun 2013 are also relatively short-lived, and possibly so the male dwarfs in Gesing 2016.

At the same time, two of these three papers find the biggest additional benefit of CR on top of dwarfism. Gesing 2016, in contrast, crossed Ames dwarfs and GHRKO mice. They do find some non-significant additive benefit of the cross on average lifespan of males when compared with the rather short-lived male dwarfs or GHRKO mice.

The studies by Bonkowski 2006 and Bonkowski 2009, in turn, managed to raise relatively long-lived controls and GHRKO dwarfs and they do not find any lifespan benefit from adding CR. Nevertheless, even under these stringent conditions, CR does significantly extend maximum lifespan of females.

And the most odd of all papers, Garcia 2008 found that CR actually shortens the lifespan of Ames dwarfs even though their controls were quite long-lived. However, the issue in this paper might have been back-crossing into the black 6 strain, producing idiosyncratic problems.

Interestingly, methionine restriction fails to extend the lifespan of both dwarf and GHRKO mice. In my mind, methionine restriction is akin to protein restriction in general which is working through growth hormone (esp. in humans protein restriction reduces IGF-1 but not CR) and partly independent of CR.

C. Conclusion and outlook
If you ask me, there is little evidence that GHRKO mice have a more robust lifespan than pituitary dwarfs. The paper by Gesing et al. (2016) suggests there may be an additive benefit on health span when crossing Ames and GHRKO mice. Some other papers find minor differences between the models. Nevertheless, the data is more or less consistent with the idea that GHRKO and Ames dwarfism engage the same pathway (GH/IGF-1 signalling) and it's immaterial where in this linear pathway you interfere. Perhaps methionine restriction depends exclusively on the GH/IGF-1 axis. In contrast, CR may work through GH signalling and additional pathways or by further modulating some pathway downstream of GH/IGF-1.

In order to clarify the controversy, future studies could use an alternative approach, trying to normalize the reduced GH/IGF-1 levels in CR mice. If this completely eliminates the lifespan differential it would support the notion that CR requires the GH axis.

There is some evidence that low GH levels during the first 10 weeks of life regulate mouse lifespan (11), but this has never been reported with CR as far as I know. How is this consistent with the idea that CR operates exclusively on the GH axis? This is another odd difference between the two paradigms. Just as I was writing "future studies should test adult onset GHRKO", I decided to check the recent literature and stumbled on a recent publiction doing just that (12), but this article is getting too long and I don't want to go into detail.

Ceiling effect, diminishing returns, "aging in spare parts"

A hypothetical example. Even if the effects of two genetic interventions are additive and/or independent, they could be masked due to segmental differences in aging rate of tssues and organs.

D. Data

Bonkowski 2006, 30% CR (2) fails to extend GHRKO lifespan:
"CR did not affect the overall survival or average or median longevity of GHRKO mice and produced a smaller (7.8% in GHRKO mice vs. 17.2% in normal mice) although statistically significant increase in the maximal longevity of GHRKO animals.[mostly limited to females]"

"we suggest that the failure of CR to increase overall, median, or average longevity in GHRKO mice is related to its failure to improve insulin sensitivity in these mutant animals" (which they later refuted in [1])

"Absence of additional extension of life by CR in GHRKO animals in the present study could have been due to the particular regimen of CR chosen or to the “ceiling effect” (i.e., maximal life extension) in GHRKO animals that were fed AL....It is exceedingly unlikely that GHRKO mutation produces the maximal possible life extension and thus that CR in animals of this genotype causes near starvation rather than a beneficial reduction in nutrient intake. "
The ceiling effect has nothing to do with the starvation response to my knowledge. It is very possible that having one mutation simply decreases the power to find a difference, because the added benefit is small. The additional benefit could be small if CR and GHRKO have partly overlapping effects on an organ system or tissue that limits natural mouse lifespan. To give a somewhat hypothetical example, let's say that CR mainly protects from cancer by reducing GH levels, but GH signalling is already highly suppressed in GHRKO. Even if there are some other benefits of adding CR to GHRKO, most mice would still die at a comparable age from neoplasia, which is an important cause of death in mice.

average ~900d vs 1150 (CR) vs 1200 (GHRKO +/- CR)
Max LS (single outlier) 1200 vs 1400 (CR or GHRKO) vs 1500 (GHRKO + CR)

Bonkowski 2009 (3), 10-15% CR fails to extend GHRKO lifespan:
Median LS: 850 vs 1000d (CR) vs 1150 (GHRKO +/- CR)
"Taken together, these studies indicate that GH resistance ablates rather than modulates (or shifts) the benefits of CR on longevity in male GHRKO mice."

Bartke 2001 (4), 30% CR robustly extends Ames life span:
Average 750d vs 900 (CR) vs 1000 (dwarf) vs 1250 (dwarf + CR)
Max LS (single outlier): 900d (wt) vs 1200 (CR) vs 1300 (dwarf) vs >1400d (dwarf + CR)
It seems clear that wildtype (wt), CR and dwarf are significantly shorter-lived than in the other studies (2, 3)

Genders are combined in the final analysis, as is the case for most of the studies mentioned here.

Gesing 2016 (1), GHRKO + Ames dwarfism:
Adiponectin was increased and so was insulin sensitivity as determined by the glucose tolerance test. Leptin and body weight were further decreased. Controls weigh around 30g, dwarf and GHRKO around 15 and double mutant only 10g. All these changes would be expected to further increase life span or health span of double mutants, but they did not. Looking at the survival curve perhaps there is a trend for decreased mid life mortality in males, but this is purely speculative.

Max lifespans (single outlier guess) and mean life spans were:

df/KO             942.3 ± 31.8
GHRKO         975.1 ± 60.6
df/df               882.6 ± 36.5
N                     726.2 ± 36.8
MaxLS: 1050 vs 1250d for all genotypes

df/KO             1219.0 ± 108.8
GHRKO         1179.0 ± 107.3
df/df               1122.0 ± 113.5
N                     700.7 ± 40.94
MaxLS: 1050 vs 1300-1400d for all genotypes

Sun 2013 (8), CR extends the lifespan of growth hormone releasing hormone KO (GHRHKO) mice:
Average: 750 vs 900 (CR) vs 1000 (GHRH-KO) vs 1100 (GHRH-KO + CR)
MaxLS (single outlier): 1000 (control) vs 1250 vs 1500 (GHRH-KO + CR)
Sexes were approx. evenly split (39 females, 58 males).

Garcia 2008 (10)
"It is possible that the C57BL/6J background under CR conditions, together with the naturally rather frail dwarf mice lead to increased mortality during the entire life span."


Gesing, A., Wiesenborn, D., Do, A., Menon, V., Schneider, A., Victoria, B., ... & Masternak, M. M. (2016). A Long-lived Mouse Lacking Both Growth Hormone and Growth Hormone Receptor: A New Animal Model for Aging Studies. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, glw193.

Bonkowski, M. S., Rocha, J. S., Masternak, M. M., Al Regaiey, K. A., & Bartke, A. (2006). Targeted disruption of growth hormone receptor interferes with the beneficial actions of calorie restriction. Proceedings of the National Academy of Sciences, 103(20), 7901-7905.

Bonkowski, M. S., Dominici, F. P., Arum, O., Rocha, J. S., Al Regaiey, K. A., Westbrook, R., ... & Bartke, A. (2009). Disruption of growth hormone receptor prevents calorie restriction from improving insulin action and longevity. PloS one, 4(2), e4567.

Bartke, A., Wright, J. C., Mattison, J. A., Ingram, D. K., Miller, R. A., & Roth, G. S. (2001). Longevity: extending the lifespan of long-lived mice. Nature, 414(6862), 412-412.

5. Gent, J., Van Kerkhof, P., Roza, M., Bu, G., & Strous, G. J. (2002). Ligand-independent growth hormone receptor dimerization occurs in the endoplasmic reticulum and is required for ubiquitin system-dependent endocytosis. Proceedings of the National Academy of Sciences, 99(15), 9858-9863.

6. Growth Horm IGF Res. 2016 Jun;28:6-10. doi: 10.1016/j.ghir.2015.06.001. Epub 2015 Jun 7.
The growth hormone receptor.

6b. Brooks, A. J., & Waters, M. J. (2010). The growth hormone receptor: mechanism of activation and clinical implications. Nature Reviews Endocrinology, 6(9), 515-525.

7. van den Eijnden, M. J., & Strous, G. J. (2007). Autocrine growth hormone: effects on growth hormone receptor trafficking and signaling. Molecular endocrinology, 21(11), 2832-2846.

7b. Harvey, S. (2010). Extrapituitary growth hormone. Endocrine, 38(3), 335-359.

8. Sun, L. Y., Spong, A., Swindell, W. R., Fang, Y., Hill, C., Huber, J. A., ... & Bartke, A. (2013). Growth hormone-releasing hormone disruption extends lifespan and regulates response to caloric restriction in mice. Elife, 2, e01098.

9. Nature. 1996 Nov 7;384(6604):33.
Dwarf mice and the ageing process.
Brown-Borg HM, Borg KE, Meliska CJ, Bartke A.

10. Mech Ageing Dev. 2008 Sep;129(9):528-33. doi: 10.1016/j.mad.2008.04.013. Epub 2008 May 13.
Effect of Ames dwarfism and caloric restriction on spontaneous DNA mutation frequency in different mouse tissues.
Garcia AM, Busuttil RA, Calder RB, Dollé ME, Diaz V, McMahan CA, Bartke A, Nelson J, Reddick R, Vijg J.

11. Bartke, A. (2016). Healthspan and longevity can be extended by suppression of growth hormone signaling. Mammalian Genome, 1-11.

12. Junnila, R. K., Duran-Ortiz, S., Suer, O., Sustarsic, E. G., Berryman, D. E., List, E. O., & Kopchick, J. J. (2016). Disruption of the growth hormone receptor gene in adult mice increases maximal lifespan in females. Endocrinology, en-2016.

13. Miquet, J. G., Muñoz, M. C., Giani, J. F., González, L., Dominici, F. P., Bartke, A., ... & Sotelo, A. I. (2010). Ames dwarf (Prop1df/Prop1df) mice display increased sensitivity of the major GH-signaling pathways in liver and skeletal muscle. Growth Hormone & IGF Research, 20(2), 118-126.

Reduced growth hormone signaling and methionine restriction: interventions that improve metabolic health and extend life span.Brown-Borg HM.Ann N Y Acad Sci. 2016 Jan;1363:40-9. doi: 10.1111/nyas.12971. Review.

Kopchick, J. J. (2016). Lessons learned from studies with the growth hormone receptor. Growth Hormone & IGF Research, 28, 21-25.