Benefits of Living Without Growth Hormone

When Roz Anderson told me about the opportunity to contribute to the Fellows’ Forum in this journal (thank you, Roz!!!), I realized that it is now exactly 25 years since Holly Brown-Borg reported that Ames dwarf mice live much longer than their normal siblings. This, of course, is “old news” now and probably ancient history to many younger readers, but for me it was an exciting beginning of work in a new field, what I sometimes call “my second career.” What I decided to attempt in this brief article is to look at this finding and at the work that followed from the perspective of the intervening years. I will summarize current information about the mechanisms that appear to be involved and the relevance of the results to human aging. As a zoologist, I am also interested in relating the findings in laboratory animals to life history traits and survival of wild animals in their natural environment. I am reminded of the words of Rich Miller, who emphasized the potential relevance of the long history of domestication of laboratory animals, and sometimes referred to mice from contemporary laboratory stocks as “mouse-like objects.”

A Bit of History…

I started working with dwarf mice in the early ’60s as a graduate student at the University of Kansas. My dissertation project was to compare the characteristics of animals homozygous for the then newly discovered Ames dwarf mutation to the phenotype of dwarf mice described by George Snell in 1929 and fairly extensively studied in the United States and in Europe. This led to the demonstration that both Snell and Ames dwarf mice, in addition to being growth hormone (GH) and thyroid-stimulating hormone deficient, also lacked pituitary prolactin (PRL) (1). I have used these animals to identify the role of PRL in the control of reproduction in females and males, which became the focus of my work for many years (2).

Some 20 years later, I started collaborating with Dr. Thomas Wagner, pioneer of transgenic (TG) technology, to characterize reproductive function of TG mice expressing high levels of human (h) or bovine (b) GH. In mice, hGH binds to both GH and PRL receptors, while bGH is GH receptor-specific (3). In the course of this work, we noticed that many characteristics of GH-TG mice resembled normal mice which were much older, and we proposed that these animals experience early and/or accelerated aging (4, 5). We wanted to explore this hypothesis and the mechanisms that we thought were involved, but our applications for support of such studies were receiving no or dismal scores.

During this time, one of my coffee break conversations with Holly Brown-Borg and Kurt Borg (who at the time were postdoctoral fellows in our lab) turned to our lack of success with these grant applications and I was asked if dwarf mice, being GH deficient, have any interesting aging-related characteristics that could be mentioned in grant proposals. I admitted that I had no idea since we always used these mice for various experiments before they had a chance to get old. We decided that we should first find out how long they live, and a few weeks later we set aside a group of Ames dwarf mice and their normal siblings for a longevity study. Two years later, we had our answer: most of the dwarf mice were still alive while most of the normal control mice had already died. Considering the short life span of GH-TG mice, we interpreted extended longevity of Ames dwarf mice as an effect of GH deficiency. It took some time for the rest of the animals in this study to die and for the paper to be written, submitted, revised etc., and in 1996, Holly Brown-Borg reported these findings in a brief scientific correspondence article in Nature (6).

Is It Justified to Ascribe the Remarkable Longevity of Ames Dwarf Mice to the Absence of GH Signals?

We suggested that GH deficiency is the key reason why Ames dwarf mice live much longer than their normal (wild-type) siblings (6), fully realizing that such a suggestion appears counterintuitive. After all, how can deletion of a major component of normal endocrine integration be beneficial for something as fundamental as staying alive? Therefore, it is important to ask if any evidence supports this conclusion. We believe that the following findings provide affirmative answer to this question:

1. Snell dwarf mice with endocrine defects and dwarfism due to a mutation of a different gene were shown by Kevin Flurkey and colleagues to be long-lived (7);

2. Animals with isolated, specific suppression of GH signaling by deletion of GH receptors, hypothalamic GH-releasing hormone (GHRH), or GHRH receptors have extended longevity and share many age-related characteristics with hypopituitary Ames and Snell dwarf mice (7–9);

3. Replacement therapy with GH alone for a brief period during early life reduces longevity of Ames dwarf mice and normalizes multiple characteristics related to mechanisms of aging (10, 11); and

4. TG mice with abnormally elevated bovine, ovine, human, or, importantly, also endogenous mouse GH have reduced longevity and symptoms of accelerated aging (5, 12).

Moreover, research in mice and other organisms provided evidence that many (likely, most) life-extending mutations, aka “longevity genes” induce loss or profound suppression of hormonal signaling or another physiological function such as mechanistic target of rapamycin or homologous signaling.

Benefits of GH Deficiency or Resistance Are Not Limited to Extended Longevity

The remarkable extension of mouse longevity by hereditary hypopituitarism, isolated GH deficiency (IGHD), or GH resistance is associated with many features of extended healthspan and retention of youthful characteristics into advanced age. Notably, this includes various measures of insulin sensitivity (13, 14), inflammation (15, 16), and cognitive function (17, 18).

Fecundity of these mutants is reduced and sexual maturation is delayed, with the age of puberty and adult fertility status depending strongly on the genetic background (19–21). Females homozygous for the Pit1^dw or Prop1^df mutations (Snell and Ames dwarf mice) can produce fertilizable eggs, but require PRL replacement therapy to become pregnant and raise their litters (1, 22). Males with the various GH-related mutations are typically fertile and can be used routinely as breeders. Interestingly, ovarian aging in these mutant mice is delayed and reproductive period appears to be extended (23).

Ames dwarf mice have increased resistance to toxins inducing lethal oxidative stress (24). This is likely related to improved antioxidant defenses (25), and alterations in the expression of stress-responsive genes (26).

Multiple Mechanisms Link Suppression of GH Signaling With Longevity

Much of our work during the last 25 years centered on the search for mechanisms responsible for extension of healthspan and life span in GH-deficient and GH-resistant mutants. We hoped, perhaps naively, to identify a specific phenotypic characteristic, a signaling pathway, or a chain of causally linked functional alterations that would explain the healthy aging and the longevity advantage of these animals. So far, a mechanism of this type has not been identified and we now strongly suspect that a single mechanism explaining the available data does not exist. Instead, there appears to be a rather long list of interconnected and interacting mechanisms (21). In other words, genetic GH deficiency or resistance leads to development of a novel, distinct, and complex phenotype with delayed and/or slower aging and extended longevity representing some of its characteristics. This is not as conceptually satisfying or as “tidy” as one would perhaps wish, but it corresponds very closely to what we have learned about the effects and the mode of action of calorie restriction (CR), another intervention with a major and consistent positive impact on longevity. Moreover, the specific features of the phenotype of long-lived mice with suppressed GH signaling overlap with the list of “Hallmarks of Aging” developed by López-Otín and colleagues (27) (Figure 1), and also with many of the features of health proposed recently by the same investigator (28). As interesting and informative as similarities of these models might be, it should be emphasized that they do not represent identity. For example, we have shown that phenotypes of hypopituitary or GH-resistant mice and wild-type mice subjected to CR, as well as alterations in the expression of various aging-related genes in these animals, are similar but not identical (29), and that CR interacts in different ways with different GH-related mutations (9, 30, 31).

Figure 1.

Characteristics of long-lived mice with growth hormone-related mutations are favorable in terms of each of the proposed “Hallmarks of Aging.” Abbreviations: BAT = Brown Adipose Tissue; Ucp-1 = Uncoupling protein 1; VSELs = Very Small Embryonic-Like stem cells; IL-1β = Interleukin 1 beta; IL-6 = Interleukin 6; tnf-α = tumor necrosis factor alpha.

In trying to interpret data related to the mechanisms of aging, it is important to emphasize that most if not all the characteristics identified as candidate mechanisms may also (or instead) represent markers of a younger biological age of the long-lived mutants being compared to wild-type controls of the same chronological age. We have attempted to address this question experimentally (32), but complete separation of the mechanisms from the consequences of slower aging will continue to be challenging. Much additional work involving genetic, molecular, and pharmacological interventions will be needed to tease out and clarify the underpinning cause–effect relationships.

Does GH Influence Human Aging?

Most of the GH mutations which were shown to extend longevity in laboratory mice correspond to rare but well-characterized genetic syndromes in humans. This includes IGHD due to mutations affecting GHRH signaling, GH resistance (Laron syndrome) due to mutations in the GH receptor gene and hypopituitarism caused by Prop1, or Pit1 mutations, and thus corresponding to Ames and Snell dwarfism in mice (21). The consequences of deletion or severe suppression of GH signaling in humans and mice are very similar and generally include slower postnatal growth, reduced adult body size with proportional dwarfism, delayed puberty, reduced fecundity, improved glucose homeostasis, and increased adiposity, paradoxically associated with reduced pro-inflammatory markers and increased adiponectin (21).

Similarly to the GH-related mouse mutants, IGHD patients from the Itabaianinha cohort in Brazil maintain a number of youthful characteristics into late life and have been described as “aging well.” Compared to their unaffected relatives and unrelated subjects from the same population, they exhibit improvements in muscle function, reduced fatigue, enhanced insulin sensitivity, and maintenance of memory into advanced age. They develop wrinkling of the skin, but essentially no graying of the hair. Moreover, they appear to be protected from cancer (except skin cancers) and have no changes in the risk of atherosclerosis in spite of increase of truncal adiposity and increased serum cholesterol (21, 33, 34). Subjects with Laron dwarfism are remarkably protected from cancer (35), and those from a large and etiologically uniform cohort in Ecuador are also fully protected from diabetes (36). Recent studies in the latter population provided evidence for reduced structural and functional symptoms of brain aging (37). Although both GH-resistant and IGHD patients have some sensory and vocal deficits as well as trends for increased risk of sleep disturbances, dizziness, and cerebrovascular disease, the overall impact of these syndromes on healthspan, quality of life, and functionality of older adults has been characterized as positive rather than negative (21, 33, 34, 38).

Predisposition to healthy (“successful”) aging and indications of extended healthspan in the Itabaianinha IGHD cohort and in the Ecuadorian Laron syndrome cohort was not associated with any obvious or consistent changes in the average life span (21, 36). The Itabaianinha cohort included one nonagenarian and one centenarian, but this appeared to be counterbalanced by a suggestive increase in early mortality. This contrasted with a report of reduced life expectancy of IGHD subjects in Switzerland (39), but resembled the findings in Prop1 mutants from the Krk island in Croatia, “the little people of Krk,” some of whom reached very advanced age (40). It should be emphasized that the numbers of individuals affected by these hereditary syndromes is very small, and therefore modest extension (or reduction) of longevity in the examined cohorts would have been difficult or impossible to detect.

Importantly, there is increasing amount of evidence that fluctuations of GH signaling within the normal range have a role in human aging and longevity. Life expectancy has been related to the amount of GH secreted during a 24-hour period and to the precision of the control of pulsatile release of this hormone in a study comparing individuals predisposed to familial longevity to their spouses or partners (41). Moreover, human longevity is negatively related to key GH-dependent traits: insulin-like growth factor 1 signaling (42–44) and height (45, 46). However, it should also be mentioned that the topic of GH and human aging is contentious and there is a controversial but persistent interest in using GH as an antiaging agent (47). Chronic elevation of circulating GH levels into the pathological range shortens healthspan and life span in patients with gigantism or acromegaly (21, 47) as it does in GH-TG mice (5, 12).

Correspondence of most but not all findings in humans to the results obtained in mice is not unexpected considering the impact of public health and medical interventions on human longevity, taxonomic (evolutionary) distance between rodents and Hominin primates, and major differences in life history traits. We speculate that interventions which slow down the rate of aging are less likely to extend longevity in a species that is exceptionally long-lived. Numerous studies dating back to the ’50s related aging and longevity to reproductive strategies and pace-of-life (48, 49). Species or populations with a fast pace-of-life have high reproductive rates, fast development, and short life span, while those with a slow pace-of-life have opposite characteristics (50). Clearly, mice and humans are near the opposite ends of the range of these life history traits. It is interesting that, compared to their normal siblings, the long-lived mice with GH-related mutations have key features of a slower pace-of-life: slower postnatal growth, delayed puberty, and reduced fecundity. Moreover, sexual maturation of dwarf mice can be accelerated (51, 52) and longevity shortened (10, 11) by GH replacement therapy. It should not be surprising that slowing down the pace-of-life by disruption of the GH signaling has more impact on longevity of animals with a fast pace-of-life (mice) than in those in which pace-of-life is already slow (humans).

The relevance of the pace-of-life-related characteristics to differences in longevity is strongly supported by studies of domestic dogs. Among dog breeds, there are huge differences in adult body size and smaller size is associated with shorter period of growth, fewer pups per litter, and longer life (53). The complex relationship of aging to mitochondrial function, energy metabolism, and metabolic rate is outside the scope of this brief commentary. However, it is intriguing that increased metabolic rate per unit of body mass is associated with extended longevity in both small dog breeds and mice with GH-related mutations (53, 54).

Would Any Characteristics of Long-Lived GH-Related Mutants Be Advantageous to Wild Mice Living in Their Natural Environment?

The remarkable extension of longevity by gene mutations or deletion detected in laboratory populations of worms, insects, and mice raised questions as to likely effects of these mutations in the natural setting. For mutant mice, we are aware of only one attempt to answer this question experimentally (55). The results suggested reduced survival of the mutants in a semi-natural setting that included a cold winter, but the results are somewhat difficult to interpret or apply to other mutants since the longevity advantage of the employed animals (56) proved difficult to reproduce (57).

We believe it is interesting to ask whether natural variation in characteristics that are associated with (and likely responsible for) extended longevity of GH-related mutants could offer any advantages “in the wild.” It seems likely that some characteristics of these animals would be disadvantageous by reducing fitness in the evolutionary sense of this term, that is the chances of producing offspring and of offspring survival and reproduction. These characteristics include slower growth, smaller adult body size, delayed puberty, and reduced fecundity. Thus, males with these characteristics would be at a disadvantage in establishing and defending their territories and gaining access to females, while females would produce fewer offspring and also likely less milk. However, these characteristics of reduced GH levels or actions are also associated with resistance to stress, improved ability to repair DNA damage, protection from cancer, extended period of ovarian function, slower and/or delayed aging, and predisposition to extended longevity (21, 23–25, 58, 59). We speculate that these characteristics could be important for the survival of the population by ensuring the presence of some individuals that are more likely to survive periods of unfavorable conditions and remain healthy and reproductively competent when the conditions improve. This could apply to winter survival in temperate and circumpolar climate (Figure 2), and also perhaps to survival of the dry seasons in the tropical/equatorial regions. In this scenario, small body size could be advantageous. In Djungarian hamsters (Phodopus sungorus), small nonhibernating mammals living in the harsh climate of central Asia, substantial reduction of body weight occurs naturally in the fall. This weight loss is believed to facilitate winter survival by reducing energy demands (60).

Figure 2.

Individuals with a slower pace-of-life may have enhanced evolutionary fitness in their natural environment. In this example, smaller mice mature later and have fewer pups but are more likely to survive winter and reproduce in the second year of life.

The evolutionary mechanisms underpinning possible persistence or elimination of growth-related mutations and the broader (and widely debated) issue of individual versus group selection will not be discussed here. However, there is evidence that genetic diversity and the corresponding range of phenotypes favor survival of populations, including threatened species and translocated animals (61, 62).

Looking Back…

Looking back at the last 25 years of work in our lab, I am still surprised how hard it was to enter a new field, establish external funding, and build a group. Special thanks to Richard Sprott, Hubert Warner, and David Finkelstein at National Institute on Aging, who helped me survive the difficult period of transition. However, there were many rewards. I found biology of aging to be a very exciting field with lots of new data, unexpected findings, heated debates, and rapid progress. It was great to meet new people and develop new collaborations and friendships. This included a thrill of exchanging information with Prof. Manuel Aguiar-Oliveira from Brazil, meeting his GH-deficient patients, and discovering their similarities to our long-living mice.

With support of our institution and colleagues interested in Alzheimer’s disease, we were able to start a series of International Symposia on Neurobiology and Neuroendocrinology of Aging which are held every other year since 1992 in Bregenz, Austria. Rich Falvo and I were “in charge” until 2008, and it is wonderful to see this series of meetings continue and prosper under the leadership of Holly Brown-Borg.

Looking Forward…

As the field of aging biology moves toward identifying and testing interventions that might slow human aging, I think that “the key” will be to find a way of altering multiple physiological characteristics to shift the balance between growth and anabolic processes and processes related to defense, repair, and resilience. Phenotypes of animals that lack GH signals, have been subjected to chronic CR, or have otherwise acquired features of slower pace-of-life, provide some indications of what this new balance should be.

As to my own plans or hopes, I am encouraged by breakfast conversations with George Martin during the last “real” meeting of AGE in Austin, Texas. We talked about getting National Institutes of Health’s funding after reaching “a certain age.” I do not think I have a realistic chance to match George’s record, but I hope we can still get some new information and maybe an answer or two to the questions that I feel are worth asking.

Andrzej Bartke received his PhD in Zoology (Genetics) from the University of Kansas and served as a Fellow, Staff Scientist, and then Senior Scientist at the Worcester Foundation for Experimental Biology in Shrewsbury, MA. In 1978, he joined the faculty at the University of Texas Health Science Center at San Antonio as an Associate Professor and continued his research studying male reproductive physiology. In 1984, he became Chair and Professor of Physiology at Southern Illinois University in Carbondale. His early contributions focused on the role of prolactin in male and female reproduction. From there, he published extensively on multiple aspects of testosterone, seasonality, and male reproduction. He also made several key contributions to the field of cannabinoids and reproductive processes in both males and females. The role of growth hormone (GH) deficiency and GH overexpression has been a running theme in Dr. Bartke’s research since 1964. In the field of mammalian aging, Dr. Bartke’s laboratory was the first to show that GH and insulin-like growth factor 1 (IGF-1) are major factors in the regulation of life span in the mammal. His laboratory showed that GH deficiency delays aging, increases lifetime insulin sensitivity, significantly reduces cancer incidence, and increases overall healthspan. These discoveries and the multitude of publications stemming from this lab clearly demonstrate that the GH/IGF/insulin signaling pathways are the primary regulators of life span in rodents. These breakthroughs are firmly supported in other species indicative of the evolutionary importance of this endocrine mechanism. Dr. Bartke served as President of the American Aging Association and President of the American Society of Andrology and served as both the Director and President of the Society for the Study of Reproduction. In addition, he has received the Robert W. Kleemeier Award from the Gerontological Society of America, the Denham Harman Research Award from the American Aging Association, the Irving Wright Award of Distinction in Aging Research from the American Federation for Aging Research, and Fred Conrad Koch Lifetime Achievement Award from the Endocrine Society.

Funding

Financial support for our studies was provided by the National Institute on Aging, Ellison Medical Foundation, Glenn Foundation, American Diabetes Association, William E. McElroy Charitable Foundation, Memorial Medical Center Foundation, and the SIU School of Medicine Geriatrics Initiative.

Conflict of Interest

None declared.