|Home | About | Journals | Submit | Contact Us | Français|
Growth hormone receptor knockout (GHRKO) mice live about 40%–55% longer than their normal (N) littermates. Previous studies of 21-month-old GHRKO and N mice showed major alterations of the hepatic expression of genes involved in insulin signaling. Differences detected at this age may have been caused by the knockout of the growth hormone receptor (GHR) or by differences in biological age between GHRKO and N mice. To address this question, we compared GHRKO and N mice at ages corresponding to the same percentage of median life span to see if the differences of gene expression persisted. Comparison of GHRKO and N mice at ~50% of biological life span showed significant differences in hepatic expression of all 14 analyzed genes. We conclude that these changes are due to disruption of GHR gene and the consequent suppression of growth hormone signaling rather than to differences in “biological age” between mutant and normal animals sampled at the same chronological age.
GENETIC and environmental control of aging and longevity are of great interest in the context of demographic projections for 21st century. Calorie restriction can effectively extend life span (1). However, it alters numerous biological factors, which complicate identification of mechanism responsible for extended longevity. Genetic manipulations provide opportunity to identify mutations of individual genes that can cause important changes in life span. One such manipulation is the knockout of the growth hormone receptor/binding protein (GHR/BP−/−, GHRKO). These mutated animals are resistant to growth hormone (GH) and have elevated GH levels in circulation. These “knockouts” also have severely reduced plasma levels of insulin-like growth factor 1 (IGF-1) and decreased levels of insulin and glucose, implying increased insulin sensitivity (2–4). GHRKO mice indeed exhibit greater responses to injected insulin and live ~40%–55% longer than normal (N, GHR+/−) mice (4–6).
The metabolic action of insulin is regulated by a cascade of molecular events initiated by insulin binding to its receptor (IR) on the cell surface, causing its autophosphorylation. This leads to phosphorylation of the insulin receptor substrates (IRS). Phosphorylated IRS1 and IRS2 bind to the p85 regulatory subunit of PI 3-kinase and cause the activation of this enzyme (7–9). We have previously analyzed insulin signaling pathways in different organs in these long-lived mice (10,11) and demonstrated major alterations in the expression of genes related to insulin signaling in the liver of the GHRKO mice (10,11). However, insulin action is known to decline with age as insulin resistance gradually develops (12,13). Most of our previous data were obtained from GHRKO and N mice at ~21 months of age. At that age, N mice have already reached ~70% of their median and 54% of their maximal life span, whereas GHRKO are at ~53% of median and ~46% of maximal life span (12–14). This raises the possibility that differences in the expression of the genes in the insulin signaling pathway between GHRKO and N mice may have been due to differences in the biological age of these animals at the time of sampling. It was therefore important to determine whether the observed differences in messenger RNA (mRNA) expression levels of the examined genes between GHRKO and N mice are present at different stages of their life history and at equivalent biological ages. To answer this question, we analyzed expression of genes related to the insulin signaling pathway in the liver from GHRKO and N mice at three different ages.
Another objective of the present study was to characterize the genetic background of the GHRKO animals. Our colony of GHRKO mice was developed by crossing founder animals with mice derived from different strains, and there is considerable evidence that the phenotypic consequences of deletion or mutation of various genes can differ depending on the genetic background. Therefore, we felt that it was of interest to characterize the genetic background of our colony of GHRKO mice.
Normal and GHRKO mice used in this study were produced in our breeding colony developed by crossing 129Ola and BALB/c N (GHR+/−) animals, generously provided by Dr. J. J. Kopchick with mice derived from crosses of C57BL/6J and C3H/J strains and maintained as a closed colony with inbreeding minimized by avoiding Brother × Sister mating. All animal protocols for this study were approved by the Southern Illinois University Laboratory Animal Care and Use Committee. The animals were housed under temperature- and light-controlled conditions (20–23°C, 12-h light–12-h dark cycle). To produce GHRKO (−/−) mice, knockouts (−/−) males were mated with heterozygous (+/−) females.
GHRKO and N male mice at the ages of 9.5 (young), 15 (adult), and 21 (old) months were selected for this study (n = 9–10 in each group, each sample represents a single animal). Within these groups, 21-month-old GHRKO and 15-month-old normal mice had a similar “biological age” corresponding to ~50% of their median life span (Table 1). All animals were fasted overnight, then anesthetized using isoflurane, bled, and killed by cervical dislocation. Blood plasma and liver tissues were collected, immediately frozen on dry ice, and stored at −80°C until analysis.
Fasting glucose levels were measured in blood collected from the tail vein using OneTouch Ultra glucose meter (Life Scan, Milpitas, CA). Plasma was obtained from blood collected by cardiac puncture and was used for assessment of insulin levels via an enzyme-linked immunosorbent assay according to manufacturer’s protocol (Linco Research Inc., St. Charles, MO and IDS, Inc., Fountain Hills, AZ). Insulin resistance (or sensitivity) was estimated by homeostasis model analysis (HOMA) according to the method described by Matthews and colleagues (15).
DNA was extracted from tail tips using phenol–chloroform standard procedures. A panel of 27 mouse single-nucleotide polymorphisms (SNPs) informative across standard laboratory strains (16) were genotyped by sequencing. We selected 44 mice for genotyping, distributed across the breeding colony (from 11 different breeder pairs). Equimolar pools of DNA from four GHR+/− mice in each pool were sequenced to estimate allele frequency in the GHRKO population at each SNP. Primers were designed using the program Primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) and synthesized by Integrated DNA Technologies (Coralville, IA). Oligonucleotide polymerase chain reaction (PCR) primer sequences are available upon request. Each PCR included 4 ng of genomic DNA and 1 μM each forward and reverse primer. The PCR cycling conditions were as follows: 2 minutes at 95°C, followed by 35 cycles of amplification, 10 seconds at 92°C, 20 seconds at 58°C, and 1 minute at 72°C and a final extension 10 minutes at 72°C. PCR products were purified (ExoSAP-IT reagent; USB Corp., Cleveland, OH) to remove unincorporated dNTPs and PCR primers and sequenced using the BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems, Foster City, CA). Sequencing products were filter purified using genCLEAN dye terminator cleanup plates (Genetix USA, Boston, MA) and read in the ABI 3730xl DNA analyzer (Applied Biosystems). Sequence traces were aligned in Sequencher v4.7 (Gene Codes Corp., Ann Arbor, MI) for genotype scoring.
Total hepatic RNA was extracted using the phenol–chloroform procedure of Chomczynski and Sacchi (17). Total RNA quantity and quality were analyzed on agarose gel using electrophoresis. The synthesis of complementary DNA (cDNA) was performed from 1 μg of total RNA using iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA) according to the manufacturer’s protocol.
The GHRKO mice used in the present study were originally generated from 129Ola embryonic stem cells and include genetic contribution from the BALB/c, C57BL/6, and C3H inbred strains. To assess the amount of background genetic variation segregating in the GHRKO mice, we genotyped 27 highly polymorphic SNP markers from a panel developed to monitor strain identity (16) in 44 GHRKO mice (Table 2). Twenty of 27 markers on 17 different chromosomes were polymorphic in GHRKO mice, with minor allele frequencies between 0.068 and 0.49, indicating a substantial amount of segregating genetic variation. For comparison, 48 mice from an unrelated Ames dwarf experimental strain typed with the same panel were only polymorphic at 7 of 27 of these loci, and 16 Idaho mice derived from a limited pool of wild-caught founders were only polymorphic at 11 of 27 of these loci.
We compared alleles of the 27 markers with those of the parental strains of GHRKO mice (Table 2). Of the 27 markers, 4 loci had been invariant among the parental strains, meaning that 3 of the 23 initially polymorphic loci have become fixed during GHRKO breeding. Because the genotyping panel is not optimized for informativeness for the parental strains, many of the parental strains carry the same alleles and relative strain contribution of the parental strains to the current generation of GHRKO mice cannot be estimated. However, the amount of allele sharing between GHRKO and the parental strains is similar and ranges from 45% for C57BL/6 (B6) to 59% for BALBc/J. The likely slightly lower contribution from B6 is due to three loci where the GHRKO population has lost the B6 allele and another three loci where the frequency of the B6 allele is low. There are also two loci for which the 129Ola allele has been lost. A marker on chromosome 18 is the only locus where a single parental strain, BALB/c, has become fixed.
These results suggest that the GHRKO mice show a significant degree of background strain polymorphism, continue to independently segregate contributions from all four parental strains and not much genetic drift or fixation of specific alleles has occurred.
Analysis of plasma insulin revealed the expected significant reduction of insulin levels in GHRKO mice when compared with their normal siblings at all ages (p < .0001; Figure 1A). Two-way analysis of variance (ANOVA) indicated no genotype or age effects on glucose levels, but there was a significant age–genotype interaction (p < .0114; Figure 1B). The calculation of HOMA score indicated significant decrease (increased insulin sensitivity) in GHRKO mice in comparison to their normal siblings (p < .0001; Figure 1C).
Analysis of the hepatic expression of 14 genes related to insulin/IGF-1 signaling and glucose regulation (Table 3) revealed differential effects of genotype and age. Expression of some of the examined genes (here after referred to as group 1) was affected primarily by genotype, whereas expression of the remaining genes (here after referred to as group 2) was subject to interactive influences of genotype and age (Figure. 2).
Consistent with GH resistance, expression of IGF-1 mRNA was severely reduced in GHRKO compared with N mice (p < .0001; Figure 2A). In contrast, expression of PPAR,, G6P, and IRS2 mRNA was significantly increased in GHRKO mice in comparison to their N counterparts (p < .0001, p < .0001, and p < .0149, respectively; Figure 2B–D). FOXO1 gene expression was higher in GHRKO (p < .0001) but also showed a significant age effect on the expression (p < .0252) with no genotype–age interaction detected (Figure 2E).
The expression of PGC1a mRNA was increased in the liver of GHRKO mice when compared with N mice (p < .0001; Figure 2F). Two-way ANOVA did not reveal age effects on the expression of PGC1α, but within N mice, there was a significant decrease in PGC1α expression with age (p < .0048 and p < .0362 for young vs adult and young vs old mice, respectively, t test). Similarly, the expression of PPARα was increased in GHRKO animals in comparison to N mice (p < .0001; Figure 2G), with no age effect (two-way ANOVA), and a significant decline in old N when compared with young N mice (p < .0018, t test).
The expression of IR, IRS1, and SOD2 was altered by genotype (p < .0001, p < .0001 and p < .0001, respectively) and age (p < .0001, p < .0017, and p < .0105, respectively; Figure 3A–C). Moreover, significant genotype–age interactions were detected (p < .0001, p < .0088, and p < .0005, respectively).
Similarly, the expression of PEPCK, GLUT2, SIRT1, and AKT2 mRNA was significantly affected by genotype (p < .0001, p < .0001, p < .004, and p < .0001, respectively) and age (p < .0.149, p < .0113, p < .004, and p < .0005, respectively; Figure 3D–G) with significant genotype–age interactions (p < .0003, p < .0242, p < .05, and p < .0001, respectively). Expression of PEPCK, GLUT2, SIRT1, and AKT2 did not differ between young GHRKO and young N mice.
Comparisons of 15-month-old N and 21-month-old GHRKO mice, that is, mice of roughly equivalent biological age (~50% of median life span), indicated that the expression of all examined genes was significantly increased in the liver of GHRKO compared with N mice (Figures 2 and and33).
The key novel finding in the present study is that the effects of genetic GH resistance in GHRKO mice on the expression of genes related to IGF-1 and insulin signaling are evident in comparison of animals of comparable biological age as well in animals of the same chronological age. The interactive effects of gene and age on the expression of the examined genes will be discussed in the second part of this section.
Selection of optimal genetic background for mutant animals used in research and, specifically, in experimental gerontology is a complex issue. Inbred strains offer advantages of well-defined characteristics, a wealth of information from previous studies, and a virtual (although not complete) absence of genetic variation. Heterogeneous genetic background and random breeding are generally associated with better fertility and viability and offer an advantage of using an experimental system that more closely resembles natural animal and human populations. However, genetic background can have major impact on the phenotypic consequences of a mutation under study, and therefore, characterizing the genetic background of our animals was of considerable interest.
Given the detected degree of genetic polymorphism, there could be significant phenotypic variability in these mice (eg, 129 vs B6 strains have numerous phenotypic differences). There may be strain-dependent modifier effects segregating in this population that may be amenable to mapping by ancestry, but this would require a number of additional markers specifically chosen for informativeness between the four parental strains of GHRKO mice. Because we did not observe unusual variability in the phenotypes we examined for this report, we did not further investigate genetic differences between individual mice from this population. However, adjustment for individual ancestry (eg, using individual ancestry estimates calculated with the program STRUCTURE) could be incorporated into multivariable models for association of each trait with GHRKO genotype to account for these differences.
Disruption of GHR- and GH-binding protein in long-lived GHRKO mice causes GH resistance, which leads to significant body mass reduction. These long-lived animals are extremely sensitive to injected insulin (4). Previously published studies indicated significant alterations of the insulin signaling pathway in the liver (10,11). Because previous studies utilized GHRKO and N animals of the same chronological and therefore presumably different biological age (10,11), we were interested in determining whether the alterations of the insulin signaling pathway observed in these studies may have been due to differences in biological age between N and GHRKO mice. The design of the present study allowed comparisons of the insulin signaling pathway in these genotypes at the same biological or the same chronological age. In our study, the comparable biological ages were 15 months for normal and 21 months for GHRKO mice (see Table 1).
In comparisons based on either biological or chronological age, GHRKO mice have a lower level of plasma insulin, and additionally, a lower HOMA score indicating that these long-lived knockouts are more insulin sensitive than N controls throughout their adult life. Analysis of the expression of genes involved in the insulin signaling pathway in the liver of GHRKO and N mice indicated significant alterations of the expression levels of these genes in GHRKO mice compared with normal mice of the same calendar age (Figures 2 and and3).3). These findings closely correspond to and confirm results obtained previously from different cohort of mice from the same strain (10,11). More importantly, the comparison between 15-month-old N and 21-month-old GHRKO mice indicated that these genes are altered in the same direction when animals of a similar biological age are compared. This implies that the differences observed between the GHRKO and the N mice of the same chronological age were not due to more advanced biological age in N animals.
As was mentioned in the Results section, statistical analysis revealed the existence of two major groups of genes. Group 1 included IGF-1, PPAR,, G6P, IRS2, FOXO1, PGC1α, and PPARα, the genes with expression affected mainly by genotype. In other words, the expression of these genes appears to be altered in GHRKO mice throughout their entire adult life. Group 2 included IR, IRS1, SOD2, PEPCK, GLUT2, SIRT1, and AKT2 genes that showed significant genotype–age interaction. This was unexpected, and further t test analysis indicated no differences in the expression of PEPCK, GLUT2, SIRT1, and AKT2 when comparing young GHRKO versus young N mice (Figure 3). However, as the mice reached the age of 15 months, the expression of these genes increased in GHRKO mice and remained upregulated in 21-month-old knockouts. Other genes from this group, IR, IRS1, and SOD2, had increased expression in young GHRKO mice, but the effect of genotype was attenuated in either the adult or old groups. This could perhaps represent an important mechanism allowing GHRKO mice to maintain efficient insulin signaling longer during their extended life span. It is known that the insulin signaling pathway is most sensitive in healthy young organisms. GHRKO mice are more insulin sensitive throughout their entire life span. The present analysis of gene expression implies that when GHRKO mice are young and healthy, the expression of some of the genes in the insulin signaling pathway is not altered, that is, the level of the expression of these genes is similar to values measured in N mice. However, as aging progresses, an unknown mechanism enhances expression of these genes in GHRKO mice resulting in the maintenance of healthy whole-body insulin sensitivity longer than it is maintained in N animals. This could represent one of the mechanisms underlying the extended longevity of GHRKO mice. Analysis of survival plots indicates that the rate of aging (as estimated from doubling time of mortality rate) is reduced in GHRKO versus N mice (23).
In summary, analysis of gene expression in GHRKO and N mice at different ages indicated that the previously reported differences in gene expression in the liver of these animals were not due to comparing animals at different stages of their biological aging. These differences between long-lived GHRKO and N mice were also present when animals of equivalent biological age were being compared. Strong interaction between the effects of genotype and age on the expression of 7 of the 14 analyzed genes suggests existence of mechanisms that allow GHRKO mice to regulate insulin/IGF-1 and glucose-regulating signaling pathways in ways that contribute to maintenance of extended health span and life span.
National Institute on Aging (AG 19899 and U19 AG023122), Ellison Medical Foundation, Southern Illinois University Geriatrics Medicine and Research Initiative, Central Research Committee of Southern Illinois University School of Medicine, and Glenn Foundation for Medical Research.
We would like to thank Steve Sandstrom for helping with the editing of the manuscript.