The antagonistic pleiotropic theory of aging posits that genes can have opposite effects on biological fitness at different ages such that their effects are beneficial early in life but are detrimental to the organism later in life (Williams, 1957). This theory is relevant to the apparent paradox of the insulin-like growth factor (IGF) system. IGFs are essential for normal fetal development and are important stimulators of cell proliferation and survival (D’Ercole, 1996). However, IGFs are also associated with cellular and chronological aging, and with the increased incidence of vascular disease and cancer that occurs with age (D’Ercole, 1996; Bayes-Genis et al., 2000; Pollak et al., 2004; Katic & Kahn, 2005). Thus, it has been shown that a reduction in IGF-I signaling achieved by mutations in receptors or intracellular receptor substrates is associated with an increase in lifespan and delayed onset of age-related disorders in diverse species (Kenyon, 2001; Tatar et al., 2003; Rincon et al., 2004; Katic & Kahn, 2005). However, a specific connection between IGF and longevity in mammals has been difficult to demonstrate. The well-characterized Ames dwarf, Snell dwarf, lit/lit and growth hormone (GH) receptor knock-out mice have primary GH-deficiency or a GH signaling defect that has a secondary effect to decrease circulating levels of IGF-I (Brown-Borg et al., 1996; Coschigano et al., 2000; Flurkey et al., 2001; Carter et al., 2002). Circulating IGF-I is produced largely by the liver under predominant GH control, but IGF-I secretion also occurs in many tissues independent of GH. Low circulating IGF-I levels as a result of caloric restriction are also associated with increased lifespan in rodents (Sohal & Weindruch, 1996; Longo & Finch, 2003; Richardson et al., 2004). Therefore, it is unclear whether IGF-I exerts its biological effects through endocrine or paracrine/autocrine mechanisms. Here, we present a novel model of local IGF-I suppression that dramatically extends lifespan in mice on an unrestricted diet and sharply reduces incidence of spontaneous tumors.
The IGF system is complex with ubiquitous ligands and with receptors present on virtually all cells. In recent years, it has become increasingly clear that IGF binding proteins (IGFBP) and IGFBP proteinases are the ultimate determinants of IGF ligand availability and, hence, response (Bunn & Fowlkes, 2003). Pregnancy-associated plasma protein A (PAPP-A) is a newly discovered metalloproteinase that has functions outside of pregnancy to degrade inhibitory IGFBP in the pericellular environment thereby increasing IGF-I bioavailability without a change in IGF-I expression (Lawrence et al., 1999; reviewed in Boldt & Conover, 2007). Conversely, interventions that decrease PAPP-A secretion or activity result in diminished local IGF-I signaling. IGF system-independent effects of PAPP-A, although possible, have yet to be identified. It has been shown that PAPP-A is important for optimal fetal growth in humans (Smith et al., 2002) and mice (Conover et al., 2004). Thus, in the absence of PAPP-A, mice are born as proportional dwarfs (Conover et al. 2004), as PAPP-A plays an essential role in determining IGF bioavailability during early embryogenesis (Bale & Conover, 2005). If PAPP-A has antagonistic pleiotropic effects through its regulation of local IGF bioavailability, then loss of PAPP-A should prolong lifespan. In this study, we tested the hypothesis that genetic deletion of PAPP-A in mice results in increased longevity.
Survival distribution of PAPP-A knock-out and wild-type siblings from matings of heterozygous mice on a mixed C57BL6 and 129SV/E background is presented in Fig. 1. We found that loss of PAPP-A expression resulted in the extension of mean lifetime survival of these mice by a striking 38% (960 ± 28 days vs. 698 ± 23 days, P < 0.0001). The data in Fig. 1 are from pooled males and females, but separation by sex showed significant increases in lifespan for both males (33%) and females (41%). Maximum lifespan, that is, average age of the last decile of surviving mice, was 872 days for wild-type and 1166 days for PAPP-A knock-out mice. Increase in lifespan was not associated with alterations in serum glucose, insulin, total protein and cholesterol, IGF-I or GH levels (Table 1), and dietary intake was not significantly different between PAPP-A knock-out and wild-type mice when expressed as ‘g/kg/h’[5.55 ± 0.412 (n = 20) and 5.04 ± 0.257 (n = 19) for PAPP-A knock-out and wild-type, respectively]. Therefore, the longevity of these animals cannot be attributed to caloric restriction, GH abnormalities, or lowered circulating IGF-I levels.
|Glucose (mmol L−1)||11.0 ± 0.7||10.9 ± 0.6|
|Insulin (ng mL−1)||1.54 ± 0.20||1.46 ± 0.14|
|Cholesterol (ng dL−1)||179 ± 18||206 ± 16|
|Protein (µg mL−1)||2.6 ± 0.3||2.7 ± 0.1|
|IGF-I (ng mL−1)||295 ± 30||229 ± 22|
|GH (ng mL−1)||14.6 ± 4.6||10.2 ± 2.4|
Pathology on 23- to 28-month-old animals (Table 2) revealed the wild-type mice to be stricken with extensive tumors (primarily liver, lung, kidney and colon) and/or enlarged spleen and lymph nodes in 12 of 17 animals (mean age of 734 days) compared to single tumors of small size in liver in only three of 20 PAPP-A knock-out mice (mean age 755 days). The two oldest PAPP-A knock-out mice in the longevity cohort (1154 and 1179 days) were tumor-free at necropsy. PAPP-A knock-out mice were also found to be resistant to the development of experimentally induced neointimal hyperplasia (Resch et al., 2006) and atherosclerosis (Harrington et al., 2007).
|Evidence of tumors||12/17||3/20|
These data indicate the PAPP-A knock-out mouse as a valuable new model for investigating molecular issues of aging that relate to IGF-I signaling, and point to PAPP-A as a possible drug target with potential to promote longevity and suppress age-related diseases, such as cancer, by moderate restraint of IGF-I bioavailability during adult life without the need for caloric restriction.