A null mutation in SERPINE1 protects against biological aging in humans

Abstract

Plasminogen activator inhibitor-1 (PAI-1) has been shown to be a key component of the senescence-related secretome and a direct mediator of cellular senescence. In murine models of accelerated aging, genetic deficiency and targeted inhibition of PAI-1 protect against aging-like pathology and prolong life span. However, the role of PAI-1 in human longevity remains unclear. We hypothesized that a rare loss-of-function mutation in SERPINE1 (c.699_700dupTA), which encodes PAI-1, could play a role in longevity and metabolism in humans. We studied 177 members of the Berne Amish community, which included 43 carriers of the null SERPINE1 mutation. Heterozygosity was associated with significantly longer leukocyte telomere length, lower fasting insulin levels, and lower prevalence of diabetes mellitus. In the extended Amish kindred, carriers of the null SERPINE1 allele had a longer life span. Our study indicates a causal effect of PAI-1 on human longevity, which may be mediated by alterations in metabolism. Our findings demonstrate the utility of studying loss-of-function mutations in populations with geographic and genetic isolation and shed light on a novel therapeutic target for aging.

Fig. 1. Association of SERPINE1 genotype status and leukocyte telomere length as a function of age in the Berne Amish kindred.

(A and B) LTL in SERPINE1 null allele carriers and noncarriers in the Berne Amish kindred as quantified by (A) polymerase chain reaction (PCR) and (B) Southern Blot. Relative LTL is shown in (A), and mean terminal restriction fragment (TRF) length is shown in (B) as a function of age stratified by SERPINE1 mutation status. P value represents difference in mean LTL and TRF by SERPINE1 mutation status (carriers versus noncarriers) after adjustment for age, sex, and family structure in Sequential Oligogenic Linkage Analysis Routines (SOLAR) (P = 0.007 and P = 0.039, respectively). Every 1-year increase in age of study participant was associated with a 0.0087 lower relative LTL (P < 0.0001) and a 30–base pair lower mean TRF (P < 0.0001).

Fig. 2. Age at death in the extended Berne Amish kindred by genotype status for SERPINE1.

On the basis of ancestral data obtained from the extended pedigree and extensive fieldwork, we identified 221 individuals with known dates of birth and death. Genotype status for SERPINE1 was ascertained by direct genotyping or obligate ascertainment in 56 family members. The mean ± SD age at death was delayed by 7 years in SERPINE1 null carriers compared with unaffected individuals (82 ± 10 versus 75 ± 12; P = 0.037 by Wilcoxon rank sum test).

Fig. 3. Association of SERPINE1 genotype status and cardiovascular measures of aging as a function of age in the Berne Amish kindred.

(A to C) Components of the aging composite scores, including brachial pulse pressure (A), e′ velocity (B), and carotid IMT (C) as a function age in Amish participants by genotype status for SERPINE1 null allele. (D to F) Composite scores of cardiovascular, cardiometabolic, and comprehensive biological aging in SERPINE1 null allele carriers and noncarriers in the Berne Amish kindred. Several cardiovascular measures that correlated strongly with chronological age with absolute correlation coefficients (|R|) greater than 0.60 (P < 0.0001) were identified. Pulse pressure [PP = 34.7 + 0.39 (age in years)] and carotid IMT [0.32 + 0.01 (age in years)] increased with age. e′ velocity [e′ = 20.3 − 0.2 (age in years)] decreased with age, as expected. These measures of vascular structure and stiffness and myocardial health were then standardized to have mean = 0 and SD = 1 (z scores) and were integrated into composite score 1 (cardiovascular age composed of brachial pulse pressure, e′ velocity, and carotid IMT), composite score 2 (cardiometabolic age composed of score 1 plus fasting insulin), and composite score 3 (comprehensive biological age composed of score 2 and LTL). Z scores were coded so that higher values corresponded to older levels of the measures of biological aging; that is, The z scores for LTL and e′ velocity, which decline with age, were reverse-coded so that higher z scores corresponded to lower levels [**P value represents difference in components and composite scores of aging by SERPINE1 mutation status (carriers versus noncarriers) after adjustment for age, sex, and family structure in SOLAR].