Mice deficient in Klotho gene expression exhibit a syndrome resembling premature human aging. To determine whether variation in the human KLOTHO locus contributes to survival, we applied two newly characterized polymorphic microsatellite markers flanking the gene in a population-based association study. In a cohort chosen for its homogeneity, Bohemian Czechs, we demonstrated significant differences in selected marker allele frequencies between newborn and elderly individuals (P < 0.05). These results precipitated a search for functional variants of klotho. We identified an allele, termed KL-VS, containing six sequence variants in complete linkage disequilibrium, two of which result in amino acid substitutions F352V and C370S. Homozygous elderly individuals were underrepresented in three distinct populations: Bohemian Czechs, Baltimore Caucasians, and Baltimore African-Americans [combined odds ratio (OR) ‫؍‬ 2.59, P < 0.0023]. In a transient transfection assay, secreted levels of klotho harboring V352 are reduced 6-fold, whereas extracellular levels of the S370 form are increased 2.9-fold. The V352͞S370 double mutant exhibits an intermediate phenotype (1.6-fold increase), providing a rare example of intragenic complementation in cis by human single nucleotide polymorphisms. The remarkable conservation of F352 among homologous proteins suggests that it is functionally important. The corresponding substitution, F289V, in the closest human klotho paralog with a known substrate, cBGL1, completely eliminates its ability to cleave p-nitrophenyl-␤-D-glucoside. These results suggest that the KL-VS allele influences the trafficking and catalytic activity of klotho, and that variation in klotho function contributes to heterogeneity in the onset and severity of human age-related phenotypes.

Abstract

An atmospheric general circulation model, which assimilates data from daily observations of temperature, humidity, wind, and sea-level air pressure, was compared with a set of observations that combines satellite and ground-based measurements of solar flux. The comparison reveals that the model underestimates by 25 to 30 watts per square meter the amount of solar energy absorbed by Earth's atmosphere. Contrary to some recent reports, clouds have little or no overall effect on atmospheric absorption, a consistent feature of both the observations and the model. Of several variables considered, water vapor appears to be the dominant influence on atmospheric absorption.