Convergent Amino-Acid Substitutions Link Primate Longevity to Male-Fertility Genes

Technical Summary: Convergent Amino-Acid Substitutions Link Primate Longevity to Male-Fertility Genes

Problem Statement
Maximum lifespan (MLS) varies significantly across mammals, yet the genomic basis for this variation remains incompletely understood. While comparative studies have linked longevity to genome maintenance (e.g., DNA repair) and reproductive trade-offs (antagonistic pleiotropy), the specific molecular mechanisms driving lifespan evolution within primates require further elucidation. Previous work by Farré et al. (2021) identified convergent amino-acid substitutions (CAAS) distinguishing long- and short-lived mammals, but a focused analysis within the primate order—leveraging denser genomic sampling—had not yet been performed to determine if it recovers a non-random component of this broader signal.

Methodology
The authors employed a phylogenetically informed comparative genomics approach using a dataset of 126 primate species matched to the Kuderna et al. (2023) phylogeny.

1. Phenotype Definition: Maximum lifespan and adult body mass were used to calculate the Longevity Quotient (LQ), defined as the observed MLS divided by the expected MLS based on a fixed mammal-wide allometric relationship ($Lifespan = 6.47 \times mass^{0.189}$). The dataset retained significant phylogenetic signal (Blomberg's $K=0.146$; Pagel's $\lambda=0.786$), necessitating an explicit phylogenetic comparison rather than treating species as independent observations.
2. Species Selection: A phenotype-shift score (PSS) framework ranked all 7,875 unordered species pairs based on phylogenetically normalized differences. The upper and lower 1% tails of these pairs were selected, followed by fixed absolute LQ thresholds to define two groups: seven "foreground" (FG) long-lived species ($LQ > 1.3$) and six "background" (BG) short-lived species ($LQ < 0.8$). These 13 species were distributed across the phylogeny to avoid domination by a single terminal clade.
3. Convergent Substitution Analysis: Using CAAStools, the authors screened 16,133 protein alignments for convergent amino-acid substitutions. They pooled comparisons across the FG and BG groups, collapsing repeated species appearances to ensure independent support counts.
   • Strict Criteria: Primary candidates required a positional $P \le 0.05$, support from at least five unique species in both FG and BG groups, and complete concordance of the dominant amino acid in the FG group.
   • Patterns: Two patterns were retained: Pattern 1 (FG and BG share distinct, non-overlapping states) and Pattern 2 (FG shares a state opposed by heterogeneous BG states).
4. Validation and Enrichment: The resulting gene set was compared against the discovery sets of Farré et al. (2021) using hypergeometric tests. Functional enrichment analysis was performed using g:Profiler against human orthologues, applying g:SCS correction for multiple testing.

Key Results

• Recurrent Substitutions: The stringent filtering process identified 1,068 recurrent phenotype-associated amino-acid positions across 934 genes. These positions were supported by repeated species concordance rather than nominal significance alone.
• Overlap with Mammalian Longevity Signals: The 934 primate-specific CAAS genes showed significant agreement with the mammal-wide longevity candidates identified by Farré et al. (2021). Specifically, 226 genes overlapped, representing a 1.88-fold enrichment over the 120 expected by chance ($P = 1.16 \times 10^{-22}$). This overlap persisted across different discovery scenarios, indicating that the primate-focused analysis recovers a non-random component of the broader mammalian longevity signal.
• Functional Enrichment: Functional analysis of the 934 genes revealed 17 significant terms. The strongest coherent signal involved male reproductive biology:
  • Male Infertility: 29 genes ($P = 3.43 \times 10^{-4}$).
  • Decreased Fertility in Males: 31 genes ($P = 7.94 \times 10^{-4}$).
  • Other significant terms included microvilli, actin-based cell projections, and extracellular-matrix receptor interactions. Notably, no Gene Ontology Biological Process terms reached significance.
• Intersection with Prior Work: Within the 31-gene set associated with decreased male fertility, 11 genes also overlapped with the Farré et al. discovery sets, and the majority contained strict Pattern-1 or Pattern-2 CAAS.

Significance and Claims
The paper claims that focused primate sampling successfully recovers a non-random genomic signal associated with lifespan evolution, validating the utility of phylogenetically informed CAAS analysis within specific taxonomic orders. The primary contribution is the identification of a strong, coherent link between longevity-associated amino-acid substitutions and male-fertility phenotypes.

The authors interpret this finding in the context of antagonistic pleiotropy, where alleles enhancing reproductive success may be associated with reduced longevity. However, the paper explicitly maintains modesty regarding causality:

• The results do not demonstrate altered fertility in the analyzed primates.
• The findings do not establish that a reproductive trade-off causally mediates longevity differences.
• The Human Phenotype Ontology memberships are annotation-based, not direct observations.

Consequently, the authors conclude that while the signal does not prove a causal mechanism, it identifies "male reproductive biology" as a promising and biologically plausible direction for future research into the evolution of lifespan. The study suggests that broader comparative analyses and functional investigations of these specific genes are required to determine the mechanisms by which male reproductive biology contributes to longevity evolution.