Adenylosuccinate lyase (ADSL) sits at the heart of purine metabolism, helping generate the molecules that fuel cells, build nucleic acids, and support signaling pathways in the body. Despite this fundamental role, the gene has taken an unusual evolutionary path: modern humans carry changes to ADSL not found in Neanderthals or Denisovans.
In a study published in PNAS, researchers have demonstrated that these modern human-specific variants lowered the enzyme’s activity and expression, leading to elevated concentrations of ADSL substrates in the brain. To uncover these effects, the team combined targeted metabolomics – using liquid chromatography–mass spectrometry (LC–MS) – with behavioral assays in humanized mouse models. Concentrations of ADSL substrates, including succinylaminoimidazole carboxamide riboside (SAICAr) and succinyladenosine (S-Ado), were measured across multiple tissues, while an automated tracking system was used to assess subtle changes in behavior.
The analyses revealed that in mice engineered to carry the modern human–like substitution, levels of ADSL substrates rose in multiple tissues, with the strongest effects seen in the brain – particularly in females. These mice also displayed behavioral differences, accessing water more efficiently than their wild-type littermates under restricted conditions.
Together, the findings suggest that natural selection acted repeatedly on ADSL during recent human evolution, reshaping purine metabolism in ways that may have influenced brain function. To learn more about the motivations, analytical challenges, and evolutionary implications of the study, we spoke with Izumi Fukunaga and Xiang-Chun Ju, co-lead authors from the Okinawa Institute of Science and Technology (OIST).
Please explain what the ADSL enzyme does, and why it matters in the brain?
ADSL is an essential gene in purine biosynthesis, encoding an enzyme that catalyzes two biochemical reactions in this pathway: conversion of SAICAr to AICAr and AMP synthesis. Purines and their derivatives serve as major cellular energy carriers (e.g. ATP and GTP), essential subunits of nucleic acids, signaling molecules (cyclic AMP and cyclic GMP) and ligands for purinergic receptors (ATP and adenosine). Purines are critical for all organs, but the enzyme expression level in the brain is particularly low. ADSL deficiency in humans can cause a number of brain disorders including psychomotor retardation, seizures and autism, underscoring its indispensable role in brain function.
What initially led your team to investigate differences in the ADSL gene between modern humans and Neanderthals/Denisovans?
In previous studies we found that humans, relative to apes and monkeys, have reduced purine biosynthesis. As purines are important molecules in our bodies – and ADSL is a key enzyme that carries a change in present-day humans not seen in Neandertals, Denisovans, apes or monkeys – we sought to better understand the effects of this change.
What was the biggest challenge you faced during the study – and how did you overcome it?
The biggest challenge was finding what effect the small amino acid change in ADSL had on behavior without prior knowledge. We overcame this using a mouse model with an automated, exploratory behavioral tracking system, which enables simultaneous screening of many aspects of behavior. To understand how the change affects humans, it’s our hope that analytical chemists could, in future, come up with brilliant, innovative methods to accurately and non-invasively measure metabolites.
What potential mechanisms or pathways do you think could link this specific change in ADSL to shifts in brain function or behavior?
We do not yet know if the underlying mechanism is related to the brain development, energy metabolism, or signaling – but we do know it is not via the adenosine receptor. Elucidating this mechanism will be fascinating.
We detected no binding between ADSL substrates and human adenosine receptors, indicating that neither substrate exerts its effects through direct interaction with these receptors. However, several other pathways directly linked to the products of ADSL-catalyzed reactions may influence ATP production and modulate the activity of purinergic receptors. For example, fumarate, one product, serves as an intermediate in the tricarboxylic acid (TCA) cycle, playing a crucial role in energy production. AICAR, another product, can directly bind to and activate AMP-activated protein kinase (AMPK) to restore cellular energy balance. Some studies have also suggested a potential role of the ADSL substrate SAICAR in regulating transcription via pyruvate kinase M2 (PKM2) in human cancer cells. Further investigations are required to elucidate the potential mechanisms of the amino acid change in brain function and behavior.
What do the two separate evolutionary changes to ADSL activity suggest to you about the selective pressures acting on early modern humans?
The two separate changes, as well as the widespread occurrence of the derived versions in modern humans, suggest strong and persistent selective pressures associated with recent human evolution.
In the study, only female mice appeared to gain a competitive advantage – what hypotheses do you suggest might have caused this sex-specific effect?
One possibility is that the reduced enzymatic activity had a stronger metabolic impact in female mice, which showed higher concentrations of SAICAR in the brain compared with male, “humanized” mice. This suggests that reduced ADSL activity may affect the female brain more strongly than the male brain. Other possibilities include sex-dependent brain development or hormonal regulation of the metabolic pathways. More clues about the underlying mechanism will further reveal why this effect appears in female mice.
What are the next steps for your team?
Uncovering how tiny genetic changes from our ancient past have helped shape the brain traits that make us human is, by nature, very exciting. Studying our ancient origins is like assembling a giant puzzle, where each genetic change in our ancestors may offer a clue to how our brains and behaviors evolved.
As we discussed in our paper, “Another important aspect illustrated by this work is that several changes may be needed to more fully model human phenotypes in an experimental system such as the mouse. Often, several changes affecting a single gene product or several gene products may be needed. In the case of ADSL, an amino acid change as well as noncoding changes affecting ADSL expression may be needed. In addition, those changes may exert their effects only in concert with other changes that affect, for example, purine metabolism or purinergic signaling. Thus, a promising approach to address modern human as well as archaic phenotypes in model organisms will be to combine several noncoding and coding changes that affect certain pathways, certain functional systems, or certain organelles in order to understand their combined effects.”
Newsletters
Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.
