A paradox in the hippocampus. Immature dentate granule cells are often described as the “plasticity reserve” of the hippocampus. They provide a pool of neurons that integrate into existing circuits, supporting learning, memory, and repair. In neurological disease, these cells have been suggested to buffer against injury or degeneration. In a recent publication, researchers showed that the hippocampus continues to generate new neurons throughout life, but that the molecular instructions for doing so vary dramatically across species. The surprising finding is this: the processes of neurogenesis are conserved, while the genes underlying these processes are often completely different. This is an important reminder that biology often converges at the level of function, even when the building blocks are not the same.

Figure 1. Gene ontology enrichment and overlapping genes in a cross-species analysis for immature granule cell (imGCs). On the left, the top biological pathways are shown for imGC-enriched genes across humans, macaques, pigs, and mice. These pathways, grouped and color-coded by Gene Ontology (GO) terms, highlighting the processes most important for immature neurons, such as neuronal development, synaptic plasticity, and ion transport. This view emphasizes that, while the precise genes differ from species to species, the biological themes remain consistent. On the right, a Venn diagram illustrates how few imGC-enriched genes are shared between species. Only a small set of genes is common to all four, with an additional subset found only in humans and macaques. The lists of shared genes in the boxes serve as concrete examples of this limited overlap, reinforcing the central message of the study: processes can be conserved even when the genes executing them diverge. [Figure modified from data provided by the authors].

From the operating room to single-cell data. I felt a particular connection to this study because our own work in biobanking surgical brain tissue contributed to it. I have written before about the value of building biobanks: the “gold of the 21^st century.” Years of carefully collecting and cataloging human samples creates opportunities that cannot be replaced by any model system. Here, hippocampal tissue obtained through epilepsy surgery—entrusted to us by our research participants—allowed us to explore questions about neurogenesis that cannot be answered in animal models alone. It was rewarding to see the connection between the biobanking world and the molecular world of single-cell sequencing.

Different genes, same processes. In our recent publication in Nature Neuroscience, Zhou and collaborators identified imGCs in humans, macaques, pigs, and mice using machine-learning approaches applied to single-cell RNA-sequencing data. What emerged was a striking pattern: while the biological processes were consistent—neurogenesis, synaptic plasticity, axonal growth—the genes responsible for these programs were almost entirely divergent between species. Only a handful of genes, like DPYSL5, overlapped in all four species. Even between humans and macaques, the overlap was less than 10%. Zhou and collaborators presented this as a simple Venn diagram (Figure 1), but it captures the essence of the paradox: divergent gene expression, convergent biology.

The human twist. Humans stood out because of a unique enrichment of proton-transporting vacuolar ATPases (v-ATPases). These genes are usually discussed in the context of lysosomal function, not neurogenesis. Yet, in human imGCs, they appear to be essential. When these ATPases were blocked in human iPSC-derived immature neurons, neurite growth and neuronal firing were reduced. This was not just a transcriptomic curiosity, it was a functional finding. The message is that immature human neurons use molecular tools that cannot be inferred from mouse studies.

Connections to disease genes. What makes this especially relevant for us in the rare disease community is that some of the human-enriched genes are already very familiar to us. We have written on this blog about CACNA1A, which spans a wide phenotypic spectrum from neurodevelopmental disorders with ataxia to hemiplegic migraine. Finding it enriched in human immature neurons suggests another angle as to why this gene has such broad clinical impact. We have also discussed SCN3A, a sodium channel gene associated with early-onset epilepsy and brain malformations that we initially identified as a disease gene in 2018. Its enrichment in human imGCs points to a role in development that may persist into adulthood. And RELN encoding reelin, which we highlighted in prior posts on neuronal migration, appears again as part of the molecular fingerprint of human imGCs. Each of these examples makes the same point: the biology of adult neurogenesis is not separate from the disease genes we see in clinic; it is deeply intertwined. At the same time, the study by Zhou and collaborators reminds us to be cautious: genetic programs tend to be stable across species, but the specific genes implementing them may not be. Insights from mouse or pig models must therefore be interpreted with care, as they may highlight the right processes but not always the right players.

Why this matters. The implications of the study by Zhou and collaborators are both evolutionary and clinical. Evolutionarily, the study supports the idea that species converge on similar solutions through different genetic programs. Clinically, we find that human-specific programs include the very genes that appear in our epilepsy genetic clinics and on genetic testing reports. This is also a practical reminder that therapeutic strategies aimed at neurogenesis require validation in human systems. Mouse data alone may not capture the complexity or specificity of the human hippocampus.

What you need to know. The recent paper by Zhou et al. in Nature Neuroscience provides one of the most comprehensive cross-species analyses of hippocampal neurogenesis to date. This study shows that immature neurons share conserved developmental and plasticity programs across species, but the underlying gene expression is highly divergent. Humans in particular use a unique molecular signature centered on vacuolar ATPases, which play functional roles in the growth and activity of immature neurons. Many genes enriched in human neurons are rare disease genes we know well, including CACNA1A, SCN3A, and RELN. For families who contribute to biobanking, this work shows why these efforts matter: they enable discoveries that reveal what makes the human brain unique, and they remind us to interpret model systems with caution when studying genes central to rare disease.

Ingo Helbig

Ingo Helbig is a child neurologist and epilepsy genetics researcher working at the Children’s Hospital of Philadelphia (CHOP), USA.