The research, conducted by an international team of investigators, presents a comprehensive comparative resource for adult human and mouse adrenal glands built on two complementary technologies: single-cell transcriptomics, which profiles the gene activity of thousands of individual cells, and spatial transcriptomics, which maps where those gene-expression programs reside within the intact architecture of the organ. By combining these approaches, the authors have produced what is effectively a high-resolution atlas of the adrenal gland in two species, capturing not only which cell types exist but also how their molecular identities and locations differ between human and mouse.

The significance of this resource lies in what it corrects. The adrenal cortex, the outer layer of the gland, is organized into zones that produce different classes of steroid hormones. In the classical textbook model, mineralocorticoids such as aldosterone are produced in the outermost zone, glucocorticoids such as cortisol in the middle zone, and androgen precursors in the innermost zone. This zonal model was largely established through work in rodents. The new single-cell and spatial data confirm that while broad zonal logic exists in both species, the underlying cellular states, the repertoire of expressed enzymes and the dynamics of cell differentiation diverge substantially between human and mouse, meaning that findings in mice cannot be assumed to apply directly to human adrenal physiology.

One of the study’s central findings concerns steroidogenic states. Steroid-producing cells do not represent a single, fixed identity; instead, they occupy a spectrum of molecular states defined by which steroidogenic enzymes they express, at what levels, and in what combinations. The comparative analysis shows that these states are shaped by species-specific programs. Human adrenal cells display enzymatic configurations and regulatory signatures that differ from their mouse counterparts in ways that affect how hormone synthesis is partitioned across the gland. For researchers studying disorders such as congenital adrenal hyperplasia, adrenal insufficiency, Cushing’s syndrome and primary aldosteronism, this is a critical caveat: therapeutic strategies validated in mouse models may engage cellular programs that human adrenal tissue either lacks or deploys differently.

The second major theme of the study is tissue turnover. The adrenal cortex is one of the most dynamic organs in the body, with its steroid-producing cells continuously replaced throughout life. The prevailing model, again derived largely from rodent studies, holds that a cap of progenitor cells near the outer surface of the gland continuously generates new steroidogenic cells that migrate inward, mature, perform their hormonal duties and eventually die, a conveyor-belt-like process of renewal. The new human data reveal that this turnover process, while conceptually conserved, proceeds with species-specific characteristics. The cellular sources of renewal, the differentiation trajectories and the pace of cell replacement differ between human and mouse, underscoring that the mouse conveyor-belt model is an approximation rather than a faithful replica of human adrenal maintenance.

Technically, the study’s strength comes from its dual-method design. Single-cell RNA sequencing excels at resolving cellular heterogeneity, distinguishing rare cell populations and reconstructing differentiation trajectories from the gene-expression fingerprints of individual cells. But dissociating an organ into single cells destroys spatial context, which is essential in an organ as architecturally organized as the adrenal gland. Spatial transcriptomics restores that context by measuring gene activity in intact tissue sections, allowing the researchers to verify that the cell states identified in the single-cell data occupy coherent anatomical positions. By applying both methods in parallel to human and mouse adrenal glands, the team could cross-validate cell-type assignments and map species differences with confidence, producing a resource designed to serve as a reference standard for the field.

The comparative framing also carries implications for regenerative medicine and drug development. Efforts to grow functional adrenal tissue in the laboratory, whether from stem cells or organoid cultures, depend on knowing which molecular programs must be activated to generate authentic steroid-producing cells. If those programs are species-specific, protocols optimized against mouse reference data may steer human cells toward the wrong developmental endpoints. Conversely, a human adrenal atlas provides a benchmark against which laboratory-grown adrenal cells can be quality-controlled, accelerating the path toward cell-based therapies for patients whose adrenal glands no longer function.

The resource is also expected to inform cancer research. Adrenocortical carcinoma is a rare but aggressive malignancy with limited treatment options, and its cellular origins remain incompletely understood. A detailed map of normal adrenal cell states in humans, including the progenitor populations and differentiation intermediates that tumors may hijack, offers researchers a framework for identifying which programs are reactivated in cancer and for designing therapies that target tumor-specific vulnerabilities while sparing normal steroid production.

Beyond disease, the study speaks to a broader lesson in modern biology: organ-level conclusions drawn from one species do not automatically transfer to another, even for organs as structurally similar as human and mouse adrenals. Evolution has conserved the gland’s essential function, the synthesis of life-sustaining steroids, but has implemented that function through partly divergent cellular and molecular means. As single-cell atlases accumulate across tissues and species, this theme recurs, and the adrenal gland now stands as a particularly clear example because its physiology is so directly tied to clinically essential hormones.

The authors have made their comparative dataset available as a community resource, allowing endocrinologists, developmental biologists and computational scientists to query cell-type markers, explore spatial gene-expression patterns and test their own hypotheses against the data. In an era when mouse models remain indispensable but increasingly scrutinized, resources of this kind provide the translational bridge that the field needs: a way to know precisely where the mouse is a faithful model of human biology, and where it is not. For the adrenal gland, at least, the answer is now written cell by cell, and it is more species-specific than anyone had fully appreciated.

Subject of Research: Comparative single-cell and spatial transcriptomic analysis of adult human and mouse adrenal glands

Article Title: Human and mouse adrenal glands are characterized by species-specific steroidogenic states and tissue turnover

Article References: Kastriti, M. E., Maksimov, D., Krupinova, J., Utkina, M., Beltsevich, D., Roslyakova, A., Glazova, O., Kaziakhmedova, S., Ryabova, A., Kuznetsova, A., Shcherbakova, A., Bondarenko, E., Antysheva, Z., Albert, E., Urusova, L., Lapshina, A., Chevais, A., Avsievich, E., Trofimov, V., … Adameyko, I. (2026). Human and mouse adrenal glands are characterized by species-specific steroidogenic states and tissue turnover. Nature Genetics, 58(9), 2270-2283. https://doi.org/10.1038/s41588-026-02737-1

Image Credits: AI Generated

DOI: 10.1038/s41588-026-02737-1

Keywords: adrenal gland, single-cell transcriptomics, spatial transcriptomics, steroidogenesis, tissue turnover, human-mouse comparison, adrenal cortex, hormone production, cell atlas, Nature Genetics, endocrinology, regenerative medicine