Circular RNAs, or circRNAs, are a curious class of molecules. Unlike the familiar linear messenger RNAs that carry instructions from DNA to the protein-making machinery of the cell, circRNAs are formed when a transcript’s ends are joined back-to-front in a process called back-splicing, producing a closed loop. That loop makes them unusually stable and resistant to the exonucleases that rapidly degrade ordinary RNA. Over the past decade, circRNAs have emerged as important regulators of gene expression, with one of their best-characterized roles being that of molecular sponges: they can bind and sequester microRNAs, the small regulatory RNAs that would otherwise suppress target messenger RNAs. This competing endogenous RNA, or ceRNA, mechanism allows circRNAs to indirectly boost the expression of genes involved in everything from development to disease. In hair follicles, where stem cells, signalling pathways and structural remodelling must be choreographed with exquisite timing, such regulation is a natural candidate for controlling how coats grow.

The research team focused on the three canonical stages of the hair follicle cycle. During anagen, the growth phase, follicular cells proliferate vigorously and the hair shaft is actively formed. Catagen is a controlled regression, in which growth ceases and the follicle remodels itself. Telogen is the resting phase, during which the follicle maintains structural stability and preserves its regenerative potential until the next cycle begins. The researchers collected skin samples from Jinlan cashmere goats representing two fleece phenotypes: one in which the coarse hair is longer than the cashmere, abbreviated CHLC, and one in which the coarse hair is shorter than the cashmere, known as CHSC. Sampling was performed at all three developmental stages, giving the team a matrix of comparisons that could reveal not only which circRNAs differ between phenotypes, but when those differences matter most during the cycle.

Using transcriptome sequencing of the skin samples, the researchers screened for differentially expressed circRNAs between the two phenotypes at each stage. The numbers themselves tell a story of escalating molecular divergence. At anagen, the team identified 35 circRNAs expressed at significantly different levels between CHLC and CHSC goats. At catagen, that figure rose to 84, and at telogen it climbed further to 115. In other words, as the follicle cycle progresses from growth through regression to rest, the two fleece phenotypes become increasingly distinguishable at the level of their circular RNA transcriptomes. This pattern suggests that the molecular roots of the visible fleece difference accumulate across the cycle rather than appearing at a single decisive moment, with the resting phase carrying the largest burden of phenotype-associated regulatory differences.

To make sense of these differentially expressed circRNAs, the team turned to the genes in which they are encoded, known as host genes, and performed functional enrichment analyses using the Gene Ontology and KEGG pathway frameworks. The results revealed strikingly stage-specific functional signatures. In the anagen stage, the enriched functions centred on substance transport and energy metabolism, precisely the activities needed to fuel the rapid proliferation of follicular cells and the morphogenesis of the growing hair shaft. Biological process terms such as cation transport and transmembrane transport featured prominently, alongside KEGG pathways including carbon metabolism and lysine degradation. The picture is of follicles in overdrive, importing ions and metabolites across membranes and burning carbon skeletons to power construction of new fibre.

The catagen stage told a different story. Here, the enriched functions clustered around the cessation of cell growth and the maintenance of immune homeostasis, processes that facilitate the controlled remodelling of the follicle as it shuts down production. Enriched biological process terms included regulation of cellular processes and biological regulation more broadly, while the KEGG analysis highlighted the cell cycle and Ras signalling pathways, both central to deciding when cells stop dividing and how tissues reorganize. By telogen, the emphasis had shifted again, this time toward structural stability and the maintenance of internal equilibrium, functions that reserve the follicle’s regenerative capacity for the next round of growth. Enriched terms included DNA metabolic processes, and pathways such as homologous recombination and tight junction appeared, hinting at genome maintenance and barrier integrity during the quiescent phase.

Beyond these stage-specific modules, the analysis also uncovered functions shared across stages, forming what the authors describe as a coordinated regulatory system that ensures orderly transitions through the follicle cycle. Signal transduction emerged as a core shared theme between the anagen and catagen stages, and ubiquitin-mediated proteolysis, the cellular machinery for tagging proteins for destruction, was co-enriched in those same two phases, consistent with the wholesale protein turnover required as a growing follicle converts into a regressing one. Lysine degradation was common to the anagen and telogen stages, while several key pathways, including tight junction assembly and the PI3K-Akt signalling cascade, a major regulator of cell survival and growth, were shared between catagen and telogen. These overlapping modules suggest that the boundaries between cycle phases are bridged by continuous regulatory programmes rather than abrupt molecular switches.

Perhaps the most consequential result came from the ceRNA network analysis performed on the catagen stage. By integrating the interactions between circRNAs, microRNAs and their target messenger RNAs, the team identified four hub candidate genes: TRIM32, IL34, IDUA and SLC16A3. Each has plausible links to follicle biology. TRIM32 is an E3 ubiquitin ligase known to be involved in maintaining the homeostasis of stem cells, including the neural stem cells from which it was first characterized, making its role in follicle stem cell maintenance a compelling hypothesis. IL34 is a cytokine implicated in immune regulation, IDUA encodes an enzyme of glycosaminoglycan degradation with potential relevance to the extracellular matrix of the skin, and SLC16A3 is a monocarboxylate transporter that shuttles lactate and other metabolites across membranes, fitting neatly with the metabolic demands of remodelling tissue. Together, the authors note, these genes are closely involved in follicle stem cell homeostasis, hair structural remodelling and skin follicle development.

The practical implications extend well beyond basic biology. Cashmere is one of the most valuable animal fibres in the world, and breeding programmes have traditionally relied on phenotypic selection, a slow process given that fleece traits only become fully apparent as animals mature and cycle through growth seasons. A molecular understanding of why some goats grow cashmere that outpaces their guard hairs could allow marker-assisted selection or even genome editing to bias flocks toward the more desirable CHSC phenotype. The circRNA catalogue assembled in this study, together with its stage-specific functional annotations and the four hub genes, provides exactly the kind of genetic resource that such programmes require. The work was supported by the China Agriculture Research System and related Shanxi provincial and institutional funds, reflecting the strategic importance of cashmere production to the region.

Like all transcriptomic studies, this one is a map rather than a mechanism. The differential expression of circRNAs and the enrichment of their host genes’ functions are correlations that now invite direct experimentation: validating the ceRNA interactions, perturbing the hub genes in follicle cultures, and testing whether manipulating specific circRNAs shifts follicles between phenotypic trajectories. Still, the study marks an important step in an emerging field, demonstrating that the dark matter of the transcriptome, the circular RNAs once dismissed as splicing noise, carries stage-specific information that tracks one of the most economically significant traits in goat husbandry. As sequencing costs fall and functional tools mature, the humble cashmere goat may become a model for how non-coding RNAs sculpt complex cyclical organs, with lessons that reach from the fleece on a plateau-dwelling herd to the follicles on a human scalp.

Subject of Research: Circular RNA regulation of hair follicle cycling and fleece phenotypes in Jinlan cashmere goats

Article Title: Identification and bioinformatic functional annotation of circRNAs mediating phenotypic differences in hair follicle cycling of Jinlan cashmere goats

Article References: Zhang, L., Guo, L., Wang, Q., Sun, H., Zhang, T., Guo, H., Han, J., & Zhao, P. (2026). Identification and bioinformatic functional annotation of circRNAs mediating phenotypic differences in hair follicle cycling of Jinlan cashmere goats. BMC Genomics. https://doi.org/10.1186/s12864-026-13401-7

Image Credits: AI Generated

DOI: 10.1186/s12864-026-13401-7

Keywords: circRNA, cashmere goats, hair follicle cycle, ceRNA, RNA-seq, transcriptomics, fleece phenotypes, anagen, catagen, telogen, molecular breeding, non-coding RNAs