Bidirectional communication between the skin and brain regulates local and global homeostasis in the body [1]. The skin is the first barrier located at the interface with the external environment, and it senses, collects, and responds to environmental stimuli (such as oxidative stress, ultraviolet (UV) and infrared radiation, and pollution) [2]. Skin converts these external stresses into cellular signals. These molecular signals are then transmitted to the brain or other organs and affect the global homeostasis of individuals [1,2,3,4,5,6,7]. UV rays are an essential component of sunlight and are considered the most important source of external stress causing extrinsic aging, which is also referred to as photo-aging [3]. Several studies suggest that UV exposure of the skin not only induces skin-related biochemical responses, but also affects signal transduction pathways in other organs such as the brain. For instance, UV exposure increases both skin and brain levels of β-endorphin in mice, which leads to opioid-related antinociception and the subsequent addiction of mice to UV light [8]. Moderate UV exposure increases the blood level of urocanic acid and enhances the learning and memory of the mouse brain through the glutamate biosynthetic pathway [9]. As the primary brain area involved in cognitive function [10], the hippocampus plays a major role in learning and memory formation [11]. Thus, the hippocampus is widely considered an essential target for the investigation of cognitive function. Previously, we reported that repeated UV exposure of the skin decreases hippocampal neurogenesis and synaptic plasticity [1]. Although mounting evidence suggests that UV exposure may induce neurobehavioral changes in the brain, the molecular mediators are largely unknown. Moreover, the cellular players involved in UV-induced neurobehavioral changes have not been explored. To specify the molecular and cellular signatures associated with UV-induced functional changes in the brain, especially in learning and memory, we analyzed the hippocampal transcriptome in chronic UV-irradiated photoaged mice.

We established a skin aging model by sequentially increasing the UV dose three times a week for 6 weeks (Fig. 1A). The dorsal skin of anesthetized mice was UV irradiated in the treated group, and the mice in the control group were anesthetized and sham-irradiated. After UV irradiation for 6 weeks, hematoxylin and eosin (H&E) staining confirmed that the epidermal layers were thickened, indicating that chronic UV irradiation induces skin aging (Fig. 1B). To investigate the effect of UV irradiation on cognitive function, mice were subjected to the object place recognition (OPR) behavior test, which depends on intact hippocampal function [12]. The discrimination index was significantly lower in UV-irradiated mice than in sham-irradiated mice (Fig. 1C, D, P = 0.049), indicating that chronic UV irradiation impaired cognitive function (Additional file 1).

Chronic skin UV irradiation induced hippocampal memory deficit and transcriptomic changes associated with microglial function. A–D Experimental schedule of UV irradiation given 3 days a week for 6 weeks. After 5 weeks of UV irradiation, the object place recognition test was performed. B H&E staining. After 5 weeks of UV irradiation, the discrimination index was calculated as the difference between C, D the exploring time to novel place object (N) and familiar object (F), divided by the total time exploring both objects (discrimination index = (N − F)/(N + F).*P < 0.05 vs. control group. Each bar represents the mean ± SEM of each group. E–G Result of differentially expressed gene (DEG) analysis. E Volcano plot of UV-DEGs. Y axis refers to the adjusted p-values in − log10 scale. X axis refers to the mean expression ratio of UV-irradiated and control mice samples in log2 scale. The colors of the dots indicate the direction of regulation (red: up-regulation, blue: downregulation, gray: non-significant regulation). Adjusted p-values lower than 1.0 × 10–120 are displayed as 2.0 × 10–120 in the figure. Top 10 most downregulated DEGs are marked by larger points and labeled. F Functional annotations for UV-DEGs showing the top 15 significantly enriched gene ontology (GO) biological pathways. Top five UV-DEGs ranked by FDR overlapping with leading edge genes are shown, with letter colors indicating the direction of regulation (red: up-regulation, blue: downregulation). G Enrichment plot of microglia markers. X axis refers to the rank of genes in descending order of log2 fold change values. The vertical black lines on the X axis are the genes overlapping with the markers. The green line shows the cumulative enrichment score, calculated by walking down the gene list, where the score increases if a gene overlaps with the markers and decreases if a gene does not. The red dashed lines indicate the maximal and minimal scores. H Representative photographs of Iba-1+ cells in the hippocampal region. I The total number of Iba-1+ cells in the hippocampus were quantified in the graph. Each bar represents the mean ± SEM of each group. The asterisks denote a significant difference (**P < 0.01 vs. control group)

To explore the molecular and cellular mechanisms underlying the memory deficit induced by chronic UV exposure to the skin, we performed RNA-seq analysis of the hippocampi from UV- and sham-irradiated mice and identified a total of 338 differentially expressed genes (DEGs), including 10 upregulated and 328 downregulated genes in UV-irradiated mice, compared to sham-irradiated controls (Fig. 1E and Additional file 2: Table S1). To elucidate biological functions perturbed by UV irradiation, we performed pre-ranked gene set enrichment analysis (GSEA) using mouse gene ontology (GO) biological pathway terms provided by the MsigDB collection [13]. Immune-related pathways were significantly enriched for downregulated DEGs, including immune response (GO: 0006955, normalized enrichment score (NES) = − 1.46, false discovery rate (FDR) = 2.2 × 10–10), regulation of immune system process (GO: 0002682, NES = − 1.33, FDR = 1.4 × 10–4), and immunoglobulin production (GO: 0002377, NES = − 1.71, FDR = 5.6 × 10–4) (Fig. 1F and Additional file 2: Table S1). Several genes, including Pla2g5, Lbp, and Wfdc2, that commonly appear in these pathways were significantly downregulated and validated in the UV-exposed mice (Additional file 3: Fig. S1). This suggests that UV exposure reduces the functionality of the neuroimmune system and putatively leads to glia-related disruption in the central nervous system. Thus, we dissected the cell-type enrichment of DEGs using a single-cell transcriptome dataset of the mouse cortex and hippocampus [14]. We collected cell-type-specific marker genes for eight major cell types: hippocampal CA1 pyramidal neurons, interneurons, microglia, astrocytes, oligodendrocytes, endothelial, ependymal, and mural cells. Our GSEA results showed that microglia (NES = − 1.31, FDR = 1.5 × 10–2) were significantly enriched for downregulated DEGs in UV-exposed mice (Fig. 1G and Additional file 2: Table S1). Taken together, this suggests that UV exposure leads to a reduction in microglia-specific gene expression and may impair microglial functionality in the neuronal system.

Microglial activation can be determined by various parameters such as cell number, morphology, gene expression, or cytokines. In the hippocampus, it was demonstrated that the extent of microglial cell activation increased with an increase in the “number of microglial cells” [15]. To further investigate microglia-mediated neuroinflammation after chronic UV exposure, we conducted immunohistochemical evaluation of microglial activation. The number of ionized calcium-binding adaptor protein-1(Iba-1)-positive microglial cells in each group were quantified. As indicated in Fig. 1H, I, the number of Iba-1-positive cells was significantly increased in the hippocampal regions of UV-irradiated mice. These results indicate that UV exposure may induce the activation of microglial cells and consequently microglia-mediated neuroinflammation in the hippocampus. It is of note that the expression of Iba-1 was not altered at 6 weeks (Additional file 2: Table S1). The discrepancy between immunostaining and mRNA assays has also been reported in previous studies. Furthermore, immunostaining that can detect changes in the number and morphology of micrmicroglia has been suggested as a sensitive marker for activated microglia [15]. Also, the correlation between downregulation of specific microglia-related genes and microglial activation requires further investigation.

In conclusion, our findings highlight the unambiguous dysfunction of cognitive and neuroimmune systems induced by chronic UV irradiation. By analyzing the hippocampal transcriptome, we first demonstrated that repeated UV exposure of the skin could reduce the expression of neuroimmune system- and microglia-related genes. Moreover, after UV irradiation, microglial cells are activated, which may cause microglia-mediated neuroinflammation in the hippocampus.