HaCaT cells were purchased from Wuxi Xinrun Biotechnology Co., Ltd. (Wuxi, China). Ectoine (≥98 %) was purchased from Shanghai Titan Scientific Co., Ltd. (Shanghai, China). Dulbecco’s modified Eagle’s medium (DMEM), TRIzol reagent, PrimeScript™ RT reagent kit with gDNA Eraser, and TB Green® Premix Ex Taq™ kit were obtained from TAKARA Holdings Inc. (Japan). Cell Counting Kit-8 was purchased from Elabscience Biotechnology Co., Ltd. (Wuhan, China). Annexin V-FITC Apoptosis Detection kit was purchased from TransGen Biotech Co., Ltd. (Beijing, China). The reactive oxygen species (ROS) test kit was purchased from Wuhan Servicebio Technology Co., Ltd. (Wuhan, China). An xMark microplate reader was bought from Bio-Rad Laboratories (USA). A CO₂ incubator (BPN-150CH) was purchased from Thermo Fisher Scientific (USA); a CytoFLEX flow cytometer was obtained from Beckman Coulter, Inc. (USA); and a LightCycler® 96 real-time PCR system was purchased from F. Hoffmann-La Roche Ltd. (Switzerland). Qubit 4.0 and NanoDrop 2000 spectrophotometers were manufactured by Thermo Fisher Scientific (USA). An Agilent 5300 Bioanalyzer was provided by Agilent Technologies Inc. (USA). A NovaSeq X Plus sequencer was purchased from Illumina, Inc. (USA).

HaCaT cells were cultured in high-glucose DMEM supplemented with 10 % fetal bovine serum (FBS) and 1 % penicillin–streptomycin at 37 °C in 5 % CO₂. Cells were observed under an inverted microscope, and experiments were performed when confluence reached 90–95 %.

HaCaT cells were seeded into 96-well plates (5 × 10³ cells well⁻¹) and incubated for 24 h at 37 °C, 5 % CO₂. After washing with PBS, cells were treated with a concentration gradient of ectoine (0.05–0.50 mg mL⁻¹; n = 5 per group) for 24 h. Cell viability was assessed with the CCK-8 assay to determine the optimal ectoine concentration.

HaCaT cells were seeded in 6-well Lab-Tek chambers with complete DMEM and grown to 80 % confluence. Cells were divided into a control group (C, 0 mg mL⁻¹) and an ectoine group (E, 0.20 mg mL⁻¹) (n = 3 per group). Cells were centrifuged at 1000 rpm for 5 min and washed with PBS. One aliquot was resuspended in 100 µL Annexin V Binding Buffer, supplemented with 5 µL Annexin V-FITC and 5 µL PI, and incubated for 15 min at room temperature in the dark. The remaining cells were incubated with 10 µmol L⁻¹ DCFH-DA in serum-free medium at 37 °C for 30 min in the dark. After two PBS washes, cells were filtered through a 300-mesh nylon membrane and analyzed on a flow cytometer to measure apoptosis and ROS levels.

Total RNA from groups C and E (n = 3 per group) was extracted using TRIzol, and integrity was verified on 1 % agarose gels. RNA purity (OD_260/280 = 1.8–2.2, OD_260/230 ≥ 2.0) and integrity (RIN ≥ 6.5, 28S:18S ≥ 1.0; >1 µg) were confirmed on a NanoDrop 2000 and Agilent 5300, respectively. Double-stranded cDNA was synthesized with the SuperScript kit and random hexamers, size-selected to ≈300 bp, and amplified by PCR. Strand-specific libraries were prepared, validated, and quantified on a Qubit 4.0 and sequenced on a NovaSeq X Plus. Adapter-containing reads, reads with >50% bases of Q ≤ 20, or containing undetermined bases were removed with fastp (v0.23.4)13. Clean reads were aligned to the reference genome using HISAT2 (v2.2.1)14 and assembled with StringTie15 .

Transcripts per million reads (TPM) was estimated with RSEM16. Differential expression was analyzed with DESeq2 (v1.46.0)17 or DEGseq (v1.26.0)18; genes with |log₂FC| ≥ 1 and FDR ≤ 0.05 (DESeq2) or FDR ≤ 0.001 (DEGseq) were considered significantly differentially expressed. GO and KEGG enrichment were performed with GOATOOLS and KOBAS, respectively; terms with a Bonferroni-corrected P ≤ 0.05 were deemed significant19.

RNA samples meeting the above quality criteria (concentration > 5 ng µL⁻¹) were reverse-transcribed using the PrimeScript™ RT kit with gDNA Eraser. The 20 µL reverse-transcription mix contained 10 µL Reaction I (5 × gDNA Eraser Buffer 2 µL, gDNA Eraser 1 µL, RNA + RNase-free H₂O to 7 µL); incubation was 42 °C for 2 min. Reaction II (PrimeScript RT Enzyme Mix I 1 µL, RT Primer Mix 1 µL, 5 × PrimeScript Buffer 2 4 µL, RNase-free H₂O 4 µL) was then added. cDNA was diluted 1:20 for qPCR. The 20 µL qPCR mix contained 10 µL TB Green Premix Ex Taq II, 0.8 µL each primer (10 µM), 2 µL cDNA, and 6.4 µL H₂O. Cycling: 95 °C 3 min; 45 cycles of 95 °C 10 s, 58 °C 20 s, 72 °C 30 s, with fluorescence acquisition during extension. PPIA served as the reference gene (n = 3 per group). Relative expression was calculated by the 2^⁻ΔΔCt method. Primers were synthesized by Sangon Biotech (Table 1 ).

Table 1

Genes name	Primer sequences (5'→3')	Length (bp)
ANGPTL4	F: GAGGCTGGACAGTAATTCAGA, R:CGTGATGCTATGCACCTTCT	134
MMP25	F: GACTGGCTGACTCGCTATGG, R:TGATGGCATCGCGCAACTT	88
NOXO1	F: ATTCAGGCAGCTCAAGACCC, R:TGACCGAGAAGCTTTGGGAG	93
CNFN	F: TTGCTCCTCTGTGCCTTGCC, R:ACGGAGCCCTGGATGTGGT	130
FoxO6	F: GTGGGGGAACCTTTCCTACG, R:TTCTGCACGCGGATGAACC	207
PPIA	F: GTCAACCCCACCGTGTTCTT, R:GTCAACCCCACCGTGTTCTT	97

Statistical analyses were performed in GraphPad Prism 8. Normality was tested by the D'Agostino–Pearson omnibus test. Two-tailed Student’s t-tests compared two groups, and one- or two-way ANOVA was used for ≥3 groups. Data are reported as mean ± SD (ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Cell viability was assessed after treating HaCaT cells with different concentrations of ectoine. Results showed that cell viability increased as ectoine concentrations rose from 0.05 mg/mL to 0.20 mg/mL (103.94 ± 1.95 and 134.00 ± 3.22, respectively), and then declined as the concentration increased from 0.20 mg/mL to 0.50 mg/mL (107.10 ± 2.47; Figure 1). Overall, cell viability was significantly increased (>100 %) in all ectoine-treated groups (Table S1), indicating that ectoine may enhance the proliferation of HaCaT cells. Consequently, based on experimental results, 0.20 mg/mL was determined to be the optimal concentration for further experiments.

Figure 1

Apoptosis rates were analyzed using flow cytometry (Figure 2ab) after treating cells with 0.20 mg/mL ectoine. The apoptosis rate in the control group was 4.65 ± 0.57 %, whereas it was significantly lower in the ectoine-treated group (1.91 ± 0.19 %; Table S2). This suggests that ectoine may regulate HaCaT cell apoptosis by exerting anti-apoptotic effects. Intracellular reactive oxygen species (ROS) levels were measured using flow cytometry (Figure 2cd). The relative DCFH-DA fluorescence intensity was significantly decreased in the ectoine-treated group, indicating a substantial reduction in intracellular ROS levels (Table S3). These findings demonstrate that ectoine effectively reduces endogenous ROS and minimizes oxidative stress.

Figure 2

The raw data were filtered and screened to obtain more than 41,204,442 clean reads per group, with Q30 > 93.99 % and Q20 > 97.90 %. The error rate was <0.03 %, and the GC content ranged from 51.22 % to 52.79 % (Table 2 ). Clean reads were aligned to the reference genome, and the mapping rates were statistically analyzed. Total mapping ranged from 97.39 % to 97.55 %, including a single-match rate >93.40 % and a multi-match rate of 3.03–4.10 %. These results indicate that the sequencing data were reliably processed and are qualified for subsequent analyses.

Table 2

Sample	Raw reads	Clean reads	Q20 (%)	Q30 (%)	GC (%)	Total mapped	Multiple mapped	Unique mapped
C1	47683022	47205576	98.02	94.29	52.36	97.48%	3.03%	94.45%
C2	43301474	42862366	97.9	93.99	51.9	97.39%	3.07%	94.32%
C3	53243942	52725454	97.97	94.17	52.79	97.55%	3.33%	94.22%
E1	41565756	41204442	97.99	94.21	51.05	97.53%	3.47%	94.06%
E2	46817134	46392670	98.05	94.4	51.49	97.52%	3.44%	94.09%
E3	52862954	52385586	98.08	94.45	51.22	97.50%	4.10%	93.40%

The expression levels of DEGs in the control (C) and ectoine (E) groups, measured as log2(FPKM + 1), were subjected to hierarchical clustering. Ectoine reversed the transcriptional differences observed between the C and E groups: clusters with low expression in the C group shifted to high expression in the E group, whereas clusters with high expression in the C group became low in the E group. More low-expression than high-expression clusters were observed in the E group, suggesting that ectoine primarily suppresses overall gene expression in HaCaT cells (Figure 3a). Differential-expression analysis (log2 fold-change ≠ 0, P ≤ 0.05) identified 292 DEGs between C and E, including 87 up-regulated and 205 down-regulated genes (Figure 3b).

Figure 3

DEGs between the C and E groups were annotated using the Gene Ontology (GO) database. In total, 292 DEGs were enriched in 43 GO terms. The top 20 enriched terms fell into three categories—biological process (BP), cellular component (CC), and molecular function (MF; Figure 4a). Within BPs, DEGs were mainly enriched in cellular processes, biological regulation, and metabolic processes. For CCs, DEGs were largely enriched in cell parts, organelles, and membrane components. Regarding MFs, DEGs were primarily linked to binding and catalytic activities.

Mapping DEGs to the KEGG pathway database grouped them into five categories (top 20 pathways shown). In Metabolism (21 DEGs), genes were mainly involved in carbohydrate and lipid metabolism. In Environmental Information Processing (32 DEGs), genes were related to signal transduction and signaling-molecule interactions. In Cellular Processes (21 DEGs), genes concerned cell growth, death, and eukaryotic communities. In Organismal Systems (64 DEGs), genes were involved in the immune, endocrine, and nervous systems. In Human Diseases (69 DEGs), genes were mainly associated with cancer and endocrine and metabolic diseases (Figure 4b).

Figure 4

GO enrichment identified the 20 most significantly enriched terms, primarily associated with molecular functions (3) and biological processes (17; Figure 5a). Five GO terms showed the greatest enrichment: glutamine catabolic process (rich factor = 0.40; 2 DEGs), glutaminase activity (0.40; 2 DEGs), interleukin-21-mediated signaling pathway (0.333; 2 DEGs), glutamate biosynthetic process (0.333; 2 DEGs), and regulation of oocyte maturation (0.333; 2 DEGs).

KEGG enrichment analysis showed that DEGs participated in 208 signaling pathways. The 20 pathways with the most significant enrichment were analyzed (Figure 5b); the Ras signaling pathway contained the largest number of enriched genes (Figure 5c). The Ras pathway is a crucial intracellular signaling mechanism involved in cell growth, differentiation, apoptosis, and metabolism. Ectoine appears to act by activating receptor tyrosine kinases (RTKs) on the cell surface, triggering downstream cascades that lead to Ras activation. Activated Ras mediates anti-apoptotic effects via the PI3K/Akt pathway: the Ras-PI3K/Akt axis suppresses the AFX transcription factor, thereby down-regulating Fas ligand (FasL) expression, and also promotes NF-κB signaling, collectively regulating apoptosis, cell survival, and growth.

Figure 5

Ectoine treatment markedly enhanced HaCaT cell viability while reducing apoptosis and ROS levels. Based on the transcriptomic data, five key DEGs associated with oxidation, apoptosis, and proliferation were selected for RT-qPCR validation (Table S4): MMP25 (extracellular-matrix degradation), NOXO1 (endogenous ROS generation), ANGPTL4 (inflammatory oxidative stress), FOXO6 (pro-apoptotic), and CNFN (keratinocyte structural protein). RT-qPCR showed that ANGPTL4 expression was significantly down-regulated (Figure 6a), with lower levels than indicated by RNA-seq. MMP25, NOXO1, and FOXO6 were also significantly down-regulated (Figure 6b–d), whereas CNFN was significantly up-regulated (Figure 6e) in ectoine-treated cells, consistent with RNA-seq findings. Overall, RT-qPCR results corroborated the RNA-seq data regarding up- and down-regulation patterns (Figure 6f). The accuracy of gene-expression quantification can be affected by primer design, instrument sensitivity, and reagent quality.

Figure 6

Apoptosis can be induced by altering the external environment, such as intense UV irradiation and desiccation, by generating ROS. ROS-induced cellular oxidative stress can cause mitochondrial damage, leading to calcium-ion imbalance and inducing the release of apoptotic proteins from the cells, which initiates the process of apoptosis and accelerates the process of skin aging20, 21. Hseu et al.9 found that after pretreatment of HaCaT cells with ectoine, the expression of antioxidant proteins (HO-1, NQO-1, γ-GCLC) and the catalytic subunit of γ-glutamylcysteine ligase (γ-GCLC) was up-regulated, thereby inhibiting the production of ROS. Cheng et al.10 observed that ectoine alleviated oxidative damage in human skin fibroblasts by up-regulating genes associated with the PI3K/AKT signaling pathway, such as COL1A1, COL1A2, FN1, IGF2, NR4A1 and PIK3R1. In skin cells, NADPH oxidases (NOXs) are endogenous sources of ROS production, while the NOX1 holoenzyme is the main source of ROS produced by ultraviolet (UV) radiation22. NOXO1, an essential subunit of NOX1 (encoded by the gene NOXO1), plays a crucial role in activating NOX1 enzymes. ROS also activates signaling pathways that promote the synthesis of matrix metalloproteinases (MMPs)23, 24. MMP25 is a key gene encoding MMPs, which are enzymes responsible for the degradation of extracellular-matrix (ECM) proteins25. ECM proteins are the main connective-tissue component of the dermis26. Degradation of the ECM leads to skin laxity and reduced elasticity, which accelerates skin aging and impairs barrier function. In the ectoine experimental group of this study, significant down-regulation of the expression of the gene NOXO1 can reduce the expression of the NOX1 holoenzyme, thus reducing the production of ROS, while significant down-regulation of the expression of the gene MMP25 can decrease the rate of ECM degradation, thereby mitigating ROS-induced damage to skin cells and delaying skin aging. Unlike Hseu et al., who focused on the induction of antioxidant proteins, we identify NOXO1 as a critical target gene of ectoine, demonstrating that its suppression disrupts the assembly of the NOX1 holoenzyme—an effect that directly inhibits ROS generation at an upstream stage rather than scavenging it downstream. While Cheng et al. reported that ectoine up-regulates COL1A1 and COL1A2 through the PI3K/AKT pathway, we uncover an additional mechanism by which ectoine down-regulates MMP25 expression, thereby slowing ECM degradation—a previously unrecognized mechanism that contributes to the preservation of skin structural integrity under oxidative stress.

Research has demonstrated that ectoine maintains the integrity of the corneal barrier by inhibiting the pro-inflammatory response and promoting the expression of cytokine IL-3727. ANGPTL4 belongs to the family of ANGPTL proteins, which are involved in vasculature generation, inflammation, oxidative stress, vascular permeability and wound healing28, 29, 30. In a psoriasis study, the gene ANGPTL4 was found to promote inflammatory responses (TNF-α, IL-1β, IL-6 and IL-17A) in keratinocytes by regulating ERK1/2- and STAT3-dependent signaling pathways31. Conversely, silencing ANGPTL4 inhibits these effects. Forkhead box protein O6 (FoxO6) is a member of the FoxO family of nuclear transcription factors, which play a crucial role in regulating a diverse array of cellular processes. These processes encompass key functions in cell survival, proliferation, apoptosis, metabolic homeostasis and aging regulation32, and their target genes include TRAIL (TNF-related apoptosis-inducing ligand), FasL (Fas ligand), Bim (Bcl-2-interacting cell death mediator), pro-apoptotic Bcl-2 family members and Bcl-633, 34. Importantly, FoxO6 phosphorylation at Ser184, mediated by Akt, has been shown to promote oxidative stress and inflammation through NF-κB activation in keratinocytes35. This finding is consistent with our observation that ectoine exerts anti-apoptotic and antioxidant effects by down-regulating FoxO6. Notably, FoxO6 knockdown has been reported to protect ARPE-19 cells from high-glucose-induced ROS accumulation36 and cardiomyocytes from hypoxia-induced damage37, further supporting its broad regulatory role in cellular stress responses. In this study, the expression of ANGPTL4 was significantly down-regulated in the ectoine experimental group, suggesting that ectoine inhibits the inflammatory response and thus maintains corneal-barrier integrity. Additionally, the expression of FoxO6 was significantly down-regulated, indicating that ectoine reduces the rate of apoptosis by modulating apoptosis-related transcription factors and by reducing oxidative stress and inflammation.

In this study, 0.2 mg ml⁻¹ ectoine significantly reduced the apoptosis rate and ROS levels in HaCaT cells, and, combined with the transcriptomics results, we speculate that ectoine may protect cells and delay aging by reducing the expression of genes associated with ROS production and apoptosis.

The epidermis is the outermost protective barrier of the human body, and this important barrier function is largely attributed to the differentiation and maturation of keratinocytes and their intercellular-adhesion properties38. Ectoine can enhance the mobility of the HaCaT-cell membrane by increasing hydrophilicity and molecular spacing, which in turn enhances the repair mechanism of the membrane and stabilizes the skin barrier39, 40. Cornifelin (CNFN) has been identified as a protein component of epidermal corneocytes41. Liu et al.42 found that CNFN deficiency contributes to cyclic alopecia and impairs the skin’s functional barrier in Zdhhc13skc4 mice. Wagner et al.43 found that CNFN deficiency results in defects in keratinocyte desmosome formation, reduced intercellular adhesion and increased susceptibility to damage in the epidermal epithelial layer. In this study, transcription-analysis results indicated that CNFN expression was significantly up-regulated in the experimental group, suggesting that ectoine may increase cell adhesion and connectivity by up-regulating CNFN, thereby enhancing the mechanical stability of cells. This represents a novel finding in elucidating ectoine’s skin-repair mechanism.

While this study focused on ectoine’s cytoprotective effects in HaCaT cells, we acknowledge the limitation of using a single cell type and propose future research with protein-level validation of key DEGs, multi-cell systems, additional markers of oxidative stress and tissue models to fully characterize its potential in skin-barrier protection, anti-apoptosis, antioxidant activity and other therapeutic applications.