研究背景

皮肤是人体最大的器官，通过将深层组织与外部环境隔离，为深层组织提供屏障，同时通过物质交换和接受刺激，确保与外部环境接触。皮肤可以抵御物理伤害（如阳光、电离辐射、红外辐射、机械应力和热损伤）、化学损伤（如腐蚀性、刺激性和致敏性物质）和生物损伤(如致病微生物)，通过表皮屏障和汗腺分泌维持水、电解质平衡，通过热感受器调节体温稳定、通过皮肤相关淋巴组织发生免疫反应，同时含有丰富的神经末梢、Pacinian小体、Meissner小体、Ruffini小体、Krause端球和Merkel细胞等，是一个重要的感觉器官。此外，它还负责物质的选择性吸收、存储、消除，参与代谢和维持体内稳态^[1]。

皮肤老化是一个自然、复杂的过程，受两种机制的影响：由时间流逝引起的内在(遗传、时间)衰老和由环境因素(包括紫外线辐射、环境污染和吸烟)引起的外在衰老(光老化)。两种因素互相叠加，且都与皮肤中活性氧和氧化应激的增加密切相关，造成表皮屏障功能紊乱、真皮-表皮交界处变平、成纤维细胞数量和活性减少、异常弹性蛋白纤维的累积以及朗格汉斯细胞功能受损等，最终表现为皮肤皱纹、弹性丧失、颜色变化、色素沉着和变色不均匀、干燥、异常表皮角化灶、毛细血管扩张、易受刺激、皮肤再生和愈合较慢等^[2]。其中，眶周皮肤由于受到日积月累的光损伤和慢性机械应力通常是最先出现明显衰老迹象的面部区域^[3]。

了解皮肤衰老的复杂机制一直是许多研究人员不断追求的目标。随着生物信息学领域的发展，已经发展出各种分析方法，如基因组学、转录组学、蛋白组学、代谢组学等。但是单一数据类型的方法对复杂机制的理解有限，结合多种数据类型可以弥补任何单一数据类型中缺失或不可靠的信息^[4]。在过去十年中，多组学方法的出现使处理老龄生物学的全部复杂分子问题成为可能，使成千上万的表观遗传标记物、转录本、蛋白质和代谢物得以定性和定量，并开始揭示复杂有机体如何在分子水平上随着年龄的增长而发生变化^{[5, 6]}。

研究目的

观察不同年龄中国女性患者眼睑皮肤的组织学差异、检测其差异表达的转录本、蛋白质及代谢物，并进行关联分析研究，探讨它们参与的生物学反应过程和相关信号通路，为皮肤抗衰的深入研究提供依据。

研究方法

第一部分：对收集到17-63岁共44例行常规重睑手术的健康女性患者眼睑皮肤，分为5个年龄组： A组：17-19岁，4例，平均年龄（18.25±0.96岁）；B组：20-29岁，13例，平均年龄（24.60±2.84）岁；C组：30-39岁，13例，平均年龄（34.00±2.40）岁；D组：40-49岁，12例，平均年龄（43.60±3.63）岁；E组：50-63岁，11例，平均年龄（54.90±4.68）岁。制作组织切片，并进行苏木精-伊红、Masson染色及羟脯氨酸测定；对冻存样本进行RNA 测序（RNA sequencing，RNA-seq），筛选差异表达基因，根据GO富集、KEGG富集、PPI互作网络分析表达差异基因参与的生物学反应和信号通路，并对关键候选转录本进行验证实验；

第二部分：对收集到的44例样本进行LC-MS蛋白组学测序，筛选差异表达蛋白，根据GO富集、KEGG富集、PPI互作网络分析表达差异蛋白参与的生物学反应和信号通路，并对关键候选蛋白进行验证实验；

第三部分：在上述44例样本基础上增加9例样本，共53例，分为5组：A组：17-19岁，4例，平均年龄（18.25±0.96岁）；B组：20-29岁，13例，平均年龄（24.90±2.66）岁；C组：30-39岁，13例，平均年龄（34.30±2.66）岁；D组：40-49岁，12例，平均年龄（43.69±3.46）岁；E组：50-63岁，11例，平均年龄（55.00±4.45）岁，对收集到的样本进行非靶向代谢组学测序，筛选差异表达代谢物，根据KEGG富集通路分析表达差异代谢物参与的信号通路，并对关键代谢物及相关基因进行验证实验；

第四部分：对前述组学研究数据中差异表达分子和富集通路进行GSEA等关联分析。

研究结果

1.      组织学研究结果显示，随着年龄增长，眼睑皮肤表皮变薄、真皮乳头层数量减少、真皮-表皮连接处变平，真皮胶原含量减少，胶原纤维变细、卷曲，总胶原蛋白含量较少。

2.      转录组学分析鉴定到829个随年龄表达上调的基因和906个随年龄下调的基因，注意到LINC01235在上调基因中相关性最强，该基因通过EMT过程在癌症转移中发挥作用；在下调基因中，以PRX相关性最强，其与GAP43、DRP2、MPZ、AKR7A2共同参与或维持神经功能，提示皮肤衰老可能与神经功能退行性变有关。表达上调基因GO及KEGG功能富集主要与RNA剪接、核糖体生物合成、细胞周期、I型单纯疱病毒感染和泛素介导的蛋白质水解等作用途径有关，PPI互作分析显示泛素介导的蛋白水解与细胞周期具有一定相关性，其中UBE2C与调控细胞周期的多个基因相关。表达下调基因GO及KEGG功能富集主要与肌动蛋白细胞骨架、胰岛素信号通路、甲状腺激素信号通路、扩张性心肌病、钙离子信号通路等有关。PPI互作网络显示下调途径中各途径间相关关系较为密切。对相关基因的qRT-PCR实验验证发现LINC01235的上调趋势与转录组结果一致，与之相关的EMT过程关键转录因子编码基因SNAI2上调，初步显示了EMT过程与皮肤衰老的相关性，并提示了LINC01235作为潜在衰老标志物的可能性。

3.      蛋白组学分析鉴定到1743个随年龄表达上调的蛋白和512个随年龄下调的蛋白。上调蛋白GO及KEGG功能主要富集在内吞作用、剪接体、百日咳、溶酶体、致病性大肠杆菌感染等作用途径中，以内吞作用最为显著，各通路间相互作用关系密切；下调蛋白GO及KEGG功能主要富集在氧化磷酸化、热产生、氧化-还原过程及肌肉收缩等相关通路，反映了能量代谢紊乱、线粒体功能受损、细胞骨架功能受损与皮肤衰老有关。对I型胶原蛋白、III型胶原蛋白进行免疫组化染色，并对其标志性编码基因COL1A1、COL3A1进行qRT-PCR实验验证，发现I型、III型胶原蛋白不同程度下降、I型/III型胶原含量比值上调，但在转录本层面趋势不同，可能与转录后调控、mRNA降解、蛋白质降解等有关，提示胶原含量变化受到多种因素叠加影响，总体上胶原降解速率大于合成速率，造成胶原含量减少。对下调蛋白COQ6进行qRT-PCR实验验证，结果显示COQ6表达下调，表明氧化-还原平衡改变在衰老皮肤中的发生。

4.      代谢组学分析检出716个代谢物，鉴定到127个随年龄表达上调的代谢物和10个随年龄下调的代谢物，注意到上调最显著的代谢物NAG是尿素循环关键酶，对尿素循环相关的代谢物进行表达丰度分析、并对关键酶基因表达进行qRT-PCR验证，结果显示尿素循环相关代谢物的累积、关键酶GLS1基因表达上调，提示衰老皮肤细胞中能量代谢发生适应性改变，细胞内尿素循环增加并形成尿素的累积。上调代谢物KEGG通路主要富集在卟啉代谢、2-氧代羧酸代谢、精氨酸合成、癌症中的中心碳代谢、蛋白质的消化和吸收等通路；表达下调的代谢KEGG富集在化学致癌物-活性氧、β-丙氨酸代谢、组氨酸代谢、胆汁分泌、嘌呤代谢、代谢途径等通路。

5.      对转录组和蛋白组学数据关联分析，在鉴定与定量层面关联到6554个转录本和蛋白质，在显著差异层面关联到270个差异表达的转录本和蛋白质，一致表达上调的蛋白或转录本主要富集在PPAR信号通路、过氧化物酶体、脂肪酸代谢途径、类固醇生物合成、缬氨酸、亮氨酸、异亮氨酸降解等通路，提示脂肪酸代谢和支链氨基酸代谢上调，表明能量代谢过程改变；一致表达下调的转录本或蛋白富集在肌动蛋白收缩或钙离子信号通路，可能说明在细胞骨架结构完整性和功能受损在衰老皮肤中的发生。

6.      对转录组、蛋白组、代谢组学数据进行关联分析，发现在30岁和45岁左右观察到年龄特异性变化。功能富集聚类热图显示，剪接体、细胞周期、核糖体的生物合成通路在转录组和蛋白组中均上调；甲状腺激素信号通路、内吞作用、胰岛素信号通路在转录组中下调，蛋白组中上调；热产生、氧化磷酸化、帕金森病、非酒精性肝硬化、糖尿病性心肌病、活性氧途径在蛋白组中同时存在上调、下调过程，以下调显著；癌症中心碳代谢在转录组和蛋白组中下调，在代谢组中上调。

研究结论

1.     随着年龄增加，皮肤表皮变薄、真皮-表皮连接处变平、真皮乳头数量减少、总胶原蛋白含量下降，I型、III型胶原蛋白含量不同程度下降、I型/III型胶原含量比值上调；且胶原表达在转录本和蛋白层面趋势不同，提示其变化受到多种因素叠加影响，总体上胶原降解速率大于合成速率，造成胶原含量减少。

2.     在转录本层面，选择性剪接与皮肤衰老显著相关，其受到营养感知信号通路的调控，提示了RNA剪接在皮肤衰老中的起始作用；泛素化介导的蛋白水解与皮肤衰老显著相关，其中UBE2C调控多个细胞周期相关基因，可能起到重要作用。

3.     在蛋白层面，内吞作用显著上调，并通过多种途径促进细胞衰老；氧化磷酸化等通路等下调表明衰老皮肤中线粒体功能受损、能量代谢紊乱；抗氧化物质COQ10的必要组成成分COQ6表达下调，表明氧化-还原平衡改变在衰老皮肤中的发生。

4.     在代谢层面，皮肤能量代谢可能发生适应型改变，如糖酵解水平下降，谷氨酰胺分解增加以适应能量需求等，其副产物尿素循环增加，并进一步损伤线粒体电子传递链；可诱导氧化应激过程的卟啉合成途径可能在皮肤衰老中发挥重要作用，随着衰老皮肤清除ROS能力减弱，抗氧化物质在转录本、蛋白质、代谢物层面均发现显著下调，发生氧化-还原失调；

5.     初步显示了LINC01235、COQ6等作为潜在衰老标志物和治疗靶点的提示意义。

6.     转录、蛋白及代谢组学关联分析表明线粒体功能紊乱、能量代谢失调、细胞骨架完整性受损等是眼睑皮肤衰老的关键因素。

7.     中国女性眼睑皮肤在30岁和45岁左右会发生剧烈的生理变化，表明眼睑皮肤在皮肤衰老，甚至是人体衰老过程中的优先进程，强调了眼睑抗衰的重要性。

Background

The skin, as the largest organ of the human body, provides a barrier that isolates the deeper tissues from the external environment while also facilitating substance exchange and reception of stimuli, ensuring the deeper tissues contact with the outside world. It offers protection against physical damage like sunlight, ionizing radiation, infrared radiation, mechanical stress, and thermal injury, as well as chemical damage from corrosive, irritating, and allergenic substances, and biological harm from pathogens. It maintains water and electrolyte balance through the epidermal barrier and sweat gland secretion, regulates body temperature stability through thermoreceptors, triggers immune responses through skin-associated lymphoid tissue, and is rich in nerve endings and sensory structures such as Pacinian corpuscles, Meissner's corpuscles, Ruffini endings, Krause end bulbs, and Merkel cells, making it a critical sensory organ. Additionally, the skin is responsible for the selective absorption, storage, and elimination of substances, and plays a role in metabolism and maintaining homeostasis within the body^[1].

Skin aging is a natural, complex process influenced by two mechanisms: intrinsic aging caused by the passage of time (genetics, time) and extrinsic aging caused by environmental factors (including UV radiation, environmental pollution, and smoking)，known as photoaging. These factors overlap and are closely associated with an increase in reactive oxygen species and oxidative stress in the skin, leading to epidermal barrier dysfunction, flattening of the dermo-epidermal junction, a decrease in number and activity of fibroblasts, accumulation of abnormal elastin fibers, and impairment of Langerhans cell function. Ultimately, this manifests as skin wrinkles, loss of elasticity, color changes, uneven pigmentation and discoloration, dryness, abnormal epidermal keratinization, capillary expansion, increased sensitivity, and slower skin regeneration and healing^[2]. The periorbital skin is often the first facial area to show obvious signs of aging due to cumulative light damage and chronic mechanical stress ^[3].

Understanding the complex mechanisms of skin aging has long been a goal that many researchers continuously strive to achieve. With the advancement in the field of bioinformatics, various analytic methods have been developed to understand the variability of complex traits, such as genomics, transcriptomics, proteomics, and metabolomics. However, the understanding of complex mechanisms is limited by single data type approaches; integrating multiple data types can compensate for missing or unreliable information within any single data type^[7]. In the past decade, the emergence of multi-omics approaches has made it possible to handle the entire complex molecular issues of aging biology, allowing for the qualitative and quantitative characterization of thousands of epigenetic markers, transcripts, proteins, and metabolites, and has begun to unveil how complex organisms change at the molecular level with age.

Objective

To observe the histological differences in eyelid skin of Chinese female patients of different ages, detecting differentially expressed transcripts, proteins, and metabolites, and conducting correlation analysis to explore the biological reactions and related signaling pathways they are involved in, providing a basis for in-depth research on skin anti-aging

Methods

Part One: Eyelid skin from 44 healthy female patients aged 17 to 63 undergoing routine blepharoplasty was collected and divided into five age groups: Group A: 17-19 years, 4 cases, average age (18.25±0.96 years); Group B: 20-29 years, 13 cases, average age (24.60±2.84 years); Group C: 30-39 years, 13 cases, average age (34.00±2.40 years); Group D: 40-49 years, 12 cases, average age (43.60±3.63 years); Group E: 50-63 years, 11 cases, average age (54.90±4.68 years). Tissue sections were prepared and stained with Hematoxylin and Eosin, Masson's trichrome, and hydroxyproline content was measured; RNA sequencing (RNA-seq) was performed on the frozen samples to screen for differentially expressed genes, which were analyzed for biological reactions and signaling pathways involvement using GO enrichment, KEGG enrichment, and PPI interaction network analysis，and validation experiments were conducted on key candidate transcripts.

Part Two: LC-MS proteomics sequencing was performed on the collected 44 samples to select differentially expressed proteins, which were analyzed for biological reactions and signaling pathways involvement using GO enrichment, KEGG enrichment, and PPI interaction network analysis，and validation experiments were conducted on key candidate proteins.

Part Three: On the basis of the aforementioned 44 samples, an additional 9 samples were added, totaling 53 cases, divided into five groups: Group A: 17-19 years, 4 cases, average age (18.25±0.96 years); Group B: 20-29 years, 13 cases, average age (24.90±2.66 years); Group C: 30-39 years, 13 cases, average age (34.30±2.66 years); Group D: 40-49 years, 12 cases, average age (43.69±3.46 years); Group E: 50-63 years, 11 cases, average age (55.00±4.45 years). Untargeted metabolomics sequencing was conducted on the collected samples to select differentially expressed metabolites, which were analyzed for signaling pathways involvement using KEGG enrichment pathway analysis，and validation experiments were conducted on key metabolites and related genes.

Part Four: Correlation analysis such as GSEA was conducted on differentially expressed molecules and enriched pathways from the previous omics studies.

Results

1.      Histological studies show that with increasing age, the eyelid skin epidermis becomes thinner, the number of dermal papillae decreases, the dermal-epidermal junction flattens, dermal collagen content decreases, collagen fibers become thinner and curly, and the total collagen content is reduced.

2.     Transcriptomic analysis identified 829 genes upregulated with age and 906 genes downregulated with age. Notably, LINC01235 was the most correlated among the upregulated genes, playing a role in cancer metastasis through the EMT process. Among the downregulated genes, PRX was the most correlated, which, along with GAP43, DRP2, MPZ, and AKR7A2, is involved in maintaining neural function, suggesting that skin aging may be associated with neurodegenerative changes. GO and KEGG functional enrichment of upregulated genes were mainly related to RNA splicing, ribosome biogenesis, cell cycle, herpes simplex virus type I infection, and ubiquitin-mediated proteolysis. PPI interaction analysis indicated a certain correlation between ubiquitin-mediated proteolysis and the cell cycle, with UBE2C associated with several genes regulating the cell cycle. GO and KEGG functional enrichment of downregulated genes were mainly related to the actin cytoskeleton, insulin signaling pathway, thyroid hormone signaling pathway, dilated cardiomyopathy, and calcium signaling pathway. PPI interaction network showed close relationships between pathways in the downregulated category. qRT-PCR experiments on related genes showed that the upregulation trend of LINC01235 was consistent with transcriptome results, with the EMT process key transcription factor encoding gene SNAI2 upregulated.

3.     Proteomic analysis identified 1,743 proteins upregulated with age and 512 proteins downregulated with age. The GO and KEGG functions of upregulated proteins were mainly enriched in endocytosis, spliceosome, pertussis, lysosome, and pathogenic Escherichia coli infection, with endocytosis being the most significant and showing close interactions between pathways. The GO and KEGG functions of downregulated proteins were mainly enriched in oxidative phosphorylation, thermogenesis, redox processes, and muscle contraction, reflecting the association of energy metabolism disorders, mitochondrial dysfunction, and cytoskeletal function impairment with skin aging. Immunohistochemical staining was performed for type I and type III collagen, and qRT-PCR was conducted to verify their marker genes COL1A1 and COL3A1. It was found that type I and type III collagen showed varying degrees of decline, and the ratio of type I/type III collagen content increased, but trends at the transcript level differed. This may be related to post-transcriptional regulation, mRNA degradation, and protein degradation, suggesting that changes in collagen content are influenced by multiple factors, with the overall collagen degradation rate exceeding the synthesis rate, leading to a reduction in collagen content. qRT-PCR validation of the downregulated protein COQ6 showed decreased COQ6 expression, indicating that changes in redox balance occur in aging skin.

4.     A total of 716 metabolites were detected in Metabolomics, 127 metabolites were up-regulated with age and 10 metabolites were down-regulated with age. Notably, the most significantly upregulated metabolite, NAG, is a key enzyme in the urea cycle. Expression abundance analysis of metabolites related to the urea cycle and qRT-PCR validation of key enzyme genes showed an accumulation of urea cycle-related metabolites and upregulation of the key enzyme GLS1 gene. This suggests adaptive changes in energy metabolism in aging skin cells, leading to an increased urea cycle and accumulation of urea. The up-regulated metabolite KEGG pathways were mainly concentrated in porphyrin metabolism, 2-Oxocarboxylic acid metabolism, arginine synthesis, central carbon metabolism in cancer, protein digestion and absorption. Metabolic KEGG with down-regulated expression were enriched in chemical carcinogens-reactive oxygen species, β-alanine metabolism, histidine metabolism, bile secretion, purine metabolism, and other pathways.

5.     Correlation analysis of transcriptomic and proteomic data identified 6554 transcripts and proteins at the identification and quantification level, with 270 differentially expressed transcripts and proteins at the significant difference level. Consistently upregulated proteins or transcripts were mainly enriched in the PPAR signaling pathway, peroxisomes, fatty acid metabolism, steroid biosynthesis, valine, leucine, and isoleucine degradation pathways, indicating an upregulation of fatty acid metabolism and branched-chain amino acid metabolism, suggesting changes in energy metabolism processes. Consistently downregulated transcripts or proteins were enriched in actin contraction or calcium signaling pathways, which may indicate the occurrence of impaired cytoskeletal integrity and function in aging skin.

6.     Correlation analysis of transcriptomic, proteomic, and metabolomic data revealed that stage-specific changes were observed at the transcriptomic, proteomic and metabolomic levels at around 30 and 45 years of age. Functional clustering heat maps showed that the biosynthetic pathways of spliceosome, cell cycle and ribosome were up-regulated in both transcriptome and proteome. Thyroid hormone signaling pathway, endocytosis and insulin signaling pathway are down-regulated in transcriptome and up-regulated in proteome. Thermogenesis, oxidative phosphorylation, Parkinson's disease, non-alcoholic fatty liver diseases, diabetic cardiomyopathy, reactive oxygen species pathway both up-regulated and down-regulated in proteome, and the down-regulated pathway was significant. Cancer center carbon metabolism is down-regulated in the transcriptome and proteome and up-regulated in the metabolome.

Conclusions

1.     As age increases, the skin epidermis becomes thinner, the dermo-epidermal junction flattens, the number of dermal papillae decreases, and the total collagen content decreases. The content of collagen types I and III decreases to varying degrees, while the ratio of collagen type I to type III increases. Collagen expression shows different trends at the transcript and protein levels, indicating that its changes are influenced by multiple factors. Overall, the rate of collagen degradation is greater than the rate of synthesis, leading to a decrease in collagen content.

2.     At the transcript level, alternative splicing is significantly associated with skin aging and is regulated by the nutrient-sensing signaling pathway, suggesting that RNA splicing plays an initiating role in skin aging. Ubiquitination-mediated protein hydrolysis is significantly associated with skin aging, with UBE2C regulating multiple cell cycle-related genes, potentially playing an important role.

3.     At the protein level, endocytosis is significantly upregulated and promotes cell aging through multiple pathways. Downregulation of pathways such as oxidative phosphorylation indicates impaired mitochondrial function and disrupted energy metabolism in aging skin. The decreased expression of COQ6, a necessary component of the antioxidant COQ10, indicates changes in the oxidative-reductive balance in aging skin.

4.     At the metabolic level, skin energy metabolism may undergo adaptive changes, such as a decrease in glycolysis levels and an increase in glutamine decomposition to meet energy demands. The increased urea cycle, a byproduct, further damages the mitochondrial electron transport chain. The heme synthesis pathway, which can induce oxidative stress, may play an important role in skin aging. As the ability of aging skin to clear ROS decreases, antioxidants are significantly downregulated at the transcript, protein, and metabolite levels, leading to oxidative-reductive imbalance.

5.     Preliminary findings suggest that LINC01235, COQ6, and others may serve as potential markers of aging and therapeutic targets.

6.     Transcriptomic, proteomic, and metabolomic correlation analysis indicates that mitochondrial dysfunction, disrupted energy metabolism, and impaired cell cytoskeleton integrity are key factors in eyelid skin aging.

7.     Chinese women experience dramatic physiological changes in eyelid skin around the ages of 30 and 45, indicating that eyelid skin ages preferentially in skin aging and even in the aging process of the whole body, emphasizing the importance of anti-aging care for the eyelids.