Role of PI3K-AKT Pathway in Ultraviolet Ray and Hydrogen Peroxide-Induced Oxidative Damage and Its Repair by Grain Ferments

UV and external environmental stimuli can cause oxidative damage to skin cells. However, the molecular mechanisms involved in cell damage have not been systematically and clearly elucidated. In our study, an RNA-seq technique was used to determine the differentially expressed genes (DEGs) of the UVA/H2O2-induced model. Gene Oncology (GO) clustering and the Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway analysis were performed to determine the core DEGs and key signaling pathway. The PI3K-AKT signaling pathway was selected as playing a part in the oxidative process and was verified by reverse transcription-quantitative polymerase chain reaction (RT-qPCR). We selected three kinds of Schizophyllum commune fermented actives to evaluate whether the PI3K-AKT signaling pathway also plays a role in the resistance of active substances to oxidative damage. Results indicated that DEGs were mainly enriched in five categories: external stimulus response, oxidative stress, immunity, inflammation, and skin barrier regulation. S. commune-grain ferments can effectively reduce cellular oxidative damage through the PI3K-AKT pathway at both the cellular and molecular levels. Some typical mRNAs (COL1A1, COL1A2, COL4A5, FN1, IGF2, NR4A1, and PIK3R1) were detected, and the results obtained were consistent with those of RNA-seq. These results may give us a common set of standards or criteria for the screen of anti-oxidative actives in the future.


Introduction
Oxidative damage to the skin can lead to skin aging, which has always been a focus of attention. Excessive reactive oxygen species (ROS) are considered to be the main factor that induces cell damage and apoptosis [1]. A large amount of research has been conducted to identify compounds that can protect against oxidative stress damage. These compounds can be derived from plants, animals, and microorganisms, as well as chemical compounds [2][3][4][5]. Cell model-based screening and mechanism exploration are common research approaches. In these studies, a large proportion of studies are based on UVA induction [6][7][8] and hydrogen peroxide induction [2,6,9,10].
As for research regarding dermatology, cosmetics, and functional foods, human skin fibroblast cells (HSFs) oxidative stress damaged models that are stimulated by either UVA or H 2 O 2 are widely used [6]. Scholars have some common knowledge both of these models, including that they can break the balance between oxidants and antioxidants and cause oxidative stress damage by increasing the oxides content [6][7][8][9][10]. There may be differences in the internal mechanism of these two models in terms of how they induce free radicals. In spite of this, we still generally believe that the effective ingredients obtained by screening

Screening for Differential Expressed Genes (DEGs) and Functional Analysis
Differential expression analysis was performed between control group and model group to identify the DEGs and their functions.
The DEGs between the control and UVA-induced model were screened using |log 2 FC| > 1.2 and P-adjust < 0.05 as the thresholds. The DEGs in the H 2 O 2 -induced oxidative damage study were screened using the criteria, |log 2 FC| > 2 and P-adjust < 0.05. The DEGs were functionally annotated using Blast2go (Version 2.5) and goatools (Version 0.6.5) for Gene Ontology (GO) annotation and clustering analysis. The clustered GO terms involving biological process (BP), molecular function (MF), and cellular component (CC) were obtained based on the criteria of the enrichment score of > 2. The significantly enriched pathways were identified using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database with p < 0.05 as the criterion. The website Database for Annotation, Visualization and Integrated Discovery (DAVID ver 6.8, http://david.ncifcrf.gov/, accessed on 1 February 2023) was also used.
Total RNA extraction and cDNA synthesis were performed using the TriQuick Reagent (TRIzol Substitute) (Beijing Solarbio Science & Technology Co., Ltd.) and the UEIris II RT-PCR System for the cDNA Synthesis SuperMix (Beijing TransGen Biotech Co., Ltd.) according to the manufacturers' instructions. All synthesized cDNA were used for the PCR amplification using the TransStart ® Top Green qPCR SuperMix kit (Beijing TransGen Biotech Co., Ltd.). The total reaction system was 20 µL. The primer sequences used were presented in Table 2 [6]. GAPDH was used as a normalization control. The cyclic parameters used were pre-denaturation at 94 • C for 30 s, followed by the PCR reaction (45 cycles of 94 • C for 15 s, 60 • C for 15 s, and 72 • C for 10 s), while fluorescence data were collected at 72 • C. The relative gene expression levels were analyzed using the 2 −∆∆CT method [31].

Statistics
All experiments were performed in triplicate at least, and the data were expressed as mean ± standard deviation (SD). The data were analyzed using SPSS 17.0 (SPSS, Armonk, New York, NY, USA) and GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA). One-way analysis of variance was used for the significance of differences between the groups (#, * p < 0.05; ##, ** p < 0.01).

The Establishment of UVA/H 2 O 2 -Induced Models and Identification of DEGs
Special criteria were used to screen for the differences between the DEGs in the model and control in each study. For the UVA-induced oxidative damage study, we collected and filtered out a total of 552 DEGs based on the criteria of |log 2 FC| > 1.2 and P-adjust < 0.05, of which 249 genes were upregulated and 304 genes were downregulated. As shown in Table 3, among the upregulated DEGs, 144 genes were involved in cell proliferation and regulation, with 22 genes encoding proteins that can produce a response to an external stimulus. A total of 15 genes participated in immune and inflammatory response, 9 genes were involved in skin barrier-related responses. In addition, 7 of the genes that encoded proteins were also found to play a role in the oxidative stress response, while the functions of the other 36 genes were unknown. Among the downregulated DEGs, 166 genes were involved in cell metabolism, proliferation, and regulation, while 24 genes participated in intercellular signaling. A total of 16 genes encoded proteins that could play a role in the oxidative stress response, and 11 genes were involved in inflammatory processes. Six genes were involved in the synthesis of the cytoskeleton. Apart from these genes, the functions of eight DEGs were uncharacterized.                 Notes: FC, fold change. The differentially expressed genes (DEGs) between the control and UVA-induced model were screened using |log 2 FC| > 1.2 and P-adjust < 0.05 as the thresholds. The DEGs were functionally annotated using Blast2go (Version 2.5) and goatools (Version 0.6.5) for gene ontology (GO) annotation and clustering analysis. UVA/H 2 O 2 can induce ROS generation, leading to a decrease in cell viability, which lead to a decline in cellular fitness and ultimately, to cell death. We determined the effects of the inducers (UVA/H 2 O 2 ) on cell viabilities ( Figure 1A,B). Dose dependency was observed in both studies. IC 50 parameters were chosen to establish damage models. UVA irradiation at a dose of 15 J/cm 2 was finally selected to be used to establish the UVA-induced oxidative stress damage model for subsequent studies. H 2 O 2 at a concentration of 1000 µmol/L H 2 O 2 for 30 min was used as the modeling condition of the H 2 O 2 -induced oxidative stress damage model (Figure 1). of the inducers (UVA/H2O2) on cell viabilities ( Figure 1A,B). Dose dependency was observed in both studies. IC50 parameters were chosen to establish damage models. UVA irradiation at a dose of 15 J/cm 2 was finally selected to be used to establish the UVA-induced oxidative stress damage model for subsequent studies. H2O2 at a concentration of 1000 µ mol/L H2O2 for 30 min was used as the modeling condition of the H2O2-induced oxidative stress damage model ( Figure 1). Likewise, as shown in Table 4, we compared the H2O2-injured Model with the Control, and a total of 607 DEGs were screened based on the criteria, |log2FC| > 2 and Padjusted < 0.05. Among them, 226 DEGs were upregulated and 381 DEGs were downregulated. Among the upregulated DEGs, 19 genes encoded proteins that regulated the oxidative stress response, 10 genes regulated interactions between cells and the surrounding environment, and 6 genes were involved in skin barrier function and lipid anabolism. In addition, there were 42 genes whose functions were unknown. The downregulated DEGs were significantly involved in pathways, cellular matrix composition entry and complement coagulation cascade. Among them, a total of 139 genes encoded proteins that participated in cell proliferation and regulation, 27 genes were responsible for regulating the immune response process and immune system, 24 genes were involved in the positive regulation process of cell chemotaxis and inflammatory response, while 9 genes were Likewise, as shown in Table 4, we compared the H 2 O 2 -injured Model with the Control, and a total of 607 DEGs were screened based on the criteria, |log 2 FC| > 2 and P-adjusted < 0.05. Among them, 226 DEGs were upregulated and 381 DEGs were downregulated. Among the upregulated DEGs, 19 genes encoded proteins that regulated the oxidative stress response, 10 genes regulated interactions between cells and the surrounding environment, and 6 genes were involved in skin barrier function and lipid anabolism. In addition, there were 42 genes whose functions were unknown. The downregulated DEGs were significantly involved in pathways, cellular matrix composition entry and complement coagulation cascade. Among them, a total of 139 genes encoded proteins that participated in cell proliferation and regulation, 27 genes were responsible for regulating the immune response process and immune system, 24 genes were involved in the positive regulation process of cell chemotaxis and inflammatory response, while 9 genes were responsible for regulating the cell response. In addition, there were 54 genes whose functions were unknown.                Notes: FC, fold change. The differential expressed genes (DEGs) in the H 2 O 2 -induced oxidative damage study were screened using the criteria of |log 2 FC| > 2 and P-adjust < 0.05. The DEGs were functionally annotated using Blast2go (Version 2.5) and goatools (Version 0.6.5) for gene ontology (GO) annotation and clustering analysis.
The above-mentioned sequencing analysis indicated that some similar functions of DEGs are grouped. There were no common genes between UVA-upregulated and H 2 O 2 -upregulated DEGs (Venn diagram shown in Figure S1 in Supplementary material), suggesting that different insight mechanisms exist. DAVID analysis showed that UVAinduced upregulated DEGs mainly focused on the regulation of transcription (enrichment score = 3.47) and protein phosphorylation (enrichment score = 1.35), which both had roles in signal transduction. While the H 2 O 2 -induced upregulated DEGs mainly participated in the processes related to the immune response (enrichment score = 3.11), inflammatory responses (enrichment score = 3.09), cell adhesion (enrichment score = 1.93) and regulation of transcription (Enrichment score = 1.11) (data not shown). While 15 common DEGs existed both in UVA and H 2 O 2 induced downregulated DEGs, 5 existed in the extracellular exosome and participated in cell migration (Data not shown).

Gene Ontology (GO) and KEGG Pathway Analysis
In the UVA induced damage model, the results of the GO enrichment analysis showed that a total of 14 GO terms of the upregulated ( Figure 1a) and downregulated (Figure 1b) DEGs were enriched. Two GO terms were enriched in the upregulated DEGs, namely cyclindependent protein kinase activity and cyclin-dependent protein serine/threonine kinase activity. On the contrary, a total of 12 categories were enriched in the downregulated DEGs. Among them, DNA-binding transcription activator activity (RNA), negative regulation of phosphorylation, and collagen-containing extracellular matrix were the top 3 GO terms with high gene ratios and significant P-adjust values, which indicated that the processes of cell proliferation, cell communication, and extracellular matrix production were inhibited. Furthermore, extracellular matrix components and their regulated genes were decreased, indicating changes in cellular morphology. At the same time, biological signal transduction induced by phosphorylation were slowed down or even inhibited.
In the H 2 O 2 oxidative damage model, there were 10 GO terms enriched in upregulated DEGs ( Figure 1c) and 12 GO terms enriched in downregulated DEGs (Figure 1d). Among these enriched terms of upregulated DEGs, receptor ligand activity, extracellular matrix organization, and response to molecule of bacterial origin, were the top three terms with high gene ratios and significant P-adjusted values. The terms related to extracellular matrix constituents or collagen production were both found in the UVA and H 2 O 2 -induced damage models, indicating the damaged changes in cellular morphology.
KEGG enrichment analysis was performed. As shown in Figure 2, the UVA-induced upregulated DEGs were mostly enriched in virus infection (herpes simplex virus 1 infection), while the downregulated DEGs were significantly enriched in the TNF signaling pathway and IL-17 signaling pathway (Figure 2A,B). It is worth noting that TNF and IL-17 can activate or regulate the PI3K-AKT signaling pathway, indicating that UVA may damage cells through that pathway. A comprehensive KEGG enrichment analysis was also conducted, as shown in Figure 2C. All the DEGs that were both upregulated and downregulated were enriched in pathway terms, such as ECM-receptor interaction and the PI3K-AKT signaling pathway.
The KEGG enrichment analysis results of the H 2 O 2 -induced damage model ( Figure 2D,E) showed that the upregulated DEGs were significantly enriched in cytokine-cytokine receptor interaction and the IL-17 signaling pathway, while the downregulated DEGs were enriched in axon guidance and complement and coagulation cascades. The H 2 O 2 -induced and UVA-induced DEGs were both involved in several common terms regardless of the method of regulation. These terms were ECM-receptor interaction, PI3K-AKT signaling pathway, focal adhesion, and TNF signaling pathway. In addition, the H 2 O 2 -induced DEGs were also involved in the other pathway terms, such as complement and coagulation cascades and cytokine-cytokine receptor interaction, which are associated with immune response ( Figure 2F). The KEGG enrichment analysis results of the H2O2-induced damage model ( Figure  2D,E) showed that the upregulated DEGs were significantly enriched in cytokine-cytokine receptor interaction and the IL-17 signaling pathway, while the downregulated DEGs were enriched in axon guidance and complement and coagulation cascades. The H2O2induced and UVA-induced DEGs were both involved in several common terms regardless of the method of regulation. These terms were ECM-receptor interaction, PI3K-AKT signaling pathway, focal adhesion, and TNF signaling pathway. In addition, the H2O2-induced DEGs were also involved in the other pathway terms, such as complement and coagulation cascades and cytokine-cytokine receptor interaction, which are associated with immune response ( Figure 2F).

Key DEGs Functioned in PI3K-AKT Pathway
According to Figure 2C,F, several common KEGG terms were found to be enriched in both UVA and H2O2 induced cell damage. The pathways involved in the occurrence, development, and response of inflammation were both enriched in these two models. TNF can combine with tumor necrosis factor receptor 1 (TNFR1) to form complex 1, which promotes the ubiquitination of E3, the ubiquitin-connected protein inhibitor of apoptosis (cIAP), which in turn activates the NF-κB signaling pathway, leading to keratinocyte inflammatory response and even keratinocyte inflammatory response. NF-κB inhibition of the keratinocytes causes RIPK1-mediated necroptosis and skin inflammation. In addition, the extracellular matrix (ECM)-related gene expression process has also been identified as an important factor that affects the level of cellular inflammation. Based on gene counts and P-value into account, we found that many DEGs induced by UVA (70) or H2O2 (55)

Key DEGs Functioned in PI3K-AKT Pathway
According to Figure 2C,F, several common KEGG terms were found to be enriched in both UVA and H 2 O 2 induced cell damage. The pathways involved in the occurrence, development, and response of inflammation were both enriched in these two models. TNF can combine with tumor necrosis factor receptor 1 (TNFR1) to form complex 1, which promotes the ubiquitination of E3, the ubiquitin-connected protein inhibitor of apoptosis (cIAP), which in turn activates the NF-κB signaling pathway, leading to keratinocyte inflammatory response and even keratinocyte inflammatory response. NF-κB inhibition of the keratinocytes causes RIPK1-mediated necroptosis and skin inflammation. In addition, the extracellular matrix (ECM)-related gene expression process has also been identified as an important factor that affects the level of cellular inflammation. Based on gene counts and P-value into account, we found that many DEGs induced by UVA (70) or H 2 O 2 (55) were involved in the PI3K-AKT pathway based on a relatively high value of -log10(P), as shown in Figure 2C,F.
Based on the findings presented here, the PI3K-AKT signaling pathway was the focus of our ongoing studies. As shown in Figure 3, the expression changes of the genes involved in the PI3K-AKT signaling pathway induced by UVA ( Figure 3A) and H 2 O 2 ( Figure 3B) are clearly demonstrated.
were involved in the PI3K-AKT pathway based on a relatively high value of -log10(P), as shown in Figure 2C,F.
Based on the findings presented here, the PI3K-AKT signaling pathway was the focus of our ongoing studies. As shown in Figure 3, the expression changes of the genes involved in the PI3K-AKT signaling pathway induced by UVA ( Figure 3A) and H2O2 (Figure 3B) are clearly demonstrated. In the UVA-induced model, most of the DEGs involved in PI3K-AKT pathway were downregulated (60 downregulated DEGs out of 70 DEGs) and mainly participated in three biological processes, such as cell adhesion, regulation of MAPK cascade, and cell proliferation and differentiation (Table S1 in Supplementary Material). The expression of receptor tyrosine (RTK)-associated protein on the membrane was downregulated, which was induced by the decrease of population stimulating factor (GF) in the extracellular matrix. RTK subsequently changed the expression of SOS and Ras, which finally induced the decrease of cell proliferation, angiogenesis, and DNA repair. Furthermore, phosphatidylinositol 3 kinase (PI3K) was also downregulated by Ras. Some cytokines and ECM in the extracellular matrix were also repressed in expression, causing a downregulation of PI3K. As previously stated, PI3K is a hub node that can influence how the expression of downstream proteins affects protein synthesis, glycolysis, apoptosis, and crosstalk with other signaling pathways. For example, the downregulation of CREB (cAMP-response element binding protein), and FOXO (Forkhead box O) in this study may affect cell survival and cell cycle progression. The levels of PKCs also decreased, affecting glucose uptake and vesicle transport. The accelerated DNA damaging-induced transcript (REDD1) and serine threonine kinase (LKB1) also resulted in an unbalancing of protein synthesis. Figure 3B showed the DEGs involved in PI3K-AKT signaling pathway in the H2O2induced oxidative damaged model, which is different from that of the UVA model. A total of 55 DEGs, including 28 upregulated DEGs and 27 downregulated DEGs, were enriched in the PI3K-AKT signaling pathway. Three GO terms were enriched (Table S2 in Supplementary Material). Cell adhesion, for example, showed a high enrichment score of 11.81, which played an important part in the inflammatory response.
In the H2O2-induced model, the downregulation of PI3K was mainly due to the downregulation of toll-like receptor (TLR2/4) and integrin (ITGB) on cell membrane and extracellular matrix (ECM), which ultimately affected cell survival, cell cycle, metabolism, and protein synthesis. Additionally, the reduced expression of Adenosine 5 '-Monophosphate (AMP)-activated protein kinase (AMPK) also influenced the synthesis of proteins. In the UVA-induced model, most of the DEGs involved in PI3K-AKT pathway were downregulated (60 downregulated DEGs out of 70 DEGs) and mainly participated in three biological processes, such as cell adhesion, regulation of MAPK cascade, and cell proliferation and differentiation (Table S1 in Supplementary Material). The expression of receptor tyrosine (RTK)-associated protein on the membrane was downregulated, which was induced by the decrease of population stimulating factor (GF) in the extracellular matrix. RTK subsequently changed the expression of SOS and Ras, which finally induced the decrease of cell proliferation, angiogenesis, and DNA repair. Furthermore, phosphatidylinositol 3 kinase (PI3K) was also downregulated by Ras. Some cytokines and ECM in the extracellular matrix were also repressed in expression, causing a downregulation of PI3K. As previously stated, PI3K is a hub node that can influence how the expression of downstream proteins affects protein synthesis, glycolysis, apoptosis, and crosstalk with other signaling pathways. For example, the downregulation of CREB (cAMP-response element binding protein), and FOXO (Forkhead box O) in this study may affect cell survival and cell cycle progression. The levels of PKCs also decreased, affecting glucose uptake and vesicle transport. The accelerated DNA damaging-induced transcript (REDD1) and serine threonine kinase (LKB1) also resulted in an unbalancing of protein synthesis. Figure 3B showed the DEGs involved in PI3K-AKT signaling pathway in the H 2 O 2 -induced oxidative damaged model, which is different from that of the UVA model. A total of 55 DEGs, including 28 upregulated DEGs and 27 downregulated DEGs, were enriched in the PI3K-AKT signaling pathway. Three GO terms were enriched (Table S2 in Supplementary Material). Cell adhesion, for example, showed a high enrichment score of 11.81, which played an important part in the inflammatory response.
In the H 2 O 2 -induced model, the downregulation of PI3K was mainly due to the downregulation of toll-like receptor (TLR2/4) and integrin (ITGB) on cell membrane and extracellular matrix (ECM), which ultimately affected cell survival, cell cycle, metabolism, and protein synthesis. Additionally, the reduced expression of Adenosine 5 '-Monophosphate (AMP)activated protein kinase (AMPK) also influenced the synthesis of proteins. The expression level of the protein encoded by phosphatase type 2 phosphatase activator (PP2A) decreased, which induced an increase in the CREB and MDM2 prototype oncogene, indirectly. The increased expression of the MDM2 inhibited tumor suppressor p53, which played a positive role in cell survival. Furthermore, the downregulation of PI3K and PP2A also affected cell metabolism, indirectly. We also found that the transmembrane 4 L six family member 1 (TM4SF1), E2F transcription factor 7 (E2F7), and BCL2 related protein A1 (BCL2A1) showed a high level of expression.
Overall, although differences were found between these two oxidative damage models, there were certain common DEGs. A total of 11 DEGs with a similar trend were identified in the pathway: PIK3R1, IGF2, ITGA4, LAMA2, LAMA5, LAMB2, NR4A1, COL1A1, COL1A2, COL4A5, and FN1. Some of them were regulatory hub nodes, while others encoded for the cellular matrix. We finally chose 6 of them in the subsequent verification (COL1A1, COL1A2, COL4A5, FN1, IGF2, and PIK3R1) by qRT-PCR, and found the consistency with the RNA-seq results.

Composition Analysis and Protective Effects of RFB, HBFB and OFB
S. commune is an edible fungus with good nutritive value that is often used in food as a functional additive. Since the functional effects of S. commune-fermented grain additives have been developed and studied in the lab for many years, we selected three kinds of fermented actives in this study to verify whether they were suitable for the criteria obtained from the above-mentioned RNA-seq results.
The physicochemical properties of the RFB, HBFB, OFB are shown in detail in Tables 5  and 6. The carbohydrate content of RFB, HBFB, and OFB was 83.9%, 73.3% and 59.2%, respectively. Among them, total sugar content of RFB was highest, about 43.36%, and the β-glucan content of OFB was the highest (2.74%). The reducing sugar was highest in RFB (3.18%). The cell viabilities to HSFs of RFB ( Figure 4A), HBFB ( Figure 4B), and OFB ( Figure 4C) were examined. RFB at a concentration of 5 mg/mL had a relatively high cell viability with a cell survival rate of 69.55% ± 2.30%. OFB at a concentration of 1.25 mg/mL led to a cell survival rate of 57.83% ± 3.81%. Apart from these results, other concentrations showed more than 80% cell survival rates, indicating almost no toxicity to cells. Furthermore, at a concentration ranging from 0.16 to 0.625 mg/mL, RFB showed the most excellent proliferation promoting effect among these three samples. We finally chose a concentration of 0.625 mg/mL of RFB, 0.625 mg/mL of HBFB, and 0.32 mg/mL OFB to be used in further studies. more than 80% cell survival rates, indicating almost no toxicity to cells. Furthermore, at a concentration ranging from 0.16 to 0.625 mg/mL, RFB showed the most excellent proliferation promoting effect among these three samples. We finally chose a concentration of 0.625 mg/mL of RFB, 0.625 mg/mL of HBFB, and 0.32 mg/mL OFB to be used in further studies. The proliferation effects on the damaged cells caused by UVA and H2O2 were then studied. As shown in Figure 4D-F, all of the samples can proliferate the decreased cell viabilities induced by UVA and H2O2 (p < 0.05), which indicated the excellent proliferation effects of RFB, HBFB and OFB.

Validation of Key DEGs by RT-qPCR
RT-qPCR was performed to validate the expression levels of the selected DEGs. The results obtained are presented in Figure 5. The PI3K-AKT pathway in the cytoplasm plays an important role in protecting cells from oxidative stress. The decreased expression levels of COL1A1, COL1A2, COL4A5, FN1, IGF2, and PIK3R1 were detected in the cells induced by UVA/H2O2 (all p < 0.01, Figure 5). RFB could significantly upregulate the expression levels of all genes that were downregulated due to UVA/H2O2 injury (all p < 0.01, compared with each model, Figure 5A). Except for COL1A2, HBFB showed similar effects on the expression of COL1A1, COL4A5, FN1, IGF2, and PIK3R1 (all p < 0.01, compared with the model separately, Figure 5B). OFB could significantly increase the expression of COL1A1, FN1, IGF2, and PIK3R1 (all p < 0.01, compared with the model separately, Figure  5C), but could not promote the expression of COL1A2 in either of the models. Furthermore, the expression of COL4A5 in the H2O2-induced model can be accelerated by OFB but was suppressed in the UVA-induced model. The proliferation effects on the damaged cells caused by UVA and H 2 O 2 were then studied. As shown in Figure 4D-F, all of the samples can proliferate the decreased cell viabilities induced by UVA and H 2 O 2 (p < 0.05), which indicated the excellent proliferation effects of RFB, HBFB and OFB.

Validation of Key DEGs by RT-qPCR
RT-qPCR was performed to validate the expression levels of the selected DEGs. The results obtained are presented in Figure 5. The PI3K-AKT pathway in the cytoplasm plays an important role in protecting cells from oxidative stress. The decreased expression levels of COL1A1, COL1A2, COL4A5, FN1, IGF2, and PIK3R1 were detected in the cells induced by UVA/H 2 O 2 (all p < 0.01, Figure 5). RFB could significantly upregulate the expression levels of all genes that were downregulated due to UVA/H 2 O 2 injury (all p < 0.01, compared with each model, Figure 5A). Except for COL1A2, HBFB showed similar effects on the expression of COL1A1, COL4A5, FN1, IGF2, and PIK3R1 (all p < 0.01, compared with the model separately, Figure 5B). OFB could significantly increase the expression of COL1A1, FN1, IGF2, and PIK3R1 (all p < 0.01, compared with the model separately, Figure 5C), but could not promote the expression of COL1A2 in either of the models. Furthermore, the expression of COL4A5 in the H 2 O 2 -induced model can be accelerated by OFB but was suppressed in the UVA-induced model.

Discussion
Oxidative stress is a disorder between oxidative molecules and insufficient defense of antioxidants, resulting in tissue damage and systemic injury. Oxidative damage leads to elevated levels of ROS in cells. ROS are an oxygen-containing small species, such as singlet oxygen ( 1 O 2 ), ozone (O 3 ), hydroxy radical (OH•), hydrogen peroxide (H 2 O 2 ), and superoxide anion radical (O 2 •−) [6]. Melatonin acts as an antioxidant by scavenging free radicals [32]. There are close relationships between oxidation and skin aging. External agents, such as ultraviolet light and environmental pollution, together with endogenous metabolic irregularities, affect skin cell oxidative damages, resulting in skin aging [33].
However, low levels of oxidative stress activate the transcription of genes encoding proteins which are involved in the defense against oxidative damage, oxidative damage repair mechanisms, and apoptosis. On the contrary, a high level of oxidative stress is a severe threat to lots of macromolecules, such as lipid peroxidation, DNA oxidative damage, protein oxidation, and monosaccharide oxidation [34].
Many efforts have been made to fight against skin aging as people's cosmetic desires increase, and many studies have examined the functional actives that can prevent oxidative damage [35,36]. UVA and H 2 O 2 are traditional inducers used to establish oxidative damage cell models in the exploration of the internal mechanism and to screen functional components. It is generally agreed that H 2 O 2 can induce the production of ROS in cells, leading to oxidative damage [37]. Exogenous H 2 O 2 treatment is a simple and feasible cell model for studying the mechanism of oxidative damage and actives screening, which can effectively simulate the process of oxidative damage. Similarly, a UVA-induced photoaging model can penetrate more deeply into the skin, elevate ROS, and thus cause DNA damage [38]. In addition, inflammatory reactions are directly associated with the pathogenesis of photoaging [39]. It has been reported that Rhodiola rosea fermented by lactobacillus plantarum, ectoin, Laminaria japonica fermentation broth, and Lacticaseibacillus paracasei Subsp. paracasei SS-01 strain exopolysaccharide all have potential roles in antioxidative damage effects [2,6,33,40], no matter which oxidative damage models are used in the studies. However, limited studies have been performed to compare different oxidative stress models in phenotypes and mechanisms.
In the study, RNA-seq technology was performed to examine the similarities and differences of the two models in the function of the involved DEGs and pathways. DEGs of UVA-induced HSFs were mainly focused on the biological processes, such as the regulation of cell proliferation, external stimulus response, and cellular inflammatory response. The negative regulation of DNA-binding transcriptional activity and phosphorylation were inhibited as shown by the GO enrichment analysis. PI3K-AKT signaling pathway was selected as one of the important terms with a high number of genes involved and a significant P-value. In the other model of this study, which was H 2 O 2 -induced, processes such as receptor ligand activity, extracellular matrix organization, and response to lipopolysaccharide were significantly upregulated, and collagen-containing extracellular matrix and extracellular matrix structural components were significantly downregulated. Although differences were obtained by the analysis of GO annotation and clustering, the same pathway (PI3K-AKT signaling pathway) obviously changed. Furthermore, the IL-17 signaling pathway was significantly influenced but with a different tendency.
The results obtained partly verified the previous conclusions. Alafiatayo et al. found that UV-irradiated HDF cells showed a lower level of proliferation, more morphological changes, and lower activity compared with the control group. Additionally, the MAPK pathway was found to have been accelerated as antioxidant-related genes were downregulated [41]. In our study, some DEGs in the UVA-induced damage model participated in the PI3K-AKT signaling pathway, the downregulated DEGs of which were mainly involved in cell adhesion, MAPK cascade regulation, and cell proliferation and differentiation. Another transcriptome deep sequencing analysis study conducted by Zheng et al. reported that the differential expression genes altered by UVA were involved in a variety of biological processes, such as the regulation of elastin and phosphoglucomutase encoding levels, and cellular inflammation. Our study found that the functional annotation of the DEGs in UVA-induced HSFs appeared in several terms, such as the regulation of cell proliferation, DNA-binding transcriptional activity, response to external stimuli, and cellular inflammation, which showed consistencies with Zheng et al. [42]. In addition, the expression of ECM and protein complexes in the extracellular matrix and the PI3K-related DEGs decreased simultaneously. As a pivotal node of this signaling pathway, the downregulation of PI3K can interfere with cell survival, apoptosis, DNA damage induction, and downstream protein synthesis, even as it affects cell cycle progression.
Many research studies have examined H 2 O 2 -induced damage to cells. In a previous report published by Barandalla et al., the overall gene expression profile of H 2 O 2 -induced HUES3 cells were analyzed using an Illumina HT-12 V4 chip, and it was found that the most affected genes were associated with RNA processing and splicing, redox reduction, and sterol metabolic processes [43]. Studies have shown that IL-17 may inhibit apoptosis through the PI3K-AKT signaling pathway, and that TNF could activate the PI3K-AKT signaling pathway [44,45]. Our study also reached a similar conclusion that, when cells were stimulated by H 2 O 2 , the upregulated DEGs were significantly enriched in the cytokine-cytokine receptor interaction and IL-17 signaling pathway, which in turn activated the PI3K signaling pathway. We also performed an enrichment analysis of the PI3K signaling pathway genes involved in the H 2 O 2 injury model. The DEGs were mainly enriched in cell adhesion, which was found to play an important role in inflammatory responses. The downregulation of the PI3K was mainly due to the downregulation of TLR2/4 and ECM.
Zou et al. systematically mapped single cell transcriptomes involved in human skin aging and revealed that the matriculated expression of transcription factors drove human skin aging. They found that the downregulation of KLF6 and HES1 were the driving force involved in skin aging [46]. In our study, the sequencing results of the UVA-induced model showed that KLF6 was significantly downregulated, too. Therefore, it can be suggested that UVA radiation can lead to skin aging by the downregulation of KLF6, and that the activation of KLF6 may be a major anti-aging approach. HES1 factor was also identified in the sequencing results of H 2 O 2 -induced model but was differentially upregulated. Therefore, it can be suggested that, upon stimulation by H 2 O 2, HSFs may activate a self-protection mechanism that activates HES1 factor to prevent cell senescence, which then plays a role in resisting changes caused by external stimuli.
In a previously published report, the underlying signaling pathway and biological processes involved in UVB-induced skin injury were identified [47]. The IL-6 gene of the interleukin family was also identified in our results. ECM is an integral component of all organs and plays a pivotal role in tissue homeostasis and repair [48]. CCN1 is a secreted ECM protein, and research has shown that blocking CCN1 function in vivo can effectively alleviate epidermal hyperplasia and inflammation in psoriasis-like mice [49,50]. In the study, the decline in ECM was a common factor for the activation of PI3K. It indicated that both light and oxidative damage could damage the secretion, synthesis, and normal operation of the cell matrix, and that the activation of the PI3K pathway was an effective regulatory pathway to restore such damage.
Our study showed that both the UVA and H 2 O 2 injury models were able to regulate the expression levels of the PI3K-AKT signaling pathway genes. The two lesions together affected 11 genes, including PIK3R1, IGF2, ITGA4, LAMA2, LAMA5, LAMB2, NR4A1, COL1A1, COL1A2, COL4A5, and FN1, with similar expression trends. The abovementioned genes were partly responsible for regulating the central node and partly for encoding the cytoplasmic matrix.
However, the regulation of PI3K-AKT pathway in the UVA and H 2 O 2 injury models was not completely consistent. In the DEGs analysis results of the UVA injury model, the main inducing factors leading to the downregulation of PI3K signaling were the decrease in the expression of cytokines, protein complexes, and population-stimulating factors in the extracellular matrix. The chain reaction caused by the decrease in PI3K signaling led to the responses in the extracellular matrix, manifested as increased transcriptional activity of DNA damage-inducing transcript (REDD1) and serine threonine kinase (LKB1), which affected normal protein synthesis and hindered cell cycle progression and survival. Although ECM factors were involved in the regulation of PI3K in the H 2 O 2 injury model, the regulation was mostly proceeded by downregulating TLR2/4 and ITGB genes. Therefore, we can infer that the main action pathway of skin aging damage caused by light and oxidation are concentrated in the PI3K-AKT signaling pathway.
Both of the damaged models suppressed PI3K gene expression and decreased the content of ECM components. H 2 O 2 was not limited to the regulation of ECM; membrane receptors and integrins were also involved. The results indicated that H 2 O 2 -induced skin oxidative aging was stronger and more complicated than light damage.
S. commune is a fungus composed of mycelium and fruiting bodies, which is rich in organic acids and polysaccharides, such as chitosan, glucosamine, and lignocellulolytic enzymes [51,52]. Studies have shown that Schizophyllum exerts anticancer, antitumor, and immunomodulatory activities [27,53,54]. Existing studies have shown that Schizophyllan can effectively restore the brain function of aging mice, increase the SOD content in serum and brain, and reduce the accumulation of malondialdehyde (MDA), thus leading to significant anti-aging activity [55]. S. commune fermented broths were also reported to have potential benefits for humans [16,26].
Rice, highland barley, and oats are some of the main food crops used worldwide and are rich in carbohydrates, such as starch, cellulose, and protein. Among them, highland barley and oats contain unique substances, such as arabinoxylan and oat polyphenols, which have been widely applied in food and pharmaceutical industries [56][57][58][59][60]. In the study, three kinds of S. commune fermented broths (RFB, HBFB, and OFB) were obtained. Previous studies have demonstrated their anti-oxidative effects in our lab (Some data shown in Figure S2 in Supplementary material). Here, RFB, HBFB, and OFB were used to verify whether the genes screened out could be accelerated by the broths, and the results were confirmed as expected. It can be concluded that RFB, HBFB, and OFB might play roles in the synthesis of insulin growth factor, cell proliferation, and DNA repair. They may also affect cell survival by promoting the formation of anti-apoptotic protein (Bcl-XL), as well as the upregulation of collagen and fibronectin, thus protecting human skin fibroblasts from UVA or H 2 O 2 oxidative damage ( Figure 6). RNA-seq analysis and verification results of our study showed that the expression of key genes involved in the PI3K-AKT signaling pathway could be the screening standard to understand the potential antioxidant actives.
There are still some limitations. Although three replicates were performed when we established these two types of models, more parallel tests are needed to obtain results RNA-seq analysis and verification results of our study showed that the expression of key genes involved in the PI3K-AKT signaling pathway could be the screening standard to understand the potential antioxidant actives.
There are still some limitations. Although three replicates were performed when we established these two types of models, more parallel tests are needed to obtain results with good stability and repeatability. The criteria used in RNA-seq analysis are empirical, which have great relationships with data analysis. We hope to determine a common screening method to identify active ingredients that can prevent oxidative damage, and plenty of standard antioxidants should be used to verify the results. Further studies should be performed to validate the results in this study.

Conclusions
In this study, oxidative stress injury models of human skin fibroblasts were established using H 2 O 2 and UVA stimulation. RNA-seq technology was used to conduct GO and KEGG functional enrichment analysis of the DEGs in the two oxidative injury models. The PI3K-AKT signaling pathway was selected as the common criteria of the two kinds of models. S. commune-fermented RFB, HBFB, and OFB were used for further mRNA validation. Results showed that these three broths can protect against oxidative damage through this pathway, indicating that it might provide a research direction for future studies on the oxidative damage mechanisms and the screening of functional actives.  Figure S2. Determination of antioxidant capacities of S. commune polysaccharide (SP) and VC in vitro. A, DPPH• scavenging effects; B, •OH free radicals scavenging effects; C, Fe 2+ chelating abilities. VC, namely ascorbic acid, was chosen as the positive control. Results were expressed as the mean ±SD (n = 3); Table S1 UVA-induced model; Table S2 H

Data Availability Statement:
The data are available from the corresponding author upon reasonable request.