Network Pharmacology-based Analysis on Treatment of Diabetic Retinopathy by Berberine

doi:10.21203/rs.3.rs-1529761/v2

Download PDF

Research Article

Network Pharmacology-based Analysis on Treatment of Diabetic Retinopathy by Berberine

https://doi.org/10.21203/rs.3.rs-1529761/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted

You are reading this latest preprint version

The diabetic retinal injury model was established. After treatment with berberine, OCT and water maze were measured, and 10 genes were screened out by 4D-DIA quantitative proteomic analysis. Heat map difference analysis and molecular docking were performed. The key words "diabetic retinal injury" were input in PubMed, and the related genes queried were combined with the results of 10 screened proteome analysis to conduct PPI protein interaction. Then GO and KEGG analyses were performed with the interacting genes. Results :OCT results showed that the outer nucleus, inner layer and outer accessory layer of RPE were thickened after berberine treatment. The results of water maze experiment showed that after berberine treatment, the learning and cognitive function of BBR group rats was significantly higher than that of DM group on day 5. By screening proteome analysis data and making Venny cross map, 10 genes could be obtained and heat map could be made. Dennd1a and Utp6 were down-regulated in diabetes group. Atp7a, Ppl, Ogn, Nsmcel, Mme, Lmo4, Ca1 and Fn1 were upregulation, and the berberine treatment group was opposite to the diabetic group. Molecular docking results showed that berberine and berberine could form stable molecular binding patterns except Fnl, Ogn and Utp6 target proteins. 343 genes were searched through PubMed keywords, and 353 genes were combined for PPI protein interaction, and 331 genes were obtained for mutual relationship, which were analyzed by GO and KEGG. Conclusion: Berberine has protective effect on diabetic retina injury.

Berberine

diabetes retinopathy

network pharmacology

bioinformatics

Diabetic retinopathy (DR) is caused by chronic hyperglycemia (Youngblood, Robinson et al. 2019). The number of patients with DR is increasing as the elderly population increases (Bai, Yang et al. 2021). DR is characterized by retinal ischemia, leakage of retinal vascular system, retinal inflammation, neovascularization and angiogenesis, etc., which can be seen in the changes of micro vessels (Zou, Zhang et al. 2020, Yu, Zhang et al. 2021). DR is also a common and devastating complication in patients with diabetes mellitus (DM) that can lead to microvascular injury (Chakravarthy and Devanathan 2018). Patients with DM have a 35% chance of developing DR, a devastating complication from which there is no effective and definitive treatment option to recovery to date (Wang and Zhang 2021). It is also the number one cause of blindness when the disease is severe (Horton and Barrett 2021). DR is a specific manifestation of severe microvascular injury and can induce some other disease factors (Skol, Jung et al. 2020). Studies have shown that DM initially affects the retinal neurons of patients, leading to neurodegeneration that occurs early as DR, after which patients develop more pronounced vascular abnormalities (Kim, Kim et al. 2017). DR not only brings great distress to the life of patients, but also imposes heavy economic burden to their families and influences the development of society (Kim, Kim et al. 2018). Therefore, there is an urgent need to explore an effective method to alleviate and treat DR in patients with DM.

Berberine is a bioactive alkaloid with high medicinal value isolated from a variety of Chinese herbal medicine, which is quaternary ammonium salt from proto-berberine group of isoquinoline alkaloid (Ortiz, Lombardi et al. 2014, Cicero and Baggioni 2016, Wang, Feng et al. 2017). Berberine has a variety of pharmacological effects, including antibacterial, anti-cancer activity, antiviral activity, anti-diabetes (Warowicka, Nawrot et al. 2020). More studies have shown that berberine has a wide range of bioactive plant chemicals, and berberine is involved in extensive metabolism through oral administration, resulting in very low plasma exposure to berberine (Imenshahidi and Hosseinzadeh 2019, Song, Hao et al. 2020). Furthermore, berberine not only improve the efficacy but also the safety of chemotherapy. It have also shown that berberine has good efficacy in many chronic diseases, such as cardiovascular and cerebrovascular diseases and chronic hypertension (Hou, He et al. 2020, Suadoni and Atherton 2021). Berberine has many medicinal values that deserve further study and it also plays a positive contribution to the treatment of metabolic diseases (Kumar, Ekavali et al. 2015).

The purpose of this study is to investigate the protective mechanism of berberine on the treatment of DR by animal experiments, network pharmacology, bioinformatics, and molecular docking.

Establishment of DR rat model and drug therapy

A total of 30 Sprague-Dawley (SD) Male rats (weighing 180-200g) were provided from the Department of Experimental Animal Science, Kunming Medical University (No. SCXK (Yunnan) K2020-0004), and were randomly divided into Sham group (n=10), DM group (n=10), and BBR group (rats with DM were treated with berberine, n=10). All procedures were performed in accordance with the guidelines and approval of the Ethics Committee of the Kunming Medical University. The animal Ethics code is KMMU20220852,The rats in DM and SAL groups were intraperitoneally injected with levomectin to establish the DM rats model. After successful modeling, the rats in BBR group were given orally with berberine (100 mg/kg/day) for 4 weeks continuously, and the rats in Sham group and DM group were given normal saline (100 mg/kg/day).

Optical coherence tomography (OCT)

OCT is a technique that provides cross-sectional images of biological structures based on differences in the optical properties of tissues.(Katkar, Tadinada et al. 2018) ，Full name: SPECTRALIS OCT, Heidelberg, Germany，In this experiment, OCT measurement was performed on rats in the control group, DM group and BBR group. Rats were anesthetized by intraperitoneal injection of 10 g/L pentobarbital sodium (0.6ml/100g) the pupils were dilated with compound topicamide eye drops and the corneas were kept moist with drops of hydroxy glucan eye drops. The manufacturer of compound Topicamide eye drops is Shentian Pharmaceutical (China) Co., LTD，Hydroxy Glucan Eye Drops, Alcon Research LLC,Then placed in front of a corneal contact lens, the angle of light was then adjusted for fundus photography and OCT examination. All images were acquired at one optic disc interval, and the retinal thickness was measured by ImageJ software.

Morris water maze

To detect the learning and cognitive function of rats, we conducted the Morris water maze experiment, and was performed at 1 day (d), 3 d, and 5 d after modeling (Bromley-Brits, Deng et al. 2011). A 10-cm diameter platform was placed in a black circular pool filled with water with a width of 120 cm and a height of 56 cm. An automatic camera system connected to a computer was set up above the pool and the pool was divided into four quadrants on the computer. The rats were then placed in the water and the number of times the rats crossed the platform, the time to climb the platform and the total distance were recorded using the SMART 3.0 tracking system (Harvard Bioscience (Shanghai) Co., Ltd.).

Proteomics analysis

After 4 weeks of berberine treatment, the rats were anesthetized by intraperitoneal injection of 3% sodium pentobarbital,and the tissues of cortex, hippocampus, lung and retina were collected and stored at -80℃ for proteomics analysis. The proteomics analysis of samples were performed by Shanghai Luming Biotechnology Co., Ltd. The four-dimensional data-independent acquisition (4D-DIA) labeled quantitative proteomic analysis was conducted. Total proteins in the sample were extracted and part of samples was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and the other part was taken for trypsin enzymolysis. After desalting the enzymolysis peptide, the samples were identified by liquid chromatography tandem mass spectrometry (LC-MS/MS). A protein spectrum database was firstly established by the traditional DDA method, and then the mass spectrometry data of each sample was collected using DIA technique. Finally, based on the DDA database, spectrum matching, quantitative information extraction and subsequent statistical analysis were carried out.

Venn diagram and heat map

The genes of berberine and DR were obtained from PubMed by searching the keywords ‘salidroside’ and ‘diabetic retinopathy’, and the shared genes between them were performed using Venny2.1.0 (https://bioinfogp.cnb.csic.es/tools/venny/index.html). The heat map of protein expression in the samples were plotted by http://www.bioinformatics.com.cn, a free online platform for data analysis and visualization.

Molecular docking

Molecular docking was performed for berberine and genes. Firstly, the protein crystal structure of the gene was searched in Protein Data Bank (PDB) (https://www.rcsb.org), and the 2D structure of berberine was downloaded from PubChem (https://pubchem.ncbi.nlm.nih.gov/). The target protein structure was dehydrated with PyMOL software to separate the ligand and receptor. With the target protein as the receptor and the active substance as the ligand, the active site of molecular docking was determined according to the coordinate of the ligand in the target protein complex. Molecular docking is performed using AutoDock Vina.

Protein-protein interaction (PPI) network analysis

The genes were searched by entering the keyword ‘diabetic retinopathy’ on PubMed (https://pubmed.ncbi.nlm.nih.gov), and the related genes were recorded. The genes related to DR and screened by Venn diagram were imported into the STRING database (https://string-db.org/) to obtain the relevant PPI network. The obtained table was sorted out, and the duplicated genes in Node1 and Node2 were deleted, and finally genes with interactions were obtained.

Biological function enrichment analysis

The genes screened from PPI network analysis were analyzed by Gene Ontology (GO) (http://www.geneontology.org) and Kyoto Encyclopedia of Genes and Genomics (KEGG) (https://www.kegg.jp/kegg/pathway.html) analyses using R 4.1.0 (http://www.r-project.org). It was required to pre-install activePerl.exe and R 4.1.0, and then the software package in the script was used for installation. After the installation, The genes were pasted into Symbol ID, the script was opened, the script was pasted into R software, the ID was converted, the ID was pasted into GO and KEGG analyses, and the script for GO and KEGG analyses was pasted into R 4.1.0. The biological process (BP), cellular component (CC), and molecular function (MF) involved in GO enrichment analysis and KEGG signaling pathways were obtained, and were also plotted by http://www.bioinformatics.com.cn.

Statistical analysis

All data were entered into SPSS (version 21.0, SPSS Inc., USA) for statistical analysis. The one-way ANOVA was used for examination of the quantitative data, and the results were expressed as mean ± standard deviation (SD). ^*P<0.05 was considered as statistically significant.

Morphological changes of fundus

As shown, there was no morphological change in the control group, and the thickness of retinal pigment epithelial (RPE) outer nucleus and intima boundary-core layer in the DM group was significantly thinner than in the control group. It was shown that the thickness of RPE outer nucleus, intima boundary-core layer and external morphology layer was thickened in BBR group, compared with control and DM groups. (Figure 1)

Results of Morris water maze

In order to detect the rats' learning and cognitive function, we,3,5 1 day after surgery the sham group, DM group, BBR water maze platform incubation experiment in rats, using the single factor analysis of variance and repeated measurement variance analysis in SPSS software to carry on the analysis, the results of the analysis concluded that the diabetes in rats after building, compared with the sham group, The learning and cognitive functions of rats in the DM group were significant, P=0.002. After berberine treatment, on day 5, as shown in Figure 2, the learning and cognitive functions of rats in the BBR group were significantly higher than those in the DM group. As shown in Figure 2, according to the results of repeated measurement anOVA, the sphericity test, P=0.138, P > 0.05, satisfies the sphericity hypothesis. The sphericity test is subject to the sphericity test. The intra-subjective effect test results show are P < 0.05, indicating that there is a certain difference between different treatment groups.

Table 1

Repeat measure ANOVA table

Repeat measure ANOVA table
Variables	DF	SS	MF	F	P
Intervene	2	40345.701	20172.850	168.293	＜0.01
Intergroup error	20	2397.346	119.867	13.077	＜0.01
Time	2	2268.910	1134.455	4.771	0.003
Time*Intervene	4	1655.623	413.906
Repeated measurement	40	3469.999	86.750

Shared genes of berberine and DR

The sequencing data were collated and the screening data were identified by protein profiling. By setting P < 0.05 and logFC absolute value > 0.5 as screening conditions, 73 proteins from the Sham group and 140 proteins from the DR-Berberine group were analyzed. There were 10 shared genes were obtained from the intersection of 73 genes in the Sham group and 140 proteins in DR-Berberine group, including Dennd1a, Utp6, Fn1, Nsmce1, Lmo4, Ca1, Ogn, Ppl, Mme and Atp7a (Figure 3, Table 2).

There were 10 shared genes were obtained from the intersection of 73 genes in the Sham group and 140 proteins in DR-Berberine group, including Dennd1a, Utp6, Fn1, Nsmce1, Lmo4, Ca1, Ogn, Ppl, Mme and Atp7a

Table 2

The raw data of proteomics analysis.

Omics raw data
Accession	ProteinDescriptions	Gene Name	MW [KDa]	Qvalue	Cscore	C3-2	C4-2	C5-2	E2-2	E3-2	E4-2	D1-2	D5-2	D7-2	P-value	log2(FC)	FC
A0A0G2K6J1	DENN domain-containing 1A	Dennd1a	120.4645	1.23E-22	0.787164	-2.19851	-1.12243	-2.32555	-2.58972	-2.61238	-1.61579	-0.19728	0.185321	-1.01048	0.016012	-1.93182	0.262098
F1M805	UTP6 small subunit processome component	Utp6	70.23084	3.97E-28	0.795101	-3.04151	-3.93692	-1.67742	-3.3605	-2.13941	-3.45177	0.425808	0.958697	0.991181	0.001211	-3.77579	0.073009
A0A096P6L8	Fibronectin	Fn1	272.4461	2.66E-75	0.843369	0.289827	2.646698	3.939096	0.734381	1.124997	0.152081	-0.07604	-1.69622	-1.60421	0.039539	1.795973	3.472495
A0A0G2JSX1	Non-structural maintenance of chromosomes element 1 homolog	Nsmce1	33.30919	0.000604	0.737861	1.070106	0.611183	0.106892	0.576944	0.215285	0.97428	-1.61671	-1.55057	-1.31589	0.00093	2.083225	4.237534
A0A0G2K919	LIM domain only 4	Lmo4	18.0077	0.000991	0.729988	-2.47686	-2.49854	-3.24286	-2.17676	-3.02981	-2.88551	-4.03335	-4.25499	-3.41988	0.029398	1.205382	2.305983
B0BNN3	Carbonic anhydrase 1	Ca1	28.29963	9.47E-59	0.82902	2.366539	-0.81989	3.750589	0.768434	1.807509	1.012363	-2.04904	-1.77009	-3.87795	0.006786	3.761795	13.56479
D3ZVB7	Mimecan	Ogn	34.06928	7.18E-42	0.812023	-1.11086	0.137761	0.377591	2.075826	3.631153	1.877639	-2.80644	-3.28461	-3.66476	0.000681	5.780139	54.95349
D4A5T8	Periplakin	Ppl	204.1472	2.45E-63	0.833184	-0.66663	0.434097	-0.93671	0.017401	0.217498	-0.62961	-2.77657	-1.95053	-2.55713	0.002957	2.296507	4.912669
P07861	Neprilysin	Mme	85.79512	7.84E-61	0.830929	-0.68506	-1.03119	-0.46674	-1.17448	-0.54597	-0.35859	-2.9762	-2.55576	-2.31194	0.003605	1.921622	3.788487
P70705	Copper-transporting ATPase 1	Atp7a	162.0929	6.54E-19	0.781901	-0.49372	-0.37961	-0.54716	-0.41664	-0.02699	-0.95078	-3.34899	-2.11648	-4.07183	0.012596	2.714295	6.562726

Note: C Table Notes: C is sham group, E is diabetic retina injury group, D is berberine administration group

Comparison of expression level of key genes

The expression level of 10 shared genes were analyzed by proteomics and shown in Figure 4. The Dennd1a and Utp6 were downregulated in the DM group, while Atp7a, Ppl, Ogn, Nsmcel, Mme, Lmo4, Ca1 and Fn1 were upregulated. However, in the BBR group, Dennd1a and Utp6 were upregulated, and Atp7a, Ppl, Ogn, Nsmcel, Mme, Lmo4, Ca1 and Fn1 were downregulated.

Molecular Docking

As shown, Atp7a, Cal, Dennd1a, Lmo4, Mme, Nsmce1, and Ppl genes were able to form a stable molecular binding pattern with, indicating that these genes had a good binding activity with berberine, however, Fnl, Ogn, and Utp6 genes were not able to dock with berberine (Figure 5).

PPI network anlysis

A total 353 genes were obtained, including the 10 shraed genes and 343 genes obtained from the PubMed by searching for the keywords ‘diabetic retinopathy’ (Table 3). The PPI network analysis was performed for 353 genes, and interactions were found between 331 genes after screening out duplicates (Table4, Figure 6). There were 5 genes that were associated with the genes queried in PubMed, which were Atp7a, Lmo4, Mme, Ogn and Ppl.

Table 3

List of 353 genes related diabetic retinopathy.

List of 353 genes related diabetic retinopathy
Ghrelin	p-AKT	ERK2	Bcl-2	IL-1 β	MCP-1	FoxOs
ROS	p-mTOR	HRMECs	HSP	mRNA	JNK	SMP30
ARPE-19	REDOX	LC3	MAPKs	Lmo4	HRPC	Akt
IL-1B	Col12a1	Pro	hRPEs	microRNA	RBY79	GSK-3β
IL-18	Chad	PRR	HIF-1α	miR-139-5p	SO-RB50	RPE-EMT
NLRP3	Bgn	Gyp17	COX-2	lncRNA H19	TNF	PEI-CeO₂
VEGF	PPP1CA	NF-κB	VEGF-A	TGF-β	5-LOX	HA-CeO₂
RPE	YAP	SERCA2a	FGF2	EndMT	IL-1	STK25
RGC	STZ	PP1	miR-140-3p	IL-1β	IL-8	RGCs
Muller	HG	AGEs	circsLC16A12	IL-17A-KO	iNOS	CCK-8
SOD	Gln	PKC	RUNX1	IL-17A	FGF	Caspase-3
TUNEL	mTORC1	syndcan-1	RNS	IL-17RA	IGF-1	p53
TNF-α	NLR	CD44	ML-SA1	IL-17RA-shRNA	GAPDH	LC3-II
IL-6	GCL	syndecan-3	MCOLN1	IL-23	GSH	LC3-I
MCP-1	caspase-1	RRMECs	TRPML1	SIRT1/NOCTh1	CCN1	p-p53
MDA	CGA	syndecan-2	rMC1	AGE-RAGE	GLP-1RA	p53
T2DM	HREC	syndecan-4	PGC1α	CUR	GLP-1	CASP8
GHSR	APRE19	glypican-1	SOD3	INS	TCM	CASP9
IL17A	M2 ipRGCs	glypican-3	CCN1	CUR + INS	Keap1-Nrf2-ARE	p38αMAPK
RMC	M3 ipRGCs	p-ERK	TLR2	IL-17	GMFB	c-Jun N
IL17RA	M1	4-HNE	DPP	ADGRE1	miR-5195-3p	JNK1
Th17	AMPK	WIF1	GLP	C57BL6	miR-124-3p	ERK1
Act1	SIRT1	Wnt	SGLT	CCL2	CCK-8	SGK1
TRAF6	NAD	pdx1	eGFR	CD45	G3BP2	PAC
IKK	NADH	MAP1LC3	RIPK1	CXCL1	p38MAPK	Hsa-miR-3184-3p
NFκB	LKB1	SQSTM1	RIPK3	DAMPs	Irs2-KO	hsa-miR-24-3p
EAAT1	TLR4	BECN1	TNFR1	DBA/2J	Irs2	hsa-miR-197-3p
Ins2	MD2	DAPI	IFNR	HSPGs	TH	BMP4
USA	NOX4	EdU	TLR	IL6	miR-195	MiR-340-5p
CA	NADPH	ERG	MLKL	PCO	ChIP	miR-18a-3p
Gibco	TAK-242	FITC	SGLT2i	PDGF	Co-IP	VEGFR2
MD	HMGB-1	CTRP3	Cx43	PDR	LPS	VEGF121
Abcam	TRX1	Claudin-5	TGFβ1	PVD	P38 MAP	eVEGF-38
ZO1	PGC-1α	Occluding	P2X7R	PVR	FASN	eVEGF-53
GFAP	NRF1	CRP	ATP	TG	NFAT2	MAPK
GS	TFAM	INS-1	NRK-52E	TGF1	p-NF-κB	Rho GTPase
SYTO13	ERK	c-jun	pCD5	TM	TRPC6	CDC42
PI	AMPK	Akt	BCX	Th1	FC-DenseNet	ROCK
Nrf2	MEK	TLR9	GAP43	Th2	MA-FC-DenseNet	VEGF165
AGE	MEK1	MTOR	NCAM	αSMA	ATG4D	RIPK3
RAGE	MEK2	EMT	HO-1	mTOR	miR-125b-5p	circEhmt1
ECM	circRNA	YY1	L6H21	miR-15b-5p	Wnt/β	NFIA
Keap1	TXNIP	VEGFA	QRT-PCR	circ_001209	RT-qPCR	SA-β-gal
NeH2	circADAM9	Snail1	SIRT2	Glu	GLP1R	P21
HO1	CARM1	Smurf2	SIRT3	DR	S1PR2	OGG1
Nox2	KLF3	LOX	SIRT4	IS	JTE-013	RGC-5
LC3B-II.	Trx-R	CaMKII	SIRT5	IL-17	hRVECs	8-OHdG
p62	FGF-1	SH-SY5Y	SIRT6	NFKB	PFK-2	ICAM-1
PI3K	p38 MAPK	Bax	SIRT7	STAT-3	cJun	JNK
Dennd1a	Utp6	Fn1	Nsmce1	Lmo4	Ca1	Ogn
Ppl	Mme	Atp7a

Table 4

Related genes in the PPI network

Related genes in the PPI network
ADHFE1	FGF2	LY96	ROS1	YY1	PRKAB1	MCL1
AGER	FN1	MAP1LC3A	RRAD	DECR1	GCLC	BCL10
AKT1	FOXM1	MAP2K1	RUNX1	SGK1	MAPK14	SDC4
AKT1S1	G3BP2	MAP2K2	SDC2	MAPK1	LOX	GLP1R
ALOX5	GABPA	MAP2K7	SDC3	CCL2	SPARC	DENND1A
APC	GAP43	MAPK1	SDC4	CRP	GHSR	ALOX5
ARHGAP18	GAPDH	MAPK14	SGK1	TLR2	SNAI1	TXN
ATG16L1	GCLC	MAPK3	SIRT1	MAPK3	BMP4	MAP1LC3A
ATG4D	GFAP	MAPK8	SIRT2	NOX4	SIRT2	IRS2
ATP7A	GH1	MCL1	SIRT3	IL1B	APC	TH
BAX	GHSR	MCOLN1	SIRT4	FGF2	YY1	IL33
BCL10	GJA1	MLKL	SIRT5	ICAM1	SMURF2	SOD3
BCL2	GLP1R	MME	SIRT6	TP53	MAP2K2	SIRT3
BECN1	GMFB	MTOR	SIRT7	EGFR	KDR	SQSTM1
BMP4	GPC1	NCAM1	SLC1A3	IL18	PIK3CA	AKT1S1
C1QTNF3	GPC3	NDUFS1	SMURF2	CXCL8	NRP1	ATG16L1
CARKD	GSS	NEIL2	SNAI1	CASP3	SOD1	GAP43
CARM1	GSTA4	NELFCD	SOD1	NLRP3	CCNA2	MAPK8
CASP1	GTF2H1	NFATC1	SOD3	TLR9	TJP1	MAP2K7
CASP3	HMOX1	NFIA	SPARC	PTGS2	GJA1	BCL2
CASP8	HSP90AA1	NLRP3	SQSTM1	JUN	BAX	CDC42
CASP9	HTRA1	NLRX1	STIM2	TLR4	MAP2K1	CLDN5
CCL2	ICAM1	NOS2	STK11	CXCL1	FASN	NFATC1
CCNA2	IFNAR1	NOX4	STK25	CASP1	STK11	CD44
CD180	IL17A	NRF1	STUB1	AKT1	PPP1CA	PTPRC
CD44	IL17RA	NRP1	TBX21	INS	NOS2	TRAF6
CDC42	IL18	NT5C	TFAM	VEGFA	SIRT7	TMEM206
CEACAM5	IL1B	OCLN	TG	TNF	CASP9	NCAM1
CGA	IL23A	OGG1	TH	IL6	HSP90AA1	TXNIP
CHRM1	IL33	OGN	TJP1	TFAM	SIRT6	FGF1
CLDN5	IL6	PIK3CA	TLR2	CYBB	PPL	S1PR2
CNPY3	INS	PRKAB1	TLR4	TNFRSF1A	FOXM1	GFAP
COL12A1	IRF6	PRRT2	TLR9	KEAP1	IL17A	SIRT5
CRP	IRS2	PTGS2	TNF	TBX21	ATP7A	WIF1
CXCL1	ISL1	PTPRC	TNFRSF1A	SIRT4	FN1	ATG4D
CXCL8	JUN	PVRL1	TP53	SIRT1	OCLN	NLRX1
CYBB	KDR	RGN	TRAF3IP2	HMOX1	CASP8	RIPK1
DENND1A	KEAP1	RIPK1	TRAF6	STUB1	MTOR	RIPK3
DSPP	LOX	RIPK3	TXN	GAPDH	BECN1	MCOLN1
DUSP26	ROS1	EMR1	FAM20C	LMO4	GH1	NFIA
EARS2	ISL1	NELFCD	GATB	PRRT2	RUNX1	MOK
EGFR	GPC1	IL23A	RRAD	STIM2	OGN	TRAF3IP2
EMR1	HTRA1	IL17RA	NPTN	TAOK2	TRPC6	LY96
ERG	SDC2	CEACAM5	G3BP2	ST3GAL4	NEIL2	MLKL
EXTL3	IRF6	PVRL1	PDX1	STK25	DOCK7	SLC1A3
FASN	DSPP	SDC3	NRF1	YY1AP1	GSTA4	RASIP1
FGF1	GPC3	MME	RPE	GSS	TG	OGG1
PPA1	IFNAR1

GO enrichment analysis

As shown, the top 10 BPs of GO enrichment analysis involved in berberine on the treatment of DR were as follows: response to oxidative stress, cellular response to chemical stress, cellular response to oxidative stress, response to molecule of bacterial origin, reactive oxygen species metabolic process, activation of protein kinase activity, response to lipopolysaccharide, cellular response to biotic stimulus, regulation of reactive oxygen species metabolic process, and response to reactive oxygen species. The first 10 CCs involved of GO enrichment analysis were raft, membrane microdomain, membrane region, Golgi lumen, focal adhesion, cell-substrate junction, lysosomal lumen, late endosome, endoplasmic reticulum lumen, and endocytic vesicle. The top 10 MFs involved in berberine on the treatment of DR were NAD+ binding, protein Serine/threonine kinase activity, cytokine receptor binding, NAD-dependent protein deacetylase activity, protein serine/threonine/tyrosine kinase activity, MAP kinase activity, NAD binding, receptor ligand activity, signaling receptor activator activity, and protein deacetylase activity. Among them, the most important biological process is Response to oxidative stress, which involves most cell groups including membrane raft, and the most important molecular function is NAD+ binding.

KEGG signaling pathways analysis

It was shown that the first ten signaling pathways involved in the berberine on the treatment of DR were as follows: lipid and atherosclerosis, salmonella infection, AGE- signaling pathway in diabetic complications, NOD-like receptor signaling pathway, fluid shear stress and atherosclerosis, Kaposi sarcoma-associated herpesvirus infection, toll-like receptor signaling pathway, TNF signaling pathway, shigellosis, and hepatitis B (Figure 8A). The corresponding genes of each signaling pathway were sorted for the first 20 KEGG pathways, and the pathways were classified. The environmental information processing pathway was TNF signaling pathway. The signaling pathways of organismal systems were as follows: Nod-like receptor signaling pathway, Toll-like receptor signaling pathway, and IL-17 signaling pathway. The signaling pathways involved in the human diseases were lipid and atherosclerosis, salmonella infection, AGE-RAGE signaling pathway in diabetic complications, fluid shear stress and atherosclerosis, Kaposi sarcoma-associated herpesvirus infection, shigellosis, hepatitis B, yersinia infection, influenza A, non-alcoholic fatty liver disease, pathogenic Escherichia coli infection, chagas disease, pertussis, proteoglycans in cancer, tuberculosis, and human cytomegalovirus infection.

Table 5

The top 10 KEGG signaling pathways.

ID	Description	pvalue	p.adjust	qvalue	Count
hsa05417	Lipid and atherosclerosis	4.53E-23	1.09E-20	3.96E-21	34
hsa05132	Salmonella infection	5.52E-22	6.62E-20	2.41E-20	35
hsa04933	AGE-RAGE signaling pathway in diabetic complications	1.34E-19	1.07E-17	3.91E-18	23
hsa04621	NOD-like receptor signaling pathway	1.40E-18	8.42E-17	3.07E-17	28
hsa05418	Fluid shear stress and atherosclerosis	1.82E-18	8.75E-17	3.19E-17	25
hsa05167	Kaposi sarcoma-associated herpesvirus infection	6.00E-18	2.11E-16	7.68E-17	28
hsa04620	Toll-like receptor signaling pathway	6.16E-18	2.11E-16	7.68E-17	22
hsa04668	TNF signaling pathway	3.35E-17	1.00E-15	3.65E-16	22
hsa05131	Shigellosis	4.85E-17	1.29E-15	4.71E-16	30
hsa05161	Hepatitis B	8.40E-17	2.02E-15	7.34E-16	25

In this paper, we established DR rats model to investigated the effects of berberine on the treatment of DR using behavioral experiments, proteomics, PPI network, GO and KEGG analysis. It was shown that there were 2 down-regulated genes, namely Dennd1a and Utp6, and 8 up-regulated genes, namely Atp7a, Ppl, Ogn, Nsmcel, Mme, Lmo4, Ca1 and Fn1 in the DM group, while the opposite was true in BBR group. Molecular docking showed that berberine could form a stable molecular binding pattern with other genes except Fnl, Ogn and Utp6， Then, a total of 331 genes with interaction relationship were analyzed by GO and KEGG analyses. The results suggested that the most important BP, CC, and MF involved in the berberine on the treatment of DR were response to oxidative stress, raft, and NAD+ binding, respectively, and the most important signaling pathway was lipid and atherosclerosis. (Chen, Lin et al. 2021)

Diabetic retinopathy (DR) is a neurovascular disease that can seriously cause blindness (Youngblood, Robinson et al. 2019). To study the protective effect of berberine on DR, we carried out animal experiments, OCT detection and Morris water maze test (Baldan Ramsey and Pittenger 2010). It has been reported that berberine can protect retinal morphology from hyperglycemic injury and reduce glycogen accumulation (Yin, Tan et al. 2021). The results of Morris water maze showed that the learning and cognitive functions of rats with DM treated with berberine were significantly improved, and the thickness of RPE outer nucleus, intima boundary-core layer and external morphology layer was thickened in DR rats treated with berberine.

We sorted out the data after proteomics measurement, and 73 proteins in sham group and 140 proteins in DR-Berberine group were analyzed. There were 10 shared genes were obtained after intersection using Venn diagram, namely Dennd1a, Utp6, Fn1, Nsmce1, Lmo4, Ca1, Ogn, Ppl, Mme and Atp7a. The polymorphism of Dennd1a gene is associated with obesity, and the complications of obesity will occur after diabetic lesions (Kadioglu, Altun et al. 2018). Utp6 is a protein that can participate in ribosomal biogenesis (Champion, Kundrat et al. 2009). Fn1 is the accumulated component of the extracellular matrix and in the case of hyperglycemia, it is also a glycoprotein that is present in plasma and extracellular matrix (Wang, Zhao et al. 2021). Nsmce1 gene has been reported to maintain genome integrity (Gong, Wang et al. 2020). We hypothesized that DR would lead to decreased learning and cognitive function in rats, and Nsmcelc had a regulatory effect on behavioral cognitive function in rats. Lmo4 can proliferate and inhibit apoptosis, oxidative stress and existing inflammatory response, and also reduce the damage of retinal pigment epithelial cells (Shao, Chen et al. 2021). After modeling of DR, the hippocampal CA1 region is extensively damaged (Zhang, Zhang et al. 2015). Ogn is a homeostasis mediator of bone and glucose and also helps regulate glucose metabolism (Lee, Ali et al. 2018). DM leads to elevated Ppl (blood lipids) after eating, and metabolic activation (Mme) cell populations associated with insulin resistance are generated in obese environments. It has been reported that Atp7a has a protective effect on DM against endothelial dysfunction through extracellular superoxide dismutase (Sudhahar, Urao et al. 2013). Differential expression of heat map showed that there were two down-regulated genes in the DM group, namely Dennd1a and Utp6, and 8 up-regulated genes, namely Atp7a, Ppl, Ogn, Nsmce1, Mme, Lmo4, Ca1 and Fn1. In the BBR group, there were 8 down-regulated genes, Atp7a, Ppl, Ogn, Nsmce1, Mme, Lmo4, Ca1 and Fn1, and 2 up-regulated genes, Dennd1a and Utp6, respectively.

Additionally, the 10 shared genes may plays an important role in the protection of berberine on the treatment of DR. We verified molecular docking with these 10 genes as receptors and berberine as ligands, and all the other 8 genes, Dennd1a, Utp6, Nsmce1, Lmo4, Ca1, Ppl, Mme and Atp7a, could dock with the drug except for Fnl and Ogn target proteins which failed to dock successfully with berberine ligands. Therefore, berberine could improve DR on the basis of the regulation of these genes.

Among the 10 proteomic genes analyzed, Atp7a, Lmo4, Mme, Ogn and Ppl were correlated with the genes queried in PubMed. However, according to the comprehensive score (Zhang, Han et al. 2021), the top 10 genes with the closest association were AKT1:HSP90AA1, AKT1:MTOR, AKT1S1:MTOR, BCL2:BECN1, CASP1:IL1B, CASP1:NLRP3, CASP8:TNFRSF1A, CASP8:RIPK3, CASP8:RIPK1, EGFR:HSP90AA1. These closely related genes may play a key role in the protection of berberine on the DR.

GO enrichment analysis shows that the most important biological process is response to oxidative stress. It has been reported that antioxidant induced by oxidative stress can improve glucose homeostasis, which plays a key role in the pathophysiology of diabetes (Yaribeygi, Sathyapalan et al. 2020). The most involved cell component is membrane raft, the microregion of plasma membrane, which can provide regulation of cell metabolism at different levels (Sviridov and Miller 2020). The most important molecular function is NAD+ binding and NAD plays an important role in cellular energy homeostasis (Chen, Bang et al. 2015). NAD can effectively intervene in diet, so as to control the pathological changes in DM (Yoshino, Mills et al. 2011).

According to the results of GO enrichment analysis, multiple genes regulate each of the first 10 pathways in BP, CC and MF, which may play a role in the protective mechanism of berberine on DR. KEGG pathway analysis showed that the first pathway was lipid and atherosclerosis, and hyperglycemia was through the deficiency of insulin secretion. Insulin is a metabolic hormone and lack of insulin leads to abnormal lipid metabolism. DM induces atherosclerosis, and DM is followed by increased blood glucose levels and dyslipidemia, and abnormal blood lipids (Poznyak, Grechko et al. 2020). Salmonella neck abscess may lead to DM, which is a risk factor for salmonella infection (Pastagia and Jenkins 2013). The third pathway is AGE-RAGE signaling pathway in diabetic complications. AGE is the end product of advanced glycation and RAGE is a transmembrane protein, and RAGE is upregulated with the increase of circulating AGE level. AGE-RAGE signal transduction can also indirectly affect the occurrence of diabetes complications (Kay, Simpson et al. 2016). NOD-like receptor signaling pathway has been identified as a key regulator of inflammation-related angiogenesis (Liu, Lu et al. 2019). The fifth pathway is fluid shear stress and atherosclerosis, which is a regulatory factor of gene expression in vascular cells that is correlated with diabetes abnormalities (Papadaki, Eskin et al. 1999). The six pathway is Kaposi sarcoma-associated herpesvirus infection, (Ye, Zeng et al. 2016). Toll-like receptor signaling pathway is one of the factors of obesity-induced inflammatory response (Rogero and Calder 2018). In the DR, the TNF molecular pathway may be the cause of its pathogenesis and progression (Dewanjee, Das et al. 2018). Shigellosis may be a complication of diabetes. By vaccinating against viral hepatitis, the burden of diabetes is reduced and the risk of developing diabetes is reduced (Karnchanasorn, Ou et al. 2016). There are corresponding enriched genes in each of the signaling pathways of KEGG, and these genes may play a key role in the protective mechanism of berberine on the DR, which needs to be further explored.

In this study, we explored the key targets and mechanisms of berberine protection against diabetic retinopathy based on proteomics and network pharmacology, and Dennd1a, Utp6, Fn1, Nsmce1, Lmo4, Ca1, Ogn, Ppl, Mme and Atp7a genes were found that they were closely associated with the protective effect of berberine on the diabetic retinopathy.

Ethical approval and consent for participation

All procedures were performed in accordance with the guidelines and approval of the Ethics Committee of the Kunming Medical University. Approved by the Animal Experiment Ethics Review Committee of Kunming Medical University, the approval number is KMMU20220852

Human and animal ethics

No human studies are involved

The animal ethics code is KMMU20220852

Public consent

I declare that all authors agree to publish.

Whether there is supporting data

I declare that the data and materials contained in this manuscript have not been published elsewhere and are available.

Conflict of interest

There is no conflict interest in this study

Money

Translational study of microrNA-target gene regulatory network in stroke and acute brain injury. Major Science and Technology Project of Sichuan Province (in the field of social development), Project No. 2020YFS0043, fund 1 million, period 2020.1-2021.12, principal: Wang Tinghua

Author's Contribution

WTH and GXL conceived and designed the study. FC and CJL conducted experiments, LN analyzed data and wrote manuscripts. All authors have read and approved the manuscript. The authors point out that all data were generated internally, without the use of paper mills.

Bai, J., et al. (2021). "Ghrelin Ameliorates Diabetic Retinal Injury: Potential Therapeutic Avenues for Diabetic Retinopathy." Oxid Med Cell Longev 2021: 8043299. https://doi.org/10.1155/2021/8043299
Baldan Ramsey, L. C. and C. Pittenger (2010). "Cued and spatial learning in the water maze: equivalent learning in male and female mice." Neurosci Lett 483(2): 148-151. https://doi.org/10.1016/j.neulet.2010.07.082
Bromley-Brits, K., et al. (2011). "Morris water maze test for learning and memory deficits in Alzheimer's disease model mice." J Vis Exp(53). https://doi.org/10.3791/2920
Chakravarthy, H. and V. Devanathan (2018). "Molecular Mechanisms Mediating Diabetic Retinal Neurodegeneration: Potential Research Avenues and Therapeutic Targets." J Mol Neurosci 66(3): 445-461. https://doi.org/10.1007/s12031-018-1188-x
Champion, E. A., et al. (2009). "A structural model for the HAT domain of Utp6 incorporating bioinformatics and genetics." Protein Eng Des Sel 22(7): 431-439. https://doi.org/10.1093/protein/gzp022
Chen, Y., et al. (2015). "Acyl-CoA-binding domain containing 3 modulates NAD+ metabolism through activating poly(ADP-ribose) polymerase 1." Biochem J 469(2): 189-198. https://doi.org/10.1042/bj20141487
Chen, Z., et al. (2021). "Network pharmacology based research into the effect and mechanism of Yinchenhao Decoction against Cholangiocarcinoma." Chin Med 16(1): 13. https://doi.org/10.1186/s13020-021-00423-4
Cicero, A. F. and A. Baggioni (2016). "Berberine and Its Role in Chronic Disease." Adv Exp Med Biol 928: 27-45. https://doi.org/10.1007/978-3-319-41334-1_2
Dewanjee, S., et al. (2018). "Molecular mechanism of diabetic neuropathy and its pharmacotherapeutic targets." Eur J Pharmacol 833: 472-523. https://doi.org/10.1016/j.ejphar.2018.06.034
Gong, M., et al. (2020). "A transcriptomic analysis of Nsmce1 overexpression in mouse hippocampal neuronal cell by RNA sequencing." Funct Integr Genomics 20(3): 459-470. https://doi.org/10.1007/s10142-019-00728-6
Horton, W. B. and E. J. Barrett (2021). "Microvascular Dysfunction in Diabetes Mellitus and Cardiometabolic Disease." Endocr Rev 42(1): 29-55. https://doi.org/10.1210/endrev/bnaa025
Hou, Q., et al. (2020). "Berberine: A Traditional Natural Product With Novel Biological Activities." Altern Ther Health Med 26(S2): 20-27.
Imenshahidi, M. and H. Hosseinzadeh (2019). "Berberine and barberry (Berberis vulgaris): A clinical review." Phytother Res 33(3): 504-523. https://doi.org/10.1002/ptr.6252
Kadioglu, E., et al. (2018). "The role of DENND1A and CYP19A1 gene variants in individual susceptibility to obesity in Turkish population-a preliminary study." Mol Biol Rep 45(6): 2193-2199. https://doi.org/10.1007/s11033-018-4380-8
Karnchanasorn, R., et al. (2016). "Viral Hepatitis and Diabetes: Clinical Implications of Diabetes Prevention Through Hepatitis Vaccination." Curr Diab Rep 16(10): 101. https://doi.org/10.1007/s11892-016-0790-y
Katkar, R. A., et al. (2018). "Optical Coherence Tomography." Dent Clin North Am 62(3): 421-434. https://doi.org/10.1016/j.cden.2018.03.004
Kay, A. M., et al. (2016). "The Role of AGE/RAGE Signaling in Diabetes-Mediated Vascular Calcification." J Diabetes Res 2016: 6809703. https://doi.org/10.1155/2016/6809703
Kim, Y. S., et al. (2018). "Alpha-lipoic acid reduces retinal cell death in diabetic mice." Biochem Biophys Res Commun 503(3): 1307-1314. https://doi.org/10.1016/j.bbrc.2018.07.041
Kim, Y. S., et al. (2017). "Metformin protects against retinal cell death in diabetic mice." Biochem Biophys Res Commun 492(3): 397-403. https://doi.org/10.1016/j.bbrc.2017.08.087
Kumar, A., et al. (2015). "Current knowledge and pharmacological profile of berberine: An update." Eur J Pharmacol 761: 288-297. https://doi.org/10.1016/j.ejphar.2015.05.068
Lee, N. J., et al. (2018). "Osteoglycin, a novel coordinator of bone and glucose homeostasis." Mol Metab 13: 30-44. https://doi.org/10.1016/j.molmet.2018.05.004
Liu, P., et al. (2019). "NOD-like receptor signaling in inflammation-associated cancers: From functions to targeted therapies." Phytomedicine 64: 152925. https://doi.org/10.1016/j.phymed.2019.152925
Ortiz, L. M., et al. (2014). "Berberine, an epiphany against cancer." Molecules 19(8): 12349-12367. https://doi.org/10.3390/molecules190812349
Papadaki, M., et al. (1999). "Fluid shear stress as a regulator of gene expression in vascular cells: possible correlations with diabetic abnormalities." Diabetes Res Clin Pract 45(2-3): 89-99. https://doi.org/10.1016/s0168-8227(99)00036-4
Pastagia, M. and S. G. Jenkins (2013). "Salmonella neck abscess as an opportunistic infection in diabetes mellitus." Case Rep Infect Dis 2013: 708419. https://doi.org/10.1155/2013/708419
Poznyak, A., et al. (2020). "The Diabetes Mellitus-Atherosclerosis Connection: The Role of Lipid and Glucose Metabolism and Chronic Inflammation." Int J Mol Sci 21(5). https://doi.org/10.3390/ijms21051835
Rogero, M. M. and P. C. Calder (2018). "Obesity, Inflammation, Toll-Like Receptor 4 and Fatty Acids." Nutrients 10(4). https://doi.org/10.3390/nu10040432
Shao, K., et al. (2021). "MicroRNA-139-5p Alleviates High Glucose-Triggered Human Retinal Pigment Epithelial Cell Injury by Targeting LIM-Only Factor 4." Mediators Inflamm 2021: 1629783. https://doi.org/10.1155/2021/1629783
Skol, A. D., et al. (2020). "Integration of genomics and transcriptomics predicts diabetic retinopathy susceptibility genes." Elife 9. https://doi.org/10.7554/eLife.59980
Song, D., et al. (2020). "Biological properties and clinical applications of berberine." Front Med 14(5): 564-582. https://doi.org/10.1007/s11684-019-0724-6
Suadoni, M. T. and I. Atherton (2021). "Berberine for the treatment of hypertension: A systematic review." Complement Ther Clin Pract 42: 101287. https://doi.org/10.1016/j.ctcp.2020.101287
Sudhahar, V., et al. (2013). "Copper transporter ATP7A protects against endothelial dysfunction in type 1 diabetic mice by regulating extracellular superoxide dismutase." Diabetes 62(11): 3839-3850. https://doi.org/10.2337/db12-1228
Sviridov, D. and Y. I. Miller (2020). "Biology of Lipid Rafts: Introduction to the Thematic Review Series." J Lipid Res 61(5): 598-600. https://doi.org/10.1194/jlr.IN119000330
Wang, F. and M. Zhang (2021). "Circ_001209 aggravates diabetic retinal vascular dysfunction through regulating miR-15b-5p/COL12A1." J Transl Med 19(1): 294. https://doi.org/10.1186/s12967-021-02949-5
Wang, K., et al. (2017). "The metabolism of berberine and its contribution to the pharmacological effects." Drug Metab Rev 49(2): 139-157. https://doi.org/10.1080/03602532.2017.1306544
Wang, Y., et al. (2021). "Identification of fibronectin 1 (FN1) and complement component 3 (C3) as immune infiltration-related biomarkers for diabetic nephropathy using integrated bioinformatic analysis." Bioengineered 12(1): 5386-5401. https://doi.org/10.1080/21655979.2021.1960766
Warowicka, A., et al. (2020). "Antiviral activity of berberine." Arch Virol 165(9): 1935-1945. https://doi.org/10.1007/s00705-020-04706-3
Yaribeygi, H., et al. (2020). "Molecular Mechanisms Linking Oxidative Stress and Diabetes Mellitus." Oxid Med Cell Longev 2020: 8609213. https://doi.org/10.1155/2020/8609213
Ye, F., et al. (2016). "High Glucose Induces Reactivation of Latent Kaposi's Sarcoma-Associated Herpesvirus." J Virol 90(21): 9654-9663. https://doi.org/10.1128/jvi.01049-16
Yin, Z., et al. (2021). "Berberine prevents diabetic retinopathy through inhibiting HIF-1α /VEGF/ NF-κ B pathway in db/db mice." Pharmazie 76(4): 165-171. https://doi.org/10.1691/ph.2021.01012
Yoshino, J., et al. (2011). "Nicotinamide mononucleotide, a key NAD(+) intermediate, treats the pathophysiology of diet- and age-induced diabetes in mice." Cell Metab 14(4): 528-536. https://doi.org/10.1016/j.cmet.2011.08.014
Youngblood, H., et al. (2019). "Proteomic Biomarkers of Retinal Inflammation in Diabetic Retinopathy." Int J Mol Sci 20(19). https://doi.org/10.3390/ijms20194755
Yu, M., et al. (2021). "Gliquidone improves retinal injury to relieve diabetic retinopathy via regulation of SIRT1/Notch1 pathway." BMC Ophthalmol 21(1): 451. https://doi.org/10.1186/s12886-021-02215-8
Zhang, L., et al. (2021). "Exploring the mechanisms underlying the therapeutic effect of Salvia miltiorrhiza in diabetic nephropathy using network pharmacology and molecular docking." Biosci Rep 41(6). https://doi.org/10.1042/bsr20203520
Zhang, Y. W., et al. (2015). "Memory dysfunction in type 2 diabetes mellitus correlates with reduced hippocampal CA1 and subiculum volumes." Chin Med J (Engl) 128(4): 465-471. https://doi.org/10.4103/0366-6999.151082
Zou, W., et al. (2020). "p38 promoted retinal micro-angiogenesis through up-regulated RUNX1 expression in diabetic retinopathy." Biosci Rep 40(5). https://doi.org/10.1042/bsr20193256

No competing interests reported.

Download PDF

Version 2

posted

You are reading this latest preprint version

Network Pharmacology-based Analysis on Treatment of Diabetic Retinopathy by Berberine

Status:

Version 2

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 2