Abstract

   Choroidal neovascularization (CNV) is a severe pathological outcome of
   age-related macular degeneration (AMD) which could lead to blindness.
   The present study was aimed at exploring the mechanisms of CNV, hoping
   to provide new clues in finding the treatment target of CNV. ITRAQ
   approach was applied to analyze the proteins in choroidal
   neovascularization of CNV and normal rats. KEGG and GO annotation were
   used to analyze the differently-expressed proteins. Western Blotting
   was used as the method to identify the differently-expressed proteins.
   TAB1, one of the differently-expressed proteins, was chosen as the
   study objective in the following research. AAV-TAB1 over-expression
   vehicle was constructed to investigate the function of TAB1 in CNV.
   IL-6 and IL-18 were tested with the use of ELISA. Western Blotting was
   used to test the expression of NF-kB pathway molecules. In terms of the
   results, expressions of 49 kinds of proteins were up-regulated in CNV
   rats while expressions of 241 kinds of proteins were down-regulated
   among the 4380 identified proteins. Overexpression of TAB1
   significantly reduced areas of CNV lesions in vivo. Cell proliferation
   significantly increased in the TAB1 over-expressed group in RPE. IL-6
   expression significantly increased in the TAB1-overexpression group
   while IL-18 expression significantly reduced. Western blotting results
   showed NF-kB pathway molecules were involved in it. The present study
   suggests TAB1 may play a protective role in CNV progression and the
   relative pathway owns the potential to be the treating target of CNV.

Supplementary Information

   The online version contains supplementary material available at
   10.1038/s41598-025-15134-1.

   Keywords: Proteomics, ITRAQ, CNV, TAB1, Inflammation

   Subject terms: Cell growth, Mechanisms of disease

Introduction

   Choroidal neovascularization (CNV) is one of the main severe
   pathological outcomes of age-related macular degeneration (AMD), and is
   a main cause leading to blindness of the elderly over 50 years in
   European countries^[34]1,[35]2. Choroidal neovascularization originates
   from the micro-vessels of choroid, breaks through Bruch’s membrane, and
   stretches into the space under RPE or the space under the
   neuroepithelium layer. It leads to vision loss eventually. Current
   golden standard therapy of anti-VEGF has gained great
   success^[36]3,[37]4, however, some drawbacks remain such as a subset of
   patients having a low response to the therapy, which alludes to finding
   new mechanisms of CNV and searching for new treating targets of it.

   Protein, as we all know, is the function executor of life. Proteomics
   is defined as the genome expression of one kind of cell, tissue, or
   whole-life entity. As the onset and progression of one disease involve
   various kinds of protein, the application of proteomics in the
   exploration of disease progression or diagnosis and treatment is
   getting hot.

   The isobaric tag for relative and absolute quantification (ITRAQ)
   technology was developed by ABI company in 2004. It was widely applied
   in exploring disease mechanisms and searching for new disease markers.
   The present research was conducted in an attempt to identify
   differently-expressed proteins, aiming to find potential mechanisms of
   CNV and new treatment targets.

   TAK1(TGF-βactivated kinase-1) is a member of the MAPK family and also
   plays a role in TGF-βsignaling pathway. TAK1 lies in the point of the
   intersection of two pathways. The activation of TAK1 relies on the
   binding of TAB1 to it. The interaction of TAK1 and TAB1 functions in
   growth and development, cell survival, immune response, and tumor
   growth^[38]3.

Results

The number of total protein species identified by iTRAQ labeling technology
and the number of differently-expressed protein types

   By analyzing and comparing the labeled peptides, a total of 4380 types
   of protein were identified in the samples of the CNV group and the
   control group (Table [39]1). Differently-expressed proteins are defined
   as proteins whose up-regulation ratio is greater than 1.2 times or
   whose down-regulation ratio is less than 0.83 times, and P < 0.05. A
   total of 290 types of differently-expressed proteins were identified in
   this study, including 49 kinds of up-regulated proteins and 241 kinds
   of down-regulated proteins (Table [40]2). The statistical results of
   protein quantification were presented in the form of a Volcano Plot
   (Fig. [41]1).

Table 1.

   Statistic result of protein identification.
   Database Total spectra Spectra (PSM) Peptides Unique peptides Protein
   groups
   Rat 376,515 51,216 22,228 20,348 4380
   [42]Open in a new tab

Table 2.

   Statistic results of protein quantification.
   Comparisons (CNV/control) Up- Down- All
   Number of protein types   49  241   290
   [43]Open in a new tab

Fig. 1.

   [44]Fig. 1
   [45]Open in a new tab

   The volcano plot. The volcano plot was made with the fold change of
   protein expression between the CNV group and the control group and the
   P value obtained by the T-test. Volcano plot was used to show the
   significant difference between the two groups. The horizontal
   coordinate indicated the fold change (logarithmic transformation with
   base 2), and the vertical coordinate indicated the significant P value
   of the fold change (logarithmic transformation with base 10). The red
   dots in the picture represented the significantly differently-expressed
   proteins (the fold change is greater than 1.2 times and the P value is
   less than 0.05), and the black dots represented the proteins with no
   significant change.

Gene ontology analysis of differently-expressed proteins

   We first analyzed gene ontology (GO) terms for biological process,
   cellular components, and molecular function with the use of the
   software Blat2Go, then performed GO functional enrichment analysis of
   the differentially-expressed proteins using Fisher’s exact test. The
   analysis results showed that the differently-expressed proteins took
   part in the biological process of mitotic nuclear division, immune
   response-regulating cell surface receptor signaling pathway, negative
   regulation of transcription, DNA-templated, negative regulation of cell
   death and cytokine production (Fig. [46]2A, D, E). Besides, they were
   involved in cellular components like extracellular matrix component,
   COPII vesicle coat, nuclear exosome (RNase complex), endopeptidase Clp
   complex, and anchored component of membrane (Fig. [47]2B, D, E).
   Additionally, the differently-expressed proteins played a role in
   molecular functions like chromatin binding, transmembrane receptor
   activity, helicase activity, kinase regulator activity, and enzyme
   inhibitor activity (Fig. [48]2C–E);

Fig. 2.

   [49]Fig. 2
   [50]Open in a new tab

   Gene ontology of differently-expressed proteins in control and CNV
   groups. Analysis results of GO terms for biological process (A),
   cellular component (B), and molecular function (C) of the
   differently-expressed proteins in CNV and control group based on the
   Blast2Go software. Analysis of biological processes, cellular
   components, and molecular functions were fused as one graph (D). The
   significance enrichment analysis of GO functional annotations is to
   evaluate the significance level of protein enrichment of a certain GO
   functional term through Fisher’s Exact Test (E). The horizontal axis in
   the picture showed the enriched GO function classification, which was
   divided into three categories: Biological Process (BP), Molecular
   Function (MF), and Cellular Component (CC). The vertical ordinate
   represented the number of different proteins in each functional
   classification; The color of the bar chart represented the significance
   of the enriched GO functional classification, that is, the P value
   calculated based on Fisher’s Exact Test. The color of the bar changed
   from orange to red, representing the gradient of the P value. The
   closer the color is to red, the smaller the P value is. The label above
   the bar showed the enrichment factor (Rich Factor ≤ 1), which
   represented the proportion of the number of differentially-expressed
   proteins annotated to the GO functional class in all identified
   proteins annotated to that GO functional class.

Enrichment analysis of KEGG pathways of the differentially expressed proteins

   Pathway analysis is a direct and necessary way to systematically
   understand the biological processes of cells, the pathogenesis of
   diseases, and the mechanisms of drug action. The present results of the
   KEGG pathway annotation of significantly differently-expressed proteins
   revealed a series of significant changes in the proteins from the cell
   surface to the nucleus, biological events, and acting factors involved
   in these processes. Besides, the top 20 KEGG pathways were shown
   (Fig. [51]3A), including the PI3K-AKT signaling pathway, human
   papillomavirus infection, herpes simplex infection, and so forth.
   Furthermore, KEGG pathway enrichment analysis was performed on
   differently-expressed proteins by Fisher’s exact test. The results
   showed that pathways of human papillomavirus infection, PI3K-Akt
   signaling, herpes simplex infection, tuberculosis, and significant
   changes that occurred in important pathways in cancer were enriched
   (Fig. [52]3B).

Fig. 3.

   [53]Fig. 3
   [54]Open in a new tab

   Top 20 KEGG pathways and the enriched KEGG pathways (A) Top 20 KEGG
   pathways including PI3K-AKT signaling pathway, human papillomavirus
   infection, and herpes simplex infection were shown in this picture. The
   horizontal axis showed the signaling pathways and the vertical axis
   showed the number of the protein types which were enriched in the
   corresponding signaling pathway. (B) The horizontal coordinate
   represents the number of differentially-expressed protein types
   contained in each KEGG pathway and the vertical axis in this picture
   represents significantly-enriched KEGG pathways. The color gradient of
   the bar graph showed the significance of the enriched KEGG pathway,
   namely the P value calculated based on Fisher’s exact test.
   Specifically, the color changed from orange to red. The closer the
   color was to red, the higher the significance level of the KEGG pathway
   enrichment was; The label beside the bar showed the enrichment factor
   (Rich Factor ≤ 1), which represented the proportion of the number of
   differentially expressed proteins involved in a KEGG pathway to the
   number of proteins involved in that pathway among all identified
   proteins.

Validation of candidate differentially-expressed proteins with the use of
western blotting

   A total of eight types of protein, namely MCM7, YES1, SEPT9, p27 kip1,
   p57 kip2, TRIM32, RPIA, and TAB1 were chosen for validation and the
   total protein in the RPE-choroidal-sclera was detected with the use of
   western blotting method. The results showed the proteins of YES1,
   SEPT9, p27 kip1, and p57 kip2 expressions were not significantly
   different between the two groups. The expressions of TAB1, MCM7,
   TRIM32, and RPIA in the CNV model group were significantly lower than
   those in the normal group (Fig. [55]4).

Fig. 4.

   Fig. 4
   [56]Open in a new tab

   Verification of the eight types of candidate protein by western
   blotting (n = 3). (A, C, E) The expressions of the candidate
   differently-expressed proteins, namely MCM7, YES1, SEPT 9, p27 Kip1,
   p57 Kip2, TRIM 32, RPIA and TAB1. (B, D, F) As the statistical result
   was shown, expressions of TRIM 32, RPIA, and TAB1 were significantly
   down-regulated while expressions of MCM7, YES1, p27 kip, SEPT 9, and
   p57 kip 2 did not significantly differ from the control. (Note, ns = no
   significance; * = P < 0.05; ** = P < 0.01; *** = P < 0.001).

TAB1 over-expression in vivo and TAB1 expression in RPE-choroidal-sclera
complex

   At day 0 (immediately after transfection), day 7, day 14, day 30, and
   day 60 after transfection with AAV- TAB1 over-expression vector, the
   total protein of the RPE-choroid-sclera complex was extracted.
   Expressions of TAB1 at the five points in time were detected
   (Fig. [57]5). The expressions of TAB1 at day 14, day 30, and day 60
   were significantly higher than those at day 0 after transfection
   (P < 0.001, P < 0.01, P < 0.01), while there was no significant
   difference between day 7 and day 0 after transfection (P > 0.05).

Fig. 5.

   [58]Fig. 5
   [59]Open in a new tab

   Western Blotting results tested on TAB1 expression after AAV-TAB1
   over-expression (A) Fundus photograph of the BN rats hinted that the
   subretinal injection of AAV-TAB1 vector was successfully performed as
   the arrow showed the swelling of the retinal neuroepithelium layer.
   TAB1 in RPE-choroidal-sclera was tested by western Blotting. (B) A
   representative picture was shown (n = 3). (C) The statistical analysis
   result showed that TAB1 expression significantly increased at day 14,
   day 30, and day 60 compared with day 0 (P < 0.001, P < 0.01, P < 0.01),
   while there was no significant difference between day 7 and day 0
   (P > 0.05). (**, P < 0.01; ***, P < 0.01; ns, P > 0.05).

Analysis of choroidal neovascularization area

   Cardiac perfusion of FITC-dextran solution was performed into the
   hearts of BN rats in the CNV group, CNV + AAV group, and CNV + AAV-TAB1
   group at 7, 14, 30, and 60 days after the CNV model was constructed and
   the RPE-choroid-sclera complexes were taken out for flat mounts. CNV
   lesions were outlined and the areas of CNV lesions in rats were
   measured (Fig. [60]6A). The results showed that the lesion areas of the
   CNV + AAV-TAB1 group significantly decreased compared with CNV group at
   day 14, day 30 and day 60 after the performance of laser coagulation
   (P < 0.01, P < 0.001, P < 0.001) while there was no significant change
   of lesion area in the CNV + AAV group and CNV group (P > 0.05)
   (Fig. [61]6B). The concentrations of IL-6 and IL-18 in the
   RPE-choroid-sclera complex homogenate of the rats in the CNV group,
   CNV + AAV group, and CNV + AAV-TAB1 group were detected by ELISA. The
   ELISA results showed the expression of IL-6 decreased in the
   CNV + AAV-TAB1 group at day 14, day 30, and day 60 (Fig. [62]6C). In
   terms of IL-18, the expression of it was significantly up-regulated in
   CNV + AAV-TAB1 group compared with the other two groups at day 14, day
   30 and day 60 (Fig. [63]6D).

Fig. 6.

   [64]Fig. 6
   [65]Open in a new tab

   Flat mounts of RPE-choroidal-sclera complex and ELISA test of IL-6 and
   IL-18 (n = 3). (A) The rats in the three groups were sacrificed and the
   eyeballs were made for RPE-choroid-sclera flat mounts. CNV lesions were
   sketched and evaluated. The representative CNV lesion pictures are
   shown in this figure. (B) The statistic results showed that the CNV
   lesions were significantly smaller in the CNV + AAV-TAB1 group than in
   the CNV group at day 14, day 30, and day 60 (P < 0.01, P < 0.001,
   P < 0.001), while there was no statistical difference between CNV and
   CNV + AAV group at any point in time (note, ns = no significance; *,
   P < 0.05; **, P < 0.01; ***, P < 0.001). (C, D) The statistical
   analysis results showed the expression variation of IL-6 and IL-18 in
   each group. (C) IL-6 expression in the RPE-choroid-sclera complex of
   rats in the CNV + AAV-TAB1 group significantly decreased compared with
   that in the CNV control group at day 14, day 30, and day 60. (P < 0.01,
   P < 0.01, P < 0.01) while there was no significant difference in IL-6
   expression between the CNV + AAV group and CNV group at any point in
   time (P > 0.05). (D) The expression of IL-18 in the RPE-choroid-sclera
   complex of rats in the CNV + AAV-TAB1 group was significantly
   up-regulated compared with that in the CNV control group at day 14, day
   30, and day 60 (P < 0.01, P < 0.001, P < 0.01). There was no
   significant difference in IL-18 expressions between the CNV + AAV group
   and the CNV group at any time point (P > 0.05).

Detection of the transfection efficiency of Adv-TAB1 over-expression vector
in rat RPE cells by western blotting

   The transfected cells were placed under a fluorescence microscope to
   observe the GFP expression after 48 h of transfection, to judge the
   transfection efficiency of adenovirus vector. The transfection
   efficiency of adenovirus was measured as 80–90% (Fig. [66]7A). In
   addition, the total protein of RPE cells in the six-well plate was
   extracted and the expression of TAB1 was detected by WB method. It was
   found that the expression of TAB1 in the TAB1-Adv group was
   significantly higher than that in the control group and the TAB1
   expression didn’t significantly change between the Control and Adv
   group (Fig. [67]7B, C).

Fig. 7.

   [68]Fig. 7
   [69]Open in a new tab

   Transfection efficiency of the adenovirus in the rat primary RPE cells
   and the expression of TAB1. (A) As the picture showed, the transfection
   efficiency was around 80–90% (the number of GFP-expressing cells
   divided by the number of the whole cells in the same visual field). (B,
   C) The western blotting results showed that the TAB1 expression was
   significantly higher than that in the control group (P < 0.001) while
   the TAB1 expressions in control and Adv were not significantly
   different (n = 3). (P > 0.05) (Note, ns = no significance; *, P < 0.05;
   **, P < 0.01; ***, P < 0.001).

Proliferation of the primary RPE cells, IL-6 and IL-18 concentration in the
cell culture supernatant, and the P52 and Rel B expression after TAB-1
overexpression in an anaerobic environment

   Cell proliferation was measured by CCK-8 assay at 12 h, 24 h, 48 h, and
   72 h after treatment of CoCl[2] in both the Adv group and the Adv-TAB1
   group. The results of the CCK-8 experiment showed that the cell
   proliferation of the Adv-TAB1 group was higher than that of the Adv
   group after CoCl[2] treatment for 24 h, 48 h, and 72 h (P < 0.05,
   P < 0.01, P < 0.01), while there was no significant difference in cell
   proliferation between the two groups at 12 h after the treatment
   (Fig. [70]8A). ELISA data showed that TAB1 over-expression
   significantly down-regulated IL-6 secretion and up-regulated IL-18
   secretion (Fig. [71]8B, C). Western Blotting was used to test the
   expression of molecules in NF-κB, namely P52 and Rel B. Representative
   pictures were shown (Fig. [72]8D). Statistical data showed that TAB1
   over-expression enhanced the expression of P52 at 24 h, 48 h, and 72 h
   and also promoted the expression of Rel B at 48 h and 72 h
   (Fig. [73]8E, F).

Fig. 8.

   [74]Fig. 8
   [75]Open in a new tab

   Cell proliferation, IL-6 & IL-18 secretion, and NF-κB signaling
   molecules expression. (A) Cell proliferations were tested by CCK-8
   assay 12, 24, 48, and 72 h after adenovirus transfection. As is shown
   in the picture, cell proliferation was up-regulated in the TAB1-Adv
   group compared with the Adv group at hour 24, hour 48, and hour 72
   (P < 0.05, P < 0.01, P < 0.01). (B) IL-6 expression was significantly
   reduced at hour 24, hour 48, and hour 72 after CoCl[2] treatment, while
   it didn’t change at hour 12. (Note, ns = no significance; *, P < 0.05;
   **, P < 0.01; ***, P < 0.001). (C) IL-18 expression was significantly
   up-regulated in the Adv-TAB1 group at hour 24, hour 48, and hour 72
   after CoCl[2] treatment, while it didn’t significantly change at hour
   12 after CoCl[2] treatment (P > 0.05). (D) Representative pictures are
   shown in this figure (n = 3). (E) P52 expression didn’t differ
   significantly at hour 12 between Adv and Adv-TAB1 groups. P52
   expression significantly increased in Adv-TAB1 compared with Adv group
   at hours 24, 48 and 72 (P < 0.05, P < 0.05, P < 0.01). (F) Rel B
   expression didn’t differ significantly at hour 12 and hour 24 between
   the Adv and Adv-TAB1 group (P > 0.05), and it significantly increased
   in Adv-TAB1 compared with the Adv group at hour 24, 48 and 72
   (P < 0.05, P < 0.01).

Discussion

   About 5%-10% of AMD patients will progress to CNV, which causes serious
   damage to central vision to the patients, making a severe physical and
   mental burden to the families and a heavy economic burden to
   society^[76]5,[77]6. Anti-VEGF therapy has become a mainstream
   treatment of CNV and has gained great success^[78]7–[79]9. However,
   some patients don’t respond to the drug very well, which hinted us some
   other factors besides VEGF could cause CNV^[80]10, and it’s urgent to
   search for new treatment targets for CNV.

   In the present proteomics study, we applied the iTRAQ method to
   identify the proteins that the RPE-Choroid-Sclera complex contained and
   also analyzed proteins that were differently expressed between normal
   BN rats and CNV BN rats. For some research, it’s important to figure
   out a specific or the targeted protein’s function, the cellular
   localization, and the biological process in which the protein takes
   part. In a high throughput omics research and a macroscopical view, the
   primary task is to find out what proteins or biological processes were
   significantly impacted by the biological treatment^[81]11. Therefore,
   we need to generalize and analyze the protein functions more
   systematically. Thus, GO annotation and pathway enrichment emerged to
   analyze the differently-expressed proteins. In the present study, GO
   analyzation and KEGG pathways enrichment results showed that, in the
   RPE-choroidal-sclera complex, differently-expressed proteins not only
   took part in the biological process or cellular functions like
   chromosome binding, DNA replication, and cell proliferation but also
   participated in some biological processes like cytokine secretion and
   immunocompetence such as immune response-regulating cell surface
   receptor as was shown in our study results.

   As is known to us, the alteration of pro- or anti-angiogenesis function
   is associated with CNV development and abnormal cell proliferation of
   endothelial was involved in the progress of CNV. VEGF was one of the
   important factors that promote CNV progression among the
   pro-angiogenesis factors, which was the theoretical foundation of the
   antibody-based therapy targeting VEGF. Despite the great success the
   anti-VEGF treatment has gained, there remains the situation that some
   patients don’t respond positively to this therapy, which implies other
   factors besides VEGF have probably participated in CNV
   development^[82]12,[83]13. Some recent research showed an abundance of
   other factors, such as immune factors, facilitate CNV
   development^[84]14,[85]15, which was consistent with our present
   proteomics results that immune-related processes were involved in CNV
   development. For instance, at the early stage of CNV development,
   complement factors deposit at the place between Bruch’s membrane and
   RPE, format a local inflammation situation, which causes direct damage
   to the retinal photoreceptors and RPEs. On the other hand, innate
   system cells, such as microglia and macrophages promote the CNV
   development by triggering VEGF secretion by other cells like
   RPEs^[86]2. These findings are consistent with our proteomics results,
   which imply that immune-related processes were accompanied by cell
   proliferation in CNV progression.

   In consideration of the change fold of differentially-expressed
   proteins and their function, eight types of differently-expressed
   proteins were chosen for the following validation assay, namely MCM7,
   YES1, SEPT9, p27 kip1, p57 kip2, TRIM 32, RPIA and TAB1, their
   functions are generally on cell proliferation and immunological
   function, which are similar to the mechanisms of CNV development. The
   validation assay result showed that MCM7, TRIM32, RPIA, and TAB1 were
   differently expressed in RPE-Choroidal-Sclera complexes between Control
   and CNV groups. TAB1 was set as our study object in the following
   research since it has not been studied in the CNV development yet and
   the past research reported TAB1 was a proliferation- and immune-related
   factor, which is probably involved in CNV since CNV is an
   over-proliferation and immune-related disease.

   TGF-β activated kinase-1(TAK1), is a member of the Mitogen-activated
   Protein Kinase (MAPK) family, and also plays a role in TGF-β associated
   pathways, participating in cell proliferation, inflammation, and
   angiogenesis^[87]16,[88]17. Protein TABs, which contain three members,
   termed TAB1, TAB2, and TAB3, were adapter proteins of TAK1^[89]18. The
   activation of TAK1 depends on the direct binding of TAB1. TAB1
   regulates body growth, cell survival, immune response, and metabolism
   by triggering MAPK and TGF-β signaling pathways via binding and
   activating TAK1^[90]19,[91]20. Some research showed that the binding of
   TAB1 to TAK1 promoted tumor development by facilitating cell
   survival^[92]21. Meanwhile, some other research showed the binding of
   TAB1 to TAK1 promotes cell death.

   In the present research, TAB1 over-expression was performed to explore
   its function in CNV development in vivo. RPE-choroidal-sclera complex
   flat mounts showed TAB1 over-expression significantly reduced CNV
   lesions compared with the control group at day 7, day 14, day 30, and
   day 60. In addition, TAB1 over-expression significantly reduced IL-6
   secretion and increased IL-18 secretion of the RPE-choroidal-sclera
   complex. The results hinted to us that TAB1 may inhibit CNV
   development. It is speculated that TAB1 may inhibit CNV progression in
   a similar way with cancers as angiogenesis shares lots of similar
   mechanisms with cancers such as over-proliferation of cells, cell
   migration, and inflammation. In the last research, TAB1 over-expression
   was demonstrated to inhibit breast cancer cell migration and invasion
   through TAB1/TAK1/p38 MAPK signaling pathway^[93]22. The TAB1
   expression level was positively associated with patients’ prognosis.
   When TAB1 was knocked down, cell migration and invasion were
   significantly enhanced. Some other research showed TAB1 was associated
   with non-small cell lung cancer^[94]23. The evidence above showed the
   relationship between TAB1 and cancers may be similar to it between TAB1
   and CNV.

   Inflammation was demonstrated in the pathogenesis of CNV progression.
   Inflammatory factors accompanied by other cytokines regulated CNV onset
   and progression by impacting vascular formation, vascular leakage, and
   retina edema. At the late stage of AMD, drusen bodies deposit
   underneath the retina, release various cytokines and chemokines create
   a chronic inflammation circumstance inducing angiogenesis, termed
   choroidal neovascularization, the severe pathological outcome of many
   diseases including AMD^[95]24,[96]25. It has been demonstrated by a
   large amount of evidence that IL-6 was a pro-inflammation cytokine that
   activated and prompted CNV progression. In a recent study, IL-6 was
   significantly elevated in laser-induced mice compared with normal
   mice^[97]26. The researchers constructed an IL-6 knockout mice to
   explore the role of IL-6 in CNV and found that IL-6 deficiency
   decreased laser-induced CNV area and exogenous IL-6 and increased
   choroidal sprouting angiogenesis. Lai et al. found that administration
   of triptolide down-regulated pro-inflammation cytokines including IL-6,
   along with reduced CNV lesions^[98]26. The study also demonstrated IL-6
   expression ties with CNV progression.

   As a complex biological process, immunoreaction may have complicated
   functions, including pathogenic and protective effects, to CNV
   progression. In the past decades, IL-18, a member of the IL-1 family,
   was demonstrated to play a protective role in CNV onset and
   progression. Doyle and colleagues administered mature IL-18 to nonhuman
   primates and found that intravitreally injection of IL-18 significantly
   reduced CNV lesions in cynomolgus monkeys. It may be applied as an
   adjuvant immunotherapy-based treatment for CNV in the future^[99]27. In
   our study, TAB1 over-expression significantly reduced IL-6 expression
   and upregulated IL-18 expression in RPE-choroidal complex and hinted to
   us that TAB1 may play an anti-inflammation role in CNV progression and
   may present a potential therapeutic target to CNV.

   In the present study, the in vivo study verified the protective role of
   TAB1 in CNV as it reduced IL-6 expression, upregulated IL-18
   expression, and lessened CNV lesions in rats. RPE cells are recognized
   as the initial site of drusen accumulation during AMD pathogenesis. The
   excessive build-up of drusen and associated inflammatory
   microenvironmental damage compromise the integrity of the RPE barrier,
   thereby facilitating CNV and subsequent retinal damage. Inflammatory
   mediators secreted by RPE cells contribute significantly to the
   progression of CNV^[100]2. Second, in our in vivo studies, we utilized
   the AAV2 subtype, which has been demonstrated to exhibit a high tropism
   for RPE cells in vivo. Thus, RPE cells were selected as the targeted
   cells of our in vitro investigations. For the further exploration of
   the effects and mechanisms of TAB1 in RPE cells in anoxic environment,
   we constructed adenovirus-mediated TAB1 gene over-expression vehicle to
   rat RPE cells in vitro to investigate the impact of TAB1 in cell
   proliferation and inflammation and its related mechanisms. CCK-8 assay
   was performed and the results showed that TAB1 over-expression
   significantly enhanced RPE proliferation, reduced IL-6 production, and
   increased IL-18 secretion in an anoxic environment. In past research,
   TAB1 was demonstrated to be related to cytokines release in a recent
   genome-wide association transethnic meta-analyses research. TAB1 was
   identified as one of 11 genes that were significantly associated with
   von Willebrand factor (VWF) levels^[101]28, in this study, TAB1-silence
   increased VWF release in the in vitro functional assay. TAB1, as an
   intracellular mediator, regulated several intracellular signaling
   pathways including TGF-β and IL-1, which hinted that TAB1 may function
   in a similar way to regulate IL-6 and IL-18 secretion in RPE cells in
   anoxic environment. Recent studies showed that TAB1 not only regulates
   cytokines secretion but also regulates cell
   proliferation^[102]28–[103]35. TAB1, as a member of
   TLR3–MyD88–IRAK1–TRAF6–TAK1 axis, was identified to mediate cell
   proliferation in breast cancer^[104]36. Besides, it was demonstrated
   that micro-RNAs participate in the regulation of TAB1. In colorectal
   cancers, TAB1 was the target gene of miR-873 and was involved in
   regulating colorectal cancer cells’ proliferation^[105]34. In addition,
   TAB1 was confirmed to be over-expressing in non-small lung cancers,
   identified as a direct target gene of miR-889, and regulated the
   proliferation and invasiveness of non-small lung cancer cells^[106]33.
   In general, TAB1 was demonstrated to regulate cytokine production and
   cell proliferation, which followed the findings in our research.

   TAB1 has been widely demonstrated involved in the NF-κB signaling
   pathway via binding to TAK1. TAB1 was demonstrated expressed in many
   kinds of cells. For example, in esophageal cancer cells, activation of
   the TAB1/TAK1/IKKβ/NF-κB signaling axis induced proliferation,
   invasion, and migration of the cells^[107]34.

   In breast cancer cells, stabilization of the XIAP-TAB1-TAK1 complex
   induced NF-kB activation and thus led to cell death^[108]37.
   Constitutive EGFR signaling promotes invasion via activation of a
   TAB1-TAK1-NF-κB-EMP1 pathway, resulting in large tumors and worse
   survival in glioblastoma orthotopic models^[109]29. This evidence
   presented the regulation of TAK1-TAB1 complex in NF-κB signaling in
   different types of cells in various conditions^[110]38–[111]42. So far,
   the role of TAB1 in CNV has not been studied yet. In the present study,
   over-expression of TAB1 in rat RPE cells in the anoxic condition
   increased IL-18 secretion, down-regulated IL-6 secretion, and
   up-regulated the expressions of Rel B and P52 expressions. TAK1 is the
   first important part of the Nf-kB cascade signaling in the cell, and
   the Nf-kB signaling pathway mainly regulates biological functions
   related to immunity and inflammation^[112]19,[113]20. Overexpression of
   TAB1 in RPE cells may induce more TAK1 activation, thereby activating
   downstream relevant signaling pathway molecules and inducing a series
   of signaling cascades. Therefore, we examined the expression of P52 and
   Rel B in the Nf-kB signaling pathway to explore the mechanism of
   inflammation protection and proliferation protection of TAB1 in BN rat
   RPE cells under a hypoxia environment. The WB results of Nf-KB showed
   that the expression levels of P52 and Rel B were increased after
   overexpression of TAB1, which may promote the secretion of IL-18 and
   inhibit the secretion of IL-6 by triggering the Nf-kB signaling pathway
   molecules. The results hinted that TAB1 may function as a protective
   regulator in CNV conditions by impacting the NF-κB signaling pathway.
   However, we have to admit that further experiments were needed to
   explain the mechanism of TAB1 overexpression in RPE cells for in-depth
   exploration.

   In the present study, the number of sample replicates was set to three,
   a decision informed by multiple considerations. First, given the highly
   controllable experimental conditions—wherein cell culture and
   operations strictly adhere to standardized protocols and the
   anticipated variation is minimal—three replicate samples are sufficient
   to reflect the experimental effect accurately. Second, in alignment
   with conventions within the field and the experimental designs of
   comparable studies, three replicate samples are widely regarded as
   acceptable for preliminary investigations. Although the sample size is
   relatively small, the findings of this study nonetheless offer valuable
   insights that inform subsequent research. Future studies will further
   validate these findings by expanding the sample size to ensure the
   robustness, reliability, and generalizability of the results.

   In general, the present study was the first proteomics research to
   analyze rat CNV RPE-choroid-sclera complex. TAB1 was found to play a
   role in CNV progression and may have a protective role in CNV by
   impacting the IL-6 and IL-18 production and cell proliferation via
   NF-κB signaling, which provided a potential target for CNV treatment.

Materials and methods

Antibodies and reagents

   The primary antibodies used were as followed: Anti-MCM7(abcam,
   ab52489), anti-YES1(abcam, ab109744), anti-SEPT9(abnova,
   H00010801-A01), anti-P27 kip1 (CST, #2552), anti-P57 kip2 (Santa cruz,
   sc-56341), anti-TRIM 32 (Abcam, ab96612), anti-RPIA (Santa Cruz,
   SC515328), Anti-TAB1 (Santa Cruz, SC166138), HRP-tagged goat
   anti-rabbit antibody (Bioss, bs-80295G-HRP); HRP tagged goat anti-mouse
   antibody (bs-40296G-HRP); anti-NF-κBp52 (Abcam,ab129097),
   anti-RelB(Abcam, ab309084). The secondary antibodies used were as
   follows: Anti-β-actin(Bioss, bs-0061R), and anti-GAPDH(Bioss,
   bs-41373R).

Animals

   Brown Norway male rats (Animal Experiment Center of Chongqing Medical
   University, Chongqing, China), weighing 150–180 g and 6–8 weeks old,
   were used for this study. This study was approved by the Ethics
   Committee of the Second Affiliated Hospital of Chongqing Medical
   University (Chongqing, China; approval no. ID: 2018-222). All
   experimental procedures were conducted in conformity with the
   institutional guidelines issued by the IACUC of NUCM and the ARRIVE
   guidelines.

CNV model construction

   Rats in the CNV group were conducted with laser photocoagulation on
   their retina after anesthetization. (Parameters: Wavelength: 532 nm;
   Power: 85 mW; Spot diameter: 50 um; Photocoagulation time: 0.05 s).
   Each photocoagulation spot was 1 to 2 PD away from the optic disc and a
   total of 5 spots were made in each eye of a rat. FFA was performed to
   evaluate CNV formation 14 days after laser coagulation (Fig [114]S2).

Subretinal injection of AAV-TAB1 over-expression and AAV vector

   The rats were anesthetized using sodium pentobarbital (350 mg/kg,
   i.p.), pupils were dilated by the compound tropicamide, and the ocular
   surface was anesthetized by oxybuprocaine hydrochloride. The cornea was
   flattened by a coverslip with ofloxacin eye ointment as a medium
   between the coverslip and the cornea for a clear vision of the
   operator. A tunnel was made at the limbus with a BD needle. Then a
   needle of the microinjector went into the tunnel until the needle tip
   penetrated the retinal nerve epithelium when the operator felt a slight
   empty sense. 6 μl of AAV vector was injected into the subretinal space.

FITC-dextran perfusion and RPE-choroidal-sclera flat mount

   Rats were anesthetized and then perfused with 1 ml of 50 mg/ml
   FITC-labeled dextran (FD2000S, sigma, the USA) dissolved in phosphate
   buffer solution (PBS, PH 7.3) through their left ventricles. The
   successful performance of the perfusion was indicated by the appearance
   of a slightly yellow color on the conjunctiva, lips, and limbs. After
   the perfusion, the rats were sacrificed, and their eyeballs were
   extracted. The eyeballs were fixed in 10% formaldehyde solution for 1 h
   and then washed slightly with PBS. The corneas, lens, and retina were
   removed from the eyeballs, while the sclera, choroid, and RPE layers
   remained for flat mount preparation. Five or six incisions were
   radically made on the flat mount, and it was finally covered by a small
   slide. The CNV lesions were outlined, and the areas of the CNV lesions
   were measured using Image J.

Protein extraction and proteolysis

   Eyeballs of the rats in both groups were taken out after sacrifice. The
   choroidal-sclera complex is obtained after the cornea, lens, and retina
   are carefully removed. Then the choroidal-sclera complexes were all
   stored in EP tubes at − 80 °C until the protein digestion process.
   Specimens from nine BN rats in each group were randomly divided into 3
   samples in three EP tubes, each EP tube contains specimens from 3 BN
   rats. Samples were smashed by ultrasonic disrupter and proteins were
   extracted with the use of SDT solution (4% (w/v) SDS, 100 Tris/HCl
   pH7.6, 0.1 M DTT). The protein then was digested by trypsin in way of
   filter-aided proteome preparation (FASP), peptides were desalted by a
   C18 cartridge and were freeze-dried. Each sample was redissolved by
   40 μL dissolution buffer and quantified.

ITRAQ labeling and scx chromatography

   Workflow for ITRAQ-labeling and analysis of differentially-expressed
   proteins (Fig [115]S1). Peptides in each sample were taken out for
   ITRAQ labeling according to the ITRAQ labeling kit protocol from AB
   SCIEX company. The labeled peptides were mixed for the following
   fractionation with AKTA Purifier 100. Buffer solution A contained 10 mM
   KH2PO4 and 25% ACN, pH 3.0. Buffer solution B contained 10 mM KH2PO4,
   500 mM KCl, 25%ACN, pH 3.0. Buffer solution A worked as an equilibrium
   liquid for the chromatographic column. Samples were loaded in a sample
   injector and then flowed to the chromatographic column with a velocity
   of 1 ml/min. For the peptide separation, linear gradients of buffer
   solution B were as follows: 0% in 25 min; 0–10% from 25 to 32 min;
   10–20% from 32 to 42 min; 20–45% from 42 to 47 min; 45–100% from 47 to
   52 min; 100% from 52 to 60 min; After 60 min, buffer solution B was
   reset to 0%. Light absorption value in 214 nm was monitored during the
   elution process. Elution fractions were collected every 1 min and each
   fraction was freeze-dried and desalted by C18 Cartridge.

LC–MS/MS data gathering

   Each graded sample was separated by a liquid phase HPLC system, Easy
   nLC. Buffer solution A was solution contained 0.1% formic acid and
   buffer solution B was solution that contained 0.1% formic acid
   acetonitrile. 95% A worked as the equilibrium solution for the
   chromatographic column. The sample was loaded to the sample column
   (Thermo Scientific Acclaim PepMap100, 100 μm*2 cm, nanoViper C18)
   through an automatic sampler, separated when passing through the
   analytical column (Thermo scientific EASY column, 10 cm, ID75 μm, 3 μm,
   C18-A2) at a rate of 300 nl/min.

Protein identification & quantification analysis

   The raw data of the mass spectrometry were exported as raw files and
   they were identified and quantitatively analyzed with software Mascot
   2.2 and Proteome Discoverer 1.4. Bioinformatics Analysis, Go Function
   Annotation: The target proteins were annotated with Blast 2 Go. It
   includes four steps which are blast, mapping, annotation, and
   annotation augmentation. KEGG Pathway Annotations: The target proteins
   were annotated with the use of a KEGG automatic annotation
   server^[116]43. Enrichment Analysis of GO and KEGG annotation: Fisher’s
   exact test was used to compare contributions of GO analysis or KEGG
   pathways in target protein set and total protein set. Then the target
   protein set was enriched by GO annotation or KEGG pathway annotation.
   Protein Cluster Analysis: Firstly, the quantitative information of the
   target protein set was normalized. Then the Complex heatmap R was used
   for classification in dimensions of protein names and expression
   levels. Finally, the hierarchical clustering heat map was made.

Western blot analysis

   The RPE-Choroidal sclera complex was taken out after the cornea, lens,
   and retina were removed. Then they were milled and smashed when dipped
   in protein lysis buffer on the ice. The RPE-Choroidal sclera complex
   was homogenized with a RIPA Protein Extraction Kit (Beyotime
   Biotechnology, C2006), and the protein concentrations were determined
   using a BCA Protein Concentration Determination Kit (Beyotime
   Biotechnology, C2006). The total protein was separated by SDS-PAGE and
   transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore,
   Bedford, MA, USA). The membranes were blocked with 5% non-fat milk for
   1 h at room temperature and subsequently incubated with primary
   antibodies at 4 °C overnight. After washing with PBST, the membranes
   were incubated with secondary antibodies conjugated to horseradish
   peroxidase for 1 h at room temperature. With chemiluminescence, the
   antibodies were visualized. The linear range of band detection was met.
   Three independent technical replicates were conducted for each sample.
   The original WB image is in the supplementary material (Figure
   [117]S3–S8).

ELISA analysis

   RPE-choroidal-sclera complexes were cut into pieces and smashed in RIPA
   Protein Extraction Kit (Beyotime Biotechnology, C2006) on ice. Samples
   and standard samples were added to each hole, and incubated for 90 min
   at 37 °C. The biotin-marked antibody was added to incubation for 60 min
   at 37 °C. Then each hole was washed softly three times with 1× buffer.
   Avidin–biotin-peroxidase complex was added, and incubated for 30 min at
   37 °C. Each hole was washed softly 5 times with 1× buffer. TMB
   (Tetramethylbenzidine) buffer was added and then each hole was
   incubated for 15 min at room temperature. A stop buffer was finally
   added.

Statistical analysis

   Statistical analysis was performed using the SPSS 26.0 software.
   Correlation analysis between groups was conducted using an unpaired
   t-test and a one-way ANOVA, followed by post-hoc Tukey’s test, to
   analyze and compare the differences between groups. The data are
   expressed as the means ± standard (SD) deviations for at least three
   independent experiments. P-values below 0.05 were considered
   statistically significant.

Supplementary Information

   Below is the link to the electronic supplementary material.
   [118]Supplementary Material 1^ (2.4MB, docx)

Acknowledgements