Open Access
Clinical and Epidemiologic Research  |   November 2024
Potential Prognostic Protein Biomarkers in Tears From Noninfectious Uveitis Patients Under Biologic Treatment as a Prelude to Personalized Medicine
Author Affiliations & Notes
  • Lorena Rodríguez-Martínez
    FarmaCHUSLab Group, Health Research Institute of Santiago de Compostela (IDIS), Santiago de Compostela, Spain
  • Carmen Antía Rodríguez-Fernández
    FarmaCHUSLab Group, Health Research Institute of Santiago de Compostela (IDIS), Santiago de Compostela, Spain
    Department of Ophthalmology, Bellvitge University Hospital, 08907 Hospitalet de Llobregat, Spain
  • Olalla Rodríguez Lemos
    Ophthalmology Department, University Clinical Hospital of Vigo (SERGAS), Vigo, Spain
  • Begoña de Domingo
    Ophthalmology Department, University Clinical Hospital of Santiago Compostela (SERGAS), Santiago de Compostela, Spain
  • Pere García Bru
    Department of Ophthalmology, Bellvitge University Hospital, 08907 Hospitalet de Llobregat, Spain
  • Jesús Mateos
    FarmaCHUSLab Group, Health Research Institute of Santiago de Compostela (IDIS), Santiago de Compostela, Spain
  • Anxo Fernández-Ferreiro
    FarmaCHUSLab Group, Health Research Institute of Santiago de Compostela (IDIS), Santiago de Compostela, Spain
    Pharmacy Department, University Clinical Hospital of Santiago de Compostela (SERGAS), Santiago de Compostela, Spain
  • Correspondence: Jesús Mateos, Health Research Institute of Santiago de Compostela (IDIS), Travesía da Choupana, Santiago de Compostela, Galicia 15706, Spain; [email protected]
  • Anxo Fernandez-Ferreiro, University Clinical Hospital Santiago Compostela, Hospital Clinico Universitario Santiago de Compostela, Calle a Choupana s/n Planta-4, Santiago de Compostela, A Coruña 15706, Spain; [email protected]
  • Footnotes
     LRM and CARF contributed equally to this work.
Investigative Ophthalmology & Visual Science November 2024, Vol.65, 29. doi:https://doi.org/10.1167/iovs.65.13.29
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Lorena Rodríguez-Martínez, Carmen Antía Rodríguez-Fernández, Olalla Rodríguez Lemos, Begoña de Domingo, Pere García Bru, Jesús Mateos, Anxo Fernández-Ferreiro; Potential Prognostic Protein Biomarkers in Tears From Noninfectious Uveitis Patients Under Biologic Treatment as a Prelude to Personalized Medicine. Invest. Ophthalmol. Vis. Sci. 2024;65(13):29. https://doi.org/10.1167/iovs.65.13.29.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: Adalimumab (ADA) is a systemic biological treatment option approved for the treatment of noninfectious uveitis (NIU); however, up to 40% of patients do not respond to the drug, either in a primary or secondary manner. Here, we evaluated the proteomic profile of patients with NIU who fail to ADA to identify proteins implicated in intraocular inflammation, as well as potential biomarkers for treatment response and novel therapeutic targets.

Methods: Cross-sectional observational study of patients with NIU under ADA treatment for six or more months. Tears were collected with microcapillary tubes and protein analyzed by data-independent acquisition/sequential window acquisition of all theoretical mass spectra. Differentially expressed proteins (DEPs) were defined based on the fold change between their expression in nonresponders (NR) and responders (R). Protein network and gene ontology analysis were performed. The χ2 test for trend and receiver operating characteristic (ROC) curves were used to evaluate potential biomarkers of treatment response.

Results: Twenty-nine DEPs, 14 upregulated and 15 downregulated, were detected in NR. These proteins were mainly related to enhanced neutrophil effector functions and redox imbalance. ROC analysis identified defensin-1,3 (DEF-1,3), biotinidase, and ATP-binding cassette transporter A1 as potential biomarkers for treatment response.

Conclusions: This is the first study on a clinical cohort of patients with noninfectious uveitis that identifies tear proteins related to neutrophil hyperactivation as drivers of the persistent intraocular inflammation observed in NR to ADA and provides evidence that targeting interleukin 6, Janus kinases, or the complement cascade could be potential alternative therapeutic strategies in these patients. Our results indicate the potential of high-throughput proteomics to provide insights into the underlying pathological mechanisms of persistent intraocular inflammation observed in patients who do not adequately respond to anti-TNF treatment and the value of tear proteomics as a tool for personalized medicine.

Uveitis is defined as the inflammation of the uvea that includes the choroid, the ciliary body, and the iris. Noninfectious uveitis (NIU) accounts for 70% to 90% of uveitis cases1 and is the third-leading cause of preventable blindness in the developed world, representing a serious health concern.2,3 
Mainstay therapy includes corticosteroids and immunosuppressants; however, the emergence of biologic agents has significantly broadened the array of treatment options.4 Biologic therapy offers notable effectiveness, favorable safety profiles, and faster onset of therapeutic benefits when compared to conventional immunosuppressive drugs. Among these biologics, adalimumab (ADA), a monoclonal antibody designed to target TNF-α, stands as the sole systemic biologic treatment option approved by the Food and Drug Administration and the European Medicines Agency for the treatment of noninfectious intermediate, posterior, and panuveitis.5 However, up to 40% of patients do not respond to the drug, either in a primary or secondary manner.6 Nonresponder patients are under therapeutic gap, because effective therapeutic options after anti-TNF therapy are scarce and administered off-label, leading to recurrent disease relapses, severe visual impairment, and, eventually, blindness. A change to another biological drug with a different target is usually recommended7 but the choice is often arbitrary, based on internal hospital protocols, physician preference, drug availability, and economic costs, lacking solid scientific evidence and standardization. 
In the last decade, “shotgun” proteomics has become a key tool in clinical research. The term “shotgun” refers to the massive identification of proteins in a sample, enabling biomarker discovery, molecular diagnosis, treatment response prediction and advancing personalized medicine.8 Through analyzing differentially expressed proteins (DEPs) and molecular pathways, proteomics provides insights into disease mechanisms and potential new therapeutic targets. In the early stages of biomarker discovery, the use of untargeted shotgun proteomics is preferred because they maximize protein coverage (identification), and no prior knowledge of the molecular mechanisms of the disease is needed. Prior studies in patients and healthy volunteers have detected several proteins in tear samples either with data-dependent analysis (DDA) or data-independent acquisition (DIA) of the raw proteomic data.911 This amount of information leads to a broader comprehensive picture of the normal/pathological state. Velez et al.12 demonstrated this in a rare disease with unknown molecular mechanisms characterized by progressive uveitis, an inflammatory vitreoretinopathy where conventional treatments such as anti-TNFα or steroids had low response rates. They detected normal TNF-α levels, explaining previous treatment failures, whereas high levels of VEGF shifted the therapeutic approach toward the use of antiangiogenic drugs, resulting in remarkable clinical improvement. Additionally, high interleukin 6 (IL-6) levels in a subset of patients guided a repositioning strategy with anti-IL-6 to target fibrosis, promoting tissue repair. 
A key factor in translational approaches is noninvasive sample collection. Blood is widely used for biomarkers because of its accessibility but may not accurately reflect ophthalmic diseases. Ocular fluids, such as tears and aqueous and vitreous humor, may be more representative of undergoing biological processes.8,13 Tears stand out as a source of biomarkers for their noninvasiveness and easy access, providing insights into the status of ocular conditions such as NIU, dry eye disease, keratoconus, and thyroid-associated orbitopathy.11,14 
Our main objective was to use shotgun DIA proteomics to characterize the proteomic signature driving ocular inflammation in NIU patients unresponsive to anti-TNF-α. We aimed to identify potential biomarkers for treatment response and novel therapeutic targets to enhance clinical outcomes. 
Material and Methods
Study Design
This was a multicenter, observational, cross-sectional study with NIU patients treated with ADA in two Spanish tertiary level hospitals (University Clinical Hospital of Santiago de Compostela and University Clinical Hospital of Vigo), recruited between June and September 2022. The study protocol was approved by the Galician Ethics Regional Committee (registration code 2021/014) and was conducted under the principles of the Declaration of Helsinki, according to the Spanish Law of Biomedical Research no. 14/2007. All patients were older than two years of age and signed the informed consent voluntarily. 
Inclusion criteria were patients diagnosed with NIU that needed ADA to control intraocular inflammation, and a minimum six-month ADA treatment was required. Diagnosis was based on clinical symptoms, ophthalmological examination, and commonly requested biochemical, immunological, and hematological analysis. Subcutaneous ADA treatment was initiated due to inefficacy of first-line therapy, and refractoriness to treatment was defined as persistence of inflammation despite the use of the highest corticosteroid dose, immunosuppressant, or the combination of both, or the presence of recurrences above a dosage threshold. 
Exclusion criteria included recent intraocular surgery (three months), prior treatment with another biologic agent (unless at least five half-lives had elapsed), and incomplete ocular examination or tear sample collection at the time of the study. Patient management was conducted by a multidisciplinary team composed of ophthalmologists, rheumatologists, and internists as per individual needs. 
Clinical Characteristics and Disease Evaluation
Patients’ characteristics were recorded including age and sex. Uveitis characteristics, such as laterality and anatomic localization (anterior, intermediate, posterior or panuveitis) were classified per the Standardization of Uveitis Nomenclature Working Group criteria.15 
Intraocular inflammation severity was classified according to Standardization of Uveitis Nomenclature criteria.15 Anterior chamber cells were graded from 0 to 4+ via slit lamp biomicroscopy before mydriatic drops, while vitreous haze was graded from 0 to 4+ through indirect ophthalmoscopy after mydriasis according to the Nussenblatt criteria.15,16 
Uveitis was deemed active with at least one active inflammatory chorioretinal or retinal lesion, anterior chamber cell grade of ≥0.5+, or vitreous haze grade of ≥0.5+. Otherwise, it was considered inactive. Patients or eyes with active disease were classified as nonresponders to ADA (NR), while those without intraocular activity were responders (R). Data were managed using the electronic system Research Electronic Data Capture (REDCap) facilitated by Fundación Sociedad Española Farmacia Hospitalaria.17 
Tear and Serum Sample Collection
Tear samples were obtained from both eyes using 3-µl Microcaps microcapillary tubes (no. 1-000-0030; Drummond Scientific, Broomall, PA, USA). Collection was done as close as possible to the next ADA administration, no longer than 48 hours before the day of the administration. Unstimulated tears (before clinical assessment or drops instillation and minimizing ocular irritation to avoid reflex tear) were collected via capillarity from the inferior temporal tear meniscus and dispensed into precooled 0.5 mL Protein LoBind tubes (Eppendorf, Madrid, Spain). Patients were instructed not to use artificial tears 24h prior to collection. To prevent keratin contamination, gloves were worn, and special attention was paid to avoid touching the eyelashes or skin with the capillary tip. Samples were kept cold throughout the collection process using the IsoPack/IsoRack set (no. 3880000.160; Eppendorf) to prevent protein degradation, then immediately stored at −20°C.18,19 Patients reported no or minimal discomfort during the procedure. Serum samples were collected to measure ADA through levels using Promonitor ELISA kit as previously described.20 
Shotgun Proteomics in Tear Samples and Data Analysis
Total tear protein content was quantified (Fig. 1) using a NanoDrop One spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and then digested to obtain peptides that were injected and analyzed by Data-Independent Acquisition/Sequential Window Acquisition of all Theoretical Mass Spectra (DIA/SWATH) using a Triple TOF 6600 system (Sciex, Framingham, MA, USA). A complete description of this section can be found in the Supplementary Material
Figure 1.
 
Schematic representation of the steps involved in the proteomic analysis. (1) Tear sample collection with microcapillary tubes and storage, (2) protein quantification, (3) running of protein samples into gel and excision of protein bands, (4) in-gel digestion of proteins with trypsin, (5) peptide analysis by DIA/SWATH, and (6) data analysis.
Figure 1.
 
Schematic representation of the steps involved in the proteomic analysis. (1) Tear sample collection with microcapillary tubes and storage, (2) protein quantification, (3) running of protein samples into gel and excision of protein bands, (4) in-gel digestion of proteins with trypsin, (5) peptide analysis by DIA/SWATH, and (6) data analysis.
Statistical Analysis
Continuous variables are depicted as the mean ± standard deviation (SD) or the median with the interquartile range (IQR) if variables were not normally distributed. Categorical variables are shown as frequency and corresponding percentage (%). Comparison of continuous variables was performed with t-test or Mann-Whitney U test. Differences in dichotomous variables were assessed with the Fisher's exact test or χ2 test as appropriate. 
MS data files were processed using Spectronaut Software 16.0 (Biognosys) with library-free directDIA pipeline's standard settings and normalized against the total peptide amount. After normalization, the ratio (NR/R) of protein expression was calculated. DEPs were defined as those with a fold change (ratio) >1.4 or <0.7 with a corresponding P value < 0.05. To focus on representative proteins, those present in less than 80% of the samples from the two subgroups were excluded (See Supplementary Fig. S1). Proteomic files are deposited and publicly available in the MassIVE dataset MSV000095892 at MassIVE repository (https://massive.ucsd.edu/). 
Protein Interaction network analysis was performed using the STRING database (https://string-db.org/), which enables the creation of a map of known interactions between DEPs identified across groups and provides information about enriched biological processes in which proteins are involved, the subcellular localization and the cellular compartment where they can be found, also known as gene ontology (GO) enrichment analysis. Bubble plot of GO enriched terms was done using SRplot.21 
The χ2 test for trend was used to assess whether there is a linear trend between protein amount of individual DEPs and the proportion of NR/R to ADA. For this analysis, protein amount was ordered from low to high and categorized in quartiles (25th quartile containing lower concentration, 100th quartile higher concentration). To evaluate the potential of DEPs as biomarkers for treatment response, receiver operating characteristic curves were plotted and the area under the curve was calculated. Data were analyzed using GraphPad Prism Software v8. Results were considered statistically significant when P < 0.05. 
Results
Patient Characteristics
Thirty-five patients (58 eyes) were included, with a mean age of 44 years (57% women). Panuveitis was the most common location (34%), followed by anterior uveitis (31%), intermediate and posterior uveitis (both 17%). Bilateral involvement was 74%. NR to ADA accounted for 49% (Table 1). 
Table 1.
 
Clinical Characteristics of NIU Patients Included in the Study
Table 1.
 
Clinical Characteristics of NIU Patients Included in the Study
Only tear samples of affected eyes were included in the analysis, 32 (55.2%) R and 26 (44.8%) NR (Fig. 2). No differences were observed in protein concentration or total protein between R and NR eyes. Only tear volume was slightly higher in the NR subgroup (See Supplementary Fig. S2). 
Figure 2.
 
Ocular fundus imaging of right and left eye of a patient with a good response to ADA. Color retinography (upper image) and autofluorescence (lower image) shown no active lesions despite severe affectation of the posterior pole in left eye. Both anterior chamber cell grade and vitreous haze grade were 0.
Figure 2.
 
Ocular fundus imaging of right and left eye of a patient with a good response to ADA. Color retinography (upper image) and autofluorescence (lower image) shown no active lesions despite severe affectation of the posterior pole in left eye. Both anterior chamber cell grade and vitreous haze grade were 0.
Differential Tear Proteomic Profile on Anti-TNF Treatment
In total, 399 proteins were quantified in tear samples (Supplementary Table 1); 29 of those met our predefined DEPs criteria (Table 2). Fourteen proteins were upregulated in NR eyes and 15 were downregulated (Fig. 3). 
Table 2.
 
Altered Tear Proteins in Patients With NIU Treated With ADA
Table 2.
 
Altered Tear Proteins in Patients With NIU Treated With ADA
Figure 3.
 
Volcano plots of DEPs in tear samples from NR eyes compared to R eyes to ADA treatment. The ratio of protein expression in NR versus R eyes is displayed in the x-axis as the log2 and the −log10 of the P value obtained from the differential abundance testing is displayed in the y-axis. Therefore proteins displayed at log2 values >0 are more abundant in NR eyes and those displayed at values <0 are more abundant in R eyes. Significantly modulated proteins are displayed at −log10 (P value) >1.3 corresponding to a P value = 0.05. Proteins excluded from the Uniprot database or outside our predefined criteria for DEPs are not shown.
Figure 3.
 
Volcano plots of DEPs in tear samples from NR eyes compared to R eyes to ADA treatment. The ratio of protein expression in NR versus R eyes is displayed in the x-axis as the log2 and the −log10 of the P value obtained from the differential abundance testing is displayed in the y-axis. Therefore proteins displayed at log2 values >0 are more abundant in NR eyes and those displayed at values <0 are more abundant in R eyes. Significantly modulated proteins are displayed at −log10 (P value) >1.3 corresponding to a P value = 0.05. Proteins excluded from the Uniprot database or outside our predefined criteria for DEPs are not shown.
Proteins are listed in decreasing order of ratio calculated by dividing the amount of protein quantified in NR eyes by that quantified in R eyes. Ratios >1 indicate protein overexpression in the NR subgroup, whereas ratios <1 indicate lower protein expression in the NR subgroup, i.e. proteins are overexpressed in the R subgroup. Accession numbers from UniProt database (https://www.uniprot.org/). DEF1;DEF3 and EF1A1;EF1A3 could not be unambiguously identified by unique peptides likely due to their high homology (sequences differ by one amino acid), assessed with the freely available resource SIM - Alignment Tool for Protein Sequences (https://web.expasy.org/sim/). Proteins excluded from the Uniprot database, such as most non-germline immunoglobulins and T-cell receptors, are not shown. 
Altered Proteins in NRs Have Immunomodulatory Roles
The STRING database for protein-protein interaction analysis (Fig. 4A) revealed one principal network of interacting proteins containing most DEPs. GO enrichment analysis showed that modulated proteins were involved in biological processes mainly related to immunomodulatory and antimicrobial activity, redox balance, and retina homeostasis (Fig. 4B) and secreted into the extracellular space by vesicles (Fig. 4C). 
Figure 4.
 
STRING Interaction Network and GO analysis of representative DEPs in the nonresponder and the responder eye subgroups. (A) Interaction network analysis. Proteins are depicted as nodes (circles), and lines represent the interactions between any two proteins either direct (physical) or indirect (functional) in nature. Upregulated proteins in NR eyes are shown in red, downregulated proteins are shown in blue. (B and (C) GO enrichment analysis. The bubble plot displays the biological processes (B) in which DEPs participate (those upregulated and downregulated in NR) at the top, and the cellular components (C) where they are located.
Figure 4.
 
STRING Interaction Network and GO analysis of representative DEPs in the nonresponder and the responder eye subgroups. (A) Interaction network analysis. Proteins are depicted as nodes (circles), and lines represent the interactions between any two proteins either direct (physical) or indirect (functional) in nature. Upregulated proteins in NR eyes are shown in red, downregulated proteins are shown in blue. (B and (C) GO enrichment analysis. The bubble plot displays the biological processes (B) in which DEPs participate (those upregulated and downregulated in NR) at the top, and the cellular components (C) where they are located.
Proteins with protective activity against reactive oxygen species (ROS) and other reactive toxic substances were decreased in the NR group, namely peroxiredoxin-1 (PRDX1), glutathione S-transferase P (GSTP1), aldehyde dehydrogenase 1 A1 (ALDH1A1) and haptoglobin (HP). Cystatin-SN (CST1) and ATP-binding cassette transporter (ABCA1), involved in tissue homeostasis, were also decreased in NR. Upregulated proteins in NR with a role in immunomodulatory or antimicrobial processes included protein S100A8 (S100A8), protein S100A9 (S100A9), α-defensins (DEF1;DEF3), anaphylatoxin C4a (C4A), and fibrinogen β (FIBB), implicated in pro-inflammatory processes. 
Potential Biomarkers for Treatment Response
These 29 identified DEPs were further investigated as potential as biomarkers for treatment response. Significant relationships between protein abundance and frequency of response to ADA were found for eight proteins (Fig. 5). A trend toward a higher percentage of NR eyes was found with increasing concentrations of DEF1;DEF3, S100A9, biotinidase, and C4A and with decreasing concentrations of ABCA1, CST1, HP and HSPB1, in line with their expression in NR (Supplementary Fig. S3). Assessment of the ability of these eight proteins to discriminate between R and NR revealed that DEF1;DEF3, biotinidase, and ABCA1 were promising as potential treatment response biomarkers, with moderate area under the curve values of 0.73, 0.68, and 0.67 (Fig. 6), respectively. 
Figure 5.
 
Relationship between the prevalence of NR and R eyes with protein abundance expressed in increasing order from the lowest (25th) to the highest (75th) quartile. Statistical analysis was done using χ2 test for trend (*P < 0.05, **P < 0.01, ***P < 0.001).
Figure 5.
 
Relationship between the prevalence of NR and R eyes with protein abundance expressed in increasing order from the lowest (25th) to the highest (75th) quartile. Statistical analysis was done using χ2 test for trend (*P < 0.05, **P < 0.01, ***P < 0.001).
Figure 6.
 
Potential tear protein biomarkers for treatment response to ADA based on receiver operating characteristic curve analysis.
Figure 6.
 
Potential tear protein biomarkers for treatment response to ADA based on receiver operating characteristic curve analysis.
Discussion
This is the first study pointing toward neutrophil hyperactivation as the underlying mechanism behind the persistent intraocular inflammation in patients with NIU despite anti-TNFα treatment, regardless of uveitis type and associated systemic disease, by using high-throughput label-free quantitative proteomics in tear samples. Our results suggested neutrophil-related innate immunity imbalance as the cause of inflammation seen in NR patients, which could inform decisions regarding biological therapy changes when dealing with multidrug-resistant uveitis. In addition, three of the altered proteins, DEF1;DEF3, biotinidase, and ABCA1, were identified as promising biomarkers of treatment response. 
Reduced expression of antioxidant proteins suggests compromised antioxidant activity. PRDX1 converts highly reactive hydrogen peroxide and hydroperoxide compounds into less toxic compounds.22 GSTP1 reacts with potentially toxic electrophilic compounds.23 ALDH1A1 functions as a retinal dehydrogenase preventing the formation of cytotoxic protein adducts by oxidizing reactive aldehydes.24 HP scavenges free hemoglobin preventing its oxidative damage.25 Impairment of antioxidant defenses can lead to ROS-mediated cell oxidative damage, known to induce sustained inflammation and tissue injury.26 In fact, an enrichment of pathways related to responses to oxygen-containing compounds was detected in NR, and neutrophils are known to produce large amounts of ROS as a defense against microbial infections. CST1 expression was also downregulated in NR, which has been proposed to have a role in eye homeostasis through its immunomodulatory activity and the inhibition of exogenous proteases.27 
One of the main sources of ROS in neutrophils is NADPH-oxidase. This enzyme is activated by the S100A8/S100A9 complex, also known as calprotectin,28 which was upregulated in NR, suggesting that inflammation could be due to calprotectin-mediated NADPH-oxidase-dependent ROS production. Calprotectin can also acts as damage-associated molecular pattern molecule and promote inflammation upon interaction with toll-like receptor 4 (TRL4) or with receptor for advanced glycation end-products, leading to recruitment and activation of neutrophils and other immune cells.2931 In patients with NIU, S100A8/S100A9 levels have been correlated with the severity of intraocular inflammation.3234 Additionally, S100A8/S100A9 can upregulate IL-6 production and activate the complement system, creating a vicious cycle of chronic inflammation-mediated tissue destruction in immune-mediated diseases.35 
Neutrophil recruitment, survival, ROS production and proinflammatory cytokine expression are modulated by fibrinogen present at sites of inflammation.3638 A link between fibrinogen alteration and neutrophil NADPH-oxidase ROS production has been proposed in the pathogenesis of Behçet disease (BD),39 which may be common to other NIU forms, as we observed increased FIBB expression in >80% of NR, including patients without BD. 
These and other modulated proteins related to neutrophil effector functions indicate neutrophil recruitment and activation in NR. The α-defensins (α-DEFs) such as DEF1 and DEF3 are antimicrobial proteins secreted by activated neutrophils that induce TNF-α and interferon γ (INF-γ)–mediated inflammation40 and have emerged as promising biomarkers of treatment response in this study. Another role of DEF-1 is the inhibition of the complement classical and lectin pathways.41 However, our observation of increased C4A expression suggests these pathways are activated. C4A expression has been related to gene copy number variations and higher IL-6 expression in BD,42 involved in neutrophil trafficking, activation, enhancing effector function and survival.43,44 C4A also inhibits monocyte chemotaxis,45 aligning with the enriched GO term for negative regulation of monocyte chemotactic protein-1 (MCP-1) production, a key regulator of monocyte recruitment. Previously, predominant neutrophil rather than monocyte infiltration in rheumatoid arthritis (RA) synovial fluid was attributed to C4A.46 Additionally, neutrophil recruitment to perivascular areas of the retina has been described in autoimmune uveitis models.47,48 Similarly, our results suggest neutrophils are the infiltrated cells driving inflammation in the eyes of NR patients; however, we have not confirmed this in our patients. 
Downregulation of ABCA1 is another evidence of neutrophil activation. This protein, involved in high density lipoprotein formation through cholesterol transport,49 can attenuate neutrophil-induced inflammation in vivo and reduce cytokine production in vitro (IL-1β, IL-6, TNF-α).4952 Remarkably, ABCA1 was identified as a potential biomarker in our study, whereas IL-1β, IL-6, and TNF-α were not detected in our samples. 
Overall, our findings point to a hyperactivation of neutrophils as the underlying mechanism behind the persistent inflammation in NIU patients unresponsive to ADA, regardless of uveitis type and associated systemic disease. Increased neutrophil infiltration, hyperactivated phenotype and aberrant responses after stimulation have already been reported in autoimmune uveitis models.47,5355 Moreover, hyperactivated neutrophils have been implicated in the pathogenesis of BD,56 juvenile idiopathic arthritis (JIA),57 RA,58 and other chronic inflammatory and autoimmune diseases.59 Importantly, this hyperactivation may be a permanent rather than transient feature, because it is still observed in stages of clinical remission,60 linking genetics to this hyperactivated phenotype.61 Because anti-TNF-α treatment has been shown to partially reverse exacerbated neutrophil function in vitro,62 it is possible that this effect occurs to a lesser extent in patients with NIU unresponsive to ADA, maybe because of subtherapeutic drug levels. Consequently, hyperactivated neutrophils may drive chronic inflammation and tissue damage, creating a destructive cycle refractory to anti-TNF therapy. However, addressing whether insufficient ADA levels contribute to this mechanism remains challenging because of the lack of a defined a therapeutic range for ADA in NIU.7 
Considering that our findings suggest neutrophil involvement in autoimmune diseases, we propose that the therapeutic target of neutrophil effector functions may be beneficial for NIU patients failing anti-TNFα therapy. This underscores the importance of evidence-based approaches over subjective clinical decision-making to improve patient management. One evident option is targeting proinflammatory cytokines or its receptors, such as IL-6 receptor, whose blockade with tocilizumab has shown promise in NIU.63 Targeting downstream cytokine signaling related to neutrophil activation with Janus kinase proteins inhibitors64 like tofacitinib (NCT03580343), baricitinib (NCT04088409), and filgotinib (NCT03207815) is another treatment strategy that is currently being evaluated in NIU. Inhibiting the complement cascade with eculizumab or ravulizumab could also be beneficial, although their efficacy in NIU remains unexplored.65 Interfering with neutrophil migration/extravasation represents an interesting therapeutic approach, for example through inhibition of CXCR1 and CXCR2,53 receptors for the neutrophil chemoattractant IL-8 or overexpression of endomucin, a glycoprotein expressed by endothelial cells that has been shown to prevent neutrophil recruitment into sites of inflammation.66 Although the possibilities are broad,67 repurposed drugs are clearly the fastest way to have safe and effective treatments available at a lower cost.68 
We acknowledge that cross-sectional design, relatively small sample size and heterogeneous ADA treatment duration are some limitations of this study. However, in the largest collection of ADA-treated patients to date, we have identified a network of modulated proteins involved in intraocular inflammation common to different uveitis types. These results need further replication, ideally in prospective studies involving NIU patients naïve to ADA, to assess the definitive value of DEF, biotinidase and ABCA1 as biomarkers in clinical practice. Nonetheless, including treatment-naïve patients remains challenging because of the disease's low prevalence. 
Conclusions
Our study is the first to use shotgun proteomics to identify modulated proteins in the tears that drive intraocular inflammation, shedding light on mechanisms underlying ADA treatment failure across various NIU types. Neutrophil hyperactivation and redox imbalance may account for the persistent eye inflammation, as evidenced by upregulation of proteins related to neutrophil recruiting, activation and proinflammatory effector functions and downregulation of proteins with antioxidant and anti-inflammatory activity in NR. DEF-1,3, biotinidase and ABCA1 were identified as potential tear biomarkers of treatment response. Additionally, IL-6, Janus kinase proteins, and the complement cascade pathways were identified as possible alternative therapeutic targets. Our work demonstrates that DIA/SWATH proteomics enables protein profiling in small and complex samples like tears, thus achieving a deeper molecular characterization of ophthalmic pathologies. This suggests the potential value of tears as a source of biomarkers and novel therapeutic targets for personalized medicine. Our findings require further validation in larger cohorts of patients to better understand the processes involved in non-response to anti-TNFα. 
Acknowledgments
The authors thank Susana Belen Bravo, the head of the Proteomic Unit from Health Research Institute of Santiago de Compostela (IDIS) in Spain, and all the team members, for the LC-MS/MS data acquisition. The authors are incredibly grateful to the patients and their families for their invaluable participation. 
Supported by Instituto de Salud Carlos III (ISCIII)–PI20/00719 co-funded by European Union (UE). L. Rodriguez-Martinez was supported by Xunta de Galicia and Galician Innovation Agency (GAIN) through the post-doctoral fellowship “Programa de axudas á etapa posdoutoral” (IN606B-2023/004) and gratefully acknowledges financial support by the Fulbright U.S. Student Program, which is sponsored by the U.S. Department of State and the U.S.-Spain Fulbright Commission. Its contents are solely the responsibility of the author and do not necessarily represent the official views of the Fulbright Program, the Government of the United States, or the U.S.-Spain Fulbright Commission. J. Mateos acknowledges the support of Xunta de Galicia (GAIN) through the Senior Talent Research grant (11_IN858A_2021_1141142). A. Fernandez-Ferreiro acknowledges the support of the Instituto de Salud Carlos III (ISCIII) research grant JR18/00014. 
Disclosure: L. Rodriguez-Martinez, None; C.A. Rodriguez-Fernandez, None; O. Rodríguez Lemos, None; B. de Domingo, None; P. García Bru, None; J. Mateos, None; A. Fernandez-Ferreiro, None 
References
Joltikov KA, Lobo-Chan AM. Epidemiology and risk factors in non-infectious uveitis: a systematic review. Front Med. 2021; 8: 695904. [CrossRef]
Thorne JE, Suhler E, Skup M, et al. Prevalence of noninfectious uveitis in the United States: a claims-based analysis. JAMA Ophthalmol. 2016; 134: 1237. [CrossRef] [PubMed]
Tsirouki T, Dastiridou A, Symeonidis C, et al. A focus on the epidemiology of uveitis. Ocul Immunol Inflamm. 2018; 26: 2–16. [CrossRef] [PubMed]
You C, Sahawneh H, Ma L, Kubaisi B, Schmidt A, Foster S. A review and update on orphan drugs for the treatment of noninfectious uveitis. Clin Ophthalmol. 2017; 11: 257–265. [CrossRef] [PubMed]
Nguyen QD, Merrill PT, Jaffe GJ, et al. Adalimumab for prevention of uveitic flare in patients with inactive non-infectious uveitis controlled by corticosteroids (VISUAL II): a multicentre, double-masked, randomised, placebo-controlled phase 3 trial. Lancet. 2016; 388(10050): 1183–1192. [CrossRef] [PubMed]
Llorenç V, Cordero-Coma M, Blanco-Esteban A, et al. Drug retention rate and causes of discontinuation of adalimumab in uveitis. Ophthalmology. 2020; 127: 814–825. [CrossRef] [PubMed]
Busto-Iglesias M, Rodríguez-Martínez L, Rodríguez-Fernández CA, et al. Perspectives of therapeutic drug monitoring of biological agents in non-infectious uveitis treatment: a review. Pharmaceutics. 2023; 15: 766. [CrossRef] [PubMed]
Ponzini E, Santambrogio C, De Palma A, Mauri P, Tavazzi S, Grandori R. Mass spectrometry-based tear proteomics for noninvasive biomarker discovery. Mass Spectrom Rev. 2022; 41(5): 842–860. [CrossRef] [PubMed]
Zhou L, Zhao SZ, Koh SK, et al. In-depth analysis of the human tear proteome. J Proteomics. 2012; 75(13): 3877–3885. [CrossRef] [PubMed]
Aass C, Norheim I, Eriksen EF, Thorsby PM, Pepaj M. Single unit filter-aided method for fast proteomic analysis of tear fluid. Anal Biochem. 2015; 480: 1–5. [CrossRef] [PubMed]
Liang A, Qin W, Zhang M, Gao F, Zhao C, Gao Y. Profiling tear proteomes of patients with unilateral relapsed Behcet's disease-associated uveitis using data-independent acquisition proteomics. Peer J. 2020; 8: e9250. [CrossRef] [PubMed]
Velez G, Bassuk AG, Colgan D, Tsang SH, Mahajan VB. Therapeutic drug repositioning using personalized proteomics of liquid biopsies. JCI Insight. 2017; 2(24): e97818. [CrossRef] [PubMed]
Bansal R, Gupta A. Protein Biomarkers in Uveitis. Front Immunol. 2020; 11: 610428. [CrossRef] [PubMed]
Hagan S, Martin E, Enríquez-de-Salamanca A. Tear fluid biomarkers in ocular and systemic disease: potential use for predictive, preventive and personalised medicine. EPMA J. 2016; 7(1): 15. [CrossRef] [PubMed]
Standardization of uveitis nomenclature for reporting clinical data. results of the First International Workshop. Am J Ophthalmol. 2005; 140: 509–516. [CrossRef] [PubMed]
Nussenblatt RB, Palestine AG, Chan CC, Roberge F. Standardizatlon of vitreal inflammatory activity in intermediate and posterior uveitis. Ophthalmology. 1985; 92: 467–471. [CrossRef] [PubMed]
Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42: 377–381. [CrossRef] [PubMed]
Gijs M, Arumugam S, Van De Sande N, et al. Pre-analytical sample handling effects on tear fluid protein levels. Sci Rep. 2023; 13: 1317. [CrossRef] [PubMed]
Rao Yu-Qing, Huang Yue, Li Jing, et al. Impact of collection methods and storage conditions on the recovery of tear proteins. Guoji Yanke Zazhi Int Eye Sci. 2018; 18: 226–230.
Toja-Camba FJ, García-Quintanilla L, Rodríguez-Martinez L, et al. Enhancing therapeutic drug monitoring in inflammatory bowel disease: a comparative analysis of rapid point-of-care infliximab, adalimumab and anti-drug antibodies’ determination against ELISA. Pharmaceutics. 2023; 15: 2615. [CrossRef] [PubMed]
Tang D, Chen M, Huang X, et al. SRplot: free online platform for data visualization and graphing. Yin Y, ed. Plos One. 2023; 18(11): e0294236. [CrossRef] [PubMed]
O'Leary PC, Terrile M, Bajor M, et al. Peroxiredoxin-1 protects estrogen receptor α from oxidative stress-induced suppression and is a protein biomarker of favorable prognosis in breast cancer. Breast Cancer Res. 2014; 16(4): R79. [CrossRef] [PubMed]
Allocati N, Masulli M, Di Ilio C, Federici L. Glutathione transferases: substrates, inhibitors and pro-drugs in cancer and neurodegenerative diseases. Oncogenesis. 2018; 7(1): 1–15. [CrossRef] [PubMed]
Chen Y, Thompson DC, Koppaka V, Jester JV, Vasiliou V. Ocular aldehyde dehydrogenases: protection against ultraviolet damage and maintenance of transparency for vision. Prog Retin Eye Res. 2013; 33: 28–39. [CrossRef] [PubMed]
Schaer CA, Deuel JW, Bittermann AG, et al. Mechanisms of haptoglobin protection against hemoglobin peroxidation triggered endothelial damage. Cell Death Differ. 2013; 20: 1569–1579. [CrossRef] [PubMed]
Hajam YA, Rani R, Ganie SY, et al. Oxidative stress in human pathology and aging: molecular mechanisms and perspectives. Cells. 2022; 11: 552. [CrossRef] [PubMed]
Fukuoka A, Matsushita K, Morikawa T, et al. Human cystatin SN is an endogenous protease inhibitor that prevents allergic rhinitis. J Allergy Clin Immunol. 2019; 143: 1153–1162.e12. [CrossRef] [PubMed]
Kerkhoff C, Nacken W, Benedyk M, Dagher MC, Sopalla C, Doussiere J. The arachidonic acid-binding protein S100A8/A9 promotes NADPH oxidase activation by interaction with p67phox and Rac-2. FASEB J. 2005; 19: 467–469. [CrossRef] [PubMed]
Roh JS, Sohn DH. Damage-associated molecular patterns in inflammatory diseases. Immune Netw. 2018; 18(4): e27. [CrossRef] [PubMed]
Scott NR, Swanson RV, Al-Hammadi N, et al. S100A8/A9 regulates CD11b expression and neutrophil recruitment during chronic tuberculosis. J Clin Invest. 2020; 130: 3098–3112. [CrossRef] [PubMed]
Guo Q, Zhao Y, Li J, et al. Induction of alarmin S100A8/A9 mediates activation of aberrant neutrophils in the pathogenesis of COVID-19. Cell Host Microbe. 2021; 29: 222–235. [CrossRef] [PubMed]
Wang Y, Zhang Z, Zhang L, et al. S100A8 promotes migration and infiltration of inflammatory cells in acute anterior uveitis. Sci Rep. 2016; 6: 36140. [CrossRef] [PubMed]
Wang YQ, Dai XD, Ran Y, et al. Circulating S100A8/A9 Levels Reflect Intraocular Inflammation in Uveitis Patients. Ocul Immunol Inflamm. 2020; 28: 133–141. [CrossRef] [PubMed]
Walscheid K, Heiligenhaus A, Holzinger D, et al. Elevated S100A8/A9 and S100A12 serum levels reflect intraocular inflammation in juvenile idiopathic arthritis-associated uveitis: results from a pilot study. Invest Ophthalmol Vis Sci. 2015; 56: 7653. [CrossRef] [PubMed]
Wang S, Song R, Wang Z, Jing Z, Wang S, Ma J. S100A8/A9 in Inflammation. Front Immunol. 2018; 9: 1298. [CrossRef] [PubMed]
Rubel C, Fernández GC, Dran G, Bompadre MB, Isturiz MA, Palermo MS. Fibrinogen promotes neutrophil activation and delays apoptosis. J Immunol. 2001; 166: 2002–2010. [CrossRef] [PubMed]
Rubel C, Gómez S, Fernández GC, Isturiz MA, Caamaño J, Palermo MS. Fibrinogen-CD11b/CD18 interaction activates the NF-κB pathway and delays apoptosis in human neutrophils. Eur J Immunol. 2003; 33: 1429–1438. [CrossRef] [PubMed]
Smiley ST, King JA, Hancock WW. Fibrinogen stimulates macrophage chemokine secretion through toll-like receptor 4. J Immunol. 2001; 167: 2887–2894. [CrossRef] [PubMed]
Becatti M, Emmi G, Silvestri E, et al. Neutrophil activation promotes fibrinogen oxidation and thrombus formation in Behçet disease. Circulation. 2016; 133: 302–311. [CrossRef] [PubMed]
Xu D, Lu W. Defensins: a double-edged sword in host immunity. Front Immunol. 2020; 11: 764. [CrossRef] [PubMed]
Groeneveld TWL, Ramwadhdoebé TH, Trouw LA, et al. Human neutrophil peptide-1 inhibits both the classical and the lectin pathway of complement activation. Mol Immunol. 2007; 44: 3608–3614. [CrossRef] [PubMed]
Hou S, Qi J, Liao D, et al. Copy number variations of complement component C4 are associated with Behçet's disease but not with ankylosing spondylitis associated with acute anterior uveitis. Arthritis Rheum. 2013; 65: 2963–2970. [CrossRef] [PubMed]
Biffl WL. Interleukin-6 Delays Neutrophil Apoptosis. Arch Surg. 1996; 131: 24. [CrossRef] [PubMed]
Borish L, Rosenbaum R, Albury L, Clark S. Activation of neutrophils by recombinant interleukin 6. Cell Immunol. 1989; 121: 280–289. [CrossRef] [PubMed]
Tsuruta T, Yamamoto T, Matsubara S, et al. Novel function of C4a anaphylatoxin. Release from monocytes of protein which inhibits monocyte chemotaxis. Am J Pathol. 1993; 142: 1848–1857. [PubMed]
Matsubara S, Yamamoto T, Tsuruta T, Takagi K, Kambara T. Complement C4-derived monocyte-directed chemotaxis-inhibitory factor. A molecular mechanism to cause polymorphonuclear leukocyte-predominant infiltration in rheumatoid arthritis synovial cavities. Am J Pathol. 1991; 138: 1279–1291. [PubMed]
Yang JM, Yun K, Jeon J, et al. Multimodal evaluation of an interphotoreceptor retinoid-binding protein-induced mouse model of experimental autoimmune uveitis. Exp Mol Med. 2022; 54: 252–262. [CrossRef] [PubMed]
Deeg CA, Kaspers B, Gerhards H, Thurau SR, Wollanke B, Wildner G. Immune responses to retinal autoantigens and peptides in equine recurrent uveitis. Invest Ophthalmol Vis Sci. 2001; 42: 393–398. [PubMed]
Yang H, Zheng S, Qiu Y, et al. Activation of Liver X Receptor Alleviates Ocular Inflammation in Experimental Autoimmune Uveitis. Invest Ophthalmol Vis Sci. 2014; 55: 2795. [CrossRef] [PubMed]
Birrell MA, Catley MC, Hardaker E, et al. Novel role for the liver X nuclear receptor in the suppression of lung inflammatory responses. J Biol Chem. 2007; 282: 31882–31890 [CrossRef] [PubMed]
Dai C, Yao X, Vaisman B, et al. ATP-binding cassette transporter 1 attenuates ovalbumin-induced neutrophilic airway inflammation. Am J Respir Cell Mol Biol. 2014; 51: 626–636. [CrossRef] [PubMed]
Tang C, Liu Y, Kessler PS, Vaughan AM, Oram JF. The macrophage cholesterol exporter ABCA1 functions as an anti-inflammatory receptor. J Biol Chem. 2009; 284: 32336–32343. [CrossRef] [PubMed]
Goldberg GL, Cornish AL, Murphy J, et al. G-CSF and neutrophils are nonredundant mediators of murine experimental autoimmune uveoretinitis. Am J Pathol. 2016; 186: 172–184. [CrossRef] [PubMed]
Weigand M, Hauck SM, Deeg CA, Degroote RL. Deviant proteome profile of equine granulocytes associates to latent activation status in organ specific autoimmune disease. J Proteomics. 2021; 230: 103989. [CrossRef] [PubMed]
Hoffmann ALC, Hauck SM, Deeg CA, Degroote RL. Pre-activated granulocytes from an autoimmune uveitis model show divergent pathway activation profiles upon IL8 stimulation in vitro. Int J Mol Sci. 2022; 23: 9555. [CrossRef] [PubMed]
Sakane T, Takeno M, Suzuki N, Inaba G. Behçet's disease. N Engl J Med. 1999; 341: 1284–1291. [CrossRef] [PubMed]
Metzemaekers M, Malengier-Devlies B, Yu K, et al. Synovial fluid neutrophils from patients with juvenile idiopathic arthritis display a hyperactivated phenotype. Arthritis Rheumatol. 2021; 73: 875–884. [CrossRef] [PubMed]
Khandpur R, Carmona-Rivera C, Vivekanandan-Giri A, et al. NETs are a source of citrullinated autoantigens and stimulate inflammatory responses in rheumatoid arthritis. Sci Transl Med. 2013; 5(178): 178ra40. [CrossRef] [PubMed]
Herrero-Cervera A, Soehnlein O, Kenne E. Neutrophils in chronic inflammatory diseases. Cell Mol Immunol. 2022; 19: 177–191. [CrossRef] [PubMed]
Brown RA, Henderlight M, Do T, et al. Neutrophils from children with systemic juvenile idiopathic arthritis exhibit persistent proinflammatory activation despite long-standing clinically inactive disease. Front Immunol. 2018; 9: 2995. [CrossRef] [PubMed]
Jarvis JN, Jiang K, Frank MB, et al. Gene expression profiling in neutrophils from children with polyarticular juvenile idiopathic arthritis. Arthritis Rheum. 2009; 60: 1488–1495. [CrossRef] [PubMed]
Parackova Z, Zentsova I, Horvath R, et al. Immunomodulation of neutrophils and platelets by TNF blockage in patients with juvenile idiopathic arthritis. Clin Immunol. 2022; 245: 109170. [CrossRef] [PubMed]
Lopalco G, Fabiani C, Sota J, et al. IL-6 blockade in the management of non-infectious uveitis. Clin Rheumatol. 2017; 36: 1459–1469. [CrossRef] [PubMed]
Tanaka Y, Luo Y, O'Shea JJ, Nakayamada S. Janus kinase-targeting therapies in rheumatology: a mechanisms-based approach. Nat Rev Rheumatol. 2022; 18: 133–145. [CrossRef] [PubMed]
Muldur S, Vadysirisack DD, Ragunathan S, et al. Human neutrophils respond to complement activation and inhibition in microfluidic devices. Front Immunol. 2021; 12: 777932. [CrossRef] [PubMed]
Zahr A, Alcaide P, Yang J, et al. Endomucin prevents leukocyte-endothelial cell adhesion and has a critical role under resting and inflammatory conditions. Nat Commun. 2016; 7: 10363. [CrossRef] [PubMed]
Németh T, Sperandio M, Mócsai A. Neutrophils as emerging therapeutic targets. Nat Rev Drug Discov. 2020; 19: 253–275. [CrossRef] [PubMed]
Xia Y, Sun M, Huang H, Jin WL. Drug repurposing for cancer therapy. Signal Transduct Target Ther. 2024; 9: 92. [CrossRef] [PubMed]
Figure 1.
 
Schematic representation of the steps involved in the proteomic analysis. (1) Tear sample collection with microcapillary tubes and storage, (2) protein quantification, (3) running of protein samples into gel and excision of protein bands, (4) in-gel digestion of proteins with trypsin, (5) peptide analysis by DIA/SWATH, and (6) data analysis.
Figure 1.
 
Schematic representation of the steps involved in the proteomic analysis. (1) Tear sample collection with microcapillary tubes and storage, (2) protein quantification, (3) running of protein samples into gel and excision of protein bands, (4) in-gel digestion of proteins with trypsin, (5) peptide analysis by DIA/SWATH, and (6) data analysis.
Figure 2.
 
Ocular fundus imaging of right and left eye of a patient with a good response to ADA. Color retinography (upper image) and autofluorescence (lower image) shown no active lesions despite severe affectation of the posterior pole in left eye. Both anterior chamber cell grade and vitreous haze grade were 0.
Figure 2.
 
Ocular fundus imaging of right and left eye of a patient with a good response to ADA. Color retinography (upper image) and autofluorescence (lower image) shown no active lesions despite severe affectation of the posterior pole in left eye. Both anterior chamber cell grade and vitreous haze grade were 0.
Figure 3.
 
Volcano plots of DEPs in tear samples from NR eyes compared to R eyes to ADA treatment. The ratio of protein expression in NR versus R eyes is displayed in the x-axis as the log2 and the −log10 of the P value obtained from the differential abundance testing is displayed in the y-axis. Therefore proteins displayed at log2 values >0 are more abundant in NR eyes and those displayed at values <0 are more abundant in R eyes. Significantly modulated proteins are displayed at −log10 (P value) >1.3 corresponding to a P value = 0.05. Proteins excluded from the Uniprot database or outside our predefined criteria for DEPs are not shown.
Figure 3.
 
Volcano plots of DEPs in tear samples from NR eyes compared to R eyes to ADA treatment. The ratio of protein expression in NR versus R eyes is displayed in the x-axis as the log2 and the −log10 of the P value obtained from the differential abundance testing is displayed in the y-axis. Therefore proteins displayed at log2 values >0 are more abundant in NR eyes and those displayed at values <0 are more abundant in R eyes. Significantly modulated proteins are displayed at −log10 (P value) >1.3 corresponding to a P value = 0.05. Proteins excluded from the Uniprot database or outside our predefined criteria for DEPs are not shown.
Figure 4.
 
STRING Interaction Network and GO analysis of representative DEPs in the nonresponder and the responder eye subgroups. (A) Interaction network analysis. Proteins are depicted as nodes (circles), and lines represent the interactions between any two proteins either direct (physical) or indirect (functional) in nature. Upregulated proteins in NR eyes are shown in red, downregulated proteins are shown in blue. (B and (C) GO enrichment analysis. The bubble plot displays the biological processes (B) in which DEPs participate (those upregulated and downregulated in NR) at the top, and the cellular components (C) where they are located.
Figure 4.
 
STRING Interaction Network and GO analysis of representative DEPs in the nonresponder and the responder eye subgroups. (A) Interaction network analysis. Proteins are depicted as nodes (circles), and lines represent the interactions between any two proteins either direct (physical) or indirect (functional) in nature. Upregulated proteins in NR eyes are shown in red, downregulated proteins are shown in blue. (B and (C) GO enrichment analysis. The bubble plot displays the biological processes (B) in which DEPs participate (those upregulated and downregulated in NR) at the top, and the cellular components (C) where they are located.
Figure 5.
 
Relationship between the prevalence of NR and R eyes with protein abundance expressed in increasing order from the lowest (25th) to the highest (75th) quartile. Statistical analysis was done using χ2 test for trend (*P < 0.05, **P < 0.01, ***P < 0.001).
Figure 5.
 
Relationship between the prevalence of NR and R eyes with protein abundance expressed in increasing order from the lowest (25th) to the highest (75th) quartile. Statistical analysis was done using χ2 test for trend (*P < 0.05, **P < 0.01, ***P < 0.001).
Figure 6.
 
Potential tear protein biomarkers for treatment response to ADA based on receiver operating characteristic curve analysis.
Figure 6.
 
Potential tear protein biomarkers for treatment response to ADA based on receiver operating characteristic curve analysis.
Table 1.
 
Clinical Characteristics of NIU Patients Included in the Study
Table 1.
 
Clinical Characteristics of NIU Patients Included in the Study
Table 2.
 
Altered Tear Proteins in Patients With NIU Treated With ADA
Table 2.
 
Altered Tear Proteins in Patients With NIU Treated With ADA
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×