Prerpints.org logo
Preprint
Article

This version is not peer-reviewed.

Precision in Tear Fluid Biomarker Discovery: Quantitative Proteomic Profiling of Small-Volume, Individual Samples Using Capillary Tube Collection

A peer-reviewed version of this preprint was published in:
Biomedicines 2025, 13(2), 386. https://doi.org/10.3390/biomedicines13020386

Submitted:

26 December 2024

Posted:

26 December 2024

You are already at the latest version

Abstract

Background/Objectives: Tear fluid, a complex biofluid that contains thousands of proteins and can be collected non-invasively, has emerged as a promising source of biomarkers for ocular and systemic health. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is currently the primary method for discovering novel biomarkers in tear fluid. However, the method of tear collection can significantly impact LC-MS/MS analysis outcomes. Tear fluid is commonly collected using either Schirmer strips or capillary tubes. While capillary tubes offer distinct advantages, such as reduced extracellular contamination and reflex tearing, most LC-MS/MS protocol development has focused on optimizing protocols for Schirmer strips. This study addresses this gap by evaluating digestion protocols for tear fluid collected with capillary tubes, focusing on biomarker discovery using small-volume samples. In this study, we evaluated multiple digestion protocols for the shotgun quantitative LC-MS/MS analysis of small-volume tear fluid samples collected using glass capillary tubes. Protocol optimization was performed using pooled samples and then compared with the analysis of individual samples. With the optimized protocol, we identified an average of 368 ± 87 proteins in pooled samples and 502 ± 127 proteins in individual small-volume tear fluid samples, using less than 1µL of total fluid volume. This protocol highlights the practicality of using glass capillary tubes for comprehensive LC-MS/MS-based tear proteomics analysis, paving the way for detailed proteomics characterization of individual tear fluid samples rather than pooled samples. By shifting from pooled to individual samples, this approach greatly accelerates tear biomarker discovery, advancing precision and personalized medicine.

Keywords: 
tear proteomics
;  
tear biomarker
;  
biomarker discovery
;  
capillary collection
;  
LC-MS/MS
Subject: 
Biology and Life Sciences  -   Biology and Biotechnology

1. Introduction

Tear fluid is a complex biological matrix that is comprised of proteins, lipids, metabolites, and small molecules that support ocular health and function [1,2]. Basal tear fluid is continuously replenished, and this fluid helps to maintain the ocular surface by providing lubrication, essential nutrients, and immunological protection [3]. Basal tear proteins are primarily produced by the lacrimal and meibomian glands, with contributions from ocular surface epithelial cells and blood plasma filtrates [4,5]. This composition results in a highly diverse proteome, with thousands of proteins that exhibit distinct protein signatures in response to various ocular [6,7,8,9] and systemic diseases [10]. Tear fluid is a highly attractive target for point-of-care diagnostic applications because the collection of tear fluid can be done non-invasively and requires minimal training [11].
However, to use tear fluid as a reliable reservoir for diagnostic testing, strategies for improving the reproducibility and consistency of fluid collection need to be identified. Challenges include the limited volume of available tear fluid on the ocular surface, the dynamic composition of basal tear fluid—which can be altered by reflex tearing during collection—and the instability of the tear fluid layer in patients with severe dry eye or in contact lens wearers.
The most common method of tear collection used in proteomic studies is Schirmer strips. Schirmer strips are small, standardized filter paper strips placed on the lower eyelid to collect tear fluid. They are commonly used to assess tear production by measuring the wetted length after a set period, and the absorbed fluid can also be analyzed for biomarkers. They are widely used for tear fluid collection because they are simple to administer, are easy to obtain, and can absorb a significant amount of tear fluid [16,17]. However, there are several drawbacks to the use of Schirmer strips for proteomic analysis. First, the contact of the Schirmer strip with the eyelid during collection induces reflex tearing [14,15]. Reflex tears are produced by the lacrimal gland in response to mechanical stimulation or eye irritation. This reflex of tears can dilute or alter the composition of the tear fluid sample [12,13]. Additionally, contact with the lid increases the risk of introducing cellular contamination [18]. Collection with Schirmer strips can also be difficult to standardize, as variations in tear flow and technique can impact results. Sample collection is typically controlled by timing the duration the strip is left in place or setting a preferred wetting length, but these methods may introduce variability, particularly in patients with very low or high tear production.
The second most common method of tear collection is glass microcapillary tubes. Glass capillary tubes are less likely to cause cellular contamination—a critical factor in proteomic analysis where cellular debris can interfere with results. Additionally, glass capillary tubes are generally less invasive and more comfortable for patients compared to Schirmer strips, making them a better option in clinical settings where patient comfort is a priority [11]. Despite these advantages, the methods for processing and analyzing tear fluid samples collected with glass capillary tubes remain underdeveloped compared to those used for Schirmer strips [19]. This is because the analysis of capillary samples presents several unique challenges for proteomic discovery. The small volumes typically obtained through capillary tubes, often less than one microliter, pose challenges for downstream processing steps such as protein quantification, desalting, and enzymatic digestion. These limitations can result in protein loss, increased variability, and reduced sensitivity in liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis [11].
The absence of standardized protocols for processing tear fluid collected by capillary tubes exacerbates these issues, as sample preparation methods optimized for larger fluid volumes, like those obtained with Schirmer strips, may not be directly applicable to the smaller samples obtained from capillary tubes [20]. Therefore, there is a critical need for the development of optimized and standardized protocols specifically tailored to the unique properties of tear fluid collected by capillary tubes [21]. Addressing these challenges is essential for realizing the full potential of tear fluid proteomics in the discovery and validation of novel biomarkers for ocular and systemic diseases. To address this, our study developed an optimized digestion protocol for proteomic analysis of tear fluid collected with glass capillary tubes, aiming to enhance biomarker discovery by improving protein identification and reducing technical variation.

2. Materials and Methods

2.1. Sample Collection

Tear fluid samples were collected using 0.5μL glass microcapillary tubes (Microcaps® Disposable Micropipettes, Drummond, PA, USA) (Figure 1A) from human volunteers. Sample collection was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki, with approval from the relevant institutional review board (IRB number Pro00072069) and informed consent obtained from all participants. The microcapillary tube was placed into the lower conjunctival sac of the eye. This area is located between the lower eyelid and the eye’s surface. The insertion is done gently and without contacting the lid to minimize reflex tearing and sample contamination by intracellular proteins, as well as to minimize discomfort. No anesthetic was applied before collection. Samples were stored at -80 °C until processing.
Tear fluid samples were extracted from the glass capillary tubes through low-speed centrifugation. Capillary tubes were placed in an Eppendorf tube and centrifuged at 200 g x 2 minutes. This method extracted the tear fluid without flushing, which can cause sample dilution.
For method optimization experiments, samples were pooled to create a single uniform sample. Fifty-three samples, collected from thirty-five individuals (mean age = 63 ± 18 years; 25 males, 27 females), were combined to form the pooled sample. The pooled sample was then divided into thirty 1μL aliquots for processing. This pooled sample served as a standardized material for assessing the number of proteins identified and the consistency of results among the six protocols, providing a robust foundation for protocol evaluation and comparison.
Additional samples were collected from fourteen healthy control subjects for individual analysis (mean age = 53 ± 23 years; 2 males, 12 females).

2.2. In-Solution Protein Digestion

Using the pooled and aliquoted tear samples (1 μL each) prepared above, we designed and systematically evaluated six tear proteomics protocols. These protocols were specifically developed to test combinations of varying digestion volumes and protein-to-trypsin ratios (Figure 1B). In addition, we established a streamlined in-solution digestion method tailored to the unique characteristics of tear fluid samples (Figure 1C).
The protein-to-trypsin ratio and total digest volume were varied across six protocols: (a) (50:1, 250μL), (b) (20:1, 250μL), (c) (50:1, 125μL), (d) (20:1, 125μL), (e) (50:1, 50μL), and (f) (20:1, 50μL) (Figure 1B & Figure 1C). Protein samples were prepared and digested as follows: 10μL of 8M urea in 50mM ammonium bicarbonate (pH 8) (Millipore Sigma, Catalog No. 09830-500G) was added to each sample tube. The mixture was vortexed and then centrifuged at 3,000 × g for 30 seconds. The 8M urea solution was prepared by dissolving 480mg of urea in 50mM ammonium bicarbonate and adjusting the final volume to 1mL. Protein concentration was determined using a Nanodrop spectrophotometer at 280nm. To break disulfide bonds, 1μL of 500 mM dithiothreitol (DTT) (Millipore Sigma, Catalog No. 3860-5GM) was added to the sample, followed by vortexing, centrifugation at 3,000 × g for 30 seconds, and incubation at 56 °C for 30 minutes. Alkylation was achieved by adding 1μL of 750mM iodoacetamide (IAA) (Millipore Sigma, Catalog No. I6125-5G), vortexing, centrifuging at 3,000 × g for 30 seconds, and incubating in the dark at room temperature for 45 minutes. The 750mM IAA solution was prepared by dissolving 13.9mg of IAA in 100μL of HPLC-grade water. Alkylation was subsequently quenched with 1μL of 500mM DTT, followed by vertexing, centrifugation, and incubation at room temperature for an additional 30 minutes. The sample was then diluted with 50mM ammonium bicarbonate to reach the total digest volume specified by each protocol, vortexed, and centrifuged. For digestion, Trypsin/Lys-C mix (Mass Spec Grade, Promega, Catalog No. V5071) was added according to the protein-trypsin ratio specified by each protocol. A trypsin stock solution was prepared by dissolving 20μg of trypsin in 100μL of 50mM ammonium bicarbonate to achieve a concentration of 0.2μg/μL and stored at -30 °C. Samples were incubated with trypsin at 37 °C for 16 hours. Digestion was stopped by adding 10% formic acid (LC-MS grade, ThermoFisher, Catalog No. 28905) to adjust the pH to 3-4, as verified with pH strips, and the digested sample was stored at 4 °C.

2.3. LC-MS/MS

MS raw data were collected using a timsTOF Pro 2 mass spectrometer (Bruker) online coupled with an Evosep One liquid chromatography system (Evosep). An 8 cm Evosep performance column (EV-1109) was used for peptide separation. The tear peptide samples were analyzed using a standard data-dependent acquisition method with parallel accumulation-serial fragmentation (DDA-PASEF) and the instrument operating in the positive mode. A full scan was first acquired with a mass range of 100-1700 m/z and a TIMS 1/k0 range of 0.60-1.60 V·s/cm2. During a full 1.17-second cycle, 10 PASEF ramps were performed, with ramp and accumulation times both set to 100ms. The following DDA-PASEF parameters were applied: precursor intensity threshold: 2.5x103; charge state: +1 to +5; dynamic exclusion: 0.4 min; target intensity for fragmentation: 2 x104; isolation window (linear): 2 m/z at 700 m/z and 3 m/z at 800 m/z; collision energy (linear): 20 eV at1/k0 of 0.60 V·s/cm2 and 59 eV at 1/k0 of 1.60 V·s/cm2.

2.4. Protein Identification and Analysis

All DDA raw datasets were searched against the Uniprot-reviewed human protein database (downloaded 3/30/2023) using MaxQuant (version 2.1, Max-Planck-Institute of Biochemistry, Germany). Peptide-spectrum matches (PSMs) were evaluated using the MaxQuant score method, reflecting the quality of the match between the observed spectrum and the theoretical peptide. Peptides were validated with a minimum score of 20 for unmodified peptides, and 40 for modified peptides. Additionally, a minimum delta score of 6 was applied for modified peptides. A False Discovery Rate (FDR) of 1% (0.01) is applied at both the PSM and protein levels to control false positives. Trypsin activity was set to be fully specific (cleaving peptide bonds on the carboxyl side of lysine (K) and arginine (R), except when either of these residues is followed by proline (P)) with two missed cleavage sites allowed. A maximum of two of the following dynamic modifications were allowed on each peptide: C [+57.021464], STY [+79.966331], and M [+15.994915]. Precursor and fragment mass tolerances were set to ± 10 ppm and ± 25 ppm, respectively. For protein identification, at least one associated peptide was identified within ± 5 ppm of its theoretical m/z value.

2.5. Statistical Analysis

The Kruskal-Wallis test was applied to evaluate overall differences in the average protein and peptide counts across the six protocols. When significant differences were observed, the Mann-Whitney U test was used for pairwise comparisons to identify specific protocols with significant differences in the number of identified proteins and peptides. To assess run-to-run variance—each run involving the processing of a single 1 μL aliquot of the pooled sample—we calculated the coefficient of variation (CV%) within each protocol. CV% was determined only for proteins that were identified in every run across all six protocols. The intra-protocol CV% for each protein was then compared using a paired t-test. A similar method for comparing variance was used to compare the variance of protein identified in the HC individual samples and the pooled sample runs for protocol (f). A t-test was used to compare the number of proteins identified in pooled runs for protocol f, and the individual samples.

2.6. Selection of Optimal Protocol

To select the optimal protocol for proteomic analysis, a comparison was made using standardized evaluation criteria. These included (i) maximizing protein and peptide yield, (ii) improving the detection of unique proteins, (iii) achieving higher detection frequency across samples, and (iv) minimizing variability in results (Figure 2). The protocol demonstrating the best overall performance across these criteria was selected for its ability to provide sensitive, consistent, and reproducible proteomic data suitable for biomarker discovery and clinical applications.

3. Results

3.1. Protein and Peptide Yield

A total of 30 processing runs were completed, with five runs conducted for each of the six experimental protocols. These runs identified 758 proteins in total. On average, protocol (f) identified the greatest number of proteins, with 368 ± 87 proteins detected per run, while protocol (a) identified the fewest proteins, with 248 ± 28 proteins (Figure 3A & Table 1). Statistical analysis revealed that protocol (a) identified a significantly lower number of proteins compared to protocols (e) (p = 0.001) and (f) (p = 0.003). However, no statistically significant differences in the number of proteins identified were observed between any of the other protocols. In terms of peptide identification, protocol (f) again performed the best, detecting an average of 1270 ± 314 peptides per run, while protocol (a) detected the fewest, with 835 ± 181 peptides (Figure 3B & Table 1). However, no statistically significant difference in the number of peptides identified was observed between protocols (p = 0.38). Protocol (f) exhibited superior performance in maximizing protein and peptide yield compared to the other protocols tested.

3.2. Unique Protein Identification

Of the proteins identified, 319 were consistently identified in at least one run per protocol, with 73 proteins reliably detected across every single run. Notably, protocol (f) emerged as the most effective in terms of the number of proteins identified, yielding 72 proteins exclusively detected using this protocol (Figure 4). This result highlights protocol (f)’s ability to maximize the proteomic coverage of tear samples, likely due to its optimized digestion volume and protein-to-trypsin ratio.

3.3. Frequency of Detection

Proteins identified in this study were categorized into four groups based on their frequency of detection across all runs in each protocol: high (present in > 75% of runs), medium (present in 50-74% of runs), low (present in 25-49% of runs), and rare (present in <25% of runs) (Figure 5 & Table 2). This classification provides critical insights into the consistency and reproducibility of protein identification across different protocols and experimental conditions. Among the identified proteins, 181 exhibited a high frequency of detection across all runs, with 73 of these proteins consistently identified in every run across all six protocols.
Such analyses are essential because identifying high-frequency proteins ensures confidence in their detection and establishes their potential as reliable candidates for further study. High-frequency proteins, especially those consistently detected across all runs, are more likely to represent core components of the tear proteome or proteins resistant to experimental variability. Notably, protocol (f) provided the largest number of high-frequency protein identification at t = 244, followed by (e) (t = 242), (d) (t = 220), (c) (t = 218), (b) (t = 213), and (a) (t = 174) (Figure 5 and Table 2). The superior performance of protocol (f) in identifying proteins with high frequency highlights its potential for achieving broader proteome coverage and greater reproducibility. By enabling the detection of proteins across a wide range of abundance levels, protocol (f) stands out as an optimal protocol for comprehensive proteomics analyses.

3.4. Protocol Variance

Understanding the variance between protocols is crucial to assess the reproducibility and consistency of protein identification across multiple runs, which directly impacts reliability. Protocol (c), observed to have the highest average run-to-run variance at 5.20 ± 2.52, exhibited significantly higher variance when compared to protocols (b) (4.06 ± 2.64, p < 0.001), (d) (4.19 ± 2.12, p < 0.001) and (f) (4.36 ± 3.26, p = 0.02) (Figure 6 & Table 3). No significant difference in variance was observed between the other protocols. Protocol (e) had the lowest average variance at 3.96 ± 3.15.

3.5. Protocol Selection

Selecting the optimal protocol requires balancing reproducibility with the ability to detect a wide range of proteins, including low-abundance and unique proteins. Despite not having the lowest variance, protocol (f) (with a total dilution volume of 50 µL and a 20:1 protein-to-trypsin ratio) emerged as the optimal choice for in-solution digestion due to several key advantages: (i) it identified the highest number of proteins and peptides, (ii) detected the greatest number of unique protein groups, and (iii) achieved the highest frequency of protein identification across all runs. Although protocol (f) did not exhibit the lowest run-to-run variance, its variance was not significantly higher than that of other protocols. Given its overall performance, protocol (f) provides the most comprehensive protein identification and is well-suited for further analyses. The reliability and effectiveness of protocol (f) make it the preferred option for tear proteomics.

3.6. Comparison of Pooled and Individual Samples

To evaluate the performance of protocol (f) in real tear proteomics studies, individual tear samples were processed using this optimized protocol and compared to the results from processing pooled tear fluid samples. A total of 1,010 unique proteins were identified in the individual samples. On average, 502 ± 127 proteins were identified across the fourteen samples; this was significantly greater than the 368 ± 87 identified on average when processing pooled samples using the same protocol (p = 0.013) (Figure 7A). In the individual sample runs, 297 proteins were identified with a high frequency (>75%), which is comparable to the number of high-frequency proteins found in the pooled sample (Figure 7B). Similar run-to-run variance was observed in both the pooled and individual sample runs. For many of the proteins consistently identified across both conditions, the individual samples didn’t exhibit significantly greater variation in run-to-run intensity than observed in the pooled samples (p = 0.302). This finding highlights that protocol (f) is a reliable and robust protocol for tear proteomics. Moreover, this protocol does not introduce significantly higher sample-to-sample variance relative to pooled samples (Figure 7C). Consequently, protocol (f) represents the optimized choice for tear proteomics and biomarker discovery.

3.5. Gene Ontology (GO) Enrichment Analysis and Tear Protein Classification

GO enrichment analysis was conducted to categorize the proteins identified in tear fluid samples based on their biological processes, cellular components, and molecular functions (Figure 8). The biological process analysis revealed key functional roles of tear fluid proteins in innate and adaptive immunity, tissue homeostasis, and inflammation regulation. These findings underscore the importance of tear fluid in pathogen defense and immune response regulation [69]. Other processes, such as lipid metabolism, cell adhesion, and inflammatory response, underscore its involvement in wound healing and tissue maintenance. Cellular component analysis linked many proteins to secreted elements, cytoplasm, cell membranes, and immunoglobulin complexes, reinforcing tear fluid’s role in immune defense and cellular signaling. Molecular function analysis showed a predominance of hydrolase and protease activities, essential for enzymatic degradation and tissue remodeling, along with oxidoreductase, actin-binding, and protease inhibitor functions, reflecting tear fluid’s role in oxidative stress regulation and structural support [60]. Notably, lactotransferrin (LTF), lysozyme (LYZ), and prolactin-inducible protein (PIP), along with immunoglobulins, were highly expressed, showcasing tear fluid’s antimicrobial and anti-inflammatory proteins [70,71,72]. Albumin, a blood protein, indicates tear fluid partly derives from blood filtrate, suggesting it may contain systemic disease biomarkers [5].

4. Discussion

Tear fluid is a rich source of protein biomarkers for both ocular and system diseases. Its non-invasive nature and ease of point-of-care collection make tear biomarkers a novel and promising approach for disease screening, diagnosis, and monitoring. LC-MS/MS-based proteomics is a powerful tool for identifying tear fluid biomarkers. However, the method of tear sample collection can significantly affect the results of shotgun proteomics analysis. Most LC-MS/MS-based proteomics analyses of tear fluid have been optimized for Schirmer strips [7,10,17,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51], with a limited focus on optimizing methods for glass microcapillary tubes [1,26,53,54,55,56,57,58,59,60,61,62,63,64]. Existing methods for capillary tubes typically require larger sample volumes (5–10 µL), which can trigger significant reflex tearing, altering the sample composition and potentially reducing test reliability. Alternatively, some methods use pooled tear fluid samples, which result in the loss of information on individual differences in tear fluid composition [1,8,26,53,54,55,56,58,59,61,62,64]. On average, across LC-MS/MS-based proteomics studies of tear fluid, Schirmer strips identify a larger number of proteins compared to studies employing glass capillary tubes—approximately 502 unique protein identifications on average for Schirmer strip studies compared to 140 for capillary studies [52]. These tubes, which mimic microfluidic processes used in point-of-care devices [65] offer several advantages over Schirmer strips [11,18]. Despite the potential benefits, the disparity in protein identification between capillary and Schirmer analysis methods significantly limits the utility of capillary collection for shot-gun discovery experiments. To date, more than 3,000 unique proteins have been identified in tear fluid, as such, the low levels of protein identification in current methods have a high likelihood of missing potential biomarkers. For shotgun proteomics analysis, the optimization of protocols to detect a high number of unique proteins reliably greatly improves the chances of identifying novel markers of disease. To gain a comprehensive understanding of the tear fluid proteome, it is essential to optimize collection protocols for smaller-volume glass capillary tubes to enable single-collection analysis of tear fluid samples while minimizing cellular contamination and reflex tearing.
In this study, we developed an improved method for in-solution digestion of individual tear fluid samples collected using small-volume glass capillaries through the systematic optimization of digestion protocols. Protocol (f) was identified as the optimal method because it identified the highest number of proteins per run on average while maintaining low run-to-run variance. Using this method, we detected 1,065 unique proteins in tear fluid across 30 pooled sample runs and 14 individual HC sample runs. We detected an average of 502 proteins in the analysis of individual healthy control samples from a total volume of only 1uL. This method significantly increases the number of unique proteins identified in capillary samples compared to the literature while also reducing the sample volume required for analysis. As tear fluid samples are known to be dynamic in composition at large volumes due to reflex tearing and the basal tear layer has an average volume of 3–10μL, the application of this method reduces the impact of reflex tearing on sample analysis. This method additionally demonstrated high run-to-run reproducibility, and is efficient, requiring only 120 minutes of processing time without the need for fractionation.
This study further demonstrates the feasibility of “single-tear” proteomic analysis using LC-MS/MS for the identification of tear fluid biomarkers. In this study, we observed a higher number of protein identifications, with a similar run-to-run variance, between pooled samples and individual samples processed using identical protocols. The increased number of protein identifications observed in individual samples can be attributed to the varying dynamic ranges and differing protein concentrations across samples [66,67]. Processing low-volume individual tear fluid samples should be preferred in exploratory research, as it more reliably detects low-abundance proteins and preserves individual variation between subjects.
Despite this improvement in analysis, the number of proteins detected remains lower compared to Schirmer strip methods, which have reported detecting as many as 1543 proteins in a single study [68]. The higher protein yield from Schirmer strips may be attributed to greater sample volumes and potential cellular contamination. Future work should focus on directly comparing capillary and Schirmer methods to understand the differences better and further optimize the capillary approach. This will help determine when one method should be preferred over the other.
This approach expands the scope of proteomics studies by aligning sample analysis with collection methods suitable for diagnostic devices. Given the significant variations in tear fluid proteomics based on collection methods, it is crucial to establish reliable techniques for analyzing tear fluid across different sampling approaches. This will provide a more comprehensive understanding of the tear fluid proteome. Additionally, optimizing capillary processing techniques will improve comparisons between Schirmer and capillary sampling methods and enhance understanding of how sampling impacts analysis. This is essential for translating early discovery work into commercial assays and point-of-care diagnostics.

Supplementary Materials

The following supporting information can be downloaded at the website of this paper posted on Preprints.org.

Author Contributions

K.F.: writing—original draft preparation, conceptualization, formal analysis, visualization. Y.F.: investigation, conceptualization, methodology. M.A.B.: writing—review and editing. K.G.: writing—review and editing, validation, supervision, methodology. J.K.: writing—review and editing. D.B.: review, editing, supervision. L.T.L.: conceptualization, writing, review and editing, supervision. K.X.: conceptualization, methodology, supervision, writing—review and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This study is partially supported by Prostate Cancer Foundation Challenge Award (to KX, 2023CHAL4223), the PA State Formula Grant (to KX, SAP #: 4100095527), and funding from The Pittsburgh Foundation, Claude-Worthington Benedum Foundation and NIH 1R21OD034476-01.

Institutional Review Board Statement

The study was approved by Advarra IRB (number Pro00072069).

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Acknowledgments

We thank Jackie Sikora of Sewickley Eye Group for assistance with tear sample collection. This work was performed as part of the AHNCI Moonshot Biomarker Program and supported in par by Highmark Health.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Li, N.; Wang, N.; Zheng, J.; Liu, X.M.; Lever, O.W.; Erickson, P.M.; Li, L. Characterization of Human Tear Proteome Using Multiple Proteomic Analysis Techniques. J. Proteome Res. 2005, 4, 2052–2061. [Google Scholar] [CrossRef] [PubMed]
  2. 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, 1–20. [Google Scholar] [CrossRef] [PubMed]
  3. Murube, J. Basal, Reflex, and Psycho-emotional Tears. Ocul. Surf. 2009, 7, 60–66. [Google Scholar] [CrossRef] [PubMed]
  4. Bron, A.J.; Tomlinson, A.; Foulks, G.N.; Pepose, J.S.; Baudouin, C.; Geerling, G.; Nichols, K.K.; Lemp, M.A. Rethinking Dry Eye Disease: A Perspective on Clinical Implications. Ocul. Surf. 2014, 12, S1–S31. [Google Scholar] [CrossRef] [PubMed]
  5. D. A. Dartt, R. R. Hodges, and D. Zoukhri, “Tears and Their Secretion,” in Advances in Organ Biology, vol. 10, J. Fischbarg, Ed., in The Biology of the Eye, vol. 10. Elsevier, 2005, pp. 21–82. [CrossRef]
  6. Aluru, S.V.; Agarwal, S.; Srinivasan, B.; Iyer, G.K.; Rajappa, S.M.; Tatu, U.; Padmanabhan, P.; Subramanian, N.; Narayanasamy, A. Lacrimal Proline Rich 4 (LPRR4) Protein in the Tear Fluid Is a Potential Biomarker of Dry Eye Syndrome. PLOS ONE 2012, 7, e51979. [Google Scholar] [CrossRef]
  7. Srinivasan, S.; Thangavelu, M.; Zhang, L.; Green, K.B.; Nichols, K.K. iTRAQ Quantitative Proteomics in the Analysis of Tears in Dry Eye Patients. Investig. Opthalmology Vis. Sci. 2012, 53, 5052–9. [Google Scholar] [CrossRef]
  8. Csősz. ; Boross, P.; Csutak, A.; Berta, A.; Tóth, F.; Póliska, S.; Török, Z.; Tőzsér, J. Quantitative analysis of proteins in the tear fluid of patients with diabetic retinopathy. J. Proteom. 2012, 75, 2196–2204. [Google Scholar] [CrossRef]
  9. Kim, H.-J.; Kim, P.-K.; Yoo, H.-S.; Kim, C.-W. Comparison of tear proteins between healthy and early diabetic retinopathy patients. Clin. Biochem. 2012, 45, 60–67. [Google Scholar] [CrossRef]
  10. Boerger, M.; Funke, S.; Leha, A.; Roser, A.-E.; Wuestemann, A.-K.; Maass, F.; Bähr, M.; Grus, F.; Lingor, P. Proteomic analysis of tear fluid reveals disease-specific patterns in patients with Parkinson's disease – A pilot study. Park. Relat. Disord. 2019, 63, 3–9. [Google Scholar] [CrossRef]
  11. R. J. Fullard and D. L. Tucker, “Evaluation of commonly used tear sampling methods and their relevance in subsequent biochemical analysis,” Invest. Ophthalmol. Vis. Sci., vol. 32, no. 8, pp. 2290–2301, Jul. 1991.
  12. Fullard, R.; Snyder, C. PROTEIN-LEVELS IN NONSTIMULATED AND STIMULATED TEARS OF NORMAL HUMAN-SUBJECTS. 1990, 31, 1119–1126.
  13. Fullard, R.J.; Tucker, D.L. Changes in human tear protein levels with progressively increasing stimulus. Invest. Ophthalmol. Vis. Sci 1991, 32, 2290–301. [Google Scholar]
  14. Krajčíková, K.; Glinská, G.; Tomečková, V. Effect of tear fluid sampling and processing on total protein quantity and electrophoretic pattern. Taiwan J. Ophthalmol. 2022, 12, 88–92. [Google Scholar] [CrossRef] [PubMed]
  15. Tham, M.L.; Mahmud, A.; Abdullah, M.; Saleh, R.M.; Razali, A.M.; Cheah, Y.K.; Taib, N.M.; Ho, K.L.; Mahmud, M.; Isa, M.M. Tear Samples for Protein Extraction: Comparative Analysis of Schirmer's Test Strip and Microcapillary Tube Methods. Cureus 2023, 15, e50972. [Google Scholar] [CrossRef] [PubMed]
  16. Nättinen, J.; Aapola, U.; Nukareddy, P.; Uusitalo, H. Clinical Tear Fluid Proteomics—A Novel Tool in Glaucoma Research. Int. J. Mol. Sci. 2022, 23, 8136. [Google Scholar] [CrossRef] [PubMed]
  17. Jones, G.; Lee, T.J.; Glass, J.; Rountree, G.; Ulrich, L.; Estes, A.; Sezer, M.; Zhi, W.; Sharma, S.; Sharma, A. Comparison of Different Mass Spectrometry Workflows for the Proteomic Analysis of Tear Fluid. Int. J. Mol. Sci. 2022, 23, 2307. [Google Scholar] [CrossRef]
  18. Vergouwen, D.P.C.; Schotting, A.J.; Endermann, T.; van de Werken, H.J.G.; Grashof, D.G.B.; Arumugam, S.; Nuijts, R.M.M.A.; Berge, J.C.T.; Rothova, A.; Schreurs, M.W.J.; et al. Evaluation of pre-processing methods for tear fluid proteomics using proximity extension assays. Sci. Rep. 2023, 13, 1–8. [Google Scholar] [CrossRef]
  19. Bachhuber, F.; Huss, A.; Senel, M.; Tumani, H. Diagnostic biomarkers in tear fluid: from sampling to preanalytical processing. Sci. Rep. 2021, 11, 1–9. [Google Scholar] [CrossRef]
  20. J. Xiao et al., “Tear fluid proteomic analysis with improved LC-MS/MS protocol,” Invest. Ophthalmol. Vis. Sci., vol. 65, no. 7, p. 6557, Jun. 2024.
  21. Zhan, X.; Li, J.; Guo, Y.; Golubnitschaja, O. Mass spectrometry analysis of human tear fluid biomarkers specific for ocular and systemic diseases in the context of 3P medicine. EPMA J. 2021, 12, 449–475. [Google Scholar] [CrossRef]
  22. Bolaños-Jiménez, R.; Navas, A.; López-Lizárraga, E.P.; de Ribot, F.M.; Peña, A.; O Graue-Hernández, E.; Garfias, Y. Ocular Surface as Barrier of Innate Immunity. Open Ophthalmol. J. 2015, 9, 49–55. [Google Scholar] [CrossRef]
  23. Green-Church, K.B.; Nichols, K.K.; Kleinholz, N.M.; Zhang, L.; Nichols, J.J. Investigation of the human tear film proteome using multiple proteomic approaches. 2008, 14, 456–470.
  24. Huang, Z.; Du, C.-X.; Pan, X.-D. The use of in-strip digestion for fast proteomic analysis on tear fluid from dry eye patients. PLOS ONE 2018, 13, e0200702. [Google Scholar] [CrossRef]
  25. Lépine, M.; Zambito, O.; Sleno, L. Targeted Workflow Investigating Variations in the Tear Proteome by Liquid Chromatography Tandem Mass Spectrometry. ACS Omega 2023, 8, 31168–31177. [Google Scholar] [CrossRef]
  26. 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. 2021, 41, 842–860. [Google Scholar] [CrossRef]
  27. Rossi, C.; Cicalini, I.; Cufaro, M.C.; Agnifili, L.; Mastropasqua, L.; Lanuti, P.; Marchisio, M.; De Laurenzi, V.; Del Boccio, P.; Pieragostino, D. Multi-Omics Approach for Studying Tears in Treatment-Naïve Glaucoma Patients. Int. J. Mol. Sci. 2019, 20, 4029. [Google Scholar] [CrossRef] [PubMed]
  28. Tong, L.; Zhou, X.Y.; Jylha, A.; Aapola, U.; Liu, D.N.; Koh, S.K.; Tian, D.; Quah, J.; Uusitalo, H.; Beuerman, R.W.; et al. Quantitation of 47 human tear proteins using high resolution multiple reaction monitoring (HR-MRM) based-mass spectrometry. J. Proteom. 2014, 115, 36–48. [Google Scholar] [CrossRef] [PubMed]
  29. Nättinen, J.; Jylhä, A.; Aapola, U.; Parkkari, M.; Mikhailova, A.; Beuerman, R.W.; Uusitalo, H. Patient stratification in clinical glaucoma trials using the individual tear proteome. Sci. Rep. 2018, 8, 12038. [Google Scholar] [CrossRef]
  30. Gerber-Hollbach, N.; Plattner, K.; O'Leary, O.E.; Jenoe, P.; Moes, S.; Drexler, B.; Schoetzau, A.; Halter, J.P.; Goldblum, D. Tear Film Proteomics Reveal Important Differences Between Patients With and Without Ocular GvHD After Allogeneic Hematopoietic Cell Transplantation. Investig. Opthalmology Vis. Sci. 2018, 59, 3521–3530. [Google Scholar] [CrossRef]
  31. Pieragostino, D.; Agnifili, L.; Fasanella, V.; D'Aguanno, S.; Mastropasqua, R.; Di Ilio, C.; Sacchetta, P.; Urbani, A.; Del Boccio, P. Shotgun proteomics reveals specific modulated protein patterns in tears of patients with primary open angle glaucoma naïve to therapy. Mol. Biosyst. 2013, 9, 1108–1116. [Google Scholar] [CrossRef]
  32. Böhm, D.; Keller, K.; Pieter, J.; Boehm, N.; Wolters, D.; Siggelkow, W.; Lebrecht, A.; Schmidt, M.; Kölbl, H.; Pfeiffer, N.; et al. Comparison of tear protein levels in breast cancer patients and healthy controls using a de novo proteomic approach. Oncol. Rep. 2012, 28, 429–438. [Google Scholar] [CrossRef]
  33. López-López, M.; Regueiro, U.; Bravo, S.B.; Chantada-Vázquez, M.d.P.; Varela-Fernández, R.; Ávila-Gómez, P.; Hervella, P.; Lema, I. Tear Proteomics in Keratoconus: A Quantitative SWATH-MS Analysis. Investig. Opthalmology Vis. Sci. 2021, 62, 30–30. [Google Scholar] [CrossRef]
  34. Li, B.; Sheng, M.; Li, J.; Yan, G.; Lin, A.; Li, M.; Wang, W.; Chen, Y. Tear proteomic analysis of Sjögren syndrome patients with dry eye syndrome by two-dimensional-nano-liquid chromatography coupled with tandem mass spectrometry. Sci. Rep. 2014, 4, srep05772. [Google Scholar] [CrossRef]
  35. Kishazi, E.; Dor, M.; Eperon, S.; Oberic, A.; Hamedani, M.; Turck, N. Thyroid-associated orbitopathy and tears: A proteomics study. J. Proteom. 2018, 170, 110–116. [Google Scholar] [CrossRef]
  36. Angeles-Han, S.T.; Yeh, S.; Patel, P.; Duong, D.; Jenkins, K.; Rouster-Stevens, K.A.; Altaye, M.; Fall, N.; Thornton, S.; Prahalad, S.; et al. Discovery of tear biomarkers in children with chronic non-infectious anterior uveitis: a pilot study. J. Ophthalmic Inflamm. Infect. 2018, 8, 17. [Google Scholar] [CrossRef] [PubMed]
  37. 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. PeerJ 2020, 8, e9250. [Google Scholar] [CrossRef] [PubMed]
  38. Pieragostino, D.; Bucci, S.; Agnifili, L.; Fasanella, V.; D'Aguanno, S.; Mastropasqua, A.; Ciancaglini, M.; Mastropasqua, L.; Di Ilio, C.; Sacchetta, P.; et al. Differential protein expression in tears of patients with primary open angle and pseudoexfoliative glaucoma. Mol. Biosyst. 2011, 8, 1017–1028. [Google Scholar] [CrossRef]
  39. Dor, M.; Eperon, S.; Lalive, P.H.; Guex-Crosier, Y.; Hamedani, M.; Salvisberg, C.; Turck, N. Investigation of the global protein content from healthy human tears. Exp. Eye Res. 2018, 179, 64–74. [Google Scholar] [CrossRef]
  40. O'Leary, O.E.; Schoetzau, A.; Amruthalingam, L.; Geber-Hollbach, N.; Plattner, K.; Jenoe, P.; Schmidt, A.; Ullmer, C.; Drawnel, F.M.; Fauser, S.; et al. Tear Proteomic Predictive Biomarker Model for Ocular Graft Versus Host Disease Classification. Transl. Vis. Sci. Technol. 2020, 9, 3–3. [Google Scholar] [CrossRef]
  41. Zhou, L.; Beuerman, R.W.; Chan, C.M.; Zhao, S.Z.; Li, X.R.; Yang, H.; Tong, L.; Liu, S.; Stern, M.E.; Tan, D. Identification of Tear Fluid Biomarkers in Dry Eye Syndrome Using iTRAQ Quantitative Proteomics. J. Proteome Res. 2009, 8, 4889–4905. [Google Scholar] [CrossRef]
  42. Aluru, S.V.; Shweta, A.; Bhaskar, S.; Geetha, K.; Sivakumar, R.M.; Utpal, T.; Padmanabhan, P.; Angayarkanni, N. Tear Fluid Protein Changes in Dry Eye Syndrome Associated with Rheumatoid Arthritis: A Proteomic Approach. Ocul. Surf. 2016, 15, 112–129. [Google Scholar] [CrossRef]
  43. Aass, C.; Norheim, I.; Eriksen, E.F.; Thorsby, P.M.; Pepaj, M. Single unit filter-aided method for fast proteomic analysis of tear fluid. Anal. Biochem. 2015, 480, 1–5. [Google Scholar] [CrossRef]
  44. López-López, M.; Regueiro, U.; Bravo, S.B.; Chantada-Vázquez, M.d.P.; Pena, C.; Díez-Feijoo, E.; Hervella, P.; Lema, I. Shotgun Proteomics for the Identification and Profiling of the Tear Proteome of Keratoconus Patients. Investig. Opthalmology Vis. Sci. 2022, 63, 12–12. [Google Scholar] [CrossRef]
  45. Winiarczyk, M.; Kaarniranta, K.; Winiarczyk, S.; Adaszek, .; Winiarczyk, D.; Mackiewicz, J. Tear film proteome in age-related macular degeneration. Graefe's Arch. Clin. Exp. Ophthalmol. 2018, 256, 1127–1139. [Google Scholar] [CrossRef]
  46. Grus, F.H.; Podust, V.N.; Bruns, K.; Lackner, K.; Fu, S.; Dalmasso, E.A.; Wirthlin, A.; Pfeiffer, N. SELDI-TOF-MS ProteinChip Array Profiling of Tears from Patients with Dry Eye. Investig. Opthalmology Vis. Sci. 2005, 46, 863–876. [Google Scholar] [CrossRef] [PubMed]
  47. Liu, Y.-C.; Yam, G.H.-F.; Lin, M.T.-Y.; Teo, E.; Koh, S.-K.; Deng, L.; Zhou, L.; Tong, L.; Mehta, J.S. Comparison of tear proteomic and neuromediator profiles changes between small incision lenticule extraction (SMILE) and femtosecond laser-assisted in-situ keratomileusis (LASIK). 2020, 29, 67–81. [CrossRef]
  48. Pieragostino, D.; Lanuti, P.; Cicalini, I.; Cufaro, M.C.; Ciccocioppo, F.; Ronci, M.; Simeone, P.; Onofrj, M.; van der Pol, E.; Fontana, A.; et al. Proteomics characterization of extracellular vesicles sorted by flow cytometry reveals a disease-specific molecular cross-talk from cerebrospinal fluid and tears in multiple sclerosis. J. Proteom. 2019, 204, 103403. [Google Scholar] [CrossRef] [PubMed]
  49. D'Souza, S.; Petznick, A.; Tong, L.; Hall, R.C.; Rosman, M.; Chan, C.; Koh, S.K.; Beuerman, R.W.; Zhou, L.; Mehta, J.S. Comparative Analysis of Two Femtosecond LASIK Platforms Using iTRAQ Quantitative Proteomics. Investig. Opthalmology Vis. Sci. 2014, 55, 3396–3402. [Google Scholar] [CrossRef]
  50. Vaajanen, A.; Nättinen, J.; Aapola, U.; Gielen, F.; Uusitalo, H. The effect of successful trabeculectomy on the ocular surface and tear proteomics—a prospective cohort study with 1-year follow-up. Acta Ophthalmol. 2020, 99, 160–170. [Google Scholar] [CrossRef] [PubMed]
  51. Funke, S.; Beck, S.; Lorenz, K.; Kotterer, M.; Wolters, D.; Perumal, N.; Pfeiffer, N.; Grus, F.H. Analysis of the effects of preservative-free tafluprost on the tear proteome. Am. J. Transl. Res. 2016, 8, 4025–4039. [Google Scholar]
  52. Wong, T.T.; Zhou, L.; Li, J.; Tong, L.; Zhao, S.Z.; Li, X.R.; Yu, S.J.; Koh, S.K.; Beuerman, R.W. Proteomic Profiling of Inflammatory Signaling Molecules in the Tears of Patients on Chronic Glaucoma Medication. Investig. Opthalmology Vis. Sci. 2011, 52, 7385–7391. [Google Scholar] [CrossRef] [PubMed]
  53. Balasubramanian, S.A.; Wasinger, V.C.; Pye, D.C.; Willcox, M.D.P. Preliminary identification of differentially expressed tear proteins in keratoconus. 2013, 19, 2124–2134.
  54. Acera, A.; Vecino, E.; Rodríguez-Agirretxe, I.; Aloria, K.; Arizmendi, J.M.; Morales, C.; A Durán, J. Changes in tear protein profile in keratoconus disease. Eye 2011, 25, 1225–1233. [Google Scholar] [CrossRef]
  55. Ananthi, S.; Prajna, N.V.; Lalitha, P.; Valarnila, M.; Dharmalingam, K. Pathogen Induced Changes in the Protein Profile of Human Tears from Fusarium Keratitis Patients. PLOS ONE 2013, 8, e53018. [Google Scholar] [CrossRef]
  56. Soria, J.; Acera, A.; Merayo-Lloves, J.; Durán, J.A.; González, N.; Rodriguez, S.; Bistolas, N.; Schumacher, S.; Bier, F.F.; Peter, H.; et al. Tear proteome analysis in ocular surface diseases using label-free LC-MS/MS and multiplexed-microarray biomarker validation. Sci. Rep. 2017, 7, 17478–17478. [Google Scholar] [CrossRef]
  57. Lee, S.H.; Park, M.-Y.; Kim, K.W.; Wee, S.W.; Kim, J.C. Zinc Finger Protein in Severe Dry Eye Syndrome. Curr. Eye Res. 2013, 39, 431–438. [Google Scholar] [CrossRef]
  58. Funke, S.; Azimi, D.; Wolters, D.; Grus, F.H.; Pfeiffer, N. Longitudinal analysis of taurine induced effects on the tear proteome of contact lens wearers and dry eye patients using a RP-RP-Capillary-HPLC–MALDI TOF/TOF MS approach. J. Proteom. 2012, 75, 3177–3190. [Google Scholar] [CrossRef] [PubMed]
  59. Ananthi, S.; Chitra, T.; Bini, R.; Prajna, N.V.; Lalitha, P.; Dharmalingam, K. Comparative analysis of the tear protein profile in mycotic keratitis patients. 2008, 14, 500–507.
  60. Yenihayat, F.; Altıntaş, .; Kasap, M.; Akpınar, G.; Güzel, N.; Çelik, O.S. Comparative proteome analysis of the tear samples in patients with low-grade keratoconus. Int. Ophthalmol. 2017, 38, 1895–1905. [Google Scholar] [CrossRef]
  61. A de Souza, G.; de Godoy, L.M.; Mann, M. Identification of 491 proteins in the tear fluid proteome reveals a large number of proteases and protease inhibitors. Genome Biol. 2006, 7, R72–R72. [Google Scholar] [CrossRef] [PubMed]
  62. Ponzini, E.; Ami, D.; Duse, A.; Santambrogio, C.; De Palma, A.; Di Silvestre, D.; Mauri, P.; Pezzoli, F.; Natalello, A.; Tavazzi, S.; et al. Single-Tear Proteomics: A Feasible Approach to Precision Medicine. Int. J. Mol. Sci. 2021, 22, 10750. [Google Scholar] [CrossRef] [PubMed]
  63. Nichols, J.J.; Green-Church, K.B. Mass Spectrometry-Based Proteomic Analyses in Contact Lens-Related Dry Eye. Cornea 2009, 28, 1109–1117. [Google Scholar] [CrossRef]
  64. Kalló, G.; Emri, M.; Varga, Z.; Ujhelyi, B.; Tőzsér, J.; Csutak, A.; Csősz. Changes in the Chemical Barrier Composition of Tears in Alzheimer’s Disease Reveal Potential Tear Diagnostic Biomarkers. PLOS ONE 2016, 11, e0158000. [Google Scholar] [CrossRef]
  65. Chen, X.; Rao, J.; Zheng, Z.; Yu, Y.; Lou, S.; Liu, L.; He, Q.; Wu, L.; Sun, X. Integrated Tear Proteome and Metabolome Reveal Panels of Inflammatory-Related Molecules via Key Regulatory Pathways in Dry Eye Syndrome. J. Proteome Res. 2019, 18, 2321–2330. [Google Scholar] [CrossRef]
  66. Technology overview | TearLab osmolarity system for diagnosing dry eye disease | Advice | NICE.” Accessed: Sep. 17, 2024. [Online]. Available: https://www.nice.org.uk/advice/mib47/chapter/technology-overview.
  67. Ritorto, M.S.; Borlak, J. A simple and reliable protocol for mouse serum proteome profiling studies by use of two-dimensional electrophoresis and MALDI TOF/TOF mass spectrometry. Proteome Sci. 2008, 6, 25–25. [Google Scholar] [CrossRef]
  68. Millioni, R.; Tolin, S.; Puricelli, L.; Sbrignadello, S.; Fadini, G.P.; Tessari, P.; Arrigoni, G. High Abundance Proteins Depletion vs Low Abundance Proteins Enrichment: Comparison of Methods to Reduce the Plasma Proteome Complexity. PLOS ONE 2011, 6, e19603. [Google Scholar] [CrossRef]
  69. Zhou, L.; Zhao, S.Z.; Koh, S.K.; Chen, L.; Vaz, C.; Tanavde, V.; Li, X.R.; Beuerman, R.W. In-depth analysis of the human tear proteome. J. Proteom. 2012, 75, 3877–3885. [Google Scholar] [CrossRef]
Preprints 144180 g001
Figure 1. Collection and processing protocols for tear fluid samples. (A) Collection of tear fluid with glass capillary tubes is non-invasive and requires minimal training. 0.5 µL glass capillary tubes were used for all collections. (B) The protein-to-trypsin ratio and total digest volume for the six protocols evaluated in this study: (a) (50:1, 250μL), (b) (20:1, 250μL), (c) (50:1, 125μL), (d) (20:1, 125μL), (e) (50:1, 50μL), and (f) (20:1, 50μL). Five runs were performed for each protocol, with each run consisting of processing a single 1 μL aliquot of the pooled sample. (C) Experimental procedures for protein denaturation, reduction, alkylation, and trypsin digestion. Protocols tested differed in steps 3 and 4 as shown. In step 3, the total digestion volume was varied. In step 4, the protein: trypsin ratio was modified. The glass capillary with 0.5μL tear sample was transferred into a Protein LoBind microcentrifuge tube, where the entire procedure was carried out to minimize sample loss and reduce errors associated with handling small volumes. Low-speed centrifugation was used to extract the tear sample from the capillary tubes without breaking.
Figure 1. Collection and processing protocols for tear fluid samples. (A) Collection of tear fluid with glass capillary tubes is non-invasive and requires minimal training. 0.5 µL glass capillary tubes were used for all collections. (B) The protein-to-trypsin ratio and total digest volume for the six protocols evaluated in this study: (a) (50:1, 250μL), (b) (20:1, 250μL), (c) (50:1, 125μL), (d) (20:1, 125μL), (e) (50:1, 50μL), and (f) (20:1, 50μL). Five runs were performed for each protocol, with each run consisting of processing a single 1 μL aliquot of the pooled sample. (C) Experimental procedures for protein denaturation, reduction, alkylation, and trypsin digestion. Protocols tested differed in steps 3 and 4 as shown. In step 3, the total digestion volume was varied. In step 4, the protein: trypsin ratio was modified. The glass capillary with 0.5μL tear sample was transferred into a Protein LoBind microcentrifuge tube, where the entire procedure was carried out to minimize sample loss and reduce errors associated with handling small volumes. Low-speed centrifugation was used to extract the tear sample from the capillary tubes without breaking.
Preprints 144180 g001
Preprints 144180 g002
Figure 2. Evaluation criteria for method optimization and selection. The goals of method optimization are to maximize protein and peptide yield, improve the detection of unique proteins, improve the frequency of detection for proteins, and minimize variance.
Figure 2. Evaluation criteria for method optimization and selection. The goals of method optimization are to maximize protein and peptide yield, improve the detection of unique proteins, improve the frequency of detection for proteins, and minimize variance.
Preprints 144180 g002
Preprints 144180 g003
Figure 3. Comparison of identified protein and peptide numbers across six different digestion protocols. (A) Comparison of identified protein numbers across six protocols. Protocol (f) yielded the highest average number of unique proteins per run. (B) Comparison of identified peptide numbers across six protocols. Protocol (f) also identified the highest average number of peptides per run.
Figure 3. Comparison of identified protein and peptide numbers across six different digestion protocols. (A) Comparison of identified protein numbers across six protocols. Protocol (f) yielded the highest average number of unique proteins per run. (B) Comparison of identified peptide numbers across six protocols. Protocol (f) also identified the highest average number of peptides per run.
Preprints 144180 g003
Preprints 144180 g004
Figure 4. Comparison of Unique Proteins Identified Across Six Digestion Protocols. A total of 319 proteins were consistently identified across all six protocols. Among the protocols, Protocol A demonstrated the lowest ability to identify unique proteins, detecting only one protein not found in any other protocol. In contrast, Protocol F showed the highest capability for identifying unique proteins, with 72 proteins exclusively detected by this protocol.
Figure 4. Comparison of Unique Proteins Identified Across Six Digestion Protocols. A total of 319 proteins were consistently identified across all six protocols. Among the protocols, Protocol A demonstrated the lowest ability to identify unique proteins, detecting only one protein not found in any other protocol. In contrast, Protocol F showed the highest capability for identifying unique proteins, with 72 proteins exclusively detected by this protocol.
Preprints 144180 g004
Preprints 144180 g005
Figure 5. The frequency of proteins identified across processing runs plotted against abundance, as indicated by log2(PSM). Total protein identifications, as well as the frequency of identification of proteins, varied between protocol groups. Protocol (f) demonstrated the highest number of proteins detected across all frequency categories. Most notably, it detected the greatest number of proteins with high frequency (>75%). Note: PSM, peptide-spectrum matches.
Figure 5. The frequency of proteins identified across processing runs plotted against abundance, as indicated by log2(PSM). Total protein identifications, as well as the frequency of identification of proteins, varied between protocol groups. Protocol (f) demonstrated the highest number of proteins detected across all frequency categories. Most notably, it detected the greatest number of proteins with high frequency (>75%). Note: PSM, peptide-spectrum matches.
Preprints 144180 g005
Preprints 144180 g006
Figure 6. Comparison of CV% for high abundance proteins between protocols. Protocol (c) demonstrated the highest run-to-run variance, significantly higher than those observed for protocols (b), (d), (e), and (f). No other significant differences in variance were observed.
Figure 6. Comparison of CV% for high abundance proteins between protocols. Protocol (c) demonstrated the highest run-to-run variance, significantly higher than those observed for protocols (b), (d), (e), and (f). No other significant differences in variance were observed.
Preprints 144180 g006
Preprints 144180 g007
Figure 7. Comparison of proteins identified in individual small-volume tear fluid samples versus pooled samples run using protocol (f). (A) Individual sample runs identified 425 more proteins that were not detected in the pooled samples processed using the same processing protocol. (B) Across the individual sample runs 297 proteins were identified with high frequency (>75%), which is at the same level as the number of high-frequency proteins identified in the pooled sample. (C) The run-to-run variance was similar between the pooled and individual sample processing runs.
Figure 7. Comparison of proteins identified in individual small-volume tear fluid samples versus pooled samples run using protocol (f). (A) Individual sample runs identified 425 more proteins that were not detected in the pooled samples processed using the same processing protocol. (B) Across the individual sample runs 297 proteins were identified with high frequency (>75%), which is at the same level as the number of high-frequency proteins identified in the pooled sample. (C) The run-to-run variance was similar between the pooled and individual sample processing runs.
Preprints 144180 g007
Preprints 144180 g008
Figure 8. Gene Ontology (GO) enrichment analysis of tear fluid proteins showing their involvement in biological processes, cellular components, and molecular function.
Figure 8. Gene Ontology (GO) enrichment analysis of tear fluid proteins showing their involvement in biological processes, cellular components, and molecular function.
Preprints 144180 g008
Table 1. Descriptive statistics comparing protein and peptide numbers between protocols. Note: STD, standard deviation.
Table 1. Descriptive statistics comparing protein and peptide numbers between protocols. Note: STD, standard deviation.
Protocol Peptides Proteins
Average STD Average STD
(a) 835 181 248 28
(b) 1088 381 306 28
(c) 1095 400 316 80
(d) 1085 253 311 57
(e) 1017 160 332 36
(f) 1270 314 368 87
Table 2. The number of proteins identified by each protocol is separated by the frequency of identification.
Table 2. The number of proteins identified by each protocol is separated by the frequency of identification.
Unique Proteins Identified Protocol
(a) (b) (c) (d) (e) (f)
High Abundance (>75%) 174 213 218 220 242 244
Medium Abundance (50%-75%) 65 70 73 73 74 90
Low Abundance (25%-49%) 72 88 89 85 80 133
Rare (< 25%) 95 137 170 138 138 162
Total 406 508 550 516 534 629
Table 3. CV% for selected highly abundant tear fluid proteins by protocol with descriptive statistics.
Table 3. CV% for selected highly abundant tear fluid proteins by protocol with descriptive statistics.
Protein Name (a) (b) (c) (d) (e) (f)
Lactotransferrin 1.09 1.83 1.96 1.68 1.04 1.67
Lysozyme C 2.08 1.91 2.55 2.79 1.95 2.29
Basement membrane-specific heparan sulfate proteoglycan core protein 2.05 1.35 3.56 2.30 1.45 2.25
Albumin 1.84 2.06 3.25 2.28 1.44 2.47
Mammaglobin-B 2.51 6.04 5.44 7.53 3.68 3.22
Cystatin-S 4.37 5.36 6.37 3.61 3.64 4.91
Proline-rich protein 4 3.25 4.55 4.31 5.81 1.39 2.70
Cystatin-C 5.08 5.29 8.15 5.04 7.96 6.04
Average 4.56 4.06 5.20 4.19 3.96 4.36
STD 3.47 2.64 2.52 2.12 3.15 3.26
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

facebook logotwitter logolinkedin logo
bsky
MDPI Initiatives
Important Links
Subscribe

Disclaimer

Terms of Use

Privacy Policy

Privacy Settings

© 2026 MDPI (Basel, Switzerland) unless otherwise stated