Sex Hormone Detection In Human Tear Film Using Enzyme-linked Immunosorbent Assays

Summary

This protocol describes an optimized enzyme-linked immunosorbent assay (ELISA) method to detect and quantify 17β-estradiol and testosterone in human tear fluid collected via microcapillary tubes. This approach enables minimally invasive measurement of tear hormone levels.

Abstract

The tear film is a complex, multi-layered fluid essential for maintaining ocular surface health, visual clarity, and protection against pathogens and environmental stressors. Composed primarily of lipids, aqueous fluid, soluble proteins, and mucins, it serves not only to lubricate and protect the eye but as a valuable diagnostic medium for ocular surface disorders such as dry eye disease (DED). Sex hormones, including androgens and estrogens, are known to influence tear film composition, but measurement of these hormones in tear fluid is limited by low sample volume and technical challenges. This protocol describes a step-by-step method for adapting commercially available enzyme-linked immunosorbent assay (ELISA) kits to detect and quantify 17β-estradiol and testosterone in basal tears collected using microcapillary tubes. The workflow incorporates optimized sample collection, dilution, and assay preparation strategies to enable analysis of low-volume tear samples. The protocol demonstrates the feasibility and reproducibility of hormone detection in tear fluid using adapted ELISA methods. Representative results are derived from pooled pilot samples and are intended to illustrate assay performance rather than biological differences. This approach may support future studies investigating tear fluid as a matrix for hormone measurement in ocular surface research.

Introduction

The tear film plays vital roles in protecting, nourishing, and lubricating the ocular surface¹. It is approximately 2–5.5 µm thick and 3–10 µL in volume, with a turnover rate of 1–2 µL per minute on a healthy eye^2,3. The lacrimal glands secrete the aqueous component of the tears, rich in water, electrolytes, proteins, and antimicrobial factors, while the meibomian glands produce lipids that slow evaporation of the tear film, and the conjunctival goblet cells release mucins that support lubrication of the ocular surface⁴. Tear production occurs in two main ways: 1) basal secretion, which maintains a stable tear film on a healthy ocular surface, and 2) reflex secretion, which is triggered by stimuli such as irritation, emotion, or bright lights. The biochemical composition differs between the two which can ultimately influence the method used to collect tears for analysis^5,6. Basal tears generally contain higher concentrations of proteins and lipids that support tear film stability and corneal health, whereas reflex tears tend to have higher aqueous volume and are more diluted to flush irritants and protect the ocular surface^7,8,9.

Because tears function to modulate inflammation, defend against pathogens, and supply nutrients to the avascular cornea¹, they have become a focal point for biomarker research¹⁰. Tears provide a minimally invasive and relatively cost-effective alternative to blood sampling for monitoring physiological changes, particularly those occurring near the site of disease⁸. To capture this data accurately, proper collection is essential¹⁰. Glass microcapillary tubes are widely used for sampling basal tears because they allow collection with minimal irritation and tear dilution⁵. This method minimizes the risk of stimulating reflex tearing, which can alter tear composition^9,11, by carefully positioning the tube to avoid contact with the bulbar conjunctiva and lid margin. Using this approach, researchers can obtain undiluted basal tears that reflect the physiological state of the ocular surface. This ensures the sample remains a reliable medium for quantifying sensitive analytes, such as sex hormones^10,12. However, the limited volume obtained presents analytical challenges for downstream biochemical assays. Basal tear collection using microcapillary tubes typically yields a few microliters per sample, which limits compatibility with standard assays.

The human ocular surface is increasingly recognized as a target tissue for sex hormones^13,14. Evidence supporting this includes the detection of mRNA transcripts for androgen receptors (AR), estrogen receptors (ER), and progesterone receptors (PR) in acinar cells of the lacrimal and meibomian glands, as well as in the epithelial cells of the cornea, conjunctiva, and conjunctival goblet cells¹⁵. Furthermore, mRNAs encoding steroidogenic enzymes have been identified in these tissues, suggesting the capacity to convert steroid precursors into biologically active sex hormones locally¹⁶. This indicates potential for in situ hormone synthesis at the ocular surface and raises a critical question as to whether these hormones are secreted into the tear film at detectable levels.

Current evidence suggests that sex hormones are integral to maintaining ocular surface homeostasis, though the precise molecular mechanisms remain under investigation^13,14. Androgens, such as testosterone, appear to stimulate lipid secretion and suppress inflammation; consequently, androgen deficiency is widely recognized as a driver of evaporative dry eye^13,14,17. In contrast, the role of estrogens is more nuanced and context dependent^13,14,18. While involved in modulating tear composition and volume, estrogen fluctuations, such as those occurring during menopause, pregnancy, or oral contraceptive use, are frequently associated with compromised aqueous secretion, decreased tear film stability, and ocular surface inflammation^19,20. As a result, individuals experiencing these hormonal imbalances represent a population at elevated risk for developing dry eye disease (DED)^13,14,20.

Despite this evidence, monitoring hormonal status in DED patients remains a challenge. Specifically, it is not yet established whether systemic serum hormone levels accurately reflect local hormone levels in the eye²¹. This distinction is critical because ocular surface tissues have been shown to possess sex hormone receptors and steroidogenic enzyme expression that support local synthesis and metabolism of sex steroids¹⁶. This intracrine capacity suggests that local hormone concentrations in the tear film may differ significantly from circulating serum levels¹³. Therefore, direct measurement of hormone concentration in tear fluid is necessary to better understand local hormonal effects on the ocular surface. However, such measures are technically challenging due to the small volume of tear samples available for analysis. Advanced analytical techniques such as liquid chromatography-mass spectrometry (LC-MS) have been used to quantify hormones in blood and tear fluids with high sensitivity, but this method requires specialized instrumentation and technical expertise that may not be readily accessible by all researchers^21,22.

The enzyme-linked immunosorbent assay (ELISA) is a widely established technique for the detection and quantification of proteins, antibodies, and hormones. For the analysis of small molecules such as sex steroids, this protocol utilizes a competitive binding format. In this method, the unlabeled hormone present in the tear sample competes with an enzyme-labeled hormone analog for a limited number of antibody binding sites. Because of this competition, the signal intensity, measured as optical density, is inversely proportional to the concentration of the hormone in the sample; effectively, a lower signal indicates a higher concentration of the target hormone^23,24.

ELISA offers several practical advantages for targeted quantification. It is cost-effective and technically accessible, allowing for implementation in standard laboratory settings without the need for specialized instrumentation. Furthermore, ELISA facilitates high-throughput screening in 96-well plate formats with minimal sample preparation, offering both time efficiency and high reproducibility²³. However, commercially available ELISA kits are typically optimized for higher-volume biological samples such as serum or plasma, and their performance in low-volume tear samples is not well defined. To address this limitation, the protocol incorporates optimized sample dilution and handling strategies to enable hormone detection within the assay’s sensitivity range.

In this study, we present a visualized protocol that adapts commercially available ELISA kits for the detection and quantification of 17β-estradiol and testosterone in low-volume human tear sample collected using glass microcapillary tubes. This protocol emphasizes practical modifications in sample collection, handling, and assay preparation to enable reproducible hormone measurement from limited tear volumes. This work focuses on establishing a feasible workflow for applying existing ELISA methodology to tear fluid analysis. This approach may serve as a foundation for future studies investigating local hormone levels at the ocular surface.

Protocol

This study was approved by the Indiana University Institutional Review Board (IRB#1908362633), and written informed consent was obtained from all participants prior to sample collection.

1. Tear Film Collection

1. Preparation
   1. Determine the tear volume to be collected, keeping in mind that 2 replicates are needed for statistical reliability of ELISA results.
   2. Prepare a 1.5 mL microcentrifuge tube for tear film storage.
   3. Select appropriate microcapillary tubes to collect tears. Recommend starting with 3.33 µL or 5 µL and changing to smaller size if collecting from a lower volume of tears on the ocular surface.
   4. Prepare a rubber squeeze bulb provided with the microcapillary tube kit, to facilitate transfer of tear fluid into a microcentrifuge tube.
2. Tear Collection
   1. Instruct the participant to look straight ahead or slightly upward.
   2. Gently place the microcapillary tube opening in the tear meniscus near the lateral canthus to collect basal tears.
   3. Minimize contact with bulbar conjunctiva and lid margin to avoid reflex tearing as much as possible.
   4. Fill the tube with tears until full.
   5. If multiple collection sessions are needed, schedule sessions within 24 h of each other and at similar times during the day to minimize variability of hormone levels.
3. Tear Storage
   1. Occlude the hole on the top of the rubber squeeze bulb then insert the microcapillary tube into the bulb while promptly placing the open end inside the microcentrifuge tube to let the tears flow out from the microcapillary tube. Additional squeezing of the bulb may be needed to get as much of the tear sample out of the microcapillary tube as possible.
   2. Close the microcentrifuge tube cap to prevent evaporation.
   3. Spin down the tube using a benchtop microcentrifuge device at 2,000 × g. for 3–5 s to let the sample be at the bottom of the tube before the storage.
   4. Store in -80 °C to preserve analyte stability prior to analysis.

2. Testosterone Assay

1. Prepare Testosterone Standards, follow protocol
   1. Follow the kit protocol, ensure reagents and equipment to prepare testosterone standards, controls, and samples are available.
   2. Reconstitute testosterone standard with deionized or distilled water to get a stock solution of 100 ng/mL. Ensure complete reconstitution and allow it to sit for minimum 15 min.
   3. Prepare testosterone standard on the day of experimentation.
   4. Prepare 6 microcentrifuge tubes for the standard, with 900 µL of calibrator diluent into the 10 ng/mL tube and 400 µL into the remaining tubes.
   5. Pipette 100 µL of stock solution into the first tube (10 ng/mL tube), mix thoroughly and change pipette tip for the next transfer.
   6. Pipette 200 µL of the first tube (10 ng/mL tube) into the next tube (3.33 ng/mL), mix thoroughly, change pipette tip for the next transfer and repeat the step until the last tube (0.041 ng/mL).
   7. Note: After completing all steps, there should be 7 concentrations of standards. Calibrator diluent serves as the zero standard (0 ng/mL). Each prepared standard is more than enough for two replications.
2. Prepare Testosterone controls: Low, Medium, and High Concentration
   1. Reconstitute each vial with the volume of deionized or distilled water indicated in the certificate of analysis (COA) of the control kit
   2. Dilute each control in 2-fold in the calibrator diluent prior to use in assay
3. Prepare samples
   1. Remove tear samples from -80 °C and place them on ice to thaw.
   2. Once samples are fully thawed, briefly centrifuge the tear samples to ensure all samples are at the bottom of the tubes at room temperature for 3–5 s.
   3. Mix 20 µL of sample with 80 µL of calibrator diluent to obtain a total volume of 100 µL per replicate.
   4. Plan for two replicates for the experiment for reliability and non-specific binding (NSB) wells for background control.
   5. Note: Each sample is analyzed in duplicate wells. Therefore, 40 µL of tear sample is required per participant to perform the assay, mixing with 160 µL of calibrator diluent to obtain a total volume of 200 µL.
4. Assay Procedure
   1. Prepare all reagents, working standards, tear samples, and controls.
   2. Remove the excess wells from the microplate frame and restore them in the desiccant pack in 4 °C.
   3. Add 50 µL of Primary Antibody Solution (blue-colored) to each well excluding wells designated for NSB wells. Cover the plate with aluminum foil.
   4. Incubate the plate at room temperature for 1 h on a horizontal orbital microplate shaker (0.12" orbit) set at 500 ± 50 rpm or at the closest available setting.
   5. Prepare sufficient wash buffer for a total of 8 washes, 400 µL/well/wash (4 washes before and 4 washes after the 3 h incubation step).
   6. Aspirate the contents of each well and wash the plate 4 times with wash buffer.
   7. Add 100 µL of calibrator diluent to the NSB wells and zero standard wells.
   8. Add 100 µL of standard, control, and prepared tear samples to the designated wells.
   9. Add 50 µL of the Testosterone Conjugate to each well and thoroughly mix by pipetting.
   10. Incubate at room temperature for 3 h on a horizontal orbital microplate shaker (0.12" orbit) set at 500 ± 50 rpm or at the closest available setting.
   11. Aspirate each well and wash 4 times using wash buffer.
   12. Prepare the substrate solution immediately before use by mixing Color Reagent A and Color Reagent B at a 1:1 ratio. Allocate 200 µL per well. Complete aspiration of residual wash buffer prior to substrate addition.
   13. Add 200 µL of Substrate Solution to each well within 15 min after preparation. Incubate for 30 min at room temperature on the benchtop. Protect from light by covering with aluminum foil.
   14. Add 50 µL of Stop Solution to each well in the same order as of Substrate Solution. Mix thoroughly each well and the color should change from blue to yellow.
   15. Measure the optical density at 450 nm and 560 nm within 30 min of adding the Stop Solution using a microplate reader.
   16. Note: The manufacturer’s instructions recommend a reference wavelength of 540 nm or 570 nm. In this protocol, a wavelength of 560 nm was used as the closest available option on the microplate reader, for background signal correction and optical imperfections.
5. Calculation
   1. Subtract the optical density measured at 560 nm (reference wavelength) from the optical density measured at 450 nm.
   2. Subtract the mean NSB optical density from each individual reading of standards, samples, and controls, then calculate the average of duplicate measurements.
   3. Generate a standard curve by fitting the data using four-parameter logistic (4PL) model.
   4. Determine testosterone concentrations by interpolating sample values from the standard curve.
   5. Adjust the calculated concentrations by applying a dilution factor of 5 to account for the 1:5 sample dilution.
   6. Verify the measurement by calculating coefficient of variation (CV) between replicates should not exceed 20%.
      1. Note: %CV = (SD/mean) × 100

3. Estradiol Assay

1. Prepare Estradiol Standards, Controls, and Samples
   1. Follow the kit protocol, ensure reagents and equipment are available to prepare standards, controls, and samples.
   2. Allow all reagents and materials to reach room temperature prior to use.
   3. Prepare fresh reagent on the day of the experiment.
   4. Standards and control are ready-to-use. Aliquot only the amount required for the current assay to minimize contamination and limit repeated temperature changes during handling.
   5. Ensure that six standards (No. 0–5) are prepared, with Standard 0 representing 0 pg/mL, and include one control sample provided with the kit.
   6. Thaw the frozen samples from -80 °C by putting them on ice.
   7. Spin down the tear samples to ensure all samples are at the bottom of the tube.
2. Assay Procedure
   1. Remove the excess wells from the microplate and restore them in the desiccant pack in 4 °C.
   2. Add 20 µL of standards, controls, and samples into wells.
   3. Note: each standard and sample should be assayed with 2 replicates.
   4. To ensure a homogeneous distribution, thoroughly mix all standards, controls, and samples immediately prior to pipetting into the wells.
   5. Add 160 µL of 17 beta Estradiol-HRP Conjugate in the testing wells.
   6. Leave a blank well for substrate blank.
   7. Cover the well with aluminum foil before putting it into the incubator. The solution in the wells can evaporate to the aluminum foil.
   8. Incubate at 37 °C for 2 h.
   9. Prepare Wash Buffer from 10X washing solution during the incubation time.
   10. When the incubation is completed, remove the aluminum foil and aspirate the liquid from the wells.
   11. Wash each well 3 times with Wash Buffer. Soak > 5 s in each washing cycle.
   12. Add 100 µL of TMB Substrate Solution into all wells.
   13. Cover the whole plate with aluminum foil and incubate at room temperature for exactly 30 min.
   14. Add 100 µL of Stop Solution into all wells in the same order and at the same rate as the previous added TMB Substrate Solution.
   15. Shake the microplate gently and the reagent is changing from blue to yellow.
   16. Measure the absorbance at 450 nm within 30 min of adding the Stop Solution.
3. Calculation
   1. Subtract the absorbance of the blank well from the readings of all standards, samples, and controls to correct background signal and optical imperfections.
   2. Calculate the average of duplicate measurements.
   3. Generate a best-fit standard curve using four-parameter logistic (4PL).
   4. Determine estradiol concentration by interpolating sample values from the standard curve..
   5. Verify the measurement by calculating coefficient of variation (CV) between replicates should not exceed 20%.
      1. Note: %CV = (SD/mean) × 100.

4. Data Analysis using R and RStudio

1. Install R and RStudio, and install the required packages listed in the Table of Materials prior to performing data analysis.
2. Import optical density measurements for standards, controls, and samples into the analysis software.
3. Preprocess the data by applying wavelength correction as specified by the assay. Then subtracting blank or NSB values and calculating the average of duplicate measurements.
4. Fit a standard curve using a four-parameter logistic (4PL) regression model with concentration as the independent variable and optical density as response variable.
5. Perform curve fitting in R using the drm() function from the drc package, with the model specified as fct = LL.4() for 4PL regression, with concentration and optical density data as input.
6. Verify the quality of the curve fit (e.g., R² > 0.99 or acceptable goodness-of-fit) before interpreting sample concentrations.
7. Calculate hormone concentrations by inverse prediction using the ED() function based on the fitted standard curve, specifying the effective dose corresponding to sample optical density values based on the fitted model.
8. Adjust calculated concentrations by applying the appropriate dilution factor (e.g., ×5 for a 1:5 dilution).
9. Report the final hormone concentrations in the appropriate units (e.g., ng/mL or pg/mL), ensuring consistency with the standard curve, and export the processed data for further analysis or visualization.

Results

Pilot Tear Film Study

A total of 20 women (45–63 years old) were included in this study. DED classification was based on symptoms and at least one clinical sign of DED. Specifically, all DED participants had an Ocular Surface Disease Index (OSDI) score > 12 and at least one eye with either fluorescein tear break-up time (TBUT) of < 5 s or corneal fluorescein staining score (CFSS) of > 3 (NEI Scale²⁵). Participants classified as non-DED had OSDI score < 12, TBUT of > 5 s in both eyes, and CFSS of ≤ 3 in both eyes. Exclusion criteria included current contact lens wearers, use of ophthalmic medications within 30 days, ocular infection, conjunctival scarring, and pregnancy or nursing by self-report. Basal tear samples were collected from both eyes of each participant using glass microcapillary tubes and pooled at the participant level to increase total sample volume. A single collection session yielded approximately 5–20 µL of tear fluid per participant, depending on individual tear production. Tears were collected prior to sodium fluorescein drop instillation. The optimized hormone ELISA protocols for basal tear analysis require 20 µL of tear fluid per well to achieve hormone detection. To account for potential variability, each sample was assayed in 2 replicates, necessitating at least 40 µL of tear fluid per sample. Due to individual tear volumes were often insufficient for analysis, samples were pooled across participants within similar age and clinical group (DED and non-DED) in order to obtain sufficient volume for ELISA testing.

Participants Demographics

There were 9 participants classified as non-DED and 11 participants classified as DED. A summary of participant demographics and clinical assessments are in Table 1.

Validation of Standard Curves

Successful execution of the competitive ELISA protocol is indicated by the generation of a standard curve with a high coefficient of determination (R² > 0.99). Since this is a competitive binding assay, the optical density (OD) values must exhibit an inverse relationship with hormone concentration. Figure 1 illustrates a representative standard curve for testosterone and Figure 2 for 17β- estradiol. High optical density corresponds to the zero-standard or low concentration samples, while low optical density corresponds to high hormone concentrations. The concentrations were extrapolated from the standard curve and were averaged among the replicates.

Detected Testosterone Concentration
The testosterone ELISA kit used in this study has a detection range of 0.04–10 ng/mL and an assay sensitivity of 0.041 ng/mL, with a minimum detectable dose of approximately 0.03 ng/mL (range: 0.012–0.041 ng/mL), as reported by the manufacturer. Tear film testosterone concentrations averaged 0.31 ng/mL ± 0.05 ng/mL in older women without DED and 0.56 ng/mL ± 0.1 ng/mL in older women with DED (Table 2). These values represent measurements obtained from pooled tear samples, and the reported mean ± SD reflects variability across technical assay replicates rather than independent biological samples. Assay reproducibility was assessed by calculating the coefficient of variation (%CV) for duplicate measurements, yielding values of 14.69% for non-DED samples and 18.67% for DED samples. For reference, the literature reports average systemic testosterone levels for similarly aged women to be 0.19–0.23 ng/mL²⁶.

Detected 17β-estradiol Concentration
The estradiol ELISA kit used in this study has a detection range of 20–2000 pg/mL and an assay sensitivity of 8.68 pg/mL, as reported by the manufacturer. Tear film 17β-estradiol concentrations averaged 66.59 ± 0.67 pg/mL in older women without DED and 47.07 ± 3.13 pg/mL in older women with DED (Table 2). As with testosterone measurements, these values were derived from pooled samples, and the reported variability reflects technical replicate measurements rather than biological variation across individuals. The %CV for estradiol measurements was 1.0% for non-DED samples and 6.65% for DED samples, further supporting the reproducibility of the assay under the described conditions. For reference, systemic blood levels reported in the literature typically range 50–65 pg/mL during early stages of menopausal transition and decline to below 20 pg/mL in postmenopause²⁷.

Based on this pilot data, we demonstrate that tear fluid analysis using optimized ELISA protocols successfully detected and quantified 17β-estradiol and testosterone. Because measurements were performed on pooled samples, these results should be interpreted as methodological feasibility rather than evidence of biological differences between groups. Raw optical density values, standard curve data, and calculated hormone concentrations are provided in supplementary data files.

Figure 1: Representative standard curve for testosterone quantification. A serial dilution of testosterone standards was analyzed using a competitive ELISA. Optical density was measured first at 450nm and then at 560nm for background signal correction. Data points represent the average of duplicate measurements. The curve was generated using a four-parameter logistic (4-PL) regression model, exhibiting a high coefficient of determination (R² = 0.999). Due to the competitive nature of the assay, an inverse relationship between testosterone concentration and optical density is observed. The displayed equation was used to interpolate the concentrations of tear film samples based on their respective OD values. Please click here to view a larger version of this figure.

Figure 2: Representative standard curve for 17β-estradiol quantification. A serial dilution of 17β-estradiol standards was analyzed using a competitive ELISA. Optical density was measured at 450nm. Data points represent the average of duplicate measurements. The curve was generated using a four-parameter logistic (4-PL) regression model, exhibiting a high coefficient of determination (R² = 0.9997). Due to the competitive nature of the assay, an inverse relationship between 17β-estradiol concentration and optical density is observed. The displayed equation was used to interpolate the concentrations of experimental samples based on their respective absorbance values. Please click here to view a larger version of this figure.

	Age (years)	OSDI	TBUT OD (sec)	TBUT OS (sec)	CFSS OD	CFSS OS
Non-DED	55.6 ± 6.2	1.6 ± 2.9	10.3 ± 1.6	11.4 ± 0.8	0.5 ±0.6	0.7 ± 0.9
(n = 9)
DED	56.4 ± 5.1	38.8 ± 21.2	5.7 ± 1.9	5.1 ± 1.2	4.6 ± 1.6	4.6 ± 1.9
(n = 11)
*Data represented as mean ± standard deviation
*Ocular surface disease index (OSDI), tear break up time (TBUT), CFSS (corneal fluorescein staining score)

Table 1: Demographic and clinical assessments for participants in both the dry eye disease (DED) and non-DED groups.

	Testosterone Concentrations (ng/mL)			17β-estradiol Concentrations (pg/mL)
	Mean	SD	%CV	Mean	SD	%CV
non-DED	0.31	0.05	14.69	66.59	0.67	1.00
DED	0.56	0.10	18.67	47.07	3.13	6.65
*SD: Standard Deviation
*%CV: Coefficient of Variation

Table 2: Average testosterone and 17β-estradiol concentrations for non-DED and DED participants. Concentrations were calculated from the standard curves generated for each hormone.

Discussion

Tear collection and analysis allow easy access for scientific assessment of ocular surface health and identification of potential biomarkers^10,22. Our protocol describes successful adaptation of commonly used ELISA methods to analyze tear film samples for sex hormones. We developed this method to address the primary limitation of tear analysis, the relatively small volume of tears available. By optimizing the workflow to accommodate these micro-volume constraints, we ensure reliable detection without compromising the sensitivity required for accurate hormonal quantification.

For this protocol, a minimum of 20 µL per aliquot is required for detection, and running samples in duplicate, recommended for reliability, necessitate at least 40 µL per participant. This often requires multiple collection sessions since typically either 15 µL of basal tears per eye²⁸ (non-dry eye) or 6 µL of basal tears per eye (dry eye) can be obtained from one session using microcapillary tubes. To minimize variability, sessions should be scheduled within 24 hours of each other and at similar times during the day. Alternative tear collection methods, such as using Schirmer’s strips, can be faster to administer, but there is a higher risk of reflex tearing and altered tear composition, even when topical anesthetic is used^3,29.

Troubleshooting steps can be followed according to each ELISA kit’s detailed instructions, depending on the identified problem. As it relates to this protocol, accurate hormone quantification in tear samples using ELISA relies on several critical steps. A thorough mixing at each stage is necessary to ensure uniform distribution of reagents and to prevent variability in optical density readings. Proper precautions are required to avoid touching the bottom of the wells during pipetting, as this can damage the coating and compromise assay performance. Additionally, residual solutions on the well walls should be gently mixed to achieve complete reagent contact. Ensuring complete removal of residual wash buffer using gentle agitation is essential to maximize reagent-surface interaction and prevent high background signal²⁴.

Because commercially available ELISA kits are optimized for high-volume samples (e.g., serum or plasma), protocol adjustments are necessary for tear fluid analysis³⁰. In this study, the sample volume and dilution factor were optimized during method development and are presented as a standardized workflow for hormone detection in tear samples. The following considerations are provided as optional guidance for adapting this protocol to other commercially available ELISA kits or different biomarkers, which may have varying sensitivity and volume requirements. When adaptation is necessary, adjustments to total volume and/or dilution factor may be explored. For example, users may evaluate one parameter at a time, such as reducing total volume while maintaining the dilution factor specified by the kit, followed by adjusting the dilution factor and maintaining the total volume^24,31. If optical density readings fall outside the standard curve range, further adjustment of the dilution factor may be considered. For instance, increasing the dilution factor for sample wells while maintaining standards the same may help bring readings within the measurable range²⁴. In such cases, the final calculated concentration should be corrected by applying the appropriate dilution factor. When selecting an assay, those requiring complex pre-treatment steps may not be suitable for tear samples due to increased handling complexity and the limited volume available³⁰.

For generation of the standard curve, manufacturer guidelines suggest a Four-Parameter Logistic (4PL) model. This non-linear regression is widely considered the industry standard for immunoassays as it accurately models the sigmoidal relationship between concentration and optical density³². Alternatively, a semi-logarithmic plot (mean absorbance against the logarithm of concentration) may be used, though this method often results in decreased accuracy at the extreme ends of the curve. Various software options (e.g., GraphPad Prism, MyAssays) are capable of executing these complex algorithms to interpolate unknown concentrations. Since different models can yield variable outcomes, it is crucial to select one method and maintain consistency throughout the analysis to ensure data comparability²⁴.

In this study, we utilized the drc (Dose-Response Curve) package within the R statistical computing environment to ensure a rigorous and uniform approach. The drm() function was employed to mathematically model the standard curve using a four-parameter log-logistic equation, followed by the ED() function for inverse prediction to determine the concentration of unknown samples (x-axis) based on their optical density (y-axis)³³. We selected the R statistical computing environment³⁴, utilized within the RStudio integrated development environment. This combination provides an open-source, no-cost solution that makes powerful analysis accessible. Although R requires a higher degree of coding proficiency compared to commercial point-and-click software, the RStudio interface significantly streamlines code management and workflow reproducibility³⁵.

Mass spectrometry-based tear film proteomics and steroid profiling offer high sensitivity and multiplexing capabilities. However, these methods require sophisticated instrumentation, complex sample processing, and specialized bioinformatic expertise, which may limit their accessibility in routine laboratory settings²². In contrast, ELISA provides a cost-effective, technically accessible alternative for targeted quantification that can be readily implemented in standard laboratory settings. Its high-throughput potential and established reproducibility make it ideal for validating specific biomarkers for large-scale studies³¹. The protocol described here focuses on adapting ELISA for low-volume tear samples, enabling practical measurement of tear hormone concentrations in a minimally invasive manner. This approach is intended to complement mass spectrometry–based techniques, particularly in settings where accessibility and workflow simplicity are important considerations⁸.

In summary, this protocol provides a practical and accessible approach for adapting ELISA-based hormone detection to low-volume tear samples. The optimized workflow demonstrated feasibility for detecting and quantifying 17β-estradiol and testosterone in tear fluid. While the current findings are based on a limited pilot dataset, the approach may support future studies investigating local hormone dynamics at the ocular surface and their role in tear film physiology and dry eye disease. In this context, tear-based hormone measurements may complement systemic assessments, although further studies are required to establish their biological and clinical relevance. Beyond applications in ophthalmology, this approach may have potential utility for exploring tear fluid as a non-invasive matrix for hormone monitoring; however, such applications remain to be validated. Overall, this protocol offers a reproducible and accessible framework for tear-based hormone analysis that may be applied in future research settings. The accessibility and simplicity of this method position it as a valuable addition to biomarker research in both clinical and translational settings.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
17 beta Estradiol ELISA Kit	Abcam	ab108667
Aluminium foil	Reynolds Wrap	10900000154	Any availble aluminium foil for kitchen to protect reagents from light
Microcapillary tube 3.33 μL	Drummond	1-000-0033	Includes a squeeze bulb for fluid transfer
Microcapillary tube 5 μL	Drummond	1-000-0050	Includes a squeeze bulb for fluid transfer
Microcentrifuge device	Benchmark Scientific	C1012
Microcentrifuge tube 1.5 mL	Fisher	05-408-129
Microplate reader	Promega	GM3000
Orbital microplate shaker	Labnet International Inc.	S2030-1000-B
R computer language	Posit Software, PBC (formerly RStudio Inc.)	R version 4.5.1	Download R language prior to Rstudio through https://cran.r-project.org/
RStudio Software	Posit Software, PBC (formerly RStudio Inc.)	RStudio 2023.06.1 Build 524 "Mountain Hydrangea"	https://posit.co/downloads/
R packages "drc", "ggplot2", "patchwork", "ggpp", "dplyr", "readr", "tidyr", "writexl"	Posit Software, PBC (formerly RStudio Inc.)		Install the package in Rstudio
Testosterone Control kit	R&D Parameter	QC165	Sold separately from the ELISA kit (KGE010)
Testosterone ELISA kit	R&D Parameter	KGE010