T-Cell Repertoire Analysis in the Conjunctiva of Murine Dry Eye Model

Abstract

Purpose: Dry eye is closely related to the activation and proliferation of immune cells, especially T cells. However, the determination of the preferential T-cell clonotypes is technically challenging. This study aimed to investigate the characterization of T-cell receptor (TCR) repertoire in the conjunctiva during dry eye.

Methods: A desiccating stress animal model was established using C57/BL6 mice (8–10 weeks, female). After 7 days of stress stimulation, the slit-lamp image and Oregon–green–dextran staining were used to evaluate the ocular surface injury. Periodic acid–Schiff staining was used to measure the number of goblet cells. Flow cytometry was used to detect the activation and proliferation of T cells in the conjunctiva and cervical lymph nodes. Next-generation sequencing was used to detect the αβ TCR repertoire of the conjunctiva.

Results: The αβ TCR diversity increased significantly in the dry eye group, including the higher CDR3 amino acid length, marked gene usage on TCR V and J gene segments, extensive V(D)J recombination, and distinct CDR3 aa motifs. More important, several T-cell clonotypes were uniquely identified in dry eye. Furthermore, these perturbed rearrangements were reversed after glucocorticoid administration.

Conclusions: A comprehensive analysis of the αβ TCR repertoire in the conjunctiva of the dry eye mouse model was performed. Data in this study contributed significantly to the research on dry eye pathogenesis by demonstrating the TCR gene distribution and disease-specific TCR signatures. This study further provided some potential predictive T-cell biomarkers for future studies.

Dry eye is a chronic self-perpetuating inflammatory illness caused by multiple intrinsic and extrinsic insults.¹ In recent years, clincal² and experimental³ studies have observed the infiltration of T cells in the conjunctiva, implying the involvement of T cells in the pathogenesis of dry eye. The consecutive stimulation toward superficial epithelial cells leads to the release of inflammatory mediators and protease, which stimulate the production of matrix metalloprotease (MMP)⁴ and inflammatory cell recruitment⁵ mainly in the conjunctiva.^6,7 Recently, several studies revealed a correlation between CD4⁺ Th cells and dry eye, especially interferon-γ–secreted T-helper type 1 (Th1)^8,9 and interleukin-17–secreted Th17^10–12 cells. Clinical evidence showed that the short-term use of glucocorticoids also improved dry eye symptoms. These findings suggest that the key is to identify the specific T-cell subset in the early stage. Drawing upon a comprehensive analysis of the T-cell receptor (TCR) repertoire of dry eye may reveal essential findings. 

T cells proliferate in response to a combination of antigens and TCR; a single TCR can recognize thousands of peptides. TCR is a heterodimer composed of two peptides. A few variable hairpin loops, named complementarity-determining regions (CDRs), interact with peptides presented by the major histocompatibility complex (MHC) molecules on the TCR β or TCR δ chain. The CDR3 is the most diverse region and directly interacts with the antigen. The diversity of CDR3 on both amino acid (AA) length and sequence rearrangement determines the functional T-cell repertoire.¹³ Besides CDR3 specificities, the TCR repertoire is effectively restricted by preferential gene expression and dominant V–J gene combinations. To be exact, random recombination happens in the variable (V), diversity (D), and joining (J) gene segments; every mature T-cell population has a unique and identical TCR. The capacity of the TCR repertoire determines host immune responses. 

Next-generation sequencing makes it possible to solve the mystery of the adaptive immune system. Comprehensive and longitudinal T-cell repertoire analyses have accelerated the understanding of clonotypes and diversity of T cells,¹⁴ bringing a revolution to immune homeostasis, new immune cell identification,¹⁵ disease diagnosis,¹⁶ immune monitoring,¹⁷ organ transplantation,¹⁸ and malignancy immunotherapy.^19,20 The heterogeneity of the TCR repertoire gets influenced in response to numerous circumstances, including infection,^21,22 metabolic disorders,²³ vaccination, organ transplantation,²⁴ malignant hematemesis, autoimmune disorders, neurodegenerative diseases,²⁵ and cancer.^26,27 

The pathophysiology of dry eye has a strong link with T cells, and predominantly expanded T-cell clones reflect local inflammation. However, the TCR characteristics in the disease are still unknown. In this study, we compared the αβ TCR repertoire in the conjunctiva between desiccating stress animal models and controls. Some highly clonally expanded T-cell clonotypes with marked gene usage were investigated. This study aimed to contribute to the early diagnosis of dry eye by exploring possible TCR detection measures. 

Methods

Animals

Healthy female C57BL/6 mice (8–12 weeks) were purchased from Xiamen University Laboratory Animal Center (Xiamen, Fujian, China). The animals were kept in a clean condition (relative humidity: 60%–70%; temperature: 23°C–26°C) under a 12-hour light/dark cycle and with unlimited food and water. Forty mice were stochastic divided into a control group, a dry eye group, and glucocorticoid-treated group (n = 6). All animal experiments were approved by the Institutional Animal Care and Use Committee of Xiamen University (Approval ID: XMULAC20180053). All studies adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Murine Dry Eye Model

The desiccating stress (DS) model^5,28–31 is one of the most widely used murine dry eye models. It combines the desiccating environmental stress model and the muscarinic receptor blockade model. To be specific, the mice were housed in a special environmental condition (less than 25% ambient humidity with flow and 23°C ± 2°C room temperature) for 7 days. Meanwhile, 0.5 mg/0.2 mL scopolamine hydrobromide (S817762-5; Macklin, Shanghai, China) was subcutaneously injected three times a day. Negative controls were hosted in a normal environment. 

Oregon–Green–Dextran Staining

Oregon–green–dextran (OGD; D7173; 70,000 molecular weight (MW); Invitrogen, Carlsbad, CA, USA) was selected for its reliability and validity to determine the corneal epithelial permeability.²⁹ Briefly, 0.5 µL OGD (50 mg/mL) was instilled onto the mice's ocular surface for 1 minute with manual blinking for several times. Subsequently, the mice were euthanized, and the cornea was flushed five times with phosphate-buffered saline. Photographs were captured with a high-dynamic resolution digital camera (63× magnification, AZ100; Nikon, Tokyo, Japan) under fluorescence excitation at 470 nm. All operations were carried out in the dark. 

Tear Secretion Test

To quantify tear secretion, the phenol red cotton thread (Zone-Quick; Yokota, Tokyo, Japan) was inserted into the lower Y fornix at approximately one-third distance of the lower eyelid from the lateral canthus for 15 second to measure the tear production in mice (n = 6 in each group). The length of wet threads represented the tear secretion capacity. 

Goblet Cell Density

A periodic acid–Schiff (PAS) staining kit (395B-1KT; Sigma-Aldrich, St. Louis, MO, USA) was used on paraffin sections to determine the density of goblet cells in the conjunctiva. After removing the wax with xylene and ethanol (100%–95%–80%–70%), the sections were stained with periodic acid for 8 minutes and Schiff's reagent for 10 minutes and counterstained using hematoxylin. The total number of PAS-positive goblet cells was counted in the superior and inferior conjunctiva. Digital images were captured using a light microscope (100× magnification; Eclipse 50i; Nikon, Tokyo, Japan). 

Flow Cytometric Immunophenotyping

Flow cytometry was chosen to determine the immune cell type involved in the dry eye model. The conjunctiva and the cervical lymph nodes were separated carefully and immediately after the mice were euthanized. The 0.1% type IV collagenase (C5138; Sigma-Aldrich) was used to digest conjunctiva at 37°C for 1 hour, and a 300-µm filter mesh was used to prepare single-cell suspensions. The lymph nodes were gently ground with metal mesh to obtain single-cell suspensions. Then, the Fc receptor was blocked with CD16/32 (clone: 2.4G2, 1:500; BD Biosciences, Franklin Lakes, NJ, USA). The live/dead cells were identified using the fixable viability dye eFluor 780 (1968231, 1:1000; Invitrogen) and stained for 20 minutes with APC (PE) anti-mouse CD3 (clone: 145-2C11, 1:100; BD Biosciences), BV421 CD45 (clone: 30-F1, BD Biosciences, USA, 1:400), BV510 CD4 (clone: RM4-5, BD Biosciences, USA, 1:400), (clone: RM4-5, 1:400; BD Biosciences), FITC TCR β (clone: H57-597, 1:400; BD Biosciences), PE-Cy7 TCR γδ (clone: GL3, 1:400; BD Biosciences), and FITC CD69 (clone: H1.2F3, 1:400; BD Biosciences) at 4°C and washed with FACS buffer (0.05% BSA in PBS), followed by detection with a CytoFlex S flow cytometer (Beckman Coulter, Brea, CA, USA). The flow cytometry data were plotted and quantified using FlowJo software version 10 (Tree Star LLC, Ashland, OR, USA). Six mice (12 eyes) were evaluated in each independent experiment. 

Immunofluorescent and Immunohistochemical Staining

The mice eyeballs were fixed in 4% paraformaldehyde for 24 hours and dehydrated in gradient ethanol (70%–80%–95%–100%–100%) and xylene. They were then embedded in paraffin, and 5-µm sections were cut. Before staining, xylene and ethanol (100%–95%–80%–70%) were used to remove the wax from the sections. 

For immunofluorescent staining, the sections were blocked with 2% fetal bovine serum, permeated with 0.2% Triton X-100 (Sigma-Aldrich), and incubated with primary antibodies, including anti-MMP3 (sc-6839; Santa Cruz Biotechnology Inc., Santa Cruz Inc., Dallas, TX, USA) and anti-MMP9 (GTX100458; GeneTex, GeneTex Inc., Irvine, CA, USA), at 4°C overnight. They were then incubated with a secondary Alexa Fluor 488 donkey anti-mouse antibody (Invitrogen, M31500) at room temperature for 1 hour. Images were obtained using a confocal laser scanning microscope (200× magnification; Zeiss Confocal Laser Scanning Microcopy 880+; Carl Zeiss, Oberkochen, Germany). 

For immunohistochemical staining, the sections were pretreated using heat-mediated antigen retrieval with sodium citrate buffer (pH 6) for 15 minutes. They were then blocked with 10% goat serum and incubated with the primary rabbit anti-CD3 (ab5690; Abcam, Inc., Cambridge, UK) at room temperature for 30 minutes and detected with a rabbit horseradish peroxidase conjugated compact polymer system (SP-9002; ZSGB-BIO, Beijing, China). The sections were then developed with DAB chromogen substrate (ab64261; Abcam, Inc.) for 2 minutes and counterstained using hematoxylin. Images were captured using a light microscope (100×/400× magnification; Eclipse 50i; Nikon). 

RNA Extraction

After euthanasia, the mouse conjunctiva was surgically excised on ice. Total RNA was isolated using an RNeasy Mini Kit (74104; Qiagen, Germantown, MD, USA) and quantified using a Nanodrop2000 spectrophotometer (Thermo Fisher Scientific, MD, USA). Then, 1000 ng total RNA was converted into first-strand cDNA by reverse transcription using a reverse transcription kit (RR047A; Takara, Shiga, Japan). 

αβ TCR Repertoire Amplification

A two-round nested amplicon arm PCR was performed as described in a previous study³² to prepare a CDR3 library of TCR α and β chains. To be specific, a Multiplex PCR Assay Kit Ver. 2 (Takara Bio, Beijing, China) using specific forward and reverse primers designed for functional V and C gene segments of the mouse TCR chain according to the International Immunogenetics Information System (IMGT) was used. 

Analysis of High-Throughput Sequencing Data

The original data provided by the Illumina HiSeq X Ten platform were transferred to raw paired-end sequencing reads by filtering the low-quality reads. V, J, and CDR3 regions of TCR β consensus sequences were identified using the BLAST Plus (Translated Query-Protein Subject BLAST 2.7.1+) in the IMGT. The TCR diversity was evaluated using the Gini coefficient, Shannon–Wiener index, and rank abundance, which have been widely used for assessing the richness and diversity of TCR as described in a previous study.³³ 

Generation of Sequence Logos

CDR3 sequences from this study were aligned with MEGA software using the ClustalW algorithm. Consensus motif analyses of the selected aligned sequences (TCR β) were performed on WebLogo,^34,35 a web-based application on the website http://weblogo.threeplusone.com/create.cgi, to generate sequence logos. The bit height and stack width corresponded to the AA identity and gap number, respectively. Blue represented most hydrophobic AAs (A, C, F, I, L, V, W, and M); green represented polar, noncharged, and nonaliphatic AAs (N, Q, S, and T); purple represented acidic AAs (D and E); and orange represented other residues (H, G, P, and Y). 

Statistical Analysis

The unpaired Student’s t-test was used to compare the distinction between the control group and the dry eye group. The Brown–Forsythe and Welch analysis of variance (ANOVA) tests were used to compare differences between the three groups (with the GC group). Statistical calculations were performed using GraphPad Prism v9.2 software (GraphPad Software, Inc., La Jolla, CA, USA). The data represented mean ± standard deviation (SD). All results were representative of two or three independent experiments, with n > 3 in each experiment. The P values were calculated using a one-way ANOVA, and P < 0.05 was considered statistically significant. 

Results

Ocular Surface Barrier Disruption and Immune Activation in the Murine DS Model

A well-known DS model was established in female C57BL/6 mice to mimic dry eye pathologic damages, such as aqueous deficiency, ocular surface irregularity, goblet cell loss, and ocular surface inflammation.³⁶ 

After 7 days of stress stimulation, the tear production decreased in the dry eye group (Fig. 1B). The OGD staining was performed to evaluate the corneal epithelial barrier function. The dry eye group exhibited extensive corneal staining with fluorescein (Figs. 1A, 1C). PAS staining was used to calculate the number of goblet cells. The number of conjunctival goblet cells in the dry eye group was significantly lower than that in the control group, as shown in Figures 1D and 1E. Constant exposure to DS also led to accumulation of inflammatory matrix metalloproteinases MMP-9 and MMP-3 (Figs. 1F, 1G) and epithelial cell death (Supplementary Fig. 3A). 

Figure 1.

 

Change in the ocular surface of the desiccating stress model. (A) Representative slit-lamp image and OGD staining image of the two groups. (B) Phenol red thread test indicating tear secretion (n = 6). (C) Statistical intensity analysis of OGD fluorescence (n = 6). (D) PAS staining shows goblet cell reduction in the conjunctiva. (E) Number of goblet cells in the conjunctiva (sum of superior and inferior conjunctiva, n = 6). (F, G) Representative immunofluorescent images of MMP-3 (green), MMP-9 (green), and DAPI (blue) in the corneal epithelium. The data are represented as mean ± SD. The P values were analyzed using the Student’s t-test. Scale bar: 200 µm (D) and 50 µm (F, G).

Next, the number of T cells (CD3⁺) in the conjunctiva was detected using immunohistochemistry. More T-cell infiltration was seen in the dry eye group (Figs. 2A, 2B). As shown in Figure 2C, the T-cell number (CD3⁺/CD45⁺) was evaluated via flow cytometry. After DS stimulation, the T-cell number increased in the conjunctiva (Figs. 2D, 2E). We next examined the draining lymph nodes. Supplementary Figure 3B confirmed that the number of T cells was higher in the dry eye group than controls. Supplementary Figures 3C, 3D, and 3E demonstrate that those expanded T cells mainly were αβ T cells, with γδ T cells also elevated in the dry eye group. The activated CD4⁺ T cells were reported as a pathogenic T-cell subset associated with the progression of immune-related dry eye disease.³⁷ Further, whether DS triggered the activation of T cells was assessed, and the CD4⁺ T cell was found to be activated (Supplementary Figs. 3F, 3G) in the draining lymph nodes of the murine DS model. 

Figure 2.

 

TCR β⁺ T cells were expanded in the DS model. (A) Representative immunohistochemistry staining images of CD3 in the conjunctiva. (B) Statistical results of CD3⁺ cells (n = 6). (C) Gating strategy for the T-cell population in the mice conjunctiva by flow cytometry. (D, E) Representative plots show frequency of CD45⁺ and CD3⁺ cells in the conjunctiva. The data are expressed as mean ± SD. The P values were calculated using the Student’s t-test. Scale bar: 20 µm (A).

T-Cell Clonal Expansion in the Murine DS Model

The T-cell clonotype affects a wide range of diseases, while TCR diversity varies significantly between healthy and diseased conditions.³⁸ Fresh conjunctiva RNA samples were isolated to perform TCR α and β chain V(D)J sequencing to distinguish the specificities of the TCR repertoire in the dry eye. Figures 3A and 3B show a clear trend of increasing clonotypes on both TCR α and β chains in the dry eye group. As shown in Figure 3C, the Gini coefficient was used to evaluate the inequality among values of a frequency distribution, demonstrating that the clonal diversity in the dry eye group was substantially more significant than that in the control group. CDR3 is the region in which TCR recognizes the antigens presented by MHC molecules, and the CDR3 aa length shows the diversity of T-cell clonality. The data in this study showed a standard Gaussian distribution in the two groups with different CDR3 lengths. 

Figure 3.

 

Diversity of TCR repertoires increased during desiccating stress. (A, B) TCR clonotypes of the TRA and TRB in the control and dry eye groups. (C) Comparison of CDR3 aa diversity using the Gini–Simpson index. (D, E) Distribution of CDR3 aa length in the TRA and TRB chains. (F, G) Top clonal clonotypes with the specific proportion between the two groups. The data are expressed as mean ± SD. The P values were calculated with the Student’s t-test (A–C) and multiple t-tests (F, G); n = 4 for each group.

Figures 3D and 3E illustrate that the CDR3 aa length in the dry eye group was significantly higher than in the control group on both TCR α and β chains. Besides, a significant difference in the clonal proportion ranking of the top 100 clonotypes was found between the two groups on the TCR α chain (Fig. 3F), which represented the decrease in shared clonal proportion compared with the control group. However, no significant difference between the two groups was evident on the TCR β chain (Fig. 3G). The data showed that TCR diversity increased in the dry eye group. 

Distinct Gene Usage Between the Two Groups

The recombination of germline gene segments generates an extensive TCR repertoire. Different gene usage varies widely between healthy and diseased individuals. The TCR α chain is made up of V (variable) regions, J (joining) regions, and C (constant) regions (Fig. 4A). The proportion of every single V and J gene segment was compared between the two groups to explore the usage bias of V and J gene segments in the dry eye group. Figures 4B–D show that the percentage of TRAV components changed significantly in the dry eye group. Compared with the control group, the frequencies of several genes, such as TRAV6-6, TRAV7-4, TRAV6-4, and TRAV4-3, specifically increased in the dry eye group, while the frequencies of TRAV6N-6 and TRAV8-1 showed a decreasing trend, compared with the control group. Figures 4E–G demonstrate the altered genes of the J segment. The frequencies of TRAJ15, TRAJ6, TRAJ11, TRA9, TRAJ45, and TRAJ35 genes predominantly increased after 7 days of DS. At the same time, the frequencies of TRAJ23 and TRAJ18 genes declined. Significant changes were observed on both V and J gene segments of the α chain when dry eye occurred. 

Figure 4.

 

Comparison of TRA gene distributions and alterations in the two groups. (A) Proposed model of V–J recombination of the α subunit. (B, C) Heatmap and volcano plot, respectively, reveal the altered V genes on the α subunit under DS. (D, E) Parts-of-whole graphs show the frequency of TRAV and TRAJ gene usage in each group. (F, G) Heatmap and volcano plot, respectively, showing the altered V gene on the α subunit under DS. The P values were calculated with multiple t-tests; n = 4 for each group.

Different from the TCR α chain, the TCR β chain consisted of D (diversity) regions between V (variable) and J (joining) regions, which significantly increased the diversity of TCR (Fig. 5A). In human lymphocytes, Nishio et al.³⁹ found that TCR repertoire distribution patterns mainly focused on the use of TRBV genes. The amounts and frequency of the dominant genes in both the V and J regions were detected in the dry eye and control groups to compare the altered genes of the TCR β chain. Figures 5B–D clearly show the phenomenal growth of TRBV13-3, TRBV15, and TRBV4, while the frequency of TRBV19 decreased. As for the J region, several genes changed in the dry eye group, but no difference greater than 0.05 was observed using the Student’s t-test (Figs. 5E–G). Under stress stimulation, the TCR diversity of the β chain increased greatly while the frequency of some genes significantly altered, especially on the V segments. However, no statistical difference between the two groups was evident in the J segments. 

Figure 5.

 

Comparison of TRB gene distributions and alterations in the two groups. (A) Proposed model of V(D)J recombination of the β subunit. (B, C) Comparison of the altered V gene on the β subunit under DS. (D, E) Parts-of-whole graphs reveal the frequency of TRBV and TRBJ gene usage in each group. (F, G) Comparison of the altered J gene on the β subunit under DS; the P values are shown in the heatmap and volcano plot. The P values were calculated using multiple t-tests; n = 4 for each group.

Dominant αβ T-Cell Clonotypes in the Dry Eye Group Compared With the Control Group

Besides analyzing the frequency of every single TCR segment, the profile of the expanded TCR repertoire was critically evaluated to find the dominant V(D)J pairings in the two groups. As expected, the dry eye group showed higher TCR clonal expansions on both α (Supplementary Figs. 1D, F) and β chains (Figs. 6B, 6D). Several differences in the V(D)J recombination, the gene count, and the CDR3 aa sequences in the two groups were observed after analyzing 15,898 CDR3 sequences and 2451 V–J rearrangements on the α chain, as well as 55,400 CDR3 sequences and 754 V–D–J rearrangements on the β chain. Some were particularly enriched in the dry eye group. 

Figure 6.

 

Usage patterns of TRB V–D–J combinations in the two groups. (A) Altered differences of V–D–J gene combinations between the two groups. The data are expressed as mean with SD. (C) Volcano plot demonstrates relevant P values among the groups. (B, D) Chord diagram exhibits abundant V–D–J pairings in the two groups. (E) TCR CDR3 aa motifs of all TRBV13-related combinations in the dry eye group. The P values were calculated using multiple t-tests; n = 4 for each group.

The top 20 V–D–J gene combinations of every sample were analyzed to find whether a unique dry eye–specific TCR existed to reveal the biased enrichment of the TCR β repertoire. Figures 6A and 6C demonstrate that many V–D–J pairings showed significant distinctions between the two groups. Figures 6B and 6D show the comparison of V–D–J pairing analysis between the two groups. The dry eye group revealed a remarkable clonal diversity compared with the control group. The results were stricken by the preferential combination (x = 1 or 2, y ranges from 1–1 to 2–7) in TRBV13-Dx-Jy. Besides the enormous alterations in gene frequency, the AA motifs also changed a lot (Fig. 6E), indicating that these specific combinations might correlate with the dry eye–specific antigens. 

Regarding TCR α repertoire, Supplementary Figures 1B and 1D show that the dry eye group had a much higher diversity of V–J pairings than the control group. When the frequency of V–J combinations was evaluated in the two groups, a large number of combinations displayed a remarkable rise in the dry eye group (Supplementary Fig. 1B). At the same time, several combinations showed a decrease in the dry eye group (Supplementary Fig. 1C). Strikingly, some unique pairings were observed only in the dry eye group, such as V8N-2-J37, V12D-2-J13, V12N-3-J40, and V16N-J40. Moreover, some combinations showed nearly equal frequencies in both groups but different AA motifs (Supplementary Fig. 1E). 

Glucocorticoids Improved the Inflationary TCR Repertoire

These results indicated that the αβ TCR repertoire exhibited greater diversity in the dry eye group. The common anti-inflammatory glucocorticoid treatment was considered to disturb these aberrations. Hence, a glucocorticoid-treated group (GC) was included, and dexamethasone was topically applied three times per day for a week. 

DS triggered a series of ocular surface dysfunction, but topical glucocorticoids promoted tear production, diminished the corneal epithelial dysfunction, and increased the number of conjunctival goblet cells (Figs. 7A–D). As for the T-cell repertoire, the clonotypes and clones were expected to decline after applying glucocorticoids (Fig. 7E). The Shannon–Wiener index was used to compare the CDR3 aa diversity, which revealed a similar tendency (Fig. 7F). Furthermore, the variations in the CDR3 aa length among the three groups were analyzed to reveal alterations in the GC group. Supplementary Figure 2A shows that specific CDR3 aa lengths declined using glucocorticoids on both TRA and TRB. Following that, using nonmetric multidimensional scaling, both V(D)J pairings and their frequency were evaluated in the three groups, revealing a substantial variation in TRA (P = 0.013) rather than TRB (P = 0.687) in all three groups (Supplementary Fig. 2B). Supplementary Figure 2C demonstrates the clone numbers in the three groups. The graph in Supplementary Figure 2C shows that 615 TRA and 57 TRB pairings were individually expressed in the dry eye group. The rank-abundance curve shows the richness and evenness of species at the same time. Figure 7G presents the distribution of TRA and TRB gene pairing abundances. The distribution curve of the dry eye group was the longest and steepest, indicating that using glucocorticoids might decrease abnormal T-cell differentiation. 

Figure 7.

 

Abundance of TCR repertoires improved after using glucocorticoids. (A) Quantification of tear production using the phenol red cotton test (n = 6). (B) Representative images of OGD staining. (C) Representative images of PAS staining in the conjunctiva. (D) Number of goblet cells in the conjunctiva (sum of superior and inferior conjunctiva, n = 6). (E) Clonotypes on TRA and clones on TRB of three groups. (F) Shannon–Wiener index demonstrates the diversity in the three groups on both TRA and TRB. (G) Rank-abundance curve of TRA and TRB repertories in the three groups. The data are expressed as mean ± SD. The P values were calculated using a one-way analysis of variance; n = 4 for each group (E–G). Scale bar: 200 µm (C).

Therefore, the pairwise overlap was used to analyze further that every single gene altered in terms of reading count, diversity, frequency, and the total number of clonotypes shared between samples on TRB. The dry eye group revealed biased enrichment, while no significant difference was found between the control and GC groups (Fig. 8A). The significantly higher combinations, including TRBV13-1-J1-4, TRBV13-3-D1-J1-2, TRBV13-3-D2-J2-3, and TRBV13-3-D2-J2-4, declined in the GC group (Fig. 8B). This implied that glucocorticoids improved abnormal T-cell clonotypes caused by DS. Different CDR3 consensus motifs might be capable of identifying individual autoimmune antigens in various diseases. The highest-ranked CDR3 aa sequences (count >100) of TCR β were clustered in the three groups (control, 11; dry eye, 37; and GC, 11), and motif analysis was performed (Fig. 8C). 

Figure 8.

 

Distribution of TRB repertoire declined in the glucocorticoid group. (A) Count, frequency, and diversity compared with the chord diagram in the three groups. A, B, and C represent the control group, the dry eye group, and the glucocorticoid group, respectively. (B) Comparison of specific V(D)J combinations in the three groups. (C) CDR3 motif analyses on TCRβ in the three groups (control, 11; dry eye, 37; and GC, 11). The data are expressed as mean ± SD. The P values were calculated using one-way analysis of variance; n = 4 for each group.

Discussion

Considering the considerable impact brought by T cells, this study assessed the T-cell responses in the conjunctiva and described the characteristics of the αβ TCR repertoire in dry eye disease for the first time. The results of the present study provided clear evidence that dry eye is an autoimmune disease with increased αβ TCR diversity. 

Dry eye is defined as a multifactorial disease characterized by tear hyperosmolarity. Continuous tear hyperosmolarity triggers the cascade of inflammatory response,^1,40 leading to the infiltration of immune cells, particularly the CD3⁺ T cells in the conjunctiva. The DS model is the most common animal model to mimic the pathophysiology of dry eye, such as decreased tear production and CD4⁺ T-cell activation in draining lymph nodes. Thus far, a limited number of studies have reported on T-cell clonotypes of relevant tissue involved in dry eye. In this study, the conjunctiva was isolated specifically, and high-throughput TCR sequencing was performed to reveal the TCR repertoire in dry eyes. 

Comparing the findings of this study with those of other studies confirmed that T-cell activation was responsible for the progression of dry eye disease.³⁶ Therefore, discovering the characteristics of the TCR repertoire in the disease and identifying the unique T-cell clonotypes were necessary. In this study, the αβ T-cell diversity of the conjunctiva was substantially higher in the dry eye group than in the control group. So far, studies on dry eye disease are lacking, but abundant findings have been reported on the TCR repertoire in other autoimmune diseases. A study performed on salivary glands suggested expanded salivary gland (SG) T cells with glandular dysfunction,⁴¹ but a limited TCR repertoire was found in infiltrating T cells of the kidneys of patients with Sjögren syndrome (SS).⁴² A study on labial SGs of primary SS detected using nested reverse transcription–PCR revealed that both their Th1 and Th17 cells showed restricted clonal diversities.⁴³ 

On the contrary, in a study on patients with SS, the expanded T-cell clones from the SG increased SG fibrosis.⁴¹ An animal experiment also suggested that oligoclonal expanded T cells were related to inflamed SG.⁴⁴ Similar results were observed in patients with rheumatoid arthritis (RA); increased TCR β repertoire of effector memory T-cell–differentiated Th17 cells significantly correlated with RA disease activity.⁴⁵ In addition, abnormally increased CD4⁺ TCR β repertoire might contribute to immunodeficiency in patients with RA.⁴⁶ Many recent studies also indicated that some tumor-specific antigens seemingly had endless diversity to exhaust lymphocytes.^26,27,47 

A high abundance of T cells has been demonstrated to lead to an immunopathologic process in many diseases. The present study proved that T cells extracted from the conjunctiva showed a significant increase in T-cell clones and clonotypes on both α and β chains with expanded CDR3 aa length, indicating that the abnormal T-cell expansion might relate to autoantigen recognition in the process of dry eye. 

In addition to TCR variety, the CDR3 aa length, V and J gene usage on different segments, predominant V(D)J combination, and aa motifs were also systematically analyzed. The result showed that the CDR3 aa length increased as dry eye progressed, focusing on 11 to 17 bp, which might be considered an important feature. Furthermore, the highly diverse TCR repertoire was distinct from individual samples, although some commonalities existed: the frequency of particular V(D)J combinations, such as TRA8N-2–TRAJ24, TRA12N-3–TRAJ40, TRA12D-2–TRAJ7, and TRBV13-3–TRBD1/2–TRBJy (y ranges from 1–1 to 2–7), was explicitly higher in the dry eye group compared with that in the control group, which was probably related to the dry eye–specific T-cell subsets. These distinct T-cell clonotypes in the conjunctiva might be used as a biomarker to distinguish between different types and stages of dry eye diseases. 

After identifying the characteristics of T-cell repertoire in the dry eye, the impossible intervention to suppress the selection and proliferation of T-cell subsets was next attempted. Dexamethasone has shown a tremendous therapeutic effect in many inflammatory diseases, including dry eye diseases. A significant decrease in the levels of inflammatory cytokines⁴⁸ and corneal barrier protective function⁴⁹ has been observed in murine dry eye models treated with methylprednisolone. However, the underlying anti-inflammation mechanism might be related to the restriction in naive T-cell proliferation and differentiation.⁵⁰ The results of this study confirmed the association between T-cell expansion and the use of glucocorticoids. Moreover, the abnormally expanded T-cell clones tended to return to normality, including the gene count, frequency, diversity, CDR3 aa length, and AA motifs. A possible explanation for these results might be related to the involvement of GC in antigen-specific selection.^51,52 

In conclusion, this study found a disease-specific αβ TCR repertoire in the conjunctiva of the DS mouse model, which added to the knowledge of the dynamics and distribution of TCR genes in the pathogenesis of dry eye. However, further investigations are needed to address the following issues: (1) the clinical TCR repertoire has been limited by the lack of patient samples. Future studies need to include a large sample size. (2) More comparisons need to be made between the different classifications of dry eye. (3) The process by which these T-cell subsets got activated and expanded needs to be investigated. 

Acknowledgments

The authors thank Haiping Zheng, Jingru Huang, and Xiang You (Core Facility of Biomedical Sciences, Xiamen University, Fujian province, China) for their technical assistance. 

Supported in part by grants from the National Key R&D Program of China (Nos. 2018YFA0107304 and 2020YFA0908100), the National Natural Science Foundation of China (Nos. 81870627, 82070931, 81770891, and 81900825), the Fundamental Research Funds for the Central Universities (No. 20720202013), and the Huaxia Translational Medicine Fund for Young Scholars (No. 2017-A-001). 

The datasets included in this study are available from the China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA: CRA005364) and are publicly accessible at https://ngdc.cncb.ac.cn/gsa. 

Disclosure: X. Bao, None; Y. Zhong, None; C. Yang, None; Y. Chen, None; Y. Han, None; X. Lin, None; C. Huang, None; K. Wang, None; Z. Liu, None; C. Li, None