Immune responses in rodent whole eye transplantation: elucidation and preliminary investigations into rejection diagnosis and monitoring

Abstract

Background: Whole Eye Transplantation (WET) offers potential for vision restoration but is hindered by the complex challenge of immune rejection. Understanding and closely monitoring these immunological responses is crucial for advancing WET. This study delves into the timeline and nature of immune responses in a rodent model of WET without immunosuppression, aiming to elucidate a detailed picture of the immune landscape post-transplantation and establish innovative diagnostic and monitoring methods.

Methods: We employed a multi-faceted approach to analyze immune responses post-WET, including assessments of gross changes in corneal transparency, thickness, and skin condition. Histopathological examinations of both ocular and surrounding skin tissues provided insights into cellular changes, complemented by ocular RT-qPCR for molecular analysis. Serological analysis was employed to quantify cytokines, chemokines, and donor-specific antibodies, aiming to identify potential biomarkers correlating with WET rejection and to validate the presence of antibody-mediated rejection. These methodologies collectively contribute to the development of non-invasive diagnostic and monitoring strategies for WET.

Results: Our study revealed a rapid and acute immune response following WET, characterized by an early innate immune response dominated by complement involvement, and infiltration of neutrophils and monocytes by post-operative day (POD) 2. This was succeeded by an acute T-cell-mediated immune reaction, predominantly involving T helper 1 (Th1) cells and cytotoxic T lymphocytes (CTLs). The presence of donor specific antibody (DSA) and indications of pyroptosis in the early phases of rejection were observed. Notably, the early elevation of serum CXCL10 by POD4, coupled with ocular CD3+ cell infiltration, emerged as a potential early biomarker for WET rejection. Additionally, corneal transparency grading proved effective as a non-invasive monitoring tool.

Conclusion: This study offers a first-time comprehensive exploration of immune responses in WET, unveiling rapid and complex rejection mechanisms. The identification of early biomarkers and the development of non-invasive monitoring techniques significantly advance our understanding of WET rejection. Additionally, these findings establish an essential baseline for future research in this evolving field.

Keywords: diagnostic strategies; immune rejection; monitoring techniques; non-invasive biomarkers; vascularized composite allotransplantation; whole eye transplantation.

Figure 1

Schematic representation of the Syn/Allo whole eye transplantation procedure. The surgical incision design includes the skin around the eye and auricle with a linear extension for exposure of the common carotid artery and external jugular vein. The transplantation procedure involves three main steps: 1) donor graft preparation, 2) simultaneous donor graft harvest and recipient preparation, and 3) anastomosis of the donor and recipient arteries and veins, coaptation of the optic nerves, and suturing of the skin. Donor ischemia time starts at the point of donor graft harvest and ends at anastomosis finish. BN, Brown-Norway rat; LEW, Lewis rat; Syn, Syngeneic transplant; Allo, Allogeneic transplant; WET, Whole Eye Transplantation, v, vein; a, artery.

Figure 2

Evaluation of ischemic response and postoperative recovery in Syn WET. (A–C) showed the ischemia in the graft skin, retinal, and limbal vasculature during surgery (Left) and demonstrated good perfusion at 30 days after Syn WET (Right). (D, E) showed Doppler OCT images, illustrating blood flow in central retina (D) and peripheral retina (E) before and 30 days after Syn WET surgery; blue and red colors indicate venous and arterial flow (white arrows), respectively. (F) presents Hematoxylin and Eosin (H&E) stained sections comparing a naïve retina with a retina from a Syn animal 30 days post-WET. The comparison reveals no significant changes in the retinal structure, except for some thinning observed, especially in the outer nuclear layer (ONL) and photoreceptor layer (PL). Scale bar: 50 μm in (F).

Figure 3

Modified corneal score system for grading rejection. The five images illustrating corneal transparency changes for grades 0-4. Changes for scores 1 and 2 are indicated by black arrows in the images.

Figure 4

Postoperative gross changes in eye and skin following WET. (A, B) Representative images of corneal transparency and limbal vasculature in Allo and Syn animals over POD2 to 8 (POD2-8). In the Allo group, limbal vessels show hyperemia (blue arrowhead), patchy perilimbal subconjunctival hemorrhage (black arrowhead), and complete disappearance of visible vessel structure indicating severe damage. (C) Corneal rejection scores for Allo and Syn animals, with a significant increase in the Allo group from POD5 onwards (n=11 per time-point for Allo, n=8 per time-point for Syn), ‘ns’ denotes not significant. (D) The fold increase in corneal thickness reveals a significant thickening in Allo animals at POD6 and POD8 when compared to both baseline (POD0) and to the Syn animals at the same time points (n=5). (E, F) Macroscopic skin changes on the Allo animals, showing progression from rash to purple discoloration, with persistent edema noted throughout the observation period. For the comparison of corneal scores, differences between groups at each time point were compared using the Wilcoxon rank sum test. For corneal thickness, independent t-tests were utilized for between-group comparisons, and repeated measures ANOVA was employed to assess overall time and group effects, accompanied by Dunnett’s Multiple Comparisons. Significance levels are denoted as: ns, not significant; **P<0.01, and ****P<0.0001.

Figure 5

Differential gene expression analysis in eye samples post-WET. (A) This clustergram presents a heatmap with dendrograms depicting patterns of gene expression across different groups, with gray indicating genes not detected. The -ΔΔCt data was used to generate the heatmap for data visualization. At POD5, the Allo group exhibits a marked increase in gene expression. (B–E) Volcano plots delineate the significance and magnitude of gene expression changes, with key genes spotlighted. The volcano plot combines a p-value statistical test with the fold regulation change enabling identification of genes with both large and small expression changes that are statistically significant. The p values are calculated based on a Student’s t-test of the replicate 2^ (- Delta CT) values for each gene in the naïve control group and treatment (Syn and Allo) groups. (n=3, Fold change >2, P<0.05; Up, increased gene expression; Down, decreased gene expression; Notsig, not significant). The Syn group displays slight transcriptional increases with TIMP1 at POD2 and MMP1B at POD5. In the Allo group at POD2, modest upregulation is noted in IL6, C3, IL5, CXCL11, IFNG, CASP1, and GZMB. By POD5, this upregulation intensifies, with 62 out of 84 rejection-related genes significantly increased. Notably, six of the initially upregulated genes show dramatic rises: CXCL11 surges from a 2.68-fold to a 688.63-fold increase, GZMB from 2.35 to 315.69-fold, IFNG from 2.55-fold to 127.76-fold, IL6 from 10.77-fold to 33.81-fold, C3 from 3.48-fold to 22.54-fold, and CASP1 from 2.46-fold to 23.67-fold. IL1B, a CASP1 substrate, also significantly increases to 28.34-fold, without initial upregulation at POD2. Additionally, CXCL9, CXCL10, and GZMA, which were not upregulated at POD2, exhibit substantial increases exceeding 100-fold by POD5. Concurrently, ITGAE (CD103) is significantly downregulated approximately 10-fold, in both Allo (0.11-fold change) and Syn (0.08-fold change) animals at POD2, with a slight decrease persisting in the Allo group at POD5 (0.48-fold change). These significant gene expression shifts underscore the immune response dynamics following transplantation.

Figure 6

Progressive inflammatory infiltration in Allo WET over time. (A–F) depict escalating mononuclear inflammatory cell presence in the conjunctiva (black arrows), and peripheral cornea (white arrowhead) on POD 2 (A–C) and 4 (D–F). The majority of these cells were MPO+ (B, E), interspersed with CD3+ lymphocytes (C, F inset). (G, H) show increased MPO+ and CD3+ staining in iris and ciliary body indicated by black arrows on day 5, alongside advancing infiltration in the conjunctiva, limbus, and peripheral cornea. (I, J) show increased cellularity in the nerve fiber layer (asterisks) at POD5, comprising predominantly CD3+ (J) over MPO+ mononuclear cells (I), despite substantial MPO+ infiltration in the choroid (I, black arrows). By POD 6, tissue necrosis becomes evident in some cases (K), characterized by corneal edema, acute inflammation (white arrowheads), and necrosis, along with necrosis in the iris (asterisk) and retina (L, asterisk). Scale bars: (A–H), K-100 μm; Inset F-20 μm; (I, J), L-50 μm.

Figure 7

Heatmap depicting inflammatory cell distribution in Allo transplanted eyes post-WET. This heatmap displays the presence of inflammatory cells identified by H&E, MPO, and CD3 staining across different tissues after Allo WET. It covers a range of post-operative days (POD2 to POD8), with each column corresponding to a specific day. The rows categorize different tissue types and stainings. The color intensity on the heatmap indicates the number of samples (out of 6) exhibiting positive inflammation for each marker, ranging from 0 (no inflammation sample) to 6 (inflammation in all samples). White squares marked with an ‘X’ denote areas where dense pigmentation hindered reliable inflammation assessment via H&E staining. The blood-retina barrier (BRB) is noted, demarcating the regions outside from those within the barrier. CB, ciliary body; BRB, blood-retina barrier.

Figure 8

Histopathological evaluation of skin graft rejection using the modified Banff VCA criteria. (A) Control skin from a naïve animal showing normal histology. (B) Syn skin at POD4, with minimal cell infiltration indicative of Grade 1 rejection. (C) Allo skin at POD4, demonstrating Grade 1 rejection with mild inflammatory infiltration. (D) Allo skin at POD4, presenting with Grade 2A rejection featuring moderate infiltration without epidermal involvement. (E) Allo skin at POD5, with Grade 2B rejection characterized by inflammation reaching the epidermis (inset: higher magnification, 80X). (F) Allo skin at POD6, displaying Grade 3A rejection with isolated keratinocyte necrosis (inset: higher magnification, 80X). (G) Allo skin at POD6, showing Grade 3B rejection with multifocal epidermal necrosis (arrows). (H) Allo skin at POD8, exhibiting Grade 4 rejection with diffuse full-thickness necrosis of the epidermis. (I) Quantitative analysis indicates a statistically significant difference in rejection grades between Allo and Syn groups from POD5 onwards (n=6/time-point for Allo, n=3/time-point for Syn). Groups were compared by Wilcoxon rank sum test, and the significance level is denoted as **P < 0.01. All skin biopsies were retrospectively assessed in a blinded manner by a board-certified veterinary pathologist. Scale bars represent 100μm for the main images and 20μm for insets.

Figure 9

Serum cytokine, chemokine, and DSAs levels in cross-sectional and longitudinal studies post-WET. (A) shows a significant rise in serum CXCL10 levels on POD4 and 5 in the Allo group compared to the Syn group and baseline, with no difference between the two days, followed by a decrease on POD6. (B) illustrates an increase in serum IFN-γ levels on POD4 from baseline, without a significant difference from the Syn group, and a further increase on POD5 with significant differences noted both intragroup and between the Allo and Syn groups, and a subsequent decrease on POD6, with significance levels indicated (n=6/time-point for Allo, n=3/time-point for Syn). (C) demonstrates that donor-specific IgM levels increased by POD4, peaking at POD5 with significant differences compared to POD0. Meanwhile, donor-specific IgG levels became detectable by POD4 and showed significant increases by POD6 compared with the levels at POD4 (n=5). ‘ns’ denotes not significant; MFI stands for mean fluorescence intensity; ‘anti-BN’ refers to anti-Brown Norway. Cytokine and chemokine results were analyzed using two-way ANOVA to assess group and timepoint interactions, followed by Tukey’s post hoc tests for multiple comparisons. DSA levels were evaluated with repeated measures ANOVA, followed by Dunnett’s Multiple Comparisons. Significance levels are denoted as ns, not significant; *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001.