Aims To evaluate the efficacy of spectral domain optical coherence tomography (SD-OCT) in monitoring the development of mouse experimental autoimmune uveoretinitis (EAU) as an animal model of endogenous uveitis, and to develop an OCT-based grading system for EAU severity.
Methods C57BL/6 mice were immunised with human interphotoreceptor retinoid-binding protein (amino acid sequence 1–20) peptide and complete Freund's adjuvant to induce EAU. The development of EAU was monitored by SD-OCT serially throughout the disease course, and the images were graded from 1 to 4 and compared with the clinical and histopathological grades.
Results SD-OCT images depicted retinal lamella structures including the inner segment/outer segment (IS/OS) line in normal mice. Retinal structural changes were observed on SD-OCT images in mice that developed EAU clinically scored as grade 1 or higher, which precisely corresponded to the pathological findings. The SD-OCT images of EAU were graded as follows: grade 1, a few infiltrating cells in the vitreous and retina; grade 2, increased vitreous cells, retinal vasculitis, and granulomatous lesion; grade 3, cell infiltration into the whole retina, disappearance of IS/OS line, and destruction of the retinal layer structure; and grade 4, disappearance of the outer retina. The SD-OCT grade of EAU based on these criteria correlated significantly with both the clinical grade (R2=0.282, p<0.005) and histopathological grade (R2=0.846, p<0.0001).
Conclusions SD-OCT is useful for evaluating the development and severity of mouse EAU. The SD-OCT scoring system we developed accurately reflects clinical and histopathological changes.
Statistics from Altmetric.com
Experimental autoimmune uveitis (EAU) in mice serves as a model for endogenous uveitis of suspected autoimmune aetiology in humans.1–4 EAU is induced by immunisation with retinal antigens such as arrestin (S-Ag) and interphotoreceptor retinoid-binding protein (IRBP),5 ,6 or by adoptive transfer of retinal antigen-specific CD4 T cells between syngeneic rodents.7 ,8 The EAU model is indispensable for elucidating basic mechanisms underlying the development of endogenous uveitis and has been used in the development of novel therapeutic modalities. The occurrence and severity of EAU are assessed by clinical and histopathological examinations, although histological findings reveal inflammatory processes in eyes with EAU in greater detail compared with clinical observation. However, the animal has to be euthanised for histological examinations, which precludes serial observation of disease progression at different time points and monitoring of therapeutic or adverse effects in the same individuals or tissues. Optical coherence tomography (OCT) is a non-invasive and reliable test, and spectral-domain OCT (SD-OCT) provides high-resolution cross-sectional images of ocular tissues in vivo and in real time. Recent studies have verified that SD-OCT provides accurate in vivo images of the mouse retina, which are compatible with or analogous to findings obtained from histological sections.9–11 In this study, we examined the utility of SD-OCT in visualising and monitoring in vivo progressive changes in mouse EAU, and developed an OCT-based grading system for EAU severity.
Materials and methods
Six- to eight-week-old female C57BL/6 (B6) mice were obtained from Japan CLEA Inc (Shizuoka, Japan). All procedures were carried out according to the Association for Research in Vision and Ophthalmology resolution on the use of animals in research.
Pentobarbital sodium was purchased from Kyorin Pharmaceutical, Tokyo, Japan. Peptide based on human IRBP amino acid sequence 1–20 (hIRBP 1–20; GPTHLFQPSLVLDMAKVLLD) was purchased from Takara Bio Co Ltd (Shiga, Japan). Complete Freund's adjuvant (CFA) and Mycobacterium tuberculosis strain H37Ra were obtained from Difco (Detroit, Michigan, USA). Purified Bordetella pertussis toxin (PTX) were purchased from Sigma-Aldrich (St Louis, Missouri, USA).
Induction of EAU
After anaesthesia by intraperitoneal injection of pentobarbital sodium (6.5 mg/mL, 0.12–0.14 mL/mouse), mice were immunised subcutaneously in the neck region with 0.2 mL of an emulsion of 200 µg of hIRBP 1–20 in CFA (1:1 wt/vol) containing 5 mg/mL M tuberculosis H37Ra. Concurrent with immunisation, 1 µg of PTX was injected intraperitoneally.12
Clinical scoring of EAU
On days 7, 14, 21 and 28 after immunisation, mice were anaesthetised by intraperitoneal pentobarbital sodium. Pupils were dilated with an eye drop containing 0.5% tropicamide and 0.5% phenylephrine hydrochloride (Mydrin-P, Santen Pharmaceutical, Osaka, Japan), and funduscopic examinations were performed. The clinical severity of EAU in each eye was assessed by three ophthalmologists, who graded severity on a scale of 0–4 according to the modified criteria for EAU in B6 mice (table 1) based on a previous report.13 When the three ophthalmologists gave different grades for the same mouse, the mean grade was calculated.
Histological scoring of EAU
Mice were euthanised and eyes were enucleated for histological analysis on days 7, 14, 21 and 28. After marking the orientation, the eyes were fixed overnight in Bouin's solution and embedded in paraffin. Sections 6 µm in size were prepared and stained with H&E. The histological severity of EAU in each eye was assessed by three ophthalmologists and scored on a scale of 0–4 in half-point increments, according to the modified criteria for EAU in B6 mice (table 1) based on previous reports.13 ,14 Briefly, the minimal criterion to score an animal as EAU positive by histopathology was the presence of inflammatory infiltration in the vitreous or retina. Progressively higher grades were assigned for the presence of discrete lesions in ocular tissues, such as vasculitis, granuloma, retinal folding and/or detachment, and photoreceptor damage. The grading system takes into account lesion type, size and number.
After anaesthesia and dilation of pupils, hydroxyethyl cellulose gel (Scopisol; Senju Pharmaceutical Co Ltd, Osaka, Japan) was dropped on the cornea, followed by placement of a Cantor Micro-M lens with base curve of 1.7 mm (CANTOR + NISSEL, Nothants, UK) to obtain better quality OCT images of the mouse retina. SD-OCT images were acquired using a Heidelberg Spectralis HRA+OCT system (Heidelberg Engineering, Heidelberg, Germany). Each two-dimensional B-scan, recorded with the equipment set at 30° field of view, consisted of 1536×496 pixels acquired at a speed of 40 000 A-scans per second. Axial resolution was 7 µm, with digital resolution reaching 3.5 µm. The combination of scanning laser retinal imaging and SD-OCT enables real-time tracking of eye movements and real-time averaging of OCT scans, which reduce speckle noise in the OCT images considerably. The severity of EAU in SD-OCT images was graded on a scale of 0–4 by the criteria shown in table 1, which we have newly established.
Correlation between grades scored by different methods was analysed by linear regression analysis and significance of the regression was tested. Statistical analysis was performed using JMP 10 (SAS Institute, Cary, North Carolina, USA). A p value less than 0.05 was considered significant.
Clinical course of EAU after hIRBP1–20 immunisation
Figure 1 shows the clinical course of occurrence and severity of EAU in individual mice after hIRBP1–20 immunisation. EAU occurred in 11 of 13 mice (84.6%) at the first week after immunisation, and was observed in all immunised mice at the second week (figure 1A). Regarding the severity of EAU, clinical grade was 0.48 at the first week, which gradually increased to 1.71 at the second week, 2.25 at the third week and 2.40 at the fourth week (figure 1B).
SD-OCT images in normal mice
Figure 2 shows a SD-OCT image and the corresponding histological section of the retina of the same normal mouse. The retinal layer structures were distinctively depicted on SD-OCT image and correlated well with the histological section. The nerve fibre layer, inner plexiform layers and outer plexiform layer were shown as highly reflective bands, and the inner and outer nuclear layers as less reflective bands. The junction between the inner segment and outer segment of the photoreceptors (IS/OS line) was depicted as a highly reflective line, similar to that observed in the human eye. The retinal pigment epithelium was seen as an even more highly reflective line outside the IS/OS line, and the border with the choroid was unclear.
Correlation of SD-OCT images with clinical and histological findings in mice with EAU
The clinical findings, SD-OCT images, and histological sections for the same mice with EAU at different grades are shown in figure 3. In mice with mildly dilated vessels from minimal vasculitis equivalent to grade 0.5 in clinical scoring (figure 3B), few infiltrating cells in the vitreous were observed in the histological section (figure 3H), but no abnormal findings were found on SD-OCT image (figure 3N). However, in mice clinically scored as grade 1 with more dilated retinal vessels and redness of optic disc (figure 3C), increased inflammatory cells in the vitreous and retinal vasculitis were found in the histological section (figure 3I). On the SD-OCT image corresponding to the area of the histological section, highly reflective dots in the vitreous and a highly reflective mass in the superficial layer of the retina were observed (figure 3O). In mice with clinical grade 2 EAU, clinical examination revealed extensive retinal vasculitis and multiple retinal exudates (figure 3D), and histological sections showed increased vitreous inflammatory cells, more severe perivasculitis, and granuloma formation in the outer retina (figure 3J). These histological findings were accurately depicted on SD-OCT image, which showed an increase in the number of highly reflective dots indicating infiltrating cells in the vitreous and around the retinal vessel, and granuloma as a highly reactive mass (figure 3P). In addition, the external limiting membrane (ELM) and IS/OS line were obscure. In clinical grade 3 EAU, clinical examination showed more severe vasculitis, extended retinal exudates and retinal haemorrhages (figure 3E), while histological sections demonstrated infiltration of many inflammatory cells into all retinal layers, multiple granulomas and disturbance of the retinal layer structure (figure 3K). SD-OCT images depicted increased number of highly reflective dots, disturbance of the retinal layer structure, multiple highly reflective masses in the outer retina and unclear outer retinal structure (figure 3Q). In addition, SD-OCT images showed disappearance of ELM and IS/OS line. In clinical grade 4 EAU, clinical examination revealed a pale optic disc and narrowed retinal vessels (figure 3F). Histological section (figure 3L) and SD-OCT image (figure 3R) showed decreased cells infiltrating the vitreous and retina, and disappearance of the retinal outer layers such as the outer nuclear layer and photoreceptor layer. These results indicated that clinical severity of EAU assessed by slit lamp microscope examination matched well with findings on SD-OCT images, and the SD-OCT findings reflected the histological findings accurately.
Correlation of SD-OCT grade of EAU with clinical and histopathological grades
The correlation between SD-OCT grade of EAU and the clinical and histopathological grades are shown in figure 4. Fourteen B6 mice were immunised with IRBP1–20. At the third, fourth or fifth weeks, mice were randomly chosen and EAU severity was scored by clinical findings and by SD-OCT findings based on the criteria shown in table 1. After clinical and SD-OCT evaluations, mice were euthanised on the same day for histopathological evaluation of EAU. The SD-OCT grade correlated significantly with the clinical grade (R2=0.282, p<0.005) and with histopathological grade (R2=0.846, p<0.0001), and significant correlation between clinical and histopathological grades (R2=0.231, p<0.05).
Acquisition of OCT images is more difficult in rodents than in humans. Because the size of the eye is smaller and anaesthesia is required, it is difficult to focus on the retina guided by scanning laser ophthalmoscope. However, since Srinivasan et al15 reported successful OCT imaging of the retina in rats and mice, OCT has been used in studies of rodent models of retinal degeneration and dysfunction.16–18 Seiler et al19 used OCT to image retinal transplants in live animals and found that OCT helps to evaluate the placement and structural quality of the transplants. In animal models of uveitis, Gadjanski et al20 first demonstrated the usefulness of OCT for observing progressive changes during the development of EAU in brown Norway rats. They reported that in the rat EAU model, severe anterior inflammation develops during the active phase, which impedes the acquisition of precise OCT images of individual retinal layers. Recently, the use of SD-OCT to evaluate mouse EAU has been reported by Chen et al21 and Chu et al.22 These two investigations, together with the present study, establish that the wide range of retinal pathology detected by OCT matches the findings of histological sections and fundus imaging. Chen et al21 examined the efficacy of SD-OCT in evaluating retinal pathology of EAU in B10RIII mice that differs in severity and time course to the EAU model of B6 mice. However, Chu et al22 demonstrated the utility of SD-OCT to investigate retinal vasculitis and leukocyte infiltration into the retina during EAU developed in C57BL/6 mice, but changes in the retinal outer layers including granuloma formation was not shown. Furthermore, to distinguish our study from the paper by Chu et al, we have indicated the correlation between OCT grades of EAU developed in B6 mice based on the criteria we have established and conventional clinical and histopathological grades in the present study. A highly significant correlation of the new SD-OCT grade with the clinical and the histological grades indicates that the SD-OCT scoring system may be an alternative tool for routine evaluation of EAU.
We found some technical limitations of SD-OCT for monitoring mouse EAU, especially in detecting subtle pathological changes during the early stage of EAU with minimal inflammation, corresponding to the clinical or histological grade 0.5. Ophthalmoscopic examination allows extensive observation of the fundus, and histological sections prepared from different regions of the eye permit detection of minimal inflammatory changes in the entire eye. Although SD-OCT images depict precisely the retinal changes shown in histological sections in EAU, the images are limited to the posterior pole of the fundus. Therefore, if inflammatory lesions are focal (such as in grade 1 EAU) and outside the posterior area, SD-OCT is not able to capture the changes. In addition, although infiltrating cells in the choroid are observed in histological sections of mice with grade 1 EAU or higher, choroidal inflammation was obscured by the high reflection of the retinal pigment epithelium on SD-OCT images. Nevertheless, these issues would be overcome by newly developed OCT technology in the near future.
In summary, this study demonstrates that SD-OCT depicts cell infiltration into the vitreous and retina, retinal vasculitis and retinal structural changes in EAU, providing information of pathological changes conventionally shown in histological sections without euthanising the mice. Thus, SD-OCT allows serial real-time monitoring of EAU development over time, and would be useful for evaluating therapeutic effects of new agents using the EAU model.
Contributors MT contributed to study design. KH, YK and MI performed the experiments. KH, YS and MT analysed the data. KH wrote the manuscript, and the remaining authors reviewed the manuscript critically.
Funding This work was supported in part by Grant-in-Aid 24592689 for Scientific Research from the Japan Society for the Promotion of Science.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.