Here, we describe an in vivo imaging technique using optical coherence tomography to facilitate the diagnosis and quantitative measurement of retinopathy in mice.
Optical coherence tomography (OCT) offers a noninvasive method for the diagnosis of retinopathy. The OCT machine can capture retinal crosssectional images from which the retinal thickness can be calculated. Although OCT is widely used in clinical practice, its application in basic research is not as prevalent, especially in small animals such as mice. Because of the small size of their eyeballs, it is challenging to conduct fundus imaging examinations in mice. Therefore, a specialized retinal imaging system is required to accommodate OCT imaging on small animals. This article demonstrates a small-animal-specific system for OCT examination procedures and a detailed method for image analysis. The results of retinal OCT examination of very-low-density lipoprotein receptor (Vldlr) knockout mice and C57BL/6J mice are presented. The OCT images of C57BL/6J mice showed retinal layers, while those of Vldlr knockout mice showed subretinal neovascularization and retinal thinning. In summary, OCT examination could facilitate the noninvasive detection and measurement of retinopathy in mouse models.
Optical coherence tomography (OCT) is an imaging technique that can provide in vivo high resolution and crosssectional imaging for tissue1,2,3,4,5,6,7,8, especially for the noninvasive examination in the retina9,10,11,12. It can also be used to quantify some important biomarkers, such as retinal thickness and retinal nerve fiber layer thickness. The principle of OCT is optical coherence reflectometry, which obtains crosssectional tissue information from the coherence of light reflected from a sample and converts it into a graphic or digital form through a computer system7. OCT is widely used in ophthalmology clinics as an essential tool for diagnosis, follow-up, and management for patients with retinal disorders. It can also provide insight into the pathogenesis of retinal diseases.
In addition to clinical applications, OCT has also been used in animal studies. Although pathology is the gold standard of morphological characterization, OCT has the advantage of noninvasive in vivo imaging and longitudinal follow-up. Furthermore, it has been shown that OCT is well correlated with histopathology in retinopathy animal models11,13,14,15,16,17,18,19,20. The mouse is the most commonly used animal in biomedical studies. However, its small eyeballs pose a technical challenge to conducting OCT imaging in mice.
Compared to the OCT first used for retinal imaging in mice21,22, OCT in small animals has now been optimized with respect to hardware and software systems. For example, OCT, in combination with the tracker, significantly reduces the signal-to-noise ratio; OCT software system upgrades allow more retinal layers to be detected automatically; and the integrated DLP beamer helps to reduce the motion artifacts.
Very-low-density lipoprotein receptor (Vldlr) is a transmembrane protein in endothelial cells. It is expressed on retinal vascular endothelial cells, retinal pigment epithelial cells, and around the outer limiting membrane23,24. Subretinal neovascularization is the phenotype of Vldlr knockout mice23. Therefore, Vldlr knockout mice are used to investigate the pathogenesis and potential therapy of subretinal neovascularization. This article demonstrates the application of OCT imaging to detect retinal lesions in Vldlr knockout mice, hoping to provide some technical reference for retinopathy research in small animal models.
The operations were performed following the Statement on the Use of Animals in Ophthalmic and Vision research from the Association for Research in Vision and Ophthalmology. The experimental design was approved by the institutional animal Ethics Committee (Medical Ethics Committee of JSIEC, EC 20171213(4)-P01). Two-month-old C57BL/6J mice and Vldlr knockout mice were used in this study. There were 7 mice in each group, all of which were female and weighed 20 g to 24 g.
1. Experimental conditions
2. Information records
3. Instrument startup and testing
4. Anesthesia
5. Application of mydriatic drops
6. Placement of the mouse
7. Confocal Scanning Laser Ophthalmoscope (cSLO)
8. Optical coherence tomography (OCT)
9. The end of the experiment (after the OCT examination)
10. Image analysis
11. Retinal stratification correction
12. Retinal lamination thickness
13. Measurement of full retinal thickness
Thanks to the high-resolution scans of OCT, the layers of the mouse retina can be observed, and abnormal reflections and their exact locations can be identified. The retinal OCT images of Vldlr knockout mice and C57BL/6J mice were compared in this study. The OCT images of all C57BL/6J mice showed various retinal layers with different reflectivity, and the demarcation was clear (Figure 8D). In contrast, all Vldlr knockout mice showed abnormal, hyperreflective lesions on the OCT images (Figure 8B).
Incomplete vitreous detachment (PVD) in Vldlr knockout mice
The OCT results showed some middle reflective bands on the retinal surfaces of Vldlr knockout mice (Figure 8B, red arrows). These middle reflective bands adhered to the retinal vessel (Figure 8B, green arrow), corresponding to the cSLO image (Figure 8A, green arrow). These features are consistent with the OCT characteristics of incomplete vitreous detachment.
Subretinal neovascularization in Vldlr knockout mice
The results showed that subretinal neovascularization had two development modes in the Vldlr knockout mice.
With involvement of the outer nuclear layer
A hyperreflective lesion, with a bottom-down triangular shape on the OCT image, appeared on the subretinal space and spread to the outer nuclear layer. The lesion did not break through the outer plexiform layer (Figure 8B, white arrow).
The OCT appearance of this type of subretinal neovascularization was consistent with the pathological findings shown in Figure 9A. The pathological section showed that neovascularization (Figure 9A, thick green arrow) broke through the RPE, photoreceptor inner/outer segments (IS/OS), and the external limiting membrane (ELM). It invaded the outer nuclear layer (ONL) but did not break through the outer plexiform layer (OPL).
Without involvement of the outer nuclear layer
A band of hyperreflective lesion appeared on the OCT image, which was located at the subretinal space (Figure 8B, yellow arrow). The cSLO image showed the corresponding location (Figure 8A, yellow arrow). The additional scans of the retina around this location (Figure 8A, yellow arrow) showed the same findings.
Consistent with the lesion (Figure 10A, thick blue arrow) in the pathological section, this subretinal neovascularization did not break through the ELM (Figure 10A, thin yellow arrow) but partially involved the photoreceptor IS/OS.
Retinal thickness results
The retinal thickness of the right eye of all mice was obtained by using the automatic stratification and thickness measurement function of OCT. The retinal thickness of Vldlr knockout mice (200.94 ± 14.64 µm) was significantly lower than that of C57BL/6J mice (217.46 ± 10.21 µm, P < 0.001, t-test, 7 right eyes/group). The comparison of retinal thickness in the four directions (temporal, nasal, superior, and inferior) of the posterior polar between the two groups is shown in Figure 11.
Figure 1: Preparation of experimental materials and animals. (A) Equipment: 1. cSLO/OCT device for small-animal retinal imaging, 2. computer and monitor, 3. Small, constant-temperature animal platform, 4. thermostat, 5. preset lens, 6. installation of the preset lens. (B) Medicines and small items: I. povidone-iodine, II. microsyringe, III. anesthetic mixture solution, IV. timer, V. mydriatic eye drops, VI. forceps, VII. medical sodium hyaluronate gel, VIII. medical cotton swab, IX. antibiotic eye ointment, X. 100 D contact lens (two). (C) Weight measurement on a digital balance. Abbreviations: cSLO = confocal scanning laser ophthalmoscope; OCT = optical coherence tomography. Please click here to view a larger version of this figure.
Figure 2: Preparation before OCT examination of mice. (A) Mydriasis eye drop application, (B) sodium hyaluronate gel coating on the cornea, (C, D) placement of a 100 D contact lens, with concave surface contacting the cornea. Abbreviation: OCT = optical coherence tomography. Please click here to view a larger version of this figure.
Figure 3: OCT examination procedures. (A) Mouse position placement, I. preset lens, II. contact lens, III. Small, constant-temperature animal platform. (B) Operation of the cSLO/OCT machine, IV. operating lever, V. tilt lever, VI. cSLO device. Abbreviations: cSLO = confocal scanning laser ophthalmoscope; OCT = optical coherence tomography. Please click here to view a larger version of this figure.
Figure 4: OCT imaging process. A. Measurement mode, B. Start Laser of the IR laser, C. eye selection (C-1-OD; C-2-OS), D. range of IR laser, E. the diopter, F. overlay of the cSLO image, G. OCT scanning start/stop laser button H. reference of OCT image, I. Range Min: 0-20, J. Range Max: 50-60, K. signal intensity of the image, L. scanning direction (e.g., vertical scan), M. scanning position selected by moving the green reference line (e.g., vertical scan through the optic papilla), N. real-time display of the OCT image, O. overlay of the OCT image, P. Shot: image acquisition, Q. SLO-OCT images that have been acquired, R. Save Examination: saving the examination result. Scale bars = 200 µm. Abbreviations: cSLO = confocal scanning laser ophthalmoscope; OCT = optical coherence tomography; IR = infrared; OD = right eye; OS = left eye. Please click here to view a larger version of this figure.
Figure 5: Automatic retinal delamination interface on OCT system. A. Load Examination button, B. Media Container, showing all the OCT images, C. OCT image being selected for analysis, D. Layer Detection button for automatic retinal layering, E. dividing line list, F. automatic delamination on the retina, G. Edit Layer button for layered correction, H. Save Examination button for saving the results. Scale bars = 200 µm. Abbreviation: OCT = optical coherence tomography. Please click here to view a larger version of this figure.
Figure 6: Layered correction (A-C) and thickness measurement (D-E). (A) Layered edit activation interface: 1. Edit Layer button, 2. dividing line list (e.g., selecting all lines), 3. activated dividing lines, 4. Spacing adjustment, 5. Limit Range adjustment. (B) Activation of a dividing line (e.g., line 3 in A), 6. line 3, the line between the inner plexiform layer and inner nuclear layer, 7. an example of layering error. (C) Layering error modification, 8. the red circle for modification. (D) An example of retinal lamellar thickness measurement, 9. Measure Marker button, 10. dividing lines of the outer nuclear layer, 11. Connect with Layer (the measurement will connect with the layer according to the dividing lines), 12. Stay Connected on Move (the measurement position is where the manual click stays), 13. the location of the result display, 14. the measurement line (perpendicular to the horizontal axis). (E) Measurement result acquisition, 15. the measurement results (red rectangle: Vert value is the thickness result), 16. Delete Marker button for measurement record deletion, 17. New Marker button for remeasurement (the new result will overwrite the original record). Scale bars = 200 µm. Please click here to view a larger version of this figure.
Figure 7: Measurement of full retinal thickness. A. Measure Marker button, B. line 1 (ILM) and C. line 7 (OS-RPE) selection for showing the boundaries of the full-thickness retina, D. Connect with Layer selection, E. Stay Connected on Move selection, F. ruler bar (vertical and horizontal ruler bars, both 200 µm in length), G. measurement lines on the retina (4 lines with 200 µm of horizontal ruler length as spacing on each side of the optic papilla), H. the measurement results (the results are differentiated by different colors and correspond to the color of the measurement lines on the retina), I. Data extraction from the Vert value in the Length in µm (tissue) row. Scale bars = 200 µm. Abbreviations: ILM = inner limiting membrane; OS-RPE = photoreceptor outer segment of retinal pigment epithelium. Please click here to view a larger version of this figure.
Figure 8: Comparison of cSLO and OCT images of Vldlr knockout and C57BL/6J mice. cSLO (A) and OCT (B) images of Vldlr knockout mice compared with the cSLO (C) and OCT (D) images of C57BL/6J mice. Characteristics of OCT in Vldlr knockout mice (B): 1) Middle reflective line (B, red arrows) on the inner surface of the retina with adhesion to the retinal vessel (B, green arrow). 2) Hyperreflective lesions, located at the subretinal space, with (B, white arrow) or without (B, yellow arrow) involvement of outer nuclear layer. The arrows on the cSLO image (A) represent the locations of the corresponding color arrows on OCT image (B). Scale bars = 200 µm. Abbreviations: cSLO = confocal scanning laser ophthalmoscope; OCT = optical coherence tomography; Vldlr = very-low-density lipoprotein receptor. Please click here to view a larger version of this figure.
Figure 9: Mode 1: retinal paraffin sections with hematoxylin-eosin staining in Vldlr knockout and C57BL/6J mouse. (A) An example of subretinal neovascularization invading the outer nuclear layer (thick green arrow), located in the middle part of the retina of a Vldlr knockout mouse. (B) Normal control, the middle part of the retina of a C57BL/6J mouse. Scale bars = 50 µm. Abbreviations: Vldlr = very-low-density lipoprotein receptor; ILM = inner limiting membrane; NFL = retinal nerve fibre layer; GCL = retinal ganglion cell layer; IPL = inner plexiform layer; INL = inner nuclear layer; OPL = outer plexiform layer; ONL = outer nuclear layer; ELM = external limiting membrane; IS = photoreceptor inner segment; OS = photoreceptor outer segment; RPE = retinal pigment epithelium layer. Please click here to view a larger version of this figure.
Figure 10: Mode 2: retinal paraffin sections with hematoxylin-eosin staining in Vldlr knockout and C57BL/6J mouse. (A) An example of subretinal neovascularization without the involvement of outer nuclear layer (thick blue arrow) and with intact ELM (thin yellow arrow), located in the middle periphery retina in a Vldlr knockout mouse. (B) Normal control, the middle periphery retina of a C57BL/6J mouse. Scale bars = 50 µm. Abbreviations: VLDR = very-low-density lipoprotein receptor; ILM = inner limiting membrane; NFL = retinal nerve fiber layer; GCL = retinal ganglion cell layer; IPL = inner plexiform layer; INL = inner nuclear layer; OPL = outer plexiform layer; ONL = outer nuclear layer; ELM = external limiting membrane; IS = photoreceptor inner segment; OS = outer photoreceptor segment; RPE = retinal pigment epithelium layer. Please click here to view a larger version of this figure.
Figure 11: Comparison of retinal thickness between C57BL/6J mice and Vldlr knockout mice (all data from the right eye). (A) Retinal thickness (µm) through the optic nerve papilla by OCT horizontal scanning. (B) Retinal thickness (µm) through the optic nerve papilla by OCT vertical scanning. The horizontal coordinate represents the measuring positions with spacing of 200 µm.*: P < 0.05, **: P < 0.01, ***: P < 0.001. Abbreviations: T = Temporal; P = Optic papilla; N = Nasal; S = Superior; I = Inferior; OCT = optical coherence tomography; VLDR = very-low-density lipoprotein receptor; OD = right eye. Please click here to view a larger version of this figure.
In this study, OCT imaging using a small-animal retinal imaging system was applied to evaluate retinal changes in Vldlr knockout mice, which demonstrate incomplete posterior vitreous detachment, subretinal neovascularization, and retinal thickness thinning. OCT is a noninvasive imaging method to examine the condition of the retina in vivo. Most OCT devices are designed for human eye examination. The size of the hardware equipment, the setting of the focal length, the setting of the system parameters, and the positioning requirements of the examinee are all based on the human eye. Modifications of the lens and system settings are required to examine small animals with human-specific OCT equipment. This paper presents small-animal OCT examination procedures.
The focal length is different during image scanning of different small animals with different sizes of eyeballs. This difference in focal length is critical and must be resolved to obtain clear and accurate fundus images. One effective method is replacing the objective lens with lenses of different curvatures. Due to its small eyeball, the mouse needs a contact lens of 100 D in front of the cornea in addition to the double-spherical 60 D preset lens of the OCT equipment.
The OCT can only provide line scans that only cover a limited region of the retina. Therefore, it is essential to standardize the protocol of OCT scans for qualitative and quantitative comparison of OCT findings in different groups. Three horizontal scans and three vertical scans were performed here. This machine provides a real-time cSLO image to monitor the location of the OCT scan so that the position of the scan can be adjusted accurately and conveniently. Additional scans can be added when an abnormal reflection is found.
The parameters of image acquisition need to be adjusted carefully. Here, it is recommended that the Range Min be 0-20 and the Range Max be 50-60 (Figure 4I, J). When the parameters are overadjusted, the signal contrast of the image would be enhanced, and the reflected signal of the retina with low reflection becomes lower or even black, and some morphological information will be lost.
The following are some tips to avoid image quality deterioration: 1. Place a contact lens in front of the eyes immediately after anesthesia to avoid cataracts; 2. Ensure that the preset lens and contact lens are clean; 3. Avoid hair entering between the cornea and the contact lens; 4. Ensure the doppler, contrast, and brightness in the OCT parameters are set appropriately.
The OCT images can be used to qualitatively detect lesions and quantitatively measure metrics such as retinal thickness. Here, a method is proposed to measure the retinal thickness at several locations, and the average can be calculated as the mean retinal thickness. This is achieved through the automatic stratification function of the OCT system. Therefore, the thickness of the retinal laminations can also be measured. The measurement method is simple and accurate (Figure 6 and Figure 7). The results showed that the retinal thickness was lower in Vldlr knockout mice than C57BL/6J mice, consistent with the literature25. The difference in retinal thickness between the two groups can be clearly shown by a graph generated from the measurements at multiple locations (Figure 11). Similar retinopathy analysis and retinal thickness measurement methods have also been reported in the Stargardt disease mouse model26. However, it is worth noting that the hyperreflective bands at the vitreous interface of the retina do not belong to the retinal tissue and should be removed during stratification. In addition, if subretinal lesions invade the retina, the thickness measurement should include the invaded portion.
This small-animal retinal imaging system has some limitations. For example, although it can provide clear images of the posterior pole within 35°, image acquisition of the peripheral retina is still challenging. In addition, cSLO forms a gray-scale image, which is not as good as a color fundus image to detect fundus lesions (pigmentation, bleeding, exudation). Hence, further improvements are needed. In summary, OCT examination by the cSLO machine could facilitate the noninvasive detection and measurement of retinopathy in mouse models.
The authors have nothing to disclose.
Project Source: Natural Science Foundation of Guangdong Province (2018A0303130306). The authors would like to thank the Ophthalmic Research Laboratory, Joint Shantou International Eye Center of Shantou University and the Chinese University of Hong Kong for funding and materials.
100-Dpt contact lens | Volk Optical,Inc, Mentor, OH | Accessory belonging to the RETImap | |
Double aspheric 60-Dpt glass lens | Volk Optical,Inc, Mentor, OH | Accessory belonging to the RETImap | |
Electric heating blanket | POPOCOLA | CW-DRT-01 | 50 x 35 cm |
Injection syringe (1 mL) | Kaile | 0.45 x 16RWLB | |
Levofloxacin Hydrochloride Eye Gel | EBE PHARMACEUTICAL Co.LTD | 5 g: 0.015 g | |
Medical sodium hyaluronate gel | Alcon | 16H01E | |
Microliter syringes | Shanghai high pigeon industry and trade co., LTD | Q31/0113000236C001-2017 | 50 µL |
Povidone iodine solution | Guangdong medihealth pharmaceutical Co.,LTD | 100 mL | |
RETImap | ROLAND CONSULT | 19-99_50-2.1_1.2E | cSLO/ERG/VEP/FA/OCT/GFP |
Small animal ear studs | OSMO POCKET OT110 | INS1005-1S | |
Tropicamide Phenylephrine Eye Drops | Santen Pharmaceutical Co.,LTD | 5 mg/mL | |
Xylazin | Sigma | X1251-5G | 5 g |
Zoletil 50 | Virbac.S.A | 7FRPA | Tiletamine 125 mg + Zolazepam 125 mg |