Imaging the Retinal Vascular Mural Cells In Vivo: Elucidating the Timeline of Their Loss in Diabetic Retinopathy
Arteriosclerosis, Thrombosis, and Vascular Biology
Abstract
BACKGROUND:
Vascular mural cells (VMCs) are integral components of the retinal vasculature with critical homeostatic functions such as maintaining the inner blood-retinal barrier and vascular tone, as well as supporting the endothelial cells. Histopathologic donor eye studies have shown widespread loss of pericytes and smooth muscle cells, the 2 main VMC types, suggesting these cells are critical to the pathogenesis of diabetic retinopathy (DR). There remain, however, critical gaps in our knowledge regarding the timeline of VMC demise in human DR.
METHODS:
In this study, we address this gap using adaptive optics scanning laser ophthalmoscopy to quantify retinal VMC density in eyes with no retinal disease (healthy), subjects with diabetes without diabetic retinopathy, and those with clinical DR and diabetic macular edema. We also used optical coherence tomography angiography to quantify capillary density of the superficial and deep capillary plexuses in these eyes.
RESULTS:
Our results indicate significant VMC loss in retinal arterioles before the appearance of classic clinical signs of DR (diabetes without diabetic retinopathy versus healthy, 5.0±2.0 versus 6.5±2.0 smooth muscle cells per 100 µm; P<0.05), while a significant reduction in capillary VMC density (5.1±2.3 in diabetic macular edema versus 14.9±6.0 pericytes per 100 µm in diabetes without diabetic retinopathy; P=0.01) and capillary density (superficial capillary plexus vessel density, 37.6±3.8 in diabetic macular edema versus 45.5±2.4 in diabetes without diabetic retinopathy; P<0.0001) is associated with more advanced stages of clinical DR, particularly diabetic macular edema.
CONCLUSIONS:
Our results offer a new framework for understanding the pathophysiologic course of VMC compromise in DR, which may facilitate the development and monitoring of therapeutic strategies aimed at VMC preservation and potentially the prevention of clinical DR and its associated morbidity. Imaging retinal VMCs provides an unparalleled opportunity to visualize these cells in vivo and may have wider implications in a range of diseases where these cells are disrupted.
Graphical Abstract
Highlights
•
Loss of retinal arteriolar smooth muscle cells occurs before any clinical manifestations of diabetic retinopathy and well before capillary pericyte loss.
•
Capillary pericyte loss is a later event in the pathophysiologic course of diabetic retinopathy and is associated with the classical clinical and microvascular manifestations of diabetic retinopathy.
•
Leakage in diabetic macular edema may be driven by both capillary pericyte loss and another potential second hit (eg, retinal nonperfusion and ischemia).
Vascular mural cells (VMCs) are a heterogenous cell population representing a range of phenotypes that reflect their distinct functional roles in various tissues and along the vascular tree.1 VMCs have a variety of functions throughout the body, including vessel stabilization, endothelial support, and maintaining the endothelial barriers in specialized organs (eg, blood-brain barrier, blood-urine barrier in the kidney, and blood-retina barrier in the eye).1 Moreover, VMCs play an important role in vascular permeability, whereby VMC abundance is correlated with permeability.1 Thus, VMC dysfunction and loss have been linked to a variety of pathological conditions including cardiovascular disease, fibrosis, and diabetic nephropathy and retinopathy. In particular, VMCs are critical to the pathogenesis of diabetes as pericyte loss in retinal capillaries and loss of podocytes (which are pericyte-like cells) from glomerular capillaries are hallmarks of diabetic retinopathy (DR) and nephropathy, respectively.2
Previously, the role of human VMCs in diabetic disease could only be studied in postmortem tissue or in nonhuman models as in vivo visualization has been precluded by the limited resolution of standard imaging techniques.3 However, the introduction of specialized retinal imaging techniques that combine adaptive optics scanning laser ophthalmoscopy (AOSLO) with a modified dark-field detection scheme has allowed high-resolution imaging of the retinal microvasculature, including the VMCs, in the living human retina.3,4 In this study, we capitalize on these advancements to image retinal VMCs as a potential window into these perivascular cells elsewhere.
While larger retinal arterioles and venules are visible with low-resolution imaging techniques, their vascular walls, as well as smaller arterioles, capillaries, and venules are harder to visualize. AOSLO facilitates cellular-level resolution by correcting the optical aberrations of the eye.5 This technology has been used for imaging individual retinal cells including photoreceptors and ganglion cells.6,7 However, the perifoveal microvascular network is challenging to image due to the complexity of the retina and the backscattering signal of structures such as the retinal nerve fiber layer.4 Here, we use an offset confocal aperture configuration to capture multiply scattered light and facilitate visualization of the retinal microvascular wall, including the VMCs, in the living human retina.3,4 The multiply scattered light AOSLO images obtained in this study are an extension of dark-field imaging, where directly backscattered light is blocked while multiply scattered light passes through the aperture, resulting in features appearing as bright on a dark background.4,8 With this AOSLO imaging technique, we are able to image vessels in the most superficial layer of the retinal vasculature (superficial vascular plexus; Figure 1A and 1B) and to visualize the outer wall of these vessels, with the VMCs along the outer wall (Figure 1C and 1D).
The 2 main types of VMCs in the retina are smooth muscle cells (SMCs), which surround arterioles and venules, and pericytes surrounding the capillaries.9 Currently, DR is the leading cause of blindness in working-age adults,10 and histopathologic reports have suggested that pericyte loss may be the earliest pathological change in the pathophysiology of DR,11 leading to weakening of the vessel wall, and predisposing those vessels to outpouching and subsequent microaneurysm formation.12 Moreover, retinal arteriolar SMCs have been shown to be compromised in DR,13,14 where they play an important role in controlling ocular blood flow and vascular tone.15 Critically, the large body of donor eye studies does not address the timing of VMC loss in relation to other signs of DR, an important unresolved knowledge gap. Hence, there is a critical need to better define the timeline of VMC demise in human DR and elucidate whether these cells, including SMCs and pericytes, are critical to the onset and progression of the vascular compromise or whether they occur in parallel or perhaps follow other pathologies in DR.
We use AOSLO in the current study to address this knowledge gap and to elucidate the timing of VMC loss during diabetes in human eyes. To do this, we compared quantitative metrics of the retinal capillary density and VMC density in a cross-sectional cohort of healthy subjects, subjects with diabetic macular edema (DME), and subjects with diabetes without diabetic retinopathy (DM no DR). Given the extensive capillary remodeling in subjects with mild or moderate DR,16 we studied retinal capillary density using the complementary motion contrast approach of optical coherence tomography angiography (OCTA). With this work, we aimed to better understand how cellular microvascular changes contribute to the pathophysiology of early, preclinical DR, as well as DME. Furthermore, as pericytes have emerged as a potential therapeutic target in DR,17 this tool could help elucidate the optimal timing of preventative strategies aimed at preserving the pericytes (and their counterparts in the entire vascular system) to effectively impact DR and other microvascular complications in diabetes.
METHODS
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Study Sample
This prospective cross-sectional study was conducted in the Department of Ophthalmology at Northwestern University in Chicago, IL from June 2021 to March 2023. The study was approved by the Institutional Review Board of Northwestern University and conducted in accordance with the regulations of the Health Insurance Portability and Accountability Act and the tenets of the Declaration of Helsinki. Written informed consent was obtained from all participants before images were taken.
We recruited a cohort of healthy subjects with no retinal pathology (healthy), subjects with DM no DR, and subjects with DME, defined as apparent retinal thickening or hard exudates.18 Inclusion criterion was age between 18 and 80 years. For the DM-no-DR and DME groups, another inclusion criterion was diagnosis of either type 1 or type 2 diabetes. Additional inclusion criteria for the DME group included clinical assessment by board-certified ophthalmologists and either extrafoveal or center-involved DME based on optical coherence tomography central subfield thickness of >260 µm. Exclusion criteria were refractive error >3 diopters, presence of glaucoma, and prediabetes, defined as HbA1c (hemoglobin A1c) within the range 5.7% to 6.4%.
Next, we ascertained that subjects clinically deemed as DM no DR did not have any clinically apparent retinopathy on retinal imaging. One trained grader (B.B.H.) evaluated all available imaging studies, including infrared photography or pseudocolor fundus photographs obtained using the ultra-wide-field scanning laser ophthalmoscope (Optomap Panoramic 200; Optos PLC, Scotland, United Kingdom), according to the International Clinical Diabetic Retinopathy Disease Severity Scale.18 In this system, eyes with microaneurysms would be considered to have mild nonproliferative DR and thus would be excluded from the DM-no-DR study group. Ambiguous cases were resolved by discussion with a more senior grader (H.F.).
Demographic and clinical data (including age, sex, history of hypertension, duration of known diabetes [if applicable], most recent HbA1c [if applicable]) were obtained from the electronic medical records. Overall, the study included 40 subjects (17 healthy, 15 DM no DR, and 8 DME) and a total of 47 eyes (20 healthy, 19 DM no DR, and 8 DME).
Image Acquisition and Processing
First, all patients underwent pupil dilation, followed by measuring the axial length using IOLMaster 700 (ZEISS, Jena, Germany), to accurately scale our metrics by accounting for the magnification of each eye. Next, we obtained capillary-level retinal imaging using OCTA to identify the individual vessels of interest and to evaluate the overall macular capillary integrity. The OCTA images were centered on the fovea and acquired using the RTVue-XR Avanti system (Optovue, Inc, Fremont, CA) with split-spectrum amplicate-decorrelation angiography software (version 2017.1.0.151).19 The specifications of the machine included an A-scan rate of 70 000 scans/s, a light source centered on 840 nm, and a bandwidth of 45 nm.
Then, for each vessel of interest, we focused on the retinal vascular wall using the Apaeros retinal imaging AOSLO system (Boston Micromachines Corporation, Cambridge, MA) with a computer-controlled aperture.20 Vessels of interest were all located within a 6×6-mm2 OCTA scan centered on the fovea. To enhance vascular wall visibility, we used an offset confocal aperture as described previously.4 In short, the positioning of the aperture controls the formation of the retinal image by capturing multiply scattered and singly scattered light. Orthogonally displacing the aperture relative to the retinal vessel improved vascular wall visibility.3 For each location of interest, 100 video frames were collected, which were aligned and registered to a single frame to eliminate the distorting effect of small eye movements.3 The aligned images were then averaged to generate the final images used for further vascular wall analysis. In total, we imaged 64 different arterioles (28 healthy, 28 DM no DR, and 8 DME), and 75 distinct arteriolar regions of interest, often including vessels and their branches (32 healthy, 35 DM no DR, and 8 DME). For the capillary pericyte analysis, a subset of 22 eyes was studied (10 healthy, 7 DM no DR, and 5 DME).
Assessing Image Quality
There is currently no standardized, accepted approach to evaluate the quality of offset confocal aperture AOSLO images. This contrasts with OCTA, where metrics such as signal strength index can be used to evaluate image quality. Therefore, we curated a library of gold standard AOSLO images and identified certain characteristics of high-quality scans that would be acceptable for further analysis. The images focused on the center point of the arteriole of interest, facilitating VMC quantification, and chosen characteristics identified included a clearly visible vessel wall, visualization of the superficial capillary plexus (SCP) at the same plane, and visible branch vessels at the same plane as the chosen vessel (Figure 1D). Furthermore, we prioritized offset aperture images where the corresponding confocal image did not show retinal nerve fibers crossing over the vessel surface, as further evidence that these images are taken at the optimal depth inside the vessel lumen (Figure 1E). We used these 4 criteria (Figure 1D and 1E) to exclude low-quality AOSLO images.
Image Analysis
After excluding poor-quality AOSLO images, 1 grader (B.B.H.) used the corresponding OCTA image to classify each vessel as either an arteriole or venule by identifying vessels with a broad periarteriolar capillary-free zone. Ambiguous cases were resolved by discussion with another senior grader (H.F.). Venules were excluded from quantification of VMCs in this study, as venular VMCs were flat in contour and less amenable for quantification. The graders were masked to the health status of the subjects, whether they were associated with the healthy, DM-no-DR, or DME category.
Next, vessel size was calculated by measuring the external diameter of each vessel using ImageJ (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available in the public domain at http://rsb.info.nih.gov/ij/index.html). Then, for each vessel location, VMC density was quantified and reported as the number of VMCs per100 µm. Previous studies have described pericytes as having a “bump on a log appearance on the abluminal walls”21 or as “structures with single and focal spatial variations along the blood vessel wall.”3 In our analysis, VMCs were defined as cells with a protruding, prominent, well-defined cell body along the outer vessel wall (Figure 1C and 1D). Image brightness and contrast were adjusted to improve VMC and vessel wall visualization. We also noted whether one image had multiple vessels or whether one vessel was found in multiple images, and the relevant VMC density was reported accordingly.
The measured ocular axial length was used to correct for the individual variations in retinal magnification.22 For each eye, the magnification factor K (the relationship between pixel coordinates and retinal dimensions expressed in millimeters) was calculated as , where the actual vitreous chamber length is given by the axial length minus the standard distance in front of the lens. The standard distance in front of the lens is 7.3 mm, the model vitreous chamber length is 16.7 mm, and the standard number of µm in one degree is 289 µm.
To validate the accuracy of our VMC density measurements, intrarater and interrater correlation analyses were performed, with graders masked to the subject diagnosis. For intrarater correlation, 1 grader (B.B.H.) measured 10 images (mixture of healthy and diabetic eyes) twice, in a randomized fashion, and compared the results. For interrater correlation, 2 independent graders (B.B.H. and A.A.F.) graded the same mixed diagnosis 10 images. Finally, to further validate our measurements, we compared the VMC density values we obtained in healthy eyes with previously reported values in living human retinas,3 as well as postmortem human brains23 and rat retinas. We also used previously published ganglion cell images of healthy subjects from a recent article to impute VMC density.24
For each eye, we analyzed the 3×3-mm2 OCTA scans for potential capillary abnormalities that correlated with VMC loss. The OCTA parameters analyzed were quality index (Q score, which represents the quality of the image), foveal avascular zone area, SCP vessel density (VD), and deep capillary plexus VD. The Q score is a quality index measure ranging from 1 to 10 that is automatically calculated by the proprietary software of the device we used, which rates the image quality based on a combination of metrics including signal strength, motion artifacts, and defocus. Using the built-in AngioVue Analytics software, default segmentation parameters were used to segment the SCP and deep capillary plexus and extract the capillary VD and other parameters.
Statistical Analysis
Statistical analysis was performed using the Python software, version 3.8.8 (Python Software Foundation, Wilmington, DE). Two-way mixed intraclass correlation coefficient was used to assess intragrader and intergrader reliability for VMC density measurements. Shapiro-Wilk tests were performed to determine whether the demographic data are normally distributed. When making comparisons between 2 study groups, Student t test was used to compare continuous demographic variables (DM duration, last HbA1c), while the Fisher exact test was performed to compare categorical variables (DM type 1 or type 2). When making comparisons across the 3 study groups, 1-way ANOVA was used to compare continuous variables (age, axial length, and vessel diameter) and χ2 test was performed to compare categorical variables (sex, diagnosis of hypertension, and race/ethnicity). Potential associations between confounding variables and AOSLO measurements were evaluated using Pearson correlations. Student t test was used to compare differences in VMC density and OCTA parameters between the healthy and DM-no-DR groups, as well as between the DM-no-DR and DME groups. This comparison was performed on an individual patient basis (averaging all VMC densities obtained per patient), a vessel basis (averaging all VMC densities obtained per vessel), and a location basis. For some subjects, we obtained images of both eyes. In this case, we averaged all VMC densities for both eyes in the per-patient analysis. P values <0.05 were considered statistically significant.
RESULTS
Study Population
Overall, among the 3 study groups (17 healthy, 15 DM no DR, and 8 DME), age and hypertension diagnosis were statistically significantly different (P=0.002 and P=0.02, respectively; Table). Furthermore, compared with DM-no-DR subjects, DME subjects had a statistically significant longer duration of DM (P=0.046) and had a higher last HbA1c (P<0.001). All other demographic and clinical parameters (sex, DM type, axial length, race/ethnicity) were comparable between the 3 groups. Moreover, from the Shapiro-Wilk tests, all demographic data were normally distributed (P>0.05) except for DM duration (P=0.006). For the arterioles included in the study, the average diameters were 95.5±25.7 µm for healthy subjects, 91.6±13.7 µm for DM-no-DR subjects, and 68.9±24.9 µm for DME subjects (1-way ANOVA, P=0.02). For the capillaries included in this study, the average diameters were 6.9±1.5 µm for healthy subjects, 10.7±2.7 µm for DM-no-DR subjects, and 7.1±1.9 µm for DME subjects (1-way ANOVA, P=0.005).
| Subject characteristics | Healthy | DM no DR | DME | P value |
|---|---|---|---|---|
| Subjects, n | 17 | 15 | 8 | … |
| Eyes, n | 20 | 19 | 8 | … |
| Vessels, n | 28 | 28 | 8 | … |
| Locations, n | 32 | 35 | 8 | … |
| Age, y; mean±SD (range) | 38.4±10.8 (18–59) | 48.4±12.14 (25–67) | 56.8±10.2 (36–72) | 0.002*† |
| Sex, n | 0.35‡ | |||
| Female | 12 | 7 | 4 | |
| Male | 5 | 8 | 4 | |
| Race/ethnicity, n | 0.08‡ | |||
| Non-Hispanic White | 14 | 9 | 3 | |
| Non-Hispanic Black | 0 | 3 | 4 | |
| Hispanic | 2 | 1 | 1 | |
| Non-Hispanic Asian | 1 | 2 | 0 | |
| DM, n | 0.62§ | |||
| Type 1 | … | 4 | 1 | |
| Type 2 | … | 11 | 7 | |
| Time since diagnosis of DM, y; mean±SD | … | 8.3±8.5 | 16.4±7.9 | 0.046*‖ |
| DR severity | … | … | 1 mild NPDR, 2 moderate NPDR, 2 severe NPDR, and 3 PDR | … |
| Last HbA1c, mean±SD | … | 6.3±0.9 | 7.9±1.5 | <0.001*‖ |
| Hypertension diagnosis | 3 (18%) | 7 (47%) | 6 (75%) | 0.02*‡ |
| Axial length, mm; mean±SD | 23.9±1.0 | 24.2±0.9 | 23.7±1.3 | 0.54† |
DM indicates diabetes; DM no DR, diabetes without diabetic retinopathy; DME, diabetic macular edema; HbA1c, hemoglobin A1c; NPDR, nonproliferative diabetic retinopathy; and PDR, proliferative diabetic retinopathy.
*
Statistically significant.
†
One-way ANOVA (comparing healthy vs DM no DR vs DME).
‡
χ2 test (comparing healthy vs DM no DR vs DME [for the χ2 test for race/ethnicity, non-Hispanic Black, Hispanic, and non-Hispanic Asian were grouped together as all are racial/ethnic minority groups]).
§
Fisher exact test (comparing DM no DR vs DME).
‖
Student t test (comparing DM no DR vs DME).
For the DME subjects, the optical coherence tomography central subfield thickness ranged from 261 to 734 µm, with an average of 372±154 µm. In terms of stage of DR severity in the DME eyes, they were widely distributed including 1 eye with mild nonproliferative DR, 2 each with moderate and severe nonproliferative DR, and 3 with proliferative DR.
Arteriolar VMC Density Analysis
The intrarater intraclass correlation coefficient for VMC density was 0.97 ([95% CI, 0.86–0.99] P=0.00001). The interrater intraclass correlation coefficient was 0.94 ([95% CI, 0.77–0.99] P=0.0001).
First, we compared arteriolar VMC density between the DM-no-DR and healthy subjects. Overall, DM-no-DR subjects had lower arteriolar VMC density than healthy subjects (5.0±2.1 versus 6.5±2.0 VMCs per 100 µm; P<0.05; Figure 2). Additional analysis based on vessels or regions of interest is shown in Figure S1. Figure 3 and Figure S2 show examples of arterioles from healthy subjects showing high VMC densities contrasting with DM-no-DR eyes and low VMC densities. Next, we compared DM-no-DR and DME patients and found that they did not have statistically significantly different arteriolar VMC density (5.0±2.1 versus 4.5±1.7 pericytes per 100 µm; P=0.60; Figure 2).
We then addressed the potential concern that age or hypertension could be confounding variables for arteriolar VMC density analysis. Using univariate analysis, we showed that age was not correlated with VMC density (Pearson correlation, P=0.11). The age range for healthy controls was 18 to 59 years, for DM-no-DR subjects was 25 to 67 years, and for DME subjects was 36 to 72 years. Furthermore, there was no statistically significant relationship between hypertension and arteriolar VMC density (t test, P=0.50; Figure S3).
Distinct VMC Morphology at Different Locations Along the Vascular Tree
In our study, we also found that VMC morphology varies along the vascular tree. Although we did not include venular locations in our VMC density analysis, we found that the mural cells were flatter and longer on venules compared with those found on arterioles (Figure S4). We also found that mural cells at sites of vascular bifurcation were larger and more globular (Figure S4).
Capillary Pericyte Density Analysis
DM-no-DR eyes did not have a significant difference in capillary pericyte density compared with healthy eyes (14.9±6.0 versus 15.6±6.6 per 100 µm; P=0.85). In contrast, DME eyes showed significant reduction in capillary pericyte density compared with DM-no-DR and healthy eyes (5.1±2.3; P=0.01 and P=0.006, respectively; Figure 4).
OCTA Analysis
Comparing healthy and DM-no-DR eyes on OCTA, we found no differences in the Q score, foveal avascular zone area, SCP VD, and deep capillary plexus VD (Figure 5; Figure S5). However, based on OCTA scans, DME eyes had a significantly lower VD in the SCP (37.6±3.8 versus 45.5±2.4; P<0.0001) and deep capillary plexus (40.4±5.7 versus 50.9±3.3; P=0.0003) compared with DM-no-DR eyes (Figure 5).
DISCUSSION
In this study, we used AOSLO and dark-field imaging to visualize and quantify retinal VMC (encompassing SMCs and pericytes) density in vivo. We focused our VMC imaging on arterioles of ≈90 µm in diameter, where we expect to image SMCs, as well as capillaries of ≈10 µm in diameter, where we expect to image capillary pericytes.1,9 Comparing healthy eyes to DM-no-DR eyes, we found that the latter had significantly lower arteriolar VMC density and this was not related to age or hypertension. We next explored whether the lower arteriolar VMC density in subjects with DM no DR was associated with capillary loss on motion contrast imaging (OCTA), but found that capillary density was comparable between the DM-no-DR and healthy eyes. Interestingly, when we evaluated VMC at the capillary level, there was no statistically significant difference between the DM-no-DR and healthy eyes. Recently, Mäe et al25 found hot spot leakage sites in mice with a blood-brain barrier impairment due to pericyte absence, which was associated with focally decreased Ang2 (angiopoietin-2) expression. To examine whether this phenomenon occurs in human eyes in vivo, we next evaluated eyes with DME, a vision compromising complication of DR with substantial intraretinal fluid leakage, and found that these eyes had significantly lower capillary VMC (pericyte) density compared with healthy and DM-no-DR eyes. Moreover, based on OCTA, these eyes also had significantly lower capillary density in both macular capillary plexuses, when compared with healthy and DM-no-DR eyes. These findings allowed us to generate an overall model for the progression of VMC compromise in human DR (Figure 6).
While the OCTA results showing capillary loss in eyes with advancing stages of DR and DME are expected,26 the finding of VMC loss primarily at the capillary level has important implications for understanding the pathogenesis of DME. Our findings would suggest that leakage may be driven not only by capillary pericyte loss but also require another potential second hit, possibly retinal nonperfusion and ischemia, both of which could be corollaries of a higher VEGF (vascular endothelial growth factor) load. Transgenic mouse studies offer relevant insights, where ablation of the retinal pericytes in adult mice did not produce any retinal vascular phenotype.27,28 Instead, additional injection of intraocular VEGF was required to elicit leakage and vascular capillary loss in these eyes. Hence, in humans with early preclinical DR, loss of VMC at the arteriolar level is not sufficient for the classic vascular manifestations of DR, leakage, hemorrhages, and microaneurysms. Instead, loss of capillary pericytes along with additional ischemic (or cumulative diabetic) insult and capillary closure may be prerequisites for apparent vascular manifestations of DR (Figure 6).
The retinal VMCs perform a variety of functions that are critical to regulating blood flow throughout the retinal microcirculation, which may be more vulnerable to injury from blood pressure fluctuations29 and changes in posture, as well as a greater need to rapidly adjust to a constantly varying metabolic demand with visual activities.30 The critical role of VMCs is highlighted by retinal pericyte coverage being the highest in the body,11 with a 1:1 endothelial cell-to-pericyte ratio, compared with 3:1 in the brain.31 Loss of arteriolar VMCs in the preclinical states of DR might explain early dysregulation of retinal blood flow, which has been reported previously.32,33
The role of pericytes in DR pathogenesis was first established in 1961, in a trypsin-digested donor eye study that showed widespread loss of capillary pericytes, as well as capillary microaneurysms and capillary closure in 55 diabetic eyes.12 Subsequently, Gardiner et al13,14 showed that loss of retinal arteriolar vascular SMCs is another salient feature of DR, being most notable in second-order arterioles and higher, as well as in precapillary arterioles in the central retina. Comparing our study to the classical findings of Kuwabara and Cogan,11 in their histological analysis, they showed loss of capillary VMCs (pericytes), whereas we analyzed both arteriolar and capillary VMCs. Furthermore, Kuwabara and Cogan studied donor eyes, with DR and microaneurysms, though they did not specify DR severity in these donor eyes, as the clinical grading of DR severity was actually a later development in the field. In contrast, we studied subjects with diabetes and 2 groups of severity: DM no DR and DME. Therefore, our study adds more nuance, while confirming the finding by Kuwabara and Cogan that capillary pericyte loss is seen in eyes with microaneurysms and established DR, and showing significant loss in eyes with leakage and DME in DR. We also add the new finding that arteriolar VMC loss occurs before capillary pericyte loss and precedes all clinical manifestations of DR.
The pathogenesis of VMC loss in DR is likely complex, as diabetes impacts a wide range of normal pericyte–endothelial cell interactions.34–37 Notably, the loss of pericytes from retinal capillaries is analogous to the loss of podocytes in glomerular capillaries in the pathogenesis of diabetes.2 Studying human donor eyes, Mizutani et al quantified apoptotic pericytes, finding significantly more in eyes with DR than healthy eyes, though these authors did not specifically report on the severity of retinopathy or the actual pericyte density in these eyes. The same group also found that 24 to 31 weeks of experimental hyperglycemia in rats led to pericyte death.38 Similarly, Hammes et al39 found that after 6 months, diabetic mice had a significant reduction in retinal pericytes. Detailed longitudinal studies of mural cells in animal models of DR are limited by semiquantitative trypsin digest techniques but have generally supported pericyte loss being an early event in DR. Hammes et al39 found that 6 months after streptozotocin ablation of the pancreas, diabetic wild-type mice had a 40% reduction in retinal pericytes and a 2.5-fold increase in acellular capillary segments compared with nondiabetic controls. Notably, these eyes, similar to human donor studies, showed coexisting capillary and pericyte loss, suggesting that pericyte loss may not be the earliest change in DR in rodents.
With the understanding of the limitations of these prior reports, our study clarifies the important gap in knowledge regarding the timeline of retinal VMC demise in humans with diabetes. Our study sets the timing of arteriolar VMC loss in humans much earlier than these prior reports, showing these cells decline even before any clinical manifestations of DR and indeed well before capillary pericyte loss in the course of DR. Our study highlights that the loss of capillary pericytes is a later event, associated with the classical clinical and microvascular manifestations of DR (Figure 6).
We acknowledge that the definition of pericytes remains a subject of debate, with recent studies suggesting these cells exist on a transcriptomic and morphological spectrum with other VMCs, without having a unique immunohistochemical marker.40 In general, the current understanding suggests a gradual transition from SMCs to pericytes, with ensheathing pericytes in precapillary arterioles serving as a hybrid of both cell types.9,25,40,41 A recent review of brain VMCs proposed that morphology and location of these cells be used as defining features, where arterioles and venules are surrounded by SMCs, while ensheathing pericytes and capillary pericytes surround the arteriole-capillary transition and capillaries, respectively.9 In our study, we, therefore, used location and morphology to define these cells. Considering our focus on arterioles of ≈90 µm in diameter and capillaries of ≈10 µm in diameter, we identified VMCs based on their protruding cell bodies.41,42 And since we studied arterioles, the arteriole-capillary transition, and capillaries, our selected cells likely represent a mixture of SMCs and pericytes, which we further defined based on vessel size.
We validated the human arteriolar VMC density metric by comparing our results in healthy eyes (6.5±2.0 pericytes per 100 µm) to previously reported values. Relevant to our choice of vessel diameters (50–99 µm), a previous study used AOSLO in the human retina in vivo, finding a mural density of 7 ±3 cells/100 µm.3 Soltanian-Zadeh et al24 used AO-OCT (adaptive optics optical coherence tomography) to study retinal ganglion cell density in subjects with glaucoma, in vivo. In these previously published images, we identified VMCs as highly reflective protuberances along the outer retinal vessel wall and calculated an arteriolar VMC density of ≈3 to 4 per 100 µm. In aggregate, these studies support the validity of our calculated arteriolar VMC density in healthy eyes. Finally, we compared our capillary images to those from Schallek et al,43 who imaged pericytes in transgenic reporter mouse eyes, showing similar cell body size (≈10 µm) and spacing to the capillary pericytes we have imaged in living human eyes.
We found that VMCs were morphologically diverse, depending on their location along the vascular tree, which has been previously demonstrated in the mouse and in human retina.3,41,44 By imaging mural cells in reporter mice (Td-Tomato mice under the NG2 [neural/glial antigen 2] promoter), the diversity of mouse retinal VMC morphology was illustrated, including sphincter VMCs encircling the prearteriolar branching vessel and spider-like appearing venular VMCs.41 Interestingly, healthy eyes in our study had 3-fold higher capillary VMC (pericyte) density than VMC (SMC) density at the arteriolar level, illustrating interesting topographical and potentially functional differences that deserve further study.
One limitation in our study was the small sample size, which was imposed by challenges in obtaining high-quality AOSLO images. Due to the pilot nature and small sample size of this study, we could not perform effect modification or adjust for multiple variables. These are important areas of future study. Another limitation was the potentially subjective nature of quantifying VMC density, in the absence of automated tools. To address this concern, we performed rigorous, masked to diagnosis, intrarater and interrater analysis (intraclass correlation coefficient) for a subset of the data, which showed high agreement with good to excellent reliability. Finally, another limitation is that we did not specifically study treatment naive DME, hence we acknowledge the potential for confounding effects of therapy on the VMCs.
In summary, capitalizing on in vivo retinal imaging to the cellular level, we show that the timing of retinal arteriolar VMC loss in diabetes may precede all other clinical manifestations of DR, occurring well before capillary pericyte loss (Figure 6). This finding highlights a variety of functions that may be dysregulated at this early stage, particularly retinal blood flow regulation at the arteriolar level. In contrast, eyes with clinically apparent DME and DR showed statistically significant decline in capillary pericyte and capillary vascular densities, suggesting a mechanistic role for capillary pericytes in the onset of leakage and clinical DR. Future studies using this approach need to establish a critical threshold of VMC loss that drives pathological changes in DR, to define the optimal timing and study the efficacy of potential therapeutics that preserve the VMC and prevent the onset of DR. Quantitative retinal imaging of VMC density with AOSLO may also be useful to study the impact of new systemic or local therapeutics on retinal VMC, as a corollary to their impact on the microvasculature in DR. We would also expect that retinal VMC density could reflect other microvascular complications of diabetes such as nephropathy, given the close link between the pericytes and podocytes, both perivascular cells that are vulnerable in diabetes.2
ARTICLE INFORMATION
Supplemental Material
Figures S1–S5
Major Resources Table
Acknowledgments
The authors would like to acknowledge Daniela Castellanos Canales and Nicole L. Decker for helping with patient recruitment and acquiring optical coherence tomography angiography and adaptive optics images. A.A. Fawzi designed the research; B.B. Huang, H. Fukuyama, and A.A. Fawzi performed the research and analyzed data; B.B. Huang, A.A. Fawzi, and S.A. Burns wrote the paper. All authors reviewed and approved the final draft.
Footnote
Nonstandard Abbreviations and Acronyms
- AOSLO
- adaptive optics scanning laser ophthalmoscopy
- DM no DR
- diabetes without diabetic retinopathy
- DME
- diabetic macular edema
- DR
- diabetic retinopathy
- HbA1c
- hemoglobin A1c
- OCTA
- optical coherence tomography angiography
- SCP
- superficial capillary plexus
- SMC
- smooth muscle cell
- VD
- vessel density
- VEGF
- vascular endothelial growth factor
- VMC
- vascular mural cell
Supplemental Material
File (atvb_atvb-2023-320169_supp1.pdf)
- Download
- 745.78 KB
REFERENCES
1.
Holm A, Heumann T, Augustin HG. Microvascular mural cell organotypic heterogeneity and functional plasticity. Trends Cell Biol. 2018;28:302–316. doi: 10.1016/j.tcb.2017.12.002
2.
Rask-Madsen C, King GL. Vascular complications of diabetes: mechanisms of injury and protective factors. Cell Metab. 2013;17:20–33. doi: 10.1016/j.cmet.2012.11.012
3.
Chui TYP, Gast TJ, Burns SA. Imaging of vascular wall fine structure in the human retina using adaptive optics scanning laser ophthalmoscopy. Invest Ophthalmol Vis Sci. 2013;54:7115–7124. doi: 10.1167/iovs.13-13027
4.
Chui TYP, VanNasdale DA, Burns SA. The use of forward scatter to improve retinal vascular imaging with an adaptive optics scanning laser ophthalmoscope. Biomed Opt Express. 2012;3:2537–2549. doi: 10.1364/BOE.3.002537
5.
Liang J, Williams DR, Miller DT. Supernormal vision and high-resolution retinal imaging through adaptive optics. J Opt Soc Am A. 1997;14:2884–2892. doi: 10.1364/josaa.14.002884
6.
Sincich LC, Zhang Y, Tiruveedhula P, Horton JC, Roorda A. Resolving single cone inputs to visual receptive fields. Nat Neurosci. 2009;12:967–969. doi: 10.1038/nn.2352
7.
Rossi EA, Granger CE, Sharma R, Yang Q, Saito K, Schwarz C, Walters S, Nozato K, Zhang J, Kawakami T, et al. Imaging individual neurons in the retinal ganglion cell layer of the living eye. Proc Natl Acad Sci USA. 2017;114:586–591. doi: 10.1073/pnas.1613445114
8.
Burns SA, Elsner AE, Sapoznik KA, Warner RL, Gast TJ. Adaptive optics imaging of the human retina. Prog Retin Eye Res. 2019;68:1–30. doi: 10.1016/j.preteyeres.2018.08.002
9.
Hartmann DA, Coelho-Santos V, Shih AY. Pericyte control of blood flow across microvascular zones in the central nervous system. Annu Rev Physiol. 2022;84:331–354. doi: 10.1146/annurev-physiol-061121-040127
10.
Yau JWY, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, Chen SJ, Dekker JM, Fletcher A, Grauslund J, et al; Meta-Analysis for Eye Disease (META-EYE) Study Group. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35:556–564. doi: 10.2337/dc11-1909
11.
Kuwabara T, Cogan DG. Retinal vascular patterns: VI. mural cells of the retinal capillaries. Arch Ophthalmol. 1963;69:492–502. doi: 10.1001/archopht.1963.00960040498013
12.
Cogan DG, Toussaint D, Kuwabara T. Retinal vascular patterns: IV. Diabetic retinopathy. Arch Ophthalmol. 1961;66:366–378. doi: 10.1001/archopht.1961.00960010368014
13.
Gardiner TA, Stitt AW. Pericyte and vascular smooth muscle death in diabetic retinopathy involves autophagy. Int J Transl Med. 2022;2:26–40. doi: 10.3390/ijtm2010003
14.
Gardiner TA, Stitt AW, Anderson HR, Archer DB. Selective loss of vascular smooth muscle cells in the retinal microcirculation of diabetic dogs. Br J Ophthalmol. 1994;78:54–60. doi: 10.1136/bjo.78.1.54
15.
Scholfield CN, McGeown JG, Curtis TM. Cellular physiology of retinal and choroidal arteriolar smooth muscle cells. Microcirculation. 2007;14:11–24. doi: 10.1080/10739680601072115
16.
Burns SA, Elsner AE, Chui TY, VanNasdale DA, Clark CA, Gast TJ, Malinovsky VE, Phan ADT. In vivo adaptive optics microvascular imaging in diabetic patients without clinically severe diabetic retinopathy. Biomed Opt Express. 2014;5:961–974. doi: 10.1364/BOE.5.000961
17.
Sahni J, Patel SS, Dugel PU, Khanani AM, Jhaveri CD, Wykoff CC, Hershberger VS, Pauly-Evers M, Sadikhov S, Szczesny P, et al. Simultaneous inhibition of angiopoietin-2 and vascular endothelial growth factor-A with faricimab in diabetic macular edema: BOULEVARD phase 2 randomized trial. Ophthalmology. 2019;126:1155–1170. doi: 10.1016/j.ophtha.2019.03.023
18.
Wilkinson CP, Ferris FL, Klein RE, Lee PP, Agardh CD, Davis M, Dills D, Kampik A, Pararajasegaram R, Verdaguer JT; Global Diabetic Retinopathy Project Group. Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. Ophthalmology. 2003;110:1677–1682. doi: 10.1016/S0161-6420(03)00475-5
19.
Jia Y, Tan O, Tokayer J, Potsaid B, Wang Y, Liu JJ, Kraus MF, Subhash H, Fujimoto JG, Hornegger J, et al. Split-spectrum amplitude-decorrelation angiography with optical coherence tomography. Opt Express. 2012;20:4710–4725. doi: 10.1364/OE.20.004710
20.
Dubra A, Sulai Y. Reflective afocal broadband adaptive optics scanning ophthalmoscope. Biomed Opt Express. 2011;2:1757–1768. doi: 10.1364/BOE.2.001757
21.
Damisah EC, Hill RA, Tong L, Murray KN, Grutzendler J. A fluoro-Nissl dye identifies pericytes as distinct vascular mural cells during in vivo brain imaging. Nat Neurosci. 2017;20:1023–1032. doi: 10.1038/nn.4564
22.
Chui TYP, Song H, Burns SA. Individual variations in human cone photoreceptor packing density: variations with refractive error. Invest Ophthalmol Vis Sci. 2008;49:4679–4687. doi: 10.1167/iovs.08-2135
23.
Fernandez-Klett F, Brandt L, Fernández-Zapata C, Abuelnor B, Middeldorp J, Sluijs JA, Curtis M, Faull R, Harris LW, Bahn S, et al. Denser brain capillary network with preserved pericytes in Alzheimer’s disease. Brain Pathol. 2020;30:1071–1086. doi: 10.1111/bpa.12897
24.
Soltanian-Zadeh S, Kurokawa K, Liu Z, Zhang F, Saeedi O, Hammer DX, Miller DT, Farsiu S, Farsiu S. Weakly supervised individual ganglion cell segmentation from adaptive optics OCT images for glaucomatous damage assessment. Optica. 2021;8:642–651. doi: 10.1364/OPTICA.418274
25.
Mäe MA, He L, Nordling S, Vazquez-Liebanas E, Nahar K, Jung B, Li X, Tan BC, Chin Foo J, Cazenave-Gassiot A, et al. Single-cell analysis of blood-brain barrier response to pericyte loss. Circ Res. 2021;128:e46–e62. doi: 10.1161/CIRCRESAHA.120.317473
26.
Sun Z, Tang F, Wong R, Lok J, Szeto SKH, Chan JCK, Chan CKM, Tham CC, Ng DS, Cheung CY. OCT angiography metrics predict progression of diabetic retinopathy and development of diabetic macular edema: a prospective study. Ophthalmology. 2019;126:1675–1684. doi: 10.1016/j.ophtha.2019.06.016
27.
Eilken HM, Diéguez-Hurtado R, Schmidt I, Nakayama M, Jeong HW, Arf H, Adams S, Ferrara N, Adams RH. Pericytes regulate VEGF-induced endothelial sprouting through VEGFR1. Nat Commun. 2017;8:1574. doi: 10.1038/s41467-017-01738-3
28.
Park DY, Lee J, Kim J, Kim K, Hong S, Han S, Kubota Y, Augustin HG, Ding L, Kim JW, et al. Plastic roles of pericytes in the blood–retinal barrier. Nat Commun. 2017;8:15296. doi: 10.1038/ncomms15296
29.
Frank RN, Dutta S, Mancini MA. Pericyte coverage is greater in the retinal than in the cerebral capillaries of the rat. Invest Ophthalmol Vis Sci. 1987;28:1086–1091.
30.
Pournaras CJ, Rungger-Brändle E, Riva CE, Hardarson SH, Stefansson E. Regulation of retinal blood flow in health and disease. Prog Retin Eye Res. 2008;27:284–330. doi: 10.1016/j.preteyeres.2008.02.002
31.
Speiser P, Gittelsohn AM, Patz A. Studies on diabetic retinopathy. 3. Influence of diabetes on intramural pericytes. Arch Ophthalmol. 1968;80:332–337. doi: 10.1001/archopht.1968.00980050334007
32.
Palochak CMA, Lee HE, Song J, Geng A, Linsenmeier RA, Burns SA, Fawzi AA. Retinal blood velocity and flow in early diabetes and diabetic retinopathy using adaptive optics scanning laser ophthalmoscopy. J Clin Med. 2019;8:1165. doi: 10.3390/jcm8081165
33.
Grunwald JE, DuPont J, Riva CE. Retinal haemodynamics in patients with early diabetes mellitus. Br J Ophthalmol. 1996;80:327–331. doi: 10.1136/bjo.80.4.327
34.
Antonetti DA, Silva PS, Stitt AW. Current understanding of the molecular and cellular pathology of diabetic retinopathy. Nat Rev Endocrinol. 2021;17:195–206. doi: 10.1038/s41574-020-00451-4
35.
Geraldes P, Hiraoka-Yamamoto J, Matsumoto M, Clermont A, Leitges M, Marette A, Aiello LP, Kern TS, King GL. Activation of PKC-δ and SHP-1 by hyperglycemia causes vascular cell apoptosis and diabetic retinopathy. Nat Med. 2009;15:1298–1306. doi: 10.1038/nm.2052
36.
Hammes HP, Lin J, Wagner P, Feng Y, vom Hagen F, Krzizok T, Renner O, Breier G, Brownlee M, Deutsch U. Angiopoietin-2 causes pericyte dropout in the normal retina: evidence for involvement in diabetic retinopathy. Diabetes. 2004;53:1104–1110. doi: 10.2337/diabetes.53.4.1104
37.
Park SW, Yun JH, Kim JH, Kim KW, Cho CH, Kim JH. Angiopoietin 2 induces pericyte apoptosis via α3β1 integrin signaling in diabetic retinopathy. Diabetes. 2014;63:3057–3068. doi: 10.2337/db13-1942
38.
Mizutani M, Kern TS, Lorenzi M. Accelerated death of retinal microvascular cells in human and experimental diabetic retinopathy. J Clin Invest. 1996;97:2883–2890. doi: 10.1172/JCI118746
39.
Hammes HP, Lin J, Renner O, Shani M, Lundqvist A, Betsholtz C, Brownlee M, Deutsch U. Pericytes and the pathogenesis of diabetic retinopathy. Diabetes. 2002;51:3107–3112. doi: 10.2337/diabetes.51.10.3107
40.
Vanlandewijck M, He L, Mäe MA, Andrae J, Ando K, Del Gaudio F, Nahar K, Lebouvier T, Laviña B, Gouveia L, et al. A molecular atlas of cell types and zonation in the brain vasculature. Nature. 2018;554:475–480. doi: 10.1038/nature25739
41.
Ivanova E, Corona C, Eleftheriou CG, Bianchimano P, Sagdullaev BT. Retina-specific targeting of pericytes reveals structural diversity and enables control of capillary blood flow. J Comp Neurol. 2021;529:1121–1134. doi: 10.1002/cne.25011
42.
Kovacs-Oller T, Ivanova E, Bianchimano P, Sagdullaev BT. The pericyte connectome: spatial precision of neurovascular coupling is driven by selective connectivity maps of pericytes and endothelial cells and is disrupted in diabetes. Cell Discov. 2020;6:1–18. doi: 10.1038/s41421-020-0180-0
43.
Schallek J, Geng Y, Nguyen H, Williams DR. Morphology and topography of retinal pericytes in the living mouse retina using in vivo adaptive optics imaging and ex vivo characterization. Invest Ophthalmol Vis Sci. 2013;54:8237–8250. doi: 10.1167/iovs.13-12581
44.
Hartmann DA, Underly RG, Grant RI, Watson AN, Lindner V, Shih AY. Pericyte structure and distribution in the cerebral cortex revealed by high-resolution imaging of transgenic mice. Neurophotonics. 2015;2:041402. doi: 10.1117/1.NPh.2.4.041402
Information & Authors
Information
Published In
Copyright
© 2023 American Heart Association, Inc.
Versions
You are viewing the most recent version of this article.
History
Received: 15 September 2023
Accepted: 13 December 2023
Published online: 28 December 2023
Published in print: February 2024
Keywords
Subjects
Authors
Disclosures
Disclosures A.A. Fawzi is a consultant at Regeneron, Roche/Genentech, Boehringer Ingelheim, RegenXbio, 3Helix, and Optos, Inc, and reports grant support from Boehringer Ingelheim. The other authors report no conflicts.
Sources of Funding
A.A. Fawzi is supported, in part, by the National Institutes of Health (NIH) grant R01EY31815. B.B. Huang is supported, in part, by the NIH grant 1T35DK126628-01, a grant from Illinois Society for the Prevention of Blindness (SP0076776), and Alpha Omega Alpha Carolyn L. Kuckein Student Research Fellowship.
Metrics & Citations
Metrics
Citations
Download Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Select your manager software from the list below and click Download.
- Characterizing Vascular Wall and Lumen Caliber in Eyes with Diabetic Retinopathy Based on Adaptive Optics Scanning Laser Ophthalmoscopy, Diagnostics, 14, 18, (2020), (2024).https://doi.org/10.3390/diagnostics14182020
- Hypertension Likely Drives Arteriolar Wall Thickening in Preclinical Diabetic Retinopathy While Diabetes Drives Wall Thickness in Clinical Retinopathy, Translational Vision Science & Technology, 13, 6, (8), (2024).https://doi.org/10.1167/tvst.13.6.8
- Perfusion Deficits in Diabetes Without Retinopathy Localize to the Perivenular Deep Capillaries Near the Fovea on OCT Angiography, Ophthalmology Science, 4, 5, (100482), (2024).https://doi.org/10.1016/j.xops.2024.100482
- Resilience to diabetic retinopathy, Progress in Retinal and Eye Research, 101, (101271), (2024).https://doi.org/10.1016/j.preteyeres.2024.101271
- Natural plant medications for the treatment of retinal diseases: The blood-retinal barrier as a clue, Phytomedicine, 130, (155568), (2024).https://doi.org/10.1016/j.phymed.2024.155568
Loading...
View Options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginPurchase Options
Purchase this article to access the full text.
eLetters(0)
eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.
Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.