Table 1.
List of antibodies and stains used for immunohistochemistry.
Fig 1.
L-HIPR induced retinal pathological changes and persistent hyaloidal vessels.
(A) Schematic of L-HIPR treatment of 65% oxygen from P0-P7 followed by room air recovery until experimental endpoints. (B) Paraffin embedded histological cross sections of P12, P17, P21 and P30 murine eyes in room air L-HIPR showing persistent hyaloidal vessels (arrow) since P12. Hyaloidal vessels associate with the central retina at P12 (asterisk) as well as the peripheral retina at P17 and P21. Retinal pathological changes started to appear at P30 including rosettes (arrowhead), detachments and intravitreal hemorrhaging. (C) Representative H&E-stained cross sections of central retina in room air and L-HIPR groups P12, P17, P21 and P30 (n = 3–4) where retinal layer thickness was quantified. Graphical representation of select layers of interest shown of the ganglion cell layer (GCL), inner plexiform layer (IPL), and inner nuclear layer (INL). * p < 0.05, one-way ANOVA with Tukey’s post hoc analysis. Scale bars equal 500μm.
Fig 2.
L-HIPR causes delayed retinal vascular development.
(A) Retina flat-mount samples (n = 3) from both room air and L-HIPR groups at P12, P17, P21 and P30 were stained with aIB4 to detect the vascular development. There is sporadic vascularization at P12 in L-HIPR and by P30 the vasculature fails to completely reach the retinal margin. L-HIPR vessels also appear tortuous and dilated, particularly at P17 and P21. XZ spinning disc confocal line scans of IB4 labeled retinal flat-mounts were acquired to visualize the delayed vascular plexuses in L-HIPR. Scans were taken near the optic nerve, mid- retina, or periphery and relevant XZ optical scan insets are denoted by green, blue, and red dashed lines, respectively. (B) Quantification of blood vessel density within the 3 vascular plexuses in L-HIPR normalized to age matched room air control. Non-significant outcomes were marked as NS, all others are statistically significant, p < 0.05, two-sided t-test. Scale bars equal 300 μm.
Fig 3.
Anomalous retinal vascular development, with two opposing angiogenic fronts, and transient circumferential vasculature is observed in L-HIPR.
(A) Alexa488 conjugated IB4 (green) was delivered in vivo via cardiac perfusion to highlight functional blood vessels. Retinal flat-mounts were prepared with Alexa568 conjugated IB4 (red) labeling all blood vessels. At P17 in L-HIPR no green signal is seen signifying that these vessels were not perfused. By P21 L-HIPR, perfusion is seen in the central retina, but peripheral vessels remain non-perfused. The light blue inset at P21 L-HIPR shows a region of interest displaying abnormal blood vessels traversing the retina circumferentially (marked with asterisk), which we suspect is persistent hyaloidal vasculature invading the retina. P30 L-HIPR shows mostly perfused retina from the optic nerve to periphery with resolution of the circumferential vasculature. (B) P21 L-HIPR flat-mounts immunolabeled with the tip-cell marker Esm1 (green) and blood vessels marked with IB4 (red). The light blue insert shows a region of interest where central and peripheral blood vessel fronts display IB4+ filopodia and Esm1+ tip-cells, indicating two angiogenic fronts that meet in the retinal mid-periphery. Scale bars equal 500 μm.
Fig 4.
The hyaloidal vessels persist and rescue the retinal vasculature in L-HIPR.
Retinal cross sections in room air and L-HIPR groups (n = 3) were stained with Isolectin B4 (IB4; red), and DAPI (white) at P12, P17 and P21. Central (yellow box) and peripheral (green box) zoomed retinal cross-sections are presented. The hyaloidal vessels were absent, with normal retinal vascular stratification by P12 in the room air eyes. In L-HIPR, the hyaloidal vessels persisted well into P21 and started to invade the peripheral retina at P12 (Asterisks). The later timepoints further illustrate how the hyaloidal vessels invade the mid-retina at P17 and P21. Scale bar 50 μm, GCL: ganglion cell, *: hyaloid, L: lens.
Fig 5.
The L-HIPR retina is inflamed.
(A) Immunolabeled cryosections at P21 and P30 of room air and L-HIPR tissue. Macrophages were labeled with anti-F4/80 (green), blood vessels with IB4 (red), astrocytes/activated Muller cells with GFAP (white), and nuclei with DAPI (blue). The room air tissues show GFAP+ astrocytes in the ganglion cell layer, however, the Muller cells become GFAP+ indicating an activated state in L-HIPR at both P21 and P30. F4/80+ macrophages can be seen in the inner and outer nuclear layers in the L-HIPR sample. Scale bars equal 50 μm.
Fig 6.
Lower pericyte count is observed in L-HIPR.
(A) Immunolabeled retinal flat-mounts at P21 and P30 of room air and L-HIPR tissue. Blood vessels were labeled with anti-CD31 (green), and pericytes with NG2 (red). Scale bars equal 50 μm. (B) Graphical representation of vessel density where CD31 vessel length after skeletonization was quantified against image area and normalized to room air values. At both time points the L-HIPR vascular density was significantly less dense. (C) Graphical representation of NG2+ pericyte counts per micron of CD31 vessel length (μm) normalized to room air values. There was not a statistically significant difference in pericyte counts between room air and L-HIPR at P21, however, there was a significant reduction at P30.
Fig 7.
L-HIPR model shows preretinal and retinal fibrosis and scarring.
(A) Masson trichrome staining of P21 and P30 L-HIPR tissue. Red inserts highlight regions of interest showing collagenous tissue (blue) in the vitreous surrounding the persistent hyaloidal vessels. (B) L-HIPR retinal flat-mount immunolabeled with α-SMA. Red inserts highlight regions of interest showing extravascular islands of α-SMA immunolabeling suggesting fibroblastic cellular proliferation on the retinal surface. Scale bars equal 300 μm.
Fig 8.
Proposed model for the progression of L-HIPR.
Course of retinal vascular development, preretinal fibrosis and hyaloidal rescue of the retinal vasculature during L-HIPR exposure.