March 2007
Volume 48, Issue 3
Free
Retinal Cell Biology  |   March 2007
Changes in Pericytes and Smooth Muscle Cells in the Kitten Model of Retinopathy of Prematurity: Implications for Plus Disease
Author Affiliations
  • Suzanne Hughes
    From the Department of Anatomy, Institute for Biomedical Research, University of Sydney, Australia;
  • Tom Gardiner
    Department of Ophthalmology and Visual Science, Queen’s University, Belfast, Ireland.
  • Louise Baxter
    From the Department of Anatomy, Institute for Biomedical Research, University of Sydney, Australia;
  • Tailoi Chan-Ling
    From the Department of Anatomy, Institute for Biomedical Research, University of Sydney, Australia;
Investigative Ophthalmology & Visual Science March 2007, Vol.48, 1368-1379. doi:10.1167/iovs.06-0850
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Suzanne Hughes, Tom Gardiner, Louise Baxter, Tailoi Chan-Ling; Changes in Pericytes and Smooth Muscle Cells in the Kitten Model of Retinopathy of Prematurity: Implications for Plus Disease. Invest. Ophthalmol. Vis. Sci. 2007;48(3):1368-1379. doi: 10.1167/iovs.06-0850.

      Download citation file:


      © 2017 Association for Research in Vision and Ophthalmology.

      ×
  • Supplements
Abstract

purpose. Dilated and tortuous vessels (plus disease) in ROP is a grim prognostic indicator of visual outcome. The purpose of this study was to determine whether alterations in pericytes and smooth muscle cells (SMCs), are associated with the pathogenesis of ROP, including plus disease.

methods. Kittens were exposed to either 4 (standard obliterative model) or 2 (modified model) days of hyperoxia, resulting in vaso-obliteration or localized vessel regression, respectively, and returned to room air. The modified model more closely resembles human ROP. Desmin and α-smooth muscle actin (SMA) immunohistochemistry and lectin labeling were used to label mural cells and vessels. The desmin ensheathment ratio (DER), a quantitative measure of vessel stability, was determined.

results. In the neovasculature of the standard model and surviving vasculature of the modified model, radial arterioles and venules were dilated and SMCs attenuated. SMA expression on venules was decreased, and the difference in desmin expression normally observed between arterioles and venules was lost, indicating altered SMC differentiation. The DER was reduced in both ROP models, consistent with highly unstable vascular plexuses, receptive to angiogenic and vascular regression signals.

conclusions. The results provide compelling evidence of significant changes in arteriolar and venular SMCs in both experimental models of ROP. The delayed differentiation and apparent dedifferentiation of SMCs during the hypoxic phases would result in an impaired ability to regulate blood flow, contributing to the vasodilation and tortuosity, hallmarks of plus disease. Vessel tortuosity was seen only in the nonobliterative model, suggesting that tortuosity may be due to increased capillary resistance resulting from capillary closure.

Retinopathy of prematurity (ROP) is a disease that affects premature infants, in which abnormal retinal vasoproliferation can be followed by extraretinal fibrovascular proliferation causing tractional detachment, which may lead to blindness. Risk factors for ROP include extreme prematurity, artificial ventilation, and very low birth weight, 1 2 and the incidence of the disease correlates with the degree of ocular immaturity at birth. 3 ROP is a major cause of childhood blindness in the developed world, where its incidence is increasing in line with the enhanced survival of premature infants under 1 kg and 26 weeks’ gestation and is an emerging cause of blindness in the developing world, due to improved neonatal care. 4 5 Depending on the level of oxygen administered to neonates, the resultant retinal hyperoxia can result in severe vessel constriction and delayed retinal vascularization. The consequent tissue hypoxia when oxygen administration is stopped and/or neuronal metabolic activity increases, results in VEGF-driven vasoproliferation and aberrant neovascularization. 6 7 8 9 10  
ROP has been classified according to the location and extent of the abnormalities and the stage of the vasoproliferative response at the vascular periphery. The severity of the disease ranges from the least severe stage, characterized by a demarcation line between the outer edge of the retinal vasculature and the avascular zone at the retinal periphery, to a severe stage, characterized by extraretinal fibrovascular proliferation and tractional retinal detachment. 11  
An additional risk factor, designated “plus disease” and characterized by tortuous arterioles and dilated veins in the posterior retina, may be superimposed on the other pathologic changes at any stage of ROP. 11 Plus disease may be accompanied by iris vessel engorgement, pupil rigidity, and vitreous haze; is a grim predictor of visual outcomes in ROP; and correlates strongly with retinal detachment. 12 13 14 15 16 Indeed, laser photocoagulation is now recommended in most cases in which plus disease, as defined by significant dilation and tortuosity of posterior retinal blood vessels, is present. 15 17 Even in milder cases in which posterior vessel dilation and tortuosity are less pronounced, a condition designated pre-plus disease, 18 the incidence of ROP progression is increased. 19 A rapidly progressive form of ROP, aggressive posterior ROP (AP-ROP), has now been recognized. It is characterized by posterior location, significant dilation and tortuosity of arteries and veins, and brushlike neovascularization at the vascular periphery. 18  
Although features of plus disease’s have been recognized since 1949, 20 and its importance as a prognostic indicator is known, our understanding of the pathogenesis of the disease is still lacking. It is thought to arise as a result of altered hemodynamics due to shunting of blood at the vascular periphery combined with retention of blood producing engorgement and vasodilation. To date, no studies of the cellular events underlying the pathogenesis of plus disease have been undertaken, and its cause has been speculative. Fluorescein angiography has shown that in the acute proliferative stages of ROP, there is often a major arteriovenous shunt, which is extremely leaky, 21 and it is thought that plus disease signifies high blood flow as a result of these shunts. However, plus disease can occur in AP-ROP without any evidence of such peripheral shunts. 18  
Aggressive posterior ROP is characterized by prominent plus disease in posterior retina in which both venules and arterioles become dilated and tortuous and thus difficult to differentiate from one another. 18  
Vascular smooth muscle cells (SMCs) are the specialized contractile cells of the arterial wall and the major effectors in control of blood pressure and flow. Pericytes represent their counterpart in the capillaries and venules, and although they show less development in their contractile apparatus, they are thought to play a role in the regulation of blood flow and to stabilize vessels (reviewed by Hughes and Chan-Ling 22 ). In particular, they are thought to prevent hyperoxia-induced vessel regression and to regulate endothelial cell proliferation. 23 24 25 26 27 28 29 A decrease in the number of mural cells (SMCs and pericytes) during development and in diseases such as diabetic retinopathy is associated with aberrant neovascularization, vessel dilation, microaneurysm formation, vascular incompetence, and possible impaired perfusion. 30 31 32 33 Earlier reports have demonstrated the key role played by astrocytes, the glial cells associated with developing retinal vasculature, 34 and VEGF 8 35 in the pathogenesis of ROP, but little is known of the role of pericytes and SMCs in this disease. 
The aberrant vasoproliferation of ROP and the ability of pericytes to regulate endothelial cell proliferation 27 suggest that these cells may play a role in the pathogenesis of ROP. We have shown that immature mural cells are present in the neovasculature in the kitten model of ROP and that these vessels have a decreased ratio of desmin filaments to endothelial cell area, or desmin ensheathment ratio (DER). 36 The low DER may reflect aberrant endothelial cell-pericyte interaction. Furthermore, the radial vessel tortuosity and vasodilation that characterizes pre-plus disease, plus disease, and AP-ROP suggests that SMCs and pericytes play a key role in these more severe manifestations of ROP, due to their functions, known and postulated, in regulating retinal blood flow. 
The vascular events in ROP can be studied in a kitten model of the disease, in which hyperoxia-induced vaso-obliteration is followed by hypoxia-driven neovascularization in both the retina and vitreous; the neovasculature is subsequently remodeled, resulting in disease regression. 10 34 36 37 In this model of ROP, virtually all the retinal vessels are lost, whereas in the human disease there is a surviving, albeit altered, retinal vasculature consisting of congested peripheral capillary plexuses supplied by radial vessels that may also be altered. Shorter exposures to hyperoxia result in partial rather than total vaso-obliteration in the kitten. 38 The purpose of our study was to determine whether pericytes and/or SMCs are altered during the various phases of the standard obliterative model of ROP: the hyperoxic phase when existing retinal vessels regress, the hypoxic neovascular phase, and the remodeling-resolving phase. In addition, we set out to determine the changes in mural cells in a modified nonobliterative kitten model of ROP in which the vasculature survives, albeit with some vessel regression and delayed vascularization and thus more closely follows the postulated early events of human ROP. 
Our findings led us to suggest that mural cell-endothelial cell relationships are severely compromised during hyperoxia and hypoxia and that changes in mural cells play a significant role in the pathogenesis of ROP—in particular, the potentially blinding form of ROP associated with plus disease. The vasculature of ROP was characterized by a massive increase in capillary density. 9 As a consequence, a higher blood pressure or greater vasoconstriction is necessary to push this larger blood volume through this immensely dense vasculature. Plus disease is a likely clinical manifestation of these changes. 
Materials and Methods
Kitten Model of ROP: Obliterative
Experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and with the permission of the appropriate local authorities. Animals were sourced from a cat-breeding colony at the University of Sydney. To induce retinal vaso-obliteration, postnatal day (P)1 kittens were placed with a lactating mother in a hyperoxic chamber (60%–70% oxygen in air) for 4 days 6 9 34 and then returned to room air for 0, 3, 7, 10, 14, and 27 days (dRA). 
Modified Kitten Model of ROP: Nonobliterative
To expose immature retinal vasculature to hyperoxia but not obliterate it, we placed P3 kittens with a lactating mother in a hyperoxic chamber for 2 days and then returned them to room air for 3 or 29 days. 
Controls kittens were raised in room air from birth for 1, 3, 6, 17, 28, and 32 days before death. Animal anesthesia and perfusion and retinal wholemount preparation for immunohistochemistry were undertaken as described previously. 39 40  
Immunochemistry and Lectin Histochemistry
Dual labeling was used to covisualize mural cells and the vasculature, as previously described. 22 40 To identify both mature and immature mural cells, antibodies against desmin and SMA were used. Briefly, retinas were incubated first with either anti-desmin (Clone D33; DakoCytomation, Sydney, Australia) or anti-SMA antibodies (clone 1A4, DakoCytomation) and then Texas red-conjugated anti-mouse immunoglobulins (GE Healthcare, Castle Hill, NSW, Australia). The endothelium was then labeled with biotinylated Griffonia simplicifolia (GS) lectin (Sigma-Aldrich, Castle Hill, NSW) followed by streptavidin conjugated with FITC (GE Healthcare). 41 All antibodies were diluted with 1% bovine serum albumin in PBS, and 0.1% Triton X-100 in 0.1 M PBS was used for all washes. Washed retinal wholemounts were mounted ganglion-cell-layer-up in glycerol-PBS (2:1, vol/vol) or antifade medium (Prolong; Invitrogen-Molecular Probes, Eugene, OR). 
Determination of DER
Fluorescently labeled retinal wholemounts were examined by confocal microscopy with a Leica argon-krypton laser mounted on an epifluorescence photomicroscope (Axiophot; Leica, Heidelberg, Germany). FITC and Texas red fluorescence were excited sequentially at 488 and 588 nm, respectively. The DER was determined as previously described. 36 With a 40× objective, a pair of desmin/GS lectin images was collected for each of 10 fields of view in the immature plexus just proximal to the leading edge of vessel formation. Each confocal image was overlaid with a 10 × 10 equally spaced grid (Photoshop ver. 5.0; Adobe Systems, Mountain View, CA). The occurrence of desmin labeling relative to lectin labeling at the 100 intersection points yielded the DER. Our earlier report 36 showed that a threshold DER of 0.9 is indicative of vascular stability, and low DERs are associated with an unstable vascular bed. 
Electron Microscopy
Retinas from control kittens aged P1, P3, and P6, and from ROP kittens exposed to 4 days’ hyperoxia and 23 dRA were examined with transmission electron microscopy, as previously described. 36 42 One preretinal membrane obtained during the recovery phase of ROP at 23 dRA was also examined. 
Results
Pericyte Changes on Nonperfused Capillary Segments after Hyperoxia in an Obliterative Kitten Model of ROP
Exposure of P1 kittens to hyperoxia for 4 days resulted in substantial vaso-obliteration, with vascular remnants that were separated from the circulation scattered throughout the central retina. To examine the effects of nonperfusion of vascular segments on pericytes, we used the vaso-obliterative kitten model of ROP. The newborn kitten retina has a trilobular vascular plexus centered on the optic nerve head with immature SMA-positive SMCs present on the major radial arteries and immature pericytes with strong desmin labeling on the immature capillaries. 36 After the kittens were exposed to 3 dRA, scattered islands of capillary plexus remnants were still apparent beyond the neovasculature. Only some remnants showed weak desmin labeling (Figs. 1A 1B) , suggesting loss or dedifferentiation of the immature pericytes on nonperfused capillaries. The extent of desmin ensheathment on these vascular remnants were significantly reduced compared with that observed during normal development. 36  
Altered Pericyte Phenotype in Neovasculature during Hypoxic and Subsequent Vascular Remodeling Phases of an Obliterative Model of ROP
To examine the effects of the hypoxic phase of the obliterative model of kitten ROP on pericytes, we assessed desmin ensheathment of the vessels. Neovascularization had begun after 3 dRA, which followed 4 days of hyperoxia. A remarkable feature of this neovasculature is that it is multilayered, as opposed to the bistratified superficial and deeper plexus seen during normal development of the kitten 41 and human 40 43 retinal vasculature. At this stage, the neovasculature consisted of a more superficial primitive capillary plexus and a second plexus located just deep to the superficial plexus that consisted of 24 to 28 radial vessels compared with the 3 radial artery-vein pairs in normally developing kitten retina. 9 38 Desmin coverage of these vessels (Figs. 1C 1D)was similar to, if not more extensive than, that in the denser capillary plexuses of control P6 retinas (Figs. 1G 1H) . Indeed, unlike the control P6 vasculature (Figs. 1I 1J) , the terminals of vascular segments extending peripherally were rounded rather than flattened, and all were ensheathed with desmin filaments (Figs. 1E 1F)
After 10 dRA the superficial capillary plexus had increased greatly in density, consisting of several capillary layers distinct from the flat single layer of capillaries present during normal development. Desmin filaments were associated with this dense, proliferating mesh; however, they were shorter than those in normal development (compare Figs. 1K 1Lwith 1G 1H ) and were absent from many of the enlarged vascular terminals at the leading edge of vessel formation (Figs. 1M 1N)
By 27 dRA, the dense capillary mesh was regressing. Short desmin filaments were associated with vessels in the early stages of regression (Figs. 1O 1P) , but rarely with the remnants of regressed vessels (Figs. 1Q 1R) . The presence of scattered points of desmin labeling on these regressed vessels suggests that the mural cells of these plexuses either dedifferentiate or die. 
Altered Glial-Pericyte-Endothelial Cell Interaction in Preretinal Membranes
The presence of a preretinal membrane in human ROP is a prognostic indicator of a poor visual outcome. For this reason, we examined the glial-pericyte-endothelial relationship in preretinal membranes in the obliterative model of kitten ROP, to determine whether they play a role in this condition. After 10 to 27 days of subsequent exposure to room air, preretinal membranes composed of dense aggregations of neovasculature and accompanying glial cells were seen in the vitreous (Fig. 2) . At the electron microscope level, the vascular component of these membranes consisted primarily of dilated capillary-like vessels, many with significant mural cell coverage (Fig. 2A) . However, mural cell coverage in these vessels was extremely irregular, some having an almost complete investment of pericyte processes and others virtually none; the extent of mural cell coverage showed no obvious correlation with vessel size. The ultrastructure of the pericytes in these vessels suggested a proliferative-actively synthetic phenotype having a cytoplasm rich in cell organelles, especially rough endoplasmic reticulum and euchromatic nuclei with large nucleoli (Fig. 2B) . Unlike the vessels of the intraretinal neovasculature endothelial fenestrations were seen, and glial contact was reduced, with glial cells found primarily at one edge of the membrane (Fig. 2C) . Another remarkable feature of preretinal vessels that differ significantly from healthy intraretinal vessels was the presence of fenestrations in the endothelial cells of some preretinal vessels (Fig 2D)
Altered Hemodynamics and Tissue Hypoxia in the Modified Nonobliterative Model of ROP
The nonobliterative model was induced by placing the kitten in hyperoxic conditions for 2 days, which did not result in widespread vaso-obliteration, with the vasculature remaining substantially intact. The difference between the obliterative and nonobliterative models is that in the nonobliterative model, we have the hypoxia-induced neovasculature incorporated within the remaining vascular remnants, whereas the obliterative model is an optic disc-centered neovascularization of an avascular tissue, driven by a massive hypoxic signal. After 3 days’ recovery in air, at P8, the vascular plexus after 2 days’ exposure to hyperoxia (Fig. 3B)resembled that in P3 control kittens (Fig. 3A)with a trilobular topography centered on the optic nerve head and the dense peripheral capillary mesh of each lobe fed by a radial artery-vein pair. However, extension of the vascular plexus into the peripheral retina was clearly delayed, being equivalent in extent of retinal vascularization to the P3 vasculature. 
The peripheral vasculature appeared similar to that at P6 and P3, with a flat, dense, immature capillary mesh (Fig. 4A) . However, there was evidence of vessel regression, not only along the arteries (Figs. 4B 4C) , which is normal at this developmental stage, but also in the capillary plexuses. Vessel regression was particularly evident in the interlobular regions (Fig. 4D)and along segments of radial veins (Figs. 4C 4E) , resulting in blind-ended “vascular stumps,” in which normal perfusion would be markedly compromised (Figs. 4C 4Farrows). 
During normal development of the kitten retina, a periarterial capillary-free space develops as a result of the higher oxygen tension of arterial blood. 44 A capillary-free space is typically absent surrounding venules during normal development of the kitten retina, due to the lower oxygen tension of venous blood (see Fig. 2 , Ref. 36 ). In marked contrast, vessel regression resulting in a perivenular, capillary-free space (Fig. 4C)was evident in the nonobliterative model of kitten ROP, indicative of a marked change in hemodynamics and oxygen tension in these retinal beds. This observation further supports desmin and SMA immunohistochemistry, which showed dedifferentiation between arterial and venous vessels in this model (described later). 
In Figure 4C , a series of small white arrowheads demarcate a region of poorly perfused capillary plexus. Furthermore, in the central retina, the radial veins were more tortuous (Fig. 4B)and in regions where capillary density was low, presumably as a result of pathologic vessel regression, vascular sprouting was apparent (Fig. 4F , arrowheads), as evidenced by tufts of filopodia at high magnification. After 29 dRA, although the vasculature had reached the retinal periphery and the outer plexus had formed, the arteriolar trees were less ordered and vascular remodeling was still widespread (Fig. 4G)compared with the vasculature of P28 control retinas (Fig. 4H)
Altered Pericyte Phenotype and Reduced DER in Surviving Vasculature of the Nonobliterative Model of ROP
Although the capillary plexus after 3 dRA resembled that in control developing retinas, the mural cells in these plexuses resembled those in the multilayered neovasculature of the vaso-obliterative model of ROP. A large number of short, scattered desmin filaments were present in the dense capillary monolayer, whereas longer filaments were present on the surface of vascular stumps and segments that had vascular sprouts (Figs. 5A 5B 5C 5D) . Earlier, we introduced a quantitative measure of the pericyte-endothelial cell interaction, the DER. Control P6 immature capillary plexuses had a DER of 0.36, whereas those of the surviving vasculature had a lower DER (0.25). 
After 29 dRA, the remodeling capillary plexuses at the vascular periphery were denser and had a larger number of rounded endothelial cell bodies than did those at P32 in control retinas (compare Figs. 5E 5Kwith Figs. 5G 5I ). Long desmin filaments were now present in some plexuses (Figs. 5F 5L)as in control plexuses (Figs. 5H 5J) ; however, the DER was at 0.38, still well below the stability threshold of 0.9. This DER is similar to that at P6 in control retinas and is indicative of vessel instability compared with 0.97 at P28 and 0.90 at P32 in control peripheral vascular plexuses—DERs associated with vessel stability. 36  
Alterations in SMCs on Radial Arterioles and Venules in Hypoxia-Induced Neovasculature in the Obliterative Model
To determine the effect of hypoxia on SMCs in radial arterioles and venules, we examined expression of αSMA and desmin on the vessels in the obliterative model. At P6 in control kitten retinas, there was a clear distinction between the radial arterioles and radial venules. The radial arterioles showed strong homogeneous staining with antibodies against SMA, consistent with continuous SMC coverage, but weak diffuse desmin labeling. In contrast, venules had a patchy SMA labeling pattern, but well-defined desmin filaments arranged in a stellate conformation on their abluminal surface (Figs. 6A 6B 6C 6D 6E 6F) . In the first week, after 4 days of hyperoxia, radial arterioles were distinguishable from radial venules in the neovasculature on the basis of their homogeneous SMA labeling; however, SMA labeling of both vessel types was weak (Figs. 6G 6H) ; SMA labeling of the arterioles was less even than in P6 control retinas and SMA labeling of venules was floccular rather than the organized structures of the P6 veins (Figs. 6I 6J)
At 10 days, SMA labeling of the neovascular radial arterioles had increased, whereas venous SMA labeling remained faint (Fig. 6K) . Venular desmin filaments appeared shorter and resembled those in the dense capillary plexus, whereas radial arterioles had more defined stellate desmin filaments when compared with P6 control retinas. Desmin labeling did not distinguish them from venules (Figs. 6M 6N) , and they resembled those in the dense capillary plexus. By 27 dRA, the radial vessels resembled those in control retinas in both SMA labeling (Fig. 6L)and desmin filaments (data not shown). 
After 23 dRA, the radial arterioles and radial venules still appeared dilated, and the arterioles had only a very thin but complete layer of SMCs, resembling control radial venules (Fig. 7A) . More peripherally, these gave rise to dilated thin-walled vessels that fed the dense capillary plexus (Fig. 7B) . Electron microscopic examination revealed that radial arteriolar walls had two layers of SMCs that uncharacteristically displayed numerous overlapping processes that in places made it appear that more than two cell layers were present (Figs. 7C 7E) . The SMCs on these vessels had extremely attenuated cell processes, scanty dense plaques at the sarcolemma and insubstantial microfilament bundles consistent with immaturity. The basement membranes surrounding the SMCs on both the radial arterioles and the smaller feeder arterioles were discontinuous, laminated, and floccular in appearance (Figs. 7D 7E) . However, the perinuclear cytoplasm of the SMCs contained an unusually large amount of rough endoplasmic reticulum, suggesting that the cells were actively engaged in the production of proteins for export, possibly additional matrix. Radial venular walls were extremely attenuated with only occasional mural cells (Figs. 7F 7G) . The glial interface with the smaller arterioles and venules was irregular, and the basal laminae of the glial cells were often separated from those of the vasculature by pockets of flocculent basement membrane-like material (Figs. 7H)
Dedifferentiation of SMCs on Arterioles and Venules That Had Survived Exposure to Hyperoxia
After 4 days’ exposure to hyperoxia, remnants of radial vessels were rare, and only some labeled weakly with SMA (Figs. 6O 6P) , consistent with smooth muscle disintegration or dedifferentiation on immature radial arteries with exposure to hyperoxia or loss of circulation. After 3 days’ exposure to room air, surviving radial vessels were seen in only one retina, identified by SMA labeling, radial orientation, extension beyond the neovasculature, and apparent connection with the optic nerve head. Unlike control P6 radial arteries, endothelial protuberances were apparent on the abluminal wall of these vessels, suggestive of endothelial cell hyperplasia and SMA labeling was uneven (Figs. 6Q 6R)
In the nonobliterative kitten model of ROP the extent of SMA labeling in the surviving vasculature resembled that at P3. SMA labeling of radial and primary arterioles declined distally, and labeling of the radial veins was weaker (Figs. 3 6S) . As in the early neovasculature compared with control vessels of P6 retinas, the SMA labeling of the radial arterioles was even less (Figs. 6S 6T) , and the SMA labeling of radial venules was lower, being barely detectable (Fig. 6U) . This pattern of SMA arteriolar labeling was distinct from the weak, even labeling in control arteriolar SMCs at the very first stage of differentiation, as judged by the expression of SMA (Fig. 6V) . As in the neovasculature, the desmin filaments on the radial venules were shorter, resembling those in the capillary plexus, and similar filaments were present on the radial arterioles (Figs. 6W 6X)
Discussion
As in previous studies, 4 days of hyperoxia resulted in vaso-obliteration, followed by aberrant vasoproliferation on return to room air, resulting in preretinal membranes and a multilayered dense retinal neovasculature, fed by an increased number of radial vessels, that was later remodeled. Our results showed that mural cell dedifferentiation and/or loss occurred in nonperfused vascular remnants. We modified this model of ROP by reducing the duration of hyperoxia exposure so that it more closely resembled the human condition of ROP. In this model, as in human ROP, retinal vascularization was retarded, with most existing vessels surviving the hyperoxic insult, in contrast to the established model in which the entire retinal vasculature is rendered nonfunctional by widespread vessel regression in response to prolonged hyperoxia. This new nonobliterative model of ROP displayed the radial vessel tortuosity and enlargement that characterizes plus disease and the newly described clinical entity of “pre-plus disease,” but does not mimic the end stages of human plus disease of extraretinal fibrovascular proliferation followed by tractional detachment. 
Electron microscopy confirmed that mural cells were present on the vessels of the preretinal membranes. Desmin labeling revealed that the immature mural cells of the surviving capillary plexuses of the nonobliterative model bore a closer resemblance to those of the aberrant multilayered and remodeling capillary neovasculature of the obliterative model, than to mural cells of the immature capillary plexuses in control animals. As in the neovasculature of the obliterative model, 36 the DER of the surviving plexus of the nonobliterative model was lower than that of the control and remained well below the threshold DER for vascular stability of 0.9, even after 29 dRA, indicating that a comparatively brief exposure to hyperoxia results in increased vessel instability and prolongs retinal vascular instability observed during neonatal development. 
During the hypoxic phase, in both the neovasculature of the obliterative model and the surviving vasculature of the nonobliterative model, the radial vessels were enlarged and their SMCs showed desmin morphology and electron microscopy features consistent with immaturity. 
Causes of Mural Cell Changes
The loss of desmin filaments on the nonperfused remnants of the capillary plexus and the loss of SMA labeling on the nonperfused remnant radial vessels in the obliterative model indicate mural cell loss and/or dedifferentiation. Mural cell survival is dependent on endothelium-derived PDGFβ during development, 32 45 and mural cell phenotype correlates with vessel hemodynamics. Thus, possible causes of mural cell loss or dedifferentiation include loss of perfusion and/or endothelial cell loss. 
An altered pericyte phenotype characterized by shortened desmin filaments was seen on both proliferating and regressing vessels. Possible causes of this phenotype include the upregulation of angiopoietin (Ang)-2 and impaired perfusion. Upregulation of Ang-2 expression is associated with both vaso-proliferation and vessel regression, 46 as occurs in the murine model of oxygen-induced ischemic retinopathy. 47 Ang-2 has been postulated to destabilize vessels by promoting the weakening of mural cell associations with the endothelium, 48 and a recent study has shown that intravitreal injection of Ang-2 causes pericyte loss in the retinal vasculature of adult rats. 49  
Effects of Altered Pericyte-Mural Cell Phenotype
We suggest that the impaired mural cell differentiation observed in this study may further decrease the stability of an already unstable, immature plexus. Mural cells have been implicated in vessel stabilization, with a decreased number of mural cells resulting in endothelial hyperplasia 31 and increased extraretinal neovascularization in rodent models of ROP. 45 50 On this issue it is significant that SMCs and pericytes in maturing vascular plexuses express Ang-1, 51 which is thought to promote vessel stabilization, with its overexpression decreasing retinal neovascularization during the hypoxic phase of a mouse model of ischemic retinopathy. 52 In contrast, immature mural cell phenotypes are associated with vessel instability. 22 36  
The DER of the surviving vasculature in ROP is initially lower than that of the developing retinal vasculature in control retinas, which are already considered unstable with DERs < 0.9. Retinal vessels in the nonobliterative model of ROP displaying vessel dilation and tortuosity had a DER comparable to that observed on highly pathologic vessels found in preretinal membranes. 36 We suggest that the lower DER is consistent with a vascular plexus even less stable than that present during normal development, which is acutely receptive to both angiogenic and vessel withdrawal signals. We postulate that the more unstable the plexus, the more vulnerable it is to fluctuations in VEGF, with increased VEGF levels exacerbating aberrant neovascularization and reduced VEGF promoting regression. Such aberrant neovascular complexes may increase capillary bed resistance at the leading edge of vessel formation and paradoxically impede the supply of nutrients and oxygen to the peripheral retina, further increasing VEGF levels. As hypoxia and VEGF increase the expression of Ang-2, 53 it is possible that the tissue environment of the peripheral retina during ROP is inhibitory to mural cell maturation. 
VEGF increases vascular permeability, and the neovasculature of the obliterative model lacks normal barrier properties, 34 and it is likely that impaired mural cell differentiation exacerbates vessel leakiness. Recent studies suggest that pericytes contribute to endothelial cell barrier properties by the paracrine secretion of Ang-1. Ang-1 decreases the permeability of endothelial junctions, apparently by affecting components of both the tight junction 54 and the adherens junction. 55 56 In these actions Ang-1 directly opposes VEGF-induced vascular leakiness. 52 57  
Finally, impaired SMC differentiation in ROP suggests a decreased ability to regulate blood flow. The underdevelopment of microfilament apparatus in the arteriolar SMCs and the decreased number of these cells in the vessel wall suggest an impaired capacity to regulate blood flow and respond to hemodynamic changes, compared with the retinal arterioles in control animals. 
Implications for ROP and Plus Disease
Increased VEGF expression cause arteriole and venule enlargement; an effect that appears to be dependent on the Ang-2 expression in the retina. 58 59 Thus, the radial vessel dilation in plus disease may be a sign of increased VEGF and Ang-2 levels in the posterior retina. 
In both models of ROP, radial vessel dilation was accompanied by impaired SMC differentiation during the hypoxic phase; thus, an impaired ability to regulate blood flow may further contribute to radial vessel dilation in these models. In both models, arterioles could no longer be distinguished from venules on the basis of desmin filament ensheathment. Similarly, in AP-ROP the distinction between arteries and veins is lost 18 although radial vessel tortuosity, the hallmark of plus disease, was seen only in the nonobliterative model. 
Increased VEGF levels cause retinal endothelial cell hypertrophy and capillary luminal narrowing which is likely to compromise blood flow. 60 In the nonobliterative model capillary bed, resistance is postulated to further increase with the decrease in connections between the capillary beds and the radial vessels, and with vasoproliferation resulting in increased capillary density, whereas in the obliterative model, although the capillary plexuses are very dense, they are fed by an increased number of radial vessels. Indeed, in the canine model of ROP where vessel regression is more widespread, radial tortuosity during the hypoxic phase is more pronounced, 61 raising the possibility that vessel tortuosity is the result of increased capillary bed resistance that may be due to either aberrant vasoproliferation or vessel withdrawal at the vascular periphery. The development of vascular shunts in human ROP could be an adaptation in an attempt to overcome pronounced capillary bed resistance and further support the similarity of the pathogenetic mechanisms underlying vessel tortuosity and dilation in human, feline, and canine ROP. 
Our observations lead us to speculate that plus disease is indicative of a highly unstable vasculature with elevated VEGF and Ang-2 levels in the posterior retina, an impaired ability to regulate retinal blood flow and severely compromised blood flow at the vascular periphery resulting in altered hemodynamics in the radial vessels. 
A remarkable finding in this study was the concordance between the clinical presentation of plus disease and the changes reported. The progression of human ROP is characterized by retinal vessel morphology, which clearly allows differentiation of artery and vein, but which becomes distorted over time by dilation and tortuosity to become plus disease. We report a similar dedifferentiation of SMCs on arterioles and venules. 
Revised indications for the treatment of ROP now support ablative therapy for eyes at any stage of plus disease. Data from this study suggest a hypoxic environment and unstable vasculature in plus disease, thus providing a pathogenetic basis and scientific rationale for this recommendation. 
Reduced Blood Flow: Adjunct in Diagnosis of Plus Disease
We have provided evidence of impaired SMC differentiation in both the obliterative and nonobliterative models of kitten ROP. The underdevelopment of microfilament apparatus in the arteriolar SMCs and the decreased number of these cells in the vessel wall suggest an impaired capacity to regulate blood flow and respond to hemodynamic changes, compared with the retinal arterioles in control animals. These observations lead us to suggest that measurement of blood flow in infants could be a suitable quantitative indicator of plus disease in addition to ophthalmoscopic examination. This suggestion is supported by Orawiec et al., 62 who found a significant reduction in maximum systolic velocity (V max) in the central retinal and ophthalmic arteries of children with threshold ROP (with plus disease) compared with children with prethreshold ROP (without plus disease). 
Indication for Combinational Therapy for Plus Disease
Our data provide compelling evidence of the contribution of a hypoxia-induced, VEGF-mediated change to the retinal vasculature and phenotypic alterations in pericytes and SMCs in the pathogenesis of plus disease in ROP. Based on these observations, we suggest that therapy combining anti-VEGF antiangiogenic therapy, Ang-1 for vessel normalization, and PDGFβ for pericyte recruitment and vessel stabilization could offer an effective combinational form of intervention in cases of ROP where plus disease, as defined by significant dilation and tortuosity of posterior retinal vessels, is present. 
Conclusion
Our observations in kitten models of ROP show that during the hypoxic phase, pericytes adopt a phenotype associated with increased vascular instability and that SMC differentiation is impaired, implying a decreased ability to regulate blood flow. Our findings are consistent with the proposal that plus disease is the manifestation of a highly unstable vasculature with impaired regulation of blood flow and highly compromised blood flow at the vascular periphery, resulting in a vascular bed that is more susceptible to fluctuations in oxygen levels and growth factors and thus prone to vicious cycles of increasing VEGF levels. Similar changes in the human condition could contribute to the pathogenesis of ROP and plus disease. 
Further, vessel tortuosity was evident only in the nonobliterative model of kitten ROP, not in the obliterative model. Significant capillary closure evident around surviving large vessels in the nonobliterative model leads us to suggest that local mechanical forces due to the unusual nature of the remodeled vascular tree are associated with arteriole dilation. 
W conclude that plus disease is associated with changes to pericytes and SMCs that may be associated with alterations in blood flow. More work is needed to verify this conclusion. Dedifferentiation of SMCs on retinal arterioles and venules, aberrant pericyte phenotypes, altered hemodynamics, and high VEGF and Ang-2 levels may all contribute to the arterial and venous dilation and tortuosity observed clinically. 
 
Figure 1.
 
Desmin filaments (red) on lectin-labeled vasculature (green): (A, B) Retinal vascular remnants after exposure to hyperoxia for 4 days followed by 3 dRA; (C–F) retinal neovasculature after exposure to hyperoxia for 4 days followed by 3 dRA central to (C, D) and at (E, F) the leading edge of vessel formation. (G–J) Retinal vasculature of control P6 kitten central to (G, H) and at (I, J) the leading edge of vessel formation. (K–N) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 10 dRA central to (K, L) and at (M, N) the leading edge of vessel formation. (O–R) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 27 dRA in the early (O, P) and late (Q, R) stages of regression. Scales listed for (A, C, E, G, I, K, M, O, Q) are the same for companion panels (B, D, F, H, J, L, N, P, R). Scale bars: (A, B, O–R) 50 μm; (C–N) 25 μm.
Figure 1.
 
Desmin filaments (red) on lectin-labeled vasculature (green): (A, B) Retinal vascular remnants after exposure to hyperoxia for 4 days followed by 3 dRA; (C–F) retinal neovasculature after exposure to hyperoxia for 4 days followed by 3 dRA central to (C, D) and at (E, F) the leading edge of vessel formation. (G–J) Retinal vasculature of control P6 kitten central to (G, H) and at (I, J) the leading edge of vessel formation. (K–N) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 10 dRA central to (K, L) and at (M, N) the leading edge of vessel formation. (O–R) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 27 dRA in the early (O, P) and late (Q, R) stages of regression. Scales listed for (A, C, E, G, I, K, M, O, Q) are the same for companion panels (B, D, F, H, J, L, N, P, R). Scale bars: (A, B, O–R) 50 μm; (C–N) 25 μm.
Figure 2.
 
Ultrastructure of preretinal vessels present after exposure to 4 days of hyperoxia followed by 23 days in room air in a kitten model of ROP. (A) Dilated preretinal vessels consisting of endothelial cells (E) with sporadic pericyte (P) ensheathment. (B) Densely packed new vessels in the preretinal membrane showing several plump pericytes (P) displaying organelle-rich cytoplasm and active euchromatic nuclei with large nucleoli. Pericyte process (p). (C) Glial cells (G) were present but were not closely associated with dilated vessels. (D) Fenestrations were apparent in the endothelial cells of some preretinal vessels.
Figure 2.
 
Ultrastructure of preretinal vessels present after exposure to 4 days of hyperoxia followed by 23 days in room air in a kitten model of ROP. (A) Dilated preretinal vessels consisting of endothelial cells (E) with sporadic pericyte (P) ensheathment. (B) Densely packed new vessels in the preretinal membrane showing several plump pericytes (P) displaying organelle-rich cytoplasm and active euchromatic nuclei with large nucleoli. Pericyte process (p). (C) Glial cells (G) were present but were not closely associated with dilated vessels. (D) Fenestrations were apparent in the endothelial cells of some preretinal vessels.
Figure 3.
 
Schematic representation of a portion of the vasculature (green) and the distribution of SMA labeling (red) in this vasculature (A), in the retina of a control kitten retina at P3 and (B) in the retina of a kitten exposed from P3 to 2 days of hyperoxia and then returned to room air for 3 days. Scale bars, 500 μm.
Figure 3.
 
Schematic representation of a portion of the vasculature (green) and the distribution of SMA labeling (red) in this vasculature (A), in the retina of a control kitten retina at P3 and (B) in the retina of a kitten exposed from P3 to 2 days of hyperoxia and then returned to room air for 3 days. Scale bars, 500 μm.
Figure 4.
 
(A–F) Regions of the retinal vasculature double labeled with lectin (green) and SMA immunohistochemistry (red) of a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 3 days. (A) A dense peripheral capillary plexus and (B) a tortuous radial venule and regressed vessel branching from a radial arteriole. (C) A dense capillary plexus, fed by an SMA-labeled arteriole (A) on the left from which several vessels have regressed resulting in vascular stumps (arrows), and drained by a radial venule (V) on the right. Vessel regression in the perivenular region was highly abnormal and indicative of altered tissue oxygenation as a result of irregular vessel hemodynamics. Small white arrowheads: a region of poorly perfused capillary plexus. The vascular tree was indicative of a dense capillary plexus with decreased arteriolar blood supply and venular drainage. (D) Regressing vessels in the interlobular capillary plexus, (E) decreased vascular density, which is evidence of vessel regression, along a radial venule, and (F) a region of decreased capillary density, evidence of vessel regression, with vascular stumps (arrow) and foci of vascular sprouting (arrowheads). (G, H) Lectin-labeled primary arterioles branching from radial arterioles in the retina of (G) a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 29 days and (H) a control P28 kitten. Scale bars: (A, B, D–H) 125 μm.
Figure 4.
 
(A–F) Regions of the retinal vasculature double labeled with lectin (green) and SMA immunohistochemistry (red) of a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 3 days. (A) A dense peripheral capillary plexus and (B) a tortuous radial venule and regressed vessel branching from a radial arteriole. (C) A dense capillary plexus, fed by an SMA-labeled arteriole (A) on the left from which several vessels have regressed resulting in vascular stumps (arrows), and drained by a radial venule (V) on the right. Vessel regression in the perivenular region was highly abnormal and indicative of altered tissue oxygenation as a result of irregular vessel hemodynamics. Small white arrowheads: a region of poorly perfused capillary plexus. The vascular tree was indicative of a dense capillary plexus with decreased arteriolar blood supply and venular drainage. (D) Regressing vessels in the interlobular capillary plexus, (E) decreased vascular density, which is evidence of vessel regression, along a radial venule, and (F) a region of decreased capillary density, evidence of vessel regression, with vascular stumps (arrow) and foci of vascular sprouting (arrowheads). (G, H) Lectin-labeled primary arterioles branching from radial arterioles in the retina of (G) a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 29 days and (H) a control P28 kitten. Scale bars: (A, B, D–H) 125 μm.
Figure 5.
 
(A–H) Desmin filaments (red) on vessels (green) of dense capillary monolayer (A, B) and on vessels with vascular stumps and vascular sprouting (C, D) in kitten retina after exposure to hyperoxia for 2 days followed by 3 dRA (modified model of kitten ROP); on vessels of a remodeled capillary plexus in a kitten retina after exposure to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP) (E, F) and on vessels of a remodeling capillary plexus in a control P32 kitten retina (G, H). (I–L) Determination of the DER. Images of a field of view of the vasculature in the peripheral retina of a control P32 kitten (I, J) and a kitten exposed to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP), (K, L) labeled with lectin (green) and anti-desmin antibodies (red). A grid with 100 intersection points was superimposed on each image and intersections with desmin or lectin labeling are shown (white dot). The ratio of desmin-positive intersections to lectin-positive intersections for a given field yields the DER. The DER of the field of view shown in (I, J) is 0.80 and the DER of the field of view shown in (K, L) is 0.44. Scale bars, 50 μm.
Figure 5.
 
(A–H) Desmin filaments (red) on vessels (green) of dense capillary monolayer (A, B) and on vessels with vascular stumps and vascular sprouting (C, D) in kitten retina after exposure to hyperoxia for 2 days followed by 3 dRA (modified model of kitten ROP); on vessels of a remodeled capillary plexus in a kitten retina after exposure to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP) (E, F) and on vessels of a remodeling capillary plexus in a control P32 kitten retina (G, H). (I–L) Determination of the DER. Images of a field of view of the vasculature in the peripheral retina of a control P32 kitten (I, J) and a kitten exposed to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP), (K, L) labeled with lectin (green) and anti-desmin antibodies (red). A grid with 100 intersection points was superimposed on each image and intersections with desmin or lectin labeling are shown (white dot). The ratio of desmin-positive intersections to lectin-positive intersections for a given field yields the DER. The DER of the field of view shown in (I, J) is 0.80 and the DER of the field of view shown in (K, L) is 0.44. Scale bars, 50 μm.
Figure 6.
 
Mural cells on radial venules (v) and arterioles (a) in the retinas of control kittens (A–F, V) and in the vaso-obliterative (G–R) and modified (S–U, W, X) kitten models of ROP. (A–D) SMA immunohistochemical labeling (red) of radial arterioles (A–C) and of radial venules (A, B, D) in the central retina of a P6 kitten. Desmin filaments (red) on a lectin-labeled (green) central radial arteriole and radial venule in a P6 retina (E, F). (G–L) SMA immunohistochemical labeling (red) of central radial arterioles and venules in retinas after 4 days’ exposure to hyperoxia followed by 3 (G, H), 7 (I, J), 10 (K), and 27 (L) dRA. (M, N) Desmin filaments (red) on lectin-labeled (green) central radial vessels of retina of kitten exposed to 4 days of hyperoxia followed by 10 dRA. (O–R) SMA immunohistochemical labeling (red) of remnants of radial vessels immediately after 4 days’ exposure to hyperoxia (O, P), and a radial arteriole that survived 4 days of hyperoxia after 7 dRA (Q, R). (S–U) SMA immunohistochemical labeling of central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 dRA, (V) SMA immunohistochemical labeling of a peripheral, less-mature segment of a radial arteriole in the retina of a control P6 kitten, and (W, X) desmin filaments (red) on the central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 days in room air. Scale bars: (A, B, E–H, K–N, W, X) 50 μm; (C, D, I, J, O–V) 25 μm.
Figure 6.
 
Mural cells on radial venules (v) and arterioles (a) in the retinas of control kittens (A–F, V) and in the vaso-obliterative (G–R) and modified (S–U, W, X) kitten models of ROP. (A–D) SMA immunohistochemical labeling (red) of radial arterioles (A–C) and of radial venules (A, B, D) in the central retina of a P6 kitten. Desmin filaments (red) on a lectin-labeled (green) central radial arteriole and radial venule in a P6 retina (E, F). (G–L) SMA immunohistochemical labeling (red) of central radial arterioles and venules in retinas after 4 days’ exposure to hyperoxia followed by 3 (G, H), 7 (I, J), 10 (K), and 27 (L) dRA. (M, N) Desmin filaments (red) on lectin-labeled (green) central radial vessels of retina of kitten exposed to 4 days of hyperoxia followed by 10 dRA. (O–R) SMA immunohistochemical labeling (red) of remnants of radial vessels immediately after 4 days’ exposure to hyperoxia (O, P), and a radial arteriole that survived 4 days of hyperoxia after 7 dRA (Q, R). (S–U) SMA immunohistochemical labeling of central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 dRA, (V) SMA immunohistochemical labeling of a peripheral, less-mature segment of a radial arteriole in the retina of a control P6 kitten, and (W, X) desmin filaments (red) on the central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 days in room air. Scale bars: (A, B, E–H, K–N, W, X) 50 μm; (C, D, I, J, O–V) 25 μm.
Figure 7.
 
(A, B) Semithin and ultrathin sections of a retina from a kitten exposed to 4 days of hyperoxia followed by 23 dRA. (A) A semithin section from the central retina showing a thin-walled radial arteriole (A) and dilated venule (V); (B) from a more peripheral block showing smaller, thin-walled, poorly differentiated vessels (arrows) that were fed by the radial vessels and directly fed into the dense capillary plexus. (C, E) Electron microscopic images of the wall of the arteriole in (A). The endothelium (E) is attenuated, in keeping with the dilated state of the vessel, whereas the smooth muscle cells (SM) show uncharacteristic overlapping processes (arrows), insubstantial microfilament bundles and an organelle-rich cytoplasm, consistent with immaturity. (F, G) Electron microscopic images of the venule shown in (A). The endothelium (E) is extremely attenuated and only partially covered by scanty mural cell processes (arrows). (D, H) Electron microscopic images of one of the smaller feeder arteriolar vessels in (B). The complete investiture of the abluminal surface of the endothelium (E) by mural cell processes is consistent with cells representing smooth muscle rather than pericytes; however, the cells are poorly differentiated with inconspicuous microfilament bundles and an organelle-rich cytoplasm (arrows). In both profiles of this vessel the glial component of the basement membrane is separated from that of the vascular cells with the intervening space filled by flocculent matrix material (Mx).
Figure 7.
 
(A, B) Semithin and ultrathin sections of a retina from a kitten exposed to 4 days of hyperoxia followed by 23 dRA. (A) A semithin section from the central retina showing a thin-walled radial arteriole (A) and dilated venule (V); (B) from a more peripheral block showing smaller, thin-walled, poorly differentiated vessels (arrows) that were fed by the radial vessels and directly fed into the dense capillary plexus. (C, E) Electron microscopic images of the wall of the arteriole in (A). The endothelium (E) is attenuated, in keeping with the dilated state of the vessel, whereas the smooth muscle cells (SM) show uncharacteristic overlapping processes (arrows), insubstantial microfilament bundles and an organelle-rich cytoplasm, consistent with immaturity. (F, G) Electron microscopic images of the venule shown in (A). The endothelium (E) is extremely attenuated and only partially covered by scanty mural cell processes (arrows). (D, H) Electron microscopic images of one of the smaller feeder arteriolar vessels in (B). The complete investiture of the abluminal surface of the endothelium (E) by mural cell processes is consistent with cells representing smooth muscle rather than pericytes; however, the cells are poorly differentiated with inconspicuous microfilament bundles and an organelle-rich cytoplasm (arrows). In both profiles of this vessel the glial component of the basement membrane is separated from that of the vascular cells with the intervening space filled by flocculent matrix material (Mx).
The authors thank Ruth-Ann Sterling for assistance with digital imaging and artwork and John Flynn and Glen Gole for helpful comments on the manuscript. 
SeiberthV, LinderkampO. Risk factors in retinopathy of prematurity a multivariate statistical analysis. Ophthalmologica. 2000;214:131–135. [CrossRef] [PubMed]
DarlowBA, HutchinsonJL, Henderson-SmartDJ, DonoghueDA, SimpsonJM, EvansNJ. Prenatal risk factors for severe retinopathy of prematurity among very preterm infants of the Australian and New Zealand Neonatal Network. Pediatrics. 2005;115:990–996. [CrossRef] [PubMed]
PhelpsD. Retinopathy of prematurity: history, classification, and pathophysiology. Neoreviews. 2001;2:E153–E166. [CrossRef]
WagnerRS. Increased incidence and severity of retinopathy of prematurity in developing nations. J Pediatr Ophthalmol Strabismus. 2003;40:193. [PubMed]
WheatleyCM, DickinsonJL, MackeyDA, CraigJE, SaleMM. Retinopathy of prematurity: recent advances in our understanding. Br J Ophthalmol. 2002;86:696–700. [CrossRef] [PubMed]
Chan-LingT, GockB, StoneJ. The effect of oxygen on vasoformative cell division evidence that “physiological hypoxia” is the stimulus for normal retinal vasculogenesis. Invest Ophthalmol Vis Sci. 1995;36:1201–1214. [PubMed]
Chan-LingT, GockB, StoneJ. Supplemental oxygen therapy basis for noninvasive treatment of retinopathy of prematurity. Invest Ophthalmol Vis Sci. 1995;36:1215–1229. [PubMed]
StoneJ, ItinA, AlonT, et al. Development of retinal vasculature is mediated by hypoxia-induced vascular endothelial growth factor (VEGF) expression by neuroglia. J Neurosci. 1995;15:4738–4747. [PubMed]
Chan-LingT, ToutS, HollanderH, StoneJ. Vascular changes and their mechanisms in the feline model of retinopathy of prematurity. Invest Ophthalmol Vis Sci. 1992;33:2128–2147. [PubMed]
Chan-LingT. Glial, neuronal and vascular interactions in the mammalian retina.OsborneN ChaderG eds. Progress in Retinal Research. 1994;357–389.Pergamon Press Oxford, UK.
ICROP. An international classification of retinopathy of prematurity. The Committee for the Classification of Retinopathy of Prematurity. Arch Ophthalmol. 1984;102:1130–1134. [CrossRef] [PubMed]
Cryotherapy for Retinopathy of Prematurity Cooperative Group. Multicenter trial of cryotherapy for retinopathy of prematurity: one-year outcome—structure and function. Arch Ophthalmol. 1990;108:1408–1416. [CrossRef] [PubMed]
SchafferDB, PalmerEA, PlotskyDF, al. Prognostic factors in the natural course of retinopathy of prematurity The Cryotherapy for Retinopathy of Prematurity Cooperative Group. Ophthalmology. 1993;100:230–237. [CrossRef] [PubMed]
SaundersRA, BluesteinEC, SinatraRB, WilsonME, O’NeilJW, RustPF. The predictive value of posterior pole vessels in retinopathy of prematurity. J Pediatr Ophthalmol Strabismus. 1995;32:82–85. [PubMed]
Early Treatment for Retinopathy of Prematurity Cooperative Group. Revised indications for the treatment of retinopathy of prematurity: results of the early treatment for retinopathy of prematurity randomized trial [see comment]. . 2003;121:1684–1694.
HardyRJ, PalmerEA, DobsonV, et al. Risk analysis of prethreshold retinopathy of prematurity. Arch Ophthalmol. 2003;121:1697–1701. [CrossRef] [PubMed]
PhelpsDL. The Early Treatment for Retinopathy of Prematurity study: better outcomes, changing strategy. Pediatrics. 2004;114:490–491. [CrossRef] [PubMed]
Anonymous . The International Classification of Retinopathy of Prematurity revisited. Arch Ophthalmol. 2005;123:991–999. [CrossRef] [PubMed]
WallaceDK, KylstraJA, ChesnuttDA. Prognostic significance of vascular dilation and tortuosity insufficient for plus disease in retinopathy of prematurity. J AAPOS. 2000;4:224–229. [CrossRef] [PubMed]
OwensWC, OwensEU. Retrolental fibroplasia in premature infants; studies on the prophylaxis of the disease; the use of alpha tocopheryl acetate. Am J Ophthalmol. 1949;32:1631–1637. [CrossRef] [PubMed]
FlynnJT, CassadyJ, EssnerD, et al. Fluorescein angiography in retrolental fibroplasia: experience from 1969–1977. Ophthalmology. 1979;86:1700–1723. [CrossRef] [PubMed]
HughesS, Chan-LingT. Characterization of smooth muscle cell and pericyte differentiation in the rat retina in vivo. Invest Ophthalmol Vis Sci. 2004;45:2795–2806. [CrossRef] [PubMed]
DarlandDC, MassinghamLJ, SmithSR, PiekE, D’AmorePA. Pericyte production of cell-associated VEGF is differentiation-dependent and is associated with endothelial survival. Dev Biol. 2003;264:275–288. [CrossRef] [PubMed]
BenjaminLE, HemoI, KeshetE. A plasticity window for blood vessel remodelling is defined by pericyte coverage of the preformed endothelial network and is regulated by PDGF-B and VEGF. Development. 1998;125:1591–1598. [PubMed]
BeckL, Jr, D’AmorePA. Vascular development: cellular and molecular regulation. FASEB J. 1997;11:365–373. [PubMed]
KondoT, HosoyaK, HoriS, TomiM, OhtsukiS, TerasakiT. PKC/MAPK signaling suppression by retinal pericyte conditioned medium prevents retinal endothelial cell proliferation. J Cell Physiol. 2005;203:378–386. [CrossRef] [PubMed]
OrlidgeA, D’AmoreP. Inhibition of capillary endothelial cell growth by pericytes and smooth muscle cells. J Cell Biol. 1987;105:1455–1462. [CrossRef] [PubMed]
DarlandDC, D’AmorePA. Cell-cell interactions in vascular development. Curr Topics Dev Biol. 2001;52:107–149.
NguyenLL, D’AmorePA. Cellular interactions in vascular growth and differentiation. Int Rev Cytol. 2001;204:1–48. [PubMed]
LindahlP, JohanssonBR, LeveenP, BetsholtzC. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science. 1997;277:242–245. [CrossRef] [PubMed]
HellstromM, GerhardtH, KalénM, et al. Lack of pericytes leads to endothelial hyperplasia and abnormal vascular morphogenesis. J Cell Biol. 2001;153:543–553. [CrossRef] [PubMed]
UemuraA, OgawaM, HirashimaM, et al. Recombinant angiopoietin-1 restores higher-order architecture of growing blood vessels in mice in the absence of mural cells [comment]. J Clin Invest. 2002;110:1619–1628. [CrossRef] [PubMed]
BetsholtzC. Insight into the physiological functions of PDGF through genetic studies in mice. Cytokine Growth Factor Rev. 2004;15:215–228. [CrossRef] [PubMed]
Chan-LingT, StoneJ. Degeneration of astrocytes in feline retinopathy of prematurity causes failure of the blood-retinal barrier. Invest Ophthalmol Vis Sci. 1992;33:2148–2159. [PubMed]
StoneJ, Chan-LingT, Pe’erJ, ItinA, GnessinH, KeshetE. Roles of vascular endothelial growth factor and astrocyte degeneration in the genesis of retinopathy of prematurity. Invest Ophthalmol Vis Sci. 1996;37:290–299. [PubMed]
Chan-LingT, PageM, GardinerT, BaxterL, RosinovaE, HughesS. Desmin ensheathment ratio as an indicator of vessel stability: evidence in normal development and in retinopathy of prematurity. Am J Pathol. 2004;165:1301–1313. [CrossRef] [PubMed]
Chan-LingT, StoneJ. Retinopathy of prematurity: origins in the architecture of the retina. Progress in Retinal Research. 1993;155–177.Pergamon Press Oxford, UK.
AshtonN, WardB, SerpellG. Effect of oxygen on developing retinal vessels with particular reference to the problem of retrolental fibroplasia. Br J Ophthalmol. 1954;38:397–432. [CrossRef] [PubMed]
Chan-LingT. Glial, vascular, and neuronal cytogenesis in whole-mounted cat retina. Microsc Res Tech. 1997;36:1–16. [CrossRef] [PubMed]
Chan-LingT, McLeodD, HughesS, et al. Astrocyte-endothelial cell relationships during human retinal vascular development. Invest Ophthalmol Vis Sci. 2004;46:2020–2032.
Chan-LingTL, HalaszP, StoneJ. Development of retinal vasculature in the cat: processes and mechanisms. Curr Eye Res. 1990;9:459–478. [CrossRef] [PubMed]
HughesS, GardinerT, HuP, BaxterL, RosinovaE, Chan-LingT. Altered pericyte-endothelial relations in the rat retina during aging: implications for vessel stability. Neurobiol Aging. 2006;27:1838–1847. [CrossRef] [PubMed]
HughesS, YangH, Chan-LingT. Vascularization of the human fetal retina: roles of vasculogenesis and angiogenesis. Invest Ophthalmol Vis Sci. 2000;41:1217–1228. [PubMed]
PhelpsDL. Oxygen and developmental retinal capillary remodeling in the kitten. Invest Ophthalmol Vis Sci. 1990;31:2194–2200. [PubMed]
Wilkinson-BerkaJL, BabicS, De GooyerT, et al. Inhibition of platelet-derived growth factor promotes pericyte loss and angiogenesis in ischemic retinopathy. Am J Pathol. 2004;164:1263–1273. [CrossRef] [PubMed]
MaisonpierrePC, SuriC, JonesPF, et al. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis [see comments]. Science. 1997;277:55–60. [CrossRef] [PubMed]
HackettSF, OzakiH, StraussRW, et al. Angiopoietin 2 expression in the retina: upregulation during physiologic and pathologic neovascularization. J Cell Physiol. 2000;184:275–284. [CrossRef] [PubMed]
HanahanD. Signalling vascular morphogenesis and maintenance. Science. 1997;277:48–51. [CrossRef] [PubMed]
HammesH-P, MartinS, FederlinK, et al. Angiopoietin-2 causes pericyte dropout in the normal retina: evidence for involvement in diabetic retinopathy. Diabetes. 2004;53:1104–1110. [CrossRef] [PubMed]
HammesHP, LinJ, RennerO, et al. Pericytes and the pathogenesis of diabetic retinopathy. Diabetes. 2002;51:3107–3112. [CrossRef] [PubMed]
SundbergC, KowanetzM, BrownLF, DetmarM, DvorakHF. Stable expression of angiopoietin-1 and other markers by cultured pericytes: phenotypic similarities to a subpopulation of cells in maturing vessels during later stages of angiogenesis in vivo. Lab Invest. 2002;82:387–401. [CrossRef] [PubMed]
NambuH, NambuR, OshimaY, et al. Angiopoietin 1 inhibits ocular neovascularization and breakdown of the blood-retinal barrier. Gene Ther. 2004;11:865–873. [CrossRef] [PubMed]
OhH, TakagiH, SuzumaK, OtaniA, MatsumuraM, HondaY. Hypoxia and vascular endothelial growth factor selectively up-regulate angiopoietin-2 in bovine microvascular endothelial cells. J Biol Chem. 1999;274:15732–15739. [CrossRef] [PubMed]
HoriS, OhtsukiS, HosoyaK, NakashimaE, TerasakiT. A pericyte-derived angiopoietin-1 multimeric complex induces occludin gene expression in brain capillary endothelial cells through Tie-2 activation in vitro. J Neurochem. 2004;89:503–513. [CrossRef] [PubMed]
GambleJR, DrewJ[b], TreziseL, et al. Angiopoietin-1 is an antipermeability and anti-inflammatory agent in vitro and targets cell junctions. Circ Res. 2000;87:603–607. [CrossRef] [PubMed]
WangY, PampouS, FujikawaK, VarticovskiL. Opposing effect of angiopoietin-1 on VEGF-mediated disruption of endothelial cell-cell interactions requires activation of PKC beta. J Cell Physiol. 2004;198:53–61. [CrossRef] [PubMed]
NambuH, UmedaN, KachiS, et al. Angiopoietin 1 prevents retinal detachment in an aggressive model of proliferative retinopathy, but has no effect on established neovascularization. J Cell Physiol. 2005;204:227–235. [CrossRef] [PubMed]
OshimaY, DeeringT, OshimaS, et al. Angiopoietin-2 enhances retinal vessel sensitivity to vascular endothelial growth factor. J Cell Physiol. 2004;199:412–417. [CrossRef] [PubMed]
WitmerAN, Van BlijswijkBC, Van NoordenCJF, VrensenGFJM, SchlingemannRO. In vivo angiogenic phenotype of endothelial cells and pericytes induced by vascular endothelial growth factor-A. J Histochem Cytochem. 2004;52:39–52. [CrossRef] [PubMed]
HofmanP, van BlijswijkBC, GaillardPJ, VrensenGF, SchlingemannRO. Endothelial cell hypertrophy induced by vascular endothelial growth factor in the retina: new insights into the pathogenesis of capillary nonperfusion. Arch Ophthalmol. 2001;119:861–866. [CrossRef] [PubMed]
McLeodDS, D’AnnaSA, LuttyGA. Clinical and histopathologic features of canine oxygen-induced proliferative retinopathy. Invest Ophthalmol Vis Sci. 1998;39:1918–1932. [PubMed]
OrawiecB, NiwaldA, GralekM, HutzW. Evaluation of blood flow in the arteries of the eye in premature neonates and in children with retinopathy of prematurity. Proceedings of World ROP Meeting: East Meets West. 2006;Vilnius Lithuania.
Figure 1.
 
Desmin filaments (red) on lectin-labeled vasculature (green): (A, B) Retinal vascular remnants after exposure to hyperoxia for 4 days followed by 3 dRA; (C–F) retinal neovasculature after exposure to hyperoxia for 4 days followed by 3 dRA central to (C, D) and at (E, F) the leading edge of vessel formation. (G–J) Retinal vasculature of control P6 kitten central to (G, H) and at (I, J) the leading edge of vessel formation. (K–N) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 10 dRA central to (K, L) and at (M, N) the leading edge of vessel formation. (O–R) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 27 dRA in the early (O, P) and late (Q, R) stages of regression. Scales listed for (A, C, E, G, I, K, M, O, Q) are the same for companion panels (B, D, F, H, J, L, N, P, R). Scale bars: (A, B, O–R) 50 μm; (C–N) 25 μm.
Figure 1.
 
Desmin filaments (red) on lectin-labeled vasculature (green): (A, B) Retinal vascular remnants after exposure to hyperoxia for 4 days followed by 3 dRA; (C–F) retinal neovasculature after exposure to hyperoxia for 4 days followed by 3 dRA central to (C, D) and at (E, F) the leading edge of vessel formation. (G–J) Retinal vasculature of control P6 kitten central to (G, H) and at (I, J) the leading edge of vessel formation. (K–N) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 10 dRA central to (K, L) and at (M, N) the leading edge of vessel formation. (O–R) Retinal neovasculature after exposure to hyperoxia for 4 days followed by 27 dRA in the early (O, P) and late (Q, R) stages of regression. Scales listed for (A, C, E, G, I, K, M, O, Q) are the same for companion panels (B, D, F, H, J, L, N, P, R). Scale bars: (A, B, O–R) 50 μm; (C–N) 25 μm.
Figure 2.
 
Ultrastructure of preretinal vessels present after exposure to 4 days of hyperoxia followed by 23 days in room air in a kitten model of ROP. (A) Dilated preretinal vessels consisting of endothelial cells (E) with sporadic pericyte (P) ensheathment. (B) Densely packed new vessels in the preretinal membrane showing several plump pericytes (P) displaying organelle-rich cytoplasm and active euchromatic nuclei with large nucleoli. Pericyte process (p). (C) Glial cells (G) were present but were not closely associated with dilated vessels. (D) Fenestrations were apparent in the endothelial cells of some preretinal vessels.
Figure 2.
 
Ultrastructure of preretinal vessels present after exposure to 4 days of hyperoxia followed by 23 days in room air in a kitten model of ROP. (A) Dilated preretinal vessels consisting of endothelial cells (E) with sporadic pericyte (P) ensheathment. (B) Densely packed new vessels in the preretinal membrane showing several plump pericytes (P) displaying organelle-rich cytoplasm and active euchromatic nuclei with large nucleoli. Pericyte process (p). (C) Glial cells (G) were present but were not closely associated with dilated vessels. (D) Fenestrations were apparent in the endothelial cells of some preretinal vessels.
Figure 3.
 
Schematic representation of a portion of the vasculature (green) and the distribution of SMA labeling (red) in this vasculature (A), in the retina of a control kitten retina at P3 and (B) in the retina of a kitten exposed from P3 to 2 days of hyperoxia and then returned to room air for 3 days. Scale bars, 500 μm.
Figure 3.
 
Schematic representation of a portion of the vasculature (green) and the distribution of SMA labeling (red) in this vasculature (A), in the retina of a control kitten retina at P3 and (B) in the retina of a kitten exposed from P3 to 2 days of hyperoxia and then returned to room air for 3 days. Scale bars, 500 μm.
Figure 4.
 
(A–F) Regions of the retinal vasculature double labeled with lectin (green) and SMA immunohistochemistry (red) of a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 3 days. (A) A dense peripheral capillary plexus and (B) a tortuous radial venule and regressed vessel branching from a radial arteriole. (C) A dense capillary plexus, fed by an SMA-labeled arteriole (A) on the left from which several vessels have regressed resulting in vascular stumps (arrows), and drained by a radial venule (V) on the right. Vessel regression in the perivenular region was highly abnormal and indicative of altered tissue oxygenation as a result of irregular vessel hemodynamics. Small white arrowheads: a region of poorly perfused capillary plexus. The vascular tree was indicative of a dense capillary plexus with decreased arteriolar blood supply and venular drainage. (D) Regressing vessels in the interlobular capillary plexus, (E) decreased vascular density, which is evidence of vessel regression, along a radial venule, and (F) a region of decreased capillary density, evidence of vessel regression, with vascular stumps (arrow) and foci of vascular sprouting (arrowheads). (G, H) Lectin-labeled primary arterioles branching from radial arterioles in the retina of (G) a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 29 days and (H) a control P28 kitten. Scale bars: (A, B, D–H) 125 μm.
Figure 4.
 
(A–F) Regions of the retinal vasculature double labeled with lectin (green) and SMA immunohistochemistry (red) of a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 3 days. (A) A dense peripheral capillary plexus and (B) a tortuous radial venule and regressed vessel branching from a radial arteriole. (C) A dense capillary plexus, fed by an SMA-labeled arteriole (A) on the left from which several vessels have regressed resulting in vascular stumps (arrows), and drained by a radial venule (V) on the right. Vessel regression in the perivenular region was highly abnormal and indicative of altered tissue oxygenation as a result of irregular vessel hemodynamics. Small white arrowheads: a region of poorly perfused capillary plexus. The vascular tree was indicative of a dense capillary plexus with decreased arteriolar blood supply and venular drainage. (D) Regressing vessels in the interlobular capillary plexus, (E) decreased vascular density, which is evidence of vessel regression, along a radial venule, and (F) a region of decreased capillary density, evidence of vessel regression, with vascular stumps (arrow) and foci of vascular sprouting (arrowheads). (G, H) Lectin-labeled primary arterioles branching from radial arterioles in the retina of (G) a kitten exposed from P3 to hyperoxia for 2 days and then returned to room air for 29 days and (H) a control P28 kitten. Scale bars: (A, B, D–H) 125 μm.
Figure 5.
 
(A–H) Desmin filaments (red) on vessels (green) of dense capillary monolayer (A, B) and on vessels with vascular stumps and vascular sprouting (C, D) in kitten retina after exposure to hyperoxia for 2 days followed by 3 dRA (modified model of kitten ROP); on vessels of a remodeled capillary plexus in a kitten retina after exposure to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP) (E, F) and on vessels of a remodeling capillary plexus in a control P32 kitten retina (G, H). (I–L) Determination of the DER. Images of a field of view of the vasculature in the peripheral retina of a control P32 kitten (I, J) and a kitten exposed to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP), (K, L) labeled with lectin (green) and anti-desmin antibodies (red). A grid with 100 intersection points was superimposed on each image and intersections with desmin or lectin labeling are shown (white dot). The ratio of desmin-positive intersections to lectin-positive intersections for a given field yields the DER. The DER of the field of view shown in (I, J) is 0.80 and the DER of the field of view shown in (K, L) is 0.44. Scale bars, 50 μm.
Figure 5.
 
(A–H) Desmin filaments (red) on vessels (green) of dense capillary monolayer (A, B) and on vessels with vascular stumps and vascular sprouting (C, D) in kitten retina after exposure to hyperoxia for 2 days followed by 3 dRA (modified model of kitten ROP); on vessels of a remodeled capillary plexus in a kitten retina after exposure to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP) (E, F) and on vessels of a remodeling capillary plexus in a control P32 kitten retina (G, H). (I–L) Determination of the DER. Images of a field of view of the vasculature in the peripheral retina of a control P32 kitten (I, J) and a kitten exposed to hyperoxia for 2 days followed by 29 dRA (modified model of kitten ROP), (K, L) labeled with lectin (green) and anti-desmin antibodies (red). A grid with 100 intersection points was superimposed on each image and intersections with desmin or lectin labeling are shown (white dot). The ratio of desmin-positive intersections to lectin-positive intersections for a given field yields the DER. The DER of the field of view shown in (I, J) is 0.80 and the DER of the field of view shown in (K, L) is 0.44. Scale bars, 50 μm.
Figure 6.
 
Mural cells on radial venules (v) and arterioles (a) in the retinas of control kittens (A–F, V) and in the vaso-obliterative (G–R) and modified (S–U, W, X) kitten models of ROP. (A–D) SMA immunohistochemical labeling (red) of radial arterioles (A–C) and of radial venules (A, B, D) in the central retina of a P6 kitten. Desmin filaments (red) on a lectin-labeled (green) central radial arteriole and radial venule in a P6 retina (E, F). (G–L) SMA immunohistochemical labeling (red) of central radial arterioles and venules in retinas after 4 days’ exposure to hyperoxia followed by 3 (G, H), 7 (I, J), 10 (K), and 27 (L) dRA. (M, N) Desmin filaments (red) on lectin-labeled (green) central radial vessels of retina of kitten exposed to 4 days of hyperoxia followed by 10 dRA. (O–R) SMA immunohistochemical labeling (red) of remnants of radial vessels immediately after 4 days’ exposure to hyperoxia (O, P), and a radial arteriole that survived 4 days of hyperoxia after 7 dRA (Q, R). (S–U) SMA immunohistochemical labeling of central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 dRA, (V) SMA immunohistochemical labeling of a peripheral, less-mature segment of a radial arteriole in the retina of a control P6 kitten, and (W, X) desmin filaments (red) on the central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 days in room air. Scale bars: (A, B, E–H, K–N, W, X) 50 μm; (C, D, I, J, O–V) 25 μm.
Figure 6.
 
Mural cells on radial venules (v) and arterioles (a) in the retinas of control kittens (A–F, V) and in the vaso-obliterative (G–R) and modified (S–U, W, X) kitten models of ROP. (A–D) SMA immunohistochemical labeling (red) of radial arterioles (A–C) and of radial venules (A, B, D) in the central retina of a P6 kitten. Desmin filaments (red) on a lectin-labeled (green) central radial arteriole and radial venule in a P6 retina (E, F). (G–L) SMA immunohistochemical labeling (red) of central radial arterioles and venules in retinas after 4 days’ exposure to hyperoxia followed by 3 (G, H), 7 (I, J), 10 (K), and 27 (L) dRA. (M, N) Desmin filaments (red) on lectin-labeled (green) central radial vessels of retina of kitten exposed to 4 days of hyperoxia followed by 10 dRA. (O–R) SMA immunohistochemical labeling (red) of remnants of radial vessels immediately after 4 days’ exposure to hyperoxia (O, P), and a radial arteriole that survived 4 days of hyperoxia after 7 dRA (Q, R). (S–U) SMA immunohistochemical labeling of central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 dRA, (V) SMA immunohistochemical labeling of a peripheral, less-mature segment of a radial arteriole in the retina of a control P6 kitten, and (W, X) desmin filaments (red) on the central radial vessels in the retina of a kitten exposed to 2 days of hyperoxia followed by 3 days in room air. Scale bars: (A, B, E–H, K–N, W, X) 50 μm; (C, D, I, J, O–V) 25 μm.
Figure 7.
 
(A, B) Semithin and ultrathin sections of a retina from a kitten exposed to 4 days of hyperoxia followed by 23 dRA. (A) A semithin section from the central retina showing a thin-walled radial arteriole (A) and dilated venule (V); (B) from a more peripheral block showing smaller, thin-walled, poorly differentiated vessels (arrows) that were fed by the radial vessels and directly fed into the dense capillary plexus. (C, E) Electron microscopic images of the wall of the arteriole in (A). The endothelium (E) is attenuated, in keeping with the dilated state of the vessel, whereas the smooth muscle cells (SM) show uncharacteristic overlapping processes (arrows), insubstantial microfilament bundles and an organelle-rich cytoplasm, consistent with immaturity. (F, G) Electron microscopic images of the venule shown in (A). The endothelium (E) is extremely attenuated and only partially covered by scanty mural cell processes (arrows). (D, H) Electron microscopic images of one of the smaller feeder arteriolar vessels in (B). The complete investiture of the abluminal surface of the endothelium (E) by mural cell processes is consistent with cells representing smooth muscle rather than pericytes; however, the cells are poorly differentiated with inconspicuous microfilament bundles and an organelle-rich cytoplasm (arrows). In both profiles of this vessel the glial component of the basement membrane is separated from that of the vascular cells with the intervening space filled by flocculent matrix material (Mx).
Figure 7.
 
(A, B) Semithin and ultrathin sections of a retina from a kitten exposed to 4 days of hyperoxia followed by 23 dRA. (A) A semithin section from the central retina showing a thin-walled radial arteriole (A) and dilated venule (V); (B) from a more peripheral block showing smaller, thin-walled, poorly differentiated vessels (arrows) that were fed by the radial vessels and directly fed into the dense capillary plexus. (C, E) Electron microscopic images of the wall of the arteriole in (A). The endothelium (E) is attenuated, in keeping with the dilated state of the vessel, whereas the smooth muscle cells (SM) show uncharacteristic overlapping processes (arrows), insubstantial microfilament bundles and an organelle-rich cytoplasm, consistent with immaturity. (F, G) Electron microscopic images of the venule shown in (A). The endothelium (E) is extremely attenuated and only partially covered by scanty mural cell processes (arrows). (D, H) Electron microscopic images of one of the smaller feeder arteriolar vessels in (B). The complete investiture of the abluminal surface of the endothelium (E) by mural cell processes is consistent with cells representing smooth muscle rather than pericytes; however, the cells are poorly differentiated with inconspicuous microfilament bundles and an organelle-rich cytoplasm (arrows). In both profiles of this vessel the glial component of the basement membrane is separated from that of the vascular cells with the intervening space filled by flocculent matrix material (Mx).
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×