Increased VEGF‐A promotes multiple distinct aging diseases of the eye through shared pathomechanisms

Abstract While increased VEGF‐A has been associated with neovascular age‐related macular degeneration (AMD), it is not known whether VEGF‐A may also promote other age‐related eye diseases. Here, we show that an increase in VEGF‐A is sufficient to cause multiple distinct common aging diseases of the eye, including cataracts and both neovascular and non‐exudative AMD‐like pathologies. In the lens, increased VEGF‐A induces age‐related opacifications that are associated with ERK hyperactivation, increased oxidative damage, and higher expression of the NLRP3 inflammasome effector cytokine IL‐1β. Similarly, increased VEGF‐A induces oxidative stress and IL‐1β expression also in the retinal pigment epithelium (RPE). Targeting NLRP3 inflammasome components or Il1r1 strongly inhibited not only VEGF‐A‐induced cataract formation, but also both neovascular and non‐exudative AMD‐like pathologies. Moreover, increased VEGF‐A expression specifically in the RPE was sufficient to cause choroidal neovascularization (CNV) as in neovascular AMD, which could be inhibited by RPE‐specific inactivation of Flk1, while Tlr2 inactivation strongly reduced CNV. These findings suggest a shared pathogenic role of VEGF‐A‐induced and NLRP3 inflammasome‐mediated IL‐1β activation for multiple distinct ocular aging diseases.

Thank you for the submission of your manuscript to EMBO Molecular Medicine. We have now heard back from the three Reviewers whom we asked to evaluate your manuscript.
As you will see, the issues raised are few but fundamental. Although I will not dwell into much detail, I would like to highlight the main points.
Firstly, while Reviewer 2 is more positive, Reviewers 1 and 3 are more reserved and with similar concerns, albeit with different emphasis and perspective. I identify two fundamental issues that require your action.
One is the perceived lack of pathophysiological relevance of your model and thus your findings. The second main concern is that the manuscript requires much more care and detail in overall presentation and description of observations. I would also like to mention that Reviewer 2 also raises an important issue, namely the justification for the reference to corneal neovascularization.
In aggregate the Reviewers also raise various other items of concern that require your attention.
In conclusion, while publication of the paper cannot be considered at this stage, given the potential interest of your findings and after internal discussion, we have decided to give you the opportunity degeneration have been hypothesized extensively. What separates this study from others is the utilization of the VEGF overexpressing mouse model, which they, in part, published on in 2013. The mouse model has the potential of being of intested to researchers within both the lens and retina field. However prior to that greater characterization of the pathology in the mice is strongly recommended.
Additional comments are below: Page 10: Several other groups prior to the Marneros 2013 paper have show using chimeric mice that infiltrating macrophages contribute to development and severity of neovascular lesions. Please appropriately reference those other groups rather than just list this one paper from your group.
Are the lesions indeed choroidal in origin? The image in figure 7 panel e for example indicates the involvement of the retina and could very well be retinal angiomatous proliferation (RAP) rather than CNV. This can only be differentiated with fluorescein angiography and examination of serial sections in which the break though Bruch's can be seen. Were serial sections evaluated?
Very interesting dry AMD-like phenotype has been shown for the VEGFA-hyper mice. Please provide a table with frequency, length, severity, etc of the deposits, RPE and PR degeneration. Clearly there are pigmentary changes evident too, which in all likelihood may indicate changes in autofluorescence. This should be commented on. Page 7:provide ref for the following sentence '(1) oxidative stress can IL-1B can induce IL-1B activation in the lens epithelial cells in vitro".
Many of the data panels do not have an associated n. Please include n for all data panels in each figure. For example Figure   Spell out all acronyms at first use. Page 3, first sentence of third paragraph should be corrected (remove 'With"....).

Referee #2 (Remarks):
This is a comprehensive study addressing important aging diseases of the eye. Several recommendations to improve the work: 1) Font sizes of some of the bar figures are very small and hard to read.
2) Figure 1d and 1e: provide age (young vs old) images for comparison.
3) Corneal neovascularization is not a common senile disease like cataract and AMD, the author needs to explain why it was investigated in this study and whether it was affected by the inhibition of NLRP3 inflammasome components.
4) It is known that anti-VEGF treatment has side effects and VEGF-A is required for normal vasculature maintenance. How about the inhibition of NLRP3 inflammasome components? Does it affect normal vasculatures? 5) Contradictory statements on Flk1 function: last sentence on page 8 vs. first two sentences on page 9.

General comments
The author has performed an extensive study of the effects of transgenic hyper-expression of VEGFA on lens and retinal cells, finding pathological changes in both circumstances. He has then crossed these mice with a variety of gene deficient mice targeting genes which are upstream or downstream of VEGFA signalling. He has then identified some predictable and unpredictable effects based on current knowledge of VEGFA signalling.
While many of these effects eg requirement for TLR2 for pathological effects in the CNV induction are of interest, whether they reflect pathophysiological processes is not clear. Additional data demonstrating similar effects in non-transgenic models using, for example, specific inhibitors / blocking antibodies or would lend support to the physiological relevance of this work.

Specific comments
The numbers of generations used for crossing the various strains should be provided.
The fixation procedures are not clear. Are they different for the lens and the posterior segment of the eye?
The details of the immunostaining are not provided.
Examination of the experimental material was performed in a blinded manner. How precisely was this i.e. who prepared the samples and who was blinded?
Details of the primers used should be provided in the supplementary materials.
Cataract grading was done: what grading system was used (eg what is meant by a moderate cataract?) and why was grading necessary (graded lenses are not discussed in the paper)?
The histological images in Figure 1  Please find below a detailed point-by-point response to the reviewers' comments: Reviewer's comments ***** Referee #1 (Comments on Novelty/Model System): Very interesting dry AMD-like phenotype has been shown for the VEGFA-hyper mice. Please provide a table with frequency, length, severity, etc. of the deposits, RPE and PR degeneration. Clearly there are pigmentary changes evident too, which in all likelihood may indicate changes in autofluorescence. This should be commented on.

Response: We included images that show autofluorescence of observed sub-RPE deposits in VEGF-
A hyper mice (Fig 5A-B). Moreover, we performed autofluorescence fundus imaging in vivo in aged VEGF-A hyper mice, which showed the presence of strongly autofluorescent deposits in the eyes of aged VEGF-A hyper mice. These changes were associated with degenerative changes of the RPE with pigment abnormalities that were observed by color fundus imaging (Fig 11B-C) as well as by light and electron microscopy ( Fig S2 and S3). Importantly, targeting the inflammasome (by genetic inactivation of CASP1/CASP11) strongly inhibited these age-dependent non-exudative AMD-like pathologies as well as fundus abnormalities.

RPE and photoreceptor degeneration occurs in all VEGF-A hyper mice in a progressive age-dependent manner. While young VEGF-A hyper mice show already degenerative changes of the RPE and retina, VEGF-A hyper mice > 12 months of age show prominent basal laminar-like sub-RPE deposits and degeneration of the RPE and photoreceptors in all mice examined (100%)
, while none of these findings were found in littermate control mice. Thus, RPE and photoreceptor degeneration occurs in aged VEGF-A hyper mice with 100% frequency. We included this information in the manuscript. We also added high-resolution light and electron microscopy images of these RPE and photoreceptor pathologies in the revised manuscript (Fig S2 and S3).
Page 7:provide ref for the following sentence '(1) oxidative stress can IL-1B can induce IL-1B activation in the lens epithelial cells in vitro".

Response:
We added this reference.

Referee #1 (Remarks):
The authors characterize the lens and retinal phenotype of mice expressing a '2-fold' increase in VEGFA. Appropriately they evaluate young and old mice. They present data suggesting that the common denominator for the associate age-related pathology is mediated through NLRP3 mechanisms supporting the hypothesis of targeting both VEGFA and the NLRP3 inflammasome for therapy. These concepts are not necessary novel and at least in the context of agerelated macular degeneration have been hypothesized extensively. What separates this study from others is the utilization of the VEGF overexpressing mouse model, which they, in part, published on in 2013. The mouse model has the potential of being of interest to researchers within both the lens and retina field. However prior to that greater characterization of the pathology in the mice is strongly recommended. Additional comments are below: Page 10: Several other groups prior to the Marneros 2013 paper have shown using chimeric mice that infiltrating macrophages contribute to development and severity of neovascular lesions. Please appropriately reference those other groups rather than just list this one paper from your group.

Response:
We have included these additional references here as well.
Are the lesions indeed choroidal in origin? The image in figure 7 panel e for example indicates the involvement of the retina and could very well be retinal angiomatous proliferation (RAP) rather than CNV. This can only be differentiated with fluorescein angiography and examination of serial sections in which the break though Bruch's can be seen. Were serial sections evaluated?  (Fig 5). This observation is also consistent with OCT imaging of CNV lesions in vivo in these mice (Fig 11A) Fig 6A). Thus, our experiments clearly demonstrate that the neovascular lesions in these genetic mouse models are of choroidal origin.
Many of the data panels do not have an associated n. Please include n for all data panels in each figure. For example Figure 3 a and b. Figure legend 4 is missing a reference to panels b and c. Figure  7: what is the age of onset of AMD-like pathologies.

Response:
The requested changes were made accordingly. All absolute numbers are shown in the figure legends (for cataracts separate tables of absolute and relative numbers per group are shown in Fig S1). Figure references for Fig 4 b and

in VEGF-A hyper mice. Should inhibition of the NLRP3 inflammasome indeed have a major effect on preventing VEGF-A-induced AMD-like pathologies, then we would expect that these pathologies would be inhibited even at an advanced age when they are fully developed in VEGF-A hyper mice. Indeed, we observed that even in 2-year old VEGF-A hyper mice genetic inactivation of NLRP3 inflammasome components could prevent the manifestation of these AMD-like pathologies. We observed that the rescue of ocular pathologies in aged VEGF-A hyper mice due to targeting the NLRP3 inflammasome also correlated with maintained visual function in these mice, as assed by ERGs in 2-
year old experimental mouse groups (Fig 10). Moreover, fundus abnormalities and autofluorescent deposits were strongly inhibited by targeting the inflammasome in VEGF-A hyper mice (Fig 11).
Minor comments: Spell out all acronyms at first use. Page 3, first sentence of third paragraph should be corrected (remove 'With"....).

Response: We have made changes accordingly.
Referee #2 (Remarks): This is a comprehensive study addressing important aging diseases of the eye. Several recommendations to improve the work: 1)Font sizes of some of the bar figures are very small and hard to read.

Response: Figure sizes and bar figures were increased. Figures have been changed to improve the presentation of the findings.
2) Figure 1d and 1e: provide age (young vs old) images for comparison. (Fig 1).

These images show that the expression pattern of VEGF-A is not affected by age. For example, the RPE, retinal cells of the inner nuclear layer (INL), and lenticular cells express VEGF-A in young and aged mice.
3) Corneal neovascularization is not a common senile disease like cataract and AMD, the author needs to explain why it was investigated in this study and whether it was affected by the inhibition of NLRP3 inflammasome components. Fig 1).

Notably, we only observed corneal neovascularization in aged VEGF-A hyper mice that also had developed cataracts. Thus, we speculate that corneal neovascularization is an indirect consequence of increased VEGF-A levels in the eye and associated with VEGF-A-induced cataract formation. However, not all VEGF-A hyper mice that formed cataracts also had corneal neovascularization. The variability of the occurrence of corneal neovascularization in aged VEGF-A hyper mice limits statistical quantifications to conclusively show in a meaningful manner whether corneal neovascularization was dependent on the NLRP3 inflammasome.
We have made changes in the manuscript accordingly to explain these finding in the proper context. Should the reviewers feel that the data on corneal neovascularization do not add to the current manuscript, then we would be willing to remove these data.

Response: We have made changes accordingly.
Referee #3 (Remarks): General comments The author has performed an extensive study of the effects of transgenic hyper-expression of VEGFA on lens and retinal cells, finding pathological changes in both circumstances. He has then crossed these mice with a variety of gene deficient mice targeting genes, which are upstream or downstream of VEGFA signaling. He has then identified some predictable and unpredictable effects based on current knowledge of VEGFA signaling. While many of these effects eg requirement for TLR2 for pathological effects in the CNV induction are of interest, whether they reflect pathophysiological processes is not clear. Additional data demonstrating similar effects in non-transgenic models using, for example, specific inhibitors / blocking antibodies or would lend support to the physiological relevance of this work.  Fig 4 and S1. Cataract grading allows for a more comprehensive evaluation of differences in cataract severity between mouse groups.

Response
The histological images in Figure 1, 2, 4 and 7 are unclear; it is difficult to see the correlation between the Western blot and the histogram in Fig 2. Similar aging changes have been reported in WT as shown for hyper VEGFA mice in Figure 7. Control data is necessary to show here.

Response:
We have improved the histological images and added additional images (also in Supplemental Figures). We have also provided additional images of age-matched control littermate mice that show that the eye abnormalities in VEGF-A hyper mice are not observed in control mice with normal ocular VEGF-A levels (Fig 9A and Fig S2). We have assessed a large number of control littermate mice and we did not observe any CNV lesions in these mice (we performed choroidal flat mounts or histological sections on >200 control mice). With progressive age control mice show mild age-appropriate changes of the RPE/choroid interface, but no AMD-like pathologies were observed in these mice (Fig 9A and Fig S2).
In summary, we have performed an extensive revision of our original manuscript and included substantial additional experimental data that strengthen the clinicopathologic correlation of the findings in our mouse model with those observed in human age-related ocular pathologies. Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed reports from the referees that were asked to re-assess it. As you will see the reviewers are now globally supportive and I am pleased to inform you that we will be able to accept your manuscript pending the following final amendments: 1) Please address the remaining set of comments from referees 1 and 3 in writing in the main text and in a rebuttal letter Please submit your revised manuscript within two weeks. I look forward to seeing a revised form of your manuscript as soon as possible.
A couple of points: 1. Figure 1, shows b-gal staining in the RPE of young mice ( Figure 1B), but in the older mice 'strong VEGF staining in the RPE' is actually not seen ( Figure 1G), so the statement that strong VEGF staining persists with age is not accurate. Also see, Page 11, first line. VEGFA expression is actually quite high in the inner retina rather than the RPE. 2. The increased VEGF levels in the literature is reported for aqueous samples and the text throughout the manuscript should be corrected to reflect this. Currently it is cited as VEGF levels are high in human tissue samples, without clearly stating that the 2 fold increase was actually measured in the aqueous.
3. Figure 2, shows VEGF levels from the lens, what about the RPE of these mice? Is there an increase in VEGF levels with age? Please show parallel data.
Referee #2 (Remarks): Previous questions have been addressed. Since corneal neovascularization shows variability and is not well studied as other phenotypes, it is recommended to remove the data from this manuscript.
Referee #3 (Remarks): I have re-checked the Figures and legend to Figure 8F. I think the legend could be more explanatory but the data are fine and impressive.. I am pleased to see these data included in the paper. I have not reservations now about the work.
1 December 21 th , 2015 Re: EMM-2015-05613 Dear Editor, We have now fully addressed all of the reviewers comments. Please see below a point-bypoint response.
Response to Editor:

3) Figures 3A and 2B:
Some irregularities have been seen in these 2 instances, in order to assess the validity of the data, we now encourage the publication of source data, particularly for electrophoretic gels and blots, with the aim of making primary data more accessible and transparent to the reader. Would you be willing to provide a PDF file per figure that contains the original, uncropped and unprocessed scans of all or key gels used in the figure? The PDF files should be labeled with the appropriate figure/panel number (1 file/figure), and should have molecular weight markers; further annotation may be useful but is not essential. The PDF files will be published online with the article as supplementary "Source Data" files. If you have any questions regarding this just contact me.

Response:
We have enclosed the uncropped full Western blots for the ERK and ERK-P experiments that are depicted in Figure 3a. Both Western blots give a clean and clear result with little background bands. The observed bands correspond to the expected bands for ERK and ERK-P (42 and 44 kDa). The RT-PCR results (Figure 2b) also correspond to the expected bands.