**4. Discussion**

Sorsby's fundus dystrophy (SFD) is a rare macular dystrophy characterized by vision loss due to persistent choroidal neovascularization [1–6]. SFD is an autosomal dominant, fully penetrant degenerative disease of the macula and is notable for its similarity in histopathological features to AMD [3–6]. The majority of SFD patients develop CNV, as well as confluent, 20–30 μm thick, amorphous deposits between the basement membrane of the RPE and Bruch's membrane. TIMP3 and/or its downstream substrates have been postulated to have a role in the pathogenesis of both SFD and AMD, because accumulation of TIMP3 has been observed in subretinal deposits in SFD [42] as

well as in AMD drusen [43–45]. In this study, we show that hyaluronan accumulates around the RPE in AMD as well as in CNV lesions of mice expressing S179C-TIMP3.

One interesting observation from our studies was the increase in HA in the plasma of patients with AMD as well as in mice lacking TIMP3 or carrying the S179C-TIMP3 mutation. Although we observe significant di fferences in the plasma levels of HA in patients with advanced AMD (GA and CNV), the number of patients analyzed (n = 49 controls) and n = 59 (AMD) is not su fficient to determine if this is of prognostic value. In addition, we did not have access to samples from patients at di fferent degrees of severity to be able to draw conclusions with this or the rate of progression of the disease. Future studies are warranted to address this question. While patients with SFD generally demonstrate disease localized to the retina, our results may be a consequence of ubiquitous expression of TIMP3 in a variety of tissues in the body [46] which could potentially explain the systemic increase in HA. Whether there is accumulation of HA in other tissues in S179C TIMP3 mice has not been evaluated.

Our studies with RPE cells sugges<sup>t</sup> that the increase in HA in these cells is likely a consequence of increased FGF. A recent study [47] suggests that TIMP proteins can control FGF-2 bioavailability in skeletal tissue and the same might be true of multiple tissues leading to systemic increase in HA in the plasma. The exact mechanism by which TIMP3 regulates FGF bioavailability in the RPE is currently unknown, but it is highly likely that the TIMP-metalloproteinase axis likely has a key role. The extracellular matrix (ECM) serves as a high capacity reservoir for FGF-2 and early studies have demonstrated that matrix metalloproteases (MMPs) have the ability to mobilize FGF-2 to a soluble phase that results in receptor activation [48]. Additional studies identifying the molecular mechanisms by which TIMP3 regulates FGF-2 bioavailability will provide insight into the pathophysiology of the disease.

FGF-2 is su fficient to increase HA accumulation and distribution in the RPE, and blocking FGF signaling in S179C-TIMP3 RPE cells brings HA levels back to normal. The mechanism by which FGF-2 increases HA accumulation is not understood. HA is endogenously synthesized by a family of membrane-integrated glycosyltransferases, called hyaluronan synthases (HAS 1-3) and is exported directly into the ECM [49,50]. Hyaluronidases (HYAL1-2) are a class of enzymes that degrade HA [51]. A balance between HA synthesizing and degrading activity keeps HA at physiological levels. In order to determine the mechanism of accumulation of HA in the RPE in SFD, we performed quantitative PCR analysis of HAS1-3 and HYAL1-2 from RPE isolated from S179C-TIMP3, TIMP3-KO mice and wildtype littermate controls. Interestingly, we observed no changes in gene expression of any of these enzymes in the mutant mice (Supplementary Figure S1). Therefore, at least in the RPE in SFD mice, the di fferences in HA content are not due to increased expression of the synthases nor decreased expression of canonical degradation enzymes. However, there are other possibilities that need to be explored in the future. HA production could be modulated by decreasing enzyme recycling from endosomes back to the cell surface as seen in keratocytes [52]. Additionally, there is a possibility that other non-canonical hyaluronidases such as KIAA1199 [53] and Tmem2 [54] could be involved. Alternatively, as previously reported degradation might be prevented by increased binding proteins on HA and leading to net accumulation [55].

HA has been shown to exhibit a diverse array of biological functions including a role in the response to tissue damage and inflammation [56]. Our studies demonstrating accumulation of HA in laser-induced CNV lesions corroborates previous studies [27]. This study also reported an increase in *CD44* and *HAS2* mRNA following laser injury [27]. It is possible that the increased accumulation of HA in laser-induced CNV lesions in S179C-TIMP3 mice might result from similar increases in mRNA transcription.

Chronic low-grade inflammation has been suggested to contribute to age-related macular degeneration [57]. In the laser-induced mouse model of CNV, inflammatory processes have been shown to play a role in the development and regression of the lesions. A number of reports link HA remodeling to the modulation of neuroinflammation with low-molecular weight HA being pro-inflammatory and high molecular weight HA being anti-inflammatory [58–60]. While we see increased deposition of HA in and around the RPE, we have not determined its physical properties such as size and molecular weight distribution in the tissue.

The receptor engagemen<sup>t</sup> of HA in the retina or it's downstream signaling under physiological or pathological conditions has not ye<sup>t</sup> been identified and will be important as we determine its exact role in the pathology of macular degenerative disease. In our study we demonstrate that primary porcine RPE cells deposited HA predominantly on the apical surface under physiological conditions similar to what had been previously reported for human RPE cells [61]. Our data revealed that FGF-2 induced HA accumulation apically as well as between cells and on the basal surface, suggesting that in addition to increased total HA content, the distribution of HA may be important for disease pathogenesis and warrants further investigation. We have recently reported that the secretion of FGF-2 by RPE cells expressing S179C-TIMP3 led to increased angiogenesis [28]. Whether HA is modified in the endothelial glycocalyx as a consequence of FGF-2 has not been studied and might provide further insight into the pathogenesis of CNV in AMD and SFD leading to the identification of novel therapeutic approaches.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4409/9/3/608/s1, Figure S1: RNA was isolated from mouse RPE using the Simultaneous RPE cell Isolation and RNA Stabilization method (SRIRS method) using the RNA Plus Mini Kit (Qiagen). Quantitative PCR was performed following reverse transcription using TaqMan probes for the mouse genes Has1 (A), Has2 (B), Has3 (No signal), Hyal1 (C), Hyal2 (D), and 18S ribosomal RNA (rRNA) (Applied Biosystems). 18S rRNA was used as endogenous control for each gene tested. mRNA expression was calculated using 2-ΔΔCt method and shown relative to expression in wildtype littermate mice.

**Author Contributions:** Conceptualization, A.W., O.W., V.H. and B.A.-A.; data curation, A.W.; formal analysis, B.A.-A.; funding acquisition, B.A.-A.; methodology, A.W., D.H., A.C., M.A., L.B., J.H.Q., R.S., J.B. and L.K.; project administration, B.A.-A.; resources, V.L.B., O.W. and H.S.; software, R.S.; supervision, B.A.-A.; validation, A.W.; writing—original draft, A.W.; writing—review and editing, A.W., H.S., V.H. and B.A.-A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported in part by US National Institute of Health EY027083 (BA-A), EY026181 (BA-A), P30EY025585(BA-A), T32EY024236 (AW), EY022768 (JHQ), EY027750 (VLB), Research to Prevent Blindness (RPB) Challenge Grant and RPB Lew Wasserman award to BA-A, Cleveland Eye Bank Foundation Grant and funds from Cleveland Clinic Foundation.

**Acknowledgments:** The authors thank the retina specialists at Cole Eye Institute from whose clinics patients with or without AMD were recruited. The authors are grateful for the generous gift of TIMP-3 null mice from Rama Khokha at the Princess Margaret Cancer Centre-Ontario Cancer Institute, Toronto, Canada and to Emma Lessieur who served as the research coordinator at the time the blood samples were collected from patients The authors also acknowledge the support of Foundation Fighting Blindness in setting up the Eye Tissue Repository. We wish to extend a sincere apology to colleagues whose work was not cited due to space limitations.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
