Local synthesis of hepcidin in the anterior segment of the eye: A novel observation with physiological and pathological implications

Abstract

PURPOSE:

The avascular cornea, trabecular meshwork (TM), and lens obtain iron, an essential biometal, from the aqueous humor (AH). The mechanism by which this exchange is regulated, however, is unclear. Recently we reported that non-pigmented ciliary epithelial cells express ferroportin (Fpn) (Ashok, 2018), an iron export protein modulated by hepcidin, the master regulator of iron homeostasis secreted mainly by the liver. Here, we explored whether ciliary epithelial and other cells in the anterior segment synthesize hepcidin, suggesting local regulation of iron exchange at this site.

METHODS:

Human and bovine eyes were dissected to isolate the ciliary body (CB), corneal endothelial (CE), TM, lens epithelial (LE), and outer epithelial cell layer of the iris. Total mRNA and protein lysates were processed to evaluate the synthesis and expression of hepcidin, the iron regulatory peptide hormone, ferroportin (Fpn), the only known iron export protein, ceruloplasmin (Cp), a ferroxidase necessary for iron export, transferrin receptor (TfR), a major iron uptake protein, and ferritin, a major iron storage protein. A combination of techniques including reverse transcription polymerase chain reaction (RT-PCR) of total mRNA, Western blotting of protein lysates, and immunofluorescence of fixed tissue sections were used to accomplish these goals.

RESULTS:

RT-PCR of isolated tissue samples revealed hepcidin-specific mRNA in the CB, TM, CE, and LE of the bovine eye. Western blotting of protein lysates from these tissues showed reactivity for hepcidin, Fpn, ferritin, and TfR. Western blotting and immunohistochemistry of similar tissues isolated from cadaveric human eyes showed expression of hepcidin, Fpn, and Cp in these samples. Notably, Fpn and Cp were expressed on the basolateral membrane of non-pigmented ciliary epithelial cells, facing the AH.

CONCLUSIONS:

Synthesis and expression of hepcidin and Fpn in the ciliary epithelium suggests local regulation of iron transport from choroidal plexus in the ciliary body to the AH across the blood-aqueous barrier. Expression of hepcidin and Fpn in CE, TM, and LE cells indicates additional regulation of iron exchange between the AH and cornea, TM, and lens, suggesting autonomous regulation of iron homeostasis in the anterior segment. Physiological and pathological implications of these observations are discussed.

1. Introduction

Hepcidin, a peptide hormone secreted mainly by the liver, maintains serum iron within a narrow range by downregulating ferroportin (Fpn), the only known iron export protein. Increase in iron saturation of serum transferrin (Tf) beyond a certain range upregulates hepcidin, which binds and degrades Fpn. This blocks uptake of iron from the intestine and release from iron stores in macrophages, hepatocytes, and other cells, reducing serum iron. The opposite scenario takes effect when the iron saturation of Tf falls below a certain range. Since serum Tf is the principal source of iron for most cells, liver hepcidin is considered the master regulator of systemic iron homeostasis (Ganz, 2013; Ganz and Nemeth, 2012; Gozzelino and Arosio, 2016; Katsarou and Pantopoulos, 2018; Sangkhae and Nemeth, 2017).

The eye is protected from systemic circulation by the outer and inner blood-retinal and blood-aqueous barriers. Liver hepcidin, although a small peptide of 24 amino acids, is unlikely to cross these barriers that regulate exchange of iron between systemic blood and the retina. Local synthesis of hepcidin by the neuroretina has been reported (Gnana-Prakasam et al., 2008), and is believed to regulate retinal iron homeostasis through cross-talk with liver hepcidin (Baumann et al., 2019). Most of our understanding of regulation of iron in the neuroretina is based on the membrane topology of Fpn. Thus, expression of Fpn on the basolateral (BL) membrane of capillary endothelial cells facing the neuroretina suggests transport of iron to the retina across the inner blood-retinal barrier, and transport out of the retina by retinal pigment epithelial (RPE) cells that form the outer blood-retinal barrier and express Fpn on the BL surface facing choroidal sinuses. Within the neuroretina, iron circulates in conjugation with transferrin (Tf-iron), and is taken up by various cells by the conventional Tf-receptor (TfR) pathway (Garcia-Castineiras, 2010; Loh et al., 2009). It is pertinent to mention here that exchange of iron across biological membranes requires ferrireductase proteins such as the prion protein and others, and ferroxidases ceruloplasmin (Cp) and/or hephaestin (Heph). The former because serum iron exists in the relatively stable ferric form in conjugation with Tf, and requires reduction to the ferrous form for transport through Fpn. Since ferrous iron is highly reactive, it is oxidized immediately to the ferric form by Cp or Heph for conjugation with Tf (Andrews and Schmidt, 2007; Singh et al., 2013; Singh, 2014). Expression of these proteins has been demonstrated in RPE and capillary endothelial cells, and the aqueous humor (AH) and vitreous humor (VH) contain soluble Cp, Heph, and significant amounts of apo-Tf (Garcia-Castineiras, 2010).

Transport of Tf-iron to AH is believed to occur along a concentration gradient from the retina, or along the lens equatorial border to lens epithelial cells for export through Fpn on their BL membrane. Released iron is oxidized by Cp and conjugates with Tf in the AH for exchange with structures in the anterior segment (Garcia-Castineiras, 2010). However, recent identification of Fpn on non-pigmented ciliary epithelial cells, the main source of AH, supports an independent source of iron across the blood-aqueous barrier as well (Ashok et al., 2018a). Moreover, the BL orientation of Fpn facing the AH suggests regulation by hepcidin in the AH, not liver hepcidin in the circulation. Interestingly, significant levels of hepcidin have been detected in the AH, though its source is unclear (Ghanem et al., 2014; Sorkhabi et al., 2010). While transport of retinal hepcidin (Gnana-Prakasam et al., 2008) from the VH to AH can be envisioned, this would preclude appropriate response of hepcidin to iron concentration in the local microenvironment, a biologically unfavorable scenario.

Here, we explored whether specific cells in the anterior segment of human and bovine eyes synthesize hepcidin locally, thus allowing stringent regulation of iron in response to stimuli in the local microenvironment. We report that hepcidin is indeed synthesized by the ciliary body, corneal endothelium, lens epithelium, trabecular meshwork, and the iris, and is upregulated in corneal endothelial cells by exposure to bacterial lipopolysaccharide (LPS), implicating iron in inflammatory conditions of the anterior segment.

2. Methods

2.1. Human and bovine samples

Three pairs of human eye globes were acquired from Lions Gift of Sight (formerly the Minnesota Lions Eye Bank) in accordance with the tenets of the Declaration of Helsinki for the use of human tissue for research. Bovine eyes were collected from a local abattoir within two hours of sacrifice. All eyes used in the current study were ostensibly healthy, without known abnormalities (Table 1).

Table 1.

List of biological samples

Human
PCoD*	Age (y)	Gender	PMI (h)
Eye # 1; Acute cardiac event	65	M	7.5
Eye # 2; Septic shock, Pneumonia	41	F	13
Eye # 3; Hodgkin’s lymphoma	29	M	9
Bovine
Mixed breed	~4	F	2

^*PMI (h)- Postmortem interval (from death till time of Enucleation) / PCoD- Primary cause of death

2.1.1. Sample preparation

The samples were either fixed in buffered formalin (1/10) for immunohistochemistry, or dissected to isolate the desired tissues. For isolating TM, iris, and CB, a sharp nick was made in the anterior cup, and the tissue was pulled out carefully using surgical grade forceps. Corneal endothelial layer was peeled using surgical angled forceps. Lens epithelial cells were scraped from the anterior surface with a sharp scalpel. Retinal layer was peeled off from the posterior cup using fine surgical grade forceps. Samples were either processed for RT-PCR or lysed and processed for Western blotting as described in sections 2.4 and 2.5 respectively.

2.2. Reagents and Antibodies

Hoechst 33342 (H3570) was from ThermoFisher Scientific, USA. Hepcidin-25 human trifluoroacetate peptide (4040671) was from BACHEM, USA. Fluoromount-G® (0100–01) was from SouthernBiotech, USA. Lipopolysaccharides (LPS) (L2630) was from Sigma Aldrich, USA.

Antibodies specific to the following proteins were used for this study and are listed in Table 2.

Table 2.

Antibodies used in this study

Primary Antibodies	Cat. No.	Host	Company	Dilution
Hepcidin antimicrobial peptide	NBP1-59337	Rb	Novus biologicals, USA	WB- 7/1000
Ferroportin (Fpn)	NBP1–21502	Rb	Novus biologicals, USA	WB-1/250
				IHC-1/100
Ceruloplasmin (Cp)	ab85237	Rb	Abcam, USA	IHC-1/100
Transferrin receptor (TfR)	13–6800	m	Invitrogen, USA	WB-1/1000
Ferritin	F5012	Rb	Sigma Aldrich, USA	WB-1/1000
β-actin	MAB1501	m	Millipore, USA	WB-1/5000
Gapdh	GT239	m	GeneTex	WB-1/2000
Rabbit IgG, polyclonal - Isotype Control	ab37415		Abcam, USA	IHC- 1/100
Secondary Antibodies
HRP-conjugated secondary anti-mouse	NA931V	Sh	GE Healthcare, USA	WB-1/10000
HRP-conjugated secondary anti-rabbit	NA934V	Dn	GE Healthcare, USA	WB-1/10000
Anti-Rabbit IgG (H+L) cross-adsorbed secondary antibody, Alexa Fluor 546	A11071	g	Invitrogen, USA	IHC- 1/1000

Rb- rabbit, m- mouse, Sh- sheep, Dn- donkey, g-goat; WB- Western blotting, IHC- immunohistochemistry

2.3. Ex vivo culture model for anterior segment of bovine eye

Corneal rims from anterior segment of bovine eyes were cultured as previously described (Hu et al., 2013; Okumura et al., 2016). In brief, freshly harvested bovine eyes were dissected and all tissues except the cornea were removed. Corneal rims were cultured with corneal endothelial side up in DMEM supplemented with 1% FBS and 1% penicillin/streptomycin at 37°C in a humidified atmosphere with 5% CO₂. After overnight equilibration, samples were exposed to 1 μg/ml of lipopolysaccharide (LPS) or solvent (control) for 3 h, and the corneal endothelial layer was peeled and the lysates were probed for hepcidin and β-actin (loading control) using Western blotting (section 2.5).

2.4. RT-PCR

RNA was extracted from CB, iris, TM, CE, LE and retina of bovine eye globes using the RNAqueous®-Micro Kit (AM1931, Invitrogen, USA) according to the manufacturer’s instructions. Samples were lysed, homogenized, and vortexed for 1 min to shear genomic DNA before loading onto mini columns, and eluted in 20 μl of buffer. The sample was treated with DNAse to remove traces of genomic DNA, and the RNA was converted to cDNA using SuperScript® IV Reverse Transcriptase Kit (18090200, Invitrogen, USA). Following this, RT-PCR was carried out using Invitrogen™ Platinum™ Taq DNA polymerase (10966–018, Invitrogen, USA) using primers specific for bovine hepcidin. β-actin was amplified in parallel for semi-quantitative analysis. The primers used were obtained from Integrated DNA technologies, USA and their sequence are provided in Table 3. The RT-PCR hepcidin band was imaged using Bio-Rad Gel Doc and the intensity was quantified with commercial software UN-SCAN-IT gels (Version 6.1) software (Silk Scientific, USA) and analyzed graphically using GraphPad Prism (Version 5.0) software (GraphPad Software Inc., USA) and provided as relative intensity following normalization with β-actin.

Table 3.

Primers for bovine hepcidin (HAMP) and actin:

bov-HAMP-F	5’-TCCTTGTCCTGCTCAGCCTG-3’
bov-HAMP-R	5’-CAGCAGAAGATGCAGATGGGA-3’
bov-β-ACTIN-F	5’-CTTCCTGGGCATGGAATCCT-3’
bov-β-ACTIN-R	5’-TTGATCTTCATTGTGCTGGGTG-3’

2.5. SDS-PAGE and Western blotting

Protein lysates prepared from different tissues were fractionated by SDS-PAGE and analyzed by Western blotting as described (Ashok et al., 2018b). In short, cells scraped from different tissues were lysed in RIPA lysis buffer (50 mM Tris-HCl pH7.4, 100 mM NaCl, 1% NP-40, 0.5% deoxycholate), boiled in reducing gel-loading buffer for 5 min at 100°C, and fractionated by reducing SDS-PAGE. Fractionated proteins were transferred to a PVDF membrane and probed for specific proteins. Quantification of hepcidin specific bands was performed by densitometry using UN-SCAN-IT gels (Version 6.1) software (Silk Scientific, USA) and analyzed graphically using GraphPad Prism (Version 5.0) software (GraphPad Software Inc., USA) and provided as fold change following normalization with β-actin or GAPDH.

2.6. Immunohistochemistry

Immunohistochemistry was performed essentially as described (Ashok et al., 2019; Ashok and Singh, 2018). In short, paraffin-fixed thin sections of three different human anterior segments were rehydrated by standard techniques, and heated at 97°C in the presence of 25 mM tris-1 mM EDTA (pH-8.5) for 40 min for antigen retrieval. Non-specific sites were blocked in 1% BSA for 1 h. The sections were reacted with primary antibodies for Fpn and Cp (Table 2) and sections were incubated with isotype and species specific Rabbit IgG as a control. Subsequently the sections were incubated with anti-rabbit Alexa Fluor 546-conjugated secondary antibody and nuclei were stained with Hoechst. Stained sections were mounted in Fluoromount-G® and imaged with Leica inverted microscope (DMi8). For each experiment, a representative image from 10 different fields were acquired.

2.7. Statistical analysis

Quantification of protein bands was performed and presented as Mean ± SEM using GraphPad Prism (Version 5.0) software. Level of significance was calculated by Student unpaired t-test between the control and experimental samples.

3. Results

3.1. Hepcidin is synthesized locally by cells of the anterior eye segment

To determine whether hepcidin and its target protein Fpn are expressed in the anterior segment, three pairs of cadaveric human eye globes (Table 1) were dissected to isolate the TM, ciliary body (CB), iris, corneal endothelium (CE), retina including the RPE cell layer (Ret), and lens epithelium (LE). Lysates from each tissue were fractionated by SDS-PAGE and analyzed by Western blotting (WB). Probing for hepcidin revealed a prominent band migrating at ~10 kDa in all tissues, consistent with the migration of pro-hepcidin. Retinal lysates served as a positive control (Figure 1 A) (Kulaksiz et al., 2004; Walker et al., 2004). Re-probing for Fpn revealed a positive signal in all tissues except for LE cells probably because of low protein concentration as indicated by the β-actin band that served as a loading control (Figure 1 A).

Figure-1. Expression of local hepcidin in human and bovine anterior eye segment and upregulation by LPS.

(A) Western blotting and probing of lysates from human TM, CB, iris, CE, Ret, and LE for hepcidin shows a band consistent with the migration of pro-hepcidin in all tissues. Re-probing for Fpn shows a positive signal in all tissues except LE cells. The membrane was re-probed for β-actin as a loading control. (B) Amplification of hepcidin from freshly harvested bovine liver and retina (Ret) by RT-PCR shows significantly more hepcidin in the liver relative to the retina (full image in Supplementary Figure S1). A similar evaluation of bovine Ret, CB, iris, TM, CE, and LE shows hepcidin-specific band in all samples. β-actin was amplified in parallel. (C) Relative abundance of hepcidin mRNA after normalization with β-actin shows ~4.8-fold higher expression in the liver than the retina. Within the anterior segment, hepcidin is ~2-fold higher in the retina relative to other tissues. (D) Probing of lysates from bovine CB, CE, iris, TM, and LE for hepcidin shows a band consistent with the migration of pro-hepcidin in all tissues. A similar band was detected in lysates from human retina processed in parallel. Recombinant hepcidin peptide migrating at ~3 kDa shows a positive reaction as expected. The membranes were re-probed for β-actin as a loading control. (E) Probing of lysates from bovine CE from ex vivo cultured corneas treated with LPS shows significant up regulation of hepcidin relative to controls (lanes 1–7). Recombinant hepcidin peptide shows a positive reaction as expected (lane 8). (F) Densitometry after normalization with β-actin shows ~2.6-fold upregulation of hepcidin by LPS relative to controls. Values are mean±SEM of the indicated n. **p<0.01. (G) Probing of membranes for Fpn shows a positive reaction in bovine TM, CB, LE, iris and CE. Lysate from human retina fractionated in parallel shows a band of similar migration. (The membrane for CE in panel D was re-probed). TM, trabecular meshwork; CB, ciliary body; CE, corneal endothelium; Ret, retina; LE, lens epithelium.

The above results were confirmed and extended by performing further analysis on ~20 bovine eyes within 2 hours of sacrifice. Tissues isolated from 5–6 different bovine eyes were processed at one time, and each experiment was repeated at least 3 times. Representative results are shown in Figure 1.

Amplification of hepcidin from bovine retina and liver tissue by semi-quantitative reverse transcription polymerase chain reaction (RT-PCR) revealed significantly more expression in the liver relative to the retina (Figure 1 B and C). A similar evaluation of bovine retina, CB, iris, TM, CE, and LE showed a band consistent with hepcidin in all samples including LE cells (Figure 1 B) (Kulaksiz et al., 2004). The hepcidin band amplified from each tissue was extracted and sequenced to confirm its identity (data not shown). Quantitative estimation of relative band intensity after normalization with β-actin showed ~2-fold higher expression of hepcidin in the retina relative to other tissues (Figure 1 C).

To evaluate the expression of hepcidin peptide, lysates from the above tissues were analyzed by Western blotting. Recombinant hepcidin peptide and lysates from human retina were fractionated in parallel as positive controls (Figure 1 D). As in human samples in Figure 1 A, a band consistent with the migration of pro-hepcidin at ~10kDa was detected in the CB, CE, iris, TM and LE, that co-migrated with pro-hepcidin from human retinal lysate (Figure 1 D). Strong reactivity was detected with recombinant hepcidin peptide as expected (Figure 1 D) though this band was not detected in tissue samples probably because of low abundance.

Since hepcidin is upregulated by interleukin-6 (IL-6) (Lee et al., 2004; Zhang et al., 2017), a cytokine associated with inflammation (Fodor et al., 2006; Ghiță et al., 2019; Yagi-Yaguchi et al., 2017), bovine ex-vivo anterior segment cultures were exposed to 1 μg/ml of lipopolysaccharide (LPS) or vehicle for 3 h, and CE cells were scraped and analyzed by Western blotting. Recombinant hepcidin peptide was fractionated in parallel as a control. Probing for hepcidin shows significant upregulation by LPS relative to controls (Figure 1 E, lanes 1–4 vs. 5–7; Figure 1 F).

Expression of Fpn, the downstream target of hepcidin, was evaluated by subjecting tissue lysates to Western blotting as above. Probing for Fpn revealed a specific band in the TM, CB, LE, iris and CE that co-migrated with Fpn from human retinal lysate (Figure 1 G).

The above results demonstrate that hepcidin is synthesized and expressed in the CB, CE, TM, and LE cells of human and bovine eyes. Although samples from the CB and iris include choroidal capillaries, identification of hepcidin mRNA suggests local synthesis in these tissues as well. All of the above tissues express Fpn, the hepcidin responsive iron export protein, suggesting hepcidin-mediated regulation of iron in the local microenvironment. Subsequent studies were directed at localizing the expression of Fpn and other iron modulating proteins in the anterior segment of the human eye.

3.2. Cellular localization of ferroportin and ceruloplasmin in the anterior eye segment

To identify the expression and polarity of Fpn in the anterior segment, fixed sections from 3 cadaveric human eyes (Table 1) were reacted with antibody specific for Fpn followed by Alexa Fluor 546-conjugated secondary antibody. A positive reaction for Fpn was detected in CE cells, though the polarity was difficult to evaluate (Figure 2 A), on the BL membrane of non-pigmented ciliary epithelial cells (Figure 2 B), on the plasma membrane of TM cells (Figure 2 C), and intracellularly in epithelial cells of the iris and LE cells (Figures 2 D & 2 E).

Figure-2. Localization of ferroportin in the human anterior eye segment.

Immunostaining of fixed tissue section from the anterior segment of human eye shows (A) intracellular reactivity for Fpn in CE cells, (B) on the BL membrane of non-pigmented ciliary epithelial cells, (C) on the plasma membrane of TM cells, and (D-E) intracellularly in the outer epithelial cell layer of the iris and LE cells (white arrowheads). Panels on the right are higher magnification images of marked areas in left panels. Sections reacted in parallel with species and isotype specific rabbit IgG control followed by anti-rabbit IgG Alexa Fluor 546-conjugated secondary antibody did not show any reactivity (Figure 3, panels 1–4). Similar results were obtained from two additional human eye globes (Supplementary Figure S2 (A–D)). NPE, non-pigmented epithelium; PE, pigmented epithelium. Scale bar: 25 μm

Immunoreaction for Cp showed a positive cytosolic reaction in CE cells (Figure 3 A), mainly on the basolateral (BL) membrane of NPE cells (Figure 3 B), on the plasma membrane of TM cells (Figure 3 C), and intracellularly in LE cells and outer cortical lens fiber cells (Figure 3 D, arrowhead and star). Minimal reaction was detected in epithelial cells of the iris (Figure 3 D). Sections reacted with species and isotype-specific IgG antibody were processed in parallel as a negative control for both Fpn (Figure 2) and Cp, and showed no reaction (Figure 3 A–D, panels 1–4).

Figure-3. Localization of ceruloplasmin in the human anterior eye segment.

Immunostaining of anterior segment of human eye shows (A) intracellular reactivity for Cp in CE cells, (B) on the BL membrane of non-pigmented ciliary epithelial cells, (C) on the plasma membrane of TM cells, and (D) intracellularly in LE cells (white arrowheads). Strong intracellular reactivity was also detected in outer lens cortical cells (*star). Minimal reactivity was detected in the iris. Right panels are higher magnification of areas marked in left panels. Sections reacted in parallel with species and isotype specific rabbit IgG control followed by anti-rabbit IgG Alexa Fluor 546-conjugated secondary antibody did not show any reaction (panels 1–4). Similar results were obtained from two additional human eye globes (Supplementary Figure S2 (E–H)). Scale bar: 25 μm

Together, these results demonstrate that Fpn and Cp, proteins necessary for iron export, are expressed on the BL surface of non-pigmented ciliary epithelial cells facing the AH, on the plasma membrane of TM cells, and in the cytosol in CE, and LE cells. Although Fpn was detected on epithelial cells of the iris, reaction for Cp was minimal.

3.3. Expression of transferrin receptor and ferritin in the anterior eye segment

To evaluate the expression of TfR and ferritin in these tissues, lysates collected from each of the above tissues were processed for Western blotting as above. Probing for TfR showed a specific band in the CB, CE, iris, TM, retina, and LE cells (Figure 4 A). Re-probing for ferritin showed a band of the expected migration in the CB, CE, iris, TM, and the retina (Figure 4 B). Lysates from LE cells were not probed for ferritin because of interference with α-crystallin that migrates in the same range as ferritin.

Figure-4. Expression of transferrin receptor and ferritin in bovine anterior eye segment.

(A) Western blotting and probing of lysates from bovine CB, CE, iris, TM, retina and LE for TfR shows a positive reaction in all samples. (B) Re-probing for ferritin shows a positive reaction in all samples. All membranes were re-probed for β-actin or GAPDH as a loading control. LE cells were not evaluated for ferritin because of interference with lens crystalline.

4. Discussion

Retinal iron dyshomeostasis is a major risk factor for age-related macular degeneration, and has stimulated extensive exploration into the underlying cause and iron chelators as a potential therapeutic option (Song et al., 2014). Similar studies for the anterior segment are lacking even though this region is in the direct path of ultraviolet light, a major source of iron-catalyzed reactive oxygen species (ROS) (Izzotti et al., 2009; Umapathy et al., 2013). Here, we demonstrate that ciliary epithelial, corneal endothelial, and TM cells express hepcidin and Fpn, suggesting the presence of a local hepcidin-Fpn-iron axis and stringent regulation of iron exchange in the anterior segment.

Previous observations indicating expression of Fpn on the BL membrane of non-pigmented ciliary epithelial cells suggested a role in iron export across the blood-aqueous barrier (Ashok et al., 2018a). This report confirms and extends these observations by demonstrating co-expression of hepcidin in these cells, suggesting autocrine and paracrine regulation of Fpn by local hepcidin. Since liver hepcidin is unlikely to access the BL membrane of these cells, it is likely that hepcidin secreted by these and other cells in the AH is the principal regulator of Fpn expression and iron efflux from non-pigmented ciliary epithelial cells. Such a mechanism is likely to respond to iron concentrations in the local microenvironment, a biologically preferred option than diffusion of retinal hepcidin, the only other source of hepcidin in the eye. Expression of TfR, Cp, and ferritin in these cells further supports independent regulation of iron uptake, storage, and export by these cells, and a dynamic exchange with the AH. Similar observations have been reported in the retina, where local hepcidin regulates iron homeostasis by modulating transport of iron through Fpn on retinal capillary endothelial cells (Baumann et al., 2019). Further studies are necessary to understand the cross-talk between local hepcidin in the posterior and anterior segments of the eye, and possible cross-talk with liver hepcidin.

Synthesis of hepcidin by lens epithelial, corneal endothelial, and TM cells is surprising because these structures are avascular, and do not require tight regulation of iron exchange with capillary blood. However, the cornea and the lens are in the direct path of ultraviolet (UV) light, and are susceptible to UV-induced ROS that is fueled by reactive iron through Fenton chemistry (Hammond et al., 2014); (Roberts, 2011); (Chen et al., 2009; Marchitti et al., 2011). Likewise, iron-induced oxidative stress in the TM is associated with primary open angle glaucoma (POAG), necessitating stringent regulation of iron in the anterior segment (Babizhayev, 2016; Bagnis et al., 2012; Lin et al., 2010). Local synthesis of hepcidin is likely to mitigate this risk, and add to the protection offered by high concentrations of ascorbic acid and apo-Tf in the AH.

Pathological implications of hepcidin are mainly related to triggers other than iron (Ganz, 2013). Pertinent to the anterior segment are IL6 and TGFβ2 that upregulate hepcidin by the STAT3 (Qian et al., 2014; Zhang et al., 2017) and Smad4 (Wang et al., 2005) pathways respectively. Our observations indicating LPS-mediated upregulation of hepcidin in corneal endothelial cells support this possibility. The resulting downregulation of Fpn in the same and surrounding cells is likely to increase intracellular iron and iron-catalyzed ROS, creating a toxic microenvironment. It is likely that association of ROS with diseases such as POAG, cataract, and corneal endothelial dystrophy is due to concomitant upregulation of hepcidin. Altered levels of hepcidin and other iron regulating proteins have been reported in the AH of glaucomatous eyes, supporting this assumption (Anders et al., 2017; Chowdhury et al., 2010; Farkas et al., 2004; Ghanem et al., 2014; Sorkhabi et al., 2010; Stasi et al., 2007).

Conclusion

Our data demonstrating the presence of a local hepcidin-Fpn-iron axis in the anterior segment suggests autonomous regulation of iron independent of the retina. Upregulation of hepcidin by triggers other than iron such as IL6 implicates iron-catalyzed ROS in the pathogenesis of inflammatory conditions of the anterior segment. Although the biochemical pathways involved in hepcidin synthesis and function are not addressed in this report, these observations are likely to stimulate further exploration on this subject. A clear understanding of the role of hepcidin in the anterior segment and cross-talk with retinal hepcidin will help in taking advantage of hepcidin agonists and antagonists currently under development for systemic disorders for the therapeutic management of ocular conditions as well (Asperti et al., 2019; Poli et al., 2017). This study is therefore both novel and clinically relevant.

Supplementary Material

1

Figure S1. Full image of cropped RT PCR gel (Figure 1B in the manuscript).

Figure S2. Immunostaining to estimate the expression of Fpn and Cp in anterior segment of additional human eyes.

Highlights.

1. Hepcidin is synthesized locally by cells of the anterior eye segment (AS).

2. Co-expression of ferroportin suggests autonomous regulation of iron in the AS.

3. Hepcidin is upregulated by IL6, implicating iron in inflammation of the AS.

Abbreviations

TM

trabecular meshwork

TGF

transforming growth factor

AH

aqueous humor

Fpn

ferroportin

Tf

transferrin

TfR

transferrin receptor

Cp

ceruloplasmin

Heph

hephaestin

CB

ciliary body

CE

corneal endothelium

LE

lens epithelium