Functional roles of bestrophins in ocular epithelia

Abstract

There are four members of the bestrophin family of proteins in the human genome, of which two are known to be expressed in the eye. The gene BEST1 (formerly VMD2) which encodes the protein bestrophin-1 (Best1) was first identified in 1998. Mutations in this gene have now been associated with four clinically distinguishable human eye diseases, collectively referred to as "bestrophinopathies". Over the last decade, laboratories have sought to understand how Best1 mutations could result in eye diseases that range in presentation from macular degeneration to nanophthalmos. The majority of our knowledge comes from studies that have sought to understand how Best1 mutations or dysfunction could induce the classical symptoms of the most common of these diseases: Best vitelliform macular dystrophy (BVMD). BVMD is a dominant trait that is characterized electrophysiologically by a diminished electrooculogram light peak with a normal clinical electroretinogram. This together with the localization of Best1 to the retinal pigment epithelium (RPE) basolateral plasma membrane and data from heterologous expression studies, have led to the proposal that Best1 generates the light peak, and that bestrophins are a family of Ca(2+) activated Cl(-) channels (CaCCs). However, data from Best1 knock-out and knock-in mice, coupled with the recent discovery of a recessive bestrophinopathy suggest that Best1 does not generate the light peak. Recently Best2 was found to be expressed in non-pigmented epithelia in the ciliary body. However, aqueous dynamics in Best2 knock-out mice do not support a role for Best2 as a Cl(-) channel. Thus, the purported CaCC function of the bestrophins and how loss of this function relates to clinical disease needs to be reassessed. In this article, we examine data obtained from tissue-type and animal models and discuss the current state of bestrophin research, what roles Best1 and Best2 may play in ocular epithelia and ocular electrophysiology, and how perturbation of these functions may result in disease.

Figure 1

Structure of hBest1. Best1 is integral membrane protein with 4-6 potential transmembrane spanning α-helices (numbered in A). There is general agreement that only 4 of these span the plasma membrane. Two models have been proposed. Model 1 is supported by the experimental data of (Milenkovic et al., 2007). Model 2 was proposed by and is supported experimentally by (Tsunenari et al., 2003). Through the years most investigators have favored model 1 because it is most frequently predicted by various protein structural software packages. There is no direct experimental data however, that should cause us to favor either model. In hBest1, there are 585 amino acids which are indicated by circles according to model 1 in B. Mutation sites and the human disease(s) they cause are indicated by colored circles. Using model 1, most disease causing mutations are clustered in regions adjacent to the cytosolic face of the 4 TM domains (B). The RFP-TM or bestrophin domain extends from the N-terminus through approximately amino acid 350 and contains all of the TM domains as well as nearly all reported disease causing mutations.

Figure 2

Characterization of anti-hBest1 antibodies and expression of hBest1 in pig tissues, RPE, and cell lines. Lysates were prepared from porcine tissues or cultured cell lines as described in Marmorstein et al., (2002). Immunoprecipiates were prepared from tissue lysates containing 1 gram of total protein as described in Marmorstein et al., (2002) and pBest1 identified in those immunoprecipitates using monoclonal antibody E6-1 (A). Note pBest1 is present only in RPE (Arrow in A) and in no other tissue. hBest1 was expressed in HEK293 cells as a control, with untransfected HEK293 cell lysates serving as a negative control. Immunoprecipitates were prepared from lysates using polyclonal anti-hBest1 antibody Pab-125 for blots shown in B-D. Lysates were prepared from human RPE (hRPE) from a single donor eye, porcine RPE (pRPE) from a single donor eye, 2 × 1 cm² monolayers of cultured fetal human RPE (fhRPE), or a confluent well from a 12-well plate of Calu-3, MCF-7, HEK293, or RPE-J. Immunoprecipitates were resolved by SDS-Page, transferred to PVDF and hBest1 identified using monoclonal anti-hbest1 antibodies E6-1 (B), E6-6 (C, E), or 1C2 (D). Authentic hBest1 is indicated by an arrowhead, and a non-specific band with a slightly higher Mr is indicated with a small arrow. We could not detect hBest1 in the human cell lines Calu-3, HEK293, MCF-7 or the rat derived RPE cell line RPE-J. The absence of the higher Mr band from negative control lanes loaded with SDS-PAGE buffer only indicates that this non-specific band is recognized by all antibodies generated against the C-terminus of hBest1. Note that 1C2 does not recognize pBest1.

Figure 3

EOG light/dark ratios in 163 eyes in families with BVMD. The ratio is indicated at the top and the total number of eyes within the boxes. From Bard and Cross (1975), Reproduced by permission of the University of Wisconsin Press.

Figure 4

Clinical progression of the fundus in BVMD. Classical vitelliform or stage IIa lesion is shown in A. Stage IIb is characterized by a visible fluid line (arrows) within the lesion (B). Partial resorption of the fluid within the lesion gives the appearance of a scrambled egg and is characteristic of Stage III or pseudohypopyon (C). Stage IV (D) is characterized by a gliotic scar accompanied by regions of hypo (stage IVa) or hyperpigmentation (stage IVb) and occasionally neovascularization (stage IVc).

Figure 5

Fluctuations in visual acuity in 3 patients with BVMD. From Bard and Cross (1975), Reproduced by permission of the University of Wisconsin Press.

Figure 6

Amplitude of EOG records obtained from a Sprague-Dawley rat (A), a Long-Evans rat (B) or a normal human subject (C) during the course of an experimental session in which the adaptation state of the eye was varied from room light level, to darkness (dark), to a steady stimulus (light). Note that the human EOG response increases during light exposure and decreases during darkness, whereas the rat responses have the opposite pattern (A) or are not modulated by light level. Data points in A and B reflect the average ± s.d. of at least 5 individual rats. Data points in C reflect individual measurements.

Figure 7

dc-ERGs recorded from a WT mouse (upper waveform) or rat (lower waveform) in response to 7-minute (mouse) or 5-minute (rat) 2.4 log cd/m² stimuli. Note that each response includes all of the major dc-ERG components, and that the overall mouse response has larger amplitude.

Figure 8

dc-ERGs recorded from WT, Best1^+/- and Best1^-/- mice in response to a 7-minute 2.4 log cd/m² stimulus. Note that each response includes all of the major dc-ERG components.

Figure 9

dc-ERGs recorded from control mice (left column) or from mice lacking one of the four VDCC β subunits (right column) in response to 7-minute 2.4 log cd/m² stimuli. Note that the LP is reduced only in lethargic mice, lacking a functional β₄ subunit.

Figure 10

Diagram of LP generation. Light-activation of rod photoreceptors alters the concentration of the unidentified LPS in the subretinal space. This change is detected by LPS receptors located on the apical RPE membrane. These receptors initiate an intracellular signal, which is modulated by Best1 and VDCCs, which ultimately increases the conductance of Cl^- channels, depolarizing the RPE cell. The depolarization event is recorded as the dc-ERG LP.

Figure 11

Effect of hBest1 on rat LP amplitudes and luminance response. RPE generated ERG components were recorded in response to increasing luminance in rats injected subretinally with replication defective adenovirus vectors with an empty expression cassette (Null) or driving expression of hBest1 (WT), hBest1^W93C (W93C) or hBest1^R218C (R218C). Data indicate mean ± s.e.m. of 4 – 12 measurements.

Figure 12

LP amplitudes in Best1^+/W93C KI mice. Dc-ERGs were recorded from Best1^+/W93C knock-in mice or Best1^+/+ littermates in response to a 7 minute light stimulus varying over a 5 log range. Note that the LP amplitude in Best1^+/W93C mice did not vary between -1 and +1 log cd/m². Significant differences (p <0.05, indicated by *) were observed at -1 log cd/m² and at 1 log cd/m². The absolute maximum was obtained from both groups at 2 log cd/m². Although the average response at 2 log cd/m² was lower in Best1^+/W93C mice than in Best1^+/+ mice, the difference was not statistically significant. Data shown are mean ± SE of the maximum LP amplitudes obtained at each stimulus luminance from 5-12 mice.

Figure 13

Localization of mBest2 to the NPE and effect on IOP. The localization of mBest2 in the mouse eye was determined using immunohistochemistry (A). Best2^-/- mice served as a control for antibody specificity. In the eye, the purple VIP reaction product indicating the presence of mBest2 was identified only in NPE cells (indicated by arrows in A) of wild type (Best2^+/+) mice. Based on this localization we compared the IOP of Best2^-/- mice with Best2^+/+ mice (B). IOP was measured via anterior chamber cannulation. Note that IOP is significantly (p <0.0001) lower in Best2^-/- mice. Data are presented as a box plot in which the line within the box marks the median IOP, and the boundaries of the box indicate the range covered by the middle 50% of measurements. Bars above and below the boxes indicate the 90th and 10th percentiles respectively. Symbols outside of the box and bars are outliers. For Best2^+/+ n = 31, for Best2^-/- n= 55.

Figure 14

Comparison of the change in intracellular Ca²⁺ concentration elicited by 50 mM ATP in RPE sheets isolated from Best1^+/+ or Best1^-/- mice. Note that the increase in Ca²⁺ is much greater in the Best1^-/- mouse than in the Best1^+/+ mouse. This effect may underlie the increased LP amplitude observed in Best1^-/- mice in response to lower intensity stimuli.

Figure 15

Hypothetical model of Best1 function in the RPE. We propose that Best1 functions to set the gain on GPCR signaling via Ca²⁺ and to maintain intracellular pH in the face of a changing gradient of CO₂/HCO₃^- flowing across the RPE in response to changes in photoreceptor respiration. Combined data from several labs indicate that Best1 interacts physically and functionally with VDCCs (panel A, 1). The kinetic effects on VDCCs are to accelerate opening and closing times, resulting in a diminished entry of Ca²⁺. Other effects on Ca²⁺ which we have identified in both Best1^W93C knock-in, and Best1^-/- mice as well as fhRPE cultures are the ability to modulate the release of Ca²⁺ stores in response to binding of a ligand to a GPCR (Panel A, 2 and 3). It is not clear yet whether this results form a block of store release (panel A, 2) or by interference with GPCR signaling (panel A, 3). Best1 may be a channel protein. Although most investigators have examined its ability to conduct Cl-, it was recently found that Best1 was as or more efficient at conducting HCO₃^-. We have found that Best1 can alter pHi by promoting NHE activity (4). These effects appear to be dependent on HCO₃^-. It is likely that the changes in pH and Ca²⁺ due to Best1 activity are also interdependent. Should these functions be disrupted, we propose that the level of dysfunction dictates the disease phenotype (panel B). The common denominator in the bestrophinopathies is accumulation of lipofuscin, which could result from altered kinetics of phagocytosis uptake (panel B, 1), maturation (panel B, 2, 4), due to changes in acidification and / or delivery of lysosomal enzymes (E) (Panel B, 2). Under normal circumstances (Panel B, 3) the phagocytosed photoreceptor outer segment (POS) is properly degraded, however, a delay or acceleration in uptake, acidification, or degradation could promote the formation of A2E or other lipofuscin components from precursors already present in the POS. Liposufscin is non-degradable (Panel B, 4) and eventually accumulates in lipofuscin granules (5). The rate and level of lipofuscin accumulation would play a major role in the severity of the disease.