Early Onset Retinal Dystrophy Due to Mutations in LRAT: Molecular Analysis and Detailed Phenotypic Study

Abstract

Purpose.: To report novel variants and characterize the phenotype associated with the autosomal recessive retinal dystrophy caused by mutations in the lecithin retinol acyltransferase (LRAT) gene.

Methods.: A total of 149 patients with Leber's congenital amaurosis (LCA) or early onset retinal dystrophy were screened for mutations in LCA-associated genes using an arrayed-primer extension (APEX) genotyping microarray (Asper Ophthalmics). LRAT sequencing was subsequently performed in this 148-patient panel. Patients identified with mutations underwent further detailed phenotyping.

Results.: APEX analysis identified one patient with a previously reported homozygous LRAT mutation. Sequencing of the panel identified three additional patients with novel homozygous LRAT mutations in exon 2. All four patients had severe progressive nyctalopia, visual field constriction, and photophilia in childhood. Visual acuity ranged from 0.22 logMAR to hand motion. Funduscopy revealed severe retinal pigment epithelial atrophy and minimal retinal pigmentation. Asteroid hyalosis and macular epiretinal fibrosis were frequent. All demonstrated reduced fundus autofluorescence. Optical coherence tomography identified disrupted retinal lamination, outer-retinal debris, and an unidentifiable photoreceptor layer in two cases. Full-field electroretinograms were undetectable or showed severe rod-cone dysfunction. Photopic perimetry revealed severe visual field constriction. Dark-adapted perimetry demonstrated markedly reduced photoreceptor sensitivity. Dark-adapted spectral sensitivity measurements identified functioning rods in two of three patients. All three had severely reduced L- and M-cone sensitivity and poor color discrimination.

Conclusions.: LRAT mutations cause a severe, early childhood onset, progressive retinal dystrophy. Phenotypic similarities to the retinal dysfunction associated with RPE-specific protein 65 kDa mutations, another visual cycle gene, suggest that LRAT deficiency may show a good response to novel therapies.

Introduction

The advent of clinical trials for novel therapies in the treatment of inherited retinal dystrophies, based upon preclinical work in animal models, has led to optimism that effective treatments may become available for these blinding disorders. ^1–7 A fundamental obstacle to the identification of suitable therapies is the genetic heterogeneity of these inherited conditions, which are caused by variants in a number of genes, some known and some as yet unidentified. One subset of this group of diseases is Leber's congenital amaurosis (LCA; Mendelian Inheritance in Man number: 204000), ⁸ the most severe form of a spectrum of retinal diseases termed early onset retinal dystrophies (EORD). To date, 14 genes and a further locus have been implicated in the pathogenesis of nonsyndromic EORD: guanylate cyclase 2D (GUCY2D) ⁹ ; RPE-specific protein 65 kDa (RPE65) ¹⁰ ; tubby-like protein 1 (TULP1) ¹¹ ; cone-rod homeobox (CRX) ¹² ; aryl-hydrocarbon-interacting protein-like 1 (AIPL1) ¹³ ; c-mer proto-oncogene tyrosine kinase (MERTK) ¹⁴ ; retinitis pigmentosa GTPase regulator interacting protein 1 (RPGRIP1) ¹⁵ ; crumbs homolog 1 (CRB1) ¹⁶ ; lecithin retinol acyltransferase (LRAT) ¹⁷ ; retinol dehydrogenase 1 (RDH12) ¹⁸ ; centrosomal protein 290 kDa (CEP290) ¹⁹ ; LCA5 ^20,21 ; spermatogenesis associated 7 (SPATA7) ²¹ ; IQ motif containing B1 (IQCB1) ²² ; and the LCA9 locus. ²³ These photoreceptor dystrophies, predominantly inherited in an autosomal recessive manner, are characterized by visual impairment noted at birth or within the first few years of life, a variable retinal appearance, and severely attenuated rod and cone responses on full-field electroretinography (ERG). ²⁴  

In 1988, evidence supporting LRAT function in the visual cycle was published by MacDonald et al. ²⁵ In 2001, the LRAT gene, comprising of three exons, was mapped to chromosome 4q31.2. ²⁶ The gene encodes a 230-amino acid enzyme that is expressed in several fetal and adult human tissues, including the retinal pigment epithelium (RPE) and liver. ²⁷ LRAT catalyzes the synthesis from retinol (vitamin A) of all-trans-retinyl esters in the RPE, and is followed in the visual cycle by the synthesis of 11-cis-retinol by RPE65, resulting in chromophore regeneration for rhodopsin and the cone photopigments. ²⁶ Deficiency of LRAT results in a lack of the 11-cis-retinal chromophore and reduced levels of functional visual pigment. ²⁸ It is evident from the Lrat^−/− mouse model that the lack of chromophore leads to photoreceptor cell death, with cone degeneration occurring rapidly and rod degeneration more slowly. ^29,30 This model displays severely attenuated rod and cone function from early in life. At this age, there is little structural change to the retina, with only rod outer segments showing around 35% shortening; the outer nuclear layer and inner retinal structures remain normal. ²⁹  

In 2001, Thompson et al. reported the first mutations in LRAT in humans, which were associated with an early onset rod cone dystrophy. ¹⁷ To date, five LRAT mutations in seven families (eight affected individuals) have been published, although there is limited information about the clinical phenotype. ^17,28,31,32  

The present report describes the detailed retinal phenotype in affected individuals from four families with mutations in LRAT and identifies three novel mutations. 

Methods

Subjects

Patients with a diagnosis of LCA or EORD were recruited from the medical retina clinics of Moorfields Eye Hospital (MEH), London, UK, and St. Mary's Hospital, Manchester, UK. Informed consent was obtained from all patients and family members involved in this study. The study complied with the tenets of the Declaration of Helsinki, and received approval from the local ethics committees. 

LRAT Molecular Analysis

Screening of patients for LRAT mutations was carried out using two methods. One patient was identified following analysis using an arrayed primer extension (APEX) genotyping microarray (LCA microarray chip; Asper Ophthalmics, Tartu, Estonia). ^33,34 Three patients were identified from a panel of 148 LCA and EORD patients using direct Sanger sequencing. This panel was enriched for novel mutations as the patients had previously been screened and excluded for known LCA variants through APEX analysis using the Asper Ophthalmics LCA chip. 

Genomic DNA was isolated from blood using a puregene kit (Gentra Puregene Blood Extraction Kit; QIAGEN, Crawley, UK). The three coding regions and intron/exon boundaries of LRAT were PCR amplified from 100 ng of genomic DNA. PCR amplification and sequencing of products was performed using previously described methods. ³⁵ Primer sequences and PCR conditions are displayed in Table 1. The likely pathogenicity of novel LRAT missense variants on protein structure and function was assessed using the following mathematical/statistical algorithms: scale-invariant feature transform (SIFT), ³⁶ PolyPhen-2, ³⁷ and pMUT. ³⁸ Exon 2 was screened for sequence variants in 96 European Collection of Cell Cultures (ECACC) control DNAs. The exome variant server (URL: http://evs.gs.washington.edu/EVS/, accessed February 2012) ³⁹ was accessed for one variant that the in silico analysis predicted not to be harmful. This is an online database, in the public domain, of the variants identified by next generation sequencing in the protein coding regions of the human genome across diverse populations. Identified mutations were confirmed bidirectionally and checked in available family members for segregation with disease. 

Table 1. 

 

Primer Sequences and PCR Conditions for LRAT Molecular Analysis

Exon	Primer Sequence	PCR Annealing Temp (°C)	MgCl2 Conc. (mM)	Amplicon Size (bp)
1F	CAATCAGTGAGCTTTCCGGGT	60	2.0	291
1R	CGGGCTGGGCAAGTTAAGCT
2FA	AGCTTAACTTGCCCAGCCCG	68	3.0	279
2RA	CCAGGATGAGACGCTTGTTG
2FB	CTATGGCATCTACCTAGGAGA	60	2.0	414
2RB	GGAATTAGATCCCTACTCGCG
3F	GTCTAGTTCTTCTGGTAGACAG	60	2.0	336
3R	GAACACAGTGTTACGGGTCACA

 Temp, temperature; Conc., concentration; mM, millimolar; bp, base pair.

Clinical Examination

The phenotype of the four patients identified with LRAT mutations was characterized following mutation identification. Best-corrected monocular visual acuity using a logMAR scale, slit-lamp biomicroscopy, and funduscopy were performed in all subjects. Color fundus photography using a retinal camera (Topcon TRC 501A; Topcon Corporation, Tokyo, Japan); optical coherence tomography (OCT) using a OCT scanner (Heidelberg SPECTRALIS Spectral domain OCT; Heidelberg Engineering, Dossenheim, Germany) or STRATUS OCT 3000 scanner (Zeiss Humphrey Instruments, Dublin, CA); and retinal autofluorescence (AF) imaging using a confocal scanning laser ophthalmoscope (Zeiss Prototype; Carl Zeiss Inc., Oberkochen, Germany) were performed in all patients. Color vision assessments were performed using Ishihara pseudoisochromatic plates, the Farnsworth-Munsell FM-100-hue tests, and Rayleigh and Moreland anomaloscope matches. The FM 100-hue was performed under CIE Standard Illuminant C from a MacBeth Easel lamp or Illuminant D50 (daylight) from a daylight lighting booth. Goldman kinetic perimetry was performed in three patients. Electrophysiological assessment including full-field ERG and pattern ERG (PERG) was performed in three patients, to incorporate the recommendations of the International Society for Clinical Electrophysiology of Vision. ^40,41 Perimetry, electrophysiological, and psychophysical testing were not available in the fourth patient. 

Psychophysical Assessments of Vision

Static Threshold Perimetry

Detailed static threshold perimetry, performed in both the light- and dark-adapted states, was conducted in three of four patients using a Humphrey visual field analyzer (Allergan Humphrey; Hertford, UK). Dark-adapted perimetry was performed following pupil dilatation (1% tropicamide and 2.5% phenylephrine hydrochloride drops) and 45 minutes dark adaptation. The Humphrey visual field analyzer was modified for dark-adapted conditions, as described previously, ^42,43 and controlled by a customized computer program (PS/2 model 50; International Business Machines, Armonk, NY). ⁴⁴  

Dark-Adapted Spectral Sensitivity

A conventional Maxwellian-view optical system with a 2-mm entrance pupil illuminated by a 75-W xenon arc lamp was used for the dark-adapted spectral sensitivity measurements. Wavelengths were selected with a monochromator with a half-maximum bandwidth of 4 nm (Jobon Yvon H-10). The radiance of each beam could be controlled by the insertion of fixed neutral density filters (Oriel) or by the rotation of a circular, variable neutral density filter (Rolyn Optics). Sinusoidal modulation was produced by the pulse-width modulation of fast, liquid crystal light shutters (Displaytech) at a carrier frequency of 400 Hz. The position of the observer's head was maintained by a dental wax impression. The experiments were under computer control. Details of the apparatus have been described previously. ⁴⁵ Observers were dark adapted for 40 minutes prior to the start of the measurements. The target stimulus was 3.5° in visual diameter and was sinusoidally flickered at 1 Hz and presented at 10° in the superior retina. Measurements were made at target wavelengths of 450, 500, 550, 600, and 650 nm. At each wavelength, the observer adjusted the radiance of the target until the flicker at 1 Hz was just at threshold. 

Critical Flicker Fusion

A second Maxwellian-view optical system with a 2-mm entrance pupil illuminated by a 900-W xenon arc lamp was used for the critical flicker fusion measurements. Wavelengths were selected with interference filters with full-width at half-maximum bandwidths of between 7 nm and 11 nm (Ealing or Oriel). Other details were as for the spectral sensitivity measurements. 

Measuring the changes in temporal sensitivity that accompany changes in light level is a particularly efficient way of characterizing the defects in light adaptation caused by molecular deficits in the transduction cascade or visual cycle. Changes in temporal acuity or resolution (also known as the critical flicker frequency or c.f.f.), as a function of light level, were measured in three of four patients. ⁴⁶ For L-cone c.f.f. measurements, a flickering target of 4° in diameter and 650 nm in wavelength was presented in the center of a 9° diameter background field of 481 nm. Fixation was central. The 481-nm background, which was fixed at 8.26 log quanta s⁻¹ deg⁻² at the cornea (1.39 log₁₀ photopic trolands), served to suppress the rods but also selectively desensitized the M-cones at lower target radiances. The 650-nm target, chosen to favor flicker detection by mainly L-cones over most of the intensity range, was varied in intensity from 6.5 to 11.0 log₁₀ quanta s⁻¹ deg⁻² (−0.63 to 3.87 log₁₀ photopic trolands). At higher target intensities, the M-cones are also likely to contribute to flicker detection. At each target intensity, the observer adjusted the flicker frequency to find the frequency at which the flicker just disappeared (the c.f.f.). For the S-cone c.f.f. measurements, a flickering target of 4° of visual angle in diameter and 440 nm in wavelength was presented in the center of a 9° diameter background field of 620 nm. Fixation was central. The 620-nm background field selectively desensitized the M- and L-cones, but had comparatively little direct effect on the S-cones. For normal observers, a 620-nm field of 11.51 log₁₀ quanta s⁻¹ deg⁻² isolates the S-cone response up to a 440-nm target radiance of about 10.5 log₁₀ quanta s⁻¹ deg⁻². ⁴⁷  

Results

Genetic Analysis

Study authors identified four families with variants in LRAT (Table 2). Figure 1 demonstrates the locations of the variants identified in this study. 

Table 2. 

 

LRAT Molecular Analysis

Patient	Family	Consanguinity	LRAT Mutation		Mutation Type	Method	Reference
1	1	Yes	c.525T>A	p.Ser175Arg	Homozygous	APEX analysis	17
2	2	Yes	c.40-41delGAinsTT	p.Glu14Leu	Homozygous	Sequencing	Novel
3	3	No	c.181T>A	p.Tyr61Asp	Homozygous	Sequencing	Novel
4	4	No	c.316G>A	p.Ala106Thr	Homozygous	Sequencing	Novel

Figure 1. 

 

LRAT gene structure showing the locations of the mutations identified in this study. Novel mutations are shown in red.

In patient 1 (family 1), APEX analysis identified the previously reported homozygous missense mutation c.525T>A, p.Ser175Arg, in exon 2 of LRAT. ¹⁷ This was the only affected individual in a multiply consanguineous Gujarati Indian family. Direct sequencing confirmed this variant in the proband and demonstrated that both parents were carriers. 

Families 2, 3, and 4 were identified by the screening of LRAT in a panel of 148 patients who had been negative for known mutations based on the Asper LCA APEX chip. The female proband of family 2 (patient 2), born to first cousin parents of British origin, was found to be homozygous for a 2-base pair deletion/insertion, c.40-41delGAinsTT, predicted to cause a substitution in the amino acid sequence: p.Glu14Leu. Both unaffected parents were found to be heterozygous for the mutation. The variant c.181T>A, p.Tyr61Asp, was identified in the homozygous state in the male proband of family 3 (patient 3), a nonconsanguineous family of Swedish origin. The mutation was present in the heterozygous state in his father. His mother was unavailable for testing. The female proband of family 4 (patient 4), a nonconsanguineous family of Caribbean origin, was identified to harbor a homozygous c.316G>A, p.Ala106Thr, variant. This patient presented with a phenotype similar to RPE65 retinopathy but was found to be negative for mutations in RPE65 prior to LRAT screening. It has not been possible to test segregation of this change, as other family members are unavailable. 

All variants identified in this study were clustered in exon 2, and were not found in 96 ECACC control DNA samples, or in the other 145 patients screened. All four variants were conserved across mammals in a protein alignment. In Silico analysis using the programs SIFT, ³⁶ PolyPhen-2, ³⁷ and pMUT ³⁸ were used to test if these variants had the potential to be disease-causing. The previously reported p.Ser175Arg variant and the novel p.Glu14Leu variant were predicted to be disease causing by all three in silico analysis programs. The novel p.Tyr61Asp variant was predicted to be disease causing in two out of three programs. Only the novel p.Ala106Thr variant was predicted to be nonpathogenic by in silico analysis. This change was not found in 192 ECACC alleles, and a study of the Exome Variant server ³⁹ showed that the p.Ala106Thr variant is not present in more than 3700 alleles derived from African Americans, who are more likely to be ethnically closer to patient 4. Therefore, this unique variant was included in this study as being potentially disease causing, particularly as the phenotype is consistent with the other patients with LRAT mutations and because the patient was not identified to harbor any mutations in RPE65. 

Clinical Findings

Two of four patients identified with LRAT mutations were female (Table 3). Onset of visual symptoms, as reported by the patients, ranged between 1 to 3 years of age. Severe nyctalopia was a universal feature, occurring at presentation in patients 1, 2, and 3. Visual field loss was evident in all patients in early childhood; and photophilia (in the form of “light staring”) in early childhood was reported in patients 1, 3, and 4. Information regarding symptom onset was ascertained directly from the patients at their first clinical assessments, in adulthood. Aside from patient 1, the parents of these subjects were unavailable to ask for their memory of their child's visual behavior in infancy. The mother of patient 1 reported normal visual behavior before age 3 years. None of the patients reported nystagmus in the first year of life, suggesting that severe visual impairment was not present from infancy. All patients reported that their quality of vision had deteriorated over time with regard to color vision, nyctalopia, visual fields, and reading vision. Patients 1–3 reported a significant increase with age in time taken to adapt from photopic to mesopic conditions. There was no significant ophthalmic family history in any patient. General health was good in all subjects except patient 2, who wore a hearing aid in her right ear for acquired conductive hearing loss secondary to otosclerosis. 

Table 3. 

 

Clinical Features of Patients with LRAT Mutations

Patient/ Sex	Ethnicity	Age at Onset (yr)	Symptoms						Age at Exam (yr)	BCVA (LogMAR)		Electroretinography
			Nyctalopia	VF Constriction	Reduced Vision	Photophilia	Poor- Color Vision	Prolonged Photopic to Mesopic Adaptation		RE	LE	Pattern ERG	Full-Field ERG
1/M	SE Asian	3	+	+	+	+	+	+	27	0.22	0.90	Some bilateral macular preservation	Undetectable
2/F	Caucasian	1	+	+	+	−	+	+	54	0.66	HM	Severe bilateral macular dysfunction	Undetectable
3/M	Caucasian	1	+	+	+	+	+	+	41	HM	1.80	Severe bilateral macular dysfunction	Severe PR dysfunction
4/F	Caribbean	2	+	+	+	+	+	Not available	31	0.80	0.50	Not available	Not available

 Yr, years; BCVA, best corrected visual acuity; VF, visual field; RE, right eye; LE, left eye; M, male; F, female; SE, Southeast; +, present; –, absent; HM, hand motion; PR, photoreceptors.

Best corrected logMAR visual acuity at first clinical assessment in the right eye was 0.22, 0.66, hand motion, and 0.80, and in the left eye was 0.90, hand motion, 1.80, and 0.50 in patients 1 to 4, respectively (age range, 27–54 years). Patient 3, with hand motion vision in the right eye, was amblyopic in this eye. Refractive errors were variable, with a spherical equivalent range of between −1.00 diopters to +3.63 diopters in the right eye and −1.00 diopters to +3.32 diopters in the left eye. Patient 1 was the only patient to have nystagmus, which was minimal and torsional in character, noted from the first clinical assessment. There is no historic data available regarding the presence of nystagmus in this patient prior to this assessment. Patients 2–4 did not have nystagmus. All four patients were unable to identify any plates on Ishihara color testing. Patients 1, 2, and 3 demonstrated low-color discrimination on the Farnsworth-Munsell 100 hue test. Their Rayleigh anomaloscope matches also suggested poor red-green color discrimination. 

Ocular examination in all patients revealed normal anterior segments. Mild nuclear sclerosis was present in the left eye of patient 2 and bilaterally in patient 4. Funduscopy in all patients revealed widespread bilateral symmetrical RPE atrophy throughout the retina with arteriolar attenuation (Fig. 2). There was minimal retinal pigmentation that, when present, was localized to the mid-peripheral retina. Within the maculae, the RPE had a mottled appearance. Additionally, in patient 1, there was a ring of retinal pallor in the macula extending to the vascular arcades (Fig. 2A). The macula appeared relatively spared in patient 3, although there were depigmented areas superior to each fovea (Fig. 2C). Patient 4 had bilateral parafoveal epiretinal membranes (ERM; Fig. 2D). Two patients had asteroid hyalosis, which was severe and bilateral in patient 2. She had undergone a left vitrectomy in an attempt to improve her vision, without any perceived benefit (Fig. 2B). In patient 3, asteroid hyalosis was present in the right eye only. 

Figure 2. 

 

Retinal features of patients with LRAT mutations identified in this study. The left eye in each patient is shown. (A) Patient 1. (B) Patient 2. (C) Patient 3. (D) Patient 4.

A reduced AF signal was detected in all patients on fundus AF (FAF) imaging (Fig. 3, normal FAF image shown for comparison in Fig. 3A). During imaging, it was necessary to utilize a large number of frames to obtain a mean image that was at the highest signal detection sensitivity, resulting in the optic nerve head and large retinal vessels appearing artefactually brighter than those obtained in normal subjects. A very small hyperautofluorescent signal at the foveolae (patient 1; Fig. 3B); patchy hyperautofluorescence relative to the optic disc (patients 2, 3, and 4; Figs. 3C, 3D, and 3E, respectively); and a diffuse hyperautofluorescent ring at the macula (patient 3; Fig. 3D), were present in those patients in whom an AF signal was detected. 

Figure 3. 

 

Fundus autofluorescence and optical coherence tomography images of the left eye of patients with LRAT mutations identified in this study. (A, F) Normal images. (B, G) Patient 1. (C, H) Patient 2. (D, I) Patient 3. (E, J) Patient 4. Red arrow, Figure 3G, saw-tooth appearance to NFL; yellow arrow, Figure 3G, preserved IS/OS junction. Yellow arrows, Figures 3H and 3I, shallow broad foveal depression. INL, inner nuclear layer; ONL, outer nuclear layer.

Spectral domain OCT (SD-OCT) imaging was performed in patients 1–3 (Fig. 3, normal SD-OCT image shown for comparison in Fig. 3F). When carrying out OCT imaging, patients were encouraged to fixate centrally and images were acquired when fixation was as central as possible. Fixation was stable in the better-seeing eye, but not as good in the eye with worse acuity. Fixation in SD-OCT imaging was monitored by means of an infrared scanning laser ophthalmoscope. In all patients, the foveal thickness was relatively preserved, although the contour of the foveal depression appeared shallow and broad in the oldest patients (patients 2 and 3, Figs. 3H and 3I respectively, yellow arrows). In patient 1 (the youngest patient), the inner segment/outer segment (IS/OS) junction appeared to be preserved throughout the macula (Fig. 3G, yellow arrow). There was a visible external limiting membrane above the IS/OS junction, and the retinal lamination appeared intact. At the fovea in patients 2 and 3, hyperreflective deposits were visible in the outer retina, which appeared to be disrupted; and the retinal lamination was less well defined with no clearly visible photoreceptor layer (Figs. 3H, 3I). In addition, the outer nuclear layer was lost centrally but possibly retained parafoveally. The nerve fiber layer (NFL) had a striking irregular “saw-tooth” appearance on SD-OCT imaging that was more prominent in the nasal parafoveal regions in patients 1–3 (demonstrated in Fig. 3G, red arrow). Below the peaks of these “teeth,” a vertical line, similar to a shadow, was observed through the lower retinal layers. There was no evidence of vitreoretinal traction. Time-domain OCT in patient 4 revealed preservation of the foveal thickness, but the foveal IS/OS junction was barely detectable (Fig. 3J). Bilateral extrafoveal ERM were visible, without loss of the foveal contour. 

Electrophysiology was performed in patients 1, 2, and 3. In patient 1, the youngest patient, residual PERGs were detectable bilaterally, in keeping with minimal residual macular function. PERGs were undetectable in patients 2 and 3. Rod-mediated full-field ERGs were undetectable in all three patients following a standard period of 25 minutes of dark adaptation. Light-adapted cone-mediated ERGs were undetectable in patient 1 and showed delayed small residual responses in patients 2 and 3. No interocular asymmetry was demonstrated in the full-field ERG responses of patient 3 following overnight dark adaptation of the left eye. 

Goldman kinetic perimetry, performed in three subjects, identified consistent responses to the brightest V4e target only (Fig. 4 shows one example, patient 2). Patients demonstrated circumferentially constricted visual fields between 5° and 40°. Patient 1 retained vision in the central 30° and demonstrated additional residual far-peripheral crescents of vision measuring 10° in the inferotemporal quadrants. Patient 2 had retained vision in the central 15° and 40° in the left (Fig. 4A) and right (Fig. 4B) eyes, respectively. She also demonstrated an additional peripheral crescent of retained vision of 20° in the inferior and temporal quadrants. Patient 3 only had retained vision within the central 5°, with additional small peripheral crescents of remaining vision. 

Figure 4. 

 

Goldman perimetry, patient 2. (A) Left eye. (B) Right eye. Red line, V4e target; green line, III4e target.

Static Threshold Perimetry

Photopic static threshold perimetry in three of four patients demonstrated severely elevated thresholds of 10–30 decibels (dB) across the central 30° (Fig. 5 shows one example, patient 3). Dark-adapted perimetry demonstrated severe loss of sensitivity in all patients of 30–40 dB. No responses were detectable to red and blue “scotopic” stimuli in the dark-adapted state in patient 1. In patients 2 and 3, only a few locations showed any measurable responses and these were within the central 9° of fixation. These results indicate severe loss of rod function; indeed, it is likely that sensitivity in these patients is mediated mainly by dark-adapted cones. The losses are comparable with the cone losses implied by the c.f.f. measurements in Figure 6B (i.e., the horizontal shifts between the patient and normal data). 

Figure 5. 

 

Full static threshold central 30° visual fields for photopic (A) and scotopic blue (B) and scotopic red (C) stimuli, left eye, patient 3. Severe threshold elevations are evident throughout with generally some central function, but elevated by 20–30 dB even at the most sensitive locations.

Figure 6. 

 

Psychophysical measurements. (A) Dark-adapted spectral sensitivities. Spectral sensitivity data for LRAT patients 1 (blue circles); 2 (green diamonds); 3 (yellow triangles); and typical normal data (red squares). The error bars are ±1 standard error between runs. The curves are standard spectral sensitivity functions vertically shifted to least squares fit each data set. The continuous black lines are shifted versions of the rod or scotopic V′(λ) luminosity function, ⁷² and the red dashed line is a shifted version of the cone or photopic V(λ) luminosity function recently adopted by the CIE. ⁷³ Agreement with V′(λ) is indicative of rod function in patients 2 and 3, but agreement with V(λ) suggests mainly cone function in patient 1. (B) Cone critical fusion frequencies. Cone c.f.f. data for LRAT patients 1, 2, and 3, and the mean data for 12 normal observers (symbols as panel A). The error bars are ±1 standard error between runs for the patient data and between observers for the normal data. Patients 1 and 3 show a substantial loss of cone temporal sensitivity, while patient 2 shows a more moderate loss.

Dark-Adapted Spectral Sensitivities

Figure 6 shows psychophysical measurements for patients 1–3. The dark-adapted spectral sensitivity data (Fig. 6A) for patients 2 (green diamonds) and 3 (yellow triangles) are consistent in shape with the rod or scotopic luminosity function (black lines), which provides clear evidence that both patients have functioning rods. However, the rod sensitivity of both patients is reduced compared with normal subjects (red squares). The data for patient 1 (blue circles) are consistent in shape with the cone or photopic luminosity function (dashed red line) and are much reduced in sensitivity, suggesting that this patient has little or no rod function. Comparison of this patient's data with those of normal subjects suggests a reduction in cone sensitivity of about 3 log units, which is consistent with the cone c.f.f. loss in Figure 6B. 

The L-cone c.f.f. measurements (Fig. 6B) for the three patients were substantially lower than those for normal subjects (red squares). Of the three patients, patient 2 (green diamonds) showed the best temporal resolution, suggesting reduced but moderately good L-cone function. In contrast, patients 1 (blue circles) and 3 (yellow triangles) exhibited a devastating loss of temporal resolution compared with normal subjects. For these patients, flicker is first seen clearly at target radiances above 9.0 log₁₀ quanta s⁻¹ deg⁻² (approximately 250 times higher than normal subjects), but resolution reaches less than 20 Hz even at the highest radiances, suggesting severely reduced L-cone function. S-cone c.f.f. measurements revealed no measureable S-cone function in all three patients. 

Discussion

The present study reports three novel mutations and extends the phenotypic characterization of LRAT retinopathy. Patients harboring mutations in LRAT were identified by genotyping microarray analysis and by direct sequencing. LRAT screening in a panel of patients with LCA and EORD, enriched for novel mutations, identified three patients, giving a frequency of 2% in this cohort. The frequency of LRAT mutations in the study's entire panel is likely to be less than 1%. This proportion is consistent with previous studies and confirms that LRAT mutations are a rare cause of LCA and EORD. ^17,28,48,49 Two of the three novel variants identified were predicted to be disease causing by in silico analysis. The pathogenicity of the p.Ala106Thr variant is uncertain; in silico analysis does not predict it to be harmful. However, it was not found in ethically matched controls and the mutation occurs in a conserved part of the gene. Furthermore, the phenotypic changes observed in this patient are consistent with the phenotype displayed in the other LRAT patients; and to date in this patient, no other mutations in LCA/EORD genes, including RPE65, have been identified. As all but one LRAT mutations identified to date cluster in exon 2 (Table 4), targeted screening of exon 2 of LRAT in patients with a phenotype suggestive of LRAT mutations may be an efficient first approach. 

Table 4. 

 

LRAT Patients and Mutations Published to Date

Subject ID	Allele 1	Exon	Allele 2	Exon	Reference
arRP186	c.525T > A, p.Ser175Arg	2	c.525T > A, p.Ser175Arg	2	Thompson et al.17
arRP824	c.525T > A, p.Ser175Arg	2	c.525T > A, p.Ser175Arg	2	Thompson et al.17
2910	c.396delAA	2	Not identified		Thompson et al.17
No ID	c.217_218delAT, p.Met73AspfsX47	2	c.217_218delAT, p.Met73AspfsX47	2	Senechal et al.28
293	c.371C > T, p.Arg109Cys	2	c.605G > A, p.Arg190His	3	Preising et al.31
27241	c.217_218delAT, p.Met73AspfsX47	2	c.217_218delAT, p.Met73AspfsX47	2	den Hollander et al.32
27266	c.217_218delAT, p.Met73AspfsX47	2	c.217_218delAT, p.Met73AspfsX47	2	den Hollander et al.32
Sister of 27266	c.217_218delAT, p.Met73AspfsX47	2	c.217_218delAT, p.Met73AspfsX47	2	den Hollander et al.32

LRAT is an enzyme that is a member of the NlpC/P60 thiol peptidase protein superfamily. ⁵⁰ Members of this superfamily share three conserved residues: Cys; His; and a polar amino acid, that are involved in the catalytic activity of the protein. In LRAT, two of these key residues are known to be Cys16 and His60. ⁵¹ The third residue may be histidine 72. ⁵² One of the novel mutations identified by the study authors, p.Tyr61Asp, lies next to this important histidine at position 60, and therefore may affect the catalytic activity of this enzyme. It is more difficult to propose a disease mechanism for the mutation p.Glu14Leu. It lies outside the truncated protein known as tLRAT, which is missing the N and C termini. ⁵³ tLRAT is catalytically active. A change in the N terminus may affect its ability to reach the endoplasmic reticulum. It is known that the C terminus has a role in trafficking of the protein, ⁵⁴ although it is not known if the N terminus has a similar role. 

The retinal disease associated with LRAT mutations in the patients identified in this study is severe, with a phenotype consisting of poor vision, nyctalopia, and visual field constriction from childhood. Light staring in early childhood is also a common feature. Over time, patients report worsening of visual acuity and contrast sensitivity, and deterioration of color perception. The electrophysiological data measured in the patients in this study are in keeping with severe generalized photoreceptor dysfunction, with rods more severely affected than cones. Previous reports ^17,28,31,32 have described a similarly severe retinal dystrophy with an undetectable electroretinogram in affected individuals. ^17,28,32 The diagnoses in other reported cases include LCA, and juvenile retinitis pigmentosa (RP). In the majority of patients, symptom onset was reported in childhood and included nyctalopia, poor vision, and nystagmus. ^28,31,32 Photophobia has been reported in some cases. ^31,32 Visual acuity ranged from 0.6 logMAR (age 6 years) to hand motion vision (age 23 years). ³² Hypermetropic refractive errors have dominated. ^17,28,31,32 The reported retinal appearance ranges from normal in infancy ²⁸ to peripheral RPE atrophy, optic disc pallor, and “perimacular retinal surface wrinkling” in adult life. ¹⁷  

All patients in this study had peripheral RPE atrophy but with little pigment migration into the retina, suggesting that photoreceptor cell death is a late feature. ⁵⁵ This is a feature in other reported cases. ^17,28,31,32 ERM formation was noted in 2 of 8 eyes (25%), more common than the reported prevalence of 1.2% identified in RP in general. ⁵⁶ Asteroid hyalosis also appears to be more common in LRAT retinopathy: it was identified in three of eight eyes (37.5%) in this cohort; and occurs more frequently than in RP (3.1% in one series ⁵⁷ ). 

None of the previous reports of patients with LRAT mutations included the results of detailed retinal imaging. FAF imaging in this study's cases demonstrated a reduced AF signal (Fig. 3), which has been reported in other retinal disorders caused by mutations in genes encoding visual cycle enzymes. ^58–60 Palczewska et al. recently published data on multiphoton excitation fluorescence microscopy, a noninvasive imaging modality that allows visualization of fluorescent biomarkers that exist within RPE cells. ⁶¹ They identified retinosomes and fluorophores such as A2E and all-trans-retinyl esters in wild type mice when their RPE had been subjected to excitation wavelengths at ∼730 nm and ∼910 nm, respectively. At these wavelengths, Lrat^−/− knockout mice appeared to lack fluorescence, which was present in wild type mice. Scanning laser ophthalmoscopy in these knockout mice identified little to no fundus autofluorescence, which would be consistent with this study's findings in human subjects with LRAT mutations. Reduced fluorophore production has also been reported in Rpe65^−/− knockout mice ⁶¹ and in humans with RPE65 mutations. ⁶² The finding of a reduced signal on FAF imaging in humans with retinal dystrophy is very suggestive of disease caused by mutations in genes involved in the visual cycle. 

Although the SD-OCT image acquisition was limited by fixation instability in some of the subjects in this study, reasonable images were obtained in most. In the youngest subject (age 27 years), the IS/OS junction remained visible at the central macula. In the two older subjects (ages 41 and 54 years), there were extensive changes within the outer retinal structures, with loss of the normal retinal lamination and “debris” within the outer nuclear layer. In patients 1–3, the retinal NFL displayed an interesting saw-tooth appearance, with no evidence of overlying vitreoretinal traction. Although OCT findings in Lrat^−/− mouse models have not yet been described, OCT studies in RPE65 patients ^63–65 and in Rpe65^−/− mice ⁶⁶ have not described this appearance either, and no funduscopic features have identified that may correlate with this NFL irregularity. Thompson et al. described a “wrinkled retinal appearance” in their cohort of LRAT patients; however, this was not correlated with OCT imaging. ¹⁷ The cause of this appearance is unclear, but given the observed higher incidence of ERM formation in LRAT retinopathy, it may indicate early epiretinal fibrosis. 

Due to the small number of patients identified with mutations in LRAT, estimation of the progression of visual loss from cross-sectional data is not possible. Patients that displayed severe photopic kinetic visual field constriction and elevated thresholds on static perimetry testing all maintained useful navigational vision that was subjectively reported to be better in ambient lighting conditions or in daylight. Dark-adapted spectral sensitivity measurements demonstrated a range of rod function from reduced to undetectable. Perimetric and c.f.f. measurements showed that cone function was also severely reduced. The losses relative to the normal in the dark-adapted spectral sensitivity measurements and those found in the dark-adapted perimetry measurements differ because the spectral sensitivity measurements more effectively isolated residual rod function in two patients. 

Lrat^−/− and Rpe65^−/− murine models display similar retinal dysfunction to each other. Rod degeneration is slow and cone degeneration is rapid in both models, and is associated with mislocalization of cone opsins. ^{29,30,62,66,67} Residual rod and cone function has been identified using psychophysical testing in human RPE65 retinopathy. ⁶⁴ From observations using Rpe65 mutant animal models in which isomerization activity continues to exist, and of RPE65 human studies in which milder phenotypes exist, it has been hypothesized that this residual vision may be due to hypomorphic allelic function. ^68,69 Another possibility is the involvement of an alternative visual cycle pathway for chromophore regeneration involving cones and Müller cells. ⁷⁰ The components of this alternative cycle and the roles of LRAT and RPE65 within it remain to be elucidated. However, the recent characterization in the cone-dominated zebrafish of RPE65c, a protein that shares 78% sequence homology to RPE65a, the orthologue of human RPE65a, the orthologue of human RPE65, as the putative isomerohydrolase involved in the cone visual cycle, supports the view that there may be an alternative pathway in cones. ⁷¹  

Studies in Lrat^−/− mice have identified two interventions that may restore retinal function. ⁵ Intraocular gene therapy and oral pharmacologic treatments with the pro-drugs 9-cis-retinyl acetate and 9-cis-retinyl succinate have led to improvements in electroretinographic and pupillary responses. Levels of visual pigment were also improved in treated mice. Such pharmacological treatment, at least in animal models, appears to be safe if administered on multiple occasions; and several low-dose treatments may show cumulative effects. Based on these studies, clinical trials for the treatment of human LRAT and RPE65 retinopathy with oral agents are underway. Initial reports suggest that administration of the oral retinoid QLT091001 is safe and may be associated with improvement in visual acuity and kinetic visual fields in patients with LCA. ⁶  

The present study described in detail the phenotype associated with patients with severe progressive EORD and variants in LRAT. Despite significant visual dysfunction from childhood, there is residual photoreceptor function in adulthood, which suggests that the window of opportunity for therapeutic intervention may extend into late childhood. 

Footnotes

 Supported by grants from the Foundation Fighting Blindness (USA), Fight for Sight, Moorfields Eye Hospital Special Trustees, the Biotechnology and Biological Sciences Research Council (BBSRC), the National Institute for Health Research UK to the Biomedical Research Centre for Ophthalmology based at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology (London, UK), and the National Institute for Health Research UK to the Manchester Biomedical Research Centre.

Footnotes

 Disclosure: A. Dev Borman, None; L.A. Ocaka, None; D.S. Mackay, None; C. Ripamonti, None; R.H. Henderson, None; P. Moradi, None; G. Hall, None; G.C. Black, None; A.G. Robson, None; G.E. Holder, None; A.R. Webster, None; F. Fitzke, None; A. Stockman, None; A.T. Moore, None