Autosomal Dominant Retinal Degeneration and Bone Loss in Patients with a 12-bp Deletion in the CRX Gene

Abstract

purpose. To define the phenotypic expression of a deletion in the gene encoding the transcription factor CRX in a large, seven-generation, white family.

methods. Fourteen affected individuals, all heterozygous for the Leu146del12 mutation in the cone–rod homeobox gene (CRX), and four nonaffected relatives from the same family were examined with visual function tests, and 10 underwent bone mineral density (BMD) measurement.

results. The ability of the mutated CRX protein to transactivate rhodopsin promoter was decreased by approximately 25%, and its ability to react synergistically with neural retinal leucine zipper (NRL) was reduced by more than 30%. The affected members of the family had an autosomal dominant ocular condition most closely resembling Leber congenital amaurosis (LCA) with severe visual impairment at an early age. Depending on age, affected members showed varying degrees of significant visual acuity loss, elevated dark-adaptation thresholds, significantly reduced cone and rod electroretinogram (ERG) amplitudes, and progressive constriction of the visual fields, in most cases leading to complete blindness. Six affected members had reduced levels of BMD in the spine and the hip (osteopenia). Four affected female members who were receiving long-term hormonal replacement therapy (HRT) demonstrated normal values of BMD.

conclusions. This large deletion of the CRX gene is associated with a severe form of autosomal dominant retinal degeneration. Affected members not receiving HRT showed reduced BMD (osteopenia). This phenotype may reflect the abnormal influence of mutant CRX on both retinal and pineal development.

Leber congenital amaurosis (LCA, Mendelian Inheritance in Man 204000) was described originally by Theodore Leber in 1869 ¹ as pigmentary retinopathy with congenital amaurosis. He also was the first to point out the familial nature of the condition and the role of consanguinity. ² Later, he recognized the existence of two forms, one in which blindness occurred in early infancy and a second, juvenile form, which manifested in puberty and did not always lead to blindness. ³  

Now it is universally recognized that LCA exhibits a wide range of clinical and genetic heterogeneity. ⁴ Mutations in six different genes have been associated with LCA, and four of those genes have been cloned. ⁵ The unknown genes are localized on chromosomal regions 6q11-q16 ⁶ and 14q24. ⁷ Three of the cloned genes include (in order of chromosomal localization): one on chromosome 1q31, the gene encoding a retinal pigment epithelium (RPE)–specific 65-kDa protein (RPE65) ⁸ and two within the chromosomal region 17p13.1. The first gene is the locus for retinal cyclic guanosine monophosphate (cGMP), producing enzyme guanylate cyclase (GUCY2D, former abbreviation RetGC), ⁹ and the second is the recently identified arylhydrocarbon-interacting receptor protein-like 1 gene (AIPL1). ^{10 11}  

The locus for a fourth cloned gene for LCA is 19q13.3, the locus encoding the cone–rod homeobox gene (CRX). In recent years, several different alterations in the structure of the CRX gene have been associated with cases of cone–rod dystrophy and LCA. ^{12 13 14 15 16}  

CRX is a photoreceptor-specific, OTX-like, homeobox gene, that regulates many proteins essential for the normal structure and function of the photoreceptor outer segments. ^{12 17 18 19} It has been demonstrated that CRX is also expressed in the pineal gland and that expression in the gland follows a daily light–dark cyclic rhythm. ^{20 21}  

LCA has been associated primarily with an autosomal recessive mode of inheritance. ²² However, dominant pedigrees have been described. ^{23 24} One of the authors (JH) published his initial observations on a dominantly inherited condition in a branch of a unique, seven-generation family with a rare, autosomal dominant, early-onset retinal degeneration that has features of LCA. ²⁵ This family was subsequently found to have a 12-bp deletion in the CRX gene that results in an in-frame deletion of residues 147-150. ²⁶ Because during subsequent visits family members reported a high rate of osteoporosis in relatives, we decided to measure the bone mineral density (BMD) in affected and nonaffected family members. 

In this report, we present an extended and updated description of this family, including measurements of visual function over 16 years of follow-up and BMD measurements in the hip, spine, and forearm. 

Materials and Methods

Subjects

All subjects were evaluated by medical, ocular, and family history and clinical ophthalmic examination. Five of them (individuals III:3, IV:2, IV:3, V:2, and V:4 in branch RFS900_I) were tested at Jules Stein Eye Institute (Los Angeles, CA) in the late 1980s. ²⁵ Those individuals have been re-evaluated during the past 3 years at the Retina Foundation of the Southwest (Dallas, TX). The proband (V:4) was initially seen at age 11 at the University of California Los Angeles and was subsequently seen at age 27 at the Retina Foundation. In addition to the original five patients, nine more affected family members also were examined at the Retina Foundation. Proceedings followed the tenets of the Declaration of Helsinki and were approved by the appropriate institutional review boards. 

Mutation Analysis

The details of the mutation analysis have been published elsewhere. ²⁶ Briefly, DNA was isolated from peripheral blood by a DNA extraction kit (Puregene; Gentra, Minneapolis, MN). Single-strand conformational polymorphism (SSCP) analysis was performed, and a second amplification was performed from a stock DNA sample, as a template for sequencing. The fragment was treated with shrimp alkaline phosphatase and exonuclease (US Biochemical, Cleveland, OH) and was sequenced with a kit (AmpliCycle; Perkin Elmer, Norwalk, CT). To confirm deletion size and location, the amplimer was also subcloned using a cloning kit (PCR-Script Amp; Stratagene, La Jolla, CA), and individual clones were sequenced using vector-specific primers. Subsequent samples from family members were either screened for the mutation by SSCP, or by PCR amplification and separation on 5% 3:1 agarose gels (NuSieve; BioWhittaker, Walkersville, MD). 

Transient Transfection Studies

A mammalian expression construct for the CRX Leu146del12 deletion mutant was generated from a human CRX_pcDNA3.1/HisC vector ¹⁹ using the a site-directed mutagenesis kit (QuickChange; Stratagene) according to the manufacturer’s instructions. The primers used for mutagenesis were 5′-CAGGTTTGGTTCAAGAACTGGAGGGCTAAATGCAGGC-3′ (forward primer) and 5′-GCCTGCATTTAGCCCTCCAGTTCTTGAACCAAACCTG-3′ (reverse). The mutation was confirmed by sequence analysis. 

Calcium phosphate–mediated transfections and luciferase andβ -galactosidase assays were performed as previously described, ¹⁹ except that transfections with glycerol shock were performed with 50% confluent, 10-cm plates of 293 cells grown in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum, and 1% penicillin-streptomycin (Gibco, Grand Island, NY). Cells were harvested 48 hours after transfection. Each transfection experiment was performed in triplicate. Aliquots of pcDNA-CRX expression construct varying from 0.1 to 1.0 μg were cotransfected with bovine rhodopsin promoter-luciferase reporter (pBR130-luc; 5.0μ g), with and without the Nrl expression plasmid pED-bNrl (1.0 μg). ²⁷ The plasmid pCMV-LacZ was included to normalize for transfection efficiency. 

Visual Function Testing

Ophthalmic examination included best corrected visual acuity, direct and indirect ophthalmoscopy and fundus photography. Visual acuity was tested on the Early-Treatment Diabetic Retinopathy Study (ETDRS) chart whenever possible. If necessary, the Distance Test Chart for the Partially Sighted (Designs for Vision, Ronkonkoma, NY), or forced-choice preferential looking tests ²⁸ were used. Dark-adapted thresholds with an 11° test were obtained on a Goldman–Weekers adaptometer (Haag-Streit, Berne, Switzerland) after 45 minutes of dark adaptation. Because patients were unable to see a fixation light, they were allowed to use any strategy to detect the test target. 

Standard full-field electroretinograms (ERGs) were elicited by methods previously described. ²⁹ Whenever detectable, rod and cone a-waves were recorded according to an established protocol. ³⁰ Briefly, a set of white flashes (2–4-log scotopic trolands [scot td]) were first presented in the dark. The cone a-waves, elicited by the same stimuli presented against a rod-saturating background, were then subtracted from the dark-adapted responses to produce rod-only a-waves. Rod-only responses and cone-only responses were fit with a computational model. ³¹  

Bone Mineral Densitometry

We measured BMD by dual-energy x-ray absorptiometry (DEXA) at various skeletal sites (lumbar spine, total hip, femoral neck, and forearm; QDR-2000; Hologic, Waltham, MA). One subject with previous spinal fusion had only hip and femoral neck BMD tested. One subject had body weight in excess to the maximum recommendations for the QDR-2000 and was tested only for BMD on the forearm. BMD was recorded in grams per square centimeter. For all locations, z-scores were calculated (as SD of BMD compared with age- and sex-matched control data from the Hologic database). BMD was expressed as a z-score and compared with the reference population of the Hologic database. Additional correction for the femoral neck analysis (provided by Hologic in 1997) was used. Four postmenopausal women had received hormonal replacement therapy (HRT) for at least 5 years. Because HRT can influence bone density, ³² these patients were analyzed separately. Differences between mean z-scores at each site in HRT-treated and untreated patients were assessed by Student’s t-test. 

Results

Pedigree and History

We were able to trace the origins of the family up to seven consecutive generations with more than 250 individuals (Fig. 1) . Labels were assigned to branches of the family in order of the age of siblings in generation II. This report focuses on individuals from three branches of the family. From those, 25 individuals were contacted, and 20 agreed to participate in the study. 

The mode of disease inheritance is clearly autosomal dominant. The pedigree has been traced back to a couple living during the second half of the 19th century in Oklahoma. According to census data, the husband was from mixed white and Native American ancestry. The wife (of white ancestry), who moved to Oklahoma from Missouri, had low vision during her adult life. According to family members, three of the nine children of this couple (generation I on Fig. 1 ) had low vision beginning in childhood. These individuals are represented by black-striped symbols in generation I of the subpedigrees on Figure 1 . There are no known cases of consanguinity in the pedigree. No systemic abnormalities apart from reduced BMD were reported in the family. There were no cases of mental retardation. 

Genetic Testing

The initial results from genetic testing of this family have been reported. ²⁶ Recently, DNA samples were obtained from an additional eight affected and six unaffected family members. Segregation of the 12-bp deletion with retinal degeneration in these samples was confirmed by PCR amplification of CRX amplimer 3b, followed by separation on 5% 3:1 agarose gels, as demonstrated in Figure 2 . 

Transactivation Activity

To explore the mechanism by which the CRX Leu146del12 mutation leads to retinal degeneration, we tested the activity of mutant CRX in a transient transfection-based transactivation assay. ¹⁹ Compared with wild-type CRX, the CRX Leu146del12 mutant demonstrated an ∼25% decrease in its ability to transactivate a bovine rhodopsin promoter–reporter construct (Fig. 3A ). It retained the ability to act synergistically with the bZIP transcription factor Nrl, ^{19 33} but its ability to do so was reduced 30% to 40% compared with wild-type CRX (Fig. 3B) . 

Fundus Appearance

Typical fundus photographs from three affected members of the family are presented in Figure 4 . The youngest family member (RFS900_I, VI:1; Fig. 4A ) showed irregular macular reflex and minimal changes in the RPE. Fundus photography of the left eye of her mother at age 27 (RFS900_I, V:2; Fig 4B ) demonstrated widespread atrophy of the RPE, irregular macular reflex, and mild vessel attenuation. Another family member at age 45 (RFS900_I, IV:3; Fig. 4C ) showed pallor of the optic disc, moderate arteriolar attenuation, diffuse RPE atrophy and a bull’s-eye lesion in the macula. 

Psychophysical and Electrophysiological Testing

Clinical results from affected family members are shown in Table 1 . Severely reduced visual acuity was documented at age 2. Visual acuity loss was gradual over the years, but in most cases ultimately led to light perception or total blindness. The only case of cone-mediated central visual acuity (better than 20/100) was observed in the left eye of RFS900_I V:4. 

Determination of the dark-adapted threshold was possible in half the affected members (7/14). With one exception (RFS900_C IV:3), the threshold was elevated by at least 4 log units, even at 8 years of age. No visual field was measurable in eight patients. The remaining six had an overall constriction of the visual field even at a young age (∼30° field with the I-4e ispoter at ages 7–11 years). In some patients (RFS900 I, V:2), the size of the visual field was even more reduced. 

Affected members of the family had greatly reduced ERG rod responses, severely attenuated cone responses (>95% amplitude loss), and delayed 30-Hz flicker responses. Patients aged more than 15 years had nondetectable or greatly reduced single cone responses. Typical ERG responses are presented in Figure 5 . 

In one of the branches, there was an exception to the general trend (IV:3, RFS900_C). The psychophysical and electrophysiological measurements in this patient revealed relatively preserved rod and cone function even at age 42, when most of the other members barely retained the ability to detect light. Although rod and cone responses were decreased in amplitude and delayed, they were an order of magnitude higher than these in any other family member, regardless of age. Furthermore, this was the only patient in the pedigree in whom a-waves were detectable with high-intensity stimulus. An analysis of the leading edge of the a-wave revealed that log S (an indication of phototransduction efficiency) was within 0.3 log units of normal, whereas log RmP3 (the maximum photoresponse amplitude) was decreased by 0.6 log units. Cone a-waves to high-intensity stimuli were not detectable (Fig. 6) . 

Bone Mineral Density

The results from the DEXA BMD and z-scores are summarized in Table 2 . Affected family members are separated into two groups: affected members without HRT and affected members with HRT. At three of four locations tested, patients without hormonal treatment had results that were more than 1 z-score below mean normal. According to the established diagnostic criteria of the World Health Organization, ³⁴ a low bone mass condition (osteopenia) is present when BMD is more than 1 SD but less than 2.5 SD below the young adult mean. For scientific purposes, the use of the age-corrected z-scores is preferable. Overall, BMD in the hip and the spine was lower in the affected members than in age-matched normal subjects or members receiving HRT, whereas BMD was not lower in the forearm (Table 2 , Fig. 7 ). The normal bone mineral content and its relationship with age differed in both genders and at different measurement sites. Changes in BMD versus age for the two most illustrative measurement locations (total spine and total hip) are demonstrated separately in Figure 7 for each gender group. 

From the data in Figure 7 , it is clear that there is a tendency for a faster than normal decrease in BMD with the age of the affected individuals. This is most evident for the BMD changes in males at the hip, where BMD loss with age of the affected individuals increased more than twice as fast with age, compared with the normal group. The BMD loss was −0.08 g/cm² per decade for the affected males, compared with −0.03 g/cm² average loss per decade for the normal male population (P < 0.0001). 

Discussion

The history of decreased vision since birth, presence of nystagmus, normal shape of the cornea, progressive visual field loss, abnormal dark adaptation, diffuse RPE changes in advanced stages and nonrecordable or greatly reduced ERG responses found in this family are consistent with the diagnosis of LCA. Diagnostically, with all the clinical findings considered, the nonrecordable or greatly reduced ERG responses in all cases differentiates them as affected by LCA rather then early-onset classic retinitis pigmentosa or cone–rod degeneration. The only feature that is not consistent with the diagnosis of LCA is the mode of inheritance. However, mutations in CRX have been associated with autosomal dominant inheritance in cases with cone–rod dystrophy. ^{17 26 35} Furthermore, two cases of de novo mutations in the CRX gene were associated with LCA, ^{13 14} suggesting new autosomal dominant mutations, and four other LCA pedigrees have been described with an autosomal dominant mode of inheritance. ^{23 24 36}  

Variable phenotypic expressivity was demonstrated in the described pedigree. By electrophysiological and psychophysical criteria, one case of incomplete penetrance (IV:3, RFS900_C) was present. In two other cases (V:4, RFS900_I and IV:6, RFS900_C), visual acuity was relatively preserved beyond puberty. Those findings are consistent with other examples of variation in the CRX-LCA phenotype. ¹⁶  

The actual molecular mechanism by which the CRX Leu146del12 deletion leads to retinal degeneration remains unclear. The four amino acids that are deleted by the Leu146del12 mutation are conserved among the murine, bovine, and human CRX proteins. Although the region affected by the mutation is outside the DNA-binding homeodomain region and the OTX tail, its proximity to the WSP motif suggested that it may significantly affect the protein’s biologic activity. Consistent with this possibility, we found that the deletion led to a 25% to 40% decrease in transactivation activity. Presumably, this transient transfection-defined abnormality translates in vivo into an alteration of photoreceptor gene expression that directly or indirectly causes retinal degeneration. In relation to this it should be noted that mutations that lead to increases in transactivating activity, as well as those that lead to decreases, could lead to photoreceptor degeneration. ³⁷ Interesting questions remain about how different mutations in CRX and other transcription factors can lead to different clinical phenotypes, and how, as shown in this study, even a single mutation can be associated with substantial clinical heterogeneity. 

An unexpected result in our study was the finding of reduced axial BMD in affected members of the RFS900 family, without other systemic abnormalities. Such an observation has not been described previously. However, osteopenia does not manifest obvious clinical symptoms and requires specialized testing to be detected. Therefore, we cannot exclude the possibility that families with other reported CRX mutations demonstrate the same abnormality. 

The tendency for osteopenia in this family is suggestive, but we were unable to demonstrate a direct link with the CRX mutation. A major confounding factor was HRT for at least 5 years in three of the five affected female members of the family who were tested for BMD. Several studies have shown that HRT has a substantial beneficial effect in preventing bone loss in postmenopausal women. ^{32 38 39} Such an effect is comparable to the difference that we observe between HRT-treated and nontreated affected family members. 

Although it is possible that a reduced level of moderate exercise can influence bone mass acquisition, ^{40 41 42} this relationship is still uncertain. ⁴³ It cannot be ruled out that, in patients with LCA, the reduced exercise level due to very low visual acuity may affect the buildup of bone mass. There is an association between low visual acuity and increased risk of hip fracture, but it can be due to other factors, such as poor visuomotor control, besides reduced levels of exercise. ⁴⁴ The effect of reduced exercise in patients with LCA and other blinding disorders on BMD has not been studied systematically and at present is unknown. Alternatively, lower bone density could involve altered pineal function. It has been shown that CRX is expressed in the pineal gland and can regulate pineal gene expression in vitro. ^{20 21}  

Melatonin (MT) is the major secretion product of the pineal gland. Totally and partially blind (light perception only) persons demonstrate increased daytime level of MT secretion ⁴⁵ and phase-advanced or phase-delayed rhythm. ⁴⁶ We did not have the opportunity to study MT levels in our patients. However, mice with a null mutation in the CRX gene demonstrated reduced expression of several pineal genes and altered photoentrainment. ⁴⁷ Therefore, it is conceivable that there was a substantial alteration in the MT secretion in the RFS900-affected family members due to abnormal pineal function in addition to any alteration due to blindness alone. Studies have shown that MT affects calcium blood levels and homeostasis, ^{48 49} which are important factors for maintaining bone mineral content. There is also recent direct evidence that MT stimulates proliferation and synthesis of type I collagen in human bone cells in vitro ⁵⁰ and that oral administration of MT increases bone mass in young growing mice. ⁵¹ It also has been demonstrated that MT directly promotes osteoblast differentiation and bone formation. ⁵² Even before the publication of these findings, a relationship between osteoporosis and altered pineal function had been proposed based on indirect evidence. ⁵³ Our findings are consistent with the possibility that altered function of the pineal gland resulting from the CRX mutation may be the cause of faster than normal age-related bone loss (especially in the hip) in the affected family members. 

 Supported by National Institutes of Health Grants EY05235 and EY07142 and by grants from the Foundation Fighting Blindness with assistance from the RGK Foundation, the George Gund Foundation, and the Hermann Eye Fund.

 Submitted for publication September 28, 2000; revised January 3, 2001; accepted January 22, 2001.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked“ advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

(A) Pedigree of RFS900 family. Circular drawing of the whole family. Roman numerals indicate the number of generations. (B, C, and D) Abbreviated presentations of the three branches of the RFS900 family with affected living members. Solid line crossing the symbol diagonally indicates a deceased person. Vertical black band crossing the symbol indicates that individual is affected by hearsay. Numbering is independent for each branch and starts at generation II of the main pedigree.

Figure 2.

 

Part of branch RFS900_I (separated with a frame on Fig. 1D ) and PCR amplification of CRX amplimer 3b, followed by separation on 5% agarose gel for the selected family members.

Figure 3.

 

The Leu146del12 deletion mutation reduces the ability of CRX to transactivate the rhodopsin promoter, both alone and in combination with NRL. (A) Ability of the indicated amounts (μg) of wild-type (CRX/pcDNA3.1) and CRX^146L-12bp/pcDNA3.1 mutant constructs to transactivate expression of the bovine −130- to +70-bp rhodopsin promoter–luciferase reporter construct in a transient transfection assay. Data shown are means ± SD. (B) Same experiment as shown in (A), but in the presence of 1 μg of NRL expression plasmid (pED-bNrl).

Figure 4.

 

Fundus photographs from members of RFS900 family showing attenuation of retinal vessels with age. (A) Posterior pole of the left eye (RFS900_I, VI:1) at age 9 (central whitening is a flash-reflection artifact); (B) posterior pole of the left eye (RFS900_I, V:2) at age 27; and (C) posterior pole of the left eye (RFS900_I, IV:3) at age 45.

Table 1.

 

Basic Clinical Findings in Family Members Carrying the Leu146del12 CRX Mutation

	Age at Initial Visit (y)	Follow-up (y)	Refraction (D)	Visual Acuity (Initial–Last Visit)	Nystagmus Since Birth	Fundus Appearance	Visual Field, † (at Age, y)	DA Threshold, ‡ at Last Visit
RFS900_C
IV:3	42	12	OD: −5.00+ 2.0× 70 OS:−4.00+ 2.0× 70	OD 20/300–20/320 OS 20/225–20/320	+	Localized macular and peripapillar atrophy, bone spicule formation	OD: 15 deg (42) OS: 45× 50 deg (42)	2.5 log masb
V:1	32	—	OD:−7.75+ 1.25× 57 OS:−9.25+ 2.75× 95	OD 20/800 OS 20/800	+	Vessel attenuation, mild to moderate granular changes of RPE, early atrophy in the macula	ND	6.6 log masb
IV:6	40	—	OD:+1.5+ 0.75× 86 OS:+1.25+ 1.25× 105	OD 20/320 OS 20/320	+	Localized macular and perimacular RPE atrophy, bone spicule formation in those areas	ND	5.7 log masb
III:15	71	—	Mild myopia (< −3.0)	PD LP OS LP		Diffuse RPE atrophy, bone spicule formation, moderate vessel attenuation, macular atrophy	ND	N/A
RFS900_E
IV:8	50	—	OD: +1.25+ 2.75× 98 OS:+0.25+ 2.25× 87	OD 5/350 OS 5/350		Diffuse RPE atrophy, bone spicule formation, moderate vessel attenuation, macular atrophy	ND	6.6 log masb
III:11	65	—	Mild hyperopic	OD NLP OS NLP	+	Moderate vessel attenuation, macular atrophy, isolated areas of bone spicule formation	ND	N/A
IV:16	25	16	N/A	OD HM–NLP OS CF1–NLP	+	N/A (subcapsular cataract in both eyes)	ND	N/A
RFS900_I
III:3	52	—	OD: −3.0 OS:−3.5	OD LP–NLP OS LP–NLP	+	Diffuse RPE atrophy, areas of bone spicule formation, moderate vessel attenuation	ND	N/A
IV:2	35	16	OD:−1.75+ 1.0× 135 OS:−1.25	OD LP–NLP OS LP–LP	+	Generalized depigmentation of the RPE with granularity, bull’s-eye lesion in the macula	ND	N/A
IV:3	30	16	OD:−4.75+ 1.75× 85 OS: −2.5+ 1.0× 75	OD 20/100–LP OS 20/200–LP	+	Generalized depigmentation of the RPE with granularity, bull’s-eye lesion in the macula	OD: 8 deg (30) OS: 12 deg (30)	N/A
V:2	11	16	OD:+3.5+ 2.0× 105 OS: +4.75+ 2.0× 75	OD CF3–NLP OS CF3–NLP	+	Vessel attenuation, mild to moderate granular changes of RPE, early atrophy in the macula	Temporal islands OU (11)	N/A
V:4	11	16	OD:−2.0 OS: N/A	OD 20/20–20/320 OS 20/60–20/80	+	Attenuated vessels, irregular macular reflex, pigment epitheliopathy outside the posterior pole	OD: 25× 15 deg (11) OS: 20 deg (11)	5.6 log masb
VI:1	5	2	OD:+3.75+ 1.25× 104 OS:+4.00+ 1.25× 90	OD 3/300–3/300 OS 3/225–3/350	+	Paramacular RPE atrophy	OD: 25× 30 (7) OS: 25× 30 (7)	5.7 log masb
VI:2	2	6	OD:+3.50+ 1.0× 90 OS:+2.50+ 1.0× 96	OD 20/290*–3/180 OS 20/290*–3.140	+	Initial signs of bull’s-eye maculopathy	OD: 30× 40 (8) OS: 30× 40 (8)	5.2 log masb

 CF1, count fingers at 1 foot; CF3, count fingers at 3 feet; HM, hand movement; LP, light perception; NLP, no light perception; ND, nondetectable; N/A, not applicable.

*  Determined by preferential looking testing.

†  Isopter IV-4.

‡  Upper limit of normal is 2.0 log microapostilbes (masb).

Figure 5.

 

Full-field rod and cone responses from one nonaffected (RFS900_C, IV:4) and two affected members of the RFS900 family. Pedigree localization, gender, and age at the time of the testing are indicated above each column of recordings. Horizontal rows show computer-averaged responses. Scotopic blue: a rod response to blue test; 30-Hz flicker: cone responses to a rapidly repeated stimulus (30-Hz white flicker). Superimposed spikes indicate stimulus timing.

Figure 6.

 

Standard full-field ERGs (A) and high-intensity a-wave series (B) from the affected family member RFS900_C IV:3. See Figure 5 for details.

Table 2.

 

BMD and z-Scores from Adult Patients Carrying the Leu146del12 CRX Mutation

	Spine BMD Spine z-Score	Total Hip BMD Total Hip z-Score	Femoral Neck BMD Femoral Neck z-Score	Total Forearm BMD Total Forearm z-Score
CRX, no HRT	0.932 ± 0.095	0.786 ± 0.152	0.689 ± 0.122	0.630 ± 0.071
	−1.32 ± 0.75*	−1.40 ± 0.78*	−1.24 ± 0.70*	−0.37 ± 0.99
CRX, HRT	1.112 ± 0.153	0.902 ± 0.060	0.848 ± 0.090	0.552 ± 0.068
	1.35 ± 1.08	0.35 ± 0.36	0.93 ± 0.86	0.42 ± 0.97

 Data are expressed in mean grams per square centimeter ± SD.

*  P < 0.05; significant difference from BMD in normal subjects.

Figure 7.

 

BMD measurements in the spine (A, men; C, women) and hip (B, men; D, women) from affected members of the RFS900 family. Solid black line: mean values of the BMD at a certain age according to the database (Hologic, Waltham, MA). Dashed lines: BMD 1 SD below and above the mean value. (B) Dashed-dotted line: Linear regression to the values for the affected family members.

The authors thank Kirsten Locke, Angela Peters, and Dianna Wheaton–Hughbanks for excellent technical assistance; Peggy Boyd–Deguay for performing the BMD tests; the Morchower Pediatric Eye Research Laboratory, Retina Foundation of the Southwest, for performing the preference looking testing; Tom Kelly from Hologic Inc. for kindly providing reference BMD data; and Tracy Neally and Sharon Aston from the National Archives in Oklahoma for collaborating on questions regarding genealogy.