Autosomal dominant retinitis pigmentosa in a large family: a clinical and molecular genetic study.

Abstract

To characterize the pedigree, visual function phenotype, and responsible mutation in a large family with autosomal dominant retinitis pigmentosa.
Pedigree data were obtained by personal interviews and corroborated with community records. One hundred twenty-eight members of the family were examined clinically, and a subset of 12 affected subjects was further studied with dark- and light-adapted static perimetry and electroretinography. The coding region of the rhodopsin gene was polymerase chain reaction (PCR) amplified and resolved by denaturing gradient gel electrophoresis. Genomic DNA samples from nine affected and five unaffected family members were analyzed by PCR amplification and restriction enzyme digestion.
A 14-generation pedigree was identified in which retinitis pigmentosa (RP) was inherited in an autosomal dominant fashion. Affected individuals reported early night blindness and showed vessel attenuation and bone spicule-like pigmentary changes. In these individuals, the rod electroretinogram (ERG) was not detectable, and the cone ERG was reduced in amplitude and delayed in timing. With dark-adapted perimetry, rod function could be detected in only one young patient, and it was markedly abnormal. Light-adapted perimetry indicated that cone sensitivity could be relatively well preserved in the central field, but it was diminished in the periphery even in the most mildly affected subjects. A valine345-to-leucine mutation was identified in the rhodopsin gene and shown to cosegregate in the heterozygous condition with the disease.
The natural history of RP in this family begins with a loss of rod function, progresses to involve the cone system, and leads eventually to a severe loss of visual function. The invariance of valine345 in all functional vertebrate visual pigments sequenced to date, and the unusually conservative nature of the valine345-to-leucine mutation suggests that the carboxy terminus of rhodopsin is involved in a highly specific interaction with one or more rod proteins.

Full text

Page 1

Articles
Autosomal Dominant Retinitis Pigmentosa in a Large
Family: A Clinical and Molecular Genetic Study
Delia J. Rosas* Alejandro]. Romany Pedro Weissbrod,*\\
Jennifer P. Macke,% and Jeremy Nathans%%
Purpose. To characterize the pedigree, visual function phenotype, and responsible mutation in
a large family with autosomal dominant retinitis pigmentosa.
Methods. Pedigree data were obtained by personal interviews and corroborated with commu-
nity records. One hundred twenty-eight members of the family were examined clinically, and a
subset of 12 affected subjects was further studied with dark- and light-adapted static perimetry
and electroretinography. The coding region of the rhodopsin gene was polymerase chain
reaction (PCR) amplified and resolved by denaturing gradient gel electrophoresis. Genomic
DNA samples from nine affected and five unaffected family members were analyzed by PCR
amplification and restriction enzyme digestion.
Results. A 14-generation pedigree was identified in which retinitis pigmentosa (RP) was in-
herited in an autosomal dominant fashion. Affected individuals reported early night blindness
and showed vessel attenuation and bone spicule-like pigmentary changes. In these individuals,
the rod electroretinogram (ERG) was not detectable, and the cone ERG was reduced in ampli-
tude and delayed in timing. With dark-adapted perimetry, rod function could be detected in
only one young patient, and it was markedly abnormal. Light-adapted perimetry indicated that
cone sensitivity could be relatively well preserved in the central field, but it was diminished in
the periphery even in the most mildly affected subjects. A valine345-to-leucine mutation was
identified in the rhodopsin gene and shown to cosegregate in the heterozygous condition with
the disease.
Conclusions. The natural history of RP in this family begins with a loss of rod function, pro-
gresses to involve the cone system, and leads eventually to a severe loss of visual function. The
invariance of valine345 in all functional vertebrate visual pigments sequenced to date, and the
unusually conservative nature of the valine345-to-leucine mutation suggests that the carboxy
terminus of rhodopsin is involved in a highly specific interaction with one or more rod pro-
teins. Invest Ophthalmol Vis Sci. 1994;35:3134-3144.
JVetinitis pigmentosa (RP) is a clinically and geneti-
cally heterogeneous group of disorders in which de-
fects in rod function are followed by visual field loss
and eventual blindness.1'2 Mutations in the genes en-
coding rhodopsin,3"5 peripherin/rds,67 and the beta
From the *Fundacion Samuel Kratz, Buenos Aires, Argentina; the •fUniversidad de
Buenos Aires, Buenos Aires, Argentina; and the Departments of%Molecular Biology
and Genetics, %Neuroscience, and ^Ophthalmology, Howard Hughes Medical
Institute, The Johns Hopkins University School of Medicine, Baltimore, Maryland.
|| PW, xuho started the study of this family, died in 1990.
Supported try the Fundacion Samuel Kratz, the Howard Hughes Medical Institute,
and the National Retinitis Pigmentosa Foundation (Baltimore, Maryland).
Submitted for publication December 14, 1993; revised February 14, 1994; accepted
February 15, 1994.
Proprietary interest category: N.
Reprint requests: Dr. Jeremy Nathans, 805 PCTB, The Johns Hopkins University
School of Medicine, 725 N. Wolfe Street, Baltimore, MD 21205.
subunit of cGMP phosphodiesterase8 have been de-
scribed in different subsets of patients with RP.
Among patients with rhodopsin gene mutations, func-
tional tests indicate that the principal defect is local-
ized to the rods.9"12 In these patients, cone function
can also be affected, presumably secondary to a loss of
rod viability.
More than 45 mutations in the rhodopsin gene
have been identified in patients with RP, and all but
one of these mutations are associated with autosomal
dominant RP (ADRP; reviewed in refs. 2 and 13). Most
of these mutations result in single amino acid substitu-
tions. Thirty-four of the mutant proteins have been
expressed in a tissue culture system and analyzed with
respect to yield, subcellular localization, and ability to
3134
Investigative Ophthalmology & Visual Science, July 1994, Vol. 35, No. 8
Copyright © Association for Research in Vision and Ophthalmology

Page 2

Rhodopsin Val-345-Leu in ADRP
3135
bind 11-cis retinal.1415 These tissue culture studies re-
vealed two classes of mutant proteins: 15% of the mu-
tant rhodopsins were observed to be indistinguishable
from the wild type, whereas 85% were defective in fold-
ing, stability, or both. One rhodopsin mutation in
which lysine206, the site of retinal attachment, was
changed to glutamate was found to be stable but con-
stitutively active.16 One of the folding and/or stability
mutants, proline23-to-histidine, has been found to
produce a progressive retinal degeneration when ex-
pressed in transgenic mice.17 Interestingly, among the
15% of mutants that resemble the wild type, most are
clustered at the extreme carboxy terminus. Recent
transgenic experiments with one of the carboxy-termi-
nal mutants, glutamine344-to-stop, have shown that
this region is involved either in the transportation of
rhodopsin to the outer segment or its retention in the
outer segment (Sung et al, submitted for publication,
1994). This observation suggests that this region of
rhodopsin is specifically recognized by one or more
cellular proteins and that other carboxy-terminal mu-
tants may also be defective in this process.
One goal of recent research on RP has been to
correlate the different clinical phenotypes with the
corresponding molecular genetic defects. Such geno-
type-phenotype correlations are useful in improving
the accuracy of individual prognoses and in elucidat-
ing the pathophysiological mechanisms involved in
RP. In the long term, the availability of large numbers
of well-characterized subjects will allow a systematic
evaluation of secondary factors, both genetic and envi-
ronmental, that alter the severity or progression of the
disease. With respect to the analysis of genetic modi-
fiers, a prerequisite for such a study is the identifica-
tion of large families that are well characterized clini-
cally and genetically. In the present study, we report
on a 14-generation family with 484 living members in
which rhodopsin mutation valine345-to-leucine causes
ADRP.
METHODS
Human Subjects
The pedigree was assembled by on-site personal inter-
views during a 3-year period. These data were corrobo-
rated where possible by comparison with community
records. For 128 subjects, the affected-unaffected
status was determined by clinical examination on site,
in the clinic, or both. For other members of the pedi-
gree, the affected-unaffected status was determined
based on the assessment of their relatives or ophthal-
mologists. Blood specimens were obtained from 14
members for genetic studies. Informed consent was
obtained after an explanation of the nature and possi-
ble consequences of the study, which followed the ten-
ets of the Declaration of Helsinki.
Clinical Examination
The appearance of the fundus was studied using direct
and indirect ophthalmoscopy. The age of onset of
night blindness was obtained from interviews with the
patients. Snellen visual acuity was measured using Bai-
ley-Lovie charts.18 Goldmann kinetic perimetry was
performed using size V (64 mm2 area) and I (0.25 mm2
area) white targets at intensity 4e (318 cd/m2) on a 10
cd/m2 white-adapting background. Targets were
moved from nonseeing into seeing areas. Kinetic vi-
sual field extents were obtained by digitizing the perim-
eter charts, mapping the extent of the visual field into
spherical coordinates, and calculating the solid angle
(stereo radians) subtended by the resulting polygon.
Results were then normalized and are reported as per-
cent of normal visual field extent.19
Dark- and light-adapted static perimetry was per-
formed with a modified automatic perimeter20 using a
size V (64 mm2 area) target and two test strategies: a
full-field test of 75 locations on a 12° grid and a pro-
file test spanning 60° of central field along the hori-
zontal meridian with 2° spacing. Two-color dark-
adapted perimetry was performed (after 45 minutes of
dark adaptation) with 500-nm and 650-nm stimuli.
The difference in sensitivity for the two stimuli indi-
cates which photoreceptor mediates detection.21 For
rod-mediated locations, rod sensitivity losses were cal-
culated by subtracting the patient data for 500 nm
from mean values with this stimulus obtained from an
age-matched control group (N = 10). Light-adapted
perimetry was performed with a 600-nm target on a 10
cd/m2 white background. Cone sensitivity losses were
obtained by comparison with data from the control
group. Gray scale maps for rod and cone sensitivity
losses were constructed using 16 levels of gray, corre-
sponding to 0 to 54 dB for rods and 0 to 30 dB for
cone.
Full-field rod- and cone-isolated ERGs were elic-
ited and recorded with a computerized system (Nicolet
Compact Four with Ganzfeld GS2000; Nicolet Instru-
ment, Madison, WI), and using Burian-Allen bipolar
contact lens electrodes. When necessary, the signal-
noise ratio was improved by digitally averaging re-
sponses. For the dark-adapted ERG, the stimuli con-
sisted of blue (Wratten 47B; 0.14 log scot td s), red
(Wratten 26), and white flashes of light (1.85 log scot
td s/2.20 log scot td s). For the light-adapted ERG,
white light stimuli (1.85 log scot td s) were flashed at 1
Hz on a 17 cd/m2 white background. A response was
considered undetectable whenever a signal could not
be discerned from the background equivalent noise
after averaging several responses. After-average equiv-

Page 3

3136
Investigative Ophthalmology & Visual Science, July 1994, Vol. 35, No. 8
alent root-mean-square noise levels were <2.5 /xV for
the aforementioned responses. A cone flicker (29 Hz)
ERG was also performed using white light stimuli
(1.85 log scot td s) on a 17 cd/m2 white background.
DNA Analysis
Genomic DNA was purified from venous blood as de-
scribed.4 Seven segments of the rhodopsin gene were
amplified by the polymerase chain reaction (PCR) us-
ing the 14 primers listed previously.4 These seven seg-
ments encompass the coding region and approxi-
mately 30 base pairs of the intron sequence adjacent
to each of the five exons. Exons 1 and 4 were each
amplified as two partially overlapping segments; exons
2, 3, and 5 were each amplified as a single segment.
One PCR primer in each pair included a 40 base, gua-
nine- and cytosine-rich segment (a guanine- and cyto-
sine-clamp) to facilitate detection of mutations by de-
naturing gradient gel electrophoresis.22 PCR products
were resolved in acrylamide gels containing a 50% to
90% gradient of denaturant as described.2:<
The exon 5 PCR product that produced a variant
pattern by denaturing gradient gel electrophoresis was
amplified using a pair of primers carrying two differ-
ent restriction enzyme cleavage sites and cloned into a
plasmid vector. Nine independent clones were se-
quenced; the mutant sequence guanine52M-to-cytosine
was observed in four clones, and the wild type se-
quence was observed in five clones.
TABLE l. Pedigree Characteristics
Total number of members
Number living
Number unaffected
Number affected
Affected males
Affected females
Father-to-son transmissions
Fraction affected among the
offspring of affected subjects
Total
1355
484
294
139
68
71
46
0.47
Examined
Clinically
128
55
73
43
30
—
—
RESULTS
Clinical and Genetic Characteristics of the
Pedigree
Evidence of hereditary visual disease was found in two
large families whose Spanish ancestors settled in the
province of Cordoba, Argentina, about the year 1700.
A genealogical study traced both families to a common
ancestor, unifying the pedigree into a single family
tree of 1355 people, 484 of whom are alive. The pedi-
gree shows several marriages between family
members, and in three cases both husband and wife
are affected. Table 1 summarizes the characteristics of
the pedigree; the complete pedigree is available from
the authors upon request. The nearly equal number of
TABLE 2. Clinical Characteristics
Patient
P2
P3
1M
P5
P6
P7
P9
NO
Pll
PI 2
Age
(years)
31
29
28
26
66
40
18
16
38
30
Snellen
Acuity
RE: 20/30
LE: 20/40
RE: 20/40
LE: 20/800
RE: 20/32
LE: 20/32
RE: 20/50
LE: 20/62
RE: 20/800
LE: 20/800
RE: 20/25
LE: 20/25
RE: 20/25
LE: 20/25
RE: 20/25
LE: 20/25
RE: 20/60
LE: 20/40
RE: 20/20
LE: 20/20
Refraction
(diopters)
Piano -0.50
Piano -0.75
-2.0; -0.75
-0.50; -0.75
-0.50; -0.75
-1.0; -1.75
-0.50; -1.50
—
—
Piano -0.50
Piano -0.50
-LOO; -1.25
-0.50; -1.50
-0.75; -1.25
-0.75; -0.75
-0.75; -1.50
-1.00; -2.00
-0.50; -1.00
-0.50; -0.50
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
90
90
170
90
90
160
80
90
90
90
90
90
95
80
90
130
25
Appearance
:§n
:§n
I I
til
:§n
§n
II
II
:§n
:§n
Kinetic
Visual Field
Extent, V4e*
(% Normal)
55
25
14
4
U
64
Normal
68
35
6
Kinetic
Visual Field
Extent, I4e*
(% Normal)
2
2
2
2
U
5
19
16
5
2
Average Rod
Sensitivity Loss
(dB)f
>47
>47
(ND)
(ND)
U
>47
>47
36.1
(ND)
(ND)
Average Cone
Sensitivity Loss
(dB)f
15.7
14.8
(ND)
(ND)
U
13.3
8.1
8.9
11.6
(ND)
ND = Not determined; U = immeasurable due to severe visual loss.
* Expressed as a % of normal mean; 2 SD below normal mean equals 90% for V-4e target and 88% for I-4e target.
t Average sensitivity loss across the visual field for rods (dark-adapted perimetry) and cones (light-adapted peri met 17).
X Cystoid macular edema.
§ Posterior subcapsular lens opacity.
|| Diffuse four quadrant bone spicule-like pigmentary changes.

Page 4

Rhodopsin Val-345-Leu in ADRP
Sill
affected males and females, the numerous examples of
father-to-son transmission, the presence of at least one
affected parent for each affected subject, and the ap-
proximately 50% probability of each child of an af-
fected subject being affected indicate that this dis-
order is inherited as an autosomal dominant trait with
high penetrance.
One hundred twenty-eight people from different
branches of the family were studied clinically. Eight
older people who retained little visual function were
interviewed and examined ophthalmoscopically on
site. One hundred twenty younger people, 55 unaf-
fected and 65 affected, traveled to Buenos Aires for
clinical examinations. All 73 affected individuals who
were examined reported night blindness beginning in
the first decade of life. Indirect ophthalmoscopy
showed diffuse, four-quadrant, bone spicule-like pig-
mentary changes and vessel narrowing in all the af-
fected subjects. Cystoid macular edema was observed
in 26 affected subjects (35%), and slit-lamp examina-
tion showed central posterior subcapsular cataracts in
36 affected subjects (49%). Both findings were seen in
14 subjects (17%). Of the 10 affected family members
who were assessed with visual function tests, nearly all
retained good visual acuity despite varying amounts of
visual field loss (Table 2).
Dark- and Light-Adapted Perimetry
Using Goldmann kinetic perimetry, a consistent pat-
tern of disease progression was observed in which a
midperipheral ring scotoma was followed by loss of the
peripheral visual field. Altitudinal visual field defects
were not observed. Static perimetry was performed on
10 affected and two unaffected subjects from three
different branches of the pedigree (Fig. 1 and Table
2). Nine of the 10 affected subjects had no detectable
B O
FIGURE i. Branches of the pedigree showing those subjects
who underwent visual function tests (PI to PI2). These
three branches are five or more generations removed from
common ancestors. DNA samples were obtained from sub-
jects marked with an asterisk.
MMMMC /%#iCCCCCCCCCCCMMMMMMMMM
20 10 0 10 20
ECCENTRICITY [degrees]
S
36 -.
24 -
12 -
0 -
12 -
24 -
36 -
48 -
I
S
36 -i
24 -
12 -
0 -
12 -
24 -
36 -
48 -
T 72 60 48 36 24 12 0 ^Z 24 36 48 N
FIGURE 2. {Top) Two-color, dark-adapted static perimetric
profiles along the central 60° of the horizontal meridian for
the left eye of subject P10. Responses to stimuli of 500 nm
(circles) and 650 nm (triangles) show cone mediation (C) in
the center and mixed rod and cone mediation (M) else-
where. Dashed lines delimit the normal range (±2 SD) for
rod-mediated detection of a 500-nm stimulus. Hatched
areas correspond to the physiological blind spot. S = supe-
rior; I = inferior; N = nasal; T = temporal. (Bottom) Full-field
maps of rod (top) and cone (bottom) sensitivity loss for the left:
eye of subject PI 0. Maps are coded in 16 levels of gray with a
0 to 30 dB range for the cone map and a 0 to 54 dB range for
the rod map. White indicates normal sensitivity, and black
indicates no detection of the stimulus.
rod function; only cone function was measurable. The
one exception, subject P10, had evidence of rod-me-
diated function, and this was in the central and midpe-
ripheral fields; average rod sensitivity loss within this

Page 5

3138
Investigative Ophthalmology & Visual Science, July 1994, Vol. 35, No. 8
8 8
P2
P11
30 20 10
10 20 30 30 20 1C
10 20 30
ECCENTRICITY
3
2
36-
24-
1 12-
1 12-
2
3
24 -
36-
4
I
48-
I
t 72 60 48 36 24 12 0 12 24 36 48 N T 72 60 48 36 24 12 0 12 24 36 48 N
48 36 24 12 0 12 24 36 48 60 72 T
FIGURE 3. Light-adapted static perimetry in three affected subjects. Visual sensitivity was
measured using a white adapting light (10 cd/m2) and a 600-nm test light for subjects P9
{left), P2 (center), and PI 1 (right). (Top) Static perimetric profiles along the horizontal merid-
ian for the central 60°. Dashed lines delimit the normal range (mean ± 2 SD), and hatched
areas correspond to the physiological blind spot. (Bottom) Gray scale maps of static cone
sensitivity loss throughout the visual field. Patients 9 and 10, left eye; patient 11, right eye.
Maps are coded in 16 levels of gray with a 30-dB range; white indicates normal sensitivity, and
black indicates no detection of the stimulus. S = superior; I = inferior; N = nasal; T =
temporal.
area was 36 dB (Fig. 2). A two-color, dark-adapted
profile along the horizontal meridian in this patient
showed that the central-most loci were cone-mediated
as indicated by the response to 650 nm stimuli (trian-
gles in Figure 2), whereas between about 10° and 30°
there was detectable rod function as indicated by the
response to 500 nm stimuli (circles in Fig. 2).
Cone sensitivity was variably affected in the pa-
tients (Table 2). Figures 2 and 3 show patients with
different degrees of cone sensitivity loss. Two patients
P I
P2
P11
i vAAAAA
i 1
FIGURE 4. Electroretinography in five family members—one normal subject (PI, left), and
four affected subjects (P9, P10, P2, and PI 1, right). (Top to bottom) Rod ERG (single dim blue
flash, dark adapted); mixed rod-cone ERG (single right white flash, dark adapted); cone ERG
(single white flash on white background); cone ERG (29 Hz white flicker on white back-
ground). Measurements, normal statistics, and calibrations are shown in the footnote to
Table 3. Vertical marks correspond to 24 /xV for the recordings from the affected subjects
(P9, P10, P2, and Pll); vertical marks correspond to 98 fiV for the recordings from the
unaffected subject (PI). Full sweeps cover 180 msecs; the stimulus onset occurred at the start
of each trace. Bar = 200 msec.