Content uploaded by Dror Sharon
Author content
All content in this area was uploaded by Dror Sharon on Sep 21, 2014
Content may be subject to copyright.
Am. J. Hum. Genet. 73:1131–1146, 2003
1131
RP2 and RPGR Mutations and Clinical Correlations in Patients
with X-Linked Retinitis Pigmentosa
Dror Sharon,
1,*
Michael A. Sandberg,
2
Vivian W. Rabe,
1
Melissa Stillberger,
2
Thaddeus P. Dryja,
1
and Eliot L. Berson
2
1
Ocular Molecular Genetics Institute and the
2
Berman-Gund Laboratory for the Study of Retinal Degenerations, Harvard Medical School,
Massachusetts Eye and Ear Infirmary, Boston
We determined the mutation spectrum of the RP2 and RPGR genes in patients with X-linked retinitis pigmentosa
(XLRP) and searched for correlations between categories of mutation and severity of disease. We screened 187
unrelated male patients for mutations, including 135 with a prior clinical diagnosis of XLRP, 11 with probable
XLRP, 30 isolate cases suspected of having XLRP, and 11 with cone-rod degeneration. Mutation screening was
performed by single-strand conformation analysis and by sequencing of all RP2 exons and RPGR exons 1–14,
ORF15, and 15a. The refractive error, visual acuity, final dark-adapted threshold, visual field area, and 30-Hz cone
electroretinogram (ERG) amplitude were measured in each patient. Among the 187 patients, we found 10 mutations
in RP2, 2 of which are novel, and 80 mutations in RPGR, 41 of which are novel; 66% of the RPGR mutations
were within ORF15. Among the 135 with a prior clinical diagnosis of XLRP, mutations in the RP2 and RPGR
genes were found in 9 of 135 (6.7%) and 98 of 135 (72.6%), respectively, for a total of 79% of patients. Patients
with RP2 mutations had, on average, lower visual acuity but similar visual field area, final dark-adapted threshold,
and 30-Hz ERG amplitude compared with those with RPGR mutations. Among patients with RPGR mutations,
those with ORF15 mutations had, on average, a significantly larger visual field area and a borderline larger ERG
amplitude than did patients with RPGR mutations in exons 1–14. Among patients with ORF15 mutations, regression
analyses showed that the final dark-adapted threshold became lower (i.e., closer to normal) and that the 30-Hz
ERG amplitude increased as the length of the wild-type ORF15 amino acid sequence increased. Furthermore, as
the length of the abnormal amino acid sequence following ORF15 frameshift mutations increased, the severity of
disease increased.
Introduction
Linkage analyses indicate that there are at least five X-
linked retinitis pigmentosa (XLRP) genes. Two of these
genes have been identified: RP2 (MIM 312600) within
Xp11.23 (Schwahn et al. 1998) and RPGR (locus RP3
[MIM 312610]) within Xp21.1 (Meindl et al. 1996;
Roepman et al. 1996). The others have been mapped
to Xp22 (locus RP23), Xq26-27 (locus RP24 [MIM
300155]), and Xp21.3-21.2 (locus RP6 [MIM 312612])
(Ott et al. 1990; Gieser et al. 1998; Hardcastle et al.
2000). Both RP2 and RPGR proteins are ubiquitously
expressed but have unknown function. The primary
structure of RP2 shows similarity to cofactor C, a pro-
tein involved in the folding of b-tubulin (Schwahn et al.
Received March 18, 2003; accepted for publication August 29,
2003; electronically published October 16, 2003.
Address for correspondence and reprints: Dr. Eliot L. Berson, Mas-
sachusetts Eye and Ear Infirmary, 243 Charles Street, Boston, MA
02114. E-mail: linda_berard@meei.harvard.edu
*Present affiliation: Department of Ophthalmology, Hadassah-He-
brew University Medical Center, Jerusalem, Israel.
䉷2003 by The American Society of Human Genetics. All rights reserved.
0002–9297/2003/7305-0015$15.00
1998). A portion of RPGR is similar to RCC1, a guanine
nucleotide exchange factor for Ran-GTPase (Meindl et
al. 1996; Roepman et al. 1996). RPGR is found in pho-
toreceptor cilia in the mouse retina, and it interacts with
another ciliary protein, RPGRIP1 (Hong et al. 2001).
Previous studies found RP2 mutations in 7%–18%
of patients with XLRP (Schwahn et al. 1998; Hardcastle
et al. 1999; Mears et al. 1999; Sharon et al. 2000; Breuer
et al. 2002). Initial studies of RPGR found mutations
in 10%–26% of patients with XLRP (Meindl et al.
1996; Buraczynska et al. 1997; Miano et al. 1999; Zito
et al. 1999; Sharon et al. 2000), but these surveys were
incomplete because they overlooked a region, called
“ORF15,” that was originally thought to be part of
intron 15 but is now known to be included in the ter-
minal exon of a major splice variant of the transcript
(Vervoort et al. 2000). When this new exon was con-
sidered, the percentage of patients with XLRP having
mutations in RPGR ranged from 30% to 60% in dif-
ferent series (Vervoort et al. 2000; Breuer et al. 2002).
Patients with RPGR mutations have been reported
to have recessive XLRP, dominant XLRP (Rozet et al.
2002), X-linked recessive atrophic macular degenera-
Table 1
Sequence Anomalies Found in the RP2 Gene in Male Patients with XLRP
Mutation Category
and DNA Change
a
Sequence Change
b
Protein Change Exon/Intron Patient(s) Previous Article(s)
Likely pathogenic mutations:
Deletions:
409delATT ATTr– –– Ile137del Exon 2 004-233 Sharon et al. 2000; Breuer et al. 2002
688delAAGAG AAG AGCr–– – C–– Lys230;Ter@232 Exon 2 004-110 Hardcastle et al. 1999; Sharon et al. 2000
798delGACA CAG ACArCA– ––– Gln266;Ter@270 Exon 3 039-109 Thiselton et al. 2000
Splice site mutations
c
:
IVS1⫹3ArT CgtaatgrCgttatg … Intron 1 004-215 Sharon et al. 2000
IVS4⫹3ArC GgtacaarGgtccaa … Intron 4 121-237
IVS4⫹3ArG GgtacaarGgtgcaa … Intron 4 004-288
Missense mutations:
G257A TGTrTAT Cys86Tyr Exon 2 004-149 Sharon et al. 2000
C284T CCCrCTC Pro95Leu Exon 2 004-229 Sharon et al. 2000
C352T CGTrTGT Arg118Cys Exon 2 121-211 Bader et al. 2003
G353A CGTrCAT Arg118His Exon 2 004-284, 004-176 Schwahn et al. 1998; Hardcastle et al. 1999; Sharon et al. 2000;
Breuer et al. 2002; Bader et al. 2003
Likely nonpathogenic polymorphism:
Missense mutation:
C844T
d
CGGrTGG Arg282Trp Exon 3 … …
a
Nucleotide positions are based on GenBank sequence (accession number NM_006915), with the A base of the first ATG in the ORF designated as base number 1.
b
Uppercase letters denote coding sequences, lowercase letters denote intron sequences, and dashes denote deleted bases.
c
Donor splice-site scores for wild-type versus mutant sequences are as follows: IVS1⫹3ArT, 0.89 versus 0.06; IVS4⫹3ArC, 0.91 versus 0.03; IVS4⫹3ArG, 0.91 versus 0.44.
d
Allele frequencies (allele counts) in X chromosomes of index patients versus controls were 0.04 (8:181) versus 0.03 (5:148), respectively.
Sharon et al.: Mutations and Clinical Correlations in XLRP 1133
Table 2
Distribution of RP2 and RPGR Mutations in Different Categories
D
IAGNOSIS
T
OTAL
N
O
.
OF
P
ATIENTS
N
O
.
WITH
M
UTATION IN
RP2
RPGR
Exons 1–14
RPGR
ORF15
XLRP 135 9 (7%) 28 (21%) 70 (52%)
Probable XLRP 11 0 2 4
Cone-rod degeneration 11 ND
a
05
Isolate males with RP 30 2 1 1
Total 187 11 31 80
a
ND pnot done.
tion (Ayyagari et al. 2002), or cone-rod degeneration
(COD1) (Demirci et al. 2002; Yang et al. 2002). All
reported patients with RP2 mutations have been de-
scribed as having recessive XLRP.
A previous study by our group, performed prior to
the discovery of ORF15, reported that patients with
mutations in RPGR have, on average, smaller electro-
retinogram (ERG) amplitudes and visual field areas than
do patients with RP2 mutations (Sharon et al. 2000).
We have now expanded our evaluation to include a
larger number of patients with clinically defined XLRP.
We also surveyed a set of males with probable XLRP
and a separate set of isolate males who were suspected
of having XLRP on the basis of ocular findings. Finally,
we analyzed a set of male patients with cone-rod de-
generation. These sets of patients were evaluated for
mutations in RP2 and RPGR, including the ORF15
region of RPGR.
Methods
Ascertainment of Patients
The present study conformed to the tenets of the Dec-
laration of Helsinki. A total of 187 male index patients
were studied. All index patients received diagnoses
through ophthalmologic examination, including ERG
testing. Most patients resided in the United States and
Canada. A total of 135 patients had a prior clinical
diagnosis of XLRP; they each had an unaffected father
and came from families with no evidence of male-to-
male transmission. Most came from families with two
or more affected male relatives, and the mother of the
index patient was examined and showed signs of the
XLRP carrier state (Berson et al. 1979). Eighty-five of
the patients with XLRP were previously screened by us
for RP2 and RPGR mutations (not including ORF15)
and were the subjects of a previous study (Sharon et al.
2000). A separate set of 11 patients had probable XLRP;
these patients had an affected brother, no other affected
relative, and a mother who was unavailable or unwilling
to have an eye examination to evaluate whether she
showed the carrier state. We also included 30 simplex
(isolate) patients suspected of having XLRP on the basis
of clinical risk factors (e.g., visual acuity ⭐20/25 and
two or more diopters of myopia [Berson et al. 1980]).
Finally, we evaluated 11 patients who had cone-rod de-
generation and who had no affected female relatives.
Patients with cone-rod degeneration had slightly reduced
rod ERG amplitudes and severely reduced cone ERG
amplitudes. Fifty-two affected relatives of index patients
who were found to have RP2 or RPGR mutations were
also clinically evaluated. We evaluated 96 unrelated con-
trol individuals (58 female and 38 male, for a total of
154 X chromosomes) with no symptoms or family his-
tory of retinal degeneration. Absence of a variant allele
in a set of 154 control chromosomes indicates with 95%
confidence that it occurs at a frequency of !2% in the
population from which the controls were derived, on the
basis of the binomial distribution (i.e.,
154
[1 ⫺0.02] p
). Informed consent was obtained from all partici-0.05
pants before they donated 10–30 ml of venous blood
for this research. Leukocyte nuclei were prepared from
the blood samples and stored at ⫺70⬚C before DNA
was purified from them.
Screening for Mutations
The SSCP technique was used to screen all five RP2
exons and RPGR exons 1–15 and 15a, as well as the
immediately flanking intron sequences, for point mu-
tations and other small-scale sequence changes. Each
exon was individually amplified from leukocyte DNA
samples by PCR, using previously published primer pairs
(Meindl et al. 1996; Schwahn et al. 1998). Seven pairs
of primers were designed to cover exon ORF15 (primer
sequences are available from the authors’ Web site).
Many of the amplified fragments from ORF15 produced
complex and nonreproducible SSCP patterns, and, there-
fore, ORF15 was directly sequenced in many patients.
Sequencing of parts of ORF15 could be performed only
in the antisense direction, because we were unable to
develop primers or sequencing conditions that allowed
sequencing of the sense direction. In a few patients, small
regions of ORF15 could not be sequenced clearly in
either direction.
PCRs were performed in the wells of 96-well microtiter
plates. Each well contained 20 ng of leukocyte DNA in
20 ml of a buffer containing 20 mM tris-HCl (pH 8.4 or
8.6), 0.25–1.5 mM MgCl
2
,50mMKCl,0.02mMdATP,
0.02 mM dTTP, 0.02 mM dGTP, 0.002 mM dCTP, 0.6
mCi [a-
33
P]dCTP (3,000 Ci/mmol), 0.1 mg/ml bovine se-
rum albumin, 0% or 10% dimethyl sulfoxide, and 0.25
UofTaq polymerase. The pH, Mg
⫹⫹
concentration, an-
nealing temperature, and presence or absence of 10%
dimethyl sulfoxide were tailored to each primer pair to
yield optimal amplification. After an initial denaturation
(94⬚C for 5 min), 25 cycles of PCR amplification were
1134
Table 3
Sequence Anomalies Found in the RPGR Gene in Male Patients with XLRP
Mutation Category
and DNA Change
a
Sequence Change
b
Protein Change Exon/Intron Patient(s) Previous Article(s)
Frameshift:
415delT TTGrT–G Leu119;Ter@131 Exon 5 004-148 Sharon et al. 2000
544-5delTT TTTrT–– Phe162;Ter@165 Exon 6 004-139 Sharon et al. 2000; Breuer et al. 2002
806delC GCCrGC– Ala249;Ter@296 Exon 7 121-196
896delT TTTrTT– Phe279;Ter@297 Exon 8 004-123 Sharon et al. 2000
897-901delCTTTT CTT TTTr–– – – –T Leu280;Ter@280 Exon 8 004-261
928delA GAGrG–G Glu290;Ter@297 Exon 8 004-105 Buraczynska et al. 2000; Sharon et al. 2000
1151-2insT GCTrTGCT Ala365;Ter@376 Exon 10 099-042 Sharon et al. 2000; Guevara-Fujita et al. 2001
1159delC CCTrC–T Pro367;Ter@380 Exon 10 004-220 Sharon et al. 2000
1435-6delTC GTCrG–– Val459;Ter@461 Exon 11 004-292, 004-165, 039-082 Sharon et al. 2000; Zito et al. 2000
1641-4delACAA ACA ATTr–– – –TT Thr528;Ter@531 Exon 14 099-056 Buraczynska et al. 2000; Sharon et al. 2000
g.ORF15⫹82_83insA
c
AATrAAAT ORF15N27;Ter@43 Exon ORF15 004-206 Sharon et al. 2000
g.ORF15⫹356_357insGAAG AAGrGAA GAA ORF15K119;Ter@183 Exon ORF15 004-237
g.ORF15⫹481_484delGAGA AGA GAArA–– ––A ORF15R160;Ter@229 Exon ORF15 004-210 Breuer et al. 2002
g.ORF15⫹483_484delGA AGAr––A ORF15E161;Ter@183 Exon ORF15 004-145, 004-153, 004-164, 004-269,
121-189
Vervoort et al. 2000; Breuer et al. 2002; Bader et al. 2003
g.ORF15⫹499_502delAGGA AAG GAGrA–– ––G ORF15K166;Ter@229 Exon ORF15 121-087 Breuer et al. 2002
g.ORF15⫹517_518delAG GAGrG–– ORF15E172;Ter@183 Exon ORF15 004-151, 004-270
g.ORF15⫹587delA AGArAG– ORF15R195;Ter@248 Exon ORF15 004-293
g.ORF15⫹614_615delAA AAArA–– ORF15K201;Ter@248 Exon ORF15 004-125, 004-180
g.ORF15⫹650_653delAGAG ACA GAGrAC– ––– ORF15T216;Ter@229 Exon ORF15 121-267
g.ORF15⫹652_653delAG GAGrG–– ORF15E217;Ter@248 Exon ORF15 001-210, 004-109, 004-146, 004-162,
004-207, 004-251, 004-253,
039-206
Vervoort et al. 2000; Breuer et al. 2002; Rozet et al.
2002; Bader et al. 2003
g.ORF15⫹659_660delAG AGA GGGrAG– –GG ORF15R219;Ter@248 Exon ORF15 004-199
g.ORF15⫹670_671delAA AAArA–– ORF15K223;Ter@248 Exon ORF15 004-111
g.ORF15⫹673_674delAG GAGrG–– ORF15E224;Ter@248 Exon ORF15 004-219, 004-221 Vervoort et al. 2000; Breuer et al. 2002
g.ORF15⫹689_692delAGAG GTA GAGrGT– – –– ORF15V229;Ter@234 Exon ORF15 004-150, 004-218 Vervoort et al. 2000; Breuer et al. 2002; Bader et al. 2003
g.ORF15⫹740_741delGG GAG GAGrGA– –AG ORF15E246;Ter@248 Exon ORF15 004-238
g.ORF15⫹746delT GGTrGG– ORF15G248;Ter@503 Exon ORF15 004-108
g.ORF15⫹752_753delGG GAG GAGrGA– –AG ORF15G250;Ter@492 Exon ORF15 004-104
g.ORF15⫹763_767del5bp GAA GGGrG–– – –– ORF15E254;Ter@492 Exon ORF15 004-134
g.ORF15⫹764_765delAG GAA GGGrGA– –GG ORF15E254;Ter@492 Exon ORF15 004-287
g.ORF15⫹818_819delAG AAA GGGrAA– –GG ORF15K272;Ter@492 Exon ORF15 004-156
g.ORF15⫹848_849delGG GGG GAGrGG– –AG ORF15G282;Ter@492 Exon ORF15 004-228, 004-231 Bader et al. 2003
g.ORF15⫹860_861delGG GGG GAGrGG– –AG ORF15G286;Ter@492 Exon ORF15 004-272
g.ORF15⫹872_873insA GGGrAGG ORF15G291;Ter@492 Exon ORF15 004-155, 004-267 Vervoort et al. 2000
g.ORF15⫹902_903delGG GGG GAGrGG– –AG ORF15G300;Ter@492 Exon ORF15 006-002, 004-143
g.ORF15⫹906_909delGGAG GGA GAAr–– – –AA ORF15G302;Ter@503 Exon ORF15 004-102
g.ORF15⫹926_927delGG GGG GAGrGG– –AG ORF15G308;Ter@492 Exon ORF15 006-007
g.ORF15⫹961_962delAA GAArG–– ORF15E320;Ter@492 Exon ORF15 004-121
g.ORF15⫹962_963insCCTC GAGrCCT CGA ORF15E321;Ter@492 Exon ORF15 004-232, 004-205
g.ORF15⫹977_978delGG GGG GAGrGG– –AG ORF15G325;Ter@492 Exon ORF15 039-358 Vervoort et al. 2000
g.ORF15⫹1010_1011delGG GGG GAGrGG– –AG ORF15G336;Ter@492 Exon ORF15 004-115 Vervoort et al. 2000
g.ORF15⫹1098_1101delGAGG GGA GGArG–– ––A ORF15E366;Ter@503 Exon ORF15 004-209
1135
g.ORF15⫹1113delG GAGr–AG ORF15E371;Ter@503 Exon ORF15 004-135
g.ORF15⫹1146delG GAAr–AA ORF15E382;Ter@503 Exon ORF15 004-167, 004-147
g.ORF15⫹1184_1185delGG GGA GGArG–– – –A ORF15G394;Ter@492 Exon ORF15 004-130 Breuer et al. 2002
g.ORF15⫹1191delG GAAr–AA ORF15E397;Ter@503 Exon ORF15 099-072
g.ORF15⫹1254_1257delGGAG GGA GAGr––– –AG ORF15G418;Ter@503 Exon ORF15 004-103
g.ORF15⫹1258_1259delAG GAGrG–– ORF15E419;Ter@492 Exon ORF15 099-059
g.ORF15⫹1339delA GAGrG–G ORF15E446;Ter@503 Exon ORF15 004-144, 004-226, 004-236 Bader et al. 2003
g.ORF15⫹1339_1340delAG GAGrG–– ORF15E446;Ter@493 Exon ORF15 240-001 Demirci et al. 2002; Yang et al. 2002
g.ORF15⫹1343_1344delGG GGGrG–– ORF15G447;Ter@493 Exon ORF15 115-010, 182-005, 289-040 Demirci et al. 2002; Yang et al. 2002
Nonsense:
C664A TCArTAA Ser202Ter Exon 6 099-029 Sharon et al. 2000
T1039G TTArTGA Leu327Ter Exon 9 004-283
G1806T GAArTAA Glu583Ter Exon 14 004-174
g.ORF15⫹327ArT AAGrTAG ORF15K109Ter Exon ORF15 004-256 Breuer et al. 2002
g.ORF15⫹369GrT GAArTAA ORF15E123Ter Exon ORF15 004-239 Breuer et al. 2002
g.ORF15⫹423GrT GAGrTAG ORF15E141Ter Exon ORF15 004-127 Breuer et al. 2002
g.ORF15⫹465GrT GAArTAA ORF15E155Ter Exon ORF15 121-845, 121-033 Rozet et al. 2000
g.ORF15⫹507GrT GAArTAA ORF15E169Ter Exon ORF15 004-173, 099-019 Breuer et al. 2002
g.ORF15⫹540GrT GAGrTAG ORF15E180Ter Exon ORF15 004-278, 004-286
g.ORF15⫹684GrT GAArTAA ORF15E228Ter Exon ORF15 004-216
g.ORF15⫹738GrT GAGrTAG ORF15E246Ter Exon ORF15 004-224
g.ORF15⫹897GrT GAArTAA ORF15E299Ter Exon ORF15 004-208
g.ORF15⫹954GrT GAGrTAG ORF15E318Ter Exon ORF15 004-181
g.ORF15⫹963GrT GAGrTAG ORF15E321Ter Exon ORF15 004-290
g.ORF15⫹1047GrT GAGrTAG ORF15E349Ter Exon ORF15 004-175
g.ORF15⫹1458GrT GAGrTAG ORF15E486Ter Exon ORF15 115-030
Splice-Site
d
:
IVS1⫹1GrA CCGgtgarCCGatga Intron 1 099-007 Zito et al. 1999; Sharon et al. 2000
IVS3⫺6TrA tatttttrtatattt Intron 3 004-202 Sharon et al. 2000
IVS4⫹1GrC CAGgtatrCAGctat Intron 4 004-169 Sharon et al. 2000; Breuer et al. 2002
IVS7⫺1GrA atagCAGrataaCAG Intron 7 004-258
IVS13⫺1GrA acagAAAracaaAAA Intron 13 004-268, 004-100 Sharon et al. 2000
Missense:
G186A GGArAGA Gly43Arg Exon 2 004-152 Sharon et al. 2000
G187A GGArGAA Gly43Glu Exon 2 004-223 Sharon et al. 2000
G238T GGCrGTC Gly60Val Exon 3 004-158, 004-133 Buraczynska et al. 1997; Fishman et al. 1998; Sharon et
al. 2000
A438G AGArGGA Arg127Gly Exon 5 004-157 Sharon et al. 2000; Breuer et al. 2002
G964A TGTrTAT Cys302Tyr Exon 8 099-008 Sharon et al. 2000
G993A GATrAAT Asp312Asn Exon 8 004-285
G993T GATrTAT Asp312Tyr Exon 8 004-279
G1017A GGArAGA Gly320Arg Exon 9 006-005
G1366A GGCrGAC Gly436Asp Exon 11 004-264 Sharon et al. 2000; Guevara-Fujita et al. 2001; Breuer et
al. 2002
a
Nucleotide positions in RPGR exons 1–14 are based on the study by Meindl et al. (1996). Nucleotide positions in ORF15 are based on the study by Vervoort et al. (2000). Please note that some
of our mutations presented here are identical to those reported elsewhere (Breuer et al. 2002), although the nomenclature is different.
b
Uppercase letters denote coding sequences; lowercase letters denote intron sequences; dashes indicate deleted bases.
c
The mutation ORF15N27;Ter@43 was previously reported by us as ORF15Asn612;Ter@628 located in exon 15 (Sharon et al. 2000). Exon 15 is now included as part of the terminal exon ORF15,
and, thus, the nomenclature has been revised.
d
Splice-site scores for wild-type versus mutant sequences are as follows: IVS1⫹1GrA, 0.96 versus 0.00 (donor); IVS3⫺6TrA, 0.96 versus 0.73 (acceptor); IVS4⫹1GrC, 0.89 versus 0.0 (donor);
IVS7⫺1GrA, 0.80 versus 0.00 (acceptor); and IVS13⫺1GrA, 0.00 versus 0.00 (acceptor).
Figure 1 Location of ORF15 mutations. Mutations that have been previously published by other groups are depicted onthe left. Mutations
reported in the present study are depicted on the right. The numbers along the ORF15 bar represent the amino acid numbers. Arrows (R)
indicate mutations not reported by other groups, asterisks (*) indicate mutations causing cone-rod degeneration, the number sign (#) indicates
a case with probable X-linked cone dystrophy (Vervoort et al. 2000), and the tilde (∼) indicates that ocular measurements were not provided
by the authors (Pusch et al. 2002).
Sharon et al.: Mutations and Clinical Correlations in XLRP 1137
Figure 2 Plots of ocular function by age for XLRP patients with RP2 mutations (blackened diamonds)orRPGR mutations (unblackened
circles). The regression lines were fitted by least-squares analysis to the RP2 data (solid lines)orRPGR data (stippled lines).
performed. Each cycle consisted of denaturation (94⬚C
for 30 s), primer annealing (50–62⬚C for 30 s), and ex-
tension (71⬚C for 30 s). The final extension was at 71⬚C
for 5 min. The amplified DNA fragments were heat de-
natured, and aliquots of the single-stranded fragments
were separated through polyacrylamide gels. Two differ-
ent gels were used for SSCP analysis of every evaluated
DNA fragment: 6% polyacrylamide in tris-borate-EDTA
(TBE) buffer and 6% polyacrylamide with 10% glycerol
in TBE buffer. Gels were run at 6–18 W for 5–18 h at
room temperature before drying and autoradiography.
Variant bands were analyzed by sequencing the corre-
sponding PCR-amplified DNA segments through use of
dye terminators (Dye terminator cycle sequencing kit;
Perkin Elmer) on ABI 373 or ABI 3100 automated
sequencers.
The numbering of the DNA bases and amino acid
residues is based on the previously reported sequences
of RP2 (Schwahn et al. 1998), RPGR (Meindl et al.
1996), and ORF15 (Vervoort et al. 2000; Vervoort and
Wright 2002). We interpreted all frameshift and non-
sense mutations as null alleles and pathogenic, unless
they were in the terminal exon, in which case they were
judged according to the criteria for missense and in-
frame changes (see below). Splice-site mutations were
considered pathogenic if they affected the canonical AG-
GT splice acceptor–splice donor sequences or if splice-
site prediction software predicted that the variant se-
quence would substantially reduce the recognition of an
existing splice site (Berkeley Drosophila Genome Project
Splice-Site Prediction Server). A missense or in-frame
change was considered pathogenic if it met the following
three criteria: (1) it was found only in our patients (and,
possibly, in patients reported by other groups) and not
in any of the 154 control chromosomes we evaluated or
in the controls from any previously reported study; (2)
every patient with the variant sequence had no other
sequence abnormality in RP2 or RPGR that obviously
created a null allele; and (3) the change was present in
all affected male relatives and no unaffected male rela-
tives who were evaluated (note that segregation analyses
were not performed in most families). In ORF15, mis-
sense changes and in-frame deletions were categorized
as nonpathogenic because neither our group nor any
other group has found evidence for such mutations in
ORF15 being pathogenic. Direct sequencing of ORF15
in normal controls was not performed. Sequence changes
that were not predicted to affect the amino acid sequence
of the encoded protein (e.g., intronic changes, isocoding
changes, etc.) and were unlikely, on the basis of splice-
1138 Am. J. Hum. Genet. 73:1131–1146, 2003
Table 4
Ocular Function for Patients with RP2 or RPGR Mutations
O
CULAR
F
UNCTION
M
UTATIONS IN
RP2 M
UTATIONS IN
RPGR
Pn Mean ⳲSEM
Geometric
Mean
a
nMean ⳲSEM
Geometric
Mean
a
ln visual acuity
b
16 ⫺2.35 Ⳳ.27 20/210 156 ⫺1.41 Ⳳ.09 20/82 .001
c
log dark-adapted threshold elevation
b
13 2.18 Ⳳ.36 … 115 2.44 Ⳳ.12 … .49
ln visual field area (in deg
2
)
b,d
15 8.20 Ⳳ.29 3,640 138 8.03 Ⳳ.10 3,071 .57
ln 30-Hz ERG amplitude (in mV)
e,f
13 .66 Ⳳ.41 1.93 137 .02 Ⳳ.12 1.02 .14
a
Geometric mean values are calculated from ln-transformed data.
b
Data adjusted for age.
c
Pbased on normalized ranks was .002 .
d
Normal visual field area is ⭓11,399 deg
2
.
e
Data adjusted for age and refractive error.
f
Normal 30-Hz ERG amplitude is ⭓50 mV.
site prediction software, to create or destroy splice sites
were considered nonpathogenic. We also considered as
nonpathogenic those sequence changes that did not co-
segregate with the disease on the basis of the results from
our previous study (Sharon et al. 2000) or previous stud-
ies performed by other groups, as well as those sequence
changes found in normal control males.
Clinical Evaluation and Statistical Analyses
We evaluated our patients and recorded the following
clinical features that reflect the severity of retinitis pig-
mentosa at a given age: visual acuity, final dark-adapted
threshold, visual field area, and 30-Hz cone ERG am-
plitude. We also measured the refractive error (recorded
here as spherical equivalent). All values were averages
between right and left eyes when both were available.
In most cases, data were collected from initial visits.
When data from an initial visit were incomplete, data
were used from the earliest subsequent visits for which
they were available (i.e., the second or third visit). In-
cluding affected relatives, the clinical sample comprised
16 patients with RP2 mutations and 156 patients with
RPGR mutations (111 of whom had ORF15 mutations).
The age at clinical evaluation ranged from 5 to 67 years,
with a of years.mean ⳲSD 28.3 Ⳳ12.0
Best-corrected visual acuities were obtained using a
projected Snellen chart. Dark-adapted thresholds were
obtained with an 11⬚white test light presented in the
Goldmann-Weekers dark adaptometer after 45 min of
dark adaptation. Kinetic visual fields were measured to
the V4e white test light of the Goldmann perimeter
against the standard background of 31.5 apostilbs,
bringing the test light from nonseeing to seeing areas.
Fields were plotted with a digitizing tablet or scanned
by custom software and converted to areas. Full-field
cone ERGs were elicited with 10-ms0.5-Hzor30-Hz
flashes of white light (0.2 candela s/m
2
) after pupillary
dilation and 45 min of dark adaptation. ERGs were
monitored with a contact lens electrode on the topically
anesthetized cornea and were differentially amplified.
Consecutive responses 110 mV in amplitude were pho-
tographed from the screen of an oscilloscope or digitized
and displayed on a computer screen. Smaller responses
were digitized, smoothed with a bandpass filter if elicited
with 30-Hz flashes, and averaged. Waveforms were
quantified with respect to trough-to-peak amplitudes;
amplitudes !1.0 mV to 0.5-Hz flashes or !0.05 mVto
30-Hz flashes were nondetectable and, for purposes of
analysis, were recoded as 1.0 mVor0.05mV, respectively.
Details of these procedures have been described else-
where (Andre´asson et al. 1988; Berson et al. 1991; Sand-
berg et al. 1995). We used the V4e white test light for
measuring visual fields and, for most analyses, the 30-
Hz white flashes for eliciting ERGs, because these con-
ditions of testing provided us with large data sets.
Visual acuities, visual field areas, and ERG amplitudes
were transformed to the log
e
scale to better approximate
normal distributions. Since the distribution of visual acu-
ities was appreciably skewed even after the logarithmic
transformation and included values (i.e., count fingers
and hand motions) that cannot be reliably expressed as
a decimal, we also converted acuities to ranks and then
to the normal form by a probit transformation (Rosner
2000). Multiple-regression analyses were performed on
all available data, with each measure of ocular function
as the dependent variable and the genetic characteris-
tic(s) and age as the independent variables. In this way,
the relationship of each dependent variable to the genetic
characteristic(s) was adjusted for patient age. For log
e
ERG amplitude as the dependent variable, the spherical
equivalent of the refractive error was included as an
additional covariate because ERG amplitude increases
with increasing positive sphere (Westall et al. 2001).
These analyses were performed after excluding outliers
for ocular function versus age identified by applying the
generalized extreme studentized residual test for linear
Sharon et al.: Mutations and Clinical Correlations in XLRP 1139
Figure 3 Plots of ocular function by age for patients with XLRP with RPGR mutations in exons 1–14 (blackened diamonds) or in ORF15
(unblackened circles). The regression lines were fitted by least-squares analysis to the exon 1–14 data (solid lines) or ORF15 data (stippled
lines).
regression (Paul and Fung 1991). Mean values are listed
with their SEs, and the mean refractive error (spherical
equivalent) was compared by genotype with the Student
ttest, after excluding outliers identified by the extreme
studentized deviate test (Rosner 2000). Data transfor-
mations and statistical analyses were performed with
JMP, version 3.2 (SAS Institute), on a Macintosh Pow-
erbook G3 computer.
Results
We screened with SSCP the DNA of 135 patients with
XLRP, 11 patients with probable XLRP, and 30 patients
with suspected XLRP for mutations in the RP2 and
RPGR genes. We also screened the RPGR gene, includ-
ing ORF15, in 11 patients with cone-rod degeneration.
In addition, we sequenced the ORF15 region in all pa-
tients with XLRP, probable XLRP, and cone-rod degen-
eration who had no mutation detected in RP2 or else-
where in RPGR.
Mutations in the RP2 Gene
We found 11 sequence changes in RP2 in our set of
patients, 10 of which are likely to be pathogenic (table
1). These 10 mutations were identified in a total of 11
unrelated patients (table 2), 6 of whom have been reported
by us previously (Sharon et al. 2000). Two of the 10
mutations were novel splice-site changes (IVS4⫹3ArC
and IVS4⫹3ArG). These novel mutations were notfound
in a screen of 96 control individuals (58 female and 38
male, for a total of 154 control chromosomes). They af-
fected the third base of the splice donor site within intron
4, changing the A in that position to a C in one patient
and to a G in a second patient.
Mutations in the RPGR Gene
We found 125 sequence changes in RPGR (table 3
and authors’ Web site). Eighty of them, found in a total
of 111 index patients (table 2), were interpreted as path-
ogenic mutations. Most of the RPGR mutations (53 of
80) were located within ORF15 (fig. 1). Forty-one of
the pathogenic mutations are novel. The mutations fell
into four groups: frameshift, nonsense, splice-site, and
missense. None of the novel mutations in exons 1–14
was found in the 96 control individuals. The novel mu-
tations in ORF15 were also not detected among the 96
controls by SSCP; although the SSCP method is of lim-
ited value for evaluating a highly repetitive sequence as
1140 Am. J. Hum. Genet. 73:1131–1146, 2003
Table 5
Ocular Function for Patients with RPGR Mutations Involving Exons 1–14 versus Exon ORF15
O
CULAR
F
UNCTION
E
XON
1–14 M
UTATIONS
E
XON
ORF15 M
UTATIONS
Pn Mean ⳲSEM
Geometric
Mean
a
nMean ⳲSEM
Geometric
Mean
a
ln visual acuity
b
45 ⫺1.27 Ⳳ.16 20/71 111 ⫺1.48 Ⳳ.10 20/88 .27
c
log dark-adapted threshold elevation
b
33 2.75 Ⳳ.22 … 82 2.32 Ⳳ.14 … .12
ln visual field area (in deg
2
)
b,d
42 7.71 Ⳳ.17 2,231 96 8.14 Ⳳ.11 3,429 .04
ln 30-Hz ERG amplitude (in mv)
e,f
39 ⫺.37 Ⳳ.23 .69 98 .16 Ⳳ.14 1.17 .06
a
Geometric mean values are calculated from ln-transformed data.
b
Data adjusted for age.
c
Pbased on normalized ranks was .58.
d
Normal visual field area is ⭓11,399 deg
2
.
e
Data adjusted for age and refractive error.
f
Normal 30-Hz ERG amplitude is ⭓50 mV.
in ORF15, we did not confirm the negative results in
controls by sequencing.
Frameshift mutations.—We identified 10 frameshift
mutations in exons 1–14, all but one of which were small
(one-, two-, four-, and five-base) deletions (table 3). All
of these mutations led to premature nonsense codons
upstream of the terminal exon (ORF15) and, thus, are
expected to produce null alleles because of nonsense-
mediated degradation of transcribed RNA. All muta-
tions but one were found in one patient each. The ex-
ception was delTC@Val459, which was carried by three
index patients (004-165, 039-082, and 004-292). These
three patients also carried a rare allele at a polymorphic
site in ORF15 (1307-1318del12), indicating that themu-
tation in these three patients was likely to be identical
by descent.
In ORF15 we found 40 frameshift mutations, 26 of
which are novel (fig. 1). The vast majority of these frame-
shift mutations in ORF15 (36 of the 40) were small
deletions of one, two, four, or five bases. The sense strand
of ORF15 is mostly composed of imperfect repeats of
purines that encode numerous glutamate residues (co-
dons GAA and GAG) and glycine residues (codons GGA
and GGG). Four of the frameshifts were inserts of one
or four bases, and three of these inserts were composed
(in the sense strand) of purines. The fourth frameshift-
insertion was a novel insertion of four pyrimidines
(insCCTC@ORF15E321); it was found in two of our
patients (004-232 and 004-205). The inserted four bases
create a palindromic sequence at the insertion point.
Nonsense mutations.—Sixteen nonsense mutations
were identified in our screen, 10 of which have not been
reported by other groups. Three of the nonsense mu-
tations were in exons 6, 9, and 14 and would be expected
to be null alleles because of nonsense-mediated decay of
the transcribed RNA. Thirteen of the 16 nonsense mu-
tations occurred within the terminal exon ORF15.
Splice-site mutations.—We identified five mutations
that potentially destroy a splice site. In four cases, a base
in the canonical dinucleotides at the splice-acceptor or
splice-donor site was mutated, whereas, in the fifth
case, the sixth base upstream of exon 7 was mutated,
IVS3⫺6TrA. Two patients (004-268 and 004-100)
shared a novel mutation, IVS13⫺1GrA. These patients
have different nonpathogenic sequence changes or poly-
morphisms elsewhere in the RPGR gene, and, thus, it is
likely that the same mutation arose independently in the
two families. One intron change, IVS4⫹7TrG, found
in an isolate male (121-229), might create a novel ac-
ceptor site, on the basis of splice-site prediction software
(probability score 0.41) (Berkeley Drosophila Ge-
nome Project Splice-Site Prediction Server); howev-
er, we tentatively categorized it as a nonpathogenic
change, because we have no other evidence supporting
its effect on RNA splicing. The splice-site prediction
software also suggests that two rare isocoding changes
(ORF15Gly488Gly and ORF15Val503Val; see the au-
thors’ Web site) might create splice-donor sites in this
terminal exon. Both of these changes were found in three
patients (004-142, 004-147, and 121-254), one of whom
(004-147) also carried a definite pathogenic mutation in
ORF15 (ORF15⫹1146delG). Thus, the two isocoding
changes are unlikely to be pathogenic.
Missense mutations.—We identified nine missense mu-
tations that were likely to be pathogenic, six of which
have been reported by us previously (Sharon et al. 2000).
The three novel missense mutations affect amino acid
residues within the RCC1 domain encoded within exons
8 and 9. Two of these, Asp312Asn and Asp312Tyr,affect
the same amino acid, and the third, Gly320Arg, affects
a nearby residue. Because these three missense changes
all affect a presumed functional domain, and because
they were not found in 154 control chromosomes, we
have categorized them as likely pathogenic mutations.
One novel missense change (ORF15Glu456Lys) was
found in ORF15. Although we are uncertain about
whether this change is pathogenic, we have included it
in the list of nonpathogenic changes (see the authors’
Sharon et al.: Mutations and Clinical Correlations in XLRP 1141
Table 6
Ocular Function for Patients with RPGR-ORF15 Mutations According to the Location of the First Mutant Codon
O
CULAR
F
UNCTION
A
LTERED
C
ODON
!445 A
LTERED
C
ODON
1445
Pn Mean ⳲSEM
Geometric
Mean
a
nMean ⳲSEM
Geometric
Mean
a
ln visual acuity
b
100 ⫺1.61 Ⳳ.11 20/100 11 ⫺.76 Ⳳ.35 20/43 .02
c
log dark-adapted threshold elevation
b
75 2.62 Ⳳ.13 … 7 ⫺.29 Ⳳ.46 … !.001
ln visual field area (in deg
2
)
b,d
87 8.02 Ⳳ.11 3,041 9 9.26 Ⳳ.34 10,509 .001
ln 30-Hz ERG amplitude (in mV)
e,f
88 ⫺.09 Ⳳ.14 .91 9 2.36 Ⳳ.45 10.6 !.001
ln .5-Hz ERG amplitude (in mV)
g
55 2.36 Ⳳ.14 10.6 9 3.80 Ⳳ.36 44.7 !.001
a
Geometric mean values are calculated from ln-transformed data.
b
Data adjusted for age.
c
Pbased on normalized ranks was .04.
d
Normal visual field area is ⭓11,399 deg
2
.
e
Data adjusted for age and refractive error.
f
Normal 30-Hz ERG amplitude is ⭓50 mV.
g
Normal .5-Hz ERG amplitude is ⭓350 mV; responses adjusted for 30-Hz ERG amplitude.
Web site) because all missense changes in ORF15 pre-
viously reported by us and other groups have been
nonpathogenic.
Nonpathogenic sequence variants.—The table on our
Web site lists the nonpathogenic sequence variants and
polymorphisms that we encountered. It should be noted
that the in-frame deletion Gln527del in exon 14 was
found in one isolate case of RP (121-847). We presume
that this change is the same as Gln526del, which has been
reported previously to be a polymorphism (Zito et al.
2000). Two nonpathogenic polymorphisms in ORF15,
ORF15Val559Ile and ORF15Asn547Asn, were found to-
gether in 11 patients, whereas only ORF15Val559Ile was
found in 1 patient. This indicates that the two sequence
variants are in linkage disequilibrium among our set of
patients. We also identified 10 ORF15 in-frame deletions
and insertions, 8 of which have been previously reported
as polymorphisms (Vervoort et al. 2000; Bader et al.
2003). We have detected one of these in-frame changes,
ORF15⫹694_708del15, in two patients with XLRP who
had no likely pathogenic mutations in either RP2 or
RPGR. We interpreted this sequence change as nonpath-
ogenic because it was found in cis with a definite path-
ogenic nonsense mutation in three previously reported
patients with XLRP (Vervoort et al. 2000; Bader et al.
2003). It should be noted though that the same in-frame
deletion was also found in a pedigree with X-linked cone-
rod dystrophy, and the authors interpreted it as a path-
ogenic mutation (Yang et al. 2002).
Clinical Findings in Patients with RP2 and RPGR
Mutations
We divided our patients into groups based on the re-
sponsible gene (RP2 vs. RPGR). Both index patients
with mutations in RP2 or RPGR and their affected rel-
atives were included in this part of the study. In all, we
clinically evaluated 16 patients with RP2 mutations and
156 patients with RPGR mutations. Within the RPGR
group, we subdivided patients according to mutation
type, location, and the predicted effect of the mutation
on the protein sequence, as detailed below. Too few pa-
tients with RP2 mutations were available to perform a
similar subdivision among mutation types.
Figure 2 compares ocular function by age, for patients
with RP2 and RPGR mutations, on the basis of cross-
sectional analyses of single visits. The least-squares re-
gression lines show that visual acuity, visual field area,
and 30-Hz ERG amplitude declined and the final dark-
adapted threshold increased (i.e., worsened) with increas-
ing age for both groups. The most striking difference
between the two groups was that patients with RP2 mu-
tations tended to have lower visual acuities across all
ages. After adjusting for age, we found that patients with
RP2 mutations ( ;
np16 mean ⳲSE age p27.0 Ⳳ
years) had, on average, a significantly lower visual
4.3
acuity than did patients with RPGR mutations (np
; years) (20/210 vs.
156 mean ⳲSE age p28.5 Ⳳ0.9
20/82, respectively), whether the results were based on
log data or normalized ranks (table 4). We found no
statistically significant difference between the patients
with RP2 and RPGR mutations in the mean log dark-
adapted threshold elevation above normal, the mean
ln visual field area, or the mean ln 30-Hz ERG am-
plitude (table 4). In addition, patients with RP2 mu-
tations were, on average, less myopic (mean spherical
; ) than pa-
equivalent ⳲSE p⫺2.65 Ⳳ0.67 np14
tients with RPGR mutations (mean spherical equiv-
alent ⳲSE p; ), but this dif-
⫺4.19 Ⳳ0.32 np154
ference was not statistically significant ( ),Pp.15
possibly because of the small number of patients with
RP2 mutations.
Figure 3 compares ocular function versus age in two
subgroups of patients with RPGR mutations: patients
1142 Am. J. Hum. Genet. 73:1131–1146, 2003
Figure 4 Plots of the dark-adapted threshold elevation (upper panels) and 30-Hz ERG amplitude (lower panels) versus the altered ORF15
codon (left panels) and the number of mutant residues (right panels). By multiple regression, each side panel controls for the relationship in
the other side panel, as well as for age and (for the lower panels) for refractive error. Both Xand Ycoordinates have been adjusted statistically
to remove the effects of these covariates on the measures of ocular function.
with exon 1–14 mutations and patients with ORF15 mu-
tations. The figure shows a tendency for patients with
ORF15 mutations to have larger visual field areas and
ERG amplitudes over most of the age range. By multiple-
regression adjusting for age and refractive error (in the
case of ERG amplitude), we found that patients with
ORF15 mutations had, on average, a larger visual field
area and ERG amplitude than did patients with RPGR
mutations in exons 1–14, although only the field differ-
ence was statistically significant ( ; table 5). ThesePp.04
comparisons indicate that patients with ORF15 muta-
tions have, on average, better panretinal function than
do patients with RPGR mutations in exons 1–14. There
was no statistically significant difference in average
visual acuity or dark-adapted threshold elevation. We
also found that patients with ORF15 mutations are
more myopic (mean spherical equivalent ⳲSE p
; ) than patients with mutations⫺4.60 Ⳳ0.39 np110
in exons 1–14 (mean spherical equivalent ⳲSE p
; ); this difference in means was⫺3.16 Ⳳ0.49 np44
statistically significant ( ).Pp.04
A previous study has reported that patients with mu-
tations in the 3
end of ORF15 (i.e., downstream of
codon 445) have the diagnosis of cone-rod dystrophy
(Demirci et al. 2002). Our data support this conclusion.
We compared ocular function according to whether a
patient had a mutation upstream versus downstream of
codon 445. Those with a mutation downstream of
amino acid 445 included three index patients with a
prior diagnosis of XLRP and five patients with a prior
diagnosis of cone-rod degeneration. These patients and
their affected male relatives had, on average, a signifi-
cantly better visual acuity, less elevated final dark-
adapted threshold, larger visual field area, and larger
30-Hz ERG amplitude than did patients with ORF15
mutations upstream of codon 445 (table 6). When con-
trolling for 30-Hz (cone) ERG amplitude, we also found
that patients with mutations downstream of codon 445
had larger 0.5-Hz (rod ⫹cone) ERG amplitudes (geo-
metric mean amplitude 44.7 mV; ) than did pa-np9
tients with mutations upstream of codon 445 (geometric
mean amplitude 10.6 mV; ). This difference innp55
means, independent of the variation in cone ERG am-
plitude, was statistically significant ( ).P!.001
We investigated whether ocular function in a given
patient depends on the position and number of mutant
Sharon et al.: Mutations and Clinical Correlations in XLRP 1143
Figure 5 The predicted effect of ORF15 mutations on translated proteins. The unblackened bars represent a normal amino acid sequence,
and the striped and blackened bars represent an aberrant amino acid sequence due to a frameshift mutation of type ⫺1(striped)or⫺2
(blackened). The Xs within the bar designate the repetitive domain. The numbers above the top bar are amino acid numbers for ORF15.
codons in ORF15. Many of the frameshift mutations in
ORF15, especially those in the second half of this exon,
create long stretches of mutant codons prior to a pre-
mature stop codon. Because ORF15 is the terminal exon,
one would predict that the transcripts with these frame-
shift mutations would not be subjected to nonsense-me-
diated decay of mRNA and thus would be translated (as
is the case in dogs with similar mutations in the canine
rpgr gene [Zhang et al. 2002]). We regressed each mea-
sure of ocular function both on the position of the mutant
codon and on the number of mutant codons between the
5
mutant codon and the downstream premature stop
codon simultaneously, to assess their independent effects.
These analyses revealed that both factors independently
affected two measures of ocular function: the dark-
adapted threshold elevation and the 30-Hz ERG ampli-
tude. Specifically, as the position of the 5
mutant codon
increased 5
to 3
, the dark-adapted threshold decreased
( ) and the ERG amplitude increased (Pp.001 P!
) (fig. 4, left panel). As the number of the intervening.001
mutant codons increased, the dark-adapted threshold in-
creased ( ) and the ERG amplitude decreasedPp.03
( ) (fig. 4, right panel). We did not find any sig-P!.001
nificant relationships for visual acuity or visual field area.
However, the refractive error became more negative as
the 5
mutant codon number increased ( ) andPp.002
showed a tendency to become more positive as the num-
ber of intervening mutant codons increased ( ).Pp.06
The encoded amino acid residues of wild-type exon
ORF15 are primarily glycine (with a polar R group)
and glutamate (a negatively charged residue). Mutations
causing a ⫺1 frameshift usually create a downstream
stretch of mutant residues, prior to a premature stop
codon, that are enriched with arginine and lysine (pos-
1144 Am. J. Hum. Genet. 73:1131–1146, 2003
itively charged residues), whereas mutations causing a
⫺2 frameshift convert most downstream residues to ar-
ginine (positively charged) and glycine (polar) (fig. 5).
We did not find a difference between the type of frame-
shift (⫺1or⫺2) and the severity of disease measured
according to any of our four visual function parameters
(data not shown).
Discussion
The present article describes 10 RP2 mutations, 2 of
which are novel, and 80 RPGR mutations, 41 of which
are novel. The majority of the RPGR mutations (53 of
80) were located in ORF15. Among 135 unrelated pa-
tients with prior clinical diagnoses of XLRP, we found
RP2 mutations in 9 patients (6.7%) and RPGR muta-
tions in 98 patients (72.9%), for a total of 79% (table
2). Of the 98 index patients with XLRP with RPGR
mutations, 70 (71.4%) had mutations in ORF15 (table
2). Results obtained in two other comprehensive studies
that included an evaluation of exon ORF15 revealed
different mutation frequencies. In one study, ORF15 mu-
tations were found in 58% of 47 patients mainly from
the United Kingdom and Ireland (Vervoort et al. 2000),
whereas, in a second study, ORF15 mutations were
found in only 30% of 91 patients from North America
(Breuer et al. 2002). In the original study analyzing
ORF15 (Vervoort et al. 2000), as well as in the present
study, many mutations were found in the most repetitive
part of ORF15 (codons 250–357); no mutations were
reported in this region by Breuer et al. (2002). Our data
are more consistent with those provided by Vervoort et
al. (2000) showing that 150% of patients with XLRP
carry pathogenic mutations in ORF15. Mutations in ei-
ther RP2 or RPGR may account for more than the ob-
served 79% of XLRP, since our methods do not detect
all mutations (e.g., some might be located deep in the
introns, in the promoter region, or in sequences 5
or 3
of the gene).
A recent study of the expression of the rpgr gene
(including orf15) in the normal mouse reported that
variable portions of the purine-rich region were spliced
out in the orf15 transcripts; however, the authors did
not exclude the possibility that a full-length transcript
of orf15 existed (Hong and Li 2002). Mutation screen-
ing data presented by others and by our group provide
evidence that mutations in this region cause XLRP, and,
thus, it is likely that this region is being transcribed in
humans.
We reported previously that patients with RP2 mu-
tations have, on average, significantly larger visual fields
and larger ERG amplitudes and a trend toward lower
visual acuities than do patients of comparable age with
RPGR mutations (Sharon et al. 2000); no patients with
ORF15 mutations were included in those comparisons.
We now report, in an analysis of a larger cohort of
patients, that patients with RP2 mutations have signif-
icantly lower visual acuity, on average, than do patients
with RPGR mutations. We could no longer detect a
significant difference in visual field area or 30-Hz ERG
amplitude; however, if we exclude patients with ORF15
mutations, patients with RP2 mutations have borderline
larger visual fields ( ) and significantly largerPp.07
ERG amplitudes ( ) than do patients with RPGRPp.04
mutations (data not shown). Despite the average clinical
differences between patients with RP2 versus RPGR
mutations, a large overlap in our measures of ocular
function was observed, making it impossible to distin-
guish, from these measures of ocular function, whether
any individual patient with XLRP has a mutation in
one gene or the other.
We observed that patients with RPGR mutations in
exons 1–14 had, on average, significantly smaller visual
fields and borderline smaller 30-Hz ERGs than did pa-
tients with ORF15 mutations. Since most of the mu-
tations in exons 1–14 are likely null, the milder clinical
findings among patients with ORF15 mutations suggest
that the expressed RPGR-ORF15 mutant proteins re-
tain some functional properties associated with less se-
vere RP.
Because of the unusually purine-rich nucleotide com-
position of ORF15, there is no stop codon in any of the
three frames in a 729-base region (243 codons) in this
terminal exon. Many frameshift mutations 5
to or within
this region result in a long stretch of mutant codons
downstream to the site of the mutation prior to a ter-
mination codon. This phenomenon is rare in other genes,
because a frameshift mutation will usually result in only
a few abnormal downstream codons followed by a pre-
mature stop codon. Among patients with ORF15 mu-
tations, we found that the longer the encoded wild-type
ORF15 amino acid sequence, the milder the disease with
respect to dark-adapted threshold and ERG amplitude,
and the longer the encoded abnormal amino acid se-
quence, the more severe the disease. We hypothesize that
the RPGR-ORF15 mutants are hypomorphic alleles, be-
cause they are associated with milder disease. This result
is supported by a study in dogs with naturally occurring
rpgr mutations causing XLRP (Zhang et al. 2002). Two
different dog strains with orf15 mutations were evalu-
ated. The retinal degeneration was less severe in the dog
strain with a very short abnormal amino acid sequence
and was more severe in the strain with a long abnormal
amino acid sequence. We could not detect any significant
differences in the severity of disease between patients with
frameshift mutations causing the two different abnormal
frames of ORF15 (see fig. 5), suggesting that the type of
the abnormal sequence does not affect the severity of
disease.
A previous study reported that patients with ORF15
Sharon et al.: Mutations and Clinical Correlations in XLRP 1145
mutations downstream of codon 445 have cone-rod de-
generation (i.e., early preferential loss of cone function
with slight-to-moderate loss of rod function) rather than
typical XLRP (Demirci et al. 2002). Three of our patients
with mutations in this region had a prior diagnosis of
XLRP, and five had cone-rod degeneration. We found
that these patients have, on average, milder disease (i.e.,
a better visual acuity, a smaller elevation of dark-adapted
threshold, and a larger visual field and ERG) than do
patients with mutations upstream of codon 445. More-
over, even among patients matched for cone ERG am-
plitude, those with downstream mutations had signifi-
cantly greater rod ERG function. These findings are
consistent with the diagnosis of cone-rod degeneration
for patients with mutations downstream of codon 445
and suggest that such mutations are more deleterious to
cones than to rods.
In summary, knowledge of which XLRP gene—RP2
or RPGR—is mutant and, if the latter, the site of the
mutation could have implications with respect to esti-
mating long-term visual prognosis. These cross-sec-
tional analyses suggest that, at a given age, patients with
RP2 mutations retain less visual acuity than do patients
with RPGR mutations and that, among patients with
RPGR mutations, those with ORF15 mutations have
milder disease than do patients with mutations in exons
1–14. It remains to be established whether patients with
XLRP with milder disease at a given age will, in fact,
have slower rates of progression over the long term than
patients with XLRP with more severe disease.
Acknowledgments
This study was supported by National Eye Institute grants
EY00169 and EY08683, The Foundation Fighting Blindness,
and The Chatlos Foundation. The authors would like to thank
Dr. Tiansen Li and Terri L. McGee, for fruitful discussions;
Dr. Alan F. Wright, Dr. Debra K. Breuer, and Dr. Anand Swa-
roop, for helpful comments regarding ORF15; and Terri L.
McGee, Scott Adams, and Jonna Grimsby, for technical
support.
Electronic-Database Information
The URLs for data presented herein are as follows:
Authors’ Web site, http://eyegene.meei.harvard.edu/OMGI/
Genes/ORF15.html (for primer sequences) and http://
eyegene.meei.harvard.edu/Genes/RPGRpolys.htm (for table
of sequence changes)
Berkeley Drosophila Genome Project Splice-Site Prediction
Server, http://www.fruitfly.org/seq_tools/splice.html (for
splice-site prediction software)
Online Mendelian Inheritance in Man (OMIM), http://www
.ncbi.nlm.nih.gov/Omim/ (for RPGR, RP2, RP6, and RP24)
References
Andre´asson SOL, Sandberg MA, Berson EL (1988) Narrow-
band filtering for monitoring low-amplitude cone electro-
retinograms in retinitis pigmentosa. Am J Ophthalmol 105:
500–503
Ayyagari R, Demirci F, Liu J, Bingham E, Stringham H, Kakuk
L, Boehnke M, Gorin M, Richards J, Sieving P (2002) X-
linked recessive atrophic macular degeneration from RPGR
mutation. Genomics 80:166
Bader I, Brandau O, Achatz H, Apfelstedt-Sylla E, Hergersberg
M, Lorenz B, Wissinger B, Wittwer B, Rudolph G, Meindl
A, Meitinger T (2003) X-linked retinitis pigmentosa: RPGR
mutations in most families with definite X linkage and clus-
tering of mutations in a short sequence stretch of exon
ORF15. Invest Ophthalmol Vis Sci 44:1458–1463
Berson EL, Rosen JB, Simonoff EA (1979) Electroretino-
graphic testing as an aid in detection of carriers of X-chro-
mosome-linked retinitis pigmentosa. Am J Ophthalmol 87:
460–468
Berson EL, Rosner B, Sandberg MA, Dryja TP (1991) Ocular
findings in patients with autosomal dominant retinitis pig-
mentosa and a rhodopsin gene defect (Pro-23-His). Arch
Ophthalmol 109:92–101
Berson EL, Rosner B, Simonoff E (1980) Risk factors for ge-
netic typing and detection in retinitis pigmentosa. Am J
Ophthalmol 89:763–775
Breuer DK, Yashar BM, Filippova E, Hiriyanna S, Lyons RH,
Mears AJ, Asaye B, Acar C, Vervoort R, Wright AF, Mu-
sarella MA, Wheeler P, MacDonald I, Iannaccone A, Birch
D, Hoffman DR, Fishman GA, Heckenlively JR, Jacobson
SG, Sieving PA, Swaroop A (2002) A comprehensive mu-
tation analysis of RP2 and RPGR in a North American co-
hort of families with X-linked retinitis pigmentosa. Am J
Hum Genet 70:1545–1554
Buraczynska M, Wu W, Fujita R, Buraczynska K, Phelps E,
Andreasson S, Bennett J, Birch DG, Fishman GA, Hoffman
DR, Inana G, Jacobson SG, Musarella MA, Sieving PA, Swa-
roop A (1997) Spectrum of mutations in the RPGR gene
that are identified in 20% of families with X-linked retinitis
pigmentosa. Am J Hum Genet 61:1287–1292
Demirci FYK, Rigatti BW, Wen G, Radak AL, Mah TS, Baic
CL, Traboulsi EI, Alitalo T, Ramser J, Gorin MB (2002) X-
linked cone-rod dystrophy (locus COD1): identification of
mutations in RPGR exon ORF15. Am J Hum Genet 70:
1049–1053
Fishman GA, Grover S, Jacobson SG, Alexander KR, Derlacki
DJ, Wu W, Buraczynska M, Swaroop A (1998) X-linked
retinitis pigmentosa in two families with a missense muta-
tion in the RPGR gene and putative change of glycine to
valine at codon 60. Ophthalmology 105:2286–2296
Gieser L, Fujita R, Goring HH, Ott J, Hoffman DR, Cideciyan
AV, Birch DG, Jacobson SG, Swaroop A (1998) A novel
locus (RP24) for X-linked retinitis pigmentosa maps to
Xq26-27. Am J Hum Genet 63:1439–1447
Guevara-Fujita M, Fahrner S, Buraczynska K, Cook J, Whea-
ton D, Cortes F, Vicencio C, Pena M, Fishman G, Mintz-
Hittner H, Birch D, Hoffman D, Mears A, Fujita R, Swaroop
A (2001) Five novel RPGR mutations in families with X-
linked retinitis pigmentosa. Hum Mutat 17:151
1146 Am. J. Hum. Genet. 73:1131–1146, 2003
Hardcastle AJ, Thiselton DL, Van Maldergem L, Saha BK, Jay
M, Plant C, Taylor R, Bird AC, Bhattacharya S (1999) Mu-
tations in the RP2 gene cause disease in 10% of families
with familial X-linked retinitis pigmentosa assessed in this
study. Am J Hum Genet 64:1210–1215
Hardcastle AJ, Thiselton DL, Zito I, Ebenezer N, Mah TS,
Gorin MB, Bhattacharya SS (2000) Evidence for a new locus
for X-linked retinitis pigmentosa (RP23). Invest Ophthalmol
Vis Sci 41:2080–2086
Hong DH, Li T (2002) Complex expression pattern of RPGR
reveals a role for purine-rich exonic splicing enhancers. In-
vest Ophthalmol Vis Sci 43:3373–3382
Hong DH, Yue G, Adamian M, Li T (2001) Retinitis pig-
mentosa GTPase regulator (RPGRr)-interacting protein is
stably associated with the photoreceptor ciliary axoneme
and anchors RPGR to the connecting cilium. J Biol Chem
276:12091–12099
Mears AJ, Gieser L, Yan D, Chen C, Fahrner S, Hiriyanna S,
Fujita R, Jacobson SG, Sieving PA, Swaroop A (1999) Pro-
tein-truncation mutations in the RP2 gene in a North Amer-
ican cohort of families with X-linked retinitis pigmentosa.
Am J Hum Genet 64:897–900
Meindl A, Dry K, Herrmann K, Manson F, Ciccodicola A,
Edgar A, Carvalho MR, Achatz H, Hellebrand H, Lennon
A, Migliaccio C, Porter K, Zrenner E, Bird A, Jay M, Lorenz
B, Wittwer B, D’Urso M, Meitinger T, Wright A (1996) A
gene (RPGR) with homology to the RCC1 guanine nucle-
otide exchange factor is mutated in X-linked retinitis pig-
mentosa (RP3). Nat Genet 13:35–42
Miano MG, Testa F, Strazzullo M, Trujillo M, De Bernardo
C, Grammatico B, Simonelli F, Mangino M, Torrente I, Ru-
berto G, Beneyto M, Antinolo G, Rinaldi E, Danesino C,
Ventruto V, D’Urso M, Ayuso C, Baiget M, Ciccodicola A
(1999) Mutation analysis of the RPGR gene reveals novel
mutations in south European patients with X-linked retinitis
pigmentosa. Eur J Hum Genet 7:687–694
Ott J, Bhattacharya S, Chen JD, Denton MJ, Donald J, Dubay
C, Farrar GJ, et al (1990) Localizing multiple X chromo-
some-linked retinitis pigmentosa loci using multilocus ho-
mogeneity tests. Proc Natl Acad Sci USA 87:701–704
Paul SR, Fung KY (1991) A generalized extreme studentized
residual multiple-outlier-detection procedure in linear re-
gression. Technometrics 33:339–348
Pusch CM, Broghammer M, Jurklies B, Besch D, Jacobi FK
(2002) Ten novel ORF15 mutations confirm mutational hot
spot in the RPGR gene in European patients with X-linked
retinitis pigmentosa. Hum Mutat 20:405
Roepman R, van Duijnhoven G, Rosenberg T, Pinckers AJ,
Bleeker-Wagemakers LM, Bergen AA, Post J, Beck A, Rein-
hardt R, Ropers HH, Cremers FP, Berger W (1996) Posi-
tional cloning of the gene for X-linked retinitis pigmentosa
3: homology with the guanine-nucleotide-exchange factor
RCC1. Hum Mol Genet 5:1035–1041
Rosner B (2000) Fundamentals of biostatistics. 5th ed. Dux-
bury Press, Boston
Rozet JM, Perrault I, Gigarel N, Souied E, Ghazi I, Gerber S,
Dufier JL, Munnich A, Kaplan J (2002) Dominant X linked
retinitis pigmentosa is frequently accounted for by truncat-
ing mutations in exon ORF15 of the RPGR gene. J Med
Genet 39:284–285
Sandberg MA, Weigel-DiFranco C, Dryja TP, Berson EL (1995)
Clinical expression correlates with location of rhodopsin
mutation in dominant retinitis pigmentosa. Invest Ophthal-
mol Vis Sci 36:1934–1942
Schwahn U, Lenzner S, Dong J, Feil S, Hinzmann B, van Duijn-
hoven G, Kirschner R, Hemberger M, Bergen AA, Rosen-
berg T, Pinckers AJ, Fundele R, Rosenthal A, Cremers FP,
Ropers HH, Berger W (1998) Positional cloning of the gene
for X-linked retinitis pigmentosa 2. Nat Genet 19:327–332
Sharon D, Bruns GA, McGee TL, Sandberg MA, Berson EL,
Dryja TP (2000) X-linked retinitis pigmentosa: mutation
spectrum of the RPGR and RP2 genes and correlation with
visual function. Invest Ophthalmol Vis Sci 41:2712–2721
Thiselton DL, Zito I, Plant C, Jay M, Hodgson SV, Bird AC,
Bhattacharya SS, Hardcastle AJ (2000) Novel frameshift
mutations in the RP2 gene and polymorphic variants. Hum
Mutat 15:580
Vervoort R, Lennon A, Bird AC, Tulloch B, Axton R, Miano
MG, Meindl A, Meitinger T, Ciccodicola A, Wright AF
(2000) Mutational hot spot within a new RPGR exon in X-
linked retinitis pigmentosa. Nat Genet 25:462–466
Vervoort R, Wright AF (2002) Mutations of RPGR in X-linked
retinitis pigmentosa (RP3). Hum Mutat 19:486–500
Westall CA, Dhaliwal HS, Panton CM, Sigesmun D, Levin AV,
Nischal KK, Heon E (2001) Values of electroretinogram re-
sponses according to axial length. Doc Ophthalmol 102:
115–130
Yang Z, Peachey NS, Moshfeghi DM, Thirumalaichary S,
Chorich L, Shugart YY, Fan K, Zhang K (2002) Mutations
in the RPGR gene cause X-linked cone dystrophy. Hum Mol
Genet 11:605–611
Zhang Q, Acland GM, Wu WX, Johnson JL, Pearce-Kelling
S, Tulloch B, Vervoort R, Wright AF, Aguirre GD (2002)
Different RPGR exon ORF15 mutations in canids provide
insights into photoreceptor cell degeneration. Hum Mol Ge-
net 11:993–1003
Zito I, Morris A, Tyson P, Winship I, Sharp D, Gilbert D,
Thiselton DL, Bhattacharya SS, Hardcastle AJ (2000) Se-
quence variation within the RPGR gene: evidence for a foun-
der complex allele. Hum Mutat 16:273–274
Zito I, Thiselton DL, Gorin MB, Stout JT, Plant C, Bird AC,
Bhattacharya SS, Hardcastle AJ (1999) Identification of
novel RPGR (retinitis pigmentosa GTPase regulator) mu-
tations in a subset of X-linked retinitis pigmentosa families
segregating with the RP3 locus. Hum Genet 105:57–62
