Reports
Individual Variation in the Size of
the Human Red and Green Visual
Pigment Gene Array
Jennifer P. Macke and Jeremy Nathans
Purpose. To determine the size variation of the X-chromosomal human red and green visual pigment gene
array in the general population using pulsed field gel
electrophoresis and Sotithern blotting.
Methods. Peripheral blood lymphocytes were prepared
from 67 anonymous males. The cells were embedded in
agarose and the genomic DNA digested with restriction
enzyme Not I. The resulting DNA fragments were resolved on a contour-clamped homogeneous electric
field gel, and the Not I fragment containing the red
and green pigment genes was visualized by Southern
blot hybridization with a human green pigment cDNA
probe.
Results. In DNA from each male, a single hybridizing
fragment was observed in Not I-digested DNA. The
lengths of the fragments from different males were observed to vary in steps of approximately 39 kilobases
(kb), consistent with earlier studies showing a visual
pigment gene repeat unit of 39 kb and a head-to-tail
tandem arrangement of the red and green visual pigment genes. In the population studied, the number of
repeat units per X-chromosome had a mean of 2.9 and
a standard deviation of 0.94.
Conclusions. The sizes of visual pigment gene arrays observed in this study resemble those determined in earlier studies based on ratios of restriction fragments resolved by conventional gel electrophoresis and visualized by whole genome Southern blotting, but differ
significantly from those determined using ratios of fragments obtained by the polymerase chain reaction. Invest Ophthalmol Vis Sci. 1997;38:1040-1043.
J. he human red and green pigment genes lie in a
head-to-tail tandem array on the X-chromosome.1"3
Variation in the number and structure of these genes
arises from unequal homologous recombination between nearly identical 39-kilobase (kb) repeat units,
each of which contains a single gene.1"3 The common
red-green dichromacies and anomalous trichroma-
From the Deportments of Molecular Biology and Genetics, Neuroscience, and
Ophthalmology, Howard Hughes Medical institute, Johns Hopkins University School
of Medicine, Baltimore, Maryland.
Supported by the Howard Hughes Medical Institute.
Submitted for publication August 2, 1996; revised November 11, 1996; accepted
November II, 1996.
Proprietary interest category: N.
lieprinl requests: /eremy Nathans, 805 PCTB, 725 North Wolfe Street, Johns
Hopkins University School of Medicine, Baltimore, MD 21205.
1040
cies found in approximately 10% of human males
arise from homologous recombination events that result in the loss of genes or the production of hybrid
genes.4'5 Differences in the number of X-linked visual
pigment genes also are found among normal trichromats and constitute one of the most common genetic
variations in the human population. 1 ' 6
The number of red and green pigment genes per
X-chromosome in the general population has been
measured by several laboratories with differing results.1'6'7 By conventional gel electrophoresis and
quantitative Southern blotting,''4~6 the average number of genes in the array was determined to be three,
with the majority of arrays having a single red pigment
gene and either one, two, or three green pigment
genes. In these studies, arrays having six or more genes
were found in, at most, a few percent of X-chromosomes. By contrast, a recent study using the polymerase chain reaction (PCR) concluded that the average
number of genes in the array was greater than four,
that visual pigment gene arrays with six or more genes
are found on 30% of X-chromosomes, and that more
than 40% of human X-chromosomes carry two or
more red pigment genes.7 In these experiments, corresponding segments from each of the highly homologous genes in the X-linked visual pigment gene array
were amplified by PCR, and the products derived from
each gene were distinguished by restriction enzyme
cleavage at sites that were known from gene sequencing to be present in only one or the other type of Xlinked visual pigment gene. The PCR-based determination, if correct, would result in a substantial revision
in current models of the evolution and physiology of
human color vision.
In this report, we describe a determination of the
size of the red and green visual pigment gene array
in a group of 67 human X-chromosomes using pulsed
field gel electrophoresis (PFGE) and Southern blotting. This method is distinguished from those used
previously in that it does not rely on quantitating ratios
of DNA fragments. In both the conventional gel electrophoresis-Southern blot method 1 ' 4 " 6 and the PCR
method,' the intensities of red and green pigment
gene fragments are compared to one another and to
an internal standard to obtain both the total number
of genes and the ratio of gene types (Figs. lAand IB).
In the PFGE-Southern blot method, the total number
of visual pigment genes is determined by sizing a Not
I fragment that carries the entire visual pigment gene
array2 (Fig. 1C). This method relies on the observation
that each 39-kb tandem repeat unit within the visual
pigment gene array contains a single visual pigment
Investigative Ophthalmology & Visual Sc
Copyright © Association for Research in
, April 1997, Vol. 38, No. 5
i, and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933200/ on 06/14/2017
Reports
1041
conventional
Southern
blotting
G:R =
B
PCR,
restriction
digestion
D:P = 2:1
G:R = 2:1
PFGE,
Southern
blotting
Not I
Not I
I
length = 1 57 kb
FIGURE l. Methods for quantitating red and green pigment gene number. The example
shown is for an array with one red and two green pigment genes. (A) Conventional gel
electrophoresis and Southern blotting; R and G represent distinguishable restriction fragments derived from the red and green pigment genes, respectively.1'4" (B) polyrnerase chain
reaction (PCR)7; distinguishable PCR products are derived from the red (R) or green (G)
pigment genes and from the most proximal (P) or from all of the distal (D) pigment genes.7
(C) Pulse field gel electrophoresis and Southern blotting.2
gene, that repeat units lack Not I sites, and that 40 kb
of nonrepeated flanking DNA also is contained within
the Not I fragment that contains the visual pigment
gene array. 1 " 3
MATERIALS AND METHODS. Blood samples
from 67 anonymous males of unknown color vision
phenotype were obtained from the Johns Hopkins
Hospital. The blood samples were collected in accordance with the regulations of the Johns Hopkins Joint
Committee on Clinical Investigation. This study adhered to the Declaration of Helsinki. An effort was
made to match the ethnic composition of the populations studied previously by Nathans et al1 and Drummond-Borg et al6 by selecting subjects with names
indicative of European ancestry. Peripheral blood lymphocytes were prepared by Ficoll-Paque purification
(Pharmacia, Piscataway, NJ), embedded in agarose,
and digested with Not I as described.8 Fragments were
resolved on a contour-clamped homogeneous electric
field gel (Bio-Rad, Melville, NY) for 14 hours at 200
V with a switch time ramping linearly from 0.2 to 22
seconds. The Not I fragment containing the red and
green pigment genes was visualized by Southern blot
hybridization using a human green pigment cDNA
clone, hs2, which is 98% identical in sequence to the
corresponding red pigment gene exons.1 In each of
the 67 DNA samples examined, a single hybridizing
Not I fragment was observed.
RESULTS. Figure 2 shows a typical Southern blot
in which genomic DNA from 12 males selected randomly was digested with Not I, resolved by PFGE, and
hybridized with a green pigment cDNA probe to visualize the Not I fragment containing the red and green
pigment gene array. As expected, a single hybridizing
fragment is seen in the DNA of each male, and the
array lengths vary in steps of approximately 39 kb.2'3
The number of repeat units is determined readily by
measuring the fragment lengths and is apparent by
visual inspection (Fig. 2). Figure 3A shows a histogram
of red and green pigment gene array lengths for 67
random males as determined by PFGE. The observed
distribution of visual pigment gene number is: 1 gene,
3 subjects; 2 genes, 22 subjects; 3 genes, 26 subjects;
4 genes, 13 subjects; and 5 genes, 3 subjects.
DISCUSSION. In this report, we apply PFGE to
determine the number of red and green pigment gene
repeat units per X-chromosome in the human population. The size distribution of DNA fragments containing the entire red and green pigment gene locus
observed by PFGE resembles closely the distributions
determined previously by conventional gel electro-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933200/ on 06/14/2017
1042
Investigative Ophthalmology & Visual Science, April 1997, Vol. 38, No. 5
291
242.5
194
145.5
97
an excess of even over odd numbers of visual pigment
genes and a complete absence of subjects with a single
green pigment gene (Figs. 3F, 3G). One explanation
for these data is that in some or all of these 12 subjects,
the ratio of PCR products used to determine the total
number of genes is in error by a factor of 2, thus
leading to a 2-fold overestimate of the total number
of visual pigment genes. In contrast, the number of
visual pigment genes in the 15 subjects7 who were
reported to have a single red pigment gene (Fig. 3E)
is consistent with the distributions obtained by the 2
other methods (Figs. 3A, 3B, and 3C).
48.5 -
FIGURE 2. Example of individual differences in red and
green pigment gene array lengths. Genomic DNA from 12
random males was digested with Not I, separated by polymerase chain reaction, transferred to nitrocellulose, and hybridized to visualize the red and green pigment gene array.4 The
lengths (in kilobases) of lambda multimer size standards are
indicated to the left. Variations in band intensities reflect
both variation in gene number and in the amount of DNA
loaded per track. The number of genes in the X-linked visual
pigment gene arrays is (from left to right): 3, 3, 2, 2, 3, 2,
4, 3, 4, 2, 3, 2.
phoresis and Southern blotting (Figs. 3A, 3B, and 3C),
but differs significantly from the size distribution obtained by PCR as reported by Neitz and Neitz7 (Fig.
3D; P < 0.001 for both untransformed or log-transformed data; Student's i-test). In particular, the distribution determined by the PCR method differs from
the other distributions by the presence of subjects with
large numbers of genes: 6 or more pigment genes
were reported in 8/27 subjects (30%) using the PCR
method, whereas in the other studies, 6 or more pigment genes were found in 0/67, 0/18, and 4/134
subjects (Fig. 3). This difference in reported gene distributions could arise either from differences in test
populations or from an error in experimental method.
One source of variability that might account for
the discrepancy in pigment gene distributions could
arise from DNA sequence variations that alter the efficiency of PCR priming or amplification. For example,
if 2 DNA segments exhibited a 3% difference in amplification efficiency per cycle, after 25 cycles of PCR,
there would be a 2-fold difference in final product
yield. The general possibility of a systematic error in
the PCR method7 is suggested by the data presented
in Table 1 of that article, which shows an unexpected
distribution of visual pigment gene number in those
12 subjects who were reported to carry more than 1
red pigment gene. Among these 12 subjects, there is
Neitz & Neitz, 1995
4.3+/-1.9
N=27
0)
1 2 3 4 5 6 7 8 9
10
.a
ic
«= 75
1 I I
Drummond-Borg et al., 1989
3.2+/-1.0
N=134
1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
SO
Neitz & Neitz, 1995
4 red pigment
genes; N=Z
2S
number of genes
FIGURE 3. Histograms showing the number of genes in the
X-linked visual pigment gene array determined by pulsed
field gel electrophoresis-Southern blotting (A, this study),
conventional gel electrophoresis-Southern blotting (B1;
C6), or PCR (D7). The mean number of genes ± the standard deviation is indicated for each distribution. The histogram in (C) represents a study of 134 random white males
of unknown color vision phenotype. 6 The arrangements determined for those red and and green pigment gene arrays
suggested that 113 of the subjects had normal trichromatic
vision and 21 had some degree of color anomaly. In (C), the
113 genotypes consistent with normal trichromatic vision are
represented by filled bars, and the 21 genotypes indicative
of dichromacy or anomalous trichromacy are represented
by open bars. For the 113 presumed normal trichromat subjects, the mean number of genes ± the standard deviation
is 3.15 ± 0.9; the figure indicates the mean number of genes
± the standard deviation for the entire set of 134 subjects.
(E to G) Histograms of the number of visual pigment genes
in the 15 subjects7 reported to carry one red pigment gene
(E), two red pigment genes (F), or four red pigment genes
(G). No subjects in that study were reported to carry three
or five red pigment genes.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933200/ on 06/14/2017
Reports
The distribution of red and green pigment gene
number determined in this report supports the conclusion that the majority of human red and green
pigment gene arrays consists of a single red pigment
gene and between one and three green pigment
genes. It now should be possible to determine how
this basic arrangement can be modified by amino acid
substitution and homologous recombination to generate the variety of red-green color vision types present
in the human population.
1043
2.
3.
4.
5.
Key Words
pulsed field gel electrophoresis, red-green color vision, visual pigments
6.
Acknowledgment
The authors thank Ms. Ingeborg Hobbs (Johns Hopkins
Hospital) for assistance in obtaining blood samples.
7.
References
8.
of human color vision: The genes encoding blue,
green, and red pigments. Science. 1986;232:193-202.
Vollrath D, Nathans J, Davis RW. Tandem array of
human visual pigment genes at Xq28. Science.
1988;240:1669-1672.
Feil R, Aubourg P, Helig R, Mandel JL. A 195-kb cosmid walk encompassing the human Xq28 color vision
pigment genes. Genomics. 1990; 6:367-373.
NathansJ, Piantanida TP, Eddy RL, Shows TB, Hogness DS. Molecular genetics of inherited variation in
human color vision. Science. 1986; 232:203-210.
Deeb SS, Lindsey DT, Sanocki E, Winderickx J, Teller
DY, Motulsky AG. Genotype-phenotype relationships
in human red/green color vision defects: Molecular
and psychophysical studies. Am J Hum Genet.
1992;51:687-700.
Drummond-Borg M, Deeb SS, Motulsky AG. Molecular patterns of X-chromosome linked color vision
genes among 134 men of European ancestry. ProcNatl
AcadSci USA. 1989;86:983-987.
Neitz M, NeitzJ. Numbers and ratios of visual pigment
genes for normal red-green color vision. Science.
1995;267:1013-1016.
Finney M. Pulse field gel electrophoresis. In: Ausubel
FM, ed. Current Protocols in Molecular Biology. New York:
1. NathansJ, Thomas D, Hogness DS. Molecular genetics
Intercellular Adhesion Molecule-1 in
Proliferative Vitreoretinopathy
G. A. Limb* A. H. Chigneltf C.J. Cole*
W. T. Green,\ L. Webster* R. D. Hollifield*
and D. C. Dumonde *
Purpose. To measure vitreous levels of the soluble intercellular adhesion molecule (sICAM-1) in eyes with rhegmatogenous retinal detachment (RRD) complicated or
uncomplicated by proliferative vitreoretinopathy (PVR)
to investigate whether levels of this molecule related to
history of previous retinal surgery or to the duration
and severity of PVR.
Methods. The authors measured vitreous sICAM-1 by
enzyme-linked immunosorbent assay in 28 eyes with
PVR and 35 eyes with uncomplicated RRD. Vitreous
from 10 eyes with macular holes and from 12 cadaveric
eye donors were used as control specimens.
From the Departments of * Immunology and t Ophthalmology, St. Thomas' Hospital,
UMDS, London, England.
Supported by the Gift of Thomas Podtlinglon Trust, the Guide Dogs for the Blind
Association, and the Iris Fund for the Prevention of Blindness.
Submitted for publication August 1, 1996; revised October 25, 1996; accepted
December 5, 1996.
Projmetary interest category: N.
Reprint requests: G. A. Limb, Ocular Inflammation Research Laboratory, The
Rayne Institute, St Thomas' Hospital, London SE1 7EH.
J. Wiley & Sons; 1995:2.5.9-2.5.14.
Results. Vitreous sICAM-1 levels were higher in the
group with RRD complicated by PVR as a whole than
in the group with RRD alone or in the control groups.
In patients with no previous retinal surgery, there was
no difference in vitreous sICAM-1 levels between the
groups with RRD alone and RRD complicated by PVR.
However, in patients who had undergone previous external surgery, those with PVR showed higher levels of
vitreous sICAM-1 than those with RRD alone. In PVR,
raised levels of sICAM-1 were associated preferentially
with a history of previous vitrectomy as well as with a
longer duration of the condition, although these levels
were not related to the grade of PVR. In eyes with RRD
alone, the levels of sICAM-1 were not enhanced with
the duration of the detachment. Despite showing high
vitreous levels of sICAM-1, patients with PVR did not
exhibit increased serum levels of this adhesion molecule.
Conclusions. The current observations suggest that those
persons in whom PVR develops may have an impairment of the mechanisms that control the inflammatory
response to retinal trauma. Persistently raised vitreous
levels of sICAM-1 point to the continued operation of
cytokine-mediated vascular reactions at the blood-retinal barrier. Invest Ophthalmol Vis Sci. 1997; 38:10431048.
X roliferative vitreoretinopathy (PVR) is a severe complication of retinal detachment (RRD) characterized
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933200/ on 06/14/2017