Color vision defects have been recognized for almost 200 years. The famous chemist Dalton reported his own color vision defect to the Manchester Literary and Philosophical Society in 1794.2 Dalton reported that an “image which others call red appears to me as little more than a shade or defect of light.” Orange, yellow, and green appeared to him as different shades of yellow. Color vision defects are sometimes referred to as daltonism. A recent study of Dalton's DNA from his preserved eyes indicated that he was a deuteranope lacking the green (middle-wave) pigment81 (see below).
Red-green color vision defects are a group of X-linked abnormalities that are common in the male European population (8 percent).82–84 Several types of color vision defects exist82–84 (Table 238-2). When the color-sensitive receptors for either the long-wave pigment (red) or the middle-wave pigment (green) are completely absent, color vision is dichromatic rather than trichromatic. Such individuals have only two, instead of the normal three, visual pigments (i.e., either green and blue or red and blue, but not green, red, and blue). Depending on whether long-wave (red) or middle-wave (green) pigment is absent, dichromats are classified as deuteranopes, who lack functional cones with green pigment (G−), or protanopes, who lack cones with red visual pigment (R−). The complete lack of the respective visual pigments in dichromats was proven first by microspectrophotometry that failed to detect absorption of the characteristic wavelength for the red pigment in protanopes (R−) and for the green pigment in deuteranopes (G−).31,32,78,85 The frequency of these traits in the Caucasian population ranges around 1 percent for protanopia as well as for deuteranopia. Molecular studies demonstrate complete absence of the green visual pigments in deuteranopes, whereas protanopia has been associated with red-green fusion genes that have the characteristics of normal green pigment.11,19
Table 238-2: Classification of Common Color Vision Defects |Favorite Table|Download (.pdf) Table 238-2: Classification of Common Color Vision Defects
|Type ||Inheritance ||Frequency (Europeans) ||Type of Color Vision ||Vision Defect |
| Red-green defects |
|Protanopia (R–) ||X-linked recessive ||~1% ||Dichromatic ||Severe red-green color confusion |
|Deuteranopia (G–) ||X-linked recessive ||~1% ||Dichromatic ||Moderate to severe red-green color confusion |
|Protanomaly (R') ||X-linked recessive ||~1% ||Trichromatic ||Mild red-green color confusion |
|Deuteranomaly (G') ||X-linked recessive ||~4–5% ||Trichromatic ||Mild red-green color confusion |
| Blue-yellow defects |
|Tritanopia* ||Autosomal dominant ||1 in 500 (or fewer) (?) ||Dichromatic* ||Blue-yellow color confusion |
|Tritanomaly* ||Autosomal dominant ||? ||Trichromatic* ||Mild blue-yellow color confusion |
The other X-linked color vision anomalies are associated with trichromatic color vision and are known as deuteranomaly (G′) and protanomaly (R′)82 (see Table 238-2). In these defects, all three visual pigments are present, but microspectrophotometric measurements showed displacement of the green and red pigment spectral sensitivity curves in deuteranomaly and protanomaly, respectively, and suggested the presence of visual pigments with altered spectral characteristics.86–88 Deuteranomalous and protanomalous subjects thus were assumed to carry variant visual pigments. Molecular studies indicated the presence of green-red fusion genes in deuteranomaly (G′)11,19 and red-green fusion genes in protanomaly (R′).11,19 However, a few deuteranomalous individuals had a complete absence of green pigment genes.19 Current molecular methodology has been unsuccessful in differentiating protanopes from protanomalous individuals, since both have red-green fusion genes (see below for details). Furthermore, green-red fusion genes are not infrequently found in persons with normal color vision19,67,71 but presumably are located in downstream locations where they are not expressed (see below).
Considerable variation in the severity of the trichromatic abnormalities (protanomaly and deuteranomaly) has been observed.82,83 Some scientists have divided the various trichromatic anomalies into mild, moderate, and severe deuteranomaly or protanomaly.89 It would be expected that a given subtype of defect generally would express with a similar degree of severity in all affected male members of a family who carry the same mutation. However, intrafamilial variability of color vision defects has been observed occasionally as in other examples of identical mutations of a variety of human genetic diseases. Further work to relate the molecular lesion to the severity of the defect needs to be done. The frequency of deuteranomaly in Europeans ranges between 4 and 5 percent, whereas the frequency of protanomaly is around 1 percent (see ref.82).
Abnormalities affecting the blue- or short-wave-sensitive pigment are known as tritanopia.90 (The origin of the terms protan, deutan, and tritan is derived from the Greek and refers to the first, second, and third variants of color defect.) The differentation of tritanopia from tritanomaly is not as clear for the green-red defects. Tritanopia defects are much rarer than deutan or protan abnormalities but have been estimated to occur as frequently as 1 in 500 individuals.91 Tritanopic individuals have problems with the perception of blue color. The defect is inherited as an autosomal dominant trait, as might be expected by the location of the blue pigment gene of chromosome 7. The molecular basis of tritan abnormalities are missense mutations, and three different amino acid substitutions (see below) have been observed.92,93 Variability in expression has been noticed in several pedigrees. In fact, some individuals with the characteristic amino acid substitution did not manifest with the phenotypic color vision defect.92
Unlike the amino acid substitutions of rhodopsin that often are associated with rod degeneration leading to autosomal dominant retinitis pigmentosa (see Chap. 235), there are no other ophthalmologic or clinical consequences associated with the various deutan, protan, and tritan types of color vision defects. Visual acuity is unaffected. The color vision defects are expressed early in life94 and remain constant throughout the life span.
Phenotypic Detection of Color Vision Defects82,95,96
A large number of different tests have been devised for the detection of color vision anomalies. Most of these have not been standardized and are not in general use. Therefore, they will not be discussed here. The tests discussed below are most useful for detection of genetic color vision defects.
Pseudoisochromatic plates are used widely for screening of color vision defects. The figures to be discriminated on these plates appear in shades of different chromatic quality. These tests use patterns of variously colored printed dots that usually are shaped as numbers. The subject is asked to read or trace a shape or number. The charts are so designed that persons with color vision defects will either miss shapes or numbers and/or will see different shapes than persons with normal color vision. The most widely used variety are the Ishihara (Japan) charts and the American Optical H-R-R (Hardy-Rand-Ritter) polychromatic plates. Illumination should be standardized at diffused daylight during testing (100-W blue daylight bulb or MacBeth easel lamp), since various pigments in the charts may vary in gloss, giving clues to a color-defective person. Ordinary tungsten bulbs may allow deuteranomalous persons to read test charts correctly. Ishihara charts have had the most use. Distinction between deutan and protan abnormalities is usually possible, but no definite reliance for subclassification should be placed on the charts' results because severe anomalous trichromats often cannot be differentiated from dichromats.
Ishihara charts do not detect tritanopia, whereas the H-R-R charts will. When color vision abnormalities are found on chart testing, anomaloscopy (see below) usually is done in genetic studies to confirm the type of color vision defect.
The Farnsworth-Munsell 100 Hue Test97 has been used widely to evaluate chromatic discrimination loss. In this test, a series of 85 color chips is arranged in their natural order of hue. Each chip has an appropriate number on its back that refers to its order in the series. Depending on the mistakes made, a standardized score is calculated and recorded on a special chart. Characteristic patterns for protans, deutans, and tritans are obtained, but differentiation between protanomaly and protanopia is difficult.
Anomaloscopy has been used widely and is based on color matching. Lord Rayleigh devised a simple test system to classify individuals with red-green color vision abnormalities. The observer views a pure yellow light (589–590 nm) on one-half of a screen, while the other half of the screen projects a mixture of red (650 nm) and green (545–550 nm) lights. The brightness or intensity of the yellow light, as well as the proportion of the green and red lights, are adjusted by the subject until both hemifields appear identical in color and brightness. Under the color condition of the Rayleigh match, color detection by the short-wave-sensitive or blue-pigment cones is negligible. The most frequently used instrument is the Nagel anomaloscope. The range of accepted matches of mixtures of green and red light against yellow is recorded, as is the midpoint of such matches. Figure 238-7 shows typical findings for normal and various color-defective persons.
Anomaloscopic Rayleigh match ranges of protan, deutan, and normal male individuals. Each horizontal line represents the range of mixtures of red and green lights that the observer could not distinguish from the standard yellow light. Identification numbers shown next to the horizontal lines refer to subjects of a large study.19 Subject 2114 (- - - -) had a variable and unreliable match range. P, Protanopic; PA, protanomaly; D, deuteranopia; DA, deuteranomaly; N, normal. (Used with permission from Deeb et al.19 )
Normal individuals accept matches in a narrow range. Dichromats such as protanopic and deuteranopic subjects will match yellow with any and all ratios of red and green, including red and green alone. Thus any sufficiently bright red-green mixture will produce a match. Dichromatic deuteranopes require much more yellow to match pure red than dichromatic protanopes, who need only a small amount of yellow to match the red field, which they perceive as of low intensity.
Protanomalous subjects produce match ranges that are shifted to the red side of the spectrum, whereas the matches of deuteranomalous subjects are displaced to the green. Subjects with severe deuteranomaly and severe protanomaly tend to have quite wide characteristically displaced match ranges, whereas those with milder anomalous defects have narrower match ranges.
Most large-scale investigations of color vision defects start with assessment of color vision by plate test followed by anomaloscopy only among those individuals who fail the test. Anomaloscopic findings in a given individual are constant and do not change. As might be expected, anomaloscopic results of different affected family members usually but not always are similar. The viewing angle conventionally is 2°. With larger viewing angles, most protanopes are classified as protanomalous, and some deuteranopes as deuteranomalous.
An objective assessment of color vision may be possible with ERG. All other test measurements are based on the observer's subjective perception of color. In ERG, a corneal electrode placed on an anesthetized dilated eye records the retinal response to standardized flashes of light. Because of the somewhat invasive nature of the test, no extensive experience of the various benign genetic color vision defects has as yet been reported. However, color vision defects characterized as deutan (36 subjects) and protan (32 subjects) (with no indication of the proportion of dichromats and trichromats) could be discriminated by the log ratio of the sensitivity at short (480-nm) and long (620-nm) wavelengths (sensitivity quotient).99 Female carrier detection was particularly successful for deutan carriers and less so for protan heterozygotes.99
Color Perception and Color Defects: Practical Implications100
Color vision defects affect perceptions of color. A large proportion of the population, therefore, lives in a different perceptual world than those with normal color vision. Normal trichromatic color vision helps to define objects in complex multicolored settings, as can be observed by comparing a colored photograph with a black-and-white rendering of the same scene. The color perception of color-defective observers has been studied in a few individuals who for unknown reasons were color defective in one eye only. One otherwise healthy young woman was deuteranopic in the left eye and color normal in the right eye (see ref.82). Her color-vision-defective eye had only three color sensations—gray, yellow, and blue—and lacked any green and red sensations. Her normal eye gave normal color vision. In a film showing the color world of this woman (MN8246, Color Vision Deficiency, Research Division, Bureau Medicine and Surgery, Department of the Navy, Washington), a room furnished entirely from gray, blue, and yellow materials had the same color appearance to her defective eye as similar objects covering the entire range of colors, including red and green. Color perception of the usual deuteranopes conforms to this general pattern. Dichromatic protanopia appears as a more severe defect than deuteranopia. Protanopes confuse not only red, yellow, and green, like deuteranopes, but also deep red, dark brown, or even black and have particular problems with red color perception. A ripe red fruit may be considered black by a protanope.
Color perception differences in anomalous trichromats are more subtle than in dichromats. Green and red are not absent but appear weakened in intensity. Deuteranomaly is considered to be the mildest anomaly. More subtle differences in color perception are seen in individuals who either have an alanine (62 percent) or a serine (38 percent) at position 180 of the red pigment gene. Those with 180 serine (≈62 percent) perceive a deeper red than those who have serine at this position. Based on Hardy-Weinberg statistics, it is expected that about 47 percent of Caucasian women are heterozygotes for both the 180 alanine and 180 serine variant. Due to X inactivation, about half the cones in most women will have alanine at position 180, and the other half have serine at that site. Such women have four types of cones—blue, green, and two types of red cones—and might have tetrachromatic vision. Further tests need to be done to determine whether such females have superior color discrimination under certain conditions as compared with trichromats.
Color-vision-defective persons, including dichromats, usually have no problem with the naming of colors. Apparently there are enough differences in color sensation to allow the denotation of the appropriate color as taught by parents and teachers.
It has been claimed by anecdotal evidence that color-defective observers can see through colored patterns that deceive normal observers.101 This advantage would be useful under military conditions and could have played a role in the evolution of color vision.102 Dichromats, in fact, perform better than normal trichromats under experimental conditions where texture was camouflaged by color.103 No anomalous trichromats were tested in this study. A recent study showed that color-blind individuals have a lower threshold for light perception than color-normal individuals.104 This would give color-blind individuals a selective advantage under low light conditions.
X-linked red-green color vision defects are cardinal examples of common genetic traits that have been used for some time to exclude applicants from a variety of industrial, marine, air, rail, and military occupations that require the ability to distinguish colors. A genetic condition per se is not reason for occupational discrimination unless such a trait makes its carriers unable to carry out a job-related task. Occupational exclusion, however, is appropriate if an affected employee's work places others at physical risk. There is no consensus as to whether the risk of physical harm to the affected individual alone is sufficient reason for job exclusion. Some would leave such decisions to the affected person, particularly if the possible damage is uncertain or may only occur in the distant future. The problems in applying these rules in a fair manner to defects of color vision comes from difficulties in interpreting abnormal color vision tests. Thus, when actual tasks requiring color discrimination instead of various artificial testing systems such as plate tests are used, many color-defective persons may perform adequately.105,106 The validity of rigid color vision standards for occupational selection, therefore, has been questioned (see ref.107). Cole108 suggested variable standards for the many different occupations that require color discrimination. It has been suggested that the specifications for practical testing and its rigor should be based on the probability, severity, and socioeconomic consequences of adverse reactions that may be caused by difficulties with color discrimination.
In general, deuteranopic and protanopic dichromats perform more poorly on color discrimination than those with anomalous trichromacy. Most, but not all, dichromats have difficulties in selection of colored articles and materials, including foods. No studies have shown that color vision defects have been the cause of aircraft accidents.109 On self-reporting and under conditions of confidentiality, 49 percent of dichromats reported color confusion at traffic lights, 33 percent found it difficult to distinguish traffic lights from street lighting, and 22 percent had difficulty detecting rear brake lights.110 It is therefore noteworthy that rear-end collisions, particularly under conditions of poor visibility, were slightly more common among protan drivers, who have more problems with perception of red rear warning lights.111 However, there is general agreement in most jurisdictions that nonprofessional automobile drivers with color vision defects should have no driving restrictions. Nevertheless, some observers are impressed with the data suggesting that color-deficient drivers have significant difficulties.112 These authors suggest shape and not just color coding of traffic lights as well as counseling color-defective drivers to exert caution at intersections and advising protans to double the usually recommended distance that separates their vehicle from the car ahead.
Molecular Genetics of Color Vision Defects
Red-Green Color Vision Defects.
Nathans and collaborators isolated and sequenced the genes that specify the three opsins responsible for normal color vision10 and showed how these genes are different in individuals with red-green color vision defects.11 Based on their study of 25 red-green color-deficient males, they concluded that, in the majority of cases, color vision defects result from unequal recombination between the highly homologous red and green pigment genes (98 percent identity in DNA sequence of exons, introns, and 3′ flanking regions). Such events lead to deletions of the green pigment genes (see Fig. 238-6) or to the formation of full-length hybrid genes consisting of portions of both red and green pigment genes (Fig. 238-8). With few exceptions, the deletion of green pigment genes was associated with deuteranopia (G−R+), 5′ green-red hybrid genes with deuteranomaly (G′R+), and 5′ red-green hybrids with either protanopia (R−G+) or protanomaly (R′G+). An interesting observation was that some males had normal green and red pigment genes in addition to the hybrids and yet tested as deutans.
Generation of red-green hybrid genes. Unequal recombination between the highly homologous red and green opsin genes generates 5′ red-green hybrids typically found in individuals with the protan class of color vision defects (protanopia and protanomaly) and 5′ green-red hybrid genes found among individuals with deuteranomaly. Filled arrows denote red gene sequences, and open arrows represent green opsin gene sequences. (Used with permission from Drummond-Borg et al.71 )
Determination of the gross structure of the red and green pigment gene arrays in males was made by quantitative Southern blot analysis, taking advantage of differences in length of DNA fragments generated from the red and green opsin genes on digestion with the restriction enzymes EcoRI, BamHI, and RsaI. Figure 238-9 shows autoradiographs of Southern blots of genomic DNA isolated from males with normal and defective color vision together with the deduced genotypes.11
Southern blot analysis of the X-linked red-green gene locus. (A) Partial restriction maps of the red and green pigment genes. E, EcoRl; B, BamHl; R, Rsal. Open boxes denote the six exons. The restriction fragments A–D used in distinguishing red from green-specific sequences are shown below the genes. The A–C fragments derived from the red opsin gene (Ar, Br, Cr) are longer than those derived from the green gene (Ag, Bg, Cg) due to a 1.5-kb insertion in the first intron (wavy line). The absence of a Rsal site in the green opsin exon 5 accounts for the larger Rsal fragment of the green opsin gene (Dg). (Adapted with permission from Nathans et al.11) (B) Autoradiograph of Southern blots of genomic DNA samples from males with normal and defective color vision digested with either a combination of EcoRl and BamHl or Rsal. The EcoRl-BamHl fragments (B and C) were detected by hybridization to a 350-bp cDNA probe encompassing exon 1 and part of exon 2, whereas the D fragments were detected by a 400-bp genomic probe from the 3′ end of intron 4 of the green opsin gene.11 Typical examples of Southern blot patterns were selected. Lane 1: An individual with protanopia who has a single red-green hybrid gene with the C and D fragments of the red gene replaced by the corresponding fragments of the green pigment gene. Lane 2: A deuteranomalous individual who has normal red pigment gene and a green-red hybrid pigment gene (see diagram in part C). Lane 3: An individual with deuteranopia who has only a normal red gene (see diagram in part C). Lane 4: A deuteranomalous individual with a gene array comprised of normal red and green as well as a green red hybrid pigment genes. Lane 5: An individual with normal color vision who has multiple green pigment genes. The unmarked lanes show patterns similar to those described above. (C) Diagrammatic representation of Southern blot patterns for males with normal and defective color vision. The color vision phenotypes and the structure of the red-green gene arrays associated with the EcoRl-BamHl, and Rsal Southern blot patterns are shown above the Southern blot. Filled and open arrows denote red and green gene sequences, respectively. The width of the solid lines representing restriction fragments reflects the relative quantity of DNA that is quantified by densitometry. In protanomaly and protanopia (protan), the Dr fragment is missing, indicating loss of the 3′ portion of the red pigment gene, while the 5′ portion of the green gene is present. In deuteranopia, Bg, Cg, and Dg fragments are missing, indicating complete deletion of the green pigment gene. In deuteranomaly, the relative proportions of fragments is shifted, indicating the presence of green-red hybrid genes.
The following important and interesting questions were raised: Is the point of fusion in red-green hybrid genes correlated with the severity of the color vision defect? Which amino acid residues contribute significantly to the difference in absorption characteristics between the red and green opsins? Are all genes in the red-green cluster equally expressed?
These questions were addressed by studying the relationship between genotype and phenotype among males who had normal color vision and others who had defective red-green color vision.19 In addition to the use of quantitative Southern blot analysis to detect deletions and hybrid red-green opsin genes, amplifications by the polymerase chain reaction (PCR) and single-strand conformation polymorphism (SSCP) were used to determine the approximate points of recombination in hybrid genes. Recombination between the red and green opsin genes would be expected to occur more frequently in introns than in exons, since introns are, on average, 10 times longer and are as homologous as exons in this gene complex.8,10,11,113 Evidence supporting this expectation was provided by analysis of hybrid genes of 64 individuals with defective color vision (see below and ref.19). Recombinations in introns 1 and 4 could be assigned with certainty, whereas those in introns 2 and 3 could not be differentiated because the sequence of exon 3 of the red is not always different from that of the green opsin gene.10,19,114 Recombinations in intron 1 would convert one pigment gene to the other (i.e., green to red), whereas those occuring in intron 5 would have no effect because exons 1 and 6 in the red and green opsin genes have identical sequences. Recombinations in exons 2, 3, and 4 would be expected to result in hybrid genes (5′ green-red or 5′ red-green) that encode six corresponding chimeric pigments with maximal sensitivities distributed between those of the normal red and green pigments (530–560 nm). These six hybrid genes (3′ red-green and 3′ green-red) are shown in Figure 238-10. However, since there are two common alleles of the red opsin gene that differ by having either Ala or Ser at position 18018 and by approximately 4 to 5 nm in absorption maxima,62 nine instead of six common types of hybrid opsin genes would be expected to exist in the population. Merbs and Nathans56 examined photobleaching difference spectra of in vitro–produced red-green hybrids commonly found in the population and showed that amino acid residues in exons 2, 3, 4, and 5 that differ between the red and green pigments produce varying degrees of spectral shifts (see Fig. 238-10).
Spectral characteristics of the hybrid visual pigments commonly found among individuals with defective color vision. The right-hand column shows the exon composition of the six possible hybrid genes resulting from unequal recombination between the red and green opsin genes occurring in introns 2, 3, and 4. Filled and open boxes denote red and green gene exons, respectively. Hybrid opsins would not be formed as a result of recombinations in introns 1 and 5, since the sequences of exons 1 and 6 are identical in the two pigment genes. The normal red and green opsin genes are included for comparison. The left-hand column gives the designations for each hybrid pigment. For example, R3G4 denotes a 5′ red-green hybrid with the first three exons derived from the red gene and the last three exons from the green gene. Note that whenever exon 3 of a hybrid is derived from the red opsin gene, two forms of the hybrid are possible depending on whether Ser or Ala is found at position 180 in exon 3. The λmax values were determined from photobleaching difference absorption spectra of recombinant pigments expressed in tissue culture cells transfected with cDNA clones encoding the various hybrid opsins.56
In a study of 64 red-green color-defective Caucasian males, the great majority of defects were associated with either deletion of the green opsin gene or the formation of 5′ red-green or 5′ green-red full-length hybrid genes.19 The results, described below, were basically in agreement with those obtained in the earlier study of 25 subjects by Nathans et al.11
Twenty-three (36 percent) of 64 color-deficient males19 were protans, and their anomaloscopic Rayleigh match ranges are given in Fig. 238-7. Figure 238-11 shows the observed gene arrays, points of fusion, and class of protan defect [protanopic (P) or protanomalous (PA)] for the same subjects. The gene arrays of all protans were characterized by the presence of 5′ red-green hybrid opsin genes instead of the normal red opsin gene. In all cases, the intron of fusion was upstream of exon 5, thus indicating that exon 5 is critical in establishing the spectral characteristics of a normal red pigment. The replacement of exon 5 of the red opsin gene with that of the green opsin produced a hybrid pigment that was sufficiently greenlike in its spectral properties that the subjects performed as protans. The significant role of exon 5 of the red opsin gene was somewhat predicted, since it contains two of the three amino acid residues (at positions 277 and 285) thought to be mainly responsible for the difference in spectral properties between the red and green photopigments (see Fig. 238-12).
Red-green color vision–pigment gene arrays found in males with protan color vision defects. Hybrid pigment genes consist of 5′ red opsin gene sequences (filled arrows) followed by green opsin gene sequences (open arrows). The subscript numbers 1–3 refers to the number of normal green opsin genes present, including subjects with 1, 2, or 3 green pigment genes. The color vision phenotype, as determined by anomaloscopy, is indicated as P, protanopia; PA, protanomaly. The assignment of intron of the fusion in hybrid genes was made on the basis of results of Southern blot analysis, PCR amplification, and sequencing of exons.19 The uncertainty in assigning the fusion point to introns 2 or 3 results from the presence of polymorphisms in exon 3, the alleles of which are shared by the red and green opsin genes; 2–3 therefore indicates fusion points in either intron 2 or 3. (Used with permission from Deeb et al.19 )
The importance of exon 5 in determining spectral sensitivities of hybrid pigments. The positions at which the red and green opsins differ by the presence of hydroxyl-bearing (boxed) versus nonpolar amino acid residues are shown. Recombination in intron 4 results in the exchange of three of these residues, two of which have been shown to account for the major difference in λmax between the normal red and green pigments. The red-green hybrid is as observed in protan subject 2203 of Fig. 238-11, and the green-red hybrid is as observed in deutan subject 1682 of Fig. 238-13. (Used with permission from Deeb et al.19 )
The Ser/Ala polymorphism in the red opsin gene plays a role in determining protan subtypes. Diagrams of X-linked gene arrays each composed of a red-green hybrid gene (fusion in intron 4) and a normal green opsin gene. The hybrid gene with Ser at position 180 differs by 5 nm in λmax (Δλmax) from the normal green opsin. A difference in λmax of this magnitude makes for protanomaly (PA), since trichromacy is preserved by this anomalous red-green pigment. In contrast, the Ala-containing hybrid has the same λmax as the normal green pigment, making the carrier a protanopic (P) dichromat.
The relationship between structure of the pigment and its spectral properties as determined by ERG was investigated in a protanope who had a 5′ red-green hybrid gene in which the point of fusion was in intron 3.115 The absorption spectrum of the pigment encoded by this hybrid gene (which encoded Ala at position 180) was very similar to that of the green pigment, suggesting that sequence differences in exons 2 and 3 contribute little to the difference in absorption between the red and green photopigments. These results are consistent with those obtained by measuring bleaching difference spectra of in vitro–produced types of red-green hybrid genes.56 They showed that the exchange of exon 5 sequences resulted in a major spectral shift in λmax (15–20 nm), whereas differences in exons 2 to 4 appeared to have smaller effects (see Fig. 238-10)
Subjects who had only the red-green hybrid gene in their arrays test as protanopes regardless of the point of fusion, whereas those who had one or more normal green opsin genes in addition to the hybrid gene test as either protanopic or protanomalous. The distribution of match ranges was quite similar for subjects with either intron 2–3 or 4 fusions. The Ser/Ala polymorphism at position 180 in the red opsin appears to underlie the preceding discrepancy between genotype and phenotype. In a study of 19 protan subjects116 who had one 5′ red-geen opsin gene as well as one or more normal green opsin gene, protanopes and protanomalous subjects could be differentiated (with one exception) on the basis of whether Ser or Ala was present at position 180 of the red portion of the hybrid gene, respectively. The presence of Ser encoded a hybrid pigment that differed by 5 nm in λmax from that of the normal green opsin, and this was associated with protanomaly. When Ala was present at position 180 of the hybrid, there was no difference in λmax, and the subject were protanopic (see above and Fig. 238-13).
The Rayleigh match ranges (see Fig. 238-7) of 41 of 64 color-deficient males placed them in the deutan series.19 The gene arrays and estimated points of fusion for 40 of these deutans are shown in Fig. 238-14. One of the deuteranomalous subjects had a grossly normal gene array but had a point mutation in the green opsin gene (see below). All deutans had a normal red opsin gene, and all subjects (except the individual with the point mutation) had major gene rearrangements that could be detected by Southern blot analysis and/or gene-specific PCR amplification across intron 4.
Color vision gene arrays found in subjects with deutan color vision defects. Explanations concerning the gene arrays are given in the legend to Fig. 238-13. D, deuteranopia; DA, deuteranomaly. Note that 3 of 13 individuals who had only a single red opsin gene unexpectedly tested as anomalous trichromats (deuteranomalous) instead of dichromats (deuteranopic). Furthermore, individual 1681 who had normal red and green opsin genes in addition to a green-red hybrid gene tested as a deuteranope, suggesting that the two normal green opsin genes are not expressed. N/A, not applicable. (Used with permission from Deeb et al.19 )
Thirteen of the 41 deutans were shown to completely lack green opsin gene sequences. Ten of the 13 subjects were classified by anomaloscopy as deuteranopes as expected, since they have only one red opsin gene. Surprisingly, the remaining 3 tested as severe deuteranomalous trichromats. Since they have only a red opsin gene, they should be completely unable to discriminate the red and green lights from the yellow standard light in the anomaloscope, yet they can do so. The pattern of a single red opsin gene in a deuteranomalous subject also has been observed in a previous study.11
Twenty-five of the 41 deutans had gene arrays characterized by the presence of one or more full-length 5′ green-red hybrid opsin genes in addition to a normal red opsin gene (see Fig. 238-14). In 18 of these subjects, one or more normal green opsin genes were found in addition to the normal red and hybrid genes. Except for two subjects who tested as deuteranopic (dichromats), these gene arrays were associated with deuteranomaly. In one of the deuteranopes (subject 1907), and most likely in the other (subject 1681), the points of fusion in their hybrid genes were in intron 1, causing the expected deuteranopia. The fusion points in all the other deuteranomalous subjects were located in introns 2 to 4.
As seen in the protan subjects, 5′ green-red hybrid genes that resulted from a crossover in intron 4 (i.e., exchanged exons 5 and 6) encoded a pigment that was essentially redlike in absorption spectrum.
The presence of more than one hybrid gene in the array does not seem to be associated with a more severe color vision defect. Neither does the presence in arrays of normal green opsin genes in addition to normal red and hybrid genes. One explanation for these observations is that not all the opsin genes in an array are expressed in the retina (see below).
Deuteranomaly in two subjects (1838 and 1927) was associated with a novel type of hybrid gene (5′ green-red-green) in which a central segment encompassing exon 4 and possibly exon 3 was exchanged between the red and green opsin genes, presumably due to a double crossover or a gene conversion event. In these cases, the two hydroxyl-bearing residues at positions 230 and 233 in exon 4 of the green opsin gene were replaced by Ile and Ala, respectively, suggesting a role for one or both of these amino acids in determining spectral sensitivities of the photopigments. The results of Merbs and Nathans56 support this hypothesis. They showed that amino acid differences in either exon 3 or 4 could produce a spectral shift of 4 nm (see Fig. 238-10).
A Point Mutation in a Single Case of Deuteranomaly.
The gene array of one of the subjects (CB 1909) with severe deuteranomaly (match range of 12–68) had no gross rearrangements. Examination of the coding sequences of his red and green opsin genes revealed that all three of his green opsin genes had a C→T transition at nucleotide 648 that translates to the substitution of Arg for Cys at position 203.117 Screening of 63 other color-defective subjects known to have major gene rearrangements revealed another deuteranomalous individual (CB 1843) who carried the same mutation in one of his three green opsin genes. The same C203R mutation had been observed in 16 unrelated families with blue cone monochromacy.118,119 In these cases, the mutation was in the green segments of 5′ red-green hybrid genes. These results suggest that the C203R mutant allele of the green opsin gene may be common in the general population. Indeed, the same mutation also was found in a green gene of 1 of 65 male subjects with normal color vision. The Cys residue is highly conserved among all visual pigments studied so far, as well as among other seven-transmembrane-segment receptors, such as the adrenergic, muscarinic, dopaminergic, and serotonergic receptors.43,120
The Cys 203 residue is believed to form a disulfide bridge with Cys 126, thus covalently linking the first and second extracellular loops of the opsins. Results of in vitro mutagenesis studies showed that the corresponding cysteine residues in bovine rhodopsin (residues 110 and 187) and in the hamster beta-adrenergic receptor (residues 106 and 187) are essential for function of these proteins.121,122 Therefore, the C203R mutation is very likely to abolish function of the green-sensitive photoreceptor. Furthermore, in analogy with the mutant rhodopsin alleles associated with autosomal dominant retinitis pigmentosa,123,124 this mutation may predispose to certain X-linked cone dystrophies as a result of accumulation of the abnormal protein. A sequence rearrangement in the red opsin gene has been found to cosegregate in one family with X-linked progressive cone degeneration.125
Based on analysis of at least 90 males with red-green color vision defects,11,19,126 inter- and intragenic recombinations between the red and green opsin genes, which result in green opsin gene deletions or shuffling of exons between the two genes, account for all but the one case of color vision defect, found to be due to a C203R mutation. Exon 5 plays a major role in spectral tuning, since it contains three of the seven residues that distinguish red from green opsins. Molecular analysis of the red and green opsin genes could classify subjects into either the protan or deutan series. Although certain trends were evident, the genotype at the level of coding sequence occasionally was not correlated with the color vision phenotype within the protan and deutan series. Nagel anomaloscopy may not provide a sufficiently accurate quantitative assessment of ability of red-green color-deficient subjects to discriminate color.13 Alternatively, severity of color deficiency may not be a function of the sequence of hybrid pigments only. Postreceptoral neural factors127,128 and variation in the amount of pigment synthesized or in the ratio of red to green cones129 were proposed to account for variation in color discrimination. Position-dependent expression of genes of the red-green locus explains why the more distal location of hybrid genes and mutant green opsin genes in addition to normal red and green opsin genes does not result in altered color vision.
Selective Expression among Multiple Green Opsin Genes: Only One Gene is Expressed.
The frequency of 5′ green-red hybrid genes was observed to be higher than the reported frequency of color vision defects among Caucasian and especially among African-American males67,71 (see below), suggesting that such hybrid genes may not always lead to color vision defects. This was proven to be the case by showing that 4 of 129 Caucasian males with anomaloscopically determined normal color vision had a 5′ green-red hybrid gene in addition to normal red and green opsin genes.19 In addition, a male with normal color vision was found to have the C203R mutation in one of his five green opsin genes,117 and another (subject 1681 in Fig. 238-14), who had a green-red hybrid opsin gene in addition to normal red and green opsin genes, tested as deuteranope, indicating that his normal green opsin genes were nonfunctional.19
The hypothesis that not all opsin genes in an array are expressed in the retina was advanced to explain these observations. Green opsin gene sequences in genomic DNA were compared with the corresponding mRNA sequences expressed in postmortem retinal tissues. Advantage was taken of a relatively common but silent polymorphism (A versus C at the third position of codon 283) in exon 5 of the green opsin gene.130 The two alleles can be differentiated by PCR amplification of exon 5 followed by SSCP analysis or digestion with EcoO109_1. Results of such a comparison in 10 male subjects who had two or more green opsin genes in their genomic DNA clearly showed that when the two alleles of exon 5 were present, only one was represented in expressed retinal mRNA,130 indicating the expression of only a single green opsin gene. In addition to the expression of a single green opsin gene, retinal RNA from the same individuals contained a single red opsin-encoding mRNA sequence.
In another set of 51 unselected postmortem retinas, three male donors of unknown color vision status had gene arrays comprised of one red, one 5′ green–red 3′ hybrid and one normal green pigment genes. We found that the expressed mRNA transcripts were encoded by the normal red and the 5′ green–red 3′ hybrid but not by the normal green pigment gene and therefore presumably reflected deuteranomalous color vision. The other two retina specimens expressed the normal red and green pigment genes but not the green-red hybrid and presumably had normal color vision.72 In another study, the lack of expression of green pigment genes carried by deutan subjects was confirmed.131 Regardless of the presence or absence of hybrid genes, it generally was observed that the red pigment gene, which occupies the most proximal position in the array, was expressed at a higher level (fourfold) than the green pigment gene.72 This was confirmed in a subsequent study.131 These results are consistent with those of molecular and ERG studies by our group on deutans who carry normal red and green and a 5′ green–red 3′ hybrid gene.132 In all seven subjects studied, the green pigment was undetectable by ERG spectral sensitivity measurements.
Neitz and Neitz133 suggested that a large proportion of males with normal color vision have a large number of red and green pigment genes with at least two expressed red pigment genes as well as at least two expressed green pigment genes. Extensive data using Southern blot analysis,19,114,130 PCR-based methodologies,72 as well as pulsed-field gel electrophoresis (which is a more direct method of estimating the total number of genes in an array from its length)134 are most consistent with a much lower average number (mean of 2) of genes in the array and with the presence of only a single red pigment gene. Furthermore, in arrays that contain more than one green pigment gene, only one is expressed in the retina.130 (See Addendum 3).
The model illustrated in Fig. 238-15 was proposed130 to explain such selective expression of one of a set of green opsin genes in an array. In this model, a locus control region (LCR) regulates expression of the opsin genes of the array in a position-dependent manner. Thus, in red cones, the LCR forms a stable, transcriptionally active complex with the red opsin promoter, whereas in green cones, the LCR forms a complex with the proximal green opsin promoter. Active complexes between the LCR and green opsin promoters located downstream of the proximal green opsin promoter are much less favored. We therefore suggest that only the most proximal green pigment gene is expressed in an array of several green pigment genes that may include 5′ green-red hybrid pigment genes. (See Addendum 4).
Model for selective expression in the X-linked red-green gene complex. Numbers denote length in kilobase pairs. (A) Red cone-specific gene transcription occurs as a result of stable coupling (mediated by DNA-binding proteins) of the LCR to the red gene promoter. (B) The LCR preferentially and stably couples to the proximal green opsin gene promoter and turns on its expression. Distal green opsin promoters are not activated presumably due to the low probability of coupling to the LCR. (Adapted with permission from Winderickx et al.130 )
The concept of an LCR was first suggested for the regulatory sequences 5′ upstream of the beta-globin gene locus. The globin LCR was shown to be essential for developmental switching of expression from fetal to adult globin genes and indicated that the order of the genes comprising this locus is important for such a transcriptional switch.135–138 Evidence for an LCR at the red-green opsin locus has been provided by Nathans et al.,118 who found that, in some instances, blue cone monochromacy, a disorder in which both red and green cone functions are absent, is associated with deletion of a regulatory sequence located 3.8 to 4.3 kb upstream of the transcription-initiation site of the red opsin gene and 43 kb upstream of the proximal green opsin gene. They subsequently showed that a region between −3.1 and −3.7 kb of the red opsin gene is required for cone-specific expression of the beta-galactosidase reporter gene in transgenic mice.139
Blue cone monochromacy (BCM) (MIM 303700), also referred to as Pi1 monochromacy or X-linked incomplete achromatopsia,100 is an extremely rare disorder (<1 in 100,000) in which both red and green cone sensitivities are absent. The physiologic functions of both rods and blue cones are preserved. Significant linkage of BCM to the red-green pigment gene locus was established by analysis of RFLP alleles at the two DNA markers, DXS15 and DXS52.140 This led Nathans et al.118 to analyze the red-green locus in individuals with BCM. Their studies on 38 families with BCM have uncovered two mechanisms for generating this phenotype (reviewed in ref.119). The first, found in 14 families, involved deletions (587 bp to 55 kb) that included a regulatory sequence located approximately 3.4 kb 5′ upstream of the transcription-initiation site of the red opsin gene. In some of these individuals, the red and green opsin genes were unaffected, whereas in others, the deletions extended into the red opsin gene. This regulatory region, referred to as an LCR (see above), was shown, in transgenic mice, to be essential for directing expression of the β-galactosidase reporter gene to both long- and short-wavelength-sensitive cones in the mouse retina.139 The second mechanism of generating BCM, found in 20 families, involved unequal homologous recombination between the green and red opsin genes that reduced the gene array to only a single red or a 5′ red-green hybrid gene. In 16 of these families, the green opsin portion of the hybrid gene had a C203R substitution, which apparently rendered the encoded hybrid opsin nonfunctional. Evidence that the C203R substitution encodes a nonfunctional photopigment was provided by Kazmi et al.,141 who expressed a recombinant human M-pigment carrying this mutation in cultured cells. The expressed opsin was misfolded, retained in the endoplasmic reticulum, and defective in interaction with the chromophore and activation of transducin. Furthermore, disruption of the same disulfide bond due a mutation (C187Y) in rhodopsin caused early and severe autosomal dominant retinitis pigmentosa in one family.142 The same C203R mutation was found to be relatively common (2 percent) in the green opsin genes of Caucasian males117 but may not always be expressed because of its distal position in the array (see below). Progressive central retinal dystrophy has been reported in some patients with BCM,118 indicating that cone degeneration may result from the accumulation of an abnormally assembled photopigment in analogy with the mutations in rhodopsin found to underlie autosomal dominant retinitis pigmentosa.123 In one kindred, BCM was shown to be due to the presence of the inactivating C203R mutation in exon 4 of both the red-green-red hybrid gene and the green pigment gene carried on the X chromosome of affected males.143 It is interesting to note that funduscopy in two older affected males in this family revealed atrophy of the retinal pigment epithelium and the chorionic choriocapillaris layer in the macula.
Other than the common C203R mutation, Nathans and colleagues observed two other photopigment inactivating substitutions associated with BCM: R247ter and P307L. Furthermore, a Danish male patient with BCM was found to have a single red pigment gene in which exon 4 is deleted and no green pigment genes.144 The deletion alters the translational reading frame and creates a stop codon. The encoded defective protein is predicted to lack the three C-terminal transmembrane helices of the red pigment. Defective red pigment gene function in the absence of green pigment genes caused BCM.
Tritanopia (MIM 190900) is a rare autosomal dominant91 disorder characterized by selective loss of blue-sensitive photoreceptor function and greatly diminished or absent chromatic discrimination in the blue region of the spectrum. A survey in the Netherlands indicates that its frequency in the population may be as high as 1 in 500.145 Three missense mutations in the gene encoding the blue pigment opsin, located at 7q31.3–q32,10 have been shown to cause tritanopia: G79R in two Japanese subjects, S214P in two Caucasian subjects, and P264S in three Caucasian subjects.92,93 The three mutant alleles cosegregate with tritanopia in an autosomal dominant fashion, but incomplete penetrance was observed in association with the G79R substitution. The mutations affect residues located in the second, fifth, and sixth transmembrane α-helical segments of the blue pigment opsin. The dominant mode of inheritance suggests that accumulation of a defective opsin within photoreceptors causes either loss of function or cell death, reminiscent of mutations in the rhodopsin and peripherin genes that cause a subset of autosomal dominant retinitis pigmentosa. Tritanopia also has been observed in association with some disorders of vision such as autosomal dominant juvenile optic atrophy.146,147
Complete achromatopsia or rod monochromacy (MIM 216900) is a rare autosomal recessive trait characterized by total loss of color vision, photophobia, nystagmus, and loss of visual acuity. A locus for complete achromatopsia has been mapped to 2q11.148 A candidate gene (CNGA3), encoding the α-subunit of the cone photoreceptor cGMP-gated cation channel, had been mapped to the same region.149 This channel plays a critical role in the light-triggered signal-transduction cascade in all three classes of cones leading to hyperpolarization of the photoreceptor membrane. Kohl et al.150 reported the presence of missense mutations in CNGA3 in five families with complete achromatopsia. (See Addendum 5).
A Single Amino Acid Polymorphism Explains Variation in Normal Color Vision
Subtle variations in color perception in the red-green region of the spectrum have been observed among individuals considered to have normal color vision. Rayleigh color matches of male subjects with normal color vision fell into two main groups17,18,151–153 and suggested a difference of several nanometers in the red pigment absorption spectra. Females with normal color vision show a third and larger group with intermediate values of match midpoints.17 A similar, independently described Rayleigh match variability in families suggested transmission by X-linked inheritance.16 These observations pointed to the presence of two common alleles of the red pigment gene. Assuming the occurrence of X-chromosome inactivation, females who are heterozygous for such a polymorphism would be expected to have patches of cones containing either one or the other of the pigment forms and therefore would show a color match distribution intermediate between those of the two homozygotes.
The subtle differences in color matching observed among individuals with normal color vision (as well as among deuteranopic dichromats, who have only a red opsin gene) have been suggested to be a reflection of small variations in the absorption maxima of the red or green photopigments.12–15,154
A common single-amino-acid polymorphism (62 percent Ser, 38 percent Ala) at residue 180 of the red photopigment was discovered in the Caucasian population.18 Fifty Caucasian males with normal color vision were tested for the hypothesis that the two major groups in the distribution of color matching could be explained by the preceding Ser/Ala polymorphism. The frequency distributions of Rayleigh match midpoints and of the deduced amino acid sequence of the red photopigment (Fig. 238-16) show that higher sensitivity to red light (i.e., requirement of less red in the mixture of red and green to match the standard yellow light) was highly correlated with the presence of Ser at position 180.18 These males therefore have a different perception of red light than those having the alanine allele at this site. Females having both the Ala and Ser alleles would be expected to have two types of red photoreceptors due to X-chromosome inactivation and thus may have tetrachromatic vision. This is analogous to the situation in New Word monkeys, who have only a single X-chromosome-encoded middle–long-wavelength pigment gene with several alleles, where females heterozygous for two alleles of this gene achieve trichromacy and males and homozygous females test as dichromats.49 The frequency of the Ser/Ala polymorphism among African-Americans is 80/20 percent (S. Deeb, unpublished observation, 1993), and among Japanese it is 84/16 percent.155
Correlation of the Rayleigh match midpoint (center of match range) with the presence of serine or alanine at position 180 of the red photopigment. The 50 subjects were Caucasian males who tested as having normal color vision. Determination of color matches was made by measuring the proportion of red in a mixture of red and green lights that was perceived to match a standard yellow light. The presence of serine at position 180 correlates with higher sensitivity to red light (i.e., less red light is used to mix with green light to match the standard yellow light). Two individuals who had Thr and Ser instead of Ile and Ala at positions 230 and 233, respectively, required more red light in the mixture in comparison with others with the same amino acid at position 180. (Used with permission from Winderickx et al.18 )
The importance of the presence of Ser or Ala at position 180 in spectral tuning of photopigments had been deduced already from studies of the visual pigments of Old and New World monkeys.59 A difference of 6 nm was observed between the peaks of maximal sensitivity (562 nm for Ser-containing pigments and 556 nm for Ala-containing pigments, as determined by ERG of the pigments of two tamarin monkeys that differed in sequence by only Ser versus Ala at position 180. The results of Merbs and Nathans62 of direct determination of the bleaching difference absorption spectra of human red pigments (expressed in tissue culture cells transfected with complementary DNA clones) that differed in sequence by Ser versus Ala at position 180 were thus in good agreement with the ERG results in the monkeys (557 and 552 nm for the Ser- and Ala-containing reconstituted pigments, respectively), as well as with the human color matching data. These results support the finding that the Ser/Ala polymorphism at position 180 of the red pigment underlies the observed variation in red-green color vision among people with normal color vision.
Other amino acid polymorphisms have been observed in both the red (eight sites) and green (five sites) opsin genes. These polymorphisms give rise in the general population to 15 and 18 different green and red opsins, respectively. Nine of these polymorphisms are located in exon 3. Alleles of these polymorphisms are shared between the red and green opsin genes, suggesting a history of relatively frequent gene conversion or unequal recombination, mainly localized to exon 3, during the evolution of the two lineages.114
Heterozygosity and Homozygosity among Females with X-Linked Color Vision Defects
Because of the high frequency of X-linked color vision defects among males, there will be a large number of female heterozygotes for these genes in the population. Among males, the gene frequency and trait frequency of an X-linked trait (p) are identical. Males have either have (p) or do not have (q) the X-linked trait under study. The expected frequency of female heterozygotes and homozygotes can be calculated by the Hardy-Weinberg law, where 2pq will be the number of heterozygotes and q2 the number of affected homozygotes. Homozygotes will be defective in color vision similar to affected males. The number of female heterozygotes (2pq) for each of the defective color vision categories is about twice the number of defective color vision male hemizygotes. Since the total frequency of color vision defects among European populations is 8 percent (p), about 15 to 16 percent (2pq) of the female population will be heterozygotes for one or another red-green color vision defect. The majority of such heterozygotes (7–10 percent) are heterozygotes for deuteranomaly, since this abnormality is the most frequent defect.
Occasional heterozygotes who express phenotypic color vision defects may have a single X chromosome such as in Turner syndrome or may come from the small proportion of females with extremely skewed X-inactivation who by chance have inactivated most of their normal X chromosome and thus express the mutant X chromosome.156,157 Skewed X-inactivation appears to be more common in one member of identical female twin pairs,158 and six pairs of heterozygote female identical twins discordant for color vision have been reported.159–163 The expected skewed inactivation of one X chromosome has been demonstrated directly in one of the twin pairs.158 Furthermore, two of five Dionne identical quintuplet girls were color blind,161 as was one of three identical triplet Japanese girls.164
The population frequency of skewed X chromosome inactivation was investigated among 400 unselected pairs of Caucasian identical female twins who included about 60 heterozygotes for red-green color vision defects (≈15 percent of 400 unselected females are expected to be carriers of red-green color vision defects). Discordant X chromosome inactivation with one twin being defective in color vision and the other having normal color vision was observed in only two twin pairs, suggesting a low population frequency (≈3 percent) of the skewed inactivation phenomenon in identical female twins (unpublished observations by the Seattle group, 1998).
As expected by the X-inactivation hypothesis, heterozygotes are mosaics for normal and abnormal color vision in their retinas. Thus, by shining a very narrow beam of red or green light into the retinas of female heterozygotes for X-linked color vision defects, patches of defective color perception were found.165,166 In other experiments, heterozygotes made more errors than controls when asked to identify the color of briefly presented stimuli under conditions that did not allow any eye movement.167
These findings are consistent with earlier data that mild abnormalities of color vision often can be detected on psychophysical testing where groups of heterozygotes are compared with normal controls.82,100,168 Such minor deviations have been observed with pseudoisochromatic plate reading, anomaloscopy, and tests that assessed hue and saturation discrimination.100 The so-called Schmidt sign is the most obvious abnormality and can be elicited by many different psychophysical techniques. It is found in protan heterozygotes. The defect consists of a reduction in the relative luminous efficiency curve compared with that in normal and affected hemizygotes.169 Similar findings have been observed in the relative spectral luminous efficiency function of deutan heterozygotes and consisted of a lessened sensitivity at short wavelengths.169
ERG has been successful in identifying a fairly large number of both protan and deutan heterozygotes by the ratio of sensitivity at short (480-nm) and long (620-nm) wavelengths for the rapid off response.99
The molecular abnormalities in heterozygotes reflect those observed in males with color vision defects. Molecular methodology that depends on Southern blots of various restriction fragments of the red and green pigment gene usually does not allow molecular detection of heterozygotes. However, when family studies are done and affected fathers and brothers are tested by the appropriate techniques, molecular genotypes of females can be determined.126 Detection of female carriers of protan defects can be made by PCR amplification without knowledge of the status of family members.170 Amplification is carried out using a forward primer that is specific for the red pigment gene promoter, with a reverse primer in intron 5 that is common to both red and green pigment genes. The amplification of a segment that contains a green pigment exon 5 would indicate that the female is a carrier of a gene array containing a 5′ red–green 3′ hybrid gene in the most proximal position. Such a gene array is characteristic for protanopia and protanomaly.
Compound heterozygotes for protan and deutan defects will be observed. The most frequent compound type are females with a deuteranomaly allele on one X chromosome and an allele for one of the other color vision defects (deuteranopia, protanopia, or protanomaly) on the other. Compound heterozygotes for both deuteranomaly and deuteranopia will manifest with deuteranomaly,82,100 as expected from the molecular findings, since a deleted green pigment gene is not expressed, whereas a visual pigment with somewhat abnormal sensitivity such as a green-red fusion gene in deuteranomaly will manifest. Similarly, protanope/protanomalous heterozygotes manifest with protanomaly, demonstrating that, among protans, the mild phenotype also is dominant over the more severe one.
In contrast, compound heterozygotes for both a deutan and protan defect will not present with color vision defects.83,100 This finding is not surprising, since normal alleles for red and green pigments are present in addition to the defects in both the protan and deutan genes. Such females, therefore, are functionally heterozygotes for both the deutan and the protan genes and therefore will have normal color vision. Compound heterozygosity for protanopia/deuteranomaly and protanomaly/deuteranomaly was studied in one large family by molecular techniques and showed no color vision abnormalities.126
With a population frequency of 1 percent for protanopia, protanomaly, and deuteranopia each and 5 percent for deuteranomaly among Caucasians, the expected frequency of homozygotes with defective color vision (p2) for protanopia, protanomaly, and deuteranopia is approximately 1 in 10,000 for each of these categories (total 3 in 10,000) and 1 in 400 for deuteranomaly homozygotes. The total frequency of color vision defects among female homozygotes is less than the squared number of the male frequency (p2 = 0.0064), since compound heterozygotes for protan and deutan defects are phenotypically normal.