Identifying human individuals with the genetic potential for tetrachromacy

The genes encoding the opsins or apoproteins of the human “red” and “green” photopigments are each composed of six exons and are arranged in a head-to-tail tandem array located on the q-arm of the X-chromosome.^1,2,3,4 Individuals with normal color vision usually have one red opsin gene in the proximal position of the gene array and one or more green opsin genes. These X-linked opsin genes have 98% identity in nucleotide sequence (including introns and 3’ flanking regions).⁵ The genes encoding long- and medium-wavelength apoproteins differ by an estimated fifteen residues, seven of these known to occur at positions which influence photoreceptor responsivity in the expressed phenotype.^6,1,7 Amino acid sequence differences across M- and L-cone genes are at codons 116, 180, 230, 233, 277, 285 and 309. Of these, seven single nucleotide substitutions (SNPs) at three particular sites (codons 180, 277 and 285, exon 3) produce substantial shifts in photopigment spectral sensitivity,⁸ and sensitivity shifts increase monotonically with substitutions.^7,9

Molecular genetic methods used by Jameson and colleagues were developed between 1997 and 2002 and are described in Wasserman, Szeszel and Jameson.¹⁰ Existing studies,^{7,11,12,13,14} which together distinguish the genomic regions of DNA sequence variation between MWS and LWS genes, provided the empirical justification for the Wasserman et al.¹⁰ method used in the human tetrachromacy research discussed here by Jameson and colleagues. The Wasserman, Szeszel & Jameson (2009) procedure uses a combination of three molecular approaches in order first to create MWS and LWS gene-specific DNA templates and then to use those templates to distinguish between their respective codon 180 sequences. A long-range polymerase chain reaction technique generates gene-specific PCR products. A PCR and a restriction digest determines MWS and LWS codon 180 genotypes (Figure 1), and DNA sequencing of each PCR template confirms this gene specificity (Figure 2). The method extends the analyses of Jameson et al. (2001) by permitting greater specificity of the identified polymorphisms and permits a more informative analysis of genotype-correlated behaviors reported by Jameson et al.¹⁵

Figure 1. Identifying the presence of protein substitutions on opsin genes illustrated by DNA radiogram (left panel) and DNA sequence trace (right panel). Fnu4HI restriction digestion assay of genomic DNA from exon 3 codon-180 of the M-cone opsin photopigment gene of seven female donors (Lanes 1-7). DNA radiogram (at left) shows the presence of the DNA sequence coding for Alanine is detected as a 160 bp band whereas the DNA sequence coding for Serine is detected as a 190 bp band. Lanes 1, 2 and 5, genomic DNA from human females with alanine at exon 3, codon-180 of the green gene. Lane 7, one female with serine at exon 3, codon. Lanes 3, 4 and 6, females with a serine-alanine dimorphism at exon 3, codon-180 of the green gene. Dimorphisms were confirmed by sequencing (graph at right) where the arrow indicates equal strength signals at a specific locus. Lane 8, DNA Ladder. Restriction gel digest products depicting analogous dimorphisms occurring at codon-180, exon 3 of the red opsin (L-cone) gene are not depicted here, but are similar to that shown for the green gene. Images courtesy of the author.

Figure 2.  Distinguishing M- from L-opsin genes. Opsin genes for M-cone photopigments were distinguished from those for L-cone photopigments using a long range polymerase chain reaction (PCR) technique. The method provided gene-specific sequences within exon 4 of the red (LWS) and green (MWS) opsin genes. DNA sequence coding was used to confirm specificity of each long-range PCR product. Amino acids and DNA bases unique to each gene are shown in bold and italics. Images courtesy of the author.

Perceptual differences associated with retinal tetrachromat genotypes.

Some human females have different M- and L-opsin genes on each X-chromosome and, as a result, the genetic potential to express more than the usual three retinal photopigment classes. These heterozygous females are putative retinal tetrachromats and may express (in addition to rods) four retinal cone classes, each with a different spectral sensitivity distribution, and the potential to experience tetrachromatic vision.¹⁶ Frequency estimates of females who are potential tetrachromats range between 15% and 50%,¹⁷ whereas less is known about the true frequency of expressing four retinal cone classes.

Although four-channel visual processing is known to occur when human trichromats simultaneously use rods and all cone classes under mesopic viewing conditions (when light conditions are low but not dark), and the expression of four retinal cone classes is accepted, still functional photopic human color tetrachromacy is debated. Color processing theory limits humans to no more than a trivariant color signal. Thus, four retinal cone classes may be a necessary (but not a sufficient) condition for full tetrachromatic color perception, since, for full tetrachromacy, four channels of cortical color signal processing also seem to be needed.¹⁸

Research has explicitly sought to demonstrate what perceptual differences (if any) are experienced by human retinal tetrachromats compared to trichromat controls with the usual 3-photopigment retinas.^{15, 16, 19, 20, 21} Still, there remains uncertainty among color vision researchers regarding whether individuals with diverse photopigment opsin genotypes should be viewed as individuals with color perception variations from normal.

What has been shown is that candidate retinal tetrachromats exhibit non-normative performance on some standardized psychophysical color vision assessment measures. As discussed by Cohn, Emmerich and Carlson²² heterozygous females fail to be detected by the use of an anomaloscope, although there are reported shifts in their anomaloscope color matches^{23, 24, 25, 26, 27} as well as shifts using flicker photometry.^{24, 28} Heterozygous females were also found to exhibit higher absolute thresholds to small spots of red light.^{26, 29} Unlike normal controls, these heterozygotes exhibit a failure of additivity of trichromatic color matches after exposure to a light bleaching of the rod system.¹⁹ However, some results, such as those described by Birch,³⁰ indicate that female compound mixed heterozygotes for protan and deutan color deficiency are usually reported to have normal, not deficient, color vision.

Compared to the earlier work in the area, Jameson and colleagues,^{15, 31,21} took a slightly different approach and aimed to demonstrate perceptual differences associated with retinal tetrachromat genotypes under more realistic viewing circumstances and stimulus formats than those typically employed. One assessment method they examined²¹ was the Farnsworth-Munsell 100-hue test (FM100).³²

The FM100 is a color vision assessment test widely used in industry and the military for screening color deficient or anomalous individuals from jobs that may critically depend on color judgments. The FM100 stimulus is a series of color samples that form a continuous hue circle, discretized into 85 color “caps” forming a smooth gradient of hue, ostensibly at a fixed level of brightness and a fixed level of saturation.

The outer color perimeter of Figure 3 approximates the 85-cap hue gradient. The test is administered to individuals as a set of randomized colors from a quadrant of the hue circle, one quadrant at a time. The observer’s task is to re-order the randomized colors until they form a perceptually smooth hue series, or to “correctly order” the color continuum so no visible transpositions in hue occur. If an individual performs with zero errors on this sorting task, then the color “caps” should be ordered without transposition errors and the transposition line traced in the polar coordinate plot of Figure 3 would resemble a smooth continuous line near the central region of graph. This is clearly not the case for the transposition line traced by the data shown in Figure 3. That is, in Figure 3 the large jagged excursions away from the inner concentric circle of the graph indicate that this heterozygote observer performed as poorly as a color deficient subject might on this sorting task. The individual’s Total Error Score equals 132, indicating a diagnosis of low color discrimination, which would in all likelihood exclude this individual from many delicate color processing scenarios in industry and the military. However, in every other respect this individual exhibited no sign of color perception deficiency, and reported no sense of diminished color experience or color confusion.

In general, this seemingly contradictory finding was seen in several of the putative tetrachromats assessed by Jameson et al.²¹ That is, several individuals with tetrachromat genotypes performed very poorly on the FM100 diagnostic, but generally experienced no color vision impairment or weakness and exhibited increased sensitivity for detecting chromatic bands in a diffracted spectrum task.¹⁵ Interpreting these results, Jameson and colleagues^15,21 suggest that the color perception of some female carriers of protan deficiencies can differ from that of female trichromat controls but not in a deficient way. Rather, in some color discrimination tasks protan carriers may be unimpaired, (detecting chromatic contrast at levels resembling those of trichromat controls ³³), while under other viewing circumstances or tasks (e.g., in a chromatic banding task) they may detect more categorical color differences compared to trichromat controls.¹⁵

An alternative explanation of the poor FM100 performance of some of these putative tetrachromats was offered by Jameson et al.²¹ That is, based on the actual performance data they suggest that such tetrachromats required a personal “correct” ordering that does not exactly follow the FM100 stimulus sequence. In this scenario, a putative tetrachromat may exhibit transpositions in the sorting task that disagree with the diagnostic’s standard sequence. This personal ordering scenario is possible if in some cases FM100 cap transpositions reflect an unimpaired individual’s non-normative just-noticeable-difference (jnd) variation along one or more color space dimensions (rather than sorting errors due to poor color sensitivity).  This is an interesting alternative interpretation of abnormal performance on a standardized test of color perception, which raises prospects for further demonstrating differences in retinal tetrachromat color processing.

Figure 3.  Polar coordinate plot of performance on the Farnsworth-Munsell 100-hue test (FM100) for a female individual with a diverse photopigment opsin genotype (reported as Subject 85 in Table 1 Jameson, Bimler & Wasserman, 2006).²¹ Genotype determined using the Wasserman et al.¹⁰ method found this individual heterozygous for both X-chromosome linked opsin genes--or a female heterozygote with a codon-180 dimorphism for the L-cone opsin (L-opsin Ser-180-Ala), and a codon-180 dimorphism for the M- cone opsin (M-opsin Ala-180-Ser). Despite otherwise excellent color perception, FM-100 compression parameter analyses showed that this individual’s patterns of FM100 confusion were displaced in a direction corresponding to a 15° axis in a polar coordinate compression parameter space.²¹ The FM100 performance shown indicates this individual performed very poorly compared to normative age-adjusted performance by an average Z value equal to 2.54 standard deviations and is likely diagnosed as false-positive deficient. This subject otherwise had unimpaired color perception, zero errors on the Ishihara pseudo-isochromatic plates, and reliably perceived 12 different chromatic bands in the Jameson et al.¹⁵ diffracted spectrum task, which is significantly greater than the average chromatic banding observed for trichromat female controls. Copyright Kimberly A. Jameson. Image courtesy of the author.

Implications for potential human tetrachromacy from other species.

Figure 6. The squirrel monkey (Saimiri sciureus) species possesses opsin genes that are ideal for attempting a transgenic cure for dichromacy.³⁹ In the squirrel monkey gene pool are three variants (or “alleles”) of the X-linked cone photopigment gene: one coding for a protein similar to the human M-photopigment (with pigment absorption maxima around 538 nm), a second coding for a protein similar to the human L-pigment (with absorption maxima around 561 nm), and a third coding for a pigment with light-absorption properties roughly midway between the first two (around 551 nm). By having two X-chromosomes, a female squirrel monkey might inherit two different longer-wavelength alleles (one on each of her X-chromosomes), and in this way she’ll acquire trichromacy (for more on these genetic mechanisms see Jacobs and Nathans 2009).⁴⁰ However, about a third of all female squirrel monkeys, will inherit the same pigment allele on both their X chromosomes and end up as dichromats, like the dichromat male squirrel monkeys. It is the latter female genotypes that were additionally missing the L-cone opsin gene that Mancuso and colleagues performed their transgenic cure for dichromacy.³⁹ Image by flickr member: mape_s.

Other terrestrial species have evolved color vision tetrachromacy in the spectral region “visible” to humans.

It is easy to think that human trichromacy in its current state is already optimized for our environment. After all, if it wasn’t optimized we’d notice, right? To understand the implications of this idea on human color procesing it helps to consider other terrestrial animals that require more than three functional photopigment classes that operate in approximately the same spectral window that humans use. That is, species that have color processing systems with operating ranges that are not hugely different from those of humans, but which have more degrees of variation than a trichromatic system. One example is the European Starling (the small to medium-sized passerine bird, Figure 7). In addition to a visual pigment that peaks in the near UV (at 362 nm), Starlings have three photopigments that roughly resemble the long-, medium- and short-wave sensitive pigments of humans. Although the European Starling UV pigment peaks outside the lower limit for the human operating range (i.e., shorter than 400 nm), one tail of the UV pigment responds considerably, and overlaps with all of the other Starling photopigment response curves, within a 400 nm to 700 nm range (Figure 8). Color discrimination performance suggests that at least some of the Starling’s other pigment curves appear to be coupling signals with the UV pigment.⁴¹  Thus, the case of the European Starling suggests that within a humanly usable range of ~400 nm to ~700 nm, tetrachromacy is clearly a viable form of color processing for these birds.

Figure 7. The European Starling (Sturnus vulgaris), a common bird native to most of temperate Europe and western Asia, is a color vision tetrachromat. Image by flickr member: daBins.

Figure 8. European Starling sensitivity (top) compared to human photopigment sensitivity (bottom). The important point to note is that although the Starling’s UV pigment peaks outside the lower limit for the human operating range (i.e., shorter than 400 nm), one tail of the UV pigment responds considerably, and overlaps with all of the other Starling photopiment response curves, within the “humanly visible” ˜400 nm to ˜700 nm range. The substantial overlap among sensitivity curves, in addition to the birds’ color discrimination performance, suggests that at least some of the Starling’s other pigment curves appear to be coupling signals with the UV pigment. Together these features suggest an achieved increase in discrimination that is of significant enough chromatic resolution to justify an evolutionary adaptation. Image adapted from Palaeontologia Electronica (http://palaeo-electronica.org/2000_1/retinal/fig_7.htm).

About the Author