Content uploaded by John Mollon
Author content
All content in this area was uploaded by John Mollon on Jun 22, 2022
Content may be subject to copyright.
Tetrachromacy: The mysterious case of extra-ordinary color vision
Gabriele Jordan and John Mollon
Recent years have witnessed a growing public interest in human tetrachromacy –
the possibility that a subpopulation of women enjoy an extra dimension of color
vision. Yet, by contrast, rigorous studies of this unusual phenotype are sparse. The
aim here is to offer the reader a guide to the facts and myths regarding this potential
‘superpower’ and to address the core methodological issues that need to be
considered when investigating it.
Introduction
Tetrachromatic color vision allows an individual to discriminate colors along a
perceptual dimension that is unavailable to the normal person. Most people’s color
perception is ‘trichromatic’, i.e. their retinas contain three types of light-sensitive cone
photopigment, absorbing maximally in the long- (L), middle- (M) and short-
wavelength (S) parts of the spectrum. Behaviorally, in a color-matching experiment
using spectral lights, they can match any color with just three variable colors: A
match between two stimulus fields can be achieved simply by equating the triplets of
cone signals in the two fields [1]. For tetrachromacy to arise, there must be four
distinct types of cone photopigment in the retina and their four independent signals –
or three independent ratios – must be available to later stages of processing.
One reason that tetrachromacy is of interest lies in its implications for plasticity within
the visual system: It is implausible that an additional chromatic pathway has evolved
to benefit a small minority of our conspecifics who are heterozygous for X-linked
cone photopigments. But how then do the retina and the cortex take advantage of
an additional, independent cone signal? Even with four types of cones, vision might
remain trichromatic in the sense that only three variables are needed in a color-
matching experiment [2]. We have called this ‘weak tetrachromacy’ [3]. The more
interesting case is ‘strong tetrachromacy’, where a woman requires four independent
stimuli in matching colors, and it is this possibility we discuss below.
In this review we address the following questions: (1) How can tetrachromacy arise
and why only in women? (2) What combinations of cone photopigments are required
for tetrachromacy? (3) What are the retinal mechanisms allowing for plasticity of the
visual system? (4) What are the methodological challenges to demonstrating the
existence of this phenotype? (5) Does tetrachromacy confer any advantages?
How can tetrachromacy arise and why only in women?
The photopigments (‘opsins’) in the L and M cones are encoded by adjacent genes
on the X-chromosome [4*]. Men inherit a single, maternal X-chromosome, whereas
women inherit one from each parent. However, to compensate for the difference in
X-chromosome dosage between men (XY) and women (XX), most genes on one X-
chromosome are inactivated in any given cell of a woman’s body [5*]. This
inactivation is random and initiated very early in embryonic development. Consider
now a woman whose X-chromosomes carry different versions (alleles) of, say, the M
opsin gene. If these alleles encode photopigments that differ in their wavelengths of
maximal sensitivity, then her retina will contain four types of cone photopigment: S,
M, L’ (or shifted M), and L [Fig 1a]. X-chromosome inactivation segregates the
alternative gene products in different cones and so provides a retinal cone mosaic
with four cone types [Fig 1b]. The size of retinal cone patches will depend on the
time of onset of X inactivation and the migration of cones in the developing retina.
The existence of tetrachromatic women was first postulated by de Vries [6]: The
women that he thought “must be tetrachromatic” were the daughters of anomalous
trichromats – for whom one photopigment (M or L) is spectrally shifted [7*]. His
hypothesis gained plausibility from the discovery in Platyrrhines - New World
monkeys - of a remarkable polymorphism of cone photopigments [4*, 8]. Most
species of Platyrrhini are basically dichromatic, combining an S-cone photopigment
with one that peaks in the middle- to long-wavelength spectral region; but there are
typically three (or more) X-chromosome alleles for the latter pigment. Males (XY)
must always be dichromatic, but a female (XX) will be trichromatic if she inherits two
different versions of the middle-/long-wave gene. If random X-chromosome
inactivation can give rise to trichromats in a basically dichromatic species, could it
not yield tetrachromacy in ours?
Which complement of photopigments gives rise to tetrachromacy?
Two alternative hypotheses.
In our species, the genes for color vision are highly polymorphic, and probably most
women are heterozygous in the sense that the arrays of opsin genes on their two X
chromosomes are not identical. But how many women are potentially
tetrachromatic?
One of the most commonly observed polymorphisms is that at site 180 in the amino-
acid sequence of the L cone opsin. In 62% of Caucasians serine (ser) is present at
this site, but in 38%, alanine (ala) [9]. The two versions of L photopigments differ by
4–5 nm in their wavelength of peak sensitivity, with ser180 being more sensitive to
longer wavelengths [10, 11]. When this ser180ala polymorphism was first described,
one of us raised the possibility that women heterozygous at site 180 might be
tetrachromatic [12]; and in a series of subsequent papers, Jameson and colleagues
have suggested that this polymorphism may indeed underlie tetrachromacy [13-15].
Tetrachromacy could then be very common: the proportion of ser180ala
heterozygotes in a Caucasian population would be 1 – (0.622 + 0.382), i.e. about
47%. The latter estimate has been confirmed in a gene expression study of nine
female donor eyes [16].
Would a separation of 4–5 nm between two L photopigments be enough to give the
potential tetrachromat a salient, extra signal? Here we can draw on earlier modelling
of ‘anomalous trichromacy’ – the mild colour vision deficiency seen in ~6% of men,
where either the normal L cone or the normal M cone is replaced by a cone shifted in
spectral position [7*]. If the two remaining cones in the middle-/long-wave region are
separated by only 4–5 nm, the estimated number of discriminable colors in real-
world hyperspectral images would increase relative to the number discriminated by a
dichromat who had only L and S cones [17]; but performance would be poorer than
that of a normal trichromat – or that of an anomalous trichromat with more widely
separated photopigments.
Since the common polymorphism at amino acid site 180 would give only a small
spectral separation, and since in practice strong tetrachromacy appears to be rare,
we ourselves have favored an alternative hypothesis, that tetrachromacy has the
best chance of arising in those carriers of anomalous trichromacy who have normal L
and M photopigments plus an anomalous photopigment that lies midway between
them. Anomalous photopigments of this type are thought to arise from hybrid genes.
The highly homologous [18] L and M genes may misalign at meiosis. If now a
crossing over occurs within the mispaired genes, the recombined outcome may
consist partly of the normal L sequence and partly of the normal M sequence. The
expressed opsin typically has a peak sensitivity intermediate between L and M. We
note that the strongest candidate for tetrachromacy identified so far (cDa29) [19] has
an estimated separation of 12 nm between her anomalous, L’ cones and her L
cones.
Anomalous trichromats vary widely in their ability to discriminate in the red–green
spectral range [7*]: Some are nearly as limited as dichromats, while a tiny minority –
the ‘Minimalanomale Trichromaten’ of Vierling [20] – perform normally on standard
tests for color deficiency and have matching ranges close to zero on the
anomaloscope. Our hypothesis is that it is the daughters of the latter men who are
most likely to become tetrachromatic.
Post-receptoral mechanisms allowing for tetrachromatic signals
In order to perceive colors, and to discriminate between them, the cone activations
from different cone types must be compared. Two independent, post-receptoral
channels have been identified for this task [21-24]. Of particular interest for
tetrachromatic vision is the phylogenetically younger, ‘red-green’ opponent channel.
Midget ganglion cells of the parvocellular pathway have receptive fields that are
divided into spatially and chromatically antagonistic centers and surrounds. In the
fovea, their center input is drawn from a single cone of one type, whereas the
surround input is likely to be drawn indiscriminately from L and M cones (mixed
surround) [22]. Such a channel lends itself to tetrachromatic vision as its chromatic
specificity is determined primarily by the type of cone feeding into the center [8].
Thus, if the signal from a single L’ cone is compared to the pooled input of other
cones, say L+L’ in the case of men with deuteranomaly (Da), a female carrier (cDa)
will have the neural basis for similar comparisons. Figure 1b shows a schematic
representation of a tetrachromatic cone mosaic with a deuteranomalous (L’ vs (L+L’))
as well as normal (L vs (L+M) and M vs (L+M)) comparisons by three potential
midget ganglion cells. If all ratios of cone excitations are capable of providing salient
input and are preserved in cortical mechanisms, then strong tetrachromacy should
result. The chromatic signal should be largest if the cone input into the surround of
the receptive field is relatively unmixed. This is more likely to occur in cases of cone
clumping and/or cone mosaics that are biased towards L-type cones [25]. Note, that
such retinas will yield spectral luminous efficiency functions (which depend on the
summed signal from L- and M-type cones) very different from normal. Thus it is
especially important to eliminate or adjust for luminance cues when testing for
tetrachromacy (see below). cDA29’s relative sensitivities to the red and green
phosphors of a Sony Trinitron monitor [25] were indicative of a retinal mosaic
dominated by L-type cones [19].
Methodological considerations
Knowledge of an individual’s receptor types, either through molecular genetic
techniques or through classifying a carrier’s first-degree relative(s), is not sufficient to
indicate the dimensionality of her color vision [19, 26*]. How can tetrachromacy be
experimentally revealed?
Color matching using spectral lights: No ordinary RGB color display is suitable for
such a test. Jordan and Mollon [3] used a specifically designed colorimeter to test
whether obligate carriers make unique color matches in the 546 to 690 nm spectral
range compared to control participants when given an additional primary in a ratio-
matching task. Only one (cDa1) out of 14 obligate carriers for simple anomalous
trichromacy made such unique matches.
Color discrimination using spectral lights: We later argued that a forced-choice
discrimination task would be a more objective test for tetrachromacy than a criterion-
based matching task [19]. As before, the primaries were in the spectral range where
S cone are insensitive. A triplet of three successive lights was presented. One light
was a red-green mixture (546 + 670 nm) that could vary in the proportion of red, and
the other two were a monochromatic yellow (590 nm). Observers had to identify the
red-green mixture. Since all normal trichromats can make a dichromatic color match
in this reduced spectral range, they were expected to fail to discriminate for some
red-green mixture and some intensity of the monochromatic yellow. In contrast, a
strong tetrachromat was predicted to always detect the mixture. Only one (cDa29)
out of 18 cDa and none of 7 cPa participants performed according to the prediction.
All control subjects failed to discriminate. Since cDa29 could not make a match with
only two primaries available in the Rayleigh region, she was classified as a strong
tetrachromat.
Color discrimination using surface colors: A good test for strong tetrachromacy is to
target carriers of deuteranomaly with a test specifically designed to reveal the
anomalous L vs (L+L’) channel [27]. Deuteranomalous men with small Rayleigh
matching ranges (good discrimination ability) are able to make discriminations not
achievable by normal observers [19, 27, 28]. If the L vs (L+L’) channel is salient
enough in cDa individuals, they ought to be able to make the same discriminations.
Luminance was jittered across the normal and deuteranomalous dimensions to
ensure that discriminations were based on chromaticities. Multidimensional scaling
(MDS) of each observer’s dissimilarity ratings was used to reconstruct their
subjective color space. Four out of nine cDa observers were found to be able to
make discriminations on this test and their MDS solution correlated significantly with
the theoretical L vs (L+L’) channel. They included cDa29 [19].
Other approaches: To “extend the stimulus complexity beyond the [reductionist]
color-matching configuration” [13, p247] and to improve the ecological validity of the
experimental tests for tetrachromacy, Jameson and colleagues have introduced two
different approaches.
First, they asked observers to delineate the perceived chromatic bands of a
projected diffraction spectrum by drawing lines on tracing paper [13]. They report a
richer color experience in 23 carriers of the L cone ser180ala genotype compared to
controls. On average, these carriers delineated 10 (range 6–15.67) spectral bands
compared to 7.6 (range 5.67–11.83) and 7.3 by normal females and males
respectively. Intriguingly, those carriers who indicated the highest number if
delineations, showed a reduced gamut (i.e. the endpoints of the spectrum) of
perceived colors compared to other participants. Smeulders et al [29] using the same
approach, report that in normal trichromats increasing the number of delineations in
the projected spectrum increases the perceived chromatic bands from 7 to 25-30.
This dramatic increase was explained as a by-product of an increase in luminance
contrast transients introduced by the lines made. We note that the bands of a
projected spectrum are not of equal luminance [29] and that carriers will have altered
sensitivities to given spectral bands owing to their altered luminous efficiency
function.
Second, the authors have linked the ser180ala genotype to the experiential
sensations of selected carriers’ and to the ability to communicate these sensations
[14]. In particular they have asked whether tetrachromacy reveals itself in the use of
color in artistic expression. Two out of four carriers were professional artists, and one
of two normal female controls was also an artist. The minimum motion technique of
Anstis and Cavanagh [30] was used to estimate the subjective isoluminance points
of 20 colors (produced on a RGB monitor) to four levels of gray (of unspecified
luminance) and one uniform background color. Results for one of two potential
tetrachromatic artists (subject CA) show different equiluminant settings for some of
the color patches compared to other observers. It was also found that her artwork,
that aims to represent what she sees, indeed differs in color use from that of the
trichromatic artist. CA’s sister’s (LA) artwork also uses color to a great effect.
Recruitment issues and population statistic. Behavioral studies rely on the
participation of volunteers, but the recruitment strategy may create biases that can
influence inferences from the study’s outcome. For example, if one recruits mothers
who have color-deficient sons [3], the sons tend to be more severely affected as they
are more aware of their deficiency. Necessarily, one misses the minimal anomalous
trichromats [20] who are likely to have the largest spectral separation between their
X-linked cone photopigments and this might partly explain the low number of
potential tetrachromats found. Using the 2012 UK Census data, we estimated a
frequency of 3.9 million carriers of hybrid genes, but only 48,585 carriers who are
expected to have the optimal spectral separation of 12 nm between their L/L’ (M/M’)
photopigments. Population screening should therefore be all inclusive and performed
with an anomaloscope or other appropriate screening test.
What are the potential benefits to a tetrachromat?
We do not yet know whether functional tetrachromacy confers any advantages, but it
seems clear that the benefits, if there are any, are subtle. On the basis of anecdotal
evidence we have speculated that there may be signals in the skin tone of
conspecifics that are linked to health or disease [31]. In such cases early detection
will be advantageous.
Will a potential tetrachromat be a more successful artist? If she uses her color
palette veridically to render her sensations, then it is not clear how she can
communicate the added richness of her private gamut to those who live in an
impoverished perceptual world.
Conclusion
Tetrachromacy, a phenotype arising from X-linked genotypic variants, is an exciting
prospect and it now seems likely that there are individuals who fit the expected
profile. However, the relationship between the existence of a fourth class of retinal
cone and the dimensionality of colour vision is more complex than previously
thought. A number of factors may determine whether functional tetrachromacy
arises, such as the spectral distance between the peak sensitivities of the cones, the
distribution and relative number of cone types across the retina, and the relative
strength or weighting of chromatic signals that are sent from the retina to the cortex.
Taking into account the sub-optimal recruiting strategies for experimental studies
and the lack of an appropriate screening tool to provide a population statistic, there
are still more questions than answers.
Conflict of interest statement
Nothing declared
References and recommended reading
* of special interest
1. Stockman A, Brainard DH: Fundamentals of color vision I: color processing in
the eye. In Handbook of Color Psychology. Edited by Elliot AJ, Fairchild MD,
Franklin, A. Cambridge: Cambridge University Press; 2015: 27-69
2. Nagy AL, MacLeod DIA, Heyneman NE, Eisner A: Four cone pigments in
women heterozygous for color deficiency. J Opt Soc Am 1981, 71: 719-722
3. Jordan G, Mollon JD: A study of women heterozygous for colour deficiencies.
Vision Res 1993, 33: 1495-1508
*4. Carvalho LS, Pessoa DMA, Mountford JK, Davies WIL, Hunt DM: The Genetic
and Evolutionary Drives behind Primate Color Vision. Front Ecol Evol 2017, 5:
34: 1-11
A recent review discussing the different pathways for trichromacy in Old World and
New World primates.
*5. Balaton BP, Dixon-McDougall T, Peeters SB, Brown CJ: The eXceptional
nature of the X chromosome. Hum Mol Genet 2018, 27: R242-R249
A review of X-chromosome inactivation and of its exceptions.
6. De Vries H: The fundamental response curves of normal and dichromatic
and trichromatic eyes. Physica 1948, 14: 367-380
*7. Bosten J: The known unknowns of anomalous trichromacy. Curr Op Beh Sci
in preparation 2019
A discussion, in this issue, of inter-individual differences in anomalous trichromacy
covering genetics, phenotypic variability and modelling.
8. Mollon JD, Bowmaker JK, Jacobs GH: Variations of colour vision in a New
World primate can be explained by polymorphism of retinal photopigments.
Proc Roy Soc London B 1984, 222: 373-399
9. Winderickx J, Lindsey DT, Sanocki E, Teller DY, Motulsky AG, Deeb SS:
Polymorphism in red photopigment underlies variation in colour matching.
Nature 1992, 356: 431-433
10. Neitz M, Neitz J, Jacobs, GH: Spectral tuning of pigments underlying red-
green color vision. Science 1991, 252: 971-974
11. Merbs SL, Nathans J: Absorption spectra of the hybrid pigments
responsible for anomalous color vision. Science 1992, 258: 464-466
12. Mollon JD: Worlds of difference. Nature 1992, 356: 378-379
13. Jameson KA, Highnote SM, Wasserman LM: Richer color experience in
observers with multiple photopigment opsin genes. Psychon Bull Rev 2001,
8:244-261
14. Jameson KA, Winkler AD, Goldfarb K: Art, interpersonal comparisons of color
experience, and potential tetrachromacy. In International Symposium on
Electronic Imaging, Human Vision and Electronic Imaging 145, Society for Imaging
Science and Technology, 2016:1-12
15. Bochko VA, Jameson KA: Investigating potential human tetrachromacy in
individuals with tetrachromat genotypes using multispectral techniques. In
International Symposium on Electronic Imaging, Human Vision and Electronic
Imaging 542, Society for Imaging Science and Technology, 2018:1-12
16. Neitz M, Kraft TW, Neitz J: Expression of L cone pigment gene subtypes in
females. Vision Res 1998, 38: 3221-3225
17. Thomas PB, Formankiewicz MA, Mollon JD: The effect of photopigment
optical density on the color vision of the anomalous trichromat. Vision Res
2011, 51: 2224-2233
18. Zhao Z, Hewett-Emmett D, Wen-Hsiung L: Frequent gene conversion between
human red and green opsin genes. J Mol Evol 1998, 46: 494-496
19. Jordan G, Deeb SS, Bosten JM, Mollon JD: The dimensionality of color vision
in carriers of anomalous trichromacy. J Vis 2010, 10:12
20. Vierling O: Die Farbensinnprüfung bei der Deutschen Reichsbahn.
Bernecker: Melsungen;1935
21. Mollon JD: “Tho’ she kneeled in that place where they grew…” The uses
and origins of primate color vision. J exp Biol 1989, 146: 21-38
22. Dacey DM: Primate retina: cell types, circuits and color opponency. Prog.
Ret Eye Res 1999, 18: 737-763
23. Dacey DM, Lee BB: The ‘blue-on’ opponent pathway in primate retina
originates from a distinct bistratified ganglion cell type. Nature 1994, 367: 731-
735
24. Lee BB, Silveira LCL: Cone Opponency: An efficient way of transmitting
chromatic information. In Human Color Vision. Edited by Kremers J, Baraas R,
Marshall N. Springer Series in Vision Research 2016: 5, 105-132
25. Hood SM, Mollon JD, Purves L, Jordan G: Color discrimination in carriers of
color deficiency. Vision Res 2006, 46: 2894-2900
*26. Jacobs GH: Photopigments and the dimensionality of animal color vision.
Neurosci Biobeh Rev 2018, 86:108-130
An excellent, comprehensive review about the link between photopigment
complements and dimensionality of colour vision in the animal kingdom.
27. Bosten JM, Robinson JD, Jordan G, Mollon JD: Multidimensional scaling
reveals a color dimension unique to 'color-deficient' observers. Curr Bio 2005,
15: R950-R952
28. Regan BC, Reffin JP, Mollon JD: Luminance noise and the rapid
determination of discrimination ellipses in colour deficiency. Vision Res 1994,
34: 1279-1299
29. Smeulders N, Campbell FW, Andrews PR: The role of delineations and spatial
frequency in the perception of the colours of the spectrum. Vision Res 1994, 34:
927-936
30. Anstis S, Cavanagh P: A minimum motion technique for judging
equiluminance. In Colour Vision: Physiology and Psychophysics. Edited by Mollon
JD, Sharpe LT. Academic Press; 1983:155-166
31. Jordan G, Mollon JD: Eine evolutionäre Interpretation des menschlichen
Farbensehens. Die Farbe 1988: 35/36, 139-170
Fig 1. Retinal structures for tetrachromacy. (a) Relative sensitivities for the three
normal cone photopigment S, M, L and two hybrid photopigments L’ and M’. (b)
Schematic representation of a foveal cone mosaic showing S, M, L and L’ cones in
blue, green, red and orange respectively. Cone inputs into three possible circular
centre-surround receptive fields of midget ganglion cells are superimposed on the
cone mosaic. For information see text.
(a)
--
L’ vs (L+L’)
L vs (L+M)
+
-
-
M vs (L+M)
-
+
-
+
(b)
Tetrachromacy: The mysterious case of extra-ordinary color vision
Gabriele Jordan and John Mollon
Conflict of interest statement
Nothing declared
