Christopher Torrens, University of Southampton, UK
Our experience of the world around us is filtered through our senses, making our interpretation of various stimuli uniquely subjective. When Claude Monet developed cataracts in his 60s, his paintings changed and reflected the world as he saw it. Those paintings from 1915 – 1923 were no longer light but rather duller with more reds and browns than before. In those paintings, we see the impact of the cataract on the artist’s vision. This author does not have cataracts, but like ~8% of Western European men, and ~0.5% of Western European women, I have a colour vision deficiency and my experience of the world is different to many.
Colour vision appears to have evolved alongside light-sensitive photoreceptors. Vision appears to take off from around the time of the Cambrian explosion (~545 – 530 million years ago) and over those 15 million years, vision starts to appear with the huge increase in diversity of life. Strikingly, the structures for vision are well conserved. The photoreceptors in prokaryotes and eukaryotic protists, such as algae, are similar to our rods and cones (Williams, 2016).
The key family of proteins in rods and cones are called the opsins. These are part of the superfamily of G-protein coupled receptors as they have seven transmembrane domains and are associated to G-proteins. The opsins in cones and rods are different, with rods for low light (scotopic) and the cones for well-lit (photopic) and, crucially, colour vision. When unstimulated, cones and rods are depolarised, with cGMP bound to cyclic nucleotide-gated Na+ channels, allowing Na+ influx and the so-called “dark current”. This “dark current” raises membrane potential and leads to glutamate release. In response to light stimulation at the appropriate wavelength, there is activation of the G protein leading to activation of phosphodiesterases and the decrease in cGMP. The subsequent closing of cyclic nucleotide-gated ion channels hyperpolarises the cell and decreases glutamate release (Kaupp and Koch, 1992).
What conveys the ability to see colour is the presence of opsins that are stimulated by different wavelengths of light. Very early in the evolution of vertebrates (~350 – 400 million years ago) four gene families of opsins in cones as well as the rod opsin were present allowing a spread of vision across a range of wavelengths (see Table 1). Indeed, there are some species of fish, birds and reptiles who are tetrachromatic and can see ultraviolet light. (As an aside, some snakes can “see” in infrared, but this is unrelated as it is not via the eye but rather the specialised pit organ that creates a separate thermal image in addition to an optic one.)
Mammals, having started off with the other vertebrates enjoying tetrachromacy, lost this along the way. Whether it is because early mammals were burrowing or nocturnal is unclear but most mammals are now dichromatic, having lost both the short-wave sensitive 1 (UV) and the medium- wave sensitive rhodopsin-like 2 (green). Interestingly, monotremes such as the platypus still have the remnant of the short-wave sensitive 1, suggesting UV sensitivity was still around in mammals at the time of monotreme divergence (~165 million years ago). While it is true that most mammals are dichromatic, primates are trichromatic, meaning we primates re-evolved trichromacy around 35 million years ago (see Table 2) (Jacobs, 2009).
This is important for the development of colour vision deficiency. Having lost the medium-wave opsin (Rh2) millions of years previously, the re-emergence of the green sensitivity comes from a gene-duplication of the long-wave opsin which has two important impacts. Firstly, in primates the overlap of wavelength between the medium- and long-wave opsins is much more than in the other vertebrates with Rh2. Secondly, the gene encoding the long-wave opsin is located on the X chromosome and so now both the long-wave and the duplicated medium-wave opsins are on the X chromosome (Dulia et al., 1999). As a result, colour vision deficiency is a recessive sex-linked trait, far more common in males (about 1 in 12) than in females (about 1 in 200; Figure 1). With both of these genes situated on the X chromosome, males only inherit one copy and consequently only need to inherit one defective copy to present with the symptoms. Since females have two X chromosomes, and because it is recessive, this would require the inheritance of two defective copies before presenting and so rates are far lower in females.
Colour vision deficiency is due to either missing or dysfunctional cones. Depending on how many and which ones are affected, colour vision deficiency can be split into either:
- Monochromatism: either no cones at all or only one type available
- Dichromatism: only two different cone types, the third one is absent
- Anomalous trichromatism: all present but with altered sensitivity in one
Depending on which cone is missing or dysfunctional, the latter two can be sub-classified based on the cone (and therefore colour) affected. If the L-Cone is affected, the ability to see red light is disrupted resulting in one form of red-green colour blindness. This is known as protanopia if the L-Cone is missing (as in dichromatism) or protanomaly if it is simply dysfunctional. If the M-Cone is affected, the ability to see green light is disrupted leading to the other form of red-green colour blindness. This is known as deuteranopia or deuteranomaly depending on whether or not the M-cone is missing or just dysfunctional. An abnormal M-cone is the most common problem leading to colour blindness and is identified in ~75% of cases.
An abnormality in either the L- or the M-cone does not simply result in deficiencies in picking up reds and greens. Since these work on a range of wavelengths (Figure 2), perception of all light within those wavelengths may be disrupted. Reds, browns, oranges, yellows and greens all fall within these wavelengths and are common problems. For my own perspective, this has limited impact on my day to day life and it is not considered a disability in the UK. That said, as a working physiologist there are considerable challenges to people with colour vision deficiency – particularly around imaging.
Given that the most common colour vision deficiency is red-green colour blindness, it is unfortunate that green and red stains predominate in microscopy. There is actually a good reason for this: while most people can distinguish between red and yellow as they are detected by the same opsin, the wavelengths are a little too close for a microscope to separate into different channels. So if you are looking for two proteins, one with a red stain and one with a yellow, you are liable to pick up some yellows in the red channel and some reds in the yellow due to the width of the emission spectra of the fluorophore. Red and green however are sufficiently far apart to allow them to be picked up in separate channels so that you can tell which of your two proteins is which. This seems fair enough, but it does not explain the lack of blue staining. Blue is further away again and would certainly be more specific as a second colour. While there are blue stains available, the vast majority appear to be red and green, which perhaps reflects a bit of historical convention (after early fluorophores, FITC and TRITC).
However, in this day and age and with such advanced digital systems, there is no reason to stick to the colours of recording channels. Software packages such as ImageJ and Adobe Photoshop have a number of functions that allow for different colour combinations to be presented. Indeed, the Color Universal Design Organization (CUDO) – a non-profit organization in Tokyo, Japan – has produced a handbook to provide the rationale and the means to make these changes (CUDO, 2006). One such idea is the replacement of red stains with magenta, easily achieved on free open-source imaging software such as ImageJ which provides step-by-step guides (Ferreira & Rasband, 2012). The advantage of magenta is that it is an equal mixture of red and blue, so that people who struggle with the red component can easily detect the blue hue, whilst the overlap of two positives appears as white.
Much of these ideas have been led from the ground up, with individuals blogging and presenting these alternatives (Cox, 2015; Keane, 2015). The journals themselves have a role to play in taking this further by including these ideas in their formatting guidelines. I shall see many images that are indecipherable to me, but there are changes beginning to creep in. Nature Medicine has the following on its instructions to authors:
“Authors are encouraged to consider the needs of colorblind readers (a substantial minority of the male population) when choosing colors for figures. Many colorblind readers cannot interpret visuals that rely on discrimination of green and red, for example. Thus, we ask authors to recolor green-and-red heatmaps, graphs and schematics for which colors are chosen arbitrarily. Recoloring primary data, such as fluorescence or rainbow pseudo-colored images, to color-safe combinations such as green and magenta, turquoise and red, yellow and blue or other accessible color palettes is strongly encouraged.”
Such advances are useful and will make a big difference to work and science, but these are of limited value in life outside of work. There are now a range of glasses commercially available to improve colour vision. The problem with most colour vision deficiency is the overlap in the wavelengths between the M- and the L-cone. These glasses work by filtering out the light at those overlapping wavelengths, giving a more accurate interpretation of the colour. What this means is that people will not see any new colours, but rather perceive the same colours in a different way. Having tried some examples of these, I did see some numbers in the Ishihara plots, but everything else had a purplish hue that was similar to wearing sunglasses indoors.
As a young man, I was told that being colour blind meant that I could never be a pilot or an electrician (neither of which is actually true) but since I had no desire to be either, it had limited impact. Adoption of the ideas proposed by Color Universal Design Organisation would make a significant improvement in my working experience as someone with colour vision deficiency. That said, the single biggest improvement in my experience would come from not asking me “what colour is this?”
Cox L (2015). Tips for designing scientific figures for color blind readers. [Online] Somersault18:24. Available at: somersault1824.com/tips-for-designing-scientific-figures-for-color-blind-readers/ [Accessed 30 May 2019]
CUDO (2006). Color Universal Design Organisation Handbook. [Online] EIZO. Available at: eizoglobal.com/products/coloredge/unicolor_pro/handbook.pdf [Accessed 30 May 2019]
Dulia KS et al. (1999). The evolution of trichromatic color vision by opsin gene duplication in New World and Old World primates. Genome Research 9, 629 – 38.
Ferreira T, Rasband W (2012). Section 9 – Color Images. [Online] Image J User Guide: IJ 1.46r Revised edition. p. 14 – 17. Available at: imagej.nih.gov/ij/docs/guide/user-guide.pdf [Accessed 30 May 2019]
Jacobs GH (2009). Evolution of colour vision in mammals. Philos Trans R Soc Lond B Biol Sci 364, 2957 – 67.
Kaupp UB, Koch KW (1992). Role of cGMP and Ca2+ in vertebrate photoreceptor excitation and adaptation. Annu Rev Phvsiol 54, 153 – 75.
Keane DR (2015). A review of color blindness for microscopists: guidelines and tools for accommodating and coping with color vision deficiency. Microsc Microanal 21, 279 – 89.
Williams DL (2016) Light and the evolution of vision. Eye 30 173 – 8.
Recommended further reading:
Birch NJ. The art of perfect presentation. New Scientist 1995; 145(1965):48-49.
Birch NJ. Communicating biology: the rule of sixes. Biologist 43, 151-153. 1996.
Birch NJ. The very last minute slide. Lancet 1994; 343:1434.