Nature's fibre optics

THE retina has an inverted structure which seems ill-suited to its function: the rod and cone cells, which are sensitive to light, and which convert light energy into electrical impulses, point backwards and are located at the back of the retina, so that light entering the eye has to pass through several layers of irregularly organized cells before it reaches them. The retina also contains nerve fibres which are positioned perpendicular to the path of light entering the eye, and many of the structures in the upper cell layers have a diameter similar to that of the wavelength of visible light. One would therefore think that light entering the eye would be subjected to a significant amount of reflection and scattering. Yet, nature somehow contrived to overcome this awkward architecture, and the retina performs its function perfectly.

As well as the various types of neurons, the retina contains specialized glial cells called Müller cells, which are arranged in parallel to each other and are oriented in the direction along which light travels through the eye. Müller cells are about 150μm (micrometres, thousandths of a millimetre) in length, and span the entire thickness of the retina, projecting from the vitreous humour (the viscous fluid in the back of the eye) to the back of the retina where light enters the rods and cones. Like other glial cells, Müller cells have been largely ignored until recently: they were thought to do little more than support and nourish retinal neurons. But in recent years it has been determined that glial cells perform other important functions, and Müller cells may overcome the retina's architectural problem, by functioning as optical fibres which transmit light through it.

Jochen Guck of the University of Cambridge and his colleagues dissected guinea pig retinae, placed them under a confocal scanning microscope, and used a light source to mimic the natural illumination to which the tissues would be exposed. They then scanned the back of the tissues to determine the patterns produced there by light entering the tissue. To their surprise, the images they obtained contained a regular pattern of bright spots which alternated with darker areas. Retinae from rabbits and humans showed the same reflection pattern. Serial sections were then cut from the back of the retina. The dark spots they had observed were found to be 2-3 μm in diameter and spaced 5-6 μm apart. When the sections were reconstructed, it was found that the dark spots formed tubular structures that ran the entire thickness of the retina, and that they had funnel-shaped structures with a diameter of approximately 15 μm at one end. The extent of back-scattering (that is, the scattering of light in the direction from which it came) in the different regions of the back of the retina was then examined. This showed that there was significant back-scattering in cell layers near the photoreceptors, but very little in the dark spots, indicating that the tube-like structures within the dark spots transmit far greater amounts of light than the surrounding tissues. The arrangement of tube-like structures was also found to correspond well to the size and spacing of Müller cells (stained red in the figure below).

muller cells retina.JPG

The researchers then investigated the optical properties of the Müller cells. Retinae were treated with enzymes that cause the cells in the tissue to dissociate from one another. A dual beam laser trap was then used to keep the cells in a fixed position, so that their ability to propagate light could be investigated. Individual cells floating in a suspension were aligned between the ends of two optic fibres (above right). One of the fibres could be used to shine light onto the cells; the other was connected to a power meter. By bringing the fibres into contact with a free-floating cell, and passing light from one fibre through the cell to the other fibre, the amount of light passing through the cell could be measured. It was found that the amount of power entering the output fibre was greatest when a Müller cell was aligned in the same orientation as the fibres; but when the cell was rotated or removed from the trap altogether, the power output decreased dramatically. An orange dye, which fluoresces when struck by light, was then injected into the Müller cells, enabling the path of light to be visualized. This confirmed that light was indeed passing straight through the cells.

This elegant set of experiments shows that the Müller cells function as conduits which guide the passage of light through the tissue of the retina. The funnel-like structures observed are the endfeet of the Müller cells, which are densely packed and form a cobblestone pattern on the membrane closest to the vitreous humour of the eye. During the laser trap experiments, one of the fibres was misaligned so that, in the absence of a Müller cell, the light did not strike the second fibre; when a Müller cell was then placed between the two fibres, it captured the light from the first fibre and guided it towards the second. Thus, the endfeet seem to be crucial in capturing divergent rays of light and guiding them towards the photoreceptors at the back of the retina. They also have a lower refractive index than other parts of the Müller cell and other cells in the retina, and serve to minimize reflection of incident light as it passes from the vitreous humour into the uppermost layers of the retina.

The way in which Müller cells transport light is similar to the mechanism by which the optical fibres in fibre optic plates carry light. Fibre optic plates consist of optic fibres bundled together, and are used instead of lenses to transfer images between distant locations; this occurs without distortion of the image or loss of image detail. Müller cells may perform the same function in the retina; each one is coupled to one cone photoreceptor and (in guinea pigs and humans) ten rods; the Müller cell arrays could therefore faithfully transmit the pattern of light falling on the front of the retina to the photoreceptors at the back of the retina, thus minimizing distortion of the image. And, because the Müller cells are funnel-shaped and narrow, they take up only 20% of the space in the retina; this leaves plenty of room within the tissue for the neural circuitry.


Franze, K., et al (2007). Müller cells are living optical fibers in the vertebrate retina. Proc. Nat. Acad. Sci. 104: 8287-8292. DOI: 10.1073/pnas.0611180104

More like this

This is the first of eight posts on evolutionary research to celebrate Darwin's bicentennial. If you were a designer tasked with creating a machine for collecting and processing light, the last thing you would come up with is the human eye. Darwin marvelled at the eye as an "organ of extreme…
New research shows that a protein found in green algae can partially restore visual function when delivered into the retina of blind mice, taking us one step further towards genetic therapy for various conditions in which the degeneration of retinal cells leads to imapired vision or complete…
tags: Ask a Science Blogger, vertebrate eye, molluscan eye Image: Wikipedia. [larger view]. The newest "Ask a Science Blogger" question is; Which parts of the human body could you design better? Since I have only 500 words or so to explain, I will discuss only one anatomical feature: I would…
A while back, I summarized a review of the evolution of eyes across the whole of the metazoa — it doesn't matter whether we're looking at flies or jellyfish or salmon or shrimp, when you get right down to the biochemistry and cell biology of photoreception, the common ancestry of the visual system…

With these dimensions, are there also diffractive effects back there? I ask because of an observation I have made: despite many attempts, I have been unable to capture a photographic image of the diffraction generated sheen on the backs of the English Kingfisher wings. There is some amount of coherence to this reflected light, because of its genesis. There are similar visual effects from looking at laser light or its reflection. I have been suspecting for some time that there is something going on in the eye that does not happen in a camera because of retinal effects.

If you haven't seen this, you will have to visit an English creek or river and observe it first hand. No description or photo can suffice. I lived on the banks of a small river just N of London for 5 years and we had a nesting pair on our property.

By Gray Gaffer (not verified) on 28 Aug 2009 #permalink

I wonder what is happening at the fovea. Light does not pass through intermediate cellular layers, and therefore bypasses the muller cells, as it reaches the fovea. Because the fiber optics "steer" light, a slight deviation in the path of the muller cell fiber may produce a distortion of the image on the retina. It makes sense that the fiber optic mechanism is present in areas such as the peripheral retina that do not process high-res images.

The macula has a dense concentration of photoreceptors, and the overlying plexiform, nuclear and ganglionic layers are markedly thinned out; even the overlying retinal vessels are absent: the photoreceptor cells get their nutrition from a plexus of high vascularity in the underlying choroid.

Mollusca have it right: their phatoreceptors form in the inner layer of the optic cup and light does not have to traverse structural layers. And they may have 15 different color sensors (instead of the three that Chordates may have:
one wonders they perceive 15 primary colors!

By Robin Datta (not verified) on 07 Sep 2009 #permalink

We're riddled with cell groups that seem to be doing nothing. That is at least until the next scientist discovers that indeed there is a purpose. I love the way that works. Maybe someday we'll figure out that we are an integral system, not just a conglomerate of parts.

P.S. I was hoping to get to your posts on wordpress.com. Loved the blog design and build.