What is the neural coding of rod and cone cells?

What is the neural coding of rod and cone cells?

We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

In Rushton's paper on the Principle of Univariance, he states:

Thus, though the rod input has two variables, wave-length and energy, the output differs only in one respect, namely 'brightness'.

However, as far as I understand, a photon can only vary according to one dimension, namely wavelength (vibrational energy), because the wavelength and energy of a photon are interdependent.

This suggests then that the input to a cone cell, for brightness to be perceived (i.e. the intensity of light), must be a beam of photons, so that the input can vary in wavelength and amplitude (does this equate to the number of quanta being absorbed by the photopigment per unit of time?).

My question, then, is how can a discrete event (the changing of shape of a photopigment) encode a continuous value (i.e. intensity)?

Indeed, brightness is a matter of photon count.

In any given rod, there are numerous photopigment molecules. There are about 1 billion rhodopsin molecules in any given rod, so although a change in conformation of any single rhodopsin molecule is a discrete event, you can consider the fraction of 11-cis- versus all-trans retinal (a photoreceptor's initial readout of brightness) to be effectively continuous, even more so because there is a <1 probability of a photon of light being absorbed by any rhodopsin molecule, so the discrete measure itself is an approximation of the 'actual' brightness/photon count.

Although it isn't perfectly analogous, because here we are talking about proportions rather than actual bitwise encoding, consider that computers are also able to represent fairly continuous values, but they are still always discrete approximations to those continuous values. A 24-bit computer monitor, for example, can actually only display 256 different intensities for each R, G, and B color stream (8 bits each).

The principle being discussed in the paper you describe is that once a rod (or cone) absorbs a photon, it loses any information about what the wavelength was, only that it was sufficient energy to activate the rhodopsin in that photoreceptor. There are several other questions and answers here on Biology.SE that discuss how wavelength (i.e. color) information can be extracted by differential activation of different cone types, for example:

Why can cones detect color but rods can't?

How do our eyes detect light at different frequencies?

More than the genes: How noncoding DNA controls cell types for vision

Non-coding RNA (ncRNA) profiling can be used to identify parts of DNA that determine how cells in the eye develop. One such region, highlighted here in green in a developing mouse retina, directs cells to grow into rods the red areas are for cones.

DNA contains the instructions for every component, function, and life cycle of each individual cell. The DNA library is expansive and vast, but all cells in our body use the same template. So, how is it that different cells within our bodies can use the same DNA, or genome, to make so many different cell types? How can the same instructions direct the cells of the heart, of the eye, and of every other part of our bodies?

New research from geneticists Carlos Perez-Cervantes and Linsin Smith in the lab of Ivan Moskowitz, MD, PhD, at the University of Chicago have developed a new way to identify the parts of DNA that control how one cell type is made instead of another. Their new approach helps to identify something called cis-regulatory elements, a noncoding part of the genome (described below) that determines the differences between cell types of the body.

Regulating your DNA

The &ldquocentral dogma&rdquo of biology is that genes in DNA are transcribed into a messenger RNA molecule, which cells can translate into proteins and then be used for a variety of functions. How then does a cell regulate which proteins to produce in each cell type? One way that cells accomplish this is by controlling which genes are turned on in a given cell through regions of DNA called cis-regulatory elements, or &ldquoenhancers.&rdquo

Enhancers are regions of DNA away from genes that are also transcribed into RNA, but these RNA molecules will not be a template for proteins and are known as non-coding RNAs. What controls the enhancers to turn on genes in some cells and off in others? Typically, a protein known as a transcription factor binds to enhancers in the DNA. These regions will produce non-coding RNAs that can help the transcription factors to start, increase, or decrease the expression of a nearby gene.

While cis-regulatory elements can help us understand how cells regulate genes, enhancers are hard to identify. Work from the Moskowitz lab and others has focused on how the non-coding RNAs made by enhancers influence what genes are being turned on or off to make a certain cell type. &ldquoPeople like to think of genes as either being on or off, like a light switch, but instead, enhancers are more like a dimmer,&rdquo said Smith, a graduate student in the Committee on Genetics, Genomics, and Systems Biology. &ldquoWe often don&rsquot know how the transcription factor controls the dimmers. Looking at the non-coding RNAs can really help us to understand this process.&rdquo

Determining an eye&rsquos fate through noncoding regions

To understand where cis-regulatory elements are and what they do, it&rsquos important to look at them in a context where you can easily test their function. In their new study, Perez-Cervantes and Smith use genetic tools to identify and look at enhancers in rods and cones, the two major cell types that provide vision. Rods are cells which provide vision in low light scenarios, so you can see at night, for example. Conversely, cone cells provide vision in bright light, such as a sunny day outside, and also color vision. In mice and humans, cone cells make up a much smaller number of cells in the mature eye, which has made it hard to study and understand how they control genes differently from rods.

Moskowitz, one of the senior authors on the new study published in the journal Development, wanted to look at how enhancers change in activity when one cell type is changed into another cell type. They decided to do this study in the mouse retina by comparing the non-coding RNA produced from enhancer regions in a rod versus a cone cell. Since the mouse retina is mostly rod cells, the researchers needed a way to study cone cells easily too. Working with collaborators Joe Corbo, MD, PhD (Washington University in St. Louis) and Connie Cepko, PhD (Harvard University) they could study a mouse with mostly cone cells by deleting a single transcription factor, Nrl. By sequencing and comparing the non-coding RNAs from retinas composed of mostly rods or mostly cones, the researchers could look for enhancers that were active in either rod or cone cells and hope to find regions that controlled one cell type to develop over the other.

Identifying noncoding RNAs was not enough. The researchers wanted to know if the non-coding RNAs they identified reflected the actual activity of the DNA regions in the retina itself. Using retinas from developing mice, they looked at whether the potential enhancers were active in either rod or cone cells. To do this, they inserted the potential enhancer in front of a marker gene that produces a protein that generates fluorescent light when imaged on a microscope. From starting with thousands of possible regions, the researchers used non-coding RNAs to find the enhancers that accurately showed cell type specific patterns: enhancers with non-coding RNAs made in rods were found to be active in rods rather than cones while those with non-coding RNAs made in cones were shown to be active in cones.

&ldquoThe non-coding RNA approach is providing unexpected insight,&rdquo Moskowitz said. &ldquoThe ability of this approach to predict enhancer function is providing opportunities for new studies. For example, we are currently examining how new enhancers turn on during arrhythmias in the heart and during diabetes in the pancreas. It&rsquos a very exciting time for the lab.&rdquo

In addition to the work being done by Moskowitz and his lab looking at non-coding RNAs and enhancers with this method, this research can help scientists to use the non-coding RNA identification method for finding enhancers in many different cell types. This may help advance how we understand the development of the eyes, and many other biological systems.

&ldquoThere&rsquos a lot of things you can learn from the noncoding transcriptome,&rdquo Perez-Cervantes, a bioinformatician in the Department of Pediatrics, said. &ldquoThis method really hones in on the functional drivers of cell type specific gene expression. I can see this being applied to broad areas of human genetics in the future.&rdquo

Plus One Zoology Chapter Wise Previous Questions Chapter 10 Neural Control and Coordination

Question 1.
Complete the given flowchart: (MARCH-2010)

a) Midbrain
b) Thalamus
c) Medulla
d) Cerebellum

Question 2.
You may have an experience of sudden withdrawal of body parts when you come in contact with objects that are extremely cold. This response occurs involuntarily without conscious efforts. (MARCH-2010)
a) Name the process behind this.
b) Construct a flowchart showing the pathway of impulses during this process.
a) Reflex action
b) Receptor → Sensory neuron → Inter neuron → Motor neuron → effector organ

Question 3.
Observe the diagram: (MARCH-2010)

a) Label the parts A and B.
b) Even though concentrated HCI is stored in the stomach, it will not generally damage the stomach wall. Why ?
a) A. Cardiac stomach
B. Pyloric stomach
b) Mucus and bicarbonates present in gastric juice protect the stomach wall.

Question 4.
Study the diagram. (SAY-2010)

Figure a and b given above are two axon terminals with synapse. Which one is conducting the impulse? Justify your answer.
Diagram a, The axon terminals contain vesicles filled with neurotransmitters it is not present in diagram b.

Question 5.
Incus, stapes and malleus are the three ear ossicles of the middle ear. Arrange them in the order in which they are present from tympanic membrane to the oval window of the cochlea. (SAY-2010)
Explain their function.
Malleus, incus and stapes
The ear ossicles increase the efficiency of transmission of sound waves to the inner ear.

Question 6.
Two types of synapses are given in the diagrams A and B. (SAY-2011)
a) Identify A and B.
b) Impulse transmission in ‘B’ is faster than that of ‘A’. Give reason.
c) Name the chemical substance that helps in the transmission of impulses in ‘A’.
a) A chemical synapse B- electrical synapse
b) Transmission of an impulse across electrical synapses is very similar to impulse conduction along a single axon. Hence impulse transmission across an electrical synapse is always faster than that across a chemical synapse.
c) neurotransmitter

Question 7.
The following diagram shows the characteristics of a phylum. (MARCH-2012)

a) Identify the phylum.
b) Label A in the diagram.
c) Mention foursalient features of the phylum.
a) Phylum chordata
b) Notochord
c) 1. Presence of Notochord
2. Dorsal hollow nerve chord
3. Pharyngeal gill stits
4. Post anal tail

Question 8.
Analyse the concept may given below and fill the gaps appropriately so as to explain the concept of brain. (SAY-2012)
a) Mid brain
b) Hindbrain
c) Inter sensory association
d) Memory & communication
e) Cardiovascular reflex
f) Gastric secretion

Question 9.
Arrange the following processes in nerve impulse con-duction in a sequential order. (MARCH-2013)
a) Bursting of synaptic vesicle
b) Development of action potential
c) Na+-K+ pump starts functioning
d) Stimulus received and influx of Na+ ions
e) Binding of neurotransmitter with postsynaptic membrane
f → d → c → b → a → e

Question 10.
Nerve impulse transmission involves. (SAY-2013)
Maintenance of resting potential
Development of action potential
Propagation of action potential
a) Diagrammatically represent the polarised and depolarised state of axon of a neuron.
b) Describe how the resting potential of a neuron is maintained.
c) “Electrical currents fade as they pass along a wire but nerve impulses do not fade as they pass along neurons”. Evaluate the statement and sub-stantiate your answer.


b) These ionic gradients are maintained by the active . transport of ions by the sodium-potassium pump which transports 3 Na+ outwards for 2 K+ into the cell . This helps to maintain the resting potential..
c) Electric current fade due to resistance of conductor. But nerve fibre do not have resistance. So the impulses pass without fade. Myelil sheath also helps to prevent the loss of charges.

Question 11.
Given below are the stages in the generation of optic nerve impulse or action potential on the retina and the role of opsin and retinal in the mechanism of vision. Arrange them in a sequential order. (SAY-2014)
a) Action potential (impulses) are transmitted by the optic nerves to the visual cortex area of the brain.
b) Light induces dissociation of retinal from opsin,
c) Generates action potential in the ganglion cells through bipolar cells.
d) Structural changes in the opsin which induce membrane permeability changes.
e) Potential differences are generated in the photoreceptor cells.
f) Neural impulses are analyzed by visual cortex area of the brain..
b → d → e → c → a → f

Question 12.
a) Prepare a pathway of an action by using the following hint. (MARCH-2015)
(Hint – Receptor, Motor neuron, Afferent neuron, Efferent neuron, Intemeuron in the spinal cord, Effector organ.)
b) Give an example of such an action.
Compare rods and cones of the retina based on the following features. ,
i) Shape
ii) Type
iii) Ability to detect colour
iv) Pigments
v) Vision
a) Receptor →Afferent neuron → lnter neuron in spinal cord → Motor neuron → Efferent neuron → Effector organ.
b) The stimulus and response forms a reflex arc in the knee jerk reflex

Rod Cone
shape rod shaped cone shaped
Type one type Green cone, red cone and blue cone
Ability to detect colour No Yes
Pigments Rhodopsin lodopsin
Vision Dim light vision Colourvision

Question 13.
Mention the functions of the following structures in human body. (SAY-2015)
(Hint: Any two each)
a) Hypothalamus
b) Axon
a) Hypothalamus- Gonadotrophin releasing hormone (GnRH) stimulates the pituitary synthesis and release of gonadotrophins. Somatostatin from the hypothalamus inhibits the release of growth hormone from the pituitary
b) Axon – The synaptic knob of axon produce neurotransmitters that helps in impulse transmission

Question 14.
Observe the diagram carefully and answer the following questions. (MARCH-2016)
a) Label the parts marked as A, B, C, D.
b) Identify the photoreceptor cells present in human eye.

a) A-lens, B-lris, c-Cornea. d-Optic nerve
b) Rodes and Cones

Question 15.
Fovea of retina in eye contains…………… (SAY-2016)
a) rod cells only
b) cone cells only
c) both roads and cones
d) rod and cones are absent
b) cone cells only

Question 16.
Complete the given table (parts of human brain) (SAY-2016)

Fore brain Hind brian
Parts Parts
i)Cerebrum i) Pons
ii) Thalamus ii) ………………………..
iii) …………………….. iii) Medulla

a) Which one of the above parts of brain that controls gastric secretions?
a) ii – cerebellum
iii – hypothalamus
b) Medulla

Question 17.
Answer the following: (MARCH-2017)
a) Cerebral hemispheres of human brain are connected by
i) association area
ii) corpus callosum
iii) corpora quadrigemina
iv) pons varolii
Observe the diagram and label A, B, C and D.

a) ii) Corpus callosum
b) A – Axon B – Synaptic vesicle
C – Synaptic cleft D – Neuro Transmitters

The Art of Seeing

2.1 The Retina

There are two types of photoreceptors in the retina: cones and rods ( Fig. 4.3 ). Cones are color selective, less sensitive to dim light than rods, and important for detailed color vision in daylight. Each cone contains one of the three kinds of photopigments, specialized proteins that are sensitive to different wavelengths of light. These wavelengths roughly correspond to our ability to distinguish red, green, and blue. When light strikes a photopigment molecule, the light energy is absorbed and the molecule then changes shape in a way that modifies the flow of electrical current in that photoreceptor neuron. Cones are densely packed into the fovea, the central part of the retina that we use to look directly at objects to perceive their fine details. In the periphery, cones are more spread out and scattered, which is why objects in the periphery appear blurrier and their colors are less vivid.

Figure 4.3 . The eye. (A) There are two types of photoreceptors in the retina: cones and rods. Cones (labeled C) are color selective, less sensitive to dim light than rods (labeled R), and important for detailed color vision in daylight. Other cells in the retina include the horizontal cell (labeled H), the flat midget bipolar (labeled FMB), invaginating midget bipolar (labeled IMB), invaginating diffuse bipolar (labeled IDB), rod bipolar (labeled RB), amacrine cell (labeled A), parasol cell (labeled P), and midget ganglion cell (labeled MG). (B) Cones are densely packed into the fovea, the central part of the retina that we use to look directly at objects to perceive their fine details. In the periphery, cones are more spread out and scattered, which is why objects in the periphery appear blurrier, and their colors are less vivid.

Source: Reid and Usrey in Squire et al., 2013.

Rods contain a different photopigment that is much more sensitive to low levels of light. Rods are important for night vision. We rely on seeing with our rods once our eyes have adapted to the darkness (dark adaptation). Curiously, there are no rods in the fovea, only cones, and the proportion of rods increases in the periphery. This is why you may have noticed when gazing at the night sky that a very faint star may be easier to see if you look slightly off to one side.

There are far more rods in the retina than cones, with roughly 120 million rods distributed throughout the retina except for the fovea, and 6–7 million cones that are concentrated in that fovea.

The signals from photoreceptors are processed by a collection of intermediary neurons, bipolar cells, horizontal cells, and amacrine cells, before they reach the ganglion cells, the final processing stage in the retina before the signals leave the eye. The actual cell bodies of ganglion cells are located in the retina, but these cells have long axons that leave the retina at the blind spot and form the optic nerve. Each ganglion cell receives excitatory inputs from a collection of rods and cones this distillation of information forms a receptive field. Ganglion cells at the fovea receive information from only a small number of cones, whereas ganglion cells in the periphery receive inputs from many rods (sometimes thousands). With so many rods providing converging input to a single ganglion cell, if any one of these rods is activated by photons of light, this may activate the ganglion cell, which increases the likelihood of being able to detect dim, scattered light. However, this increase in sensitivity to dim light is achieved at the cost of poorer resolution rods provide not only more sensitivity but also a “blurrier” picture than the sharp daytime image provided by cone vision.

Retinal ganglion cells receive both excitatory and inhibitory inputs from bipolar neurons, and the spatial pattern of these inputs determines the cell's receptive field ( Fig. 4.4A ). A neuron's receptive field refers to the portion of the visual field that can activate or strongly inhibit the response of that cell. Retinal ganglion neurons have center-surround receptive fields. For example, a cell with an on-center off-surround receptive field will respond strongly if a spot of light is presented at the center of the receptive field. As that spot of light is enlarged, responses will increase up to the point where light begins to spread beyond the boundaries of the on-center region. After that, the response of the ganglion cell starts to decline as the spot of light gets bigger and stimulates more and more of the surrounding off-region. Similarly, a cell with an off-center on-surround receptive field will respond best to a dark spot presented in the center of the receptive field.

Figure 4.4 . Center-surround receptive fields. (A) Schematic example of a center-surround cell's response to different-sized patches of light. Notice that the biggest spiking response (shown by the lines on the right) occurs for the intermediate-sized center light patch. The spot of light has to be just the right size to get the maximum response out of that particular neuron. (B) A model of how a center-surround receptive field might be achieved by the collaboration and competition between different connective neurons in the retina.

Source: Frank Tong, with permission.

How can the behavior of retinal ganglion cells be understood? A key concept is that of lateral inhibition ( Kuffler, 1953 ). Lateral inhibition means that the activity of a neuron may be inhibited by inputs coming from neurons that respond to neighboring regions of the visual field. For example, the retinal ganglion cell in Fig. 4.4B receives excitatory inputs from cells corresponding to the on-center region and inhibitory inputs from the off-center region. The strengths of these excitatory and inhibitory inputs are usually balanced, so if uniform light is presented across both on- and off-regions, the neuron will not respond to uniform illumination.

Lateral inhibition is important for enhancing the neural representation of edges, regions of an image where the light intensity sharply changes. These sudden changes indicate the presence of possible contours, features, shapes, or objects in any visual scene, whereas uniform parts of a picture are not particularly informative or interesting. Fig. 4.5 shows a picture of a fox in original form and after using a computer to filter out just the edges (right picture) so that the regions in black show where ganglion cells would respond most strongly to the image. Lateral inhibition also leads to more efficient neural representation because only the neurons corresponding to the edge of a stimulus will fire strongly other neurons with receptive fields that lie in a uniform region do not. Because the firing of neurons takes a lot of metabolic energy, this is much more efficient. This is an example of efficient neural coding only a small number of neurons need to be active at any time to represent a particular visual stimulus.

Figure 4.5 . The edges hold most information. An example of how most of the information in the picture comes from the edges of objects. Figure on the left is the original, on the right is the information from the edges only—taken from the image using a computer algorithm.

Source: Frank Tong, with permission.

Lateral inhibition also helps to ensure that the brain responds in a similar way to an object or a visual scene on a cloudy day and on a sunny day. Changes in the absolute level of brightness will not affect the pattern of activity on the retina very much at all it is the relative brightness of objects that matters most. An example of this is you see a friend wearing a red shirt. The absolute level of brightness of that shirt when you see your friend outside your house on a sunny day versus inside your house in a sheltered room will differ, but this will not affect the pattern of activity on the retina. On the other hand, the relative brightness of the shirt compared with other nearby objects or the background scene will make a difference on the retinal activity. Finally, lateral inhibition at multiple levels of visual processing, including the retina, lateral geniculate nucleus (LGN), and visual cortex, may lead to interesting visual illusions such as the Hermann grid illusion ( Fig. 4.6 ). We will discuss more about this type of illusion in Section 3 .

Figure 4.6 . Hermann grid illusion. Take a careful look at the collection of black squares in the figure. Do you notice anything unusual? Do you have the impression of seeing small dark circles in between the black squares in the periphery? Do not be alarmed, this is completely normal. This is a great example of receptive fields with lateral inhibition at work ( Herman, 1870 ). In the rightmost matrix of squares, some possible receptive fields are shown. A receptive field that falls between the corners of four dark squares will have more of its inhibitory surround stimulated by the white parts of the grid than a receptive field that lies between just two of the dark squares. As a result, neurons with receptive fields positioned between four dark squares will fire more weakly, leading to the impression of small dark patches at these cross points. At the fovea, receptive fields are much smaller so the illusion is only seen in the periphery.

Source: Frank Tong, with permission.

Transduction of Light

The rods and cones are the site of transduction of light to a neural signal. Both rods and cones contain photopigments. In vertebrates, the main photopigment, rhodopsin , has two main parts Figure (PageIndex<4>)): an opsin, which is a membrane protein (in the form of a cluster of &alpha-helices that span the membrane), and retinal&mdasha molecule that absorbs light.

Figure (PageIndex<4>): (a) Rhodopsin, the photoreceptor in vertebrates, has two parts: the trans-membrane protein opsin, and retinal. When light strikes retinal, it changes shape from (b) a cis to a trans form. The signal is passed to a G-protein called transducin, triggering a series of downstream events.

When light hits a photoreceptor, it causes a shape change in the retinal, altering its structure from a bent (cis) form of the molecule to its linear (trans) isomer. This isomerization of retinal activates the rhodopsin, starting a cascade of events that ends with the closing of Na + channels in the membrane of the photoreceptor. Thus, unlike most other sensory neurons (which become depolarized by exposure to a stimulus) visual receptors become hyperpolarized and thus driven away from threshold (Figure (PageIndex<5>)).

Figure (PageIndex<5>): When light strikes rhodopsin, the G-protein transducin is activated, which in turn activates phosphodiesterase. Phosphodiesterase converts cGMP to GMP, thereby closing sodium channels. As a result, the membrane becomes hyperpolarized. The hyperpolarized membrane does not release glutamate to the bipolar cell.

Illustration of the rods and cones in the eye

That is much more than the rods and cones. That is an awesome cross-section of the retina including optic nerve fibers, amacrine cells, horizontal cells, ganglion cells etc. I could have used this picture two days ago in my class!

SAME! I just linked this on my course's Moodle page. Quite an excellent image.

Can someone walk me through what's what in this image? Like maybe starting from the red cells on the far left?

We had to learn the histology of the eye in my Anatomy final. I still have my flashcards. Here's my sketch and labeling of the layers of the retina.

This was my accompanying text. It's just notes, and it might not all make sense to you, but perhaps it helps!

3 concentric layers:
I. Fibrous tunic: cornea and sclera,
II. Vascular tunic: iris, ciliary body, choroid,
III. Neural tunic: retina: photosensitive and nonphotosenitive

III. Anterior part of non-photosensitive retina: 2 parts:

pars iridica retinae on posterior surface of eyeball. Has 2 epithelial layers: outer less pigmented comes from outer wall of optic cup → differentiate into myoepithelium of dilator pupillae. Inner heavy pigmented.

pars ciliaris retinae lines internal surface of ciliary body: outer pigmented and inner non-pigmented epithelial layers (inner = continuation of neural retina)
- photosensitive retina: begins at ora serrate until papilla of optic nerve on choroid layer
o 10 layers: (except at fovea and papilla)
1. pigmented epithelium on Bruch’s membrane simple cuboidal retinal-blood barrier
2. Rods and cones: perpendicular to plane make up regular striation
3. Outer limiting membrane: thin eosinophillic line not membrane → is junction of photoreceptor cellsand
4. Müller cells (supporting glial cells)
5. outer nuclear layer: perikarya of rod and cone photoreceptor cells densely packed, very basophilic
6. outer plexiform layer: lightly stained has axons of rods and cones, dendrites and axons of bipolar and horizontal cells
7. inner nuclear layer: all bodies of horizontal, bipolar and amacrine and ganglion cells – densely packed, very basophil
8. inner plexiform layer: synapses of bipolar and amacrine cells with ganglion cells
9. ganglion cell layer: nuclei of ganglion (some amacrine) cells
10. nerve fiber layer: unmyelinated axons of retinal ganglion cells → converge to optic disk → optic nerve
11. Inner limiting membrane: end feet of Müller cells = basal lamina to separate from vitreous body


Differentiation of patient and control retinal organoids

Fibroblasts from a patient with a homozygous null mutation, a frame shift and premature stop (c.223dupC, p.L75Pfs*19) in the NRL gene, were reprogrammed to hiPSCs 32 . Three independent, karyotypically normal L75Pfs clones (Supplementary Fig. 1) were compared with WA09 and two wildtype (WT) hiPSC lines (1013 and 1581) 19 . All lines were indistinguishable in their ability to make stage 1 organoids, characterized by a phase bright, neuroepithelial appearance and the presence of ganglion cells and proliferative retinal progenitor cells (Supplementary Fig. 2) 19 . CRX+/RCVN+ photoreceptor precursor production was comparable between early stage 2 WT and L75Pfs organoids, when photoreceptor subtype specification begins (Fig. 1 compare Fig. 1b, c, merge in e, to g, l, q, h, m, r, merges in j, o and t) 19 . However, NRL+ rod photoreceptors were never detected in L75Pfs organoids (compare Fig. 1d to i, n, s). As photoreceptors matured and formed outer segments (the “hair-like” surface projections in Fig. 2c, i), L75Pfs organoids showed a striking S-opsin dominant photoreceptor phenotype (Fig. 2) 19 . Unlike WT organoids, which possess a single layer of ML-opsin+ cones and rare S-opsin+ cones along the outermost aspect of the outer nuclear layer (ONL) (Fig. 2a), L75Pfs organoids contained S-opsin expressing cells throughout the ONL and in greater abundance than ML-opsin expressing cells (Fig. 2b also compare Fig. 2e, g to k, m). Quantification of ARR3+ cones and NR2E3+ rods as a percent of total nuclei in the ONL (Fig. 2d–h, j–n) revealed a dramatic reduction in rods and an increase in cones in the L75Pfs ONL (Fig. 2o). Interestingly, while WT organoids had rare ARR3-/NR2E3- nuclei in the ONL,

20% of the L75Pfs ONL nuclei expressed neither marker. Since ARR3 is normally expressed >60 days after cone progenitors are detected, these ARR3-/NR2E3- cells may represent rod progenitor-derived cells that either have not committed to a cone fate or do not yet express ARR3. We quantified the ML- or S-opsin expressing cells as a fraction of the total ARR3+ cells and detected a 38-fold shift in the ML:S-opsin cone ratio, from 19:1 in WT to 1:2 in L75Pfs organoids (Fig. 2p). Additional analyses of rod and cone gene expression by RT-qPCR revealed that rod developmental genes were downregulated in L75Pfs organoids relative to WT organoids, while S-opsin expression was significantly increased in the L75Pfs organoids (Fig. 2q, p < 0.005, Mann–Whitney test). Thus, in L75Pfs human retinal organoids, rods appeared shifted toward an S-cone fate, consistent with the Nrl−/− mouse phenotype 3,33 . Finally, we examined the inner nuclear layer (INL) of L75Pfs organoids and found it indistinguishable from WT organoids (Supplementary Fig. 3a–n), including comparable production of PKCα+ rod bipolar cells (consistent with the Nrl−/− mouse phenotype) 34 . However, in contrast to the Nrl−/− mouse, L75Pfs organoids displayed an intact outer limiting membrane (OLM) with no increase in rosette formation compared to WT organoids (Supplementary Fig. 3o–y) 3,35 .

at Confocal images of d100 (stage 2) organoids from a WT line (ae) or three individual clonal lines of the L75Pfs mutant (ft) showing photoreceptors immunostained for RCVN (b, g, l, q), CRX (c, h, m, r), or NRL (d, i, n, s). a, f, k, p: nuclei (blue) e, j, o merge in t: merge). Scale bars = 50 μm.

a, b Confocal images from stage 3 organoids (i.e., presence of photoreceptor outer segments) showing a single layer of cones with few S-cones (green) in WT organoids (a) versus an abundance of S-cones (green) distributed throughout the ONL in L75Pfs organoids (b). ML-cones are shown in orange. Scale bars = 25 μm. cn Photoreceptor characterization of WT and L75Pfs retinal organoids. Bright field (c, i) and confocal (dh and jn) images showing S-opsin+/ARR3+ cones (k, l) distributed throughout the ONL of L75P(fs) organoids that do not express the rod marker NR2E3 (m) (a transcription factor whose expression is controlled by NRL). This finding is in contrast to WT organoids that display ordered expression of cones (e, f) along the outermost ONL with a multicellular layer of NR2E3+ rod nuclei (g) internal to the cone layer, as well as an overall low number of S-opsin+ (e) cones. Scale bars: c, I = 250 microns dh and jn = 25 μm. o, p Quantification of photoreceptors in confocal images of stage 3 organoids from 3 WT lines and 3 L75Pfs clones. o, p Quantification of photoreceptors in confocal images of stage 3 organoids from 3 WT lines and 3 L75Pfs clones. o NR2E3+ rod and ARR3 + cone abundance as a percentage of total nuclei in the ONL: 15 images from 5 organoids per line or clone were counted. p ML- and S-cone abundance as a percentage of total ARR3+ cones in the ONL: 11 WT images from 4 organoids per line and 15 L75Pfs images from 5 organoids per clone were counted. q RT-qPCR from stage 2–3 organoids showing a reduction in rod transcripts and an increase in S-opsin transcripts in L75Pfs organoids relative to WT organoids. p < 0.005, Mann–Whitney test.

Reintroduction of WT NRL restores rod formation

To confirm that the observed L75Pfs phenotypes were indeed due to lack of NRL, we introduced functional NRL to determine whether NRL expression could rescue the phenotype. We ectopically expressed WT NRL in d90 L75Pfs retinal organoids using a lentivirus expression cassette. Of note, d90 corresponds to the onset of NRL protein detection in WT organoids 19 . After 100 additional days in culture, organoids transduced with virus containing either a control GFP expression cassette (without NRL) or a WT NRL expression cassette were examined by immunocytochemistry for NRL, rhodopsin (RHO), and S-opsin expression. Figure 3a–d shows co-expression of GFP in ARR3+ cones of control cassette-treated L75Pfs NRL organoids, which remained NRL-/RHO- and expressed S-opsin throughout the ONL (Fig. 3e–l), similar to untreated L75Pfs organoids. In contrast, WT NRL expression cassette-treated organoids showed NRL protein in patches of nuclei within the ONL (Fig. 3m–t). Furthermore, all cells with restored NRL expression did not express S-opsin (Fig. 3m–t). Additionally, rare RHO+ cells (Fig. 3q, w), which were never observed in untreated or pgkGFP-transduced (Fig. 3i–l) L75Pfs organoids, were observed and were uniformly negative for S-opsin (Fig. 3t) and ARR3 (Fig. 3x). Of note, the localization of RHO to outer segments in some lenti-pgkNRL transduced cells (Fig. 3w) is reminiscent of RHO immunostaining in WT organoids (Supplementary Fig. 4a–d). Thus, restoring NRL protein expression to L75Pfs photoreceptor precursor cells restricted S-opsin expression and could promote, although at low efficiency, RHO expression.

ah Confocal images from L75Pfs organoids transduced with a control lentivirus carrying a pgkGFP expression cassette. GFP (c) was found in ARR3+ (b) cones NRL (g) was not detected, and S-opsin+ cones (f) were localized throughout the ONL and no RHO (k) expression was detected, as expected in the absence of ectopic WT NRL expression. mx Confocal images from L75Pfs organoids transduced with lentivirus carrying a pgkNRL expression cassette reveal patches of NRL expression within the ONL (o and s) and no NRL co-expression with S-opsin (n and r merges in p and t). However, ectopically expressed NRL does co-express with RHO (q merge in t), a rod marker that was never present in control transduced L75Pfs organoids. In pgkNRL-transduced organoids, RHO (w) did not co-localize with the cone marker ARR3 (v merge in x). Scale bars = 25 μm.

ScRNAseq to identify and analyze organoid cell types

After establishing that the observed L75Pfs phenotype was due to NRL loss, we sought to identify and transcriptomically analyze the cell populations in WT and L75Pfs retinal organoids. We performed scRNAseq via the Dropseq platform on WT and L75Pfs organoids differentiated to early (100–103 days) or late (170 days) stage 2 19,21 . At the earlier time, 4 WT and 4 L75Pfs organoids yielded transcriptional profiles of 5294 and 4787 cells, respectively. Cells were clustered by t-distributed stochastic neighbor embedding (tSNE) using the Seurat R package 36 . The even distribution of cells classified either by number of genes expressed or number of unique molecular identifiers (UMIs) throughout the clusters confirmed that these factors were not driving clustering (Supplementary Fig. 5). Rather, based on known marker genes (Supplementary Table 1, Supplementary Fig. 6), the clusters represent stereotypical retinal populations present in both WT and L75Pfs organoids (Fig. 4a). Both rod and cone photoreceptors were present, with almost all NR2E3 expressing cells being WT (Fig. 4b). Spearman correlations were performed between WT cells of each population and published fetal and adult retinal scRNAseq datasets (Supplementary Figs. 7 and 8) 23,24 . This analysis revealed d100 organoids yielded amacrine, horizontal, and retinal ganglion cells more similar to fetal retinal populations. Rods most closely resembled adult peripheral rods, while cones and Müller glia more closely resembled adult foveal cells. Differential gene expression tests were performed between WT and L75Pfs cells of each cluster, and genes with significantly different expression and an average natural log fold change greater than 0.5 (

1.6 fold) are summarized in Supplementary Data 1. Of genes enriched in L75Pfs cells of the rod cluster, the presence of the cone transducin, GNGT2, indicates that this population is acquiring a cone-like profile.

a tSNE plot showing cell populations present in all d100 organoids. b Expression of NR2E3 in WT (top) compared to L75Pfs (bottom) cells, showing that nearly all expression is in WT cells. c tSNE plot showing cell populations present in all d170 organoids. d Scatter plot of OPN1SW and OPN1MW expression at d170 indicating six co-expressing cells. e Scatter plot showing the number of UMIs and genes expressed by the six co-expressing cells from D. f Violin plots showing specific enrichment of novel cone marker genes across WT ML-cones, ML/S-cones, and S-cones. g, h Heatmap of genes differentially expressed in d170 rod (g) and S-cone (h) clusters between WT and L75Pfs cells. i Comparison of expression of OPN1SW, OPN1MW, and NR2E3 by cell population of WT and L75Pfs organoids, and total expression of OPN1SW, OPN1MW, and NR2E3 within all WT and L75Pfs cells.

At d170, 3 WT, and 6 L75Pfs organoids yielded 8920 and 15,447 single-cell transcriptomes, respectively. Cell populations identified by marker gene expression (Supplementary Table 1) showed that mature retinal cells were captured, including bipolar cells and opsin-expressing photoreceptors (Fig. 4c, Supplementary Fig. 9). Loss of retinal ganglion cells was also observed. Possible explanations for retinal ganglion cell loss include microfluidic bias favoring other cell types or death of retinal ganglion cells due to the lack of vasculature in retinal organoids. Notably, age-dependent retinal ganglion cell loss has been reported in retinal organoids 19 . Again, Spearman correlations were performed between WT cells of each population and published fetal and adult retinal scRNAseq datasets (Supplementary Figs. 10 and 11) 23,24 . Like d100, organoids at d170 yielded amacrine and horizontal cells more similar to fetal cells. Rods and bipolar cells more closely resembled adult peripheral cells, and, similarly to d100, cones and Müller glia were more highly correlated with adult foveal cells. Two cone opsin-expressing populations were identified, one that expressed both ML-opsin and S-opsin and the other consisting of cells that primarily express S-opsin. Due to the identity of the M- and L-opsin 3′ UTRs that were captured via our analysis, we could not distinguish between M- and L-opsin transcripts. Interestingly, 1.5% of WT cone opsin-expressing cells co-expressed both ML- and S-opsin (Fig. 4d). The number of UMIs and genes expressed by these cells suggests they are not doublets (Fig. 4e).

We performed differential gene expression analysis on these 3 cone opsin-expressing populations (ML-, ML/S-, or S- expressing) to identify novel markers of developing cone subtypes. As expected, there were only minimal differences in gene expression between the cone populations. We identified MYL4 as a possible ML-cone marker and CCDC136 and DCT as possible S-cone markers (Fig. 4f). CCDC136 is preferentially expressed in mouse S- and S/M-cones and recently has been shown to be enriched in primate S-cones 37,38 . Interestingly, Peng et al. identified MYH4, the heavy chain complement to MYL4, as a transcript distinguishing ML-cones from S-cones 38 . NUP93, SLC12A6, PDRG1, and TRAPP2CL were significantly enriched in the ML/S-cone population compared to the ML or S- expressing cell populations. Further studies are necessary to determine if this ML/S-co-expressing cone population exists in vivo.

To determine altered transcriptional profiles at d170, within each cell population we identified differentially expressed genes with an average natural log fold change greater than 0.5 (

1.6-fold) between WT and L75Pfs cells (Supplementary Data 2). Within the rod cluster, WT cells had significantly higher expression of many rod-specific genes including GNAT1, ROM1, SAMD7, NR2E3, CNGB1, GNB1, and PDE6G, while cone-specific phosphodiesterase, PDE6H (Fig. 4g), was enriched in L75Pfs cells, suggesting a more cone-like character of these cells. Despite loss of NRL protein, the NRL transcript is still detectable in L75Pfs cells, possibly due to the presence of transcripts that have yet to be removed by nonsense mediated decay. Of differentially expressed genes in the S-opsin expressing population (Fig. 4h), L75Pfs S-opsin expressing photoreceptors were enriched for OPN1SW and GNGT1, a rod-enriched transducin (Fig. 4g). Despite enrichment of MYL4 in WT compared to L75Pfs S-cones, this gene exhibited substantially higher expression in ML-cones compared to S-cones, supporting its designation as enriched in ML-cones (Fig. 4f). To identify whether NRL loss alters the distribution of photoreceptor subtypes, we compared expression of OPN1SW, OPN1MW, and NR2E3 within each cell population of d170 WT and L75Pfs organoids, as well as total expression of these genes across both genotypes (Fig. 4i). OPN1SW was detected in more cells and more clusters in L75Pfs organoids compared to WT, while OPN1MW was primarily detected in only the cone cluster of both the WT and L75Pfs organoids. Since NRL activates NR2E3 transcription, L75Pfs organoids had (as expected) lower expression of NR2E3 and fewer NR2E3 expressing cells compared to WT. Quantification of OPN1SW, OPN1MW, and NR2E3 expression levels and percentage of expressing cells in d170 organoids can be found in Table 1. Although WT and L75Pfs organoids had significantly different relative numbers of cells expressing OPN1SW vs OPN1MW, on an individual cell basis the OPN1SW and OPN1MW expressing cells expressed comparable levels of OPN1SW and OPN1MW. However, for NR2E3, both the percentage of expressing cells and the expression level within individual expressing cells was significantly lower in L75Pfs cells. Taken together, this data suggests that NRL loss has a profound effect on rod development, shifting them towards an S-cone identity.

Trajectory reconstruction of WT photoreceptor development

After identifying retinal populations, we used the 5144 WT photoreceptors identified from both time points to create a pseudotemporal trajectory of WT photoreceptor development 39 . While the population contained contaminating bipolar cell precursors (44/5144 cells with VSX1 or VSX2 expression), this small population is unlikely to impact trajectory construction (we could have removed these cells, but felt that selectively removing small subpopulations of cells was more likely to produce an artifact than leaving them in). To determine the gene set for ordering the trajectory, we performed a semi-supervised differential gene expression test for genes varying by age and assigned cell type within the WT photoreceptor dataset. After removing mitochondrial and ribosomal genes, the top 780 genes by p-value were used for ordering (Supplementary Data 3). Importantly, neither VSX1 nor VSX2 were present in this list, verifying that the contaminating bipolar cells did not affect the trajectory reconstruction. The resulting WT trajectory had one node separating rod and cone photoreceptors, with OPN1SW or OPN1MW expressing cells in state 2 and NR2E3/SAG expressing cells in state 3 (Fig. 5a–d). Five hundred and ninety genes were significantly differentially expressed at this node and the top 100 non-ribosomal genes were used to create a heatmap of genes enriched along the rod versus cone branches (Supplementary Data 4, Fig. 5e). While many of these genes are known as rod- or cone-specific, we identified some novel cone- or rod-enriched genes. In addition to MPP4 and CC2D2A, genes already shown to be enriched in human fetal cones, we identified GNAI3, CA2, MAP4, MYL4, MCF2, KIF2A, and KIF21A as cone-enriched, and PTPRZ1, CABP5, IRX6, B2M, and PRUNE2 as rod-enriched (Supplementary Fig. 12) 40 . We checked published adult human scRNAseq data and confirmed significant enrichment of MAP4, MYL4, MCF2, and KIF2A in cones, and CABP5 and IRX6 in rods (p < 0.05, one-sided T-test) 24 . We performed similar trajectory analyses using adult photoreceptor data to compare the organoid trajectory to in vivo photoreceptors (Supplementary Fig. 13) 24 . The resulting trajectory separated rods and cones, and the top 100 non-ribosomal differentially expressed genes between the state 1 rods and state 4 cones were used to create a heatmap for comparison to organoid development. Thirty genes were differentially expressed in both datasets, with all but MALAT1, TMSB10, and EEF1A1 exhibiting the same enrichment pattern. The expression patterns of known markers confirm that our trajectory accurately recapitulates photoreceptor development and validate its utility for analysis of perturbations occurring in NRL null photoreceptors.

ac Trajectory of 5144 WT photoreceptors colored by state (a), pseudotime (b), and age (c). d Expression of photoreceptor markers used to distinguish the identity of each branch of the trajectory. e Heatmap of the top 100 non-ribosomal differentially expressed genes at the node separating rods and cones. Genes are hierarchically clustered into four clusters based on expression pattern. The center of the heatmap is the beginning of pseudotime, with cell maturity moving horizontally to the left (cones) and right (rods).

Reconstruction of the combined WT and L75Pfs trajectory

After creating a WT trajectory that accurately represented photoreceptor development, we applied the same parameters to create a trajectory of 13,317 combined WT and L75Pfs photoreceptors to elucidate the shift in development resulting from the absence of NRL (Fig. 6a–d). Again, there were few contaminating bipolar cell precursors (254/13,317 cells with VSX1 or VSX2 expression) that likely did not affect trajectory reconstruction. The combined trajectory indicated nine cell states, compared to three states of the WT-only trajectory. We sought to characterize these states by their gene expression and gene ontology (GO). After the start of pseudotime with state 1, the first node separated two populations of immature photoreceptors. To characterize the photoreceptors in state 9, we input genes enriched in this state to a GO tool 41 . The enriched GO terms included various cell differentiation and cell stress processes (Supplementary Fig. 14). The remaining portions of the trajectory included developing photoreceptors (states 2, 3, and 4), and four branches corresponding to 2 rod and 2 cone cell fates. States 2, 3, and 4 were defined as developing photoreceptors due to expression of CRX, OTX2, and RCVRN and absence of WT cells with substantial rod or cone gene expression patterns (Supplementary Fig. 15b–d). These states have some cells expressing NRL or ARR3, but the low numbers and level of expression, compared to the more clearly-defined rod and cone populations of states 5, 6, 7, and 8, suggests they are developing photoreceptors (Fig. 6e). Expression of NR2E3 and SAG in WT cells identified states 7 and 8 as rod/rod-like cell fates and OPN1MW expression defined states 5 and 6 as cone fates. To differentiate between cone states 5 and 6, we utilized GO analysis on gene sets enriched in each fate (Supplementary Fig. 15e). State 5 cones were enriched for GO terms relating to electron transport chain and guanylate cyclase activity, whereas state 6 cones had enrichment for retinal development and photoreceptor differentiation terms, suggesting that state 6 cones may be less mature than the more metabolically active cones of state 5. Despite apparent differences in maturity and metabolism, cells in these states were combined into a single cone population for downstream analyses due to confidence in their identity as cone photoreceptors. We performed similar analyses on states 7 and 8 to distinguish between the two rod/rod-like populations (Supplementary Fig. 15f). While both populations were enriched for terms related to photoreceptor activity, state 8 cells appear to have stronger rod profiles due to enrichment in rhodopsin signaling terms. As with the cone states, cells of these two rod states were combined into one rod population for downstream analysis. This analysis identified the L75Pfs photoreceptor populations and their WT counterparts, thus allowing for further comparison of mature L75Pfs photoreceptor populations to WT rods and cones.

ad Trajectory of combined WT and L75Pfs photoreceptors colored by state (a), genotype (b), pseudotime (c), or age (d). e Expression of rod and cone marker genes by state, with L75Pfs cell populations on the left and WT cell populations on the right. WT expression of NR2E3 and SAG was used to assign states 7 and 8 as rods, and expression of OPN1MW to assign states 5 and 6 as cones. f Expression levels of cone markers in states 5 and 6 by genotype, with the position of the largest dot indicating the average level for each marker. All genes are expressed at comparable levels except ARR3, GRK7, OPN1SW, and PDE6H.

Characterization of L75Pfs S-opsin expressing cells

Because previous murine studies described Nrl−/− photoreceptors as possible “cods”, we sought to determine if this would hold for human NRL null photoreceptors 3 . To better characterize the L75Pfs photoreceptors, we determined genes differentially expressed compared to WT cells in each photoreceptor subset. Comparison of genes enriched in either WT or L75Pfs cone states (combined states 5 and 6) revealed 184 genes, few of which were rod- or cone-specific genes (Supplementary Data 5). Importantly, nearly all cone-specific genes showed no significant differential expression between WT and L75Pfs cones (Fig. 6f). Exceptions included ARR3, GRK7, and OPN1SW, which were more highly expressed in L75Pfs cells, and PDE6H, which showed slight enrichment in WT cells. Lower OPN1SW expression in WT cones was expected as ML-cones are the dominant cone subtype in WT organoids (Fig. 2p). The low number of differentially expressed genes and comparable expression levels of most cone genes between WT and L75Pfs cones suggests that this population of L75Pfs cones is essentially normal.

To characterize L75Pfs photoreceptors in the rod/rod-like branches (states 7 and 8), we compared them to both WT rods and WT cones by performing differential gene expression analysis according to the schematic in Fig. 7a (L75Pfs rod-like red cells vs WT rod purple cells L75Pfs rod-like red cells vs WT cone blue cells). L75Pfs photoreceptors of the rod-like states had high OPN1SW expression, while also exhibiting WT rod levels of OPN1MW expression and WT cone levels of NR2E3/SAG expression (Fig. 7b). In comparing L75Pfs and WT rods/rod-like cells, 397 genes were differentially expressed (Supplementary Data 6), with L75Pfs photoreceptors expressing significantly lower levels of the rod-specific genes CNGB1, GNAT1, GNB1, GNGT1, NR2E3, NRL, PDE6G, ROM1, and SAG than their WT counterparts (Fig. 7c). The expression levels of these rod genes in L75Pfs rod-like cells were comparable to their expression levels in WT cones (Supplementary Fig. 16a). The low expression of these genes compared to WT rods is likely due to the loss of NRL. In comparing L75Pfs rod-like cells to WT cones, 791 genes were differentially expressed, including many cone-specific genes (Supplementary Data 7). Compared to WT cones, L75Pfs rod-like cells had significantly lower expression of ARR3, CNGB3, GNAT2, GNB3, GNGT2, GUCA1C, PDE6C, and PDE6H (Fig. 7d). Except for ARR3 and PDE6H, all of these genes were expressed at comparable levels in L75Pfs rod-like cells and WT rods (Supplementary Fig. 16b). Interestingly, L75Pfs rod-like cells also had significantly higher expression of GNGT1, a rod transducin also associated with foveal cones 38 . The high OPN1SW expression, rod levels of expression of cone genes, cone levels of expression of rod genes, and degree of differential gene expression compared to WT rods and cones suggests that L75Pfs rod-like cells are human analogs of “cods”.

a Plot depicting the differential expression analysis. Gene expression in the cells highlighted in red (L75Pfs rod-like cells) was separately compared to expression in the cells highlighted in blue (WT cones) and the cells highlighted in purple (WT rods). b Expression levels of NR2E3, OPN1MW, OPN1SW, and SAG across the three cell groups (red, blue, and purple) used for differential analysis. c, d Expression levels of rod- and cone-specific genes differentially expressed between L75Pfs rod-like cells and WT rods (c), or WT cones (d). The position of the largest dots indicate average gene expression levels for be. Expression level of MEF2C in cells in the cone states (5 and 6) on the left versus the rod states (7 and 8) on the right, with cells colored by genotype (black = L75Pfs, teal = WT). f, g Percentage of MEF2C-regulated genes expressed in a minimum of 10 cells (f) and differentially expressed between the red (L75Pfs rod-like cells) and blue (WT cones) groups (g) versus random genes. h Confocal images demonstrating MEF2C (green) co-expression with ARR3 (purple) in cone photoreceptors of WT (left) and L75Pfs (right) organoids at d160. NR2E3+ nuclei (red) are present in WT but not in L75Pfs organoids. Scale bar = 10 μm.

MEF2C as a candidate regulator of cone cell fate

Because the cod expression pattern is not completely consistent with either rods or cones, the presence of this population suggests that NRL is not the only transcription factor directing rod versus cone photoreceptor fate specification. We postulated that other transcription factor(s) with higher expression in cones than L75Pfs cods could play a previously unappreciated role in regulating cone development. We identified 75 transcription factors of the 791 genes differentially expressed between WT cones and L75Pfs cods, with MEF2C as a potential candidate due to its higher expression level in both WT and L75Pfs cones compared to L75Pfs cods and WT rods (Fig. 7e). Interestingly, MEF2C has been shown to act downstream of Nrl in mice however, in our dataset MEF2C showed significant enrichment in cone photoreceptors, suggesting a potential difference between human and murine photoreceptor fate determination 42 . To determine if MEF2C could be involved in human photoreceptor gene regulation, we queried human retinal ATAC-seq data for regions of open chromatin with MEF2C binding sites 43 . After identifying 475 genes with MEF2C binding sites, we tested for enrichment of MEF2C-regulated genes in our dataset. Three hundred and seventy five genes (80%) were expressed in at least ten cells in this photoreceptor dataset, a significant enrichment compared to the expected 50% for random genes (Fig. 7f). Furthermore, 34 genes (7.2%) were differentially expressed between L75Pfs cods and WT cone cells, a significant enrichment compared to 4.2% expected for random genes (Fig. 7g). This enrichment in expressed genes and differentially expressed genes suggests a possible role for MEF2C in photoreceptor development, and specifically rod versus cone cell fate specification. Consistent with our transcriptome data, MEF2C protein was detected exclusively in ARR3+ photoreceptors in both WT and L75Pfs retinal organoids at d160 (Fig. 7h, Supplementary Fig. 17a–j). MEF2C expression was also detected in d122 human fetal retina within a single-cell layer adjacent to NR2E3+ developing rod nuclei, consistent with cone localization (Supplementary Fig. 17k–n). Strong MEF2C expression was also detected in peripheral adult monkey retina in infrequent cells next to NR2E3+ rod nuclei, also corresponding to cones (Supplementary Fig. 17o–r), and within a line of INL nuclei likely corresponding to Müller glia. Additionally, faint expression was detectable in rod nuclei (Supplementary Fig. 17q). Furthermore, analysis of published adult human scRNAseq data confirmed enrichment of MEF2C expression in cones relative to rods (p = 0.04489, one-sided T-test) 24 . This protein localization profile, adult human photoreceptor expression analysis, and enrichment both in expressed genes and differentially expressed genes in our dataset suggests a possible role for MEF2C in human cone photoreceptor development or maturation.

Trichromatic Coding

Figure 3. Human rod cells and the different types of cone cells each have an optimal wavelength. However, there is considerable overlap in the wavelengths of light detected.

There are three types of cones (with different photopsins), and they differ in the wavelength to which they are most responsive, as shown in Figure 3. Some cones are maximally responsive to short light waves of 420 nm, so they are called S cones (“S” for “short”) others respond maximally to waves of 530 nm (M cones, for “medium”) a third group responds maximally to light of longer wavelengths, at 560 nm (L, or “long” cones). With only one type of cone, color vision would not be possible, and a two-cone (dichromatic) system has limitations. Primates use a three-cone (trichromatic) system, resulting in full color vision.

The color we perceive is a result of the ratio of activity of our three types of cones. The colors of the visual spectrum, running from long-wavelength light to short, are red (700 nm), orange (600 nm), yellow (565 nm), green (497 nm), blue (470 nm), indigo (450 nm), and violet (425 nm). Humans have very sensitive perception of color and can distinguish about 500 levels of brightness, 200 different hues, and 20 steps of saturation, or about 2 million distinct colors.

Decoding smell

Since the beginning of the pandemic, a loss of smell has emerged as one of the telltale signs of COVID-19. Though most people regain their sense of smell within a matter of weeks, others can find that familiar odors become distorted. Coffee smells like gasoline roses smell like cigarettes fresh bread smells like rancid meat.

This odd phenomenon is not just disconcerting. It also represents the disruption of the ancient olfactory circuitry that has helped to ensure the survival of our species and others by signaling when a reward (caffeine!) or a punishment (food poisoning!) is imminent.

Scientists have long known that animals possess an inborn ability to recognize certain odors to avoid predators, seek food, and find mates. Now, in two related studies, researchers from the Yu Lab at the Stowers Institute for Medical Research show how that ability, known as innate valence, is encoded. The findings, published in the journals Current Biology and eLife, indicate that our sense of smell is more complicated -- and malleable -- than previously thought.

Our current understanding of how the senses are encoded falls into two contradictory views -- the labeled-line theory and the pattern theory. The labeled-line theory suggests that sensory signals are communicated along a fixed, direct line connecting an input to a behavior. The pattern theory maintains that these signals are distributed across different pathways and different neurons.

Some research has provided support for the labeled-line theory in simple species like insects. But evidence for or against that model has been lacking in mammalian systems, says Ron Yu, PhD, an Investigator at the Stowers Institute and corresponding author of the reports. According to Yu, if the labeled-line model is true, then the information from one odor should be insulated from the influence of other odors. Therefore, his team mixed various odors and tested their impact on the predicted innate responses of mice.

"It's a simple experiment," says Qiang Qiu, PhD, a research specialist in the Yu Lab and first author of the studies. Qiu mixed up various combinations of odors that were innately attractive (such as the smell of peanut butter or the urine of another mouse) or aversive (such as the smell of rotting food or the urine of a predator). He then presented those odor mixtures to the mice, using a device the lab specially designed for the purpose. The device has a nose cone that can register how often mice investigate an odor. If mice find a particular mixture attractive, they poke their nose into the cone repeatedly. If they find the mixture aversive, they avoid the nose cone at all costs.

To their surprise, the researchers discovered that mixing different odors, even two attractive odors or two aversive odors, erased the mice's innate behavioral responses. "That made us wonder whether it was simply a case of one odor masking another, which the perfume industry does all the time when they develop pleasant scents to mask foul ones," says Yu. However, when the team looked at the activity of the neurons in the olfactory bulb that respond to aversive and attractive odors, they found that was not the case.

Rather, the patterns of activity that represented the odor mixture were strikingly different from that for individual odors. Apparently, the mouse brain perceived the mixture as a new odor identity, rather than the combination of two odors. The finding supports the pattern theory, whereby a sensory input activates not just one neuron but a population of neurons, each to varying degrees, creating a pattern or population code that is interpreted as a particular odor (coyote urine! run!). The study was published online March 1, 2021, in Current Biology.

But is this complicated neural code hardwired from birth, or can it be influenced by new sensory experiences? Yu's team explored that question by silencing sensory neurons early in life, when mice were only a week old. They found that the manipulated mice lost their innate ability to recognize attractive or aversive odors, indicating that the olfactory system is still malleable during this critical period of development.

Interestingly, the researchers found that when they exposed mice during this critical period to a chemical component of bobcat urine called PEA, the animals no longer avoided that odor later in life. "Because the mice encountered this odor while they were still with their mothers in a safe environment and found that it did not pose a danger, they learned to not be afraid of it anymore," says Yu. This study was published online March 26, 2021, in eLife.

Though the COVID-19 pandemic has warped the sense of smell in millions of people, Yu does not predict that it will have significant implications for most adults who recover from the disease. However, he thinks this altered sensory experience could have a major impact on affected infants and children, especially considering the role that many odors play in social connections and mental health.

"The sense of smell has a strong emotional component to it -- it's the smell of home cooking that gives you a feeling of comfort and safety," says Yu. "Most people don't recognize how important it is until they lose it."

Other co-authors from Stowers include Yunming Wu, PhD Limei Ma, PhD, Wenjing Xu, PhD, Max Hills, and Vivekanandan Ramalingam, PhD.

The work was funded by the Stowers Institute for Medical Research and the National Institute on Deafness and Other Communication Disorders of the National Institutes of Health (award numbers R01DC008003, R01DC014701, and R01DC016696). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Visual Processing

Visual signals are processed in the brain through several different pathways.

Learning Objectives

Describe the complexity of visual processing in the brain

Key Takeaways

Key Points

  • The magnocellular pathway carries information about form, movement, depth, and differences in brightness the parvocellular pathway carries information on color and fine detail.
  • The optic chiasma allows us to coordinate information between both eyes and is produced by crossing optical information across the brain.
  • Visual signals move from the visual cortex to either the parietal lobe or the temporal lobe.
  • Some signals move to the thalamus, which sends the visual signals to the primary cortex.
  • Visual signals can also travel from the retina to the superior colliculus, where eye movements are coordinated with auditory information.
  • Visual signals can move from the retina to the suprachiasmatic nucleus (SCN), the body’s internal clock, which is involved in sleep/wake patterns and annual cycles.

Key Terms

  • superior colliculus: the primary area of the brain where eye movements are coordinated and integrated with auditory information
  • optic chiasma: found at the base of the brain and coordinates information from both eyes
  • suprachiasmatic nucleus: cluster of cells that is considered to be the body’s internal clock, which controls our circadian (day-long) cycle

Higher Processing

The myelinated axons of ganglion cells make up the optic nerves. Within the nerves, different axons carry different parts of the visual signal. Some axons constitute the magnocellular (big cell) pathway, which carries information about form, movement, depth, and differences in brightness. Other axons constitute the parvocellular (small cell) pathway, which carries information on color and fine detail. Some visual information projects directly back into the brain, while other information crosses to the opposite side of the brain. This crossing of optical pathways produces the distinctive optic chiasma (Greek, for “crossing”) found at the base of the brain and allows us to coordinate information from both eyes.

Once in the brain, visual information is processed in several places. Its routes reflect the complexity and importance of visual information to humans and other animals. One route takes the signals to the thalamus, which serves as the routing station for all incoming sensory impulses except smell. In the thalamus, the magnocellular and parvocellular distinctions remain intact there are different layers of the thalamus dedicated to each. When visual signals leave the thalamus, they travel to the primary visual cortex at the rear of the brain. From the visual cortex, the visual signals travel in two directions. One stream that projects to the parietal lobe, in the side of the brain, carries magnocellular (“where”) information. A second stream projects to the temporal lobe and carries both magnocellular (“where”) and parvocellular (“what”) information.

Another important visual route is a pathway from the retina to the superior colliculus in the midbrain, where eye movements are coordinated and integrated with auditory information. Finally, there is the pathway from the retina to the suprachiasmatic nucleus (SCN) of the hypothalamus. The SCN is a cluster of cells that is considered to be the body’s internal clock, which controls our circadian (day-long) cycle. The SCN sends information to the pineal gland, which is important in sleep/wake patterns and annual cycles.

The suprachiasmatic nucleus (SNC): The presence of light and darkness influences circadian rhythms and related physiology and behavior through the SCN.

Watch the video: Prof. Sando Mussa-Ivaldi - The Neural Coding of Force Fields (July 2022).


  1. Pedar

    I thank for the help in this question, now I will know.

  2. Voodoora

    I apologize, but in my opinion you are wrong. Write to me in PM, we will handle it.

  3. Krocka

    I congratulate, by the way, this magnificent thought falls

  4. Amani

    Bravo, what a phrase ... great thought

  5. Gadi

    It doesn't bother me.

  6. Mikanos

    haha patstalom)))))

Write a message