The photoexcitation of retinal is a one-photon process, so I have no doubt that one photon will be enough to activate one rhodopsin molecule.
The no-so-obvious part is whether activation of a single rhodopsin molecule can lead to enough hyperpolarization to trigger an action potential.
The wikipedia article has a good description of the transduction pathway.
In summary:
The photoreceptor cell is kept at -40 mV due to a balance of outward potassium flux and inward sodium flux. The inward sodium flux is regulated by ion channels that depend on cyclic GMP to stay open. At this level of polarization, voltage-gated calcium channels allow calcium to flow into the cell. The calcium in the cell promotes the release of inhibitory glutamate molecules out of the cell.
The action of a photon is to transform retinal into a catalytic form. This catalytic form then activates a protein called transducin, which, by binding, activates a phosphodiesterase that can break down the cyclic GMP molecules that open the sodium channels. Closing of the sodium channel means that now there is only an outward potassium flow, so the voltage drops to ~ -60 mV, closing the voltage-gated ion Calcium channels. No calcium means that the inhibitory glutamate is no longer released, so the neuron connected to the receptor becomes activated.
Then, the question is, can a single active rhodopsin molecule lead to enough hyperpolarization?
For the first step, I have managed to find this reference, in which they show that a single activated rhodopsin molecule is able to activate approximately 1,000 transducin molecules.
The phosphodiesterase that transducin activates is PDE6, which is a tetramer with two regulatory subunits. Since activation of PDE6 requires the transducing to remain bound, and I am not sure whether one or two are needed, we can say that as a maximum one event will trigger the activation of 1,000 PDE6 molecules. I have found a recent overview of PDE6 here.
The catalytic rate of PDE6 is 6,000 - 8,000 cGMP molecules per second (source). So shortly after the absorption of one photon we will have about 8 million cGMP molecules being broken per second.
The concentration of cGMP in dark-adapted photoreceptor is about 3.5 μM (source). A rod cell has a volume of about 300 μm^3 (source), so there are approximately 0.3 picoliters of liquid in a cell - and approximately 600,000 cGMP molecules.
So it would take approximately 75 ms of PDE6 being active to completely deplete the cGMP - assuming that the rate is not dependent on concentration (which it is).
An image can be seen in as little as 13 ms (source), but it is probably not necessary that all of the cGMP is depleted for enough hyperpolarization to occur. I did not find a source with an actual threshold for this.
So, my personal conclusion from this quick and dirty back-of-the-envelope analysis is that it does seem physically reasonable that a single photon can trigger an action potential. It is interesting as this is not what I expected.
The researchers found that about 90 photons had to enter the eye for a 60% success rate in responding. Since only about 10% of photons arriving at the eye actually reach the retina, this means that about 9 photons were actually required at the receptors. Since the photons would have been spread over about 350 rods, the experimenters were able to conclude statistically that the rods must be responding to single photons, even if the subjects were not able to see such photons when they arrived too infrequently.
Piecing these and your findings together, it hints to an interesting sub-question, what do we really mean when we ask the original question?
Can the human eye physically detect it? Seems like…yes?
Which suggests the subsequent physiological thresholds involved, various human signal processing chains etc. What a fascinating topic.
The choice of a 60% success rate is an interesting one, too.
If we could consciously see single photons, we would experience too much visual “noise” in very low light, and so this filter is a necessary adaptation, not a weakness.
From what I understand, the rhodopsin protein will sometimes become spontaneously activated without having to absorb light. Since a single activation is enough to trigger the next cell to fire, this makes the system so sensitive that it is noisy when it is dark. So it makes sense that rather than considering every triggering event as a positive signal, some later step in the signal processing chain would require a certain number photoreceptor cells to activate before it lets the signal pass through into our consciousness.
I have not looked further into the processing chain, so I don’t know at what level this threshold is applied. It could be that multiple neurons need to activate to provide enough of an impulse to activate the optic nerve - or it could be that the signal from a single event does propagate all the way to the visual cortex, and the visual cortex filters these low signals. I suspect that this is well known, but I shouldn’t allow myself to spend time on this today. I will try to find out when I do have time, as this is a very interesting topic!
The photoexcitation of retinal is a one-photon process, so I have no doubt that one photon will be enough to activate one rhodopsin molecule.
The no-so-obvious part is whether activation of a single rhodopsin molecule can lead to enough hyperpolarization to trigger an action potential.
The wikipedia article has a good description of the transduction pathway.
In summary:
The photoreceptor cell is kept at -40 mV due to a balance of outward potassium flux and inward sodium flux. The inward sodium flux is regulated by ion channels that depend on cyclic GMP to stay open. At this level of polarization, voltage-gated calcium channels allow calcium to flow into the cell. The calcium in the cell promotes the release of inhibitory glutamate molecules out of the cell.
The action of a photon is to transform retinal into a catalytic form. This catalytic form then activates a protein called transducin, which, by binding, activates a phosphodiesterase that can break down the cyclic GMP molecules that open the sodium channels. Closing of the sodium channel means that now there is only an outward potassium flow, so the voltage drops to ~ -60 mV, closing the voltage-gated ion Calcium channels. No calcium means that the inhibitory glutamate is no longer released, so the neuron connected to the receptor becomes activated.
Then, the question is, can a single active rhodopsin molecule lead to enough hyperpolarization?
For the first step, I have managed to find this reference, in which they show that a single activated rhodopsin molecule is able to activate approximately 1,000 transducin molecules.
The phosphodiesterase that transducin activates is PDE6, which is a tetramer with two regulatory subunits. Since activation of PDE6 requires the transducing to remain bound, and I am not sure whether one or two are needed, we can say that as a maximum one event will trigger the activation of 1,000 PDE6 molecules. I have found a recent overview of PDE6 here.
The catalytic rate of PDE6 is 6,000 - 8,000 cGMP molecules per second (source). So shortly after the absorption of one photon we will have about 8 million cGMP molecules being broken per second.
The concentration of cGMP in dark-adapted photoreceptor is about 3.5 μM (source). A rod cell has a volume of about 300 μm^3 (source), so there are approximately 0.3 picoliters of liquid in a cell - and approximately 600,000 cGMP molecules.
So it would take approximately 75 ms of PDE6 being active to completely deplete the cGMP - assuming that the rate is not dependent on concentration (which it is).
An image can be seen in as little as 13 ms (source), but it is probably not necessary that all of the cGMP is depleted for enough hyperpolarization to occur. I did not find a source with an actual threshold for this.
So, my personal conclusion from this quick and dirty back-of-the-envelope analysis is that it does seem physically reasonable that a single photon can trigger an action potential. It is interesting as this is not what I expected.
What a fantastic read! Well done!
Did you see this in the OP article?
Piecing these and your findings together, it hints to an interesting sub-question, what do we really mean when we ask the original question?
Can the human eye physically detect it? Seems like…yes?
Which suggests the subsequent physiological thresholds involved, various human signal processing chains etc. What a fascinating topic.
The choice of a 60% success rate is an interesting one, too.
Thanks! I appreciate you reading it :)
The first paragraph points out the following:
From what I understand, the rhodopsin protein will sometimes become spontaneously activated without having to absorb light. Since a single activation is enough to trigger the next cell to fire, this makes the system so sensitive that it is noisy when it is dark. So it makes sense that rather than considering every triggering event as a positive signal, some later step in the signal processing chain would require a certain number photoreceptor cells to activate before it lets the signal pass through into our consciousness.
I have not looked further into the processing chain, so I don’t know at what level this threshold is applied. It could be that multiple neurons need to activate to provide enough of an impulse to activate the optic nerve - or it could be that the signal from a single event does propagate all the way to the visual cortex, and the visual cortex filters these low signals. I suspect that this is well known, but I shouldn’t allow myself to spend time on this today. I will try to find out when I do have time, as this is a very interesting topic!