Sensory neurons as living control systems

[Matti]

[From MK (2018.12.02.1330 CET)]

Control all the way down... and up.

···

----
Temperature (Austin). 2015 Jul-Sep; 2(3): 336�337.

Published online 2015 Oct 12. doi: [10.1080/23328940.2015.1050157]
PMCID: PMC4843904
PMID: 27227043
Letter on Kobayashi's view of cutaneous thermoreceptors and their role
in thermoregulation
Douglas S Ramsay,1,2,3,* Karl J Kaiyala,1 and Stephen C Woods4

Dear Editor-in-Chief,

The article by Shigeo Kobayashi1 reiterates2,3 a novel concept for
thermoregulation. The concept is provocative, as Kobayashi and
colleagues argue against a fundamental protocol of the canonical
thermoregulatory control scheme, namely that temperature is measured
and encoded by thermal sensors to provide input for the homeostatic
control of body temperature. Classical models of thermoregulation are
envisioned in terms of an “engineering-style�? central controller that
receives, decodes and compares afferent temperature information to a
reference signal (set-point) as a basis for actuating a coordinated
set of effector responses that efficiently, even “wisely,�? defend
normothermia in the face of thermal challenges. This classical model
localizes the “comparator�? to the central nervous system (CNS). By
contrast, Kobayashi proposes that temperature-sensitive receptor
molecules (thermo-TRP channels) located in cutaneous nerve endings are
the actual comparators, being triggered at characteristic threshold
temperatures so as to generate error inputs that actuate CNS-mediated
effector responses. This model both precludes requirements for
temperature encoding-decoding and thermoeffector coordination via a
discrete CNS comparator. Accordingly, Kobayashi has repositioned the
‘thermostat’ from the brain to myriad ‘thermostats’ residing in the
interface with the thermal environment. Moreover, according to this
model, input from the thermoreceptors is conveyed to other brain areas
to evoke temperature sensation (e.g., “cold in the skin�?).
----
Prog Neurobiol. 1989;32(2):103-35.
Temperature-sensitive neurons in the hypothalamus: a new hypothesis
that they act as thermostats, not as transducers.
Kobayashi S1.
https://www.sciencedirect.com/science/article/pii/0301008289900129?via%3Dihub
https://doi.org/10.1016/0301-0082(89)90012-9
----

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4843900/

Temperature (Austin). 2015 Jul-Sep; 2(3): 346�351.

Published online 2015 Apr 27. doi: [10.1080/23328940.2015.1039190]
PMCID: PMC4843900
PMID: 27227048
Temperature receptors in cutaneous nerve endings are thermostat
molecules that induce thermoregulatory behaviors against thermal load
Shigeo Kobayashi*

"When skin temperature falls below a set-point, mammals experience
“cold in the skin�? and exhibit heat-seeking behaviors for error
correction. Physiological thermostats should perform the behavioral
thermoregulation, and it is important to identify the thermostats. A
classical model of the sensory system states that thermoreceptors
(e.g., thermoTRPs) in skin nerve endings are sensors that transform
temperature into the firing rate codes that are sent to the brain,
where the codes are decoded as “cold�? by a labeled line theory.
However, the view that the temperature code is transformed into “cold�?
(not temperature) is conflicting. Another model states that a
thermostat exists in the brain based on the view that a skin
thermo-receptor is a sensor. However, because animals have no
knowledge of the principle of temperature measurement, the brain is
unable to measure skin temperature with a thermometer calibrated based
on a code table of each sensor in the skin. Thus, these old models
cannot identify the thermostats. We have proposed a new model in which
temperature receptors in a nerve ending are molecules of the
thermostats. When skin temperature falls below a set-point, these
molecules as a whole induce impulses as command signals sent to the
brain, where these impulses activate their target neurons for “cold�?
and heat-seeking behaviors for error correction. Our study challenges
the famous models that sensory receptor is a sensor and the brain is a
code processor."
----
https://www.ncbi.nlm.nih.gov/pubmed/3942861

Brain Res. 1986 Jan 1;362(1):132-9.
Warm- and cold-sensitive neurons inactive at normal core temperature
in rat hypothalamic slices.
Kobayashi S.
Abstract
Electrical activities of thermosensitive neurons were recorded
extracellularly in slices of rat preoptic area and anterior
hypothalamus. Of 63 spontaneously firing neurons found at high
searching temperature (37-40 degrees C), 33% were warm-sensitive, 8%
were cold-sensitive and the remaining 59% were thermally insensitive.
In particular, 6 warm-sensitive neurons were active only above 38
degrees C of rat normal core temperature. In contrast, of 38
spontaneously firing neurons found at low searching temperature (32-36
degrees C), 8% were warm-sensitive, 29% were cold-sensitive and the
remaining 63% were thermally insensitive. Furthermore, all these
cold-sensitive neurons were active only below 38 degrees C. Therefore,
the warm- and cold-sensitive neurons active at 38 degrees C would be
functioning for narrow band control and the remaining warm- and
cold-sensitive neurons inactive at 38 degrees C would be recruited for
wide band control when core temperature was changed critically from 38
degrees C. Their firing rate activities often showed obvious threshold
responses, large hysteresis of the threshold responses and remarkable
transient responses to slice temperature changes. From aspects of
automatic control theory, these warm- and cold-sensitive neurons
themselves may be thermostats to regulate the brain temperature rather
than thermosensors to monitor it.
----
M

[Matti]

[From MK (2018.12.11.1330 CET)]

Bruce Nevin 2018-12-07_14:27:54 UTC--

"Where it says 'error inputs' I think we must read 'perceptual
signal', experienced as a perception of 'cold' at the location of the
sensor (or 'hot', presumably, for the opposing error). The structure
of the sensor is homologous with the input side of a control loop
(input function, perceptual input signal, reference signal,
comparator) but what would be the error signal is instead a perceptual
signal representing the deviation of sensed temperature from the
reference value of the reference signal. "

Does this not constitute a tweak to what the perceptual signal "stands
for"? If the neuron is a living control system its generated activity
is not a function of something in the environment; but a function of
internal comparison done within the neuron.

The iotic notion of 'sensory input --> 'peripheral neuron' -->
neuronal output' appears to be as wrong as the Input --> organism -->
output model of organisms.

The question is then, as always, not what the neuron "senses" but what
the neuron controls and with respect to which internal reference. Does
the notion of 'sensing' not start to collapse if one begins to think
about the individual periphery neuron as a control systems?

That neuron controls 'cold' with respect to that internal reference;
that other neuron controls 'heat' with respect to that internal
reference; and so on.

M

[MartinT]

[Martin Taylor 2018.12.11.13.21]

[From MK (2018.12.11.1330 CET)]

Bruce Nevin 2018-12-07_14:27:54 UTC--

"Where it says 'error inputs' I think we must read 'perceptual
signal', experienced as a perception of 'cold' at the location of the
sensor (or 'hot', presumably, for the opposing error). The structure
of the sensor is homologous with the input side of a control loop
(input function, perceptual input signal, reference signal,
comparator) but what would be the error signal is instead a perceptual
signal representing the deviation of sensed temperature from the
reference value of the reference signal. "

Does this not constitute a tweak to what the perceptual signal "stands
for"? If the neuron is a living control system its generated activity
is not a function of something in the environment; but a function of
internal comparison done within the neuron.

I would say that it is a function of both. Forget PCT for a moment, and think only of sensory adaptation as you perceive in, say, when you move from bright sunlight into a relatively dark house. At first you see almost nothing, but after a while you see just as well as you did in the sun outside. It's easy to see this as the sensitivity of the neuron being controlled so as always to produce the same average firing rate over a long time. But a perfect control loop does not exist. There is always some effect of the disturbance on the perception, and we can tell whether we are in a bright sunlit garden or a moderately brightly lit room.

If the neuron is a control system supplied with some external reference value, where might that value come from? Is it intrinsic to the neuron, like any top-level controlled variable, or is is supplied with an external reference value from somewhere? The effect in visual rods and cones coincides with depletion of the relevant molecules by the absorption of photons and the resultant neuron firing. Neurons in the brain have many thousands of inputs. If each is a control system in itself, what is it controlling? PCT doesn't say. PCT talks about the gross changes in large combined "bundles" of neurons when something changes in the sensory environment. Does each neuron similarly have a bundle of "reference" synaptic inputs, a bundle of "disturbance" sensory inputs and internal processing that result in a microsecond by microsecond decision whether to fire? Who knows.

The iotic notion of 'sensory input --> 'peripheral neuron' -->
neuronal output' appears to be as wrong as the Input --> organism -->
output model of organisms.

Yes. (What does "iotic" mean?)

The question is then, as always, not what the neuron "senses" but what
the neuron controls and with respect to which internal reference.

I don't think that question is properly posed. As soon as you ask "what" is controlled by any control system or in any control loop, you presuppose that there is a real-reality answer. If you say that "brightness" is controlled, you assert that there is some "brightness" property of your environment. You might go further, and assert that this "brightness" property is a flux of things that are both waves with no location and particles with a specific location and momentum that are theorized to exist in one version of what real-reality consists of. The best you can say to someone who asks the "what" question is something along the lines of "When I do this, I would say I am changing the ixfodling. Would you say that it was changing the ixfodling for you, too?" For "ixfodling", you could substitute "brightness" or "warmth" or "loudness" or even "the legality of the situation", but you couldn't usefully say "photon flux" or "average molecular velocity".

  Does
the notion of 'sensing' not start to collapse if one begins to think
about the individual periphery neuron as a control systems?

Not the way I see it. If you accepted that premise, you would also have to say that no level of the hierarchy controls perception of anything related to the environment, whereas the fact that our ancestors survived long enough to produce us is a demonstration that at least some of these perceptions were useful in keeping them alive for a while -- including behaving in ways that were sufficiently close to legal that they were not executed as juveniles. The "legality of an action" is just as much a perception of your environment as is "brightness".

That neuron controls 'cold' with respect to that internal reference;
that other neuron controls 'heat' with respect to that internal
reference; and so on.

Maybe so, but anyone who lives in a temperate zone of the world has experienced that a temperature of, say, 10°C will feel cold in September but hot in March (select the temperature appropriate to your own environment--10°C is appropriate for Toronto). The environment is not "warm" or "cold". You are, but that does not stop you from moving into a place where you feel comfortably warm, whatever the temperature of "comfortably warm" might be at a particular moment.

Martin

[Matti]

[From MK (2018.12.12.1815 CET)]

Martin Taylor 2018.12.11.13.21--

Yes. (What does "iotic" mean?)

Ioticy:

Refers to roughly the system concept that preceded the system concept
implied by PCT in the investigation of biological and physical
systems. My provocative popularized shorthand way to speak about "the
input-output approach" that does not involve the notion of control.
The assumption of "ioticy" concerns one's underlying system concept
when one chooses to investigate a system where the presence of control
is physically plausible; not one's actual chosen modelling approach.
The iotic view was the dominant view in the treatment of organisms for
a long time period; it still is in some undergraduate literature and
it is the most common viewpoint in the treatment of e.g. neurons (the
neuron is considered to 'take in' input and _transform_ that input to
output). So ioticy is just my term for "S-->O --> R" or in the in case
of neurons; "I--> N --> O".

Since most people lack an understanding of control, and I'd actually
place myself in that category despite my exposure to Powers' writings,
the iotic worldview is still the dominant view of in the life
scientific literature; the latter interpreted in the broad sense i.e.
everything from biochemistry and upwards.

M