Thursday 5 July 2012

Julio Martinez: Voluntary Attention and Working Memory in The Primate Brain: Recording from Single Cells


      Abstract: Neurons in the dorsolateral prefrontal cortex (dlPFC) of primates respond to visual stimuli and are selective for attributes such as location, object identity and motion direction (Zaksas and Pasternak, 2006). Furthermore, this selectivity persists when stimuli are removed and their attributes maintained in working memory. Previous studies have suggested that some dlPFC neurons preferentially represent the current sensory input, while others represent the contents of working memory (Fuster, 2000; Pasternak and Greenle, 2005). To investigate this issue we recorded the spiking activity of 155 dlPFC neurons from two rhesus monkeys while they performed different tasks in which they compared the motion direction of a sample random-dot pattern to that of a subsequent test pattern. In the Memory task, the sample and test presentations were separated by a delay period during which the monkey was required to remember the sample direction. In the No-memory task, the sample remained present during the entire trial, thus eliminating the working memory requirement. For each of the two tasks, the ability of each neuron to represent the motion direction of the sample during the delay period was quantified using signal detection theory. In approximately half of the direction-selective neurons, representations were stronger when the sample remained present (No-memory task) than when it was remembered (Memory task). Interestingly, in the remaining neurons, the sample direction was more strongly encoded when it was remembered than when it remained perceptually available. This suggests that while the former neurons preferentially encode sensory input, the latter may serve a specific role in working memory maintenance. The ability of the entire population of recorded neurons to represent the sample direction was quantified using a linear discriminant analysis. In both tasks classification performance remained well above chance throughout the entire delay period. These results demonstrate that in the dlPFC the strength of visual representations during working memory is temporally robust and comparable to that of representations driven by sensory input.

Fuster JM. Memory networks in the prefrontal cortex. Prog Brain Res.
2000;122:309-16. Review http://bit.ly/FusterPFC
Zaksas D, Pasternak T. Directional signals in the prefrontal cortex and in
area MT during a working memory for visual motion task. J Neurosci. 2006
Pasternak T, Greenlee MW. Working memory in primate sensory systems. Nat Rev
Neurosci. 2005 Feb;6(2):97-107. Review. Erratum in: Nat Rev Neurosci. 2005

Comments invited

13 comments:

  1. Dr. Martinez mentioned that, partly for ethical reasons, macaques used for single-cell recordings are Rhesus, and not Chimpanzee. I thought that comment, even if done in passing, was directly relevant to some of the broad questions we are addressing in this summer institute. Why is it 'ethical' to implant electrodes on a Rhesus macaque and not a Chimp? Are such questions solely answered using classical phylogeny? Or are attributions of some form of 'consciousness' being made to non-human great-apes (chimps, along with gorillas and bonobos)?

    ReplyDelete
    Replies
    1. Just to clarify (and this does not change the validity of your question), my interpretation of what he said was that we don't implant chimps for certain types of questions. There might be other questions where the rationale for using chimps is sufficient (I don't actually know).

      With the disclaimer that I have had to apply for REB approval in animal research, I can imagine that there are reasons other than phylogeny for not working with chimps. For example, maybe they are more likely to exhibit distress in laboratory living conditions.

      Delete
    2. This true, the reasons for approving a protocol are not only phylogeny. Other reasons are : is the research going to contribute to our knowledge of the human brain so we can use this knowledge to generate treatment strategies or new drugs and improve the life of humans? is is possible to use another model? does the procedure cause unnecessary pain or discomfort? is modelling an alternative? All of these are carefully examined by animal care committees that recommend the protocol for approval or not. Primates are consider level D protocols in many institutions, even if they demonstrate that the procedure does not cause pain and they minimize discomfort. Why not chimps? I do not have a good answer but if you can do it with rhesus why using chimps? They are larger animals, they are more aggressive and difficult to handle and more expensive to keep in colonies. I have never come across a protocol that proposes working with chimps at least doing electrophysiological experiments in chimps. At the beginning of the last century some scientists did lesson experiments in chimps but as far as I know they don't do that any longer. To finish, currently some groups conduct electrophysiological experiments in humans that undergo neurosurgical procedures, of course they need approval from the subjects and ethics committees. These are patients that undergo brain surgery for other reasons and during the procedure scientists and physicians map different brain areas and use different types of stimuli. Unfortunately these are not done very often and the controls for these experiments are not the best (for example eye movements and visual stimuli), so we cannot replace monkey experiments by human experiments. But the latter tells you the Phylogeny is not the only criteria.

      Delete
  2. I agree with Mr. Sheppard-Jones. Conscious pain perception seems to apply to all mammals, because of the well-established anatomy of the thalamus and neocortex. The brain anatomy of birds is now being interpreted in very similar ways.

    Behaviorally, the signs of pain perception are extremely widely shared among species. Writhing, distress cries, and the long evolutionary history of the sensory pain system all suggest that there are major ethical questions whenever we inflict tissue damage on an animal.

    Chimps and macaques are primates, and are hard to differentiate from humans with respect to pain perception. Notice that we routinely study pain in these creatures, and that researchers currently use the term "conscious" to refer to waking animals of all kinds. Whether that can be proven rigorously is another question. There can be no reasonable doubt in mammals, because of the shared brain anatomy, physiology, and behavioral signs.

    Depression is currently studied in rats, using a "social defeat" paradigm. Socially rejected rats (it's hard to be rejected by an attractive peer!) act depressed, their posture is depressed, their eyes are downcast, I'll bet their vocalizations sound depressed, and of course they neurochemistry looks stressed and depressed. Testosterone goes down, they give up sooner in fights.

    A search of PubMed.gov will turn up all kinds of evidence.

    David Edelman and others are making the case that cephalopods may have striking similarities, in spite of their very different evolutionary history.

    NOTE: I take this evidence to mean that animal experimentation raises ETHICAL QUESTIONS, not that we must avoid all animal experimentation. There are tradeoffs. It is also true that animal experiments have been crucial in saving millions of human lives.

    I also feel very strongly that humans who want to eliminate all animal experiments are immature and ethically wrong.

    There really are very difficult ethical tradeoffs. There are no simple answers.

    ReplyDelete
    Replies
    1. Dr. Gary Comstock's talk was a good follow-up to these issues. Indeed, no simple answers.

      Delete
  3. Martinez demonstrated that endogenous attention enhances the response of MT neurons to their preferred stimulus. Endogenous attention is a conscious process, which (I imagine) involves suppressing attention to distractors. Sounds a bit like the conflicts that Morsella talked about in Supramodular Interaction Theory.

    ReplyDelete
    Replies
    1. Not sure about the theory, perhaps you can expand on that. Endogenous attention seems to be under voluntary control since you can direct it voluntarily. The way we think about it is that working memory representations (which seem to be conscious) become attentional template signals that travel from frontal and parietal areas toward sensory areas to enhance or inhibit sensory processing. I would like also to add that many times it is difficult to map concepts into physiological mechanisms but I am not sure whether this is the case.

      Delete
  4. Did I remember correctly Dr. Martinez saying that here is a 5x expansion of frontal cortex between macaque monkey to human? If so, that sounds contradictory or at least exaggerated in comparison to Dr. Finlay's report that brain parts increase in size proportionally with the whole brain size across species. She did mention that telencephalon increase has a steeper slope than other brain parts but 5x sounds like a lot between pretty similar species!

    ReplyDelete
    Replies
    1. The expansion of the prefrontal cortex in humans is well treated in the book "The prefrontal cortex" by Joaquin Fuster. He states that in humans is about 30% of the neocortex, in chimps about 18% and in rhesus about 11%. If you look at rodents is probably less than 5% and it is not as well defined as in primates. This is a good reason to use primates a s models to explore the function of the prefrontal cortex. My personal opinion is that you cannot treat the cortex as an organ. The cortex has areas and these areas vary in size and function within and between species. For example, animals that do not rely on vision to explore the world do not need a large visual cortex (between species). Also if you have recorded at some point from areas in visual cortex of animals that have it, you will realize how different an MT neurones and a V4 neuron are in terms of responses properties and likely in terms of connectivity. For the case of the prefrontal cortex it seems has expanded in humans according to Fuster. Also there are anatomical criteria (see petrides and Pandya) that differentiate amongst areas of the prefrontal cortex in macaques and also humans.

      Delete
  5. Attention and consciousness was a topic that had been tackled several times during the whole summer school, and I think this is a topic that is worth paying more attention on it (if I can say so!) and I was wondering if the fact that all what is conscious seems to be at some point attented could be a good possibility to explore to explain the adaptive function of consciousness. Why are we constantly perceiving stimuli from the environment whereas we attend (we are conscious) only of a tiny proportion of them. And why some stimuli can reach the level of consciousness whereas other will always be unconscious. Why would we need to distinguish between attented things between unattended, and whyhaving both unsconscious mechanisms and conscious mechanisms? It seems that was is primarily vital is always unconscious (eg. heart beating, axtivation of the imune system, digestive processes, etc...) and only depends on the internal state of the body, wheras what is conscious are functions that requires adjustment with the external environment (eg. predator avoidance, adjustments to climate, etc...) or functions that require external ressources to adjust the internal needs (eg. feeding, breathing, etc...). That being said, would consciousness be necassary for organisms living in a complex environment which needs to easily distinguish what is relevant in the environment by attending these relevant environmental elements?

    ReplyDelete
    Replies
    1. Why does the capacity to distinguish what is relevant in a complex environment need to be felt rather than just done? Today's toy robots can already distinguish what is relevant in a simple environment to some extent: How complex does the environment have to be to require feeling? T3? And why?

      Delete
  6. Dr Harnad, thinking about your ever-present objection to any speakers’ proposition as to what consciousness is for (… but this is just doing, not feeling!…), I’m starting to believe that defining the function of feeling is an ill-posed problem, for a simple reason.

    For something to have a function, it needs to act upon something else.

    Well, if something acts upon something else, you would consider it to be a “doing”, not a feeling.

    Then the question is: how can feelings have any function if they are not doing anything?

    Hence, any attempt to explain the function of consciousness would result in attributing a doing to feeling, and would inevitably lead to your famous objection.

    If feeling has a function, feeling is a doing.

    ReplyDelete
    Replies
    1. ILL-POSED PROBLEM OR ILL-FATED SOLUTIONS?

      "Doings" are dynamic states structures: things that move, that you can touch, measure, observe, record.

      Feelings have correlated doings, but they are not themselves doings. They don't move, and the only one who can observe them is the feeler. Otherwise all one can observe of them is their correlated doings.

      So it's perfectly natural to ask: what are the feelings for? The doings seem to be enough.

      i too am, of course, a monist. So I have no doubt that the brain (doings) causes feelings, somehow. I just want to know how; and why?

      Correlations are not the answer.

      (By the way, if telekinetic dualism had been true, and feelings had turned out to be the 5th fundamental force, the hard problem would have been solved. Alas, all evidence contradicts telekinetic dualism: The other 4 forces can do the whole job (of doing) fine, and there's no causal room left/)

      Delete