Maps without frontiers

Mapping differences
1. Visual system
  1. In addition to projections to the tectum (superior colliculus), mammalian retinal ganglion cells (RGCs) also project to the lateral geniculate nucleus (LGN), which relays these inputs to the cortex:
    1. Interestingly, the mapping of RGC axons onto the LGN is also topographic and is set up using similar gradients of Ephrins
      1. (Feldheim et al., 1998, Neuron, 21 1303)
      2. Unlike the tectum, the LGN receives inputs from both eyes
        allows stereoscopic vision to be integrated
        Terminals from, say, the temporal retina of the left eye (blue) are located in adjacent layers of the same region of the LGN as those from the temporal right eye (red)
        Guidance of RGC axons to these layers is predetermined.
2. Olfactory system
  1. How do you represent a non-spatial sense?
    1. 1000 receptors but each neuron expresses only one!
      1. “One neuron – one OR” principle
    2. Receptor expression dispersed in nasal epithelium, but axons become organised in the olfactory bulb (OB)
      1. How is this done?
        Mapping from epithelium to bulb
        Critically different from RT mapping:
        Glomeruli responding to related odorants are clustered in OB
        Receptor expression governs guidance
        Receptor swap experiments demonstrate that where axons go is determined by which receptor is expressed.
        OR activity determines guidance response state
        Olfactory receptors (ORs) are 7 TM GPCR-like molecules
        In the absence of ligand (odour), each receptor has a characteristic basal activity. Early guidance is activity-independent.
        Neurons expressing the same OR have similar cAMP signalling levels (adenylate cyclase dependent;ACIII)
        This determines the level of transcription of familiar guidance cues (Robo/Slit, Eph/Ephrin, Neuropilin/Sema).
        This results in receptor/cue protein levels characteristically associated with expression of a particular OR, which, in turn, determines mapping in olfactory bulb (OB)
        Disruption of guidance cue expression (e.g. Neuropilin2) disrupts regional mapping in OB
        Conversion from continuous to discrete map
        Axons entering OB are pre-sorted due to cue/receptor interactions
        Cue expression switches with time (e.g. from Robo to Nrp/Sema) so that early entering axons then guide later entering axons.
        Sorting into glomeruli is activity-dependent
        Activity drives higher cAMP levels which turns on expression of homophilic adhesion molecules (Kirrels and contactins), and Ephs and Ephrins (again!)
        These interactions sort axons expressing same ORs into groups to form the glomeruli.
      2. And, does this mean there is also spatial organisation of olfactory info in the cortex?
Olfactory system experiments
1. Choi et al. Cell (2011) vol. 146 (6) pp. 1004-15
  1. Experimental strategy
    1. Introduce ‘channelrhodospin’ (ChR2) into subset of PC neurons
      1. ChR2 is a light-activated cation channel that stimulates action potentials upon exposure to light
        ie can ‘fire’ PC neurons independent of mitral cell input.
      2. Stimulate the ChR2+ subset of neurons with light, paired with either an aversive or appetitive (appetite inducing) stimulus in naïve (unconditioned) animals (classic associative learning).
        After conditioning, test whether light stimulus alone can elicit the appropriate behavioural response.
      3. Introducing channelrhodopsin to PC neurons
        3 ways (all using viruses)
        Simplest: use synapsin promoter
        Hits 50% of cells at injection site
        Infect floxed* Chr2 into mouse in which cre driven from Emx1 promoter (excitatory neuron-restricted)
        * These lox sites are in ‘flip’ orientation so cre will invert the gene not delete it
        Also hits 50%, but only excitatory neurons
        Infect floxed* Chr2 at same time as virus containing synapsin driving cre
        Much lower Chr2 expression rate (10%)
  2. ChR2 activation can condition aversive behaviour
    1. Photostimulation (PS) of ChR2-expressing neurons in the piriform cortex - the conditioned stimulus (CS) - was paired with foot shock – the unconditioned stimulus (US) - on only one side of the chamber to condition the animals (10 pairings).
      1. Animals then exhibited flight behaviour to PS alone, but only when ChR2 was present in piriform neurons (and a minimum of 200 had to be infected with ChR2).
      2. Conditioning with odorants and PS together, showed that subsequently either PS or odorants could elicit flight.
  3. ChR2 photostimulation can also drive appetitive behaviours
    1. Mice trained to take water in response to odorant, could be re-trained to respond instead to PS
    2. Male mice also could be trained to associate presence of a female with either an odour as the CS or with PS as the CS
  4. Piriform cortex neurons are plastic in their associative capability
    1. The same set of ChR2-expressing PC neurons can be re-trained in either direction
    2. Distinct sets of ChR2-expressing PC neurons can be trained and retrained to elicit different behaviours
    3. ie the piriform cortex is a very plastic substrate
  5. Does this prove that the PC is the site of odorant learning?
    1. No, just shows that PC can be used for associative learning. (What would prove it?)
    2. However, does show that PC is very plastic: apparently any group of ~200 neurons can be used to elicit diverse behavioural associations, reversibly.
      1. NB similar experiments in other regions of the cortex (e.g. somatosensory) elicited a specific behavioural output according to location (ie topographically constrained). Huber et al., 2008 Nature v451, p61
    3. Nonetheless, strongly suggests that random connections from OB into PC are used to associate odours with particular experiences.
Responses to odorants
1. Learned
  1. Piriform cortex
    1. Beyond the bulb…
      1. As in the visual system, olfactory signals are relayed from the bulb to multiple higher centres (e.g. SC, LGN, then to cortex in visual system)
        However, unlike the visual system, mitral cell axons projecting to the piriform cortex (PC) do not exhibit any spatial organisation
        Correspondingly, individual odorants activate subpopulations of neurons distributed across the PC
        NB Individual PC neurons respond to multiple, structurally dissimilar odorants
    2. Is the piriform cortex the site of olfactory learning?
      1. Choi et al., (2011) test this using optogenetic activation of arbitrary subsets of PC neurons…..
  2. How does the brain know which odorant is which?
  3. In mammals, the majority of odors only drive behaviour after learning. ie the significance of odors is learnt by association.
    1. However, it is not known which brain regions are involved.
2. Innate
  1. Are all responses to odorants learned?
    1. A small subset of odours elicit innate responses
      1. e.g. trimethyl-thiazoline (TMT) from fox elicits fear (in mice!)
        There are spatially invariant projections from OB to cortical amygdala that may be involved.
        Similar bifurcation in flies……
  2. Drosophila olfactory system: conserved function

Next up

Maps without frontiers

Description

Resource summary

Media attachments

Similar

	Created by Kristi Brogden about 10 years ago