Announcements Jamming Avoidance Response (JAR) Writing assignment - - PDF document

announcements
SMART_READER_LITE
LIVE PREVIEW

Announcements Jamming Avoidance Response (JAR) Writing assignment - - PDF document

9/30/2011 Neuronal Implementation of the Announcements Jamming Avoidance Response (JAR) Writing assignment due Monday (W22) Df = f neighbor - f own Check links on website for lecture related supplements No discussion on Wed after


slide-1
SLIDE 1

9/30/2011 1

Neuronal Implementation of the Jamming Avoidance Response (JAR)

402 Hz Df = -4 Hz 398 Hz Df = +4 Hz 390 Hz Df = fneighbor - fown EOD Frequency (Hz)

Announcements

  • Writing assignment due Monday (W22)
  • Check links on website for lecture related

supplements

  • No discussion on Wed after fall break
  • Wiki-titles due Monday, Oct. 3

Neuronal Implementation of the Jamming Avoidance Response (JAR)

402 Hz Df = -4 Hz 398 Hz Df = +4 Hz 410 Hz 390 Hz Df = fneighbor - fown EOD Frequency (Hz)

4

Walter Heiligenberg (1937-1994) Pacemaker

Answer: No! The JAR relies solely on sensory information

Fish (S1) Stimulus (S2)

Fish’s discharge silenced by curare

Question: Does the fish compare the stimulus with its own pacemaker?

Pacemaker Stimulus (S2)

JAR

X

Fish (S1) Stimulus (S2) Stimulus (S2) EOD Replacement (S1)

S1 JAR S2

Answer: Yes! The JAR only occurs under differential geometry

Question: Does the JAR depend

  • n stimulus geometry?

Fish’s discharge silenced by curare

S1 + S2 JAR

X

slide-2
SLIDE 2

9/30/2011 2

(1) Two signals: S1 + S2 (2) |Df| < 5 Hz (3) Differential geometry Minimal Stimulus Conditions for JAR Combining Two Periodic Signals Results in a Beat (Amplitude Modulation)

S1=300 Hz S2= 302 Hz +Df S1=300 Hz S2= 298 Hz

  • Df

+ +

|| ||

Two sine waves combine to produce a complex wave (amplitude modulation and phase modulation).

Combined Signal is Also Modulated in Phase (Timing)

S1 S2 S1 + S2

Relationship Between Amplitude and Phase Depends on the Sign of Df

S1 S1 + S2

+Df

  • Df

Fish’s own EOD Neighbor’s EOD

Note: go to “links” page to see Flash Video of JAR

slide-3
SLIDE 3

9/30/2011 3

+Df

  • Df

Amplitude Increases / Phase Delays Amplitude Decreases / Phase Advances Amplitude Increases / Phase Advances Amplitude Decreases / Phase Delays

Peripheral Electroreceptors T-units T-units

Note: go to links page to download flash video of T-unit

P-units P-units

Go go links page to download P-unit video

slide-4
SLIDE 4

9/30/2011 4

Together, T-units and P-units provide the necessary information for executing the JAR

+Df

  • Df

Neuroanatomy of the JAR

Heiligenberg (1991) In: Neural Nets in Electric Fish

Neuroanatomy of the ELL T-units converge onto spherical cells

  • Combining inputs from

several T-afferents results in even more precise action potential times

Neuroanatomy of the ELL P-units project to pyramidal cells

inhibitory interneurons

  • E-units respond to

amplitude increases

  • I-units respond to

amplitude decreases

slide-5
SLIDE 5

9/30/2011 5

Neuroanatomy of the JAR The torus is a laminated structure Spherical cells and pyramidal cells project to different laminae

  • Pyramidal cells

project to several laminae in the torus (3, 5, 7, and 8)

  • Spherical cells

project exclusively to lamina 6

1) Giant cells receive direct excitatory input from spherical cells 2) Spherical cells also send fibers to tiny dendrites of small cells 3) Small cells receive excitatory inputs from giant cells onto their soma 4) Small cells receive convergent timing input from different parts of the body surface 5) Small cells respond to differences in timing between different parts of the body surface

Phase comparisons are made in lamina VI Convergence of amplitude and phase information within the torus

1) Small cells in layer 6 project to other layers of the torus 2) Phase sensitive neurons are found in layers 5, 7, 8a, 8b, 8c, and 9 3) E and I inputs from ELL project to layers 5, 7, 8a, and 8c 4) Phase and amplitude information converge in layers 5, 7, and 8

Responses of torus neurons to jamming stimuli

AM DPM

  • Df

+Df

  • Df

+Df

  • Df

+Df

Amplitude-sensitive cell (E-unit) Phase-sensitive cell (adv-unit) Sign-selective cell (E/adv-unit)

delay advance

slide-6
SLIDE 6

9/30/2011 6

Model of a sign-selective cell

I unit Phase advance unit E unit

  • Df-selective

neuron +Df-selective neuron

Neuroanatomy of the JAR

  • Df-selective

E/adv I/del

nE integrates input from sign-selective neurons in the torus and drives prepacemaker nuclei

nE nE PPn sPPn Pn

+Df-selective E/del I/adv

Neuroanatomy of the JAR

sPPn

The African fish Gymnarchus also performs the JAR The electrosensory systems of Gymnarchus and Eigenmannia evolved independently

Gymnarchus Eigenmannia

electroreception electroreception EOD EOD

slide-7
SLIDE 7

9/30/2011 7

Neuronal implementation of the JAR in Gymnarchus

Masashi Kawasaki

  • Like Eigenmannnia, Gymnarchus does

not have a corollary discharge

  • Gymnarchus uses the same algorithm
  • f combining information about

amplitude and differential phase

Gymnarchus also has separate amplitude- and time-coding primary afferents

S-unit O-unit

Amplitude-sensitive cell in ELL

  • Gymnarchus has E-units

and I-units in ELL, just like Eigenmannia

Phase-sensitive cell in ELL

  • Eigenmannnia

calculates differential phase in the torus

  • Gymnarchus

calculates differential phase in ELL

Sign-selective cells in the torus

  • Gymnarchus has sign-

selective neurons in the torus, just like Eigenmannia

1) It is sometimes advantageous to study novel behaviors because they are well-suited to neurophysiological study 2) Task-specific functions are encoded in separate, often parallel, channels. Time and amplitude channels serve specialized functions (similar to barn owl) 3) Recognition units emerge at higher levels in the sensory

  • hierarchy. The complex recognition properties arise from

successive analysis of features 4) Recognition units may drive motor output 5) Identical computational algorithms may evolve in unrelated species by convergent evolution 6) Neuronal substrates for similar functions may involve very different neuronal subtypes

What have we learned from studying the JAR?