Exploration of larval zebrafish internal states using temperature

Presented by

Guillaume Le Goc

Supervised by

Georges Debrégeas

Raphaël Candelier

October 1st 2021

Internal states

"Dynamic, interconnected communication loops distributed across body, brain, and time"

[Kanwal et al., Integr. Comp. Biol. 2021]

Going for a party ?

Yay !

Nope

Mood

All good

Stressed

Internal states

"Dynamic, interconnected communication loops distributed across body, brain, and time"

[Kanwal et al., Integr. Comp. Biol. 2021]

Stress

Hunger

Arousal

Fatigue

...

[Kanwal et al., Integr. Comp. Biol. 2021]

Internal states

"Dynamic, interconnected communication loops distributed across body, brain, and time"

[Kanwal et al., Integr. Comp. Biol. 2021]

Stress

Hunger

Arousal

Fatigue

...

[Kanwal et al., Integr. Comp. Biol. 2021]

Model : Zebrafish larva

[Lister et. al., Development 1999]

Small & transparent vertebrate,

Mutant & transgenic lines,

Rich behavioral repertoire

Calcium reporter expressed in nearly all neurons : brain-wide activity recordings

[LJP]

100µm

5-7 days post-fertilization larvae

External sensory environment : Temperature

[Haesemeyer et. al., Cell Syst. 2015]

Map of heat responsive neurons

Timescale of heat perception

[Haesemeyer et. al., Neuron 2018]

Zebrafish in the wild : 18-38°C

Exploration of larval zebrafish internal states with temperature

Study the effect of uniform temperature on behavior & brain activity

 

Define, measure & study internal states

Outline

Introduction

  1. Behavior
    1. Setup & larval locomotion
    2. Thermal modulation of exploratory behaviors
  2. Calcium imaging
    1. Spontaneous neural activity : the case of the ARTR
    2. Data-driven Ising model

Summary & perspectives

Outline

Introduction

  1. Behavior
    1. Setup & larval locomotion
    2. Thermal modulation of exploratory behaviors
  2. Calcium imaging
    1. Spontaneous neural activity : the case of the ARTR
    2. Data-driven Ising model

Summary & perspectives

1.1 Setup & larval locomotion

1.1 Setup & larval locomotion

[Gallois & Candelier, PLoS Comp. Biol. 2021]

Offline tracking with FastTrack

Discrete swim bouts

1.1 Larval locomotion

Discrete swim bouts

Short-term kinematic parameters

Interbout interval

\delta{t}_n = t_{n+1}-t_n

1.1 Larval locomotion

Discrete swim bouts

Short-term kinematic parameters

Interbout interval

\delta{t}_n = t_{n+1}-t_n

Displacement

d_n = \left\lVert \vec{r}_{n+1} - \vec{r}_{n} \right\lVert

1.1 Larval locomotion

Discrete swim bouts

Short-term kinematic parameters

Interbout interval

\delta{t}_n = t_{n+1}-t_n

Displacement

d_n = \left\lVert \vec{r}_{n+1} - \vec{r}_{n} \right\lVert

Turn angle

\delta\theta_n = \delta\theta_{n+1} - \delta\theta_n \textrm{(mod } 2\pi\textrm{)}

1.1 Larval locomotion

p_{turn} = \frac{\textrm{left turns} + \textrm{right turns}}{\textrm{number of bouts}}

1.1 Larval locomotion

p_{turn}

3 short-term kinematic parameters

Interbout interval

\delta{t}

Displacement

d

Turn amplitude

\left|\delta\theta\right|

+ 1 trajectory-based parameter

Turn probability

1.2 Thermal modulation of behavior

[Le Goc et. al., BMC Biol. 2021]

10 batches of 10 fish, 30 minutes

5 temperatures :

T = 18, 22, 26, 30, 33°C

1.2 Parameters are thermally modulated

[Le Goc et. al., BMC Biol. 2021]

1.2 Parameters are thermally modulated

[Le Goc et. al., BMC Biol. 2021]

1.2 Reorientation dynamics

[Dunn, Mu et. al., eLife 2016]

1.2 Orientational persistence

[Dunn, Mu et. al., eLife 2016]

[Karpenko et. al., eLife 2020]

1.2 Orientational persistence

[Le Goc et. al., BMC Biol. 2021]

[Karpenko et. al., eLife 2020]

Ternarized reorientation

\Delta

Left turn     →  +1

Forward     →   0

Right turn  →   -1

\left\langle\Delta_{n+1}\right\rangle = p_{turn}(1-2p_{flip})\Delta_n

1.2 Orientational persistence

[Le Goc et. al., BMC Biol. 2021]

[Karpenko et. al., eLife 2020]

Ternarized reorientation

\Delta

Left turn     →  +1

Forward     →   0

Right turn  →   -1

\left\langle\Delta_{n+1}\right\rangle = p_{turn}(1-2p_{flip})\Delta_n

1.2 Parameters statistical coupling

[Le Goc et. al., BMC Biol. 2021]

5 thermally modulated parameters :

\delta{t}
d
\delta\theta
p_{turn}
k_{flip}

Sufficient to describe trajectories

1.2 Parameters statistical coupling

[Le Goc et. al., BMC Biol. 2021]

5 thermally modulated parameters :

\delta{t}
d
\delta\theta
p_{turn}
k_{flip}

1.2 Parameters statistical coupling

Nicoguaro CC BY 4.0

1.2 Well-defined behavioral space

[Le Goc et. al., BMC Biol. 2021]

1.2 Behavioral space exploration

[Le Goc et. al., BMC Biol. 2021]

Single-fish experiments, 26°C, 2h long, N = 18

1.2 Inter- & intra-individual variability

[Le Goc et. al., BMC Biol. 2021]

1 color = 1 fish

T = 26°C

1.2 Diffusive modulation of the behavior

[Le Goc et. al., BMC Biol. 2021]

Behavioral modulation time scale ~ 30min

Fish #13 trajectories

1. Summary

Larval zebrafish :

  • has a constrained accessible locomotor repertoire
  • explores it on a time scale of 30min
  • controls the behavioral space
  • is a convenient external drive of the internal states

Difficulty to be in a stationnary regime

Temperature :

Outline

Introduction

  1. Behavior
    1. Setup & larval locomotion
    2. Thermal modulation of exploratory behaviors
  2. Calcium imaging
    1. Spontaneous neural activity : the case of the ARTR
    2. Data-driven Ising model

Summary & perspectives

2. Calcium imaging

2.1 Spontaneous activity

T = 18, 22, 26, 30, 33°C

30min recordings, ~ 6 volumes/s

~ 50 000 neurons/recordings

2.1 The ARTR

[Dunn, Mu et. al., eLife 2016]

[Wolf et. al., Nat. Commun. 2017]

Anterior Rhombencephalic Turning Region

2.1 The ARTR

Tuned with turn direction (Dunn, Mu, et. al. eLife 2016)

Tuned with eyes orientation (Wolf, et. al., Nat. Commun. 2017)

[Wolf et. al., Nat. Commun., 2017]

Bilateral circuit with persistent activity

[Wolf et. al., Nat. Commun. 2017]

[Karpenko et. al., eLife 2020]

m_L
m_R

2.1 Locomotor reorientation dynamics

S(f) = \frac{4Ik_{flip}}{4k_{flip}^2 + (2\pi{}f)^2}

[Wolf, Le Goc et. al., in prep.]

Random telegraph signal

2.1 ARTR switching dynamics

Behavior

ARTR

[Wolf, Le Goc et. al., in prep.]

2.1 ARTR switching dynamics

Behavior

ARTR

[Wolf, Le Goc et. al., in prep.]

Behavior/Neural activity consistency

2.1 ARTR persistence

[Wolf, Le Goc et. al., in prep.]

With S. Wolf, S. Cocco & R. Monasson (LPENS)

2.2 Data-driven Ising model

[Wolf, Le Goc et. al., in prep.]

?

2.2 Data-driven Ising model

P\left(\bm{s}\right) = \frac{1}{Z}\exp\left(-E(\bm{s}) \right)
E(\bm{s}) = -\sum_{i}h_i s_i - \sum_{i,j} J_{ij}s_i s_j

[Schneidman et. al., Nature 2006]

h_i
J_{ij}

local fields

coupling between neurons i and j

[Wolf, Le Goc et. al., in prep.]

Less constrained model (max. entropy)

\bm{s}

2.2 Model parameters inference

[Wolf, Le Goc et. al., in prep.]

2.2 Real vs synthetic data

Real

Synthetic

Monte Carlo Metropolis Hastings sampling of

[Wolf, Le Goc et. al., in prep.]

P(\bm{s})

2.2 Model captures persistence features

[Wolf, Le Goc et. al., in prep.]

2.2 Model captures persistence features

[Wolf, Le Goc et. al., in prep.]

Energy landscape

defines dynamics

2. Summary

ARTR thermal modulation consistent with behavior despite immobilisation

Time-independent, energy-based model reproduces neural dynamics

Persistent activity can emerge from the network

Model captures persistence times variability and thermal modulation

Outline

Introduction

  1. Behavior
    1. Setup & larval locomotion
    2. Thermal modulation of exploratory behaviors
  2. Calcium imaging
    1. Brain-scale response to water puffs
    2. Spontaneous neural activity : the case of the ARTR
    3. Data-driven Ising model

Summary & perspectives

Summary

  • Temperature drives locomotor space
  • Continuous and diffusive-like modulation of internal states
  • Framework to treat inter- & intra-individual variability
  • Orientational internal state neural substrate identified : ARTR
  • Persistence properties emerge from network
  • Interpretable data-driven model captures circuit features
  • Find upstream input of ARTR
  • Use the model in stimulation context
  • Mean-field theory
  • Make use of whole-brain data
  • Infinite space navigation
  • Use internal state as a basis to study reaction to stimulation

Future work

Acknowledgements

Georges Debrégeas

Raphaël Candelier

Volker Bormuth

Ghislaine Morvan-Dubois

Sophia Karpenko

Geoffrey Migault

ulie Lafaye

Natalia Belén Beiza Canelo

Thomas Panier

Benjamin Gallois

Fish facility staff

Abdelkrim Mannioui

Alex Bois

Stéphane Tronche

Marie Breau

Marion Baraban

Rémi Monasson

Simona Cocco

Sébastien Wolf

Hugo Trentesaux

Hippolyte Moulle

Sharbatanu Chatterjee

Mattéo Dommanget-Kott

Alexandre Nauleau

Antoine Hubert

Leonardo Demarchi

Malika Pierrat

Anissa Zerouklane

Acknowledgements

Georges Debrégeas

Raphaël Candelier

Volker Bormuth

Ghislaine Morvan-Dubois

Rémi Monasson

Simona Cocco

Sébastien Wolf

Thanks !

Sophia Karpenko

Geoffrey Migault

ulie Lafaye

Natalia Belén Beiza Canelo

Thomas Panier

Benjamin Gallois

Hugo Trentesaux

Hippolyte Moulle

Sharbatanu Chatterjee

Mattéo Dommanget-Kott

Alexandre Nauleau

Antoine Hubert

Leonardo Demarchi

Malika Pierrat

Anissa Zerouklane

Fish facility staff

Abdelkrim Mannioui

Alex Bois

Stéphane Tronche

Marie Breau

Marion Baraban

1.2 Thermal gradient

10 x 10 fish, 45min

1.2 Temperature preference

1.2 Temperature preference

1.2 Temperature effect on kinematics

1.2 Temperature effect on IBI

1.2 Temperature effect on displacement

1.2 Temperature effect on turn amplitude

2.1 Response to water puffs

2.1 Response to water puffs

2.1 Response to water puffs

Z-projected trial-averaged mean reponse of 1 fish

hot (n=8)

cold (n=12)

Qualitative response map

2.1 Activity space drives persistence

[Wolf, Le Goc et. al., in prep.]

2.2 Real vs synthetic data

Real

Synthetic

[Wolf, Le Goc et. al., in prep.]

2.2 Real vs synthetic data

[Wolf, Le Goc et. al., in prep.]