pith. sign in

arxiv: 1907.11609 · v1 · pith:6EHK4UTWnew · submitted 2019-07-26 · 🧬 q-bio.NC

Correlation transfer by layer 5 cortical neurons under recreated synaptic inputs in vitro

Daniele Linaro , Gabriel K. Ocker , Brent Doiron , Michele Giugliano This is my paper

Pith reviewed 2026-05-24 15:13 UTC · model grok-4.3

classification 🧬 q-bio.NC
keywords correlation transferlayer 5 neuronsdynamic clamppyramidal neuronsinterneuronscortical microcircuitslinear response theorysynaptic inputs
0
0 comments X

The pith

Layer 5 pyramidal neurons and interneurons transfer correlated inputs with cell-type-specific gain and timescales.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper examines how pyramidal cells and two classes of GABAergic interneurons in layer 5 convert correlated synaptic inputs into correlated outputs. Researchers applied biophysically realistic correlated inputs to these cells in neocortical slices using dynamic clamp. Physiological differences between the cell types produced distinct patterns of correlation transfer. Linear response theory and computational modeling connected single-cell membrane properties to both the strength and the timing of the transferred correlations. The results tie cellular features directly to the distinct functional roles of these neuron classes in cortical microcircuits.

Core claim

The physiological differences between pyramidal neurons and two classes of GABAergic interneurons of layer 5 manifest unique features in their capacity to transfer correlated inputs, with linear response theory and computational modeling showing how cellular properties determine both the gain and timescale of correlation transfer.

What carries the argument

Dynamic clamp delivery of biophysically realistic correlated inputs, quantified through linear response theory to extract gain and timescale of correlation transfer in each cell type.

If this is right

  • Pyramidal neurons and interneurons make functionally distinct contributions to correlated activity within cortical networks.
  • Membrane properties of each cell type set both the amplitude and the temporal filtering of transferred correlations.
  • Network models that treat all layer 5 cells uniformly will miss cell-type-specific effects on population correlations.
  • The observed differences supply a cellular basis for the specialized roles of these neuron classes in microcircuit computations.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Models of cortical population activity could be refined by inserting cell-type-specific transfer functions rather than generic correlation rules.
  • Testing the same dynamic-clamp protocol under neuromodulatory conditions would reveal whether correlation transfer changes with brain state.
  • Pairing these in vitro results with simultaneous recordings from connected pairs in vivo could test whether the measured transfer functions predict actual circuit behavior.

Load-bearing premise

The correlated inputs generated with dynamic clamp in vitro accurately recreate the statistical structure of synaptic bombardment that layer 5 neurons experience in vivo.

What would settle it

Direct measurement showing that the gain and timescale of correlation transfer recorded in vitro under dynamic clamp do not match the correlation structure observed in the same cell types during natural in vivo activity.

Figures

Figures reproduced from arXiv: 1907.11609 by Brent Doiron, Daniele Linaro, Gabriel K. Ocker, Michele Giugliano.

Figure 1
Figure 1. Figure 1: Electrophysiological intrinsic properties of pyramidal cells and two types of interneurons. (A) Typical membrane potential of pyramidal cells in response to the injection of hyperpolarizing and depolarizing current steps. Inset: magnification of the first action potential fired by the cells. (B) Same as (A) but for fast-spiking interneurons with non-accommodating firing pattern and (C) for low-threshold-sp… view at source ↗
read the original abstract

Correlated electrical activity in neurons is a prominent characteristic of cortical microcircuits. Despite a growing amount of evidence concerning both spike-count and subthreshold membrane potential pairwise correlations, little is known about how different types of cortical neurons convert correlated inputs into correlated outputs. We studied pyramidal neurons and two classes of GABAergic interneurons of layer 5 in neocortical brain slices obtained from rats of both sexes, and we stimulated them with biophysically realistic correlated inputs, generated using dynamic clamp. We found that the physiological differences between cell types manifested unique features in their capacity to transfer correlated inputs. We used linear response theory and computational modeling to gain clear insights into how cellular properties determine both the gain and timescale of correlation transfer, thus tying single-cell features with network interactions. Our results provide further ground for the functionally distinct roles played by various types of neuronal cells in the cortical microcircuit.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript claims that physiological differences among layer 5 pyramidal neurons and two classes of GABAergic interneurons produce distinct features in their transfer of correlated inputs when stimulated in vitro with dynamic-clamp currents designed to be biophysically realistic. Linear response theory and computational modeling are used to relate cellular properties to the gain and timescale of correlation transfer, thereby connecting single-cell features to network-level interactions.

Significance. If the dynamic-clamp inputs faithfully reproduce the statistical structure of in-vivo synaptic bombardment, the work supplies a mechanistic link between cellular biophysics and pairwise correlations in cortical circuits. The explicit use of linear response analysis to predict both gain and temporal filtering constitutes a clear strength, moving beyond purely phenomenological descriptions.

major comments (2)
  1. [Abstract] Abstract (final sentence) and Methods (dynamic-clamp protocol): the central claim that cell-type differences manifest 'unique features in their capacity to transfer correlated inputs' rests on the unverified premise that the applied currents reproduce the pairwise correlations, amplitude distributions, and 1–100 Hz spectral content of layer-5 synaptic bombardment in vivo. No quantitative match to in-vivo recordings is supplied; any systematic mismatch in this frequency band would render the reported distinctions specific to the artificial ensemble rather than reflective of cortical microcircuit function.
  2. [Results] Results (linear-response section): the statement that linear-response predictions 'match the recorded traces' is load-bearing for the modeling conclusions, yet the manuscript provides neither the fitting procedure, the number of free parameters, nor quantitative error metrics (e.g., mean-squared error or coherence) that would allow the reader to judge whether the match is predictive or post-hoc.
minor comments (2)
  1. [Methods] Sample sizes, exclusion criteria, and statistical power for each cell class should be stated explicitly in the Methods or a dedicated table.
  2. Figure legends would benefit from explicit indication of which statistical tests were used for the cell-type comparisons shown in the main figures.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and valuable comments on our manuscript. We address the two major comments below. Where the comments identify missing details, we have revised the manuscript to include additional information and clarifications.

read point-by-point responses

  1. Referee: [Abstract] Abstract (final sentence) and Methods (dynamic-clamp protocol): the central claim that cell-type differences manifest 'unique features in their capacity to transfer correlated inputs' rests on the unverified premise that the applied currents reproduce the pairwise correlations, amplitude distributions, and 1–100 Hz spectral content of layer-5 synaptic bombardment in vivo. No quantitative match to in-vivo recordings is supplied; any systematic mismatch in this frequency band would render the reported distinctions specific to the artificial ensemble rather than reflective of cortical microcircuit function.

    Authors: The dynamic clamp protocol was designed using statistics from the literature on in vivo synaptic inputs to layer 5 neurons, including amplitude distributions and spectral content from studies such as those by Destexhe et al. and others on cortical bombardment. However, we agree that a direct quantitative comparison is not provided in the current manuscript. We will add a paragraph in the Methods section and a supplementary figure showing the power spectra and correlation values used, with references to the in vivo data sources they are based on. This will clarify that the inputs are intended to be representative rather than an exact replica of a specific recording. revision: yes

  2. Referee: [Results] Results (linear-response section): the statement that linear-response predictions 'match the recorded traces' is load-bearing for the modeling conclusions, yet the manuscript provides neither the fitting procedure, the number of free parameters, nor quantitative error metrics (e.g., mean-squared error or coherence) that would allow the reader to judge whether the match is predictive or post-hoc.

    Authors: We appreciate this point. The linear response model was fitted to the experimentally measured frequency response functions using a least-squares optimization in the frequency domain. Each cell type had a two-parameter model (gain and cutoff frequency). We will revise the Results section to describe the fitting procedure explicitly, report the best-fit parameter values with confidence intervals, and include quantitative metrics such as the mean squared error and the frequency-dependent coherence between the model prediction and the data. These additions will allow readers to assess the quality of the match. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper reports experimental measurements of correlation transfer in different layer-5 cell types under dynamic-clamp stimulation, then invokes linear-response theory and modeling as explanatory tools to link cellular properties to observed gain and timescales. No quoted equations or steps reduce a claimed prediction or uniqueness result to a parameter fitted on the same data, a self-citation chain, or a definitional renaming. The central claim rests on physiological differences producing distinct transfer features, which is presented as an empirical outcome rather than a quantity forced by construction from the inputs. The modeling is described as providing insight, not as generating the recorded values. This is the common case of a self-contained experimental study whose conclusions do not collapse into their own fitted parameters.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper rests on standard domain assumptions about synaptic input statistics and linear response applicability; no free parameters or invented entities are introduced in the abstract.

axioms (1)
  • domain assumption Linear response theory can be used to predict gain and timescale of correlation transfer from single-cell properties
    Invoked to gain insights into how cellular properties determine correlation transfer.

pith-pipeline@v0.9.0 · 5690 in / 1280 out tokens · 19267 ms · 2026-05-24T15:13:17.618285+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

25 extracted references · 25 canonical work pages

  1. [1]

    itatory and inhibitory synaptic conductance inputs: 7 (2) 𝐼9(𝑡)=𝐺Z,9(𝑡)(𝐸Z−𝑉9(𝑡))+𝐺],9(𝑡)(𝐸]−𝑉9(𝑡))𝐼=(𝑡)=𝐺Z,=(𝑡)(𝐸Z−𝑉=(𝑡))+𝐺],=(𝑡)(𝐸]−𝑉=(𝑡)) where 𝐺Z,9(𝑡),𝐺Z,=(𝑡) (𝐺],9(𝑡),𝐺],=(𝑡)) are randomly fluctuating excitatory (inhibitory) synthetic synaptic conductance waveforms, whose apparent reversal potential is chosen as 𝐸Z=0 𝑚𝑉 (𝐸]=−80 𝑚𝑉) (Chance et al., 20...

    work page 2002

  2. [2]

    In analogy to the current-driven synaptic inputs (Eq

    vanishes: (5) 0=〈𝐼(𝑡)〉≈𝑔Z∙(𝐸Z−𝑉g)+𝑔]∙(𝐸]−𝑉g)⟺ 𝑅Z=[g]abc∙𝜏]∙𝑅]∙(𝐸]−𝑉g)]/[gZabc∙𝜏Z∙(𝐸Z−𝑉g)] By this definition, acting on the value of a single parameter 𝑉g allows one to change the ratio between excitatory and inhibitory inputs and thus to alter the output firing rate of the patched neuron. In analogy to the current-driven synaptic inputs (Eq. 1), where 𝜇,...

    work page 2014

  3. [3]

    we initially also probed the steady-state frequency-current curve of the neuron as in (Chance et al., 2002). Specifically, we injected 1 𝑠-long DC depolarizing current steps increasing amplitude, superimposed to conductance-driven recreated synaptic inputs, and measured the evoked firing rate as the number of emitted spikes divided by 0.9 𝑠, after discard...

    work page 2002

  4. [4]

    Output correlation depends on the fraction of common inputs

    (6) 𝐴= 𝑁z€z−4−1‚M9∙ƒ ISI}−ISI}M9‚/ ISI}+ISI}M9‚†‡ˆ‰‡M9 }Š‹ where 𝑁z€z is the number of spikes evoked by a constant depolarizing 1 𝑠-long step of current, ISI} is the 𝑞-th inter-spike interval (i.e. ISI}=𝑡}9−𝑡}), and where the first four 11 spikes were always discarded from the analysis (i.e. the sum starts from 𝑞=4). The amplitude of the step of curren...

    work page 2007

  5. [5]

    as (13) Cov(𝑛9,𝑛=)(𝜏)=Cov(𝑦9,𝑦=)(𝜏)∗Δ©(𝜏) where Δ©(𝜏) is an even function defined as Δ©(𝜏)=∫𝑤©(𝑠)𝑤©(𝜏+𝑠)𝑑𝑠ÄMÄ taking the value of 𝑇−|𝜏|,for 𝜏∈[−𝑇 ;𝑇], and otherwise zero. By the Wiener-Khinchin theorem, the cross-covariance function Cov(𝑦9,𝑦=)(𝜏) is the Fourier anti-transform of the cross-spectrum of the spike trains CovÆ YÈ9,YÈ=‚(𝜔):: (14) Cov(𝑦9,𝑦=)(𝜏)...

    work page 2007

  6. [6]

    For these quantities, a full experimental characterization was demonstrated previously (Higgs and Spain, 2009, Ilin et al., 2013, Kondgen et al., 2008, Linaro et al., 2018)

    14 (15) YÈ9(𝜔)≅YÈ9,g(𝜔)+AÈ9(𝜔)⋅QÈ(𝜔) YÈ=(𝜔)≅YÈ=,g(𝜔)+AÈ=(𝜔)⋅QÈ(𝜔) where QÈ(𝜔) is the Fourier transform of 𝑞(𝑡), YȪ,g(𝜔) corresponds to the baseline spike train (when 𝑞(𝑡)=0, i=1,2), and where AÈ9(𝜔) and AÈ=(𝜔) are the dynamical response functions of the two neurons. For these quantities, a full experimental characterization was demonstrated previously (H...

    work page 2009

  7. [7]

    and approximate the cross-spectrum by Eq. 15 and obtain (16) CovÆ YÈ9,YÈ9‚(𝜔)=〈YÈ9∗(𝜔)⋅YÈ=(𝜔)〉≅AÈ∗9(𝜔)⋅AÈ=(𝜔)⋅SÑ(𝜔) where AÈ∗9(𝜔) is the complex conjugate of AÈ9(𝜔), and SÑ(𝜔)=〈QÈ∗(𝜔)⋅QÈ(𝜔)〉 is the power spectrum of the common input 𝑞(𝑡). Finally, estimating the covariance of the spike counts 𝑛9(𝑡) and 𝑛=(𝑡), requires evaluating Eq. 13 in 𝜏=0 and substitu...

    work page 2002

  8. [8]

    and their complete estimate, across a sufficiently wide range of firing rates, is incompatible with the limited duration of each experiment. Neuron models For modeling different cortical cell types, we used an adaptive exponential integrate-and-fire model neuron (Brette and Gerstner, 2005), modified to incorporate a slow voltage-gated adaptation current 𝐼...

    work page 2005

  9. [9]

    𝜏Z,𝜏]→0) (Tuckwell, 1989), and 𝑅]=10 𝑘𝐻𝑧, while changing the value of 𝑅Z in order to change the output firing rates

    For simplicity however, we set gZabc=g]abc=0.06 𝑚𝑆/𝑐𝑚=, and considered instantaneous synaptic coupling (i.e. 𝜏Z,𝜏]→0) (Tuckwell, 1989), and 𝑅]=10 𝑘𝐻𝑧, while changing the value of 𝑅Z in order to change the output firing rates. Such a choice for 𝑅] , used for Figure 8C-H, reflects the firing rate of the summed activity of the modeled presynaptic inhibitory ...

    work page 1989

  10. [10]

    noise-free

    and extended it by dynamic-clamp to the case of conductance inputs, in addition to conventional current inputs. We examined in the details how altering the fraction of common inputs changes the similarity of their output spike trains, in pairs of unconnected neurons. 19 Cell types and electrophysiological responses. We recorded in vitro from 𝑛=47 pyramida...

    work page 2013

  11. [11]

    The results of this first set of experiments are shown in Fig

    and to facilitate the comparison across cell types. The results of this first set of experiments are shown in Fig. 3: similarly to what was described previously for pyramidal cells under a current-clamp stimulation (de la Rocha et al., 2007), 𝜌© increases monotonically with 𝑐 21 under conductance-clamp stimulation, while always remaining smaller than the ...

    work page 2007

  12. [12]

    and summarized in the Materials and Methods section is suited for interpreting the experimental data. Spike-count covariance depends on cell pairs type By definition, evaluating the spike-count correlation involves computing the covariance between the spike counts from neuron pairs and then normalizing by their respective variances. Linear response theory...

    work page 2014

  13. [13]

    For sufficiently irregular spike trains the absolute magnitude of spike-count covariation increases with the length of the window 𝑇

    to span (geometric) mean firing rates of the pairs of cells over the range [0,25] 𝐻𝑧. For sufficiently irregular spike trains the absolute magnitude of spike-count covariation increases with the length of the window 𝑇. It is then convenient to examine the values of spike-count covariance for a given value of 𝑇 while comparing two conditions, e.g., differe...

    work page 2007

  14. [14]

    operating point

    (see Methods). For large values of 𝑇, the window over which the spike counts 𝑛9©,𝑛=© are estimated, this theory reduces to: (1) lim©→ÄCov(𝑛9©,𝑛=©)𝑇≈𝑐𝜎=∙gain9∙gain= where gain9 and gain= are the slopes of the frequency-current curves of the two neurons. Indicating by 𝜇9 and 𝜇= the mean amplitudes of the current inputs experienced by the two neurons, this i...

    work page 2008

  15. [15]

    8B) (see Methods for details)

    were matched to the passive and firing rate characteristics of the experimentally measured pyramidal, FS, and NON-FS cells (Fig. 8B) (see Methods for details). Mimicking the experiments, we employed a stochastic bombardment of presynaptic excitatory and inhibitory conductance inputs to drive stochastic spike train responses (Fig. 8B). The simplicity of th...

    work page 2001

  16. [16]

    put forth a theory that related the gain of single neuron input-output transfer to how correlated input fluctuations are transferred by neuron pairs to output spike-count correlations (Eq. 23). However, in that study, as well as subsequent ones, in vitro experiments were only qualitatively compared to theory. For instance, while the dependence of spike-co...

    work page 2008

  17. [17]

    or excitability class (Barreiro et al., 2010, Hong et al.,

    work page 2010

  18. [18]

    blanket of inhibition

    of membrane integration determines the timescales over which correlations are transferred, nevertheless experimental tests were only qualitative in nature (e.g., an increase or decrease in correlation as membrane properties are changed). The difficulty with a quantitative test of a linear response theory of correlations is that a systematic measurement of...

    work page 2014

  19. [19]

    and references therein). Interestingly, it has recently been shown that a class of GABAergic cells in the prefrontal cortex sends long-range projections to subcortical areas (Lee et al., 2014): it would be of great interest to investigate whether these cells present correlation transfer properties that differ from what we have described here. 30 In conclu...

    work page 2014

  20. [20]

    Phys Rev Lett, 86, 2186-9

    Effects of synaptic noise and filtering on the frequency response of spiking neurons. Phys Rev Lett, 86, 2186-9. 31 Buzsáki G and Mizuseki K (2014) The log-dynamic brain: how skewed distributions affect network operations. Nature Review Neuroscience 15(4): 264–78. Chance, F. S., Abbott, L. F. & Reyes, A. D

    work page 2014

  21. [21]

    Neural Comput, 14, 2057-110

    Dynamics of the firing probability of noisy integrate-and-fire neurons. Neural Comput, 14, 2057-110. Freund, T. F. & Katona, I

    work page 2057

  22. [22]

    Nature, 491, 599-602

    Synaptic amplification by dendritic spines enhances input cooperativity. Nature, 491, 599-602. Hengen KB, Lambo ME, Van Hooser SD, Katz DB, Turrigiano GG (2013) Firing rate homeostasis in visual cortex of freely behaving rodents. Neuron 80(2):335-42. Henze, D. A. & Buzsaki, G

    work page 2013

  23. [23]

    Cereb Cortex, 18, 2086-97

    The dynamical response properties of neocortical neurons to temporally modulated noisy inputs in vitro. Cereb Cortex, 18, 2086-97. Konig, P., Engel, A. K. & Singer, W

    work page 2086

  24. [24]

    Nat Rev Neurosci, 5, 793-807

    Interneurons of the neocortical inhibitory system. Nat Rev Neurosci, 5, 793-807. Hanno S. Meyer Verena C. Wimmer M. Oberlaender Christiaan P.J. de Kock Bert Sakmann Moritz Helmstaedter (2010) Number and Laminar Distribution of Neurons in a Thalamocortical Projection Column of Rat Vibrissal Cortex. Cerebral Cortex 20(10):2277–86. Naud, R. & Gerstner, W

    work page 2010

  25. [26]

    Same analysis as that performed in Fig

    Correlation between the product of the gain of the f-I curves and the measured covariance values, in mixed cell pairs. Same analysis as that performed in Fig. 9, but for pairs of cells composed of different cell types. Also in this case there is a very strong correlation between the experimentally measured values of covariance and the product of the slope...

    work page arXiv