Axons compensate for biophysical constraints of variable size to uniformize their action potentials

János Brunner; Antónia Arszovszki; Gergely Tarcsay; János Szabadics

doi:10.1371/journal.pbio.3002929

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

DeepDyve requires Javascript to function. Please enable Javascript on your browser to continue.

Axons compensate for biophysical constraints of variable size to uniformize their action potentials

Brunner, János;Arszovszki, Antónia;Tarcsay, Gergely;Szabadics, János 2024-12-02 00:00:00 a1111111111 AU Active : Plea conductances seconfirmthata tune llhethe adingl kinetics evelsarere of axonal presente action dcorrect potentials ly: (APs) to support special- a1111111111 ized functions of neuron types. However, the temporal characteristics of voltage signals strongly depend on the size of neuronal structures, as capacitive and resistive effects slow down voltage discharges in the membranes of small elements. Axonal action potentials are particularly sensitive to these inherent biophysical effects because of the large diameter var- OPENACCESS iabilities within individual axons, potentially implying bouton size-dependent synaptic Citation: Brunner J, Arszovszki A, Tarcsay G, effects. However, using direct patch-clamp recordings and voltage imaging in small hippo- Szabadics J (2024) Axons compensate for biophysical constraints of variable size to campal axons in acute slices from rat brains, we demonstrate that AP shapes remain uni- uniformize their action potentials. PLoS Biol form within the same axons, even across an order of magnitude difference in caliber. Our 22(12): e3002929. https://doi.org/10.1371/journal. results show that smaller axonal structures have more Kv1 potassium channels that locally pbio.3002929 re-accelerate AP repolarization and contribute to size-independent APs, while they do not Academic Editor: Eunjoon Kim, Institute for Basic preclude the plasticity of AP shapes. Thus, size-independent axonal APs ensure consistent Science, REPUBLIC OF KOREA digital signals for each synapse within axons of same types. Received: April 19, 2024 Accepted: November 6, 2024 Published: December 2, 2024 Peer Review History: PLOS recognizes the benefits of transparency in the peer review Introduction process; therefore, we enable the publication of Axons broadcast information about neuronal activity in the form of digital action potentials all of the content of peer review and author responses alongside final, published articles. The (APs), whose characteristics align with the specific functions of neuronal cell types. Broad and 2+ editorial history of this article is available here: narrow APs differentially trigger Ca -influx at synapses, thereby contributing to the cell type- https://doi.org/10.1371/journal.pbio.3002929 specific strength and plasticity of synaptic responses [1–3]. These precisely tuned presynaptic Copyright:© 2024 Brunner et al. This is an open APs are usually different from somatic APs and they rely on cell type-specific functions of axo- access article distributed under the terms of the nal ionic conductances [4–14]. The diversity of axonal K channels plays particularly impor- Creative Commons Attribution License, which tant roles in cell type-specific tuning, determining the speed of repolarization, which controls permits unrestricted use, distribution, and 2+ 2+ the duration of synaptic Ca channel opening and the driving force for Ca influx [4,15,16]. reproduction in any medium, provided the original However, in addition to the active components of APs, passive membrane properties also author and source are credited. influence AP waveforms. Fast voltage changes are constrained by biophysical principles deter- Data Availability Statement: Raw source data and mined by the small and variable axon diameter [17–20]. Axons are not simple RC circuits, codes are available in publicly available where smaller membrane capacitance compensates for larger input resistance, because the var- repositories. Data: https://repo.researchdata.hu/ dataset.xhtml?persistentId=hdl:21.15109/ARP/ iable surface-to-diameter ratio causes an insistent impedance mismatch. As a result, APs are PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 1 / 21 PLOS BIOLOGY Uniform axonal action potentials BTK8A4 Codes: https://zenodo.org/records/ expected to be more heavily filtered in smaller axonal structures. The consequential limitations 13977231; https://github.com/brunnerjanos/ of these filtering effects on the rise and propagation of axonal APs have been recognized for axonal-AP/tree/1.0. All other relevant data are decades, and the accepted view now is that they are compensated by local regulation of Na within the manuscript and supporting Information channels and optimal morphology [2,18,21–25] or circumvented by myelin sheaths [26,27]. files. However, Na channels have little contribution to the repolarization phase that is the most Funding: This project has received funding from critical for the digital output functions. the European Research Council (ERC) under the 2+ The duration and decay of APs determine Ca influx into the pre-synapse, initiating the European Union’s Horizon 2020 research and translation of digital APs into analog synaptic responses. Due to the fast voltage-dependent innovation programme (Grant agreement No. 2+ 772452, nanoAXON to J.S.). This work was also operation and efficient activation of synaptic Ca channels, AP repolarization-dependent 2+ supported by the Ja´nos Bolyai Research deactivation and driving force control the amount of Ca influx, which exponentially deter- Scholarship of the Hungarian Academy of Sciences mines synaptic release [5,8,28–30]. Hence, even minor changes in AP shape exert a large (J.B. and J.S.). The funders had no role in study impact on the evoked responses [2,3,16]. design, data collection and analysis, decision to However, the variability in size and geometry of individual synaptic boutons often depends publish, or preparation of the manuscript. on properties or states that are not directly related to synaptic efficacy, such as the presence of Competing interests: The authors have declared organelles, position along main and terminal transportation routes [31], accumulation of vesi- that no competing interests exist. cles [21], sudden changes in the membrane surface due to bulk endocytosis [32,33] or excess Abbreviations:A AP,U action : Anabpotential; breviatioDTX, nlisthasbeencompiledforthoseusedinthetext:Pleaseverifythatallentriesarecorrect: ionic changes during strong activity [34]. In the synaptic target area, unmyelinated axonal dendrotoxin-I; LMFB, large mossy fiber bouton; diameters typically range between 0.1 and 5 μm, while synaptic boutons within the same axons sMF, small mossy fiber. also exhibit several-fold differences, resulting in quadratically larger membrane surfaces. Con- sequently, size variability can substantially complicate synaptic effects by uncontrolled proper- ties that passively influence AP shapes. Therefore, synapse-to-synapse variability in AP decay would not only compromise the cell-type specificity of axonal signaling but also raise questions about whether variable APs within a single axon can truly be considered as real digital signals. Early theoretical studies predicted that due to the different biophysical properties caused by variable size, the electrotonically isolated axon compartments elicit APs with variable repolari- zation speed [17,19], implying variable trigger signals for each synapse. However, this question was not experimentally addressed previously. Thus, it remained unknown whether the vari- ability of AP decay between individual boutons of single axons is compensated or provides an additional level of synaptic diversity. Therefore, here we employed direct patch clamp record- ings and voltage imaging at unprecedented temporal, morphological, and spatial resolution in various types and sizes of hippocampal axons to understand the size dependence of axonal AP signaling. We focused on the decay phase of APs because it has a fundamental impact on out- 2+ put function of axons by regulating Ca influx and has a different correlation to axon size than the extensively studied rising phase, which is crucial for reliable axonal AP propagation [2,18,21–25,35]. Our results provide strong evidence for the maintenance of size-independent AP uniformity within single axons of various types. Among them, in hippocampal mossy fiber axons, Kv1 channel density varies in boutons of different sizes in a manner that compensates for changes in AP shape otherwise expected from capacitive and resistance effects. The varying density of Kv1 channels may be related to distinct anchoring mechanisms and unusual distri- bution of these channels. Results Despite variable biophysics similar AP shape is maintained at variable caliber axons First, we explored the biophysical behavior of variable-caliber axons by simulating APs by sodium and potassium conductances that were homogeneously distributed in the morphology of a real mossy fiber axon originating from a hippocampal granule cell (Figs 1A and S1). As expected from simple morphology simulations [17–20], the passive biophysical properties and PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 2 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 1. SimilarAU AP :shapes Abbrevi inatio sMFs nlisand tshaveb LMFBs eencomp despite iledfor thetho different seusedinFi biophysic gs1to5: alPlproperties. easeverifythat (A)alle Simulati ntriesa ons recorr in real ect: MF morphology with uniform I and I predict that AP kinetics, measured as AP area, is variable due to the size- K Na dependent passive membrane properties. In general, smaller boutons have slower AP. A more detailed and extended versions of these simulations are shown in S1 Fig. (B) The recorded sMFs (upper image) were identified with subsequent in situ confocal imaging (lower image) or post hoc anatomy (see Methods). Inset shows an LMFB along the sMF-recorded axon that was used for its identification. LMFBs were recorded in the same conditions as sMFs. (C) Their diameters segregated the recorded structures into 2 populations. sMFs include small en passant boutons and main axons. Small dots show individual measurements, large black symbols are the mean, black vertical lines are the median, and diamonds and whiskers show 25%–75% and 10%–90% ranges, respectively. (D) The sMFs had smaller membrane capacitance, larger input resistance and slower membrane time constant compared to LMFBs (n = 50 sMF and 64 LMFBs). (E) APs were evoked by small current injections. (F) AP amplitudes and width at half-maximum (HW) were similar both in the raw recordings and after elimination of the instrumental filtering effect by post hoc simulation of the recording equipment and biological structures [38] (n = 15 sMF and 11 LMFBs). (G) The integral area of APs was similar in sMFs and LMFBs and did not correlate with the capacitance of individual boutons, which measures membrane surface and size of the recorded structures (n = 50 sMF and 64 LMFBs). Source data and codes available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g001 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 3 / 21 PLOS BIOLOGY Uniform axonal action potentials the waveforms of active conductance-evoked APs changed with the size of the axons. Both the rise and repolarization were generally slower in smaller structures. However, local AP shapes were also influenced by adjacent axonal regions, as sudden widening of the diameter can act as a capacitive sink. While the effects of size and capacitance mismatches on AP propagation (i.e., AP rise/sodium channel activation) have been previously investigated [18,21–26], how these biophysical constraints affect AP repolarization remains unknown. Modeling the subsequent functional steps, the differences in AP shapes were multiplied by the nonlinear correlations 2+ 2+ between AP shape and Ca influx, and between Ca influx and synaptic release, resulting in a 40% variability in release efficacy. The results showed that—consistent with previous predic- tions [17,19,20]—the biophysical consequences of variable axon diameter can lead to changes 2+ in AP repolarization in a range that can affect Ca influx, which can add complexity to neuro- nal outputs. Although this simplified simulation did not capture all possible details, it provides a testable prediction: AP repolarization and shape are generally slower in smaller axons, and they are also variable due to mismatch effects at diameter changes. Next, we tested AP shapes in mossy fibers using direct patch clamp recordings from small axon terminals. These axons form small boutons and connecting segments, whose sizes match most cortical axons, and exceptionally large boutons [3,36]. Mossy fibers lack complex arbori- zation and myelin sheaths, which would complicate their biophysical properties. Only unequivocally identified and intact mossy fibers were analyzed (Fig 1B), categorized either as small mossy fibers (sMFs, including both small boutons and main axons) or large mossy fiber boutons (LMFBs) (Fig 1C). Cut ends, called blebs [1], were not included. As expected from the smaller size, the membrane capacitance was smaller in sMFs, whereas their input resistance was larger. The measured membrane time constant was also slower in sMFs at resting condi- tions (Fig 1D and S1 Table) that would suggest that APs are also size dependent. However, the amplitude and kinetics of locally evoked APs were surprisingly similar in the raw record- ings (Fig 1E and 1F and S1 Table). This may be the consequence of the comparable size of the structures and recording pipettes [1,37–39], which were similarly small in all recordings. As these parameters can be influenced by instrumental filtering caused by the small pipette required for patching axons, we quantified the integral area of APs. Our analysis showed that the integral area of APs was not correlated with bouton identity or the effective size of the recorded mossy fiber structure, measured by local capacitance (Fig 1G). Interference between small axons and experimental instruments affects these recordings [1,37–39]. Therefore, in a subset of experiments, we corrected these signals to investigate the native, instrument-free AP shapes by correcting current clamp errors through composite sim- ulation of the morphology, currents, and the recording instruments [38]. Importantly, the cor- rected APs of sMFs and LMFBs remained indistinguishable (corrected data in Fig 1, S1 Table). The smaller variance of the corrected data compared to the raw data confirms minimal biological variability of APs. In contrast to the models that predicted variable AP decay when active conductances were homogeneously distributed, our results using corrected patch clamp data suggested that despite the complex biophysical consequences of variable axon size, APs in sMFs are similar to LMFB APs. Additionally, the corrected voltage data confirmed that AP area measurements are insensitive to instrumental filtering (Fig 1F). Simultaneous imaging of sMFs and LMFBs confirms AP uniformity within individual axons To test the AP shapes simultaneously in various axonal compartments, we used fast voltage- sensitive dye imaging which allows simultaneous monitoring of APs in different structures of the same axons at sufficient temporal resolution. Voltage imaging of distal axons was PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 4 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 2. Voltage-sensitive dye imaging confirms similar AP shapes in sMFs and LMFBs. (A) JPW1114 (50 μM) was loaded via recording of an LMFB. The pipette was retracted and after equilibration (�1 h), fluorescence signal changes were monitored at 10 kHz in the vicinity of the recording site, while APs were evoked by distal stimulation (>100 μm). (B) Electrophysiological traces were recorded at the beginning of the recordings, independent of the imaging data. Gray traces show individual experiments, with colored average APs from all sMFs and LMFBs (n = 11 experiments). Black trace shows the average signal from pixels right next to the imaged axons. (C) AP parameters from simultaneously imaged LMFBs and sMFs. (D) In the second imaging approach, the loading pipette was kept on the LMFBs allowing the over-sampling of imaged signals by the high-resolution recordings (n = 18 experiments). Amplitudes are not comparable because the presence of the excess dyes in LMFBs reduce signals by increasing non- responding dye fraction (i.e., F ). (E) Voltage imaging revealed no difference between the local APs of 2 distinct sMF structures, the main axon (main) and the filopodial extensions (filo) of the same axons. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g002 challenging because achieving sufficient concentrations of the membrane bound dye at dis- tances of 100 to 1,000 micrometers results in toxically high concentrations at the somatic load- ing sites. To overcome these limitations, we directly loaded JPW1114 into LMFBs and their neighboring axon segments with sMFs. This dye is fast enough to resolve the rapid time course of axonal APs [1,11]. First, we investigated naturally propagating APs evoked by distal stimula- tion. The dye-loading pipette was subsequently retracted (Fig 2A). The shapes of the simulta- neously imaged APs of sMFs and LMFBs were indistinguishable, and they also matched the previously recorded APs (n = 11; Fig 2B and 2C and S1 Table). Because this measurement is limited by the 10 kHz imaging rate, in the next experiments, we used higher temporal resolution with the guide of simultaneous electrophysiological signals (Fig 2C). We kept the dye-loading pipette on the LMFBs during imaging and oversampled the JPW1114 signals of locally evoked APs by aligning individual traces to the 100 kHz electro- physiology signals before averaging for distinct axonal compartments. This arrangement reli- ably reports the temporal profile of APs, although the amplitudes are not comparable due to the presence of the excess, non-responding dye in the pipette on the LMFB (i.e., F ). Impor- tantly, this measurement also confirmed uniform AP shapes in sMFs and LMFBs (n = 18, S1 Table). Furthermore, simultaneous imaging revealed similar AP shapes in 2 distinct types of PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 5 / 21 PLOS BIOLOGY Uniform axonal action potentials sMF structures: the filopodia emanating from LMFBs and main axons with small en passant boutons (Fig 2E and S1 Table). Furthermore, we did not detect differences in the APs of bas- soon-positive and -negative small axonal segments (identified by correlated post hoc immuno- labeling, S2 Fig) suggesting that AP shape is similar in synaptic and non-synaptic segments. Altogether, we concluded from our independent measurements that AP shape is uniform within individual mossy fibers despite the large intrinsic variabilities of their diameters and biophysical properties. Cell type-specific but uniform APs in various axons Next, we tested whether the AP uniformity applies to other cell types. We measured the APs and biophysical properties in diverse types of axons, representing the main cortical axon types, including subcortical and locally originated glutamatergic and GABAergic types. The axons were rigorously identified by using their morphological properties and immunolabeling for 4 to 5 markers on each axon (see Figs 3A, 3B, S3, S4 and S5). First, we confirmed in these axons that the membrane properties and AP shapes are axon type specific (Fig 3C and S2 Table), val- idating the principle that axonal functions are cell type-specifically tuned. Axons from hypo- thalamic supramammillary nucleus (SuMa) elicited wider APs than mossy fibers (average AP area: 25.8 ± 1.2 mV*ms, n = 16), while the axonal APs of hilar mossy cells (21.6 ± 3.6 mV*ms, Fig 3. AP uniformity in diverse types of hippocampal axons. (A) A representative recording from a morphologically identified VGluT2+ SuM axon in the DG (further details of identification are shown in S3 Fig). (B) A representative recording from a morphologically identified VGluT1+ mossy cell axon (see also S4 and S5 Figs). (C) Average electrical −5 −16 properties of axons are cell type dependent (ANOVA: F(4) = 4.45, 7.5, and 27.4, p = 0.002, 1.6*10 and 1*10 , Bonferroni pairwise comparison is shown in S2 Table). (D) Within cell types, the AP shapes did not show correlation with the local size and membrane time constant of the axons (statistical details of linear fits are shown in S3 Table). Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential. https://doi.org/10.1371/journal.pbio.3002929.g003 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 6 / 21 PLOS BIOLOGY Uniform axonal action potentials n = 7, MCa) and CB1R-expressing GABAergic axons (18.4 ± 1.8 mV*ms, n = 8, CB1a) were between SuMa and mossy fibers (LMFB: 15.5 ± 0.4 mV*ms, n = 64, sMF: 14.8 ± 0.5 mV*ms, n = 50). The cell type-specific APs also demonstrate that the observed uniformity of mossy fiber APs was not due to the insensitivity of recording methods. These axons, like most in the CNS, have variable diameter due to varicosities. However, within anatomically homogeneous cell types, the AP shape did not depend on the size and pas- sive membrane properties of the recorded axonal structure (Fig 3D and S3 Table). These results suggest that all tested axon types employ uniform APs in spite of their variable size and biophysical properties. Size-independent AP plasticity Axonal AP waveforms are not always the same within individual boutons and their shapes exhibit functionally relevant plasticity, as axonal APs typically show activity-dependent broad- ening during sustained activities influencing their synaptic effects [5,6,8,12–14,30,40,41]. To assess whether LMFB and sMF APs maintain uniformity during various dynamic conditions, we conducted direct recordings and VSD imaging (Fig 4A, 4B, 4D and 4E). We found that sMF APs, similarly to previous data from LMFBs [5], display frequency-dependent broadening during AP trains. Importantly, the extent of AP plasticity was similar in sMFs and LMFBs and did not depend on the size of the structure (Fig 4C and 4F). Thus, AP uniformity persists dur- ing dynamic conditions allowing for similar activity-dependent modulation of axonal output. Active conductances underlie uniform APs There must be mechanisms that compensate for the passive biophysical constraints caused by axon size. In the simulations that recreated original recordings of single APs by adjusting the amounts of active currents in the detailed model of an axon with instruments (see Fig 1F), both sMFs and LMFBs required a similar amount of sodium currents (I ) on average. How- Na ever, sMFs needed more potassium currents (I ) than LMFBs. This predicts a lower I /I K Na K ratio in sMFs (Fig 5A). To test this prediction, we experimentally measured the I /I ratio in Na K outside-out patch recordings, where both I and I were measured in each patch using 2 dif- Na K ferent voltage steps (Fig 5B). Consistent with the model prediction, while I densities were Na similar, the total I were larger in patches pulled from sMFs, resulting in larger I /I ratio K Na K (n = 46 sMFs and 47 LMFBs, for statistical details see S1 Table and Fig 5C). As outside-out patches sample ionic channels from local membranes only, we concluded that sMFs have more I , which speed up their AP repolarization and promote AP uniformity. Among axonal Kv channels, Kv1s are known to be responsible for fine-tuning APs for spe- cific functions of various axons [4,6–9,13,14,40,42], whereas Kv3 and BK channels usually pro- vide the backbone of repolarization in mossy fibers [43] and Kv7 channels set axonal excitability that also influence spiking properties [44]. However, the contribution of different Kv channels changes during complex firing patterns. Interestingly, electron microscopic stud- ies noted that Kv1.1 channel distribution was not homogeneous in mossy fibers and the expression was more prominent in small axon compartments [45]. Among these channels, we focused on the potential role of Kv1 channels in uniformizing single axonal APs, we inhibited Kv1.1 and Kv1.2 channels with dendrotoxin-I (DTX), which resulted in wider APs in mossy fibers. However, the effects of DTX were larger on AP repolarization in sMFs than on LMFBs APs (38.6% versus 21.5% at the population level, Fig 5D), indicating that sMFs employ more Kv1 channels to achieve fast repolarization. In the presence of DTX, axon size and AP shape showed a similar correlation as predicted in axon models with homogeneous conductances (S1E Fig). This observation suggests the non-uniform distribution of AP-relevant potassium PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 7 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 4. AP shape plasticity affects sMFs and LMFBs similarly. (A) AP trains were evoked at 3 different frequencies during direct patch clamp recording of sMFs and LMFBs. The example traces show 40 Hz stimulation, and 3 ms-long current injections pulses were used to avoid the contamination of the repolarization phase by the capacitance artifact. (B) Progressive and frequency-dependent broadening of sMF- and LMFB-APs (see also [5]). (C) Summary data show similar broadening in sMFs and LMFBs at different frequencies (left panel). Furthermore, the broadening of APs did not depend on the functional size of the boutons (right panel). (D) The 100 APs were evoked at 40 Hz in axons during VSD imaging experiments. To avoid photo damage, the axons were illuminated and imaged only during the first 11 and the last 13 APs. (E) Example APs obtained by 2 different imaging approaches with extracellular stimulation (upper panels) and with simultaneous intracellular recording, which allowed oversampling of the imaging signals (see Fig 2). (F) Summary data show similar broadening in sMFs and LMFBs obtained with the 2 approaches. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g004 currents explains the difference between the measured and predicted size-dependent behavior of axonal APs. It also suggests an important role for Kv1 channels in accelerating APs in smaller MF compartments. Furthermore, the size-dependent DTX effects on APs indepen- dently confirmed our patch data about the contribution of larger I to the APs of sMFs. Indeed, in the presence of DTX, the I density and I /I ratio were similar in LMFBs and K Na K sMFs (S6 Fig). The resting membrane potential was not significantly affected by DTX (sMF control versus DTX: −75 ± 0.8 mV versus −72.7 ± 2.6 mV, LMFB control versus DTX: −- 74.4 ± 1.2 mV versus −75.1 ± 2.5 mV, ANOVA: F(3) = 0.441, p = 0.72, n = 54, 17 and 36, 17), indicating little Kv1 channel activation at resting conditions in MF axons. In contrast to the PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 8 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 5. Active conductances compensate for the biophysical constraints of smaller axon size. (A) Recreating native sMF APs (see Fig 1) in simulations required similar I but more I than in LMFBs, resulting in larger I /I ratio in Na K Na K the latter. (B) Average of recorded I and I current densities in membrane patches pulled from sMFs and LMFBs Na K (n = 48 and 40 recordings) using 2 voltage-steps in the same recordings (0 mV for 0.4 ms for I , +70 mV for I , from Na K −110 mV). (C) The average I density was similar between LMFB and sMF recordings; however, sMF membranes Na had larger I densities, resulting in larger I /I ratio in LMFBs. (D) Selective inhibition of Kv1-mediated I by 100 K Na K K nM DTX had size-dependent effects on axonal APs with larger effects on smaller sMFs. Four representative recordings are shown on the right. (E) In contrast, selective partial Kv3 inhibition by TEA (100 μM) had similar effects on APs of sMFs and LMFBs. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ ARP/BTK8A4. AP, action potential; DTX, dendrotoxin-I; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g005 DTX effects, partial inhibition of Kv3 channels by 100 μM of TEA had similar effects on sMF and LMFB APs (16.9% and 19.5%, Fig 5E). These results show that the contributions of at least 2 AP-relevant channels are differently related to axon size. While Kv3s have similar contribu- tions to single APs, the contribution of Kv1 correlates with axon size compensating the bio- physical consequences of small and variable axon size. Discussion We demonstrated that individual axon types elicit size-independent uniform APs. While theo- retical considerations and our initial simulations predicted that AP repolarization would be distorted by the local biophysical environment of varying axon size, our direct axonal patch- clamp recordings and voltage imaging showed that AP repolarization is remarkably resistant to these effects. We also found that Kv1 channels play a crucial role in maintaining axonal PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 9 / 21 PLOS BIOLOGY Uniform axonal action potentials uniform AP shapes regardless of size. These channels accelerate the repolarization phase, com- pensating for the slower membrane time constant in smaller axonal structures, where their contribution is greater than in larger axons. Having a uniform AP signal in all synapses throughout the axon of an individual cell has several fundamental advantages and conse- quences. Firstly, with coordinated uniform trigger signals, synapses can individually diversify their output, relying on the known diversity of synaptic machineries that allows pre- and post- synaptic cell alignment-, state-, or learning-dependent communications. If AP shapes varied with the size of the presynaptic boutons, the release machinery would have to adjust for the actual size of each synapse. Homeostatic regulation of the density of a single channel according to the size of boutons and uniform APs can eliminate the need for such complex mechanisms. Notably, Kv1 channels are anchored by molecular pathways that are independent from other AP-mediating channels (see below for details) and these channels have a sufficiently fast acti- vation to accelerate repolarization [7,9,13,43]. Secondly, without uniform axonal APs, the previously described cell type-specific rules of axonal signaling would be compromised. Diverse shapes of axonal APs evolved to fulfill cell type-specific functions. For example, narrow APs cope well with high-frequency firing in the cerebellum [8,12], whereas wider APs mediate reliable release in circuits, where sub-millisec- ond precision is not required [10,41]. However, the effect size of the biophysical consequences of variable axon diameter on AP kinetics are comparable to differences of APs in various cell types. Thus, if passive consequences of variable size were not compensated, cell type-specific AP shapes could make less significant functional contributions. We showed that AP unifor- mity applies to diverse cell types serving distinct physiological functions through specific AP shapes. Therefore, it is likely that axonal AP uniformity is a general feature of axons. These results are also important from a technical standpoint, because they confirm that our measure- ments are sufficiently sensitive for small differences in AP kinetics, and these recordings revealed previously unknown axonal AP properties, including those of hippocampal axons from supramammillary nucleus and hilar mossy cells. Thirdly, AP uniformity is also important for consistent effects of AP shape plasticity. Activ- ity-dependent AP broadening also falls within the same range as the expected passive effects caused by the variability of bouton size within individual axons. We found that AP uniformity applies not only to solitary APs, but small and large boutons showed similar broadening dur- ing AP trains. In this regard, one of the key aspects of AP decay in mossy fibers is the involve- ment of multiple potassium channels. Repolarization is primarily mediated by Kv3 with additional activity-dependent support from BK channels in LMFBs [43].Kv7 plays a significant role in setting the intrinsic excitability, consequentially impacting AP signaling, especially dur- ing complex firing patterns [44]. The exact role of Kv1 channel in mossy fibers was not known. Our findings show that Kv1 channels significantly contribute to AP repolarization, particularly in sMFs. Unlike Kv3, Kv1 functions correlate with bouton size or type during sin- gle APs. Kv3 and Kv7 channels share the ankyrin-spectrin pathway for axonal anchoring with sodium channels [46,47]. In contrast, Kv1 channels are anchored through the PSD93-A- DAM22-LGI1 pathway and are associated with Caspr2 proteins location dependently [48–50]. Kv1 transport is regulated by unique cargo-tag mechanisms involving CDK5 [51]. The direct interactions of PSD-93 and CDK5 with the cytoskeleton [52] may offer a potential molecular mechanism linking bouton volume to Kv1 density [53,54]. Future studies should also explore the roles of these pathways together with the structure size-dependent stoichiometry and sub- axonal localization of other potassium channels. Importantly, at least 1 type of potassium channel has size-dependent contributions, while other channels support repolarization in a size-independent manner. Although we focused on Kv1 and Kv3 channels, it is possible that other channels are also regulated by axon size and PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 10 / 21 PLOS BIOLOGY Uniform axonal action potentials their combined contribution is necessary for uniform axonal AP under various physiological conditions. The cell type-specificity of axonal signaling also implies that the mechanisms main- taining uniform axonal APs are likely diverse. In mossy fibers, Kv1 channels play a key role in this homeostatic AP uniformization. This finding is further supported by previous electron microscopic analysis showing that Kv1.1 channels are present in small axons but are generally absent from large terminals [45]. Generally, Kv1 channels are often employed for fine-tuning axonal APs [6,7,12–14,55,56]. Notably, cephalopods overcome slower biochemical mechanisms in low-temperature environments using Kv1 channels with faster opening and closing kinetics by elaborated RNA-editing mechanisms depending on the geographical latitude [57]. We measured APs only from unmyelinated segments of axon. Direct patch clamp record- ings are not possible from the myelinated axons. While SuM, MC, CB1R axons are known to have myelinated segments, in the recorded regions they were mostly unmyelinated as also sug- gested by large number of synaptic boutons. While APs can be measured in myelinated seg- ments with VSD imaging, we employed this method only for mossy fibers, which do not have myelin sheaths. APs are known to be different in the myelin-sheath-insulated and exposed membranes [26], but the insulated segments do not contribute directly to the output functions of axons [2]. Therefore, AP uniformity is not relevant in these segments. Importantly, several components of Kv1 anchoring and regulatory complexes are directly implicated in diseases. Specifically, several forms of mutations in the Kv1-anchoring LGI1 pro- teins lead to autosomal dominant temporal lobe epilepsies [48,58,59]. Furthermore, auto-anti- bodies against this extracellular protein initiate autoimmune encephalitis [60,61], which also inhibit their interaction with the ADAM22-family proteins [59]. Caspr2 proteins are also auto- immune targets, causing periferial neuromyotonia [58]. Episodic ataxias are associated with various mutations in the Kv1.1 gene [62], while loss of Kvβ, which is involved in the normal targeting of Kv1 channels, was observed in patients with severe seizures [58,63]. Axonal swell- ings are biomarkers of Alzheimer’s disease, and they potentially contribute to the development of abnormal neuronal functions in this disease [64,65]. Because previously it was not known that these proteins and features are important for maintaining uniform axonal AP signaling, future studies should clarify the contribution of disrupted size-independent AP uniformity to the development of these diseases. Methods Ethics statement All experiments were performed in accordance with the Institutional Ethical Codex, Hungar- ian Act of Animal Care and Experimentation (1998, XXVIII, section 243/1998) and the Euro- pean Union guidelines (directive 2010/63/EU), and with the approval of the National Scientific Ethical Committee on Animal Experimentation (NEBIH, PE/EA/48-2/2020) and Institutional Animal Care and Use Committee. Slice electrophysiology Hippocampal slices were prepared from young rats (P23-35, deeply anesthetized with isoflur- ane) in ice-cold artificial cerebrospinal fluid (85 mM NaCl, 75 mM sucrose, 2.5 mM KCl, 25 mM glucose, 1.25 mM NaH PO , 4mM MgCl , 0.5 mM CaCl , and 24 mM NaHCO , Leica 2 4 2 2 3 VT1200 vibratome). We used an upright microscope (Eclipse FN-1; Nikon) equipped with a high numerical aperture objective (25× NA1.1W objective, Nikon) and with a confocal scan- ning system (C1 Plus, Nikon). Recording solution was composed of 126 mM NaCl, 2.5 mM KCl, 26 mM NaHCO , 2 mM CaCl , 2 mM MgCl , 1.25 mM NaH PO , and 10 mM glucose. 3 2 2 2 4 Recordings were performed using an intracellular solution containing 90 mM potassium PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 11 / 21 PLOS BIOLOGY Uniform axonal action potentials gluconate, 43.5 mM KCl, 1.8 mM NaCl, 1.7 mM MgCl , 0.05 mM EGTA, 10 mM HEPES, 2 mM Mg-ATP, 0.4 mM Na -GTP, 10 mM phosphocreatine, 8 mM biocytin, and 20 μM Alexa Fluor 594. The temperature was kept at 34 to 36˚C during the recordings. The axonal structures in the stratum lucidum of CA3 region or in the inner layer of stratum moleculare of DG region were patched under the guidance of IR-DIC optics by using capillaries pulled from borosilicate tubing (1.5 mm outer, 0.75 mm inner diameter, Sutter Instruments). Data was collected using a MultiClamp700B amplifier and digitized by a Digidata 1440 A/D converter at 250 kHz sampling rate (Molecular Devices). Capacitance neutralization was carefully adjusted close to its possible maximum. After the recording, we captured the position of the recording electrode together with the in situ morphological characteristics of the recorded axon based on the fluorescent Alexa Fluor 594 signal by using the confocal scanning head. For detailed visualization of the morphology, we employed post hoc fluorescent streptavidin reaction. Only identified axons were included (see below) and recordings close to the cut end of the axons were excluded. For the measurement of the local electrical parameters of axons, 100 ms-long pulses from −70 mV to −90 mV were used in voltage clamp. First, we recorded the signals in cell-attached configuration, which allowed the estimation of the seal resistance in each measurement (12.1 ± 0.6 GΩ for sMF, 13.5 ± 0.6 GΩ for LMFBs, and 12.5 ± 0.9 GΩ for other axons). For the quantification of local axonal capacitance, the average of the cell-attached signal was sub- tracted from the whole-cell recorded trace to eliminate residual uncompensated instrumental capacitance. We fitted the capacitive current transients with the sum of 2 exponential functions in a 3 to 7 ms time window after the peak assuming that a two-compartmental electrical equiv- alent circuit is an adequate representation of recorded axons [66,67]. In this model, the faster exponential describes the electrical behavior of the local membrane at the recording site, whereas the slower one is the composite representation of the entire axonal arbor. To validate this two-compartmental simplification, we systematically fitted the traces with mono-, bi-, or tri-exponential functions in a subset of the experiments. The two-compartmental model signif- icantly outperformed the fit of the mono-exponential model assessed (sum of squared error: 2 2 2 2 sMF: 7.81 nA versus 2.37 nA ; LMFB: 10.58 nA versus 2.18 nA for mono- and bi-exponen- tial fits, respectively). Furthermore, fitting the traces with the sum of 3 exponentials resulted in only marginal improvement compared to the biexponential model (sMF: 2.37 nA versus 2.16 2 2 2 nA ; LMFB: 2.18 nA versus 2.04 nA for bi- and tri-exponential fits, respectively) indicating that this simplified two-compartmental model is necessary and sufficient to characterize the passive electrical properties of the recorded axons. Thus, the structure capacitance derived from the fast component of the fit and it represented the biophysically effective size of each recorded bouton. Input resistance was measured both in voltage- and current clamp record- ings to check the consistency of the recording, but for quantification we used the voltage clamp data. The membrane time constant data derived from independent current clamp recordings. To calculate the AP area in electrophysiological recording data, we first deter- mined the start of the AP at the voltage value where the upstroke velocity reached 100 mV/ms. Then, we computed the AP area within the region between the start of the AP and the point where the membrane potential returned to the same voltage value. Membrane patches were pulled from the recorded axons after the recording of whole-bou- ton properties of the structure, including the structure capacitance. The membrane area was estimated from the correlation between pipette resistance and the capacitance of the outside- out patches. To obtain this correlation, we precisely determined the capacitance of the pulled patches in the subset of experiments by pushing the pipette tip into a Sylgard ball at the end of the recording. This area was used to normalize the recorded I and I from each bouton. The Na K 2 2 average estimated areas of outside-out patches were 3.69 ± 0.14 μm and 3.84 ± 0.17 μm for sMFs and LMFBs. PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 12 / 21 PLOS BIOLOGY Uniform axonal action potentials Modeling of axonal recordings with realistic pipettes and amplifier circuits (Fig 1F) Computer simulations were performed in the NEURON simulation environment (v7.5) [68] using a detailed model that captures the experimentally observed behavior of the measuring system as described in Olah and colleagues [38]. Pipette parameters and the compensatory mechanisms of the amplifier were set according to the actual measurements using modified model-cell circuits or short-circuiting components of the amplifier. Furthermore, the pipette parameters were quantified using partial submerging, tip-breaking, and direct microscopic measurements of their physical dimensions. The pipette models were adjusted for each record- ing pipette to recreate capacitive signals obtained in open and on-cell situations. The amplifier and adjusted pipette models allowed to recreate the characteristic capacitive signals during each current clamp recording. To incorporate realistic morphologies in the model, we used a 140 μm-long detailed recon- struction of the recorded part for each axon. The reconstruction was performed semi-automat- ically by the Vaa3D software [69] using the confocal images taken right after the recordings. Remote axonal parts were implemented as simplified dummy compartments. To obtain the passive background of the axon, we fitted model responses to the experimentally recorded sub- threshold voltage signals by optimizing axonal passive parameters and the access resistance of the recording. These fundamental parameters were similar for models which were based on -2 LMFBs and sMFs (C = 0.90 ± 0.08 and 0.89 ± 0.05 μF*cm , R = 151.49 ± 18.81 and m a 134.45 ± 15.66 O*cm and R = 117.81 ± 7.63 and 185.26 ± 34.31 M).Ω R was held con- access m stant at 50 k*Ω cm . To predict the axonal AP waveforms, we incorporated active sodium and potassium con- ductance models with adjustable voltage dependence and kinetic parameters onto the model and the current clamp output was fitted to the experimental data by freely varying all channel parameters. For each axon, 10 APs from 4 different initializations were fitted and the average of the conductance parameters yielded by these 40 independent optimizations was considered as the best fit options. Fits were performed using the PRAXIS algorithm of the NEURON’s built-in Multiple Run Fitter. In addition to recreating the recorded AP signals and instrument related capacitive artifacts, these fits predicted the underlying g and g properties as well. Na K Finally, predicted native AP signals were obtained by running these constrained g and g in Na K realistic axonal morphology without the inclusion of the pipette and amplifier circuit models. To generate experimentally testable currents, these g and g were used to generate currents Na K in response to voltage steps in individual axons. NEURON codes are available at Zenodo: https://zenodo.org/records/13977231 and Github: https://github.com/brunnerjanos/axonal-AP/tree/1.0. VSD imaging The tip of the pipettes was backfilled with normal intracellular solution (see above) and the voltage sensitive dye (JPW1114 also called di-2-ANEPEQ, 10 to 50 μM [70]) containing intra- cellular solution was carefully loaded thereafter. LMFBs were patched within 4 min after the loading of the pipette. After forming whole cell configuration, the dye was allowed to diffuse into the axon at least for 30 min before starting the imaging experiment. The upright micro- scope (Nikon FN-1) was equipped with a 1.1 numerical aperture 25× water-immersion objec- tive and a 1.5× magnifier lens in front of the camera. For imaging voltage signal, we used a single mode 532 nm laser source (CNI Changchun MSL-FN-532S-400mW), which was cou- pled via a bifurcated SMA-905 fiber guide (0.4 mm, NA 0.39) to the epifluorecent attachment of the microscope, which was equipped a filter cube (ZT532rdc dichroic mirror and 542 nm PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 13 / 21 PLOS BIOLOGY Uniform axonal action potentials long-pass emission filter). When the axonal objects needed to be illuminated for longer time, for example during focusing and searching, we used a diode laser illuminator (89North LDI-7) via the second branch of the fiber guide. The excitation intensity was regulated by the direct control of the output power and/or by the use of ND filters. Optical signals were acquired with a CMOS camera (Redshirt Imaging, Da Vinci-1K) with a frame rate of 10 kHz (64 × 46 pixels, pixel size: 0.79 μm) and digitized simultaneously with the electrophysiological signals by the AD board of the imaging system. At the end of the experiment, the VSD fluorescence was used also for morphological identification. Analysis of VSD signals Data were discarded if the half-width of the APs showed more than 20% increase of its initial value. A custom Fiji macro was used to correct for photo-bleaching in individual pixels of the image stacks using independent double exponential fits. The pixels were clustered into 8 bins for the fitting. To select the active axonal compartments within bleach corrected image stacks, we generated aΔF/F image, where the intensities around (±0.2 ms) the AP peaks were sub- tracted and divided with the baseline values of each pixel on the image. The identification was helped by the multifocal morphological images that were recorded after the imaging session. In the first imaging approach in which the naturally propagating APs were imaged, the 10 kHz imaging traces were segmented and aligned with custom Python scripts. For the second imag- ing approach that intended to increase the temporal resolution of the imaging signals, we implemented a shift and mean algorithm (written in Python) to increase the temporal resolu- tion of the recorded signals (adopted from [71]). In brief, the higher sampling rate of the electrophysiology signals registered simultaneously on the DigiData AD-board allowed us to align the spikes with 250 kHz temporal resolution. The oversampled imaging signals were pooled, aligned according to these more precise AP peaks and then resampled by averaging multiple (typically 8) data points resulting in 24.5 to 27.5 kHz final resolution. For demonstra- tion purposes, signals were low-pass filtered using a bidirectional digital Chebyshev-II lowpass filter (f_stop = 8 kHz, Microcal OriginPro). If the imaging experiment included longer axons, VSD signals were averaged in short axonal segments (<5 μm). The signals from these seg- ments were averaged for each axon for comparison of sMF and LMFB. VSD signals are repre- sented and were analyzed without filtering, except for the half-width measurements, where the signals were low-pass filtered (Butterworth) in 2 steps: 10 kHz and 100 kHz for over-sampled signals, 2 kHz and 20 kHz for extracellular stimulation-evoked APs. Area for background mea- surements (Fig 2B) was selected 3 to 7 pixels distance from the imaged LMFB, which had the brightest emission. Because of the lack of AP signals and intensity changes in the background area, we concluded that neighboring structures (i.e., LMFB and sMF) can be independently analyzed, which were selected with similar distances. AP area was determined as the integral area in a 0.6 ms window starting from 0.05 or 0.1 ms from the peak of peak-normalized AP sig- nals. Normalization was needed because VSD signal amplitude depends on the background intensity, which is influenced by other factors in some cases (e.g., pipette dye load). Anatomical identification of whole-bouton recorded axons Recorded axons were loaded via the pipette with biocytin and in most cases with Alexa Fluor 594. The fluorescent dye allowed in situ imaging of the axons after their recordings and identi- fication of the recordings in subsequent morphological analysis (Nikon C1 scanning confocal head or RedShirtImaging CMOS Da Vinci-1K camera with eplifluorescent Sutter LED illumi- nation). In case of sMF recordings, axons were identified by unbranching axon that run paral- lel in the stratum lucidum where they form typical LMFBs. Biocytin labeling and subsequent PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 14 / 21 PLOS BIOLOGY Uniform axonal action potentials fluorescent streptavidin reactions allowed a more detailed visualization of the morphology and immunolabeling. Slices were fixed in 2% paraformaldehyde and picric acid solution for at least 20 h and then they were resectioned (40 to 60 μm). Recorded axons were stained for biocytin using Alexa-conjugated streptavidin (1:500, Thermo Fisher, S11227). The recorded axons were typically present in 3 to 5 sections. Short axons were excluded from the analysis. To maximize the capacity for immunolabeling for the recorded axons, each section was individually stained with different combinations of 2 or 3 primary antibodies for 72 h: anti-VGluT2 (guinea pig, 1:2,000, Millipore, AB2251, RRID: AB_1587626), anti-VGluT1 (guinea pig, 1:1,000, Millipore, AB5905, RRID: AB_2301751), anti-VGAT (rabbit, 1:1,000, Synaptic System, 131–003, RRID: AB_887869), anti-CB1R (rabbit, 1:1,000, Cayman, 10006590, RRID: AB_10098690), anti-par- valbumin (goat, 1:2,000, Swant, PVG-213, RRID: AB_2650496), anti-parvalbumin (guinea pig, 1:2,000, Swant, GP72, RRID: AB_2665495), anti-calretinin (goat, 1:2,000, Swant, CG1, RRID: AB_10000342), anti-bassoon (mouse, 1:3,000, Abcam, ab82958, RRID: AB_1860018). Fluores- cent dye-conjugated (AlexaFluor488, AlexaFluor647, DyLight405) secondary Donkey antibod- ies were incubated overnight. Sections were mounted in Vectashield or Prolong Glass. The slices were examined with confocal microscopy (Nikon C2 scanning confocal). Image stacks were acquired initially at 0.08 μm/px (XY) and 0.125 μm (Z direction) resolution (objective 60×, 1.4 NA) and then deconvolved (Huygens software). The resolution was lower for axonal reconstructions (0.12 μm/px XY and 0.3 μm/px Z), which were performed with Neurolucida software (version 2020.2.2, MBF Bioscience LLC, Williston, Vermont, USA, RRID: SCR_001775). Bassoon spots were manually assigned as synaptic sites in individual pixels of VSD-imaged mossy fibers using 3D analysis in Nikon NIS Elements software (version AR 4.13.04). 2+ Modeling of axonal APs, Ca influx, and release rate (Fig 1A and S1 Fig) To model the relationship between axonal morphology and the AP properties and the evoked synaptic release, a propagating AP was simulated in a complete granule cell reconstruction that had intact mossy fiber trajectory in the CA3 field. This morphology was used previously [38], but here we implemented a more complete electrical simulation on this structure. Both -2 2 the passive parameters (C = 1 μF*cm , R = 150 O*cm, and 50 k*Ω cm ) and active conduc- m a -2 -2 tances (290 mS*cm and 17 mS*cm for the Hodgkin–Huxley type sodium and potassium channels, respectively) were distributed homogeneously in the entire cell. Equilibrium poten- tials for the leak, I and I were set to −80 mV, −90 mV, and 70 mV, respectively; whereas, K Na 2+ 2+ voltage dependence of the Ca current followed the Goldman–Hodgkin–Katz [72]. Ca influx was modeled as a compound current produced by the mixture of P/Q, N, and R-type 2+ channels with a ratio of 6.5:2.5:1, respectively, as described in LMFBs [73]. Ca channel kinet- ics were adopted with a temperature coefficient of Q = 2.5 to account for the higher tempera- 2+ ture used in this model (35˚C). Release was calculated from the Ca influx assuming a conservative 2.5-power relationship. A single AP was triggered through somatic current injec- 2+ tion and parameters related to the propagating AP and the evoked Ca influx was quantified at every micrometer along the axon. As there is no explicit voltage threshold for a propagating AP, the duration of the AP was quantified as the width at −20 mV. This simulation was per- formed with a temporal resolution of 250 kHz. Supporting information 2+ S1 Fig. Simple simulations predict large variability of AP shape, axonal Ca influx, and synaptic release along the variable caliber axons if homogeneous conductances are present. (A) Theoretical rules—derived from passive RC properties of neuronal membranes—defining PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 15 / 21 PLOS BIOLOGY Uniform axonal action potentials the dependence of the speed of neuronal signals, presumably including axonal AP shape, on the size of the structure. (B) APs were simulated in a 3D reconstructed mossy fiber by using homogeneously distributed Hodgkin–Huxley sodium and potassium conductances and shown along the axon with variable diameter. Only the data from the middle section of the 2+ 2+ axon is shown. Ca influx was simulated by using a realistic set of Ca conductance [73]. 2+ Synaptic release related to the integrated Ca influx on the 2.5× power. (C) Variable AP kinet- ics measured as rate of rise and decay depending on the biophysical size of simulated axon compartment. The rise and decay phases of APs show the expected correlation at population level; however, this correlation is variable between individual boutons due to the influence of neighboring axonal regions. (D) Area of simulated APs along the entire axon demonstrate the large variability and axon size-dependence of AP shape. Simulated APs were measured at every 1 μm and were plotted against the membrane capacitance of the same spots. AP area was the integral of the repolarization phase only to avoid contamination of the measurement by 2+ effects on the rise phase of the APs [17–20]. (E) AP area and simulated Ca influx at 12 bou- tons that potentially form synaptic contacts within a short segment (linear fits AP area: R : 0.249, ANOVA p > F: 0.057, slope: −4.37. I : R : 0.287 ANOVA p > F 0.042, slope: 0.261). Ca Colors of traces scales with the capacitance of the boutons. (F) To demonstrate the reliability of AP area measurements compared to classical AP kinetic parameters, we compared raw electrophysiological recordings of APs with their corrected waveforms (see Fig 1F). The corre- lation between the half-width of APs (HW) and AP area was weaker in raw AP measurements compared to the corrected APs (R : 0.389 vs. 0.814, left panels). Furthermore, in theory, HW and AP area values should linearly correlate. However, this correlation was significant only in 2 2 the case of corrected APs (R : 0.47), but not in the raw data (R : 0.017). Raw data and simula- tions are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ ARP/BTK8A4. (TIF) S2 Fig. AP shapes in axon segments that form synapses and connecting parts. (A) Synaptic regions were identified by presynaptic marker, bassoon (BSN) positivity. Individual pixels were assigned by matching the post hoc reconstruction of biocytin labeled segments that were subject of VSD imaging. Only neighboring segments were included to avoid the contamina- tion by conduction delay. LMFBs were excluded. (B) Confocal image analysis of the bassoon puncta in the biocytin labeled axon. (C) Average APs from 4 experiments with correlated syn- apse identification. (D) Comparison of the half-widths of APs from sMF segments identified as synaptic or non-synaptic based on BSN staining. Source data are available at https://repo. researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S3 Fig. Identification of axons patched in the DG originating from the supramammillary nucleus (SuMa). Two examples are shown with the reconstructions of the biocytin-labeled axons and immunolabeling for typical SuM markers and proteins that are present in other axons. SuM axons restricted to the granule cell layer and inner moleculare layer of the DG (3D-represented by light and dark gray areas). Original images are available at https://repo. researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S4 Fig. Identification of axons patched in the DG originating from the hilar mossy cells (MCa). Two examples are shown with the reconstructions of the biocytin-labeled axons and immunolabeling for typical MC markers and proteins that are present in other axons. MC axons restricted to the inner molecular layer of the DG. Original images are available at PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 16 / 21 PLOS BIOLOGY Uniform axonal action potentials https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S5 Fig. Identification of axons originating from the GABAergic CB1-receptor expressing cells (CB1a). Two examples are shown with the reconstructions of the biocytin-labeled axons and immunolabeling for typical GABAergic cell markers and proteins that are present in other axons. Original images are available at https://repo.researchdata.hu/dataset.xhtml? persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S6 Fig. When Kv1 components are eliminated, sMFs and LMFBs have similar I currents and I /I ratio. (A) Average of I current densities in membrane patches pulled from sMFs Na K K and LMFBs in the presence of DTX (gray traces) and in control conditions. (B) The average I and I densities and I /I ratio were similar in LMFB and sMF recordings in the presence Na K Na K of DTX. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId= hdl:21.15109/ARP/BTK8A4. (TIF) S1 Table. Statistical results of comparison of passive and AP properties of sMFs and LMFBs. Mean and median data with SEM and Q1–Q3 ranges are shown in figures together with individual data points. (PDF) S2 Table. Statistical results of entire population (ANOVA) and pairwise (Bonferroni) com- parison of AP area data of different axon types. Mean data are shown in Fig 3C. (PDF) S3 Table. Statistical data on the correlation between size and membrane time constant with AP shapes. Linear fits are shown in Fig 3D. (PDF) Acknowledgments ´ ´ We thank Balazs Hangya, Arpad Mike, and Ivan Soltesz for their valuable comments on this manu- ´ ´ ´ script. We also thank Andrea Szabo for the excellent technical assistance and Dr. Laszlo Barna and ´ ´ Dr. Pal Vagi and the Nikon Microscopy Center at IEM for kindly providing microscopy support. Author Contributions Conceptualization: Ja ´nos Brunner, Ja ´nos Szabadics. Formal analysis: Ja ´nos Brunner, Anto ´nia Arszovszki, Gergely Tarcsay, Ja ´nos Szabadics. Investigation: Ja ´nos Brunner, Anto ´nia Arszovszki, Gergely Tarcsay. Supervision: Ja ´nos Szabadics. Writing – original draft: Ja ´nos Szabadics. Writing – review & editing: Ja ´nos Brunner, Anto ´ nia Arszovszki, Gergely Tarcsay. References 1. Mateus JC, Sousa MM, Burrone J, Aguiar P. Beyond a Transmission Cable-New Technologies to Reveal the Richness in Axonal Electrophysiology. J Neurosci. 2024; 44(11). https://doi.org/10.1523/ JNEUROSCI.1446-23.2023 PMID: 38479812 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 17 / 21 PLOS BIOLOGY Uniform axonal action potentials 2. Rama S, Zbili M, Debanne D. Signal propagation along the axon. Curr Opin Neurobiol. 2018; 51:37–44. https://doi.org/10.1016/j.conb.2018.02.017 PMID: 29525575 3. Vandael D, Jonas P. Structure, biophysics, and circuit function of a “giant” cortical presynaptic terminal. Science. 2024; 383(6687):eadg6757. https://doi.org/10.1126/science.adg6757 PMID: 38452088 4. Dodson PD, Forsythe ID. Presynaptic K+ channels: electrifying regulators of synaptic terminal excitabil- ity. Trends Neurosci. 2004; 27(4):210–217. https://doi.org/10.1016/j.tins.2004.02.012 PMID: 15046880 5. Geiger JR, Jonas P. Dynamic control of presynaptic Ca(2+) inflow by fast-inactivating K(+) channels in hippocampal mossy fiber boutons. Neuron. 2000; 28(3):927–939. https://doi.org/10.1016/s0896-6273 (00)00164-1 PMID: 11163277 6. Gonzalez Sabater V, Rigby M, Burrone J. Voltage-Gated Potassium Channels Ensure Action Potential Shape Fidelity in Distal Axons. J Neurosci. 2021; 41(25):5372–5385. https://doi.org/10.1523/ JNEUROSCI.2765-20.2021 PMID: 34001627 7. Hoppa MB, Gouzer G, Armbruster M, Ryan TA. Control and plasticity of the presynaptic action potential waveform at small CNS nerve terminals. Neuron. 2014; 84(4):778–789. https://doi.org/10.1016/j. neuron.2014.09.038 PMID: 25447742 8. Kawaguchi SY, Sakaba T. Control of inhibitory synaptic outputs by low excitability of axon terminals revealed by direct recording. Neuron. 2015; 85(6):1273–1288. https://doi.org/10.1016/j.neuron.2015. 02.013 PMID: 25728570 9. Kole MH, Letzkus JJ, Stuart GJ. Axon initial segment Kv1 channels control axonal action potential waveform and synaptic efficacy. Neuron. 2007; 55(4):633–647. https://doi.org/10.1016/j.neuron.2007. 07.031 PMID: 17698015 10. Liu C, Cai X, Ritzau-Jost A, Kramer PF, Li Y, Khaliq ZM, et al. An action potential initiation mechanism in distal axons for the control of dopamine release. Science. 2022; 375(6587):1378–1385. https://doi. org/10.1126/science.abn0532 PMID: 35324301 11. Popovic M, Vogt K, Holthoff K, Konnerth A, Salzberg BM, Grinvald A, et al. Imaging Submillisecond Membrane Potential Changes from Individual Regions of Single Axons, Dendrites and Spines. Adv Exp Med Biol. 2015; 859:57–101. https://doi.org/10.1007/978-3-319-17641-3_3 PMID: 26238049 12. Ritzau-Jost A, Delvendahl I, Rings A, Byczkowicz N, Harada H, Shigemoto R, et al. Ultrafast action potentials mediate kilohertz signaling at a central synapse. Neuron. 2014; 84(1):152–163. https://doi. org/10.1016/j.neuron.2014.08.036 PMID: 25220814 13. Shu Y, Yu Y, Yang J, McCormick DA. Selective control of cortical axonal spikes by a slowly inactivating K+ current. Proc Natl Acad Sci U S A. 2007; 104(27):11453–11458. https://doi.org/10.1073/pnas. 0702041104 PMID: 17581873 14. Zbili M, Rama S, Benitez MJ, Fronzaroli-Molinieres L, Bialowas A, Boumedine-Guignon N, et al. Homeostatic regulation of axonal Kv1.1 channels accounts for both synaptic and intrinsic modifications in the hippocampal CA3 circuit. Proc Natl Acad Sci U S A. 2021; 118(47). https://doi.org/10.1073/pnas. 2110601118 PMID: 34799447 15. Kawaguchi SY. Dynamic Factors for Transmitter Release at Small Presynaptic Boutons Revealed by Direct Patch-Clamp Recordings. Front Cell Neurosci. 2019; 13:269. https://doi.org/10.3389/fncel.2019. 00269 PMID: 31249514 16. Sabatini BL, Regehr WG. Timing of synaptic transmission. Annu Rev Physiol. 1999; 61:521–542. https://doi.org/10.1146/annurev.physiol.61.1.521 PMID: 10099700 17. Goldstein SS, Rall W. Changes of action potential shape and velocity for changing core conductor geometry. Biophys J. 1974; 14(10):731–757. https://doi.org/10.1016/S0006-3495(74)85947-3 PMID: 18. Luscher HR, Shiner JS. Computation of action potential propagation and presynaptic bouton activation in terminal arborizations of different geometries. Biophys J. 1990; 58(6):1377–1388. https://doi.org/10. 1016/S0006-3495(90)82484-X PMID: 2275958 19. Manor Y, Koch C, Segev I. Effect of geometrical irregularities on propagation delay in axonal trees. Bio- phys J. 1991; 60(6):1424–1437. https://doi.org/10.1016/S0006-3495(91)82179-8 PMID: 1777567 20. Swadlow HA, Kocsis JD, Waxman SG. Modulation of impulse conduction along the axonal tree. Annu Rev Biophys Bioeng. 1980; 9:143–179. https://doi.org/10.1146/annurev.bb.09.060180.001043 PMID: 21. Chereau R, Saraceno GE, Angibaud J, Cattaert D, Nagerl UV. Superresolution imaging reveals activity- dependent plasticity of axon morphology linked to changes in action potential conduction velocity. Proc Natl Acad Sci U S A. 2017; 114(6):1401–1406. https://doi.org/10.1073/pnas.1607541114 PMID: PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 18 / 21 PLOS BIOLOGY Uniform axonal action potentials 22. Cox CL, Denk W, Tank DW, Svoboda K. Action potentials reliably invade axonal arbors of rat neocorti- cal neurons. Proc Natl Acad Sci U S A. 2000; 97(17):9724–9728. https://doi.org/10.1073/pnas. 170278697 PMID: 10931955 23. Lang-Ouellette D, Gruver KM, Smith-Dijak A, Blot FGC, Stewart CA, de Vanssay de Blavous P, et al. Purkinje cell axonal swellings enhance action potential fidelity and cerebellar function. Nat Commun. 2021; 12(1):4129. https://doi.org/10.1038/s41467-021-24390-4 PMID: 34226561 24. Waxman SG, Black JA, Kocsis JD, Ritchie JM. Low density of sodium channels supports action poten- tial conduction in axons of neonatal rat optic nerve. Proc Natl Acad Sci U S A. 1989; 86(4):1406–1410. https://doi.org/10.1073/pnas.86.4.1406 PMID: 2537496 25. Zang Y, Marder E, Marom S. Sodium channel slow inactivation normalizes firing in axons with uneven conductance distributions. Curr Biol. 2023; 33(9):1818–24 e3. https://doi.org/10.1016/j.cub.2023.03. 043 PMID: 37023754 26. Cohen CCH, Popovic MA, Klooster J, Weil MT, Mobius W, Nave KA, et al. Saltatory Conduction along Myelinated Axons Involves a Periaxonal Nanocircuit. Cell. 2020; 180(2):311–22 e15. https://doi.org/10. 1016/j.cell.2019.11.039 PMID: 31883793 27. Ford MC, Alexandrova O, Cossell L, Stange-Marten A, Sinclair J, Kopp-Scheinpflug C, et al. Tuning of Ranvier node and internode properties in myelinated axons to adjust action potential timing. Nat Com- mun. 2015; 6:8073. https://doi.org/10.1038/ncomms9073 PMID: 26305015 28. Bischofberger J, Geiger JR, Jonas P. Timing and efficacy of Ca2+ channel activation in hippocampal mossy fiber boutons. J Neurosci. 2002; 22(24):10593–10602. https://doi.org/10.1523/JNEUROSCI.22- 24-10593.2002 PMID: 12486151 29. Borst JG, Sakmann B. Effect of changes in action potential shape on calcium currents and transmitter release in a calyx-type synapse of the rat auditory brainstem. Philos Trans R Soc Lond B Biol Sci. 1999; 354(1381):347–355. https://doi.org/10.1098/rstb.1999.0386 PMID: 10212483 30. Pattillo JM, Artim DE, Simples JE Jr, Meriney SD. Variations in onset of action potential broadening: effects on calcium current studied in chick ciliary ganglion neurones. J Physiol. 1999;514(Pt 3):719– 728. https://doi.org/10.1111/j.1469-7793.1999.719ad.x PMID: 9882744 31. Perge JA, Niven JE, Mugnaini E, Balasubramanian V, Sterling P. Why do axons differ in caliber? J Neu- rosci. 2012; 32(2):626–638. https://doi.org/10.1523/JNEUROSCI.4254-11.2012 PMID: 22238098 32. Delvendahl I, Vyleta NP, von Gersdorff H, Hallermann S. Fast, Temperature-Sensitive and Clathrin- Independent Endocytosis at Central Synapses. Neuron. 2016; 90(3):492–498. https://doi.org/10.1016/j. neuron.2016.03.013 PMID: 27146271 33. Wu W, Wu LG. Rapid bulk endocytosis and its kinetics of fission pore closure at a central synapse. Proc Natl Acad Sci U S A. 2007; 104(24):10234–10239. https://doi.org/10.1073/pnas.0611512104 PMID: 34. Hill DK. The volume change resulting from stimulation of a giant nerve fibre. J Physiol. 1950; 111(3– 4):304–27. https://doi.org/10.1113/jphysiol.1950.sp004481 PMID: 14795441 35. Kamiya H. Modeling Analysis of Axonal After Potential at Hippocampal Mossy Fibers. Front Cell Neu- rosci. 2019; 13:210. https://doi.org/10.3389/fncel.2019.00210 PMID: 31139051 36. Claiborne BJ, Amaral DG, Cowan WM. A light and electron microscopic analysis of the mossy fibers of the rat dentate gyrus. J Comp Neurol. 1986; 246(4):435–458. https://doi.org/10.1002/cne.902460403 PMID: 3700723 37. Magistretti J, Mantegazza M, Guatteo E, Wanke E. Action potentials recorded with patch-clamp amplifi- ers: are they genuine? Trends Neurosci. 1996; 19(12):530–534. https://doi.org/10.1016/s0166-2236 (96)40004-2 PMID: 8961481 38. Olah VJ, Tarcsay G, Brunner J. Small Size of Recorded Neuronal Structures Confines the Accuracy in Direct Axonal Voltage Measurements. eNeuro. 2021; 8(4). https://doi.org/10.1523/ENEURO.0059-21. 2021 PMID: 34257077 39. Ritzau-Jost A, Tsintsadze T, Krueger M, Ader J, Bechmann I, Eilers J, et al. Large, Stable Spikes Exhibit Differential Broadening in Excitatory and Inhibitory Neocortical Boutons. Cell Rep. 2021; 34(2):108612. https://doi.org/10.1016/j.celrep.2020.108612 PMID: 33440142 40. Begum R, Bakiri Y, Volynski KE, Kullmann DM. Action potential broadening in a presynaptic channelo- pathy. Nat Commun. 2016; 7:12102. https://doi.org/10.1038/ncomms12102 PMID: 27381274 41. Klyachko VA, Ahern GP, Jackson MB. cGMP-mediated facilitation in nerve terminals by enhancement of the spike afterhyperpolarization. Neuron. 2001; 31(6):1015–1025. https://doi.org/10.1016/s0896- 6273(01)00449-4 PMID: 11580900 42. Rama S, Zbili M, Fekete A, Tapia M, Benitez MJ, Boumedine N, et al. The role of axonal Kv1 channels in CA3 pyramidal cell excitability. Sci Rep. 2017; 7(1):315. https://doi.org/10.1038/s41598-017-00388-1 PMID: 28331203 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 19 / 21 PLOS BIOLOGY Uniform axonal action potentials 43. Alle H, Kubota H, Geiger JR. Sparse but highly efficient Kv3 outpace BKCa channels in action potential repolarization at hippocampal mossy fiber boutons. J Neurosci. 2011; 31(22):8001–8012. https://doi. org/10.1523/JNEUROSCI.0972-11.2011 PMID: 21632922 44. Martinello K, Giacalone E, Migliore M, Brown DA, Shah MM. The subthreshold-active K(V)7 current reg- ulates neurotransmission by limiting spike-induced Ca(2+) influx in hippocampal mossy fiber synaptic terminals. Commun Biol. 2019; 2:145. https://doi.org/10.1038/s42003-019-0408-4 PMID: 31044170 45. Wang H, Kunkel DD, Schwartzkroin PA, Tempel BL. Localization of Kv1.1 and Kv1.2, two K channel proteins, to synaptic terminals, somata, and dendrites in the mouse brain. J Neurosci. 1994; 14 (8):4588–4599. https://doi.org/10.1523/JNEUROSCI.14-08-04588.1994 PMID: 8046438 46. Rasband MN. Clustered K+ channel complexes in axons. Neurosci Lett. 2010; 486(2):101–106. https:// doi.org/10.1016/j.neulet.2010.08.081 PMID: 20816921 47. Xu M, Cao R, Xiao R, Zhu MX, Gu C. The axon-dendrite targeting of Kv3 (Shaw) channels is determined by a targeting motif that associates with the T1 domain and ankyrin G. J Neurosci. 2007; 27(51):14158– 14170. https://doi.org/10.1523/JNEUROSCI.3675-07.2007 PMID: 18094255 48. Fukata Y, Lovero KL, Iwanaga T, Watanabe A, Yokoi N, Tabuchi K, et al. Disruption of LGI1-linked syn- aptic complex causes abnormal synaptic transmission and epilepsy. Proc Natl Acad Sci U S A. 2010; 107(8):3799–3804. https://doi.org/10.1073/pnas.0914537107 PMID: 20133599 49. Ogawa Y, Horresh I, Trimmer JS, Bredt DS, Peles E, Rasband MN. Postsynaptic density-93 clusters Kv1 channels at axon initial segments independently of Caspr2. J Neurosci. 2008; 28(22):5731–5739. https://doi.org/10.1523/JNEUROSCI.4431-07.2008 PMID: 18509034 50. Schulte U, Thumfart JO, Klocker N, Sailer CA, Bildl W, Biniossek M, et al. The epilepsy-linked Lgi1 pro- tein assembles into presynaptic Kv1 channels and inhibits inactivation by Kvbeta1. Neuron. 2006; 49 (5):697–706. https://doi.org/10.1016/j.neuron.2006.01.033 PMID: 16504945 51. Vacher H, Yang JW, Cerda O, Autillo-Touati A, Dargent B, Trimmer JS. Cdk-mediated phosphorylation of the Kvbeta2 auxiliary subunit regulates Kv1 channel axonal targeting. J Cell Biol. 2011; 192(5):813– 824. https://doi.org/10.1083/jcb.201007113 PMID: 21357749 52. Huang CY, Rasband MN. Axon initial segments: structure, function, and disease. Ann N Y Acad Sci. 2018; 1420(1):46–61. https://doi.org/10.1111/nyas.13718 PMID: 29749636 53. Kim SH, Ryan TA. CDK5 serves as a major control point in neurotransmitter release. Neuron. 2010; 67 (5):797–809. https://doi.org/10.1016/j.neuron.2010.08.003 PMID: 20826311 54. Zhang C, Rasband MN. Cytoskeletal control of axon domain assembly and function. Curr Opin Neuro- biol. 2016; 39:116–121. https://doi.org/10.1016/j.conb.2016.05.001 PMID: 27203619 55. Ishikawa T, Nakamura Y, Saitoh N, Li WB, Iwasaki S, Takahashi T. Distinct roles of Kv1 and Kv3 potas- sium channels at the calyx of Held presynaptic terminal. J Neurosci. 2003; 23(32):10445–10453. https://doi.org/10.1523/JNEUROSCI.23-32-10445.2003 PMID: 14614103 56. Rowan MJ, Tranquil E, Christie JM. Distinct Kv channel subtypes contribute to differences in spike sig- naling properties in the axon initial segment and presynaptic boutons of cerebellar interneurons. J Neu- rosci. 2014; 34(19):6611–6623. https://doi.org/10.1523/JNEUROSCI.4208-13.2014 PMID: 24806686 57. Garrett S, Rosenthal JJ. RNA editing underlies temperature adaptation in K+ channels from polar octo- puses. Science. 2012; 335(6070):848–851. https://doi.org/10.1126/science.1212795 PMID: 22223739 58. Stathopoulos P, Alexopoulos H, Dalakas MC. Autoimmune antigenic targets at the node of Ranvier in demyelinating disorders. Nat Rev Neurol. 2015; 11(3):143–156. https://doi.org/10.1038/nrneurol.2014. 260 PMID: 25623793 59. Yamagata A, Fukai S. Insights into the mechanisms of epilepsy from structural biology of LGI1- ADAM22. Cell Mol Life Sci. 2020; 77(2):267–274. https://doi.org/10.1007/s00018-019-03269-0 PMID: 60. Kornau HC, Kreye J, Stumpf A, Fukata Y, Parthier D, Sammons RP, et al. Human Cerebrospinal Fluid Monoclonal LGI1 Autoantibodies Increase Neuronal Excitability. Ann Neurol. 2020; 87(3):405–418. https://doi.org/10.1002/ana.25666 PMID: 31900946 61. Sell J, Rahmati V, Kempfer M, Irani SR, Ritzau-Jost A, Hallermann S, et al. Comparative Effects of Domain-Specific Human Monoclonal Antibodies Against LGI1 on Neuronal Excitability. Neurol Neu- roimmunol Neuroinflamm. 2023; 10(3). https://doi.org/10.1212/NXI.0000000000200096 PMID: 62. Rajakulendran S, Schorge S, Kullmann DM, Hanna MG. Episodic ataxia type 1: a neuronal potassium channelopathy. Neurotherapeutics. 2007; 4(2):258–266. https://doi.org/10.1016/j.nurt.2007.01.010 PMID: 17395136 63. Heilstedt HA, Burgess DL, Anderson AE, Chedrawi A, Tharp B, Lee O, et al. Loss of the potassium channel beta-subunit gene, KCNAB2, is associated with epilepsy in patients with 1p36 deletion PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 20 / 21 PLOS BIOLOGY Uniform axonal action potentials syndrome. Epilepsia. 2001; 42(9):1103–1111. https://doi.org/10.1046/j.1528-1157.2001.08801.x PMID: 11580756 64. Stokin GB, Lillo C, Falzone TL, Brusch RG, Rockenstein E, Mount SL, et al. Axonopathy and transport deficits early in the pathogenesis of Alzheimer’s disease. Science. 2005; 307(5713):1282–1288. https:// doi.org/10.1126/science.1105681 PMID: 15731448 65. Yuan P, Zhang M, Tong L, Morse TM, McDougal RA, Ding H, et al. PLD3 affects axonal spheroids and network defects in Alzheimer’s disease. Nature. 2022; 612(7939):328–337. https://doi.org/10.1038/ s41586-022-05491-6 PMID: 36450991 66. Novak P, Gorelik J, Vivekananda U, Shevchuk AI, Ermolyuk YS, Bailey RJ, et al. Nanoscale-targeted patch-clamp recordings of functional presynaptic ion channels. Neuron. 2013; 79(6):1067–1077. https://doi.org/10.1016/j.neuron.2013.07.012 PMID: 24050398 67. Oltedal L, Morkve SH, Veruki ML, Hartveit E. Patch-clamp investigations and compartmental modeling of rod bipolar axon terminals in an in vitro thin-slice preparation of the mammalian retina. J Neurophy- siol. 2007; 97(2):1171–1187. https://doi.org/10.1152/jn.01010.2006 PMID: 17167059 68. Hines ML, Carnevale NT. The NEURON simulation environment. Neural Comput. 1997; 9(6):1179– 1209. https://doi.org/10.1162/neco.1997.9.6.1179 PMID: 9248061 69. Peng H, Bria A, Zhou Z, Iannello G, Long F. Extensible visualization and analysis for multidimensional images using Vaa3D. Nat Protoc. 2014; 9(1):193–208. https://doi.org/10.1038/nprot.2014.011 PMID: 70. Antic S, Zecevic D. Optical signals from neurons with internally applied voltage-sensitive dyes. J Neu- rosci. 1995; 15(2):1392–1405. https://doi.org/10.1523/JNEUROSCI.15-02-01392.1995 PMID: 7869106 71. Rama S. Shift and Mean Algorithm for Functional Imaging with High Spatio-Temporal Resolution. Front Cell Neurosci. 2015; 9:446. https://doi.org/10.3389/fncel.2015.00446 PMID: 26635526 72. Clay JR. Novel description of ionic currents recorded with the action potential clamp technique: applica- tion to excitatory currents in suprachiasmatic nucleus neurons. J Neurophysiol. 2015; 114(1):707–716. https://doi.org/10.1152/jn.00846.2014 PMID: 26041831 73. Li L, Bischofberger J, Jonas P. Differential gating and recruitment of P/Q-, N-, and R-type Ca2+ chan- nels in hippocampal mossy fiber boutons. J Neurosci. 2007; 27(49):13420–13429. https://doi.org/10. 1523/JNEUROSCI.1709-07.2007 PMID: 18057200 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 21 / 21 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS Biology Public Library of Science (PLoS) Journal http://www.deepdyve.com/lp/public-library-of-science-plos-journal/axons-compensate-for-biophysical-constraints-of-variable-size-to-eE01Gi96PD

Axons compensate for biophysical constraints of variable size to uniformize their action potentials

Brunner, János; Arszovszki, Antónia; Tarcsay, Gergely; Szabadics, János

PLoS Biology , Volume 22 (12) – Dec 2, 2024

Download PDF

Share Full Text for Free

21 pages

Loading...

Page 2

Loading...

Page 3

Loading...

Page 4

Loading...

Page 5

Loading...

Page 6

Loading...

Page 7

Loading...

Page 8

Loading...

Page 9

Loading...

Page 10

Loading...

Page 11

Loading...

Page 12

Loading...

Page 13

Loading...

Page 14

Loading...

Page 15

Loading...

Page 16

Loading...

Page 17

Loading...

Page 18

Loading...

Page 19

Loading...

Page 20

Loading...

Page 21

References (73)

Changliang Liu, Xintong Cai, Andreas Ritzau-Jost, Paul Kramer, Yulong Li, Zayd Khaliq, S. Hallermann, P. Kaeser (2022)
An action potential initiation mechanism in distal axons for the control of dopamine release
Science, 375
I. Delvendahl, Nicholas Vyleta, H. Gersdorff, S. Hallermann (2016)
Fast, Temperature-Sensitive and Clathrin-Independent Endocytosis at Central Synapses
Neuron, 90
H. Vacher, Jae-Won Yang, Oscar Cerda, A. Autillo‐Touati, B. Dargent, J. Trimmer (2011)
Cdk-mediated phosphorylation of the Kvβ2 auxiliary subunit regulates Kv1 channel axonal targeting
The Journal of Cell Biology, 192
Rahima Begum, Y. Bakiri, K. Volynski, D. Kullmann (2016)
Action potential broadening in a presynaptic channelopathy
Nature Communications, 7
H. Alle, H. Kubota, Jörg Geiger (2011)
Sparse But Highly Efficient Kv3 Outpace BKCa Channels in Action Potential Repolarization at Hippocampal Mossy Fiber Boutons
The Journal of Neuroscience, 31
Maarten Kole, Johannes Letzkus, G. Stuart (2007)
Axon Initial Segment Kv1 Channels Control Axonal Action Potential Waveform and Synaptic Efficacy
Neuron, 55
J. Borst, B. Sakmann (1999)
Effect of changes in action potential shape on calcium currents and transmitter release in a calyx-type synapse of the rat auditory brainstem.
Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 354 1381
Andreas Ritzau-Jost, I. Delvendahl, Annika Rings, N. Byczkowicz, Harumi Harada, R. Shigemoto, J. Hirrlinger, J. Eilers, S. Hallermann (2014)
Ultrafast Action Potentials Mediate Kilohertz Signaling at a Central Synapse
Neuron, 84
H. Heilstedt, D. Burgess, A. Anderson, A. Chedrawi, Barry Tharp, Olivia Lee, C. Kashork, D. Starkey, Yuan-qing Wu, J. Noebels, L. Shaffer, S. Shapira (2001)
Loss of the Potassium Channel β‐Subunit Gene, KCNAB2, Is Associated with Epilepsy in Patients with 1p36 Deletion Syndrome
Epilepsia, 42
Charles Cox, W. Denk, D. Tank, Karel Svoboda, Karel Svoboda (2000)
Action potentials reliably invade axonal arbors of rat neocortical neurons.
Proceedings of the National Academy of Sciences of the United States of America, 97 17
Josefine Sell, V. Rahmati, Marin Kempfer, S. Irani, Andreas Ritzau-Jost, S. Hallermann, C. Geis (2023)
Comparative Effects of Domain-Specific Human Monoclonal Antibodies Against LGI1 on Neuronal Excitability
Neurology® Neuroimmunology & Neuroinflammation, 10
P. Stathopoulos, H. Alexopoulos, M. Dalakas (2015)
Autoimmune antigenic targets at the node of Ranvier in demyelinating disorders
Nature Reviews Neurology, 11
Katiuscia Martinello, Elisabetta Giacalone, M. Migliore, D. Brown, M. Shah (2019)
The subthreshold-active KV7 current regulates neurotransmission by limiting spike-induced Ca2+ influx in hippocampal mossy fiber synaptic terminals
Communications Biology, 2
Hanchuan Peng, Alessandro Bria, Zhi Zhou, G. Iannello, Fuhui Long (2014)
Extensible visualization and analysis for multidimensional images using Vaa3D
Nature Protocols, 9
Y. Manor, C. Koch, Idan Segev (1991)
Effect of geometrical irregularities on propagation delay in axonal trees.
Biophysical journal, 60 6
Mickaël Zbili, Sylvain Rama, M. Benítez, L. Fronzaroli-Molinieres, Andrzej Bialowas, Norah Boumedine-Guignon, J. Garrido, D. Debanne (2021)
Homeostatic regulation of axonal Kv1.1 channels accounts for both synaptic and intrinsic modifications in the hippocampal CA3 circuit
Proceedings of the National Academy of Sciences, 118
G. Stokin, C. Lillo, T. Falzone, R. Brusch, E. Rockenstein, Stephanie Mount, R. Raman, P. Davies, E. Masliah, David Williams, L. Goldstein (2005)
Axonopathy and Transport Deficits Early in the Pathogenesis of Alzheimer's Disease
Science, 307
J. Clay (2015)
Novel description of ionic currents recorded with the action potential clamp technique: application to excitatory currents in suprachiasmatic nucleus neurons.
Journal of neurophysiology, 114 1
D. Vandael, Peter Jonas (2024)
Structure, biophysics, and circuit function of a “giant” cortical presynaptic terminal
Science, 383
Shin-ya Kawaguchi, T. Sakaba (2015)
Control of Inhibitory Synaptic Outputs by Low Excitability of Axon Terminals Revealed by Direct Recording
Neuron, 85
Marko Popovic, Kaspar Vogt, Knut Holthoff, Arthur Konnerth, B. Salzberg, A. Grinvald, S. Antic, Marco Canepari, D. Zecevic (2015)
Imaging Submillisecond Membrane Potential Changes from Individual Regions of Single Axons, Dendrites and Spines.
Advances in experimental medicine and biology, 859
Sylvain Rama (2015)
Shift and Mean Algorithm for Functional Imaging with High Spatio-Temporal Resolution
Frontiers in Cellular Neuroscience, 9
Wei Wu, Ling-Gang Wu (2007)
Rapid bulk endocytosis and its kinetics of fission pore closure at a central synapse
Proceedings of the National Academy of Sciences, 104
Matthew Rowan, Elizabeth Tranquil, Jason Christie (2014)
Distinct Kv Channel Subtypes Contribute to Differences in Spike Signaling Properties in the Axon Initial Segment and Presynaptic Boutons of Cerebellar Interneurons
The Journal of Neuroscience, 34
J. Perge, J. Niven, E. Mugnaini, V. Balasubramanian, P. Sterling (2012)
Why Do Axons Differ in Caliber?
The Journal of Neuroscience, 32
Charles Cohen, Marko Popovic, J. Klooster, M. Weil, W. Möbius, K. Nave, Maarten Kole (2019)
Saltatory Conduction along Myelinated Axons Involves a Periaxonal Nanocircuit
Cell, 180
Hans-Rudolf Lüscher, J. Shiner (1990)
Computation of action potential propagation and presynaptic bouton activation in terminal arborizations of different geometries.
Biophysical journal, 58 6
U. Schulte, J. Thumfart, N. Klöcker, C. Sailer, W. Bildl, M. Biniossek, Doris Dehn, T. Deller, S. Eble, Karen Abbass, Tanja Wangler, H. Knaus, B. Fakler (2006)
The Epilepsy-Linked Lgi1 Protein Assembles into Presynaptic Kv1 Channels and Inhibits Inactivation by Kvβ1
Neuron, 49
Sung Kim, T. Ryan (2010)
CDK5 Serves as a Major Control Point in Neurotransmitter Release
Neuron, 67
T. Ishikawa, Yukihiro Nakamura, N. Saitoh, Wen-Bin Li, S. Iwasaki, Tomoyuki Takahashi (2003)
Distinct Roles of Kv1 and Kv3 Potassium Channels at the Calyx of Held Presynaptic Terminal
The Journal of Neuroscience, 23
S. Waxman, J. Black, J. Kocsis, J. Ritchie (1989)
Low density of sodium channels supports action potential conduction in axons of neonatal rat optic nerve.
Proceedings of the National Academy of Sciences of the United States of America, 86 4
P. Novak, J. Gorelik, U. Vivekananda, A. Shevchuk, Y. Ermolyuk, Russell Bailey, A. Bushby, Guy Moss, D. Rusakov, D. Klenerman, D. Kullmann, K. Volynski, Y. Korchev (2013)
Nanoscale-Targeted Patch-Clamp Recordings of Functional Presynaptic Ion Channels
Neuron, 79
Paul Dodson, I. Forsythe (2004)
Presynaptic K+ channels: electrifying regulators of synaptic terminal excitability
Trends in Neurosciences, 27
V. Oláh, Gergely Tarcsay, János Brunner (2021)
Small Size of Recorded Neuronal Structures Confines the Accuracy in Direct Axonal Voltage Measurements
eNeuro, 8
H. Kamiya (2019)
Modeling Analysis of Axonal After Potential at Hippocampal Mossy Fibers
Frontiers in Cellular Neuroscience, 13
B. Claiborne, D. Amaral, W. Cowan (1986)
A light and electron microscopic analysis of the mossy fibers of the rat dentate gyrus
Journal of Comparative Neurology, 246
J. Mateus, M. Sousa, J. Burrone, P. Aguiar (2024)
Beyond a Transmission Cable—New Technologies to Reveal the Richness in Axonal Electrophysiology
The Journal of Neuroscience, 44
A. Yamagata, S. Fukai (2019)
Insights into the mechanisms of epilepsy from structural biology of LGI1–ADAM22
Cellular and Molecular Life Sciences, 77
D. Hill (1950)
The volume change resulting from stimulation of a giant nerve fibre
The Journal of Physiology, 111
S. Rajakulendran, S. Schorge, D. Kullmann, M. Hanna (2007)
Episodic ataxia type 1: A neuronal potassium channelopathy
Neurotherapeutics, 4
Jörg Geiger, P. Jonas (2000)
Dynamic Control of Presynaptic Ca2+ Inflow by Fast-Inactivating K+ Channels in Hippocampal Mossy Fiber Boutons
Neuron, 28
Ronan Chéreau, G. Saraceno, Julie Angibaud, D. Cattaert, U. Nägerl (2017)
Superresolution imaging reveals activity-dependent plasticity of axon morphology linked to changes in action potential conduction velocity
Proceedings of the National Academy of Sciences, 114
Andreas Ritzau-Jost, T. Tsintsadze, M. Krueger, Jonas Ader, I. Bechmann, J. Eilers, B. Barbour, Stephen Smith, S. Hallermann (2020)
Large, Stable Spikes Exhibit Differential Broadening in Excitatory and Inhibitory Neocortical Boutons
Cell Reports, 34
Yousheng Shu, Yuguo Yu, Jing Yang, D. McCormick (2007)
Selective control of cortical axonal spikes by a slowly inactivating K+ current
Proceedings of the National Academy of Sciences, 104
Y. Ogawa, I. Horresh, J. Trimmer, D. Bredt, E. Peles, M. Rasband (2008)
Postsynaptic Density-93 Clusters Kv1 Channels at Axon Initial Segments Independently of Caspr2
The Journal of Neuroscience, 28
S. Antic, D. Zecevic (1995)
Optical signals from neurons with internally applied voltage-sensitive dyes
, 15
Chuansheng Zhang, M. Rasband (2016)
Cytoskeletal control of axon domain assembly and function
Current Opinion in Neurobiology, 39
H. Kornau, J. Kreye, Alexander Stumpf, Y. Fukata, Daniel Parthier, Rosanna Sammons, B. Imbrosci, S. Kurpjuweit, A. Kowski, M. Fukata, H. Prüss, D. Schmitz (2020)
Human Cerebrospinal Fluid Monoclonal LGI1 Autoantibodies Increase Neuronal Excitability
Annals of Neurology, 87
S. Goldstein, W. Rall (1974)
Changes of action potential shape and velocity for changing core conductor geometry.
Biophysical journal, 14 10
Sandra Garrett, J. Rosenthal (2012)
RNA Editing Underlies Temperature Adaptation in K+ Channels from Polar Octopuses
Science, 335
H. Swadlow, J. Kocsis, S. Waxman (1980)
Modulation of impulse conduction along the axonal tree.
Annual review of biophysics and bioengineering, 9
M. Hines, N. Carnevale (1997)
The NEURON Simulation Environment
Neural Computation, 9
Jacopo Magistretri, M. Mantegazza, E. Guatteo, E. Wanke (1996)
Action potentials recorded with patch-clamp amplifiers: are they genuine?
Trends in Neurosciences, 19
Y. Fukata, K. Lovero, Tsuyoshi Iwanaga, A. Watanabe, Norihiko Yokoi, K. Tabuchi, R. Shigemoto, R. Nicoll, M. Fukata (2010)
Disruption of LGI1–linked synaptic complex causes abnormal synaptic transmission and epilepsy
Proceedings of the National Academy of Sciences, 107
Marc Ford, O. Alexandrova, L. Cossell, Annette Stange-Marten, James Sinclair, C. Kopp‐Scheinpflug, M. Pecka, D. Attwell, B. Grothe (2015)
Tuning of Ranvier node and internode properties in myelinated axons to adjust action potential timing
Nature Communications, 6
Mingxuan Xu, R. Cao, Rui Xiao, M. Zhu, C. Gu (2007)
The Axon–Dendrite Targeting of Kv3 (Shaw) Channels Is Determined by a Targeting Motif That Associates with the T1 Domain and Ankyrin G
The Journal of Neuroscience, 27
J. Bischofberger, Jörg Geiger, P. Jonas (2002)
Timing and Efficacy of Ca2+ Channel Activation in Hippocampal Mossy Fiber Boutons
The Journal of Neuroscience, 22
M. Hoppa, Géraldine Gouzer, Moritz Armbruster, T. Ryan (2014)
Control and Plasticity of the Presynaptic Action Potential Waveform at Small CNS Nerve Terminals
Neuron, 84
C. Huang, M. Rasband (2018)
Axon initial segments: structure, function, and disease
Annals of the New York Academy of Sciences, 1420
Sylvain Rama, Mickaël Zbili, D. Debanne (2018)
Signal propagation along the axon
Current Opinion in Neurobiology, 51
Victoria Sabater, M. Rigby, J. Burrone (2020)
Voltage-Gated Potassium Channels Ensure Action Potential Shape Fidelity in Distal Axons
The Journal of Neuroscience, 41
Peng Yuan, Mengyang Zhang, Lei Tong, Thomas Morse, R. McDougal, Hui Ding, D. Chan, Yifei Cai, J. Grutzendler (2022)
PLD3 affects axonal spheroids and network defects in Alzheimer’s disease
Nature, 612
Shin-ya Kawaguchi (2019)
Dynamic Factors for Transmitter Release at Small Presynaptic Boutons Revealed by Direct Patch-Clamp Recordings
Frontiers in Cellular Neuroscience, 13
Daneck Lang-Ouellette, Kim Gruver, Amy Smith-Dijak, François Blot, Chloe Stewart, Pauline Blavous, Connie Li, Carter Eitrem, Charlotte Rosen, P. Faust, M. Schonewille, A. Watt (2021)
Purkinje cell axonal swellings enhance action potential fidelity and cerebellar function
Nature Communications, 12
Sylvain Rama, Mickaël Zbili, Aurélie Fékété, Mónica Tapia, M. Benítez, Norah Boumedine, J. Garrido, D. Debanne (2017)
The role of axonal Kv1 channels in CA3 pyramidal cell excitability
Scientific Reports, 7
B. Sabatini, W. Regehr (1999)
Timing of synaptic transmission.
Annual review of physiology, 61
V. Klyachko, G. Ahern, M. Jackson (2001)
cGMP-Mediated Facilitation in Nerve Terminals by Enhancement of the Spike Afterhyperpolarization
Neuron, 31
Hao Wang, Dennis Kunkel, P. Schwartzkroin, B. Tempel (1994)
Localization of Kv1.1 and Kv1.2, two K channel proteins, to synaptic terminals, somata, and dendrites in the mouse brain
, 14
J. Pattillo, D. Artim, J. Simples, S. Meriney (1999)
Variations in onset of action potential broadening: effects on calcium current studied in chick ciliary ganglion neurones
The Journal of Physiology, 514
M. Rasband (2010)
Clustered K+ channel complexes in axons
Neuroscience Letters, 486
Liyi Li, J. Bischofberger, P. Jonas (2007)
Differential Gating and Recruitment of P/Q-, N-, and R-Type Ca2+ Channels in Hippocampal Mossy Fiber Boutons
The Journal of Neuroscience, 27
Yunliang Zang, E. Marder, S. Marom (2022)
Sodium channel slow inactivation normalizes firing in axons with uneven conductance distributions
Current Biology, 33
L. Oltedal, S. Mørkve, M. Veruki, E. Hartveit (2007)
Patch-clamp investigations and compartmental modeling of rod bipolar axon terminals in an in vitro thin-slice preparation of the mammalian retina.
Journal of neurophysiology, 97 2

Publisher: Public Library of Science (PLoS) Journal
Copyright: Copyright: © 2024 Brunner et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: Raw source data and codes are available in publicly available repositories. Data: https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4 Codes: https://zenodo.org/records/13977231; https://github.com/brunnerjanos/axonal-AP/tree/1.0. All other relevant data are within the manuscript and supporting Information files. Funding: This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 772452, nanoAXON to J.S.). This work was also supported by the János Bolyai Research Scholarship of the Hungarian Academy of Sciences (J.B. and J.S.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Abbreviations: AP, action potential; DTX, dendrotoxin-I; LMFB, large mossy fiber bouton; sMF, small mossy fiber
ISSN: 1544-9173
eISSN: 1545-7885
DOI: 10.1371/journal.pbio.3002929
Publisher site: See Article on Publisher Site

Abstract

a1111111111 AU Active : Plea conductances seconfirmthata tune llhethe adingl kinetics evelsarere of axonal presente action dcorrect potentials ly: (APs) to support special- a1111111111 ized functions of neuron types. However, the temporal characteristics of voltage signals strongly depend on the size of neuronal structures, as capacitive and resistive effects slow down voltage discharges in the membranes of small elements. Axonal action potentials are particularly sensitive to these inherent biophysical effects because of the large diameter var- OPENACCESS iabilities within individual axons, potentially implying bouton size-dependent synaptic Citation: Brunner J, Arszovszki A, Tarcsay G, effects. However, using direct patch-clamp recordings and voltage imaging in small hippo- Szabadics J (2024) Axons compensate for biophysical constraints of variable size to campal axons in acute slices from rat brains, we demonstrate that AP shapes remain uni- uniformize their action potentials. PLoS Biol form within the same axons, even across an order of magnitude difference in caliber. Our 22(12): e3002929. https://doi.org/10.1371/journal. results show that smaller axonal structures have more Kv1 potassium channels that locally pbio.3002929 re-accelerate AP repolarization and contribute to size-independent APs, while they do not Academic Editor: Eunjoon Kim, Institute for Basic preclude the plasticity of AP shapes. Thus, size-independent axonal APs ensure consistent Science, REPUBLIC OF KOREA digital signals for each synapse within axons of same types. Received: April 19, 2024 Accepted: November 6, 2024 Published: December 2, 2024 Peer Review History: PLOS recognizes the benefits of transparency in the peer review Introduction process; therefore, we enable the publication of Axons broadcast information about neuronal activity in the form of digital action potentials all of the content of peer review and author responses alongside final, published articles. The (APs), whose characteristics align with the specific functions of neuronal cell types. Broad and 2+ editorial history of this article is available here: narrow APs differentially trigger Ca -influx at synapses, thereby contributing to the cell type- https://doi.org/10.1371/journal.pbio.3002929 specific strength and plasticity of synaptic responses [1–3]. These precisely tuned presynaptic Copyright:© 2024 Brunner et al. This is an open APs are usually different from somatic APs and they rely on cell type-specific functions of axo- access article distributed under the terms of the nal ionic conductances [4–14]. The diversity of axonal K channels plays particularly impor- Creative Commons Attribution License, which tant roles in cell type-specific tuning, determining the speed of repolarization, which controls permits unrestricted use, distribution, and 2+ 2+ the duration of synaptic Ca channel opening and the driving force for Ca influx [4,15,16]. reproduction in any medium, provided the original However, in addition to the active components of APs, passive membrane properties also author and source are credited. influence AP waveforms. Fast voltage changes are constrained by biophysical principles deter- Data Availability Statement: Raw source data and mined by the small and variable axon diameter [17–20]. Axons are not simple RC circuits, codes are available in publicly available where smaller membrane capacitance compensates for larger input resistance, because the var- repositories. Data: https://repo.researchdata.hu/ dataset.xhtml?persistentId=hdl:21.15109/ARP/ iable surface-to-diameter ratio causes an insistent impedance mismatch. As a result, APs are PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 1 / 21 PLOS BIOLOGY Uniform axonal action potentials BTK8A4 Codes: https://zenodo.org/records/ expected to be more heavily filtered in smaller axonal structures. The consequential limitations 13977231; https://github.com/brunnerjanos/ of these filtering effects on the rise and propagation of axonal APs have been recognized for axonal-AP/tree/1.0. All other relevant data are decades, and the accepted view now is that they are compensated by local regulation of Na within the manuscript and supporting Information channels and optimal morphology [2,18,21–25] or circumvented by myelin sheaths [26,27]. files. However, Na channels have little contribution to the repolarization phase that is the most Funding: This project has received funding from critical for the digital output functions. the European Research Council (ERC) under the 2+ The duration and decay of APs determine Ca influx into the pre-synapse, initiating the European Union’s Horizon 2020 research and translation of digital APs into analog synaptic responses. Due to the fast voltage-dependent innovation programme (Grant agreement No. 2+ 772452, nanoAXON to J.S.). This work was also operation and efficient activation of synaptic Ca channels, AP repolarization-dependent 2+ supported by the Ja´nos Bolyai Research deactivation and driving force control the amount of Ca influx, which exponentially deter- Scholarship of the Hungarian Academy of Sciences mines synaptic release [5,8,28–30]. Hence, even minor changes in AP shape exert a large (J.B. and J.S.). The funders had no role in study impact on the evoked responses [2,3,16]. design, data collection and analysis, decision to However, the variability in size and geometry of individual synaptic boutons often depends publish, or preparation of the manuscript. on properties or states that are not directly related to synaptic efficacy, such as the presence of Competing interests: The authors have declared organelles, position along main and terminal transportation routes [31], accumulation of vesi- that no competing interests exist. cles [21], sudden changes in the membrane surface due to bulk endocytosis [32,33] or excess Abbreviations:A AP,U action : Anabpotential; breviatioDTX, nlisthasbeencompiledforthoseusedinthetext:Pleaseverifythatallentriesarecorrect: ionic changes during strong activity [34]. In the synaptic target area, unmyelinated axonal dendrotoxin-I; LMFB, large mossy fiber bouton; diameters typically range between 0.1 and 5 μm, while synaptic boutons within the same axons sMF, small mossy fiber. also exhibit several-fold differences, resulting in quadratically larger membrane surfaces. Con- sequently, size variability can substantially complicate synaptic effects by uncontrolled proper- ties that passively influence AP shapes. Therefore, synapse-to-synapse variability in AP decay would not only compromise the cell-type specificity of axonal signaling but also raise questions about whether variable APs within a single axon can truly be considered as real digital signals. Early theoretical studies predicted that due to the different biophysical properties caused by variable size, the electrotonically isolated axon compartments elicit APs with variable repolari- zation speed [17,19], implying variable trigger signals for each synapse. However, this question was not experimentally addressed previously. Thus, it remained unknown whether the vari- ability of AP decay between individual boutons of single axons is compensated or provides an additional level of synaptic diversity. Therefore, here we employed direct patch clamp record- ings and voltage imaging at unprecedented temporal, morphological, and spatial resolution in various types and sizes of hippocampal axons to understand the size dependence of axonal AP signaling. We focused on the decay phase of APs because it has a fundamental impact on out- 2+ put function of axons by regulating Ca influx and has a different correlation to axon size than the extensively studied rising phase, which is crucial for reliable axonal AP propagation [2,18,21–25,35]. Our results provide strong evidence for the maintenance of size-independent AP uniformity within single axons of various types. Among them, in hippocampal mossy fiber axons, Kv1 channel density varies in boutons of different sizes in a manner that compensates for changes in AP shape otherwise expected from capacitive and resistance effects. The varying density of Kv1 channels may be related to distinct anchoring mechanisms and unusual distri- bution of these channels. Results Despite variable biophysics similar AP shape is maintained at variable caliber axons First, we explored the biophysical behavior of variable-caliber axons by simulating APs by sodium and potassium conductances that were homogeneously distributed in the morphology of a real mossy fiber axon originating from a hippocampal granule cell (Figs 1A and S1). As expected from simple morphology simulations [17–20], the passive biophysical properties and PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 2 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 1. SimilarAU AP :shapes Abbrevi inatio sMFs nlisand tshaveb LMFBs eencomp despite iledfor thetho different seusedinFi biophysic gs1to5: alPlproperties. easeverifythat (A)alle Simulati ntriesa ons recorr in real ect: MF morphology with uniform I and I predict that AP kinetics, measured as AP area, is variable due to the size- K Na dependent passive membrane properties. In general, smaller boutons have slower AP. A more detailed and extended versions of these simulations are shown in S1 Fig. (B) The recorded sMFs (upper image) were identified with subsequent in situ confocal imaging (lower image) or post hoc anatomy (see Methods). Inset shows an LMFB along the sMF-recorded axon that was used for its identification. LMFBs were recorded in the same conditions as sMFs. (C) Their diameters segregated the recorded structures into 2 populations. sMFs include small en passant boutons and main axons. Small dots show individual measurements, large black symbols are the mean, black vertical lines are the median, and diamonds and whiskers show 25%–75% and 10%–90% ranges, respectively. (D) The sMFs had smaller membrane capacitance, larger input resistance and slower membrane time constant compared to LMFBs (n = 50 sMF and 64 LMFBs). (E) APs were evoked by small current injections. (F) AP amplitudes and width at half-maximum (HW) were similar both in the raw recordings and after elimination of the instrumental filtering effect by post hoc simulation of the recording equipment and biological structures [38] (n = 15 sMF and 11 LMFBs). (G) The integral area of APs was similar in sMFs and LMFBs and did not correlate with the capacitance of individual boutons, which measures membrane surface and size of the recorded structures (n = 50 sMF and 64 LMFBs). Source data and codes available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g001 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 3 / 21 PLOS BIOLOGY Uniform axonal action potentials the waveforms of active conductance-evoked APs changed with the size of the axons. Both the rise and repolarization were generally slower in smaller structures. However, local AP shapes were also influenced by adjacent axonal regions, as sudden widening of the diameter can act as a capacitive sink. While the effects of size and capacitance mismatches on AP propagation (i.e., AP rise/sodium channel activation) have been previously investigated [18,21–26], how these biophysical constraints affect AP repolarization remains unknown. Modeling the subsequent functional steps, the differences in AP shapes were multiplied by the nonlinear correlations 2+ 2+ between AP shape and Ca influx, and between Ca influx and synaptic release, resulting in a 40% variability in release efficacy. The results showed that—consistent with previous predic- tions [17,19,20]—the biophysical consequences of variable axon diameter can lead to changes 2+ in AP repolarization in a range that can affect Ca influx, which can add complexity to neuro- nal outputs. Although this simplified simulation did not capture all possible details, it provides a testable prediction: AP repolarization and shape are generally slower in smaller axons, and they are also variable due to mismatch effects at diameter changes. Next, we tested AP shapes in mossy fibers using direct patch clamp recordings from small axon terminals. These axons form small boutons and connecting segments, whose sizes match most cortical axons, and exceptionally large boutons [3,36]. Mossy fibers lack complex arbori- zation and myelin sheaths, which would complicate their biophysical properties. Only unequivocally identified and intact mossy fibers were analyzed (Fig 1B), categorized either as small mossy fibers (sMFs, including both small boutons and main axons) or large mossy fiber boutons (LMFBs) (Fig 1C). Cut ends, called blebs [1], were not included. As expected from the smaller size, the membrane capacitance was smaller in sMFs, whereas their input resistance was larger. The measured membrane time constant was also slower in sMFs at resting condi- tions (Fig 1D and S1 Table) that would suggest that APs are also size dependent. However, the amplitude and kinetics of locally evoked APs were surprisingly similar in the raw record- ings (Fig 1E and 1F and S1 Table). This may be the consequence of the comparable size of the structures and recording pipettes [1,37–39], which were similarly small in all recordings. As these parameters can be influenced by instrumental filtering caused by the small pipette required for patching axons, we quantified the integral area of APs. Our analysis showed that the integral area of APs was not correlated with bouton identity or the effective size of the recorded mossy fiber structure, measured by local capacitance (Fig 1G). Interference between small axons and experimental instruments affects these recordings [1,37–39]. Therefore, in a subset of experiments, we corrected these signals to investigate the native, instrument-free AP shapes by correcting current clamp errors through composite sim- ulation of the morphology, currents, and the recording instruments [38]. Importantly, the cor- rected APs of sMFs and LMFBs remained indistinguishable (corrected data in Fig 1, S1 Table). The smaller variance of the corrected data compared to the raw data confirms minimal biological variability of APs. In contrast to the models that predicted variable AP decay when active conductances were homogeneously distributed, our results using corrected patch clamp data suggested that despite the complex biophysical consequences of variable axon size, APs in sMFs are similar to LMFB APs. Additionally, the corrected voltage data confirmed that AP area measurements are insensitive to instrumental filtering (Fig 1F). Simultaneous imaging of sMFs and LMFBs confirms AP uniformity within individual axons To test the AP shapes simultaneously in various axonal compartments, we used fast voltage- sensitive dye imaging which allows simultaneous monitoring of APs in different structures of the same axons at sufficient temporal resolution. Voltage imaging of distal axons was PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 4 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 2. Voltage-sensitive dye imaging confirms similar AP shapes in sMFs and LMFBs. (A) JPW1114 (50 μM) was loaded via recording of an LMFB. The pipette was retracted and after equilibration (�1 h), fluorescence signal changes were monitored at 10 kHz in the vicinity of the recording site, while APs were evoked by distal stimulation (>100 μm). (B) Electrophysiological traces were recorded at the beginning of the recordings, independent of the imaging data. Gray traces show individual experiments, with colored average APs from all sMFs and LMFBs (n = 11 experiments). Black trace shows the average signal from pixels right next to the imaged axons. (C) AP parameters from simultaneously imaged LMFBs and sMFs. (D) In the second imaging approach, the loading pipette was kept on the LMFBs allowing the over-sampling of imaged signals by the high-resolution recordings (n = 18 experiments). Amplitudes are not comparable because the presence of the excess dyes in LMFBs reduce signals by increasing non- responding dye fraction (i.e., F ). (E) Voltage imaging revealed no difference between the local APs of 2 distinct sMF structures, the main axon (main) and the filopodial extensions (filo) of the same axons. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g002 challenging because achieving sufficient concentrations of the membrane bound dye at dis- tances of 100 to 1,000 micrometers results in toxically high concentrations at the somatic load- ing sites. To overcome these limitations, we directly loaded JPW1114 into LMFBs and their neighboring axon segments with sMFs. This dye is fast enough to resolve the rapid time course of axonal APs [1,11]. First, we investigated naturally propagating APs evoked by distal stimula- tion. The dye-loading pipette was subsequently retracted (Fig 2A). The shapes of the simulta- neously imaged APs of sMFs and LMFBs were indistinguishable, and they also matched the previously recorded APs (n = 11; Fig 2B and 2C and S1 Table). Because this measurement is limited by the 10 kHz imaging rate, in the next experiments, we used higher temporal resolution with the guide of simultaneous electrophysiological signals (Fig 2C). We kept the dye-loading pipette on the LMFBs during imaging and oversampled the JPW1114 signals of locally evoked APs by aligning individual traces to the 100 kHz electro- physiology signals before averaging for distinct axonal compartments. This arrangement reli- ably reports the temporal profile of APs, although the amplitudes are not comparable due to the presence of the excess, non-responding dye in the pipette on the LMFB (i.e., F ). Impor- tantly, this measurement also confirmed uniform AP shapes in sMFs and LMFBs (n = 18, S1 Table). Furthermore, simultaneous imaging revealed similar AP shapes in 2 distinct types of PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 5 / 21 PLOS BIOLOGY Uniform axonal action potentials sMF structures: the filopodia emanating from LMFBs and main axons with small en passant boutons (Fig 2E and S1 Table). Furthermore, we did not detect differences in the APs of bas- soon-positive and -negative small axonal segments (identified by correlated post hoc immuno- labeling, S2 Fig) suggesting that AP shape is similar in synaptic and non-synaptic segments. Altogether, we concluded from our independent measurements that AP shape is uniform within individual mossy fibers despite the large intrinsic variabilities of their diameters and biophysical properties. Cell type-specific but uniform APs in various axons Next, we tested whether the AP uniformity applies to other cell types. We measured the APs and biophysical properties in diverse types of axons, representing the main cortical axon types, including subcortical and locally originated glutamatergic and GABAergic types. The axons were rigorously identified by using their morphological properties and immunolabeling for 4 to 5 markers on each axon (see Figs 3A, 3B, S3, S4 and S5). First, we confirmed in these axons that the membrane properties and AP shapes are axon type specific (Fig 3C and S2 Table), val- idating the principle that axonal functions are cell type-specifically tuned. Axons from hypo- thalamic supramammillary nucleus (SuMa) elicited wider APs than mossy fibers (average AP area: 25.8 ± 1.2 mV*ms, n = 16), while the axonal APs of hilar mossy cells (21.6 ± 3.6 mV*ms, Fig 3. AP uniformity in diverse types of hippocampal axons. (A) A representative recording from a morphologically identified VGluT2+ SuM axon in the DG (further details of identification are shown in S3 Fig). (B) A representative recording from a morphologically identified VGluT1+ mossy cell axon (see also S4 and S5 Figs). (C) Average electrical −5 −16 properties of axons are cell type dependent (ANOVA: F(4) = 4.45, 7.5, and 27.4, p = 0.002, 1.6*10 and 1*10 , Bonferroni pairwise comparison is shown in S2 Table). (D) Within cell types, the AP shapes did not show correlation with the local size and membrane time constant of the axons (statistical details of linear fits are shown in S3 Table). Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential. https://doi.org/10.1371/journal.pbio.3002929.g003 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 6 / 21 PLOS BIOLOGY Uniform axonal action potentials n = 7, MCa) and CB1R-expressing GABAergic axons (18.4 ± 1.8 mV*ms, n = 8, CB1a) were between SuMa and mossy fibers (LMFB: 15.5 ± 0.4 mV*ms, n = 64, sMF: 14.8 ± 0.5 mV*ms, n = 50). The cell type-specific APs also demonstrate that the observed uniformity of mossy fiber APs was not due to the insensitivity of recording methods. These axons, like most in the CNS, have variable diameter due to varicosities. However, within anatomically homogeneous cell types, the AP shape did not depend on the size and pas- sive membrane properties of the recorded axonal structure (Fig 3D and S3 Table). These results suggest that all tested axon types employ uniform APs in spite of their variable size and biophysical properties. Size-independent AP plasticity Axonal AP waveforms are not always the same within individual boutons and their shapes exhibit functionally relevant plasticity, as axonal APs typically show activity-dependent broad- ening during sustained activities influencing their synaptic effects [5,6,8,12–14,30,40,41]. To assess whether LMFB and sMF APs maintain uniformity during various dynamic conditions, we conducted direct recordings and VSD imaging (Fig 4A, 4B, 4D and 4E). We found that sMF APs, similarly to previous data from LMFBs [5], display frequency-dependent broadening during AP trains. Importantly, the extent of AP plasticity was similar in sMFs and LMFBs and did not depend on the size of the structure (Fig 4C and 4F). Thus, AP uniformity persists dur- ing dynamic conditions allowing for similar activity-dependent modulation of axonal output. Active conductances underlie uniform APs There must be mechanisms that compensate for the passive biophysical constraints caused by axon size. In the simulations that recreated original recordings of single APs by adjusting the amounts of active currents in the detailed model of an axon with instruments (see Fig 1F), both sMFs and LMFBs required a similar amount of sodium currents (I ) on average. How- Na ever, sMFs needed more potassium currents (I ) than LMFBs. This predicts a lower I /I K Na K ratio in sMFs (Fig 5A). To test this prediction, we experimentally measured the I /I ratio in Na K outside-out patch recordings, where both I and I were measured in each patch using 2 dif- Na K ferent voltage steps (Fig 5B). Consistent with the model prediction, while I densities were Na similar, the total I were larger in patches pulled from sMFs, resulting in larger I /I ratio K Na K (n = 46 sMFs and 47 LMFBs, for statistical details see S1 Table and Fig 5C). As outside-out patches sample ionic channels from local membranes only, we concluded that sMFs have more I , which speed up their AP repolarization and promote AP uniformity. Among axonal Kv channels, Kv1s are known to be responsible for fine-tuning APs for spe- cific functions of various axons [4,6–9,13,14,40,42], whereas Kv3 and BK channels usually pro- vide the backbone of repolarization in mossy fibers [43] and Kv7 channels set axonal excitability that also influence spiking properties [44]. However, the contribution of different Kv channels changes during complex firing patterns. Interestingly, electron microscopic stud- ies noted that Kv1.1 channel distribution was not homogeneous in mossy fibers and the expression was more prominent in small axon compartments [45]. Among these channels, we focused on the potential role of Kv1 channels in uniformizing single axonal APs, we inhibited Kv1.1 and Kv1.2 channels with dendrotoxin-I (DTX), which resulted in wider APs in mossy fibers. However, the effects of DTX were larger on AP repolarization in sMFs than on LMFBs APs (38.6% versus 21.5% at the population level, Fig 5D), indicating that sMFs employ more Kv1 channels to achieve fast repolarization. In the presence of DTX, axon size and AP shape showed a similar correlation as predicted in axon models with homogeneous conductances (S1E Fig). This observation suggests the non-uniform distribution of AP-relevant potassium PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 7 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 4. AP shape plasticity affects sMFs and LMFBs similarly. (A) AP trains were evoked at 3 different frequencies during direct patch clamp recording of sMFs and LMFBs. The example traces show 40 Hz stimulation, and 3 ms-long current injections pulses were used to avoid the contamination of the repolarization phase by the capacitance artifact. (B) Progressive and frequency-dependent broadening of sMF- and LMFB-APs (see also [5]). (C) Summary data show similar broadening in sMFs and LMFBs at different frequencies (left panel). Furthermore, the broadening of APs did not depend on the functional size of the boutons (right panel). (D) The 100 APs were evoked at 40 Hz in axons during VSD imaging experiments. To avoid photo damage, the axons were illuminated and imaged only during the first 11 and the last 13 APs. (E) Example APs obtained by 2 different imaging approaches with extracellular stimulation (upper panels) and with simultaneous intracellular recording, which allowed oversampling of the imaging signals (see Fig 2). (F) Summary data show similar broadening in sMFs and LMFBs obtained with the 2 approaches. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. AP, action potential; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g004 currents explains the difference between the measured and predicted size-dependent behavior of axonal APs. It also suggests an important role for Kv1 channels in accelerating APs in smaller MF compartments. Furthermore, the size-dependent DTX effects on APs indepen- dently confirmed our patch data about the contribution of larger I to the APs of sMFs. Indeed, in the presence of DTX, the I density and I /I ratio were similar in LMFBs and K Na K sMFs (S6 Fig). The resting membrane potential was not significantly affected by DTX (sMF control versus DTX: −75 ± 0.8 mV versus −72.7 ± 2.6 mV, LMFB control versus DTX: −- 74.4 ± 1.2 mV versus −75.1 ± 2.5 mV, ANOVA: F(3) = 0.441, p = 0.72, n = 54, 17 and 36, 17), indicating little Kv1 channel activation at resting conditions in MF axons. In contrast to the PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 8 / 21 PLOS BIOLOGY Uniform axonal action potentials Fig 5. Active conductances compensate for the biophysical constraints of smaller axon size. (A) Recreating native sMF APs (see Fig 1) in simulations required similar I but more I than in LMFBs, resulting in larger I /I ratio in Na K Na K the latter. (B) Average of recorded I and I current densities in membrane patches pulled from sMFs and LMFBs Na K (n = 48 and 40 recordings) using 2 voltage-steps in the same recordings (0 mV for 0.4 ms for I , +70 mV for I , from Na K −110 mV). (C) The average I density was similar between LMFB and sMF recordings; however, sMF membranes Na had larger I densities, resulting in larger I /I ratio in LMFBs. (D) Selective inhibition of Kv1-mediated I by 100 K Na K K nM DTX had size-dependent effects on axonal APs with larger effects on smaller sMFs. Four representative recordings are shown on the right. (E) In contrast, selective partial Kv3 inhibition by TEA (100 μM) had similar effects on APs of sMFs and LMFBs. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ ARP/BTK8A4. AP, action potential; DTX, dendrotoxin-I; LMFB, large mossy fiber bouton; sMF, small mossy fiber. https://doi.org/10.1371/journal.pbio.3002929.g005 DTX effects, partial inhibition of Kv3 channels by 100 μM of TEA had similar effects on sMF and LMFB APs (16.9% and 19.5%, Fig 5E). These results show that the contributions of at least 2 AP-relevant channels are differently related to axon size. While Kv3s have similar contribu- tions to single APs, the contribution of Kv1 correlates with axon size compensating the bio- physical consequences of small and variable axon size. Discussion We demonstrated that individual axon types elicit size-independent uniform APs. While theo- retical considerations and our initial simulations predicted that AP repolarization would be distorted by the local biophysical environment of varying axon size, our direct axonal patch- clamp recordings and voltage imaging showed that AP repolarization is remarkably resistant to these effects. We also found that Kv1 channels play a crucial role in maintaining axonal PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 9 / 21 PLOS BIOLOGY Uniform axonal action potentials uniform AP shapes regardless of size. These channels accelerate the repolarization phase, com- pensating for the slower membrane time constant in smaller axonal structures, where their contribution is greater than in larger axons. Having a uniform AP signal in all synapses throughout the axon of an individual cell has several fundamental advantages and conse- quences. Firstly, with coordinated uniform trigger signals, synapses can individually diversify their output, relying on the known diversity of synaptic machineries that allows pre- and post- synaptic cell alignment-, state-, or learning-dependent communications. If AP shapes varied with the size of the presynaptic boutons, the release machinery would have to adjust for the actual size of each synapse. Homeostatic regulation of the density of a single channel according to the size of boutons and uniform APs can eliminate the need for such complex mechanisms. Notably, Kv1 channels are anchored by molecular pathways that are independent from other AP-mediating channels (see below for details) and these channels have a sufficiently fast acti- vation to accelerate repolarization [7,9,13,43]. Secondly, without uniform axonal APs, the previously described cell type-specific rules of axonal signaling would be compromised. Diverse shapes of axonal APs evolved to fulfill cell type-specific functions. For example, narrow APs cope well with high-frequency firing in the cerebellum [8,12], whereas wider APs mediate reliable release in circuits, where sub-millisec- ond precision is not required [10,41]. However, the effect size of the biophysical consequences of variable axon diameter on AP kinetics are comparable to differences of APs in various cell types. Thus, if passive consequences of variable size were not compensated, cell type-specific AP shapes could make less significant functional contributions. We showed that AP unifor- mity applies to diverse cell types serving distinct physiological functions through specific AP shapes. Therefore, it is likely that axonal AP uniformity is a general feature of axons. These results are also important from a technical standpoint, because they confirm that our measure- ments are sufficiently sensitive for small differences in AP kinetics, and these recordings revealed previously unknown axonal AP properties, including those of hippocampal axons from supramammillary nucleus and hilar mossy cells. Thirdly, AP uniformity is also important for consistent effects of AP shape plasticity. Activ- ity-dependent AP broadening also falls within the same range as the expected passive effects caused by the variability of bouton size within individual axons. We found that AP uniformity applies not only to solitary APs, but small and large boutons showed similar broadening dur- ing AP trains. In this regard, one of the key aspects of AP decay in mossy fibers is the involve- ment of multiple potassium channels. Repolarization is primarily mediated by Kv3 with additional activity-dependent support from BK channels in LMFBs [43].Kv7 plays a significant role in setting the intrinsic excitability, consequentially impacting AP signaling, especially dur- ing complex firing patterns [44]. The exact role of Kv1 channel in mossy fibers was not known. Our findings show that Kv1 channels significantly contribute to AP repolarization, particularly in sMFs. Unlike Kv3, Kv1 functions correlate with bouton size or type during sin- gle APs. Kv3 and Kv7 channels share the ankyrin-spectrin pathway for axonal anchoring with sodium channels [46,47]. In contrast, Kv1 channels are anchored through the PSD93-A- DAM22-LGI1 pathway and are associated with Caspr2 proteins location dependently [48–50]. Kv1 transport is regulated by unique cargo-tag mechanisms involving CDK5 [51]. The direct interactions of PSD-93 and CDK5 with the cytoskeleton [52] may offer a potential molecular mechanism linking bouton volume to Kv1 density [53,54]. Future studies should also explore the roles of these pathways together with the structure size-dependent stoichiometry and sub- axonal localization of other potassium channels. Importantly, at least 1 type of potassium channel has size-dependent contributions, while other channels support repolarization in a size-independent manner. Although we focused on Kv1 and Kv3 channels, it is possible that other channels are also regulated by axon size and PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 10 / 21 PLOS BIOLOGY Uniform axonal action potentials their combined contribution is necessary for uniform axonal AP under various physiological conditions. The cell type-specificity of axonal signaling also implies that the mechanisms main- taining uniform axonal APs are likely diverse. In mossy fibers, Kv1 channels play a key role in this homeostatic AP uniformization. This finding is further supported by previous electron microscopic analysis showing that Kv1.1 channels are present in small axons but are generally absent from large terminals [45]. Generally, Kv1 channels are often employed for fine-tuning axonal APs [6,7,12–14,55,56]. Notably, cephalopods overcome slower biochemical mechanisms in low-temperature environments using Kv1 channels with faster opening and closing kinetics by elaborated RNA-editing mechanisms depending on the geographical latitude [57]. We measured APs only from unmyelinated segments of axon. Direct patch clamp record- ings are not possible from the myelinated axons. While SuM, MC, CB1R axons are known to have myelinated segments, in the recorded regions they were mostly unmyelinated as also sug- gested by large number of synaptic boutons. While APs can be measured in myelinated seg- ments with VSD imaging, we employed this method only for mossy fibers, which do not have myelin sheaths. APs are known to be different in the myelin-sheath-insulated and exposed membranes [26], but the insulated segments do not contribute directly to the output functions of axons [2]. Therefore, AP uniformity is not relevant in these segments. Importantly, several components of Kv1 anchoring and regulatory complexes are directly implicated in diseases. Specifically, several forms of mutations in the Kv1-anchoring LGI1 pro- teins lead to autosomal dominant temporal lobe epilepsies [48,58,59]. Furthermore, auto-anti- bodies against this extracellular protein initiate autoimmune encephalitis [60,61], which also inhibit their interaction with the ADAM22-family proteins [59]. Caspr2 proteins are also auto- immune targets, causing periferial neuromyotonia [58]. Episodic ataxias are associated with various mutations in the Kv1.1 gene [62], while loss of Kvβ, which is involved in the normal targeting of Kv1 channels, was observed in patients with severe seizures [58,63]. Axonal swell- ings are biomarkers of Alzheimer’s disease, and they potentially contribute to the development of abnormal neuronal functions in this disease [64,65]. Because previously it was not known that these proteins and features are important for maintaining uniform axonal AP signaling, future studies should clarify the contribution of disrupted size-independent AP uniformity to the development of these diseases. Methods Ethics statement All experiments were performed in accordance with the Institutional Ethical Codex, Hungar- ian Act of Animal Care and Experimentation (1998, XXVIII, section 243/1998) and the Euro- pean Union guidelines (directive 2010/63/EU), and with the approval of the National Scientific Ethical Committee on Animal Experimentation (NEBIH, PE/EA/48-2/2020) and Institutional Animal Care and Use Committee. Slice electrophysiology Hippocampal slices were prepared from young rats (P23-35, deeply anesthetized with isoflur- ane) in ice-cold artificial cerebrospinal fluid (85 mM NaCl, 75 mM sucrose, 2.5 mM KCl, 25 mM glucose, 1.25 mM NaH PO , 4mM MgCl , 0.5 mM CaCl , and 24 mM NaHCO , Leica 2 4 2 2 3 VT1200 vibratome). We used an upright microscope (Eclipse FN-1; Nikon) equipped with a high numerical aperture objective (25× NA1.1W objective, Nikon) and with a confocal scan- ning system (C1 Plus, Nikon). Recording solution was composed of 126 mM NaCl, 2.5 mM KCl, 26 mM NaHCO , 2 mM CaCl , 2 mM MgCl , 1.25 mM NaH PO , and 10 mM glucose. 3 2 2 2 4 Recordings were performed using an intracellular solution containing 90 mM potassium PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 11 / 21 PLOS BIOLOGY Uniform axonal action potentials gluconate, 43.5 mM KCl, 1.8 mM NaCl, 1.7 mM MgCl , 0.05 mM EGTA, 10 mM HEPES, 2 mM Mg-ATP, 0.4 mM Na -GTP, 10 mM phosphocreatine, 8 mM biocytin, and 20 μM Alexa Fluor 594. The temperature was kept at 34 to 36˚C during the recordings. The axonal structures in the stratum lucidum of CA3 region or in the inner layer of stratum moleculare of DG region were patched under the guidance of IR-DIC optics by using capillaries pulled from borosilicate tubing (1.5 mm outer, 0.75 mm inner diameter, Sutter Instruments). Data was collected using a MultiClamp700B amplifier and digitized by a Digidata 1440 A/D converter at 250 kHz sampling rate (Molecular Devices). Capacitance neutralization was carefully adjusted close to its possible maximum. After the recording, we captured the position of the recording electrode together with the in situ morphological characteristics of the recorded axon based on the fluorescent Alexa Fluor 594 signal by using the confocal scanning head. For detailed visualization of the morphology, we employed post hoc fluorescent streptavidin reaction. Only identified axons were included (see below) and recordings close to the cut end of the axons were excluded. For the measurement of the local electrical parameters of axons, 100 ms-long pulses from −70 mV to −90 mV were used in voltage clamp. First, we recorded the signals in cell-attached configuration, which allowed the estimation of the seal resistance in each measurement (12.1 ± 0.6 GΩ for sMF, 13.5 ± 0.6 GΩ for LMFBs, and 12.5 ± 0.9 GΩ for other axons). For the quantification of local axonal capacitance, the average of the cell-attached signal was sub- tracted from the whole-cell recorded trace to eliminate residual uncompensated instrumental capacitance. We fitted the capacitive current transients with the sum of 2 exponential functions in a 3 to 7 ms time window after the peak assuming that a two-compartmental electrical equiv- alent circuit is an adequate representation of recorded axons [66,67]. In this model, the faster exponential describes the electrical behavior of the local membrane at the recording site, whereas the slower one is the composite representation of the entire axonal arbor. To validate this two-compartmental simplification, we systematically fitted the traces with mono-, bi-, or tri-exponential functions in a subset of the experiments. The two-compartmental model signif- icantly outperformed the fit of the mono-exponential model assessed (sum of squared error: 2 2 2 2 sMF: 7.81 nA versus 2.37 nA ; LMFB: 10.58 nA versus 2.18 nA for mono- and bi-exponen- tial fits, respectively). Furthermore, fitting the traces with the sum of 3 exponentials resulted in only marginal improvement compared to the biexponential model (sMF: 2.37 nA versus 2.16 2 2 2 nA ; LMFB: 2.18 nA versus 2.04 nA for bi- and tri-exponential fits, respectively) indicating that this simplified two-compartmental model is necessary and sufficient to characterize the passive electrical properties of the recorded axons. Thus, the structure capacitance derived from the fast component of the fit and it represented the biophysically effective size of each recorded bouton. Input resistance was measured both in voltage- and current clamp record- ings to check the consistency of the recording, but for quantification we used the voltage clamp data. The membrane time constant data derived from independent current clamp recordings. To calculate the AP area in electrophysiological recording data, we first deter- mined the start of the AP at the voltage value where the upstroke velocity reached 100 mV/ms. Then, we computed the AP area within the region between the start of the AP and the point where the membrane potential returned to the same voltage value. Membrane patches were pulled from the recorded axons after the recording of whole-bou- ton properties of the structure, including the structure capacitance. The membrane area was estimated from the correlation between pipette resistance and the capacitance of the outside- out patches. To obtain this correlation, we precisely determined the capacitance of the pulled patches in the subset of experiments by pushing the pipette tip into a Sylgard ball at the end of the recording. This area was used to normalize the recorded I and I from each bouton. The Na K 2 2 average estimated areas of outside-out patches were 3.69 ± 0.14 μm and 3.84 ± 0.17 μm for sMFs and LMFBs. PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 12 / 21 PLOS BIOLOGY Uniform axonal action potentials Modeling of axonal recordings with realistic pipettes and amplifier circuits (Fig 1F) Computer simulations were performed in the NEURON simulation environment (v7.5) [68] using a detailed model that captures the experimentally observed behavior of the measuring system as described in Olah and colleagues [38]. Pipette parameters and the compensatory mechanisms of the amplifier were set according to the actual measurements using modified model-cell circuits or short-circuiting components of the amplifier. Furthermore, the pipette parameters were quantified using partial submerging, tip-breaking, and direct microscopic measurements of their physical dimensions. The pipette models were adjusted for each record- ing pipette to recreate capacitive signals obtained in open and on-cell situations. The amplifier and adjusted pipette models allowed to recreate the characteristic capacitive signals during each current clamp recording. To incorporate realistic morphologies in the model, we used a 140 μm-long detailed recon- struction of the recorded part for each axon. The reconstruction was performed semi-automat- ically by the Vaa3D software [69] using the confocal images taken right after the recordings. Remote axonal parts were implemented as simplified dummy compartments. To obtain the passive background of the axon, we fitted model responses to the experimentally recorded sub- threshold voltage signals by optimizing axonal passive parameters and the access resistance of the recording. These fundamental parameters were similar for models which were based on -2 LMFBs and sMFs (C = 0.90 ± 0.08 and 0.89 ± 0.05 μF*cm , R = 151.49 ± 18.81 and m a 134.45 ± 15.66 O*cm and R = 117.81 ± 7.63 and 185.26 ± 34.31 M).Ω R was held con- access m stant at 50 k*Ω cm . To predict the axonal AP waveforms, we incorporated active sodium and potassium con- ductance models with adjustable voltage dependence and kinetic parameters onto the model and the current clamp output was fitted to the experimental data by freely varying all channel parameters. For each axon, 10 APs from 4 different initializations were fitted and the average of the conductance parameters yielded by these 40 independent optimizations was considered as the best fit options. Fits were performed using the PRAXIS algorithm of the NEURON’s built-in Multiple Run Fitter. In addition to recreating the recorded AP signals and instrument related capacitive artifacts, these fits predicted the underlying g and g properties as well. Na K Finally, predicted native AP signals were obtained by running these constrained g and g in Na K realistic axonal morphology without the inclusion of the pipette and amplifier circuit models. To generate experimentally testable currents, these g and g were used to generate currents Na K in response to voltage steps in individual axons. NEURON codes are available at Zenodo: https://zenodo.org/records/13977231 and Github: https://github.com/brunnerjanos/axonal-AP/tree/1.0. VSD imaging The tip of the pipettes was backfilled with normal intracellular solution (see above) and the voltage sensitive dye (JPW1114 also called di-2-ANEPEQ, 10 to 50 μM [70]) containing intra- cellular solution was carefully loaded thereafter. LMFBs were patched within 4 min after the loading of the pipette. After forming whole cell configuration, the dye was allowed to diffuse into the axon at least for 30 min before starting the imaging experiment. The upright micro- scope (Nikon FN-1) was equipped with a 1.1 numerical aperture 25× water-immersion objec- tive and a 1.5× magnifier lens in front of the camera. For imaging voltage signal, we used a single mode 532 nm laser source (CNI Changchun MSL-FN-532S-400mW), which was cou- pled via a bifurcated SMA-905 fiber guide (0.4 mm, NA 0.39) to the epifluorecent attachment of the microscope, which was equipped a filter cube (ZT532rdc dichroic mirror and 542 nm PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 13 / 21 PLOS BIOLOGY Uniform axonal action potentials long-pass emission filter). When the axonal objects needed to be illuminated for longer time, for example during focusing and searching, we used a diode laser illuminator (89North LDI-7) via the second branch of the fiber guide. The excitation intensity was regulated by the direct control of the output power and/or by the use of ND filters. Optical signals were acquired with a CMOS camera (Redshirt Imaging, Da Vinci-1K) with a frame rate of 10 kHz (64 × 46 pixels, pixel size: 0.79 μm) and digitized simultaneously with the electrophysiological signals by the AD board of the imaging system. At the end of the experiment, the VSD fluorescence was used also for morphological identification. Analysis of VSD signals Data were discarded if the half-width of the APs showed more than 20% increase of its initial value. A custom Fiji macro was used to correct for photo-bleaching in individual pixels of the image stacks using independent double exponential fits. The pixels were clustered into 8 bins for the fitting. To select the active axonal compartments within bleach corrected image stacks, we generated aΔF/F image, where the intensities around (±0.2 ms) the AP peaks were sub- tracted and divided with the baseline values of each pixel on the image. The identification was helped by the multifocal morphological images that were recorded after the imaging session. In the first imaging approach in which the naturally propagating APs were imaged, the 10 kHz imaging traces were segmented and aligned with custom Python scripts. For the second imag- ing approach that intended to increase the temporal resolution of the imaging signals, we implemented a shift and mean algorithm (written in Python) to increase the temporal resolu- tion of the recorded signals (adopted from [71]). In brief, the higher sampling rate of the electrophysiology signals registered simultaneously on the DigiData AD-board allowed us to align the spikes with 250 kHz temporal resolution. The oversampled imaging signals were pooled, aligned according to these more precise AP peaks and then resampled by averaging multiple (typically 8) data points resulting in 24.5 to 27.5 kHz final resolution. For demonstra- tion purposes, signals were low-pass filtered using a bidirectional digital Chebyshev-II lowpass filter (f_stop = 8 kHz, Microcal OriginPro). If the imaging experiment included longer axons, VSD signals were averaged in short axonal segments (<5 μm). The signals from these seg- ments were averaged for each axon for comparison of sMF and LMFB. VSD signals are repre- sented and were analyzed without filtering, except for the half-width measurements, where the signals were low-pass filtered (Butterworth) in 2 steps: 10 kHz and 100 kHz for over-sampled signals, 2 kHz and 20 kHz for extracellular stimulation-evoked APs. Area for background mea- surements (Fig 2B) was selected 3 to 7 pixels distance from the imaged LMFB, which had the brightest emission. Because of the lack of AP signals and intensity changes in the background area, we concluded that neighboring structures (i.e., LMFB and sMF) can be independently analyzed, which were selected with similar distances. AP area was determined as the integral area in a 0.6 ms window starting from 0.05 or 0.1 ms from the peak of peak-normalized AP sig- nals. Normalization was needed because VSD signal amplitude depends on the background intensity, which is influenced by other factors in some cases (e.g., pipette dye load). Anatomical identification of whole-bouton recorded axons Recorded axons were loaded via the pipette with biocytin and in most cases with Alexa Fluor 594. The fluorescent dye allowed in situ imaging of the axons after their recordings and identi- fication of the recordings in subsequent morphological analysis (Nikon C1 scanning confocal head or RedShirtImaging CMOS Da Vinci-1K camera with eplifluorescent Sutter LED illumi- nation). In case of sMF recordings, axons were identified by unbranching axon that run paral- lel in the stratum lucidum where they form typical LMFBs. Biocytin labeling and subsequent PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 14 / 21 PLOS BIOLOGY Uniform axonal action potentials fluorescent streptavidin reactions allowed a more detailed visualization of the morphology and immunolabeling. Slices were fixed in 2% paraformaldehyde and picric acid solution for at least 20 h and then they were resectioned (40 to 60 μm). Recorded axons were stained for biocytin using Alexa-conjugated streptavidin (1:500, Thermo Fisher, S11227). The recorded axons were typically present in 3 to 5 sections. Short axons were excluded from the analysis. To maximize the capacity for immunolabeling for the recorded axons, each section was individually stained with different combinations of 2 or 3 primary antibodies for 72 h: anti-VGluT2 (guinea pig, 1:2,000, Millipore, AB2251, RRID: AB_1587626), anti-VGluT1 (guinea pig, 1:1,000, Millipore, AB5905, RRID: AB_2301751), anti-VGAT (rabbit, 1:1,000, Synaptic System, 131–003, RRID: AB_887869), anti-CB1R (rabbit, 1:1,000, Cayman, 10006590, RRID: AB_10098690), anti-par- valbumin (goat, 1:2,000, Swant, PVG-213, RRID: AB_2650496), anti-parvalbumin (guinea pig, 1:2,000, Swant, GP72, RRID: AB_2665495), anti-calretinin (goat, 1:2,000, Swant, CG1, RRID: AB_10000342), anti-bassoon (mouse, 1:3,000, Abcam, ab82958, RRID: AB_1860018). Fluores- cent dye-conjugated (AlexaFluor488, AlexaFluor647, DyLight405) secondary Donkey antibod- ies were incubated overnight. Sections were mounted in Vectashield or Prolong Glass. The slices were examined with confocal microscopy (Nikon C2 scanning confocal). Image stacks were acquired initially at 0.08 μm/px (XY) and 0.125 μm (Z direction) resolution (objective 60×, 1.4 NA) and then deconvolved (Huygens software). The resolution was lower for axonal reconstructions (0.12 μm/px XY and 0.3 μm/px Z), which were performed with Neurolucida software (version 2020.2.2, MBF Bioscience LLC, Williston, Vermont, USA, RRID: SCR_001775). Bassoon spots were manually assigned as synaptic sites in individual pixels of VSD-imaged mossy fibers using 3D analysis in Nikon NIS Elements software (version AR 4.13.04). 2+ Modeling of axonal APs, Ca influx, and release rate (Fig 1A and S1 Fig) To model the relationship between axonal morphology and the AP properties and the evoked synaptic release, a propagating AP was simulated in a complete granule cell reconstruction that had intact mossy fiber trajectory in the CA3 field. This morphology was used previously [38], but here we implemented a more complete electrical simulation on this structure. Both -2 2 the passive parameters (C = 1 μF*cm , R = 150 O*cm, and 50 k*Ω cm ) and active conduc- m a -2 -2 tances (290 mS*cm and 17 mS*cm for the Hodgkin–Huxley type sodium and potassium channels, respectively) were distributed homogeneously in the entire cell. Equilibrium poten- tials for the leak, I and I were set to −80 mV, −90 mV, and 70 mV, respectively; whereas, K Na 2+ 2+ voltage dependence of the Ca current followed the Goldman–Hodgkin–Katz [72]. Ca influx was modeled as a compound current produced by the mixture of P/Q, N, and R-type 2+ channels with a ratio of 6.5:2.5:1, respectively, as described in LMFBs [73]. Ca channel kinet- ics were adopted with a temperature coefficient of Q = 2.5 to account for the higher tempera- 2+ ture used in this model (35˚C). Release was calculated from the Ca influx assuming a conservative 2.5-power relationship. A single AP was triggered through somatic current injec- 2+ tion and parameters related to the propagating AP and the evoked Ca influx was quantified at every micrometer along the axon. As there is no explicit voltage threshold for a propagating AP, the duration of the AP was quantified as the width at −20 mV. This simulation was per- formed with a temporal resolution of 250 kHz. Supporting information 2+ S1 Fig. Simple simulations predict large variability of AP shape, axonal Ca influx, and synaptic release along the variable caliber axons if homogeneous conductances are present. (A) Theoretical rules—derived from passive RC properties of neuronal membranes—defining PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 15 / 21 PLOS BIOLOGY Uniform axonal action potentials the dependence of the speed of neuronal signals, presumably including axonal AP shape, on the size of the structure. (B) APs were simulated in a 3D reconstructed mossy fiber by using homogeneously distributed Hodgkin–Huxley sodium and potassium conductances and shown along the axon with variable diameter. Only the data from the middle section of the 2+ 2+ axon is shown. Ca influx was simulated by using a realistic set of Ca conductance [73]. 2+ Synaptic release related to the integrated Ca influx on the 2.5× power. (C) Variable AP kinet- ics measured as rate of rise and decay depending on the biophysical size of simulated axon compartment. The rise and decay phases of APs show the expected correlation at population level; however, this correlation is variable between individual boutons due to the influence of neighboring axonal regions. (D) Area of simulated APs along the entire axon demonstrate the large variability and axon size-dependence of AP shape. Simulated APs were measured at every 1 μm and were plotted against the membrane capacitance of the same spots. AP area was the integral of the repolarization phase only to avoid contamination of the measurement by 2+ effects on the rise phase of the APs [17–20]. (E) AP area and simulated Ca influx at 12 bou- tons that potentially form synaptic contacts within a short segment (linear fits AP area: R : 0.249, ANOVA p > F: 0.057, slope: −4.37. I : R : 0.287 ANOVA p > F 0.042, slope: 0.261). Ca Colors of traces scales with the capacitance of the boutons. (F) To demonstrate the reliability of AP area measurements compared to classical AP kinetic parameters, we compared raw electrophysiological recordings of APs with their corrected waveforms (see Fig 1F). The corre- lation between the half-width of APs (HW) and AP area was weaker in raw AP measurements compared to the corrected APs (R : 0.389 vs. 0.814, left panels). Furthermore, in theory, HW and AP area values should linearly correlate. However, this correlation was significant only in 2 2 the case of corrected APs (R : 0.47), but not in the raw data (R : 0.017). Raw data and simula- tions are available at https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ ARP/BTK8A4. (TIF) S2 Fig. AP shapes in axon segments that form synapses and connecting parts. (A) Synaptic regions were identified by presynaptic marker, bassoon (BSN) positivity. Individual pixels were assigned by matching the post hoc reconstruction of biocytin labeled segments that were subject of VSD imaging. Only neighboring segments were included to avoid the contamina- tion by conduction delay. LMFBs were excluded. (B) Confocal image analysis of the bassoon puncta in the biocytin labeled axon. (C) Average APs from 4 experiments with correlated syn- apse identification. (D) Comparison of the half-widths of APs from sMF segments identified as synaptic or non-synaptic based on BSN staining. Source data are available at https://repo. researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S3 Fig. Identification of axons patched in the DG originating from the supramammillary nucleus (SuMa). Two examples are shown with the reconstructions of the biocytin-labeled axons and immunolabeling for typical SuM markers and proteins that are present in other axons. SuM axons restricted to the granule cell layer and inner moleculare layer of the DG (3D-represented by light and dark gray areas). Original images are available at https://repo. researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S4 Fig. Identification of axons patched in the DG originating from the hilar mossy cells (MCa). Two examples are shown with the reconstructions of the biocytin-labeled axons and immunolabeling for typical MC markers and proteins that are present in other axons. MC axons restricted to the inner molecular layer of the DG. Original images are available at PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 16 / 21 PLOS BIOLOGY Uniform axonal action potentials https://repo.researchdata.hu/dataset.xhtml?persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S5 Fig. Identification of axons originating from the GABAergic CB1-receptor expressing cells (CB1a). Two examples are shown with the reconstructions of the biocytin-labeled axons and immunolabeling for typical GABAergic cell markers and proteins that are present in other axons. Original images are available at https://repo.researchdata.hu/dataset.xhtml? persistentId=hdl:21.15109/ARP/BTK8A4. (TIF) S6 Fig. When Kv1 components are eliminated, sMFs and LMFBs have similar I currents and I /I ratio. (A) Average of I current densities in membrane patches pulled from sMFs Na K K and LMFBs in the presence of DTX (gray traces) and in control conditions. (B) The average I and I densities and I /I ratio were similar in LMFB and sMF recordings in the presence Na K Na K of DTX. Source data are available at https://repo.researchdata.hu/dataset.xhtml?persistentId= hdl:21.15109/ARP/BTK8A4. (TIF) S1 Table. Statistical results of comparison of passive and AP properties of sMFs and LMFBs. Mean and median data with SEM and Q1–Q3 ranges are shown in figures together with individual data points. (PDF) S2 Table. Statistical results of entire population (ANOVA) and pairwise (Bonferroni) com- parison of AP area data of different axon types. Mean data are shown in Fig 3C. (PDF) S3 Table. Statistical data on the correlation between size and membrane time constant with AP shapes. Linear fits are shown in Fig 3D. (PDF) Acknowledgments ´ ´ We thank Balazs Hangya, Arpad Mike, and Ivan Soltesz for their valuable comments on this manu- ´ ´ ´ script. We also thank Andrea Szabo for the excellent technical assistance and Dr. Laszlo Barna and ´ ´ Dr. Pal Vagi and the Nikon Microscopy Center at IEM for kindly providing microscopy support. Author Contributions Conceptualization: Ja ´nos Brunner, Ja ´nos Szabadics. Formal analysis: Ja ´nos Brunner, Anto ´nia Arszovszki, Gergely Tarcsay, Ja ´nos Szabadics. Investigation: Ja ´nos Brunner, Anto ´nia Arszovszki, Gergely Tarcsay. Supervision: Ja ´nos Szabadics. Writing – original draft: Ja ´nos Szabadics. Writing – review & editing: Ja ´nos Brunner, Anto ´ nia Arszovszki, Gergely Tarcsay. References 1. Mateus JC, Sousa MM, Burrone J, Aguiar P. Beyond a Transmission Cable-New Technologies to Reveal the Richness in Axonal Electrophysiology. J Neurosci. 2024; 44(11). https://doi.org/10.1523/ JNEUROSCI.1446-23.2023 PMID: 38479812 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 17 / 21 PLOS BIOLOGY Uniform axonal action potentials 2. Rama S, Zbili M, Debanne D. Signal propagation along the axon. Curr Opin Neurobiol. 2018; 51:37–44. https://doi.org/10.1016/j.conb.2018.02.017 PMID: 29525575 3. Vandael D, Jonas P. Structure, biophysics, and circuit function of a “giant” cortical presynaptic terminal. Science. 2024; 383(6687):eadg6757. https://doi.org/10.1126/science.adg6757 PMID: 38452088 4. Dodson PD, Forsythe ID. Presynaptic K+ channels: electrifying regulators of synaptic terminal excitabil- ity. Trends Neurosci. 2004; 27(4):210–217. https://doi.org/10.1016/j.tins.2004.02.012 PMID: 15046880 5. Geiger JR, Jonas P. Dynamic control of presynaptic Ca(2+) inflow by fast-inactivating K(+) channels in hippocampal mossy fiber boutons. Neuron. 2000; 28(3):927–939. https://doi.org/10.1016/s0896-6273 (00)00164-1 PMID: 11163277 6. Gonzalez Sabater V, Rigby M, Burrone J. Voltage-Gated Potassium Channels Ensure Action Potential Shape Fidelity in Distal Axons. J Neurosci. 2021; 41(25):5372–5385. https://doi.org/10.1523/ JNEUROSCI.2765-20.2021 PMID: 34001627 7. Hoppa MB, Gouzer G, Armbruster M, Ryan TA. Control and plasticity of the presynaptic action potential waveform at small CNS nerve terminals. Neuron. 2014; 84(4):778–789. https://doi.org/10.1016/j. neuron.2014.09.038 PMID: 25447742 8. Kawaguchi SY, Sakaba T. Control of inhibitory synaptic outputs by low excitability of axon terminals revealed by direct recording. Neuron. 2015; 85(6):1273–1288. https://doi.org/10.1016/j.neuron.2015. 02.013 PMID: 25728570 9. Kole MH, Letzkus JJ, Stuart GJ. Axon initial segment Kv1 channels control axonal action potential waveform and synaptic efficacy. Neuron. 2007; 55(4):633–647. https://doi.org/10.1016/j.neuron.2007. 07.031 PMID: 17698015 10. Liu C, Cai X, Ritzau-Jost A, Kramer PF, Li Y, Khaliq ZM, et al. An action potential initiation mechanism in distal axons for the control of dopamine release. Science. 2022; 375(6587):1378–1385. https://doi. org/10.1126/science.abn0532 PMID: 35324301 11. Popovic M, Vogt K, Holthoff K, Konnerth A, Salzberg BM, Grinvald A, et al. Imaging Submillisecond Membrane Potential Changes from Individual Regions of Single Axons, Dendrites and Spines. Adv Exp Med Biol. 2015; 859:57–101. https://doi.org/10.1007/978-3-319-17641-3_3 PMID: 26238049 12. Ritzau-Jost A, Delvendahl I, Rings A, Byczkowicz N, Harada H, Shigemoto R, et al. Ultrafast action potentials mediate kilohertz signaling at a central synapse. Neuron. 2014; 84(1):152–163. https://doi. org/10.1016/j.neuron.2014.08.036 PMID: 25220814 13. Shu Y, Yu Y, Yang J, McCormick DA. Selective control of cortical axonal spikes by a slowly inactivating K+ current. Proc Natl Acad Sci U S A. 2007; 104(27):11453–11458. https://doi.org/10.1073/pnas. 0702041104 PMID: 17581873 14. Zbili M, Rama S, Benitez MJ, Fronzaroli-Molinieres L, Bialowas A, Boumedine-Guignon N, et al. Homeostatic regulation of axonal Kv1.1 channels accounts for both synaptic and intrinsic modifications in the hippocampal CA3 circuit. Proc Natl Acad Sci U S A. 2021; 118(47). https://doi.org/10.1073/pnas. 2110601118 PMID: 34799447 15. Kawaguchi SY. Dynamic Factors for Transmitter Release at Small Presynaptic Boutons Revealed by Direct Patch-Clamp Recordings. Front Cell Neurosci. 2019; 13:269. https://doi.org/10.3389/fncel.2019. 00269 PMID: 31249514 16. Sabatini BL, Regehr WG. Timing of synaptic transmission. Annu Rev Physiol. 1999; 61:521–542. https://doi.org/10.1146/annurev.physiol.61.1.521 PMID: 10099700 17. Goldstein SS, Rall W. Changes of action potential shape and velocity for changing core conductor geometry. Biophys J. 1974; 14(10):731–757. https://doi.org/10.1016/S0006-3495(74)85947-3 PMID: 18. Luscher HR, Shiner JS. Computation of action potential propagation and presynaptic bouton activation in terminal arborizations of different geometries. Biophys J. 1990; 58(6):1377–1388. https://doi.org/10. 1016/S0006-3495(90)82484-X PMID: 2275958 19. Manor Y, Koch C, Segev I. Effect of geometrical irregularities on propagation delay in axonal trees. Bio- phys J. 1991; 60(6):1424–1437. https://doi.org/10.1016/S0006-3495(91)82179-8 PMID: 1777567 20. Swadlow HA, Kocsis JD, Waxman SG. Modulation of impulse conduction along the axonal tree. Annu Rev Biophys Bioeng. 1980; 9:143–179. https://doi.org/10.1146/annurev.bb.09.060180.001043 PMID: 21. Chereau R, Saraceno GE, Angibaud J, Cattaert D, Nagerl UV. Superresolution imaging reveals activity- dependent plasticity of axon morphology linked to changes in action potential conduction velocity. Proc Natl Acad Sci U S A. 2017; 114(6):1401–1406. https://doi.org/10.1073/pnas.1607541114 PMID: PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 18 / 21 PLOS BIOLOGY Uniform axonal action potentials 22. Cox CL, Denk W, Tank DW, Svoboda K. Action potentials reliably invade axonal arbors of rat neocorti- cal neurons. Proc Natl Acad Sci U S A. 2000; 97(17):9724–9728. https://doi.org/10.1073/pnas. 170278697 PMID: 10931955 23. Lang-Ouellette D, Gruver KM, Smith-Dijak A, Blot FGC, Stewart CA, de Vanssay de Blavous P, et al. Purkinje cell axonal swellings enhance action potential fidelity and cerebellar function. Nat Commun. 2021; 12(1):4129. https://doi.org/10.1038/s41467-021-24390-4 PMID: 34226561 24. Waxman SG, Black JA, Kocsis JD, Ritchie JM. Low density of sodium channels supports action poten- tial conduction in axons of neonatal rat optic nerve. Proc Natl Acad Sci U S A. 1989; 86(4):1406–1410. https://doi.org/10.1073/pnas.86.4.1406 PMID: 2537496 25. Zang Y, Marder E, Marom S. Sodium channel slow inactivation normalizes firing in axons with uneven conductance distributions. Curr Biol. 2023; 33(9):1818–24 e3. https://doi.org/10.1016/j.cub.2023.03. 043 PMID: 37023754 26. Cohen CCH, Popovic MA, Klooster J, Weil MT, Mobius W, Nave KA, et al. Saltatory Conduction along Myelinated Axons Involves a Periaxonal Nanocircuit. Cell. 2020; 180(2):311–22 e15. https://doi.org/10. 1016/j.cell.2019.11.039 PMID: 31883793 27. Ford MC, Alexandrova O, Cossell L, Stange-Marten A, Sinclair J, Kopp-Scheinpflug C, et al. Tuning of Ranvier node and internode properties in myelinated axons to adjust action potential timing. Nat Com- mun. 2015; 6:8073. https://doi.org/10.1038/ncomms9073 PMID: 26305015 28. Bischofberger J, Geiger JR, Jonas P. Timing and efficacy of Ca2+ channel activation in hippocampal mossy fiber boutons. J Neurosci. 2002; 22(24):10593–10602. https://doi.org/10.1523/JNEUROSCI.22- 24-10593.2002 PMID: 12486151 29. Borst JG, Sakmann B. Effect of changes in action potential shape on calcium currents and transmitter release in a calyx-type synapse of the rat auditory brainstem. Philos Trans R Soc Lond B Biol Sci. 1999; 354(1381):347–355. https://doi.org/10.1098/rstb.1999.0386 PMID: 10212483 30. Pattillo JM, Artim DE, Simples JE Jr, Meriney SD. Variations in onset of action potential broadening: effects on calcium current studied in chick ciliary ganglion neurones. J Physiol. 1999;514(Pt 3):719– 728. https://doi.org/10.1111/j.1469-7793.1999.719ad.x PMID: 9882744 31. Perge JA, Niven JE, Mugnaini E, Balasubramanian V, Sterling P. Why do axons differ in caliber? J Neu- rosci. 2012; 32(2):626–638. https://doi.org/10.1523/JNEUROSCI.4254-11.2012 PMID: 22238098 32. Delvendahl I, Vyleta NP, von Gersdorff H, Hallermann S. Fast, Temperature-Sensitive and Clathrin- Independent Endocytosis at Central Synapses. Neuron. 2016; 90(3):492–498. https://doi.org/10.1016/j. neuron.2016.03.013 PMID: 27146271 33. Wu W, Wu LG. Rapid bulk endocytosis and its kinetics of fission pore closure at a central synapse. Proc Natl Acad Sci U S A. 2007; 104(24):10234–10239. https://doi.org/10.1073/pnas.0611512104 PMID: 34. Hill DK. The volume change resulting from stimulation of a giant nerve fibre. J Physiol. 1950; 111(3– 4):304–27. https://doi.org/10.1113/jphysiol.1950.sp004481 PMID: 14795441 35. Kamiya H. Modeling Analysis of Axonal After Potential at Hippocampal Mossy Fibers. Front Cell Neu- rosci. 2019; 13:210. https://doi.org/10.3389/fncel.2019.00210 PMID: 31139051 36. Claiborne BJ, Amaral DG, Cowan WM. A light and electron microscopic analysis of the mossy fibers of the rat dentate gyrus. J Comp Neurol. 1986; 246(4):435–458. https://doi.org/10.1002/cne.902460403 PMID: 3700723 37. Magistretti J, Mantegazza M, Guatteo E, Wanke E. Action potentials recorded with patch-clamp amplifi- ers: are they genuine? Trends Neurosci. 1996; 19(12):530–534. https://doi.org/10.1016/s0166-2236 (96)40004-2 PMID: 8961481 38. Olah VJ, Tarcsay G, Brunner J. Small Size of Recorded Neuronal Structures Confines the Accuracy in Direct Axonal Voltage Measurements. eNeuro. 2021; 8(4). https://doi.org/10.1523/ENEURO.0059-21. 2021 PMID: 34257077 39. Ritzau-Jost A, Tsintsadze T, Krueger M, Ader J, Bechmann I, Eilers J, et al. Large, Stable Spikes Exhibit Differential Broadening in Excitatory and Inhibitory Neocortical Boutons. Cell Rep. 2021; 34(2):108612. https://doi.org/10.1016/j.celrep.2020.108612 PMID: 33440142 40. Begum R, Bakiri Y, Volynski KE, Kullmann DM. Action potential broadening in a presynaptic channelo- pathy. Nat Commun. 2016; 7:12102. https://doi.org/10.1038/ncomms12102 PMID: 27381274 41. Klyachko VA, Ahern GP, Jackson MB. cGMP-mediated facilitation in nerve terminals by enhancement of the spike afterhyperpolarization. Neuron. 2001; 31(6):1015–1025. https://doi.org/10.1016/s0896- 6273(01)00449-4 PMID: 11580900 42. Rama S, Zbili M, Fekete A, Tapia M, Benitez MJ, Boumedine N, et al. The role of axonal Kv1 channels in CA3 pyramidal cell excitability. Sci Rep. 2017; 7(1):315. https://doi.org/10.1038/s41598-017-00388-1 PMID: 28331203 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 19 / 21 PLOS BIOLOGY Uniform axonal action potentials 43. Alle H, Kubota H, Geiger JR. Sparse but highly efficient Kv3 outpace BKCa channels in action potential repolarization at hippocampal mossy fiber boutons. J Neurosci. 2011; 31(22):8001–8012. https://doi. org/10.1523/JNEUROSCI.0972-11.2011 PMID: 21632922 44. Martinello K, Giacalone E, Migliore M, Brown DA, Shah MM. The subthreshold-active K(V)7 current reg- ulates neurotransmission by limiting spike-induced Ca(2+) influx in hippocampal mossy fiber synaptic terminals. Commun Biol. 2019; 2:145. https://doi.org/10.1038/s42003-019-0408-4 PMID: 31044170 45. Wang H, Kunkel DD, Schwartzkroin PA, Tempel BL. Localization of Kv1.1 and Kv1.2, two K channel proteins, to synaptic terminals, somata, and dendrites in the mouse brain. J Neurosci. 1994; 14 (8):4588–4599. https://doi.org/10.1523/JNEUROSCI.14-08-04588.1994 PMID: 8046438 46. Rasband MN. Clustered K+ channel complexes in axons. Neurosci Lett. 2010; 486(2):101–106. https:// doi.org/10.1016/j.neulet.2010.08.081 PMID: 20816921 47. Xu M, Cao R, Xiao R, Zhu MX, Gu C. The axon-dendrite targeting of Kv3 (Shaw) channels is determined by a targeting motif that associates with the T1 domain and ankyrin G. J Neurosci. 2007; 27(51):14158– 14170. https://doi.org/10.1523/JNEUROSCI.3675-07.2007 PMID: 18094255 48. Fukata Y, Lovero KL, Iwanaga T, Watanabe A, Yokoi N, Tabuchi K, et al. Disruption of LGI1-linked syn- aptic complex causes abnormal synaptic transmission and epilepsy. Proc Natl Acad Sci U S A. 2010; 107(8):3799–3804. https://doi.org/10.1073/pnas.0914537107 PMID: 20133599 49. Ogawa Y, Horresh I, Trimmer JS, Bredt DS, Peles E, Rasband MN. Postsynaptic density-93 clusters Kv1 channels at axon initial segments independently of Caspr2. J Neurosci. 2008; 28(22):5731–5739. https://doi.org/10.1523/JNEUROSCI.4431-07.2008 PMID: 18509034 50. Schulte U, Thumfart JO, Klocker N, Sailer CA, Bildl W, Biniossek M, et al. The epilepsy-linked Lgi1 pro- tein assembles into presynaptic Kv1 channels and inhibits inactivation by Kvbeta1. Neuron. 2006; 49 (5):697–706. https://doi.org/10.1016/j.neuron.2006.01.033 PMID: 16504945 51. Vacher H, Yang JW, Cerda O, Autillo-Touati A, Dargent B, Trimmer JS. Cdk-mediated phosphorylation of the Kvbeta2 auxiliary subunit regulates Kv1 channel axonal targeting. J Cell Biol. 2011; 192(5):813– 824. https://doi.org/10.1083/jcb.201007113 PMID: 21357749 52. Huang CY, Rasband MN. Axon initial segments: structure, function, and disease. Ann N Y Acad Sci. 2018; 1420(1):46–61. https://doi.org/10.1111/nyas.13718 PMID: 29749636 53. Kim SH, Ryan TA. CDK5 serves as a major control point in neurotransmitter release. Neuron. 2010; 67 (5):797–809. https://doi.org/10.1016/j.neuron.2010.08.003 PMID: 20826311 54. Zhang C, Rasband MN. Cytoskeletal control of axon domain assembly and function. Curr Opin Neuro- biol. 2016; 39:116–121. https://doi.org/10.1016/j.conb.2016.05.001 PMID: 27203619 55. Ishikawa T, Nakamura Y, Saitoh N, Li WB, Iwasaki S, Takahashi T. Distinct roles of Kv1 and Kv3 potas- sium channels at the calyx of Held presynaptic terminal. J Neurosci. 2003; 23(32):10445–10453. https://doi.org/10.1523/JNEUROSCI.23-32-10445.2003 PMID: 14614103 56. Rowan MJ, Tranquil E, Christie JM. Distinct Kv channel subtypes contribute to differences in spike sig- naling properties in the axon initial segment and presynaptic boutons of cerebellar interneurons. J Neu- rosci. 2014; 34(19):6611–6623. https://doi.org/10.1523/JNEUROSCI.4208-13.2014 PMID: 24806686 57. Garrett S, Rosenthal JJ. RNA editing underlies temperature adaptation in K+ channels from polar octo- puses. Science. 2012; 335(6070):848–851. https://doi.org/10.1126/science.1212795 PMID: 22223739 58. Stathopoulos P, Alexopoulos H, Dalakas MC. Autoimmune antigenic targets at the node of Ranvier in demyelinating disorders. Nat Rev Neurol. 2015; 11(3):143–156. https://doi.org/10.1038/nrneurol.2014. 260 PMID: 25623793 59. Yamagata A, Fukai S. Insights into the mechanisms of epilepsy from structural biology of LGI1- ADAM22. Cell Mol Life Sci. 2020; 77(2):267–274. https://doi.org/10.1007/s00018-019-03269-0 PMID: 60. Kornau HC, Kreye J, Stumpf A, Fukata Y, Parthier D, Sammons RP, et al. Human Cerebrospinal Fluid Monoclonal LGI1 Autoantibodies Increase Neuronal Excitability. Ann Neurol. 2020; 87(3):405–418. https://doi.org/10.1002/ana.25666 PMID: 31900946 61. Sell J, Rahmati V, Kempfer M, Irani SR, Ritzau-Jost A, Hallermann S, et al. Comparative Effects of Domain-Specific Human Monoclonal Antibodies Against LGI1 on Neuronal Excitability. Neurol Neu- roimmunol Neuroinflamm. 2023; 10(3). https://doi.org/10.1212/NXI.0000000000200096 PMID: 62. Rajakulendran S, Schorge S, Kullmann DM, Hanna MG. Episodic ataxia type 1: a neuronal potassium channelopathy. Neurotherapeutics. 2007; 4(2):258–266. https://doi.org/10.1016/j.nurt.2007.01.010 PMID: 17395136 63. Heilstedt HA, Burgess DL, Anderson AE, Chedrawi A, Tharp B, Lee O, et al. Loss of the potassium channel beta-subunit gene, KCNAB2, is associated with epilepsy in patients with 1p36 deletion PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 20 / 21 PLOS BIOLOGY Uniform axonal action potentials syndrome. Epilepsia. 2001; 42(9):1103–1111. https://doi.org/10.1046/j.1528-1157.2001.08801.x PMID: 11580756 64. Stokin GB, Lillo C, Falzone TL, Brusch RG, Rockenstein E, Mount SL, et al. Axonopathy and transport deficits early in the pathogenesis of Alzheimer’s disease. Science. 2005; 307(5713):1282–1288. https:// doi.org/10.1126/science.1105681 PMID: 15731448 65. Yuan P, Zhang M, Tong L, Morse TM, McDougal RA, Ding H, et al. PLD3 affects axonal spheroids and network defects in Alzheimer’s disease. Nature. 2022; 612(7939):328–337. https://doi.org/10.1038/ s41586-022-05491-6 PMID: 36450991 66. Novak P, Gorelik J, Vivekananda U, Shevchuk AI, Ermolyuk YS, Bailey RJ, et al. Nanoscale-targeted patch-clamp recordings of functional presynaptic ion channels. Neuron. 2013; 79(6):1067–1077. https://doi.org/10.1016/j.neuron.2013.07.012 PMID: 24050398 67. Oltedal L, Morkve SH, Veruki ML, Hartveit E. Patch-clamp investigations and compartmental modeling of rod bipolar axon terminals in an in vitro thin-slice preparation of the mammalian retina. J Neurophy- siol. 2007; 97(2):1171–1187. https://doi.org/10.1152/jn.01010.2006 PMID: 17167059 68. Hines ML, Carnevale NT. The NEURON simulation environment. Neural Comput. 1997; 9(6):1179– 1209. https://doi.org/10.1162/neco.1997.9.6.1179 PMID: 9248061 69. Peng H, Bria A, Zhou Z, Iannello G, Long F. Extensible visualization and analysis for multidimensional images using Vaa3D. Nat Protoc. 2014; 9(1):193–208. https://doi.org/10.1038/nprot.2014.011 PMID: 70. Antic S, Zecevic D. Optical signals from neurons with internally applied voltage-sensitive dyes. J Neu- rosci. 1995; 15(2):1392–1405. https://doi.org/10.1523/JNEUROSCI.15-02-01392.1995 PMID: 7869106 71. Rama S. Shift and Mean Algorithm for Functional Imaging with High Spatio-Temporal Resolution. Front Cell Neurosci. 2015; 9:446. https://doi.org/10.3389/fncel.2015.00446 PMID: 26635526 72. Clay JR. Novel description of ionic currents recorded with the action potential clamp technique: applica- tion to excitatory currents in suprachiasmatic nucleus neurons. J Neurophysiol. 2015; 114(1):707–716. https://doi.org/10.1152/jn.00846.2014 PMID: 26041831 73. Li L, Bischofberger J, Jonas P. Differential gating and recruitment of P/Q-, N-, and R-type Ca2+ chan- nels in hippocampal mossy fiber boutons. J Neurosci. 2007; 27(49):13420–13429. https://doi.org/10. 1523/JNEUROSCI.1709-07.2007 PMID: 18057200 PLOS Biology | https://doi.org/10.1371/journal.pbio.3002929 December 2, 2024 21 / 21

Journal

PLoS Biology – Public Library of Science (PLoS) Journal

Published: Dec 2, 2024

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Axons compensate for biophysical constraints of variable size to uniformize their action potentials

Axons compensate for biophysical constraints of variable size to uniformize their action potentials

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

References (73)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies