Kutsarova E et al. (2023), BDNF signaling in correlation-dependent structu...

XB-ART-59549

PLoS Biol 2023 Apr 01;214:e3002070. doi: 10.1371/journal.pbio.3002070.

Show Gene links Show Anatomy links

BDNF signaling in correlation-dependent structural plasticity in the developing visual system.

Kutsarova E , Schohl A , Munz M , Wang A , Zhang YY , Bilash OM , Ruthazer ES .

???displayArticle.abstract???
During development, patterned neural activity instructs topographic map refinement. Axons with similar patterns of neural activity converge onto target neurons and stabilize their synapses with these postsynaptic partners, restricting exploratory branch elaboration (Hebbian structural plasticity). On the other hand, non-correlated firing in inputs leads to synapse weakening and increased exploratory growth of axons (Stentian structural plasticity). We used visual stimulation to control the correlation structure of neural activity in a few ipsilaterally projecting (ipsi) retinal ganglion cell (RGC) axons with respect to the majority contralateral eye inputs in the optic tectum of albino Xenopus laevis tadpoles. Multiphoton live imaging of ipsi axons, combined with specific targeted disruptions of brain-derived neurotrophic factor (BDNF) signaling, revealed that both presynaptic p75NTR and TrkB are required for Stentian axonal branch addition, whereas presumptive postsynaptic BDNF signaling is necessary for Hebbian axon stabilization. Additionally, we found that BDNF signaling mediates local suppression of branch elimination in response to correlated firing of inputs. Daily in vivo imaging of contralateral RGC axons demonstrated that p75NTR knockdown reduces axon branch elongation and arbor spanning field volume.

???displayArticle.pubmedLink??? 37011100
???displayArticle.pmcLink??? PMC10101647
???displayArticle.link??? PLoS Biol
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: bdnf ctrl ntrk2 plg
GO keywords: regulation of synaptic plasticity [+]

???attribute.lit??? ???displayArticles.show???

	Fig 1. Blocking distinct components of BDNF signaling differentially affects correlation-dependent branch dynamics in ipsilaterally projecting RGC axons. (A) Retinal co-electroporation of EGFP and MO to visualize ipsi axons with receptor knockdown. Two-photon z-series projection of ipsi axon with EGFP (cyan) and lissamine-tagged Ctrl-MO (magenta). In magenta, 2 microglia are also visible. For TrkB-Fc experiments, intraventricular injection was performed >1 h prior to imaging. (B) Visual stimulation (10 ms light flash; 0.5 Hz) delivered via optic fibers positioned by the tadpoles’ eyes. Ipsi axons were imaged every 10 min: 1 h Darkness, 1.5 h or 2 h Asynchronous, 1.5 h or 2 h Synchronous stimulation (DAS: shorter protocol used for intraventricular TrkB-Fc). (C) Reconstructed arbors showing added, lost, and transient branches (added and lost within the summarized period). Drawings summarize branch dynamics over 1 h. Time course (20 min average) of branch (D) additions and (E) losses normalized to the average in darkness, DAS (1 h, 1.5 h, 1.5 h). Average branch (F) addition in the first and (G) loss rates in the second half of stimulation. Significant interactions in the two-way mixed design model: (F) p = 0.0189; (G) p < 0.001. (Within-subject factor: stimulation; between-subject factor: BDNF manipulation); post hoc tests corrected for multiple comparisons (BKY two-stage linear step-up procedure): p ≤ 0.05, *p ≤ 0.01. Graphs show mean ± SEM. (D–G) Control (n = 12), TrkB-Fc (n = 7), p75-MO (n = 7), TrkB-MO (n = 7). Scale bars: (A, C) 20 μm; (B) 500 μm. The data used to generate Fig 1D and 1E can be found in S1 Data and Fig 1F and 1G in S2 Data. BDNF, brain-derived neurotrophic factor; DAS, Darkness-Asynchronous-Synchronous; EGFP, enhanced green fluorescent protein; MO, morpholino oligonucleotide; RGC, retinal ganglion cell.
	Fig 2. Role of BDNF signaling in local branch elimination induced by synchronous stimulation in ipsilaterally projecting RGC axons. (A) Three consecutive reconstructed arbors from a control ipsi axon during synchronous stimulation. Addition and elimination event locations (for time point “t”) are defined as the coordinates of the branch point of the newly added axonal branch (≥1.5 μm) between t-10 min and t (blue), or terminal point of a lost branch between t and t+10 min (magenta), respectively. (B) All elimination events during DAS (1 h, 1.5 h, 1.5 h) are superimposed on the reconstructed arbor (last time point of each stimulation period). (C) Illustration of distances between an elimination event and all the other elimination events (pair distances). (D) All the elimination event pair distances for the axon from (B). Means are denoted as yellow crosses. Normalized branch (E) elimination and (F) addition event pair distances. To correct for changes in arbor size, each point represents the ratio of observed-to-randomized (S4 Fig) mean event pair distances, normalized to darkness for each axon. Interaction in the two-way mixed design model: (E) p < 0.001; (F) p = 0.0562. Post hoc tests corrected for multiple comparisons (BKY two-stage linear step-up procedure): p ≤ 0.05, *p ≤ 0.01. (E, F) Data are shown as mean + SEM. (E, F) Control (n = 12), (E) TrkB-Fc (n = 6), (F) TrkB-Fc (n = 7), (E, F) p75-MO (n = 7), (E, F) TrkB-MO (n = 7). (B) Scale bar: 20 μm. The data used to generate Fig 2E and 2F can be found in S3 Data. BDNF, brain-derived neurotrophic factor; DAS, Darkness-Asynchronous-Synchronous; RGC, retinal ganglion cell.
	Fig 3. Effects of retinal TrkB-MO and p75-MO on contralaterally projecting RGC axonal arbor elaboration over days. (А–C) Two-photon z-series projections over 4 days for axons co-electroporated with EGFP and (A) Ctrl-MO, (B) p75-MO, (C) TrkB-MO. (D) Reconstructed control arbor from day 4 showing axonal skeleton (black), terminal segments (blue), and terminal points (orange). (E, F) Morphometric analysis including (E) number of terminal points and (F) skeleton length, normalized to day 1. (G) Distribution of terminal segments (percent of total per axon) binned by length. Significant interactions in the two-way mixed design model: (E, F) p < 0.001; (G) p = 0.04394 (<5 μm bin); p = 0.01981 (5–10 μm bin). Post hoc tests corrected for multiple comparisons (BKY two-stage linear step-up procedure): (E, F) p ≤ 0.05, *p ≤ 0.01, and (G) comparison between day 1 and day 4, (G) comparison between day 2 and day 4: #p ≤ 0.05, ##p ≤ 0.01, ###p ≤ 0.001, (G) comparison between day 3 and day 4: ‡p ≤ 0.05, ‡‡p ≤ 0.01. Data represent (E–G) mean + SEM. (E–G) Control (n = 15), p75-MO (n = 10), TrkB-MO (n = 12). Scale bar: 20 μm applies to all images. The data used to generate Fig 3E and 3G can be found in S4 Data. MO, morpholino oligonucleotide; RGC, retinal ganglion cell.
	Fig 4. Arbor span compactness of contralaterally projecting RGC axons is affected by retinal TrkB-MO and p75-MO. (A–C) Three-dimensional spans of reconstructed RGC axon arbors over 4 days, co-electroporated with EGFP and MO: (A) Ctrl-MO, (B) p75-MO, and (C) TrkB-MO. (D) Significant interaction in the two-way mixed design model: p = 0.00119. (E) TrkB-MO leads to a faster increase in arbor spanning field volume compared to p75-MO [Arbor spanning field expansion index = (day 4—day 1)/(day 1 + day 4)], analyzed by Kruskal–Wallis test, p = 0.0175, and pairwise post hoc tests corrected for multiple comparisons: p ≤ 0.01. Data represent mean + SEM. (D, E) Control (n = 15), p75-MO (n = 10), TrkB-MO (n = 12). The data used to generate Fig 4D can be found in S5 Data and Fig 4E in S6 Data. MO, morpholino oligonucleotide; RGC, retinal ganglion cell.
	Fig 5. Proposed model of BDNF signaling in correlation-dependent structural remodeling. (A) Summary schematic of structural remodeling of an RGC axon of interest (red) instructed by patterned activity based on our current and previous studies: correlated firing (left) and non-correlated firing (right) of axon with neighboring inputs and postsynaptic partner. In conditions of poorly correlated firing, there is an increase in branch dynamics (addition and loss) and exploratory growth. Correlated firing results in a decrease in branch dynamics addition and loss, occurring within part of the axonal arbor. Protection against branch elimination occurs locally in parts of the arbor, strongly connected to tectal neurons (blue) via synapses in which postsynaptic NMDARs are strongly activated. Zoom-in of proposed molecular and cellular mechanisms underlying (B) Stentian and (C) Hebbian structural plasticity in the developing visual system. (B) Uncorrelated firing between an RGC axon (red) and its neighboring axons (gray), leads to a failure to activate postsynaptic NMDARs. A signal that promotes branch initiation and growth binds to p75NTR on the axon of interest (red) may be released by the neighboring axons (gray) or the postsynaptic partner (blue). We propose that this signal may be proBDNF. (C) Correlated firing of the RGC axon of interest (red) and its neighboring axons (gray) results in an activation of postsynaptic NMDARs, leading to an increase in the concentration of BDNF via release or cleavage of proBDNF (e.g., via MMP-9 or tPA). BDNF binds to postsynaptic TrkB that initiates a retrograde stabilization signal (of unknown identity) leading to suppression of branch addition and targeted decrease of branch loss. BDNF, brain-derived neurotrophic factor; MMP-9, Matrix metalloproteinase-9; NMDAR, N-methyl-D-aspartate receptor; NT, neurotrophin; RGC, retinal ganglion cell; tPA, tissue plasminogen activator.

References [+] :

Alsina, Visualizing synapse formation in arborizing optic axons in vivo: dynamics and modulation by BDNF. 2001, Pubmed, Xenbase

	Fig 1. Blocking distinct components of BDNF signaling differentially affects correlation-dependent branch dynamics in ipsilaterally projecting RGC axons. (A) Retinal co-electroporation of EGFP and MO to visualize ipsi axons with receptor knockdown. Two-photon z-series projection of ipsi axon with EGFP (cyan) and lissamine-tagged Ctrl-MO (magenta). In magenta, 2 microglia are also visible. For TrkB-Fc experiments, intraventricular injection was performed >1 h prior to imaging. (B) Visual stimulation (10 ms light flash; 0.5 Hz) delivered via optic fibers positioned by the tadpoles’ eyes. Ipsi axons were imaged every 10 min: 1 h Darkness, 1.5 h or 2 h Asynchronous, 1.5 h or 2 h Synchronous stimulation (DAS: shorter protocol used for intraventricular TrkB-Fc). (C) Reconstructed arbors showing added, lost, and transient branches (added and lost within the summarized period). Drawings summarize branch dynamics over 1 h. Time course (20 min average) of branch (D) additions and (E) losses normalized to the average in darkness, DAS (1 h, 1.5 h, 1.5 h). Average branch (F) addition in the first and (G) loss rates in the second half of stimulation. Significant interactions in the two-way mixed design model: (F) p = 0.0189; (G) p < 0.001. (Within-subject factor: stimulation; between-subject factor: BDNF manipulation); post hoc tests corrected for multiple comparisons (BKY two-stage linear step-up procedure): p ≤ 0.05, *p ≤ 0.01. Graphs show mean ± SEM. (D–G) Control (n = 12), TrkB-Fc (n = 7), p75-MO (n = 7), TrkB-MO (n = 7). Scale bars: (A, C) 20 μm; (B) 500 μm. The data used to generate Fig 1D and 1E can be found in S1 Data and Fig 1F and 1G in S2 Data. BDNF, brain-derived neurotrophic factor; DAS, Darkness-Asynchronous-Synchronous; EGFP, enhanced green fluorescent protein; MO, morpholino oligonucleotide; RGC, retinal ganglion cell.
	Fig 2. Role of BDNF signaling in local branch elimination induced by synchronous stimulation in ipsilaterally projecting RGC axons. (A) Three consecutive reconstructed arbors from a control ipsi axon during synchronous stimulation. Addition and elimination event locations (for time point “t”) are defined as the coordinates of the branch point of the newly added axonal branch (≥1.5 μm) between t-10 min and t (blue), or terminal point of a lost branch between t and t+10 min (magenta), respectively. (B) All elimination events during DAS (1 h, 1.5 h, 1.5 h) are superimposed on the reconstructed arbor (last time point of each stimulation period). (C) Illustration of distances between an elimination event and all the other elimination events (pair distances). (D) All the elimination event pair distances for the axon from (B). Means are denoted as yellow crosses. Normalized branch (E) elimination and (F) addition event pair distances. To correct for changes in arbor size, each point represents the ratio of observed-to-randomized (S4 Fig) mean event pair distances, normalized to darkness for each axon. Interaction in the two-way mixed design model: (E) p < 0.001; (F) p = 0.0562. Post hoc tests corrected for multiple comparisons (BKY two-stage linear step-up procedure): p ≤ 0.05, *p ≤ 0.01. (E, F) Data are shown as mean + SEM. (E, F) Control (n = 12), (E) TrkB-Fc (n = 6), (F) TrkB-Fc (n = 7), (E, F) p75-MO (n = 7), (E, F) TrkB-MO (n = 7). (B) Scale bar: 20 μm. The data used to generate Fig 2E and 2F can be found in S3 Data. BDNF, brain-derived neurotrophic factor; DAS, Darkness-Asynchronous-Synchronous; RGC, retinal ganglion cell.
	Fig 3. Effects of retinal TrkB-MO and p75-MO on contralaterally projecting RGC axonal arbor elaboration over days. (А–C) Two-photon z-series projections over 4 days for axons co-electroporated with EGFP and (A) Ctrl-MO, (B) p75-MO, (C) TrkB-MO. (D) Reconstructed control arbor from day 4 showing axonal skeleton (black), terminal segments (blue), and terminal points (orange). (E, F) Morphometric analysis including (E) number of terminal points and (F) skeleton length, normalized to day 1. (G) Distribution of terminal segments (percent of total per axon) binned by length. Significant interactions in the two-way mixed design model: (E, F) p < 0.001; (G) p = 0.04394 (<5 μm bin); p = 0.01981 (5–10 μm bin). Post hoc tests corrected for multiple comparisons (BKY two-stage linear step-up procedure): (E, F) p ≤ 0.05, *p ≤ 0.01, and (G) comparison between day 1 and day 4, (G) comparison between day 2 and day 4: #p ≤ 0.05, ##p ≤ 0.01, ###p ≤ 0.001, (G) comparison between day 3 and day 4: ‡p ≤ 0.05, ‡‡p ≤ 0.01. Data represent (E–G) mean + SEM. (E–G) Control (n = 15), p75-MO (n = 10), TrkB-MO (n = 12). Scale bar: 20 μm applies to all images. The data used to generate Fig 3E and 3G can be found in S4 Data. MO, morpholino oligonucleotide; RGC, retinal ganglion cell.
	Fig 4. Arbor span compactness of contralaterally projecting RGC axons is affected by retinal TrkB-MO and p75-MO. (A–C) Three-dimensional spans of reconstructed RGC axon arbors over 4 days, co-electroporated with EGFP and MO: (A) Ctrl-MO, (B) p75-MO, and (C) TrkB-MO. (D) Significant interaction in the two-way mixed design model: p = 0.00119. (E) TrkB-MO leads to a faster increase in arbor spanning field volume compared to p75-MO [Arbor spanning field expansion index = (day 4—day 1)/(day 1 + day 4)], analyzed by Kruskal–Wallis test, p = 0.0175, and pairwise post hoc tests corrected for multiple comparisons: p ≤ 0.01. Data represent mean + SEM. (D, E) Control (n = 15), p75-MO (n = 10), TrkB-MO (n = 12). The data used to generate Fig 4D can be found in S5 Data and Fig 4E in S6 Data. MO, morpholino oligonucleotide; RGC, retinal ganglion cell.
	Fig 5. Proposed model of BDNF signaling in correlation-dependent structural remodeling. (A) Summary schematic of structural remodeling of an RGC axon of interest (red) instructed by patterned activity based on our current and previous studies: correlated firing (left) and non-correlated firing (right) of axon with neighboring inputs and postsynaptic partner. In conditions of poorly correlated firing, there is an increase in branch dynamics (addition and loss) and exploratory growth. Correlated firing results in a decrease in branch dynamics addition and loss, occurring within part of the axonal arbor. Protection against branch elimination occurs locally in parts of the arbor, strongly connected to tectal neurons (blue) via synapses in which postsynaptic NMDARs are strongly activated. Zoom-in of proposed molecular and cellular mechanisms underlying (B) Stentian and (C) Hebbian structural plasticity in the developing visual system. (B) Uncorrelated firing between an RGC axon (red) and its neighboring axons (gray), leads to a failure to activate postsynaptic NMDARs. A signal that promotes branch initiation and growth binds to p75NTR on the axon of interest (red) may be released by the neighboring axons (gray) or the postsynaptic partner (blue). We propose that this signal may be proBDNF. (C) Correlated firing of the RGC axon of interest (red) and its neighboring axons (gray) results in an activation of postsynaptic NMDARs, leading to an increase in the concentration of BDNF via release or cleavage of proBDNF (e.g., via MMP-9 or tPA). BDNF binds to postsynaptic TrkB that initiates a retrograde stabilization signal (of unknown identity) leading to suppression of branch addition and targeted decrease of branch loss. BDNF, brain-derived neurotrophic factor; MMP-9, Matrix metalloproteinase-9; NMDAR, N-methyl-D-aspartate receptor; NT, neurotrophin; RGC, retinal ganglion cell; tPA, tissue plasminogen activator.