BDNF/TrkB.T1 signaling is a novel mechanism for astrocyte morphological maturation

Brain derived neurotrophic factor (BDNF) is a critical growth factor involved in the maturation of neurons, including neuronal morphology and synapse refinement. Herein, we demonstrate astrocytes express high levels of BDNF’s receptor, TrkB (in the top 20 of protein-coding transcripts), with nearly exclusive expression of the truncated isoform, TrkB.T1 which peaks in expression during astrocyte morphological maturation. Using a novel culture paradigm, we show that astrocyte morphological complexity is increased in the presence of BDNF and is dependent upon BDNF/TrkB.T1 signaling. Deletion of TrkB.T1 in vivo revealed morphologically immature astrocytes with significantly reduced volume and branching, as well as dysregulated expression of perisynaptic genes associated with mature astrocyte functions, including synaptogenic genes. Indicating a role for functional astrocyte maturation via BDNF/TrkB.T1 signaling, TrkB.T1 KO astrocytes do not support normal excitatory synaptogenesis. These data suggest a significant role for BDNF/TrkB.T1 signaling in astrocyte morphological maturation, a critical process for CNS development.


40
Astrocyte maturation is crucial developmental processes for normal CNS function. In the 41 rodent cortex, astrocyte maturation takes place largely during the first 2-4 postnatal weeks.

51
The maturation period of astrocyte morphogenesis coincides with neuronal synaptic refinement 52 (Freeman,  distinguishable given that the full-length receptor contains exons for the tyrosine kinase domain, 91 while the truncated TrkB.T1 receptor lacks this domain but has an additional exon (exon 12) not 92 found in the full length receptor (Fig. 1A). Therefore, isoform-specific transcript expression was 93 analyzed from the isolated astrocytes and corresponding whole cortex. This analysis revealed 94 that PND28 astrocytes predominately express the truncated isoform, with nearly 90% of all Ntrk2 95 expression in cortical astrocytes attributed to TrkB.T1 (151.91 +/-9.18 FPKM for NM_008745, 96 19.30 +/-4.45 FPKM for NM_001025074) (Fig. 1b). Figure 1. Astrocytes express high levels of truncated TrkB during astrocyte morphogenesis. A Cartoon representation of Ntrk2 isoform exons and primers utilized. Total TrkB was probed with exons across both isoforms, with isoform-specific exons utilized to probe TrkB.Fl (blue) and TrkB.T1 (teal) B Total and isoform-specific Ntrk2 RNA expression in astrocytes and whole cortex from juvenile animals. The majority of Ntrk2 isoform expression in astrocytes is attributed to the truncated TrkB.T1 receptor isoform. Quantitative PCR analysis of acutely isolated CNS populations from juvenile mice for C overall Ntrk2, D TrkB.FL, or E TrkB.T1 mRNA expression, normalized to matched whole cortex. F Ntrk2 receptor isoform mRNA expression in astrocytes across development. Expression of TrkB.T1 is highest in juvenile animals, when G Bdnf mRNA is highest in cortical tissues. Data represented as mean +/-SEM, n = 3-6 animals. This result prompted us to evaluate total, full length, and truncated Ntrk2 mRNA 98 expression in astrocytes relative to other CNS cell populations. Sequential isolation of 99 oligodendrocytes, microglia, astrocytes, and neurons was performed as we have previously 100 described (Holt and Olsen, 2016) in late juvenile mice. Cellular purity was confirmed via qPCR by 101 evaluating cell type specific gene expression (Fig S1). Oligodendrocytes were excluded for 102 subsequent analysis due to lack of cellular purity (Fig. S1). QPCR analysis of total and isoform-103 specific Ntrk2 mRNA expression indicated total TrkB (primer detects both isoforms, grey in Fig   104   1A) was most highly expressed in astrocytes, relative to neurons or microglia. TrkB.FL is the 105 predominant isoform expressed by neuronal populations (Fig. 1d). As indicated by the above RNA 106 sequencing data, astrocytes predominately expressed the truncated TrkB.T1 expression ( Fig.   107 1e). Expression of the truncated receptor is highest during the juvenile period (PND 28, Fig. 1f) 108 relative to young (PND 8) and adult (PND 60) astrocytes when total availability of BDNF peaks in 109 the cortex (Fig. 1g)  Novel serum-free primary astrocyte culture yields morphologically complex astrocytes. A Cartoon representation of magnetic separation of astrocytes for culture in serum-free, defined media. B Quantitative PCR data from cultured astrocytes at 7 and 14 DIV demonstrates purity of cultured cells. C mRNA expression of Ntrk2 isoforms in cultured astrocytes compared to agematched acutely isolated astrocytes demonstrates a developmental upregulation of TrkB.T1 expression. D Representative image of GFAP and TrkB immunofluorescence shows localization of TrkB to astrocytic membrane. E Shape index analysis and F cumulative frequencies of astrocytes cultured in serum-containing or serum-free media and membranous immunolabeling demonstrates increased cellular complexity in serum-free conditions. f Representative images of GFAP+ astrocytes cultured for 14DIV in the presence or absence of serum, and representatives images of membrane (Glast)/GFAP staining. Data represented as mean +/-SEM, n = 3-6 cultures, with 2 biological replicates per culture, *p < 0.05. (Kondo et al., 1995) and Ezrin-a member of the ERM protein family-links the plasma 150 membrane to the actin cytoskeleton, and has been previously demonstrated to be localized to

157
BDNF induces an increase in astrocyte morphological complexity.

158
Given the high levels of astrocytic TrkB expression during a period of astrocyte 159 morphological maturation we next evaluated a role for BNDF on astrocyte morphology.

160
Astrocytes were isolated and cultured as described above, and experiments performed after 14

216
In vivo loss of TrkB.T1 decreases astrocyte morphogenesis

217
The experiments above established that BDNF signaling through the TrkB.

381
We demonstrated that scavenging BDNF from the media within an hour of exposure 382 prevented the increase in cellular complexity 24 hours later, suggesting that BDNF must be 383 actively present to elicit an increase in astrocyte morphological complexity. This experiment is 384 particularly interesting given that following synaptogenesis, BDNF secretion from neurons is 385 largely targeted to synaptic zones and is secreted in an activity-dependent manner (Park and

498
25V for 7 min), followed by 1 hour blocking with LI-COR blocking buffer at a 1:1 ratio with TBS.

499
Primary antibodies, including concentration and incubation times, are given in Table 2. All 500 secondary antibodies were LI-COR, and incubated at 1:10,000 for 1 hour at room temperature.

501
Imaging was performed on a LI-COR Odyssey machine on both the 680 and 800 channels.