• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Curr Biol. Author manuscript; available in PMC Jun 5, 2008.
Published in final edited form as:
PMCID: PMC2175069
NIHMSID: NIHMS25352

Genetic modulation of BDNF signaling affects the outcome of axonal competition in vivo

Summary

Background

Activity-dependent competition that operates on branch stability or formation, plays a critical role in shaping the pattern and complexity of axonal terminal arbors. In the mammalian central nervous system (CNS), the effect of activity-dependent competition on axon arborization and the assembly of sensory maps is well established. However, the molecular pathways that modulate axonal branch stability or formation in competitive environments remain unknown.

Results

We establish an in vivo axonal competition paradigm in the mouse olfactory system, by employing a genetic strategy that permits suppression of neurosecretory activity in random subsets of olfactory sensory neurons (OSNs). Long-term follow up confirmed that this genetic manipulation triggers competition by revealing a bias towards selective stabilization of active arbors and local degeneration of synaptically silent ones. Using a battery of genetically modified mouse models we demonstrate that a decrease either in the total levels or the levels of activity-dependent secreted BDNF (due to a val66met substitution), rescues silent arbors from withering. We show that this effect may be mediated, at least in part, by p75(NTR).

Conclusions

We establish and experimentally validate a genetic in vivo axonal competition paradigm in the mammalian CNS. Using this paradigm we provide evidence for a specific effect of BDNF signaling on terminal arbor pruning under competition in vivo. Our results have implications for the formation and refinement of the olfactory and other sensory maps, as well as for neuropsychiatric diseases and traits modulated by the BDNF val66met variant.

Introduction

The formation of functional neural networks requires precisely orchestrated regulation of the topographic projections of axons, as well as of the growth and branching of their terminal arbors, a process often modulated by neural activity [1]. In particular, activity-dependent competition that appears to operate by affecting branch stability or formation of new branches, plays a critical role in shaping the pattern and complexity of axonal terminal arbors in the mammalian peripheral motor axons [2] and the visual system of lower vertebrates, such as the Xenopus [3] and zebrafish optic tectum [4]. In the mammalian central nervous system (CNS), the effect of activity-dependent competition on the assembly of sensory systems is well established [5] and an understanding of this effect at the level of individual arbor arborization has been gained in considerable detail, especially in the visual system [6]. Experiments involving reconstruction of single afferent arbors revealed that differential growth and selective pruning of afferent arbors is likely to underlie, at least in part, the functional effects of activity based-competition in the visual cortex during the critical period [7-9]. In the olfactory system, a number of recent genetic experiments began to reveal a prominent role of activity-based competition on OSN survival and targeting [10-13]. However, whether and how competition functions in vivo during the growth and branching of the olfactory terminal arbors is unknown.

The molecular pathways and retrograde signals modulate branch stability or formation of new branches in a competitive milieu remain largely unknown. Establishment of robust in vivo competition paradigms that allow investigation at the individual arbor level and can reliably dissociate genuine competition-driven effects from other non-specific ones, are important in efforts to identify molecular pathways that mediate competition-driven axonal arborization. Here, we establish such an in vivo competition paradigm in the mouse olfactory system, by employing a previously described genetic strategy [11] that permits temporally and spatially controlled suppression of neurosecretory activity in random subsets of OSNs by the expression of tetanus toxin light chain (TeTxLC). We analyze the long-term consequences of blocking neurotransmitter release in individual olfactory axons, and demonstrate the morphological consequences of such activity perturbation at the level of individual axon arbors and their branches. Using a battery of genetically modified mouse models we demonstrate that a decrease either in the total levels or the levels of activity-dependent secreted BDNF (due to a val66met substitution), as well as a decrease in the levels of the neurotrophin receptor p75(NTR) rescues silent arbors from withering within this competitive environment.

Results

A genetic strategy to trigger competition

We employ a genetic strategy [14] that permits temporally and spatially controlled suppression of neurosecretory activity in subsets of mature OSNs by the expression of TeTxLC. We have previously described a mouse line that, when crossed to an Olfactory Marker Protein (OMP)-IRES-tTA line, can direct conditional expression of TeTxLC and tau-lacZ under the control of the tetO operator (tetO-TeTxLC-IRES-tau-lacZ) in the majority (~70-80%) of mature OSNs. Using this line we established that widespread expression of TeTxLC in OSNs results in depletion of synaptobrevin (VAMP2), efficient blocking of synaptic vesicle recycling and efficient blocking of postsynaptic mitral cell activation in the olfactory bulb (OB) [11]. We will refer to this line as High Frequency expressing line (HF-TeTxLC). We have now generated tetO-TeTxLC-IRES-tau-lacZ mouse lines that, when crossed to the same OMP-IRES-tTA line, can direct expression of TeTxLC and tau-lacZ at low frequency, and in a sparse mosaic pattern, in a subset (≤ 1%) of OSNs (Fig. 1A, see also Supplementary Experimental Procedures). Expression of the axonal marker tau-lacZ in a small number of OSNs generates a ‘Golgi-like’ pattern that allows visualization of the complexity of the terminal arbors (Fig. 1). We will refer to these lines as Low Frequency expressing lines (LF-TeTxLC). In addition, we established several control founder transgenic lines permitting tetO-dependent expression of tau-lacZ at low frequency (LF-tauLacZ) or high frequency (HF-tauLacZ). It is not possible to directly assess VAMP2-clevage and synaptic transmission in individual silent arbors in the LF- lines, but comparison with the expression level in the HF-TeTxLC lines shows that although the number of expressing neurons is drastically lower in the LF-TeTxLC lines, the level of transgene expression per neuron (Supplementary Figure 1) is identical.

Figure 1
Terminal arbor complexity in LF-and HF-TeTxLC lines

Reduction of terminal arbor complexity in LF-TeTxLC but not in HF-TeTxLC lines

First, we examined the consequences of inhibition of synaptic release on terminal arborization during the postnatal development of olfactory synapses. Axon arbors were evaluated by the number of branch points and the total branch length [15]. Consistent with previous results in rats [15], analysis of individual axons in control LF-tauLacZ lines revealed a rapid increase in the complexity of terminal arbors from P0 to P6 (not shown) followed by a plateauing of the variables to the adult values, without any evidence of pruning an initially exuberant arborization (Fig. 1B, top panel). At this point, OSN axons give rise to a complex arbor of branches that is limited in the number of branch points (6.6 ± 0.19) per axon arbor and in total length (231.6 ± 11.9 μm) (Fig. 1C, D). The intraglomerular region supplied by each labeled terminal arbor is generally randomly distributed within the glomerular structure, but spatially restricted. Similar analysis of individual axons in LF-TeTxLC lines showed that the number of branches and the total length of branches also matured to normal adult levels from P0 to P6 (not shown). After P6, however, the terminal arbor complexity started to decrease dramatically resulting by day P14 in a relatively simple arborization pattern (Fig. 1B, bottom panel) with 1.96 ± 0.17 branches per axon arbor (P = 5.67E-8) and an average total length of 153.29 ± 13.16 μm (P = 0.002) (Fig. 1C, D). Moreover, as expected, the majority of wild-type active axons from LF-TeTxLC lines labeled with the lipophilic dye DiI demonstrated a wild-type pattern of terminal arbor growth (Fig. 1B, insert).

We also measured the arbor growth of activity-suppressed cells in the HF-TeTxLC line where the synaptic activity of nearby axons is also suppressed (see Supplementary Information). In contrast to the LF-TeTxLC lines this analysis failed to reveal any decrease in the pattern of complexity of terminal arbors at any point during the postnatal development of the olfactory system in the HF-TeTxLC line (Fig. 1E). Indeed, inhibition of synaptic release in a non-competitive environment resulted in a modest increase of the number of terminal branches compared to that of the control (Fig. 1F-G; see also Supplementary Information). Thus, pre-synaptically inactive axons fare differently when confronting many other normally active axons than when mingled with other neighboring weak inputs. This observation confirms the presence of activity-based competition in LF-TeTxLC lines.

A permissive effect of NMDA receptor (NMDAR) activity on competition among terminal arbors

We asked whether the reduction of inactive branches in LF-TeTxLC lines is merely a passive consequence of presynaptic inactivity of the losing terminals, or rather, requires formation of functional synapses between “winning” active axons and post-synaptic neurons. We addressed this question indirectly, by investigating whether the effect of activity-based competition on olfactory axon branching is governed by postsynaptic NMDAR-dependent activity as has been demonstrated in lower vertebrates [16]. To this end, we reconstructed terminal arbors from LF-TeTxLC, LF-tauLacZ and HF-TeTxLC animals treated during early postnatal life with MK-801, a blood-brain barrier permeable noncompetitive NMDAR antagonist that selectively blocks open NMDAR channels (see Supplementary Information). MK-801 at 0.25 mg/kg [17], or saline, were administered systemically during P6 - P10, a period that coincides with the onset of inactive branch destabilization observed in the LF-TeTxLC mice. Analysis of individual axons in LF-TeTxLC lines treated with MK-801, or vehicle, showed that, as described above, all tested arbor complexity parameters plateaued at normal adult levels at around P6. At this point, however, analysis of individual axonal arbors in MK-801-treated mice failed to reveal the dramatic decrease in complexity observed in saline-treated or untreated LF-TeTxLC mice (Fig. 2A). Rather, following the termination of MK-801 administration, the number of branches remained stable at normal levels over a period of at least four days (P10 - P14) (Fig. 2B, C). By contrast, four weeks after cessation of MK-801 administration and clearance of the drug, simple degenerating terminals re-appeared at very high frequencies (Fig. 2A, B, C), consistent with our finding that mature synapses are also susceptible to the effects of competition (LC, JAG unpublished). Thus, activation of postsynaptic NMDAR by “winning” active axons can influence or help mediate competition-induced withering of inactive axon arbors. Therefore, by allowing competition-induced presynaptic rearrangements to unfold, postsynaptic NMDAR-dependent signaling appears to play a permissive role in this process.

Figure 2
NMDA receptor activity plays a permissive role in competition among terminal arbors

In contrast to the dramatic effects on the stability of presynaptic axonal branches in LF-TeTxLC lines, systemic treatment with the NMDAR blocker was generally insufficient to influence substantially axon arbor morphology in control LF-tauLacZ mice (Fig. 2A). Specifically, similar systemic treatment with MK-801 during P6 - P10 of control LF-tauLacZ mice resulted in a modest increase in the number of branches as compared to saline-treated control lines (from ~7 to ~9 branches per axon, P = 0.029) (Fig. 2B). The total branch length did not change significantly (P = 0.93) (Fig. 2C) due to a concomitant decrease of the average branch length (from 33.15 ± 2.48 μm to 25.44 ± 2.48 μm, P = 0.017). Notably, this response of the control mice to global postsynaptic blockade is analogous to the one observed following global presynaptic inhibition in HF-TeTxLC lines. This observation strongly suggests that the effect observed in the latter case is likely mediated via decreased activation of postsynaptic NMDA receptors. Consistent with this interpretation, analysis on the effects of systemic MK-801 administration to the HF-TeTxLC line failed to reveal any significant additional changes in the arbor complexity (Fig. 2A, B, C).

Activity-based competition results in local degeneration of inactive branches

Genetic manipulations that result in a competitive environment in the olfactory epithelium and bulb may eventually affect neuronal health and survival [10, 12, 13] and thus indirectly affect terminal arbor rearrangements. However, we found no evidence of increased cell death the olfactory epithelium of LF-TeTxLC mice (see Supplementary Information and Supplementary Fig. 3). We therefore asked whether instead of cell death, activity-based competition leads to local elimination of inactive branches and if so, what form it takes (i.e. degeneration or branch tip retraction [18]. The formation of smooth large axonal varicosities (LAV, 2.5-7 μm in length) due to accumulation of organelles and disorganized cytoskeleton, has been proposed as one of the morphological signatures of axon degeneration [19]. Therefore, we initially evaluated the number of LAVs in LF-TeTxLC and LF-tauLacZ axon arbors at the light microscopic level (Fig. 3A). There is a marked difference in the frequency of LAVs between LF-TeTxLC and LF-tauLacZ axon arbors (Fig. 3B). Moreover, in comparing to that of control LF-tauLacZ lines (1.84 ± 0.18 at age P14 and 2.00 ± 0.18 per 100 μm branch in adult) we found that the average LAV density is significantly higher in LF-TeTxLC axon arbors (4.58 ± 0.79 at age P14, P = 0.019; and 5.87 ± 0.57 per 100 μm branch at P28, P = 0.00024) (Fig. 3C). Correlated light and electron microscopical analysis showed that in labeled axons in control LF-tauLacZ lines, small round terminals 1-2 μm in diameter were positioned en passant along axon arbor branches [Fig. 3Da (c indicated), c; e (f indicated), f]. Ultrastructurally, such normally active terminals had abundant and uniform spherical vesicles, as well as an asymmetrical membrane thickening on the postsynaptic dendrite [20]. By contrast, the LAVs, more prominent on LF-TeTxLC axons, commonly contain membranous organelles, multivesicular bodies and engulfed structures [(compare LAV on LF-tauLacZ line axon in Fig. 3D a(d indicated), d with LAVs in Fig. 3D e(g indicated), g; h(i indicated), i]. The small terminals on the LF-TeTxLC axons have the cytology of the small terminals on LF-tauLacZ lines (Fig. 3D e(f indicated), f)), but also show signs of degeneration (Fig. 3D h(j,k indicated).

Figure 3
Competition results in local degeneration of inactive branches

Structural rearrangements in synaptic connectivity, such as the ones observed here, are often modulated by neurotrophins, including BDNF, via a mechanism that influences both synapse and axon branch stability [21]. We therefore asked whether BDNF signaling plays a role in shaping the pattern and complexity of axonal terminal arbors in a competitive environment.

Reduction in BDNF levels rescues silent arbors from pruning

Mouse BDNF appears to be absent from OSNs and their axon terminals, but it is present in the soma and dendrites of mitral and periglomerular cells [22] and Fig. 4A. To investigate if BDNF signaling plays a role in the structural rearrangement described here, we re-analyzed the fate of terminal arbors in LF-TeTxLC lines in a BDNF-deficient environment. We used a previously described strain where the mouse BDNF gene is inactivated by homologous recombination [23] and a series of standard crosses to generate mice heterozygous for the BDNF mutation that also carry the OMP-IRES-tTA gene and the LF-tetO-TeTxLC-IRES-tau-lacZ transgene. Because BDNF homozygous mutant mice die at high rates within the first two postnatal weeks [23] we restricted our analysis to BDNF heterozygous mutant mice. Reconstruction of olfactory terminals from BDNF (+/-) ; LF-TeTxLC and wild-type BDNF (+/+) ; LF-TeTxLC littermates revealed that decrease in complexity observed in LF-TeTxLC mice at the end of the second postnatal week, is prevented in the background of BDNF (+/-) mice (Fig. 4B). Instead, the arbor complexity in BDNF (+/-) ; LF-TeTxLC mice is maintained at the wild-type level, that is ~7 branches per axon and average branch length of 249.37 ± 33.05 μm (Fig. 4C, D). Moreover, there is a sharp decrease in the density of LAVs (1.92 ± 0.14 per 100 μm branch, P = 0.00051 compared to that of LF-TeTxLC) to the levels observed in wild-type tauLacZ control mice (Fig. 4E). This suggests that reduction of endogenous BDNF level by half [24] is sufficient to reduce the competitive advantage of the active axons in LF-TeTxLC mice.

Figure 4
BDNF levels influence competition-driven terminal axon arborization

Contrary to the dramatic effects on the stability of axonal branches in LF-TeTxLC lines, BDNF-deficiency did not influence substantially axon arbor morphology in control LF-tauLacZ. OSNs from BDNF (+/-) ; LF- tauLacZ mice showed a modest increase in the number of branches as compared to BDNF (+/+) ; LF-tauLacZ (from ~7 to ~9, P = 0.015) (Fig. 4C). Again, the total branch length did not change significantly due to a concomitant decrease of the average branch length (from 34.3 ± 1.86 to 26.3 ± 2.71 μm, P = 0.016). Finally, comparison of the arbor complexity between BDNF (+/-) ; HF-TeTxLC and BDNF (+/+) ; HF-TeTxLC lines (Fig. 4B, C, D), or between MK-801 treated BDNF (+/-) or BDNF (+/+) mice (data not shown) failed to reveal any significant additional changes in arbor complexity. This finding suggests that the effect of BDNF on competition-induced pruning is likely downstream of OSN synaptic activity. Reduction of sensory activity during the first two postnatal weeks does not affect BDNF levels in mitral cells in rats [22]. Consistent with these findings, we do not see any significant changes in BDNF levels in mitral cells from HF-TeTxLC line at P14 (data not shown). Thus, activity-dependent modulation of total BDNF levels does not appear to be important and instead regulated (activity-dependent) release of BDNF may be involved.

A knocked-in BDNFmet variant deficient in regulated secretion rescues silent arbors from pruning

Measuring directly regulated release of BDNF in the glomeruli is not feasible. However, a recently identified primate-specific variant in the BDNF prodomain (val66met) that has shown to impair regulated secretion both in vitro and in vivo [25] has provided us with a genetic tool to investigate whether regulated release of BDNF plays a role in the structural rearrangement described here. We generated two mouse strains where the mouse BDNF coding sequence was substituted by the corresponding human BDNF sequence carrying either a met (BDNF met) or a val (BDNFval) allele (see Fig. 5A and Experimental Procedures).

Figure 5
BDNFmet variant influences the competition-driven terminal axon arborization

Using a series of standard crosses, we re-analyzed the fate of terminal arbors of the LF-TeTxLC lines in homozygous human BDNF(met/met) and human BDNF(val/val) background, as well as in wild-type mouse BDNF (+/+) littermate mice. Reconstruction of olfactory terminals from BDNF(met/met) ; LF-TeTxLC, BDNF(val/val) ; LF-TeTxLC and BDNF (+/+) ; LF-TeTxLC littermates revealed that the decrease in complexity observed in LF-TeTxLC mice at the end of the second postnatal week, is prevented in the background of BDNF(met/met) mice but not in the background of BDNF(val/val), or wild-type BDNF (+/+) littermate controls (Fig. 5B). Instead, the arbor complexity in BDNF(met/met) ; LF-TeTxLC mice is maintained at ~6.5 branches per axon and the total branch length at 208.74 ± 10.37 μm (Fig. 5C, D). Moreover, there is a sharp decrease in the density of LAVs (1.83 ± 0.37 per 100 μm branch, P < 0.001) compared to that of LF-TeTxLC to the levels observed in LF-tauLacZ control mice (6.0 ± 0.6, Fig. 5E). This suggests that reduction in regulated BDNF release is sufficient to reduce the competitive advantage of the active axons in LF-TeTxLC mice. Contrary to the dramatic effects on the stability of presynaptic axonal branches in LF-TeTxLC lines, BDNFmet does not influence substantially axon arbor morphology in control LF-tauLacZ (Fig. 5B). OSNs from BDNF(met/met) ; LF- tauLacZ mice showed almost the same number of branches as compared to BDNF(val/val) ; LF-tauLacZ and the total branch length did not change significantly (Fig. 5C, D). This discrepancy with the BDNF (+/-) ; LF- tauLacZ line (Fig. 4C), if not due to experimental variation, it may reflect differences in the degree of release reduction [25].

Reduction in p75(NTR) levels rescues silent arbors from pruning

A good candidate to mediate the inhibitory effects of BDNF is the neurotrophin receptor p75(NTR), which can promote inhibition in certain contexts by integrating diverse growth-inhibitory cues, including those from neurotrophins and myelin proteins [26]. p75(NTR) is localized to OB glomeruli [26], the ensheathing glia [27], as well as to periglomerular cells, but not to mitral cells (Fig. 6A).

Figure 6
p75(NTR) levels influence competition-driven terminal axon arborization

We used a previously described strain where the p75(NTR) gene is mutated in exon III by homologous recombination [28] to generate mice heterozygous or homozygous for the p75(NTR) mutation that also carry the OMP-IRES-tTA gene and the LF-tetO-TeTxLC-IRES-tau-lacZ transgene. Reconstruction of olfactory terminals from heterozygous p75(NTR) (+/-), homozygous p75(NTR) (-/-) and wild-type p75(NTR) (+/+) ; LF-TeTxLC littermates revealed that that the decrease in complexity observed in LF-TeTxLC mice at the end of the second postnatal week, is prevented in gene-dosage dependent manner in the background of p75(NTR) deficient mice (Fig. 6B, C, D). Arbor complexity in heterozygous p75(NTR)(+/-) ; LF-TeTxLC mice is maintained at ~4 branches per axon, whereas arbor complexity in homozygous p75(NTR) (-/-) ; LF-TeTxLC mice is maintained at the wild-type level of ~7 branches per axon (Fig. 6C). Unlike the effect observed in heterozygous BDNF (+/-) ; LF-TeTxLC mice, p75(NTR) deficiency did not restore the total branch length to wild-type levels (Fig. 6D) and only partially abolished emergence of LAVs (Fig. 6E). Thus, reduction or elimination of endogenous full-length p75(NTR) is sufficient to reduce the competitive advantage of the active axons in LF-TeTxLC mice, but not to the extent observed following reduction in BDNF levels (Fig. 6 C-E) (see also Supplementary Information). Finally, comparison of the arbor complexity between p75(NTR)(-/-) ; HF-TeTxLC and p75(NTR) (+/+) ; HF-TeTxLC lines (Fig. 6B, C, D), failed to reveal any significant additional changes in arbor complexity, although the positive effect of global pre-synaptic inactivity on the total branch length appeared to be somewhat attenuated in p75(NTR)(-/-) mice.

Discussion

Here we present a genetic approach in which individual axon terminals can be visualized at high resolution, and synaptic vesicle release can be manipulated in small random subset of OSN axons to mimic a genuine competitive process. To our knowledge, this is the first genetic in vivo competition paradigm at single axon resolution in the olfactory system and in the mammalian CNS, in general. We use this competition paradigm to begin identifying the molecular and cellular substrates upon which activity-based competition may operate.

At the molecular level, we provide evidence for a specific effect of BDNF signaling on terminal arbor pruning under competition in vivo. What are the potential explanations for the observed neurotrophin effects? According to the “trophic” models that are traditionally evoked to explain outcomes of axonal competition, synaptic activity enhances axonal growth mediated by a limiting trophic factor. Therefore, one possibility is that BDNF is one of the elusive limiting trophic factors which is “soaked-up” by active and synaptically stabilized terminals leading to a “starvation”-induced death of the synaptically inactive ones. However, it appears counterintuitive that reduction of the already limiting trophic factor levels, will be sensed by competing terminals as less restrictive than normal levels. Another, conceptually related possibility is that neurotrophin deficiency indirectly alters presynaptic function [29], or interferes with maturation of postsynaptic neurons or synapses between OSNs and postsynaptic neurons [30] and, as a result, active OSN axons are prevented from “expressing” their competitive advantage. A third possibility is that BDNF or pro-BDNF is a ‘punitive’ signal or, more generally, modulates the signaling cascades involved in “punitive processes”, perhaps acting through p75(NTR) leading to active retraction of inactive axons. This possibility is also supported by the finding that p75(NTR) can promote inhibition by integrating diverse growth-inhibitory cues [31] and most importantly by a recent independent observation that p75(NTR) can mediate the inhibitory effects of BDNF on axon growth in a competitive environment in vitro [32]. In addition, p75(NTR) was recently shown to mediate local activation of caspases at the presynaptic terminals of OSNs following NMDA-mediated excitotoxic death of OSN target neurons in the OB [26]. Given that the biologically active neurotrophin pro-forms bind to p75(NTR) with approximately 1000-fold higher affinity compared to the mature form [33], it is tempting to speculate that presynaptic activity- and NMDAR activity-dependent secretion of pro-BDNF by postsynaptic mitral and periglomerular neurons induces axon branch regression via p75(NTR)-mediated signaling [33]. Interestingly, recent studies have established that a significant proportion of the processing of proBDNF to mature BDNF occurs extracellularly by proteases on the cell surface or extracellular space, which determine the balance of BDNF forms (pro or mature) produced [34]. Therefore, TeTxLC-expressing neurons with impaired synaptic vesicle turnover may be more vulnerable because, for example, they fail to secrete extracellular proteases that trigger proteolytic cleavage of proBDNF [35], or because of local upregulation of p75(NTR) protein levels and signaling at the presynaptic terminal. It should be noted, however, that such interpretations are tempered by the fact that in addition to olfactory terminals, p75(NTR) is also expressed in periglomerular cells and ensheathing glia, and therefore the site of p75(NTR) action and whether it exerts a direct or indirect effect on presynaptic OSN terminals is still unclear. In addition, although reduction in both BDNF and p75(NTR) levels rescue inactive arbors from pruning, there are important differences in the pattern and extent of rescue. Whether such differences are due to involvement of additional neurotrophin receptors and ligands will be the subject of future investigation.

At the cellular level, our analysis demonstrated that branch elimination rather than cell death is the most proximal effect of competition, with disabled OSNs with fewer terminal branches and enlarged terminals lingering for extended periods of time. The axonal degeneration observed here appears to be mechanistically related to the local degeneration events (short branch withdrawal) observed in cortical axons [18] and in the Drosophila mushroom body neurons [36], which is partly mediated by the ubiquitin-proteasome system [37].

Our results may have important implications for understanding the topographic formation and refinement of the olfactory, as well as of other sensory maps [38]. In the olfactory system, in particular, previous work revealed that activity-based competition modulates the OSN projection pattern [11]. Here we show that similar to the visual system, competition also modulates the branching of the olfactory terminal arbors. Whether these two effects are causally linked is still unknown and will be the focus of future experiments. The possibility that BDNF signaling contributes to the initial formation of the olfactory bulbar map has been convincingly excluded [39] but its potential role on map refinement [12] is largely unexplored. By contrast, p75(NTR) has already been implicated in normal topographic projections in the olfactory system, based on the appearance of ectopic and disorganized glomeruli in the p75(NTR) (-/-) mice [40].

Finally, our results also provide novel insights into neuropsychiatric diseases and traits modulated by the val66met BDNF variant. It is worth noting that one of the most consistent effects of this variant is in modulating the risk of childhood onset psychiatric disorders [41, 42]. In these cases the effect of the met66 allele appears to be almost invariably protective. In that regard, it is tempting to speculate that this allele may protect against these disorders, at least in part, by restricting the level of activity-dependent pruning of afferents in key brain regions during development depending on the local expression pattern of neurotrophins and their receptors. Consequently, the BDNFmet and BDNFval strains could become a useful genetic tool to test this hypothesis.

Experimental Procedures

Generation of Transgenic Mice

Generation of tetO-TeTxLC-IRES-tau-lacZ transgenic mice has been described in detail previously [11] and is described in more detail in Supplementary Experimental Procedures. The BDNF and p75(NTR) mutants are also described in Supplementary Experimental Procedures.

Generation of BDNFmet and BDNFval knock-in alleles

A 370 bp fragment, located 5′ of the mouse BDNF coding region, was amplified from genomic mouse DNA using the following primers: 5′-ACAGATGTAGTAAAACGTTGGAG-3′, 5′-TTACTGATCCACTCCAGCTGC-3′. This fragment was subcloned into pGEMT (Promega) using T-A cloning and was used as a probe against a 129 Sv/Ev BAC library. One of the identified BAC clones was expanded for use in the generation of targeting constructs. Flanking sequence 11 kb upstream (long arm) and 3 kb downstream (short arm) of the BDNF gene was PCR-ed from the BAC clone using the following sets of primers: 5′-GCGGCCGCCAGGCTCTATTTGATTATAAAATAG- 3′, 5′-GGCCGGCCATGTGCACTGAATTTCAGTTCAG-3′; 5′-ACGCGTCGACTGACT GCCTGCGACAAACTT-3′, 5′-GGCGCGCCTCAGCC CTGGTTCATGGATCCTG-3′). Additionally, 274 bp of sequence including the 5′UTR and a portion of the coding sequence was amplified from human DNA extracted from a Val/Val or a Met/Met individual using the following primers: 5′-ACCAGGTGAGAAGAG TGATGACCATCCTTTTCCTTAC-3′, 5′-CACCCGGGA CGTGTACAAGTC-3′. The long arm was further modified into the long armval and the long armmet fragments by excising the mouse sequence located between unique XmaI and SexAI sites, aligned with the human 274 base pair fragment, and replacing it with sequence from the Val/Val or the Met/Met individuals. Finally the BDNFmet and BDNFval knock-in constructs were assembled, by cloning the corresponding long arms into the AscI site of the pACNIII targeting construct [14] (5′ of an self-excisable neo cassette) and the short arm into the NotI and FseI sites, located at the 3′ of this cassette. All PCR reactions related to the generation of this construct were conducted using the Expand High Fidelity kit (Roche), and subcloning steps involved the use of the Rapid Ligation Kit (Roche) or the Infusion Kit (BD Biosciences). 70 μg of each plasmid (BDNFmet and BDNFval) was electroporated into 129 Sv/Ev embryonic stem cells. Twenty-four hours after electroporation, neomycin selection was applied and approximately 400 clones were picked for each construct after five days of selection. These clones were analyzed using a PCR approach, expanded and confirmed with Southern blot analysis. The restriction enzyme MfeI was used to digest ES cell DNA, and a probe located outside the targeting construct (at the 3′ end) generated a 5.9 kb targeted band and a 3.8 kb wild-type band in correctly targeted clones. Both constructs recombined at a rate of ~1%. Upon germline transmission, DNA extracted from tail biopsy samples of both lines of mice was genotyped using the human primers described above followed by a diagnostic restriction enzyme analysis with either PmlI (BDNFval) or BmgBI (BDNFmet).

Tissue Preparation

Mice were transcardially perfused with cold 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer saline (PBS) and the brains were removed and post-fixed in 4% PFA overnight at 4°C. Fixed brains were used for immunochemistry or dye embedding.

Immunohistochemistry

as described in Supplementary Experimental Procedures.

Dye labeling

as described in Supplementary Experimental Procedures.

Evaluation of Axon Arbors

Axonal arbors were evaluated as described previously [15] for number of branch points and total length of branches. The total length of branches (in micrometers) was measured as the summed length of all branches from the first branch point within the glomerulus onward. All variables were statistically evaluated using the Mann-Whitney Test. A minimum of 10 axonal arbors from each animal (n = 5-6 animals per age group/per genotype) were evaluated. In the LF- lines the majority of the labeled glomeruli (>70%) contained one or two synaptically-inactive axons and the rest contained three or more axons. Importantly, we did not observe any spatial bias in the distribution and bulbar projections of expressing OSNs in the LF lines. In order to account for some potential confounding influences related to the location of the examined OSNs in the olfactory epithelium and the location of their projections to the OB, we sampled from the entire bulb without any spatial bias. Because of the high background associated with DiI, the number of glomeruli with one or two terminal was relatively low (~30%) but we made every effort to sample from several animals and bulbar locations to minimize any spatial bias.

Electron Microscopy

As described in Supplementary Experimental Procedures..

In vitro Tracer Injections

As described in Supplementary Experimental Procedures.

Drug Treatment

As described in Supplementary Experimental Procedures.

Western Blotting

As described in Supplementary Experimental Procedures.

Supplementary Material

01

Acknowledgments

The authors thank Richard Axel and Jane Dodd for comments on the manuscript. We also thank Dionne Swor for technical assistance, Megan Sribour for assistance with the generation and maintenance of the mouse colony and Jennifer Power, Ben Shykind, ChangRon Yu, Pei-Ken Hsu, Maggie O’Meara, Jonathan Rockey and Eugene Fayerberg for help and reagents. This work was supported by the National Institutes of Health (DC006292), a Burroughs Wellcome Fund Career Award in the Biomedical Sciences, a grant from the Whitehall Foundation and NARSAD (to JAG) and EY012736 (to CAM).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. Goda Y, Davis GW. Mechanisms of synapse assembly and disassembly. Neuron. 2003;40:243–264. [PubMed]
2. Buffelli M, Burgess RW, Feng G, Lobe CG, Lichtman JW, Sanes JR. Genetic evidence that relative synaptic efficacy biases the outcome of synaptic competition. Nature. 2003;424:430–434. [PubMed]
3. Ruthazer ES, Akerman CJ, Cline HT. Control of axon branch dynamics by correlated activity in vivo. Science. 2003;301:66–70. [PubMed]
4. Hua JY, Smear MC, Baier H, Smith SJ. Regulation of axon growth in vivo by activity-based competition. Nature. 2005;434:1022–1026. [PubMed]
5. Torborg CL, Feller MB. Spontaneous patterned retinal activity and the refinement of retinal projections. Prog Neurobiol. 2005;76:213–235. [PubMed]
6. Antonini A, Fagiolini M, Stryker MP. Anatomical correlates of functional plasticity in mouse visual cortex. J Neurosci. 1999;19:4388–4406. [PMC free article] [PubMed]
7. Bence M, Levelt CN. Structural plasticity in the developing visual system. Prog Brain Res. 2005;147:125–139. [PubMed]
8. Antonini A, Stryker MP. Plasticity of geniculocortical afferents following brief or prolonged monocular occlusion in the cat. J Comp Neurol. 1996;369:64–82. [PubMed]
9. Antonini A, Stryker MP. Rapid remodeling of axonal arbors in the visual cortex. Science. 1993;260:1819–1821. [PubMed]
10. Zhao H, Reed RR. X inactivation of the OCNC1 channel gene reveals a role for activity-dependent competition in the olfactory system. Cell. 2001;104:651–660. [PubMed]
11. Yu CR, Power J, Barnea G, O’Donnell S, Brown HE, Osborne J, Axel R, Gogos JA. Spontaneous neural activity is required for the establishment and maintenance of the olfactory sensory map. Neuron. 2004;42:553–566. [PubMed]
12. Zou DJ, Feinstein P, Rivers AL, Mathews GA, Kim A, Greer CA, Mombaerts P, Firestein S. Postnatal refinement of peripheral olfactory projections. Science. 2004;304:1976–1979. [PubMed]
13. Nakatani H, Serizawa S, Nakajima M, Imai T, Sakano H. Developmental elimination of ectopic projection sites for the transgenic OR gene that has lost zone specificity in the olfactory epithelium. Eur J Neurosci. 2003;18:2425–2432. [PubMed]
14. Gogos JA, Osborne J, Nemes A, Mendelsohn M, Axel R. Genetic ablation and restoration of the olfactory topographic map. Cell. 2000;103:609–620. [PubMed]
15. Klenoff JR, Greer CA. Postnatal development of olfactory receptor cell axonal arbors. J Comp Neurol. 1998;390:256–267. [PubMed]
16. Rajan I, Witte S, Cline HT. NMDA receptor activity stabilizes presynaptic retinotectal axons and postsynaptic optic tectal cell dendrites in vivo. J Neurobiol. 1999;38:357–368. [PubMed]
17. Kakizawa S, Yamasaki M, Watanabe M, Kano M. Critical period for activity-dependent synapse elimination in developing cerebellum. J Neurosci. 2000;20:4954–4961. [PubMed]
18. Portera-Cailliau C, Weimer RM, De Paola V, Caroni P, Svoboda K. Diverse modes of axon elaboration in the developing neocortex. PLoS Biol. 2005;3:e272. [PMC free article] [PubMed]
19. Coleman M. Axon degeneration mechanisms: commonality amid diversity. Nat Rev Neurosci. 2005;6:889–898. [PubMed]
20. Kasowski HJ, Kim H, Greer CA. Compartmental organization of the olfactory bulb glomerulus. J Comp Neurol. 1999;407:261–274. [PubMed]
21. Hu B, Nikolakopoulou AM, Cohen-Cory S. BDNF stabilizes synapses and maintains the structural complexity of optic axons in vivo. Development. 2005;132:4285–4298. [PubMed]
22. McLean JH, Darby-King A, Bonnell WS. Neonatal olfactory sensory deprivation decreases BDNF in the olfactory bulb of the rat. Brain Res Dev Brain Res. 2001;128:17–24. [PubMed]
23. Ernfors P, Lee KF, Jaenisch R. Mice lacking brain-derived neurotrophic factor develop with sensory deficits. Nature. 1994;368:147–150. [PubMed]
24. Chen ZY, Ieraci A, Teng H, Dall H, Meng CX, Herrera DG, Nykjaer A, Hempstead BL, Lee FS. Sortilin controls intracellular sorting of brain-derived neurotrophic factor to the regulated secretory pathway. J Neurosci. 2005;25:6156–6166. [PMC free article] [PubMed]
25. Chen ZY, Jing D, Bath KG, Ieraci A, Khan T, Siao CJ, Herrera DG, Toth M, Yang C, McEwen BS, et al. Genetic variant BDNF (Val66Met) polymorphism alters anxiety-related behavior. Science. 2006;314:140–143. [PMC free article] [PubMed]
26. Carson C, Saleh M, Fung FW, Nicholson DW, Roskams AJ. Axonal dynactin p150Glued transports caspase-8 to drive retrograde olfactory receptor neuron apoptosis. J Neurosci. 2005;25:6092–6104. [PubMed]
27. Kafitz KW, Greer CA. Olfactory ensheathing cells promote neurite extension from embryonic olfactory receptor cells in vitro. Glia. 1999;25:99–110. [PubMed]
28. Lee KF, Li E, Huber LJ, Landis SC, Sharpe AH, Chao MV, Jaenisch R. Targeted mutation of the gene encoding the low affinity NGF receptor p75 leads to deficits in the peripheral sensory nervous system. Cell. 1992;69:737–749. [PubMed]
29. Lu B, Pang PT, Woo NH. The yin and yang of neurotrophin action. Nat Rev Neurosci. 2005;6:603–614. [PubMed]
30. Lu B. BDNF and activity-dependent synaptic modulation. Learn Mem. 2003;10:86–98. [PubMed]
31. Kaplan DR, Miller FD. Axon growth inhibition: signals from the p75 neurotrophin receptor. Nat Neurosci. 2003;6:435–436. [PubMed]
32. Singh KK, Miller FD. Activity regulates positive and negative neurotrophin-derived signals to determine axon competition. Neuron. 2005;45:837–845. [PubMed]
33. Lee R, Kermani P, Teng KK, Hempstead BL. Regulation of cell survival by secreted proneurotrophins. Science. 2001;294:1945–1948. [PubMed]
34. Chen ZY, Patel PD, Sant G, Meng CX, Teng KK, Hempstead BL, Lee FS. Variant BDNF (Met66) alters the intracellular trafficking and activity-dependent secretion of wild-type BDNF in neurosecretory cells and cortical neurons. J Neurosci. 2004;24:4401–4411. [PubMed]
35. Pang PT, Teng HK, Zaitsev E, Woo NT, Sakata K, Zhen S, Teng KK, Yung WH, Hempstead BL, Lu B. Cleavage of proBDNF by tPA/plasmin is essential for long-term hippocampal plasticity. Science. 2004;306:487–491. [PubMed]
36. Watts RJ, Hoopfer ED, Luo L. Axon pruning during Drosophila metamorphosis: evidence for local degeneration and requirement of the ubiquitin-proteasome system. Neuron. 2003;38:871–885. [PubMed]
37. Hoopfer ED, McLaughlin T, Watts RJ, Schuldiner O, O’Leary DD, Luo L. Wlds protection distinguishes axon degeneration following injury from naturally occurring developmental pruning. Neuron. 2006;50:883–895. [PubMed]
38. Cabelli RJ, Shelton DL, Segal RA, Shatz CJ. Blockade of endogenous ligands of trkB inhibits formation of ocular dominance columns. Neuron. 1997;19:63–76. [PubMed]
39. Nef S, Lush ME, Shipman TE, Parada LF. Neurotrophins are not required for normal embryonic development of olfactory neurons. Dev Biol. 2001;234:80–92. [PubMed]
40. Tisay KT, Bartlett PF, Key B. Primary olfactory axons form ectopic glomeruli in mice lacking p75NTR. J Comp Neurol. 2000;428:656–670. [PubMed]
41. Hall D, Dhilla A, Charalambous A, Gogos JA, Karayiorgou M. Sequence variants of the brain-derived neurotrophic factor (BDNF) gene are strongly associated with obsessive-compulsive disorder. Am J Hum Genet. 2003;73:370–376. [PMC free article] [PubMed]
42. Geller B, Badner JA, Tillman R, Christian SL, Bolhofner K, Cook EH., Jr. Linkage disequilibrium of the brain-derived neurotrophic factor Val66Met polymorphism in children with a prepubertal and early adolescent bipolar disorder phenotype. The American journal of psychiatry. 2004;161:1698–1700. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links