Delivery of different genes into presynaptic and postsynaptic neocortical neurons connected by a BDNF-TrkB synapse

Highlights

• Different genes were delivered into neurons connected by BDNF-TrkB synapses.

• A synthetic peptide neurotransmitter containing BDNF and the His tag was developed.

• This peptide neurotransmitter is designed to bind to TrkB on postsynaptic neurons.

• Antibody-mediated, targeted gene transfer to postsynaptic neurons used anti-His tag.

• The roles of neurons connected by BDNF-TrkB synapses may now be studied.

Abstract

Brain-Derived Neurotrophic Factor (BDNF) signaling through TrkB receptors has important roles in synapse formation, synaptic plasticity, learning, and specific diseases. However, it is challenging to relate BDNF-TrkB synapses to circuit physiology or learning, as BDNF-TrkB synapses are embedded in complex circuits that contain numerous neuron and synapse types. Thus, analyzing the physiology of neurons connected by BDNF-TrkB synapses would be advanced by a technology to deliver different genes into presynaptic and postsynaptic neurons, connected by a BDNF-TrkB synapse. Here, we report selective gene transfer across BDNF-TrkB synapses: The model system was the large projection from rat postrhinal to perirhinal cortex. The first gene transfer, into presynaptic neurons in postrhinal cortex, used a virus vector and standard gene transfer procedures. This vector expresses a synthetic peptide neurotransmitter composed of three domains, a dense core vesicle sorting domain, BDNF, and the His tag. Upon release, this peptide neurotransmitter binds to TrkB receptors on postsynaptic neurons. The second gene transfer, into connected postsynaptic neurons in perirhinal cortex, uses antibody-mediated, targeted gene transfer and an anti-His tag antibody, as the synthetic peptide neurotransmitter contains the His tag. Confocal microscope images showed that using untargeted gene transfer, only 10–15% of the transduced presynaptic axons were proximal to a transduced postsynaptic dendrite. But using targeted gene transfer, ∼70% of the transduced presynaptic axons were proximal to a transduced postsynaptic dendrite. This technology may support studies on the roles of neurons connected by BDNF-TrkB synapses in circuit physiology and learning.

Introduction

Signaling across specific synapses by Brain-Derived Neurotrophic Factor (BDNF) release and subsequent binding to the TrkB receptor plays critical roles in the formation of synapses, synaptic plasticity, long-term potentiation (LTP), learning, and specific diseases (Panja and Bramham, 2014, Yoshii and Constantine-Paton, 2010, Zhao et al., 2017). Of note, in the hippocampus or neocortex, BDNF to TrkB signaling is required for specific types of synaptic plasticity, including LTP, and is associated with specific learning tasks (Egan et al., 2003, Lu, 2003, Patterson et al., 1996). These observations suggest that BDNF to TrkB signaling across specific synapses has an essential role in the synaptic plasticity that supports learning specific tasks. But studies on these issues are challenging because BDNF-TrkB synapses are embedded in complex forebrain circuits, and only a fraction of the synapses are BDNF-TrkB synapses. Forebrain areas contain hundreds to thousands of neuron types, and each type forms precise connections with other neuron types, to support circuit physiology (Dudai, 1989). More specifically, a neocortical column contains at least tens, and probably hundreds, of neuron types, that form precise synaptic connections both within the column and to other areas, and each type of connection likely supports specific physiological functions (Peters and Jones, 1984, Sugino et al., 2006). Of note, current genetic approaches to analyzing circuits and behaviors, including optogenetics or activation of specific signaling or transcriptional pathways, are typically applied to an entire circuit (Dymecki and Kim, 2007, Fenno et al., 2011, Luo et al., 2008, Zhang et al., 2005). Thus, elucidating the role of BDNF to TrkB signaling for synaptic plasticity and learning would benefit from a gene transfer technology that can selectively deliver different genes into the presynaptic and postsynaptic neurons that form a BDNF-TrkB synapse.

For the model system to develop gene transfer selectively across BDNF-TrkB synapses, we used the large projection from postrhinal (POR) cortex to perirhinal (PER) cortex (Burwell and Amaral, 1998b), as this synapse plays an important role in visual object learning, and likely uses BDNF to TrkB signaling across some synapses. POR cortex is required for this learning (Murray et al., 2007, Zhang et al., 2010a), receives afferents from visual areas, and sends efferents to multiple neocortical areas, including PER cortex (Agster and Burwell, 2009, Burwell and Amaral, 1998a). We developed a genetic intervention that targets some of the essential information for this learning to the genetically-modified circuit: Activation of protein kinase C (PKC) pathways in several hundred glutamatergic and GABAergic neurons in POR cortex (via a virus vector) increases activation-dependent release of GABA and glutamate, and supports more accurate learning of new visual object discriminations (Zhang et al., 2005, Zhang et al., 2012c). Some of the essential information for this learning is encoded in the genetically-modified circuit: After gene transfer and learning, small neurochemical lesions that ablate the genetically-modified circuit selectively reduce performance for discriminations learned after gene transfer (Zhang et al., 2010a). Correlatively, this circuit is preferentially activated during this learning (Zhang et al., 2005, Zhang et al., 2010a).

Established genetic technologies for mapping circuits or visualizing synapses support numerous informative experiments, but lack the capability to selectively deliver different genes across a BDNF-TrkB synapse. Anterograde or retrograde projections, across one, or a series, of synapses have been mapped using specific viruses, including Sindbis Virus, Vesicular Stomatitis Virus, Pseudorabies Virus, Rabies Virus, and Herpes Simplex Virus (HSV-1) (reviewed in (Lo and Anderson, 2011)). These technologies deliver the same genes into the presynaptic and postsynaptic neurons, as they use a single virus that spreads across the affected synapses. Synapses between specific neurons have been visualized using mGRASP, but this technology does not selectively deliver genes into connected neurons, as it uses untargeted gene transfer, and lacks synapse type specificity (Feng et al., 2014, Kim et al., 2011). A Rabies Virus-based technology can deliver a gene into postsynaptic neurons, and then deliver a different gene into all the presynaptic neurons (Osakada et al., 2011), but requires a specific transgenic mouse and a specific rabies virus, lacks specificity for synapse type, and physiologically useful expression is limited to 5–11 days after gene transfer, as Rabies Virus is neurotoxic (Osakada et al., 2011). We developed a technology that can deliver a gene across specific glutamatergic synapses; the presynaptic neurons express a synthetic peptide neurotransmitter that binds to NMDA receptors, and antibody-mediated targeted gene transfer delivers a gene selectively into neurons that bind the synthetic peptide neurotransmitter (Zhang et al., 2012b).

In rat visual cortex, a subset of the glutamatergic neurons also express TrkB, as shown by costaining with an anti-TrkB antibody and antibodies that recognize specific NMDA receptor subunits (Tongiorgi et al., 1999). These results suggest that in PER cortex, a subset of the glutamatergic synapses are also BDNF-TrkB synapses. Thus, it appears likely that targeting gene transfer across BDNF-TrkB synapses would provide additional specificity compared to targeting gene transfer across all glutamatergic synapses.

Here, we report a gene transfer technology that can deliver one gene into presynaptic neurons and a second gene into neurons that are connected via a BDNF-TrkB synapse, by using a synthetic peptide neurotransmitter that binds to TrkB receptors. This technology was established by delivering a first gene into presynaptic neurons in POR cortex, and then delivering a second gene selectively across BDNF-TrkB synapses into postsynaptic neurons in PER cortex. This technology should be useful for analyzing the role of BDNF-TrkB synapses in synaptic plasticity and learning.