Laser-evoked synaptic transmission in cultured hippocampal neurons expressing channelrhodopsin-2 delivered by adeno-associated virus

https://doi.org/10.1016/j.jneumeth.2009.06.024Get rights and content

Abstract

We present a method for studying synaptic transmission in mass cultures of dissociated hippocampal neurons based on patch clamp recording combined with laser stimulation of neurons expressing channelrhodopsin-2 (ChR2). Our goal was to use the high spatial resolution of laser illumination to come as close as possible to the ideal of identifying monosynaptically coupled pairs of neurons, which is conventionally done using microisland rather than mass cultures. Using recombinant adeno-associated virus (rAAV) to deliver the ChR2 gene, we focused on the time period between 14 and 20 days in vitro, during which expression levels are high, and spontaneous bursting activity has not yet started. Stimulation by wide-field illumination is sufficient to make the majority of ChR2-expressing neurons spike. Stimulation with a laser spot at least 10 μm in diameter also produces action potentials, but in a reduced fraction of neurons. We studied synaptic transmission by voltage-clamping a neuron with low expression of ChR2 and scanning a 40 μm laser spot at surrounding locations. Responses were observed to stimulation at a subset of locations in the culture, indicating spatial localization of stimulation. Pharmacological means were used to identify responses that were synaptic. Many responses were of smaller amplitude than those typically found in microisland cultures. We were unable to find an entirely reliable criterion for distinguishing between monosynaptic and polysynaptic responses. However, we propose that postsynaptic currents with small amplitudes, simple shapes, and latencies not much greater than 8 ms are reasonable candidates for monosynaptic interactions.

Introduction

The cloning of channelrhodopsin-2 (ChR2) and subsequent expression in mammalian cells promised to revolutionize neurophysiology because it enabled optical stimulation of neurons in a spatially localized and temporally precise fashion (Boyden et al., 2005, Nagel et al., 2003, Wang et al., 2007). ChR2 has been used to identify presynaptic partners of an electrophysiologically recorded postsynaptic neuron (Arenkiel et al., 2007, Petreanu et al., 2007, Wang et al., 2007). Other applications include mapping neuronal circuits, probing synaptic function in genetically defined populations of neurons, and inducing plasticity at single synapses (Atasoy et al., 2008, Liewald et al., 2008, Wang et al., 2007, Zhang et al., 2008, Zhang and Oertner, 2007). In principle, ChR2 could also be used to study the responses of networks to complex spatiotemporal patterns of stimulation.

Expression of ChR2 in neurons has been achieved by mouse transgenesis (Arenkiel et al., 2007, Wang et al., 2007); in utero electroporation (Petreanu et al., 2007), lentivirus (Boyden et al., 2005) and recombinant adeno-associated virus (rAAV) (Bi et al., 2006). A major drawback of lentivirus is that its DNA integrates into the host genome. Therefore the transgene is potentially susceptible to integration-induced epigenetic silencing (Ellis, 2005, Xia et al., 2007). Furthermore, a host gene could be disrupted by lentivirus DNA integration, which could affect normal neuronal function. Although mouse transgenesis by oocyte DNA injection is a powerful tool, integration of exogenous DNA at specific sites can lead to integration-induced gene silencing (Clark et al., 1997, Robertson et al., 1996) and position-effect variegation (Robertson et al., 1995) with gene expression in some cells but not others. Although a plasmid delivered by in utero electroporation remains extrachromasomal, which may alleviate the silencing problem, transfection of early progenitor cells leads to mosaic gene expression in neuronal populations of the postnatal brain (Borrell et al., 2005, Hatanaka et al., 2004).

Recombinant adeno-associated virus gene delivery has been successfully used to express ChR2 in mouse retinal neurons, and expression was reported to be stable for a year (Bi et al., 2006). We chose rAAV for introducing ChR2 in cultured hippocampal neurons for several reasons. First and foremost, genetic modules introduced into rAAV are less prone to epigenetic gene silencing. Second, long-term expression, from months to years, is achievable. Due to high rate of infectivity, rAAV can be used to introduce multiple genes into the same neurons in pre-selected brain regions (Shevtsova et al., 2005) without epigenetic silencing (Zhu et al., 2007). This broadens the experimental possibilities so that other genes whose products act as biosensors for different signaling systems, such as for calcium (Miyawaki, 2003, Palmer and Tsien, 2006, Wallace et al., 2008) and neurotransmitter release (Miesenbock et al., 1998), could also be introduced into the same neuron using rAAV as the delivery method. This would make it possible to optically record functional neuronal connectivity without the need to use patch pipettes. Moreover, by gene selective knockdown of endogenous protein levels using small interfering RNA (siRNA) (Fountaine et al., 2005), especially under control of the tetracycline-controlled systems (Hasan et al., 2004, Sprengel and Hasan, 2007), it should be possible to correlate how changes in gene activity affect neuronal circuits and animal behavior (Grillner, 2006, Kandel, 2001).

Additionally, the rAAV gene delivery method allows targeting of selective brain regions, which makes it especially powerful for in vivo applications. ChR2 has also been targeted to genetically defined populations of neurons through cell-type specific promoters and Cre-dependent constructs to examine neural circuits based on cell types (Atasoy et al., 2008, Liewald et al., 2008). Availability of different AAV serotypes provides additional means to selectively target different neuron types (Burger et al., 2004, Shevtsova et al., 2005, Tenenbaum et al., 2004).

Another feature of rAAV is that it shows low immunogenicity over a long time span (Sun et al., 2002), a key safety criterium that has made rAAV gene delivery the method of choice for therapeutic treatment of animal diseases, including humans. Therefore, rAAV-mediated delivery of ChR2 should not only help to investigate functional brain circuits in living animals but it may also provide a plausible approach to treat neurological diseases which require deep brain stimulation (Gradinaru et al., 2007, Mehrkens et al., 2008, Obeso et al., 1997).

The preceding considerations are general reasons for using rAAV to introduce the ChR2 gene into neurons. Our goal in this paper was to develop the ChR2 method specifically for studying evoked synaptic transmission in mass cultures of dissociated hippocampal neurons.

Dual intracellular recording from pairs of dissociated hippocampal or neocortical neurons is a widely accepted method of studying synaptic plasticity (Arancio et al., 1995, Bi and Poo, 1998, Goda and Stevens, 1996, Kaplan et al., 2003). Such experiments are often done with low density cultures. In one culture method, the substrate that the neurons grow on is sprayed in a mist onto the coverslips, making dots that are less than 1 mm in diameter. Then neurons are plated at low density. The microdots physically limit the neurons’ growth so that some microislands end up with just a few neurons or even just two (Bekkers and Stevens, 1991, Bekkers and Stevens, 1995, Segal and Furshpan, 1990). Within such a microisland, the probability of connection is high, so that it is straightforward to find connected pairs of neurons (Kaplan et al., 2003). In another culture method, neurons and glia are plated at the same time at low density, and this also leads to growth of isolated pairs of neurons (Wilcox et al., 1994).

While the microisland technique makes it easier to record from pairs, culturing healthy neurons becomes more challenging at lower densities. Furthermore, pairs of neurons on microislands tend to be strongly coupled, probably by multiple synaptic contacts (Segal and Furshpan, 1990). Evoked postsynaptic currents are typically hundreds of picoamperes or more, while spontaneous postsynaptic currents (“mini”s) are tens of picoamperes or less (Wilcox et al., 1994). A low calcium solution combined with microperfusion of a high calcium, hypertonic solution (Bekkers and Stevens, 1995) has been used to reduce the amplitude of evoked postsynaptic currents by permitting activation of only a small subset of synapses between two neurons. But without this kind of manipulation, postsynaptic currents evoked in microisland cultures are much larger than those recorded in brain slice experiments.

For these reasons, we have been interested in using mass cultures for synaptic plasticity experiments. These types of cultures are relatively easy to grow and keep healthy, because neurons can be cultured at higher densities. But in our experience, it is difficult to find connected pairs by intracellular recording of randomly chosen neurons, because the probability of connection is low. ChR2 could potentially solve this problem, by allowing the screening of many candidate neurons to find presynaptic partners of a single postsynaptic neuron. We could do this by expressing ChR2 in the culture, and then stimulating presynaptic neurons with a laser while patch recording from a single neuron. If a stimulated neuron is monosynaptically or polysynaptically connected to the recorded neuron, a synaptic response should be observed.

For this purpose, we needed a delivery method for ChR2 that reliably resulted in viable cells, adequate expression levels, and expression during the right time window. We also needed a method of optical stimulation that was spatially precise and reliably evoked action potentials.

In dissociated cultured neurons, experiments on synaptic physiology are typically done between one and three weeks in vitro (Arancio et al., 1995, Bekkers et al., 1990, Bekkers and Stevens, 1995, Bi and Poo, 1998, Goda and Stevens, 1996, Gomperts et al., 2000, Kaplan et al., 2003). Earlier than one week, there is little or no synaptic transmission (Gomperts et al., 2000). After three weeks, there is typically spontaneous synchronous bursting (Pasquale et al., 2008, van Pelt et al., 2004a, van Pelt et al., 2004b), which could interfere with plasticity experiments. For this reason, we were interested in characterizing the expression of ChR2 during this time window. Based on the fluorescence of a ChR2-YFP fusion protein, expression starts at about one week, and is strong after about two weeks in vitro. Infected cells looked as healthy as uninfected cells, as they could not be distinguished from each other in phase contrast images.

To characterize the effectiveness of optical stimulation, we performed patch clamp recordings of neurons while simultaneously illuminating them transiently. After two weeks in vitro, about 80% of infected cells could be stimulated to generate action potentials using wide-field illumination. Stimulation by laser was more difficult, and depended on the size of the illuminated spot. A 40 μm diameter spot stimulated approximately one third of infected neurons, whereas a 10 μm diameter spot stimulated only one quarter. This demonstrated a trade-off between the fraction of cells that can be stimulated, and the accuracy of spatial localization of stimulation. The luminance of the spot had little effect, if it was above a threshold value. If a cell could be stimulated to fire an action potential, then this response was highly reliable every time it was illuminated.

Our next goal was to investigate the best means of using ChR2 to study synaptic transmission. We performed patch clamp recordings of neurons that were not expressing ChR2, or only weakly, in order to reduce the possibility or magnitude of direct stimulation by light. Then we scanned the laser across multiple locations arranged in a grid. Many types of responses were recorded in the patch clamped neuron, which appeared to be direct, monosynaptic, or polysynaptic. Based on classification of these responses, we propose a criterion for identifying possible monosynaptic pairs of neurons using amplitude, shape, and latency of the recorded currents.

Section snippets

Preparation of rAAV

The viral expression construct rAAV-PhSYN–ChR2-YFP was constructed by subcloning the ChR2-YFP fragment (Boyden et al., 2005) into an adeno-associated (serotype-2) viral expression cassette with the human synapsin promoter (PhSYN), a woodchuck posttranscriptionalregulatory element (WPRE), and a bovine growth hormone (BGH) polyadenylation sequence (Shevtsova et al., 2005). rAAV was prepared by transfecting HEK293 cells with the plasmid rAAV-PhSYN–ChR2-YFP together with helper plasmids (Grimm et

rAAV mediated delivery of ChR2-YFP into neurons

To monitor the expression levels of ChR2 in neurons, the ChR2 gene was fused to a yellow fluorescent indicator protein (YFP) (Boyden et al., 2005) and cloned into a rAAV expression plasmid with the human synapsin promoter (PhSYN) (Shevtsova et al., 2005) driving ChR2-YFP expression (rAAV-PhSYN–ChR2-YFP) (Fig. 1A). Virus with capsid proteins of the serotype 1 and 2 was prepared as described previously (Zhu et al., 2007). The culture was infected with the virus, rAAV-PhSYN–ChR2-YFP, 1–2 days

Discussion

We employed rAAV gene delivery to transfer ChR2 to cultured hippocampal neurons. In wide-field stimulation experiments we found that action potentials could be reliably evoked in many neurons after two weeks in vitro, as was previously reported for ChR2 delivered via lentivirus (Boyden et al., 2005, Schoenenberger et al., 2008). We further quantified the fraction of neurons that could be reliably stimulated, as a function of weeks in vitro.

We then experimented with laser stimulation, which has

Acknowledgement

The authors are grateful for support from the Howard Hughes Medical Institute, the Max Planck Society, and the 2008 NSF Emerging Frontiers in Research and Innovation (EFRI) program. We would like to thank Seungeun Oh for assistance with optics, Jeannine Foley for providing hippocampal culture, Rolf Sprengel and Winfriend Denk for support.

References (62)

  • M.O. Heuschkel et al.

    A three-dimensional multi-electrode array for multi-site stimulation and recording in acute brain slices

    J Neurosci Methods

    (2002)
  • A. Miyawaki

    Fluorescence imaging of physiological activity in complex systems using GFP-based probes

    Curr Opin Neurobiol

    (2003)
  • V. Pasquale et al.

    Self-organization and neuronal avalanches in networks of dissociated cortical neurons

    Neuroscience

    (2008)
  • D.L. Pettit et al.

    Chemical two-photon uncaging: a novel approach to mapping glutamate receptors

    Neuron

    (1997)
  • J. van Pelt et al.

    Longterm stability and developmental changes in spontaneous network burst firing patterns in dissociated rat cerebral cortex cell cultures on multielectrode arrays

    Neurosci Lett

    (2004)
  • O. Arancio et al.

    Activity-dependent long-term enhancement of transmitter release by presynaptic 3′,5′-cyclic GMP in cultured hippocampal neurons

    Nature

    (1995)
  • D. Atasoy et al.

    A FLEX switch targets Channelrhodopsin-2 to multiple cell types for imaging and long-range circuit mapping

    J Neurosci

    (2008)
  • A. Auricchio et al.

    Isolation of highly infectious and pure adeno-associated virus type 2 vectors with a single-step gravity-flow column

    Hum Gene Ther

    (2001)
  • J.M. Bekkers et al.

    Origin of variability in quantal size in cultured hippocampal neurons and hippocampal slices

    Proc Natl Acad Sci U S A

    (1990)
  • J.M. Bekkers et al.

    Excitatory and inhibitory autaptic currents in isolated hippocampal neurons maintained in cell culture

    Proc Natl Acad Sci U S A

    (1991)
  • J.M. Bekkers et al.

    Quantal analysis of EPSCs recorded from small numbers of synapses in hippocampal cultures

    J Neurophysiol

    (1995)
  • G.Q. Bi et al.

    Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and postsynaptic cell type

    J Neurosci

    (1998)
  • E.S. Boyden et al.

    Millisecond-timescale, genetically targeted optical control of neural activity

    Nat Neurosci

    (2005)
  • E.M. Callaway et al.

    Photostimulation using caged glutamate reveals functional circuitry in living brain slices

    Proc Natl Acad Sci U S A

    (1993)
  • A.J. Clark et al.

    Mammalian cDNA and prokaryotic reporter sequences silence adjacent transgenes in transgenic mice

    Nucleic Acids Res

    (1997)
  • J. Ellis

    Silencing and variegation of gammaretrovirus and lentivirus vectors

    Hum Gene Ther

    (2005)
  • T.M. Fountaine et al.

    Delivering RNA interference to the mammalian brain

    Curr Gene Ther

    (2005)
  • S.N. Gomperts et al.

    Distinct roles for ionotropic and metabotropic glutamate receptors in the maturation of excitatory synapses

    J Neurosci

    (2000)
  • V. Gradinaru et al.

    Targeting and readout strategies for fast optical neural control in vitro and in vivo

    J Neurosci

    (2007)
  • D.J. Hagler et al.

    Properties of synchronous and asynchronous release during pulse train depression in cultured hippocampal neurons

    J Neurophysiol

    (2001)
  • M.T. Hasan et al.

    Functional fluorescent Ca2+ indicator proteins in transgenic mice under TET control

    PLoS Biol

    (2004)
  • Cited by (11)

    • Methamphetamine modulates glutamatergic synaptic transmission in rat primary cultured hippocampal neurons

      2014, Brain Research
      Citation Excerpt :

      Dissecting the mechanism of how METH influences the GLU-mediated synaptic transmission in neurons will help us understand the complexity of its neurotoxicity. It is known that synaptic interactions between cultured hippocampal neurons can be very strong, especially in low density cultured neurons (Wang et al., 2009; Segal and Furshpan, 1990). To electrophysiological recording, density of cultured neuron was low in our preparation.

    • Acat1 knockdown gene therapy decreases amyloid-β in a mouse model of Alzheimer's disease

      2013, Molecular Therapy
      Citation Excerpt :

      We chose to administer Acat1 gene knockdown using a recombinant adeno-associated virus (AAV) mediated gene therapy approach. AAVs are popular gene therapy vectors15 because they produce long-term (at least 1 year),16 stable, episomal expression of the viral transgene, provoke only a minimal inflammatory response, have minimal oncogenic potential, can infect and transduce postmitotic cells (for instance, neurons),17,18 and have been used successfully in many studies to transduce central nervous system tissues19,20,21 and other tissues.22 Recombinant AAVs with hybrid serotype 1/2 efficiently target and transduce neurons.23

    • Optogenetic approaches to characterize the long-range synaptic pathways from the hypothalamus to brain stem autonomic nuclei

      2012, Journal of Neuroscience Methods
      Citation Excerpt :

      We observed an average failure rate of 4.5 ± 2.06% (n = 7, ranging from 0 to 15.6%). Typical average latencies for monosynaptic transmission using optogenetic stimulation of axons in other studies range between 3.6 and 17.3 ms and polysynaptic latencies typically vary between 18 and 25.6 ms (Petreanu et al., 2007, 2009; Ren et al., 2011; Toni et al., 2008; Wang et al., 2009; Xiong et al., 2012). In support of a monosynaptic connection, the low jitter (0.83 ± 0.09 ms, n = 7) of the synaptic responses in this study falls in the reported range for monosynaptic pathways (Petreanu et al., 2007, 2009; Xiong et al., 2012).

    • Local production of corticotropin-releasing hormone in prefrontal cortex modulates male-specific novelty exploration

      2022, Proceedings of the National Academy of Sciences of the United States of America
    View all citing articles on Scopus
    View full text