Neurons and glia are the functional units of brains. Consequently, normal brain function relies on the correct development and operating of these cells. With regard to CNS function and development, calcium channels are among the most pleiotropic genes. Firstly, calcium influx influences basically every step of neuronal differentiation, and secondly, calcium channels significantly contribute to the specific membrane excitability of neurons.
Figure 1: Voltage gated calcium channels are expressed in all neuronal compartments. Depending on channel biophysical properties and localization voltage gated calcium channels may serve different functions.
Therefore, the analysis of calcium channel function contributes to our understanding of brain development and operation. Calcium channel function depends on the channels’ biophysical properties, but also critically where in the neuron these ion channels are localized. For example, in the presynaptic terminal, calcium channels serve synaptic vesicle release, in dendrites, they may boost excitatory postsynaptic input, in the axon they may activate calcium dependent potassium channels and thereby allowing maximal firing frequencies, and in the soma they may translate activity into gene expression (Fig. 1).
We focus on the properties and the localization of voltage gated calcium channels (VGCCs) for normal brain function (Kadas et al., 2017 J Neurosci 37(45):10971-10982; Ryglewski et al., 2012 J Physiol 590(4):809-825) and development (Ryglewski et al., 2017 Neuron 93(3):632-645; Ryglewski et al., 2014 PNAS 111(50):18049-18054). We use the genetic model Drosophila melanogaster with its versatile tool kit and a relatively small but, nevertheless, sufficiently complex central nervous system to study basic principles of neuronal performance. We employ electrophysiological techniques such as sharp electrode but also patch clamp recordings, nerve stimulation with simultaneous muscle recordings, calcium imaging experiments, opto- and thermogenetic activation or silencing of neurons, immunohistochemical methods including confocal microscopy, but also intracellular dye fills and molecular techniques as well as cell culture with heterologous expression of ion channels and patch clamp recordings of those.
In vertebrate genomes 10 genes encode VGCCs that are categorized in three VGCC families: Cav1.1-1.4, Cav2.1-2.3, and Cav3.1-3.3. In contrast, the Drosophila genome contains only one homolog each, namely Dmca1D (Cav1), Dmca1A, also known as cacophony (cac) or nightblind (nbA; Cav2), and DmαG or Dmα1T (Cav3).
Table 1: Vertebrate and Drosophila voltage gated calcium channels
Cav1 and Cav2 as well as their Drosophila homologs mediate high voltage activated (HVA) VGCCs, whereas Cav3 and its homolog mediate true low voltage activated (LVA) VGCCs with transient calcium currents (table 1). We have found that in adult Drosophila wing motoneurons most voltage gated calcium current in the somatodendritic domain is mediated by cacophony. Interestingly, cacophony channels mediate not only sustained HVA but also transient LVA currents, challenging the view that LVA calcium currents are always mediated by Cav3 channels (Ryglewski et al., 2012 J Physiol 590(4):809-825, Fig. 2). Moreover, cacophony channels are also localized to all Drosophila motoneuron axon terminals and are essential for synaptic vesicle release. Therefore, one calcium channel gene, namely cacophony, mediates different voltage gated calcium currents, localizes to different sub-neuronal compartments, and underlies several fundamentally different functions during development and adult neural circuit function. Our current research focuses on the mechanisms that regulate cacophony channel biophysical properties and localization, and thus, function.
Figure 2: In Drosophila one gene, cacophony, gives rise to at least four different voltage gated calcium channels with different biophysical properties such as activation voltages and kinetics in the same neuron: adult sustained HVA (A, D) and transient LVA (B) as well as pupal HVA (I) and presynaptic HVA (J) voltage gated calcium channels (Ryglewski et al., 2012 J Physiol 590(4):809-825) as revealed by targeted RNAi (E), pharmacology (F), and temperature sensitive mutants (G, H).
In principle, three different mechanisms come to mind:
1) alternative splicing
2) pre-mRNA editing
3) co-assembly with accessory subunits
First, the cacophony gene carries several alternative exon pairs, two of which being at places that may explain the different channel properties that we found in adult wing motoneurons (Fig. 3).
Figure 3: Cacophony features two alternative exon pairs, IS4a/b and I-IIa/b (top), that encode part of the voltage sensor of the first homologous repeat or part of the intracellular loop between the first and second homologous repeats that contains the β-subunit binding site (bottom). Schematics adapted after Kawasaki et al. 2002.
One of these mutually exclusive exon pairs encodes part of the voltage sensor of the first homologous repeat possibly explaining the different activation voltages that we found (HVA vs. LVA). The second pair codes for part of the site of β-subunit binding (BID; β-interacting domain) maybe playing a role in channel kinetics (sustained vs. transient, Fig. 3). Second, the cacophony transcript undergoes A-I pre-mRNA editing. Third, HVA VGCCs are known to associate with accessory subunits such as α2δ-, β, and γ-subunits. Accessory subunits may play a role in trafficking, localization, and surfacing of VGCCs, but also in affecting activation voltages, and activation and inactivation kinetics to name but a few. Two Ph.D. theses in the lab are currently trying to shed light on the post-transcriptional modification of cacophony:
1) alternative splicing: Lukas Kilo is following the question whether alternative splicing of cacophony may result in the expression of cacophony-based sustained HVA and transient LVA calcium currents (Ryglewski et al., 2012 J Physiol 590(4):809-825, Fig. 2), or whether different isoforms are used for presynaptic (axon terminal) or postsynaptic (dendritic) calcium channel function. Cacophony features two alternative exon pairs (5/6 or IS4a/b and 10/11 or I-IIa/b; nomenclature according to Peixoto et al., 1997 and Kawasaki et al., 2002; Fig. 3) that correspond to the voltage sensor in the fourth transmembrane domain of the first homologous repeat of the channel and the β-subunit binding site located on the intracellular loop between the first and second homologous repeats, respectively. Lukas takes two different approaches. First, he clones different cacophony isoforms from cDNA to then express these as UAS-constructs in Drosophila for functional testing. He aims at employing a knock-in strategy in cacophony null mutants. Expression of the only available UAS-cacophony-transgene (UAS-cac1 carrying alternative exons IS4b/I-IIb; Kawasaki et al., 2002) in all neurons in a cacophony null mutant background rescues cacophony null lethality. Lukas is collaborating with B. Altenhein, University of Cologne, and M. Heine and U. Thomas, LIN Magdeburg, creating other cDNA versions of cacophony (IS4a/I-IIa; IS4a/I-IIb; IS4b/I-IIa; Fig. 4) to express them in flies either in a cacophony null mutant background, or, in case we face lethality, in combination with UAS-cac1 since we know its expression rescues cac null lethality.
Figure 4: Alternative splicing of IS4a/b or I-IIa/b (top) mediates 4 different annotated cacophony transcripts (bottom).
The new cac cDNA constructs will also be tested in heterologous expression systems (co-expressed with α2δ- and β-subunits needed for normal calcium channel function) to record the resulting calcium currents using the whole cell patch clamp technique. The second approach is removing one of the alternative exons at a time in vivo using CRISPR/cas9 (Fig. 5) and then assessing flight and courtship song and recording the resulting calcium currents in vivo. The possibility of differential targeting of the different cacophony splice variants will be tested with calcium imaging from identified neurons in situ. This work will contribute to the understanding how differential splicing may be used as a means to multiply and regulate ion channel biophysical properties and localization.
Figure 5: Genomic removal of the alternative exon IS4b by CRISPR/cas9 methodology results in the loss of this particular splice variant (A). The removal of this particular 950 base pairs long piece of DNA was confirmed by PCR (B) and subsequent sequencing (C). The presence of a respective transcript was confirmed with RT-PCR (D).
2) pre-mRNA editing: The cacophony transcript is known to undergo A-I pre-mRNA editing. We currently test the role of editing for the regulation of cacophony channel biophysical properties and localization. Our first findings indicate interesting roles of ADAR mediated A-I editing for specific calcium channel properties, such as current amplitudes and activation voltages.
3) co-assembly with accessory subunits: Laurin Heinrich is asking the question whether the interaction of VGCC α-subunits with accessory α2δ-subunits follow a specific code. As in vertebrates, the Drosophila genome contains four different genes encoding α2δ-subunits 1-4.
Figure 6: HVA voltage gated calcium channels associate with α2δ-subunits. Vertebrate genomes contain 7 different genes for HVA voltage gated calcium channels plus 4 each for α2δ- and β-subunits. Together this gives rise to 112 possible α- / α2δ- / β-subunit combinations instead of just 8 in Drosophila.
Together with only one gene for β- and two genes for HVA α-subunits, this makes 8 possible HVA α- / α2δ- / β-subunit combinations, instead of 112 in vertebrates (Fig. 6), thus allowing for a chance to crack the code. Laurin is studying the expression pattern of α2δ-subunits in the Drosophila ventral nerve cord using antibodies and protein traps resulting in the native protein being tagged with a fluorescent marker such as GFP or mCherry (Fig. 7). Moreover, she is using targeted RNAi knock down of α2δ-subunits to assess their roles for calcium channel biophysical properties and localization, as well as for structural and functional differentiation of neurons.
Figure 7: α2δ-1 and stj are differentially expressed in the Drosophila CNS. Genomic tagging of Drosophila α2δ-1 with GFP reveals localization predominantly in the larval and the adult central nervous system neuropil as well as somata (A). Expression of membrane-bound GFP (UAS-mCD8::GFP) under the control of the stj-promoter reveals expression predominantly in motoneurons (B, left) .α2δ-3 (stj) is expressed mainly in motoneuron somata and at the larval neromuscular junction as revealed by tagging a stj-HA fusion transgene with an HA-antibody (B, right). Panel B, (bottom right) from Kurshan et al., 2009.