Defective glycinergic synaptic transmission in zebrafish motility mutants

Abstract

Glycine is a major inhibitory neurotransmitter in the spinal cord and brainstem. Recently, in vivo analysis of glycinergic synaptic transmission has been pursued in zebrafish using molecular genetics. An ENU mutagenesis screen identified two behavioral mutants that are defective in glycinergic synaptic transmission. Zebrafish bandoneon (beo) mutants have a defect in glrbb, one of the duplicated glycine receptor (GlyR) beta subunit genes. These mutants exhibit a loss of glycinergic synaptic transmission due to a lack of synaptic aggregation of GlyRs. Due to the consequent loss of reciprocal inhibition of motor circuits between the two sides of the spinal cord, motor neurons activate simultaneously on both sides resulting in bilateral contraction of axial muscles of beo mutants, eliciting the so-called 'accordion' phenotype. Similar defects in GlyR subunit genes have been observed in several mammals and are the basis for human hyperekplexia/startle disease. By contrast, zebrafish shocked (sho) mutants have a defect in slc6a9, encoding GlyT1, a glycine transporter that is expressed by astroglial cells surrounding the glycinergic synapse in the hindbrain and spinal cord. GlyT1 mediates rapid uptake of glycine from the synaptic cleft, terminating synaptic transmission. In zebrafish sho mutants, there appears to be elevated extracellular glycine resulting in persistent inhibition of postsynaptic neurons and subsequent reduced motility, causing the 'twitch-once' phenotype. We review current knowledge regarding zebrafish 'accordion' and 'twitch-once' mutants, including beo and sho, and report the identification of a new alpha2 subunit that revises the phylogeny of zebrafish GlyRs.

Keywords: behavior; glycine; motility; receptor; synapse; transporter; zebrafish.

Figure 1

Zebrafish embryos display three early behaviors. (A) A 19 hpf wild-type embryo exhibits spontaneous coiling. (B) At 24 hpf, a wild-type embryo responds to mechanosensory stimulation with two fast, alternating contractions. (C) A bandoneon (beo) mutant embryo with a defect in glrbb responds to mechanosensory stimulation with bilateral axial muscle contractions that causes the trunk to shorten and bend dorsally. (D) At 48 hpf, wild-type embryos swim away following tactile stimulation.

Figure 2

The output of the CNS is normal in accordion (atp2a1) mutants but abnormal in bandoneon (glrbb) mutants. (A) Schematic summary of the experimental procedure. (B) Wild-type muscle at 48 hpf responds to tactile stimulation with rhythmic depolarizations, representing cyclic muscle contractions during fictive swimming. (C) accordion (atp2a1) mutant muscle respond to tactile stimulation with a similar rhythmic pattern, indicating that the outputs from the CNS are normal. (D) bandoneon (glrbb) mutant muscle responds to tactile stimulation with a short and large voltage response devoid of rhythmicity, suggesting that the CNS outputs are aberrant.

Figure 3

Sequence alignments of zebrafish GlyR subunits with avian and mammalian counterparts. (A) Sequence alignment of human (GenBank accession: NP_000162), rat (NP_037265), mouse (NP_065238) GlyR α1 with zebrafish (NP_571477) GlyR α1. The four membrane-spanning domains are represented as M1-M4. The 2’ residues in M2 are highlighted by grey box. Signal peptides are denoted by negative numbering. (B) Protein sequence alignment of human (NP_002054), rat (NP_036700), mouse (NP_906272), chick (XP_001234291) GlyR α2 with zebrafish (GQ406228) GlyR α2. (C) Protein sequence alignment of human (NP_006520), rat (NP_446176), mouse (NP_536686), chick (XP_420527) GlyR α3 with zebrafish (NP_694497) GlyR α3. (D) Protein sequence alignment of rat (XP_346351), mouse (NP_034427), chick (XP_001232995) with zebrafish GlyR GlyR α4a (GQ406229) and GlyR α4b (AAH85599). (E) Protein sequence alignment of human (NP_000815), rat (NP_445748), mouse (NP_034428) and chick (XP_420379) GlyR β with zebrafish GlyR βb (NP_001003587) and GlyR βa (XP_683646). Position of mutations identified in the three beo alleles are represented by arrowheads.

Figure 4

Simultaneously contraction of bilateral axial muscles in bandoneon (glrbb) mutants due to the loss of glycinergic synaptic transmission. (A) Superimposed voltage responses of muscles evoked by mechanosensory stimulation. Arrows indicate the time of stimulation. The latency of the muscle response to contralateral stimulation was shorter than that to ipsilateral stimulation in wild-type, whereas the latency to ipsilateral and contralateral stimulation was comparable in beo mutants. Histograms show that the latency to half-maximal amplitude of the first depolarization was shorter in contralateral stimulation compared to ipsilateral stimulation in wild-type. The latency of the response to tactile stimulation in strychnine-treated wild-type muscles was comparable to latency in beo mutants. (B) Spontaneous synaptic currents recorded from a wild-type motor neuron in the presence of TTX were decreased in frequency following block of NMDA and AMPA receptors by application of CNQX and APV, respectively. The non-glutamatergic currents in wild-type are eliminated by further application of strychnine, showing that they are glycinergic currents. In beo, non-glutamatergic currents in the presence of CNQX and APV are missing, indicating that glycinergic synaptic currents are absent. (C) A puff of exogenous glycine induced a current in a wile-type motor neuron and a smaller current in a beo mutant motor neuron.

Figure 5

The aberrant motor response of shocked (slc6a5) mutants defective in GlyT1 is due to high external glycine. (A) Muscle voltage recording from a wild-type embryo showed normal fictive swimming in response to mechanosensory stimulation. Arrows indicate the time of stimulation. (B) Muscle recording from a sho mutant embryo displaying a large, nonrhythmic depolarization. (C) Muscle recording from a sho mutant after the hindbrain was exposed and perfused with glycine-free solution demonstrated rhythmic depolarizations similar to fictive swimming. (D) Muscle recording from the same sho embryo as in (C) after switching the perfusion from glycine-free solution to saline containing 0.2 mM glycine again exhibited the aberrant response characteristic of sho mutants.