Our current understanding of insect phototransduction is based on a small number of species, but insects occupy many different visual environments. We created the retinal transcriptome of a nocturnal insect, the cockroach, Periplaneta americana to identify proteins involved in the earliest stages of compound eye phototransduction, and test the hypothesis that different visual environments are reflected in different molecular contributions to function. We assembled five novel mRNAs: two green opsins, one UV opsin, and one each TRP and TRPL ion channel homologs. One green opsin mRNA (pGO1) was 100–1000 times more abundant than the other opsins (pGO2 and pUVO), while pTRPL mRNA was 10 times more abundant than pTRP, estimated by transcriptome analysis or quantitative PCR (qPCR). Electroretinograms were used to record photoreceptor responses. Gene-specific in vivo RNA interference (RNAi) was achieved by injecting long (596–708 bp) double-stranded RNA into head hemolymph, and verified by qPCR. RNAi of the most abundant green opsin reduced both green opsins by more than 97% without affecting UV opsin, and gave a maximal reduction of 75% in ERG amplitude 7 days after injection that persisted for at least 19 days. RNAi of pTRP and pTRPL genes each specifically reduced the corresponding mRNA by 90%. Electroretinogram (ERG) reduction by pTRPL RNAi was slower than for opsin, reaching 75% attenuation by 21 days, without recovery at 29 days. pTRP RNAi attenuated ERG much less; only 30% after 21 days. Combined pTRP plus pTRPL RNAi gave only weak evidence of any cooperative interactions. We conclude that silencing retinal genes by in vivo RNAi using long dsRNA is effective, that visible light transduction in Periplaneta is dominated by pGO1, and that pTRPL plays a major role in cockroach phototransduction.

Octopamine is a chemical relative of noradrenaline providing analogous neurohumoral control of diverse invertebrate physiological processes. There is also evidence for direct octopaminergic innervation of some insect peripheral tissues. Here, we show that spider peripheral mechanoreceptors are innervated by octopamine-containing efferents. The mechanosensory neurons have octopamine receptors colocalized with synapsin labeling in the efferent fibers. In addition, octopamine enhances the electrical response of the sensory neurons to mechanical stimulation.Spider peripheral mechanosensilla receive extensive efferent innervation. Many efferent fibers in the legs of Cupiennius salei are GABAergic, providing inhibitory control of sensory neurons, but there is also evidence for other neurotransmitters. We used antibody labeling to show that some efferents contain octopamine and that octopamine receptors are concentrated on the axon hillocks and proximal soma regions of all mechanosensory neurons in the spider leg. Synaptic vesicles in efferent neurons were concentrated in similar areas.Octopamine, or its precursor tyramine, increased responses of mechanically stimulated filiform (trichobothria) leg hairs. This effect was blocked by the octopamine antagonist phentolamine. The octopamine-induced modulation was mimicked by 8-Br-cAMP, a cAMP analog, and blocked by Rp-cAMPS, a protein kinase A inhibitor, indicating that spider octopamine receptors activate adenylate cyclase and increase cAMP concentration.Frequency response analysis showed that octopamine increased the sensitivity of the trichobothria neurons over a broad frequency range. Thus, the major effect of octopamine is to increase its overall sensitivity to wind-borne signals from sources such as flying insect prey or predators.

Spider sensory neurons with cell bodies close to various sensory organs are innervated by putative efferent axons from the central nervous system (CNS). Light and electronmicroscopic imaging of immunolabeled neurons has demonstrated that neurotransmitters present at peripheral synapses include γ-aminobutyric acid (GABA), glutamate and octopamine. Moreover, electrophysiological studies show that these neurotransmitters modulate the sensitivity of peripheral sensory neurons. Here, we undertook immunocytochemical investigations to characterize GABA and glutamate-immunoreactive neurons in three-dimensional reconstructions of the spider CNS. We document that both neurotransmitters are abundant in morphologically distinct neurons throughout the CNS. Labeling for the vesicular transporters, VGAT for GABA and VGLUT for glutamate, showed corresponding patterns, supporting the specificity of antibody binding. Whereas some neurons displayed strong immunolabeling, others were only weakly labeled. Double labeling showed that a subpopulation of weakly labeled neurons present in all ganglia expresses both GABA and glutamate. Double labeled, strongly and weakly labeled GABA and glutamate immunoreactive axons were also observed in the periphery along muscle fibers and peripheral sensory neurons. Electron microscopic investigations showed presynaptic profiles of various diameters with mixed vesicle populations innervating muscle tissue as well as sensory neurons. Our findings provide evidence that: (1) sensory neurons and muscle fibers are innervated by morphologically distinct, centrally located GABA- and glutamate immunoreactive neurons; (2) a subpopulation of these neurons may co-release both neurotransmitters; and (3) sensory neurons and muscles are innervated by all of these neurochemically and morphologically distinct types of neurons. The biochemical diversity of presynaptic innervation may contribute to how spiders filter natural stimuli and coordinate appropriate response patterns.

. The mechanosensilla in spider exoskeleton are innervated by bipolar neurons with their cell bodies close to the cuticle and dendrites attached to it. Numerous efferent fibers synapse with peripheral parts of the mechanosensory neurons, with glial cells surrounding the neurons, and with each other. Most of these efferent fibers are immunoreactive to ␥-aminobutyric acid (GABA), and the sensory neurons respond to agonists of ionotropic GABA receptors with a rapid and complete inhibition. In contrast, little is known about metabotropic GABA B receptors that may mediate longterm effects. We investigated the distribution of GABA B receptors on spider leg mechanosensilla using specific antibodies against 2 proteins needed to form functional receptors and an antibody that labels the synaptic vesicles on presynaptic sites. Both anti-GABA B receptor antibodies labeled the distal parts of the sensory cell bodies and dendrites but anti-GABA B R1 immunoreactivity was also found in the axons and proximal parts of the cell bodies and some glial cells. The fine efferent fibers that branch on top of the sensory neurons did not show GABA B receptor immunoreactivity but were densely labeled with anti-synapsin and indicated synaptic vesicles on presynaptic locations to the GABA B receptors. Intracellular recordings from sensory neurons innervating the slit sensilla of the spider legs revealed that application of GABA B receptor agonists attenuated voltageactivated Ca 2ϩ current and enhanced voltage-activated outward K ϩ current, providing 2 possible mechanisms for controlling the neurons' excitability. These findings support the hypothesis that GABA B receptors are present in the spider mechanosensilla where their activation may modulate information transmission.

G-protein-coupled octopamine (OA) receptors mediate their effects by Ca²(+) signaling or adjusting intracellular cAMP levels. Depending on OA concentration and cell type, activation of OA receptors in excitable cells triggers excitatory or inhibitory effects, but the mechanisms by which Ca²(+) or cAMP mediates these effects are not well understood. We investigated signaling mechanisms that are potentially activated by OA, and OA effects on excitability and frequency sensitivity in mechanosensory neurons innervating the VS-3 slit sensilla on the patella of the spider Cupiennius salei. These neurons are directly innervated by octopaminergic efferents, and possess OA receptors that were immunoreactive to an antibody against an OA receptor highly expressed in mushroom bodies. OA application enhanced VS-3 neuron sensitivity, especially at high stimulation frequencies. This enhancement lasted for at least 1 h after OA application. Changes in sensitivity were also detected when the Ca²(+) ionophore ionomycin or the cAMP analog 8-Br-cAMP was applied. However, the cAMP pathway was unlikely to mediate the OA effect, as the protein kinase A inhibitor RP-cAMPS did not diminish this effect. In contrast, the OA-induced sensitivity enhancement was significantly reduced by KN-62, an inhibitor of Ca²(+) /calmodulin-dependent protein kinase II (CaMKII), and by the Ca²(+) chelator BAPTA-AM. OA depolarized the neurons by 3.8 mV from resting potential, well below the threshold for opening of voltage-activated Ca²(+) channels. OA also reduced the amplitudes of voltage-activated K(+) currents. We propose that OA receptors in VS-3 neurons activate inositol 1,4,5-trisphosphate, leading to Ca²(+) release from intracellular stores. The Ca²(+) surge switches on CaMKII, which modulates voltage-activated K(+) channels, resulting in persistent enhancement in excitability.