Methods for examining neurons

back

In the study of the insect brain, the traditional strategy is to assemble detailed information concerning the constituent neurons that can be integrated to understand the system level. More recently, imaging techniques, molecular genetic methods, theory, simulation, and other engineering methods have been added to electrophysiological recording. For example, experimental data can be merged to implement a simulation that can be evaluated using a robot. This section introduces common methods used in the study of insect nervous systems.

■Intracellular recording and staining
■(Functional) imaging
■Genetic tools
■Methods from engineering


■Intracellular recording and staining

First, we describe the analysis of the morphology of individual neurons. In the insect brain, the somata from clusters at the surface of the brain (Figure 1A). For better access, the brain is removed from the head capsule and immobilised on a glass slide while being immersed in Ringer's solution. This way, observations can be made with water immersion lenses (LUMPlanFl/IR, 40x, N.A. 0.80, working distance 3.4 mm or 60x，N.A. 0.90, working distance 2.0 mm) on a fixed stage microscope (BX51WI，Olympus) using near-infrared light (>775 nm) to improve visibility. With such a setup, somata can be visualised by enhanced videomicroscopy (Figures 1B, C, using a Hamamatsu C2741-79 CCD camera ). Cells were impaled with a motorised 3D micromanipulator (ONU-31P，Olympus) using micropipettes（10～100MΩ impedance)．

Pipette filling solutions contain markers that can be injected into cells using current. We use the fluorescent dyes Lucifer Yellow CH，Alexa Fluor 568 or Neurobiotin（which can be visualised by binding with Cy-3 conjugated avidin)．For tracing neural circuits, two or even more neurons can be labeled with different colors. Functional information can be obtained by labeling for candidate neurotransmitters in addition (Figures 1D, E)．

　Impaling neurons with micropipettes allows both the recording of neural activity and the identification of neuronal morphology by tracer injection and subsequent fluorescent detection (Figures 1B, C). Combining such and other information, a database of neural structure and function can be established. In the silkmoth, this approach led to a better understanding of how pheromone information is processed in the brain and which brain structures are involved in this process.




Figure 1: Silkmoth brain and single neurons. (A) Frontal (transversal) view of the silkmoth brain. (B,C) Details of the brain seen in (A) as viewed in differential interference contrast with near-infrared enhanced videomicroscopy (>775nm). Somata can be visualised and it is possible to impale single neurons with a micropipette (arrowhead) mounted on a micromanipulator. Fluorescent dye can be injected into individual neurons. (D) Labeled individual neurons. (E) Neuron labeled in (D, green) in combination with immunohistochemical counterstaining for a neurotransmitter (GABA, in magenta). Scale bars: 1mm (A), 0.1mm (B-E).

back


■Functional imaging

How to investigate the activity of many neurons simultaneously, that is spatio-temporal patterns of neural activity? One possibility is to use fluorescent compounds that allow functional imaging. We focused on the primary olfactory center of the brain, the antennal lobe (AL) in order to investigate odorant information processing. An example of results obtain from such imaging epxeriment is shown in Figure ?.

A widely used method to image neural activity is using calcium-sensitive fluorescent compounds. These can bind calcium ions with some binding constant and binding goes along with changes in fluorescence intensity. Calcium concentration in neurons changes in relation to neural activity, for example by synaptic activity and accompanying action potentials. It is possible to stain neural tissue with dyes incorporating an AM (acetoxymethyl) ester moiety by bath application as these dyes can cross cell membranes but are then rendered impermeant by intracellular enzymes (esterases). Calcium changes observed with this very summary method are slow, in the order of seconds. An example of a calcium response in the antennal lobe following olfactory stimulation is shown (Figure ?).

More recently, genetically encoded calcium-sensitive probes have been designed, for example GCaMP. Such fluorescent proteins can be introduced into organisms to enable imaging of specific cell populations. For example, transgenic silkmoths have been produced in which such probe is expressed in neurons expressing pheromone rececptor proteins (see below).

Voltage-sensitive dyes associate with cell membranes and allow the monitoring of transmembrane voltage (membrane potential) by correlated fluorescence changes. An advantage is that negative and positive changes of transmembrane voltage can be recorded, however, fractional changes in fluorescence intensity are small (in the order of 0.2% maximally). We have successfully imaged antennal lobe activation following electrical stimulation of the antennal nerve and have shown that activation is increased by application of the neuromodulator serotonin (Fig. ?). Such experiments help to understand the mechanisms at the base of state-dependent changes in sensitivity associated with serotonin (5HT) levels in the brain.

back


■Genetic engineering methods

Advances in large-scale DNA sequencing technology have made it possible to provide complete genome sequence information for a number of species. The silk moth genome project is near its completion [10,11]. Such information is essential to allow the use of techniques like detection of specific mRNAs by in situ hybridisation, construct well-specified antigens to generate antibodies for immunohistochemical methods, and producing transgenic stains in which for example fluorescent markers are expressed under the control of chosen promoters.

While in situ hybridisation and immunohistochemistry can only be applied in fixed tissued, transgenic animals can be use alive and allow for visualisation of neurons and many other applications. For instance, it is possible to genetically control the function of specific cells, to manipulate excitation and inhibition experimentally in vivo or to completely knock out some gene of interest. As a result, genetic approaches are a most important tool in gaining insight into nervous systems and behaviour today.

The analysis of an insect central nervous systems using genetic approaches has so far been carried farthest in the fruit fly（Drosophila melanogaster）but in 2000, Tamura has also introduced similar methods to the work with silkmoths (12). We are currently also using genetic approaches in our studies of the silkmoth brain.

In the silkmoth, a transgenic method has been established using the piggyBac transposon (12). For expression of foreign genes in transgenic animals, one commonly uses a promoter known to drive the expression of genes in the target cells (Figure 7A). A reporter gene is also expressed for confirmation, this is usually a fluorescent protein such as green fluorescent protein (GFP) or a red fluorescent protein (DsRed). Using the 3xP3 promoter, photoreceptors were labeled with DsRed in silkmoth embryos and larvae, for example. The reporter gene product was relatively evenly distributed throughout the cytoplasm, facilitating detection.


　　　　　　　　　　　　　　　

Fig. 7 Schematic diagram of reporter gene expression systems in transgenic silkmoths. (A) In the promoter-reporter system, promoter sequence is placed immediately upstream of reporter gene and directly activates reporter gene expression. (B) In the GAL4-UAS system, reporter gene expression is activated via GAL4-UAS yeast transcriptional system.

back


In thefruit fly, signal sequences determining subcellular localisation of the gene products have also been added such that specific regions of cells such as synaptic terminals could be labeled. Such approaches are under development in the silkmoth, also. Using the genetically encoded calcium sensor G-CaMP as a reporter gene, it is possible to record neural activity form the targeted cells. Individuals of transgenic lines are genetically uniform because they have been generated from a few individuals initially and are then imbred, thus the same neurons are labeled in all individuals of a strain.

From yeast genetics, the transcription activator protein GAL4 and its promoter target UAS (upstream promoter sequence), have been introduced to transgenic work in Drosophila. GAL4 is placed under the control of a promoter for a gene expressed in desired target cells in one strain whereas a desired reporter gene is placed under the control of UAS in another strain. As such, both strains are easy to maintain because only in the GAL4 strain, some GAL4, which does nothing in flies normally, is produced. Crossing the two strains combines the information and GAL4 activates the transcription of the gene under UAS control (Figure 7B). A key advantage is that UAS lines lines can be crossed with any number of GAL4 lines in order to express the gene product of interest in different populations of target cells of choice. Due to the combinatorial approach, the work required is much less than what whould be needed if every promoter/gene combination would have to be produced de novo.

In the silkmoth, the GAL4/UAS system was first succesfully used by Imamura in 2003 (13). We first demonstated the use of GAL4/UAS in the silkmoth nervous system using promoters for two types of peptidergic neurons to express GFP (Yamagata et al). Thus, labeling brain neurons with genetically fluorescent proteins is possible in the silkmoth.

In order to determine the spatial input pattern from the antennae to the brain upon pheromone stimulation, we used promoters for the odorant receptors for the major component of the pheromone blend of the silkmoth, bombykol (receptor gene BmOR114) and the minor component bombykal (receptor gene BmOR315). Moths were produced in which these promoters control GAL4 expression and they were crossed with an UAS-GFP stain for confirmation of the specific target regions of odorant receptor neurons expressing the two pheromone receptors. Then, crossed were made with a UAS-GCaMP stain, allowing us to do calcium imaging in the antennal lobe at the level of the axon terminals of the odorant receptor neurons (Sakurai et al, in preparation).

It has thus become possible to apply molecular genetic tools to brain research in the silkmoth. In the future, we will employ these tools to specifically ablate neurons, to alter neurotransmission in specific circuits, and to start efforts to image neural activity in higher brain areas related to pheromone information processing. In particular the possibility of specifically interfering with or controlling neural activity in known target neurons holds a great promise for progress in understanding the neural basis of pheromone orientation behaviour in the silkmoth.

back


■Engineering methods

Useful robots are expected to become technical systems that should be able to perform tasks of interest in noisy, naturalistic environments. From another point of view, they can also be testbeds for models of the generation mechanisms of insect behaviour. These models in turn can become useful when they are capable of performing reliably in natural environments, as insects can do effortlessy. Through appropriate interfaces, robots and insects can be merged into systems operating with components of both. This approach is promising for propelling our understanding of the relation between neural circuits and behaviour by allowing specific manipulations during behaviour. We are currently working on three systems in which insect behaviour, nervous system, and robotics are involved:

・ Robot control through a silkmoth brain model

・ Insect-steered robot

・ Robot control through silkmoth brain activity

back