FP CILIA. Customized Intelligent Life-Inspired Arrays. Integrated Project

Transcription

1 FP CILIA Customized Intelligent Life-Inspired Arrays Integrated Project Information Society Technolgies Future & Emerging Technologies Proactive Initiative BIO-I3 DELIVERABLE: D Public E LECTROPHYSIOLOGICAL P ROPERTIES OF N EURONS I NVOLVED IN S IGNAL D ETECTION Actual submission date: September 8, 2006 Start day of project: September 1st, 2005 Duration: 48 months Copyright Members of the CILIA Consortium See for details on the copyright holders. CILIA ( Customized Intelligent Life-Inspired Arrays ) is a project funded by the European Union. For more information on the project, its partners and contributors please see The information contained in this document represents the views of CILIA as of the date they are published. CILIA does not guarantee that any information contained herein is error-free, or up to date. CILIA MAKES NO WARRANTIES, EXPRESS, IMPLIED, OR STATUTORY, BY PUBLISHING THIS DOCUMENT.

2 CONTENT 1. EXECUTIVE SUMMARY TERMINOLOGY RESPONSES OF LATERAL LINE AFFERENT NERVE FIBRES TO MOVING OBJECTS INTRODUCTION METHODS QUALITY AND NUMBER OF RECORDINGS Signal-to-noise ratio of electrophysiological recordings Number of cells recorded And ongoing activity RESPONSE PATTERNS AND DISCHARGE RATES RESPONSE PATTERNS EFFECTS OF Object motion direction EFFECTS of object size Effects of object distance STIMULUS CHARACTERIZATION Water velocities measured with hot-wire anemometry pressure changes measured with a hydrophone water Surface Waves CONCLUSIONS COMPARISON ALLN - PLLN RESPONSES TO MOVING OBJECTS IN RUNNING WATER REFERENCES FP Members of CILIA consortium PUBLIC 2 / 18

3 1. EXECUTIVE SUMMARY Using extracellular single cell recordings, we investigated how fibres in the anterior lateral line nerve of goldfish, Carassius auratus, respond to the water motions generated by moving objects of different size that approach the fish from different directions and at different lateral distances. We found different but distinct patterns of neural activity in response to an object that was moved alongside the fish. The patterns included (i) monophasic responses that were characterized by single transient increase or decrease in discharge rate, (ii) biphasic responses that were characterized by an increase followed by a decrease in discharge rate or vice versa, and (iii) triphasic responses that were characterized by an increase followed by a decrease and again an increase in discharge rate, or by the inverse pattern, i.e. a decrease followed by an increase and another decrease in discharge rate. About one third of the fibres responded to a reversal of object motion direction with an inversion of the response pattern, an effect that can be explained by the intrinsic directional sensitivity of lateral line hair cells. Fibres increased their discharge rate with increasing object size (object distance constant) and decreased discharge rate with increasing object distance (object size constant). With respect to the types of response patterns and the response behaviour to different object sizes and distances the results are in fair agreement with previous studies in which the responses of posterior lateral line nerve fibres to moving objects were recorded. However, the relative proportions of the recorded response patterns were different from those found in the posterior lateral line nerve, and a much smaller proportion of the fibres responded with an inversion of the response pattern to a reversal of object motion direction. These differences suggest that location and orientation of lateral line neuromasts on the fish head are different from those on the fish trunk. We are presently studying neuromast orientation in WPSet 1 using different microscopic techniques. FP Members of CILIA consortium PUBLIC 3 / 18

4 2. TERMINOLOGY Abbreviations ALLN PLLN AP PA E I Anterior Lateral Line Nerve Posterior lateral Line Nerve Anterior-posterior Posterior-anterior Excitation (Increase in discharge rate) Inhibition (Decrease in discharge rate) FP Members of CILIA consortium PUBLIC 4 / 18

5 3. RESPONSES OF LATERAL LINE AFFERENT NERVE FIBRES TO MOVING OBJECTS 3.1. INTRODUCTION The aim of this part of the project was to analyse the neural responses of afferent nerve fibres in the anterior lateral line nerve (ALLN) of goldfish to the water motions generated by moving objects. In particular, we were interested in how information about motion direction, object size and distance is represented. The data that we gathered were compared with existing data from the posterior lateral line nerve (PLLN) in order to determine commonalities and differences. The obtained data will be used for task "Database for computer simulations of neuronal networks" and task Computational modelling. The results will support the interpretation of data obtained from recordings of central lateral line neurons METHODS Data were collected from 31 goldfish, ranging in length from 8.5 cm to 12 cm. Fish were acquired from commercial dealers and were maintained in 250-l aquaria at ambient temperature on a daily h light-dark cycle. For the experiments standard surgical and electrophysiological methods were used. In brief, fish were anaesthetized with MS 222 and immobilized with Pancuronium bromide. The skull was opened to expose the lateral line nerves at the point where they enter the brainstem. Finally, fish were positioned in an experimental tank (50x50 cm) that rested on a vibration-isolated table. Fish were firmly attached to stainless-steel holder and artificially respirated. The lateral line was stimulated with water motions generated by a moving object using the same set-up described in Mogdans and Bleckmann (1998). A rectangular plexiglass object oriented with its long axis vertically was moved on a circular orbit (radius 13.5 cm) along the side of the fish. The object was mounted onto the edge of a disc that was positioned on the bottom of the experimental tank. Thus, the object protruded upright from below the animal across most of its dorsoventral extent but did not penetrate the water surface. The disc was turned under water by driving its center axis (30 mm diameter plexiglass cylinder at a distance of 14.5 cm from the fish) with a DC motor from above the water surface. The motor-disc assembly was attached to a sliding bar which allowed to displace the object orbit perpendicular to and along the length of the fish. The fish was positioned outside of the orbit, tangential to its perimeter (see Fig. 1 in Müller et al. 1996). The position of the fish relative to the orbit of the moving object was such that a minimum lateral distance (1 cm) between object and fish was reached at eye level. Parameters that were varied during the experiments were object motion direction (AP: anterior-posterior, PA: posterior-anterior), object size (0.1x0.1 up to 4 cm cross section) and object distance (1 cm, 3 cm and 6 cm). Object velocity was constant at 15 cm/s. The water motions generated by objects moving in different directions, by objects of different size and by different object distances were characterized by measuring water velocity with a hot-wire anemometer and by measuring pressure waves with a hydrophone. Neural activity was recorded with glass micropipettes filled with 3 mol/l KCl (impedance MΩ, tip size 1-2 µm). This method yielded recordings with excellent signal-to-noise ratios (see below). However, due to the small tip size recordings were mechanically unstable. Thus many fibres were lost before the entire stimulus protocol could be completed. Action potentials were amplified (10x), low-pass filtered (1 or 10 khz), displayed on an oscilloscope and stored on a computer. For analysis, raster plots and peri stimulus time histograms were generated. FP Members of CILIA consortium PUBLIC 5 / 18

6 Figure 1: Photograph of the experimental set-up. Left: Experimental tank and equipment used to move objects. Right: Close-up view of a fish in the fish holder. P Plate on water surface, H Holder, M Motor, T Tubus to hold an object, O Object, LS Light diodes to control object location, F.-H Fish holder, o St.-P. Upper plexiglass plate (in air), u St.-P Lower plexiglass plate (under water) 3.3. QUALITY AND NUMBER OF RECORDINGS SIGNAL-TO-NOISE RATIO OF ELECTROPHYSIOLOGICAL RECORDINGS With the recording technique used, high signal-to-noise ratios were obtained. Background noise of the electrophysiological recordings varied between 30 µv and 40 µv. Typically the amplitudes of the action potentials were in the mv-range and thus exceeded noise levels many times (for an example see Fig. 1 in D7) resulting in signal-to-noise levels of 20:1 and greater. In many particularly good recordings the amplitudes of the action potentials were so high that the background noise of the recording system was virtually non-existent, resulting in recordings that were of a quality that resembled the quality of intracellular recordings NUMBER OF CELLS RECORDED AND ONGOING ACTIVITY Single cell recordings were made from 77 fibres in the left (ipsilateral) ALLN of 31 goldfish. Of these, 62 responded to a moving object with a change in discharge rate and/or discharge pattern. The remaining 15 fibres did not respond to a moving object. The average ongoing activity of the 62 responsive fibres was 21.9 spikes*s -1 (median, range 0.7 to 69.1 spikes*s -1 ) RESPONSE PATTERNS AND DISCHARGE RATES RESPONSE PATTERNS While an object passed alongside the fish, different fibres responded with different patterns of discharge. Independent of motion direction the following patterns occured (i) monophasic responses that were characterized by single transient increase (E, n=29) or decrease (I, n=10) in discharge rate (Fig. 2), (ii) biphasic responses that were characterized by an increase followed by a decrease (EI, n=37) or a decrease followed by an increase (IE, n=24) in discharge rate (Fig. 3), and (iii) triphasic responses that were characterized by an increase followed by a decrease and again an increase (EIE, n=8) in discharge rate, or by the inverse pattern, i.e. a decrease followed by an increase and another decrease in discharge rate (IEI, n=9) (Fig. 4). FP Members of CILIA consortium PUBLIC 6 / 18

7 Figure 2: Examples two primary afferent fibres in the ALLN that exhibited monophasic responses to a moving object (size 1.0x1.0x9.5 cm, lateral distance 1 cm, speed 15 cm/s AP anterior-posterior, PA posterior-anterior object motion direction). Left: Monophasic excitation. Right: Monophasic inhibition. The figure shows raster diagrams of ten successive stimulus presentations (each marker represents one action potential) and the corresponding peri-stimulus-time histograms (bin width 20 ms). Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the fish (open arrowhead). Figure 3: Examples of two primary afferent fibres in the ALLN that exhibited biphasic responses to a moving object (size 1.0x1.0x9.5 cm, lateral distance 1 cm, speed 15 cm/s AP anterior-posterior, PA posterior-anterior object motion direction). Left: Biphasic responses that inverted when object motion direction was reversed. Right: Biphasic responses that were comparable for both motion directions. The figure shows raster diagrams of ten successive stimulus presentations (each marker represents one action potential) and the corresponding peri-stimulus-time histograms (bin width 20 ms). Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the fish (open arrowhead). FP Members of CILIA consortium PUBLIC 7 / 18

8 Figure 4: Example of two primary afferent fibres in the ALLN that exhibited triphasic responses to a moving object (size 1.0x1.0x9.5 cm, lateral distance 1 cm, speed 15 cm/s AP anterior-posterior, PA posterior-anterior object motion direction). Left: IEI pattern to AP motion direction and EIE pattern to PA motion direction. Right: EIE pattern to AP motion direction and IEI pattern to PA motion direction. The figure shows raster diagrams of ten successive stimulus presentations (each marker represents one action potential) and the corresponding peri-stimulus-time histograms (bin width 20 ms). Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the fish (open arrowhead). Figure 5: Average evoked discharge rates of 62 ALLN fibres in response to AP and PA object motion (object size 1.0x1.0x9.5 cm, lateral distance 1 cm, speed 15 cm/s). The regression line on the data (Y=0.9X+2.2, R=0.89) corresponds closely to the bisector. FP Members of CILIA consortium PUBLIC 8 / 18

9 EFFECTS OF OBJECT MOTION DIRECTION In most fibres (n=53, 85%) the pattern of discharge depended on object motion direction. In 23 of these fibres the response pattern to one motion direction was a mirror image of the response pattern to the opposite motion direction, i.e. EI patterns changed to IE patterns and EIE patterns changed to IEI patterns when motion direction was reversed (see examples on the left side of Figures 2, 3 and 4). This effect can be attributed to the intrinsic directional sensitivity of the hair cells within the neuromasts. Among the other 30 fibres the response pattern depended on object motion direction but not in a predictable way. Finally, in 9 fibres (15%) response patterns were the same to both motion directions (see Figure 3 right). In terms of discharge rate no difference was found between the average response rates measured while the object passed the fish in AP or PA motion direction (Fig. 5). Thus, across the population of fibres recorded, directional sensitivity was not apparent EFFECTS OF OBJECT SIZE Responses to objects of at least four different sizes were obtained from 25 afferent fibres. An example of the responses of one fibre is shown in Fig. 6. Note that this fibre started to discharge bursts of spikes after the object had passed the fish. These bursts most likely were elicited by pressure fluctuations and/or movements of the water column present in the object s wake. Figure 6: Example of the responses of a primary afferent fibre in the ALLN to moving objects of different sizes (lateral distance 1 cm, speed 15 cm/s, AP motion direction). The figure shows peri-stimulus-time histograms (bin width 20 ms) calculated across ten stimulus presentations. In nearly all fibres discharge rates increased with increasing object size and reached saturation with the biggest objects used (Fig. 7). In some fibres discharge rates did not FP Members of CILIA consortium PUBLIC 9 / 18

10 increase in a systematic way with increasing object size or decreased with the largest object used. We assume that these fibres changed their ongoing rates during the course of the experiment. Figure 7: Normalized discharge rates of primary afferent fibre in the ALLN to moving objects of different size (lateral distance 1 cm, speed 15 cm/s, AP motion direction) EFFECTS OF OBJECT DISTANCE Responses to objects that were moved at different distances alongside the fish were obtained from 13 afferent fibres. In 8 fibres evoked discharge rates decreased with increasing lateral distance between object and fish. However, the degree of rate decrease was different between individual fibres which is indicative of different sensitivities. Unexpectedly, in 5 fibres evoked rates increased with increasing object distance. Most likely this was due to the fact that in these fibres ongoing rates and/or the quality of the recording changed over the course of the experiment, since our stimulus measurements showed that water velocity and pressure changes caused by the moving object decreased with increasing object distance STIMULUS CHARACTERIZATION WATER VELOCITIES MEASURED WITH HOT-WIRE ANEMOMETRY The water velocities caused by a moving object were measured with a hot-wire anemometer (DANTEC CTA 90C10) while a fish was in the experimental tank. The measuring probe was placed as close as possible to the fish skin between eye and operculum. As in a previous study (Mogdans and Bleckmann 1998), the moving object caused highly reproducible changes in water velocity that consisted of a slow rise while the object approached the measuring sensor and a steep rise at the time when the object passed the hot-wire. This pattern was highly reproducible across stimulus repetitions (Fig. 8). FP Members of CILIA consortium PUBLIC 10 / 18

11 Figure 8: Time course of the water velocities caused by a moving object (size 1x1 cm, velocity 15 cm/s, AP motion direction). Five consecutively recorded traces are superimposed. Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the fish (open arrowhead). The general pattern of water velocity changes was comparable for AP and PA motion direction. However, object motion in the PA direction resulted in slightly higher velocity amplitudes than object motion in the AP direction (Fig. 9). We attribute this to the asymmetry of the fish head that is narrow at the snout and broad at the operculum. Consequently there was less space between object and fish when the object approached from the fish from posterior than when it approched the fish from anterior. This would result in higher water velocities. Figure 9: Effect of motion direction (AP versus PA) on the time course of the water velocities caused by a moving object (size 1x1 cm, velocity 15 cm/s). Each trace shows the average of 10 consecutive measurements. Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the fish (open arrowhead). With increasing object size both the amplitude and steepness of the rise of the water velocities measured as the object passed the sensor increased. In addition, water velocities generated by the wake of the object, i.e. measured after it had passed the hot-wire, increased with increasing size of the object (Fig. 10). With increasing lateral distance between object and fish the amplitude of the water velocities that were measured while the object passed the sensor decreased. In contrast, water velocities generated by the wake of large objects, i.e. that were measured after the object had passed the hot-wire, were measured even at 6 cm object distance (Fig. 10). FP Members of CILIA consortium PUBLIC 11 / 18

12 Figure 10: Effect of object size and distance on the time course of the water velocities caused by a moving object (object velocity 15 cm/s). Each trace shows an individual measurement. Numbers to the right of each trace indicate lateral distance between object and fish. Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the fish (open arrowhead). FP Members of CILIA consortium PUBLIC 12 / 18

13 PRESSURE CHANGES MEASURED WITH A HYDROPHONE The changes in pressure caused by a moving object were measured with a hydrophone (Bruel&Kjaer 8103). Due to the large diameter of the hydrophone (9mm) the measurements were made without fish in the tank and the hydrophone was place at the location where normally the fish head would reside. In terms of their relative amplitudes and time courses the pressure changes matched the results obtained with the hot-wire anemometer. As expected (Mogdans and Bleckmann 1998), the moving object caused highly reproducible changes in pressure at the time when the object passed the hydrophone. The general temporal course of the pressure change was comparable for AP and PA motion direction. However, object motion in the PA direction resulted in slightly larger pressure change than object motion in the AP direction. With increasing object size the amplitude of the pressure changes measured as the object passed the hydrophone increased. In addition, pressure changes generated by the wake of the object, i.e. measured after it had passed the hot-wire, increased with increasing size of the object. The largest objects that we used (3.5 and 4 cm) produced pressure changes in their wakes that were about as large as the pressure changes measured as the object passed directly alongside the hydrophone. Moreover, these pressure changes lasted for a very long time (Fig. 11). With increasing lateral distance between object and fish the amplitude of the pressure waves measured as the object passed the sensor decreased. In contrast, pressure changes generated by the wake of the objects were large even at 6 cm object distance (Fig. 11) WATER SURFACE WAVES The moving object caused water surface waves that were measured with a Wheatstone bridge (custom built) at the location of the fish head but without a fish in the tank. The time course and amplitude changes of the water surface waves were in agreement with the measurements made with the hot-wire anemometer and the hydrophone. Measurements were highly reproducible and comparable for AP and PA motion direction with slightly larger surface waves in the PA direction than in the AP direction. With increasing object size the amplitude of the surface waves measured as the object passed the sensor increased. In addition, surface waves generated by the wake of the object, i.e. measured after it had passed the sensor, increased with increasing size of the object. The largest objects that we used (3.5 and 4 cm) produced pressure changes in their wakes that were about as large as the surface waves that were measured when the object passed directly alongside the hydrophone. Moreover, these surface waves lasted for a very long time (Fig. 12). With increasing lateral distance between object and fish the amplitude of the surface waves measured as the object passed the sensor decreased, at least with the smaller objects. In contrast, surface changes generated by the wake of the objects were large even at 6 cm object distance (Fig. 12). FP Members of CILIA consortium PUBLIC 13 / 18

14 Figure 11: Effect of object size and distance on the time course of the pressure waves caused by a moving object (object velocity 15 cm/s). Each trace shows a single measurement. Numbers to the right of each trace indicate lateral distance between object and fish. Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the hydrophone (open arrowhead). FP Members of CILIA consortium PUBLIC 14 / 18

15 Figure 12: Effect of object size and distance on the time course of the surface waves caused by a moving object (object velocity 15 cm/s). Each trace shows a single measurement. Numbers to the right of each trace indicate lateral distance between object and fish. Arrows indicate start of object motion (filled arrowhead) and the moment when the object passed the surface wave sensor (open arrowhead). FP Members of CILIA consortium PUBLIC 15 / 18

16 4. CONCLUSIONS 4.1. COMPARISON ALLN - PLLN The present data were gathered in order to characterize the representation of moving source by lateral line afferent fibres in the ALLN and to compare them with previous data gathered from the PLLN (Mogdans and Bleckmann 1998). For this reason we used the same experimental set-up as was used in that previous study. The results from the measurements that we made to characterize the hydrodynamic events generated by objects moving in the tank were in almost all aspects identical to those measured in the previous study. This indicates that the stimuli delivered to the lateral line were the same in both studies, a necessary prerequisite to compare the neural data. The central finding of our study is that all of the response patterns that we observed in the ALLN were also found previously among PLLN fibres. However, the relative proportions of the patterns differed between ALLN and PLLN (Table 1). Table 1: Proportion of response patterns to a moving object recorded from nerve fibres in the ALLN and PLLN of goldfish. Nerve ALLN PLLN Monophasic Responses (%) 31 2 Biphasic Responses (%) Triphasic Responses (%) Other Responses (%) 6 2 We suggested for the PLLN that biphasic responses originated from fibres innervating canal neuromasts whereas triphasic responses originated from fibres innervating superficial neuromasts (for a detailed discussion see Mogdans and Bleckmann 1998). Among other arguments this interpretation was based on the fact that neuromasts respond in proportion to pressure gradients (Coombs et al. 1996, Coombs and Conley 1997) and that pressure gradient patterns and consequently neural response patterns of lateral line afferent fibres depend on the relative orientation of receiver (neuromast) and source (moving object). The same rationale may also apply for the ALLN. In the present study we found a substantially larger proportion of monophasic and biphasic responses compared to the PLLN. Moreover, substantially more fibres did not respond with a reversal of discharge pattern when object motion direction was reversed. Most likely these differences are due to the fact that the spatial organization and perhaps the physiological properties of the head lateral line differ from that of the trunk lateral line. Clearly the spatial arrangement of lateral line canals on the fish head differs from that on the trunk in that canals for most of their length do not run parallel to the long axis of the fish (e.g. Webb 1989). In addition, even though the goldfish possesses about as many superficial neuromasts on the head as on the trunk (Puzdrowski 1989, own studies), their spatial orientation relative to the long axis of the fish may be different from that on the trunk. Unfortunately no study has ever described in detail the axes of sensitivity of superficial neuromasts. Therefore we are presently investigating this aspect in WP 1.3. In addition to differences in lateral line morphology there are also differences in geometry between fish head and trunk. As discussed in section 3.5 this may have resulted in slightly different hydrodynamic stimulation depending on object motion direction. This may contribute FP Members of CILIA consortium PUBLIC 16 / 18

17 to the fact that many ALLN fibres exhibited different response patterns to different object motion directions. With increasing object size, evoked responses of ALLN fibres increased. In the PLLN the effect of object size on fibre responses was not investigated systematically. However, the measurements made to characterize the hydrodynamic events caused by the moving object suggest that the observed result is largely due to an increase in stimulus strength with increasing object size. With increasing object distance, evoked responses of ALLN fibres decreased. A similar effect was found in the PLLN. Again this effect is largely due to the fact that stimulus strength decreased with increasing object distance RESPONSES TO MOVING OBJECTS IN RUNNING WATER In the study presented here we recorded responses to moving objects exclusively under still water conditions for two reasons. First, the main aspect of the experiments was to compare responses of ALLN and PLLN fibres, an important aspect because the spatial organization of the lateral line on the goldfish head may differ from that on the trunk. We therefore used the same experimental setup that was used previously for PLLN recordings. This setup was not designed for and did not allow the generation of a background water flow. Second, the effects of running water on neural responses to moving objects have been studied previously in goldfish and trout (Engelmann et al. 2003). In this study, recordings were made from PLLN fibres. The data showed that in general, running water affected the responses of goldfish nerve fibres more strongly than the responses of trout fibers indicating that the lateral line systems in these two species are adapted to different hydrodynamic conditions. However, measurements of changes in pressure and water velocity caused by the moving object indicated that the observed effects can largely be explained by peripheral hydrodynamic effects. Therefore, we do not expect any additional or different effects of running water on the responses of ALLN fibres compared to those observed in the PLLN. FP Members of CILIA consortium PUBLIC 17 / 18

18 5. REFERENCES Coombs S, Conley RA (1997) Dipole source localization by mottled sculpin II. The role of lateral line excitation patterns. J Comp Physiol A 180: Coombs S, Hastings M, Finneran J (1996) Modeling and measuring lateral line excitation patterns to changing dipole source locations. J Comp Physiol A 178: Engelmann J, Kröther S, Bleckmann H and Mogdans J (2003). Effects of running water on lateral line responses to moving objects. Brain Behav Evol 61: Mogdans J and Bleckmann H (1998) Responses of the goldfish trunk lateral line to moving objects. J Comp Physiol A 182: Puzdrowski RL (1989) Peripheral distribution and central projections of the lateral-line nerves in goldfish, Carassius auratus. Brain Behav Evol 34: Webb JF (1989) Developmental constraints and evolution of the lateral line system in teleost fishes. In: Coombs S, Görner P, Münz H (eds) The Mechanosensory Lateral Line. Neurobiology and Evolution, Springer, Berlin, Heidelberg, New York, pp FP Members of CILIA consortium PUBLIC 18 / 18