2008-2009 Abstracts from the Lab

• Mountcastle, A.M., Daniel, T.L. (2009). Wing stiffness affects mean advective flows of Manduca sexta, with wing overlap a potential contributor. SICB, Boston, MA. [Abstract]

  Mountcastle, A.M., Daniel, T.L. (2009). Wing stiffness affects mean advective flows of Manduca sexta, with wing overlap a potential contributor. SICB, Boston, MA.
  
  Many insects have wings that bend and twist during flight, often with dramatic deformations. The pattern and extent of deformation are dependent on wing flexural stiffness and the boundary conditions that govern actuation. Prior work has shown that the extent of deformation during hovering in Manduca can vary between strokes. The aerodynamic consequences of wing compliance, however, remain largely unknown. In this study, we examined the effects of wing stiffness on the overall induced flow in the wings of the hawkmoth, Manduca sexta. We subjected moth wings to robotic actuation in their dominant plane of rotation at the natural wing beat frequency of 25 Hz. We used digital particle image velocimetry at high temporal resolution (2,100 fps) to assess the influence of wing stiffness on the mean advective flows of three wings, each tested in a fresh, flexible state and a desiccated, stiff state (overall spanwise flexural stiffness increased 2-2.5x). We find that flexible wings yield mean advective flows with total magnitudes 2-4x those of their stiff wing counterparts, and vertical (lift-favorable) components that are 7-31x those of stiff wings. If flight forces are sensitive to wing deformation, then any mechanism that alters deformation is a potential source of flight control. We show that the overlap between forewing and hindwing can vary by 15% during ventral stroke reversals in Manduca. Flexural stiffness tests on extracted wing pairs reveal that overall spanwise stiffness can increase by a factor of 1.3 for a similar change from min. to max. wing overlap. Our results show that wing compliance may play a critical role in the production of insect flight forces, and suggest the possibility that wing overlap may affect compliance.

Williams,C David; Regnier, Mike and Daniel,T.L. 1427-Pos Simulating The Effect Of Lattice Spacing On The Frank-Starling Mechanism Biophys. J. 2008 94: 1427. [Abstract]

C David Williams, Mike Regnier, and Thomas L. Daniel 1427-Pos Simulating The Effect Of Lattice Spacing On The Frank-Starling Mechanism Biophys. J. 2008 94: 1427.

The spatial relationship between thin and thick filaments is one mechanism regulating myocardial performance. It is thought that the interaction of sarcomere length and lattice spacing in constant volume contraction is responsible for regulation of cross-bridge recruitment and alterations of force generation such as the Frank-Starling mechanism. Previous spatially explicit models operate in a one dimensional space that permits no alteration of filament spacing, and thus are unable to consider the role of lattice spacing in force development. Here, we develop a two-dimensional spatially explicit model of actin-myosin interaction. We relax the assumption that cross-bridges behave as simple linear springs aligned with the axes of the myofilaments and instead employ a coupled linear/torsional spring mechanism for each myosin’s mechanics. Thus we more accurately reflect the lever-arm model of force generation. The system’s kinetics use two-dimensional, spatially explicit, stochastically-driven cross bridge cycling with a three state binding model. Cross-bridge recruitment is driven by constrained diffusion of myosin heads, and an iterative solution method is used at each time-step to balance all forces on each point along the myofilaments. Results address how the interaction of filament spacing and compliant realignment of binding sites can at least partially account for the Frank-Starling relationship, showing an increase in force generated at smaller filament spacings. This supports the idea that increased force development at longer sarcomere lengths is, at least partially, due to changes in lattice spacing. [Supported by HL65497 and funds from the Komen Endowed Chair]

George, N.T., Daniel, T.L. (2009). Temperature gradients in the dorsolongitudinal flight muscles of Manduca sexta may yield functional gradients. SICB, Boston, MA. [Abstract]

C George, N.T., Daniel, T.L. (2009). Temperature gradients in the dorsolongitudinal flight muscles of Manduca sexta may yield functional gradients. SICB, Boston, MA.

During muscle contraction, heat is produced as chemical energy is converted into mechanical work. Many large insects with active flight muscles use this byproduct to elevate flight muscle temperature, thereby achieving higher mechanical power output during flight. Contractile heat production paired with convective and radiative heat loss necessarily lead to a temperature gradient, but the functional consequences of such a gradient remain unknown. Because force generation of muscle depends on temperature, subunits experiencing lower temperatures could function differently than those at higher temperatures. This is particularly relevant to the flight musculature of Manduca sexta. The dominant flight muscles (dorsolongitudinal muscles: DLMs) are divided into five subunits, each separately innervated. We measured two important aspects of temperature dynamics in the DLMs during tethered flight: (1) using a hypodermic thermocouple probe, we showedthat there is a strong temperature gradient in the dorso-ventral direction, with a mean difference of 8.8 °C (a max of 10 °C; n = 7) across 5 mm, (2) using standard electromyography, we showed that the 5 subunits of the DLMs are activated nearly simultaneously (max time difference = 2 ms, 5% of cycle time; n = 3). Therefore, the muscle bundles do not appear to employ a spatial offset in timing to correct for the thermal gradient that they generate. Taken together, the observation of simultaneous activation and a strong thermal gradient suggest that the dorsal-most subunits may function differently from warmer, more ventrally located units.

Fox, JL and Daniel, TL. (2009). Estimation of information transfer rates in highly precise sensory afferents. SICB, Boston, MA. [Abstract]

C Fox, JL and Daniel, TL. (2009). Estimation of information transfer rates in highly precise sensory afferents. SICB, Boston, MA.

To coordinate their motion, animals rely on sensory systems to acquire, process, and transmit necessary information from the environment. Dipteran insects use specialized structures known as halteres to detect forces that occur as a result of body motions during flight. The primary afferents extending from haltere mechanoreceptors respond to stimuli with extremely high timing precision, suggesting that they are capable of transmitting information at high rates (Fox and Daniel 2008). Given this high degree of precision, we sought to directly measure the mutual information between a stimulus and the resulting spike train. We recorded the activity of more than 30 haltere primary afferent cells while mechanically stimulating the haltere with band-limited Gaussian white noise. In doing so, we directly measure the rate of information transfer. We found that many haltere primary afferent cells are able to transmit information at a rate of at least 60 bits per second and up to 133 bits per second, significantly higher than the rate found in many visual systems. Additionally, we used the measured jitter in response to repeated sine waves (n = 15 cells) to estimate the bit rate as a function of frequency, allowing us to create a neural tuning curve in the common currency of bits per second. By using this modality-independent metric of neural encoding, we can assess the sensory conduction of haltere primary afferents in the context of other information-processing systems.

Fox, JL and Daniel, TL. (2008). Information processing in a biological gyroscope. Annual meeting of the Society for Neuroscience, Washington, DC.[Abstract]

C Fox, JL and Daniel, TL. (2008). Information processing in a biological gyroscope. Annual meeting of the Society for Neuroscience, Washington, DC.

Sensory systems must acquire and process information on a timescale that is relevant to the animal’s behavior. For flying insects, this timescale can be as short as the period of a single wingbeat, requiring a sensory modality operating on the scale of tens of milliseconds. While flying, flies (Diptera) obtain mechanosensory information with extremely low latency from reduced hindwings known as halteres. The halteres contain several fields of campaniform sensilla that respond to externally-imposed forces, the largest of which is the Coriolis force that occurs during body rotations. This force causes lateral deflections of the haltere and contains components at twice the frequency of oscillation. The information gathered by haltere primary afferent neurons is sent monosynaptically to a wing-steering motoneuron (as well as to other locations in the CNS); therefore, there are very few intermediate steps and extremely limited processing time for the incoming mechanosensory information to be modified before it is used to direct behavior. In contrast to multi-synaptic sensory systems, the haltere mechanosensory system offers a rare opportunity to simultaneously measure information at both the earliest sensory and immediate premotor stages. As such, an analysis of the information capacity in the primary afferent layer is essential for understanding of the haltere’s function in flight control.Our current work shows that haltere primary afferent neurons respond rapidly to mechanical stimuli, and that they do so with extremely high precision. At the behaviorally relevant frequencies, neurons fire with high vector strength between 0.935 and 0.999. Similarly, neurons respond with sub-millisecond jitter (0.130 to 0.927 ms) at the relevant frequencies. Thus, we predicted that primary afferents are able to process information at high rates. To measure the neuronal information-processing capacity, we recorded from single cells while oscillating the haltere with band-limited Gaussian noise. Raster plots of responses to a repeated Gaussian stimulus showed that firing patterns were highly reproducible with low jitter. We used the methods of Strong et al. (1998) and Brenner et al. (2000) to directly calculate the cells’ bit rates by subtracting conditional entropy (measured from responses to repeated stimuli) from the entropy of the spike train. We also used covariance analysis (Fairhall et al., 2006) to determine the relevant stimulus features that are detected by primary afferents. These methods provided a quantitative measure of signal transduction in this unique system that rapidly and precisely guides a complex behavior with minimal stimulus processing.

Hinterwirth, AJ and Daniel, TL. (2009). Antennae mediate an abdominal flexion response to body rotations in the hawkmoth Manduca sexta SICB, Boston, MA. [Abstract]

Hinterwirth, AJ and Daniel, TL. (2009). Antennae mediate an abdominal flexion response to body rotations in the hawkmoth Manduca sexta SICB

The crepuscular moth Manduca sexta does not rely on vision alone for flight control. It uses mechanosensory information from its antennae to mediate rapid responses to aerodynamic disturbances (Sane et al. 2007). Specifically, antennae may act as vibrational gyroscopes detecting Coriolis forces that occur when a moth undergoes rotational motion. However, there are no data that clearly show moths respond to pure rotational motion stimuli in absence of any other inputs, such as visual or wind stimuli. To address the role of antennae as sensors of body rotations, we developed an experimental setup that allows us to investigate the respective influences of the visual and rotational mechanosensory systems. We tether a moth in an LED visual arena that can be mechanically rotated (after Sherman & Dickinson, 2003). Visual or mechanical rotations can thereby be presented independently, or in any arbitrary phase with respect to each other. At the same time, we monitor two behavioral responses: abdominal flexion and wing trajectory. Our results show that antennae mediate abdominal flexion as a response to pure mechanical rotations. The angle of flexion increases when the rotational velocity is increased (up to 250 deg/s). Moreover, removing the antennal flagellum diminishes the response significantly, in many cases completely (Mean reduction to ~18% of control group’s gain, N=14 animals). The abdominal response, however, can be rescued by gluing back a flagellum on each antennal base, as long as mechanoreceptors on the scape and pedicel are left intact. (Rescue to ca. 70% of the control group’s gain in 7 of 8 animals in which antennal re-attachment was performed.) These results thus provide strong evidence for the antennal gyroscope hypothesis.

Tse J, Jong P, Hinterwirth AJ, and Daniel TL (2009). Stimulating antennal muscles leads to path changes in a moth’s flight trajectory. SICB, Boston, MA. [Abstract]

C Tse J, Jong P, Hinterwirth AJ, and Daniel TL (2009). Stimulating antennal muscles leads to path changes in a moth’s flight trajectory. SICB, Boston, MA.

Previous work has shown that mechanoreceptors at the base of the antennae in the crepuscular hawkmoth Manduca sexta are necessary for stable flight control. They mediate flight responses by acting as vibrational gyroscopes, detecting Coriolis forces that appear during body rotations (Sane et al. 2007). We hypothesize that artificially introducing a force that is applied out of the antennae’s vibrational plane should lead to compensatory reactions and therefore to changes in a moth’s flight path. To test this hypothesis, we created small tungsten electrodes that are implanted in the dorso-medial side of the scape, targeting specific extrinsic muscles that lead to antennal retraction (out of the natural vibrational plane). We use a low current 5 V square wave stimulus generated by an Arduino board at 60 Hz with a duty cycle tuned to elicit an antennal retraction without suppressing flight behavior. The stimulus is delivered to the implanted electrodes via a cable of ultrafine stainless steel wire. Thus moths are loosely tethered to the stimulator and permitted to fly. Preliminary experiments show that, in 4 out of 14 animals in which electrode implantation was attempted, stimulation of the left antennal muscles leads to a high probability of left turns immediately following the stimulus. The direction of the elicited response agrees with what is expected if antennae act as gyroscopic sensors. These results provide further evidence that moth antennae are crucial components of a mechanosensor-mediated flight control circuit .

Loudon, S., Aldworth, Z., and Daniel TL (2009). Perturbing flight paths in Lepidoptera by inducing abdominal flexion.. SICB, Boston, MA. [Abstract]

Loudon, S., Aldworth, Z., and Daniel TL (2009). Perturbing flight paths in Lepidoptera by inducing abdominal flexion.. SICB, Boston, MA.

Insects use rapid body rotations during flight control. These rotations are generally considered to be initiated by the wings. Insects can also use the movement of other body parts, such as the head, legs or abdomen, to influence their flight path. A recent study in the crepuscular hawkmoth, Manduca sexta, has suggested that abdominal movements are used during tracking flight to cancel pitching and yawing motions initiated by the wings. However, it is unknown to what extent abdominal movements themselves can be used to initiate changes in the flight path. To address this issue we elicited abdominal movement in Manduca through electrical stimulation during loosely tethered flight in order to determine the extent of association between abdominal movement and changes in flight trajectory. We used multiple high-speed cameras to film the animals’ flight in a closed chamber under low-light conditions. Stimulation current was delivered using tungsten electrodes inserted through the ventral cuticle of the thorax. Four stereotyped body points were digitized from the flight sequences, and used to extract 3-dimensional flight parameters such as yaw and pitch, as well as the dorsal-ventral and lateral flexion rates of the abdomen. We then used correlation analysis to determine the relationship between the abdominal movement and the observed changes to the flight path. We consistently found abdominal movements and changes in flight posture following stimulation, and flexion data were correlated with the attitude changes at delays of between 500-750 ms. These results demonstrate that abdominal movements can be used in insect flight control.

Aldworth,Z. and Daniel TL (2009). Wing mechanosensors can transmit bending information at high bit rates. SICB, Boston, MA. [Abstract]

Aldworth,Z. and Daniel TL (2009). Wing mechanosensors can transmit bending information at high bit rates. SICB, Boston, MA.

All insects are equipped with mechanosensory structures in their wings (campaniform sensilla), many of which encode wing bending or strain. In large insects, such as Manduca sexta, wings deform significantly during flight. Moreover such deformation can be an important determinant of the aerodynamic forces generated by such wings. Additionally, the bending waves seen in Manduca wings may travel at high speeds, well in excess of the time associated with a single wing flap. The extent to which the nervous system can encode such information, however, remains unknown. To address this issue we measured the information transfer rates (bit rates) of wing mechanosensors. We used a band limited Gaussian white noise mechanical stimulus applied to wings with simultaneous intracellular recording from primary sensory neurons. From the statistics of the signal and the emergent spike train we were able to compute the information carried by each spike, and the information transfer rate (bits/s) as a function of stimulus frequency. We found that the jitter (standard deviation of spike occurrence time) is extremely low (200 ms) and that the bit rate of information transfer exceeds 100 bit/s and is maximal at a frequency of at least 200 Hz. These results suggest that mechanosensory neurons can transmit strain information at rates that are sufficiently fast to detect bending waves of flapping wings.