The authors have declared that no competing interests exist.
The escape behavior of a cockroach may not occur when it is either in a quiescent state or after being stung by the jewel wasp (
The Jewel Wasp uses live cockroaches,
Interestingly, cockroaches display an innate quiescent state that is correlated to daytime and that behaviorally resembles the venom-induced lethargic state [
The CX is known to receive multiple sensory inputs, including visual [
Cockroaches (
Adult male cockroaches with intact antennae were placed in an arena 5 hours after having being stung or un-stung. To induce quiescence, we used a protocol described in Watson and Ritzmann (1994). By positioning a plastic petri dish over the cockroach that touched both antennae (referred to as 'antennal-contact'), several attempts to induce quiescence were made. The duration of induced quiescence of both free-ranging and tethered cockroaches was measured using a stopwatch. Tactile stimuli to the last abdominal segments (3 seconds interval stimuli) with a no.1 artists brush were used to assess the level of responsiveness of the animal, and the stimulus was repeated until the cockroach either regained alertness and performed a stereotypic escape response or initiated spontaneous locomotion. For electrophysiological recordings, we used either stung cockroaches or cockroaches that entered quiescence in more than 80% of the trials.
Cockroaches were cold-anesthetized and immobilized in modelling clay. An electromyogram (EMG) bipolar electrode, made of two 38μm formvar coated nichrome wires, was inserted manually into the coxal depressor muscle of the metathoracic leg (muscle 177d). This muscle is controlled by only two excitatory motoneurons, the slow and fast coxal depressor motoneurons (Ds and Df motoneuron, respectively) [
To test the involvement of the head ganglia and specifically the CX in the control of the Ds motoneuron activity, EMG recordings from the coxa were combined with focal procaine injection into the CX. Procaine is a reversible sodium voltage-dependent channel blocker which non-selectively prevents action potential in the site of injection when injected in the cockroach’s central nervous system [
The EMG recording was amplified with a high-gain differential amplifier (AM systems model 1700), and acquired with a Micro 1401 (CED) data acquisition unit (DAQ), recorded as waveform data and saved directly to a computer. The onset of Stimulus was also recorded through the DAQ.
At the end of the experiment, the head was severed and placed in formalin (Sigma) overnight. The brain was then excised from the head, embedded in agar (6% agar in saline, Sigma), and sliced with a vibratome (Leica VT 1000S). The location of the Janus Green tracer in the brain was identified using a light microscope. Only trials in which the CX was confirmed as the injection site were used for further analysis.
Behavioral data regarding the response to tactile stimuli was analyzed by averaging the number of stimuli needed to induce an escape response for each of the three groups of cockroaches ("Awake", "Quiescence" and "Stung"). Only trials that resulted in an escape response were considered for the statistical analysis.
Waveform data from the electrophysiological recordings was analyzed offline using Spike2 software (CED) and then exported to Excel. Only spikes with amplitude of at least twice that of the background noise were considered. For all EMG recordings, the number of spikes per second (spike/s) was calculated and then averaged for each group of cockroaches. In addition, the same data was analyzed using an interval histogram analysis, at a 0.1 seconds Interval Length. This data was normalized by dividing the number of intervals in each bin by the total number of intervals for each cockroach and then averaged among cockroaches from the same group.
All statistical tests were performed using SigmaPlot 13.0 software. The statistical significance was determined by using t-test, one-way ANOVA, Kruskal-Wallis one-way ANOVA on Ranks or Friedman Repeated Measures ANOVA on Ranks. For 'All Pairwise Multiple Comparison' procedures, the test recommended by 'SigmaPlot' was used.
Quiescence induction (
Quiescent cockroaches remained motionless and did not engage in locomotory activities such as grooming or walking. The body was positioned parallel to the ground with a low posture, head down with the mouth parts and antennae motionless; the latter were resting limply on the surface. Such a posture has been described in other insects as well [
The responsiveness to stimuli in awake, quiescent and stung cockroaches (Awake; n = 14, 42 trials; Quiescence; n = 20, 76 trials and Stung; n = 7, 29 trials) was measured by the number of stimuli required to induce an escape response. Awake cockroaches reacted with an escape response to the first stimulus in every trial (42 trials). Quiescent cockroaches generally reacted to repeated stimuli (54 trials); however some quiescent cockroaches escaped after the first stimulus (21 trials) or did not escape at all (1 trial). None of the stung cockroaches performed an escape response after the first stimulus and most of the stung cockroaches did not escape even after repetitive stimuli (26 trials), except in 3 trials when escape did occur. The average number of stimuli±SEM required to induce an escape response (only of the trials that resulted in escape) was found to be significantly different among the three groups (P<0.001, Awake = 1.0±SEM, Quiescence = 3.2±0.2 and Stung = 6.0±1.0).
The fact that quiescent cockroaches reacted to repetitive stimuli with an escape response shows that quiescent cockroaches have an elevated arousal threshold and also that the quiescent state is reversible with stimulation. Similarly to the quiescent cockroach, the stung cockroach also displayed a heightened threshold for responding to stimuli. However, in contrast to quiescence, the number of stimuli required to induce an escape was higher and in general, even repeated stimuli did not induce an escape response. In addition, the venom-induced lethargy is long-lasting and reverses only after 5 to 7 days [
Both the stung lethargic state and the quiescent state have some similarities to a naturally occurring rest state in cockroaches that is correlated to daytime [
Using EMG recordings from the coxal depressor muscle of the metathoracic leg, the transition between awake and quiescent states of the cockroach was explored (
(A) Transition between awake and quiescence. Top trace: Change in Ds activity before, during, and after the transition between awake and quiescent states in the same animal. The transition from awake state to quiescent state following antennal-contact induced quiescence is accompanied by a decrease in Ds firing rate. The large amplitude spike is an artifact occurring during antennal contact. Bottom trace: The transition from quiescent state to awake state in the same animal occurs spontaneously with a short burst followed by an increase in Ds firing rate. (B) Representative EMG recording traces of Ds ongoing activity in awake (top trace), quiescent (middle trace) and stung (bottom trace) immobilized cockroaches. (C) Each bar represents the average spikes/second±SEM (Awake = 18.2±3.2; Quiescence = 8.4±2.0 and Stung = 2.3±0.4). The average value for each group was significantly different from the other two groups (P<0.001): the Awake group displayed the highest muscle tone and the Stung group the lowest (n = 12 for each group). (D) Interval histograms of Ds spikes in awake (top histogram), quiescent (middle histogram) and stung (bottom histogram) cockroaches. Each bar represents the normalized number of spike intervals in each time bin (interval prevalence) ±SEM. Data points labeled with different letters are significantly different from each other (P<0.001 for 'A' label and P<0.05 for 'B-D' labels). (n = 12 for each group).
Moreover, the relatively-constant spike intervals in awake cockroaches became more variable in quiescent and stung cockroaches (
The Jewel Wasp stings cockroaches directly into their brain in order to control their ability to move. Such behavioral manipulation is achieved by the injection of venom inside and around the central complex (CX), a neuropil in the supra-esophageal ganglion (SupEG or “brain”). The CX is known to be involved in sensory integration and pre-motor processing [
(A) Representative EMG recording traces of Ds ongoing activity after procaine injection to the CX. (B) Each bar represents the average spikes/second ±SEM. The average values of the t = 10–50 min time point was significantly different for the two groups (P<0.05). (Procaine-CX; n = 6; Saline-CX; n = 6).
Regarding the fast motoneuron, Schaefer and Ritzmann (2001) have examined Df recruitment in decapitated cockroaches [
To summarize, both the quiescent and the stung states differ significantly from the awake state with regard to: 1) an increased threshold for initiation of walking and 2) decreased postural Ds motoneuron activity. These changes might be mediated by the head ganglia pre-motor circuits since injection of Procaine to the CX decreases Ds motoneuron activity in a manner similar to the wasp venom injection in the CX. These results suggest that the venom-induced lethargic state in cockroaches might represent an extreme version of the quiescent state and that both states might be controlled by the head ganglia. If true, this implies that, in the course of the coevolved arms race between a parasite and its host, the parasite has found ways to tap into an existing neuronal circuit which is part of the quiescence-regulating network for its own benefit.
The video displays the normal escape response of a cockroach after a tactile stimulus to the abdomen with a brush. Next, the altered escape response of a stung cockroach (marked with a white dot on back) is shown.
(WMV)
An adult male cockroach is placed in an arena and a plastic petri dish is positioned over the cockroach and touching both antennae. The antennal contact induces quiescence in the cockroach, which exhibits a change in posture and lack of antennal and mouth-parts movement. Repetitive tactile stimuli with a brush does not evoke an escape response and only after waking, marked by the antennal movement, does locomotion occur.
(WMV)
We would like to thank Gal Haspel and Maayan Kaiser for commenting on and editing an earlier version of the manuscript and Tova Rapoport for editing the final version of the manuscript. This work is currently supported by the United States-Israel Binational Sciences Foundation (2015161). The experiments performed comply with Principles of Animal Care, NIH publication no. 86–23, revised in 1985, and with the current laws of the State of Israel.