Fig 1.
The two different types of leg removal utilized.
Fig 2.
A. Side view of “RoachScope” custom 3D printed microscope with integrated 10x loupe lens. B. Placement of smartphone on microscope to take stable photographs.
Fig 3.
Example of measurement of cockroach leg photographs using 10 mm reference guides on microscope base and the NIH ImageJ analysis program.
Fig 4.
Observation of cockroach legs at various time points.
Note that at 43 days a well-formed regenerated leg exists in the autotomy cockroach, but the coxa-cut cockroach only reveals a healed coxa and lacks a regenerated leg. At 125 days, the autotomy-removed leg has fully regrown, but in the coxa-cut leg, the regenerated leg is still quite small. Images over time are not necessarily of the same cockroaches.
Fig 5.
Percentage of new molts with a regenerated leg.
Y–axis refers to the percentage of molts observed at the indicated time point that revealed a regenerated leg. After day 43 post-leg removal, in the autotomy group, almost all molts contained a regenerated leg. Similar observations were delayed in the coxa-cut group. The dip at day 110 in the autotomy group (12 of 15 observations showing a regenerated leg) and at day 170 in the coxa-cut group (4 of 7 observations showing a regenerated leg) is likely due to molt failure, with the already regenerating leg self-autotomizing during the molt.
Fig 6.
Size of regenerated leg in comparison to contralateral uncut control leg.
In molts where a regenerated leg was observed, the size of the femur and tibia were measured on both the regenerated leg and the control leg. Femur and tibia length were then added together, and the size of the regenerated leg as a percentage of the length of the contralateral control leg was calculated. Autotomy-removed legs regrow earlier and are larger than coxa-cut legs.
Fig 7.
Neural Recordings in Different Electrode Configurations.
Electrodes were first placed in the coxa and tibia tarsus joint; spontaneous rate was recorded for 10 seconds, followed by 10 seconds of lightly touching the tibial barbs with a plastic probe. The coxa electrode was then moved to the femur, and a similar recording of spontaneous activity and tactile activity done. Finally, the coxa was removed with forceps, and the measurements of neural signals repeated. The average root mean square, a measure of signal amplitude, for all 8 legs is shown underneath the traces and is reported in arbitrary A/D (analog to digital) conversion units.
Fig 8.
Firing Rate Analysis in Different Electrode Configurations.
Firing rate was measured during 10 seconds of spontaneous activity followed by 10 seconds of evoked spiking (by manually deflecting the tibial barbs) during all three conditions. Firing rate average and standard deviation from all 8 legs measured is shown (same data set as shown as Fig 7). Spontaneous Firing Rate and Barb Deflection Firing Rate was significantly higher when one electrode was in the coxa versus when one electrode was in the femur. No differences were observed between the coxa attached and coxa removed femur electrode conditions. * = p<0.05, Student’s t-test.
Fig 9.
Microstimulation Thresholds for Eliciting Cockroach Leg Deflection.
Electrodes were first placed in the coxa and femur and connected to modified audio cables plugged into an iPhone 5s running the “Tone Generator” App. The volume necessary (threshold) to cause visible leg movement at two electrical sine wave frequencies was then measured. The coxa electrode was subsequently moved alongside the femur electrode, and thresholds tested again. Finally, the coxa was removed with forceps, and thresholds measured similarly. Thresholds were significantly higher with both electrodes in the femur (bipolar configuration) versus one electrode in the coxa and the other in the femur (monopolar configuration). There was not a significant difference in thresholds between the coxa present and coxa absent conditions with both electrodes in the femur. Thresholds were measured in 8 legs during the three conditions and then averaged. Voltage output of iPhone was measured with an oscilloscope and is also reported next to threshold values. * = p<0.05.
Fig 10.
Number of molts observed during each measurement period in the four groups.
Differences between groups was not apparent, though the “group synchronization” of peak molting at day 100 and 150 may deserve further examination in future work.