Transient heat stress during the critical period leads to an unstable network.

A) Acute increase of ambient temperature leads to increased network activity, as measured with the fluorescent calcium indicator GCaMP8f expressed in segmentally repeated aCC motoneurons in isolated central nervous systems; while constant 25°C leads to reduced activity over time. Left: micrograph of an isolated ventral nerve cord with segmentally repeated aCC motoneurons expressing GCaMP8f. Centre: traces of GCaMP fluorescence peaks measured at different temperatures. Right: quantification of fluorescence peak frequency changes over time (control/sham -bottom trace) and in response to temperature increase (top trace). B) Heat stress, when experienced during the embryonic CP of locomotor network development, leads to reduced network stability. Left: Schematic representation of embryonic and larval development and the paradigm of temperature manipulation during embryogenesis. Embryos were kept at the control temperature of 25°C, either continuously (control) or moved to 32°C for 2 hours, either before (13-15 h, 15-17 h AEL), during (17-19 h AEL) or after (19-21 h AEL) the locomotor network CP, then returned to 25°C until the late larval stage, when subjected to seizure induction by electroshock. Right: quantification of seizure recovery times. n=30, data are shown with mean ± SEM, ANOVA, ****p < 0.00001, ‘ns’ indicates statistical non-significance.

Morphological changes at the neuromuscular junction after critical period perturbation.

A) Schematic of temperature manipulation during embryogenesis and location of dorsal acute muscle 1 (DA1) within a fillet dissected larva. Embryos collected over a 6h period were incubated at 32°C for 24h to guarantee heat stress exposure during the locomotor network CP. B) This results in a change of subsequent NMJ development, which manifests as a significant overgrowth of aCC presynaptic terminals (green, visualised with anti-HRP-AlexaFluor488). The increase in bouton number and NMJ area is paralleled by concomitant increases in active zones (magenta), resulting in a normal active zone density. C) Postsynaptically, embryonic heat stress causes in a significant decrease of glutamate receptors containing the larger conductance GluRIIA subunit, while those with the GluRIIB subunit remain unaffected. Thus, the postsynaptic GluRII receptor composition is significantly changed following a transient embryonic heat stress, as reflected in the ratio of receptors containing GluRIIA vs GluRIIB. Data are shown with mean ± SEM, unpaired t-test, ***p < 0.0001, ****p < 0.00001, ‘ns’ indicates statistical non-significance.

Heat stress pertubations during distinct periods of development differentially affect postsynaptic GluRII receptor composition vs NMJ size.

A) Schematic of temperature manipulations during embryogenesis. Embryos collected over one hour were incubated at 32°C for different phases of embryogenesis. B) Resultant changes to the development of the aCC NMJ on muscle DA1 were quantified at the late wandering third instar larval stage. Reduced expression of the GluRIIA subunit occurred following heat stress experienced during 13-16 hours AEL, comparable to GluRIIA changes caused by heat stress during all of embryogenesis, but not in response to heat stress experienced before or after this 3-hour window. GluRIIB levels are not affected by heat stress during any period or all of embryogenesis. NMJ size, in contrast, is affected by heat stress experienced during earlier phases of muscle development, from 9-13 hours AEL. Data are shown with mean ± SEM, ANOVA, *p < 0.01, **p < 0.001, ***p < 0.0001, ****p < 0.00001, ‘ns’ indicates statistical non-significance. Sample sizes: for GluRIIA and GluRIIB measurements, each data point represents a unique specimen. NMJ size was measured in the same specimens for up to four muscle DA1 NMJs in abdominal segments 3-5 sampled per larva.

Physiological properties of the late larval NMJ following an embryonic 32°C perturbation.

A) Representative traces of control two electrode voltage clamp recordings across a gradient of external calcium concentrations. B) Postsynaptic current amplitude is not altered between control and embryonic heat stress-manipulated animals at any given external calcium concentration (n=5). C) Plotting postsynaptic current amplitude against its variance to calculate quantal size (q) and the number of release sites being used (N), also did not show a significant difference between control and embryonic heat stress-manipulated animals (n=5). D) To measure readily releasable pool size, a 60Hz stimulus train was applied for 1s where we could observe a slight increase in facilitation over the first 10 pulses in specimens that in their embryonic stage had experienced heat stress of 32°C (n=9) as compared to controls (n=11). E) Cumulative amplitude was plotted over time and a linear regression was back extrapolated through the last 20 pulses (same data as for E). F) The y intercept was divided by the Mini size of the same trace to calculate the readily releasable pool (RRP) size. G) NMJs in heat stress-manipulated specimens have normal EPSC and mEPSC amplitudes and quantal content. Acute application of the GluRIIA blocker PhTx reliably induced presynaptic homeostatic potentiation in controls and 32°C manipulated animals. Data in F & G are shown with mean ± SEM, ANOVA, *p < 0.01, **p < 0.001, ***p < 0.0001, ‘ns’ indicates statistical non-significance.

Larval motoneurons display reduced excitability and increased synaptic input after an embryonic 32°C heat stress experience.

A) Embryos were subjected to 32°C heat stress, then reared to the late larval stage under control conditions of 25°C. Whole cell patch clamp recordings show action potentials generated by aCC motoneurons isolated from excitatory synaptic input by mecamylamine addition in response to current injection. Inset shows linear regression centred around 32-72 pA current injections, showing significant reduction in excitability in heat stress manipulated specimens relative to controls. B) Whole cell patch clamp recordings from the aCC motoneuron in larvae: Bi) reveal an increase in the duration of spontaneous rhythmic currents (SRCs) in larvae that had previously experienced 32°C heat stress during embryonic development. Current density and frequency of SRCs remain comparable to controls. Bii) Targeted optogenetic activation (using Chronos, for 1 second) of one premotor excitatory interneuron, A27h, showed A27h synaptic input has increased as a result of a previous embryonic 32°C heat stress experience. Data are shown with mean ± SEM, unpaired t-test, *p < 0.01, ****p < 0.00001, ‘ns’ indicates statistical non-significance.

Spontaneous larval motoneuron and muscle firing properties are unchanged following embryonic heat stress.

A) Monitoring spontaneous motoneuron firing by cell-attached patch clamp recordings from the aCC motoneuron in a late third instar larval nerve cord found no significant differences between embryonically manipulated specimens and controls, regarding burst rate per second, burst duration or number of spikes per burst (n=10). B) Measuring muscle output generated by motoneuron spontaneous firing with current clamp recordings from the DA1 muscle in late third instar larvae showed that larvae are able to produce normal motoneuron firing output to the muscle, irrespective of embryonic experience (n=8 for controls, n=6 for experimentals with 32°C embryonic experience). Data are shown with mean ± SEM, unpaired t-test, ‘ns’ indicates statistical non-significance.

Embryonic heat stress manipulation causes reduced speed of both, activity wave propagation and larval crawling.

A) Functional imaging of A27h premotor interneurons using GCaMP8f in isolated CNSs from late larvae that previously had experienced transient 32°C heat stress during embryonic development. Activity wave propagation across abdominal segments is slowed down, i.e. intersegment duration has increased, notably between anterior segments, A3 to A1. B) In intact animals, locomotor network output is similarly altered: embryonic heat stress (32°C) leads to a significant reduction in larval crawling speed, as compared to controls (25°C). Data are shown with mean ± SEM, unpaired t-test, **p < 0.001, ****p < 0.00001, ‘ns’ indicates statistical non-significance. For B) only the crawling speed of each larva was sampled up to three times, once per unique 5-minute crawling interval.

Larval crawling speed is determined by the CP of central locomotor network.

A) Embryos were subjected to heat stress (32°C) during different windows of development. Crawling speed of resultant third instar larvae was affected only when the CP for the central locomotor network (17-19hrs after egg laying [AEL]) was manipulated. B) Embryonic experience of heat stress leads to larvae that retain their ability to increase crawling speed in response to acute relative increases in ambient temperature, to a higher absolute temperature (35°C) than controls (29°C). Data are shown with mean ± SEM, ANOVA, *p < 0.01, **p < 0.001, ****p < 0.00001, ‘ns’ indicates statistical non-significance. Each larva was sampled up to three times, once per unique 5-minute crawling interval.

Model of distinct critical periods for body wall muscles and the central locomotor circuitry.

Embryonic heat stress (32°C) leads to adjustments of postsynaptic receptor fields, when applied during an earlier developmental window of 13-16 hrs after egg laying. Subsequently, from 17-19 hrs after egg laying the central locomotor circuitry is sensitive to perturbations, leading to an increase in synaptic drive of the premotor network, potentially including central pattern generating circuits (CPGs). In response, motoneurons decrease their excitable properties, thus maintaining normal locomotor output to the body wall muscles. Nevertheless, changes in the premotor network lead to a reduction in activity wave progression and thus reduced larval crawling speed, also manifesting a reduction in overall network stability.