Extensor motoneurone properties are altered immediately before and during fictive locomotion in the adult decerebrate rat.

This study examined motoneurone properties during fictive locomotion in the adult rat for the first time. Fictive locomotion was induced via electrical stimulation of the mesencephalic locomotor region in decerebrate adult rats under neuromuscular blockade to compare basic and rhythmic motoneurone properties in antidromically identified extensor motoneurones during: (1) quiescence, before and after fictive locomotion; (2) the 'tonic' period immediately preceding locomotor-like activity, whereby the amplitude of peripheral flexor (peroneal) and extensor (tibial) nerves are increased but alternation has not yet occurred; and (3) locomotor-like episodes. Locomotion was identified by alternating flexor-extensor nerve activity, where the motoneurone either produced membrane oscillations consistent with a locomotor drive potential (LDP) or did not display membrane oscillation during alternating nerve activity. Cells producing LDPs were referred to as such, while those that did not were referred to as 'idle' motoneurones. LDP and idle motoneurones during locomotion had hyperpolarized spike threshold (Vth ; LDP: 3.8 mV; idle: 5.8 mV), decreased rheobase and an increased discharge rate (LDP: 64%; idle: 41%) during triangular ramp current injection even though the frequency-current slope was reduced by 70% and 55%, respectively. Modulation began in the tonic period immediately preceding locomotion, with a hyperpolarized Vth and reduced rheobase. Spike frequency adaptation did not occur in spiking LDPs or firing generated from sinusoidal current injection, but occurred during a sustained current pulse during locomotion. Input conductance showed no change. Results suggest motoneurone modulation occurs across the pool and is not restricted to motoneurones engaged in locomotion.

Plasticity of rat motoneuron rhythmic firing properties with varying levels of afferent and descending inputs.

Hindlimb motoneuron excitability was compared among exercise-trained (E), sedentary (S), and spinal cord transected (T) Sprague-Dawley rats by examining the slope of the frequency-current (F/I) relationship with standard intracellular recording techniques in rats anesthetized with ketamine-xylazine. The T group included spinal transected and spinal isolated rats; the E animals were either spontaneously active (exercise wheel) or treadmill trained; and rats in the S group were housed in pairs. An analysis of motoneuron initial [1st interspike interval (ISI)], early (mean of 1st three ISIs), and steady-state (mean of last 3 ISIs) discharge rate slopes resulting from increasing and decreasing 500-ms injected square-wave depolarizing current pulses was used to describe rhythmic motoneuron properties. The steepest slope occurred in the S group (55.3 ± 22.2 Hz/nA), followed by the T group (35.5 ± 15.3 Hz/nA), while the flattest slope was found in the E group (25.4 ± 10.9 Hz/nA). The steepest steady-state slope occurred in the S group but was found to be similar between the T and E groups. Furthermore, a spike-frequency adaptation (SFA) index revealed a slower adaptation in motoneurons of the E animals only (∼40% lower). Finally, evidence for a secondary range of firing existed more frequently in the T group (41%) compared with the S (12%) and E (31%) groups. The lower F/I slope and lower SFA index of motoneurons for E rats may be a result of an increase in Na(+) conductance at the initial segment. The results show that motoneuronal rhythmic firing behavior is plastic, depending on the volume of daily activation and on intact descending pathways.

Predicting subcellular localization of proteins using machine-learned classifiers.

MOTIVATION: Identifying the destination or localization of proteins is key to understanding their function and facilitating their purification. A number of existing computational prediction methods are based on sequence analysis. However, these methods are limited in scope, accuracy and most particularly breadth of coverage. Rather than using sequence information alone, we have explored the use of database text annotations from homologs and machine learning to substantially improve the prediction of subcellular location. RESULTS: We have constructed five machine-learning classifiers for predicting subcellular localization of proteins from animals, plants, fungi, Gram-negative bacteria and Gram-positive bacteria, which are 81% accurate for fungi and 92-94% accurate for the other four categories. These are the most accurate subcellular predictors across the widest set of organisms ever published. Our predictors are part of the Proteome Analyst web-service.