Differential activity-dependent development of corticospinal control of movement and final limb position during visually guided locomotion

Although we understand that activity- and use-dependent processes are important in determining corticospinal axon terminal development in the spinal cord, little is known about the role of these processes in development of skilled control of limb movements. In the present study we determined the effects of unilateral motor cortex activity blockade produced by muscimol infusion during the corticospinal axon terminal refinement period, between postnatal weeks 5-7, on visually guided locomotion. We examined stepping and forepaw placement on the rungs of a horizontal ladder and gait modifications as animals stepped over obstacles during treadmill walking. When cats traversed the horizontal ladder, the limb contralateral to inactivation was placed significantly farther forward on the rungs than the ipsilateral limb, indicating defective endpoint control. Similarly, when animals stepped over obstacles on a treadmill, the contralateral limb was placed farther in front of the obstacle, but only when it was the first (i.e., leading) limb to step over the obstacle, not when it was the second (i.e., trailing) limb. This is also indicative of an endpoint control deficit. In contrast, neither during ladder walking, nor when stepping over obstacles on the treadmill, was there any consistent evidence for a major impairment in limb trajectory. These results point to distinct and possibility independent corticospinal mechanisms for movement trajectory control and endpoint control. Although corticospinal activity during early postnatal development is needed to refine circuits for accurate endpoint control, this activity-dependent refinement is not needed for movement trajectory control.

Contributions of the reticulospinal system to the postural adjustments occurring during voluntary gait modifications

To test the hypothesis that reticulospinal neurons (RSNs) are involved in the formation of the dynamic postural adjustments that accompany visually triggered, voluntary modifications of limb trajectory during locomotion, we recorded the activity of 400 cells (183 RSNs; 217 unidentified reticular cells) in the pontomedullary reticular formation (PMRF) during a locomotor task in which intact cats were required to step over an obstacle attached to a moving treadmill belt. Approximately one half of the RSNs (97/183, 53%) showed significant changes in cell activity as the cat stepped over the obstacle; most of these cells exhibited either single (26/97, 26.8%) or multiple (63/97, 65.0%) increases of activity. There was a range of discharge patterns that varied in the number, timing, and sequencing of the bursts of modified activity, although individual bursts in different cells tended to occur at similar phases of the gait cycle. Most modified cells, regardless of the number of bursts of increased discharge, or of the discharge activity of the cell during unobstructed, control, locomotion, discharged during the passage of the lead forelimb over the obstacle. Thus, 86.9% of the modified cells increased their discharge when the forelimb ipsilateral to the recording site was the first to pass over the obstacle, and 72.2% when the contralateral limb was the first. Approximately one quarter of the RSNs increased their discharge during the passage of each of the four limbs over the obstacle in both the lead (27.1%) and trail (27.9%) conditions. In general, in any one cell, the number and relative sequencing of the subsequent bursts (with respect to the lead forelimb) was maintained during both lead and trail conditions. Patterns of activity observed in unidentified cells were very similar to the RSN activity despite the diverse population of cells this unidentified group may represent. We suggest that the increased discharge that we observed in these reticular neurons reflects the integration of afferent activity from several sources, including the motor cortex, and that this increased discharge signals the timing and the relative magnitude of the postural patterns that accompany the voluntary gait modification. However, based on the characteristics of the patterns of neuronal activity in these cells, we further suggest that while individual RSNs probably contribute to the selection of different patterns of postural activity, the ultimate expression of the postural response may be determined by the excitability of the locomotor circuits within the spinal cord.

Vestibulospinal and reticulospinal neuronal activity during locomotion in the intact cat. II. Walking on an inclined plane

The experiments described in this report were designed to determine the contribution of vestibulospinal neurons (VSNs) in Deiters' nucleus and of reticulospinal neurons (RSNs) in the medullary reticular formation to the modifications of the walking pattern that are associated with locomotion on an inclined plane. Neuronal discharge patterns were recorded from 44 VSNs and 63 RSNs in cats trained to walk on a treadmill whose orientation was varied from +20 degrees (uphill) to -10 degrees (downhill), referred to as pitch tilt, and from 20 degrees roll tilt left to 20 degrees roll tilt right. During uphill locomotion, a majority of VSNs (25/44) and rhythmically active RSNs (24/39) showed an increase in peak discharge frequency, above that observed during locomotion on a level surface. VSNs, unlike some of the RSNs, exhibited no major deviations from the overall pattern of the activity recorded during level walking. The relative increase in discharge frequency of the RSNs (on average, 31.8%) was slightly more than twice that observed in the VSNs (on average, 14.4%), although the average absolute change in discharge frequency was similar (18.2 Hz in VSNs and 21.6 Hz in RSNs). Changes in discharge frequency during roll tilt were generally more modest and were more variable, than those observed during uphill locomotion as were the relative changes in the different limb muscle electromyograms that we recorded. In general, discharge frequency in VSNs was more frequently increased when the treadmill was rolled to the right (ear down contralateral to the recording site) than when it was rolled to the left. Most VSNs that showed significant linear relationships with treadmill orientation in the roll plane increased their activity during right roll and decreased activity during left roll. Discharge activity in phasically modulated RSNs was also modified by roll tilt of the treadmill. Modulation of activity in RSNs that discharged twice in each step cycle was frequently reciprocal in that one burst of activity would increase during left roll and the other during right roll. The overall results indicate that each system contributes to the changes in postural tone that are required to adapt the gait for modification on an inclined surface. The characteristics of the discharge activity of the VSNs suggest a role primarily in the overall control of the level of electromyographic activity, while the characteristics of the RSNs suggest an additional role in determining the relative level of different muscles, particularly when the pattern is asymmetric.

Vestibulospinal and reticulospinal neuronal activity during locomotion in the intact cat. I. Walking on a level surface

To examine the function of descending brain stem pathways in the control of locomotion, we have characterized the discharge patterns of identified vestibulo- and reticulospinal neurons (VSNs and RSNs, respectively) recorded from the lateral vestibular nucleus (LVN) and the medullary reticular formation (MRF), during treadmill walking. Data during locomotion were obtained for 44 VSNs and for 63 RSNs. The discharge frequency of most VSNs (42/44) was phasically modulated in phase with the locomotor rhythm and the averaged peak discharge frequency ranged from 41 to 165 Hz (mean = 92.8 Hz). We identified three classes of VSNs based on their discharge pattern. Type A, or double peak, VSNs (20/44 neurons, 46%) showed two peaks and two troughs of activity in each step cycle. One of the peaks was time-locked to the activity of extensor muscles in the ipsilateral hindlimb while the other occurred anti-phase to this period of activity. Type B, or single pause, neurons (13/44 neurons, 30%) were characterized by a tonic or irregular discharge that was interrupted by a single pronounced and brief period of decreased activity that occurred just before the onset of swing in the ipsilateral hindlimb; some type B VSNs also exhibited a brief pulse of activity just preceding this decrease. Type C, or single peak, neurons (9/44 neurons, 23%) exhibited a single period of increased activity that, in most cells, was time-locked to the burst of activity of either extensor or flexor muscles of a single limb. The population of RSNs that we recorded included neurons that showed phasic activity related to the activity of flexor or extensor muscles [electromyographically (EMG) related, 26/63, 41%], those that were phasically active but whose activity was not time-locked to the activity of any of the recorded muscles (13/63, 21%) and those that were completely unrelated to locomotion (24/63, 38%). Most of the EMG-related RSNs showed one (15/26) or two (11/26) clear phasic bursts of activity that were temporally related to either flexor or extensor muscles. The discharge pattern of double-burst RSNs covaried with ipsilateral and contralateral flexor muscles. Peak averaged discharge activity in these EMG-related RSNs ranged from 4 to 98 Hz (mean = 35.2 Hz). We discuss the possibility that most VSNs regulate the overall activity of extensor muscles in the four limbs while RSNs provide a more specific signal that has the flexibility to modulate the activity of groups of flexor and extensor muscles, in either a single or in multiple limbs.

Calcium channel agonists and dystonia in the mouse

Systemic administration of the L-type calcium channel agonists +/-Bay K 8644 or FPL 64176 causes a characteristic pattern of motor dysfunction in normal C57BL/6J mice that resembles generalized dystonia. There is no associated change in the electroencephalogram, confirming that the motor disorder does not reflect epileptic seizures. However, the electromyogram reveals an increase in baseline motor unit activity with prolonged phasic discharges consistent with dystonia. The duration and severity of dystonia is dependent on the dose administered and the age of the animal at testing. The effects are transient, with the return of normal motor behavior 1-4 hours after treatment. Similar effects can be provoked by intracerebral administration of small amounts of the drugs, indicating a centrally mediated response. Dystonia can be attenuated by co-administration of dihydropyridine L-type calcium channel antagonists (nifedipine, nimodipine, and nitrendipine) but not by non-dihydropyridine antagonists (diltiazem, verapamil, and flunarizine). These results implicate abnormal function of L-type calcium channels in the expression of dystonia in this model.

Calcium channel activation and self-biting in mice

The L type calcium channel agonist (+/-)Bay K 8644 has been reported to cause characteristic motor abnormalities in adult mice. The current study shows that administration of this drug can also cause the unusual phenomenon of self-injurious biting, particularly when given to young mice. Self-biting is provoked by injecting small quantities of (+/-)Bay K 8644 directly into the lateral ventricle of the brain, suggesting a central effect of the drug. Similar behaviors can be provoked by administration of another L type calcium channel agonist, FPL 64176. The self-biting provoked by (+/-)Bay K 8644 can be inhibited by pretreating the mice with dihydropyridine L type calcium channel antagonists such as nifedipine, nimodipine, or nitrendipine. However, self-biting is not inhibited by nondihydropyridine antagonists including diltiazem, flunarizine, or verapamil. The known actions of (+/-)Bay K 8644 as an L type calcium channel agonist, the reproduction of similar behavior with another L type calcium channel agonist, and the protection afforded by certain L type calcium channel antagonists implicate calcium channels in the mediation of the self-biting behavior. This phenomenon provides a model for studying the neurobiology of this unusual behavior.

Effects of red nucleus microstimulation on the locomotor pattern and timing in the intact cat: a comparison with the motor cortex

Effects of red nucleus microstimulation on the locomotor pattern and timing in the intact cat: a comparison with the motor cortex. To determine the extent to which the rubrospinal tract is capable of modifying locomotion in the intact cat, we applied microstimulation (cathodal current, 330 Hz; pulse duration 0.2 ms; maximal current, 25 microA) to the red nucleus during locomotion. The stimuli were applied either as short trains (33 ms) of impulses to determine the capacity of the rubrospinal tract to modify the level of electromyographic (EMG) activity in different flexors and extensors at different phases of the step cycle or as long trains (200 ms) of pulses to determine the effect of the red nucleus on cycle timing. Stimuli were also applied with the cat at rest (33-ms train). This latter stimulation evoked short-latency (average = 11.8-19.0 ms) facilitatory responses in all of the physiological flexor muscles of the forelimb that were recorded; facilitatory responses were also common in the elbow extensor, lateral head of triceps but were rare in the physiological wrist and digit extensor, palmaris longus. Responses were still evoked in most muscles when the current was decreased to near threshold (3-10 microA). Stimulation during locomotion with the short trains of stimuli evoked shorter-latency (average = 6.0-12.5 ms) facilitatory responses in flexor muscles during the swing phase of locomotion and, except in the case of the extensor digitorum communis, evoked substantially smaller responses in stance. The same stimuli also evoked facilitatory responses in the extensor muscles during swing and produced more complex effects involving both facilitation and suppression in stance. Increasing the duration of the train to 200 ms modified the amplitude and duration of the EMG activity of both flexors and extensors but had little significant effect on the cycle duration. In contrast, whereas stimulation of the motor cortex with short trains of stimuli during locomotion had very similar effects to that of the red nucleus, increasing the train duration to 200 ms frequently produced a marked reset of the step cycle by curtailing stance and initiating a new period of swing. The results suggest that whereas both the motor cortex and the red nucleus have access to the interneuronal circuits responsible for controlling the structure of the EMG activity in the step cycle, only the motor cortex has access to the circuits responsible for controlling cycle timing.

Kinetic and energetic patterns for hindlimb obstacle avoidance during cat locomotion

The safe control of walking over different terrains requires appropriate adaptations in the dynamic and kinematic limb patterns. To date, the study of locomotor dynamics in the cat has been confined to level, unobstructed walking. The present study extends the work of Lavoie et al. by applying linked segment analyses to estimate muscle contributions to torque and mechanical power at the hindlimb joints of two female cats during both unobstructed walking and obstacle avoidance. Data during obstacle avoidance were analyzed both when the hindlimb led in clearance and was farthest from the obstacle, and when it trailed in clearance and was closest or near to the obstacle. It was found that, in both the Far and Near obstructed conditions, the cats cleared the obstacles primarily by increasing the knee flexor torque already used during unobstructed gait. Contributions from the hip and ankle muscle groups were more variable. There was more emphasis on the hip extensors in mid to late stance, and the hip flexors generated a small amount of energy at paw-lift in the Far condition. In the Near condition, the hip extensors were employed to control hip flexion. We suggest that hip flexor generation power in mid-swing contributes to the clearance of the upcoming obstacle in the Far condition while, in the Near condition, hip flexion advances the already extended limb ahead of the obstacle. The ankle was actively dorsiflexed in the Near condition but was maintained in extension in the Far condition. The emphasis on active knee flexor control by the cat to avoid obstacles, as well as the dependence of ankle control on obstacle proximity, is similar to strategies seen for humans. However, the knee flexor strategy is innate to the cat's normal level walking control, whereas in humans active knee flexion at toe-off requires a reorganization from level, non-obstructed gait.

Improved control of gas humidity in neonatal intensive care ventilators

Difficulties experienced by nursing staff in controlling humidity in neonatal intensive care ventilators are shown to originate in the use of a single control, namely a heater, to condition two parameters, namely temperature and humidity, at the inlet to the inspiratory tube. It is shown that better control can be achieved by separating the gas flow through the humidifier into two streams, only one of which is humidified. It is proposed that the temperature along most of the inspiratory tube should be maintained about 3-5 degrees C higher than the target delivery temperature. This eliminates the risk of condensation in the tube, a second commonly occurring problem in existing systems. It is shown why the control of ventilator systems should be based on absolute humidity, not relative humidity, even though the latter may be preferred for monitoring purposes.

Corticoreticular pathways in the cat. II. Discharge activity of neurons in area 4 during voluntary gait modifications

We propose that the descending command from area 4 that is responsible, in part, for the change in limb trajectory required to step over an obstacle in one's path also plays a role in triggering the anticipatory postural modifications that accompany this movement. To test this hypothesis, we recorded the discharge characteristics of identified classes of corticofugal neurons in area 4 of the cat. Neurons were identified either as: pryamidal tract neurons (PTNs) if their axon projected to the caudal pyramidal tract (PT) but not to the pontomedullary reticular formation (PMRF); as corticoreticular neurons (CRNs) if their axon projected to the PMRF but not to the PT; and as PTN/CRNs if their axon projected to both structures. Altogether, the discharge properties of 212 corticofugal neurons (109 PTNs, 66 PTN/CRNs, and 37 CRNs) within area 4 were recorded during voluntary gait modifications. Neurons in all three classes showed increases in their discharge frequency during locomotion and included groups that increased their discharge either during the swing phase of the modified step, during the subsequent stance phase, or in the stance phase of the cycle preceding the step over the obstacle. A slightly higher percentage of CRNs (39%) discharged in the stance phase prior to the gait modification than did the PTNs or PTN/CRNs (20% and 17% respectively). In 37 electrode penetrations, we were able to record clusters of 3 or more neurons within 500 micro(m) of each other. In most cases, PTN/CRNs recorded in close proximity to PTNs had similar receptive fields and discharged in a similar, but not identical, manner during the gait modifications. Compared with adjacent PTNs, CRNs normally showed a more variable pattern of activity and frequently discharged earlier in the step cycle than did the PTNs or PTN/CRNs. We interpret the results as providing support for the original hypothesis. We suggest that the collateral branches to the PMRF from corticofugal neurons with axons that continue at least as far as the caudal PT provide a signal that could be used to trigger dynamic postural responses that are appropriately organized and scaled for the movements that are being undertaken. We suggest that the more variable and earlier discharge activity observed in CRNs might be used to modify the postural support on which the movements and the dynamic postural adjustments are superimposed.

Computer-controlled endoscopic performance assessment system

We have devised an advanced computer-controlled system (ADEPT) for the objective evaluation of endoscopic task performance. The system's hardware consists of a dual gimbal mechanism that accepts a variety of 5.0-mm standard endoscopic instruments for manipulation in a precisely mapped and enclosed work space. The target object consists of a sprung base plate incorporating various tasks. It is covered by a sprung perforated transparent top plate that has to be moved and held in the correct position by the operator to gain access to the various tasks. Standard video endoscope equipment provides the visual interface between the operator and the target-instrument field. Different target modules can be used, and the level of task difficulty can be adjusted by varying the manipulation, elevation, and azimuth angles. The system's software is designed to (a) prompt the surgeon with the information necessary to perform the task, (b) collect and collate data on performance during execution of specified tasks, and (c) save the data for future analysis. The system was alpha and beta tested to ensure that all functions operated correctly.

Corticoreticular pathways in the cat. I. Projection patterns and collaterization

This paper summarizes and compares the projection patterns and the receptive fields of cortical neurons in areas 4 and 6 that project to the pontomedullary reticular formation (PMRF). A total of 326 neurons were recorded in area 4 and 129 in area 6 in four awake, unrestrained cats that were chronically implanted with arrays of electrodes in the PMRF and the pyramidal tract (PT). In area 4, 47% of the neurons projected to the caudal PT but not to the PMRF (PTNs); 19% were activated only from the PMRF [corticoreticular neurons (CRNs)], whereas 27% were activated from both the PT and the PMRF (PTN/CRNs). More PTN/CRNs conducted at velocities >20 m/s (82%) than did CRNs (23%). In area 6, only 19% of the neurons were identified as PTNs, 12% were PTN/CRNs and 31% were CRNs; a further 38% could not be activated from either structure. Collateral branches within the PMRF conducted at maximum velocities of 20 m/s (average = 6.5 m/s). No significant differences in the conduction velocities of the collateral branches were found either between fast and slow PTNs or between area 4 and area 6 neurons. A large proportion of neurons in area 4 (85/173, 49%) were activated by passive manipulation of the more distal, contralateral forelimb, with approximately equal numbers being classed as PTNs, PTN/CRNs and CRNs. Most neurons in area 6 for which a receptive field could be found were excited by lightly touching or tapping the face and neck; a receptive field could not be determined for 39% of the area 6 neurons compared with only 5% of those in area 4. Finally, there was evidence that neurons in quite widespread areas of the pericruciate cortex, including both areas 4 and 6 projected onto similar, restricted regions of the PMRF. The fact that the cortical projection from area 4 to the PMRF includes a high percentage of fast PTNs with a receptive field on the distal forelimb is consistent with the view that this projection may serve to integrate movement and the dynamic postural adjustments that accompany them. The fact that the cortical projection from area 6 to the PMRF is primarily from slow PTNs with receptive fields on the face, neck and back is consistent with a role for this cortical area in adjusting the general posture of the animal on which movements are superimposed.

Total radiated power, infrared output, and heat generation by cold light sources at the distal end of endoscopes and fiber optic bundle of light cables

BACKGROUND

Skin burns and ignition of drapes have been reported with the use of cold light sources. The aim of the study was to document the temperature generated by cold light sources and to correlate this with the total radiated power and infrared output.

METHODS

The temperature, total radiated power, and infrared output were measured as a function of time at the end of the endoscope (which is inserted into the operative field) and the end of the fiber optic bundle of the light cable (which connects the cable to the light port of the endoscope) using halogen and xenon light sources.

RESULTS

The highest temperature recorded at the end of the endoscope was 95 degrees C. The temperature measured at the optical fiber location of the endoscope was higher than at its lens surface (p < 0.0001). At the end of the fiber optic bundle of light cables, the temperature reached 225 degrees C within 15 s. The temperature recorded at the optical fiber location of all endoscopes and light cables studied rose significantly over a period of 10 min to reach its maximum (p <0.0001) and then leveled off for the duration of the study (30 min). The infrared output accounted only for 10% of the total radiated power.

CONCLUSIONS

High temperatures are reached by 10 min at the end of fiber optic bundle of light cables and endoscopes with both halogen and xenon light sources. This heat generation is largely due to the radiated power in the visible light spectrum.

Organization of the projections from the pericruciate cortex to the pontomedullary brainstem of the cat: a study using the anterograde tracer Phaseolus vulgaris-leucoagglutinin

The anterograde tracer Phaseolus vulgaris-leucoagglutinin (PHA-L) was used to study the distribution and density of the projections that originate from four identified subdivisions of the pericruciate cortex (namely, the forelimb and hind limb representations of area 4, area 6a beta, and area 6a gamma) and that terminate in the pontomedullary brainstem in the cat. Injections of PHA-L in all areas of the pericruciate cortex labelled numerous fibers and their terminal swellings in the brainstem. The major target regions of all four cortical areas were the pontine nuclei and the pontomedullary reticular formation (PMRF). Injections into both the forelimb and hind limb representations of area 4 and into area 6a beta resulted in a dense pattern of terminal labelling in restricted regions of the medial and lateral parts of the ipsilateral pontine nuclei. The labelling following the area 6a beta injection was spatially distinct from that seen following the area 4 injections. Injections into the forelimb representation of area 4 as well as into area 6a beta and 6a gamma resulted in the labelling of numerous terminal swellings bilaterally in the PMRF; in contrast, there were few labelled terminal swellings in the PMRF following injections into the hind limb representation of area 4. Terminal swellings on individual corticoreticular fibers were far less densely aggregated than those in the pontine nuclei. The dense pattern of innervation to restricted regions of the pontine nuclei supports previous suggestions that the corticopontine projections retain a high degree of topographical specificity that could be used in the control of discrete voluntary movements. In contrast, the more diffuse pattern of the projections to the PMRF may facilitate the selection and activation of the complex postural patterns that accompany voluntary movement.

Organization of the projections from the pericruciate cortex to the pontomedullary reticular formation of the cat: a quantitative retrograde tracing study

Dextran-amines were used as retrograde tracers to investigate the organization of cortical projections to different cytoarchitectonic regions of the pontomedullary reticular formation of the cat. Injections into the nucleus reticularis pontis oralis resulted in labelling of neurones in the proreus cortex and area 6a beta of the premotor cortex, with little labelling in the motor cortex (area 4). This labelling was predominantly ipsilateral to the injection site. In contrast, injections into the nucleus reticularis pontis caudalis (NRPc), nucleus reticularis gigantocellularis (NRGc), and nucleus reticularis magnocellularis (NRMc) resulted in bilateral labelling--primarily in areas 6a beta, 6a gamma, and in the rostromedial region of area 4--with little labelling in the proreus cortex. In general, the cortical projections to the caudal NRGc and the NRMc were larger than those to the NRPc. More than 25% of the total projections to each of the latter three reticular regions arose from the medial part of area 4. Labelling in the hindlimb regions of area 4 was largest following the NRMc injections and smallest after injections in the NRPc. The projections to the NRPc originated from more medial parts of areas 4 and 6 than did the projections to the caudal region of the NRGc. These results suggest that areas 4 and 6 may be able to differentially activate different regions of the pontomedullary reticular formation depending on the movement that is made and perhaps also on the context of that movement.

Psychomotor skills for endoscopic manipulations: differing abilities between right and left-handed individuals

OBJECTIVE

The objective of this study was to compare the psychomotor aptitudes relevant to endoscopic manipulations between right-handed and left-handed subjects.

SUMMARY BACKGROUND DATA

There has been little research on the psychomotor performance in relation to minimal access surgery and there are no psychomotor tests to evaluate aspects of psychomotor abilities relevant to endoscopic manipulations.

METHODS

A microprocessor-controlled psychomotor tester was developed for objective evaluation of endoscopic performance. The task involved negotiating ten target holes with a probe under videoscopic imaging. Subjects consisted of two groups of 10 medical students: right- and left-handed. After a prestudy familiarization session, each subject performed two test runs with one hand, followed by two runs with the other hand. These test runs were repeated 1 week later. The outcome measures were the total execution time, force on backplate, angular deviations, error rate, and first-time accuracy.

RESULTS

A significant difference in the error rate and first time accuracy was observed between subjects (p < 0.001 and p < 0.001, respectively) and between the dominant and nondominant hands (p < 0.001 and p < 0.025, respectively), with no significant change with practice. Right-handed subjects performed better with either hand in terms of error rate (p < 0.001) and first time accuracy (p < 0.001). Practice improved the execution time (p < 0.001) and the degree of angular deviations (p < 0.02).

CONCLUSIONS

Right-handed subjects perform less errors and exhibit better first time accuracy. The parameters that improve with practice reflect the positive effect of training, whereas others, such as errors rate and first time accuracy which do not, reflect innate abilities.