A neural network approach to the acoustic startle reflex and prepulse inhibition

Prepulse inhibition (PPI) is the normal suppression of the startle reflex when an intense stimulus is preceded by a weak non-startling prestimulus. PPI is widely used as a model for sensorimotor gating processes and has been shown to be impaired in various neuropsychiatric disorders, including schizophrenia. We have reproduced startle-like behavior and basic PPI modifications with a neural network. The network design was constrained by the attempt (1) to use as few connections as possible and (2) to relate neuroanatomical structures to the simulated network. Performance of the network was evaluated by the behavior of the simulated motor neurons in response to prepulse and pulse stimuli presented with various lead intervals and prepulse intensities. A delayed inhibitory pathway via the pedunculopontine nucleus (PPTg) to the caudal pontine reticular nucleus was found to be a necessary but insufficient requirement to reproduce basic PPI output patterns. Additional requirements included (a) a low threshold at or below the caudal pontine reticular formation, (b) signal amplification in the inhibitory pathway and (c) prolongation of activity in the inhibitory pathway. On the grounds of the most appropriate output patterns of the simulations, we propose a mechanism of sustained activation in the PPTg due to recursive connections. Relations between stimuli, behavior (motor output) and the underlying architecture are discussed. Potentially, this modeling technique can be extended to investigate the impact of drugs and higher brain regions on PPI.     

Primary and secondary neural networks of auditory prepulse inhibition: a functional magnetic resonance imaging study of sensorimotor gating of the human acoustic startle response

Feedforward inhibition deficits have been consistently demonstrated in a range of neuropsychiatric conditions using prepulse inhibition (PPI) of the acoustic startle eye-blink reflex when assessing sensorimotor gating. While PPI can be recorded in acutely decerebrated rats, behavioural, pharmacological and psychophysiological studies suggest the involvement of a complex neural network extending from brainstem nuclei to higher order cortical areas. The current functional magnetic resonance imaging study investigated the neural network underlying PPI and its association with electromyographically (EMG) recorded PPI of the acoustic startle eye-blink reflex in 16 healthy volunteers. A sparse imaging design was employed to model signal changes in blood oxygenation level-dependent (BOLD) responses to acoustic startle probes that were preceded by a prepulse at 120 ms or 480 ms stimulus onset asynchrony or without prepulse. Sensorimotor gating was EMG confirmed for the 120-ms prepulse condition, while startle responses in the 480-ms prepulse condition did not differ from startle alone. Multiple regression analysis of BOLD contrasts identified activation in pons, thalamus, caudate nuclei, left angular gyrus and bilaterally in anterior cingulate, associated with EMG-recorded sensorimotor gating. Planned contrasts confirmed increased pons activation for startle alone vs 120-ms prepulse condition, while increased anterior superior frontal gyrus activation was confirmed for the reverse contrast. Our findings are consistent with a primary pontine circuitry of sensorimotor gating that interconnects with inferior parietal, superior temporal, frontal and prefrontal cortices via thalamus and striatum. PPI processes in the prefrontal, frontal and superior temporal cortex were functionally distinct from sensorimotor gating.     

Sensory gating of the P13 midlatency auditory evoked potential and the startle response in the rat

The human P1/P50 midlatency auditory evoked potential and the startle response (SR) have been used as measures of sensory and sensorimotor gating, respectively. In the present study, both prepulse and paired stimulus paradigms were used in order to investigate the relationship between sensory gating mechanisms of the P13 potential, the putative rodent equivalent of the P1 potential, and those of the SR. In addition, these were compared to the properties of the N40 potential, another measure of sensory gating. Simultaneous recordings from the vertex (P13 potential and N40 potential) and neck musculature (SR) showed that (1) in a prepulse paradigm, increasing the intensity of the prepulse or decreasing the interstimulus interval resulted in increased inhibition of the P13 potential, N40 potential (to a lesser degree) and the SR (to a greater degree), (2) when using a low signal-to-noise ratio between the prepulse intensity and the background level, prepulse inhibition of the SR was reduced or absent while that of the P13 potential was present, (3) the amplitude of the 'prepulse evoked' P13 potential was significantly correlated with prepulse inhibition of the P13 potential, the N40 potential and the SR, (4) in a paired identical stimulus paradigm, decreasing the interstimulus interval resulted in increased habituation of the P13 potential, N40 potential (to a lesser degree) and the SR, and (5) increasing the intensity of the paired stimulation resulted in increased habituation of the P13 potential and the N40 potential (to a lesser degree), but not of the SR. These results demonstrate the presence of prepulse inhibition of the P13 potential, the N40 potential and the SR in a parallel manner, but show certain specific differences in their responses to parametric changes.     

GABAergic projection from nucleus accumbens to ventral pallidum mediates dopamine-induced sensorimotor gating deficits of acoustic startle in rats

Previous studies have demonstrated that increased mesolimbic dopamine (DA) activity disrupts sensorimotor gating as measured by prepulse inhibition (PPI) of the acoustic startle response (ASR) in rats. Other behavioral changes following mesolimbic DA activation are mediated through GABAergic efferent projections from the nucleus accumbens (NAC) to the ventral pallidum (VP). In this experiment, we examined whether PPI deficits in rats following mesolimbic DA activation are mediated through these same GABAergic substrates. PPI was significantly disrupted in rats following infusion of DA (40 micrograms) into the NAC, and this effect was reversed by infusion of a low dose (10 ng) of the GABA agonist muscimol into the VP. In a second experiment, we tested the hypothesis and the loss of PPI following intra-NAC DA infusion results from a disruption of GABAergic activity within the VP. Consistent with this hypothesis, infusion of the GABA antagonist picrotoxin (0-0.2 micrograms) into the VP caused a significant loss of PPI. These findings strongly suggest that the accumbens-ventral pallidal GABAergic circuitry is a substrate for the decrease in sensorimotor gating induced by mesolimbic DA overactivity.     

The acoustic startle reflex in ischemic stroke

The authors recorded the acoustic startle response in 32 patients with stroke, 6 patients with incomplete cervical cord lesions, and 26 controls. Increased startle occurred in about one quarter of both stroke and spinal cord injury patients. The response in biceps demonstrated the greatest deviation from normal, with less marked changes in tibialis anterior. Increased startle in spinal cord injury suggests that changes at the segmental level may contribute. Symptomatic increased startle occurred only in pontine lesions.     

Interaction between acoustic and electric sensitization of the acoustic startle response in rats.

Sensitization is the general increase of responsiveness observed after aversive stimulation. Usually footshocks are used as aversive stimuli. According to the 'Dual Process Theory' by Groves and Thompson. Psychol. Rev. 1970;77:419-450, not only additional aversive stimuli but also the response-eliciting stimuli themselves have a sensitizing effect, the degree of sensitization depending upon the stimulus intensity. We tested this suggestion in the footshock sensitization paradigm of the acoustic startle response (ASR): (1) High SPL (sound pressure level) acoustic stimuli (119 dB SPL) presented instead of footshocks also elicited strong sensitization. (2) While footshocks presented after startle stimuli with low SPL (95 dB) were able to produce a strong further sensitization of the ASR, footshocks presented after startle stimuli with high SPL (110 dB) only caused a minor sensitization of the ASR. (3) Diazepam (3 mg/kg i.p.) decreased ASR to high SPL (115 dB) stimuli. In this case footshocks elicited significant sensitization of the ASR despite intense startle stimuli. The present results support the 'Dual Process Theory'. Furthermore we could show that acoustic and footshock sensitization interact. We therefore suggest that both, acoustic and footshock sensitization, are mediated partly via the same neural circuitry.     

Appetitive odor-cue conditioning attenuates the acoustic startle response in rats

We here show that a neutral odor previously paired with a positive emotional context is an effective stimulus for attenuating the acoustic startle response (ASR) in rats. Olfactory cues can, therefore, be effectively used in the startle probe procedure for appetitive conditioning. This cue-induced reduction in ASR is not related to attentional alterations or a more general arousal by odor presentation, the conditioned olfactory cue rather elicits a pleasant emotional state during which the ASR is inhibited. This odor conditioned "pleasure" attenuation of the startle response might, therefore, provide a new effective operational measure for the hedonic aspects of reward.     

Impaired prepulse inhibition of acoustic startle in schizophrenia.

BACKGROUND: Schizophrenics show deficits in sensorimotor gating, as measured by prepulse inhibition of acoustic startle (PPI). The goal of this investigation is to further characterize PPI and habituation deficits in schizophrenia, and to examine whether differing subgroups of schizophrenics would show comparable PPI deficits. METHODS: PPI was measured in 24 male schizophrenic subjects (9 acutely decompensated inpatients and 15 stable outpatients) and in 20 age-matched normal control subjects. Schizophrenic subjects were rated for positive and negative symptoms at the time of testing. RESULTS: Schizophrenic subjects showed deficits in prepulse inhibition and habituation as compared to normal subjects. Similar latency facilitation was produced by the prepulse in both groups. Acutely decompensated inpatients and stable outpatients did not differ in percent PPI. PPI did not correlate with severity of positive or negative symptoms. CONCLUSIONS: These results suggest that schizophrenic subjects have impaired central inhibitory mechanisms as measured by PPI, and support the hypothesis that periods of relative clinical remission are not accompanied by normalization of sensorimotor gating.     

Prepulse inhibition of acoustic startle in spontaneously hypertensive rats

Prepulse inhibition is modulated by dopaminergic drugs and is disrupted in attention-deficit hyperactivity disorder, as well as mental illnesses such as schizophrenia. Spontaneously hypertensive rats (SHR) have been proposed as an animal model of attention-deficit hyperactivity disorder and show marked alterations of dopaminergic regulation of behaviour. SHR showed significantly lower startle amplitude than Wistar-Kyoto (WKY) rats and Sprague-Dawley (SD) rats, but no difference in startle habituation. Baseline percentage prepulse inhibition was higher in SHR and WKY rats than in SD rats. Treatment with amphetamine caused significant disruption of prepulse inhibition in SHR and WKY rats, but not SD rats. In contrast, treatment with apomorphine caused prepulse-dependent disruption of prepulse inhibition in SD rats only. Both MK-801 and 8-OH-DPAT treatment caused disruption of prepulse inhibition in all three rat strains. This study shows differential changes in startle level and prepulse inhibition in SHR, however these rats are not uniformly different from either WKY rats or SD rats and WKY rats differ in a number of respects from SD rats. In conclusion, these data further reveal altered dopaminergic regulation of behaviour in SHR, but also shows that caution is needed about the control strain used to compare these animals with.     

Lipopolysaccharide does not affect acoustic startle reflex in mice

Bacterial endotoxin (lipopolysaccharide; LPS) evokes in rodents an adaptive sickness behavior. It also produces changes in stress hormones secretion and activity of brain serotonergic and noradrenergic systems that have been implicated in stress responses, fear, and anxiety. Acoustic startle reflex (ASR) is regarded as a protective behavioral response that is enhanced in threatening situations or following an aversive event, and it can be modulated by physiological and emotional state of an animal. Effects of intraperitoneal injections of LPS on ASR, prepulse inhibition (PPI), locomotor activity in open field, and blood plasma corticosterone concentration were studied in lines of mice that display high (HA line) or low (LA line) swim stress-induced analgesia and also differ in emotional behaviors, including the magnitude of ASR. In both lines LPS produced robust sickness behavior, as evidenced by a decrease in locomotion and body weight, and an increase in corticosterone concentration. However, in neither line LPS injections affected responses to acoustic stimuli as assessed by the ASR and PPI magnitudes. The findings suggest that in sickness behavior induced by LPS the protective responses to salient environmental stimuli are not impaired. The significance of this finding for the concept of sickness behavior is discussed.     

Impaired prepulse inhibition of acoustic startle in obsessive-compulsive disorder.

BACKGROUND: Animal and clinical studies suggest that impaired sensorimotor gating, as assessed with the prepulse inhibition (PPI) paradigm, may result from dysfunctional frontostriatal brain circuits and from neurochemical alterations which are also implied in the pathophysiology of obsessive-compulsive disorder (OCD). However, there is only preliminary evidence about impaired PPI in OCD so far. METHODS: Acoustic PPI was measured in 30 OCD patients and 30 matched healthy controls with a paradigm using different prepulse intensities. Psychopathology assessment included ratings for obsessions, compulsions, and depression. RESULTS: PPI was reduced in OCD patients, and this deficit was most pronounced for most intense (16 dB(A)) prepulses, where mean PPI was 39.6% in unmedicated patients (n = 4), 45.8% in medicated patients, and 58.9% in controls. No group differences were observed with regard to the habituation of acoustic startle magnitude. Startle measures were generally not associated with clinical measures, although such associations may have been obscured by medication effects. CONCLUSIONS: The present study confirms deficient central inhibitory functioning in patients with OCD and supports the model of deficient frontostriatal circuits in OCD. The relationship of PPI deficits to pharmacological and behavioral treatment and to possible subtypes of OCD merits further study.     

Structural modeling of functional neural pathways mapped with 2-deoxyglucose: effects of acoustic startle habituation on the auditory system

This paper describes the first application of structural modeling to neuroscience. Structural modeling (also known as path analysis) is a method to assess the relative impact of directional links in a system and how these interrelations may change under different conditions. The objective was to demonstrate how structural modeling can be used to determine the functional interrelationships between brain structures that form the auditory system. Using structural modeling, changes in auditory system 2-DG uptake were examined during long- and short-term habituation of the acoustic startle reflex. Models were based on the anatomical connections between central auditory system structures. Using functional 2-DG data, the correlations between these structures were calculated and numerical weights were computed for each anatomical link. The analysis revealed that the lemniscal path was dominant during short-term habituation, while during long-term habituation this influence was modified through extra-lemniscal pathways. The models are discussed in the context of previous findings to demonstrate how structural modeling can not only complement, but also extract more information from 2-DG mapping experiments.     

Sensitization and habituation of the acoustic startle reflex in patients with schizophrenia

Assessments of prepulse inhibition and habituation of the acoustic startle response have proved to be valuable tools for assessing deficits of sensorimotor gating and information processing in schizophrenia patients. Recent studies, however, have reported inconsistent results regarding startle habituation deficits in schizophrenia using block-to-block analyses. Some of these inconsistencies may be due to abnormal initial sensitization effects to startle-eliciting stimuli. In a longitudinal study during the course of an acute psychotic episode, 34 medicated inpatients were examined with regard to sensitization and habituation effects in a trial-by-trial analysis and compared with 18 normal control subjects. On two examinations--10 days after admission and after psychopathological improvement 2-3 weeks later--schizophrenia patients exhibited an exaggerated magnitude increment across the first few startle-eliciting stimuli and habituation deficits that were evident when the effect of sensitization was removed from analysis. In the present study, both increased sensitization and reduced habituation appeared to be trait markers of schizophrenic psychoses. The enhanced sensitization effect--presumably due to an abnormal arousal modulation--reflects abnormal stimulus processing in schizophrenia, i.e. the diminished ability to learn the irrelevance of simple identical stimuli. In addition, the present data have important implications for designing startle studies to assess sensitization, habituation and prepulse inhibition in one session.     

Footshock-induced sensitization of the acoustic startle response in two strains of mice

It has been shown before that unconditioned footshocks can augment the acoustic startle response in rats. In the present study, male mice of two strains, C57Bl/6N and BALB/c, were compared with regard to footshock-induced sensitization of the acoustic startle response. Presentation of footshocks did not affect the acoustic startle response in C57Bl/6N mice, while in contrast, footshock-induced sensitization was apparent in the BALB/c strain. Shocked C57Bl/6N mice, but not BALB/c mice, displayed robust conditioning to the startle context when re-tested the next day. These findings indicate that mice may exhibit footshock-induced sensitization of the acoustic startle response, but that the effects of footshocks on the acoustic startle are strain- and time-dependent.     

The effect of tail pinch on the acoustic startle response in rats

A series of 3 experiments was carried out to evaluate the effect of tail pinch (TP) on the amplitude of the acoustic startle response (ASR) in rats. There is a consistent group of pharmacological findings which support the view that the amplitude of the ASR is facilitated by transmission in both dopaminergic and noradrenergic neural systems. It has recently been reported that TP increases cortical norepinephrine release and pars compacta unit activity. It might therefore be expected that TP facilitate the amplitude of the ASR. This hypothesis was tested in the first experiment. Surprisingly, it was found that TP significantly depressed startle amplitude. In the second experiment if was found that this TP-induced depression in startle amplitude was reduced by damage to the nucleus accumbens and that the amount of reduction correlated with the extent of damage. In the third experiment the generality of the effect of TP on sensorimotor reactivity was evaluated by testing its effect on footshock threshold and airpuff-elicited startle response. TP also depressed responsiveness in these tests. These results are consistent with other observations that the nucleus accumbens plays a role in the modulation of sensorimotor reactivity.     

Electromyographical differentiation between the acoustic blink and startle reflex. Implications for studies investigating startle behavior

Recent evidence suggests that electromyographic activity in the orbicularis oculi muscle occurring in response to sudden acoustic stimuli consists of two overlapping components: the blink and the startle reflex. The aim of the present study was to identify these two components in acoustically elicited eyeblink responses and to analyze their differential modulation by weak acoustic prepulses. The prevalence, latency and amplitude characteristics of double EMG peaks in pulse-alone and prepulse-pulse trials (PP) with 30 ms and 100 ms interstimulus intervals were assessed in 16 healthy volunteers. EMG responses with two peaks were registered in 42.6 % of the pulse-alone trials and in 56.2 % of the PP30 and 48.7 % of the PP100 trials, respectively. Prepulse inhibition of the amplitude was greater for the second peak (14.2 % (P2) vs. -11.5 % (P1) in PP30 trials; 62.6 % (P2) vs. 32.3 % (P1) in PP100 trials), resulting also in higher P1/P2 amplitude ratios in prepulse-pulse trials (P1/P2: 62.9 % in pulse-alone, 92.6 % in PP30 and 100.1 % in PP100 trials). In conclusion, double peaks are a common phenomenon in human studies of acoustically elicited blink responses. It is postulated that the first peak represents the auditory blink reflex, whereas the second peak corresponds to the startle reflex, which may be more susceptible to prepulse inhibition. This complexity should be taken into account in clinical studies of the modulation of the startle reflex. Received: 15 November 2001 / Accepted: 14 June 2002     

Dopamine gating of forebrain neural ensembles

Dopamine may exert different actions depending on a number of factors. A common view is that D1 receptors may be responsible for excitatory actions whereas D2 receptors are involved in inhibitory actions. However, this position cannot be reconciled with several findings indicating otherwise. The role of dopamine on forebrain neural ensembles may be better understood in the light of functional states of the system. Pyramidal cortical neurons and striatal medium spiny neurons alternate between two membrane potential states ('up' and 'down') that could shape dopamine actions. It is proposed that D1 receptors can act as state-stabilizers by sustaining up states and thereby facilitating plasticity mechanisms by providing postsynaptic depolarization and increasing NMDA function. In this way, dopamine can sustain activity in depolarized units. This action is accompanied by a decrease in cell firing (perhaps mediated by D2 receptors), which renders the cells responsive only to strong stimuli. The result would be a net increase in signal-to-noise ratio in a selected assembly of neurons.     

The acoustic startle response as an effective model for elucidating the effect of genes on the neural mechanism of behavior in mice.

One current approach in investigating the neural basis of behavior is to use mutant mice with specific genetic alterations which affect neural functions. We are convinced that this approach is only effective if a behavioral model with sufficiently known underlying neuronal mechanisms is used. We present a model system which is well-suited for the above approach. Because the neural basis is known in great detail, in the startle system behavioral results can be very well interpreted. This is demonstrated here by using footshock sensitization of the acoustic startle response (ASR) as an example. Sensitization is elicited by aversive stimuli such as electric footshocks and causes an increase in ASR amplitude. The present experiment showed that this ASR increase is not due to a drop in the startle threshold but to increased gain in the response to suprathreshold stimuli. This makes it possible to draw conclusions about the neuronal site of the startle threshold in the startle pathway and the synapse at which the gain shift during sensitization occurs. The possibility of interpreting behavioral output on a well known neural basis (as demonstrated here) makes the ASR a promising model system for investigating (neuro-) genetic influences of behavior.     

Neuronal substrates of sensory gating within the human brain.

BACKGROUND: For the human brain, habituation to irrelevant sensory input is an important function whose failure is associated with behavioral disturbances. Sensory gating can be studied by recording the brain's electrical responses to repeated clicks: the P50 potential is normally reduced to the second of two paired clicks but not in schizophrenia patients. To identify its neural correlates, we recorded electrical traces of sensory gating directly from the human hippocampus and neocortex. METHODS: Intracranial evoked potentials were recorded using hippocampal depth electrodes and subdural strip and grid electrodes in 32 epilepsy patients undergoing invasive presurgical evaluation. RESULTS: We found evidence of sensory gating only in the hippocampus, the temporo-parietal region (Brodmann's areas 22 and 2), and the prefrontal cortex (Brodmann's areas 6 and 24); however, whereas neocortical habituating responses to paired clicks were peaking around 50 msec, responses within the hippocampus proper had a latency of about 250 msec. CONCLUSIONS: Consistent with data from animal studies, our findings show that the hippocampus proper contributes to sensory gating, albeit during a time window following neocortical habituation processes. Thus, sensory gating may be a multistep process, with an early phase subserved by the temporo-parietal and prefrontal cortex and a later phase mediated by the hippocampus.