Startle Reflex

I. Acoustic

A. Neuroanatomy

1. Primary Reflex

a) Overall circuit

• primary acoustic startle pathway consists of only 3 synapses onto (a) cochlear root neurons; (b) neurons in the nucleus reticularis pontis caudalis, and (c) motoneurons in the facial motor nucleus (pinna reflex) or spinal cord (whole body startle). (2)
• "electrolytic lesions of the area around the ventral nucleus of the lateral lemniscus may have eliminated startle by destroying the axon collaterals which project heavily to the NRPC as well as via retrograde degeneration of cochlear root neurons caused by severing axons passing through this area en route to more rostral areas." (2; p. 529).

Ear---Cochlear Root Neurons---Nucleus Reticularis Pontis Caudalis---Spinal Cord---Muscles

b) Sensory circuits (from cochlear nuclei to the reticular formation)

c) Motor circuits (in the reticular formation and spinal cord)

• Acoustic startle is not blocked by lesions of the midbrain reticular formation, but is completely blocked by lesions of the caudal pontine or medullary reticular formation; thus, it appears that midbrain reticular formation is not an essential element of the primary acoustic startle ciruit. (1)
• The most critical sites in the pontine reticular formation for excitotoxic lesions in rats are in the ventral 1/3 of the RPC, about 1-1.5 mm lateral to the midline, near the fibers of the VI nerve. The largest giant cells in the entire reticular formation of newborn rats or cats are in the RPC at this level. Very low thresholds for electrical stimulation of startle are found in these RPC sites. In freely moving cats, acoustically driven units are found widely in the reticular formation; the units activated at latencies of 3-7 ms (i.e., those units possibly involved in the primary acoustic startle response) are concentrated in RPC. The majority of the RPC cells that could mediate the short-latency acoustic startle response have reticulospinal axons that conduct long distances at extraordinarily high velocities. Since very large cells would be required to support the large volume of these large-caliber, long axons, the largest cells in RPC are the most likely candidates to mediate the acoustic startle response in cats. (1)
• RPC giant neurons have terminals along the length of the ipsilateral spinal cord and in several brain stem nuclei including facial nucleus, abducens nucleus, nucleus parvocellularis, and nucleus gigantocellularis of the medullary reticular formation. These brain stem motor nuclei may mediate startle response components in head and neck muscles. (1)
• Giant RPC neurons in rats are usually excited by acoustic stimuli, with a mean spike latency of 5.2 ms. Koch et al. attempted to selectively lesion RPC giant neurons with the excitotoxin, quinolinate. The ability of lesions to kill all RPC cells correlated with the loss of acoustic startle (r=+.58), but the correlation with the loss of RPC giant cells was slightly higher (r=+.73). A maximum of about 60 giant RPC neurons on each side of the rat brain is likely to be responsible for most of the startle reflex. (1)
• In summary, activation of the startle response in rats appears to occur primarily in 20-60 ventrocaudal RPC giant neurons that act like 'command neurons." These RPC giant neurons, in turn, distribute the activation to many hundreds of motoneurons the length of the spinal cord and brain stem. (1)

Questions to Answer:

1) What are the physiological and behavioral functions of the pontine reticular formation, as compared with the midbrain reticular formation?

2) Does the primary pinna reflex pathway include the RPC?

References:

1) Yeomans, J. S., & Frankland, P. W. (1996). The acoustic startle reflex: Neurons and connections. Brain Reseach News, 21, 301-314.

2) Davis, M. (1996). Differential roles of the amygdala and bed nucleus of the stria terminalis in conditioned fear and startle enhanced by corticotropin-releasing hormone. In T. Ono, B. L. McNaughton, S. Molotchnikoff, E. T. Rolls, & H. Nishijo (Eds.), Perception, memory, and emotion: Frontiers in neuroscience (pp. 525-548). Oxford: Elsevier. (LL #4987)

B. Modulation Effects

1. Fear-Related

a) Explicit Cue Potentiation

- refers to classic potentiation during presentation of an aversive CS (e.g., light previously paired with shock).

- lesions of the amygdala, but not the the bed nucleus of the stria terminalis (BNST), block explicit cue potentiation.

Questions to Answer:

1)

References:

1) Davis, M. (1996). Differential roles of the amygdala and bed nucleus of the stria terminalis in conditioned fear and startle enhanced by corticotropin-releasing hormone. In T. Ono, B. L. McNaughton, S. Molotchnikoff, E. T. Rolls, & H. Nishijo (Eds.), Perception, memory, and emotion: Frontiers in neuroscience (pp. 525-548). Oxford: Elsevier. (LL #4987)

b) Contextual Fear Potentiation

- refers to a gradual increase in baseline startle amplitude over aversive conditioning sessions, assessed prior to shock-cue pairings (i.e., before any shocks or lights presented). (1)

- lesions of the amygdala block both explicit cue and contextual conditioning, whereas lesions of the bed nucleus of the stria terminalis (BNST) block contextual conditioning, but not explicit cue conditioning. (1)

- humans - Davis (1) believes that the general increase in baseline startle that occurs when shock electrodes are attached to human subjects (alerting them to the threat of shock) is analogous to contextual conditioning in rats: "In humans, the conditioned aspect of this measure of anxiety is based not on an actual conditioning experiment in the laboratory, but rather general knowledge that shock is dangerous, conditioned over the course of a lifetime."

- PTSD patients - Davis and colleagues (1) have repeatedly found normal baseline startle in Vietnam veterans in the absence of experimental stress, but increased startle throughout experiments that involved stressful procedures. "These results suggest that Vietnam veterans with PTSD display normal fear to explicit stimuli, but abnormal contextual conditioning."

- Davis (1) - model of the possible relationships between the amygdala, BNST, and CRH in fear and anxiety: "Highly processed explicit cue information (lights, tones, touch, smells) may activate the amygdala which in turn activates hypothalamic and brainstem target areas involved in specific signs of fear. Somewhat less explicit information, such as contextual information, may activate the BNST which in turn activates hypothalamic and brainstem target areas involves in specific signs of fear. However, because the nature of this information may be less specific than that produced by an explicit cue, activation of the BNST may be more akin to anxiety than fear. Moreover, the long-lasting anxiogenic effects of CRH may be mediated via activation of the BNST, which in turn would activate hypothalamic and brainstem target areas involved."

Questions to Answer:

1) What is relationship between hippocampus and BNST? Is BNST involved in hypervigilance/generalized attention? Does BNST activation depend upon continued cognitive processing (rumination), which might be attenuated in brain-injured patients or as a function of alcohol or other drugs?

References:

1) Davis, M. (1996). Differential roles of the amygdala and bed nucleus of the stria terminalis in conditioned fear and startle enhanced by corticotropin-releasing hormone. In T. Ono, B. L. McNaughton, S. Molotchnikoff, E. T. Rolls, & H. Nishijo (Eds.), Perception, memory, and emotion: Frontiers in neuroscience (pp. 525-548). Oxford: Elsevier. (LL #4987)

c) CRH-Enhanced Startle

- intraventricular (ICV) infusion of CRH increases the acoustic startle reflex. (1)

- effect is blocked by the following drugs: (i) benzodiazepine chlordiazepoxide (also blocks fear-potentiated startle?); (ii) CRH antagonist a-helical CRH9-41 (does not block fear-potentiated startle). (1)

- administration of ganglionic blockers or prior adrenalectomy did not block the excitatory effect of CRH on startle, indicating a central site of action (1).

- either electrolytic or chemical lesions of either the ventral hippocampus or the bed nucleus of the stria terminalis (BNST) block the excitatory effect of CRH on startle; hence, we now believe that lesions of the medial septal area block CRH-enhanced startle by destruction of the fornix, which carries fibers projecting from the ventral hippocampus to the BNST. (1)

- lesions that block contextual conditioning also block CRH-enhanced startle (e.g., lesions of the hippocampus or BNST), whereas lesions that block explicit cue conditioning do not (e.g., lesions of the amygdala). This suggests that CRH-enhanced startle and fear-potentiated startle are additive and hence may be independent. Indeed, CRH given to rats trained in a fear-potentiated startle paradigm causes a significant increase in baseline startle, but has no effect on the magnitude of fear-potentiated startle. (1).

Questions to Answer:

1)

References:

1) Davis, M. (1996). Differential roles of the amygdala and bed nucleus of the stria terminalis in conditioned fear and startle enhanced by corticotropin-releasing hormone. In T. Ono, B. L. McNaughton, S. Molotchnikoff, E. T. Rolls, & H. Nishijo (Eds.), Perception, memory, and emotion: Frontiers in neuroscience (pp. 525-548). Oxford: Elsevier. (LL #4987)

II. Tactile

II. Vestibular

II. Visual

Return to LABWEB HomePage

created by C. Patrick

updated on 2-6-97 by J. Curtin