The midbrain periaqueductal grey (PAG), together with the hypothalamus, are key brain regions mediating the physiological responses to a wide range of stressors (1-3). The PAG consists of four subregions arranged as longitudinal columns (the dorsomedial, dorsolateral, lateral and ventrolateral columns). These subregions differ greatly with respect to their anatomical connections and chemical properties. The dorsolateral PAG (dlPAG) in particular is very distinctive, in that it contains a high density of neurons immunoreactive for nitric oxide synthase whereas the adjacent lateral PAG (lPAG) and dorsomedial PAG (dmPAG) do not (1). In addition, the dlPAG, but not the other subregions, receives major inputs from brain regions mediating visual, auditory and olfactory stimuli (1-3). Furthermore, neurons in the dlPAG but not the adjacent subregions are activated by acute threatening stimuli (e.g. the odour of a predator (3)). Finally, the output pathways from the these PAG subregions are also distinctly different. For example, the lPAG and dm PAG have direct descending projections to the lower brainstem, while the dlPAG have direct ascending projections to the dorsomedial hypothamus, another key brain region known to mediate stress-evoked physiological responses, that are characterised by an increase in sympathetic and respiratory activity.

All these observations suggest that the dlPAG is a key centre that integrates the physiological response to an acute threatening stimulus. In this study we tested this hypothesis by comparing the pattern of cardiovascular and respiratory responses evoked by activation of neurons in the dlPAG with those evoked from surrounding regions. Arterial pressure and heart rate, renal sympathetic nerve activity (RSNA) and phrenic nerve activity (PNA) were recorded simultaneously in urethane-anaesthetized rats. Microinjections of very small amounts of the neuroexcitatory agent D,L-homocysteic acid (DLH, 750 pmol in 15 nl) were made into sites throughout the dmPAG, dlPAG and lPAG. The results showed that significant increases (>30%) in RSNA were evoked from all three PAG subregions. In contrast, significant increases in PNA burst rate (respiratory rate) and amplitude were evoked only from a highly circumscribed subregion that corresponded to the dlPAG at an intermediate rostrocaudal level. Within this subregion, the evoked increases in RSNA and respiratory activity were highly correlated (r=0.86, P < 0.001), suggesting the possibility that a common population of “command neurons” within the dlPAG generate both the sympathetic and respiratory responses from this subregion.

Taken together with the results of previous studies (1-4), the results suggest that the dlPAG may play a unique role in generating the cardiovascular and respiratory response to an acute threat, via ascending connections to neurons in the dorsomedial hypothalamus that in turn project down to sympathetic and respiratory nuclei in the lower brainstem. Repeated activation of these pathways may lead to sustained sympathoexcitation and hypertension.

1. Bandler R, Shipley MT. Trends in Neuroscience 17, 379-389, 1994.
2. Carrive P. Behavioural Brain Research 58, 27–47, 1993.
3. Keay KA, Bandler R. Neuroscience Biobehavioual Reviews 25, 669-678, 2001.
4. Horiuchi J, McDowall LM, Dampney RAL. Journal of Physiology. 587,5149-5162, 2009.