. 40
( 51 .)


studies using Brattleboro rats (Ixart et al., 1982), CRH knockout mice (Muglia et al.,
1997), and CRH immunoneutralization (Ixart et al., 1985) showing that diurnal
corticosterone release in intact animals is driven almost exclusively by CRH release.
Given that corticosterone alone cannot account for the daily variations of CRH
and AVP hnRNA, a major component in the integrative process that controls CRH
221 Glucocorticoids and the ups and downs of neuropeptide gene expression

and AVP gene expression in CRH neurons must be the large group of afferent sets
encoding the extero- and interosensory information. Considering the fact that
changes in corticosterone secretion are not suf¬cient to drive these nocturnal tran-
scriptional episodes, it would seem likely that at least one of these afferent sets
plays a signi¬cant role in activating gene transcription during the dark phase.
Currently, the detailed architecture of the afferent systems responsible for shaping
CRH neuroendocrine function across the day in the absence of stress is poorly
understood. However, the circadian timing system controlled by the SCH should
be considered as one potential controller of CRH and AVP gene activation in these
circumstances. The SCH provides the principal timing signal for daily surges of
plasma ACTH and corticosterone (Moore and Eichler, 1972; Szafarczyk et al., 1979;
Cascio et al., 1987; Buijs et al., 1993a; 1998).
Some SCH efferents clearly innervate the PVHmp, but these are more sparse
than those that innervate other nearby targets (Vrang et al. 1995a, b; Watts et al.,
1987; Leak and Moore, 2001) particularly the DMH, which heavily innervates the
PVHmp and is heavily implicated in in¬‚uencing circadian corticosterone output
(Watts et al., 1987; Buijs et al., 1993b; 1998; Vrang et al., 1995a, b; Kalsbeek et al.,
1996; Thompson et al., 1996; Leak and Moore, 2001; Chou et al., 2003). The exact
nature of the afferent set controlled by the SCH remains to be established.

Like the daily variations associated with feeding and the responses to negative energy
balance, the physiological consequences of the HPA motor response to stress are in
many respects more anticipatory than reactive events. Although ACTH and corti-
costerone secretion are obviously triggered by the stressor, they are not motor
responses that act to remove the immediate consequences of the stress, as occurs
with reactive homeostatic motor events. The target actions of increased circulating
corticosterone (mediated in part by its interactions with leptin, insulin, and the
thyroid hormones) anticipate the possibility that the stressor will lead to a debili-
tating sequence of events, particularly the catabolic effects of negative energy bal-
ance (see also Sapolsky et al., 2000). In comparison, the more re¬‚ex homeostatic
motor components of the stress response are exempli¬ed by the consequences
of sympathetic activation that counteract the immediate effects of the stress; for
example, hypotension or hypoglycemia are stress-derived effects that can be rap-
idly negated by reactive sympathetic responses (e.g. vasoconstriction, increased heart
rate, hyperglycemia) that have evolutionarily ancient homologs (Watts, 2000).
To provide the framework for examining the regulatory actions of glucocorti-
coids on PVH neuroendocrine peptide gene expression during stress, we should
consider two issues within this context. First, what precisely is the function of
increased CRH and AVP (because of its importance to ACTH secretion during
222 A. G. Watts

stress) gene expression in neuroendocrine CRH neurons during the ACTH
response to stress? The most likely answer is that rather than impacting ongoing stress
events, stress-associated augmentation of mRNA levels are recuperative mecha-
nisms that provide the peptide synthetic mechanisms of the neuroendocrine neuron
with the ability to sustain future secretory activity. Second, what is the temporal
organization of the gene-regulatory response to stress? This sequence is obviously
quite complex, and is best considered when broken down into phases, each of which
is likely to be differentially regulated by corticosterone in a manner that may not
necessarily operate with an inhibitory action.
Constrained by this perspective, I will now evaluate glucocorticoid action on
CRH and AVP gene expression within the context of four components of the HPA
secretory response to stress:

(1) How the preceding corticosterone environment affects the onset of gene
expression (Initiation).
(2) The duration of gene activation (Dynamics).
(3) Which genes are activated (Which genes?).
(4) Corticosterone™s ongoing actions on secretogogue gene expression during the
secretory response itself (Negative feedback).

To approach these questions we have taken advantage of a viscerosensory stressor
(sustained hypovolemia) that has four qualities useful for data interpretation that
are not found with many other commonly used stressors. First, its sensory trans-
duction and physiological mechanisms are very well understood; second, its phys-
iological onset can be determined accurately; and third, because this occurs some
time after the stress of handling and injection, any effects resulting from the stressor
are easily distinguished from these initial non-speci¬c effects; and ¬nally, its intensity
increases linearly for up to 4 h (Tanimura et al., 1998).

Two questions concerning the initiation of gene transcription can be addressed
using sustained hypovolemia. First, how do the processes in the CRH neuron that
initiate secretogogue gene transcription temporally interact with those that con-
trol secretogogue release? Second, are these interactions dependent on cortico-
sterone? Answers to these questions should provide clues about the nature of the
intracellular mechanisms that link peptide synthesis to those that control peptide
release at the neuroendocrine terminal in the ME.
We have shown that ACTH secretogogue release from neuroendocrine terminals
in the ME during the early part sustained hypovolemia can occur in the absence of
an accompanying episode of CRH or AVP gene transcription (Figures 7.10“7.12).
223 Glucocorticoids and the ups and downs of neuropeptide gene expression

Stimulus Secretion CRH gene AVP gene
transcription transcription
Plasma volume deficit Plasma ACTH CRH hnRNA AVP hnRNA
* 100
* * 100
100 100

50 50 50
Response (% maximum)

0 0 0
0 2 3 0 1 3 0 1 2 0 1 2 3
1 2 3

* * 100 * *
100 100
50 50
50 50
* *

0 0 0
0 2 3 0 1 3 0 2 3 0 2 3
1 2 1 1
Time (h)

Figure 7.10 The presence of corticosterone has profound effects on the ability of sustained
hypovolemic stress to initiate and sustain CRH gene transcription. (a) in intact animals
sustained hypovolemia activates CRH gene transcription 3 h after stress onset; (b) but this
occurs earlier and transiently in ADX animals. Note that both the development of the
stimulus (as measured plasma volume de¬cit) and the secretory response of the CRH
neuron (as measured by ACTH release) are unaffected by the presence or absence of
corticosterone. Finally, AVP gene expression only occurs in the absence of corticosterone.
Black circled times on the X-axes indicated when a signi¬cant effect is ¬rst detected.
Data adapted from Tanimura et al. (1998) and Tanimura and Watts (2000). Open circles,
injected corticosterone; solid circles, sustained hypovolemia; * indicates p 0.05

This result shows clearly that the mechanisms responsible for gene transcription
are dissociable from those initiating activity-dependent secretogogue release.
Secretogogue gene transcription is activated only if release is maintained for a
signi¬cantly longer period as the stressor increases in intensity (Tanimura et al.,
1998). Importantly, these data show that increased gene transcription does not
invariably accompany an ongoing secretory event, and emphasize that the cellular
events that activate gene expression are very likely different from those responsible
for secretion. However, secretogogue release and ACTH secretogogue gene trans-
cription both occur together in the absence of corticosterone during sustained
hypovolemia (Tanimura and Watts, 2000).
224 A. G. Watts

Saline PEG Saline PEG




(a) (b)

Figure 7.11 Photomicrographs showing CRH mRNA, CRH hnRNA, and proenkephalin (pENK) mRNA
in situ hybridization signals in the PVH of (a) intact and (b) ADX animals 5 h after a
subcutaneous saline or polyethylene glycol (PEG) injection. Note that the absence of
corticosterone is associated with a lower CRH mRNA and hnRNA response to PEG
compared to the saline controls. However, the pENK mRNA response remains intact.
Data from Tanimura and Watts (1998)

Dynamics of transcriptional activation
What effect does corticosterone have on determining the dynamics of CRH gene
activation during sustained hypovolemia? In intact animals CRH gene transcrip-
tion occurs (as evidenced by measuring CRH hnRNA levels in the PVHmp) only
when a certain stress intensity threshold is reached (Figure 7.12; Tanimura et al.,
1998). Once this happens, transcription is then maintained in the presence of ele-
vated plasma corticosterone for up to 5 h. However, in two experiments we have
demonstrated that the detailed dynamics of this response are critically dependent
upon the corticosterone environment to which the CRH neuron has been exposed
before the stressor occurs.
First, we examined the effect of manipulating preceding circulating cortico-
sterone concentrations on the magnitude of the CRH mRNA response 5 h into the
stress, when the CRH mRNA response to the stressor is at its greatest (Tanimura
225 Glucocorticoids and the ups and downs of neuropeptide gene expression

Plasma ACTH concentration (pg/ml)

CRH hnRNA level (cell number)
500 100
10 0 10 20 30 40
10 0 10 20 30 40
(a) (b) Plasma volume deficit (%)

Figure 7.12 The relationship between stimulus intensity (percent of plasma volume de¬cit) and the
response of (a) ACTH secretion and (b) CRH hnRNA following injections of polyethylene
glycol (PEG) to intact animals. Note that ACTH secretion is activated at a lower stimulus
intensity than is CRH gene transcription suggesting that the mechanisms controlling these
two process can be uncoupled. Data adapted from Tanimura et al. (1998)

and Watts, 1998). In ADX animals with no corticosterone replacement, we found
that instead of increasing CRH transcription and mRNA levels, these were actually
lower at this time in stressed animals than in the unstressed controls. The magni-
tude of CRH gene response to stress returned to that seen in intact animals when
preceding plasma corticosterone were clamped at levels seen in intact animals.
However, the magnitude of the CRH mRNA response was enhanced compared to
intact animals, when plasma corticosterone levels were clamped at levels lower than
those required to normalize thymus weights. Activation of the pre-proenkephalin
gene (which is also increased by this stressor) was unaffected by these manipulations
of corticosterone. In a second experiment, we looked in more detail at how the
absence of corticosterone affected the temporal response of CRH gene expression
to stress (Tanimura and Watts, 2000). We found that the reason for the reversal
seen at 5 h in ADX animals was not because they could not initiate CRH gene tran-
scription, but because an initial and, compared to that in intact animals, premature
transcriptional episode could not be maintained (Figure 7.10). Collectively, these
data show that corticosterone has a profound effect on directing how the CRH
gene responds to the stressor, and that at very low plasma concentrations, corti-
costerone acts as a facilitatory agent that supports CRH gene transcription in the
face of a sustained stressor. Where this facilitatory action occurs is unknown, but
could target anywhere from the afferent sets encoding hypovolemia to mechanisms
in the CRH neuron itself.
226 A. G. Watts

Which genes?
Whether the CRH or AVP gene is activated during sustained hypovolemia depends


. 40
( 51 .)