. 39
( 51 .)


AVP hnRNA (% sham ADX)

AVP mRNA (% sham ADX)
CRH hnRNA (% sham ADX)

400 *
* * 900
200 600
200 600

100 300 **
** ** **
** **
0 0
15 30 60 120 240 15 30 60 120 240
Sham ADX
Sham ADX
(a) (b)
Time from corticosterone injection (min) Time from corticosterone injection (min)

Figure 7.7 Supra-physiological doses of corticosterone acutely injected suppress CRH and AVP gene
expression in the PVH with different time scales. AVP gene expression is downregulated
more rapidly than CRH. Data from Ma et al. (1997)

showed that reductions in AVP hnRNA occurred within minutes, while CRH
hnRNA was much less responsive (Figure 7.7). These data suggest that cortico-
sterone uses different mechanisms to control CRH and AVP gene expression.
However, this type of experiment makes it very dif¬cult to determine the actions of
corticosterone, since the i.p. injection itself acts as a stressor. This means that the
administered corticosterone is acting in non-basal (i.e. stressed) conditions and
that other factors will make interpretation dif¬cult. Due to these problems deter-
mining exactly how long physiological levels of corticosterone take to alter CRH
gene expression in unstressed animals remains unclear, but the consensus is that
it is slow and of the order of hours. Certainly if circulating corticosterone shifts
above or below the circadian mean for a signi¬cant period (during chronic stress
or following adrenalectomy with or without exogenous corticosterone treatment)
resultant changes in CRH gene expression take at least 12 h to become measurable
(Swanson and Simmons, 1989; Ma and Aguilera, 1999). Considered together, these
data suggest that one component responsible for the sluggish response of CRH
gene expression to changes in circulating corticosterone is mediated by mecha-
nisms that modify either the rate of increase or decline in CRH gene transcription,
depending on whether corticosterone is decreasing or increasing across the day
(Watts et al., 2004). The fact that the time when transcriptional activation/decline
occurs is constrained within daily time windows implies that CRH gene expression
is, like ACTH secretion (Akana et al., 1986), differentially sensitive to cortico-
sterone across a 24-h period.
218 A. G. Watts

Physiological states

In the previous section I examined four domains in which glucocorticoids act to
control CRH gene expression in neuroendocrine neurons. I will now consider how
this control is manifest during two major physiological states: basal conditions,
when glucocorticoids exert important regulatory actions on metabolism (Dallman
et al., 2000); and stress, when glucocorticoid secretion is stimulated to control the
effects of perturbations away from the basal state.

Basal conditions
It has been known for many years that in most mammals, including humans,
ACTH and glucocorticoid secretion rates are not constant throughout the day.
Both hormones exhibit daily variations where maximum secretion occurs around
the time that maximum activity begins and minimum secretion around the time
that general activity slows (see Watts et al., 2004, for references).
To drive this daily secretory rhythm, signals from the circadian clock in the SCH
schedule CRH and, to lesser extent, AVP release from neuroendocrine terminals. In
turn, releasable pools of CRH and AVP in neuroendocrine terminals are sustained
by synthetic mechanisms in the PVHmp, a critical component of which involves
transcribing primary (hn) RNA transcripts from their cognate genes. Considering
the interaction between glucocorticoids and CRH and AVP gene expression we
discussed earlier, the question arises whether this relationship is manifest across
the day in the absence of stress? Is ACTH secretogogue synthesis maintained by
continuous low-level transcription, or are there signi¬cant episodes of CRH or
AVP gene transcription?
We recently showed that in intact rats there is a prominent increase in CRH
hnRNA levels that occurs at night when rats are most active (Figure 7.8). This
strongly suggests that, like the secretory components of the HPA axis, CRH gene
transcription is not constant across a 24-h period, but increases and decreases in a
simple rhythm (Watts et al., 2004). Interestingly, the rate of CRH gene transcrip-
tion is completely out of phase with ACTH secretion (Figure 7.8), suggesting that
separate mechanisms control secretogogue gene transcription and release at the
ME, and these mechanisms are only loosely coupled. I will return to this point in a
later section.
In this same study we showed that the ¬‚uctuating levels of circulating cortico-
sterone normally seen across the day are not required for daily rhythm of CRH
gene expression. However, varying levels of circulating corticosterone do have a
signi¬cant effect on the overall level of CRH gene transcription (Figure 7.9). In the
absence of stress the overall level of transcription is signi¬cantly higher in ADX
rats with no corticosterone replacement compared to intact animals (Watts et al.,
219 Glucocorticoids and the ups and downs of neuropeptide gene expression

40 120
120 CRH

CRH hnRNA hybridization signal

AVP hnRNA hybridization signal

Mean ( SEM) plasma ACTH
100 100

concentration (pg/ml)

(arbitrary units)
(arbitrary units) 80 80
20 60

15 AVP
40 40
20 20

0 24 0
20 24 4 8 12 16 20

(a) Zeitgeber time (h)

05.00 h 11.00 h 18.00 h 24.00 h






Figure 7.8 (a) CRH gene transcription (as indexed by CRH hnRNA levels) exhibits a marked daily
variation in intact animals. In contrast AVP gene transcription shows no change over the
day. Maximum CRH hnRNA levels are seen around the time of lights on, while the lowest
levels are around the time of lights off. The time of lights off is shown by the gray boxes.
Also note that the pattern of CRH gene transcription is completely out of phase with ACTH
secretion, suggesting that CRH gene transcription and release are uncoupled in the
absence of stress and (b) Corresponding photomicrographs at selected times of the day
are shown. Data adapted from Watts et al. (2004)

2004). Thus, the amount of CRH hnRNA present at both the nadir and peak of the
cycle is a crucial target of corticosterone™s long-term actions on CRH synthesis in
the PVHmp. It is likely that different mechanisms determine the values of each of
these parameters, which in turn involve the integration of neural information
220 A. G. Watts

700 70 120

hybridization signal (grayscale units)
hybridization signal (grayscale units)
600 60 100
Mean ( SEM) plasma ACTH

Mean ( SEM) CRH hnRNA

Mean ( SEM) AVP hnRNA
concentration (pg/ml)

500 50
400 40
300 30
200 20

100 10

0 0
2 6 10 14 18 22 2 2 6 10 14 18 22 2 2 6 10 14 18 22 2
Zeitgaber time (h) Zeitgaber time (h) Zeitgaber time (h)
(a) (b) (c)
ADX rats Intact animals

Figure 7.9 The absence of corticosterone in ADX rats profoundly increases the overall level of plasma
ACTH (a) and CRH hnRNA (b), and AVP hnRNA (c) compared to intact animals. However,
signi¬cant daily variations in plasma ACTH (a) and CRH hnRNA (b) still occur in both
intact and ADX rats, whereas a daily rhythm of AVP hnRNA (c) is only seen in ADX rats

encoded by sets of PVHmp afferents and humoral agents, of which corticosterone
is the most important (Lightman and Harbuz, 1993; Watts, 1996; Sawchenko et al.,
2000; Watts and Sanchez-Watts, 2002).
The pattern of ¬‚uctuating CRH hnRNA levels we see in intact animals contrasts
sharply with AVP hnRNA, which is expressed at very low levels in the PVHmp of
intact animals, and shows no variations across the day (Figure 7.8). However, there
are signi¬cant daily variations of AVP hnRNA levels in the PVHmp of ADX ani-
mals (Figure 7.9). Given that the actions of corticosterone on AVP gene expression
are rapid and involve direct actions on the gene, and that the AVP gene is exqui-
sitely sensitive to circulating corticosterone (Burke et al., 1997; Ma et al., 1997;
Kovács et al., 2000; Watts and Sanchez-Watts, 2002), it would seem that the
amounts of corticosterone circulating in intact rats during the latter part of the
light period and early dark period are suf¬cient to suppress completely the mech-
anism that drives AVP gene expression. This suggests that, unlike CRH gene tran-
scription, the dynamics of circulating corticosterone in intact animals are
important for blunting daily variations in AVP gene transcription. The constant
and very low levels of AVP hnRNA we see throughout the day is consistent with


. 39
( 51 .)