Vasopressin Gene Expression in the Rodent Hypothalamus: Transcriptional and Posttranscriptional Responses to Physiological Stimulation
David Murphy and David Carter Neuropeptide Laboratory Institute of Molecular and Cell Biology National University of Singapore Singapore 0511, Republic of Singapore
The neuropeptide vasopressin (VP) is expressed in the supraoptic nucleus, a discrete group of neurons in the hypothalamus that respond to osmotic stimuli. In the rat the pattern of expression of VP mRNA changes in two ways as a consequence of the physiological stimulation of these neurons. Firstly, there is an accumulation of VP mRNA, and secondly, the poly(A) tail of the VP mRNA increases in length. We asked whether the increase in VP mRNA level is a consequence of transcriptional or posttranscriptional mechanisms. We present evidence from nuclear run-on assays that increases in the transcription of the rat VP gene are sufficient to account for the accumulation of VP mRNA observed in chronically stimulated animals. However, we note that in acutely stimulated animals there are rapid and relatively large increases in VP gene transcription that do not correlate with increases in the VP mRNA level, but coincide with the appearance of a homogeneous class of VP mRNAs with elongated poly(A) tails. We suggest that immediately after the onset of an acute osmotic stimulus, there is a rapid destruction of preexisting VP mRNAs and their replacement with new transcripts bearing longer poly(A) tails. We have also addressed the question of the function of the elongated VP mRNA poly(A) tail. It is unlikely that the poly(A) tail extension is involved in RNA stability; the transcriptional changes observed are sufficient to account for the increase in VP mRNA level, and we show that in the mouse similar increases in VP mRNA level are observed without concomitant changes in poly(A) tail length. We did not observe a change in the polysome distribution of the VP mRNA after osmotic stimulation. The elongated poly(A) tail of the VP mRNA may be involved in translational regulation or intracellular compartmentalization. (Molecular Endocrinology 4: 10511059, 1990)
INTRODUCTION
The brain peptide vasopressin (VP) is a major component of neuroendocrine systems that regulate salt and water balance in mammals. Circulating VP, released from stores in posterior pituitary nerve terminals, promotes water conservation by stimulating reabsorbtion of water by kidney collecting ducts in dehydrated animals. The VP gene is expressed in two classes of neurons in the mammalian hypothalamus: magnocellular neurons, which make up the supraoptic nucleus (SON) and a discrete magnocellular portion of paraventricular nucleus (PVN), and parvocellular neurons, found principally in the suprachiasmatic nucleus (SCN) and the parvocellular part of the PVN. Magnocellular cells of the SON and PVN innervate the posterior pituitary. Vasopressin is synthesised as part of a precursor preprohormone that is cleaved and processed during its passage from the neuronal cell bodies in the SON and PVN to the magnocellular axon terminals in the posterior lobe of the pituitary (1, 2). The mature VP peptide is stored in posterior lobe axon terminals until it is released into the peripheral circulation in response to an appropriate physiological stimulus, resulting in depletion of stored hormone (3). Presumably to compensate for the loss of stored material and to anticipate further demand, there is a concomitant increase in VP precursor synthesis in the hypothalamic magnocellular neurons (4, 5). The physiological regulation of VP synthesis in magnocellular neurons is not well understood. At the level of expression of VP mRNA in rat hypothalamus, two effects have been observed after an osmotic stimulus. Firstly, there is an increase in the abundance of VP RNA (6). It has not been determined whether this increase is a consequence of an increase in the transcription of the VP gene or an increase in the stability of the VP RNA. Secondly, the poly(A) tail of the VP mRNA increases dramatically in length (7-9). The function of the poly(A) tail of eukaryotic mRNAs, and how it ulti-
0888-8809/90/1051-1059S02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
1051
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 7
MOL ENDO-1990 1052
mately affects gene expression, is still not understood. However, there is support for the hypotheses that poly(A), possibly through its interaction with a specific poly(A)-binding protein, is involved in RNA stability (10, 11) and/or the control of translation (12, 13). Thus, we can speculate that the increase in the length of the VP mRNA poly(A) tail after an osmotic stimulus could result in an increase in synthesis of the VP preprohormone through two possible mechanisms. The larger poly(A) tail could stabilize the VP mRNA, resulting in an increase in the available pool of template for translation, and this could explain the observed increase in steady state levels of VP RNA observed in osmotically stimulated rats. Alternatively, the increased length of the poly(A) tract could directly regulate the translation or compartmentalization of VP mRNA. We have exploited three experimental models to investigate the regulation of VP mRNA after hyperosmotic stimulation. These models are 1) substitution of drinking water with a solution of 2% (wt/vol) NaCI for 5 days (salt loading); this is a chronic hyperosmotic stimulus of VP secretion; 2) complete fluid deprivation for periods of up 3 days (dehydration); dehydration is a preferable stimulus to salt loading for investigating the early events of an osmotic stimulus, as the former is not dependent upon drinking behavior (9); and 3) the ip injection of hypertonic (1.5 M) saline; this is an acute hyperosmotic isovolemic stimulus of VP secretion (15). We have previously shown (14, 16) that both dehydration and ip hypertonic saline injection result in an extremely rapid (within 1 - 2 h) increase in the poly(A) tail length of the VP mRNA. However, dehydration does not result in an increase in the abundance of the VP mRNA until 48 h after withdrawal of fluid (17, 18). We have sought to assess the relative contributions of transcription, RNA stability, and translational control to the physiological regulation of the VP gene during hyperosmotic stimuli in two rodent species: rat and mouse. We show that in both species a chronic osmotic stimulus results in an increase in hypothalamic VP mRNA abundance. However, only in the rat is there an increase in VP RNA poly(A) tail length. We have used nuclear run-on transcription assays to directly measure changes in the level of transcription of the rat VP gene as a consequence of the chronic osmotic stimuli of salt loading for 5 days and 3 days of dehydration. Both stimuli resulted in increases in the level of VP gene transcription sufficient to account for the concomitant increases in VP mRNA abundance. However, when we investigated the early transcriptional events that follow the onset of dehydration or the ip injection of hypertonic saline we were surprised to observe relatively large increases in the rate of VP gene transcription that do not correlate with changes in steady state levels of VP mRNA. Finally, we show by polysome analysis that the change in poly(A) tail length has no effect on the number of ribosomes associated with the VP mRNA.
RESULTS Changes in the Pattern of Expression of VP mRNA in the Rat SON and Mouse Hypothalamus after Osmotic Stimuli We have compared the changes in the pattern of expression of VP mRNA in the rat SON and the mouse hypothalamus after the onset of two osmotic stimuli: salt loading and dehydration. Five days of salt loading result in modest but reproducible increases in the steady state levels of both rat SON and mouse hypothalamic VP mRNA. This is illustrated by typical Northern blot results in Fig. 1b and is quantitatively summarized in Fig. 1a. Figure 1, a and b, also illustrate that there is an clear increase in the size of rat SON VP transcript after 5 days of salt loading. Its migration through the gel relative to the a-tubulin mRNA is decreased by 16% in salt-loaded animals compared to that in control animals. This is a consequence of an increase in the length of the poly(A) tail (7-9); when the poly(A) tract is enzymatically removed, the RNAs from control and stimulated animals comigrate (data not shown). However, there is no such increase in the size of murine VP mRNA after 5 days of salt loading (Fig. 1, a and b). Similarly, there is no increase in the length of the murine hypothalamic VP mRNA poly(A) tail after dehydration; a Northern blot of a time course of dehydration over a period of 2 days is presented in Fig. 2. A progressive increase in VP mRNA abundance is observed, but no increase in mRNA size can be identified. In the rat SON subjected to an identical stimulus, the change in the abundance of VP RNA is temporally distinct from the increase in the length of the VP RNA poly(A) tail (14). An increase in the steady state level of VP RNA was not observed until day 2, but the increase in poly(A) tail length can be seen as early as 2 h after the withdrawal of drinking water (14,18). This contrasts to the situation in the mouse hypothalamus, where there is a more rapid accumulation of VP RNA (an increase is observed by 12 h after water removal), and no increase in poly(A) tail length is apparent over the whole duration of the stimulus (Fig. 2). Changes in Transcription of the VP Gene after Chronic Hyperosmotic Stimuli We have used nuclear run-on transcription assays to assess the level of transcription of the VP gene in the rat SON after 5 days of salt loading and 3 days of dehydration (Fig. 3; tabulated in Table 1). After a 5-day salt loading, transcription of the VP gene increases 2fold (Fig. 3, Exp 1). In this experiment, SON were collected from 45 rats for each group. This increase in transcription is consistent with the similarly modest increase in steady state levels of VP mRNA seen in the rat SON after 5 days of salt loading (Fig. 1a). Three days of dehydration also resulted in a 2-fold increase in VP gene transcription (Fig. 3, Exp 2; eight animals per group). Again, this is consistent with the 2-fold increase
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
VP Gene Regulation
3
1053
VP mRNA ABUNDANCE
% Control animals
•I C
2
12 C
3 4
2
MIGRATION OF VP RNA
••if
% Control animals
RAT
Fig. 2. Time Course of the Accumulation of VP RNA in the Hypothalamus of Mice Subjected to Dehydration Northern blotting was performed on RNA (20 ^g/track) extracted from hypothalami of mice subjected to dehydration for 0.5 days (track 2), 1 day (track 3), and 2 days (track 4) and compared to that from control mice (track 1).
EXPERIMENT 1
PUC12
Fig. 1. Changes in the Abundance and Poly(A) Tail Length of
Rat SON and Mouse Hypothalamic VP mRNA after Salt Loading and Dehydration a, Quantitative analysis of VP mRNA abundance and poly(A) tail length in rat SON and mouse hypothalamus before (C) and after (2) salt loading. The level of VP RNA was determined relative to the level of an a-tubulin internal control and is expressed as a percentage of the mean of the control groups (mean ± SE). The length of the poly(A) tail is expressed as a function of the migration of the VP RNA (V) compared to the tubulin RNA (T). The migration (V - T) is expressed as a percentage of the mean of the control groups (mean ± SE). Note that an RNA with a longer poly(A) tail will have a shorter migration. Rats, n = 5, two animals per group; mice, n = 9, one animal per group, b, Representative autoradiograms of Northern analysis of rat SON (2 ^g/track) and mouse hypothalamic (25 Mg/track) RNA from control (C) and 2% NaCI-loaded (2) animals hybridized simultaneously with probes against atubulin (T) and VP RNAs. Mouse and rat RNAs were not fractionated on the same gel, and therefore, RNA sizes cannot be compared between species.
VP Thy-1
'•mi' GFAP
EXPERIMENT 2
Thy-1 PUC12 «•» VP GFAP
in VP mRNA levels found in these animals (18). Note that in both experiments there was no increase in the level of transcription of the Thy-1 or GFAP genes, and there was little nonspecific association between probe and the negative control pUC12 DNA. Transcriptional Responses of the VP Gene to Acute Stimuli We have previously shown that in the rat, the magnocellular vasopressinergic neurons of the SON undergo
Fig. 3. Transcriptional Changes in Rat SON Subjected to Chronic Stimuli Exp 1: Nuclei from control SON (1) and salt-loaded SON (2) were subjected to nuclear run-on analysis. In vitro labeled RNA was hybridized to 5 nq cloned plasmids containing VP genomic DNA (VP), glial fibrillary acidic protein cDNA (GFAP), and Thy1 genomic DNA. pUC12 DNA was used as a negative control. Exp 2: Nuclei from control (1) and 3-day dehydrated (2) rat SON were subjected to nuclear run-on analysis.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 1054
Vol 4 No. 7
Table I. Changes in the level of transcription in the SON of the rat VP gene in response to chronic and acute osmotic stimuli No. of nais/
Chronic 5-Day salt loading 3-Day dehydration Acute 12-h dehydration (A) 12-h dehydration (B) 24-h dehydration (A) 24-h dehydration (B) ip hypertonic NaCI, 0.5 h ip hypertonic NaCI, 1 h
% of Control Level
Croup
VP
Thy-1
GFAP
45 8
195 211
121 133
85 61
10 8 10 8
488 245 211 440
8
586
10
249
95 92 87 132 126 77
135 109 60 90 99 77
Nuclear run-on experiments were performed as described in Materials and Methods. Results are expressed as a percentage of the level of transcription found in parallel determinations of SON from unstimulated (control) SON.
GFAP
EXPERIMENT 1
Thy-1 PUC12 mm
Vp
4 EXPERIMENT 2
in VP gene transcription 0.5 and 1 h after the ip introduction of hypertonic saline. These data are tabulated in Table 1.
mm Thy-1 * mm
PUC12 Vp
*>- GFAP Fig. 4. Transcriptional Changes in Rat SON Subject to Acute Stimuli Exp 1. Run-on analysis was performed on rat SON nuclei isolated from control (1) animals and animals subjected to the following: 0.5-day dehydration (3), 1-day dehydration (4), and 1 h of stimulus after an ip injection of hypertonic saline (2). Exp 2: Run-on analysis was performed on rat SON nuclei isolated from control (1) animals and animals subjected to the following: 0.5-day dehydration (3), 1-day dehydration (4), and 0.5 h of stimulus after an ip injection of hypertonic saline (2).
a rapid response to acute stimuli of VP secretion (14, 16, 18). The stimuli examined were dehydration (a hyperosmotic, hypovolaemic stimulus), ip injection of hypertonic saline (a hyperosmotic isovolemic stimulus), and ip injection of polyethylene glycol (an isosmotic hypovolemic stimulus). Within 2 h of the onset of the stimuli, an increase in the length of the VP mRNA poly(A) tail was measurable (14,16,18). An increase in the level of VP mRNA is not a component of these early responses; rather, accumulation of VP RNA appears to be a chronic response in the rat (17, 18). We investigated the transcriptional changes occurring during the early acute response to dehydration and ip hypertonic saline (Fig. 4). We observed surprisingly large increases in VP gene transcription 12 and 24 h after the withdrawal of water. Similarly, we observed large increases
Polysome Distribution of VP mRNA in Control and Salt-Loaded Rat SON One possible function of the longer VP mRNA poly(A) tail in osmotically stimulated rats is in the stimulation of translation (13). We tested this possibility by determining whether the distribution of the VP mRNA in polysome gradients changed after a salt load. Two experiments were performed. In the first, cytoplasmic SON extracts from groups of 25 control or salt-loaded animals were fractionated through 5-50% (wt/vol) sucrose gradients containing Mg 2+ ions, which maintains the integrity of the polysomes. RNA was extracted from gradient fractions and subjected to Northern blotting (Fig. 5a). In a second experiment, cytoplasmic SON extracts from control or salt-loaded rats were subjected to two types of sucrose gradient centrifugation. Extracts from control or salt-loaded rats (45 animals/ group) were fractionated on 10-40% (wt/vol) sucrose gradients containing Mg 2+ ions, while identical extracts were fractionated on 10-40% (wt/vol) sucrose gradients containing EDTA. The EDTA chelates the divalent cations required for polysome integrity, resulting in dissociation of the polysomes and the release of mRNA, which can now be found only in lighter gradient fractions (19, 20). RNA extracted from these gradients was also subjected to Northern analysis (Fig. 5b). Figure 5, a and b (i), illustrate that in both experiments the distribution of VP mRNA in the gradients of intact polysomes of control and salt-loaded SON extracts is identical. The increase in poly(A) tail length does not apparently increase the overall number of ribosomes associated with the VP mRNA. As anticipated, EDTA has the effect of completely dissociating the polysomes from control rats (Fig. 5b; compare i, tracks 1-10, with ii, tracks 1-10). However, although the vast majority of VP mRNA appears in light fractions, the extracts from salt-loaded rats contain a class of VP mRNA associated with heavy cellular components that are apparently resistant to EDTA dissociation (Fig. 5bii, tracks 16-20). That the heavy polysomes are intact and quantitatively recovered was shown by reprobing the Northern filter in Fig. 5bi with a probe directed against the c-fos mRNA. c-fos mRNA was associated only with the heaviest gradient fractions from both control and salt-loaded rats (data not shown).
DISCUSSION
As a consequence of a chronic osmotic stimulus, expression of VP RNA in rat SON is subject to two sorts of change that are potentially important in the adaptive response of the animal. Firstly, there is an increase in the abundance of the VP mRNA (6). Sec-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
1055
VP Gene Regulation
50% 5%
50%
a
Total RNA
1 2 3 4 5 6 7 8 9 10 1112 13 14 15 16 17 18 19 20
b i)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
ii)
CONTROL
SALT LOADED
Fig. 5. Polysomal Distribution of VP RNA from SON of Control and Salt-Loaded Rats Distribution of VP mRNA in sucrose gradients from control (tracks 1-10) and salt-loaded rats (tracks 11-20). a, Five to 50% sucrose gradients containing Mg z+ ; b, 10-40% sucrose gradients containing i) Mg2+ and ii) EDTA. Heavier fractions (from the bottom of the gradient) are on the right of the Northern autoradiograms.
ondly the poly(A) tail of the VP mRNA increases in length (7-9). The increased abundance of VP mRNA could be achieved by either increased transcription of the VP gene or increased stability of the VP mRNA. Further, an increase in the stability of the VP mRNA could be a
consequence of the increase in poly(A) tail length (10, 11). This appears to be unlikely, however, as we have now described five lines of evidence that point to these two effects being separately regulated and functionally distinct. Firstly, the accumulation of VP mRNA in the osmotically stimulated SON is dependent upon seroton-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 1056
inergic functions, whereas the increase in VP mRNA poly(A) tail length appears to be independent of serotonin (9). Secondly, while there is no change in the amount of VP mRNA in the parvocellular cells of the SCN after an osmotic stimulation, the poly(A) tail length of the SCN VP mRNA increases in parallel with that of the magnocellular cells (18). This indicates that the systems up-regulating the VP mRNA level are unique to the magnocellular neurons, while the increase in VP mRNA poly(A) tail is a consequence of the activation of a pathway common to all hypothalamic vasopressinergic cells. Thirdly, the two modes of regulation are temporally distinct; after the onset of an acute osmotic stimulus, such as complete water deprivation, the increase in SON VP mRNA poly(A) tail length is extremely rapid (measurable within 2 h of the onset of the stimulus), whereas the increase in the SON VP mRNA level is not measurable until more than 24 h after the withdrawal of drinking water (17, 18). Fourthly, the two modes of regulation are developmentally distinct; the systems activating the osmotically stimulated increase in VP mRNA abundance are effective even before birth (21), the mechanisms that regulate VP mRNA poly(A) tail length are not mature until postnatal day 11 (KumFai Chooi, D. C , and D. M., manuscript in preparation). Fifthly, the increase in VP mRNA poly(A) tail length is species specific. There is no such effect in the mouse, although there is an accumulation of VP mRNA after an osmotic stimulus similar to that in the rat. To summarize these data, we have described circumstances where an increase in VP mRNA abundance can occur without an increase in VP mRNA poly(A) tail length, and circumstances where an increase in poly(A) tail length does not result in an increase in VP mRNA abundance. Thus, the osmotically induced accumulation of VP mRNA must either be a consequence of increased transcription or the activation of stability systems not involving the poly(A) tail. We have used nuclear run-on assays to measure changes in VP gene transcription in the rat SON after osmotic stimuli. The increases in transcription observed after the chronic stimuli of 5 days of salt loading and 3 days of dehydration correlate well with the observed changes in steady state VP mRNA levels after these stimuli. Thus, it is not necessary to invoke a stability mechanism to account for the accumulation of VP mRNA in the chronically stimulated SON. Using nuclear run-on assays we have also examined the early transcriptional response of the VP gene to acute stimuli. We subjected rats to two acute stimuli, complete water deprivation and ip injection of hypertonic saline. Both of these stimuli result in a rapid increase in the length of the VP mRNA poly(A) tail; a change is measurable within 1 - 2 h of the onset of the stimulus. However, no increase in VP mRNA abundance was observed at these early time points and, indeed, it cannot be detected until after 2 days of dehydration (18). We were surprised, therefore, to observe rapid and relatively large increases in VP gene transcription 12 and 24 h after water deprivation and
Vol 4 No. 7
0.5 and 1 h after hypertonic saline injection. This early activation of VP gene transcription does not correlate with increased VP mRNA abundance. Indeed, in dehydration experiments we have observed decreases in VP mRNA levels during the first 16 h of the stimulus (14). In the case of ip hypertonic saline injection experiments, we have consistently observed a decrease in VP mRNA abundance 1 h after application of the stimulus (14). Although these early transcriptional activation events do not correlate with increases in VP mRNA abundance, they do coincide with the increases in VP mRNA poly(A) tail length (14, 16, 18). The poly(A) tail length changes in a population of VP mRNAs are very homogenous; all of the molecules in a population have the same size poly(A) tail, and separate populations of elongated and unelongated mRNAs do not exist (7-9, 14,16,18). Two possible mechanisms, therefore, could mediate the increase in poly(A) tail length. Firstly, a cytoplasmic mechanism may act upon all VP mRNAs, both preexisting and newly synthesized. Such a mechanism could involve a cytoplasmic poly(A) polymerase (22), or it may be the result of an increased protection of the 3' end of the VP transcript from nucleolytic attack as a consequence of an increased number of ribosomes associated with a translationally stimulated VP mRNA. This latter possibility would appear unlikely, as our data indicate that the size of the VP mRNA polysomes does not change with salt loading. The second mechanism would involve the rapid degradation of preexisting VP RNAs with shorter poly(A) tails and their replacement with newly synthesised molecules with longer poly(A) tails. We have shown that there is a rapid increase in VP gene transcription, but a decrease in VP mRNA steady state levels, after application of an acute stimulus. We, therefore, suggest that the preexisting population of VP mRNA molecules is subject to degradation after an acute stimulus, to be rapidly replaced, through increased VP gene transcription, by a new population of molecules bearing an elongated poly(A) tail. The longer poly(A) tail could be added in at the primary site of polyadenylation in the nucleus or in the cytoplasm by the action of a cytoplasmic poly(A) polymerase. Alternatively, the cytoplasmic shortening of the poly(A) tail of VP mRNA recently transported from the nucleus may be prevented after salt loading (13). The precise role of the 3' poly(A) tract of eukaryotic mRNAs in the regulation of gene expression remains to be determined. Similarly, we still cannot attribute a function to the physiologically induced increase in length of the hypothalamic VP mRNA poly(A) tail that follows an osmotic stimulus. There is evidence to suggest that the poly(A) tail is involved in the stability of mRNAs (10, 11). However, given the transcriptional changes observed during a chronic osmotic stimulation, it is not necessary to invoke a stability mechanism to explain the concomitant increases in VP mRNA abundance. However, poly(A) removal, which precedes mRNA degradation, is thought to be a translation-dependent process (12). An increase in the translation of VP mRNAs after an osmotic stimulus might, therefore, lead to a
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
1057
VP Gene Regulation
decrease (rather than an increase) in the pool of translatable template were it not for a compensatory increase in poly(A) tail length. Recent evidence from yeast (13) has further implicated the poly(A) tail in association with the poly(A)-binding protein in the control of translation initiation. It is, thus, conceivable that the increase in poly(A) tail length may directly stimulate translation of the VP mRNA and consequent accumulation of VP precursor. We tested this hypothesis by analyzing the polysome distribution of the VP mRNA in extracts of control and salt-loaded SON. If the VP mRNA is subject to an increased rate of translation initiation as a consequence of an increase in the length of its poly(A) tail, then this may result in an increase in the number of ribosomes bound. The VP mRNA was found to be associated mainly with heavy polysomal fractions in control animals, and this distribution did not change with salt loading. The number of ribosomes associated with the VP mRNA populations in the SON of control and salt-loaded rats is identical. However, it should be noted that the VP mRNA is small and is subject to a high rate of translation even in control animals. Given that a small message can only accommodate a limited number of ribosomes, this may represent a physical limit. Thus, the rates of translation initiation, elongation, and termination, none of which can be measured by the polysome analysis described here if the physical size of the RNA is limiting ribosome number, may be influenced by the length of the poly(A) tail. The influence of poly(A) tail changes on translation of the VP mRNA could possibly be analyzed using in vitro translation systems prepared from SON. Additionally, we cannot rule out the possibility that the increased poly(A) tail length has effects on systems other than those affecting stability and translation. For example, the longer poly(A) tail could be involved in cellular compartmentalization. It has recently been demonstrated that the VP mRNA in hypothalamic cell bodies of the chronically osmotically stimulated Brattleboro rat (1, 2) has a different intracellular distribution than that in normal animals (23). In the latter, VP mRNA is found throughout the cytoplasm, but in the Brattleboro rat, the VP mRNA is mainly confined to the cytoplasmic periphery. The VP mRNA poly(A) tail of the Brattleboro rat is elongated compared to that of normal animals (24) as a consequence chronic dehydration. The poly(A) tail extension might, therefore, act as an intracellular localization signal. Similarly, we have previously described the presence of VP RNA in the posterior lobe of the pituitary. The abundance of this RNA increases dramatically with a hyperosmotic stimulus (25). We still do not know if this RNA is made locally in posterior pituitary astrocytes (pituicytes) or if it is transported down the pituitary stalk to posterior pituitary nerve terminals from vasopressinergic perikarya in the hypothalamus (24). If the posterior lobe VP RNA proves to be derived from the hypothalamus, then perhaps the increase in the length of the poly(A) tail forms part of a signal or facilitatory mechanism for the transportation of this RNA. However, this is unlikely, as the mouse, too, has VP RNA in its posterior pituitary,
which also increases in abundance with salt loading (25). However, the VP mRNA poly(A) tail in the osmotically stimulated mouse hypothalamus does not increase in length. Another possible function of the long poly(A) tail in osmotically stimulated rats is that it may act as a signal for the storage of VP RNA. We have noticed that a class of heavy VP mRNA-containing fractions is present in EDTA-containing sucrose gradients of salt-loaded, but not control, extracts. Such conditions dissociate polysomes, so the heavy fractions may contain storage, or perhaps even transport, particles. Studies in this laboratory are now focussing on identifying the VP gene sequences responsible for mediating the transcriptional and posttranscriptional changes that occur as a consequence of an osmotic stimulus. Further, we are attempting to identify protein factors that interact with these sequences and ascertain the role of intracellular signalling systems, such as cAMP (26), in the control of the activity of these factors.
MATERIALS AND METHODS Animals Groups of male Sprague-Dawley rats (200-300 g) and male CBA/J mice (8-10 weeks of age) were obtained from the Laboratory Animal Centre (National University of Singapore) and were maintained in constant conditions of lighting (lights on, 0600-1800 h) and temperature. Food was freely available. Physiological stimuli consisted of the following: 1) salt loading: drinking water was replaced with 2% NaCI (wt/vol) at 0800 h on day 0, and samples were taken at 0800 h 5 days later; 2) dehydration: for the 12 h time point, water was withdrawn at 2000 h, and samples were taken at 0800 h on the following morning. For the 24 h and 3 day time points, water was withdrawn at 0800 h, and samples were taken at 0800 h, 24 and 72 h later; and 3) ip hypertonic saline injection (1.5 M; 1.8 ml/100 g): a single injection was administered at 0700 h, with samples taken 0.5 and 1 h later. Rats and mice were killed by decapitation. Brains were removed and rinsed in ice-cold PBS. Rat SON and mouse hypothalami were dissected and either frozen on dry ice for subsequent RNA analysis or placed into ice-cold lysis buffer for immediate processing for nuclear runon or polysome analysis. RNA Analysis Total cellular RNA was isolated and analyzed on Northern blots, as previously described (25). Northern filters were probed either simultaneously or separately with 32P-labeled oligonucleotides directed against the VP RNA (27) and the
David Murphy and David Carter Neuropeptide Laboratory Institute of Molecular and Cell Biology National University of Singapore Singapore 0511, Republic of Singapore
The neuropeptide vasopressin (VP) is expressed in the supraoptic nucleus, a discrete group of neurons in the hypothalamus that respond to osmotic stimuli. In the rat the pattern of expression of VP mRNA changes in two ways as a consequence of the physiological stimulation of these neurons. Firstly, there is an accumulation of VP mRNA, and secondly, the poly(A) tail of the VP mRNA increases in length. We asked whether the increase in VP mRNA level is a consequence of transcriptional or posttranscriptional mechanisms. We present evidence from nuclear run-on assays that increases in the transcription of the rat VP gene are sufficient to account for the accumulation of VP mRNA observed in chronically stimulated animals. However, we note that in acutely stimulated animals there are rapid and relatively large increases in VP gene transcription that do not correlate with increases in the VP mRNA level, but coincide with the appearance of a homogeneous class of VP mRNAs with elongated poly(A) tails. We suggest that immediately after the onset of an acute osmotic stimulus, there is a rapid destruction of preexisting VP mRNAs and their replacement with new transcripts bearing longer poly(A) tails. We have also addressed the question of the function of the elongated VP mRNA poly(A) tail. It is unlikely that the poly(A) tail extension is involved in RNA stability; the transcriptional changes observed are sufficient to account for the increase in VP mRNA level, and we show that in the mouse similar increases in VP mRNA level are observed without concomitant changes in poly(A) tail length. We did not observe a change in the polysome distribution of the VP mRNA after osmotic stimulation. The elongated poly(A) tail of the VP mRNA may be involved in translational regulation or intracellular compartmentalization. (Molecular Endocrinology 4: 10511059, 1990)
INTRODUCTION
The brain peptide vasopressin (VP) is a major component of neuroendocrine systems that regulate salt and water balance in mammals. Circulating VP, released from stores in posterior pituitary nerve terminals, promotes water conservation by stimulating reabsorbtion of water by kidney collecting ducts in dehydrated animals. The VP gene is expressed in two classes of neurons in the mammalian hypothalamus: magnocellular neurons, which make up the supraoptic nucleus (SON) and a discrete magnocellular portion of paraventricular nucleus (PVN), and parvocellular neurons, found principally in the suprachiasmatic nucleus (SCN) and the parvocellular part of the PVN. Magnocellular cells of the SON and PVN innervate the posterior pituitary. Vasopressin is synthesised as part of a precursor preprohormone that is cleaved and processed during its passage from the neuronal cell bodies in the SON and PVN to the magnocellular axon terminals in the posterior lobe of the pituitary (1, 2). The mature VP peptide is stored in posterior lobe axon terminals until it is released into the peripheral circulation in response to an appropriate physiological stimulus, resulting in depletion of stored hormone (3). Presumably to compensate for the loss of stored material and to anticipate further demand, there is a concomitant increase in VP precursor synthesis in the hypothalamic magnocellular neurons (4, 5). The physiological regulation of VP synthesis in magnocellular neurons is not well understood. At the level of expression of VP mRNA in rat hypothalamus, two effects have been observed after an osmotic stimulus. Firstly, there is an increase in the abundance of VP RNA (6). It has not been determined whether this increase is a consequence of an increase in the transcription of the VP gene or an increase in the stability of the VP RNA. Secondly, the poly(A) tail of the VP mRNA increases dramatically in length (7-9). The function of the poly(A) tail of eukaryotic mRNAs, and how it ulti-
0888-8809/90/1051-1059S02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
1051
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 7
MOL ENDO-1990 1052
mately affects gene expression, is still not understood. However, there is support for the hypotheses that poly(A), possibly through its interaction with a specific poly(A)-binding protein, is involved in RNA stability (10, 11) and/or the control of translation (12, 13). Thus, we can speculate that the increase in the length of the VP mRNA poly(A) tail after an osmotic stimulus could result in an increase in synthesis of the VP preprohormone through two possible mechanisms. The larger poly(A) tail could stabilize the VP mRNA, resulting in an increase in the available pool of template for translation, and this could explain the observed increase in steady state levels of VP RNA observed in osmotically stimulated rats. Alternatively, the increased length of the poly(A) tract could directly regulate the translation or compartmentalization of VP mRNA. We have exploited three experimental models to investigate the regulation of VP mRNA after hyperosmotic stimulation. These models are 1) substitution of drinking water with a solution of 2% (wt/vol) NaCI for 5 days (salt loading); this is a chronic hyperosmotic stimulus of VP secretion; 2) complete fluid deprivation for periods of up 3 days (dehydration); dehydration is a preferable stimulus to salt loading for investigating the early events of an osmotic stimulus, as the former is not dependent upon drinking behavior (9); and 3) the ip injection of hypertonic (1.5 M) saline; this is an acute hyperosmotic isovolemic stimulus of VP secretion (15). We have previously shown (14, 16) that both dehydration and ip hypertonic saline injection result in an extremely rapid (within 1 - 2 h) increase in the poly(A) tail length of the VP mRNA. However, dehydration does not result in an increase in the abundance of the VP mRNA until 48 h after withdrawal of fluid (17, 18). We have sought to assess the relative contributions of transcription, RNA stability, and translational control to the physiological regulation of the VP gene during hyperosmotic stimuli in two rodent species: rat and mouse. We show that in both species a chronic osmotic stimulus results in an increase in hypothalamic VP mRNA abundance. However, only in the rat is there an increase in VP RNA poly(A) tail length. We have used nuclear run-on transcription assays to directly measure changes in the level of transcription of the rat VP gene as a consequence of the chronic osmotic stimuli of salt loading for 5 days and 3 days of dehydration. Both stimuli resulted in increases in the level of VP gene transcription sufficient to account for the concomitant increases in VP mRNA abundance. However, when we investigated the early transcriptional events that follow the onset of dehydration or the ip injection of hypertonic saline we were surprised to observe relatively large increases in the rate of VP gene transcription that do not correlate with changes in steady state levels of VP mRNA. Finally, we show by polysome analysis that the change in poly(A) tail length has no effect on the number of ribosomes associated with the VP mRNA.
RESULTS Changes in the Pattern of Expression of VP mRNA in the Rat SON and Mouse Hypothalamus after Osmotic Stimuli We have compared the changes in the pattern of expression of VP mRNA in the rat SON and the mouse hypothalamus after the onset of two osmotic stimuli: salt loading and dehydration. Five days of salt loading result in modest but reproducible increases in the steady state levels of both rat SON and mouse hypothalamic VP mRNA. This is illustrated by typical Northern blot results in Fig. 1b and is quantitatively summarized in Fig. 1a. Figure 1, a and b, also illustrate that there is an clear increase in the size of rat SON VP transcript after 5 days of salt loading. Its migration through the gel relative to the a-tubulin mRNA is decreased by 16% in salt-loaded animals compared to that in control animals. This is a consequence of an increase in the length of the poly(A) tail (7-9); when the poly(A) tract is enzymatically removed, the RNAs from control and stimulated animals comigrate (data not shown). However, there is no such increase in the size of murine VP mRNA after 5 days of salt loading (Fig. 1, a and b). Similarly, there is no increase in the length of the murine hypothalamic VP mRNA poly(A) tail after dehydration; a Northern blot of a time course of dehydration over a period of 2 days is presented in Fig. 2. A progressive increase in VP mRNA abundance is observed, but no increase in mRNA size can be identified. In the rat SON subjected to an identical stimulus, the change in the abundance of VP RNA is temporally distinct from the increase in the length of the VP RNA poly(A) tail (14). An increase in the steady state level of VP RNA was not observed until day 2, but the increase in poly(A) tail length can be seen as early as 2 h after the withdrawal of drinking water (14,18). This contrasts to the situation in the mouse hypothalamus, where there is a more rapid accumulation of VP RNA (an increase is observed by 12 h after water removal), and no increase in poly(A) tail length is apparent over the whole duration of the stimulus (Fig. 2). Changes in Transcription of the VP Gene after Chronic Hyperosmotic Stimuli We have used nuclear run-on transcription assays to assess the level of transcription of the VP gene in the rat SON after 5 days of salt loading and 3 days of dehydration (Fig. 3; tabulated in Table 1). After a 5-day salt loading, transcription of the VP gene increases 2fold (Fig. 3, Exp 1). In this experiment, SON were collected from 45 rats for each group. This increase in transcription is consistent with the similarly modest increase in steady state levels of VP mRNA seen in the rat SON after 5 days of salt loading (Fig. 1a). Three days of dehydration also resulted in a 2-fold increase in VP gene transcription (Fig. 3, Exp 2; eight animals per group). Again, this is consistent with the 2-fold increase
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
VP Gene Regulation
3
1053
VP mRNA ABUNDANCE
% Control animals
•I C
2
12 C
3 4
2
MIGRATION OF VP RNA
••if
% Control animals
RAT
Fig. 2. Time Course of the Accumulation of VP RNA in the Hypothalamus of Mice Subjected to Dehydration Northern blotting was performed on RNA (20 ^g/track) extracted from hypothalami of mice subjected to dehydration for 0.5 days (track 2), 1 day (track 3), and 2 days (track 4) and compared to that from control mice (track 1).
EXPERIMENT 1
PUC12
Fig. 1. Changes in the Abundance and Poly(A) Tail Length of
Rat SON and Mouse Hypothalamic VP mRNA after Salt Loading and Dehydration a, Quantitative analysis of VP mRNA abundance and poly(A) tail length in rat SON and mouse hypothalamus before (C) and after (2) salt loading. The level of VP RNA was determined relative to the level of an a-tubulin internal control and is expressed as a percentage of the mean of the control groups (mean ± SE). The length of the poly(A) tail is expressed as a function of the migration of the VP RNA (V) compared to the tubulin RNA (T). The migration (V - T) is expressed as a percentage of the mean of the control groups (mean ± SE). Note that an RNA with a longer poly(A) tail will have a shorter migration. Rats, n = 5, two animals per group; mice, n = 9, one animal per group, b, Representative autoradiograms of Northern analysis of rat SON (2 ^g/track) and mouse hypothalamic (25 Mg/track) RNA from control (C) and 2% NaCI-loaded (2) animals hybridized simultaneously with probes against atubulin (T) and VP RNAs. Mouse and rat RNAs were not fractionated on the same gel, and therefore, RNA sizes cannot be compared between species.
VP Thy-1
'•mi' GFAP
EXPERIMENT 2
Thy-1 PUC12 «•» VP GFAP
in VP mRNA levels found in these animals (18). Note that in both experiments there was no increase in the level of transcription of the Thy-1 or GFAP genes, and there was little nonspecific association between probe and the negative control pUC12 DNA. Transcriptional Responses of the VP Gene to Acute Stimuli We have previously shown that in the rat, the magnocellular vasopressinergic neurons of the SON undergo
Fig. 3. Transcriptional Changes in Rat SON Subjected to Chronic Stimuli Exp 1: Nuclei from control SON (1) and salt-loaded SON (2) were subjected to nuclear run-on analysis. In vitro labeled RNA was hybridized to 5 nq cloned plasmids containing VP genomic DNA (VP), glial fibrillary acidic protein cDNA (GFAP), and Thy1 genomic DNA. pUC12 DNA was used as a negative control. Exp 2: Nuclei from control (1) and 3-day dehydrated (2) rat SON were subjected to nuclear run-on analysis.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 1054
Vol 4 No. 7
Table I. Changes in the level of transcription in the SON of the rat VP gene in response to chronic and acute osmotic stimuli No. of nais/
Chronic 5-Day salt loading 3-Day dehydration Acute 12-h dehydration (A) 12-h dehydration (B) 24-h dehydration (A) 24-h dehydration (B) ip hypertonic NaCI, 0.5 h ip hypertonic NaCI, 1 h
% of Control Level
Croup
VP
Thy-1
GFAP
45 8
195 211
121 133
85 61
10 8 10 8
488 245 211 440
8
586
10
249
95 92 87 132 126 77
135 109 60 90 99 77
Nuclear run-on experiments were performed as described in Materials and Methods. Results are expressed as a percentage of the level of transcription found in parallel determinations of SON from unstimulated (control) SON.
GFAP
EXPERIMENT 1
Thy-1 PUC12 mm
Vp
4 EXPERIMENT 2
in VP gene transcription 0.5 and 1 h after the ip introduction of hypertonic saline. These data are tabulated in Table 1.
mm Thy-1 * mm
PUC12 Vp
*>- GFAP Fig. 4. Transcriptional Changes in Rat SON Subject to Acute Stimuli Exp 1. Run-on analysis was performed on rat SON nuclei isolated from control (1) animals and animals subjected to the following: 0.5-day dehydration (3), 1-day dehydration (4), and 1 h of stimulus after an ip injection of hypertonic saline (2). Exp 2: Run-on analysis was performed on rat SON nuclei isolated from control (1) animals and animals subjected to the following: 0.5-day dehydration (3), 1-day dehydration (4), and 0.5 h of stimulus after an ip injection of hypertonic saline (2).
a rapid response to acute stimuli of VP secretion (14, 16, 18). The stimuli examined were dehydration (a hyperosmotic, hypovolaemic stimulus), ip injection of hypertonic saline (a hyperosmotic isovolemic stimulus), and ip injection of polyethylene glycol (an isosmotic hypovolemic stimulus). Within 2 h of the onset of the stimuli, an increase in the length of the VP mRNA poly(A) tail was measurable (14,16,18). An increase in the level of VP mRNA is not a component of these early responses; rather, accumulation of VP RNA appears to be a chronic response in the rat (17, 18). We investigated the transcriptional changes occurring during the early acute response to dehydration and ip hypertonic saline (Fig. 4). We observed surprisingly large increases in VP gene transcription 12 and 24 h after the withdrawal of water. Similarly, we observed large increases
Polysome Distribution of VP mRNA in Control and Salt-Loaded Rat SON One possible function of the longer VP mRNA poly(A) tail in osmotically stimulated rats is in the stimulation of translation (13). We tested this possibility by determining whether the distribution of the VP mRNA in polysome gradients changed after a salt load. Two experiments were performed. In the first, cytoplasmic SON extracts from groups of 25 control or salt-loaded animals were fractionated through 5-50% (wt/vol) sucrose gradients containing Mg 2+ ions, which maintains the integrity of the polysomes. RNA was extracted from gradient fractions and subjected to Northern blotting (Fig. 5a). In a second experiment, cytoplasmic SON extracts from control or salt-loaded rats were subjected to two types of sucrose gradient centrifugation. Extracts from control or salt-loaded rats (45 animals/ group) were fractionated on 10-40% (wt/vol) sucrose gradients containing Mg 2+ ions, while identical extracts were fractionated on 10-40% (wt/vol) sucrose gradients containing EDTA. The EDTA chelates the divalent cations required for polysome integrity, resulting in dissociation of the polysomes and the release of mRNA, which can now be found only in lighter gradient fractions (19, 20). RNA extracted from these gradients was also subjected to Northern analysis (Fig. 5b). Figure 5, a and b (i), illustrate that in both experiments the distribution of VP mRNA in the gradients of intact polysomes of control and salt-loaded SON extracts is identical. The increase in poly(A) tail length does not apparently increase the overall number of ribosomes associated with the VP mRNA. As anticipated, EDTA has the effect of completely dissociating the polysomes from control rats (Fig. 5b; compare i, tracks 1-10, with ii, tracks 1-10). However, although the vast majority of VP mRNA appears in light fractions, the extracts from salt-loaded rats contain a class of VP mRNA associated with heavy cellular components that are apparently resistant to EDTA dissociation (Fig. 5bii, tracks 16-20). That the heavy polysomes are intact and quantitatively recovered was shown by reprobing the Northern filter in Fig. 5bi with a probe directed against the c-fos mRNA. c-fos mRNA was associated only with the heaviest gradient fractions from both control and salt-loaded rats (data not shown).
DISCUSSION
As a consequence of a chronic osmotic stimulus, expression of VP RNA in rat SON is subject to two sorts of change that are potentially important in the adaptive response of the animal. Firstly, there is an increase in the abundance of the VP mRNA (6). Sec-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
1055
VP Gene Regulation
50% 5%
50%
a
Total RNA
1 2 3 4 5 6 7 8 9 10 1112 13 14 15 16 17 18 19 20
b i)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
ii)
CONTROL
SALT LOADED
Fig. 5. Polysomal Distribution of VP RNA from SON of Control and Salt-Loaded Rats Distribution of VP mRNA in sucrose gradients from control (tracks 1-10) and salt-loaded rats (tracks 11-20). a, Five to 50% sucrose gradients containing Mg z+ ; b, 10-40% sucrose gradients containing i) Mg2+ and ii) EDTA. Heavier fractions (from the bottom of the gradient) are on the right of the Northern autoradiograms.
ondly the poly(A) tail of the VP mRNA increases in length (7-9). The increased abundance of VP mRNA could be achieved by either increased transcription of the VP gene or increased stability of the VP mRNA. Further, an increase in the stability of the VP mRNA could be a
consequence of the increase in poly(A) tail length (10, 11). This appears to be unlikely, however, as we have now described five lines of evidence that point to these two effects being separately regulated and functionally distinct. Firstly, the accumulation of VP mRNA in the osmotically stimulated SON is dependent upon seroton-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 1056
inergic functions, whereas the increase in VP mRNA poly(A) tail length appears to be independent of serotonin (9). Secondly, while there is no change in the amount of VP mRNA in the parvocellular cells of the SCN after an osmotic stimulation, the poly(A) tail length of the SCN VP mRNA increases in parallel with that of the magnocellular cells (18). This indicates that the systems up-regulating the VP mRNA level are unique to the magnocellular neurons, while the increase in VP mRNA poly(A) tail is a consequence of the activation of a pathway common to all hypothalamic vasopressinergic cells. Thirdly, the two modes of regulation are temporally distinct; after the onset of an acute osmotic stimulus, such as complete water deprivation, the increase in SON VP mRNA poly(A) tail length is extremely rapid (measurable within 2 h of the onset of the stimulus), whereas the increase in the SON VP mRNA level is not measurable until more than 24 h after the withdrawal of drinking water (17, 18). Fourthly, the two modes of regulation are developmentally distinct; the systems activating the osmotically stimulated increase in VP mRNA abundance are effective even before birth (21), the mechanisms that regulate VP mRNA poly(A) tail length are not mature until postnatal day 11 (KumFai Chooi, D. C , and D. M., manuscript in preparation). Fifthly, the increase in VP mRNA poly(A) tail length is species specific. There is no such effect in the mouse, although there is an accumulation of VP mRNA after an osmotic stimulus similar to that in the rat. To summarize these data, we have described circumstances where an increase in VP mRNA abundance can occur without an increase in VP mRNA poly(A) tail length, and circumstances where an increase in poly(A) tail length does not result in an increase in VP mRNA abundance. Thus, the osmotically induced accumulation of VP mRNA must either be a consequence of increased transcription or the activation of stability systems not involving the poly(A) tail. We have used nuclear run-on assays to measure changes in VP gene transcription in the rat SON after osmotic stimuli. The increases in transcription observed after the chronic stimuli of 5 days of salt loading and 3 days of dehydration correlate well with the observed changes in steady state VP mRNA levels after these stimuli. Thus, it is not necessary to invoke a stability mechanism to account for the accumulation of VP mRNA in the chronically stimulated SON. Using nuclear run-on assays we have also examined the early transcriptional response of the VP gene to acute stimuli. We subjected rats to two acute stimuli, complete water deprivation and ip injection of hypertonic saline. Both of these stimuli result in a rapid increase in the length of the VP mRNA poly(A) tail; a change is measurable within 1 - 2 h of the onset of the stimulus. However, no increase in VP mRNA abundance was observed at these early time points and, indeed, it cannot be detected until after 2 days of dehydration (18). We were surprised, therefore, to observe rapid and relatively large increases in VP gene transcription 12 and 24 h after water deprivation and
Vol 4 No. 7
0.5 and 1 h after hypertonic saline injection. This early activation of VP gene transcription does not correlate with increased VP mRNA abundance. Indeed, in dehydration experiments we have observed decreases in VP mRNA levels during the first 16 h of the stimulus (14). In the case of ip hypertonic saline injection experiments, we have consistently observed a decrease in VP mRNA abundance 1 h after application of the stimulus (14). Although these early transcriptional activation events do not correlate with increases in VP mRNA abundance, they do coincide with the increases in VP mRNA poly(A) tail length (14, 16, 18). The poly(A) tail length changes in a population of VP mRNAs are very homogenous; all of the molecules in a population have the same size poly(A) tail, and separate populations of elongated and unelongated mRNAs do not exist (7-9, 14,16,18). Two possible mechanisms, therefore, could mediate the increase in poly(A) tail length. Firstly, a cytoplasmic mechanism may act upon all VP mRNAs, both preexisting and newly synthesized. Such a mechanism could involve a cytoplasmic poly(A) polymerase (22), or it may be the result of an increased protection of the 3' end of the VP transcript from nucleolytic attack as a consequence of an increased number of ribosomes associated with a translationally stimulated VP mRNA. This latter possibility would appear unlikely, as our data indicate that the size of the VP mRNA polysomes does not change with salt loading. The second mechanism would involve the rapid degradation of preexisting VP RNAs with shorter poly(A) tails and their replacement with newly synthesised molecules with longer poly(A) tails. We have shown that there is a rapid increase in VP gene transcription, but a decrease in VP mRNA steady state levels, after application of an acute stimulus. We, therefore, suggest that the preexisting population of VP mRNA molecules is subject to degradation after an acute stimulus, to be rapidly replaced, through increased VP gene transcription, by a new population of molecules bearing an elongated poly(A) tail. The longer poly(A) tail could be added in at the primary site of polyadenylation in the nucleus or in the cytoplasm by the action of a cytoplasmic poly(A) polymerase. Alternatively, the cytoplasmic shortening of the poly(A) tail of VP mRNA recently transported from the nucleus may be prevented after salt loading (13). The precise role of the 3' poly(A) tract of eukaryotic mRNAs in the regulation of gene expression remains to be determined. Similarly, we still cannot attribute a function to the physiologically induced increase in length of the hypothalamic VP mRNA poly(A) tail that follows an osmotic stimulus. There is evidence to suggest that the poly(A) tail is involved in the stability of mRNAs (10, 11). However, given the transcriptional changes observed during a chronic osmotic stimulation, it is not necessary to invoke a stability mechanism to explain the concomitant increases in VP mRNA abundance. However, poly(A) removal, which precedes mRNA degradation, is thought to be a translation-dependent process (12). An increase in the translation of VP mRNAs after an osmotic stimulus might, therefore, lead to a
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 17:36 For personal use only. No other uses without permission. . All rights reserved.
1057
VP Gene Regulation
decrease (rather than an increase) in the pool of translatable template were it not for a compensatory increase in poly(A) tail length. Recent evidence from yeast (13) has further implicated the poly(A) tail in association with the poly(A)-binding protein in the control of translation initiation. It is, thus, conceivable that the increase in poly(A) tail length may directly stimulate translation of the VP mRNA and consequent accumulation of VP precursor. We tested this hypothesis by analyzing the polysome distribution of the VP mRNA in extracts of control and salt-loaded SON. If the VP mRNA is subject to an increased rate of translation initiation as a consequence of an increase in the length of its poly(A) tail, then this may result in an increase in the number of ribosomes bound. The VP mRNA was found to be associated mainly with heavy polysomal fractions in control animals, and this distribution did not change with salt loading. The number of ribosomes associated with the VP mRNA populations in the SON of control and salt-loaded rats is identical. However, it should be noted that the VP mRNA is small and is subject to a high rate of translation even in control animals. Given that a small message can only accommodate a limited number of ribosomes, this may represent a physical limit. Thus, the rates of translation initiation, elongation, and termination, none of which can be measured by the polysome analysis described here if the physical size of the RNA is limiting ribosome number, may be influenced by the length of the poly(A) tail. The influence of poly(A) tail changes on translation of the VP mRNA could possibly be analyzed using in vitro translation systems prepared from SON. Additionally, we cannot rule out the possibility that the increased poly(A) tail length has effects on systems other than those affecting stability and translation. For example, the longer poly(A) tail could be involved in cellular compartmentalization. It has recently been demonstrated that the VP mRNA in hypothalamic cell bodies of the chronically osmotically stimulated Brattleboro rat (1, 2) has a different intracellular distribution than that in normal animals (23). In the latter, VP mRNA is found throughout the cytoplasm, but in the Brattleboro rat, the VP mRNA is mainly confined to the cytoplasmic periphery. The VP mRNA poly(A) tail of the Brattleboro rat is elongated compared to that of normal animals (24) as a consequence chronic dehydration. The poly(A) tail extension might, therefore, act as an intracellular localization signal. Similarly, we have previously described the presence of VP RNA in the posterior lobe of the pituitary. The abundance of this RNA increases dramatically with a hyperosmotic stimulus (25). We still do not know if this RNA is made locally in posterior pituitary astrocytes (pituicytes) or if it is transported down the pituitary stalk to posterior pituitary nerve terminals from vasopressinergic perikarya in the hypothalamus (24). If the posterior lobe VP RNA proves to be derived from the hypothalamus, then perhaps the increase in the length of the poly(A) tail forms part of a signal or facilitatory mechanism for the transportation of this RNA. However, this is unlikely, as the mouse, too, has VP RNA in its posterior pituitary,
which also increases in abundance with salt loading (25). However, the VP mRNA poly(A) tail in the osmotically stimulated mouse hypothalamus does not increase in length. Another possible function of the long poly(A) tail in osmotically stimulated rats is that it may act as a signal for the storage of VP RNA. We have noticed that a class of heavy VP mRNA-containing fractions is present in EDTA-containing sucrose gradients of salt-loaded, but not control, extracts. Such conditions dissociate polysomes, so the heavy fractions may contain storage, or perhaps even transport, particles. Studies in this laboratory are now focussing on identifying the VP gene sequences responsible for mediating the transcriptional and posttranscriptional changes that occur as a consequence of an osmotic stimulus. Further, we are attempting to identify protein factors that interact with these sequences and ascertain the role of intracellular signalling systems, such as cAMP (26), in the control of the activity of these factors.
MATERIALS AND METHODS Animals Groups of male Sprague-Dawley rats (200-300 g) and male CBA/J mice (8-10 weeks of age) were obtained from the Laboratory Animal Centre (National University of Singapore) and were maintained in constant conditions of lighting (lights on, 0600-1800 h) and temperature. Food was freely available. Physiological stimuli consisted of the following: 1) salt loading: drinking water was replaced with 2% NaCI (wt/vol) at 0800 h on day 0, and samples were taken at 0800 h 5 days later; 2) dehydration: for the 12 h time point, water was withdrawn at 2000 h, and samples were taken at 0800 h on the following morning. For the 24 h and 3 day time points, water was withdrawn at 0800 h, and samples were taken at 0800 h, 24 and 72 h later; and 3) ip hypertonic saline injection (1.5 M; 1.8 ml/100 g): a single injection was administered at 0700 h, with samples taken 0.5 and 1 h later. Rats and mice were killed by decapitation. Brains were removed and rinsed in ice-cold PBS. Rat SON and mouse hypothalami were dissected and either frozen on dry ice for subsequent RNA analysis or placed into ice-cold lysis buffer for immediate processing for nuclear runon or polysome analysis. RNA Analysis Total cellular RNA was isolated and analyzed on Northern blots, as previously described (25). Northern filters were probed either simultaneously or separately with 32P-labeled oligonucleotides directed against the VP RNA (27) and the
