Address correspondence and reprint requests to Matthew D. Whim, Department of Cell Biology and Anatomy, LSU Health Sciences Center, Medical Education Building (MEB 6142), 1901 Perdido Street, New Orleans, LA 70112, USA. E-mail: firstname.lastname@example.org
Neuropeptide Y is a co-transmitter that is synthesized by chromaffin cells in the adrenal medulla. During the fight-or-flight response these cells release NPY in addition to epinephrine and norepinephrine. Following the stress-induced reflex, the levels of NPY are increased as part of a homeostatic response that modulates catecholaminergic signaling. Here, we examined the control of NPY expression in mice after brief exposure to the cold water forced swim test. This treatment led to a shift in NPY expression between two populations of chromaffin cells that reversed over the course of 1 week. When NPY(GFP) BAC transgenic animals were exposed to stress, there was an increase in cytoplasmic, non-secretable GFP, indicating that stress increased NPY promoter activity. In vivo blockage of Y2 (but not Y1 or Y5) receptors increased basal adrenal NPY expression and so modulated the effects of stress. We conclude that release of NPY mediates a negative feedback loop that inhibits its own expression. Thus, the levels of NPY are determined by a balance between the potentiating effects of stress and the tonic inhibitory actions of Y2 receptors. This may be an efficient way to ensure the levels of this modulator do not decline following intense sympathetic activity.
A characteristic feature of the response to stress is an increase in the activity of endocrine cells in both the adrenal cortex and medulla. Activation of the hypothalamic-pituitary-adrenal (HPA) axis induces the release of steroid hormones from the adrenal cortex while exposure to stress evokes secretion of norepinephrine and epinephrine from chromaffin cells in the adrenal medulla. This change in release is rapidly followed by a compensatory change in secretory capacity. Thus, stressors such as restraint, cold stress, and hypoglycemia all lead to an increase in the expression of tyrosine hydroxylase, the rate limiting enzyme for catecholamine synthesis (Chuang and Costa 1974; Fluharty et al. 1983; Nankova et al. 1994; Vietor et al. 1996). Stress also regulates the cortical expression of P450 11β-hydroxylase and aldosterone synthetase which are required for corticosterone and aldosterone production, respectively (Aguilera et al. 1995; Engeland et al. 1997). These coordinated changes in catecholamine release and synthesis are coupled processes, giving rise to the term ‘stimulus-secretion-synthesis coupling’ (Thoenen et al. 1969a, b; Eiden et al. 1984; Mahata et al. 2003). They presumably ensure that the adrenal response does not become refractory after exposure to an acute stressor.
Although this activity-dependent increase in release capacity appears to be a failsafe mechanism, positive feedback loops carry a potential risk because increased filling of the secretory pool of granules could lead to an overly vigorous stress response. For example repeated exposure to stress can lead to a facilitated release of catecholamines and glucocorticoids (Lilly et al. 1986; Figueiredo et al. 2003; Kuzmiski et al. 2010) suggesting that a change in secretory capacity can have physiological consequences. Chronic diseases including type II diabetes and hypertension are also associated with altered sympathoadrenal activity and catecholamine secretion (Guyenet 2006; McCrimmon and Sherwin 2010; Ramanathan and Cryer 2011) but whether disregulated secretory capacity contributes to these conditions is not known. Under normal circumstances what prevents an acute stressor from leading to an unrestrained increase in secretory capacity? We have recently found that the adrenal expression of tyrosine hydroxylase is tonically suppressed by neuropeptide Y and that loss of NPY leads to an elevated secretory capacity (Wang et al., submitted). Because NPY is a co-transmitter with the catecholamines in chromaffin cells and can be co-released (Varndell et al. 1984; Henion and Landis 1990; Whim 2006) this suggests that the peptide could play the restraining role outlined above. Co-transmitters (in particular neuropeptides) may be an effective way to provide feedback regulation because their release generally requires high frequency bursts of neuronal activity (Bloom et al. 1988; Leng and Ludwig 2008) - precisely those conditions under which feedback modulation occurs.
The idea that NPY might be involved in the adaptive response to stress is consistent with the finding that the adrenal levels of NPY mRNA and peptide are raised after exposure to an acute stressor. This transcriptionally dependent response is prevented either by adrenal denervation or by blockade of nicotinic ACh receptors indicating that pre-ganglionic activity plays an important role in determining NPY expression (Hiremagalur et al. 1994; Jahng et al. 1997). Because the adrenal medulla contains several NPY receptor subtypes (Cavadas et al. 2006) this raises the possibility that the expression of NPY itself could also be controlled by an autocrine feedback loop. Nevertheless, there is disagreement about which cells in the adrenal medulla express NPY (Varndell et al. 1984; Lundberg et al. 1986; Henion and Landis 1990) and the role of the different Y receptor subtypes in modulating NPY expression has not been examined. In this study, we use a molecular and transgenic approach to track NPY promoter activity and peptide expression after exposure to an acute stressor and then determine whether Y receptors are involved in regulating the adrenal expression of NPY in vivo.
Materials and methods
Animals and stress paradigm
Wild type C67BL/6J mice and NPY(GFP) animals (van den Pol et al. 2009) were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). Male animals were used in most experiments with the exception of some of the experimental groups in Figs 1, 2c and 5. Animals were individually housed before (3–5 days) and after stress exposure. The cold water forced swim test was performed as described (Saal et al. 2003). Briefly, mice were placed in an 18 cm diameter cylinder containing 3 L of cold water (4–5°C) for 5 to 6 min. After recovery under an infra red heat lamp (15–25 min) mice were returned to their home cages. Control animals were left undisturbed. Mice were killed by decapitation, 1 day or 1 week later, as noted in the text. Littermates were used in all experiments (P24–26). All experiments involving animals were approved by the Animal Care and Use Committees at Louisiana State University Health Sciences Center and Pennsylvania State University.
In vivo injection
Animals were injected (i.p.) with Y receptor antagonists; BIBP 3226 (1 mg/kg), BIIE 0246 (1 mg/kg) or L-152,804 (10 mg/kg) 15 min before the forced swim test. Control animals were injected with the same volume (~200 μL) of sterile saline.
Cell culture and immunocytochemistry
Adrenal medullae, in most case from P24–26 mice, were dissociated and chromaffin cell cultures were prepared on poly-d-lysine coated coverslips as described (Whim 2006). Two hours after plating the cultures were fixed with 4% paraformaldehyde then stained (Mitchell et al. 2008). Primary antibodies were rabbit anti-NPY (1:40000, T-4070, Peninsula Labs, San Carlos, CA, USA) and guinea pig anti-PNMT (1 : 100–500, EUD7001, Acris, San Diego, CA, USA). Secondary antibodies were donkey anti-rabbit Alexa 488 (1:200, Invitrogen, Grand Island, NY, USA) and donkey anti-guinea pig DyLight 549 (1:100, Jackson ImmunoResearch, West Grove, PA, USA).
Adrenal glands were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) overnight at 24°C then washed with PBS, rapidly frozen in 2-methylbutane on dry ice, embedded in cryomatrix medium and kept at −80°C until use. Frozen sections (30 μm) were washed with PBS, re-fixed with 4% paraformaldehyde for 20 min then permeabilized in 0.3% Triton X-100 in PBS for 15 min. Sections were then incubated in 3% H2O2 for 45–60 min to quench endogenous peroxidase activity and transferred to blocking reagent (in Tris buffered saline; TBS, Perkin Elmer, Waltham, MA, USA) for 30 min. Sections were incubated in primary antibody (rabbit anti-NPY; 1:10000, Peninsula Labs, T-4070) at 4°C overnight. Sections were washed with PBS containing 0.05% Tween-20 and incubated with a secondary antibody conjugated to peroxidase (1:500 donkey anti-rabbit; Jackson ImmunoResearch) for 30 min at 24°C, washed with PBS containing 0.05% Tween-20, incubated in TSA-FITC (Perkin Elmer) for 10 min, then sequentially washed with PBS containing 0.05% Tween-20, rinsed with distilled water and mounted in Vectashield.
Images were obtained with a Nikon TE2000U microscope (Melville, NY, USA) and 10X and 60X oil immersion (1.4 numerical aperture) objectives and a Retiga 1300 monochrome camera (Surrey, BC, Canada). Image-Pro Plus 5.1 (Media Cybernetics, Rockville, MD, USA), Origin Pro7, Excel and Minitab 15 were used for data analysis. To quantify the degree of staining in tissue sections, the medulla minus any cell-free areas was outlined and the mean pixel intensity was calculated (i.e. an area-independent measurement). Images were then background subtracted using AOI's from the adrenal cortex and 7–10 sections from each animal were measured. In each figure open circles indicate the mean value from each animal and black lines link the matched control and stressed animals from each independent experiment.
For analysis of cells in vitro, images of single chromaffin cells were taken at the level of the nucleus to maintain the same focal plane between cells. The mean pixel intensity was calculated as above, but background subtraction was not used except for the data shown in Figs 2d and 3d. In all staining experiments involving the quantification of the stress response, the slides were blinded until the analysis was complete.
Reverse transcription PCR
mRNA was prepared from the adrenal medulla using TRIzol (Invitrogen) following the manufacturer's instructions. Samples were treated with DNase (Ramamoorthy and Whim 2008), purified using RNeasy (Qiagen, Valencia, CA, USA) and cDNA was obtained by reverse transcription. For semi-quantitative RT-PCR, sequential dilutions of cDNA's at 1:4, 1:16 and 1:64 were used for PCR amplification and the results were quantified only within the linear range. NPY expression was normalized by actin expression. Primers and PCR protocols were as follows; NPY; 5′ CACGATGCTAGGTAACAAG 3′ and 5′ CACATGGAAGGGTCTTCAAG 3′ (294 bp, annealing Tm 55°C, 35 cycles); actin; 5′ GCCAACCGTGAAAAGATGAC 3′ and 5′ CAACGTCACACTTCATGATG 3′ (523 bp, annealing Tm 55°C, 35 cycles). The latter two primers have been previously described (van den Pol et al. 2009).
Statistical significance was assessed using the Student's t-test. Comparisons between three or more groups were made with a general linear model anova (post hoc Tukey's paired comparison). Population distributions were fit using OriginPro 7. As the absolute level of staining varied between individual experiments comparisons were made between paired control and experimental groups (which were processed in parallel and generated from littermates in each independent experiment). In the figures individual data points are plotted as filled symbols. In histograms with error bars black lines connect the average values (open symbols) from paired control and experimental animals. The Kolmogorov–Smirnov test was used for analyzing statistical significance in the cumulative fraction data sets.
Acute stress leads to a sustained increase in the adrenal expression of neuropeptide Y
Previous studies have shown that a variety of chronic stressors including restraint, cold, and hypoxia lead to an increase in the adrenal expression of NPY (Hiremagalur et al. 1994; Han et al. 2005; Raghuraman et al. 2011). To determine whether the peptidergic synthetic capacity of the sympathoadrenal system is also responsive to an acute stress, we exposed mice to the cold water forced swim test (FST) for 5 min. Control and stressed mice were then killed 1 day or 1 week later and the level of NPY-immunoreactivity (NPY-ir) was examined in adrenal cryosections. In all cases NPY-ir was restricted to the medulla as expected for a co-transmitter that is expressed by chromaffin cells (Fig. 1a). Quantification of NPY-ir indicated that 1 day after the FST there was a significant increase in NPY expression compared with matched control animals (69.9 ± 25.6 vs. 203.4 ± 50.6 arbitrary intensity units; mean ± SEM, n = 3, p < 0.01) and that the levels had returned toward baseline 1 week after the stress exposure (Fig. 1b).
To examine the stress-induced change in NPY expression at the level of single cells, the effect of the FST was measured from individual chromaffin cells isolated in vitro (Fig. 2a). In cells from control animals the NPY-ir intensity distribution was skewed and after stress the data were well fit with the sum of two Gaussian distributions. This suggested that the in vivo treatment had led to a shift in NPY expression between two populations of cells, with stress resulting in an increase in the proportion of cells expressing high levels of NPY and a corresponding decline in the number of low-expressing cells (Fig. 2b). Group data confirmed that stress resulted in a significant increase in the level of NPY-ir (Fig. 2c) and led to a large rightward shift in the cumulative distribution of intensities (Fig. 2c).
The presence of two populations of NPY-ir cells raised the question of whether these corresponded to epinephrine and norepinephrine chromaffin cells. To address this issue, acutely isolated cells were co-stained for NPY and PNMT. The latter is a selective marker of epinephrine-synthesizing cells (Ebert et al. 2004). NPY-ir was observed in both PNMT positive and negative chromaffin cells (Fig. 2d). This is in contrast to previous studies in which NPY was reported to be restricted to either epinephrine (Lundberg et al. 1986; Henion and Landis 1990) or norepinephrine-synthesizing chromaffin cells (Varndell et al. 1984). When the levels of immunoreactivity were quantified no examples of PNMT positive/NPY negative cells were found (Fig. 2d middle panel) consistent with the ubiquitous expression of NPY in chromaffin cells. Finally, the correlation coefficient between the levels of NPY- and PNMT-ir in single cells was very low (Fig. 2d, right panel; R = 0.52). Thus, the two populations of NPY-expressing chromaffin cells do not appear to be related to their ability to synthesize either epinephrine or norepinephrine.
The effect of acute stress involves an increase in NPY synthesis
Neuropeptide Y enters the regulated secretory pathway and is released via exocytosis. Thus, unlike the stress-induced increase in the enzyme tyrosine hydroxylase (Thoenen 1970; Nankova et al. 1994), the change in NPY-ir could be because of an increase in synthesis or a reduction in secretion (leading to elevated levels of intracellular NPY). To determine whether the change in NPY-immunoreactivity reflected a genuine increase in peptide synthesis, we examined the effects of stress on GFP expression in NPY(GFP) BAC mice. In these animals GFP is produced as a cytoplasmic protein whose expression is controlled by regulatory elements in the NPY gene (van den Pol et al. 2009). Therefore, we reasoned that an increase in GFP expression would reflect a transcriptional activation of the NPY gene by stress. In the adrenals of NPY(GFP) mice, GFP expression was limited to tyrosine hydroxylase-expressing cells consistent with the restriction of NPY expression to chromaffin cells (Wang et al. submitted). When the levels of GFP were measured from chromaffin cells in vitro we found that GFP expression was significantly higher in cells isolated from stressed animals (Fig. 3a). In the control animals the intensity distribution was well fit with the sum of two Gaussian distributions again indicating the presence of two populations of NPY-expressing chromaffin cells (Fig. 3b). Stress led to a decrease in the proportion of low GFP-expressing cells and an increase in the number of cells expressing high levels of GFP (Fig. 3b). Group data showed that stress resulted in a significant increase in the level of GFP (Fig. 3c, left panel) and led to a large rightward shift in the cumulative distribution of intensities confirming the increase in GFP expression (Fig. 3c, right panel). These results indicate that stress leads to an increase in NPY gene transcription. When acutely isolated cells were stained for PNMT (Fig. 3d, left panel) we found that although all chromaffin cells expressed NPY(GFP), a fraction of chromaffin cells were PNMT negative consistent with NPY synthesis by both epinephrine and norepinephrine chromaffin cells. The correlation coefficient was again low (Fig. 3d, middle panel; R = 0.41) implying that NPY expression is not related to the ability to synthesize epinephrine. The normalized PNMT-ir intensity distribution was clearly skewed and well fit by the sum of two Gaussian distributions (Fig. 3d, right panel) consistent with the idea that chromaffin cells release either epinephrine or norepinephrine (Petrovic et al. 2010). Finally semi-quantitative RT-PCR experiments were used to confirm that the levels of NPY mRNA were indeed significantly higher in the stressed animals compared with matched controls (Fig. 3e).
Y2, but not Y1 or Y5 receptors, regulate the expression of adrenal NPY in vivo
What is the activity-dependent signal that mediates the increase in NPY expression? Previous studies have shown that preventing splanchnic nerve activity and the activation of nicotinic receptors eliminates the stress-induced increase in adrenal NPY mRNA (Hiremagalur et al. 1994; Jahng et al. 1997). However, the adrenal medulla contains mRNA encoding the Y1, Y2 and Y5 neuropeptide Y receptors (Cavadas et al. 2006). Because chromaffin cells can co-release the catecholamines and NPY (Cavadas et al. 2001; Whim 2006) one possibility is that NPY could locally regulate its own expression - this would also have been prevented by splanchnic nerve blockade. To test the involvement of Y receptors we individually blocked the Y1, Y2, and Y5 receptors by i.p. injection of antagonists before exposure to the forced swim test. These experiments showed that BIIE 0246 (1 mg/kg), a Y2 antagonist (Doods et al. 1999), led to a significant increase in the basal levels of NPY-ir (Fig. 4a; 277.2 ± 25.9 vs. 429.0 ± 50.2 arbitrary intensity units; control vs. BIIE0246 injection; mean ± SEM, n = 3, p < 0.05). In the BIIE 0246-injected animals, stress did not lead to an additional elevation in NPY-ir perhaps because of the effect of BIIE0246 on basal NPY expression (Fig. 4b; 491.5 ± 64.7 vs. 484.6 ± 63.8 arbitrary intensity units; FST vs. FST after BIIE0246 injection; mean ± SEM, n = 3, not significant).
In contrast although germline Y1 knockout mice have elevated levels of NPY (Cavadas et al. 2006) we did not detect an effect of acute Y1 receptor blockade on NPY-ir. Thus, injection of BIBP3226 (1 mg/kg),a Y1 antagonist (Doods et al. 1996) did not alter either the basal expression of NPY (Fig. 5a) or prevent the potentiating effects of acute stress (Fig. 5b). Because we have found that this concentration of BIBP3226 leads to a significant increase in the adrenal expression of tyrosine hydroxylase (Wang et al. submitted) and is sufficient to block an NPY-induced increase in blood pressure in rats (Doods et al. 1996), the lack of effect on NPY expression is unlikely to be because of an ineffective concentration of antagonist. In similar experiments L-152,804 (10 mg/kg), a Y5 antagonist (Kanatani et al. 2000; Schroeder et al. 2005) also did not alter either the basal expression of NPY (Fig. 5c) or prevent the potentiating effects of acute stress (Fig. 5d). Mice with a germline deletion of the Y5 receptor also show no change in the whole brain levels of NPY mRNA (Marsh et al. 1998). From these experiments we conclude that rather than being responsible for the effects of stress, NPY actually negatively regulates its own expression via the Y2 receptor.
A large and rapid increase in catecholamine secretion from sympathetic neurons is a canonical feature of the stress response. To compensate for this change in release, exposure to a stressor is usually followed by a sustained elevation in the expression of tyrosine hydroxylase, the rate limiting enzyme for catecholamine synthesis. When the stress-induced change in tyrosine hydroxylase expression is lost, the ability to adequately respond to a stressor is also altered, consistent with a functional role for stimulus-secretion-synthesis coupling (Hamelink et al. 2002). However, while the cellular and molecular mechanisms that regulate the initiation of the sympathoadrenal stress response are well known (for review see Kvetnansky et al. 2009), the factors that limit its extent are less clear. Here, we examined the effect of stress on the adrenal expression of neuropeptide Y which is co-released with the catecholamines and negatively regulates tyrosine hydroxylase expression (Cavadas et al. 2006).
Using a variety of approaches we find that an acute stressor that lasts only 5 min leads to a large increase in adrenal NPY expression. Thus, both brief and extended exposure to stress can regulate the adrenal expression of NPY (Hiremagalur et al. 1994; Jahng et al. 1997; Han et al. 2005). Quantitative analysis of NPY expression using immunocytochemistry or transgenic expression of NPY(GFP) indicates that there are two populations of chromaffin cells that differ in their relative expression of NPY and that acute stress shifts the balance between the two. Unexpectedly these two populations do not correspond to epinephrine- and norepinephrine-containing chromaffin cells as assessed by PNMT-ir. Thus, our findings differ from previous study in which NPY was localized either to epinephrine (Lundberg et al. 1986; Henion and Landis 1990) or norepinephrine chromaffin cells (Varndell et al. 1984). It is now appreciated that epinephrine- and norepinephrine-synthesizing chromaffin cells are functionally distinct. Thus, hypoglycemia leads preferentially to epinephrine secretion while a fall in blood pressure activates the norepinephrine-synthesizing chromaffin cells (Scheurink and Ritter 1993; Morrison and Cao 2000; Verberne and Sartor 2010). The fact that acute stress does not lead to a uniform increase in NPY expression is consistent with the adrenal containing a heterogeneous mixture of chromaffin cells.
What is the functional significance of the rise in NPY expression? Because NPY inhibits the adrenal expression of tyrosine hydroxylase (Cavadas et al. 2006) it appears that NPY is an endogenous regulator of the stress response. By limiting tyrosine hydroxylase expression and thus the adrenal secretory capacity, NPY may act as a brake thus ensuring that the potentiating effect of stress on catecholamine secretion and synthetic capacity does not lead to a pathologically overactive stress response. Although Y receptors are widely distributed throughout the peripheral nervous system the NPY that is released from the adrenal medulla is thought to act locally because the plasma levels of NPY show little change following adrenalectomy (Castagne et al. 1987; Morris et al. 1987; Bernet et al. 1998) but see (Evequoz et al. 1995). Consistent with a restricted site of action, functional NPY receptors are present on chromaffin cells (Zheng et al. 1997; Zhang et al. 2000; Cavadas et al. 2001; Rosmaninho-Salgado et al. 2007). Stimulation of these receptors has been variously reported to increase (Cavadas et al. 2001; Rosmaninho-Salgado et al. 2007) or decrease catecholamine release (Higuchi et al. 1988; Hexum et al. 1994) suggesting that NPY could also have a short-term feed-forward role in the stress response. Because the forced swim test leads to activation of chromaffin cells as monitored by changes in c-fos-ir (Wang et al., submitted) it seems likely that adrenal NPY contributes to the modulation we have described. However, we do not eliminate the possibility that other neuronal sources of NPY are also involved.
Using the in vivo injection of antagonists we find that Y2 receptors selectively regulate the adrenal expression of NPY. Thus, injection of BIIE0246 increased the basal levels of NPY-ir. Y2 receptors are often found pre-synaptically (Qian et al. 1997; Sun et al. 2001; Fu et al. 2004; Stanic et al. 2006) but because a Y2-dependent increase in intracellular calcium has been reported for human chromaffin cells in vitro (Cavadas et al. 2001), it is possible that these receptors are also located post-synaptically in the adrenal gland. Nevertheless, our results do not exclude the possibility that the relevant receptors are located outside the adrenal. This subtype specificity is curious because the adrenal medulla contains at least three classes of Y receptors and these are generally thought to couple to the same downstream signaling pathways (see Michel et al. 1998). In conditional, but not germline, Y2 knockout mice there is an increase in the level of NPY mRNA (Naveilhan et al. 1999; Sainsbury et al. 2002) which is consistent with a negative feedback mechanism controlling NPY expression. Of particular interest is a study in which Y2 receptors were selectively deleted from NPY-expressing hypothalamic neurons. In these animals Y2 knockout led to an increase in the level of NPY mRNA in the arcuate nucleus (Shi et al. 2010). Our finding that acute inhibition of Y2 receptors leads to an elevation in adrenal NPY-ir suggests that a Y2-mediated autocrine regulation of NPY expression may be a widespread mechanism that modulates the synthesis of this peptide.
In conclusion, we find that a feedback loop alters the level of NPY expression in the adrenal gland under both basal and stress-induced conditions. This appears to be an efficient system that ensures the levels of this peptidergic modulator are maintained even in the face of sustained sympathetic activity.
We thank Dr. June Liu for critically reading the manuscript and Dr. Ya-Ping Tang for the use of equipment. MDW current address; Department of Cell Biology and Anatomy, LSU Health Sciences Center, 1901 Perdido Street, New Orleans, LA 70112. QW current address; Division of Hypothalamic Research, University of Texas Southwestern Medical Center, Dallas, TX 75390-9077. This study was supported by a grant from the National Institutes of Health (RO1DK080441) to MDW. The authors have no conflicts of interests to declare.