1. Neuroscience
Download icon

Antagonistic modulation of NPY/AgRP and POMC neurons in the arcuate nucleus by noradrenalin

  1. Lars Paeger
  2. Ismene Karakasilioti
  3. Janine Altmüller
  4. Peter Frommolt
  5. Jens Brüning
  6. Peter Kloppenburg  Is a corresponding author
  1. University of Cologne, Germany
  2. Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases, Germany
  3. Max Planck Institute for Metabolism Research, Germany
  4. University Hospital Cologne, Germany
Research Article
  • Cited 15
  • Views 2,132
  • Annotations
Cite this article as: eLife 2017;6:e25770 doi: 10.7554/eLife.25770

Abstract

In the arcuate nucleus of the hypothalamus (ARH) satiety signaling (anorexigenic) pro-opiomelanocortin (POMC)-expressing and hunger signaling (orexigenic) agouti-related peptide (AgRP)-expressing neurons are key components of the neuronal circuits that control food intake and energy homeostasis. Here, we assessed whether the catecholamine noradrenalin directly modulates the activity of these neurons in mice. Perforated patch clamp recordings showed that noradrenalin changes the activity of these functionally antagonistic neurons in opposite ways, increasing the activity of the orexigenic NPY/AgRP neurons and decreasing the activity of the anorexigenic POMC neurons. Cell type-specific transcriptomics and pharmacological experiments revealed that the opposing effect on these neurons is mediated by the activation of excitatory α1A - and β- adrenergic receptors in NPY/AgRP neurons, while POMC neurons are inhibited via α2A – adrenergic receptors. Thus, the coordinated differential modulation of the key hypothalamic neurons in control of energy homeostasis assigns noradrenalin an important role to promote feeding.

https://doi.org/10.7554/eLife.25770.001

Introduction

Substantial progress has been made identifying central neuronal circuits that are crucial for controlling feeding behavior and energy homeostasis. Understanding the cellular and molecular mechanisms mediating information processing in neuronal circuits that regulate energy homeostasis is essential to devise specific pharmacological strategies, which assist obese patients in starting and maintaining a program of weight loss with no or minimized side effects on other neuronal systems (Gouni-Berthold et al., 2013; Holes-Lewis et al., 2013; Misra, 2013). Important microcircuits in control of energy balance are localized in the arcuate nucleus of the hypothalamus (ARH), which contains two main neuron populations with antagonistic functions: Neuropeptide Y and agouti-related peptide (NPY/AgRP) co-expressing neurons signaling hunger, promoting food intake, and pro-opiomelanocortin (POMC) - expressing neurons which signal satiety (Sohn et al., 2013; Varela and Horvath, 2012).

Many important fuel-sensing endocrine and metabolic factors, such as insulin, leptin, glucose, serotonin, dopamine, free fatty acids and uridine diphosphate (Gao and Horvath, 2007; Jo et al., 2009; Jordan et al., 2010; Sohn et al., 2011; Steculorum et al., 2015; Varela and Horvath, 2012; Zhang and van den Pol, 2016) have been identified and various immediate actions of these modulators and nutrient components on the POMC and AgRP neurons are well understood (Belgardt et al., 2008; Claret et al., 2007; Jo et al., 2009; Parton et al., 2007; Spanswick et al., 1997, 2000). However, the effect of the catecholamine noradrenalin (NA) on the ARH is not well defined. Noradrenalin is associated with a number of important CNS functions (Szabadi, 2013) including energy homeostasis and food intake (Commins et al., 1999; Ste Marie et al., 2005; Thomas and Palmiter, 1997; Wellman, 2005, 2000). Potentially, NA could be a direct modulator of the energy homeostasis-regulatory POMC- and AgRP neurons in the ARH, since noradrenergic projections from the hindbrain (A1, A2 noradrenergic cell groups) and the locus coeruleus (LC) to the ARH have been described in rats (Fraley, 2006; Fraley and Ritter, 2003; Yoon et al., 2013) and manipulation of the noradrenergic system in the hypothalamus has been found to modify feeding behavior and food intake. Impairment of the dorsal or the ventral noradrenergic bundle, which contain efferent projections from the LC to the forebrain, caused hyper- or hypophagia, respectively (Ahlskog and Hoebel, 1973; Hoebel et al., 1989). In line with these lesion experiments, local NA injections into defined hypothalamus regions either elicits or inhibits feeding in rodents (Booth, 1967; Leibowitz, 1978). These findings not only demonstrate noradrenergic modulation of food intake, but also indicate that the NA effect on feeding behavior is versatile and site-specific. Additional evidence that NA is associated with the regulation of the homeostatic system in the ARH came from extracellular recordings of unidentified ARH neurons in rats, which responded to NA application (Kang et al., 2000). Furthermore, NA increases the expression of Agrp and Npy mRNA in the rat ARH in response to glucoprivation (Fraley and Ritter, 2003). This effect is abolished if NA terminals in the ARH are eliminated by the injection of saporin-conjugated anti-dopamine-β-hydroxylase (anti-dβh), a selective NA immunotoxin. Specifically, the immunotoxin also eliminated a large number of neurons in the ventrolateral medulla (A1/C1 cell groups), which have been identified as mediators of glucoprivic feeding. Together, these findings provide strong evidence that NA modulates eating behavior by changing functional properties of ARH neurons.

Recent work by the Bouret lab provides a plausible hypothesis in regard to the question in which behavioral context the noradrenergic LC projections into the ARH play a key role and what response they facilitate. Based on in vivo electrophysiological recordings in primates, it is hypothesized that the LC plays a key role in mobilizing the required energy to face anticipated physical challenges, such as to obtain rewards (Bouret and Richmond, 2015; Varazzani et al., 2015). In this model, the LC activity is directly related to the energetic need that is required to master the challenge. NA release in the ARH might then promote compensatory energy intake.

On the cellular level NA action can be mediated by a variety of ARs that can inhibit or excite neurons. NA and adrenalin activate three major classes of G protein-coupled receptors (GPCRs): α1-, α2- and β-ARs (Hein, 2006). These receptor types can be further subdivided in a number of different isoforms. The intracellular signaling cascades of ARs are well investigated. While excitatory α1-ARs are coupled to Gq/11 and mediate an increase in intracellular Ca2+ levels and closure of G protein-coupled inwardly rectifying potassium channels (GIRKs), inhibitory α2-ARs couple to Gi/o leading to inhibition of voltage-gated Ca2+ channels (VGCCs) and opening of GIRKs. β-ARs mediate excitation via activation of Gs proteins and subsequent elevation of internal Ca2+ concentrations and opening of VGCCs (Marzo et al., 2009).

In this study, we assessed if NA can directly modulate orexigenic NPY/AgRP and anorexigenic POMC neurons, which are key components of the control circuits in ARH that regulate food intake and energy homeostasis. In electrophysiological recordings, NA had clear opposite effects on these neurons, increasing the activity of NPY/AgRP neurons and inhibiting POMC neurons. Cell type-specific transcriptomics and pharmacological experiments revealed that the activation of NPY/AgRP neurons is predominantly mediated by α1A-ARs, while POMC neurons are inhibited via α2A-ARs. Collectively, our data indicate a strong orexigenic influence of NA on the ARH circuitry that controls energy balance.

Results

Here we asked, if and how NA directly modulates the activity of NPY/AgRP and POMC expressing neurons of the ARH. This question was addressed in three steps. First, electrophysiological recordings were performed to analyze the modulatory effect of NA on these neurons. Second, we used cell type-specific transcriptomics to build an inventory of adrenergic receptors that are expressed in NPY/AgRP and POMC neurons and potentially could mediate the modulation of these neurons. Third, pharmacological tools were used to test which adrenergic receptors are functionally expressed and how they contribute to the net modulatory effect of NA.

Noradrenalin modulates electrophysiological activity differentially in NPY/AgRP and POMC neurons

In the first set of experiments, we probed if NA modulates the electrophysiological activity of NPY/AgRP and POMC neurons. To address this question directly, we analyzed identified NPY/AgRP-GFP and POMC-GFP neurons of the ARH in acute brain slices of adult mice. Recordings were performed in the perforated patch clamp configuration to minimize effects on intracellular signaling pathways.

As described previously, most NPY/AgRP and POMC neurons were spontaneously active under control conditions (NPY/AgRP neurons: (Baver et al., 2014; Steculorum et al., 2015; Tsaousidou et al., 2014; Wei et al., 2015); POMC neurons: (Ernst et al., 2009; Koch et al., 2015; Newton et al., 2013; Plum et al., 2006; Zhang and van den Pol, 2013). Bath-applied 10 µM NA had clear opposite effects on these functionally antagonistic neuron types. NA depolarized orexigenic NPY/AgRP neurons and increased their action potential frequency from 2.0 ± 0.9 Hz to 5.0 ± 1.0 Hz (Figure 1A,B; NPY/AgRP neurons, Wilcoxon matched pairs signed rank test, n = 8, p=0.0012). In contrast, NA clearly hyperpolarized anorexigenic POMC neurons and drastically reduced or even abolished spontaneous generation of action potentials from 3.5 ± 1.1 to 0.7 ± 0.7 Hz (Figure 1A,B; POMC neurons, Wilcoxon matched pairs signed rank test, n = 8, p=0.0156). In both neuron types, the NA effect had a rapid onset and was readily reversible. To better define and understand the cause-effect relationship of the NA modulation, we measured concentration-response curves for both neuron types.

Noradrenalin differentially modulates functionally antagonistic NPY/AgRP and POMC neurons.

NA (10 µM) excited orexigenic NPY/AgRP neurons and inhibited anorexigenic POMC neurons. Original recordings (A) and averaged responses (B) of NPY/AgRP (n = 8) and POMC neurons (n = 8) during NA application. The mean response is expressed as change in action potential frequency. The boxplots show the absolute change in action potential frequency for both neuron populations. **p<0.01, Wilcoxon matched pairs signed ranks test.

https://doi.org/10.7554/eLife.25770.002

NPY/AgRP neurons, NA concentration-response relation

NA depolarized the membrane potential and increased the action potential frequency starting at a concentration of 100 nM (Figure 2A). The maximal responses were observed at 10 µM where NA increased the average firing frequency from 2.2 ± 0.8 Hz to 4.7 ± 0.8 Hz (Figure 2B; paired t-test, n = 8, p<0.0001) and depolarized the membrane by 5.1 ± 1.0 mV (Figure 2C; paired t-test, n = 8, p=0.0011). The normalized concentration-response relations for the membrane depolarization and increase in firing frequency were well fit with a sigmoidal relation (Equation 1). The concentration-response relations had EC50s of 1.9 µM (1.1–3.2 µM; n = 8) and 1.5 µM (0.8–2.8 µM; n = 8), respectively. The effective concentration range is in line with previous studies that analyzed NA modulation in other cell types (Hayar et al., 1997). To rule out that the observed NA effect is caused by secondary modulation, which is mediated by neurons that are presynaptic to NPY/AgRP neurons, we analyzed the NA effect on NPY/AgRP neurons in which GABAergic and glutamatergic input was pharmacologically blocked. Synaptically isolated NPY/AgRP neurons responded similarly to the application of NA as without synaptic blockers (Figure 2D,E), demonstrating that the excitatory NA effect is mediated by cell-intrinsic ARs and the applied GABA and glutamate antagonists do not directly interfere with the AR.

Figure 2 with 1 supplement see all
Noradrenergic modulation of NPY/AgRP neurons is concentration dependent.

(A) Recording of a NPY/AgRP neuron demonstrating the effect of increasing NA concentrations. (B) and (C) Concentration - response relations showing the NA effect on action potential frequency (B) and membrane potential (C). The curves fit to a sigmoidal relation (Equation 1). NA had an EC50 of 1.9 µM (1.1–3.2 µM; n = 8) for the AP frequency and 1.5 µM (0.8–2.8 µM; n = 8) for the membrane potential, respectively. (D) and (E) The NA modulation of NPY/AgRP neurons is direct and not dependent on synaptic input. Original recording (D) and averaged effect on action potential frequency (E) showing that the NA effect on NPY/AgRP neurons is not changed when glutamatergic and GABAergic synaptic input is blocked. Experiments in (E) were performed consecutively with the same set of neurons (n = 8). Control 1 and control two refers to the different starting conditions, that is preincubation with or without synaptic blockers. ***p<0.001, one-way ANOVA with post hoc Tukey analysis.

https://doi.org/10.7554/eLife.25770.003

POMC neurons, NA concentration-response relation

10 µM NA clearly inhibited POMC neurons (Figure 1A–C). In neurons with low control action potential frequencies, this often caused an all or nothing effect. Therefore, we used the membrane potential and membrane conductance densities to quantify the NA effect and to establish concentration-response relations.

The NA inhibition started at 100 nM and the maximal inhibitory effect was reached at 10 µM where NA hyperpolarized the membrane potential by 18.6 ± 3.2 mV (Figure 3B; paired t-test, n = 8; p=0.0007) and increased the conductance from 647 ± 106 pS to 1546 ± 266 pS (Figure 3C; paired t-test, n = 8; p=0.0029). The NA effect developed rapidly and was reversible. However, after application of high concentration, the wash out of NA induced a slow, sustained (~10 min) rebound excitation. NA concentrations higher than 10 µM did not further increase the effect.

Noradrenergic modulation of POMC neurons is concentration dependent.

(A) Recording of a POMC neuron demonstrating the effect of increasing NA concentrations. (B) and (C) Concentration - response relations showing the NA effect on membrane potential (B) and conductance density (C). The curves are fits to a sigmoidal relation (Equation 1). NA had an EC50 of 0.9 µM (0.6–1.5 µM; n = 8) for the membrane potential and 1.3 µM (1.0–1.9 µM; n = 8) for the conductance density, respectively. (D) and (E) The NA modulation of POMC neurons is direct and not dependent on synaptic input. Original recording (D) and averaged effect on conductance density (E) showing that the NA effect on POMC neurons is not changed when glutamatergic and GABAergic synaptic input is blocked. Experiments shown in (E) were performed consecutively with the same set of neurons (n = 4). Control 1 and control two refers to the different starting conditions, that is pre-incubation with or without synaptic blockers. **p<0.01; ***p<0.001; one-way ANOVA with post hoc Tukey analysis.

https://doi.org/10.7554/eLife.25770.005

The normalized concentration-response relations for conductance increase and membrane hyperpolarization were well fit with a sigmoidal relation (Equation 1), with EC50s of 0.9 µM (0.6–1.5 µM; n = 8) and 1.3 µM (1.0–1.9 µM; n = 8) for the hyperpolarization and the conductance density, respectively. The effective concentration range is in line with previous studies that analyzed NA modulation in other cell types (Jurgens et al., 2007).

In experiments in which the GABAergic and glutamatergic input of the POMC neurons was pharmacologically blocked, the NA response was slightly decreased (~16%) compared to recordings without synaptic blockers present, suggesting a small presynaptic contribution by NA responsive neurons, which project on POMC neurons (Figure 3D,E). Given the excitatory effect of NA on NPY/AgRP neurons, this effect is likely caused by their mono-directional inhibitory synaptic innervation of POMC neurons.

Adrenergic receptors are expressed in NPY/AgRP and POMC neurons of the ARH

To understand how NA modulates NPY/AgRP and POMC neurons at the molecular level, we performed cell type-specific transcriptomics to analyze AR gene expression in these neurons. Using enhanced green fluorescent protein (eGFP)-based fluorescence-activated cell sorting (FACS), we isolated eGFP-expressing NPY/AgRP or POMC neurons from coronal slices, which contained the hypothalamus, of 6-week-old mice (four mice per sample; NPY-GFP: n = 3; POMC-GFP: n = 2). Subsequently, we performed total RNA isolation and whole transcriptome analysis from these neuronal populations. Comparing POMC versus NPY transcriptional profiles yielded 4482 significantly differentially expressed genes (FC≥2, p≤0.01). Within this pool, Pomc was enriched 147-fold in the corresponding population, with Npy and Agrp being higher expressed in NPY-derived samples, but not in a statistically significant way (Figure 4A, supplemental file 1). NPY/AgRP neurons expressed all subtypes of ARs, except α1D- and α2B- ARs. POMC neurons only expressed α1A- and α2A- ARs, albeit to a lower magnitude (Figure 4A).

Figure 4 with 1 supplement see all
Cell type-specific expression of ARs.

(A) Log expression levels of adrenergic receptor genes in Pomc- and Npy-expressing cell populations. Deeper red colours indicate higher expression levels in the respective cell population. RPKM: reads per kilobase per million mapped reads. (B) Images from RNA in situ hybridizations against Adra1a (upper panels, green dots) and Adra2a (lower panels, green dots) in Npy- and Pomc- expressing (red) neurons. White arrowheads indicate doubly-labeled cells. Higher magnification indexes are representative of doubly-labeled cells. (C) Quantification of double positive cells for Adra1a and Adra2a in Npy- and Pomc- expressing neurons.

https://doi.org/10.7554/eLife.25770.006

To verify our gene expression data related to the ARs expressed in these neuronal populations, we performed RNA in situ hybridizations, using probes against Adra1a, Adra2a, Adrb1 and Adrb2. Both NPY/AgRP and POMC neurons were positive for Adra1a and Adra2a expression, with Adra1a being expressed in more NPY/AgRP neurons, compared to POMC neurons (Figure 4B,C). Adra2a was however expressed to a similar extent between the two populations (Figure 4C). Adrb1 and Adrb2 were expressed only in NPY/AgRP neurons (Figure 4—figure supplement 1). Taken together, our data suggest that ARs are differentially expressed between NPY and POMC neurons within the hypothalamus, possibly mediating any distinct modulatory effects of NA on these populations.

NA excites NPY/AgRP neurons predominantly via the activation of α1A-ARs

While NPY/AgRP neurons expressed excitatory α1A,B- and β-ARs, and inhibitory α2A -AR, NA clearly excited all recorded NPY/AgRP neurons. To assess the AR classes that contribute to the excitatory electrophysiological response we used pharmacological tools, which were selective for AR subclasses.

In the first series of experiments α-ARs were blocked by the α1-AR antagonist prazosine (5 µM) and the α2-AR antagonist yohimbine (5 µM; Figure 5A,B). During the combined application of prazosine and yohimbine the excitatory NA effect was completely blocked in 44% (4 of 9; Figure 5A) and significantly (~60%) attenuated in 55% (5 of 9; Figure 5B) of the NPY/AgRP neurons. Together, these experiments suggest that the excitatory NA effect in NPY/AgRP neurons can be mediated solely by α1-ARs or by the co-activation of α1-ARs and β-ARs.

The excitatory noradrenergic effect on NPY/AgRP neurons is predominantly mediated by α1A-ARs.

(A) and (B) In 44% (4 out of 9) of the NPY/AgRP neurons the NA (10 µM) effect was completely blocked by the α1-AR antagonist prazosine (5 µM) and α2-AR antagonist yohimbine (5 µM) (A), while in 56% (5 out of 9) of the NPY/AgRP neurons the NA effect was markedly reduced (~60%) but not completely blocked (B). The series of experiments shown in (A) and (B) were each performed consecutively with the same set of neurons. n values are given in brackets. Control 1 and control two refers to the different starting conditions, that is pre-incubation with different AR antagonist. **p<0.01; ***p<0.001; one-way ANOVA with post hoc Tukey analysis.

https://doi.org/10.7554/eLife.25770.008

Next, we asked which α1-AR subtype(s) mediates the excitatory NA effect. The selective α1A-AR agonist A-61603 (1 µM) depolarized the membrane potential and increased the AP frequency in NPY/AgRP neurons similarly as a subsequent NA application (Figure 6A,B). Both the A-61603 and the NA effect were fully reversible. In line with these data, we found that the selective α1A-AR antagonist WB4101 (100 nM) abolished the excitatory NA response (Figure 6C,D). Application of WB4101 resulted in an inhibitory effect, suggesting baseline activation of α1A-AR. The subtype-specific α1B,D-AR antagonist chloro-ethyl-clonidine (CEC; 100 nM) did not block or alter the NA action (Figure 6E,F), suggesting that α1B,D-ARs do not contribute to the excitation. Taken together, these data show that NA excites NPY/AgRP neurons via activation of α1A-ARs.

Figure 6 with 1 supplement see all
NPY/AgRP neurons are excited by α1A-AR and inhibited by α2A-AR when α1A-AR are blocked.

(A) Rate histogram with its original recording showing that the α1A-AR agonist A61603 (1 µM) had similar excitatory effects on action potential frequency (B) as NA. (C) and (D) The selective α1A-AR antagonist WB4101 (100 nM) blocked the excitatory NA effect. In the presence of WB4101 NA had an inhibitory effect on the membrane potential, which was eliminated by the α2A-AR antagonist BRL44408 (10 µM). (E) The α1BD-AR blocker CEC (100 nM) did not change the NA effect on NPY/AgRP neurons. *p<0.05; **p<0.01; ***p<0.001; one-way ANOVA with posthoc Tukey analysis.

https://doi.org/10.7554/eLife.25770.009

Of note, the α1A-AR antagonist WB4101 not only blocked the NA excitation (as described above). In the presence of WB4101 NA even had an inhibitory effect on the NPY/AgRP membrane potential (Figure 6C,D). A subsequent block of inhibitory α2A-ARs by the specific antagonist BRL44408 reversed the inhibition to a small excitation. These data show that the inhibitory α2A-ARs are functional in NPY/AgRP neurons and also support the notion that excitatory β-ARs are activated by NA, as indicated in the first set of experiments.

Collectively, our data indicate that all receptors, which were identified by cell-type specific transcriptomics, are functional in NPY/AgRP neurons. The excitatory NA effect that we observed in virtually all neurons is predominantly mediated by α1A-ARs and might be in some neurons co-mediated by β-ARs. The activation of excitatory AR interferes with the activation of inhibitory α2A-ARs.

NA inhibits POMC neurons via the activation of α2A-ARs

POMC neurons express excitatory α1A-ARs and inhibitory α2A-ARs, and were markedly inhibited by NA. First, we tested, if the inhibitory NA effect is mediated by α2A-AR, as suggested by our transcriptomics data. The α2A-AR antagonist BRL 44408 (10 µM) blocked the NA mediated inhibitory effect. Application of 10 µM BRL 44408 alone, did not significantly alter the recorded electrophysiological properties, suggesting no baseline activity of α2A-ARs (Figure 7A). The specific α2B-AR antagonist ARC 239 (1 µM) did not change the inhibitory effect of NA, which suggests that the inhibitory NA action is solely mediated by α2A-ARs (Figure 7B).

The inhibitory noradrenalin effect on POMC neurons is mediated by α2A-AR.

(A) and (B) The α2A-antagonist BRL 44408 (10 µM) blocked the inhibitory NA (5 µM) effect (A), while the α2A-antagonist ARC239 (1 µM) did not affect the NA action (B). (C–E) Inhibitory NA effect (C). Blocking α2-AR by yohimbine revealed an excitatory NA effect (D), which was blocked by the α1-AR antagonist prazosine (E). The series of experiments shown in (A) and (B) were each performed consecutively with the same set of neurons. Control 1 and control two refers to the different starting conditions, that is preincubation with different AR antagonist. n values are given in brackets. *p<0.05; **p<0.01; ***p<0.001. (A) and (B) one-way ANOVA with post hoc Tukey analysis; (C–E) paired t-test.

https://doi.org/10.7554/eLife.25770.011

Next, we asked if the excitatory α1A-ARs are functional in POMC neurons (Figure 7C–E). When the inhibitory α2-ARs were blocked by yohimbine (5 µM), NA excited POMC neurons (Figure 7D; paired t-test, n = 10, p=0.0168). This excitation was completely abolished by the α1-AR antagonist prazosine (5 µM), showing the functionality of α1-ARs in POMC neurons (Figure 7E; paired t-test, n = 6, p=0.2467).

Taken together, our results show that the NA-induced inhibition of POMC neurons is mediated by α2A-AR and can mask the activation of excitatory α1-ARs.

Discussion

NA neurons of the brainstem give rise to a complex system of fiber tracts innervating most of the forebrain including the hypothalamus and the ARH (Szabadi, 2013). Because of the extensive and complex projection pattern, the NA system is associated with a number of important CNS functions including control of wake-sleep-cycles, nociception and memory consolidation (Berridge and Waterhouse, 2003; Samuels and Szabadi, 2008a, 2008b). One important role that has been clearly linked to the NA system by a combination of anatomical, pharmacological, lesion and behavioral studies is the modulation of energy homeostasis and food intake (Wellman, 2005, 2000)

Here, we show for the first time that NA directly modulates NPY/AgRP and POMC neurons of the ARH, which are key components of the neuronal circuits that control food intake and energy homeostasis. Interestingly, NA changes the activity of these functionally antagonistic neurons in opposite ways, increasing the activity of the orexigenic NPY/AgRP neurons and decreasing the activity of the anorexigenic POMC neurons. The strong differential modulation, which was reproducible in virtually all recorded neurons, is remarkable, since other important fuel sensing modulators like insulin, leptin, glucose, oleic acid and UDP typically affect only a subpopulation of these neurons (Jo et al., 2009; Parton et al., 2007; Steculorum et al., 2015; Williams et al., 2010). On the network and systems level, this clear differential NA modulation of functionally antagonistic neuron types should generate a strong orexigenic drive on the homeostatic circuits of the ARH. This conclusion is consistent with early behavioral experiments from Clark et al. (1988) showing that injection of NA into the third ventricle increased food intake. Furthermore, the authors showed that blockade of α2-AR attenuated the NA-induced feeding behavior, which is in line with our finding that satiety signaling POMC- neurons are inhibited by NA via α2-AR.

Since the modulatory effects of NA were observed in synaptically isolated neurons, the data suggest that NA directly activated intrinsically expressed ARs. Our cell type-specific transcriptomic data showed expression of excitatory and inhibitory receptor subtypes in both NPY/AgRP and POMC neurons, which is in line with recent data from Henry et al. (2015). NPY/AgRP neurons expressed excitatory α1A-, β-ARs and inhibitory α2A-ARs, while POMC neurons only express excitatory α1A-ARs and inhibitory α2A-ARs. Expression of other AR-subtypes was not detected in these neurons.

To unravel which of the expressed AR-subtypes contribute to the cell type-specific physiological effects of NA on NPY/AgRP and POMC neurons, we used pharmacological tools that have been previously shown to be selective for AR subtypes. While in NPY/AgRP the robust excitation is mediated by excitatory α1A- and β-ARs, α2A-ARs mediated the strong inhibition in POMC neurons.

The functional consequences of these diverse expression and activation profiles on the systems and behavioral level is not entirely clear and has to be investigated in future studies. Previous studies in other neuron types revealed similar phenomena and suggested that simultaneous expression of excitatory and inhibitory (NA) receptors might derive from developmental effects or localized, differential expression of ARs in different functional compartments.

Simultaneous expression of excitatory and inhibitory adrenergic receptors has been found in neurons of the LC. In juvenile mice, mRNA for excitatory α1A,B,D- and inhibitory α2A,B,C-ARs has been detected and application of NA caused excitation, while in adult mice NA hyperpolarizes LC neurons. Here, α1-ARs attenuate the NA-induced hyperpolarization (Osborne et al., 2002). As in LC neurons, POMC neurons might change their response to NA during development. In the ARH developmental changes of electrophysiological responses to leptin have been described for NPY/AgRP neurons (Baquero et al., 2014; Nilsson et al., 2005). Simultaneous expression of excitatory and inhibitory ARs has also been found in dentate neurons of the human cerebellum (Schambra et al., 2005). Here, α2B-ARs are thought to modulate excitatory glutamatergic input, while α1A,B-ARs may play a role in timing and computation of inhibitory input. Previous studies in dentate neurons of the cerebellum suggested that α1-ARs are expressed on the cell body while α2-ARs are expressed on the terminals, in order to locally regulate transmitter release (Milner et al., 2000).

This raises the question regarding the NA source and under which physiological conditions NA is active to modulate food intake. While the blood brain barrier in adult animals is considered mostly impermeable for catecholamines, there is evidence from early work that catecholamines can cross the blood brain barrier in the hypothalamus (Weil-Malherbe et al., 1959). However, the site specific effects of local lesions and local NA injections into the hypothalamus might argue for local, site specific NA release from neurons. In rats, projections from the hindbrain/brainstem and the LC to the ARH have been identified (Fraley, 2006; Fraley and Ritter, 2003; Yoon et al., 2013). Specific deletion of projections from the ventro-lateral medulla (A1/C1 cell group) is sufficient to eliminate an increase in AgRP and NPY mRNA expression in response to glucoprivation induced by 2-DG. Furthermore, homeostatic feeding in response to local hypoglycemia in the hindbrain was decreased (Fraley and Ritter, 2003). In this context, recent work in primates from the Bouret group suggested that the LC plays a key role in mobilizing the required energy in order to face anticipated physical challenges, such as obtaining rewards (Bouret and Richmond, 2015; Varazzani et al., 2015). In this model, the LC activity is directly related to the energetic need that is required to master the challenge. Consistent with this hypothesis, our data suggest that NA release into the ARH upon LC activation generates an orexigenic drive to compensate for the required energy. In line with this hypothesis, previous studies revealed that NA injections into the hypothalamus elicit food intake (Booth, 1967; Leibowitz, 1978).

Together, these behavioral experiments are strong evidence for the orexigenic nature of NA signaling in the control of food intake. In this context, the orexigenic hormones insulin and leptin may modulate NA signaling in the ARH. While insulin selectively decreased α2-AR expression in the ARH and dorsomedial hypothalamus (Levin et al., 1998), leptin inhibits NA release in rats (Brunetti et al., 1999; Francis et al., 2004), Both effects would augment the net anorexigenic signal. In support of the latter relation, obese mice that lack the gene to produce leptin (Lepob/ob) exhibit increased NA levels within the hypothalamus, which further decreases the anorexigenic signal (Oltmans, 1983).

Several drugs that were developed to modify food intake, act on the monoaminergic transmitter system in the brain. Since the molecular and cellular targets that mediate the effects of these drugs are often not clearly identified, they often cause strong side effects (Hainer et al., 2006). To successfully support obese patients to manage their food intake with minimized side effects, it is important to precisely define the targets of these drugs. Since the current findings reveal cell type-specific pathways that mediate NA modulation in the ARH they might help to better define specific targets for anti-obesity drugs.

Materials and methods

Animal care

Request a detailed protocol

All animal procedures were conducted in compliance with guidelines approved by local government authorities (Bezirksregierung Köln, Cologne, Germany) and were in accordance with NIH guidelines. Mice were housed at 22–24°C with a 12 hr light/12 hr dark cycle. Animals had access to water and chow ad libitum. All experiments have been performed with adult male and female NPY/AgRP-hrGFP (Lowell 2009) or POMC-EGFP (Cowley 2001) mice. The mice for the electrophysiological recordings were 15–20 weeks old. To facilitate tissue dissociation 6 weeks old animals were used for RNA sequencing.

Neuronal sorting and RNA sequencing

Request a detailed protocol

At 6 weeks of age, NPY/AgRP-hrGFP or POMC-EGFP mice were sacrificed by decapitation and the brain area corresponding to the hypothalamus was micro-dissected under a brightfield stereoscope. Each sample contained hypothalami from four mice (males and females) and tissue samples were processed in ice-cold HBSS, until enzymatic treatment. Samples were centrifuged to remove HBSS and were subsequently incubated for 15 min in pre-incubation buffer, at 37°C. Enzyme mix was added and each sample was manually dissociated by gentle pipetting, throughout a total duration of 35 min and according to manufacturer’s instructions (Neural tissue dissociation kit, Miltenyi Biotec, Germany). Subsequently, samples were centrifuged at 500 x g, 10 min, 4°C, and the cell pellet was re-suspended in cold PBS, containing 5% fetal calf serum.

NPY/AgRP-hrGFP or POMC-EGFP-expressing neurons were gated and sorted with a blue laser. GFP-positive neurons were sorted directly into total RNA extraction buffer (PicoPure RNA isolation kit, ThermoFisher Scientific, USA). Samples were lysed for 30 min by incubation at 42°C and subsequently frozen to −80°C, to allow for simultaneous processing among replicates. Total RNA was extracted according to manufacturer’s instructions (PicoPure RNA isolation kit, ThermoFisher Scientific, USA) and all samples were DNAse-treated (ThermoFisher Scientific, USA). RNA quality was assessed with an Agilent 2100 bioanalyzer and samples with low degradation (RIN >7.5) were further processed for library preparation and sequencing. Due to low amount of input material, pre-amplification using the Ovation RNASeq System V2 was performed. Total RNA was used for first strand cDNA synthesis, using both poly(T) and random primers, followed by second strand synthesis and isothermal strand-displacement amplification. For library preparation, the Illumina Nextera XT DNA sample preparation protocol (Part # 15031942 Rev. C) was used, with 1 ng cDNA input. After validation (Agilent 2200 TapeStation) and quantification (Invitrogen Qubit System) all five transcriptome libraries were pooled. The pool was quantified using the Peqlab KAPA Library Quantification Kit and the Applied Biosystems 7900HT Sequence Detection. One lane and a paired-end read of 2 × 100 bp on the Illumina**HiSeq 2000 sequencer using v3 chemistry resulted in 39-44Mreads/sample (7.9–9.7 Gb) and a ratio of bases above Q30 of 90.2%. RNA-seq data were analyzed using the QuickNGS pipeline, described elsewhere (Wagle et al., 2015) (n = 3 for NPY/AgRP- and n = 2 for POMC-expressing neurons).

Statistically significant (p<0.01) altered genes between the two neuronal populations were analyzed with the Ingenuity Pathway Analysis (IPA) software for gene ontology (GO) enrichment, using either the up-regulated or the down-regulated enrichment mode of the core IPA analysis (POMC-neurons were used as controls – down-regulated pool – and NPY-neurons as samples – up-regulated pool).

Electrophysiology

Brain slice preparation

Request a detailed protocol

The animals were lightly anesthetized with isoflurane (B506; AbbVie Deutschland GmbH and Co KG, Ludwigshafen, Germany) and subsequently decapitated. The brain was rapidly removed and a block of tissue containing the hypothalamus or brainstem was immediately cut out. Coronal slices (270–300 μm) were cut with a vibration microtome (HM-650 V; Thermo Scientific, Walldorf, Germany) under cold (4C), carbogenated (95% O2 and 5% CO2), glycerol-based modified artificial cerebrospinal fluid (GaCSF; Ye et al., 2006 ) to enhance the viability of neurons. GaCSF contained (in mM): 250 Glycerol, 2.5 KCl, 2 MgCl2, 2 CaCl2, 1.2 NaH2PO4, 10 HEPES, 21 NaHCO3, 5 Glucose and was adjusted to pH 7.2 with NaOH resulting in an osmolarity of ~310 mOsm. Brain slices were transferred into carbogenated artificial cerebrospinal fluid (aCSF). aCSF contained (in mM): 125 NaCl, 2.5 KCl, 2 MgCl2, 2 CaCl2, 1.2 NaH2PO4, 21 NaHCO3, 10 HEPES, and 5 Glucose and was adjusted to pH 7.2 with NaOH resulting in an osmolarity of ~310 mOsm. First, they were kept for 20 min in a 35°C’ recovery bath’ and then stored at room temperature (24°C) for at least 30 min prior to recording. For the recordings, slices were transferred to a Sylgard-coated (Dow Corning Corp., Midland, MI, USA) recording chamber (~3 ml volume) and, if not mentioned otherwise, continuously perfused with carbogenated aCSF at a flow rate of ~2 ml · min−1. Recordings in the ARC were made at 24°C using an inline solution heater (SH27B; Warner Instruments, Hamden, CT, USA) operated by a temperature controller (TC-324B; Warner Instruments).

Patch clamp recordings

Request a detailed protocol

Current-clamp recordings of NPY/AgRP- and POMC-expressing neurons in the ARH of 15–20 week old mice were performed in the perforated patch clamp configuration. Neurons were visualized with a fixed stage upright microscope (BX51WI, Olympus, Hamburg, Germany) using 40× and 60× water-immersion objectives (LUMplan FL/N 40×, 0.8 numerical aperture, 2 mm working distance; LUMplan FL/N 60×, 1.0 numerical aperture, 2 mm working distance, Olympus) with infrared differential interference contrast optics (Dodt and Zieglgänsberger, 1990) and fluorescence optics. POMC and NPY/AgRP neurons were identified by their anatomical location in the ARH and by their GFP fluorescence that was visualized with an X-Cite 120 illumination system (EXFO Photonic Solutions, Ontario, Canada) in combination with a Chroma 41001 filter set (EX: HQ480/40x, BS: Q505LP, EM: HQ535/50m, Chroma, Rockingham, VT, USA). Electrodes with tip resistances between 4 and 6 MΩ were fashioned from borosilicate glass (0.86 mm inner diameter; 1.5 mm outer diameter; GB150-8P; Science Products) with a vertical pipette puller (PP-830; Narishige, London, UK). All recordings were performed with an EPC10 patch-clamp amplifier (HEKA, Lambrecht, Germany) controlled by the program PatchMaster (version 2.32; HEKA) running under Windows. In parallel, data were recorded using a micro1410 data acquisition interface and Spike 2 (version 7) (both from CED, Cambridge, UK). Data were sampled at 10 kHz and low-pass filtered at 2 kHz with a four-pole Bessel filter. Whole-cell capacitance was determined by using the capacitance compensation (C-slow) of the EPC10. Cell input resistances (RM) were calculated from voltage responses to hyperpolarizing current pulses. The calculated liquid junction potential of 14.6 mV between intracellular and extracellular solution was compensated or subtracted offline (calculated with Patcher’s Power Tools plug-in from http://www.mpibpc.mpg.de/groups/neher/index.php?page=software for IGOR Pro 6 [Wavemetrics, Lake Oswego, OR, USA]).

Perforated patch clamp recordings

Request a detailed protocol

Perforated patch experiments were conducted using protocols modified from Horn and Marty (1988) and Akaike and Harata (1994). Recordings were performed with ATP and GTP free pipette solution containing (in mM): 128 K-gluconate, 10 KCl, 10 HEPES, 0.1 EGTA, 2 MgCl2 adjusted to pH 7.3 with KOH resulting in an osmolarity of ~300 mOsm. ATP and GTP were omitted from the intracellular solution to prevent uncontrolled permeabilization of the cell membrane (Lindau and Fernandez, 1986). The patch pipette was tip filled with internal solution and back filled with internal solution, which contained the ionophore to achieve perforated patch recordings and 0.02% tetraethylrhodamine-dextran (3000 MW, D3308, Invitrogen, Eugene, OR, USA) to monitor the stability of the perforated membrane.

Amphotericin B (A4888; Sigma) was dissolved in dimethyl sulfoxide (DMSO; D8418, Sigma) following the protocols of Rae et al. (1991) and Kyrozis and Reichling (1995). The used DMSO concentration (0.1–0.3%) had no obvious effect on the investigated neurons. The ionophore was added to the modified pipette solution shortly before use. The final concentration of amphotericin B was ~200 μg · ml−1. During the perforation process access resistance (Ra) was constantly monitored and experiments were started after Ra had reached steady state (~15–20 min) and the action potential amplitude was stable.

Noradrenalin experiments

Request a detailed protocol

Noradrenalin (noradrenaline-bitartrate; I9278, Sigma) was bath-applied at concentrations between 10 nM and 100 μM for ~5 min until the NA effect had reached a steady state. Membrane potentials and AP frequencies were measured as averages from time periods (at least 30 s) when the NA effect had reached steady state. The input resistance of a neuron was determined as the mean from three sets of small hyperpolarizing current pulses (−5 to −20 pA). Conductance densities were determined as the inverse of the input resistance divided by the whole-cell capacitance. Concentration-response relations were normalized to the maximal response and fit with the sigmoidal relation (Equation 1)

(1) y=11+10(LogEC50X)*slope)

Drugs

Noradrenaline-bitartrate(10 nM - 100 μM; I9278, Sigma), the α1-AR antagonist prazosine (5 µM; P7791, Sigma), the α2-AR antagonist yohimbine (5 µM; Y3125, Sigma), the α2A-AR antagonist BRL 44408 (10 μM, C5776, Sigma), the α2B-AR antagonist ARC 239, the α1A-AR antagonist WB 4101, the α1C, D-AR antagonist CEC (1 nM – 1 µM, Q102, Sigma) and the α1A-AR agonist A61603 (100 nM; A5861, Sigma) were added to the normal aCSF.

To reduce glutamatergic and GABAergic synaptic input to the recorded neurons, 10 µM CNQX (6-cyano-7-nitroquinoxaline-2,3-dione, C127, Sigma-Aldrich), 50 µM DL-AP5 (DL-2-amino-5-phosphonopentanoic acid, BN0086, Biotrend), and 100 µM PTX (picrotoxin, P1675, Sigma Aldrich) was added to the extracellular saline.

RNA in situ hybridizations

Request a detailed protocol

In situ hybridizations were performed using the RNAscope 2-plex detection chromogenic kit, according to manufacturer’s instructions (Advanced Cell Diagnostics, USA). Briefly, wild-type C57BL/6 mice were intra-cardially perfused with 4% paraformaldehyde, post-fixed overnight in 4% PFA at room temperature and subsequently cryo-protected in 25% sucrose in 1x PBS. Coronal slices containing the hypothalamus were boiled at 98°C for antigen retrieval, hybridized with gene-specific probe sets against Adra1a, Adra2a, Adrb1 or Adrb2 and Npy or Pomc, to indicate the corresponding neuronal populations. Staining was evaluated using a brightfield Leica DM5500 B microscope. Per staining, 3 samples with three sections each were used to manually quantify double-positive cells. Per NPY- or POMC- positive cell, only cells containing at least two green (gene-specific) dots were considered as double-positive.

Data analysis

Request a detailed protocol

Data analysis was performed with Spike2 (version 7; Cambridge Electronic Design Ltd., Cambridge, UK), Igor Pro 6 (Wavemetrics, Portland, OR, USA) and Graphpad Prism (version 5.0b; Graphpad Software Inc., La Jolla, CA, USA). If not stated otherwise, all calculated values are expressed as means ± SEM (standard error of the mean). EC50 values are expressed as mean followed by their 95% confidence interval (CI) values. The ‘+’ signs in the box plots show the mean, the horizontal line the median of the data. The whiskers were calculated according to the ‘Tukey’ method. For comparisons of dependent and independent data paired and unpaired t-tests were used, respectively. For multiple comparisons ANOVA with Tukey post hoc analysis was performed. A significance level of 0.05 was accepted for all tests. Significance levels were: *p<0.05, **p<0.01, ***p<0.001. In the figures n values are given in brackets. Exact p-values are reported if p>0.0001.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69

Decision letter

  1. Joel K Elmquist
    Reviewing Editor; University of Texas Southwestern Medical Center, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Antagonistic modulation of NPY/AgRP and POMC neurons in the arcuate nucleus by noradrenalin" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Richard Aldrich as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission

Summary:

In this submission, Paeger et al., elucidate the effects of exogenously applied NA on the electrical output of energy-balance-regulating NPY and neurons in the arcuate nucleus of the hypothalamus. They determine that NA activates NPY neurons but inhibits POMC neurons, and go on to elucidate the receptors involved by a combination of molecular biology and functional pharmacology combined with electrophysiology. The authors believe this activity to be dependent upon α1A – and β- adrenergic receptors in NPY/AgRP neurons and α2A – adrenergic receptors in POMC neurons. The finding are interesting, especially in the context of recent work by Buret linking NA-releasing neurons to signaling energy need. The experiments are well performed and the data support the conclusions of the authors. However, there are some concerns of the methodology that limit enthusiasm.

Essential revisions:

Figures 12. While the NPY/AgRP cells demonstrate an increased APF in response to NA, a change in membrane potential is not apparent. This appears opposite to the effects in POMC neurons. Does NA alter the RMP of NPY/AgRP neurons? If not, how does one reconcile this with the WB4101 data (Figure 6C and D) suggesting an α2A-AR dependent regulation of RMP in NPY/AgRP neurons?

Figure 4 appears of limited advancement over previous work published by the Sternson group. Also, the authors use POMC gene expression as a control, however why have the authors not examined NPY or AgRP gene expression to validate the cell type-specificity?

Figure 5 and Figure 6 – The authors use various agonists/antagonists to examine the requirement of ARs in the acute activation of NPY/AgRP neurons. While Figure 6 legend is missing, it's indicated in the Results section that the experiments in Figure 6 (similar to Figures 2, 3, and 5) were performed using sequential applications in the same cell. A primary concern with these experiments is whether there is desensitization of the first application prior to the second test. This point is raised given that Figure 6A suggests that such desensitization may exist as the amplitude of the depolarization/frequency of events in response to A61603 followed by NA appear quite different. The authors should demonstrate subsequent application of all agonists is similar in the first and second application.

Figure 6 and 7 – Is it possible that the dosages used are confounding the interpretation of the current results? In particular, in the current study the authors used 1µM A-61603. A-61603 is highly selective for α1A-AR when compared to α1B-AR or α1D-AR by almost 100 fold or more. However, early studies suggested that A-61603 is selective at binding α1A-AR over α2-AR by as little as 10 fold – Ki 8.89nM for α1A-AR and Ki 100nM for α2-AR (Knepper et al., 1995). Due to the dosage used, how can the authors be certain that the effects observed are not dependent upon α2-ARs.

Similarly, WB4101 has been shown to label platelet α-2 adrenergic receptors (Sundaresan et al., 1983) and WB4101 appears to bind to a nonadrenergic low-affinity site in both calf and human brain (Weinreich and Seeman, 1981; Lyon and Randall, 1980).

Also, BRL 44408 is selective for α2A-AR over α2B-AR (Ki = 1.7 nM and 144.5 nM), however the authors used BRL 44408 at 10µM likely eliminating this selectivity. These points warrant further experimentation if the authors are to conclude AR specific effects.

Given the authors focus on the role of NPY/AgRP and POMC neurons in feeding behavior, how might the current results fit with earlier work which suggested that α2-AR blockade attenuates feeding behavior induced by both NPY and epinephrine (Clark et al., 1988).

Why are the resting membrane potentials of POMC neurons so hyperpolarized in Figure 7? Results, subsection “Noradrenalin modulates electrophysiological activity differentially in NPY/AgRP and POMC neurons” suggests that POMC neurons were spontaneously active under control conditions. Were these neurons spontaneously active? A search of the literature (including work from the same authors e.g. L. Plum et al., 2006, AJ Newton et al., 2013) reports RMP of POMC neurons from ~-40mV to ~-60mV. RMPs of -70mV or -80mV would appear to exceed this range and these cells appear quite different from those reported in Figure 3. Could the data be susceptible to cell selection biases?

https://doi.org/10.7554/eLife.25770.013

Author response

Essential revisions:

Figures 12. While the NPY/AgRP cells demonstrate an increased APF in response to NA, a change in membrane potential is not apparent. This appears opposite to the effects in POMC neurons. Does NA alter the RMP of NPY/AgRP neurons? If not, how does one reconcile this with the WB4101 data (Figure 6C and D) suggesting an α2A-AR dependent regulation of RMP in NPY/AgRP neurons?

The reviewers raise an important point. The NPY/AgRP neurons increased APF and depolarized in response to NA, which was stated in Results, subsection “NPY/AgRP neurons, NA concentration-response relation” (5.1 ± 1.0 mV, paired t-test, n = 8, p = 0.0011). Compared to the hyperpolarization of the POMC neurons the depolarization of the NPY/AgRP neurons is smaller. Given the scaling in Figures 1 and 2 (where we show the full amplitude of the AP), this depolarization is less obvious. In the revised Figure 2A we have added a line, which indicates the control membrane potential to better visualize the NA induced depolarization. Note that Figure 1A and Figure 2D only show the initial phase of the NA response.

Figure 4 appears of limited advancement over previous work published by the Sternson group. Also, the authors use POMC gene expression as a control, however why have the authors not examined NPY or AgRP gene expression to validate the cell type-specificity?

We agree with the reviewers – in hindsight. The gene expression analysis was performed because our initial electrophysiological experiments revealed a clear differential modulation: excitation in NPY/AgRP and inhibition in POMC neurons. Accordingly, we expected clear differential expression of excitatory and inhibitory ARs, which did not match the data by the Sternson lab. To clarify this discrepancy, we repeated the gene expression analysis, which confirmed the data from the Sternson lab, and unequivocally revealed that ARs are differentially expressed between NPY and POMC neurons.

Figure 5 and Figure 6 – The authors use various agonists/antagonists to examine the requirement of ARs in the acute activation of NPY/AgRP neurons. While Figure 6 legend is missing, it's indicated in the Results section that the experiments in Figure 6 (similar to Figures 2, 3, and 5) were performed using sequential applications in the same cell. A primary concern with these experiments is whether there is desensitization of the first application prior to the second test. This point is raised given that Figure 6A suggests that such desensitization may exist as the amplitude of the depolarization/frequency of events in response to A61603 followed by NA appear quite different. The authors should demonstrate subsequent application of all agonists is similar in the first and second application.

General response: We apologize for the errors in Figure 6. We now provide the right legend for Figure 6. In addition we have corrected an error in Figure 6A: the bar indicating the NA application is shifted to the correct position.

We absolutely agree with the reviewers that repeated drug applications potentially can cause desensitization. However, note that this experimental scheme allowed us to confirm that each recorded cell was responsive to NA in the first place. We consider this type of “intrinsic control” as a real experimental advantage. In our opinion, this did not compromise our conclusions (see below).

Specific response: As suggested by the reviewer we tested if desensitization for NA and A61603 occurs. Under our experimental conditions, the effect of both substances is reversible and did not cause desensitization (NA: Figure 2—figure supplement 1 and A61603: Figure 6—figure supplement 1).

NA: The effect of NA is reversible and we did not observe desensitization using sequential application. In our opinion this supports the validity of the conclusions that we have drawn from the data of Figure 2E, Figure 3E and Figure 5A and B. In Figure 2E NA induced the same response in control conditions and in synaptic blockers, suggesting that the excitatory NA effect is fully mediated by cell-intrinsic ARs (Results, subsection “NPY/AgRP neurons, NA concentration-response relation”). These data alone show the lack of desensitization. In Figure 3E NA induced a smaller response during the NA application in synaptic blockers. Since we have shown that the reduced response is not caused by desensitization (Figure 2—figure supplement 1), we conclude that the NA effect is intrinsic and in (a small) part mediated by presynaptic effects (Results, subsection “POMC neurons, NA concentration-response relation”). Since the same line of arguments is valid for the data shown in Figure 5, we conclude that the complete (Figure 5A) and the partial block (Figure 5B) of the NA response is caused by the AR blockers (Results, subsection “NA excites NPY/AgRP neurons predominantly via the activation of α1A-Ars”) and not by desensitization.

A61603: The effect of A61603 is reversible and we did not observe desensitization using sequential application. Figure 6A showed that the α1A-AR agonist A61603 can reproduce the excitatory NA effect on the APF in NPY/AgRP-neurons. The main purpose of the following NA applications was to confirm that the investigated neurons were responsive to NA in the first place – which they were. Between the A61603- and the NA response was no significant difference (Figure 6B).

Nevertheless, the quantification showed that the NA response tended to be slightly smaller than the A61603 response. This result, however, has to be seen in the context of the subsequent experiments (Figure 6C and D). The analysis on a finer scale revealed that NA not only activated excitatory α1-AR, but also inhibitory α2-AR. Together this caused a smaller net excitation than A61603 alone. This issue is addressed in more detail next (and in Results, subsection “NA excites NPY/AgRP neurons predominantly via the activation of α1A-Ars”).

Figure 6 and 7 – Is it possible that the dosages used are confounding the interpretation of the current results? In particular, in the current study the authors used 1µM A-61603. A-61603 is highly selective for α1A-AR when compared to α1B-AR or α1D-AR by almost 100 fold or more. However, early studies suggested that A-61603 is selective at binding α1A-AR over α2-AR by as little as 10 fold – Ki 8.89nM for α1A-AR and Ki 100nM for α2-AR (Knepper et al., 1995). Due to the dosage used, how can the authors be certain that the effects observed are not dependent upon α2-ARs.

We agree with the reviewers that the drug concentrations used in this type of experiments can play an important role for interpreting the data. In this context, it is important to note that it is not straightforward “to transfer” Ki values between preparations. Therefore, our experiments and conclusions are typically based on the application of more than one drug, e.g., the application of agonists and antagonists.

In case of the NPY/AgRP neurons we observed that the net NA effect was excitatory. However, as already mentioned above, a more detailed analysis revealed that theses neurons are excited by α1A-AR and inhibited by α2A-AR when α-1A-AR are blocked. This conclusion is based on the following reasoning: The α1A-AR A61603 agonist had a similar excitatory effect as NA, while the α1A-AR antagonist WB4101 blocked the excitatory NA effect. Interestingly, in the presence of WB4101 (when the excitatory action is blocked) NA had an inhibitory effect, which could be blocked by the α2A-AR antagonist BRL44408. In summary, the experiments show that NA activates both excitatory α1-AR and inhibitory α2-AR, causing net excitation. This dual effect of NA on NPY/AgRP neurons is explained in the legend of Figure 6 and in text (Results, subsection “NA excites NPY/AgRP neurons predominantly via the activation of α1A-Ars”).

Similarly, WB4101 has been shown to label platelet α-2 adrenergic receptors (Sundaresan et al., 1983) and WB4101 appears to bind to a nonadrenergic low-affinity site in both calf and human brain (Weinreich and Seeman, 1981; Lyon and Randall, 1980).

We thank the reviewers for pointing out these references.

Also, BRL 44408 is selective for α2A-AR over α2B-AR (Ki = 1.7 nM and 144.5 nM), however the authors used BRL 44408 at 10µM likely eliminating this selectivity. These points warrant further experimentation if the authors are to conclude AR specific effects.

We agree with the reviewers that the experiments with the α2A-AR blocker BRL 44408 alone are not sufficient to differentiate between α2A-AR or α-2B-AR inhibitory action. Therefore, we want to refer to the complete series of experiments (Figure 7). The BRL 44408 experiments (Figure 7A) and the yohimbine experiments (Figure 7D) strongly suggest that the inhibitory NA effect on POMC-neurons is caused by α2-AR. Following the argument of the reviewers, these experiments alone cannot differentiate between α2A-AR and α2B-AR effects. However, since α2B-AR antagonist ARC239 did not block the NA inhibitory NA response, we conclude that the inhibitory NA effect is indeed mediated by α2A-AR. In addition, and very importantly, the transcriptome of the POMC- neurons did show α2A-AR, but no α2B-AR expression (Results, subsection “Αdrenergic receptors are expressed in NPY/AgRP and POMC neurons of the ARH”).

Given the authors focus on the role of NPY/AgRP and POMC neurons in feeding behavior, how might the current results fit with earlier work which suggested that α2-AR blockade attenuates feeding behavior induced by both NPY and epinephrine (Clark et al., 1988).

Our results fit very well with the behavioral data form Clark et al., 1988. Here we show that NA markedly inhibits satiety signaling POMC neurons via α2-AR. This alone (even independently from the excitatory NA action on the hunger signaling NPY/AgRP neurons) would increase the orexigenic drive from the NPY/AgRP – POMC system and should lead to increased food intake. The α2-AR antagonist yohimbine blocks the strong inhibitory effect of NA on the POMC neurons, increasing their anorexigenic effect, which in turn should reduce feeding. We have included this Discussion in the revised manuscript (Discussion section).

Given the state of research at the time, Clark et al., 1988 suggested cellular mechanisms, which focused exclusively on the role of NPY. Since, these mechanisms were formulated relatively vaguely at the time (For example: “Yohimine attenuation of NPY and adrenergic transmitter-evoked feeding behavior may involve α2-adrenoceptors located on presynaptic elements, or, as suggested by a number of studies (e.g. [18,19]), postsynaptic elements, or both.”) our results neither support nor strictly contradict their hypotheses.

Why are the resting membrane potentials of POMC neurons so hyperpolarized in Figure 7? Results, subsection “Noradrenalin modulates electrophysiological activity differentially in NPY/AgRP and POMC neurons” suggests that POMC neurons were spontaneously active under control conditions. Were these neurons spontaneously active? A search of the literature (including work from the same authors e.g. L. Plum et al., 2006, AJ Newton et al., 2013) reports RMP of POMC neurons from ~-40mV to ~-60mV. RMPs of -70mV or -80mV would appear to exceed this range and these cells appear quite different from those reported in Figure 3. Could the data be susceptible to cell selection biases?

First of all, we want to apologize for not having stated the age of the mice that were used for the electrophysiological recordings. The recordings have been performed on 15 – 20 weeks old animals, which is now clearly stated in ‘Materials and methods, subsection “Animal care” and subsection “Electrophysiology” of the revised manuscript. This is important to consider, since previous work suggests that the POMC neurons hyperpolarize with increasing age (Yang et al., 2012 Neuron 75:425-436; Newton et al., 2013). Yang et al., report a mean RMP of ~-65 mV in 24 weeks old mice (estimated from their Figure 1A). The authors do not provide absolute values, but the boxplot shows several neurons that are clearly more hyperpolarized. In Newton et al., we report a mean RMP of -60 mV (SEM: 1.1; SD: 5.2; Range: -68 mV to -50 mV) in 15-week-old mice.

In the current manuscript, the boxplot (Figure 7C) shows the whole range of the measured RMP with a mean of 65.3 mV (SEM: 3.4; SD: 10.2). Based on the previous reports this is in a range, which we would expect for this age and it is also consistent with data in Figure 3. The boxplot (Figure 7C) also reveals that the mean is influenced by 3 very hyperpolarized neurons, while the other 6 neurons are more depolarized. Since the recordings of the hyperpolarized neurons were of high quality, we had no reason to classify them as outliers. The boxplot also shows that NA affects all neurons in the same way, independently of the control RMP. Thus, the NA effect is not limited to the hyperpolarized neurons or stronger in these neurons.

In Figures 7A and B we replaced the example, which showed a very hyperpolarized RMP, with an example in which the RMP is close to the mean RMP.

https://doi.org/10.7554/eLife.25770.014

Article and author information

Author details

  1. Lars Paeger

    1. Biocenter, Institute for Zoology, University of Cologne, Cologne, Germany
    2. Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases, Cologne, Germany
    Contribution
    LP, Conceptualization, Data curation, Formal analysis, Investigation, Writing—original draft, Writing—review and editing
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8716-3483
  2. Ismene Karakasilioti

    1. Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases, Cologne, Germany
    2. Max Planck Institute for Metabolism Research, Cologne, Germany
    3. Center of Molecular Medicine Cologne, University of Cologne, Cologne, Germany
    4. Center for Endocrinology, Diabetes and Preventive Medicine, University Hospital Cologne, Cologne, Germany
    Contribution
    IK, Formal analysis, Investigation, Writing—review and editing
    Competing interests
    The authors declare that no competing interests exist.
  3. Janine Altmüller

    1. Center of Molecular Medicine Cologne, University of Cologne, Cologne, Germany
    2. Cologne Center for Genomics, University of Cologne, Cologne, Germany
    Contribution
    JA, Formal analysis, Investigation
    Competing interests
    The authors declare that no competing interests exist.
  4. Peter Frommolt

    Bioinformatics Facility, Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases, Cologne, Germany
    Contribution
    PF, Formal analysis, Investigation
    Competing interests
    The authors declare that no competing interests exist.
  5. Jens Brüning

    1. Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases, Cologne, Germany
    2. Max Planck Institute for Metabolism Research, Cologne, Germany
    3. Center of Molecular Medicine Cologne, University of Cologne, Cologne, Germany
    4. Center for Endocrinology, Diabetes and Preventive Medicine, University Hospital Cologne, Cologne, Germany
    Contribution
    JB, Resources, Formal analysis, Supervision, Funding acquisition, Writing—review and editing
    Competing interests
    The authors declare that no competing interests exist.
  6. Peter Kloppenburg

    1. Biocenter, Institute for Zoology, University of Cologne, Cologne, Germany
    2. Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases, Cologne, Germany
    Contribution
    PK, Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Validation, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    peter.kloppenburg@uni-koeln.de
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4554-404X

Funding

Deutsche Forschungsgemeinschaft (TR-SFB 134/TP A03)

  • Peter Kloppenburg

Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases

  • Peter Kloppenburg

Max-Planck-Gesellschaft

  • Jens Brüning

EMBO (Marie Curie Long-Term Fellowship)

  • Ismene Karakasilioti

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

The authors wish to thank Ali Abdallah, Christoph Goettlinger, Eva Tsaousidou, and Helmut Wratil for excellent technical assistance.

Reviewing Editor

  1. Joel K Elmquist, University of Texas Southwestern Medical Center, United States

Publication history

  1. Received: February 6, 2017
  2. Accepted: May 26, 2017
  3. Version of Record published: June 20, 2017 (version 1)

Copyright

© 2017, Paeger et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,132
    Page views
  • 369
    Downloads
  • 15
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Neuroscience
    María Fernanda López-Gutiérrez et al.
    Research Article

    Previous studies have related pair bonding in Microtus ochrogaster, the prairie vole, with plastic changes in several brain regions. However, the interactions between these socially-relevant regions have yet to be described. In this study, we used resting state magnetic resonance imaging to explore bonding behaviors and functional connectivity of brain regions previously associated with pair bonding. Thirty-two male and female prairie voles were scanned at baseline, 24h and 2 weeks after the onset of cohabitation By using network based statistics, we identified that the functional connectivity of a cortico-striatal network predicted the onset of affiliative behavior, while another predicted the amount of social interaction during a partner preference test. Furthermore, a network with significant changes in time was revealed, also showing associations with the level of partner preference. Overall, our findings revealed the association between network-level functional connectivity changes and social bonding.

    1. Developmental Biology
    2. Neuroscience
    Alessia Caramello et al.
    Research Article Updated

    During embryonic development, radial glial cells give rise to neurons, then to astrocytes following the gliogenic switch. Timely regulation of the switch, operated by several transcription factors, is fundamental for allowing coordinated interactions between neurons and glia. We deleted the gene for one such factor, SOX9, early during mouse brain development and observed a significantly compromised dentate gyrus (DG). We dissected the origin of the defect, targeting embryonic Sox9 deletion to either the DG neuronal progenitor domain or the adjacent cortical hem (CH). We identified in the latter previously uncharacterized ALDH1L1+ astrocytic progenitors, which form a fimbrial-specific glial scaffold necessary for neuronal progenitor migration toward the developing DG. Our results highlight an early crucial role of SOX9 for DG development through regulation of astroglial potential acquisition in the CH. Moreover, we illustrate how formation of a local network, amidst astrocytic and neuronal progenitors originating from adjacent domains, underlays brain morphogenesis.