**Contents**


## **About the Special Issue Editor**

**Albertino Bigiani** (Professor) Degree in Biological Sciences (University of Modena, Modena, Italy); Ph.D. in Neuroscience (University of Pisa, Pisa, Italy); Postdoctoral fellow with Prof. Stephen D. Roper (Department of Anatomy and Neurobiology, Colorado State University, Fort Collins, Colorado, USA). He is now Professor of Physiology and has a tenured position at the University of Modena and Reggio Emilia (Italy). He received training in membrane electrophysiology, and has expertise in the field of cell physiology. His current research interests are the mechanisms used by taste cells to detect sodium ions (Na+), which are responsible for the taste of table salt (salty). Website: https://www.researchgate.net/profile/Albertino Bigiani.

### *Editorial* **Salt Taste, Nutrition, and Health**

#### **Albertino Bigiani**

Dipartimento di Scienze Biomediche, Metaboliche e Neuroscienze, Università di Modena e Reggio Emilia, 41125 Modena, Italy; albertino.bigiani@unimore.it; Tel.: +39-059-205-5349

Received: 19 May 2020; Accepted: 22 May 2020; Published: 25 May 2020

The sodium ion (Na+) is essential for life. Na<sup>+</sup> is the main cation in the extracellular fluid bathing all our cells, and it is also a key element in many body secretions. In addition to determining the extracellular fluid osmolality, Na<sup>+</sup> is involved in several physiological processes that would be impaired by its deficiency—without Na+, neurons and muscle cells would not be able to generate electrical impulses, the intestinal adsorption of nutrients would be undermined, and the kidneys would not work properly. To maintain stable levels of Na<sup>+</sup> in our body, losses through kidneys, the gastrointestinal tract, and sweating have to be balanced by the ingestion of foods containing this mineral. Our ability to recognize Na<sup>+</sup> relies mainly on the activity of the taste system. Usually, we consume sodium in the form of sodium chloride (NaCl), the common table salt. Moderate concentrations of salt induce a specific "salty" sensation that is appealing and appetitive for us. However, saltiness can be perceived as pleasant or unpleasant according to the salt concentration and the medium in which salt occurs. For example, we dislike salt concentrations >150 mM in aqueous solution, but we found them palatable when salt is in foodstuffs. The positive hedonic tone (liking) of salt taste may expose us to salt overconsumption. This is further exacerbated by the widespread use of NaCl as preservative and as flavoring agent in processed foods. When consumed in excess, salt is detrimental to health: in fact, excessive sodium intake is linked to the development of hypertension and related pathologies.

This Special Issue provides a contribution to our understanding of salt taste and of its impact on nutrition and health. Although Na<sup>+</sup> is a simple chemical, the biological processes underlying its handling in our body are quite complex. Figure 1 shows a simplified scheme of sodium balance mechanisms. In the following paragraphs, I will discuss some of the new findings with relation to the processes highlighted in Figure 1: Na<sup>+</sup> detection, eating behavior, blood pressure regulation, and Na<sup>+</sup> output regulation.

**Figure 1. Sodium balance**. An adequate extracellular concentration of Na<sup>+</sup> (*[Na*+*]ext*) is vital for the functioning of our cells. Diet provides the daily amount of Na<sup>+</sup> to balance off losses through kidney, intestine, and sweating (not shown). Food Na<sup>+</sup> is detected mainly by taste buds, which contained sensory cells able to sense this cation through sodium receptors (*ENaC* and others not yet identified). Taste information is fundamental for recognizing the quality and the intensity of saltiness. However, large salt concentration may also activate trigeminal nerve endings, which contain another "salt" receptor (*TRPV1*). This sensory pathway is believed to provide information on stimulus intensity in supra-threshold salt concentration range. Chemosensory signals are processed in the brain for conscious perception and for sodium appetite regulation. The outcome of central processing guides salt intake (Na<sup>+</sup> input). Note that oral chemosensory input provides feed-forward signals alerting central neurons on the presence of sodium-containing foods before Na<sup>+</sup> absorption has occurred in the gut. Once absorbed, ingested Na<sup>+</sup> is confined mainly to the extracellular space, and changes in its extracellular concentration affect blood volume and blood pressure, which in turn influence the renin-angiotensin-aldosterone system (*RAAS*). RAAS controls the amount of Na<sup>+</sup> lost by kidney and colon (Na<sup>+</sup> output), and also provides feedback to neural centers and taste buds to regulate sodium appetite and salt sensitivity, respectively. In turn, the brain modulates renin secretion through baroreceptor reflexes. Papers in this Special Issue touch upon some of the processes (Na<sup>+</sup> detection, eating behavior, blood pressure regulation, Na<sup>+</sup> output regulation) that are associated with the handling of food Na<sup>+</sup> by our body. Note that, for simplicity, other factors involved in sodium balance, such as the atrial natriuretic peptide, are not shown.

#### **1. Na**<sup>+</sup> **Detection**

The principal detectors of food Na<sup>+</sup> are the taste cells, specialized epithelial cells clustered in the taste buds of the oral mucosa. Na<sup>+</sup> activates these sensory cells by interacting with specific membrane proteins (sodium receptors) located in their apical portion, where they reach contact with the saliva. Taste cells then relay sensory information to nerve fibers, which transmit electrical impulses to the brain (Figure 1). The use of the diuretic drug, amiloride, has allowed researchers to distinguish two parallel signaling routes in most mammals: the amiloride-sensitive (AS) pathway and the amiloride-insensitive (AI) pathway. The epithelial sodium channel (ENaC), which is blocked by amiloride, works as a sodium receptor in the AS pathway. Sodium detection by the AI pathway is not affected by amiloride. The AS mechanism is highly specific for Na<sup>+</sup> detection, whereas the AI one is more broadly tuned to detect also other cations. However, the identity of the AI salt receptor(s) is(are) still unknown. These gustatory pathways transmit information on the stimulus quality (saltiness) and its intensity (salt concentration). Four papers provide data on the mechanisms underlying Na<sup>+</sup> detection by taste cells. Bigiani [1] reviews our current understanding on the possible involvement of ENaC in the initial events of Na<sup>+</sup> detection in humans. Although this oligomeric protein works as low-salt receptor in laboratory mammals, available data "seem to favor a role for ENaC downstream of the initial receptive events" in human salt taste. Shigemura et al. [2] provide compelling evidence that some components of the renin-angiotensin-aldosterone system (RAAS) occur in mouse taste buds. Circulating RAAS plays a key role in the regulation of sodium balance by controlling sodium excretion and by providing feedback to specific neural circuits and taste buds to regulate sodium appetite and salt sensitivity, respectively (Figure 1). The existence of components of RAAS in taste buds suggest that the activity of salt-sensitive cells could be also modulated locally in response to "random perturbations in the oral cavity during feeding and drinking", and this could affect information sent to central neurons. The authors suggest that local RAAS could be involved in "short-term feedforward regulation predicting changes in body fluid composition". Cattaneo et al. [3] found that the abundance of certain bacterial taxa on the tongue dorsum of healthy volunteers was negatively correlated with salt taste sensitivity (assessed by recognition threshold for NaCl solutions): namely, the more abundant those taxa, the lower the salt taste threshold. This result underscores the importance of the so-called perireceptor events in salt detection and possibly in defining the inter-individual differences in salt taste perception. The authors suggest that "oral microbiota composition deserves to be considered as an influencing variable when investigating perireceptor events involved in chemosensory processes". Lossow et al. [4] studied the effect of changes in dietary salt content on mRNA expression for ENaC subunits in mouse taste buds. They found that mRNA levels were not affected either by low-salt or by high-salt diet. Thus, taste function does not seem to play a major role in body adjustment to sodium imbalance, at least on the basis of the mRNA expression levels for the sodium receptor in mice.

In addition to taste cells, food Na<sup>+</sup> can be detected by trigeminal nerve endings, which are widely distributed throughout the lingual mucosa and also around taste buds. This sensory pathway is activated by high salt concentrations and likely provides information on salt concentration in supra-threshold range and also to avoid ingestion of hyperosmotic salt that could impair extracellular osmolality. Trigeminal nerve endings express the transient receptor potential vanilloid 1 receptor (TRPV1), an ion channel proposed to work as "salt" receptor. Rhyu et al. [5] provide interesting data on the involvement of trigeminal pathway in salt detection. These authors found that some "kokumi" peptide fractions isolated from Ganjang, a typical Korean soy sauce, were able to increase the perceived salt taste intensity in human volunteers, that is, they worked as a salt taste enhancer. Kokumi refers to tasteless compounds able to improve persistency and mouthfulness (mouth-filling sensation) as well as to enhance some of basic taste qualities, including salty taste, as demonstrated here. By recording the activity of chorda tympani (CT) taste nerve in rats, Rhyu et al. [5] found that these taste active peptides did not affect the ENaC-mediated AS pathway, but the AI one. However, the enhancing effect was likely due an interaction between trigeminal nerve endings containing TRPV1 and taste cells of the AI pathway. These results suggest "a novel relationship between trigeminal system and salt taste perception".

#### **2. Eating Behavior**

Eating behavior is a complex activity that has evolved to assure the proper ingestion of chemicals for body metabolism and homeostasis. In humans, the consumption of salty foods is driven not only by the integrated neuroendocrine mechanisms regulating the activity of central appetite neurons (Figure 1), but also by the combination of several other factors, including genetic context and non-homeostatic influences, such as learning, cultural factors, and personal habits. All these factors significantly affect the preference for salty foods, and therefore salt intake, determining individual variability. Three papers address human eating behavior in terms of possible association between salt taste perception and salt preference or dietary habits. Cattaneo et al. [3] found a correlation between decreased salt taste sensitivity and consumption of bakery and salty baked products in healthy, normal-weight, 18-30-year old volunteers. These findings indicate that inter-individual variation in salt perception may affect habitual salt consumption. As reported above, they also found a correlation between relative abundance of certain bacterial taxa and salt taste sensitivity. Thus, oral microbiota may influence food preference. The possible association between salt taste genotype and eating behavior has been studied by Ferraris et al. [6] in a large-sized (*n* = 536) and well-characterized elderly cohort in Australia. They evaluated the association between the occurrence of single nucleotide polymorphisms (SNPs) for the TRPV1-encoding gene and individual salt intake. TRPV1 is a salt receptor expressed in trigeminal nerve endings, which may contribute to the perception of NaCl, especially when salt concentration is high. Their findings indicate that there is no association between TRPV1 SNP and salt intake in the analyzed elderly cohort, suggesting that, at least for people aged 65 years or older, the TRPV1 genotype is not crucial in defining salt consumption. Veˇcek et al. [7] performed a cross-sectional study on general population of Dalmatia, Croatia (*n* = 2798 subjects) to determine possible association between salt taste perception, Mediterranean diet, and Metabolic Syndrome (MetS). They found that there were no differences in the overall Mediterranean diet compliance between subjects with different salt taste threshold. However, they found that "subjects with higher salt threshold added salt to their food more frequently compared to subjects with both lower and intermediate threshold". This means that individuals with higher salt sensitivity (lower taste threshold) could have a reduced salt intake. Interestingly, they found that these subjects also showed lower prevalence of MetS.

#### **3. Blood Pressure Regulation**

Nervous and endocrine mechanisms assure a proper level of hydraulic pressure in the large arteries to sustain an adequate blood perfusion throughout the body. Salt intake may affect blood pressure. Ingested Na<sup>+</sup> is confined into the extracellular compartment. Here, through water retention, Na<sup>+</sup> sets the overall volume, including the blood volume. In turn, this affects blood pressure (Figure 1). It is therefore clear that one of the consequences of an increased salt intake is likely an increase in blood pressure. Several studies underscore the role of excessive salt consumption in the development of hypertension, which represents a component of MetS. The association between salt intake and blood pressure is addressed in this Special Issue. Ferraris et al. [6] found (see above) that SNP for the gene encoding the salt receptor, TRPV1, was not associated with a variation in salt intake in an elderly cohort. Consistently, neither systolic nor diastolic blood pressure varied by genotype. The authors acknowledge that "as the cohort was 65 years and older, the results are not necessarily generalizable to the wider adult or youth population" since aging may affect genetic expression. In their cross-sectional study on the general population of Dalmatia (Croatia), Veˇcek et al. [7] found that "subjects with higher salt taste threshold were on average older than those with lower threshold". This means that the ability to recognize salty stimuli blunts with age. In addition, they found that "age was also negatively correlated with salt taste intensity perception", evaluated with supra-threshold testing. As a whole, these results suggest that aging affects salt taste performance. Veˇcek et al. [7] also found that high blood pressure, a MetS component, was more common among subjects with higher salt recognition threshold, that is, those with lower salt sensitivity. Mun et al. [8] studied the effect of Doenjang, a traditional Korean seasoning with a high salt content obtained by soybean fermentation, on the blood pressure in rats, which were fed a high-salt diet with or without Doenjang. Interestingly, blood pressure was significantly lower in the first group, although the administered salt content was similar in the two groups. In addition, RAAS was also affected: renin and aldosterone levels were decreased in mice fed with Doenjang. Likely, other chemicals occurring in Doenjang as well as its microbial community offset the effect of salt intake on blood pressure. Thus, their results suggest that "eating traditional salty fermented food is not a direct cause of hypertension, and the intake of Doenjang in normal healthy animals improved blood pressure". In line with the goal of reducing salt consumption without affecting

the palatability of foodstuffs, Rhyu et al. [5] identified kokumi active peptides that could be used as salt taste enhancers. Clearly, this would be beneficial for controlling blood pressure, because less salt would be necessary to have the same saltiness perception when these peptides are present in foods.

#### **4. Na**<sup>+</sup> **Output Regulation**

Renal excretion and losses through the gastrointestinal tract represent the output of the homeostatic system controlling the extracellular concentration of Na<sup>+</sup> ([Na<sup>+</sup>]ext). Variations in salt intake lead to changes in [Na<sup>+</sup>]ext, which in turn affect, through expansion/reduction of extracellular volume to correct osmolality, blood volume and RAAS (Figure 1). RAAS targets the renal nephron and the colon, where aldosterone affects sodium reabsorption through ENaC to match changes in sodium intake. Lossow et al. [4] found that changes in dietary salt content in mice affected ENaC mRNA expression levels in kidney and distal colon. As reported above, they also analyzed ENaC expression in taste buds, since in this mammal ENaC works as sodium taste receptor. They data clearly indicate that "colon and kidney seem to be of greater importance to compensate imbalanced sodium homeostasis than gustatory tissue based on the monitored ENaC expression levels". These results underscore the importance of an adequate regulation of sodium output to compensate changes in [Na<sup>+</sup>]ext due to variations in salt intake. In this regard, it is worth noting that salt taste works as a feedforward mechanism that can reduce sodium appetite to prevent overconsumption. On the contrary, Na<sup>+</sup> output is the expression of a homeostatic regulation, which uses compensatory feedback mechanisms to stabilize [Na+]ext (Figure 1).

#### **5. Final Note**

I would like to thank all the authors in this Special Issue for providing their new research data on salt taste mechanisms and on the role of salt taste in nutrition and health. I am sure their valuable contributions will be appreciated by the readership of *Nutrients*, as well as by the scientific community.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **Does ENaC Work as Sodium Taste Receptor in Humans?**

#### **Albertino Bigiani**

Dipartimento di Scienze Biomediche, Metaboliche e Neuroscienze, Università di Modena e Reggio Emilia, 41125 Modena, Italy; albertino.bigiani@unimore.it; Tel.: +39-059-205-5349

Received: 29 March 2020; Accepted: 21 April 2020; Published: 24 April 2020

**Abstract:** Taste reception is fundamental for the proper selection of food and beverages. Among the several chemicals recognized by the human taste system, sodium ions (Na+) are of particular relevance. Na<sup>+</sup> represents the main extracellular cation and is a key factor in many physiological processes. Na<sup>+</sup> elicits a specific sensation, called salty taste, and low-medium concentrations of table salt (NaCl, the common sodium-containing chemical we use to season foods) are perceived as pleasant and appetitive. How we detect this cation in foodstuffs is scarcely understood. In animal models, such as the mouse and the rat, the epithelial sodium channel (ENaC) has been proposed as a key protein for recognizing Na<sup>+</sup> and for mediating preference responses to low-medium salt concentrations. Here, I will review our current understanding regarding the possible involvement of ENaC in the detection of food Na<sup>+</sup> by the human taste system.

**Keywords:** sodium taste; sodium receptor; salt taste; amiloride; taste transduction

#### **1. Introduction**

The sodium ion (*Na*+) is an essential mineral for our body because it regulates the osmolality of the extracellular fluid and plays a key role in many physiological processes, from the generation of nerve impulses to renal function. Na<sup>+</sup> is lost continuously through urine, feces, and sweat. Thus, to maintain proper bodily balance, we need to replace losses by the ingestion of food containing this cation. Our ability to detect Na<sup>+</sup> in foodstuffs relies on the taste system: Na<sup>+</sup> elicits a specific sensation called salty taste [1,2] that guides the intake of this important mineral [3,4]. Table salt (NaCl) represents a prototypic chemical substance evoking salty taste. It is well established that Na<sup>+</sup> is responsible for the perceived saltiness and for the pleasantness of low to medium concentrations of table salt [4,5].

The early events in taste reception typically include the interaction of the chemical stimulus with a membrane receptor in taste cells, specialized epithelial cells clustered in sensory end-organs called taste buds [6]. Most of studies on sodium taste reception have been performed on laboratory rodents (mouse and rat). In these mammals, the epithelial sodium channel (ENaC) works as low-salt receptor, mediating acceptance responses to low-medium salt concentrations and driving salt consumption [5]. The obvious question is then: does ENaC play any role in sodium detection in humans? In this review, I will discuss current information supporting or arguing against the possible involvement of this ion channel in human salt taste. I will evaluate whether data from human studies fit the model for the detection of Na<sup>+</sup> based on ENaC as sodium receptor localized at the apical membrane of taste cells; that is, where these cells contact the saliva in the so-called taste pore region of taste buds [6,7].

#### **2. Psychophysics**

In laboratory rodents, a pharmacological feature of taste ENaC is its sensitivity to amiloride, a diuretic drug, which selectively blocks the channel at submicromolar concentrations [8]. Since ENaC is inhibited by amiloride, application of this drug during stimulation with NaCl blunts the taste

response and the attractiveness of low sodium concentrations [9,10]. Following the same methodological approach, the involvement of the ENaC pathway in human taste reception has been investigated over 15 years by sensory evaluation of the amiloride effect on salt taste. Unexpectedly, findings have been very controversial. Some authors found that indeed the presence of amiloride reduced the perceived saltiness of the NaCl solutions, although to varying degrees [11–15]. On the contrary, other investigators found that amiloride had negligible or no effect on salt taste [16–21]. The discrepancy among studies might be due to differences in the experimental design. For example, perception arising from stimulation of the anterior tongue seems to be affected by amiloride [11–13], whereas perception from whole-mouth stimulation is not [16]. This raises the possibility that in the whole-mouth protocol, additional sensory inputs from the oral cavity may "obscure" the information conveyed by amiloride-sensitive pathway at the level of central processing [22]. However, other studies in which stimuli were delivered only to the anterior dorsal surface of the tongue failed to find a significant effect of amiloride on saltiness perception [17,20,21]. Of note, amiloride strongly reduces the sour side taste of salt solutions when subjects can use not only one response category (saltiness), but all taste qualities (saltiness, sweetness, sourness, bitterness) to describe their perception [17,19].

Further factors affecting the outcome of the sensory assays might include the impact of amiloride bitterness in establishing the taste quality of salt solutions as well as the amiloride concentration tested. It is worth noting that some studies showing the amiloride effect used high drug concentrations (500 μM and 1 mM) [12–14], whereas studies claiming a negligible effect of amiloride on salt perception adopted smaller concentrations of the drug (10–100 μM) [17,18,20,21]. The specificity of amiloride for ENaC is true only for low drug concentration (half-maximal inhibition in submicromolar range; [8]). At higher concentrations, amiloride also affects other cellular proteins, including ion channels, transporters, and receptors [3,23]. A few examples of the molecular targets other than ENaC are shown in Table 1. It is worth noting the same proteins are also found in the taste cells of laboratory rodents (Table 1, rightmost column). As nicely pointed out by Lindemann [3] "if concentrations above 10 μM are needed for half-maximal inhibition, effects of amiloride other than blockage of Na<sup>+</sup> channels need to be considered". Thus, it is possible that human testing may have yielded contrasting results due to nonspecific effects of amiloride. However, it is also possible that ENaC in humans displays a lower sensitivity to amiloride than typical ENaC, as indicated by molecular biology studies (see below).


**Table 1.** Molecular targets, other than ENaC, for the inhibitory effect of amiloride expressed by the inhibition constant (*Ki* = amiloride concentration producing 50% inhibition). Data refer to nontaste tissues. The occurrence of these molecular targets in rodents taste cells is also referenced.

#### **3. Electrophysiology**

In laboratory animals, application of NaCl solution to the tongue mucosa produces a transepithelial current due to the movement of ions across the epithelium [9,34,35]. This current is believed to be sustained mainly by Na<sup>+</sup> entering taste cells via the apical ENaC since amiloride strongly reduces it. Obviously, during NaCl stimulation, a voltage drop between mucosal and serosal side of the tongue develops, and this voltage drop can be measured with adequate instrumentation [9,34,35]. This approach has been applied to human volunteers to evaluate the contribution of the amiloride-sensitive pathway to the perceived saltiness. Electrophysiological recordings of lingual surface potential (LSP) in response to focal NaCl stimulation have provided evidence that, in some individuals, amiloride (100

μM) was able to reduce the voltage drop caused by NaCl application [36]. However, the effect was highly variable among individuals, ranging from 0% to 42% inhibition. Further studies demonstrated a positive correlation between LSP and the perceived intensity of saltiness [37]. It was also found that in expert salt tasters, amiloride (10 μM) impaired the ability to distinguish between two different salt concentrations (100 and 300 mM NaCl) [37]. Again, not all subjects exhibiting a LSP during NaCl stimulation were sensitive to amiloride, underscoring the high variability of the amiloride effect across individuals observed previously [36]. Nonetheless, these electrophysiological studies seem to confirm the psychophysical findings suggesting suppression of perceived saltiness by amiloride in some individuals when a small area of the tongue surface is stimulated [11–13].

#### **4. Molecular Biology and Immunohistochemistry**

In laboratory rodents, ENaC is an oligomeric protein made of three nonidentical subunits, named α-, β-, and γ-ENaC [38,39]. Expression cloning studies with *Xenopus* oocytes have clearly indicated that the α-subunit is required to induce channel activity, whereas the presence of the β- and γ-subunit allows maximal expression of sodium current [40]. Although it is still unknown how these subunits assemble to form ENaC in taste cells [41], all of them have been detected in taste tissues from laboratory animals by using molecular and immunohistochemical techniques [42–46]. It is important to underscore that there is also direct evidence that amiloride-sensitive taste cells do have ENaC subunits, whereas amiloride-insensitive cells do not [46]. By applying the same techniques on tissue samples from human subjects, it has been possible to establish that α-, β-, and γ-ENaCs occur in human taste papillae [47,48]. In man, however, taste tissues also express an additional δ-subunit, which is missing in rodents [48,49]. The δ-subunit is analog to the α-subunit in that its presence is necessary to form a Na+-permeable channel [50]. It is then possible that in human taste tissues, ENaC may include either an α- or δ-subunit. This subunit change may have an impact on the amiloride sensitivity, since replacement of α-subunit by a δ-subunit makes the channel 50-fold less sensitive to amiloride [50–52]. If this is the case, then the negligible effect of amiloride observed in some psychophysical studies (e.g., [17,18]) could be attributed to variations in the molecular composition of the ENaC protein [53]. It is worth noting that the expression level of ENaC subunits may vary significantly among subjects and that ENaC mRNAs are also detected in nontaste epithelium [48].

The detection of ENaC subunits in human taste tissues does not necessarily imply that this channel is involved in the initial events of sodium detection. According to the model of sodium taste detection proposed for rodents, ENaC should be found at the apical membrane of taste cells to work as a sodium receptor [3,8]. Immunohistochemical localization of ENaC subunits in human lingual epithelium has revealed that δ-ENaC is exclusively restricted to the taste pore region in both fungiform and circumvallate taste buds [48]. However, it has not been possible to establish whether this subunit localized to the apical membrane of taste cells or to tight junctions surrounding the apical ends of these cells. Unexpectedly, the other ENaC-subunits were found in the basolateral compartment of taste cells, which is involved in later stages of the sensory transduction and in intercellular communications with nerve endings and adjacent cells [5]. This raises the possibility that ENaC may serve other functions in taste buds. Recent findings indicate that ENaC plays a key role in the regulation of adult neurogenesis [54]. It is then tempting to speculate that ENaC might be involved in taste cell development since these cells continuously turnover [55,56]. Of note, another study found that δ-ENaC immunoreactivity was distributed over both the apical and basolateral ends of fungiform taste cells [49]. Although it is not clear how to reconcile these conflicting results on labeling pattern, both findings support the notion that δ-ENaC is expressed in human taste buds.

Although the model of sodium detection involving apically located ENaC explains several experimental observations, ENaC-subunits localized to the basolateral membrane could also mediate sodium reception by sensing Na<sup>+</sup> leaked through tight junctions around taste cells [5]. Thus, this paracellular pathway could be responsible for activation of taste cells even in the absence of apical ENaC. Since the basolateral compartment of taste cells is exposed to an extracellular solution

containing about 150 mM Na+, a significant diffusion of this cation through tight junctions could occur only if Na<sup>+</sup> concentration in the mucosal surface is much higher than 150 mM. It is then possible that basolateral ENaCs may be relevant for sodium detection when salt concentration in the stimulating solution exceeds plasma tonicity. The basolateral localization of ENaC subunits may be a further factor in determining the variable amiloride sensitivity observed in human studies (see above).

#### **5. Genetics**

Single nucleotide polymorphisms (SNPs) in the gene coding for the ENaC β-subunit (*SCNN1B*) are somehow associated with changes in suprathreshold taste sensitivity for NaCl solutions, but not with salt taste threshold [57]. The β-subunit does not play a role in pore formation of the channel protein, but it is assumed to modulate channel activity and to be important for channel trafficking to the cell membrane [40,58,59]. Thus, these findings indicate that variations in the β-ENaC genes may contribute to differences in salt taste perception among individuals through a possible effect on the expression of ENaC in the taste cell membrane. Interestingly, these data imply that ENaC may be involved in the recognition of NaCl at concentrations that would have an impact on the actual consumption of dietary salt, that is, at suprathreshold concentrations. As pointed out by Contreras [60], in general, people do not add salt to food in order to be able to just detect it, but do so to a preferred suprathreshold level. It is noteworthy that amiloride tends to reduce suprathreshold intensities of perceived NaCl in adult volunteers [15]. However, a significant difference in taste intensity ratings between individuals with SNPs was found only for large concentrations of NaCl, such as1M[57]. Aqueous solutions of salt above ~150 mM are not preferred by humans [61]. In animals, salt levels exceeding tonicity of blood plasma are normally not accepted [62]. High salt concentrations activate other sensory pathways in addition to the ENaC-mediated one, including an amiloride-insensitive taste pathway and trigeminal nerve endings [5,22]. These components of salt reception mediate aversion responses and work as warning mechanisms to avoid the ingestion of hyperosmotic salt solutions [5,10,63]. Thus, the findings on SNPs of the ENaC β-subunit seemingly do not fit the model involving ENaC as low-salt receptor.

#### **6. Salt Taste Enhancers**

Chemicals able to increase the sensation evoked by NaCl without being salty themselves, the so-called "salt taste enhancers", have attracted the attention of researchers for many years [64]. The reason is that these substances may be used to reduce salt content in processed foods to prevent excessive sodium intake, which is linked to the development of hypertension and subsequent pathologies [65,66]. Research on the mechanisms underlying the action of salt taste enhancers has provided some clues on the peripheral events leading to salt taste perception in humans.

Studies on human αβγ- or δβγ-ENaC functionally expressed in *Xenopus* oocytes have shown that sodium current through ENaCs is activated by salt-taste-modulating substances, such as L-arginine (Arg) [48]. By monitoring the changes of intracellular calcium levels in cultured human fungiform taste (HBO) cells, Xu et al. [67] found recently that some arginyl dipeptides, which proved to work as potent salty taste enhancers (up to 20% increase in perceived saltiness [68]), induced a significant increase in the number of cultured cells responding to NaCl. They also found that the effect required the presence of either α-ENaC or δ-ENaC. These results clearly indicate that salt taste enhancers target the human sodium receptor ENaC in both the αβγ or δβγ form. Unexpectedly, Arg was unable to stimulate cultured human taste cells [67], although it enhances the perceived saltiness in human sensory evaluations [48,68].

There are some aspects of the above studies that require keen consideration. In particular, it is remarkable how the effect of Arg may be affected by the cell system used to express human αβγ- or δβγ-ENaC. In *Xenopus* oocytes, Arg increases the sodium current through ENaCs [48], whereas in HBO cells, it is ineffective in changing intracellular calcium levels [67]. It is possible that this discrepancy may derive from the different experimental and methodological approach adopted. However, the finding that Arg potentiates ion currents through ENaC is, by itself, quite surprising. Both Arg and amiloride

bear a guanidinium group (Figure 1), which occurs as a cation in physiological conditions (pH 7.4). It is believed that the positive charge-bearing guanidinium group of amiloride penetrates part of the ENaC channel pore, causing channel blockage, whereas the pyrazine group interacts with the outer mouth of the channel [69]. The chemical similarity may suggest that Arg, like amiloride, could affect ENaC directly from the extracellular space. Indeed, Ogawa et al. [70] suggested that "the guanidinium group of Arg may interact with sodium channels in taste bud membranes". Yet, Arg enhances the current through ENaC, whereas amiloride reduces it.

**Figure 1.** Structure of L-arginine (top) and amiloride (bottom). Both chemicals bear a guanidinium group (red), which is protonated in physiological conditions. This group is believed to interact with the ENaC channel pore from the extracellular space.

#### **7. Salivary Proteins**

Proteins represent an important component of the saliva [71], the medium carrying Na<sup>+</sup> to the apical, chemosensitive tips of taste cells. Recent studies on human subjects have found a correlation between salivary serine-type endoprotease activity and sensitivity to NaCl [72]. Since serine proteases increase the activity of ENaC through proteolytic cleavage [73,74], the authors have proposed that endoproteases of the saliva might affect salt taste sensitivity by modifying ENaC functioning. Stolle et al. [75] have identified a tetrapeptide that is likely released from salivary proteins by serine-type endoprotease activity and that is able to enhance salt taste perception. This means that in the saliva of salt sensitive subjects, an endogenous salt enhancer might be produced by enzymatic cleavage. It was also found that the abundance of two salivary proteins, lipocalin-1 and lysozyme C, could be related to individuals' low sensitivity to NaCl. The authors have put forward the hypothesis that electrostatic interaction of these proteins with ENaC in taste cells may reduce the accessibility of sodium ions to ENaC [75]. These findings are clearly fascinating, but do not prove that ENaC actually functions as a sodium receptor in humans.

#### **8. Discussion**

The possible involvement of the ion channel, ENaC, in human taste reception has been investigated with different approaches. To date, however, it is not possible to provide a definitive answer as to the role of the ENaC pathway in producing salty sensations due to inconsistent findings. Perhaps the more conflicting results are from psychophysical studies involving the use of amiloride, a pharmacological probe for ENaC, to challenge saltiness perception. There are several issues regarding the adopted methodology that might be responsible for the observed discrepancies. For instance, the amiloride concentration used in sensory tests is not always adequate to avoid side effects on other ion channels and transporters. Thus, the apparent effect of amiloride in some studies (e.g., [12–14]) might be somehow misleading. Nonetheless, electrophysiological studies suggest that the application of salt solution on the human tongue induces a voltage drop across the mucosa that is similar to the one observed in laboratory animals. However, ENaC mRNAs are found also in nontaste epithelium [44,48,76], raising the possibility that the ion current crossing the mucosa could also be due to Na<sup>+</sup> diffusion through epithelial cells.

According to the model proposed for rodents, ENaC should be found at the apical membrane of taste cells, which stick out into the taste pore bathed by saliva containing taste stimuli [3,8]. Available data suggest that only the δ-subunit localizes to the taste pore region in human taste buds, whereas other ENaC subunits seem to be segregated in the basolateral compartment, beneath the apical tips of taste cells. It is not yet known whether all the subunits are required to form a functional sodium receptor [41]. Differential expression of ENaC-subunits has been described in transporting epithelia [77], suggesting that endogenous channel in vivo may require only one or two subunits to work properly. δ-ENaC expressed alone in *Xenopus* oocytes is able to mediate a membrane current [50]. Thus, the occurrence of the δ-subunit in the taste pore region of human taste buds seems to suggest that it may function as a sodium receptor. Unfortunately, the microscopic analysis has not allowed establishing with confidence whether this subunit lies in the apical end of taste cells or in the tight junctions connecting adjacent taste cells just below the taste pore [48]. Indirect evidence supporting a role for ENaC in human salt taste has been provided by genetic studies [57] and by in vitro assays on cells expressing the human αβγ- or δβγ-ENaC [48,67], although there is some inconsistency among these data. Recent analysis of the correlation between the salivary proteome and the salt sensitivity in human volunteers are seemingly consistent with a role of ENaC [72,75].

In conclusion, the available data are suggestive of possible involvement of ENaC in human sodium detection, although it is not clear whether this occurs at the beginning of the reception process (interaction between sodium receptor and Na<sup>+</sup> at the apical membrane of taste cells) or later on, after Na<sup>+</sup> has been detected. The lack of the amiloride effect in some psychophysical studies [16–18,20,21] and the presence of α-, β-, and γ-subunit only in the basolateral portion of taste buds [48] seem to favor a role for ENaC downstream of the initial receptive events. Consistent with this hypothesis is the finding that SNPs in the gene coding for the ENaC β-subunit affect suprathreshold sensitivity to salt solutions, that is, at concentration levels above the detection/recognition threshold [57]. Clearly, further research is required to obtain a coherent and thorough comprehension of the early events of sodium detection in human taste cells. This information represents the premise for understanding interindividual variability in the function of sodium taste receptors and its potential implications for eating behavior.

**Funding:** This research received no external funding.

**Acknowledgments:** This work was supported in part by Università di Modena e Reggio Emilia (FAR Dipartimentale 2018, 2019).

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Expression of Renin-Angiotensin System Components in the Taste Organ of Mice**

**Noriatsu Shigemura 1,2,\*, Shingo Takai 1, Fumie Hirose 1,3, Ryusuke Yoshida 1,4, Keisuke Sanematsu 1,2 and Yuzo Ninomiya 1,2,5,\***


Received: 2 September 2019; Accepted: 15 September 2019; Published: 19 September 2019 -

**Abstract:** The systemic renin-angiotensin system (RAS) is an important regulator of body fluid and sodium homeostasis. Angiotensin II (AngII) is a key active product of the RAS. We previously revealed that circulating AngII suppresses amiloride-sensitive salt taste responses and enhances the responses to sweet compounds via the AngII type 1 receptor (AT1) expressed in taste cells. However, the molecular mechanisms underlying the modulation of taste function by AngII remain uncharacterized. Here we examined the expression of three RAS components, namely renin, angiotensinogen, and angiotensin-converting enzyme-1 (ACE1), in mouse taste tissues. We found that all three RAS components were present in the taste buds of fungiform and circumvallate papillae and co-expressed with αENaC (epithelial sodium channel α-subunit, a salt taste receptor) or T1R3 (taste receptor type 1 member 3, a sweet taste receptor component). Water-deprived mice exhibited significantly increased levels of renin expression in taste cells (*p* < 0.05). These results indicate the existence of a local RAS in the taste organ and suggest that taste function may be regulated by both locally-produced and circulating AngII. Such integrated modulation of peripheral taste sensitivity by AngII may play an important role in sodium/calorie homeostasis.

**Keywords:** taste; sodium taste; renin; angiotensin II; angiotensinogen; angiotensin-converting enzyme

#### **1. Introduction**

The renin-angiotensin system (RAS) is a major hormone system involved in body fluid and sodium homeostasis [1]. Angiotensin II (AngII), an octapeptide hormone, is the most powerful biologically active product of the RAS and plays important roles in the regulation of vascular tone, cardiac function, and renal sodium re-absorption. AngII is also thought to be a potent stimulator of sodium appetite and preference. For example, intracerebroventricular or intravenous infusion of AngII in the rat produces dose-dependent salt appetite and stimulates sodium intake over a range of concentrations that are normally rejected [2,3]. The gustatory system provides critical information about the quality and nutritional value of food before it is ingested. Thus, changes in sodium taste sensitivity might

contribute to the ingestive behaviors induced by AngII. We recently addressed this hypothesis and revealed that AngII suppresses amiloride-sensitive salt taste responses and enhances the responses to sweet compounds via the AngII type 1 receptor (AT1) expressed in taste cells, without any effects on the amiloride-insensitive salt, sour, bitter, or umami responses [4]. These results suggest that the taste organ is a peripheral target of AngII and that AngII may function to increase sodium intake through the specific reduction in amiloride-sensitive salt taste sensitivity and increase energy intake through the enhancement of sweet responses. However, the molecular mechanisms underlying the modulation of taste function by AngII remain uncharacterized.

It is generally known that AngII is produced by the classical or circulating RAS [1]. Renin is a proteolytic enzyme released primarily from the juxtaglomerular cells of the kidney in response to a decrease in arterial blood pressure or sodium chloride level in the nephron [5]. Angiotensinogen is secreted constitutively, mainly by hepatic cells, into the circulation. Renin cleaves angiotensinogen at the N-terminus to form the decapeptide, angiotensin I (AngI). AngI is converted to AngII through the removal of two C-terminal residues by circulating angiotensin-converting enzyme-1 (ACE1), which is found in various organs, including the lung and kidney [6]. AT1, the main receptor subtype for circulating AngII, is widely distributed throughout the body including vascular smooth muscle, kidney, heart, brain, and taste organ [4,7]. AngII is degraded into smaller active peptides, AngIII, AngIV, and Ang (1–7), by endopeptidases or carboxypeptidases such as ACE2, which is a homolog of ACE1 [8].

In addition to the circulating RAS, it is now recognized that tissues such as the kidneys, brain, heart, adrenal glands and vasculature each have an organ-specific RAS [9,10]. For example, all the RAS components are present in the kidneys, and intrarenal AngII is produced independently of the circulating RAS to function as a paracrine factor via AT1. Inappropriate activation of the intrarenal RAS contributes to the pathogenesis of hypertension and renal injury [11,12]. RAS components are also present in cardiac myocytes and fibroblasts, where they synthesize AngII intracellularly [13,14]. Hyperglycemia selectively upregulates the intracellular RAS system in cardiac myocytes and vascular smooth muscle cells [15], and this is associated with cardiomyocyte apoptosis, oxidative stress, and fibrosis in diabetic rats [16]. The above findings raise the possibility that an organ-specific RAS might also exist in taste tissues and that AngII might be produced locally in response to changes in the peripheral oral environment (e.g., a change in the sodium concentration or osmolality of the saliva or the presence of chemical compounds in foods).

To explore this possibility, we utilized reverse transcription-polymerase chain reaction (RT-PCR), in situ hybridization, and double-staining immunohistochemistry to investigate the expression patterns of renin, angiotensinogen, and ACE in the taste tissues of mice under hydrated and dehydrated conditions.

#### **2. Materials and Methods**

All experimental procedures were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals established by the National Institutes of Health and approved by the Committee for Laboratory Animal Care and Use at Kyushu University, Japan. The ethical approval code for the animal experiments was A27-009.

#### *2.1. Animals*

This study used male and female C57BL/6NCrj mice (B6; Charles River, Tokyo, Japan), taste receptor type 1 member 3 (T1R3)-green fluorescent protein (GFP) mice [17] and glutamate decarboxylase 67 (GAD67)-GFP mice [18] aged 8–16 weeks of age and weighing 21–30 g. The mice were housed at a constant temperature (24 ± 1 ◦C) under a 12 h–12 h light-dark cycle (lights on at 08:00) and given access to food and water ad libitum.

#### *2.2. Reverse Transcription-Polymerase Chain Reaction (RT-PCR)*

RT-PCR was performed as described previously [4,19–21]. Mouse taste buds in the peeled epithelium were individually removed from fungiform or circumvallate papillae by aspiration with a transfer pipette. The RNeasy Plus Micro kit (Qiagen, Stanford, CA, USA) was used to purify RNAs from 100 fungiform or circumvallate taste buds from three mice or from a 1 mm × 1 mm block of epithelial tissue without taste buds. cDNAs were synthesized by RT [oligo(dT)12–18 primer] using the SuperScript pre-amplification system (Invitrogen, Carlsbad, CA, USA). Two protocols were used to prevent genomic DNA from contributing to the signal: (1) primers were chosen to span one or more introns to distinguish the PCR products from genomic DNA; The primer sequences are shown in Supplementary Table S1, and (2) RNA was handled in parallel in the presence and absence of reverse transcriptase. PCR was performed using the following conditions: 95 ◦C for 5 min (one cycle); 94 ◦C for 15 s, 58 ◦C for 30 s, 68 ◦C for 40–80 s (25–40 cycles); and 75 ◦C for 5 min (one cycle). Each 20 μL of PCR solution contained 0.5 U of Taq DNA polymerase (TaKaRa Ex TaqHS; Takara Bio, Kusatsu, Japan), 2 <sup>μ</sup>L of 10<sup>×</sup> PCR buffer containing 20 mmol/L Mg2<sup>+</sup>, 0.2 mmol/L of each deoxyribonucleotide triphosphate, and 0.6 μmol/L of each primer pair. The resulting amplification products were visualized in a 2% agarose gel with 0.5 μg/mL ethidium bromide.

#### *2.3. In Situ Hybridization*

In situ hybridization experiments were performed as described previously [4,21–24]. PCR products amplified using specific primer pairs for RAS genes (Supplementary Table S1) were purified and cloned into the pGEM T-easy vector (Promega, Madison, WI, USA) and confirmed by direct sequencing. Digoxigenin-labeled antisense RNA probes were synthesized by in vitro transcription using the digoxigenin-ribonucleic acid (DIG-RNA) Labeling Mix and T7 or SP6 RNA polymerase (Roche, Mannheim, Germany). Frozen blocks of the dissected anterior parts of the tongue embedded in optimum-cutting temperature (OCT) compound (Sakura Fine technical, Tokyo, Japan) were sectioned into 6-μm thick slices, which were mounted on silane-coated glass slides. The cryosections were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 10 min at room temperature, treated two times with 0.1% diethylpyrocarbonate in PBS for 15 min, washed with 5× saline sodium citrate buffer (SSC) for 15 min at room temperature, and then prehybridized in a hybridization buffer consisting of 50% formamide, 5× SSC, 5× Denhardt's solution, 500 μg/mL denatured salmon testis DNA and 250 μg/mL denatured baker's yeast tRNA for 1 h at room temperature. Hybridization was performed for 18 h at 58 ◦C in a hybridization buffer that included 200 ng/mL antisense (or sense) RNA probe. After hybridization, the sections were washed two times in 5× SSC for 5 min each and two times in 0.2× SSC for 30 min each at 65 ◦C. Subsequently, the sections were immersed in Tris-buffered saline (TBS) consisting of 50 mmol/L Tris-HCl (pH 7.5), and 150 mmol/L NaCl for 5 min at room temperature, treated with a blocking solution containing 0.5% blocking reagent (Roche) in TBS for 30 min, and applied with anti-digoxigenin Fab fragments-conjugated to alkaline phosphatase (AP; 1:400 dilution; Roche) in blocking solution for 60 min at room temperature. After three washes of 5 min each in Tris-NaCl-Tween 20 (TNT) buffer consisting of 50 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl and 0.05% Tween 20, the sections were immersed in AP buffer comprising 100 mmol/L Tris-HCl (pH 9.5), 100 mmol/L NaCl, and 50 mmol/L MgCl2 for 5 min. The signals were developed using 5-bromo-4-chloro-3-indolylphosphate and nitroblue-tetrazolium chloride as chromogenic substrates. Next, the reaction was stopped by washing the slides with Tris-ethylenediaminetetraacetic acid (EDTA) buffer, after which the slides were mounted. The signal specificity of the mRNA for each gene in the taste tissues was tested using a sense RNA probe as a negative control.

#### *2.4. Immunohistochemistry*

Immunohistochemistry was performed as described previously [4,20]. The dissected tongues of B6, T1R3-GFP or GAD67-GFP mice were fixed in 4% paraformaldehyde in PBS for 45 min at 4 ◦C. For immunohistochemical analyses for rennin, mice were deprived of water for 47 h (23 h water deprivation, 1 h water drinking, and 23 h water deprivation) before tongue dissection. After dehydration with sucrose solution (10% for 1 h, 20% for 1 h, and 30% for 3 h, at 4 ◦C), the tongue frozen block was embedded in OCT compound (Sakura Fine technical) and sectioned into 8-μm thick slices, which were mounted on silane-coated glass slides and air-dried. The sections were rinsed with TNT buffer, exposed to 1% blocking reagent (Roche) for 1 h at room temperature, and applied overnight at 4 ◦C with primary antibodies targeting renin (1:100 dilution; sc-27318; Santa Cruz Biotechnology, Santa Cruz, CA, USA), angiotensinogen (1:100; 11992-1-AP; Protein tech, Chicago, IL, USA), ACE1 (1:100; sc-12187; Santa Cruz Biotechnology), ecto-nucleoside triphosphate diphosphohydrolase-2 (ENTPDase2; 1:100; AF5797; R&D Systems, Minneapolis, MN, USA), epithelial sodium channel α-subunit (αENaC; 1:100; AB3530P; Millipore, Darmstadt, Germany), AT1 (1:100, sc-1173; Santa Cruz Biotechnology) or Gα-gustducin (gustducin; 1:100; sc-395; Santa Cruz Biotechnology) in 1% blocking reagent. After washing with TNT buffer, the tissues were incubated for 2 h at room temperature with peroxidase or alkaline phosphatase-conjugated secondary antibodies (1:500–1000; Jackson ImmunoResearch Laboratories, Philadelphia, PA, USA) in 1% blocking reagent, and this was followed by incubation for 30 min at room temperature with tyramide-Alexa 568 (for GFP-mice) or tyramide-Alexa 488 substrates (TSA kit; Invitrogen) for the detection of renin, ACE1 or angiotensinogen. After washing with TNT, the tissues were incubated with AP buffer for 5 min at room temperature followed by HNPP/FastRed AP substrate (HNPP fluorescent detection kit; Roche) for 40 min at room temperature to detect the signals of the counterpart. The immunofluorescence of labeled cells and GFP fluorescence were observed using a confocal laser scanning microscope (Fluoview FV-1000; Olympus Corp., Tokyo, Japan) and accompanying software. Nomarski images were also obtained in order to visualize individual cells in the taste buds.

To evaluate the number of cells expressing renin, angiotensinogen, ACE1, T1R3-GFP, GAD67-GFP, αENaC, ENTPDase2 and AT1, Nomarski images were overlaid with immunofluorescence (or GFP) images, we then counted the number of positive cells displaying apparent apical processes and/or perinuclear region in each taste bud in horizontal sections of fungiform papillae and circumvallate papillae. Image-Pro Plus v4.0 (Media Cybernetics, Rockville, MD, USA) was used to exclude artifactual signals: cells were considered positive if their signal density was greater than the mean plus two standard deviations (SDs) of that of taste cells in the negative control (primary antibodies omitted). The same cells found on contiguous sections were counted only once.

#### *2.5. Quantitative Densitometric Analysis*

To examine whether the expression of renin, a principal initiator of the RAS cascade, in taste cells was upregulated in response to dehydration, B6 mice were deprived of water for 47 h (23 h water deprivation, 1 h water drinking, and 23 h water deprivation). The mice had free access to food throughout the procedure. Dissected tongues from water-deprived and non-water-deprived mice (n = 3, each) were fixed at the same time in 4% paraformaldehyde in PBS for 45 min at 4 ◦C. After dehydration, embedding, and sectioning at 8 μm, both tongue slices were mounted on a single silane-coated glass slide. Then, renin signals were detected by single-molecule immunohistochemistry using CF568-conjugated anti-goat IgG (1:200, Biotium, Hayward, CA, USA) as a secondary antibody instead of a tyramide amplification system, in order to avoid excessive catalyzed reporter deposition of tyramide. All procedures, including image capture, were performed under the same conditions (incubation periods, temperatures, reagent volumes/concentrations, and exposure values for image capture). The taste cell in each section was delineated by comparison with adjacent sections counterstained with gustducin (a type II taste cell marker proposed to be a bitter taste-related G-protein in mouse circumvallate papillae) [25].

Measurement of the total area of a positive cell, and the mean signal intensity was performed using Adobe Photoshop CS6 (Adobe Systems, San Jose, CA, USA) [26]. Regions of interest were drawn around immuno-positive cells in the captured images, and the mean total area and the mean and median pixel intensities of the signals were determined using the Measurement tool in Photoshop. The pixel intensities were normalized against the background by subtracting the value obtained from an area outlying the positive cells in the same section. Two sections of the circumvallate papillae at regular intervals were analyzed for each mouse.

#### *2.6. Quantitative PCR (qPCR)*

To further examine whether the expression of renin in taste cells was upregulated in response to dehydration, the relative abundance of renin mRNAs in circumvallate taste papillae was examined by using a quantitative PCR method as previously described [24]. The isolated taste buds from each circumvallate papillae of each mouse were pooled (n = 4–7 water-deprived mice, n = 3–7 non-water-deprived mice). As a positive control for renin expression, a 1 mm3 block of the kidney was also collected from each mouse (n = 7 and 7, water-deprived and non-water-deprived mice, respectively). Total RNA extraction was performed as described in RT-PCR section. The RNA concentration was measured using NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, MA, USA). The purity of nucleic acids was assessed by calculating the A260/A280 absorbance ratio. SuperScript VILO Master Mix (cat. no. 11755050, Thermo Fisher Scientific) was used for cDNA synthesis. For quantitative real-time PCR, Fast SYBR Green Master Mix (Applied Biosystems, CA, USA) was used. PCR was performed as follows: 95 ◦C for 20 s (one cycle); 95 ◦C for 3 s, 60 ◦C for 30 s (40 cycles); and 95 ◦C for 15 s, 60 ◦C for 1 min, 95 ◦C for 15 s (one cycle for melting curve analysis) using the ABI StepOnePlus system (Applied Biosystems). Data were analyzed with the StepOne Software (ver. 2.3, Applied Biosystems). The presence of a single amplicon was verified by melting curve analysis, and by agarose gel electrophoresis. Data were obtained from at least three independent experiments, and all reactions were run in triplicate. The quantitative PCR data were normalized using the ΔΔCt method with GAPDH (glyceraldehyde-3-phosphate dehydrogenase) in each sample as reference. ΔΔCt values were calculated by subtracting the average ΔCt of the non-water-deprived samples from each ΔCt of both the non-water-deprived and the water-deprived samples. Fold change from non-water-deprived to water-deprived conditions was calculated as 2(−ΔΔCT). All primer pairs for renin, gustducin, keratin 8, and GAPDH were chosen such that the primers are in separate exons. The PCR primers used for each gene are presented in Supplementary Table S1.

#### *2.7. Statistical Analysis*

All values are given as the mean ± standard error of the mean. The data were statistically analyzed using Student's t-test for unpaired samples (Excel; Microsoft Corp., Redmond, WA, USA). *p* < 0.05 was considered statistically significant.

#### **3. Results**

#### *3.1. RT-PCR Reveals Renin-Angiotensin-Related Gene Expression in Mouse Taste Buds*

The expressions of renin, angiotensinogen, ACE1, and ACE2 mRNAs in the taste cells of B6 mice were examined by RT-PCR. As shown in Figure 1A, bands of the correct size (460 bp for renin, 318 bp for angiotensinogen, 398 bp for ACE1, and 303 bp for ACE2) were evident in taste papillae. Renin, angiotensinogen, and ACE1 mRNAs were expressed in fungiform and circumvallate papillae but not in tongue epithelium devoid of taste buds. Similarly, RT-PCR products for a taste cell marker, transient receptor potential channel M5 (TRPM5; 368 bp) [27–30], were also found in fungiform and circumvallate papillae but not in tongue epithelium devoid of taste buds. ACE2 mRNA was expressed in all tissues. As a positive control, β-actin mRNA (360 bp) was also detected in all tissues. All control experiments in which the reverse transcriptase enzyme was omitted (RT-) yielded negative results.

**Figure 1.** Renin, angiotensinogen (Agt) and angiotensin-converting enzyme-1 (ACE1) mRNAs are expressed in mouse taste bud cells. (*A*) Reverse transcription-polymerase chain reaction (RT-PCR) amplification of renin, Agt, ACE1, ACE2, transient receptor potential channel M5 (TRPM5), and β-actin mRNAs from fungiform papillae (FP), circumvallate papillae (VP), and tongue epithelium devoid of taste cells (ET). RT+ and RT− conditions are, respectively, with and without reverse transcriptase. M (bp): 100 bp marker ladder. (**B**) In situ hybridization detection of renin, Agt, ACE1, epithelial sodium channel α subunit (αENaC), taste receptor type 1 member 3 (T1R3), and polycystic kidney disease 2-like 1 (PKD2L1) in FP and VP of B6 mice. The sense probes served as a negative control. Dotted lines indicate the outlines of taste buds: scale bar, 50 μm.

#### *3.2. Renin-Angiotensin-Related Genes Are Localized to a Subset of Taste Bud Cells*

In situ hybridization experiments detected renin, angiotensinogen, and ACE1 mRNA in a subset of cells in the fungiform and circumvallate papillae of mice but not in surrounding epithelial cells (Figure 1B). Comparable results were obtained for three markers of taste cells: αENaC (an amiloride-sensitive salt taste receptor subunit candidate) [31], T1R3 (a sweet/umami taste receptor component) [32,33] and polycystic kidney disease 2-like 1 (PKD2L1; a sour taste-related molecule) [34,35] (Figure 1B). Control hybridizations using sense probes for renin, angiotensinogen, ACE1, αENaC, T1R3, and PKD2L1 were negative. These results, together with the RT-PCR data, strongly suggest that renin, angiotensinogen, and ACE1 are expressed in mouse taste bud cells of both the anterior and posterior tongue.

#### *3.3. Renin-Angiotensin-Related Proteins Are Co-Expressed with T1R3 or* α*ENaC in Taste Bud Cells*

Immunohistochemistry experiments detected renin, ACE1, and angiotensinogen in some spindle-shaped taste cells of the fungiform and circumvallate papillae of the mouse but not in surrounding tissues or gustatory nerves. Notably, co-expression of renin and taste cell markers was observed (Figure 2 and Table 1). In both the fungiform and circumvallate papillae, a subset of renin-positive cells expressed αENaC (ENaC/renin: 83.9% in fungiform papillae and 88.1% in circumvallate papillae) and T1R3 (as marked by T1R3-GFP; T1R3/renin: 54.7% in fungiform papillae and 49.2% in circumvallate papillae). Renin-expressing cells also showed immunoreactivity for AT1 (AT1/renin: 72.0% in fungiform papillae and 80.0% in circumvallate papillae) [4], but renin expression was not found in sour/type III cells as marked by GAD67-GFP (GAD/renin: 0% in both fungiform and circumvallate papillae).

The co-expression of ACE1 and taste cell markers was also examined (Figure 3 and Table 2). ACE1-positive cells expressed αENaC (ENaC/ACE1: 71.9% in fungiform papillae and 91.3% in circumvallate papillae) and T1R3 (T1R3/ACE1: 70.0% in fungiform papillae and 57.5% in circumvallate papillae). ACE1-expressing cells also exhibited positivity for AT1 (AT1/ACE1: 81.3% in fungiform papillae and 79.7% in circumvallate papillae). ACE1 expression was rarely observed in GAD67-expressing cells (GAD/ACE1: 0% in fungiform papillae and 2.7% in circumvallate papillae).

Figure 4 and Table 3 present data regarding the co-expression of angiotensinogen and taste cell markers. The majority of angiotensinogen-positive cells expressed T1R3 (T1R3/angiotensinogen: 73.0% in fungiform papillae and 70.7% in circumvallate papillae) and renin (renin/angiotensinogen: 75.8% in fungiform papillae and 70.6% in circumvallate papillae). A subset of cells expressing ENTPDase2, a type I taste cell marker [36], showed positive signals for angiotensinogen (ENTPDase2/angiotensinogen: 37.5% in fungiform papillae and 34.1% in circumvallate papillae). Angiotensinogen was not observed in GAD67-expressing cells (GAD/angiotensinogen: 0% in both fungiform and circumvallate papillae).

The inverse co-expression ratios (renin, ACE1, and angiotensinogen/taste cell markers) are shown in Tables 1–3. A summary of the expression patterns of renin, ACE1, angiotensinogen, and taste cell markers in the fungiform and circumvallate papillae is shown in Figure 5.

**Figure 2.** Co-expression of renin with taste receptor family 1 member 3 (T1R3), glutamate decarboxylase 67 (GAD), epithelial sodium channel α subunit (αENaC) or angiotensin II type 1 receptor (AT1) in taste bud cells. (**A**) Expression of renin in fungiform papillae (FP) and circumvallate papillae (VP) of T1R3-green fluorescent protein (GFP) mice. (**B**) Expression of renin in FP and VP of GAD67-GFP mice. (**C**) Co-expression of renin with αENaC in FP and VP of B6 mice. (**D**) Co-expression of renin with AT1 in FP and VP of B6 mice. Immunostaining for renin is shown in red (**A**,**B**) or green (**C**,**D**). GFP fluorescence in T1R3-GFP and GAD67-GFP mice is shown in green (**A**,**B**). Immunostaining for αENaC and AT1 is shown in red (**C**,**D**). The last merge panels: Nomarski images were overlaid with immunofluorescence (or GFP) merge images. Arrows indicate renin-expressing taste cells that co-express T1R3, αENaC, or AT1. Dotted lines indicate the outlines of taste buds: scale bar, 25 μm.

**Table 1.** Co-expression ratio of Renin and taste cell markers in fungiform (FP) and circumvallate papillae (VP) in mice.


AT1: angiotensin II type 1 receptor; ENaC: epithelial sodium channel α-subunit; GAD: glutamate decarboxylase; T1R3: taste receptor type 1 member 3. The number of protein1 + protein2 doubly labeled cell/number of protein2 positive cell and n = number of taste buds examined are shown in parentheses.

**Figure 3.** Co-expression of angiotensin-converting enzyme-1 (ACE1) with taste receptor type 1 member 3 (T1R3), glutamate decarboxylase 67 (GAD), epithelial sodium channel α subunit (αENaC) or angiotensin II type 1 receptor (AT1) in taste bud cells. (**A**) Expression of ACE1 in fungiform papillae (FP) and circumvallate papillae (VP) of T1R3-green fluorescent protein (GFP) mice. (**B**) Expression of ACE1 in FP and VP of GAD67-GFP mice. (**C**) Co-expression of ACE1 with αENaC in FP and VP of B6 mice. (**D**) Co-expression of ACE1 with AT1 in FP and VP of B6 mice. Immunostaining for ACE1 is shown in red (**A**,**B**) or green (**C**,**D**). GFP fluorescence in T1R3-GFP and GAD67-GFP mice is shown in green (**A**,**B**). Immunostaining for αENaC and AT1 is shown in red (**C**,**D**). The last merge panels: Nomarski images were overlaid with immunofluorescence (or GFP) merge images. Arrows indicate ACE1-expressing taste cells that co-express T1R3, αENaC, or AT1. Dotted lines indicate the outlines of taste buds: scale bar, 25 μm.

**Table 2.** Co-expression ratio of ACE1 and taste cell markers in fungiform (FP) and circumvallate papillae (VP) in mice.


ACE1: angiotensin-converting enzyme-1; AT1: angiotensin II type 1 receptor; ENaC: epithelial sodium channel α-subunit; GAD: glutamate decarboxylase; T1R3: taste receptor type 1 member 3. The number of protein1 + protein2 doubly labeled cell/number of protein2 positive cell, and n = number of taste buds examined are shown in parentheses.

**Figure 4.** Co-expression of angiotensinogen (Agt) with taste receptor type 1 member 3 (T1R3), glutamate decarboxylase 67 (GAD), nucleoside triphosphate diphosphohydrolase-2 (ENTPDase2) and renin in taste bud cells. (**A**) Expression of Agt in fungiform papillae (FP) and circumvallate papillae (VP) of T1R3-green fluorescent protein (GFP) mice. (**B**) Expression of Agt in FP and VP of GAD67-GFP mice. (**C**) Co-expression of Agt with ENTPDase2 in FP and VP of B6 mice. (**D**) Co-expression of Agt with renin in FP and VP of B6 mice. Immunostaining for Agt is shown in red. GFP fluorescence in T1R3-GFP and GAD67-GFP mice and immunostaining for ENTPDase2 and renin are shown in green. The last merge panels: Nomarski images were overlaid with immunofluorescence (or GFP) merge images. Arrows indicate Agt-expressing taste cells that co-express T1R3, ENTPDase2, or renin. Dotted lines indicate the outlines of taste buds: scale bar, 25 μm.

**Table 3.** Co-expression ratio of Agt and taste cell markers in fungiform (FP) and circumvallate papillae (VP) in mice.


Agt: angiotensinogen; ENTPDase2: ecto-nucleoside triphosphate diphosphohydrolase-2; GAD: glutamate decarboxylase; T1R3: taste receptor type 1 member 3. The number of protein1 + protein2 doubly labeled cell/number of protein2 positive cell, and n = number of taste buds examined are shown in parentheses.

**Figure 5.** Summary of the patterns of co-expression between renin, angiotensinogen (Agt), angiotensin-converting enzyme-1 (ACE1), angiotensin II type 1 receptor (AT1) and type-specific markers of taste cells [nucleoside triphosphate diphosphohydrolase-2 (ENTPDase2) for cell type I, taste receptor type 1 member 3 (T1R3) for cell type II, and glutamate decarboxylase 67 (GAD) for cell type III] in both the fungiform and circumvallate papillae of mice.

#### *3.4. Renin Expression Is Upregulated in the Taste Buds Cells of Water-Deprived Mice*

To examine whether water deprivation leads to changes in the abundance of renin, we performed immunohistochemistry and quantitative image analyses using circumvallate papillae from non-deprived and water-deprived mice. In non-deprived mice (used as the control), immunoreactivity for renin was observed in the apical process regions of taste cells with weaker expression in the cell body regions (Figure 6A). In water-deprived mice, immunoreactivity for renin was observed throughout some cells, i.e., from apical to basal regions (Figure 6B). There were no significant differences between non-deprived and water-deprived mice in the mean number of positive cells per bud [2.75 ± 0.2 (n = 28) vs. 2.85 ± 0.2 (n = 27)] or the mean area (pixels) of a positive cell [576.7 ± 25.0 (n = 28) vs. 563.2 ± 27.4 (n = 28)] (Figure 6C,D). However, significant differences were observed between control and water-deprived mice in both the mean pixel intensity (arbitrary units) [14.5 ± 0.9 (n = 28) vs. 19.2 ± 1.5 (n = 28), p < 0.05] and the median pixel intensity (arbitrary units) [10.3 ± 0.6 (n = 28) vs. 15.4 ± 1.4 (n = 28), p < 0.01] of the immuno-positive cells (Figure 6E,F).

**Figure 6.** The level of renin expression in taste cells is upregulated after water deprivation. (**A**) In non-deprived (control) mice, renin was expressed at the apical process regions of taste cells and more weakly in the cell body regions (lower image: higher magnification of the upper image). (**B**) In mice deprived of water for 47 h, renin expression was observed throughout some cells, i.e., from apical to basal regions. Water deprivation did not change the mean number of renin-positive cells per taste bud [n = 28 taste buds from three non-water deprived mice (ND), n = 27 taste buds from three water-deprived mice (WD)] (**C**) or the mean area of renin-positive cells (n = 28 taste cells from ND, n = 28 taste cells from WD) (pixels: pxs) (**D**). However, significant differences were detected in the mean pixel intensity (**E**) and median pixel intensity (**F**) of renin-positive cells (arbitrary unit: AU). Gustducin immunoreactivity was detected throughout positive cells from the apical to the basal regions in both non-water-deprived mice (**G**) and water-deprived mice (**H**). Water deprivation did not change the mean number of gustducin-positive cells per taste bud (n = 23 taste buds from ND, n = 23 taste buds from WD) (**I**), mean area of gustducin-positive cells (n = 32 taste cells from ND, n = 38 taste cells from WD) (pixels: pxs) (**J**), mean pixel intensity of gustducin-positive cells (**K**) or median pixel intensity (**L**) of gustducin-positive cells (arbitrary unit: AU). (**M**) Quantification of mRNA expression of renin, gustducin and keratin 8 by real-time PCR in the circumvallate taste papilla (CV) and Kidney (as the control tissue for renin expression) from ND and WD. Data were obtained from at least three independent experiments (n = 3–7 mice) per group each PCR assays were performed in triplicate. The quantitative PCR results were normalized using the ΔΔCt method with GAPDH (glyceraldehyde-3-phosphate dehydrogenase) in each sample as the reference, and shown as fold change of mRNA expression compared to ND. \* p < 0.05, \*\* p < 0.01 for comparisons between ND and WD groups (Student's t-test). All data are presented as the mean ± standard error of the mean. Dotted lines indicate the outlines of taste buds: scale bar, 25 μm.

We also examined differences in the abundance of gustducin protein (a taste-specific G-protein mostly expressed in bitter taste cells in mouse circumvallate papillae) in taste cells between control and water-deprived mice. Immunoreactivity for gustducin was observed throughout the apical-to-basal regions of a subset of taste cells in both the non-deprived and water-deprived groups (Figure 6G,H). We observed no significant differences between the non-deprived and water-deprived groups in

the mean number of positive cells per taste bud [6.13 ± 0.3 (n = 23) vs. 6.09 ± 0.3 (n = 23)], the mean area of a positive cell [548.5 ± 20.4 (n = 32) vs. 532.8 ± 20.5 (n = 38)], the mean pixel intensity of a positive cell [22.8 ± 1.1 (n = 32) vs. 23.8 ± 1.2 (n = 38)] or the median pixel intensity of a positive cell [19.6 ± 1.1 (n = 32) vs. 20.6 ± 1.2 (n = 38)] (Figure 6I–L).

Next, we asked whether renin mRNA expression in circumvallate taste bud cells is upregulated after water-deprivation by using quantitative PCR. mRNA levels of genes were normalized for GAPDH as the endogenous reference gene and shown as fold change of mRNA expression compared to non-water-deprived groups (Figure 6M). Statistical analysis by *t*-test revealed a significant difference in relative renin mRNA expression in circumvallate taste buds between control and water-deprived mice [1.36 ± 0.62 (n = 3 mice) vs. 4.60 ± 0.46 (n = 4), p < 0.01]. In the Kidney as the control tissue for renin expression, renin mRNA was significantly increased in the water-deprived group compared with the non-water-deprived group [1.11 ± 0.22 (n = 7) vs. 5.53 ± 0.51 (n = 7), p < 0.01) (Figure 6M). We also performed quantitative PCR analysis of gustducin and keratin 8 (a pan-taste cell marker) in the circumvallate taste bud cells, and observed no significant differences in the mRNA expression between control and water-deprived groups [gustducin: 1.14 ± 0.33 (n = 5) vs. 1.02 ± 0.20 (n = 7), keratin 8: 1.37 ± 0.43 (n = 6) vs. 1.51 ± 0.52 (n = 6)] (Figure 6M). Together, these results suggest that renin expression is upregulated in taste cells in response to dehydration.

#### **4. Discussion**

In the present study, we found that three RAS components, namely renin, ACE1, and angiotensinogen, were present in the taste buds of fungiform and circumvallate papillae and co-expressed with αENaC (a salt taste receptor), T1R3 (a sweet taste receptor component) [31–33] and AT1 [4]. Furthermore, significantly increased levels of renin expression were observed in taste cells after water deprivation of the mouse. These results suggest that the taste organ contains a local RAS that may be capable of producing AngII within the taste buds. AngII may function as a modulator of amiloride-sensitive salt and sweet taste sensitivities via AT1 in an autocrine and paracrine manner.

In animals with AngII-dependent hypertension, the intrarenal AngII level is higher than that which can be explained on the basis of equilibration with circulating AngII [12]. This suggests that AngII can be synthesized within the kidney. Indeed, all components of the RAS are present in the distal nephron [11,37]. Principal cells in the connecting tubule and the cortical collecting duct abundantly express renin, a key initiator of the RAS, which cleaves the N-terminal end of angiotensinogen to generate AngI in the tubular fluid. Renin immunolocalizes predominately to the apical side of the cytoplasm of principal cells. Quantitative histological analysis revealed that mice exhibited minimal renin immuno-staining (below the detection level) in the connecting tubule under high sodium diets, in contrast, overnight sodium restriction led to a marked increase in the number of the renin-positive connecting tubule cells. Furthermore, the amount of renin mRNA was shown to increase in response to sodium restriction [11,37]. In our study, renin immunostaining predominated at the apical regions of taste cells, which face the oral cavity and are available to sense various chemical compounds in saliva or foods (Figure 6A). Moreover, significantly increased levels of renin in taste cells were observed in water-deprived mice (Figure 6E,F,M), which may be consistent with the results observed in connecting tubule cells in the kidney [11]. These results indicate that renin synthesis initiating the RAS cascade occurs in taste cells and may be regulated in response to factors in the oral environment such as sodium concentration (osmolality) in the saliva or chemical compounds in foods, which are sensed at the apical regions of the taste cells.

A previous study showed that amiloride-sensitive NaCl taste responses were suppressed 10–30 min after intraperitoneal injection of AngII, suggesting that systemic AngII produced by the circulating RAS is able to modulate NaCl taste responses in taste cells [4]. Thus, there may be two pathways regulating taste, namely the circulating RAS and the local RAS. What would be the advantage of the co-existence of these two RASs? One possibility is that the local RAS assists in maintaining the constancy of body NaCl balance through acute changes in AngII production in response to fasting or random perturbations in the oral cavity during feeding and drinking. In other words, the local RAS could help to avert decreases in body sodium levels or prevent excessive sodium/calorie consumption independently of changes in body NaCl balance. However, once NaCl balance has been changed, the distal nephron NaCl concentration would serve to regulate the circulating RAS, which supports the maintenance of prolonged deviations of body NaCl concentration from the normal set point [5]. Such temporally integrated regulation of taste sensitivity by local AngII (short-term feedforward regulation predicting changes in body fluid composition) and systemic AngII (long-term negative feedback regulation in response to changes in body salt/water balance) may play an important role in sodium/energy homeostasis. ACE2, which degrades Ang II to Ang (1–7) to oppose the actions of Ang II, is present not only in fungiform and circumvallate papillae but also in the tongue epithelium (Figure 1A). This observation suggests that AngII generated locally in taste buds can be rapidly degraded by ACE2, which would support the hypothesis of short-term regulation of taste sensitivity by a local RAS.

Multiple lines of evidence from studies using molecular approaches indicate that the basic taste qualities (sweet, salty, bitter, sour, and umami) are mediated by distinct taste cells expressing unique taste receptors [38]. Sweet, umami and bitter substances activate G-protein coupled receptors (T1R2+T1R3 for sweet [32,33], T1R1+T1R3 for umami [33] and T2Rs for bitter [39]) and subsequent common signaling pathways involving phospholipase C-β [28], the type 3 inositol 1,4,5-trisphosphate receptor [40] and TRPM5 [27–30]. Salty and sour substances are believed to activate channel-type receptors (ENaC for amiloride-sensitive salt taste [31,41] and PKD2L1/1L3 for sour taste [34,35]). Each of these taste receptors is expressed in a different set of taste cells [38]. The present study demonstrated that all three of the RAS components studied are co-expressed with αENaC or T1R3 but not GAD67, suggesting that taste-regulated AngII production by the local RAS may be mediated by αENaC-expressing, amiloride-sensitive salt taste cells, and T1R3-expressing sweet taste cells. The expression patterns of the RAS components in the taste organ may relate to previous observations that AngII suppresses amiloride-sensitive salt taste responses and enhances sweet taste responses without any effects on bitter, umami and sour responses [4]. The relationship between salt and sweet preferences via AngII signaling, analogous to that between reduced salt-sensitive neural responses and increased sugar-sensitive neural responses to dietary NaCl in the rat nucleus of the solitary tract after intracerebroventricular renin infusion on chronic deoxycorticosterone acetate (DOCA, the precursor of aldosterone) treatment [42], or that between sodium and glucose absorption via sodium-glucose cotransporter-1 in the small intestine [43], may optimize sodium and calorie intake via the taste system.

The macula densa is a group of 15–20 epithelial cells in the distal convoluted tubule of the kidney. These cells play a critical role in sensing changes in tubular fluid composition and sending signals to the juxtaglomerular apparatus that controls renin release [44]. It has been shown that increasing the NaCl concentration at the macula densa suppresses renin release, whereas reducing the NaCl concentration results in a prompt stimulation of renin release [45]. NaCl sensing by the macula densa involves apical NaCl transport mechanisms, including the furosemide-sensitive Na+/K+/2Cl−cotransporter (NKCC), which is the primary NaCl entry mechanism [46,47]. The apical membranes of macula densa cells also express Na+/H<sup>+</sup> exchanger (NHE), which participates in Na<sup>+</sup> transport and the regulation of intracellular pH and cell volume [48,49]. It has been reported that NHE1 and NHE3 were detected in taste receptor cells [50]. It is possible that NKCC or NHE may participate in sensing changes in NaCl concentration at the cell membrane of taste cells.

It is generally accepted that an elevation in the circulating level of AngII inhibits renin secretion from the juxtaglomerular apparatus of the kidney. However, intrarenal AngII increases renin mRNA and protein levels in the distal nephron [51], indicating positive-feedback regulation of intrarenal RAS by AngII. In experiments involving renal cross-transplantation between global AT1-knockout mice and wild-type controls, AngII was shown to cause hypertension through stimulation of AT1 receptors in the kidney [52]. Overexpression of renin in the collecting duct caused spontaneous hypertension [53]. The Aldosterone/NaCl-induced RAS functional impairment also, not only caused a reduction of the salt

taste sensitivity, but also salt-sensitive hypertension in rat [54]. These results suggest that intrarenal RAS contributes to the pathogenesis of hypertension. The existence of such a positive-feedback mechanism in the taste organ potentially would have broad implications, since continuous activation of the taste organ RAS might be associated with hypertension induced by excessive salt consumption through sustained low salt taste sensitivity.

#### **5. Conclusions**

We have demonstrated that the taste organ has three major components of the RAS, namely renin, angiotensinogen, and ACE1, which would enable AngII to be produced locally in the taste buds. Expression analyses showed that the RAS components are co-expressed with αENaC or T1R3 in a subset of taste cells. Renin immunoreactivity was detected at the apical regions of taste cells, which face the oral cavity and thus are exposed to various chemical compounds in saliva or food. Furthermore, renin synthesis in taste tissue was significantly upregulated in response to water deprivation. Taken together, these results suggest the existence of a previously unidentified local RAS in the taste organ. The specific reduction of amiloride-sensitive salt taste sensitivity and enhancement of sweet taste sensitivity may be mediated by both locally-produced AngII (temporal feedforward regulation) and circulating AngII (continuous negative feedback regulation). Such an integrated regulation of peripheral taste sensitivity by AngII may play an important role in sodium/calorie homeostasis.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6643/11/9/2251/s1, Table S1: PCR primers.

**Author Contributions:** Conceptualization—N.S.; Methodology—N.S.; Validation—S.T. and K.S.; Investigation—N.S., S.T., F.H., R.Y. and K.S.; Data curation—N.S., S.T., F.H., K.S. and R.Y.; Writing—Original Draft Preparation—N.S.; Writing—Review & Editing—N.S. and Y.N.

**Funding:** This research was supported in part by Grants-in-Aid for Scientific Research 15K11044, 18K19652 and 19H03818 (N.S.), 18077004 and 18109013 (Y.N.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. The authors thank OXMEDCOMMS (www.oxmedcomms.com) for writing assistance.

**Acknowledgments:** The authors thank S. Damak (Nestlé Research Center, Lausanne, Switzerland) and R.F. Margolskee (Monell Chemical Senses Center, Philadelphia, PA, USA) for providing the original stock of T1R3-GFP mice and Y. Yanagawa (Gunma University Graduate School of Medicine, Gunma, Japan) for providing the original stock of GAD67-GFP mice.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
