Chronic exposure to liquid sucrose and dry sucrose diet have differential effects on peripheral taste responses in female rats

Abstract

Sugar-sweetened beverages are the major source of added calories in the Western diet and their prevalence is associated with obesity and metabolic disruption. Despite the critical role of the taste system in determining food selection and consumption, the effects of chronic sucrose consumption on the peripheral taste system in mammals have received limited attention. We offered female Sprague Dawley rats free access to water and one of three diets for up to 40 days: (1) sucrose-free chow or “NS” diet; (2) a high-sucrose dry diet or “HS”; or (3) 30% sucrose solution and the NS diet, designated “LiqS” diet. Sucrose consumption by LiqS rats gradually increased and by day 14 was equal to that of HS rats. Food intake decreased in LiqS rats, but their energy intake remained higher than for NS or HS rats. There was no significant difference in weight gain of the groups during the study. Recordings from the chorda tympani nerve (CT), which innervates taste buds on the anterior tongue, revealed decreased responses to 1 M sucrose in both LiqS and HS rats and to acesulfame K and salt tastants in LiqS rats after 40 days on diet. Umami, bitter, and acid response magnitudes were unchanged in both groups. These results demonstrate that chronic sucrose exposure inhibits taste responses to higher concentrations of sweet stimuli. More surprisingly, CT responses to NaCl and 0.5M NaAc were significantly reduced in rats on the LiqS diet. Thus, the physical form of the diet influences taste responsiveness to salt and sweet taste function. These data suggest that taste buds are previously unappreciated targets of chronic sucrose consumption.

1. INTRODUCTION

Sugar sweetened beverage (SSB) consumption has been implicated in the development of metabolic disease and obesity (Flegal et al. 2010; Malik et al. 2010; Stanhope 2016; Sundborn et al. 2019) although the association remains controversial (Momin and Wood 2018; Trumbo and Rivers 2014). Currently, 50% of the US population drinks at least one SSB per day and 7% consume three or more (Rosinger et al. 2017). As recently reviewed, the consumption of SSBs is associated with greater weight gain compared to solid sugars (Sundborn et al. 2019).

The detection of sweetness in foods and drinks preferred by humans and rodents is initiated in oral taste buds (Fernstrom et al. 2012; Roper and Chaudhari 2017) and transmitted to the CNS through three cranial nerves. Taste buds are heterogeneous signaling organs composed of taste receptor cells sensitive to sweet, umami, bitter, sour, salt and likely fat stimuli (Besnard et al. 2016; Gilbertson and Khan 2014; Pittman 2010). Type I taste cells are glial-like, while Type II taste receptor cells bear G-protein coupled receptors activated by complex stimuli mediating appetitive sweet and umami and aversive bitter taste qualities. Sour taste is transduced by a decreased pH in Type III taste cells, while salt taste is generated by the passage of sodium ions through epithelial sodium channels (ENaC) and at least one additional, unidentified pathway (reviewed in Roper and Chaudhari 2017). The chorda tympani (CT) nerve, a branch of cranial nerve VII, innervates taste receptor cells located in fungiform taste buds on the anterior tongue. The recording of CT responses is useful for testing the function of a large population of taste buds in response to multiple taste qualities. CT nerve activity contributes to sweet intake and preference as demonstrated by differences in CT responses from sweet-sensitive vs. sweet-insensitive (i.e. C57BL/6ByJ vs. 129P3/J) mouse strains and behavioral changes after nerve sectioning (Spector et al. 1996; Vigorito et al. 1987). Post-ingestive input plays a greater role in the daily consumption of sugar solutions (Glendinning et al. 2010), though chronically increased sugar intake could affect dietary choice by changing neural responsivity to sweet or non-sweet stimuli.

Despite demonstration of a critical role for taste in modulating food choice and intake (Smeets et al. 2012; Spetter et al. 2014), the effects of chronic sucrose consumption on the peripheral taste system are largely unexplored. In one study, reported almost 30 years ago, F344 rats were divided into groups receiving one of six concentrations of sucrose solution or water for 4 days a week over their lifespan. The intake of sucrose solution gradually increased in groups offered either 0.5 or 1.0 M sucrose solution as did body weight compared to rats receiving water alone. In periodic 23-hr tests, lick rate per bout and the percentage of daytime bouts were elevated in rats consuming higher concentrations of sucrose solution, suggesting an elevated preference for sucrose (Smith and Wilson 1989). In another study, lingual expression of the sweet receptor subunit, T1r3, was reduced in alcohol non-preferring rats offered 5% sucrose for three weeks compared to those maintained on water alone, but this treatment did not alter CT responses to NaCl or sucrose in alcohol-preferring or non-preferring rats (Coleman et al. 2011).

Rats consumed higher amounts of 20% sucrose in solution and gained more weight over 7 days than when previously offered granular sucrose for 7 days (Sclafani and Kramer 1983). Togo et al. (2019) recently reported that C57BL/6 mice offered 30% sucrose solution plus a low sucrose diet for 8 weeks were fatter than those offered a dry high sucrose diet. The mice offered sucrose solution consumed large volumes of sucrose each day even when they also had access to water. Lingual expression of the sweet taste receptor subunits, T1r2 and T1r3, was elevated in mice offered dry sucrose diet, but not in those drinking sucrose solution. Thus, there is evidence from rodent studies that short and long-term consumption of sucrose solution is preferred and induces more weight gain than sucrose in solid form but the effects on peripheral taste responsivity are not well understood.

Obesity is generally associated with decreased peripheral taste responsiveness. Taste bud gene expression, proliferation, and density are decreased in obese rodents (Chen et al. 2010; Duca et al. 2014; Kaufman et al. 2018) and humans (Archer et al. 2019; Kaufman et al. 2019). Calcium responses to the artificial sweeteners, acesulfame (ace)K and saccharin are also reduced in isolated taste receptor cells from obese mice (Maliphol et al. 2013). CT responses to sugars, however, are elevated in diabetic (Ninomiya et al. 1995; Shimizu et al. 2003) and non-diabetic (Shimizu et al. 2003) rodent models of obesity. Obese rodents generally exhibit decreased behavioral responses to artificial sweeteners (Chen et al. 2010; Maliphol et al. 2013), sucrose (Duca et al. 2014; Johnson 2012), and fat (Pittman et al. 2015). In sum, obesity distinctly alters taste buds and CT responses resulting in reduced behavioral responsiveness in rodents. Contradictory changes, or no change, in psychophysical and hedonic responses are observed, further highlighting the complexity of the relationship between taste and obesity (Bartoshuk et al. 2006; Donaldson et al. 2009).

We have previously shown that male and female Sprague Dawley rats voluntarily consume a significant percentage of their calories from 30% sucrose solution (i.e. 0.88M) but reduce their intake of dry diet so that energy intake is unchanged (Harris 2019; 2018). Although the sucrose intake of rats offered the sucrose solution is matched by rats given free access to a dry high sucrose diet, meal patterns differ in that rats drink sucrose solution in frequent small meals during the light and dark periods, while rats on high-sucrose dry diet eat most meals during the dark period (Harris 2018). Sucrose preference has not been tested in these animals, but the rats offered 30% sucrose solution maintain a high level of intake even though they also have access to a water bottle, suggesting that they continue to find the sucrose preferable to water (Harris 2018). An advantage of this model for testing peripheral taste responsiveness is that rats chronically consuming 30% sucrose solution do not have significantly increased energy intake or weight gain compared to controls fed normal chow and water (Harris 2018).

We determined whether neurophysiological taste responses to sweet, umami, salty, bitter or sour stimuli change in rats consuming high levels of sucrose. Neural taste responses were measured in rats offered a high sucrose diet presented as 30% (0.88M) sucrose solution or dry diet with 67% calories from sucrose. This level of sucrose in the dry diet was selected because in a previous study with both male and female rats we found that the consumption of a 30% sucrose solution gradually increased and after 25 days stabilized at a level equal to that of rats offered the dry high sucrose diet (Harris 2018). Female rats were used because females offered LiqS diet become leptin resistant faster than males (Harris and Apolzan 2015; Harris 2018) which suggests that diet induced changes in taste response may also occur more rapidly in females.

METHODS

2.1. Animals and diets

Female Sprague Dawley rats (Harlan Sprague Dawley, Indianapolis, IN), weighing 198–212 g at the start of the experiment, were housed individually in hanging wire cages with a Nylabone as enrichment and had free access to food and water throughout the experiment. Room temperature was maintained at 22 ±1 °C with a 12 hour light: dark cycle starting at 6:00 a.m. The Institutional Animal Care and Use Committee at the Medical College of Georgia at Augusta University approved all procedures which followed guidelines set by the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

2.2. Diets

Weight matched groups of animals were offered one of three diets. Control rats consuming no sucrose had free access to sucrose-free chow (“NS”; D 11724, Research Diets Inc., New Brunswick, NJ). A second group had free access to chow containing 66.6% kcal as sucrose (“HS”; DL 11725 Research Diets Inc.), while the experimental group had access to NS diet and 30% sucrose solution (“LiqS”; Kroger Sugar, Hood Packing Corporation, Hamlet, NC). Dry diets were equicaloric as previously detailed (Harris 2019). Thirty-four rats were included in the initial pilot experiment and two rats per day began dietary treatment to make it technically feasible to obtain neural recordings at staggered endpoints of 40 days after initiating diets. Body weights, chow intakes corrected for spillage, and sucrose intakes were measured each morning starting at 7:30 a.m. Spillage of sucrose solution was not measured as previous experience has found this to be negligible. Electrophysiological results indicated that dietary effects were not apparent until Day 40, therefore a second group of 18 rats (6 rats/diet) were included to allow for a more in-depth analysis of taste responses at this time point.

2.3. Electrophysiology

Rats were anesthetized with ketamine (50 mg/kg) and xylazine (10 mg/kg) i.p. then placed on a water-circulating heating pad to maintain body temperature at 36–39 °C. Supplemental anesthetic was injected as needed in response to toe pinch. Rats were tracheotomized and the hypoglossal nerves sectioned bilaterally to stop tongue movement. The right or left CT nerve was dissected using a lateral approach as previously reported (Shi et al. 2012; Steen et al. 2010). The desheathed CT nerve was placed on a platinum electrode with the indifferent electrode in nearby tissue. Neural activity was amplified (Grass Instruments, Warwick, RI), integrated with a time constant of 1.5 sec, and the summated signal monitored with PowerLab hardware and software (AD Instruments, Colorado Springs, CO).

We recorded CT responses during stimulation of the anterior tongue with NaCl (0.10 and 0.50M), sodium acetate (NaAc; 0.10M and 0.5M), sucrose (0.1, 0.5 and 1.0M), saccharin (0.05 and 0.1M), aceK (0.025 and 0.05M), 0.3M monopotassium glutamate (MPG), 1.0M glucose, 16% polycose, 0.01M quinine hydrochloride (QHCl), and 0.01M HCl mixed in distilled water. These stimulus concentrations were tested because they elicit intermediate and maximal CT responses in mice (Danilova and Hellekant 2003; Zhu et al. 2014). Chemicals were purchased from Sigma (St. Louis, MO). Approximately 3 ml of each stimulus were applied to the tongue by hand with a syringe. Approximately 25 sec later the tongue was rinsed with distilled water for at least 1 minute. We also recorded responses to NaCl dissolved in and rinsed with 50 μM amiloride.

Taste response magnitudes were calculated by measuring the height of the summated, integrated traces at 20 sec after stimulus application and subtracting the baseline. Test responses were expressed relative to mean responses to 0.5M NH₄Cl applied at the beginning and end of the stimulus series. Only series bracketed by 0.5M NH₄Cl responses that deviated by <10% were included in analyses.

2.4. Type II taste cell immunofluorescence and analysis

Tongues were collected from a subset of rats on NS (n=3) or LiqS (n=4) diet following nerve recordings at day 40 post-treatment. Tongues were removed immediately anterior to the intermolar eminence, frozen in OCT mounting medium (Thermo Fisher Scientific, Waltham, MA) and cryosectioned at 8 μm. We identified type II taste cells expressing the transient receptor potential channel subfamily M member 5 (TRPM5) within taste buds immunopositive for keratin 19, a marker for mature taste receptor cells in rat (Wong et al. 1994). Coronal sections were fixed in acetone then rinsed, blocked with 2% goat serum, and incubated in Rabbit TRPM5 (1:500; catalog #18027–1-AP; Proteintech, Rosemont, IL) and mouse K19 (1:400; catalog #C7159; Sigma-Aldrich, St. Louis, MO) overnight at 4 °C. We incubated slides in goat Alexa Fluor 488 and 594 secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) for 1 hour at room temperature, counterstained with DAPI (Life Technologies, Foster City, CA, USA) and coverslipped with Fluoro-Mount G (Southern Biotech, Birmingham, AL, USA). Non-specific staining in the absence of TRPM5, K19, or both antibodies was minimal.

Images were acquired at 90x using a digital camera (Cool Snap, Roper Scientific, Tucson, AZ, USA) mounted on an Olympus BX50 microscope equipped with epifluorescence (Olympus). We analyzed fungiform taste buds on the anterior third of the tongue where taste bud density is highest and taste bud size is more uniform compared to caudal regions of the fungiform lingual field (Krimm and Hill 1998). The center section of the taste bud was identified for analysis, either by the presence of a taste pore or by tracking taste buds through serial sections. We used MetaMorph Software (Universal Imaging Corporation / Molecular Devices; Downington, PA, USA) to digitally threshold TRPM5+ pixels within taste buds (n=1–6/rat) defined by K19 immunoreactivity. More specifically, K19 taste buds were outlined, the red channel was turned off, and pixels marked by green immunofluorescence were thresholded to measure TRPM5+ pixels / taste bud.

2.4. Statistical analyses

Hypotheses that were tested, the experimental design, and the statistical analyses were planned in advance. Statistica software version 9.0 (StatSoft, Tulsa, Oklahoma) was used to perform a repeated-measures ANOVA analysis of daily body weight, energy intake and sucrose intake data. Tukey’s HSD post-hoc test was used to identify differences between individual groups on specific days when the ANOVA indicated a significant overall effect or interaction.

Prism software (version 8.1.2; Graphpad Software, Inc., La Jolla, CA, USA) was used to analyze CT responses and TRPM5 expression. Two-way, mixed-model ANOVAs followed by Tukey’s tests were used to analyze responses to series of NaCl with and without amiloride, sucrose, saccharin, and AceK. Responses to single concentrations of stimuli were compared with one-way ANOVAs with Tukey’s multiple comparison tests. Absolute NH₄Cl values and the percent of TRPM5+ pixels / taste buds in NS vs. LiqS rats was compared with Student’s t tests. The result from one taste bud in the NS group was identified as a significant outlier using Grubb’s test and not included in the comparison between dietary groups. Data are reported as means ± sem and an a level of P ≤ 0.05 was considered statistically significant.

3. RESULTS

3.1. Body weight and intake

Data from all of the female rats that were on diet for 40 days (n=8 or 9) was used to test for dietary effects on body weight and sucrose and energy intakes. The LiqS rats consumed significant amounts of 30% sucrose solution throughout the experiment. Initially the amount of sucrose consumed by LS rats was lower than that of HS rats, but intake gradually increased and by day 13 energy consumed as sucrose was similar in HS and LiqS rats (Figure 1A, Table S1). Total sucrose intake during the experiment was comparable for the two groups of rats (LiqS rats: 1412 ± 20 kcal/40 d. HS rats: 1137 ± 115 kcal/40 d). The LS rats reduced their intake of the dry NS diet to partially compensate for the calories consumed as sucrose solution. However, this drop in intake did not result in a precise compensation and the daily energy intake of the LiqS rats was greater than that of HS and NS rats on a majority of days of the experiment (Figure 1B, Table S1). The total energy intake during the 40 days of the experiment was about 25% greater (df, 2, F, 36, p<0.0001) for LiqS than HS or NS rats (NS rats: 2167 ± 42 kcal/40 d, HS rats: 2120 ± 30 kcal/40 d, LiqS rats: 2665 ± 59 kcal/40 d). By the end of the experiment the LiqS rats weighed significantly more than HS (p<0.03), but not NS (p<0.09) rats (Figure 1C, Table S1) (NS rats: 248 ± 4 g, HS rats: 242 ± 17 g, LiqS rats: 259 ± 7 g), however, this may, at least in part, have reflected a small, non-significant difference in weight of the groups at the start of the experiment (NS rats: 202 ± 3 g, HS rats: 201 ± 5 g, LiqS rats: 205 + 3 g) as there was no significant difference in the amount of weight gained during the experiment (Figure 1C insert).

Figure 1:

Daily sucrose intake, energy intake, body weight and weight gain of rats that remained in the experiment for 40 days. Data are means ± sem for 8 or 9 rats and were analyzed using repeated-measures ANOVAs followed by Tukey’s tests. Asterisk indicate a significant difference between LiqS and HS rats and # indicates a difference between LiqS and HS and NS rats.

3.2. CT responses after 12 and 26 days on high-sucrose diets

In pilot studies we recorded CT responses 12 or 26 days after initiating the NS, HS, or LiqS diets based on the dynamics of sucrose intake. At day 12, sucrose consumption was significantly elevated in HS vs. LiqS rats. By day 26, sucrose intake was similar between the two groups and had plateaued in the LiqS group (Fig. 1A). When multiple concentrations of NaCl, sucrose or saccharin were tested, however, CT responses to NaCl (Fig. 2A), sucrose (Fig. 2B) or saccharin (Fig. 2C) appeared similar. Moreover, mean CT responses to single concentrations of glucose, the non-sweet carbohydrate-containing dietary supplement, polycose, umami, or bitter quinine stimuli (Fig. 2D) did not exhibit striking differences between groups. Thus, we did not perform further testing at day 12 and 26 and instead provided longer access to the diets in additional groups.

Figure 2:

Mean CT nerve ± sem responses to (A) NaCl, (B) sucrose, (C) saccharin, (D) glucose, polycose, MPG or QHCl after 12 or 26 days of sucrose solution, high-sucrose chow, or a sucrose-free diet (n=3 rats / group). Since responses appeared similar between groups in these pilot studies, we maintained diets and recorded at day 40 in separate groups.

3.3. CT responses after 40 days on high-sucrose diets

We assessed CT responses in rats maintained on sucrose-free or high-sucrose diets for 40 days. While the CT responded robustly to the standard stimulus, NH₄Cl, in all groups, differences in NaCI and sweet responses emerged (Fig. 3). Electrophysiological results from mixed-effects ANOVAs are shown in Table S2. There was a significant effect of sucrose concentration on CT responses, as expected. The effect of diet approached, but did not reach significance, although responses to 1.0 M sucrose were significantly lower in HS (P=0.036) and LiqS (P=0.006) rats vs. NS controls in post-tests (Fig. 4A). Diet significantly affected responses to the artificial sweetener, AceK, due to decreased response magnitudes in the LiqS vs. NS groups (Fig. 4B; P=0.004). Overall, high-sucrose diets decreased neural responses to upper concentrations of sucrose and Ace K by 38–68%, respectively. Sucrose in solution impacted sweet taste responsiveness more than high sucrose chow. By contrast, there were significant main effects of stimulus concentration but not diet on CT responses to saccharin (Fig. 4C, Table 2S). Responses to single concentrations of glucose and polycose were similar between dietary treatment groups (Fig. 4D, Table 3S).

Figure 3:

Representative CT response traces from rats receiving either NS, HS, or LiqS diet recorded after 40 days of dietary treatment. Response magnitudes to salt and some sweet stimuli were reduced, particularly in LiqS vs. NS rats. Responses to test stimuli are bracketed by responses to 0.5M NH₄Cl. Dots under responses denote stimulus application, and stimulus concentrations are shown (M unless otherwise indicated). Time stamp bars under the first NH₄Cl response represent 10 sec.

Figure 4:

Mean CT nerve responses to sweet stimuli after 40 days of dietary treatment. Response magnitudes were analyzed by mixed-effects model ANOVAs followed by Tukey’s tests. Responses to (A) 1.0M sucrose were significantly reduced by the LiqS or HS diets. (B) CT responses to 0.05N AceK were decreased by the LiqS diet, while there were no significant effects of diet on responses to (C) saccharin, (D) glucose, or polycose. N=5 (HS) or 10 rats (NS and LiqS)/ group.

As shown in Fig. 5, CT nerve responses to salt were also inhibited in LiqS rats that chronically consumed sucrose solution. There were significant main effects of stimulus and diet on taste responses to NaCl. LiqS rats had significantly lower responses to both 0.1M (Table 2S, P=0.002) and 0.5M NaCl (P=0.043) compared to NS controls. NaAc responses were similarly affected by concentration and diet, and CT responses to 0.5M NaAc were reduced in LiqS vs. NS rats (Fig. 5B, Table 2S, P=0.021). We recorded neural responses to NaCl mixed in the ENaC blocker, amiloride (Fig. 5C). The antagonist eliminated differences in taste responses between groups, implicating the ENaC sodium-sensing pathway in these effects. In contrast to the changes in sodium responses, high-sucrose diets did not alter umami, bitter, or acid responses (Fig. 5D, Table 3S). We also analyzed absolute NH₄Cl response magnitudes but there were no significant differences between the dietary groups (Table 3S).

Figure 5:

Mean CT nerve responses to salt and non-sweet stimuli after 40 days of dietary treatment. Response magnitudes were analyzed by mixed-effects model ANOVAs followed by Tukey’s tests. Responses to (A) NaCl and (B) 0.5M NaAc were significantly decreased by LiqS treatment. (C) Lingual application of the ENaC antagonist, amiloride, as a rinse and diluent for tastants abolished group differences. (D) CT responses to umami, bitter, and sour stimuli did not significantly differ between groups. N=5 (HS) or 10 rats (NS and LiqS)/ group.

3.4. Type II taste cell analysis

We recorded decreased CT responses to some sweet tastants, which are transduced by a subset of type II taste cells (Roper and Chaudhari 2017). The expression of some type II taste-related genes, including TRPM5 and PLCβ2, (Perez et al. 2002; Roper and Chaudhari 2017), is reduced in obese mice (Kaufman et al. 2019) and humans (Archer et al. 2019). LiqS rats, while not obese, become leptin-resistant (Harris 2018). We tested whether type II cells in fungiform taste buds were decreased by chronic LiqS in parallel with lowered sweet taste function. As shown in Fig. S1, there were similar levels of TRPM5 immunofluorescence within taste buds defined by K19 expression (NS: 55.6±2.87; LiqS: 55.4±2.27).

4. DISCUSSION

The objective of this study was to test whether chronic consumption of sucrose either as part of a dry diet or as a preferred solution had any impact on taste responsiveness in rats. We have previously shown that the LiqS paradigm results in rats consuming a very high percentage of their calories as sucrose, but due to a compensatory reduction in intake of dry diet, LiqS rats do not gain any more weight than their controls within the time frame considered here (Harris 2018). Consistent with previous studies the LiqS rats voluntarily consumed the same amount of energy as sucrose as rats offered the HS diet and this allowed us to control for changes in taste caused by sucrose consumption versus consumption of a calorie containing preferred solution (Harris 2019; 2018). Female LiqS rats develop leptin resistance within four weeks (Harris 2018), implying that this time interval is adequate for diet-induced changes in metabolism even when the rats are not obese. This information, together with the time-related changes in consumption of the sucrose solution guided our decision on which time points to investigate.

Neural responses to both palatable and non-palatable concentrations of salt and higher concentrations of sweet stimuli (Sclafani et al. 1987; Weiner and Stellar 1951), including sucrose and AceK, were reduced in LiqS rats. Yet responses to the sweet tastants saccharin, glucose and polycose were not significantly inhibited by dietary treatment, and umami, bitter, and sour responses were unaffected. The divergence in sweet stimuli modulated by diet could be partially explained by distinct molecular targets on taste buds. The carbohydrate, polycose, can be discriminated from natural sugars and activates taste pathways independently of the canonical sweet receptors on taste cells (Sclafani et al. 1987; Treesukosol et al. 2011; Zukerman et al. 2009). There is also molecular, cellular, electrophysiological, and behavioral evidence for T1 R-independent glucose transduction (Damak et al. 2003; Schier et al. 2019; Schier and Spector 2016; Yee et al. 2011). Another possibility is that dietary effects extend to glucose and saccharin but a “floor effect” due to the mediocre sweet responsiveness of the CT nerve masks neurophysiological changes. For example, average glucose response magnitudes were only 10–17% of 0.5 M NH₄Cl responses in the CT nerve. We note that the 0.05 and 0.10M concentrations of saccharin used here to stimulate sizable CT responses also activate bitter taste responses (Dess 1993; Smith and Sclafani 2002).

The LiqS diet decreased normally robust CT responses to NaCl and NaAc. This was unexpected, as prior studies using chronic high sucrose paradigms were limited to behavioral or neural responses to sweet stimuli (May et al. 2019; Smith and Wilson 1989). Reports on the acute effects of exposure to sucrose or glucose solutions have also been largely confined to sweet taste behavior. In one exception, sweet or NaCl responses recorded from the first central taste relay, the nucleus of the solitary tract (NTS), were unchanged in mice that drank 0.5M (i.e. 16.2%) sucrose solution for three days (McCaughey and Glendinning 2013). This result together with our CT recordings performed at day 4 or 26 indicate that >26 days of sucrose drinking are required to modulate neural sweet and salt responsiveness. The higher concentration of sucrose solution offered in our study might also be required to inhibit salt taste function.

Diet formulation was a factor in taste alterations since sweet and salt responses were reduced in LiqS rats while only 1M sucrose responses were significantly reduced in HS rats. Metabolism is also differentially disrupted by sucrose in solution vs. chow. We have previously reported that rats offered 30% sucrose solution become leptin-resistant within 40 days, whereas those that eat the HS diet do not (Harris 2018). This resistance develops independent of an increase in body fat mass. However, it is possible that if the period of exposure to sucrose solution was extended, then the LiqS rats would gain more fat than their controls as Ackroff et al. (2007) reported that female rats offered chronic access to a 32% sucrose solution gained significantly more weight than rats offered only water and we previously found that 54 days access to 30% sucrose solution for 54 days increased in fat content of female LiqS rats compared with HS, but not NS rats (Harris 2018). In addition to the metabolic difference between the two groups consuming high levels of sucrose, there is a difference in meal pattern. LiqS rats consume sucrose solution as small meals throughout the light and dark period, in comparison with leptin-responsive HS animals which consume a majority of their food during the dark period (Harris 2018).

We demonstrate that the proportion of TRPM5+ type II taste cells was unaffected by the LiqS diet and is unlikely to underlie the inhibition of sweet CT responses. Another potential explanation, however, is that the subset of sweet-sensing T1r2+T1r3 taste cells decreased while bitter-sensing T1r2 and/or umami-sensing T1r1+T1r3 cells increased resulting in similar proportions of TRPM5-positive taste cells in LiqS rats. CT responses to umami and bitter stimuli were not significantly affected by the LiqS diet, suggesting that sweet and salt transduction might be suppressed by alternative mechanisms. For example, diet-induced changes in endocrine factors that regulate metabolism and feeding could alter CT responses without taste receptor cell gain or loss (Behrens and Meyerhof 2019). The maintenance of normal umami, bitter, and sour responses in LiqS rats narrows potential candidates known to modulate taste function. Leptin levels are increased in rats chronically exposed to 30% (Harris 2018) or 10% (Sanchez-Tapia et al. 2019) sucrose solution, suggesting a possible mechanism for decreasing sweet responses. Leptin signaling inhibits taste receptor cell, neural, behavioral, and psychophysical responses to sweet stimuli (Kawai et al. 2000; Meredith et al. 2014; Nakamura et al. 2008; Niki et al. 2015; Ohta et al. 2003; Shigemura et al. 2004; Yoshida et al. 2015), though CT responses are unchanged (Glendinning et al. 2015) or elevated (Lu et al. 2012) by the hormone in other studies. Changes in taste responsiveness are unlikely to occur through straightforward leptin-dependent mechanisms. First, leptin levels were not increased by the HS diet in a prior study (Harris 2018) though neural taste responses to sucrose were decreased in HS rats here. Second, CT responses to NaCl in mice were similar before and after the administration of exogenous leptin (Kawai et al. 2000).

The decreased salt responses in LiqS rats may relate to elevated levels of the proinflammatory cytokine, tumor necrosis factor (TNF)-α, reported in rats chronically consuming 30% sucrose in another study (Aguilera et al. 2004). TNF-α suppresses sodium flux through ENaC in polarized taste buds (Kumarhia et al. 2016). Ghrelin, a hormone which stimulates food intake, is another potential regulator of salt responses in this model, though the taste phenotypes of ghrelin pathway knockout mice are complex. Ghrelin KO and ghrelin receptor (i.e. growth hormone secretagogue receptor) KO mice increased licking to aversive concentrations of NaCl in brief-access tests, which the authors interpret as decreased salt taste responsiveness (Cai et al. 2013; Shin et al. 2010). Exogenous ghrelin also increased saccharin consumption in preference tests (Disse et al. 2010). Mice lacking the ghrelin O-acyltransferase (GOAT) enzyme which acetylates ghrelin reduced licking to lower concentrations of NaCl, reported as potentiated salt taste, and increased ENaC-gamma subunit expression in taste buds (Cai et al. 2013). Whether ghrelin is altered by the chronic consumption of sucrose solution is unknown. Ghrelin concentrations were lower in adolescents with obesity in selfreported medium and high SSB groups (i.e. 1 and ≥2 servings/day, respectively) vs. the low SSB group consuming ≤1 serving / day though the sample size was small in the correlational study (Shearrer et al. 2016).

Studies on the neurobiology of drinking SSB have generally focused on post-ingestive effects and central reward pathways (Naneix et al. 2016; Veldhuizen et al. 2017), bypassing taste buds (Ren et al. 2010). Recently, however, it was reported that Drosophila fed 30% sucrose for 10 days exhibited reduced behavioral responses and chemosensory synaptic activity to sugar (May et al. 2019). In the current study, we demonstrate that a high-sucrose dry diet decreased neurophysiological responses to 1M sucrose, while chronic sucrose drinking decreased neurophysiological responses to the highest concentration of sweet and 0.1M and 0.5M salt in the peripheral taste system, indicating the disruption of gustatory input to the CNS. CT nerve sectioning studies demonstrate the importance of this nerve’s activity for the ingestion of and preference for sodium (Barry et al. 1996; 1993; Breslin et al. 1993; 1995; Frankmann et al. 1996; Shires et al. 2011; St John and Spector 1998) and sweet solutions (Spector et al. 1996; Vigorito et al. 1987). Reduced CT responsiveness may therefore predict blunted taste-elicited behavioral responses to salt and sweet, though the relationship between peripheral taste activity and behavior in the context of high-energy diets is complex (Maliphol et al. 2013; Ninomiya et al. 1995; Shimizu et al. 2003; Treesukosol et al. 2018). Nevertheless, chronic SSB consumption may reduce neural output from taste buds and associated nerve fibers and alter salt and sugar intake.

Supplementary Material

Figure S1

Figure S1: Type II taste receptor cells in fungiform papillae of rats fed sucrose-free chow or sucrose solution. TRPM5 positive pixels (green) were used as a measure of type II taste receptor cells within K19-positive taste buds. Nuclei were stained with DAPI. (A) NS or (B) LS rats treated for 40 days had similar levels of TRPM5 expression when analyzed by Student’s t tests. Arrows indicate TRPM5 positive type II receptor cells with nuclei (blue). Arrowheads show TRPM5 positive and TRPM5+K19 immunofluorescence (yellow). Bar in (B) = 20 μm.

Abbreviations

NS

sucrose-free chow

HS

high-sucrose chow

LiqS

30% sucrose solution plus sucrose-free chow

aceK

acesulfame K

CT

chorda tympani nerve

HFD

high-fat diet

ENaC

epithelial sodium channel

NaAc

sodium acetate

MPG

monopotassium glutamate

QHCl

quinine hydrochloride

PLCβ2

phospholipase Cβ2

TRPM5

transient receptor potential channel subfamily M member 5

K19

keratin 19

TNF-α

tumor necrosis factor-α