Abstract Chemotherapy is one of the most common treatments for cancer; however, a side effect is often altered taste. This study examined how cyclophosphamide, a chemotherapy drug, affects salt taste in mice. On the basis of previous findings, it was predicted that cyclophosphamide-induced disruptions in salt taste would be observed near days 2–4, 8–12, and 22–24 posttreatment, and that multiple, smaller doses would cause more severe disruptions to taste. To test these predictions, two experiments were performed, one using brief access testing to measure appetitive qualities, and another using operant conditioning to measure detection thresholds. After a single 100 mg/kg cyclophosphamide injection, peak alterations in brief access lick rates were seen near days 5–8 and 15 posttreatment, whereas peak alterations in detection thresholds were seen days 6, 14, and 20 posttreatment. After five 20 mg/kg injections of cyclophosphamide, brief access lick rates revealed disruptions only on postinjection day 8 whereas thresholds appeared to cycle, gradually increased to and decreased from peak elevations on posttreatment days 4, 10, 15, 20, and 23. Although salt taste functions were disrupted by cyclophosphamide, the patterns of these disruptions were less severe and shorter than expected from cell morphology studies, suggesting a functional adjustment to maintain behavioral accuracy. Fractionation of cyclophosphamide dosing had minimum effect on brief access responses but caused longer, cyclic-like disruptions of detection thresholds compared to single-dose administration. animal psychophysics, brief access, chemotherapy, mice, salt taste, taste cell renewal Introduction Chemotherapy is one of the most common treatments for cancer patients. Depending upon the treatment regime, as many as 80% of these patients may suffer a disturbance in taste and/or smell functioning (Bernhardson et al. 2008; Boltong and Keast 2012). Changes in taste are generally unpleasant and disruptive to food selection and intake of patients, negatively impacting the quality of their lives and potentially their prognosis (Bernstein 1978; Wickham et al. 1999; Lindemann 2001; Davidson et al. 2012; Aaldriks et al. 2013). In a quantitative study of taste and smell functioning during chemotherapy, patients reported changes in one or more of the basic tastes assessed (sweet, salty, and bitter), but the one most affected was salty taste (Steinbach et al. 2009). One of the most commonly used chemotherapy drugs is cyclophosphamide (CYP), a nitrogen mustard that is metabolized into acrolein and phosphoramide mustard in the liver by the cytochrome P450 enzyme complex (Anderson et al. 1995). Both metabolites are cytotoxic, but phosphoramide mustard is an alkylating agent, which attacks open DNA strands. CYP is used for chemotherapy and immunosuppression because it targets replicating cells that are most often cancer cells, but it will also target other fast replicating cells, such as hair follicles (Ahmed and Hombal 1984). As with many chemotherapeutic drugs, it is often given as a smaller dose, repeated over time (referred to as dose fractionation) to reduce the side effects of the drug and to increase the duration of the toxic effects of the drug on cancerous cells (Ozols and Young 1984; Levin and Hryniuk 1987; Lamar et al. 2016). Previous work with CYP suggests that reported changes in taste functions are due to tissue damage to taste epithelium (Mukherjee and Delay 2011; Mukherjee et al. 2013; Mukherjee et al. 2017) as well as through conditioned taste aversions (Bernstein 1978). A key structure in taste epithelium is the taste bud, comprising 3 main types of taste sensory cells (TSCs): Type I, Type II, and Type III (Chaudhari and Roper 2010). Sweet, bitter, and umami substances are detected by Type II TSCs, and sour substances are detected by Type III TSCs (Nelson et al. 2001; Ishimaru et al. 2006; Chaudhari and Roper 2010). Salt taste is thought to be detected by all 3 cell types and 2 basic categories of receptors: amiloride-sensitive and amiloride-insensitive receptors (Heck et al. 1984; Vandenbeuch et al. 2008; Yoshida et al. 2009; Roper 2015; Lewandowski et al. 2016). Type I TSCs have been implicated in amiloride-sensitive salt taste transduction (Vandenbeuch et al. 2008). Cellular and behavioral investigations have shown that Type II TSCs respond to salt stimuli and may contribute to amiloride-insensitive salt detection (Tordoff et al. 2014; Price et al. 2016; Roebber et al. 2017). Type III TSCs appear to respond as amiloride-insensitive and osmotically sensitive salt taste in two subpopulations, distinguished by their sensitivity to the anion of the salt (Lewandowski et al. 2016; Zocchi et al. 2017).Although specific receptors and TSCs underlying salt taste have not been fully characterized (Roper 2015), it is clear that multiple TSC types and receptor proteins are involved in generating the sensation that results in the perception of salt taste. Each type of TSC has relatively short, albeit slightly different, lifespans. Perea-Martinez et al. (2013) used birth date labeling to determine that the half-life is 8 days for Type II TSCs and 22 days for Type III TSCs. Type I TSCs were believed to have 2 subpopulations with different half-lives (8 and 24 days); however, Type I and undifferentiated cells could not be distinguished reliably. Because of these relatively short lifespans, taste buds require a continuous stream of replacement cells to maintain the effectiveness of the system. A population of basal cells located in a layer just ventral to the taste bud serves as progenitors that constantly generate replacement cells that migrate to taste buds and eventually differentiate into functional TSCs (Miyamoto et al. 1996). These same progenitor cells are highly susceptible to the effects of CYP (Mukherjee et al. 2013, 2017). This study sought to examine the effects of CYP on salt taste in a mouse model when CYP is administered either as a single dose or as a fractionated dose. To understand how CYP might affect salt taste, it is helpful to understand the model developed from behavioral, morphological, and immunohistochemical data reported previously by Mukherjee and Delay (2011) and Mukherjee et al. (2013). These studies have reported a 2-phase increase in umami and sweet taste detection thresholds (both involving Type II TSCs) that extended the effects of a single dose of CYP for up to 15 days (Mukherjee and Delay 2011; Mukherjee et al. 2013). In the 1st phase, elevated thresholds were seen in the initial 3–4 days after injection, probably due to the drug’s impact on fungiform papillae. Fungiform papillae are reduced by nearly half within 3–4 days and many of the remaining taste buds appear disorganized (Mukherjee and Delay 2011). Even though fungiform papillae did not return to control levels until 12–14 days after injection, behavioral thresholds for umami and sucrose substances appeared to return to normal 5–7 days after administration, then began to increase again. In this 2nd phase, the elevations in thresholds were more severe than in the 1st phase and lasted up to day 15 after injection for sucrose and nearly the same for umami. The increase in thresholds in the 2nd phase coincided closely in time with the onset of a rapid decrease in the cell population and number of circumvallate taste buds, which further reduced the functional population of TSCs. By 12–14 days after injection, both fungiform and circumvallate taste buds appeared to be recovering and behavioral thresholds were returning to normal. This 2nd phase of threshold shifts appears to be related to the effects of interrupting the TSC-renewal cycle by CYP (Mukherjee et al. 2013, 2017). Normal cell cycling of the progenitor cells in the basal layer of the tongue epithelium dropped to a low level within 24 h and did not resume until 4–6 days after injection. This leaves a window of time in which a minimum of new replacement cells are born and creates a gap in ongoing maturation and differentiation of cells destined to replace aging TSCs when they die (e.g., Type II cells responsible for detection of sweet, bitter, and umami). Once cell cycling resumes and the new wave of replacement cells mature and differentiate, taste bud populations are restored and normal taste functions can resume. It is notable that the appearance and magnitude of each shift in thresholds in both studies corresponded to a drop in the cellular populations of taste buds (Mukherjee and Delay 2011; Mukherjee et al. 2013). The 1st followed the loss of fungiform papillae, but the mice appeared to adjust to the reduced signal within a few sessions, presumably utilizing sensory input from surviving TSCs located mostly in the posterior portion of the tongue and possibly the soft palate. However, when the aging TSCs in the circumvallate also began to die without replacement, the population of TSCs was reduced even more, resulting in a larger and longer elevation in taste thresholds. Sensitivity and, thus, behavioral measures of detection thresholds were restored only after, it appeared, a wave of new TSCs (Type II cells) were in place and functional. It was these findings, coupled with the half-life of Type II cells (Perea-Martinez et al. 2013), that guided our expectations regarding the effects of a single-dose CYP on salt thresholds and brief access lick rates in this study. However, we also anticipated some variation in the specific days for these effects. The model assumes turnover of each cell type is equivalent over days, which is probably unrealistic. If not equivalent, then it is also likely that cell renewal cycles of individual mice are out of synchrony with each other, especially if multiple cell types are involved in these taste functions. In addition, we do not know how much of a TSC population can be reduced before deficits in taste functions are detectable. Disruptions in salt taste were expected to be most obvious when a cell type was approaching its population half-life without replacement. As all 3 types of TSCs, each with a different life span, appear to play a role in salt detection (Vandenbeuch et al. 2008; Tordoff et al. 2014; Lewandowski et al. 2016; Roebber et al. 2017), it was predicted that a single injection of CYP would disrupt salt taste within 3 approximate time frames, 2 of which correspond to the life spans of each cell type. The 1st time frame would be within 2–4 days after injection due to initial cytotoxicity. The 2nd would be about 8–12 days after injection (half-life of Type II and some Type I TSCs), and the 3rd would be about 22–24 days after injection (half-life of Type III and some Type I TSCs). Although these time spans were predicted from known half-life data, we also were open to variations in the temporal appearance of taste deficits. It is not clear how much of a TSC population can be reduced before changes in different NaCl taste functions can be detected. Moreover, any disruption in salt taste behavior might be of relatively short duration because a portion of the total salt sensation would still be present and thus could be used by mice to compensate in the behavioral task. Previous studies have used a single injection of CYP to pinpoint the time frame for the subsequent effects of the drug. However, the clinical use of chemotherapy drugs often involves dose-fractionation in which the total drug dose is divided into portions, each administered at a different time. The expectation is that spreading the dosing over time can increase the duration of the drug’s effects on rapidly dividing cells, including those in the taste system (Ozols and Young 1984; Lamar et al. 2016; Socia et al. 2017). Because of its clinical relevance, we also thought it important to determine if fractionation of CYP dosing would have a similar effect on salt taste. It was predicted that smaller doses of CYP (fractionated), repeated over days, might have a weaker but more prolonged disturbance in taste function. In general, the results of these experiments, especially detection threshold experiments, supported these hypotheses. General methods Subjects For these experiments, 6- to 8-week-old male C57BL/6J mice (total N= 45; n=31 for experiment 1, n= 14 for experiment 2) were obtained from Jackson Laboratory (Stock No: 000664; https://www.jax.org/strain/000664). The mice were kept on 22.5-h water deprivation, housed in groups of 2–4, and given ad libitum access to food. The colony room was on a 12-h light:dark cycle with the lights turned on at 7 AM. All procedures were approved by the University of Vermont IUCAC Board (Protocol 14-003). Before beginning the experiment, the mice were given a week to acclimate to the colony room and water deprivation. They were also handled during this time to ensure they were habituated to their handlers. Chemical reagents Cyclophosphamide monohydrate (97%, CYP) was obtained from Acros Organics. Sodium chloride (99%) was obtained from Fisher Chemicals. Sucrose was obtained from Domino Foods. All taste solutions were prepared fresh daily with Millipore-filtered water (Millipore, Burlington, MA, USA). CYP solutions were prepared with bacteriostatic water (NDC 63323-249-30, APP Pharmaceuticals). Experiment 1: brief access testing To begin investigating the effects of CYP on salt taste preference, brief access methodology was used. Typically, water-deprived C57BK/6J mice will ingest NaCl at low concentrations but begin to show avoidance of NaCl above 100 mM (Duncan 1962). It was hypothesized that the damage to the taste system caused by CYP would lead to alterations in ingestion of NaCl solutions relative to water, which should be more apparent at higher concentrations. That is, CYP should reduce the aversive qualities of NaCl. Methods Apparatus. To measure the appetitive qualities of NaCl, licks of various solutions were counted using 4 computer-operated MS 160 Davis Rig gustometers (DiLog Instruments). Each rig consisted of a chamber (30 cm long × 15 cm wide × 23 cm high) and a computer-controlled mobile tray in which tubes containing test solutions were mounted. The tray was located at one end of the chamber. An oval-shaped opening was in the chamber wall next to the tray. A shutter covering the opening could be lowered by a computer to give the mouse access to the lick spout of one of the tubes. The diameter of the opening at the tip of the spouts was 2.5 mm. The computer moved the tray to position the assigned tube with its lick spout behind the shutter. Once the shutter was opened, the computer recorded a lick each time the mouse contacted the lick spout. Procedure. Prior to the beginning of the experiment the mice were habituated to the Davis Rigs and trained to lick from the spouts for 5–7 days. During training, the stimulus tubes were filled with water. To initiate a trial, a stimulus tube was positioned behind the opening and the shutter was opened to give the mouse access to the lick spout. Timing for the trial began when the mouse contacted the lick spout and ended when the shutter was closed over the opening. The period between the 1st lick of a spout and the closure of the shutter was gradually reduced to 7.5 s, the same duration of the trials during test sessions. During a 10-s inter-presentation interval, the next solution was moved into place behind the shutter. If the mouse did not lick, the shutter remained open until a lick was recorded or the time limit was reached for the 20-min session. This training continued until the performance of the mice stabilized and each mouse consistently emitted more than 10 licks in 3 out of every 4 trials during each session. After this criterion was met, the next stage of training was initiated by presenting the mice with 50, 100, 175, and 300 mM NaCl solutions along with 2 water tubes and a 100-mM sucrose tube using the same procedures as would be used posttreatment. NaCl concentrations were selected for 2 reasons. First, if CYP alters NaCl taste as severely as patients had reported or as in earlier work with umami taste, CYP would likely have an effect on lick rates of one or more of these concentrations and, second, the higher concentrations of this range were known to be mildly aversive (Eddy et al. 2009; Steinbach et al. 2009). CYP can cause nausea that can induce conditioned taste aversions, confounding the results and altering interpretation of the findings (Bernstein 1978; Wayner et al. 1978; Logue 1979). Pretreatment exposure to NaCl was expected to reduce the possibility that mice might develop a conditioned taste aversion to NaCl after an injection of CYP. Nevertheless, we included the sucrose solution to assess possible CYP-induced conditioned aversion tendencies. A randomized block procedure using a series of Latin squares was used to establish the order of stimulus presentations. Each mouse was assigned a different stimulus order each day and this order was different from those assigned to the other mice that day. Mice were trained until they were licking consistently for most trials (mean = 5 training sessions). After the training was complete, all mice were rehydrated for 24 h and then cages were randomly assigned to 1 of 3 treatment groups until group size was met for the 5 consecutive days of injections. One treatment group (n = 10) received saline injections (1 mL/kg body weight) the 1st 4 days and 1 100 mg/kg CYP injection on the 5th day (hereafter referred to as the 1*100 CYP condition). A 2nd treatment group (n = 11) received an injection of 20 mg/kg CYP on each of the 5 days (hereafter referred to as the 5*20 CYP condition). The saline control group (n = 10) received a 1 mL/kg saline injection in each of the 5 days. All injections were given intraperitoneally. Cages were changed daily. Dosages were chosen to be equivalent to a moderate level for human cancer treatment while not reaching levels known to cause inflammation of the bladder or nephrotoxicity, which might alter motivational state beyond what was induced by the deprivation schedule (Ozols and Young 1984; Vizzard 2000; Weiner and Cohen 2002). Following the last injection, the mice were given a 24-h recovery period, then returned to the 22.5-h water deprivation schedule prior to the initiation of posttreatment testing. Daily test sessions were conducted until 26 days after the last injection. The mice were presented with the same concentration range of NaCl and 100 mM sucrose in the Davis Rigs daily at 12 PM. One hour after the test session, the mice were given additional water in the home cage for 5 min to rehydrate them from the intake of salt to prevent gastric malaise. Statistical methods. Lick rates were normalized to water lick rates by dividing the mean lick count for each concentration of NaCl by the mean lick count for water trials during that session. This corrects for variable motivational states and differences in inherent lick rates between mice. The linear mixed-model analysis of variance (ANOVA) used to analyze lick rates had the following factors: drug treatment (3 levels: saline, 1*100 CYP, and 5*20CYP) treated as a between-subject variable, concentration of NaCl (4 levels: 50, 100, 175, and 300 mM NaCl), and days posttreatment (24 levels: days 2–25) treated as within-subject variables. Covariance was treated as first-order autoregressive or AR(1), assuming no sphericity of the data. To identify significant group differences, the data were then partitioned to further evaluate each group using ANOVA simple effects tests, and post hoc t-test with Sidak α-correction procedures (Howell 2016). All statistical analyses were performed with IBM SPSS Statistics 24 Software. Graphs were created with GraphPad Prism 7 (GraphPad Software Inc.). Results The 3-way ANOVA revealed significant main effects for drug treatment (F(2,690) = 14.68, P < 0.0005), day posttreatment (F(23, 1148) = 4.38, P < 0.0005), and NaCl concentration (F(3, 2545) = 7.11, P < 0.0005). There were significant interactions of drug treatment by concentration (F(6, 2545) = 5.39, P < 0.0005) and day by concentration (F(69, 2730) = 3.14, P < 0.0005). All groups licked sucrose at a higher rate than water (a ratio consistently above 1.0), indicating there was no evidence of a conditioned taste aversion. The mean normalized sucrose lick rate ± SEM for each group was: 1) saline mice = 1.32 ± 0.03, 2) 1*100 CYP mice = 1.47 ± 0.03, and 3) 5*20 CYP mice = 1.44 ± 0.03. Pretreatment, all 3 groups showed a comparable concentration-dependent aversion to NaCl, but not posttreatment (Figure 1A, B). Saline-treated mice generally avoided NaCl, with a lick ratio ranging from 0.5 to 0.8 licks of NaCl per lick of water during the pretreatment baseline period and throughout the posttreatment period. To determine if there were shifts in baseline between pre/posttreatment conditions, a mixed-model ANOVA was performed on the lick rates of the saline group, comparing pre/posttreatment lick rates for the 4 concentrations collapsed across days. Lick rates were significantly lower pretreatment than posttreatment for the saline mice (F(2, 48) = 26.88, P < 0.0005). In addition, the main effect for concentration was significant (F(3, 74) = 6.00, P = 0.001). Sidak α-corrected post hoc testing found significantly higher lick rates for 50 mM compared to 175 and 300 mM NaCl (P values < 0.004 and 0.001, respectively). The interaction between concentration and pre/post lick rates did not reach significance (F(3, 74) = 2.53, P = 0.064). Figure 1. View largeDownload slide Drug treatment by concentration of NaCl on lick rates pooled over days pre- and posttreatment. In both the pretreatment (A) and the posttreatment (B) graphs, the x axis is the concentration of NaCl (in mM) and the y axis is the mean (±SEM) normalized lick rate of that concentration (see text) from mice given 5 days of saline injections (saline, filled box), mice given 4 injections of saline and a single 100 mg/kg CYP injection (1*100 CYP, open box), and mice given 5 injections of 20 mg/kg (5*20 CYP, filled circle). These normalized lick rates were then averaged across all days of testing. All groups demonstrated an aversion to all concentrations of NaCl (no aversion = 1.0).In pretreatment, there was a significant decrease in lick rates of 175 and 300 mM NaCl compared to 50 mM NaCl for all groups (*: P < 0.05). There were no significant differences between groups. Posttreatment lick rates for the 2 CYP groups were significantly higher than saline mice. ###, 1*100 CYP mice significantly greater than saline mice (P < 0.001); ++, Saline group significantly less than 5*20 CYP (P < 0.01); and ***,1*100 CYP (P < 0.001) groups. Figure 1. View largeDownload slide Drug treatment by concentration of NaCl on lick rates pooled over days pre- and posttreatment. In both the pretreatment (A) and the posttreatment (B) graphs, the x axis is the concentration of NaCl (in mM) and the y axis is the mean (±SEM) normalized lick rate of that concentration (see text) from mice given 5 days of saline injections (saline, filled box), mice given 4 injections of saline and a single 100 mg/kg CYP injection (1*100 CYP, open box), and mice given 5 injections of 20 mg/kg (5*20 CYP, filled circle). These normalized lick rates were then averaged across all days of testing. All groups demonstrated an aversion to all concentrations of NaCl (no aversion = 1.0).In pretreatment, there was a significant decrease in lick rates of 175 and 300 mM NaCl compared to 50 mM NaCl for all groups (*: P < 0.05). There were no significant differences between groups. Posttreatment lick rates for the 2 CYP groups were significantly higher than saline mice. ###, 1*100 CYP mice significantly greater than saline mice (P < 0.001); ++, Saline group significantly less than 5*20 CYP (P < 0.01); and ***,1*100 CYP (P < 0.001) groups. After CYP treatment, the 1*100 CYP mice had periods of increased lick rates compared to saline mice on postinjection days 2, 5–8, and day 15 (Figure 2A). The 5*20 CYP-treated mice generally followed saline-treated mice, except an increase in lick rates on day 8 (Figure 2B). Simple-effects tests were performed by partitioning the data by days and concentration since this interaction was significant, then performing a 1-way ANOVA with treatment condition as the grouping variable. After this analysis, alpha-corrected Sidak posthoc tests were performed. The following days’ ANOVAs found significantly altered lick rates (reported as day (concentration)): 2 (175 and 300 mM), 5 (50, 175, and 300 mM), 6 (50 mM), 8 (50, 100, 175, and 300 mM), and 20 (50 and 175 mM), F(2, 3041) ≥ 3.09, all P values < 0.045. Posthoc testing found significant increases (P < 0.05) in lick rates for the 1*100 CYP group compared to the saline-treated mice on all days but day 20. On this day, saline-treated mice had significantly higher lick rates of 50 mM NaCl than 5*20 CYP-treated mice (P = 0.007). On day 20, 1*100 CYP mice had significantly increased lick rates of 175 mM NaCl compared to 5*20 CYP-treated mice (P = 0.023). Thus, in general, only the 1*100 CYP group showed elevated intake compared to saline group, but the CYP mice did not show a consistent concentration-dependent effect on lick rate over days. Therefore, to simplify the analysis, the concentration factor was collapsed by averaging the normalized lick rates for all 4 concentrations during each session to generate a single lick rate score for each animal for each session. These lick rates were then analyzed with a 2-way ANOVA with drug treatment and day posttreatment (identical to those variables detailed in the methods) as the only factors. This ANOVA found significant main effects for treatment (F(2, 717) = 14.62, P < 0.0005), day posttreatment (F(23, 1570) = 5.81, P < 0.0005), and the interaction of these factors (F(2, 1580) = 1.40, P = 0.042). Figure 2. View largeDownload slide Average normalized lick rates of NaCl solutions over days during brief access testing. The x axis represents the days past the last day of treatment. The y axis represents the mean (±SEM) of normalized lick rates (see text) for each day. (A) The 1*100 mg/kg CYP group (open circles) saw significant increases in NaCl lick rates compared to controls on days 5 and 8 posttreatment compared to saline mice (filled circles). (B) The 5*20 mg/kg CYP group (open circles) saw significantly increased lick rates compared to saline controls (filled circles) on day 8. The data for the saline mice are presented in both panels to make visual comparisons easier. * indicates a comparison between the CYP group and saline-treated animals. * = P < 0.05, ** = P < 0.01, *** = P < 0.001. Figure 2. View largeDownload slide Average normalized lick rates of NaCl solutions over days during brief access testing. The x axis represents the days past the last day of treatment. The y axis represents the mean (±SEM) of normalized lick rates (see text) for each day. (A) The 1*100 mg/kg CYP group (open circles) saw significant increases in NaCl lick rates compared to controls on days 5 and 8 posttreatment compared to saline mice (filled circles). (B) The 5*20 mg/kg CYP group (open circles) saw significantly increased lick rates compared to saline controls (filled circles) on day 8. The data for the saline mice are presented in both panels to make visual comparisons easier. * indicates a comparison between the CYP group and saline-treated animals. * = P < 0.05, ** = P < 0.01, *** = P < 0.001. Because the interaction of the 2 variables was significant, simple-effects tests with Sidak α-correction were completed to determine when lick rates were significantly different between treatment conditions. These simple-effects tests were performed by partitioning the data by days and performing a 1-way ANOVA with treatment condition as the grouping variable. Significant group differences in lick rates were found on days 2, 5, 6, 8, 15, and 20 (F(2, 1533) ≥ 3.91, all P values < 0.02) (Figure 2). To determine how the treatment conditions compared, Sidak α-corrected posthoc tests were performed on each of the days identified by the simple-effects tests. When comparing 1*100 CYP-treated animals to saline animals, higher lick rates for NaCl were observed for the CYP group on postinjection days 2, 5, 6, 8, and 15 (P ≤ 0.045, Figure 2A). The 1*100 CYP mice also licked NaCl at significantly higher rates compared to 5*20 CYP mice on days 5, 6, 15, and 20 (P ≤ 0.050). The 5*20 CYP-treated mice had higher lick rates of NaCl compared to saline mice on postinjection day 8 (P = 0.038, Figure 2B). There were no days when 5*20 CYP mice had significantly higher lick rates than 1*100 CYP mice. Discussion Typically, rodents show concentration-dependent aversion of NaCl above 100 mM (Duncan 1962; Glendinning et al. 2002; Ruiz et al. 2006; Eddy et al. 2009). Although no group exhibited a strong concentration-dependent aversion gradient, all 3 groups found all concentrations of NaCl at least mildly aversive throughout the experiment. Saline mice showed a mild concentration-dependent decrease in lick rates for NaCl before saline injection, comparable to those reported previously for brief-access testingfor these concentrations (Glendinning et al. 2002; Eddy et al. 2009). A similar concentration-dependent decline in lick rates was observed after saline injection, although their lick rates were generally higher after treatment. On the other hand, even though CYP-injected mice continued to lick NaCl at lower rates than water after drug treatment, their lick rates were higher than that of the saline group. This suggests that CYP effectively reduced the aversive properties of NaCl, especially for the 1*100 CYP mice. However, there are other potential explanations. A reduction in the aversive properties of NaCl has been reported when mice have prior experience with the taste of 75 mM NaCl or higher (Beauchamp and Fisher 1993), such as the general increase observed in posttreatment lick rates of the saline mice. Mice of the strain C57BL used in this study typically demonstrate little or no aversion or even a mild preference for NaCl solutions when the concentration of test solutions is at or below the lowest concentration used in this experiment (Beauchamp and Fisher 1993; Bachmanov et al. 1996; Eddy et al. 2009). The leveling of the normal aversion curve for the CYP mice may also have been caused, at least in part, by animals not being salt deprived in their diet (Bertino et al. 1982; Berridge et al. 1984; Zocchi et al. 2017). Physiological need has been demonstrated to influence taste (Scott 1990), and salt deprivation would make a saline solution more acceptable at higher concentrations as the animal needs salt. However, unless CYP had some physiological effect causing salt deprivation, this possibility seems unlikely. Regardless of whether either of these possibilities had an impact on posttreatment lick rates, CYP significantly elevated lick rates relative to those for saline control mice. When concentration was collapsed for each session, the analyses revealed disruptions to lick rates of CYP-treated animals, primarily the 1*100 CYP animals. These animals demonstrated higher lick rates for NaCl immediately after treatment, as seen on day 2, and later on days 5, 6, and 8 posttreatment. The CYP effects in the later cluster of days (days 5–8) preceded the predicted time span of 8–12 days and were shorter in duration and less severe than anticipated from previous behavioral work (Mukherjee and Delay 2011; Mukherjee et al. 2013). In addition, CYP treatment appeared to have little effect on lick rates at the furthest predicted time span (days 22–24). When trying to understand what TSC cell types were essential for this task, it is important to consider the latter 2 time spans when behavioral deficits were expected. The onset and duration of these temporal spans were estimates of when the gap in replacement cells, a gap resulting from the CYP-induced loss of progenitor cells, fails to adequately replace mature cells at the end of their lifespan (Mukherjee et al. 2017). The results of this experiment suggest a subpopulation of Type I and/or Type II TSCs (but not Type III cells) might play a significant role in transducing the appetitive aspects of salt taste in this behavioral task. The 5*20 CYP mice only demonstrated increased lick rates of NaCl when compared to saline animals on one day, suggesting that fractionating the dose of CYP did not disrupt salt taste as much as a single, full dose. Five smaller doses spread over days may not have caused cellular populations to drop significantly below a critical point that would disrupt identification of the appetitive qualities of NaCl. This might be because the CYP mice could detect the suprathreshold concentrations used in this experiment and that these mice could assess taste qualities from licks emitted over the entire 7.5-s trial period (Ruiz et al. 2006). Alternatively, there may have been fewer direct cytotoxic effects on the taste system or a smaller interruption of the cell renewal process induced by fractionated doses than induced by a single, full dose of CYP. A general limitation of this experiment was that it could not determine if the disrupted behavior was caused by impaired NaCl sensitivity or altered qualities of the taste signal. Experiment 2 measured detection thresholds to determine if at least some of these disruptions in behavior were caused by impaired NaCl sensitivity. Experiment 2: detection thresholds Previous studies found that CYP elevated sucrose and umami detection thresholds (Mukherjee and Delay 2011; Mukherjee et al. 2013). In those studies, the 2-phase loss of sensitivity for these substances directly corresponded to the loss of Type II TSCs first in fungiform papillae, followed by the loss of these cells in circumvallate taste buds. Experiment 2 was performed to determine if CYP affected detection threshold sensitivity for NaCl and if this loss might follow a temporal pattern associated with the half-life of one or more of the TSC populations. An operant conditioning task was used which required the mice to focus on identifying the stimulus relative to water. This discrete trial task tested discrimination ability and taste sensitivity, rather than appetitive qualities, for NaCl to determine detection thresholds. Methods Apparatus. To evaluate detection thresholds, 5 identical Knosys Ltd (Knosys Inc.; Brosvic and Slotnick 1986) computer-controlled lickometers were used, following established procedure (Delay et al. 2006; Ruiz et al. 2006; Mukherjee and Delay 2011; Mukherjee et al. 2013; Jewkes et al. 2017). The lickometers consisted of a Plexiglass operant chamber 17 cm high, 12 cm long, and 12 cm wide, in which the mice were placed. A fan was mounted in the ceiling for positive-pressure airflow within the operant chamber. A 1-cm circular opening was centered 2.5 cm above the floor, through which a stainless-steel lick spout was accessible. This lick spout had 9 smaller, stainless-steel capillary tubes within it. Each capillary tube was connected via Flexible C-flex tubing (ID: 0.031 in; #06424-60; Cole-Parmer) to a single 3-mL syringe barrow in which one of the test solutions was stored. The tips of these capillary tubes were recessed 2 mm from the tip of the lick spout to prevent the mouse from gaining spatial cues during delivery of the test solution. The syringe barrels containing taste solutions were mounted on a separate Plexiglas rack, 7.5 cm above the lick spout, and facing away from the operant chamber to minimize visual cues. The 9th capillary tube was connected to a 5-mL syringe barrel containing water for reinforcement. Olfactory cues were minimized by the fan blowing air through the chamber and out of the lick spout hole, as well as the brief access periods, the narrow tubing, small sample sizes, and the recessed delivery tubes within the lick spout. Pinch valves (P/N 075P2-S1013; Bio-Chem Fluidics Inc.) kept the tubing from dispensing water or stimulus solutions until opened by the computer controlling the lickometer. Although these valves are designed to operate quietly, they were close enough to the chamber that the mouse might be able to hear their operation. Therefore, an independent, audible solenoid mounted directly above the lick spout was opened and closed simultaneously to mask possible auditory spatial cues from their operation. Upon licking the tube, the mouse completed a circuit with a stainless-steel grate on the floor of the cage, allowing for 1 lick to be counted when a 60 µA current passed through the circuit. The stainless-steel grate was placed in such a way that the mouse could only lick while standing on it. A Piezo buzzer (Jameco Electronics), mounted in the ceiling above the lick spout, produced a continuous 2.9 kHz tone inside the test chamber at 80–90 dB when activated. Procedure. After acclimatizing to water deprivation and habituation period detailed in experiment 1, the mice were first trained to lick from the lick spout, then trained to the final discrete trial sequence over the next 14 days. To initiate a trial, the mouse first had to complete a variable ratio 18 lick response requirement (range 3–33), which initiated a brief 5.5-µL water rinse. This was followed by another variable ratio 18 lick response requirement. When this 2nd ratio schedule was met, a trial began with the delivery of a 7.5-µL aliquot of either water (S+) or NaCl (S−). Once the stimulus was delivered, the mouse had 2 s to identify the stimulus and had to alter its response accordingly by the last 0.4 s of the 2-s period. There were 4 stimulus–response outcomes possible. If the mouse did not lick during the last 0.4 s of the 2-s period, it identified the stimulus as an S− (NaCl). If the mouse continued to lick during the last 0.4 s of the 2-s period, it identified the stimulus as an S+ (water).Incorrect identification of a stimulus resulted in a time out and tone punishment. Correct identification of the S+ yielded a water reward and correct identification of the S- prevented the occurrence of the time-out-tone punisher. The training strategy was to begin with easily detected concentration of NaCl and gradually introduce lower concentrations of NaCl until the mouse was responding consistently to a combination of supra- and subthreshold concentrations. Training started with water (S+) solutions on all trials. After 2–3 days, the 1st S− solution (300 mM NaCl) was introduced. A session lasted 40 min with a minimum of 70 trials, and mice were rehydrated for 10 min beginning 1 h after each session to counteract any dehydration from salt intake. Once the mouse could correctly identify and respond to 300 mM NaCl at least 75% of the trials, a 175 mM NaCl solution was added to the stimulus array. Once this concentration was detected accurately in 75% of trials, 100 mM NaCl was added to the array. This was repeated, usually every 2–3 days, with 50 and 25 mM until the mouse was able to detect 25 mM accurately. When 50 mM NaCl was added to the array, 300 mM was removed from the training solutions. Once the mice were accurately detecting 25 mM NaCl, 2.5 mM was introduced for 1–2 sessions; then the final array of concentrations was introduced: 0.1, 2.5, 25, 50, 100, and 175 mM NaCl. These concentrations were selected based on reported NaCl thresholds in mice using a similar methodology and included at least 1 concentration well below the anticipated detection thresholds (defined as the lowest concentration of NaCl detected in 50% of the presentations) of C57BL control mice (Ruiz et al. 2006). Once the mice were being presented with the final concentrations, the mice were given a session daily until their threshold estimates were stable (defined as within 5 mM of the previous day) for 3 consecutive days. Throughout training and testing approximately half of the trials had an S+ presentation to ensure no advantage could be gained by choosing only 1 response. Latin squares were used to randomize the sequence of S− presentations and to ensure that each concentration of NaCl was presented in approximately the same proportion as all other concentrations. Different sequences were assigned to a mouse each day and different sequences were assigned to each mouse within a session. After all pretreatment thresholds were determined, the mice were taken off water deprivation and allowed to rehydrate for 24 h before beginning their assigned drug treatment. The 3 groups of mice received 1 of 3 treatments (the same as in experiment 1): 1) 5 injections of saline (control, n = 5), 2) 4 injections of saline followed by 1 injection of 100 mg/kg CYP (1*100 CYP, n = 4), and 3) 5 injections of 20 mg/kg CYP (5*20 CYP, n = 5). Intraperitoneal injections were given once per day, and cages were changed each day to remove any toxic CYP metabolites in urine (Ahmed and Hombal 1984). After the final injection, mice were given a day to recover and then placed on water deprivation for 24 h before resuming testing with the full array of stimuli. This testing spanned 25 days. During threshold testing, results of mice were monitored daily to ensure that stimulus control was maintained. The responses to 175 mM NaCl and the S− stimuli were used as indicators of stimulus control. Detection of lower concentrations of NaCl could be impaired in CYP-treated mice so these were not used as part of the criteria. Loss of stimulus control can be seen if the mice are randomly responding to suprathreshold concentrations. The mice included in analyses did not exhibit evidence of losing stimulus control. Statistical methods. Thresholds were used as the dependent variable and the factors of the linear mixed-model ANOVA were drug treatment (3 levels: saline, 5*20 CYP, 1*100 CYP; between subjects), and day postinjection (25 levels; within subjects). The data were analyzed using an approach similar to that described for experiment 1. In addition, post/pre differences (calculated as [threshold for day – average of pretreatment threshold scores]) were analyzed using the same linear mixed-model ANOVA as described for the threshold scores. This was a control performed to ensure that individual skill levels of the mice did not influence the results. Results Prior to any analyses, 1 mouse in the 5*20 CYP group was removed from the experiment because estimated detection thresholds were likely too high to measure accurately on multiple days. The 2-factor mixed-model ANOVA found that detection thresholds were significantly affected by drug treatment (F(2,92) = 35.47, P < 0.0005), but not day posttreatment (P = 0.183) or the interaction of drug treatment and day posttreatment (P = 0.205). The main effect of drug treatment was also analyzed with post hoc Sidak tests (α-corrected) that found significant differences between saline and each of the CYP groups (1*100 CYP = P < 0.0005, and 5*20 CYP = P < 0.0005).There was no significant difference between the 2 CYP groups (P = 0.247). Saline controls generally maintained very consistent thresholds near those seen during training, with a mean threshold ± SEM of 3.7 ± 3.0 mM NaCl (Figure 3). The 1*100 CYP mice had higher mean detection thresholds than saline mice, averaging 32.0 ± 3.4 mM NaCl with brief periods of more severe disruptions peaking at days 6, 14, and 20 (Figure 3A). The thresholds for the 5*20 CYP group were slightly higher over the study, averaging 40.0 ± 3.4 mM. Interestingly, average detection thresholds of the 5*20 CYP group reached a maximum disruption roughly every 4–5 days, with gradual rises and falls in thresholds before and after (Figure 3B). In general, both CYP groups had much higher detection thresholds when compared to saline controls. Figure 3. View largeDownload slide NaCl detection thresholds over days posttreatment. The x axis is days posttreatment. The y axis is the mean (±SEM) detection threshold concentration (in mM), plotted as a logarithmic scale. Both CYP groups showed sustained increases in NaCl detection thresholds. (A) The 1*100 CYP group had more severe, more transient elevations of salt detection than the 5*20 CYP group, with 3 days of peak increases, 6, 14, and 20. (B) The thresholds of 5*20 CYP mice were generally less severely impaired than 1*100 CYP thresholds, but their average threshold was slightly higher and appeared more cyclic over the course of the experiment. The same data for the saline mice are presented in both panels to make visual comparisons easier. * = P < 0.05, ** = P < 0.01, *** = P < 0.001. Figure 3. View largeDownload slide NaCl detection thresholds over days posttreatment. The x axis is days posttreatment. The y axis is the mean (±SEM) detection threshold concentration (in mM), plotted as a logarithmic scale. Both CYP groups showed sustained increases in NaCl detection thresholds. (A) The 1*100 CYP group had more severe, more transient elevations of salt detection than the 5*20 CYP group, with 3 days of peak increases, 6, 14, and 20. (B) The thresholds of 5*20 CYP mice were generally less severely impaired than 1*100 CYP thresholds, but their average threshold was slightly higher and appeared more cyclic over the course of the experiment. The same data for the saline mice are presented in both panels to make visual comparisons easier. * = P < 0.05, ** = P < 0.01, *** = P < 0.001. Although the ANOVA did not detect a significant day effect, there appeared to be some days when CYP-treated mice were more severely affected relative to control mice that might have been obscured by the error variance. Consequently, the data were partitioned to examine the effects of drug treatment on NaCl thresholds each day following the completion of drug treatments. These ANOVAs found significant increases in thresholds due to the drug treatment condition on days 4, 6, 10–12, 14–15, 20, and 23 postinjection (F(2,240) ≥ 3.17, all P values < 0.03). To determine which groups were affected on these days, simple-effects testing (Howell 2016) of treatment condition by day posttreatment found significantly higher thresholds for the 1*100 CYP compared to saline-injected controls on days 5–6, 12, 14, and 20 (P < 0.05; Figure 3A and Table 1). NaCl thresholds were particularly elevated on day 6 posttreatment for 1*100 CYP mice, with estimated thresholds averaging 109.4 ± 17.2 mM NaCl ( X¯ ± SEM) compared to 4.7 ± 15.4 mM NaCl for saline-injected mice. The 5*20 CYP mice showed significantly higher thresholds compared to saline-injected controls on days 4, 9–11, 15–16, 20, and 23 posttreatment (all P values < 0.05; Figure 3B and Table 1).The disruptions to detection thresholds observed in this group were much more cyclic than the 1*100 CYP mice. When compared to the 5*20 CYP group, the 1*100 CYP group had significantly higher thresholds on days 6 and 14, whereas the 5*20 CYP group had significantly higher thresholds than the 1*100 CYP group on days 10 and 11 posttreatment (P < 0.05). Table 1. P values from difference scores of experiment 2. Significant values are indicated by asterisk Day Saline vs. 1*100 CYP Saline vs. 5*20 CYP 5*20 CYP vs. 1*100 CYP 2 0.988 0.642 0.882 3 0.216 0.180 0.999 4 0.599 0.010* 0.219 5 0.022* 0.075 0.630 6 <0.0005* 0.160 0.042* 7 0.998 0.883 0.951 8 0.773 0.696 0.999 9 0.686 0.066 0.465 10 0.763 0.003* 0.043* 11 0.870 0.018* 0.140 12 0.049* 0.064 0.609 13 0.618 0.967 0.889 14 0.002* 0.986 0.010* 15 0.840 0.040* 0.271 16 0.878 0.962 0.994 17 0.974 0.133 0.332 18 0.693 0.425 0.975 19 0.975 1.000 0.975 20 0.034* 0.021* 0.882 21 0.717 0.630 0.999 22 0.980 0.235 0.474 23 0.498 0.011* 0.303 24 0.814 0.881 0.999 25 0.698 0.553 0.996 26 0.999 0.999 0.999 Day Saline vs. 1*100 CYP Saline vs. 5*20 CYP 5*20 CYP vs. 1*100 CYP 2 0.988 0.642 0.882 3 0.216 0.180 0.999 4 0.599 0.010* 0.219 5 0.022* 0.075 0.630 6 <0.0005* 0.160 0.042* 7 0.998 0.883 0.951 8 0.773 0.696 0.999 9 0.686 0.066 0.465 10 0.763 0.003* 0.043* 11 0.870 0.018* 0.140 12 0.049* 0.064 0.609 13 0.618 0.967 0.889 14 0.002* 0.986 0.010* 15 0.840 0.040* 0.271 16 0.878 0.962 0.994 17 0.974 0.133 0.332 18 0.693 0.425 0.975 19 0.975 1.000 0.975 20 0.034* 0.021* 0.882 21 0.717 0.630 0.999 22 0.980 0.235 0.474 23 0.498 0.011* 0.303 24 0.814 0.881 0.999 25 0.698 0.553 0.996 26 0.999 0.999 0.999 View Large Table 1. P values from difference scores of experiment 2. Significant values are indicated by asterisk Day Saline vs. 1*100 CYP Saline vs. 5*20 CYP 5*20 CYP vs. 1*100 CYP 2 0.988 0.642 0.882 3 0.216 0.180 0.999 4 0.599 0.010* 0.219 5 0.022* 0.075 0.630 6 <0.0005* 0.160 0.042* 7 0.998 0.883 0.951 8 0.773 0.696 0.999 9 0.686 0.066 0.465 10 0.763 0.003* 0.043* 11 0.870 0.018* 0.140 12 0.049* 0.064 0.609 13 0.618 0.967 0.889 14 0.002* 0.986 0.010* 15 0.840 0.040* 0.271 16 0.878 0.962 0.994 17 0.974 0.133 0.332 18 0.693 0.425 0.975 19 0.975 1.000 0.975 20 0.034* 0.021* 0.882 21 0.717 0.630 0.999 22 0.980 0.235 0.474 23 0.498 0.011* 0.303 24 0.814 0.881 0.999 25 0.698 0.553 0.996 26 0.999 0.999 0.999 Day Saline vs. 1*100 CYP Saline vs. 5*20 CYP 5*20 CYP vs. 1*100 CYP 2 0.988 0.642 0.882 3 0.216 0.180 0.999 4 0.599 0.010* 0.219 5 0.022* 0.075 0.630 6 <0.0005* 0.160 0.042* 7 0.998 0.883 0.951 8 0.773 0.696 0.999 9 0.686 0.066 0.465 10 0.763 0.003* 0.043* 11 0.870 0.018* 0.140 12 0.049* 0.064 0.609 13 0.618 0.967 0.889 14 0.002* 0.986 0.010* 15 0.840 0.040* 0.271 16 0.878 0.962 0.994 17 0.974 0.133 0.332 18 0.693 0.425 0.975 19 0.975 1.000 0.975 20 0.034* 0.021* 0.882 21 0.717 0.630 0.999 22 0.980 0.235 0.474 23 0.498 0.011* 0.303 24 0.814 0.881 0.999 25 0.698 0.553 0.996 26 0.999 0.999 0.999 View Large To evaluate shifts in thresholds independent of the skill level of each mouse, post/pre differences in thresholds, calculated as the average of preinjection threshold estimates subtracted from the threshold estimates for the day, were analyzed using a 3-factor mixed ANOVA. A main effect was found for drug treatment (F(2, 83) = 30.83, P < 0.0005). The saline group posttreatment scores showed they were consistent with pretreatment thresholds, with a mean difference score ± SEM of −2.3 ± 1.7 mM NaCl. However, contrary to raw threshold scores, the 5*20 CYP group maintained a larger average difference score throughout the experiment, with a mean ± SEM of 32.4 ± 12.1 mM NaCl compared to 26.6 ± 10.2 mM NaCl for 1*100 CYP-treated animals, although this difference was not significant. As with raw threshold estimates, main effects were not found for day posttreatment or the interaction of drug treatment and day posttreatment (P = 0.168, 0.254, respectively). Difference scores were further partitioned to compare group differences by day posttreatment, since previous analyses of raw threshold estimates had uncovered significance with this partitioning. These ANOVAs found significant group differences on days 4, 6, 10–11, 14, 20, and 23 posttreatment (F(2, 240) ≥ 3.28, all P values ≤ 0.039).The comparisons between the 1*100 CYP group and the saline group mirrored the results of threshold testing, with significantly larger difference scores in the CYP group seen on days 5, 6, 12, 14, and 20 posttreatment (P < 0.05; Figure 4A), indicating impaired salt taste consistent with raw threshold estimates. The magnitude of these disruptions also remained the same, with a mean difference score ± SEM on day 6 of 109.1 ± 12.0 mM NaCl. Comparing difference scores of the 5*20 CYP group to the saline group, significantly larger differences (higher thresholds) were observed for the 5*20 CYP group on days 4, 9–11, 15, 17, 20, and 23 posttreatment (P < 0.05; Figure 4B). Comparisons between the 1*100 and 5*20 CYP groups found significantly higher thresholds for the 1*100 CYP group on days 6 and 14 (P < 0.05).The difference scores of 5*20 CYP mice also mirrored the raw threshold estimates in the timing and magnitude of disruptions. Generally, the results of the difference scores nearly matched those of calculated detection thresholds, verifying the threshold findings and suggesting that individual skill of mice did not play an important role in the trends observed. Figure 4. View largeDownload slide Post/pre threshold difference scores over days posttreatment. The x axis represents days posttreatment. The y axis represents the mean (±SEM) of post/pre threshold difference scores as millimolar concentrations, calculated by subtracting the average pretreatment threshold from the estimated threshold by day for a single mouse. (A) The threshold difference scores for the 1*100 CYP and the saline mice over days are shown. (B) The threshold difference scores for the 5*20 CYP and the saline mice over days are shown. The same data for the saline mice are presented in both panels to make visual comparisons easier. * = P < 0.05, ** = P < 0.01, *** = P < 0.001. Figure 4. View largeDownload slide Post/pre threshold difference scores over days posttreatment. The x axis represents days posttreatment. The y axis represents the mean (±SEM) of post/pre threshold difference scores as millimolar concentrations, calculated by subtracting the average pretreatment threshold from the estimated threshold by day for a single mouse. (A) The threshold difference scores for the 1*100 CYP and the saline mice over days are shown. (B) The threshold difference scores for the 5*20 CYP and the saline mice over days are shown. The same data for the saline mice are presented in both panels to make visual comparisons easier. * = P < 0.05, ** = P < 0.01, *** = P < 0.001. Discussion Saline-injected mice maintained relatively consistent detection thresholds over the experiment with estimates around 1–5 mM NaCl, except the last 4 days. These threshold estimates are consistent with previously reported thresholds in C57BL/6J mice and are near those reported for rats (Geran and Spector 2000; Ruiz et al. 2006; Ishiwatari and Bachmanov 2012). The thresholds in this experiment were determined with an operant task using reinforcement and punishment, neither of which directly affected the tongue. Unlike some previous work using this methodology (Mukherjee and Delay 2011), no shock was used as a punisher during testing or training, making the only alterations to the lingual tissue from the CYP treatment. Both CYP groups show substantial increases in salt detection thresholds when compared to the saline-injected mice. The 1*100 CYP group had elevated thresholds every 6–8 days that decreased in severity each time. These disruptions were close to the postinjection periods hypothesized for the 1*100 CYP group. As anticipated, NaCl detection thresholds in the 5*20 CYP group were also significantly elevated and appeared to have cyclic-like characteristics. In general, it is clear that detection thresholds were compromised in both CYP groups and the effects of CYP on NaCl taste functions in this experiment were much more pronounced than those seen in the 1st experiment. General discussion Earlier research had shown that CYP elevated detection thresholds, a measure of taste sensitivity, for sucrose and umami substances and disrupted suprathreshold discrimination between 2 umami substances (Mukherjee and Delay 2011; Mukherjee et al. 2013). These experiments extended the study of the effects of this chemotherapy agent by showing that CYP also negatively affected salt taste. What was surprising, however, was that the drug treatment regimen had a different impact in each experiment. Brief-access testing after a single dose of 100 mg/kg revealed a lessening of the aversive properties of NaCl, detectable in lick rates of suprathreshold concentrations near days 5–9 posttreatment. Fractionation of the same amount of drug over days (5*20 CYP) had little impact on taste behavior in the same brief-access test. On the other hand, detection thresholds for NaCl were elevated significantly by both methods of dosing, although the magnitude and posttreatment duration of these effects depended somewhat upon the method of dose administration. Even though CYP and its metabolites (acrolein and phosphoramide mustard) are cleared within hours after administration (Ahmed and Hombal 1984), the metabolites are cytotoxic and can kill taste cells directly or induce apoptosis in TSCs and the more susceptible progenitor cells that generate replacement cells for taste buds (Mukherjee et al. 2017). Using the model developed from previous work by our lab (detailed in the Introduction), periods of disrupted behavior should correlate with damage to populations of TSCs (Mukherjee and Delay 2011; Mukherjee et al. 2013, 2017). Based on knowledge of cellular level disruptions after CYP treatment, we estimated when disruptions to behavior based on salt taste function might occur. The 1*100 CYP groups are the most appropriate groups for assessing the magnitude and temporal patterns of the effects of CYP since the mice received the full dose of the drug in 1 exposure. In experiment 1, saline-injected mice consistently showed a general aversion to NaCl relative to water. In contrast, brief-access lick rates in the 1*100 CYP group were generally higher than saline controls through the 1st 15 days after CYP injection, especially days 5–10, suggesting that these mice found the NaCl solutions less aversive than controls. A jump in their normalized lick rates to slightly over 1.0 on day 8 suggests that NaCl had lost most or all of its aversive qualities that day. The time frame of these effects coincides with the loss and recovery of fungiform papillae previously documented (Mukherjee and Delay 2011), where a substantial portion of the salt signal is generated. Moreover, the spike in lick rates 8 days after injection aligns with the potential half-lives of Type I and Type II TSCs. By 8 days postinjection it would be expected that many of these TSCs would have reached the end of their normal life span but would not yet be replaced because of the interruption of the TSC renewal process. It is possible that the remaining oral sensation, potentially reinforced by postingestive effects of salt intake, might have reinforced the mice’s attention to the weakened salt signal, enabling them to compensate the next day to reduce their sodium intake. The differences in CYP-induced changes in behavior between experiments 1 and 2 may also be due to differential effects of CYP treatment on the taste sensory signal needed to perform the behavior (Spector and Grill 1992; Mukherjee and Delay 2011; Dana and McCaughey 2015; Mukherjee et al. 2017). The timing and magnitude of these effects may differ when measuring appetitive qualities that are dependent upon identifying suprathreshold qualities of NaCl, or detection thresholds that are dependent upon distinguishing low concentrations of NaCl from water. Elevations in taste thresholds in this study are likely to reflect a loss of relevant TSCs. The 1*100 CYP group in experiment 2 exhibited losses in salt sensitivity more consistent with our working model but require a more complex explanation as it is likely that all 3 TSC types are affected by CYP. It is important to note that at this time there is no direct evidence of Type I TSC death after CYP treatment due to difficulties in accurate quantification. However, because entire fungiform papillae and circumvallate taste buds are lost after CYP treatment, and there is no evidence to date that one cell type is more resistant to the effects of CYP than the other taste cell types, it seems likely that Type I cells are also vulnerable to the effects of CYP (Mukherjee and Delay 2011). In addition, the 1*100 CYP mice had their largest disruption in behavior on day 8 posttreatment in experiment 1, characterized by increased licking of NaCl solutions relative to water. Increased detection thresholds for these CYP mice, although consistently higher than controls, were the most severe at 3 time periods in experiment 2: days 5–6, 12, 14, and 20. The disruptions on days 5–6 and 8 in experiments 1 and 2, respectively, approximate the half-life of Type II TSCs and/or a subpopulation of Type I TSCs, whereas further disruptions seen in experiment 2 at day 20 precede the half-life of Type III and/or a subpopulation of Type I TSCs (Perea-Martinez et al. 2013). These TSC types have all been shown to respond to salt stimuli, and Type I cells are known to detect amiloride-sensitive salt taste signals in mice (Vandenbeuch et al. 2008; Tordoff et al. 2014; Lewandowski et al. 2016; Price et al. 2016; Roebber et al. 2017). Although there is no reported TSC with a half-life that appears to account for the increase in detection thresholds on days 12 and 14 in experiment 2, this may represent the maximum combined loss of fungiform and circumvallate taste buds that occurs just before the number of taste buds and the number of mature, functioning TSCs within taste buds begin to recover (Mukherjee and Delay 2011; Mukherjee et al. 2013, 2017). Both experiments detected disruptions in salt taste near the proposed lifespans of a subpopulation of Types I and II TSCs, but only experiment 2 detected disrupted behavior near the half-life of Type III TSCs and a possible subpopulation of Type I TSCs. With their much longer life span, Type III cells have a different turnover rate after CYP challenge and might need to be replaced at a different rate after injection than the other cell types. Moreover, their role in salt transduction is not fully known in terms of detection and/or encoding salt qualities. Further research is needed to elucidate the behavioral role of Type III TSCs in salt taste and how their destruction by CYP affects salt taste functions. Fractionating the CYP dose yielded some unique findings. Fractionated dosing only minimally affected ingestive behavior during the brief-access test. However, fractionating the dose significantly elevated detection thresholds in experiment 2. The 1*100 CYP group had 3 posttreatment periods when spikes in detection threshold elevations occurred, followed by relatively rapid improvements in performance accuracy. As seen in previous studies, the rapid elevation in threshold occurs when there is a rapid loss in key populations of the TSCs generating the sensation that cues the mouse’s response in the threshold task. When that signal is altered by the loss of TSCs, the mouse must try to identify an alternative sensation that enables it to optimize reinforcement and minimize punishment. It appears that this task may have been more difficult for animals in the 5*20 CYP group, especially when compared to the results of the 1st experiment, where suprathreshold stimuli may have provided useable sensory signals to support appetitive behavior. Moreover, the 1st 2 threshold shifts occurred 1–2 days earlier than those of the 1*100 CYP group. Unpublished data in our lab found that a single administration of doses as low as 18.75 mg/kg can cause detectable changes in these tissues, although the extent of the tissue damage is much less than that by higher doses. Preliminary work also indicates that not only the fractionated doses of 20 mg/kg result in loss of Type II and III cells, but the period of depressed cell proliferation lasts longer than a single dose of 100 mg/kg CYP (Socia et al. 2017).The elevation of thresholds of 5*20 CYP-treated animals had some disruptions as severe as 1*100 CYP mice especially toward the latter portion of experiment 2. This suggests that although fractionated doses may not have initially killed as many TSCs, these effects on cell renewal cumulate with repeated administration and may result in taste deficits that are initially milder but are longer in duration. There are also other factors that might have influenced these results. Because of the longitudinal nature of these experiments, olfaction could not be reliably removed, and was still available to the mice, especially during brief-access testing. The control procedures in the threshold experiment were designed to minimize olfactory cues for control mice as well as CYP-treated mice. In addition, olfaction may not play a salient role for CYP-treated mice ashuman and murine data suggest that there are also disruptions to olfaction during and after chemotherapy treatment (Joseph et al. 2017; Walliczek-Dworschak et al. 2017). CYP groups in both experiments may have also experienced xerostomia, a side effect of chemotherapy that has been shown to affect gustatory sensitivity (Pico et al. 1998; Temmel et al. 2005). This may have worsened already existing taste impairments observed in the threshold experiment, but does not mediate CYP-induced taste deficits in mice (Mukherjee and Delay 2011). It is unlikely that the dose of CYP used in this study, singly or fractionated, damaged neurons of the central nervous system and altered cognition as it has difficulty crossing the blood–brain barrier (Neuwelt et al. 1983). However, although unlikely, CYP might have reduced taste input to the nucleus of the solitary tract enough to cause changes in response profiles of these neurons (Di Lorenzo et al. 1997). Although further research may be needed to eliminate some of these factors, in general, it seem likely that the observed drug-induced changes in behavior are taste mediated. This study also raises important questions for clinical investigation. Even though we did not measure direct damage to the tongue from the multiple injection condition (5*20 CYP), it was clearly harder for the mice to identify salty stimuli. This mirrors clinical reports that salt taste is the most affected among cancer patients (Bernhardson et al. 2008; Steinbach et al. 2009), who often receive fractionated doses of chemotherapeutics (Ozols and Young 1984; Levin and Hryniuk 1987; Lamar et al. 2016). The 5*20 CYP group’s cyclicity in detection thresholds may suggest that this fractionated treatment makes it even more difficult to compensate for changes in salt taste. Although mice are not directly comparable to humans, many of the observed injuries (Mukherjee and Delay 2011; Mukherjee et al. 2013) have been similarly observed in humans (Vacha et al. 2003; Just et al. 2005; Srur et al. 2011). Taste is an important quality of making eating enjoyable, which is a problem for cancer patients who often suffer from malnutrition during the course of their treatment (Davidson et al. 2012). This study suggests that there is a compensatory mechanism for salt taste loss, and that it may be easier to adjust to the insult from a single dose rather than the more clinically common fractionated treatment. In summary, this study has shown that CYP treatment results in multiple, potentially severe disruptions to salt taste. These disruptions differed depending on whether salt taste was being tested for detection thresholds or appetitive qualities. There were some periods of overlap between single-injection groups. This suggests periods where cell populations critical to either detection or acceptability, or possibly both qualities are damaged. These data suggest that multiple TSC types are likely involved in normal salt taste and that all are susceptible to the effects of CYP, either from the direct cytotoxic effects of the drug or from the disruption in the cell renewal system responsible for maintaining the cellular populations within taste buds. Regional (papillae-based) differences in salt taste may also have contributed to the increased sensitivity of detection thresholds to CYP treatment as the fungiform taste buds are more susceptible to the effects of CYP than circumvallate taste buds. Although the effects of a moderate dose of CYP given as a single injection on the taste have been characterized, the effects of dose fractionation appear to be somewhat different and a promising area for further investigation of processes underlying taste cell renewal. Funding This work was supported by the National Institutes of Health [R01DC012829 to E.R.D.], as well as the University of Vermont Undergraduate Research Program [Brennan Summer Undergraduate Research Fellowship awarded to B.C.J.], and the University of Vermont College of Arts and Sciences [APLE grants to M.G.G. and B.C.J.]. Conflicts of Interest The authors declare they have no conflicts of interest. Acknowledgement The authors thank Alan Howard of the University of Vermont for invaluable input to the article. References Aaldriks AA , van der Geest LG , Giltay EJ , le Cessie S , Portielje JE , Tanis BC , Nortier JW , Maartense E . 2013 . Frailty and malnutrition predictive of mortality risk in older patients with advanced colorectal cancer receiving chemotherapy . J Geriatr Oncol . 4 : 218 – 226 . Ahmed AR , Hombal SM . 1984 . Cyclophosphamide (Cytoxan). A review on relevant pharmacology and clinical uses . J Am Acad Dermatol . 11 : 1115 – 1126 . Anderson D , Bishop JB , Garner RC , Ostrosky-Wegman P , Selby PB . 1995 . Cyclophosphamide: review of its mutagenicity for an assessment of potential germ cell risks . Mutat Res . 330 : 115 – 181 . Bachmanov AA , Tordoff MG , Beauchamp GK . 1996 . Ethanol consumption and taste preferences in C57BL/6ByJ and 129/J mice . Alcohol Clin Exp Res . 20 : 201 – 206 . Beauchamp GK , Fisher AS . 1993 . Strain differences in consumption of saline solutions by mice . Physiol Behav . 54 : 179 – 184 . Bernhardson BM , Tishelman C , Rutqvist LE . 2008 . Self-reported taste and smell changes during cancer chemotherapy . Support Care Cancer . 16 : 275 – 283 . Bernstein IL . 1978 . Learned taste aversions in children receiving chemotherapy . Science . 200 : 1302 – 1303 . Berridge KC , Flynn FW , Schulkin J , Grill HJ . 1984 . Sodium depletion enhances salt palatability in rats . Behav Neurosci . 98 : 652 – 660 . Bertino M , Beauchamp GK , Engelman K . 1982 . Long-term reduction in dietary sodium alters the taste of salt . Am J Clin Nutr . 36 : 1134 – 1144 . Boltong A , Keast R . 2012 . The influence of chemotherapy on taste perception and food hedonics: a systematic review . Cancer Treat Rev . 38 : 152 – 163 . Brosvic GM , Slotnick BM . 1986 . Absolute and intensity-difference taste thresholds in the rat: evaluation of an automated multi-channel gustometer . Physiol Behav . 38 : 711 – 717 . Chaudhari N , Roper SD . 2010 . The cell biology of taste . J Cell Biol . 190 : 285 – 296 . Dana RM , McCaughey SA . 2015 . Gustatory responses of the mouse chorda tympani nerve vary based on region of tongue stimulation . Chem Senses . 40 : 335 – 344 . Davidson W , Teleni L , Muller J , Ferguson M , McCarthy AL , Vick J , Isenring E . 2012 . Malnutrition and chemotherapy-induced nausea and vomiting: implications for practice . Oncol Nurs Forum . 39 : E340 – E345 . Delay ER , Hernandez NP , Bromley K , Margolskee RF . 2006 . Sucrose and monosodium glutamate taste thresholds and discrimination ability of T1R3 knockout mice . Chem Senses . 31 : 351 – 357 . Di Lorenzo PM , Lemon CH , Kawamoto MD . 1997 . Changes in response profiles of taste cells in the brain stem parallel those observed in somatosensory thalamic cells following anesthetization of part of their receptive fields . In: Bower JM , editor. Computational neuroscience: trends in research . Boston, MA: Springer . p. 71 – 73 . Duncan CJ . 1962 . Salt preferences of birds and mammals . Physiol Zool . 35 : 120 – 132 . Eddy MC , Eschle BK , Barrows J , Hallock RM , Finger TE , Delay ER . 2009 . Double P2X2/P2X3 purinergic receptor knockout mice do not taste NaCl or the artificial sweetener SC45647 . Chem Senses . 34 : 789 – 797 . Geran LC , Spector AC . 2000 . Amiloride increases sodium chloride taste detection threshold in rats . Behav Neurosci . 114 : 623 – 634 . Glendinning JI , Gresack J , Spector AC . 2002 . A high-throughput screening procedure for identifying mice with aberrant taste and oromotor function . Chem Senses . 27 : 461 – 474 . Heck GL , Mierson S , DeSimone JA . 1984 . Salt taste transduction occurs through an amiloride-sensitive sodium transport pathway . Science . 223 : 403 – 405 . Howell DC . 2016 . Fundamental statistics for the behavioral sciences . Belmont (CA) : Wadsworth Publishing . Ishimaru Y , Inada H , Kubota M , Zhuang H , Tominaga M , Matsunami H . 2006 . Transient receptor potential family members pkd1l3 and pkd2l1 form a candidate sour taste receptor . Proc Natl Acad Sci USA . 103 : 12569 – 12574 . Ishiwatari Y , Bachmanov AA . 2012 . NaCl taste thresholds in 13 inbred mouse strains . Chem Senses . 37 : 497 – 508 . Jewkes BC , Barlow LA , Delay ER . 2017 . Effect of radiation on sucrose detection thresholds of mice . Chem Senses . 43 : 53 – 58 . Joseph K , Awadallah N , Delay ER , Delay RJ . 2017 . Co-localization of basal and proliferative cells in the olfactory epithelium after cyclophosphamide injury . Bonita Springs (FL) : AChems XXXIX . Just T , Pau HW , Bombor I , Guthoff RF , Fietkau R , Hummel T . 2005 . Confocal microscopy of the peripheral gustatory system: comparison between healthy subjects and patients suffering from taste disorders during radiochemotherapy . Laryngoscope . 115 : 2178 – 2182 . Lamar ZS , Fino N , Palmer J , Gruber L , Morris BB , Raetskaya-Solntseva O , Kennedy L , Vaidya R , Hurd D , Zamkoff K . 2016 . Dose-adjusted etoposide, prednisone, vincristine, cyclophosphamide, and doxorubicin (EPOCH)with or without rituximab as first-line therapy for aggressive non-Hodgkin lymphoma . Clin Lymphoma Myeloma Leuk . 16 : 76 – 81 . Levin L , Hryniuk WM . 1987 . Dose intensity analysis of chemotherapy regimens in ovarian carcinoma . J Clin Oncol . 5 : 756 – 767 . Lewandowski BC , Sukumaran SK , Margolskee RF , Bachmanov AA . 2016 . Amiloride-insensitive salt taste is mediated by two populations of Type III taste cells with distinct transduction mechanisms . J Neurosci . 36 : 1942 – 1953 . Lindemann B . 2001 . Receptors and transduction in taste . Nature . 413 : 219 – 225 . Logue AW . 1979 . Taste aversion and the generality of the laws of learning . Psychol Bull . 86 : 276 . Miyamoto T , Miyazaki T , Okada Y , Sato T . 1996 . Whole-cell recording from non-dissociated taste cells in mouse taste bud . J Neurosci Methods . 64 : 245 – 252 . Mukherjee N , Carroll BL , Spees JL , Delay ER . 2013 . Pre-treatment with amifostine protects against cyclophosphamide-induced disruption of taste in mice . PLoS One . 8 : e61607 . Mukherjee N , Delay ER . 2011 . Cyclophosphamide-induced disruption of umami taste functions and taste epithelium . Neuroscience . 192 : 732 – 745 . Mukherjee N , Pal Choudhuri S , Delay RJ , Delay ER . 2017 . Cellular mechanisms of cyclophosphamide-induced taste loss in mice . PLoS One . 12 : e0185473 . Nelson G , Hoon MA , Chandrashekar J , Zhang Y , Ryba NJ , Zuker CS . 2001 . Mammalian sweet taste receptors . Cell . 106 : 381 – 390 . Neuwelt EA , Barnett PA , Glasberg M , Frenkel EP . 1983 . Pharmacology and neurotoxicity of cis-diamminedichloroplatinum, bleomycin, 5-fluorouracil, and cyclophosphamide administration following osmotic blood-brain barrier modification . Cancer Res . 43 : 5278 – 5285 . Ozols RF , Young RC . 1984 . Chemotherapy of ovarian cancer . Semin Oncol . 11 : 251 – 263 . Perea-Martinez I , Nagai T , Chaudhari N . 2013 . Functional cell types in taste buds have distinct longevities . PLoS One . 8 : e53399 . Pico JL , Avila-Garavito A , Naccache P . 1998 . Mucositis: its occurrence, consequences, and treatment in the oncology setting . Oncologist . 3 : 446 – 451 . Price J , Tordoff MG , Knox J , Aleman TR , Margolskee R , McCaughey SA . 2016 . Role of alpha-gustducin and itpr3 in salty taste in mice . Bonita Springs (FL) : AChems XXVIII . Roebber JK , Roper S , Chaudhari N . 2017 . Which cells detect salty taste in fungiform taste buds? [abstract] . Bontia Springs (Fl.) : AChems XXXIX . Roper SD . 2015 . The taste of table salt . Pflugers Arch . 467 : 457 – 463 . Ruiz C , Gutknecht S , Delay E , Kinnamon S . 2006 . Detection of NaCl and KCl in TRPV1 knockout mice . Chem Senses . 31 : 813 – 820 . Scott TR . 1990 . The effect of physiological need on taste. Taste, experience, and feeding . Washington (DC, USA) : American Psychological Association . p. 45 – 61 . Socia SH , Girardin JL , King JH , Delay ER . 2017 . Effects of cyclophosphamide and amifostine on the taste cell replacement cycle [abstract] . Bonita Springs (FL) : AChems XXXIX . Spector AC , Grill HJ . 1992 . Salt taste discrimination after bilateral section of the chorda tympani or glossopharyngeal nerves . Am J Physiol . 263 : R169 – R176 . Srur E , Pau HW , Just T . 2011 . Changes in taste bud volume during taste disturbance . Auris Nasus Larynx . 38 : 512 – 515 . Steinbach S , Hummel T , Böhner C , Berktold S , Hundt W , Kriner M , Heinrich P , Sommer H , Hanusch C , Prechtl A , et al. 2009 . Qualitative and quantitative assessment of taste and smell changes in patients undergoing chemotherapy for breast cancer or gynecologic malignancies . J Clin Oncol . 27 : 1899 – 1905 . Temmel AF , Quint C , Schickinger-Fischer B , Hummel T . 2005 . Taste function in xerostomia before and after treatment with a saliva substitute containing carboxymethylcellulose . J Otolaryngol . 34 : 116 – 120 . Tordoff MG , Ellis HT , Aleman TR , Downing A , Marambaud P , Foskett JK , Dana RM , McCaughey SA . 2014 . Salty taste deficits in CALHM1 knockout mice . Chem Senses . 39 : 515 – 528 . Vacha P , Fehlauer F , Mahlmann B , Marx M , Hinke A , Sommer K , Richter E , Feyerabend T . 2003 . Randomized phase III trial of postoperative radiochemotherapy +/- amifostine in head and neck cancer. Is there evidence for radioprotection ? Strahlenther Onkol . 179 : 385 – 389 . Vandenbeuch A , Clapp TR , Kinnamon SC . 2008 . Amiloride-sensitive channels in type I fungiform taste cells in mouse . BMC Neurosci . 9 : 1 . Vizzard MA . 2000 . Changes in urinary bladder neurotrophic factor mRNA and NGF protein following urinary bladder dysfunction . Exp Neurol . 161 : 273 – 284 . Walliczek-Dworschak U , Gudziol V , Mitzschke C , Froehner M , Hummel T . 2017 . Testicular cancer patients undergoing cisplatin based chemotherapy exhibit temporary olfactory threshold scores changes . Eur Arch Otorhinolaryngol . 274 : 2813 – 2818 . Wayner EA , Flannery GR , Singer G . 1978 . Effects of taste aversion conditioning on the primary antibody response to sheep red blood cells and Brucella abortus in the albino rat . Physiol Behav . 21 : 995 – 1000 . Weiner HL , Cohen JA . 2002 . Treatment of multiple sclerosis with cyclophosphamide: critical review of clinical and immunologic effects . Mult Scler . 8 : 142 – 154 . Wickham RS , Rehwaldt M , Kefer C , Shott S , Abbas K , Glynn-Tucker E , Potter C , Blendowski C . 1999 . Taste changes experienced by patients receiving chemotherapy . Oncol Nurs Forum . 26 : 697 – 706 . Yoshida R , Horio N , Murata Y , Yasumatsu K , Shigemura N , Ninomiya Y . 2009 . NaCl responsive taste cells in the mouse fungiform taste buds . Neuroscience . 159 : 795 – 803 . Zocchi D , Wennemuth G , Oka Y . 2017 . The cellular mechanism for water detection in the mammalian taste system . Nat Neurosci . 20 : 927 – 933 . © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: email@example.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
Chemical Senses – Oxford University Press
Published: May 19, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera