Taste dysfunction in BTBR mice due to a mutation of Itpr3, the inositol triphosphate receptor 3 gene

Michael G Tordoff; Hillary T Ellis

doi:10.1152/physiolgenomics.00092.2013

. 2013 Jul 16;45(18):834–855. doi: 10.1152/physiolgenomics.00092.2013

Taste dysfunction in BTBR mice due to a mutation of Itpr3, the inositol triphosphate receptor 3 gene

Michael G Tordoff ^1,^✉, Hillary T Ellis ¹

PMCID: PMC3783815 PMID: 23859941

Abstract

The BTBR T⁺ tf/J (BTBR) mouse strain is indifferent to exemplars of sweet, Polycose, umami, bitter, and calcium tastes, which share in common transduction by G protein-coupled receptors (GPCRs). To investigate the genetic basis for this taste dysfunction, we screened 610 BTBR × NZW/LacJ F₂ hybrids, identified a potent QTL on chromosome 17, and isolated this in a congenic strain. Mice carrying the BTBR/BTBR haplotype in the 0.8-Mb (21-gene) congenic region were indifferent to sweet, Polycose, umami, bitter, and calcium tastes. To assess the contribution of a likely causative culprit, Itpr3, the inositol triphosphate receptor 3 gene, we produced and tested Itpr3 knockout mice. These were also indifferent to GPCR-mediated taste compounds. Sequencing the BTBR form of Itpr3 revealed a unique 12 bp deletion in Exon 23 (Chr 17: 27238069; Build 37). We conclude that a spontaneous mutation of Itpr3 in a progenitor of the BTBR strain produced a heretofore unrecognized dysfunction of GPCR-mediated taste transduction.

Keywords: NZW/LacJ, sweet, umami, bitter, QTL

the BTBR T⁺ tf/J (BTBR) mouse strain was initially developed to study embryogenesis and infertility because its “tufted” (tf) hair pattern identified animals carrying the lethal T-complex (reviews Refs. 50, 55).¹ More recently, it has been used to investigate the physiological mechanisms underlying abdominal obesity and insulin-resistant diabetes (14, 22, 42) and as a model for low sociability, autism, and neurological abnormalities including the absence of a corpus callosum (31, 39, 59, 76)]. We became interested in the BTBR strain because it had unusually high taste preferences for calcium solutions (68); it ranked second out of 40 mouse strains in its avidity for calcium [after the PWK/PhJ strain (68, 70)]. Here, we report a series of experiments culminating in the identification of the gene underlying this phenotype. During the course of this work, we discovered that the taste phenotype was not specific to calcium but instead involved a wider taste dysfunction, which we describe below.

METHODS

All procedures involving animals were approved by the Animal Care and Use Committee of the Monell Chemical Senses Center. Taste compounds were purchased from Sigma Chemical (St. Louis, MO) except for calcium lactate (CaLa), which was purchased from Fisher Scientific, and trisodium pyrophosphate (Na₃HP₂O₇), which was provided by SPF-DIANA (Elven, France). All were dissolved in deionized water except capsaicin, which was first dissolved in 1 ml of 95% ethanol, and this was then dissolved in deionized water to produce the appropriate concentration. For brevity, we use “taste solution” to refer to the fluids used in two-bottle choice and brief-access tests but recognize that their olfactory, trigeminal, and/or postingestive actions may influence intake and preference.

Subjects and Maintenance

All experiments, except those with transgenic mice (see below), involved mice bred in-house from stock purchased from The Jackson Laboratory (Bar Harbor, ME) [BTBR, stock no. 002282; NZW/LacJ (NZW), stock no. 001058]. We used the NZW strain as a control parental strain for comparisons with the BTBR strain primarily because it had lower calcium preferences (68) and higher sodium preferences (69), thus avoiding the possibility that the BTBR strain's avidity for calcium was due to a nonspecific appetite for all salts or all taste compounds. Additional considerations were the divergent phylogenetic origins of the two strains (40, 74), which presumably captures more genetic diversity, and a desire to avoid rediscovering the calcium preference quantitative trait loci (QTLs) found in earlier work with a C57BL/6J × PWK/PhJ cross (70).

Mice were maintained in a vivarium at 23°C with a 12:12 h light/dark cycle (lights off at 7 PM). They were housed in plastic “tub” cages (26.5 cm × 17 cm × 12 cm) with stainless steel wire lids and wood shavings scattered on the floor. The mice were transferred to clean cages with fresh bedding once every week while waiting to be tested and once every 9–10 days during two-bottle choice tests (so as not to disturb them during a test). The cage lids included space for pelleted food and a water bottle (see Ref. 66 for details). The food was AIN-76A, a nutritionally complete semisynthetic diet that contains by weight: 20% protein (casein), 65% carbohydrate (sucrose and cornstarch), 5% fat (corn oil), and 10% fiber (cellulose), minerals, and vitamins (no. 100000; Dyets, Bethlehem, PA). Minerals in this diet include 5.2 g/kg calcium, 0.51 g/kg magnesium, 3.6 g/kg potassium, and 1.02 g/kg sodium. When mice were not being tested, deionized water was available from an inverted 300 ml glass bottle with a neoprene stopper and a stainless steel drinking spout. A detailed description of mouse husbandry, housing conditions, and other procedures is available online (66).

Mice were weaned at 21–23 days old and housed in groups of up to six of the same sex. When they were aged 6–8 wk old they were transferred to individual cages for taste phenotyping. Tests began after the mice had adapted to their new housing for at least 7 days.

Phenotyping

For reasons described below, there were four series of mice that were tested at different times in separate experiments: 1) BTBR and NZW inbred strains, 2) BTBR × NZW F₂ hybrids, 3) Chr 17 BTBR/NZW congenics and BTBR/BTBR littermate controls, and 4) Itpr3 knockout (KO) and wild-type (WT) littermates. Because of constraints due to breeding, vivarium space, and equipment, tests were made of cohorts of 16–129 mice. Each cohort was matched for the number of mice of each sex and included roughly equal numbers of littermates from each genetic group when this was pertinent (i.e., 50% congenics and 50% controls, or 50% KOs and 50% WTs). Group sizes and the constitution of each cohort are presented in Tables 1–4.

Table 1.

Analyses of parental strain two-bottle choice test preference scores

			Two-way ANOVA
Compound	Cohort and Order	Group Composition	Strain	Concentration	Strain × Concentration	Concentrations Supporting Strain Differences
Saccharin	5b	8 F	F(1,14) = 8.16, P = 0.0127	F(5,70) = 8.90, P < 0.0001	F(5,70) = 7.08, P < 0.0001	3.2, 10, 32 mM
Sucrose	2d	8 M	F(1,14) = 6.82, P = 0.0205	F(7,98) = 35.8, P < 0.0001	F(7,98) = 4.75, P = 0.0001	4, 8%
Maltose	10b	6 M, 4 F	F(1,18) = 17.3, P = 0.0006	F(7,126) = 35.3, P < 0.0001	F(7,126) = 1.21, P = 0.3013	none
Polycose	9a	8 M	F(1,14) = 5.94, P = 0.0287	F(7,98) = 61.6, P < 0.0001	F(7,98) = 4.76, P = 0.0001	1, 2, 4, 8%
MSG	10a	6 M, 4 F	F(1,18) = 20.7, P = 0.0002	F(6,108) = 20.3, P < 0.0001	F(6,108) = 7.65, P < 0.0001	100, 178, 320, 562 mM
IMP	4b	8 F	F(1,14) = 9.19, P = 0.0090	F(8,112) = 22.6, P < 0.0001	F(8,112) = 6.32, P < 0.0001	3.2, 10, 32, 100 mM
Na₃HP₂O₇	6a	16 M, 16 F	F(1,30) = 27.9, P < 0.0001	F(6,180) = 50.0, P < 0.0001	F(6,180) = 6.30, P < 0.0001	3.2, 10, 32, 56 mM
HCl	8a	8 F	F(1,14) = 5.78, P = 0.0307	F(5,70) = 13.0, P < 0.0001	F(5,70) = 1.17, P = 0.3303	none
Citric acid	4d	8 F	F(1,14) = 2.94, P = 0.1086	F(5,70) = 6.01, P = 0.0001	F(5,70) = 0.65, P = 0.6652	none
Denatonium	2a, 3a	16 M	F(1,29) = 11.0, P = 0.0024	F(6,174) = 18.4, P < 0.0001	F(6,174) = 2.90, P = 0.0102	0.32, 1.0, 3.2 mM
QHCl	2b	8 M	F(1,14) = 6.46, P = 0.0235	F(3,42) = 7.91, P = 0.0003	F(3,42) = 3.67, P = 0.0195	0.032, 0.1 mM
Caffeine	2c, 10a	16 M	F(1,30) = 32.5, P = 0.0001	F(4,120) = 56.4, P < 0.0001	F(4,120) = 6.27, P = 0.0001	1, 3.2, 10 mM
ZnCl₂	4a	8 F	F(1,14) = 0.97, P = 0.3404	F(3,42) = 33.0, P < 0.0001	F(4,42) = 0.67, P = 0.5771	none
Capsaicin	3b	8 M	F(1,13) = 0.02, P = 0.8840	F(4,52) = 41.8, P < 0.0001	F(4,52) = 2.25, P = 0.0758	none
CaCl₂	1a	10 M, 10 F	F(1,38) = 107.4, P < 0.0001	F(3,114) = 14.9, P < 0.0001	F(3,114) = 51.1, P < 0.0001	7.5, 25, 75 mM
CaLa	1b	10 M, 10 F	F(1,38) = 17.9, P = 0.0001	F(3,114) = 5.52, P = 0.0014	F(3,114) = 7.99, P < 0.0001	7.5, 25, 75 mM
MgCl₂	5a, 6c, 7d	24 M, 32 F	F(1,54) = 0.60, P = 0.4400	F(4,216) = 31.4, P < 0.0001	F(4,216) = 0.60, P = 0.6599	none
NaCl	5c	8 F	F(1,14) = 1.48, P = 0.2434	F(5,70) = 6.26, P < 0.0001	F(5,70) = 3.30, P = 0.0099	100, 178 mM
NaLa	10c	6 M, 4 F	F(1,18) = 1.81, P = 0.1942	F(5,90) = 63.1, P < 0.0001	F(5,90) = 2.93, P = 0.0170	320 mM
KCl	7b	8 M, 8 F	F(1,30) = 3.64, P = 0.0658	F(5,150) = 11.8, P < 0.0001	F(5,150) = 0.85, P = 0.5158	none
NH₄Cl	7c	8 M, 8 F	F(1,29) = 3.01, P = 0.0932	F(5,145) = 18.3, P < 0.0001	F(5,145) = 0.30, P = 0.9103	none

Open in a new tab

Compounds are listed in the order presented in Figs. 3–5. Cohort and order provides batch of mice (1–10) and compounds tested in alphabetical order (a–d), with “a” being tested first, “b” second, and so on. Group composition gives number of male (M) and female (F) mice of each strain tested (there were usually equal numbers of BTBR and NZW mice although a mouse was occasionally excluded from data analysis because it spilled fluid). Concentrations supporting strain differences = P < 0.05 according to post hoc Fisher least significant difference (LSD) tests. Saccharin, sodium saccharin; MSG, monosodium glutamate; IMP, inosine monophosphate (umami taste); Na₃HP₂O₇, trisodium pyrophosphate; denatonium, denatonium benzoate; QHCl, quinine hydrochloride; CaLa, calcium lactate; NaLa, sodium lactate.

Table 4.

Analyses of brief-access test lick rates

			Two-way ANOVA
Compound	Cohort and Order	Group Composition	Strain	Concentration	Strain × Concentration	Concentrations Supporting Strain Differences
Parental strain (NZW vs. BTBR)
Sucrose	1c	8 F	F(1,14) = 19.1, P< 0.0001	F(4,56) = 37.0, P< 0.0001	F(4,56) = 11.5, P< 0.0001	320, 1,000 mM
Polycose	2c	4 M, 4 F	F(1,14) = 24.4, P< 0.0001	F(4,56) = 1.23, P= 0.3084	F(4,56) = 4.18, P= 0.0051	32, 68, 145, 320 mM
HCl	1e, 2f	4 M, 12 F	F(1,29) = 2.62, P= 0.1166	F(4,116) = 36.1, P< 0.0001	F(4,116) = 4.84, P= 0.0012	10, 32 mM
Citric acid	2a	4 M, 4 F	F(1,14) = 7.87, P= 0.1402	F(4,56) = 73.9, P< 0.0001	F(4,56) = 11.9, P< 0.0001	0, 10, 32, 100, 320 mM
Denatonium	1e	8 F	F(1,14) = 17.8, P< 0.0001	F(4,56) = 48.9, P< 0.0001	F(4,56) = 31.9, P< 0.0001	0, 0.32, 1.78, 10 mM
QHCl	1a	8 F	F(1,14) = 0.23, P= 0.6340	F(4,56) = 14.7, P< 0.0001	F(4,56) = 14.4, P< 0.0001	0,^* 0.032,^* 0.32, 1 mM
Caffeine	2g	4 M, 4 F	F(1,14) = 1.62, P= 0.2240	F(4,56) = 39.3, P< 0.0001	F(4,56) = 5.51, P= 0.0008	10, 32 mM
Capsaicin	2b	4 M, 4 F	F(1,13) = 0.12, P= 0.7385	F(4,52) = 34.3, P< 0.0001	F(4,52) = 1.06, P= 0.3814	none
CaCl₂	1d	8 F	F(1,14) = 4.65, P= 0.0489	F(4,56) = 24.7, P< 0.0001	F(4,56) = 12.0, P< 0.0001	320 mM
MgCl₂	2e	4 M, 4 F	F(1,14) = 1.05, P= 0.3233	F(4,56) = 52.3, P< 0.0001	F(4,56) = 2.35, P= 0.0649	none
NaCl	1b	8 F	F(1,14) = 3.25, P= 0.0929	F(4,56) = 11.6, P< 0.0001	F(4,56) = 5.10, P= 0.0014	600, 1,000 mM
KCl	2d	4 M, 4 F	F(1,13) = 4.15, P= 0.0623	F(4,52) = 28.5, P< 0.0001	F(4,52) = 2.66, P= 0.0424	150, 600 mM
Chr 17 BTBR/NZW congenic and BTBR/BTBR control mice
Sucrose	1c	5 M, 3 F	F(1,12) = 3.27, P= 0.0953	F(4,48) = 12.7, P< 0.0001	F(4,48) = 6.31, P= 0.0003	320, 1,000 mM
Polycose	2c	4 M, 5 F	F(1,12) = 6.98, P= 0.0215	F(4,48) = 2.47, P= 0.0568	F(4,48) = 3.30, P= 0.0182	145, 320 mM
HCl	1e, 2f	9 M, 8 F	F(1,29) = 0.74, P= 0.3980	F(4,116) = 27.8, P< 0.0001	F(4,116) = 1.62, P= 0.1737	none
Citric acid	2a	4 M, 5 F	F(1,14) = 2.59, P= 0.1297	F(4,56) = 26.2, P< 0.0001	F(4,56) = 0.48, P= 0.7492	none
Denatonium	1e	5 M, 3 F	F(1,11) = 11.1, P= 0.0066	F(4,56) = 16.6, P< 0.0001	F(4,56) = 4.55, P= 0.0125	0.32, 1.78, 10 mM
QHCl	1a	5 M, 3 F	F(1,14) = 6.00, P= 0.0280	F(4,56) = 5.15, P= 0.0013	F(4,56) = 1.57, P= 0.1949	0.1, 0.32, 1 mM^*
Caffeine	2e	4 M, 5 F	F(1,14) = 0.06, P= 0.8146	F(4,56) = 42.7, P< 0.0001	F(4,56) = 0.65, P= 0.6280	none
Capsaicin	2b	4 M, 5 F	F(1,13) = 0.04, P= 0.8435	F(4,52) = 36.9, P< 0.0001	F(4,52) = 0.93, P= 0.4550	none
CaCl₂	1d	5 M, 3 F	F(1,13) = 6.80, P= 0.0217	F(4,52) = 22.9, P< 0.0001	F(4,52) = 4.38, P= 0.0040	100, 320 mM
MgCl₂	2e	4 M, 5 F	F(1,13) = 22.4, P= 0.0004	F(4,52) = 54.0, P< 0.0001	F(4,52) = 5.67, P= 0.0007	32, 100, 320 mM
NaCl	1b	5 M, 3 F	F(1,14) = 2.05, P= 0.1742	F(4,56) = 15.5, P< 0.0001	F(4,56) = 4.16, P= 0.0051	1,000 mM
KCl	2d	4 M, 5 F	F(1,13) = 17.0, P= 0.0012	F(4,52) = 13.9, P< 0.0001	F(4,52) = 3.12, P= 0.0224	150, 300, 600 mM

Open in a new tab

These analyses are based on the means shown in Fig. 6. Cohort and order provides batch of mice (1st or 2nd) and compounds tested in alphabetical order (a–f), with “a” being tested first, “b” second, and so on. Group composition gives number of M and F mice of each strain; for cohort 2 there were 4 M + 4 F in the BTBR/BTBR group. Concentrations supporting strain differences = P < 0.05 according to post hoc Fisher LSD tests.

Post hoc tests are not strictly justified in this case because the interaction was not significant.

Two-bottle choice: 96 h screening tests.

All BTBR × NZW F₂ hybrids and cohorts of each of the six other groups were tested using a screen involving 96 h two-bottle choice tests. We used 96 h tests, rather than the more usual 48 h tests, because they provide significantly more sensitivity (64), which is crucial for genetic analyses involving the phenotypes of individual animals. Each mouse was weighed (to the nearest 0.1 g), and then it received a series of 10 or 11 tests. During each test, its regular water bottle was replaced with two graduated drinking tubes. For the first test, the two drinking tubes contained deionized water. Subsequent tests involved one drinking tube of deionized water and one of the following solutions in the order listed: 50 mM CaCl₂, 50 mM CaLa, 50 mM MgCl₂, 100 mM KCl, 100 mM NH₄Cl, 100 mM NaCl, 5 mM citric acid, 0.03 mM quinine hydrochloride (QHCl) and 2 mM sodium saccharin. An additional (11th) test solution, 0.1 mM QHCl was included at the end of the test series for the inbred, congenic, and knockout cohorts when it became clear that NZW controls did not avoid 0.03 mM QHCl. To reduce the possibility that consumption of a taste solution might influence consumption of the next in the series, there was a day with a single water bottle available after each test except the first one (which involved just water).

This series was used in a previous experiment (70) and was selected here for the following reasons: Our earlier strain surveys suggested that 50 mM CaCl₂ was a representative calcium solution; it was well above the threshold of detection but not completely avoided by most strains (2, 68). To assess whether group differences involving this solution were specific to calcium, to divalent chlorides, or to all mineral salts in general we tested comparable concentrations of CaLa, MgCl₂, NH₄Cl, KCl, and NaCl. To provide a general assessment of taste preferences we also tested QHCl, citric acid, and saccharin as representative bitter, sour, and sweet compounds.

At the start of each 96 h test, two graduated drinking tubes were provided. The tube on the left (closest to the cage wall) contained water and the tube on the right (closest to the food hopper) contained taste solution (see Ref. 5 or 66 for illustrations and details of drinking tube construction). The position of the tubes was switched every 24 h (to control for side preferences). The volume of fluid remaining (to the nearest 0.1 ml) was recorded at 48 and 96 h.

Fluid spillage and evaporation were ignored. Previous work with identical procedures found they account for <0.5 ml over 48 h (65, 67). Intakes during the 96 h tests were divided by four to obtain average daily intakes. Preference scores were determined as the percentage of taste solution consumed relative to total fluid consumed (i.e., water intake + taste solution intake).

Two-bottle choice: 48 h tests with ascending concentrations.

The 96 h two-bottle screen was excellent for obtaining robust taste preference phenotypes from individual animals and small groups, but it was limited in that it provided information about single concentrations of only a small set of taste compounds. To obtain a more detailed characterization of the taste preferences of inbred, congenic, and Itpr3 KO mice, we used 48 h tests with 21 taste compounds presented in ascending concentrations, which is the standard procedure for testing groups of animals.

The compounds were chosen as examples of each basic taste modality or as members of classes that could be similar to calcium (for example, mineral salts). Each mouse was tested with two to six taste compounds. The order of tests was determined partly by our interests and partly to reduce carry-over effects [e.g., sucrose and Polycose were tested separately (79)]. Concentrations were chosen with the goal of covering the range from indifference to strong avoidance (or strong preference for the sweeteners and Polycose). In most cases, concentrations increased progressively in half-log steps (e.g., 0.316, 1, 3.16, 10 mM) but 1) for NaCl and Na₃HP₂O₇, additional concentrations were tested to straddle the expected peak in preference (34, 69), and 2) for sucrose, maltose, and Polycose concentrations progressively doubled to be consistent with earlier work. Test series that produced results with high variability were repeated using additional cohorts of mice (see Tables 1–3).

Table 3.

Analyses of Itpr3 WT and KO mice two-bottle choice test preference scores

			Two-way ANOVA
Compound	Cohort and Order	Group Composition	Group	Concentration	Group × Concentration	Concentrations Supporting Strain Differences
Saccharin	5b	9 WT, 8 KO (M)	F(1,15) = 23.5, P = 0.0002	F(6,90) = 5.09, P = 0.0002	F(6,90) = 2.97, P = 0.0108	1, 3.2, 10, 32 mM
Sucrose	1b	11 WT, 11 KO (F)	F(1,20) = 15.7, P = 0.0007	F(7,140) = 45.9, P < 0.0001	F(7,140) = 3.50, P = 0.0017	1, 2, 4%
Maltose	4b	8 WT, 8 KO (F)	F(1,14) = 23.8, P = 0.0002	F(7,98) = 29.2, P < 0.0001	F(7,98) = 3.92, P = 0.0008	1, 2, 4, 8%
Polycose	3b	9 WT, 9 KO (M)	F(1,16) = 16.6, P = 0.0008	F(7,112) = 27.6, P < 0.0001	F(7,112) = 4.66, P = 0.0001	0.5, 1, 2, 4, 20%
MSG	6a, 8a	18 WT, 16 KO (F)	F(1,32) = 2.33, P = 0.1368	F(6,192) = 32.70, P < 0.0001	F(6,192) = 1.75, P = 0.1117	10, 32 mM^*
IMP	6b	9 WT, 8 KO (F)	F(1,15) = 2.69, P = 0.1214	F(8,120) = 16.4, P < 0.0001	F(8,120) = 3.10, P = 0.0033	10, 32 mM
Na₃HP₂O₇	7a	8 WT, 8 KO (M)	F(1,14) = 19.9, P = 0.0008	F(10,140) = 23.2, P < 0.0001	F(10,140) = 6.05, P < 0.0001	5.6, 10, 18, 32, 56 mM
HCl	2c	9 WT, 10 KO (M)	F(1,17) = 353.6, P < 0.0001	F(5,85)= 10.7, P < 0.0001	F(5,85) = 1.16, P = 0.3360	none
Citric acid	1d	11 WT, 11 KO (F)	F(1,19) = 4.33, P = 0.0512	F(5,95) = 18.0, P < 0.0001	F(5,95) = 0.63, P = 0.6744	none
Denatonium	1c, 7d	8 WT, 9 KO(M) 11 WT, 10 KO (F)	F(1,36) = 14.6, P = 0.0005	F(6,216) = 21.7, P < 0.0001	F(6,216) = 3.98, P = 0.0009	0.032, 0.1, 0.32, 1 mM
QHCl	2b	9 WT, 10 KO (M)	F(1,17) = 34.6, P < 0.0001	F(3,51) = 3.62, P = 0.0191	F(3,51) = 1.33, P = 0.2760	all^*
Caffeine	3c	9 WT, 9 KO (M)	F(1,16) = 0.79, P = 0.3859	F(4,64) = 27.0, P < 0.0001	F(4,64) = 2.18, P = 0.0815	none
ZnCl₂	7c	8 WT, 9 KO (F)	F(1,15) = 9.68, P = 0.0071	F(4,60) = 49.8, P < 0.0001	F(4,60) = 4.05, P = 0.0057	3.2, 10, 32 mM
Capsaicin	2d	9 WT, 10 KO (M)	F(1,17) = 5.77, P = 0.0280	F(4,68) = 22.0, P < 0.0001	F(4,68) = 0.80, P = 0.5280	none
CaCl₂	1a	11 WT, 11 KO (F)	F(1,20) = 59.3, P < 0.0001	F(4,80) = 18.6, P < 0.0001	F(4,80) = 10.3, P < 0.0001	10, 32, 100 mM
CaLa	2a	9 WT, 10 KO (M)	F(1,17) = 53.6, P < 0.0002	F(4,68) = 1.46, P = 0.2233	F(4,68) = 3.81, P = 0.0075	1,3.2, 10, 32, 100 mM
MgCl₂	3a, 8a	18 WT, 18 KO (M)	F(1,34)= 23.2, P < 0.0001	F(4,136) = 10.9, P < 0.0001	F(4,136) = 1.27, P = 0.2861	10, 32 mM^*
NaCl	4a	8 WT, 8 KO (F)	F(1,14) = 0.58, P = 0.4569	F(5,70) = 22.8, P < 0.0001	F(5,70) = 0.61, P = 0.6922	none
NaLa	5a	9 WT, 8 KO (M)	F(1,15) = 0.70, P = 0.4150	F(5,75) = 25.4, P < 0.0001	F(5,75) = 1.20, P = 0.3162	none
KCl	4d, 5c	9 WT, 8 KO (M)	F(1,31) = 7.95, P = 0.0083	F(4,124) = 18.1, P < 0.0001	F(4,124) = 3.03, P = 0.0201	100, 178 mM
		8 WT, 8 KO (F)
NH₄Cl	3d, 7b, 8b	17 WT, 17 KO (M)	F(1,48) = 6.55, P = 0.0137	F(5,240) = 26.9, P < 0.0001	F(5,240) = 2.85, P = 0.0162	3.2, 32, 100 mM
		9 WT, 9 KO (F)

Open in a new tab

Compounds are listed in the order presented in Figs. 3–5. Cohort and order provides batch of mice (1–8) and compounds tested in alphabetical order (a–d), with “a” being tested first, “b” second, and so on. Group composition gives number of wild-type (WT) and knockout (KO) mice tested; each cohort was either all M or all F. Concentrations supporting strain differences = P < 0.05 according to post hoc Fisher LSD tests.

Post hoc tests not strictly justified given the nonsignificant interaction (although a significant main effect of group). The analysis of KCl preferences omitted results of the test with 562 mM KCl because 18 of 33 mice spilled this concentration.

In each test series, the mice first received two drinking tubes containing deionized water for 48 h and then a choice between deionized water and ascending concentrations of the taste compound, with each test lasting 48 h. The positions of the two bottles were switched every 24 h. Body weights were measured at the beginning of each experiment. Daily water intakes were determined from the sum of intakes when the mice received two bottles of water at the start of each concentration series. A single value was obtained from each mouse by averaging across all water vs. water tests it received, and this was used in subsequent analyses of water intake.

We initially used analyses of variance (ANOVAs) with factors of strain and sex to compare parental strain body weights and consumption of fluids (with an additional factor of Concentration when appropriate) during two-bottle choice tests. Consistent with other results from our laboratory (e.g., Refs. 68–70), there were no sex differences in preferences for any of the taste compounds that were presented to both male and female mice. Consequently, we combined the results from both sexes for analyses of preference scores. Post hoc least significant difference (LSD) tests were used to assess differences between the groups in consumption of specific concentrations of a taste compound and to determine differences in response of each group to individual concentrations of each taste compound. One-sample t-tests were used to determine whether the preference scores of a group differed significantly from indifference (i.e., 50% preference). All analyses were conducted using a criterion for significance of P < 0.05 (Statistica 10; Stat Soft, Tulsa, OK).

Brief-access taste tests.

Long-term two-bottle choice tests provide a measure of taste solution preference, but this is not necessarily mediated by taste. To obtain a measure of taste solution acceptance that was less prone to “contamination” by postingestional effects, we used gustometers to measure licking responses during brief-exposure taste tests. The methods were modeled on those used by other groups (15, 24, 57) with differences described below. Two batches of inbred mice and two batches of congenic mice were tested, with at least eight mice in each group and the test order as specified in Table 4.

Each MS160-mouse gustometer consists of a 14.5 × 30 × 15 cm test chamber with a motorized shutter that controls access to a taste solution. Bottles of taste solution are mounted on a rack that is positioned so that any of eight different taste solutions can be presented to the mouse. The drinking spout of each bottle is part of a high-frequency alternating current contact circuit so that each lick the mouse makes is detected and recorded. Details of construction and other technical information are available elsewhere (24, 57). To avoid any undue influence of subtle differences between the gustometers we used, we ensured that each mouse was always tested in the same gustometer and that equal numbers of each group were tested in each gustometer.

The mice were weighed daily, immediately before being placed into a gustometer. To train a mouse to sample taste solutions, it was first water-deprived for 22.5 h and then placed in a gustometer with the shutter open, allowing access to the water spout. During this first training session, the mouse had continuous access to water for 25 min from the time it first licked the drinking spout. It was then returned to its home cage and given water for 1 h. On the following 2 days, this procedure was repeated, except the shutter allowing access to water was closed 5 s after each time the mouse began to lick, and it was reopened after a 7.5 s interval. Once again, after 25 min, the mouse was returned to its home cage and given water for 1 h. By the 2nd test using these procedures, all mice had learned to obtain water during the 5 s access periods.

The mice then began test sessions with various taste compounds (listed in Table 4). Only one taste compound was used during a session; it was presented in five different concentrations (including water). The deprivation regimen used to investigate the response to sucrose and Polycose differed from that used to investigate the response to the other taste compounds because the mice need less deprivation to consume the hedonically positive tastes (24). Prior to a session with sucrose or Polycose, each mouse received free access to food and water for 24 h. It then received 1 g of food and 2 ml of water, and the session began 24 h later. After these sessions, the mouse had a recovery day with free access to food and water for 24 h. Its water was then removed for 22.5 h to prepare it for the next session. After sessions with hedonically negative compounds, each mouse had ad libitum access to food but received water for only 1 h in its home cage; it was then water-deprived for ∼22.5 h in preparation for the next session.

When the two hedonically positive compounds were being investigated, the session began with a single test of the highest concentration available to kindle the mouse's interest in the drinking spout. After this, repeated series of five concentrations (including water) were presented in a quasirandom order such that a concentration could appear only once in a series of five tests. For each exposure, the shutter was open for 5 s, during which licks of the drinking spout were counted. This was followed by 7.5 s with the shutter closed, during which a new taste solution was positioned ready for the next presentation. For sessions with hedonically negative compounds, repeated series of five concentrations (including water) were presented in randomized order. Additional 1 s “washout” trials with water were interposed between each test trial. Thus, a mouse received access to a hedonically negative taste solution for 5 s followed by 7.5 s with the shutter closed, then access to water for 1 s followed by 7.5 s with the shutter closed, followed by the next taste solution for 5 s, and so on. We think the 1 s washout trials with water have the effect of cleansing the mouse's palate and help prevent it from quitting because it expects only bad-tasting solutions. All test sessions consisted of 180 spout presentations (90 of which were washout tests for the hedonically negative compounds) although all mice stopped drinking well before the session ended.

For statistical analyses, we obtained the mean number of licks made by each mouse in response to each concentration of each taste compound by averaging together the results from identical exposures. Mice that did not respond to any of the 18 presentations of a particular concentration of a taste compound were not included in statistical analyses of that compound. Analyses involved mixed-design two-way ANOVAs with factors of strain and concentration, followed by post hoc LSD tests to assess strain differences in lick rates for specific concentrations of taste solution and to assess differences in response of each strain to individual concentrations of each taste compound relative to water lick rates.

Generation and Testing of BTBR × NZW F₂ Hybrids

Mating scheme.

Ten male and 10 female mice from the BTBR and NZW strains were bred in our facility to produce BTBR × NZW F₁ mice, and these were mated brother-to-sister to produce F₂ mice. A total of 610 F₂ hybrids (309 male, 301 female) were bred and successfully tested.

Phenotyping.

Phenotyping was conducted using the 96 h two-bottle choice tests developed for screening (see above). All 610 F₂ mice were tested (in 10 batches of 17–129 mice), but data from 26 out of 5,983 tests were lost due to spilled drinking tubes or other technical problems. Consequently, analyses were based on the results of 606–610 mice.

DNA extraction and genotyping.

Genomic DNA was extracted and purified from mouse tails by a sodium hydroxide method (73). Genotyping was conducted by The Center for Inherited Disease Research (CIDR, Johns Hopkins University). There were 626 informative (polymorphic) markers, with the average distance between markers of 4 Mb (2.8 cM). DNA samples purchased from The Jackson Laboratory as well as parental and F₁ DNA from our laboratory were included as controls in the genotyping analysis. As a second quality control, blind duplicate DNA samples were genotyped. After the typing was completed, the code was broken and duplicate samples were matched and the data compared, with 99.996% agreement among duplicates (50,016 of 50,018 paired genotypes).

Linkage analysis.

Genome-wide screens of the F₂ mice were conducted using markers from all autosomes. Linkage between individual traits and genotypes was assessed using algorithms implemented by the R/qtl 1.04–53 package of R (10). Genotype probabilities and genotyping errors were estimated using the “calc.genoprob” function. Interval mapping by maximum likelihood estimation was conducted to screen for main-effect QTLs using the “scanone” function (normal model, EM scan method). The significance of each marker regression result was established from 1,000 permutations of the observed data using the “n.perm” function.

Because several of the frequency distributions were skewed (see Frequency distributions, below), in initial analyses we compared the interval maps obtained with data that were untransformed to those obtained with data normalized by log or arcsin transformation. The transformations produced very little difference in the resulting interval maps so, for simplicity, we present the untransformed results here. We also looked for sex-related effects. We used R/qtl to calculate logarithm of odds (LOD) scores under Model_additive and Model_sex interaction (56) and the difference in LOD score was calculated using the “arithscan” function. To quantify sex-by-genotype interactions, we compared the fit of the two models using the difference in LOD scores as a metric (ΔLOD). With a ΔLOD ≥ 3.1 as a criterion (21), there were no sex-related effects in any analysis so sex was not included as a factor in the results reported here.

Breeding and Testing of BTBR.NZW-Drinksac6 Congenic Mice (BTBR/NZW) and BTBR/BTBR Controls

Our linkage analyses revealed a locus influencing taste preferences on Chr 17 (see below), so to isolate this we produced a congenic line. BTBR.NZW F₁ mice were backcrossed to the BTBR strain. The offspring were phenotyped, starting when 6–8 wk old, with 1) a 96 h choice between two bottles of water, 2) a 96 h choice between water and 2 mM saccharin solution, and 3) a 96 h choice between water and 50 mM CaCl₂ solution. The position of the bottles was switched every 24 h, and intakes were recorded every 48 h. Mice with felicitous recombinations and appropriate phenotypes were used as parents for subsequent generations, which were produced by backcrossing to BTBR mice. As backcrossing progressed, the panel of single nucleotide polymorphism (SNP) markers was extended to more precisely localize the regions where recombination occurred. Recombinations that narrowed the congenic interval occurred in the 4th, 6th, 8th, and 10th backcross generations, so these mice were used to start new lines, and once successful breeding was established, lines with larger congenic intervals were abandoned. In the 11th backcross generation, the congenic interval supporting the phenotypes involved a 0.8 Mb (max) region on Chr 17 bounded by a recombination between rs47196150 and rs33434357 (26.66–26.70 Mb) proximally, and rs33071006 and rs3656446 (27.26–27.48 Mb) distally. This region contains 14 known and seven unknown (predicted) genes (Table 5) The phenotypic values for these mice (n = 14) relative to littermates homozygous for BTBR in this region (n = 24) were as follows: total water intake, controls = 5.7 ± 0.2 ml/day, congenics = 4.7 ± 0.3 ml/day; 2 mM saccharin preference, controls = 45 ± 3%, congenics = 76 ± 4%; 50 mM CaCl₂ preference, controls = 44 ± 2%, congenics = 20 ± 5%. Thus, the phenotypes were preserved.

Table 5.

List of known genes in the congenic interval between rs47196150 and rs3656446 on Chr 17

Start	End	Gene	Name
26698471	26792295	Ergic1	endoplasmic reticulum-golgi intermediate compartment 1
26813363	26836589	Atp6v0e	ATPase, H⁺ transporting, lysosomal V0 subunit E
26918024	26929510	Bnip1	BCL2/adenovirus E1B interacting protein 1
26975610	26978510	Nkx2-5	NK2 transcription factor related, locus 5
27054036	27069524	Kifc1	kinesin family member C1
27070072	27074835	Phf1	PHD finger protein 1
27074918	27076423	Cuta	cutA divalent cation tolerance homolog
27094363	27107590	Syngap1	synaptic Ras GTPase activating protein 1
27110112	27113150	Zbtb9	zinc finger and BTB domain containing 9
27156755	27165954	Bak1	BCL2-antagonist/killer 1
27110162	27173323	Ggnbp1	gametogenetin binding protein 1
27194249	27259168	Itpr3	inositol 1,4,5-triphosphate receptor 3
27280914	27304709	Ip6k3	inositol hexaphosphate kinase 3
27326702	27340693	Lemd2	LEM domain containing 2

Open in a new tab

Start and end refers to bp position on Chr 17 (NCBI build 37). Table does not include 7 predicted genes in the congenic interval.

According to standard nomenclature (29), this new congenic strain is described based on the captured QTL as BTBR.NZW-Drinksac6 or, based on the physical limits of the congenic interval as BTBR.NZW-(rs47196150-rs3656446)/Mon. Mice of the 11th generation were backcrossed to produce sufficient N₁₂ mice to conduct a detailed investigation of the taste preferences associated with the congenic line (described above). For brevity, we refer to the mice as “BTBR/NZW congenics” and “BTBR/BTBR controls” here. Note that the “control” mice (carrying BTBR/BTBR alleles in the congenic interval) show the abnormal phenotype; the congenic mice (carrying BTBR/NZW alleles) have preferences akin to those of the NZW strain, which are similar to those observed in most other strains.

Production of Itpr3 KO Mice

Our studies with Chr 17 congenic mice led us to engineer mice with a KO of a gene, Itpr3, in the congenic interval. To do this, we purchased C57BL/6 ES cell clones from the North American Conditional Mouse Mutagenesis project (NorCOMM, ES cell line MFGC N01293P1_W239C5). The clones involve deletion of 300 bp spanning Exon 2 of Itpr3 (Chr 17: 27194249–272194548) and were verified and isolated by the University of Pennsylvania Gene Targeting Service (Dr. Tobias Raabe). They were injected into BALB/c blastocysts at the Transgenic and Chimeric Mouse Facility at the University of Pennsylvania (Dr. Jean Richa). The first batch of clones resulted in 12 male and six female offspring, which ranged between 10 and 95% chimeric (judged by coat color: white = BALB/c, black = B6). Males with the greatest chimerism (dark coat color) were mated with C57BL/6J females; the offspring with all black coats were mated brother-to-sister to produce homozygous KOs. Germ line transmission was verified by genotyping for the presence of Lac Z and Neo vectors (in-house and Transnetyx).

Sequencing the NZW and BTBR Forms of Itpr3

According to the reference sequence based on the C57BL/6NJ mouse strain, Itpr3 has 58 exons (http://www.ensembl.org/Mus_musculus/Transcript/Exons?db=core;g=ENSMUSG00000042644;r=17:27057304–27122223;t=ENSMUST00000049308). We sequenced all 58 exons of Itpr3 from BTBR, NZW, and C57BL/6J mice. Genomic DNA purchased from the Jackson Laboratory was diluted to 10 ng/μl, and primers flanking each exon were designed using Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) or Primer3 (frodo.wi.mit.edu/). The DNA was amplified by PCR using HotMaster Taq DNA Polymerase (5 PRIME) under the recommended conditions, and PCR products were visualized on 1% agarose gels (90 V, 30 min). DNA bands of the appropriate sizes were excised from the gel and purified using the QIAquick Gel Extraction Kit (QIAGEN). Purified DNA and primers were analyzed using Sanger Sequencing (ABI 3730) by the University of Pennsylvania DNA Sequencing Facility. The resulting sequences were aligned and analyzed using Sequencher 5.0 software.

Exon 20 proved intransigent to sequencing (i.e., none of the purified DNA from this region returned sequences) so it was approached more intensively. We first selected eight primer pairs spanning the ∼2,000 bases of this exon. Sequencing of seven of these segments was successful. To sequence the remaining stubborn 200 bp region, a primer walk was conducted by taking 5 bp steps starting upstream of the closest working primer distal to exon 20. All of the primers used in the primer walk produced bands, which were purified and sequenced as described above.

The exonic sequences of NZW and C57BL/6J forms of Itpr3 were identical to the C57BL/6NJ reference sequence, except for five synonymous SNPs. The BTBR form of Itpr3 was also identical to the reference sequence except for a 12 bp deletion in Exon 23 (Chr 17: 27238069, National Center for Biotechnology Information Build 7), which codes for amino acids 983–986. To confirm the sequencing results, DNA from BTBR and NZW mice was amplified by PCR using primers that closely flanked the deletion. PCR products were then visualized on a 4% agarose gel (90 V, 90 min). This revealed bands that were 12 bp smaller in the BTBR than NZW DNA. The 12 bp deletion appeared to be unique to the BTBR strain in that it was not present in any of the 18 strains sequenced by the Sanger Mouse Genomes Project (http://www.sanger.ac.uk/resources/mouse/genomes).

A SNP (rs108851476) between B6 and BALB/c mice in the 335th amino acid of Itpr3 is responsible for a Pro-Leu polymorphism that influences receptor function and intracellular calcium signaling (30). However, this SNP is not polymorphic between the BTBR and NZW strains so cannot be responsible for the differences observed here.

RESULTS

Comparison of Parental Strains

Body weight.

Male BTBR mice were slightly but significantly heavier than were male NZW mice; there was no difference in the body weights of females of the two strains. For example, at the start of 96 h choice tests, when mice were aged 43–60 days (average 53 ± 0.8 days), the four sex-strain groups weighed (means ± SE, g): BTBR male = 28.5 ± 0.5, NZW male = 25.2 ± 0.7, BTBR female = 24.0 ± 0.6, NZW female = 23.0 ± 0.3 [strain × sex interaction, F(1,60) = 9.13, P = 0.0037].

Water intake.

The BTBR mice drank significantly more water than did the NZW mice [Fig. 1; F(1,174) = 8.87, P = 0.0033] with no difference between the sexes and no strain × sex interaction [means ± SE (n), ml/day; BTBR male = 4.9 ± 0.2 (45), NZW male = 4.1 ± 0.2 (46), BTBR female = 5.3 ± 0.2 (43), NZW female = 4.6 ± 0.2 (44)].

TWO-BOTTLE CHOICE TESTS.

In 96 h two-bottle choice tests, the BTBR mice had significantly higher preference scores than did the NZW mice for 50 mM CaCl₂ and 50 mM CaLa; they had significantly lower preference scores for 2 mM saccharin and 5 mM citric acid. Preference scores of the NZW strain for all solutions were significantly different from 50% (i.e., indifference). In contrast, the BTBR strain's response to 50 mM CaCl₂, 50 mM CaLa, 50 mM MgCl₂, 100 mM KCl, 2 mM saccharin, and 0.1 mM QHCl did not differ from 50% (Fig. 2).

Fig. 2. — Results of 4 experiments yielding preference scores for 9 taste solutions in 4-day 2-bottle choice tests. Inbred = NZW and BTBR parental strains; F₂ = BTBR × NZW F₂ hybrid mice categorized by alleles at rs3693494, a marker on chromosome 17; Chr 17 = chromosome 17 BTBR/NZW congenic mice and BTBR/BTBR controls; KO = *Itpr3* WT and KO mice (on a C57BL/6 background). Legend refers to alleles of *Itpr3* each group possesses.

In 48 h choice tests, relative to NZW mice, BTBR mice had significantly higher preferences for all three concentrations of CaCl₂ and CaLa, 0.32–3.2 mM denatonium, 0.032 mM QHCl, and 1–10 mM caffeine. The BTBR mice had significantly lower preferences for 100 and 178 mM NaCl, 3.2–56 mM Na₃HP₂O₇, 3.2–32 mM inosine monophosphate (IMP), 3.2–32 mM saccharin, and 4 and 8% sucrose. The BTBR stain also had lower preferences for HCl overall and tended to have lower preferences for citric acid (Figs. 3–5 and Table 1).

Fig. 3. — Two-bottle choice preference scores of inbred (*left*), congenic (*middle*), and *Itpr3* KO (*right*) mice. Horizontal dotted lines show indifference (50% preference). Results of statistical analyses are given in Table 1.

Fig. 4. — Sour, bitter, and other 2-bottle choice preference scores of inbred (*left*), congenic (*middle*), and *Itpr3* KO (*right*) mice. Horizontal dotted lines show indifference (50% preference). Results of statistical analyses are given in Table 2.

Fig. 5. — Mineral salt 2-bottle choice preference scores of inbred (*left*), congenic (*middle*), and Itpr3 KO (*right*) mice. Horizontal dotted lines show indifference (50% preference). Results of statistical analyses are given in Table 3.

Relative to water, the NZW strain preferred at least one concentration of saccharin, sucrose, Polycose, IMP, and Na₃HP₂O₇ (Table 6). The BTBR strain also preferred some concentrations of saccharin, sucrose, Polycose, and Na₃HP₂O₇, and also denatonium, caffeine, and the two calcium salts. Relative to the NZW strain, the BTBR strain had lower thresholds for behavioral avoidance of Na₃HP₂O₇, HCl, and NaCl; it had higher thresholds for denatonium, QHCl, caffeine, MgCl₂, and KCl, and unlike the NZW strain, it did not avoid any concentration of CaCl₂ or CaLa (Figs. 3–5, Table 6).

Table 6.

Concentrations of each taste compound that produced a significant change in preference scores relative to indifference (50%)

	Avidity (Significantly >50%)						Avoidance (Significantly <50%)
Taste Compound	NZW	BTBR	BTBR/NZW Congenic	BTBR/BTBR Control	Itpr3 WT	Itpr3 KO	NZW	BTBR	BTBR/NZW Congenic	BTBR/BTBR Control	Itpr3 WT	Itpr3 KO
Saccharin	1+	10	1+	−	0.32+	−	−	−	−	−	−	−
Sucrose	4%+	8%+	1%+	16%+	1%+	8%+	−	−	−	−	−	−
Maltose	0.5%+	4%+	0.5%+	4%+	1%+	4%+	−	−	−	−	−	−
Polycose	1%+	2%+	0.5%+	16%+	0.5%+	4%+	−	−	−	0.5–2%	−	1%
MSG	10–178	32	10–320	10–100	10–178	100–178	562	178+	562	562	316+	316+
IMP	3.2–32	−	1–100	1–100	1–32	−	100	100	100+	100+	316	316
Na₃HP₂O₇	3.2–32	10	1–10	1,10	5.6–56	−	100	56+	56+	32+	100+	56+
HCl	−	−	0.32	−	0.1–3.2	0.32–3.2	3.2+	1+	3.2+	1+	10	10
Citric acid	−	−	0.32, 3.2	0.1–0.32	0.1–0.32	0.1–0.32	0.1, 3.2+	0.1+	10	10	10	10
Denatonium	−	0.01	−	0.01, 0.32–1	−	0.03–0.32	0.32+	3.16	0.32+	−	0.32+	3.2
QHCl	−	−	−	−	−	0.1	0.01+	0.1	0.032+	−	0.01+	−
Caffeine	−	1–3.2	−	−	3.16	−	10+	32	10	10	1, 10+	10+
ZnCl₂	−	−	−	−	3.2–10	−	3.2+	3.2+	10+	10+	100	32+
Capsaicin	−	−	−	−	−	−	0.001+	0.001+	0.001+	0.001+	0.001+	0.003+
CaCl₂	−	7.5–25	−	10–32	−	10–32	7.5+	−	32+	100	10+	100
CaLa	−	7.5–75	−	10–32	−	3.2+	7.5+	−	10+	−	3.2+	−
MgCl₂	−	−	−	−	−	−	10+	100	100	100	10+	100
NaCl	−	−	32–178	32–178	−	−	316+	178	316+	316+	316+	178+
NaLa	32–178	32–178	32–100	−	32–178	100–178	562	320+	320+	178+	562	562
KCl	−	−	32–178	−	32–178	−	10, 100+	316	316+	178+	316+	316+
NH₄Cl	−	−	3.2–100	32	−	32	316	316	316	316	316	316

Open in a new tab

Values are determined from two-bottle choice experiments, according to one-sample t-tests. Values are concentrations in millimoles or percent (%). −, No concentration was preferred (left columns) or avoided (right columns); +, all higher concentrations tested also influenced preferences.

BRIEF-ACCESS TESTS.

The NZW and BTBR mice licked water at similar rates, with the exception of the test involving QHCl, where the NZWs licked more than did the BTBRs (Fig. 6), perhaps due to successive contrast effects in the NZWs (see Ref. 25). The NZW mice had lick rates that would be expected of members of a sweet “taster” strain (28, 43). Relative to licking water, they avidly licked high concentrations of sucrose and Polycose and reluctantly licked high concentrations of denatonium, QHCl, capsaicin, CaCl₂, and NaCl [similar to C57BL6/J or various wild-type mice (15, 24, 26, 53)]. The BTBR mice showed a different response pattern: They licked all concentrations of sucrose, Polycose, QHCl, and NaCl at rates that were statistically similar to water. They also showed significantly less reduction of licking than did the NZW mice to high concentrations of citric acid, denatonium, and CaCl_2. The two strains had similar capsaicin concentration-response functions (Fig. 6, Table 4).

Fig. 6. — Lick rates of NZW and BTBR inbred mice (*left* panel of each pair) and Chr 17 BTBR/NZW congenic and BTBR/BTBR control mice (*right* panel of each pair) given 5 s to drink various concentrations of 12 taste solutions. *P < 0.05 relative to other group according to Fisher LSD tests. Results of statistical analyses are given in Table 4. Polycose concentrations are in mM based on a molecular weight of 1,000 g.

F₂ Segregating Hybrid Mice

Sex and reciprocal cross effects.

We investigated the contribution of sex and parentage to the body weight and taste phenotypes of BTBR × NZW F₂ mice by three-way ANOVAs with factors of sex, strain of maternal grandmother (F₀), and strain of paternal grandmother. For these tests, we used a criterion for significance of P < 0.01, to provide some protection from errors due to the multiple tests involved.

Males were significantly heavier than females [body weight (g) at start of testing males, 32.4 ± 0.2 g (n = 306), females = 25.5 ± 0.2 g (n = 296)]. However, females drank significantly more water than did males (males = 5.6 ± 0.1 ml/day, females = 6.7 ± 0.1 ml/day), and this was true also of intakes of most taste solutions (MgCl₂, KCl, NaCl, citric acid, QHCl, and saccharin; see Table 7). In general, females drank more water than did males in two-bottle choice tests, so the net result was for males and females to have similar preferences. Preference scores for CaLa, NH₄Cl, and QHCl were significantly higher in males than females, and preference scores for KCl were significantly higher in females than males. However, these effects were tiny (a difference of 3–4%) except for NH₄Cl preference scores, where males had markedly higher values than did females [43 ± 0% vs. 34 ± 0%, Table 7, similar to results we obtained with a B6 × PWK F₂ cross (70)].

Table 7.

Taste solution intakes and preferences of BTBR × NZW F₂ mice, arranged by sex

	Taste Solution Intakes, ml/day		Preference Scores, %
Taste Compound	M	F	M	F
Water	5.6 ± 0.1	6.7 ± 0.1^‡
50 mM CaCl₂	1.8 ± 0.1	2.1 ± 0.1	31 ± 1	28 ± 1
50 mM CaLa	2.1 ± 0.1	2.3 ± 0.1	39 ± 1	35 ± 1^†
50 mM MgCl₂	1.8 ± 0.1	2.1 ± 0.1^†	36 ± 1	32 ± 1
100 mM KCl	2.6 ± 0.1	3.6 ± 0.1^‡	47 ± 0	50 ± 0^†
100 mM NH₄Cl	2.2 ± 0.1	2.3 ± 0.1	43 ± 0	34 ± 0^‡
100 mM NaCl	4.7 ± 0.1	6.9 ± 0.2^‡	76 ± 0	76 ± 0
5 mM citric acid	1.6 ± 0.0	2.0 ± 0.1^‡	34 ± 0	31 ± 1
30 μM QHCl	1.5 ± 0.0	1.9 ± 0.1^‡	36 ± 0	32 ± 1^‡
2 mM saccharin	4.5 ± 0.1	6.5 ± 0.2^‡	76 ± 0	78 ± 0

Open in a new tab

M/F differences assessed from 3-way ANOVAs main effect, with 1 and approximately 595 df:

^†

P < 0.005,

^‡

P < 0.001, Values are means ± SE, of 309 M and 301 F (slightly fewer if a mouse spilled). Body weights at start of testing were M = 32 ± 0.2 g, F = 26 ± 0.2 g.

There were some complex effects related to 1) the strain of paternal grandmother interacting with the strain of maternal grandmother influencing citric acid intake (Table 8), 2) the sex of a mouse interacting with the strain of its parental grandmother affecting KCl intake, saccharin intake, and saccharin preference (Table 9), and 3) the strain of the paternal grandmother affecting QHCl intake [BTBR paternal grandmother = 1.9 ± 0.1 ml/day (n = 208), NZW paternal grandmother = 1.6 ± 0.1 ml/day(n = 395), F(1,595) = 8.06, P = 0.0047].

Table 8.

Results of analyses of reciprocal cross (maternal and paternal strain effects) on taste solution intakes and preferences of BTBR × NZW F₂ mice: influence of strain of maternal and paternal grandmother on citric acid intake

Maternal Grandmother	Paternal Grandmother	Group Size, n	Citric Acid Intake, ml/day
BTBR	BTBR	152	1.8 ± 0.1
BTBR	NZW	105	1.9 ± 0.1
NZW	BTBR	56	2.2 ± 0.1
NZW	NZW	288	1.6 ± 0.1

Open in a new tab

Interaction, F(1,593) = 7.89, P = 0.0051.

Table 9.

Results of analyses of reciprocal cross (maternal and paternal strain effects) on taste solution intakes and preferences of BTBR × NZW F₂ mice: interactions of strain of paternal grandmother with sex

Sex	Paternal Grandmother	Group Size, n	KCl Intake, ml/day	Saccharin Intake, ml/day	Saccharin Preference Score, %
F	BTBR	91	4.2 ± 0.2	7.3 ± 0.3	82 ± 2
F	NZW	205	3.4 ± 0.1	6.0 ± 0.2	77 ± 1
M	BTBR	116	2.6 ± 0.2	4.5 ± 0.2	75 ± 2
M	NZW	191	2.6 ± 0.1	4.8 ± 0.2	79 ± 1
Interaction		F(1,595)	7.08, P= 0.0080	11.0, P= 0.0010	7.78, P= 0.0055

Open in a new tab

Frequency distributions.

To visualize the shape of the distribution and test for normality, the scores for each trait were assigned to 25 equal-sized bins of 4% preference each. Separate distributions for solution intake and preference for each sex were plotted, but, for brevity, only the combined preference distributions of both sexes are shown here (Fig. 7). Normality was assessed by Lilliefors tests. All distributions deviated significantly from normality, except for NH₄Cl preferences (Fig. 7).

Fig. 7. — Distributions of preference scores of ∼610 BTBR × NZW F₂ mice.

Interval Mapping

Locus on Chr 17.

The most remarkable results involved a set of 14 linkages to Chr 17 (Table 10, Fig. 8). Highly significant linkages were present with peaks at one of two adjacent markers (either rs3672065 or rs3693494 at Chr 17: 14.3 or 29.9 Mb). These included intakes of water and seven taste solutions and preferences for six taste solutions (Table 10). Extraordinarily strong linkages were present for saccharin preference (LOD = 100.8, accounting for 31% of the phenotypic variance) and the intake and preference of CaCl₂ and CaLa (LODs > 34.0, accounting for >17% of the phenotypic variance; Table 10). In every case, mice with the NZW/NZW and BTBR/NZW genotypes at this locus had statistically identical phenotypes, which differed from mice with the BTBR/BTBR genotype, implying dominance of the NZW allele. The action of this allele was to decrease intake of water, CaCl₂, CaLa, MgCl₂, NH₄Cl, and NaCl but increase intake of saccharin. It decreased preference scores for CaCl₂ and CaLa but increased preference scores for NaCl, citric acid, saccharin, and Na₃HP₂O₇. The apparent contradiction of the same allele decreasing NaCl intake but increasing NaCl preference can be explained by consideration of the effect of the NZW allele on water intakes (Fig. 9).

Table 10.

QTLs on Chr 17 influencing water and taste solution intake and preference, with means ± SEs for each genotype

		Intake, ml/day					Preference Score, %
QTL ID	Fluid Ingested	LOD Score	% Variance	BTBR/BTBR (n = 134)	BTBR/NZW (n = 314)	NZW/NZW (n = 162)	LOD Score	% Variance	BTBR/BTBR (n = 134)	BTBR/NZW (n = 314)	NZW/NZW (n = 162)
Drinkwater2	water	15.9	9.2	7.4 ± 0.2	5.9 ± 0.1^a	5.5 ± 0.1^a
Drinkcacl25	CaCl₂	56.7	26.8	3.5 ± 0.1	1.6 ± 0.1^a	1.3 ± 0.1^a	44.6	22.6	48 ± 2	27 ± 1^a	22 ± 1^a
Drinkcala4	CaLa	58.6	25.2	3.9 ± 0.1	1.8 ± 0.1^a	1.7 ± 0.1^a	34.3	17.2	55 ± 2	34 ± 1^a	31 ± 1^a
Drinkmgcl26	MgCl₂	13.0	8.3	2.6 ± 01	1.9 ± 0.1^a	1.7 ± 0.1^a	2.5	1.8	38 ± 1	35 ± 1	31 ± 1^a
Drinkkcl6	KCl	4.5	2.5	3.5 ± 0.1	3.1 ± 0.1^a	2.8 ± 0.1^a	1.0	0.2	48 ± 1	50 ± 1	47 ± 1
Drinknh4cl3	NH₄Cl	6.5	3.7	2.7 ± 0.1	2.2 ± 0.1^a	2.0 ± 0.1^a	0.1	0.1	39 ± 2	39 ± 1	38 ± 1
Drinknacl3	NaCl	5.5	3.6	6.8 ± 0.3	5.7 ± 0.1^a	5.3 ± 0.2^a	3.8	1.6	73 ± 1	78 ± 1^a	78 ± 1^a
Drinkcitric1	citric acid	1.5	1.1	1.6 ± 0.1	1.8 ± 0.1	1.9 ± 0.1	10.5	6.4	24 ± 1	34 ± 1^a	37 ± 1^a
Drinkqhcl3	QHCl	1.8	1.2	1.9 ± 0.1	1.7 ± 0.1	1.6 ± 0.1	1.2	0.3	32 ± 1	36 ± 1	35 ± 1
Drinksac6	saccharin	9.0	3.6	4.2 ± 0.2	5.9 ± 0.2^a	5.7 ± 0.2^a	100.8	31.4	57 ± 1	83 ± 1^a	84 ± 1^a
Drinkpyro1	Na₃HP₂O₇	0.1	0.0	4.4 ± 0.2	4.5 ± 0.1	4.5 ± 0.2	33.2	19.4	60 ± 2	78 ± 1^a	81 ± 1^a

Open in a new tab

The peak of all quantitative trait loci (QTLs) was at rs3693494 (29.9 Mb), or the adjacent marker rs3672065 (14.3 Mb) for KCl intake, NH₄Cl intake and preference, and NaCl preference. Logarithm of odds (LOD) scores in boldface are significant (P < 0.01). Group sizes (n) were slightly smaller for some tests because of measurement errors or missing genotypes.

P < 0.01 different from mice with BTBR/BTBR genotype. The percentage of phenotypic variance (% variance) by variation in the locus was calculated as r² from the correlation of number of NZW alleles (0, 1, or 2) with phenotype value.

Fig. 8. — Interval maps for intake and preference of water and taste solutions by 610 BTBR × NZW F₂ mice. Faint horizontal lines are genome-wide significance levels for P < 0.05 (*lower* line) and P < 0.01 (*upper* line) based on preference scores. Note that the y-axes for CaCl₂, CaLa, saccharin, and trisodium pyrophosphate (Na₃HP₂O₇) are broken to accommodate the very high peaks on Chr 17.

Fig. 9. — Stacked bar graphs showing contribution of taste solution and water to total daily fluid intake of BTBR × NZW F₂ mice arranged by genotype at *rs3693494* on Chr 17 (29.9 Mb). Percentages above bars give solution preference scores, which are plotted in Fig. 2. NN = NZW/NZW, NR = BTBR/NZW, RR = BTBR/BTBR.

Although they had marked differences in water intakes, there was no difference among mice with the three rs3672065 genotypes in body weight (e.g., at the end of behavioral tests, BTBR/BTBR = 29.9 ± 0.6 g, BTBR/NZW = 32.1 ± 0.5 g, NZW/NZW = 29.9 ± 0.6 g).

Other loci.

A cluster involving 13 linkages, with peaks at eight markers, was present on Chr 8. The most pronounced effects in this region were on preferences for CaCl₂, CaLa, and MgCl₂. The QTL peaks for these three taste solution preference scores did not coincide exactly, and there were two distinct QTL peaks for CaLa and three distinct peaks for MgCl₂ (Fig. 8). On the other hand, confidence intervals of the QTLs (1.0 LOD drops from the peak) overlapped, and the mode of inheritance for each of these peaks was additive, with the NZW allele increasing preference scores. Thus, there may be a single locus influencing intake of calcium and magnesium salts or several closely spaced loci with similar functions. There was no evidence that this locus interacted with the major locus on Chr 17 (see Table 11). This region also had smaller, albeit significant, linkages to citric acid and QHCl preference, each with the same additive mode of inheritance as the more prominent linkages involving calcium and magnesium salts. This also supports the conclusion that there are several closely linked loci in this region.

Table 11.

Independent effects on 50 mM CaCl₂ preference scores of loci on chromosome 8 and 17 in BTBR × NZW F₂ cross

	Genotype on Chr 8, rs13479776
Genotype on Chr 17, rs3693494	NZW/NZW	BTBR/NZW	BTBR/BTBR	Combined
NZW/NZW	23 ± 2 (48)	24 ± 1 (75)	17 ± 2 (39)	22 ± 1 (162)^a
BTBR/NZW	30 ± 1 (89)	27 ± 1 (152)	21 ± 2 (73)	27 ± 1 (314)^b
BTBR/BTBR	57 ± 4 (34)	46 ± 2 (76)	45 ± 3 (23)	49 ± 2 (133)^c
Combined	34 ± 2 (171)^a	31 ± 1 (303)^a	24 ± 1 (135)^b	30 ± 1 (609)

Open in a new tab

Values are means ± SE (n) of CaCl₂ preference scores, %. Similar differences were obtained with preferences for CaLa and MgCl₂. Superscripts show differences between marginal means according to Fisher LSD post hoc tests. Effect of Chr 8: F(2,600) = 11.8, P < 0.0001. Effect of Chr 17: F(2,600) = 116.9, P < 0.0001; Chr 8 × Chr 17 interaction: F(4,600) = 1.94, P = 0.1016 (not significant).

Most of the remaining linkages had an LOD score of <5.0 and accounted for <3.5% of the phenotypic variance. The exceptions were linkages to saccharin intake on Chr 7 (at 64.8 Mb, LOD = 6.26), and both saccharin and water intake on distal Chr 11 (at ∼100 Mb; LOD = 6.17 and 5.50, respectively; Table 12). Of the 36 linkages in total, the NZW genotype provided the plus allele for 26, the BTBR was the plus allele for eight, and two displayed heterosis (i.e., the heterozygous mice did not fall between the two homozygous groups).

Table 12.

Other QTLs influencing water intake and taste solution intake and preference, with means ± SEs for each genotype

QTL ID	Chr	Mb	Marker	Fluid	Measure	LOD Score	% Variance	BTBR/BTBR (n = 134)	BTBR/NZW (n = 314)	NZW/NZW (n = 162)
Drinkqhcl4	1	66.7	rs13475902	QHCl	intake	3.76	2.2	1.6 ± 0.1	1.6 ± 0.1	1.9 ± 0.1^a^b
Drinkcitric2	1	118.5	rs3695581	citric acid	intake	4.14	3.1	1.5 ± 0.1	1.8 ± 0.1	2.0 ± 0.1^a
					preference	3.10	2.2	29 ± 1	33 ± 1^a	36 ± 1^a^b
Drinkcacl26	2	73.2	CEL-2_73370728	CaCl₂	intake	3.26	0.5	2.3 ± 0.1	1.8 ± 0.1^a	1.8 ± 0.1^a
Drinkwater3	2	77.4	gnf02.076.311^*	water	intake	4.06	2.9	6.6 ± 0.2	6.1 ± 0.1^a	5.6 ± 0.2^a^b
Drinkkcl7	2	77.4	gnf02.076.311^*	KCl	intake	4.67	3.4	3.5 ± 0.1	3.1 ± 0.1^a	2.6 ± 0.1^a^b
Drinkwater4	6	132.1	rs3722480	water	intake	3.32	0.2	5.6 ± 0.1	6.4 ± 0.1^a	5.9 ± 0.2^b
Drinkqhcl5	7	31.3	rs13479191	QHCl	preference	4.33	0.7	34 ± 1	37 ± 1^a	30 ± 1^a^b
Drinksac7	7	31.3	rs13479191	saccharin	intake	4.31	2.8	5.0 ± 0.2	5.3 ± 0.2	6.3 ± 0.2^a^b
Drinksac8	7	51.8	mCV23672419	saccharin	intake	6.26	3.8	5.0 ± 0.2	5.3 ± 0.1	6.5 ± 0.3^a^b
Drinkmgcl27	8	30.1	rs4227096	MgCl₂	preference	6.09	4.4	30 ± 2	34 ± 1^a	40 ± 1^a^b
Drinkcala5	8	51.4	CEL-8_51607005	CaLa	preference	7.76	5.1	29 ± 2	39 ± 1^a	42 ± 1^a
Drinkcitric3	8	58.0	rs3707439	citric acid	preference	3.30	2.3	29 ± 1	32 ± 1^a	37 ± 1^a^b
Drinkcacl27	8	60.3	rs3672639^*	CaCl₂	intake	3.14	1.3	1.5 ± 0.1	2.1 ± 0.1^a	2.0 ± 0.1^a
					preference	6.25	3.7	23 ± 1	32 ± 1^a	33 ± 1^a
Drinkmgcl28	8	84.3	rs13479871	MgCl₂	intake	3.62	3.6	1.6 ± 0.1	2.0 ± 0.1^a	2.2 ± 0.1^a
					preference	7.47	5.4	28 ± 2	35 ± 1^a	40 ± 1^a^b
Drinkmgcl29	8	86.9	rs13479884	MgCl₂	intake	5.02	3.4	1.6 ± 0.1	2.0 ± 0.1^a	2.2 ± 0.1^a
					preference	7.31	5.2	28 ± 2	35 ± 1^a	39 ± 1^a^b
Drinkcala6	8	96.7	rs6287320	CaLa	intake	3.68	1.9	1.8 ± 0.1	2.3 ± 0.1^a	2.4 ± 0.1^a
					preference	6.07	2.9	31 ± 2	39 ± 1^a	41 ± 1^a
Drinkqhcl6	8	101.6	rs3669235	QHCl	intake	3.88	2.5	1.5 ± 0.1	1.6 ± 0.1	1.9 ± 0.1^a^b
					preference	3.22	2.2	31 ± 1	34 ± 1^a	38 ± 1^a^b
Drinknacl4	9	103.6	rs13480386	NaCl	intake	3.62	2.7	5.2 ± 0.2	5.9 ± 0.1^a	6.5 ± 0.2^a^b
Drinkcitric4	9	103.6	rs13480386	citric acid	preference	3.66	2.4	37 ± 2	31 ± 1^a	30 ± 2^a
Drinkkcl8	10	17.9	rs3712394	KCl	intake	3.76	2.8	2.7 ± 0.1	3.1 ± 0.1^a	3.5 ± 0.1^a^b
Drinkcitric5	10	67.9	rs13480630	citric acid	intake	4.60	3.3	2.0 ± 0.1	1.8 ± 0.1	1.5 ± 0.1^a^b
					preference	3.76	2.8	37 ± 1	33 ± 1^a	28 ± 1^a^b
Drinkpyro2	10	69.5	rs13480638^*	Na₃HP₂O₇	preference	4.36	3.3	78 ± 1	76 ± 1	70 ± 2^a^b
Drinksac9	11	96.4	rs13481173	saccharin	intake	6.17	4.5	4.8 ± 0.2	5.4 ± 0.1^a	6.3 ± 0.2^a^b
Drinkwater5	11	103.4	rs6393948^*	water	intake	5.50	4.0	5.5 ± 0.2	6.1 ± 0.1^a	6.7 ± 0.2^a^b
Drinkpyro3	11	103.4		Na₃HP₂O₇	intake	3.31	2.8	4.1 ± 0.1	4.6 ± 0.1^a	4.8 ± 0.1^a
Drinknacl5	12	50.7	rs3700857	NaCl	intake	3.75	2.6	6.3 ± 0.2	5.9 ± 0.2^a	5.1 ± 0.2^a^b
Drinknh4cl4	12	101.8	rs13481632	NH₄Cl	preference	3.66	0.1	42 ± 1	36 ± 1^a	40 ± 1^b
Drinkpyro4	13	87.3	rs6288319	Na₃HP₂O₇	preference	3.52	3.0	72 ± 2	74 ± 1	80 ± 1^a^b
Drinknacl6	17	78.8	gnf17.082.284	NaCl	preference	3.54	2.6	73 ± 1	77 ± 1^a	79 ± 1^a

Open in a new tab

All LOD scores are significant (P < 0.01). Group sizes (n) were slightly smaller for some tests because of measurement errors or missing genotypes. The peak of DrinkNaCl6 on Chr 17 is distinct from the rs3693494 peak (Table 10).

Adjacent to previous marker listed, suggesting there may be a common locus. Genetic coordinates are based on NCBI Build 35. Values are means ± SE, in ml/day for intakes or % for preference scores.

P < 0.01 different from mice with BTBR/BTBR genotype,

P < 0.01 different from mice with BTBR/NZW genotype. The percentage of phenotypic variance (% variance) accounted for by variation in the locus was calculated as r² from the correlation of number of NZW alleles (0, 1, or 2) with phenotype value.