商鞅变法内容:Gustatory responses of pigs to various natural and arti?cial compounds known to be sweet in man

来源：百度文库编辑：中财网时间：2024/05/02 01:37:10

Food Chemistry 68 (2000) 375±385

Rapid communication

www.elsevier.com/locate/foodchem

Gustatory responses of pigs to various natural and articial compounds known to be sweet in man

D. Glaser a,*, M. Wanner b, J.M. Tinti c, C. Nofre c

aAnthropological Institute and Museum, University of Zuèrich-Irchel, Winterthurerstrasse 190, 8057 Zuèrich, Switzerland

bInstitute of Animal Nutrition, University of Zuèrich, Switzerland

cFaculte de Medecine R.T.H. Laennec, Universite Claude Bernard, Lyon, France

Received 4 June 1999; received in revised form; accepted 16 September 1999

Abstract

The gustatory preferences in pigs towards 33 compounds known to be sweet in humans were evaluated through a specic two- choice preference method. All the 14 carbohydrates tested are preferred over water, sucrose being the most e€ective. Sucrose and fructose response intensities are identical in pigs and humans but lactose, maltose, d-glucose and d-galactose are two times less ecient in pigs. The molar order of e€ectiveness is sucrose > d-fructose > maltose=lactose > d-glucose > d-galactose, roughly similar to humans. As in humans, d-glucose, l-glucose and methyl a-d-glucopyranoside display equal potency, while methyl b-d- glucopyranoside is ine€ective. The 7 polyols tested are attractive; xylitol is the preferred one, being as e€ective as sucrose. Out of 12 intense sweeteners tested, 7 are ine€ective (aspartame, cyclamate, monellin, NHDC, P-4000, perillartine, thaumatin), and 5 are attractive (acesulfame-K, saccharin, alitame, dulcin, sucralose), but with a much weaker eciency (acesulfame, 18 less; saccharin,

1. Introduction

The aim of the present study was to investigate how the pig (Sus scrofa domesticus) responds to various compounds known to be sweet in man via a method derived from the well-known `two-bottle preference' test originated by Richter (see Richter, 1942). By means of this behavioural method, with some modications to adapt the test to the pig, we were able to determine, semi-quantitatively, the preference of this animal for various compounds, and, at least for the compounds described as sweet by man, and to infer that the com- pounds which are clearly attractive to the pig should also be perceived as `sweet' to this animal. Thanks to the Richter-type drinking test, it is in fact already known that pigs exhibit, over water, a strong preference for aqueous solutions of sucrose (the most strongly pre- ferred sugar by pigs), glucose (Baldwin, 1976; Kare, Pond

& Campbell, 1965; Kennedy & Baldwin, 1972), lactose (Kare et al., 1965), and sodium saccharin (Baldwin, 1976; Kennedy & Baldwin, 1972), but not for aqueous solutions

* Corresponding author.

of sodium cyclamate (Baldwin, 1976; Kennedy & Baldwin,

1972). Further, through electrophysiological measure-

ments, it has been shown that several other compounds

tasting sweet to humans, such as monellin, thaumatin

(Hellekant, 1976), aspartame or superaspartame (Helle-

kant & Danilova, 1996), do not elicit any signicant

neural responses in the chorda tympani nerve of pigs.

From these data, it was concluded that these compounds

do not taste sweet to pigs (Hellekant & Danilova, 1996).

The purpose of the current work was to go deeper into

our understanding of the responses of pigs to various

compounds sweet to humans, by analysing the gustatory

behaviour of pigs towards 15 carbohydrates, seven

polyols, and 12 various natural or articial compounds

and some commercially used as sweetening agents for

humans.

2. Animals, method and materials

2.1. Animals

Seventy-ve pigs (39 males and 36 females, 2±4 months old) were used for this study. Experiments were

376 D. Glaser et al. / Food Chemistry 68 (2000) 375±385

carried out over a period of 2 years with eight di€erent groups. During the test periods, pigs were housed in individual cages (2 m 3 m), each of them being equip- ped with an automatic water supply freely accessible.

2.2. Method

The method employed is an adapted Richter-type drinking test, derived from the two-bottle preference test previously used by Kare et al. (1965) and Kennedy and Baldwin (1972) in pigs. Two containers D one containing plain water, the other the compound to be tested dissolved in water D are supplied to the animal. The consumption of the tested solution is then mea- sured relative to that of water. In order to test the greatest number of compounds, the problem encoun- tered with pigs is the volume of the solutions ingested (several litres per diem for a strongly preferred solu- tion), these compounds often being very expensive.

To overcome these technical and nancial obstacles, the original Richter procedure was modied by carrying out a preliminary training session before the testing session proper, so that the pigs would acquire the habit of sampling before drinking. Every naive pig was thus trained to make a choice between two containers (buckets), one with tap water, the other with a 100 g/l sucrose solution which is highly attractive to pigs. This training is easily performed thanks to the innate pre- ference of pigs for sugar and to their aptitude for being easily conditioned. In fact, pigs are very quick to locate the sweet-tasting solution (by using a few licks, without drinking, to evaluate the taste quality) and to drink its total volume (250 ml) (in less than 1 min), while the volume of the water control remains practically unchanged. To avoid any forced choice and favour a real preference, the animals always had free access to their usual automatic water supply, even during the testing sessions. As previously observed with primates (Steiner & Glaser, 1984, 1995), various other beha- vioural clues (postural positions, movements of the head, frequency of licking, etc.) were also observed in association with consumption of the sweet solution. The main taste-induced hedonic behaviour expressions eli- cited in pigs by the sweet-tasting solution are the head oriented towards the stimulus, eager drinking, a quick swallow and sucking-smacking, as illustrated in an available video tape (Glaser, Tinti, Nofre & Wanner,

1997); with a bitter-tasting solution (a quinine hydro- chloride solution at a concentration of 49 mg/l), pigs show a typical behaviour of rejection: no consumption of the solution and several behavioural clues associated with the bitter taste, such as the head withdrawn from the stimulus and head shake, as illustrated in the same video tape (Glaser et al., 1997).

Thus, after the training session, each pig knows that one of the two buckets may contain an attractive `sweet'

substance. It was observed that a time generally of about 10±20 s is sucient for a trained pig to make a rapid choice, through two or three licks, between the two options, and to drink the preferred ˉuid greedily, or to move away denitively (with no further interest in the experiment) if an appealing solution has not been detected. The standard duration for each tasting experience was thus xed at only 1 min. Moreover, this brief-exposure procedure, which minimizes the ˉuid consumption, also has the advantage of avoiding any possible postingestional factor, such as caloric regula- tion or physiological aversion. The consumption di€er- ences between the water control and the preferred sapid solution are always important: generally a few millilitres for the water control versus the total volume (250 ml) for the preferred ˉuid. The responses to ascending con- centrations of the tested substances (the concentrations usually progressed such that each level was twice as great as the one before) are denoted by a `+' sign (strong preference) if the tested solution represents at least 80% of the `percentage intake' (volume of test solution consumed/volume of total ˉuid consumed from both test solution and water control 100), or by a `' sign in the other cases, which can then denote a weak preference for, an indi€erence to, or a rejection of the test solution (see Fig. 1).

To validate the results, each experiment was generally duplicated with two di€erent groups out of the eight groups of pigs used during the 2-year period of this study, except for some expensive and/or weakly e€ective compounds which, to avoid excessive costs, were tested on only one group of pigs and/or on a limited number of animals.

Fig. 1. Data analysis: a summarized diagrammatic presentation of the relationships between taste stimulant concentrations and gustatory responses (adapted from Goatcher & Church, 1970). Note that `per- centage intake' means:

Test solution intake

Total fluid intake 100:

The gustatory responses are considered as positive (+) for a percen- tage intake of 80±100% (in the zone of strong preference), and nega- tive () below 80% (in the zones of weak preference, indi€erence or rejection).

D. Glaser et al. / Food Chemistry 68 (2000) 375±385 377

2.3. Materials

Special stainless steel containers, with a conical shape,

Table 1

Gustatory responses of pigs to seven monosaccharides known to be

sweet in humans

were designed to allow the pig full access to the water

control and to the tested solution. This specic shape

makes it possible to reduce the quantity of the ingested

solutions to an acceptable volume (250 ml), which limits

Carbohydratesa Concentrations Number of pigs

mmol/l g/l

Pentoses

Gustatory responsesb

the pig's insatiability towards `sweet' solutions and the

cost of trials with expensive compounds. The two con-

tainers are attached with brackets to a wall of the cage,

in a random left±right position to prevent choices that

could be based on the place of containers. Experiments

started in the morning (at about 9.00 a.m.) and lasted

about half-an-hour. During this period, no more than

two or three trials were carried out with each animal.

All the chemicals tested were of commercial origin (see footnotes `a' in Tables 1, 2, 3 and 5 below), except for alitame and P-4000 which were synthesized as by Brennan and Hendrick (1981) and Verkade, Van Dijk and Meerburg (1946), respectively.

3. Results

Pigs have a gustatory preference for all the 15 carbo- hydrates tested over water (Tables 1 and 2); sucrose is the most preferred carbohydrate (Table 2). Pigs also have a marked preference for all the seven polyols examined versus water; xylitol is the most preferred polyol, being approximately as e€ective as sucrose on a molar basis (Table 3).

A comparison (on a molar basis with regard to sucrose) between the sweetness potencies of these com- pounds in humans and their preferences in pigs shows that their relative e€ectiveness order in pigs closely par- allels their relative potencies in humans, except for xyli- tol which shares the rst place with sucrose, and for sorbitol, d-galactose, d-xylose and d-ribose which appear to be in a higher rank in pigs (Table 4). In humans as in pigs, d-fructose is, on a molar basis, half as potent as sucrose (Table 4). Further, the d- and l- enantiomeric forms of glucose display an equal e€ec- tiveness, both in humans and in pigs (Table 4).

However, the results obtained with 12 articial or natural compounds known to be sweet in humans are more disparate (Table 5). Only ve compounds D i.e. acesulfame-K, alitame, dulcin, saccharin and sucralose D are able to elicit a preference in pigs; the seven others D i.e. aspartame, cyclamate, monellin, NHDC, P-4000, perillartine and thaumatin D do not elicit any appeal in pigs, even for solutions several tens of times more concentrated than needed to induce an explicit sweet perception in humans (except for P-4000 which is too poorly soluble to test concentrated solutions).

d-Ribose 233.79 35.10 4 4+

150.00 22.52 2 2

d-Xylose 116.89 17.55 2 2+

58.44 8.77 2 2

Hexoses

d-Fructose 29.14 5.25 10 10+

15.54 2.62 12 8,4+

7.27 1.31 16 7,9+

3.60 0.65 4 4

87.42 15.75 10 3,7+

58.28 10.50 4 4

d-Glucose 116.56 21.00 12 12+

87.42 15.75 10 4,6+

58.28 10.50 13 10,3+

29.14 5.25 15 12, 3+

14.57 2.62 2 2

l-Glucose 116.56 21.00 5 5+

87.42 15.75 2 2

58.28 10.50 6 6

d-Mannose 116.56 21.00 2 2+

58.28 10.50 2 2

from Fluka; d-fructose, d-galactose and d-glucose, from Merck.

b + Indicates a preference when the tested solution represents 80%

or more of total ˉuid intake from both test solution and water control;

, an indi€erence or a rejection in the other cases.

Note that the acesulfame and saccharin molecules, which share common molecular structural features, are both preferred in pigs, but that alitame and aspartame induce opposite taste responses in pigs, alitame being undoubtedly attractive, aspartame not being so. For all the compounds attractive to pigs, the comparison (always on a molar basis relative to sucrose), between their preferences in pigs and their sweetness potencies in man, shows that their e€ectiveness in pigs is markedly lower (from 25 times for sucralose to 120 times for dulcin) than that necessary in humans for matching the sweetness level of a 2% (58.4 mmol/l) sucrose solution (Table 6).

4. Discussion

The present data highlight several basic similarities between the gustatory responses of pigs and of humans to various carbohydrates and polyols.

Thus, the lowest concentration of sucrose clearly pre- ferred in all the animals tested ( 14 mmol/l) (Table 2) is

378 D. Glaser et al. / Food Chemistry 68 (2000) 375±385

very close to the detection and recognition thresholds of sucrose in humans, which are about 10 and 17 mmol/l, respectively (Amerine, Pangborn & Roessler, 1965a).

The relative molar order of the carbohydrate e€ec- tiveness in pigs roughly mirrors the relative molar

Table 2

Gustatory responses of pigs to seven oilgosaccharides and to one het- eroside known to be sweet in humans

sweetness potency order in humans (Table 4), except for d-galactose, d-xylose and d-ribose. Thus, if we consider the main nutritive carbohydrate sweeteners, the order of e€ectiveness, on a molar basis, is for humans: sucrose > d- fructose > maltose=lactose > d-glucose > d-galactose (Table 4), while this order for pigs is: sucrose > d-fructose

> maltose=lactose > d-glucose=d-galactose (Table 4). Note that in rats, this order is somewhat di€erent: maltose

> d-fructose=lactose > sucrose > d-glucose > d- galactose (Richter & Campbell, 1940; Tonosaki & Beid-

Disaccharides

mmol/l g/l

of pigs

responsesb

Amazingly, d-glucose (in fact, d-glucopyranose) has the same e€ectiveness as its enantiomeric form, l-glu- cose (l-glucopyranose), both in pigs and in humans

Lactose.H2O 100.00 36.03 2 2+

58.42 21.05 10 2,8+

43.71 15.75 10 8,2+

29.14 10.50 4 4

Maltose.H2O 100.00 36.03 24 24+

58.42 21.05 24 10,14+

43.71 15.75 12 9,3+

29.14 10.50 12 12

Melibiose 116.85 40.00 4 4+

58.42 20.00 4 2,2+

43.82 15.00 4 2,2+

29.21 10.00 4 3,1+

14.60 5.00 4 4

Sucrose 14.60c 5.00 12 12+

7.30c 2.50 18 8,10+

5.84 2.00 12 12

(Table 4), which is in favour of a similar symmetrical

arrangement of the recognition sites which bind both

these chiral (dissymmetrical) mirror-image molecules in

the porcine receptor as well as in the human receptor.

As a result, as postulated for the human sweetness

receptor by two of us (Nofre & Tinti, 1996), the porcine

receptor interaction sites of these two molecules with

opposite handedness are possibly a lysine residue NH+

group associated with two aspartate or glutamate residue

Table 3

Gustatory responses of pigs to seven polyols known to be sweet in

humans

Trehalose 116.85 40.00 4 4+

87.64 30.00 6 4,2+

43.82 15.00 4 4

Trisaccharides

Polyolsa Concentrations Number

of pigs mmol/l g/l

Tetrols

Gustatory

responsesb

Melezitose.H2O Ranose.5H2O

dl-Threitol 116.27 14.20 6 6+

87.20 10.65 2 2

Erythritol 234.19 28.60 4 4+

175.64 21.45 2 2

Pentols

Heterosided

Methyl a-d-gluco-

pyranoside

a Lactose, maltose, sucrose, trehalose, ranose and methyl a-d- glucopyranoside are compounds from Fluka; melibiose and melezi- tose, from Sigma.

b + Indicates a preference when the tested solution represents 80% or more of total ˉuid intake from both test solution and water control; , an indi€erence or a rejection in the other cases.

c Note that these values in pigs are close to the detection and recognition thresholds of sucrose as obtained in humans, which are about 10 and 17 mmol/l respectively (Amerine et al., 1965a). This

observation argues in favour of a roughly similar mechanism in the

interaction of sucrose with the pig and the human sweetness receptors, and substantiates our choice of adopting sucrose as the sweetness

standard preference in pigs.

d Note that methyl b-d-glucopyranoside, which is unsweet in

humans (unlike its a anomer), does not elicit any preference in pigs,

these animals being indi€erent towards concentrations of 58.39 to

116.78 mmol/l of this compound.

a dl-Threitol, d-arabitol, ribitol, xylitol and sorbitol are com- pounds from Sigma; erythritol, from Fluka; mannitol, from Merck.

b + Indicates a preference when the tested solution represents 80%

or more of total ˉuid intake from both test solution and water control;

, an indi€erence or a rejection in the other cases.

D. Glaser et al. / Food Chemistry 68 (2000) 375±385 379

Table 4

Comparison (on a molar basis relative to sucrose) between the sweet-

ness potencies in humans and the preferences in pigs for 23 various polyhydroxy compounds (carbohydrates and polyols)

Table 5

Gustatory responses of pigs to 12 compounds described as sweeteners

in humans

Carbohydrates and polyols Potencies

in humansa

Preferences in pigsc

Compoundsa Concentrations Number

of pigs mmol/l g/l

Custatory

responsesb

Sucrose 1.00 1.00

d-Fructose 0.50 0.50

Melezitose 0.35 0.25

Lactose 0.33 0.146

Maltose 0.33 0.146

Xylitol 0.30 1.00

d-Glucose 0.25 0.125

l-Glucose 0.25 0.125

d-Mannose 0.25b 0.125

Melibiose 0.25 0.125

Trehalose 0.25 0.125

Ranose 0.25 0.125

Methyl a-d-glucopyranoside 0.25 0.125

dl-Threitol 0.25 0.125

Erythritol 0.25 0.062

d-Arabitol 0.25b 0.062

Ribitol 0.25 0.062

Mannitol 0.25 0.062

Acesulfame-K 1.73 0.35 4 4+

0.24 0.05 4 2,2+

0.04 0.01 4 4

Alitame 0.30 0.10 4 4+

0.15 0.05 4 4

Aspartame 16.98 5.00 5 5

10.19 3.00 4 4

5.09 1.50 4 4

Cyclamate 99.39 20.00 4 4

(sodium salt) 49.69 10.00 4 4

24.84 5.00 4 4

Dulcin 13.31 2.40 4 4+

6.65 1.20 4 3,1+

3.32 0.60 4 4

Monellin 0.018 0.20 4 4

Sorbitol 0.25 0.25

d-Galactose 0.20b 0.125

d-Xylose 0.20 0.125

Neohesperidin dihydrochalcone

(NHDC)

0.097 0.60 4 4

relative to a 2% sucrose solution) were evaluated (or re-evaluated) by

c The approximate relative preferences in pigs were estimated from

the animals of the same experimental group (14.60 mmol/l) divided by the lowest concentration of the tested compound which is able to induce a preference in all the animals of the same group (e.g. 29.14 mmol/l for d-fructose).

CO2 groups symmetrically arranged in space relatively to the lysine ammonium group (Fig. 2).

Another interesting analogy between the porcine and the human responses towards carbohydrates is that pigs have an equal preference for d-glucose (Table 1) and for methyl a-d-glucopyranoside (Table 2), a heterosidic carbohydrate (Fig. 3a ); likewise, both these molecules have similar sweetness potencies in humans (Table 4). Moreover, pigs are indi€erent towards methyl b-d-gluco- pyranoside (Fig. 3b), the anomeric form of methyl a-d- glucopyranoside; similarly, methyl b-d-glucopyranoside is unsweet to humans (see Table 2, footnote `d', and Table 4). The unsweetness of methyl b-d-glucopyrano- side can be explained, both in man and in pigs, as the result of a `steric clash' between the methyl group of its b-methoxy substituent and the methyl group of a threonine residue (denoted Thr-6) of the receptor, lead- ing to a mistted adaptation of the molecule to the

a All the compounds cited are described in The Merck Index (12th ed.) as sweeteners. Aspartame, monellin, neohesperidin dihydro- chalcone, perillartine and thaumatin, are compounds from Sigma; acesulfame-K, from Supelco; cyclamate, from Merck; dulcin, from Interchim, France; Saccharin, from Fluka; sucralose from Redpath, Canada; alitame and P-400 were synthesized after Brennan and Hen- drick (1981) and Verkade et al. (1946), respectively.

b + Indicates a preference when the tested solution represents 80% or more of total ˉuid intake from both test solution and water control; , an indi€erence or a rejection in the other cases.

receptor. With methyl a-d-glucopyranoside, this steric hindrance does not occur owing to the di€erent spatial orientation of its a-methoxy group, which allows a sui- table docking of the molecule into the receptor, in the same way as d-glucopyranose (Fig. 2a). Although the relative sequence of the responses in pigs towards car- bohydrates mirrors the sequence of their potencies in humans, their relative response intensities (by compar- ison with sucrose, our standard reference) are very dif- ferent (between the porcine and the human responses) for all the carbohydrates tested (except for d-fructose and sucrose) (see Table 4 and Fig. 4). For example, lactose

380 D. Glaser et al. / Food Chemistry 68 (2000) 375±385

Table 6

Comparison (on a molar basis relative to sucrose) between the sweet-

ness potencies in humans and the preferences in pigs for various com- pounds described as sweeteners in humans

Compounds Potencies

in humansa

Preferences in pigsb,c

Ratio potency in humans/ preference

in pigsd

Monellin 100 000 ± Thaumatin 100 000 ±

Neohesperidin

dihydrochalcone (NHDC)

5-Nitro-2-propoxyaniline

(P-4000)

3600 ±

2300 ±

Alitame 1900 48.66 40

Sucralose 1160 47.09 25

Perillartine 370 ±

Saccharin 215 3.34 65

Aspartame 155 ±

Acesulfame-K 150 8.43 18

Dulcin 130 1.09 120

Cyclamate (Na) 17.6 ±

a The approximate sweetness potencies in humans (on a molar basis relative to a 2% sucrose solution) were evaluated (or re-evaluted) by six trained panellists of our laboratory through the paired-comparison (two-sample) test (see Amerine et al., 1965b).

b The approximate relative preferences in pigs were estimated from the lowest concentration of sucrose able to induce a preference in all the animals of the same experimental group (14.60 mmol/l) divided by the lowest concentration of the tested compound which is able to induce a preference in all the animals of the same group (e.g. 4.36 mmol/l for saccharin).

c - Indicates an indi€erence or a rejection.

d The value of the ratio indicates how many times, on a molar basis, the studied compound is approximately less `sweet' in pigs than in humans.

Fig. 2. The six dominant electrostatic interactions (through six ioni- cally-assisted hydrogen bonds, indicated by dotted lines) between the human sweetness receptor and the molecules of (a) d-glucopyranose and (b) l-glucopyranose, as postulated by Nofre and Tinti (1996). Note that the functional groups of the three ionic recognition sites (the so-called `ionic triad', denoted Asp-1 or Glu-1, Lys-2, and Asp-3 or Glu-3), which are assumed to be implicated in these interactions, are symmetrically arranged in space; this enables us to understand why the d- and l-enantiomers of glucopyranose elicit similar responses in humans (Nofre & Tinti, 1996), and, by inference, in pigs.

and maltose are approximately six times less appre- ciated by pigs than sucrose, while in man both these compounds display a sweetness potency which is about one third that of sucrose; likewise, d- and l- glucose, d-mannose, d-galactose, melibiose, trehalose, ranose and methyl a-d-glucopyranoside are approxi- mately eight times less preferred by pigs than sucrose, while in man all these compounds exhibit a sweetness potency which is about a quarter that of sucrose.

Concerning the polyols (Tables 3 and 4), xylitol is the most potent of these compounds in pigs and in humans; but, while xylitol is, in humans, about one third less sweet than sucrose on a molar basis ( 0.30 sucrose on a molar basis, 0.70 on a weight basis according to our own assessment), it is roughly as preferred as sucrose on a molar basis in pigs (see Fig. 4). Sorbitol is, just after xylitol, the most favoured polyol in pigs; it is approximately four times less preferred in pigs than sucrose or xylitol, but about twice as preferred as d- glucose, while in man sorbitol is isosweet with d-glu- cose. Note that sorbitol is common in many fruits (see Wang & van Eys, 1981), often at a concentration of about 10±30 g/l of fresh fruit juice (see Dwivedi, 1986).

Although slowly absorbed by the intestine, this polyol may be considered as an e€ective energetic sweetener, being metabolically converted into d-fructose at the hepatic level (see, e.g. Dwivedi, 1986; Sicard, 1982). dl- Threitol, a tetrol, is isosweet with d-glucose in pigs and in humans. For the other polyols D namely, erythritol, d-arabitol, ribitol and mannitol D these compounds are about 16 times less preferred than sucrose or xylitol and twice less than d-glucose in pigs, while they are approximately four times less sweet than sucrose and are isosweet with d-glucose in man.

Among these results on the carbohydrates and poly-

ols, we particularly highlight the amplication of the

human response to d-glucose by comparison with the

porcine response, and, conversely, the reduction of the

human response to xylitol compared with the porcine

response (Fig. 4). From a phylogenetic point of view,

the di€erence between the responses of pigs and humans

towards d-glucose and xylitol is possibly a consequence

of an evolutionary adaptation of the human (and, more

generally, of the catarrhine) sweetness receptor to a

keener detection of d-glucose, a highly-energetic free car-

bohydrate which is common, with sucrose and d-fructose,

D. Glaser et al. / Food Chemistry 68 (2000) 375±385 381

in various foods of plant origin (see, e.g. Astrup & Raben, 1996; Frostell, 1980; Guesry & Secretin, 1991). Concerning xylitol, which is also rather common in various fruits and vegetables (see Wang & van Eys,

1981) at concentrations of about 0.1±0.4 g per kg of

fresh weight (see Makinen & Soderling, 1980), note that

this compound possesses a weak physiological interest

as a result of a slow and incomplete intestinal absorp-

tion (approximately one-third of the ingested portion of

xylitol is absorbed, the rest being actively metabolized

by intestinal ˉora) and of a dual metabolic pathway in

Fig. 3. (a) Methyl a-glucopyranoside, which matches d-glucose both in humans and in pigs, and (b) methyl b-glucopyranoside, which is ine€ective both in humans and in pigs. Note that the unsweetness of methyl b-d-glucopyranoside in humans is assigned to a `steric clash' between the methyl substituent of the equatorially-oriented methoxy group of this heteroside and the side chain of a threonine residue of the receptor, Thr-6 (see Fig. 6 hereafter for further details), which induces a mist of the ligand into the receptor; by inference, we assume that the pig disinterest in this compound could be due to the same steric hindrance between this molecule and the porcine receptor, and that Thr-6 is consequently retained in the porcine receptor. On the other hand, the methyl substituent of the axially-oriented methoxy group of methyl a-glucopyranoside does not collide with the Thr-6 residue according to a simulated molecular interaction of this molecule with the Nofre/Tinti model of the sweetness receptor, which should explain why this molecule is isosweet with d-glucose in humans as in pigs, since it is able to interact with the receptor through the same dominant electrostatic interactions as those postulated for d-glucose (see Fig. 2a).

the liver through relatively secondary routes (see, e.g. Bar, 1986; Levine, 1986; Schi€man & Gatlin, 1993; Sicard, 1982). The minor interest of xylitol in mammals might explain why the free access to the sweetness receptor of this molecule D which, through its sweet- ness, should normally interfere with the food selection D is partly hindered in the most `advanced' receptors, such as in the catarrhine ones.

Out of the 12 additionally-tested compounds which are also well known to taste sweet to man (Table 5), pigs show no preference for seven of them, namely aspartame, sodium cyclamate, monellin, NHDC, P-4000, perillartine and thaumatin (see Fig. 5).

The indi€erence of pigs towards aspartame is not surprising as all the mammals tested so far D with the only exception of Catarrhini (Old World primates, including man) (Glaser, Tinti & Nofre, 1995) D do not give any explicit `sweet' gustatory responses to aspar- tame, as observed in hamsters (Danilova, Hellekant, Roberts, Tinti & Nofre, 1998; Nowlis, Frank, Pfa€man,

1980), gerbils (Jakinovich, 1981), rats (Nowlis et al.,

1980, Hellekant & Walters, 1992), dogs, cows and

horses (Glaser, Tinti & Nofre, unpublished results),

Prosimii (prosimians) and Platyrrhini (New World

monkeys) (Glaser et al., 1995; Glaser, Tinti & Nofre,

1996). According to the multipoint attachment (MPA)

theory as proposed by Nofre and Tinti (1996), the

human sweetness receptor appears to be formed of at

least eight recognition (`binding') sites arranged around

the central cavity of the receptor; these sites are assumed

to be made up of: an aspartate or a glutamate residue

(termed Asp-1 or Glu-1), a lysine residue (Lys-2),

another aspartate or glutamate residue (Asp-3 or Glu-

3), four threonine residues (Thr-4, Thr-5, Thr-6, Thr-7),

and a serine residue (Ser-8) (Fig. 6). As no di€erence in

the gustatory responses of diverse nonhuman catarrhine

primates towards various articial sweeteners has been

detected so far, it has been inferred that these primates

hold the same key recognition sites in their sweetness

receptors as those of humans (Glaser et al., 1996; Nofre,

Fig. 4. The relative e€ectiveness in pigs (on a molar basis) of the main carbohydrates and polyols found in foods compared to the relative sweetness potencies (on a molar basis) of the same compounds in humans.

Fig. 5. The relative e€ectiveness in pigs (on a molar basis) of the main articial sweeteners compared to the relative sweetness potencies (on a molar basis) of the same compounds in humans.

382 D. Glaser et al. / Food Chemistry 68 (2000) 375±385

Tinti & Glaser, 1996). For prosimian and platyrrhine primates, which do not taste aspartame, unlike catar- rhine primates (Glaser et al., 1995), it has been pro- posed, from structure±activity relationships, that this distinctive character between these primates could be due to the presence, in the noncatarrhine sweetness receptors, of a serine or alanine residue (Ser-5 or Ala-5) in place of the Thr-5 site of the catarrhine receptors (Glaser et al., 1996; Nofre et al., 1996). This substitution of Ser-5 (or Ala-5) for Thr-5 makes impossible an e€ective steric t of the phenyl ring of aspartame between Thr-5 and Thr-7, this steric t being apparently crucial for the activation of the receptor by aspartame (Nofre & Tinti, 1996; Nofre et al.). By analogy with the non- catarrhine primates, we infer that the indi€erence to aspartame of pigs (and, more generally, of all the non- catarrhine mammals) is the result of the replacement of Thr-5 by Ser-5 (or Ala-5) (Fig. 7), which makes this sweetener ine€ective.

Just as aspartame, sodium cyclamate, which is sweet to all the catarrhine primates tested until now (Nofre et al., 1996), is `unsweet' to pigs (Kennedy & Baldwin,

1972; Baldwin, 1976; Glaser et al. in the present work), and to all the mammals studied so far, such as hamsters (Danilova, Hellekant, Roberts et al., 1998; Danilova, Hellekant, Tinti & Nofre, 1998; MacKinnon, Frank & Rehnberg, 1996; Rehnberg, Hettinger & Frank, 1990), gerbils (Jakinovich, 1981), rats (Murray, Wells, Kohn & Miller, 1953), cats (Bartoshuk, Jacobs, Nichols, Ho€ &

Fig. 6. Model of the human sweetness receptor according to the mul- tipoint attachment (MPA) theory (Nofre & Tinti, 1996). The spheres of the model represent the approximate spatial positions of the di€er- ent functional groups that may be involved in the interactions of the human receptor with various natural or articial sweeteners. Note that the MPA model has recently been re-examined (Nofre & Tinti, unpublished work); particularly, it has been inferred, from a compre- hensive structure±activity relationship study, that an additional Thr

Ryckman, 1975, Beauchamp, Maller & Rogers, 1977), tree shrews, and noncatarrhine primates (Nofre et al.,

1996). According to a recent improvement of the MPA theory (Nofre & Tinti, unpublished work), it appears that the sweet stimulus induced by cyclamate in man may be partly due (in addition to several electrostatic interactions between the NHSO3 group of cyclamate and some recognition sites of the receptor) to a steric t of the cyclamate cyclohexyl group between Thr-6 and a valine residue (provisionally termed Val-10, as indicated in the caption of Fig. 6) located behind Thr-4 (and under Thr-5) in the MPA model (see Fig. 6). From this re-examined version of the model, it is argued that, in the porcine receptor (and possibly in all the noncatar- rhine mammalian sweetness receptors), Val-10 could be replaced by an Ala residue, which should suppress any possibility of activation of the receptor through a steric

t of the cyclamate cyclohexyl group (Fig. 8).

For the ve other compounds sweet to humans but

`unsweet' to pigs (monellin, NHDC, P-4000, perillartine

and thaumatin), we believe that the indi€erence of pigs

towards these various compounds is also the result of

the absence of one (or more) steric interaction(s) or

steric t(s) between these molecules and the porcine

receptor, as has been postulated for aspartame or

cyclamate.

The ve other articial sweeteners tested in the pre-

sent study (acesulfame-K, saccharin, alitame, dulcin,

and sucralose) elicit clear `sweet' responses in pigs

Fig. 7. Aspartame (150 sucrose in man on a molar basis, but not

`sweet' to pigs) and its putative main steric interactions (indicated with double-headed arrows) with the human sweetness receptor, according to the MPA theory (Nofre & Tinti, 1996), or with the porcine receptor,

as inferred from detailed structure±activity studies on primates (Glaser

et al., 1996; Nofre et al., 1996); from these studies, we suggest, by analogy, that the presumed Thr-5 recognition site of the human receptor could be replaced by a Ser-5 or an Ala-5 residue in the por-

cine receptor. For clarity, the putative electrostatic interactions

between aspartame and the receptor have not been indicated in this diagram; from the MPA theory, it is assumed that these interactions

recognition site (denoted Thr-9 in the diagram) must exist above Asp-

mainly occur between, on the one hand, the CO2, NH3

and COOCH3

1/Glu-1 and before Thr-7, and a valine site (denoted Val-10 in the

diagram) behind Thr-4 and under Thr-5.

groups of aspartame, and, on the other hand, the ionic triad and Thr-4

of the receptor.

D. Glaser et al. / Food Chemistry 68 (2000) 375±385 383

Fig. 8. Cyclamate (17.6 sucrose in man on a molar basis, but not

`sweet' to pigs) and the putative steric t (indicated with double-

headed arrows) of its cyclohexyl moiety between Thr-6 and Val-10 (see the caption of Fig. 6) of the human receptor. As Thr-6 appears to be retained in pigs (see the caption of Fig. 3), only Val-10 should be changed, possibly into an alanine (Ala-10) residue. For clarity, the

electrostatic interactions between cyclamate and the receptor have not

been represented; from the MPA theory, it is assumed that these

interactions take place between the NHSO3 part of cyclamate and the ionic triad of the receptor.

(Table 5), but much weaker than in humans (from 18 to 120 times less intense according to the sweetener employed) (Table 6). The weakness of the pig responses is attributed to the lack of some steric interaction (or steric t) aptitudes of the porcine receptor with regard to the steric interaction (or steric t) capabilities of the human receptor. To illustrate this view, we shall take two examples, acesulfame-K and saccharin, on account of their importance as commercial sweeteners. For con- venience, the other less-known sweeteners will be the subject of separate publications.

The sweetness potency of acesulfame (Fig. 9a) is in humans 150 times that of sucrose and its e€ectiveness in pigs is 10 times that of sucrose on a molar basis (see Fig. 5 and Table 6). According to data from structure- activity relationship studies (Nofre & Tinti, unpublished work), it is assumed that acesulfame should interact with the human receptor, in addition to several electro- static interactions, through one steric interaction which occurs between the 6-methyl group of acesulfame and the assumed Thr-9 recognition site (see the caption of Fig. 6). Furthermore, it is known that the unsubstituted oxathiazinone dioxide ring (Fig. 9b) is only 10 times sweeter than sucrose on a molar basis in humans (Clauss & Jensen, 1973); the low potency of this com- pound is interpreted, through the views of the MPA theory, by the impossibility, for this molecule, of con- tracting a steric interaction with the Thr-9 recognition site. As acesulfame has a relative e€ectiveness of 10 times sucrose in pigs (just as the unsubstituted oxathia- zinone dioxide ring in humans), it is inferred that Thr-9 is not retained in pigs (see Fig. 9a), and that this residue could be, for example, an alanine (Ala) or a serine (Ser) residue in the porcine receptor.

Concerning saccharin (Fig. 10), which is justly regar- ded as a very close structural analogue of acesulfame, its sweetness potency is in humans of 215 times that of

Fig. 9. (a) Acesulfame (6-methyloxathiazinone dioxide): the sweetness potency of this compound is in humans 150 sucrose, and its e€ec- tiveness in pigs 10 sucrose on a molar basis; (b) unsubstituted oxa- thiazinone dioxide: its sweetness potency in humans is about

10 sucrose on a molar basis (Clauss & Jensen, 1973). These values indicate that the steric interaction of acesulfame, as assumed in the human sweetness receptor (indicated by a double-headed arrow) and assigned to a putative Thr-9 residue (see the caption of Fig. 6), does not exist in the porcine receptor. As a consequence, the porcine receptor must behave with acesulfame just as the human receptor with the unsubstituted oxathiazinone dioxide, i.e. without formation of a steric interaction between the receptor and the acesulfame methyl group. For clarity, the electrostatic interactions have not been repre- sented in the diagram; these interactions involve (i) the acesulfame CONHSO2 moiety and (ii) the receptor ionic triad and the Thr-6 resi- due according to the MPA theory.

Fig. 10. Saccharin: the sweetness potency of this compound is in humans 215 sucrose, and its e€ectiveness in pigs 3.3 sucrose on a molar basis. It is assumed that saccharin interacts with the human receptor through two steric interactions (represented by two double- headed arrows in the diagram): one between the Thr-6 site and the 4- position of the benzo ring of saccharin, the other between the Thr-9 site and the 6-position of the benzo ring. This generates an ecient steric t of the molecule of saccharin onto the receptor. In the porcine receptor, while Thr-6 looks retained (see the caption of Fig. 3), Thr-9 appears to be missing, as inferred from the pig responses to acesulfame (see Fig. 9), which prevents any steric t possibility of the saccharin molecule. For clarity, the electrostatic interactions have not been represented.

sucrose, and its e€ectiveness in pigs of 3.3 times that of sucrose on a molar basis (see Fig. 5 and Table 6). From structure±activity relationship studies (Nofre & Tinti, unpublished work), it is now assumed that sac- charin should interact with the human sweetness receptor, in addition to several electrostatic interactions, through the steric t of its benzo aromatic ring between the methyl groups of (i) Thr-6 (via the 4-CH of the sac- charin benzo ring) and (ii) Thr-9 (via the 6-CH of the

384 D. Glaser et al. / Food Chemistry 68 (2000) 375±385

saccharin benzo ring) (Fig. 10). Through the concepts of the MPA theory, since it appears that Thr-6 should be maintained in pigs (see the caption of Fig. 3) but not Thr-9 (see the caption of Fig. 9), the steric t of the saccharin molecule, which is highly ecient in the human receptor, should be missing in the porcine receptor. This could explain why saccharin is about 65 times less e€ective in pigs than in humans.

If the presence or absence of Thr-9 in receptors is really the source of the disparities between species in their gustatory responses to saccharin (or acesulfame), it may be supposed that its presence or absence in a receptor could also be at the origin of the substantial individual variations often encountered with these sweeteners within species (e.g. through erratic results in the gustatory responses, through tendencies towards bimodal distributions of the sweetened ˉuid intake, or, in rodents, via animals selectively bred for high versus low saccharin consumption). Such individual variations have been observed, e.g. in rats (Badia-Elder, Kiefer & Dess, 1996; Dess, 1993; Giza, McCaughey, Zhang & Scott, 1996; Nachman, 1974), guinea pigs (Jacobs,

1978), Virginia opossums (Pressman & Doolittle, 1966), hedgehogs (Ganchrow, 1976), squirrel monkeys (Dua- Sharma & Smutz, 1977; Fisher, Pfa€mann & Brown,

1965), or even in pigs (Kare et al., 1965).

The genetic origin of these within-species variations in

the responses to saccharin (or acesulfame) has been

particularly well documented in mice, in which clear-cut

dichotomous di€erences have been demonstrated

between various inbred mouse strains (Beauchamp et

al., 1998; Capretta, 1970; Fuller, 1974; Lush, 1989;

Lush, Hormigold, King & Stoye, 1995; Ninomiya,

Higashi, Katsukawa, Mizukoshi & Funakoshi, 1984;

Pelz, Whitney & Smith, 1973; Ramirez & Fuller, 1976).

For example, it is recognized that C57BL/6 mice

strongly prefer saccharin solution to water, while DBA/

2 mice show a much lower preference for this sweetener

(Capretta; Fuller; Lush); this e€ect is even more marked

with acesulfame (Lush; Lush et al.). This strain di€er-

ence appeared to be due to a single gene called Sac, the

C57BL/6 allele having been designated Sacb, the DBA/2

allele, Sacd (Fuller); these ndings were conrmed by

Lush, who localized this gene on mouse chromosome 4

(Chr 4), mapping it near the telomeric end of the chro-

mosome, between the D4Smh6b and Tel4q regions, at

8.13.4 cM distal to Nppa (Lush; see Mock & Hirano,

1998, for the latest report on mouse chromosome 4).

From these ndings, it is tempting to speculate that

the molecular di€erence among the animals having a

strong preference for saccharin or acesulfame and those

having a weak preference for these sweeteners lies only

in the presence or in the absence of a threonine residue

in their sweetness receptors.

Acknowledgements

The authors are grateful for kind assistance in carry- ing out the experiments to Roland Liechti, Helena Cio- larro, Dr. med. vet. Urs Huwyler, Dr. med. vet. Annette Liesegang, Dr. med. vet. Claudia Lutolf, Dr. med. vet. Andrea Zabka and med. vet. Ingo Zehne, and thank Edith Le Bredonchel for her contribution to this work. The present research was supported by the NutraSweet Company and is a contribution to the EU-project: `The Mechanistic Understanding of the Sweetness Response' AIR3-CT94-2107) and was nancially supported from the `Bundesamt fur Bildung und Wissenschaft' (BBW Nr. 94.0156) Bern, Switzerland.

References

Amerine, M. A., Pangborn, R. M., & Roessler, E. B. (1965a). Princi- ples of sensory evaluation of food. New York: Academic Press (p.

88).

Amerine, M. A., Pangborn, R. M., & Roessler, E. B. (1965b). Princi-

ples of sensory evaluation of food. New York: Academic Press (pp.

328±331).

Astrup, A., & Raben, A. (1996). Mono- and disaccharides: nutritional

aspects. In A.-C. Eliasson, Carbohydrates in foods (pp. 159±189).

New York: Marcel Dekker.

Badia-Elder, N., Kiefer, S. W., & Dess, N. (1996). Taste reactivity in

rats selectively bred for high vs. low saccharin consumption. Phy-

siology & Behavior, 59, 749±755.

Baldwin, B. A. (1976). Quantitative studies on taste preferences in

pigs. Proceedings of the Nutrition Society, 35, 69±73.

Bar, A. (1986). Xylitol. In L. O. Nabors, & R. C. Gelardi, Alternative

sweeteners (pp. 185±216). New York: Marcel Dekker.

Bartoshuk, L. M., Jacobs, H. L., Nichols, T. L., Ho€, L. A., &

Ryckman, J. J. (1975). Taste rejection of nonnutritive sweeteners in

cats. Journal of Comparative and Physiological Psychology, 89, 971±

975.

Beauchamp, G. K., Bachmanov, A. A., Reed, D. R., Inoue, M.,

Ninomiya, Y., Tordo€, M. G., & Price, R. A. (1998). Marker-assisted selection of a high saccharin-preferring 129.B6-Sac congenic mouse

strain. Chemical Senses, 23, 644 (Abstr.).

Beauchamp, G. K., Maller, O., & Rogers, J. G. Jr. (1977). Flavor

preferences in cats (Felix catus and Panthera sp.). Journal of Com-

parative and Physiological Psychology, 91, 1118±1127.

Brennan, T. M., & Hendrick, M. E. (1981). Branched amides of l-

aspartyl-d-amino acid dipeptides and compositions thereof. Eur-

opean Patent Application 0034876 (Sep. 2, 1981).

Capretta, P. J. (1970). Saccharin and saccharin-glucose ingestion in

two inbred strains of Mus musculus. Psychonomic Science, 21, 133±

135.

Clauss, K., & Jensen, H. (1973). Oxathiazinone dioxides D a new

group of sweetening agents. Angewandte Chemie, International Edi-

tion in English, 12, 869±876.

Danilova, V., Hellekant, G., Roberts, T., Tinti, J. M., & Nofre, C. (1998). Behavioral and single chorda tympani taste ber responses

in the common marmoset, Callithrix jacchus jacchus. Annals of the

New York Academy of Sciences, 855, 160±164.

Danilova, V., Hellekant, G., Tinti, J. M., & Nofre, C. (1998). Gusta-

tory responses of the hamster Mesocricetus auratus to various com-

pounds considered sweet by humans. Journal of Neurophysiology,

80, 2102±2112.

D. Glaser et al. / Food Chemistry 68 (2000) 375±385 385

Dess, N. K. (1993). Saccharin's aversive taste in rats: evidence and implications. Neuroscience and Biobehavioral Reviews, 17, 359±372. Dua-Sharma, S., & Smutz, E. R. (1977). Taste acceptance in squirrel monkeys (Saimiri sciureus). Chemical Senses and Flavor, 2, 341±352. Dwivedi, B. K. (1986). Polyalcohols: sorbitol, mannitol, maltitol, and hydrogenated starch hydrolysates. In L. O. Nabors, & R. C. Gelardi,

Alternative sweeteners (pp. 165±183). New York: Marcel Dekker. Fisher, G. L., Pfa€mann, C., & Brown, E. (1965). Dulcin and sac-

charin taste in squirrel monkeys, rats and men. Science, 150, 506±

507.

Frostell, G. (1980). Natural and added sweet carbohydrates in foods

and diets. In P. Koivistoinen, & L. Hyvonen, Carbohydrate sweet-

eners in foods and nutrition (pp. 1±14). London: Academic Press.

Fuller, J. L. (1974). Single-locus control of saccharin preference in

mice. Journal of Heredity, 65, 33±36.

Ganchrow, J. R. (1977). Consummatory responses to taste stimuli in

the hedgehog (Erinaceus europaeus). Physiology & Behavior, 18,

447±453.

Giza, B. K., McCaughey, S. A., Zhang, L., & Scott, T. R. (1996). Taste responses in the nucleus of the solitary tract in saccharin-pre-

ferring and saccharin-averse rats. Chemical Senses, 21, 147±157.

Glaser, D., Tinti, J. M., & Nofre, C. (1995). Evolution of the sweetness receptor in primates. I. Why does alitame taste sweet in all prosi- mians and simians, and aspartame only in Old World simians?

Chemical Senses, 20, 573±584.

Glaser, D., Tinti, J. M., & Nofre, C. (1996). Gustatory responses of non-human primates to dipeptide derivatives or analogues, sweet in

man. Food Chemistry, 56, 313±321.

Glaser, D., Tinti, J. M., Nofre, C. & Wanner, M. (1997). Gustatory

responses in pigs to compounds sweet in man. Video tape, SVHS, 5.5 min.

Goatcher, W. D., & Church, D. C. (1970). Taste responses in rumi- nants. I. Reactions of sheep to sugars, saccharin, ethanol and salts.

Journal of Animal Science, 30, 773±783.

Guesry, P. R., & Secretin, M.-C. (1991). Sugars and nonnutritive

sweeteners. In N. Kretchmer, & E. Rossi, Sugars in nutrition (pp.

33±51). New York: Raven Press.

Hellekant, G. (1976). On the gustatory e€ects of monellin and thau-

matin in dog, hamster, pig and rabbit. Chemical Senses and Flavor,

2, 97±105.

Hellekant, G., & Danilova, V. (1996). Species di€erences toward

sweeteners. Food Chemistry, 56, 323±328.

Hellekant, G., & Walters, D. E. (1992). An example of phylogenetic di€erences in sweet taste: sweetness of ve high-potency sweeteners

in rats. In M. Mathlouthi, J. A. Kanters, & G. G. Birch, Sweet-taste

chemoreception (pp. 373±386). London: Elsevier.

Jacobs, W. W. (1978). Taste responses in wild and domestic guinea

pigs. Physiology & Behavior, 20, 579±588.

Jakinovich Jr., W. (1981). Stimulation of the gerbil's gustatory recep-

tors by articial sweeteners. Brain Research, 210, 69±81.

Kare, M. R., Pond, W. C., & Campbell, J. (1965). Observations on the

taste reactions in pigs. Animal Behaviour, 13, 265±269.

Kennedy, J. M., & Baldwin, B. A. (1972). Taste preference in pigs for

nutritive and non-nutritive sweet solutions. Animal Behaviour, 20,

706±718.

Levine, R. (1986). Monosaccharides in health and disease. Annual

Review of Nutrition, 6, 211±224.

Lush, I. E. (1989). The genetics of tasting in mice. VI. Saccharin, ace-

sulfame, dulcin and sucrose. Genetical Research, 53, 95±99.

Lush, I. E., Hornigold, N., King, P., & Stoye, J. P. (1995). The genetics of tasting in mice. VII. Glycine revisited, and the chromosomal loca-

tion of Sac and Soa. Genetical Research, 66, 167±174.

MacKinnon, B. I., Frank, M. E., & Rehnberg, B. G. (1996). Sweetener similarity in hamsters as determined by generalization of condi- tioned taste aversions. Chemical Senses, 21, 637±638.

Makinen, K. K., & Soderling, E. (1980). A quantitative study of mannitol, sorbitol, xylitol and xylose in wild berries and commercial

fruits. Journal of Food Science, 45, 367±371, 374.

Mock, B. A., & Hirano, M. C. (1998). Mouse chromosome 4. Mam-

malian Genome, 8, S68±S90.

Murray, E. J., Wells, H., Kohn, M., & Miller, N. E. (1953). Sodium sucaryl: a substance which tastes sweet to human subjects but is

avoided by rats. Journal of Comparative and Physiological Psychol-

ogy, 46, 134±137.

Nachman, M. (1974). The inheritance of saccharin preference. Journal

of Comparative and Physiological Psychology, 52, 451±457.

Ninomiya, Y., Higashi, T., Katsukawa, H., Mizukoshi, T., & Funa- koshi, M. (1984). Qualitative discrimination of gustatory stimuli in

three di€erent strains of mice. Brain Research, 322, 83±92.

Nofre, C., & Tinti, J. M. (1996). Sweetness reception in man: the

multipoint attachment theory. Food Chemistry, 56, 263±274.

Nofre, C., Tinti, J. M., & Glaser, D. (1996). Evolution of the sweetness receptor in primates. II. Gustatory responses of non-human pri-

mates to nine compounds known to be sweet in man. Chemical

Senses, 21, 747±762.

Nowlis, G. H., Frank, M. E., & Pfa€mann, C. (1980). Specicity of

acquired aversions to taste qualities in hamsters and rats. Journal of

Comparative and Physiological Psychology, 94, 932±942.

Pelz, W. E., Whitney, G., & Smith, J. C. (1973). Genetic inˉuences

on saccharin preference of mice. Physiology & Behavior, 10, 263±

265.

Pressman, T. G., & Doolittle, J. H. (1966). Taste preferences in the

Virginia opossum. Psychological Reports, 18, 875±878.

Ramirez, I., & Fuller, J. L. (1976). Genetic inˉuence on water and

sweetened water consumption in mice. Physiology & Behavior, 16,

163±168.

Rehnberg, B. G., Hettinger, T. P., & Frank, M. E. (1990). The role of sucrose-sensitive neurons in ingestion of sweet stimuli by hamsters.

Physiology & Behavior, 48, 459±466.

Richter, C. P. (1942). Total self-regulatory functions in animals and

human beings. Harvey Lectures, 38, 63±103.

Richter, C. P., & Campbell, K. H. (1940). Taste thresholds and taste

preferences of rats for ve common sugars. Journal of Nutrition, 20,

31±46.

Schi€man, S. S., & Gatlin, C. A. (1993). Sweeteners: state of

knowledge review. Neuroscience and Biobehavioral Reviews, 17,

313±345.

Sicard, P. J. (1982). Hydrogenated glucose syrups, sorbitol, mannitol

and xylitol. In G. G. Birch, & K. J. Parker, Nutritive Sweeteners (pp.

145±170). London: Applied Science Publishers.

Steiner, J. E., & Glaser, D. (1984). Di€erential behavioral responses to

taste stimuli in non-human primates. Journal of Human Evolution,

13, 709±723.

Steiner, J. E., & Glaser, D. (1995). Taste-induced facial expressions in

apes and humans. Human Evolution, 10, 97±105.

Tonosaki, K., & Beidler, L. M. (1989). Sugar best single chorda tym-

pani nerve ber responses to various sugar stimuli in rat and ham-

ster. Comparative Biochemistry and Physiology, 94A, 603±605.

Verkade, P. E., Van Dijk, C. P., & Meerburg, W. (1946). Researches

on the alkoxy-amino-nitrobenzenes. Recueil des Travaux Chimiques

des Pays-Bas, 65, 346±360.

Wang, Y.-M., & van Eys, J. (1981). Nutritional signicance of fructose

and sugar alcohols. Annual Review of Nutrition, 1, 437±475.

we are appreciative of your e-mail responses dated June 5 and June 12 to our request for an 8% red