A bio-cultural approach to the study of food choice: The contribution of taste genetics, population and culture

Accepted Manuscript A bio-cultural approach to the study of food choice: The contribution of taste genetics, population and culture Davide S. Risso, Cristina Giuliani, Marco Antinucci, Gabriella Morini, Paolo Garagnani, Sergio Tofanelli, Donata Luiselli PII:

S0195-6663(16)30516-5

DOI:

10.1016/j.appet.2017.03.046

Reference:

APPET 3407

To appear in:

Appetite

Received Date: 4 October 2016 Revised Date:

23 March 2017

Accepted Date: 30 March 2017

Please cite this article as: Risso D.S., Giuliani C., Antinucci M., Morini G., Garagnani P., Tofanelli S. & Luiselli D., A bio-cultural approach to the study of food choice: The contribution of taste genetics, population and culture, Appetite (2017), doi: 10.1016/j.appet.2017.03.046. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

1

A bio-cultural approach to the study of food choice: the contribution of taste genetics,

2

population and culture

3

Davide S. Rissoa-b*, Cristina Giulianib*, Marco Antinuccic, Gabriella Morinid, Paolo Garagnanie,

5

Sergio Tofanellic^, Donata Luisellib^

RI PT

4

6 7

a

8

Washington 98195, USA

9

b

SC

Department of Genome Sciences, University of Washington School of Medicine, Seattle,

Department of Biological, Geological and Environmental Sciences (BiGeA), Laboratory of

Molecular Anthropology and Centre for Genome Biology, University of Bologna, via Selmi 3,

11

40126 Bologna, Italy

12

c

Department of Biology, University of Pisa, Via Ghini 13, 56126 Pisa, Italy

13

d

University of Gastronomic Sciences, Piazza Vittorio Emanuele 9, Bra, Pollenzo 12042, CN, Italy

14

e

Department of Experimental, Diagnostic and Specialty Medicine (DIMES), University of

15

Bologna, Via San Giacomo 12, 40126 Bologna, Italy

M AN U

10

TE D

16

*, ^ These authors contributed equally

18

* Corresponding authors: [email protected], [email protected]

20 21 22 23 24 25 26

AC C

19

EP

17

27 28 29

1

ACCEPTED MANUSCRIPT

30 Abstract

32

The study of food choice, one of the most complex human traits, requires an integrated approach that

33

takes into account environmental, socio-cultural and biological diversity. We recruited 183 volunteers

34

from four geo-linguistic groups and highly diversified in terms of both genetic background and food

35

habits from whom we collected genotypes and phenotypes tightly linked to taste perception. We

36

confirmed previous genetic associations, in particular with stevioside perception, and noted significant

37

differences in food consumption: in particular, broccoli, mustard and beer consumption scores were

38

significantly higher (Adjusted P=0.02, Adjusted P<0.0001 and Adjusted P=0.01, respectively) in North

39

Europeans, when compared to the other groups. Licorice and Parmesan cheese showed lower

40

consumption and liking scores in the Sri Lankan group (Adjusted P=0.001 and Adjusted P<0.001,

41

respectively). We also highlighted how rs860170 (TAS2R16) strongly differentiated populations and was

42

associated to salicin bitterness perception. Identifying genetic variants on chemosensory receptors that

43

vary across populations and show associations with taste perception and food habits represents a step

44

towards a better comprehension of this complex trait, aimed at improving the individual health status.

45

This is the first study that concurrently explores the contribution of genetics, population diversity and

46

cultural aspects in taste perception and food consumption.

47 48 49

TE D

M AN U

SC

RI PT

31

Keywords

51

Food choice, Nutrition, Population Diversity, Taste, Genetics

53 54 55 56 57

AC C

52

EP

50

58 59 60 61 2

ACCEPTED MANUSCRIPT

1. Introduction

63

Food choice is a highly complex human trait whose study requires an integrated multidisciplinary

64

approach (Anon 2000), which takes into account environmental, socio-cultural and biological factors

65

(Rozin 1982; Turner et al. 2013; Armelagos 2014; Pieroni et al. 2016). Food selection allows humans to

66

fulfill the vital function of nutrition, constituting the deepest connection with the environment, and

67

being a relevant factor to define human communities. During the course of human evolution the sense

68

of taste, together with the other chemical senses smell and chemesthesis, has played a fundamental

69

role in food choice, by ensuring an efficient discrimination between edible sources of nutrients and

70

potentially toxic substances (Glendinning 1994; Breslin 2013). Plant secondary metabolites trigger

71

environmental adaptation, providing defense mechanisms against pathogens and making plants

72

unpalatable to predators. Phenols, flavonoids, terpenes, alkaloids and glucosinolates are in fact bitter or

73

irritant plant compounds: for this reason, variants at genes encoding taste receptors have undergone

74

adaptive changes in relation to eating habits (Fischer et al. 2005; Campbell et al. 2012; Risso et al. 2014a;

75

Perry et al. 2015). Most phytochemicals also exhibit a wide array of biological properties (Kabera et al.

76

2014; Wink 2015) that humans have learnt to exploit for their therapeutic effects (Petrovska 2012). This

77

has added interest to the study of the genetic bases of taste preferences and health status in relation to

78

the introduction of phytochemicals in the diet (Tepper 1998; Dinehart et al. 2006; des Gachons et al.

79

2009; Mennella 2014; Mennella et al. 2015). TAS2R genes, which codify for bitter taster receptors, are

80

also expressed in the gastrointestinal tract and in other extra-oral tissues such as gut, lungs and testis

81

(Finger and Kinnamon 2011), where they modulate systemic functions of tastants either endogenously

82

produced or contained in food (Santa-Cruz Calvo et al. 2015; Clark et al. 2015; Shaik et al. 2016). The

83

most studied taste-related gene is TAS2R38, which encodes the bitter taste receptor mediating the

84

ability to be a “taster” (PAV haplotype) or a “non-taster” (AVI haplotype) for different bitter compounds,

85

including the well-known phenylthiocarbamide (PTC) and 6-n-propylthiouracil (PROP) (Kim et al. 2003,

86

Boxer and Garneau 2015). Correlations have been identified between PROP/PTC taster status and

87

dietary intake (Turnbull and Matisoo-Smith 2002; Tepper 2008; Feeney et al. 2011). This, in addition to

88

the particular worldwide distribution of TAS2R38 haplotypes, has suggested that natural selection may

89

have been acted on this gene, since the earlier stages of human evolution (Wooding et al. 2004;

90

Campbell et al. 2012; Risso et al. 2016).

91

Nonetheless, correlations between genetic variations and food preferences have been recognized hard

92

to detect, since each food is constituted by several different chemical compounds that may be able to

93

activate different receptors (Meyerhof et al. 2010; Roudnitzky et al. 2011, 2015). In addition, food

AC C

EP

TE D

M AN U

SC

RI PT

62

3

ACCEPTED MANUSCRIPT

choice is the result of the integration with other sensory inputs in the brain, and it is further complicated

95

by different cultural backgrounds such as learning, memory and emotion (Bertino et al. 1983; Shepherd

96

2012; Williams et al. 2016). For these reasons, food choice and habits have been recently explored

97

under a different perspective, where the co-variation of taste phenotypes and genotypes is addressed

98

within the context of population diversity in eating behaviors (Kumanyika 2008; Robino et al. 2014;

99

Pirastu et al. 2015). Under this approach, inter-populations variability is taken into account when

RI PT

94

performing genotype-phenotype association studies investigating the relationships between SNPs,

101

structural variations, and taste perception (Campbell et al. 2012, 2014; Allen et al. 2013; Roudnitzky et

102

al. 2015, 2016).

103

In this explorative study we explored an integrated approach and recruited a total number of 183

104

volunteers belonging to four different geographical regions and highly diversified in terms of both

105

genetic background and food habits: Italy, North Europe, Maghreb and Sri Lanka. After collecting bitter,

106

sweet and umami taste-phenotypes using both natural and synthetic compounds, a food habits

107

questionnaire on common bitter, sweet and umami-tasting foods was presented to the participants. A

108

panel of 37 SNPs located in 14 genes involved in taste perception was also analyzed in order to explore

109

potential associations between genetic variants, food habits, taste perception and population

110

differences.

111

TE D

M AN U

SC

100

2. Materials and Methods

113

2.1. Studied population

114

An overall number of 183 individuals were recruited and enrolled in the study. Subjects (81 females and

115

102 males with an average age of 42.71 ± 15.89) did not report any food allergies, were not following

116

any prescribed diet or using drugs that might interfere with taste perception. Most of the participants

117

(N=111) were Italians, with the remaining subjects coming from the Maghreb region (N=18), Sri Lanka

118

(N=26) and Northern Europe (N=28) but recruited in Italy. Detailed information of samples included in

119

the study is shown in Table 1. A written informed consent was obtained from all the volunteers and all

120

the experimental protocols were in accordance with the ethical standards of the Ethics Committee of

121

the University of Bologna and with the Helsinki Declaration of 1975, as revised in 2000.

AC C

EP

112

122 123 4

ACCEPTED MANUSCRIPT

124 125 Table 1. Characteristics of the study participants in the four analyzed populations.

127 128

B.M.I. Body mass index.

Population

N (males/females)

Mean age (SD)

Italy

111 (55/56)

47.64 (16.06)

Maghreb

18 (18/0)

38.56 (13.80)

N. Europe

28 (6/22)

21.50 (6.35)

21.50 (2.93)

Sri Lanka

26 (23/3)

38.59 (13.12)

25.43 (2.34)

All

183 (102/81)

42.71 (15.89)

24.14 (3.47)

131 132 133

24.27 (3.52)

25.48 (3.31)

SC

130

Mean B.M.I. (SD)

M AN U

129

RI PT

126

134 2.2 DNA collection

136

Saliva samples were collected from participants using Oragene saliva collection kits (Genotek Inc.

137

Kanata, Ontario, Canada). DNA was extracted according to the manufacturer’s protocol and checked for

138

quantity/quality by spectrophotometric analyses (GeneQuantTMRNA/DNA Calculator, Amersham

139

Biosciences, UK). We genotyped 37 SNPs selected from a panel of 14 taste-related genes (Table 2) using

140

the SequenomMassARRAY technology (Sequenom, San Diego, CA, USA).

141

PCR fragments were analyzed by MALDI-TOF mass spectrometry (Gabriel et al. 2009) and spectrograms

142

were checked individually, in order to evaluate the presence of calling errors. As additional quality

143

control, we checked that no significant deviations from Hardy-Weinberg equilibrium (p-value<0.001)

144

were present in the analyzed SNPs and that their call rate was >0.95. Primer sequences are shown in

145

Supplementary Table 1.

147 148

EP

AC C

146

TE D

135

149 150

5

ACCEPTED MANUSCRIPT

151 152 Table 2. List of the selected genes and single nucleotide polymorphisms.

Gene

Product function

SNPs

Chromosome

TAS2R1

Bitter receptor

rs2234233

5

Bitter receptor

rs2234001

7

Bitter receptor

rs11610105, rs3741843, rs7138535

12

TAS2R16

Bitter receptor

rs860170

7

TAS2R38

Bitter receptor

rs10246939, rs1726866, rs713598

7

TAS2R50

Bitter receptor

rs10772397, rs1376251

CA6

Carbonate dehydratase

rs2274333

GNAT3

Bitter, sweet and umami taste transduction

rs6467192, rs6467217, rs1524600, rs1107657, rs940541

TRPM5

Taste transduction

141478574

11088981, 11091432, 11091693 122635024

141672604, 141672705,141673345 11138683, 11138852

1

9017204

7

80107798, 80138178, 80138303, 80150018, 80150594

M AN U

12

rs2301696, rs800344, rs800345, rs2301698

2426984, 2429653, 2429733, 2437425, 2444094

11

rs2301699, rs800347, rs800348, rs2074234, rs886277 rs7534618, rs4920564, rs35874116, Sweet receptor

TAS1R2

9629529

SC

TAS2R4 TAS2R14

Positions (GRCh37)

RI PT

153

2430597, 2430865, 2432964, 2439767 19170078, 19179104, 19181393

1

rs9701796, rs3935570, rs4073538, rs4920566 Sweet and Umami receptor

rs35424002

GLUT2

Glucose transporter

rs5400

TAS1R1

Umami receptor

rs34160967

mGluR1

Glutamate receptor

TE D

TAS1R3

rs6923492

154 155

19186129, 19167371, 19172341, 19179824

1

1269986

3

170732300

1

6635306

6

146755324

2.3 Food habits questionnaire

157

Participants completed a questionnaire on individual liking and consumption about 12 common foods:

158

seven bitter-tasting (“Broccoli”, “Radicchio”, “Artichoke”, “Arugula”, “Licorice”, “Mustard” and “Beer”),

159

three sweet-tasting (“Date”, “Honey” and “Torrone”) and two umami-tasting foods (“Parmesan cheese”

160

and “Bouillon cube”). Subjects were asked to rate each item on a 3-point liking scale: “Dislike” (score 1),

161

“Sort of” (score 2) and “Like extremely” (score 3). In addition, these items were rated on a 3-points

162

consumption scale: “Yearly” (score 1), “Monthly” (score 2) and “Weekly” (score 3). A “Never tasted”

163

(score 0) option was also included. The liking and consumption scores were then pooled together to

164

have a single 1-to-6 liking + consumption score. Supplementary Table 2 summarizes mean and standard

165

deviation for each item score. Lastly, weight (in kg) and height (in m) were collected in order to calculate

166

the body mass index (BMI).

167

2.4. Assessment of taste phenotypes

AC C

EP

156

6

ACCEPTED MANUSCRIPT

Taste perception phenotypes were collected for the three examined taste qualities (e.g. umami, sweet

169

and bitter). Volunteers were asked to refrain from eating and drinking for at least three hours before the

170

beginning of the session and to rinse their mouth with room temperature deionized water before each

171

tasting step. Subjects were informed that they may receive stimuli eliciting more than one taste quality.

172

They were asked to hold the presented solutions in their mouth for ten seconds and subsequently

173

choose among different tastes such as bitter, sweet, umami and/or tasteless, as well as to rate their

174

perceived intensity separately on multiple Labeled Magnitude Scales (LMSs, Green et al. 1996), one for

175

each taste quality.

176

For the umami taste phenotypes, two cups containing different concentrations (3.13 mM and 12.5 mM,

177

respectively) of monosodium glutamate dissolved in water (MSG, Sigma Aldrich S.r.l.) were presented to

178

the subjects to estimate individual’s MSG thresholds. According to the recognized solution, they were

179

classified in MSG non-sensitives (no recognition of umami taste), medium-sensitives (recognition of the

180

umami taste only in the second tasting) and sensitives (recognition of umami taste in both tastings).

181

Regarding sweet taste, volunteers were asked to give their preference for a series of three cups

182

containing different sucrose (Sigma Aldrich S.r.l) water solutions (200, 400 and 800 mM, respectively) on

183

a seven-point category scale ranging from “Dislike extremely” to “Like extremely” (Supplementary

184

Figure 1). Subjects were then classified in sucrose non-likers, medium-likers and likers, after analysis of

185

the “liking-curve” of different sucrose tastings (Lundgren et al. 1978).

186

Lastly, bitter taste-related phenotypes were collected by quantifying the ability of tasting three different

187

bitter compounds (salicin, PROP and stevioside) using the LMS. A single water solution of stevioside 1.26

188

mM (Nastevia, Stevia Italia s.r.l. Italy) was presented to the subjects. The PROP-taster status was

189

assessed using cotton swabs dipped in 50 mM 6-n-propylthiouracil solution (Sigma Aldrich S.r.l.)

190

modifying the protocol for filter paper discs described in Zhao et al. (2003) as previously shown (Risso et

191

al. 2014b). Cotton swabs dipped in supra-threshold solutions of salicin 15mM (Sigma Aldrich S.r.l) were

192

also given to the participants. LMS scores were then used as quantitative variables for logistic and linear

193

regression analyses.

194

2.5. Genetic and Statistical analyses

195

Arlequin v.3.5 (Excoffier et al. 2010) was used to compute summary statistics such as FST values,

196

estimated heterozygosity (EH), gene diversity, number of polymorphic loci and the analysis of molecular

197

variance (AMOVA). FST values and MAFs were then compared to the ones calculated in population-

198

matching groups of the 1000 Genomes Project (1KGenomes) (1000 Genomes Project Consortium et al.

199

2015) and the Human Genome Diversity Project (HGDP) (Cann et al. 2002). Distributions of food habit

AC C

EP

TE D

M AN U

SC

RI PT

168

7

ACCEPTED MANUSCRIPT

scores and differences in taste perception were evaluated though the analysis of variance (ANOVA)

201

using R statistical analysis software (http://www.R-project.org/). Tukey’s post-hoc tests were run to

202

evaluate differences between means. Since the number of subjects in the Italian group was 4 to 6 times

203

bigger than the other groups, all the statistical analyses were replicated on a subsample of 30 individuals

204

obtained by 1000 bootstrap resampling. Due to the size of the other population samples, we were

205

unable to replicate the genotype-phenotype associations independently in each groups. For this reason,

206

we explored genotype-phenotype associations and interactions between genotypes and food habits

207

scores through logistic regressions in PLINK v.1.07 (Purcell et al. 2007), using sex, age, BMI and ethnicity

208

as covariates in the entire group (N=183). When multiple tests were performed under the same null

209

hypothesis, the statistical significance was also adjusted according to the Bonferroni correction (i.e.

210

Adjusted P= nominal P-value x number of individual tests).

211

3. Results

212

3.1. Genetic variability

213

The analyzed SNPs showed similar diversity patterns across the different groups: the overall pairwise FST

214

values were indeed low (average value of 0.024 with a standard deviation of 0.01), but significant in all

215

the comparisons (Supplementary Table 3 A-B). Average gene diversity (+/- SD), number of polymorphic

216

sites and estimated heterozygosity were indeed similar across different populations (Supplementary

217

Table 4). This means that most of the variance is between individuals within populations (98%) rather

218

than among populations (2%). As shown in Figure 1, the distribution of the calculated FST values

219

highlights the overall low level of differentiation of these SNPs across the studied populations.

220

However, three SNPs, namely rs1376251 (TAS2R50), rs800345 (TRPM5) and rs860170 (TAS2R16),

221

showed modest to intermediate differentiation in our sample, falling in the last 5th percentile of the FST

222

distribution. Table 3 shows the the observed minor allele frequencies at these SNPs and those calculated

223

from public datasets are comparable with TAS2R50 rs1376251 C, TRPM5 rs800345 T and TAS2R16

224

rs860170 G alleles found to be more frequent in populations of Asian ancestry. In the same way, the

225

calculated pairwise FST values better highlighted the different distribution of the analyzed SNPs among

226

our population groups (Supplementary Figure 2).

AC C

EP

TE D

M AN U

SC

RI PT

200

227 228 229

8

ACCEPTED MANUSCRIPT

230 Figure 1. Distribution of the analyzed SNPs according to pairwise FST values (X-axis) and Minor Allele

232

Frequencies (MAFs). Arrows indicate the SNPs showing the highest FST values, falling in the upper 5th

233

percentile of the distribution (dashed horizontal line).

RI PT

231

234 235

SC

236 237

M AN U

238 239 240 241 242

Table 3. Allele frequencies of the highest differentiated SNPs with the highest FST values in our groups

244

(shown in italics) and population-matched groups from the 1000 Genomes project Phase III

245

(1KGenomes) and Human Genome Diversity Project (HGDP).

246

MZB, Mozabite from Algeria (HGDP); EUR, Europeans (1KGenomes); TSI, Tuscans from Italy

247

(1KGenomes); STU, Sri Lankan Tamil from the UK (1K Genomes).

EP

TE D

243

rs860170 (TAS2R16)

rs800345 (TRPM5)

rs1376251 (TAS2R50)

A

G

C

T

C

T

AC C

Population MZB

0.82

0.18

0.87

0.13

0.50

0.50

Maghreb

0.86

0.14

0.83

0.17

0.50

0.50

EUR

0.69

0.31

0.92

0.08

0.64

0.36

N. Europe

0.67

0.33

0.89

0.11

0.43

0.57

TSI

0.69

0.31

0.93

0.07

0.58

0.42

Italy

0.67

0.33

0.92

0.08

0.59

0.41

STU

0.63

0.37

0.70

0.30

0.33

0.67

Sri-Lanka

0.59

0.41

0.65

0.35

0.77

0.23

248 249 9

ACCEPTED MANUSCRIPT

250 3.2. Genotype-phenotype association

252

Three statistically significant associations were found between SNPs and food habit ratings. TAS2R38

253

rs10246939, rs1726866 and rs713598 were found to be associated with Broccoli score (Adjusted P-

254

values=0.02, 0.03 and 0.04 respectively) in all the studied populations. In particular, the taster haplotype

255

(PAV) was inversely correlated to the reported Broccoli score (R=-0.71, P<0.05).

256

PROP perceived bitterness was, as expected, associated with the TAS2R38 PAV taster haplotype in all

257

the studied populations (Adjusted P=1.05E-15). The CCG genotype (PAV haplotype) was in fact more

258

frequent in individuals who rated higher scores of PROP bitter perception (Supplementary Figure 3A).

259

The perceived bitterness of stevioside was positively associated with TAS2R4 rs2234001 G allele. When

260

separating individuals who could perceive stevioside bitterness (N=135) from individuals who couldn’t

261

(N=48), rs2234001 G allele was in fact more abundant in the former (55% vs 34%; OR=2.69, Adjusted P=

262

1.24E-03). In the same way, TAS2R4 rs2234001 CC homozygotes (N= 48) showed lower average levels

263

(30.83 +/- 4.74) of perceived stevioside bitterness when compared to GG homozygotes (49.10 +/- 4.16)

264

(Adjusted P-value= 4.95E-03; Supplementary Figure 3B). The perception of salicin bitterness was

265

associated with TAS2R16 rs860170 A allele (Adjusted P=0.01). At the genotypic level, a notable trend can

266

be indeed observed, with average salicin bitterness levels decreasing from 55.15 +/- 25.45 (AA

267

genotype) to 39.37 +/- 24.07 (GG genotype) (Adjusted P=0.01; Supplementary Figure 3C). Lastly, we

268

failed to detect any evidence of correlation between sweet and umami taste-related gene

269

polymorphisms and the perception of sucrose and umami solutions (all P’s>0.05) as well as between

270

CA6 rs2274333 and perceived PROP bitterness (P>0.05). A summary of the results is reported in

271

Supplementary Figure 4.

272

3.3 Population-based food preferences

273

Most of the liking/consumption scores showed a similar distribution across populations. However,

274

significant different food habit scores were found. Table 4 recapitulates these findings, showing how the

275

score of three bitter foods (broccoli, mustard and beer) was significantly higher in North Europeans

276

(5.18 +/- 1.04, 4.79 +/- 1.66 and 4.54 +/- 1.59, respectively) when compared to Maghrebis (4.50 +/- 1.46,

277

2.33 +/- 1.72 and 3.17 +/- 2.31, respectively), Lankans (4.31 +/- 1.52, 2.58 +/- 1.83 and 2.65 +/- 1.85,

278

respectively) and Italians (4.08 +/- 1.40, 2.85 +/- 1.65 and 4.11 +/- 1.68, respectively) (Adjusted P=0.02,

279

Adjusted P=7.40E-06 and Adjusted P=0.01, respectively). In addition, one bitter item (licorice) showed

280

lower ratings in the Sri Lankan group (1.77 +/- 1.30) with respect to Maghrebis (3.33 +/- 2.16), North

281

Europeans (3.89 +/- 1.71) and Italians (3.44 +/- 1.47) (Adjusted P=0.001). Parmesan cheese showed

AC C

EP

TE D

M AN U

SC

RI PT

251

10

ACCEPTED MANUSCRIPT

lower scores in Lankans (3.91 +/- 1.77) and Maghrebis (3.72 +/- 1.81), with respect to Italians (5.24 +/-

283

1.19) and North Europeans (5.25 +/- 0.80) (Adjusted P=3.00E-03). No significant correlations (All P-

284

values > 0.05) between BMI and food preferences were observed (Supplementary Figure 5).

285

Table 4. Liking/consumption scores showing significant differences across population samples.

RI PT

282

All (SD)

Italians (SD)

Maghreb (SD)

North-Europe (SD)

Sri Lanka (SD)

Adjusted P-value

Broccoli

4.32 (1.41)

4.08 (1.40)

4.5 (1.46)

5.18 (1.04)

4.31 (1.52)

0.02

Beer

3.87 (1.84)

4.11 (1.68)

3.17 (2.31)

4.54 (1.59)

2.65 (1.85)

0.01

Licorice

3.26 (1.68)

3.44 (1.47)

3.33 (2.16)

3.89 (1.71)

1.77 (1.30)

0.001

Mustard

3.06 (1.79)

2.85 (1.65)

2.33 (1.72)

4.79 (1.66)

2.58 (1.83)

7.40E-06

Parmesan cheese

4.92 (1.42)

5.24 (1.19)

3.72 (1.81)

5.25 (0.80)

3.91 (1.77)

3.00E-03

M AN U

286

SC

Food

3.4. Population-based taste phenotypes

288

The distribution of MSG perception varied across different groups. Maghrebis and Lankans, in fact,

289

showed a lower percentage of MSG sensitives (6% and 8%, respectively) when compared to Italians and

290

North Europeans (44% and 64%, respectively; P=1.72E-08). Conversely, a higher percentage of MSG non-

291

sensitives was noted in Maghrebis and Lankans (39% and 46%, respectively), in comparison to Italians

292

and North Europeans (13% and 7%, respectively; P=1.72E-08) (Figure 2A).

293

Regarding sucrose preference, we found more non-likers in the North European group (71%), in

294

comparison to Maghrebis (44%), Italians (36%), and Lankans (27%) (P=0.01; Figure 2B).

295

Bitter taste perception varied across populations as well: the perceived bitterness of stevioside was in

296

fact higher in the North European group (60.54 +/- 22.11), with lower levels in Lankans (39.62 +/- 21.53),

297

Italians (38.60 +/- 25.22) and Maghrebis (14.67 +/- 20.67) (P=0.001). However, the perceived levels of

298

salicin and PROP bitterness did not vary across different groups (P=0.61 and P=0.38, respectively).

300 301

EP

AC C

299

TE D

287

302 303

11

ACCEPTED MANUSCRIPT

304

Figure 2. Monosodium-Glutamate perception (A) and sucrose preference (B) in the examined

305

populations. Standard deviation values in the Italian group are based on 1000 bootstrap replicates of 30

306

individuals. * indicates statistical significance.

RI PT

307 308 309

SC

310 311

M AN U

312 313 314 315

317 318 319

EP

TE D

316

4. Discussion

321

In the present study, we propose an integrated approach where the variance of genotypes and

322

phenotypes of human taste chemoreception is partitioned into geo-linguistic groups. Genotypes of 37

323

taste-related loci showed that three SNPs at TAS2R50 (rs1376251), TRPM5 (rs800345) and TAS2R16

324

(rs860170) genes had particular high levels of differentiation among our studied populations. The MAFs

325

of these SNPs were indeed found to be much higher in populations of Asian ancestry. We did not find

326

any correlations between two of these SNPs and the investigated phenotypes, neither biological nor

327

cultural, although a recent study showed that TAS2R50 rs1376251 is associated with a decreased

328

dietary fiber and vegetable intake (Schembre et al. 2013). We cannot exclude the possibility that the

AC C

320

12

ACCEPTED MANUSCRIPT

lack of association can be linked to the limited number of samples considered in this study, highlighting

330

the importance of further investigations on potential associations between these two SNPs and taste

331

phenotypes.

332

Other studies reported an association between this SNP, myocardial infarction and coronary heart

333

disease, suggesting a possible relationship between this polymorphism and unbalanced food habits

334

(Shiffman et al. 2008; Akao et al. 2012). These correlations are however still debated, since an

335

independent study of 3657 patients with myocardial infarction and 1211 controls failed to replicate this

336

association (Koch et al. 2011). TRPM5 rs800345 was also found to be significantly associated with pre-

337

diabetic phenotypes (Ketterer et al. 2011), indicating a potential link between genetic variability across

338

populations and glucose tolerance.

339

When examining potential associations between taste genetic variants and food habits, we found a

340

significant association between TAS2R38 SNPs and broccoli consumption, in agreement with previous

341

results (Niewind et al. 1988; Jerzsa-Latta et al. 1990) and, more generally, with an inverse relationship

342

between TAS2R38-mediated sensitivity to bitterness and vegetable intake (Dinehart et al. 2006; Duffy et

343

al. 2010, Colares-Bento et al. 2012). As expected from the literature, we found that higher scores of

344

PROP bitterness perception were associated with the TAS2R38 PAV taster haplotype (Kim et al. 2003;

345

Duffy et al. 2004; Allen et al. 2013; Fischer et al. 2014). However, we could not confirm the association

346

between CA6 rs2274333 and perceived PROP bitterness (Padiglia et al. 2010), thus supporting more

347

recent studies (Feeney and Hayes 2014; Barbarossa et al. 2015). The perceived bitterness of stevioside

348

was positively associated with TAS2R4 rs2234001 G allele, replicating previous results from our group on

349

a lower sample (Risso et al. 2014b) and in part confirming data published by a recent study (Hayes et al.

350

2011). Hayes and colleagues showed that individuals carrying one or two copies of a specific more

351

responsive haplotype containing this SNP (i.e. TGAG: rs765007, rs224001, rs2234012, rs2227264)

352

experienced twice as much coffee bitterness when compared to individuals homozygous for the less

353

responsive haplotype (CCGT). However, it is important to note that this less responsive haplotype was

354

also found to be associated with higher bitterness ratings of both ethanol and capsaicin (Nolden et al.

355

2016), indicating the need to replicate these findings in order to elucidate the role of this SNP/haplotype

356

in bitterness perception, in particular considering genetic ancestry, ethnicity, gender and age.

357

On the contrary, the association with TAS2R14 rs3741843 reported in the same study was not

358

confirmed here (Risso et al. 2014b). The bitterness elicited by stevioside was far higher in North

AC C

EP

TE D

M AN U

SC

RI PT

329

13

ACCEPTED MANUSCRIPT

Europeans than in the Maghreb population, alerting caution towards a non population-oriented

360

marketing of Stevia rebaudiana extracts.

361

More generally, it is worth noticing that the peaks of frequency difference among groups and the

362

associations with taste phenotypes involved different SNPs, suggesting that present-day genetic

363

variance was not caused by unbalanced positive selection upon the investigated phenotype patterns.

364

The only exception is rs860170 (TAS2R16), with its ancestral allele (A) being highly predominant in north

365

African individuals and associated with salicin bitterness perception. This result is in line with those of

366

Soranzo and colleagues (2005) detecting signatures of positive selection at TAS2R16 surveying the

367

geographic pattern of its variants. Interestingly, an haplotype including TAS2R16 rs860170-A was found

368

associated with longevity in humans (Campa et al. 2012), thus opening new prospects on the interplay

369

between human evolution, taste perception and health. In this paper the authors have hypothesized

370

that salicin could have similar effects to aspirin (acetylsalicylic acid), acting as an anti-inflammatory

371

agent and therefore favoring healthy ageing. This consideration is also important for human

372

evolutionary studies and we speculate that the rs860170 derived allele could increase the tolerance

373

level for salicin bitterness and other similar substances used as analgesic, anti-inflammatory, and

374

antipyretic. Inflammation and foods that modulate the release of pro-inflammatory mediators might be

375

considered the mechanism crucial for both human evolution and human pathology (chronic

376

inflammation) (Franceschi and Campisi 2014; Straub 2014). The link between inflammation, ageing and

377

nutrition has been addressed (Szarc vel Szic et al. 2015) and most plant-derived dietary phytochemicals

378

able to modulate oxidative stress and inflammatory signaling through epigenetic mechanism are also

379

taste-active. However, the role and contribution of taste perception in this process still needs to be

380

elucidated.

381

In this paper, we reported for the first time population-related differences in taste perception and

382

eating habits in native groups with well-defined cultural and geographic features.

383

Broccoli, mustard and beer liking/consumption scores were significantly higher in North Europeans

384

when compared to the other groups. In addition, licorice and Parmesan cheese showed lower

385

consumption and liking scores in the Sri Lankan group. Similarly, Lankans showed a lower percentage of

386

MSG sensitives: this is in agreement with previous observations (Singh et al. 2010) that studied

387

variations in umami taste perception in Europeans. Lastly, we here reported more sucrose non-likers in

388

Northern Europe, in comparison to African or Asian populations, supporting previous studies observing

389

that life experiences may contribute to ethnic differences in sweet preferences (Pepino and Mennella

390

2005). Other authors reported in fact that cultural practices common in Africa, such as feeding sugar

AC C

EP

TE D

M AN U

SC

RI PT

359

14

ACCEPTED MANUSCRIPT

water to infants, resulted in an increased sweet preferences during the second year of life of the child

392

(Beauchamp and Moran 1984). This indicates that traditions, experience and historical background could

393

have a significant impact on food preferences from the first period of life (Kobayashi and Kennedy 2002,

394

Mennella 2014). This study, in accordance with some previous ones, suggests that taste responsiveness

395

could be related to the ethnic background (Ahn et al. 2011; Williams et al. 2016). However, an adequate

396

classification from an ethno-anthropological point of view is still needed in order to avoid misleading

397

interpretations.

RI PT

391

398 5. Conclusion

400

Identifying genetic variants on chemosensory receptors that vary across populations and show

401

associations with taste perception and food habits is of high importance. Although based on a small

402

number of samples, this exploratory and multidisciplinary study constitutes an innovation in the fields,

403

since it represents an attempt to explore and combine different aspects (genetics, population variability

404

and phenotypic associations) that together might have an impact in determining individuals’ food

405

preferences.

406

Conflict of Interest

407

The authors declare no conflicts of interest.

408

Source of funding

409

This work was supported by the Fondazione Del Monte ‘‘Taste Genetic Geography’’ Grant to DL and by

410

University of Pisa ex60% grants to ST.

411

Acknowledgments

412

The authors would like to thank the ‘‘Slow Food Communities,’’ Silvano Zaccone, Luisella Verderi and all

413

the people who agreed to participate to this study.

414

References

415 416

1000 Genomes Project Consortium et al. 2015. A global reference for human genetic variation. Nature. 2015;526(7571):68-74

417

Ahn, Y.-Y. et al. 2011. Flavor network and the principles of food pairing. Scientific reports, 1, p.196

418 419

Allen AL, McGeary JE, Knopik VS, Hayes JE. 2013. Bitterness of the non-nutritive sweetener acesulfame potassium varies with polymorphisms in TAS2R9 and TAS2R31. Chem Senses. 38(5):379-89

AC C

EP

TE D

M AN U

SC

399

15

ACCEPTED MANUSCRIPT

Akao H et al. 2012. KIF6, LPA, TAS2R50, and VAMP8 genetic variation, low density lipoprotein cholesterol lowering response to pravastatin, and heart disease risk reduction in the elderly. Atherosclerosis. 2012 Feb;220(2):456-62

423 424

Allen AL, McGeary JE, Knopik VS, Hayes JE. 2013. Bitterness of the non-nutritive acesulfame potassium varies with polymorphisms in TAS2R9 and TAS2R31. Chem Senses 38(5):379–389

425 426

Armelagos, G.J. 2014. Brain evolution, the determinates of food choice, and the omnivore's dilemma. Critical Reviews in Food Science and Nutrition, 54, pp.1330-1341

427 428 429

Barbarossa, I.T. et al. 2015. The gustin (CA6) gene polymorphism, rs2274333 (A/G), is associated with fungiform papilla density, whereas PROP bitterness is mostly due to TAS2R38 in an ethnically-mixed population. Physiology & behavior, 138, pp.6–12

430 431

Beauchamp, G.K. and Moran, M. 1984. Acceptance of sweet and salty tastes in 2-year-old children. Appetite, 5(4), pp.291–305

432 433

Bertino M. Beauchamp G.K. Jen K.C. 1983. Rated taste perception in two cultural groups. Chemical Senses. 8, pp.3-15

434 435

Boxer EE, Garneau NL. Rare haplotypes of the gene TAS2R38 confer bitter taste sensitivity in humans. Springerplus. 17;4:505

436 437

Breslin, P.A.S. 2013. An Evolutionary Perspective on Food and Human Taste. Current Biology, 23(9), pp.R409–R418

438

Campa, D. et al. 2012. Bitter taste receptor polymorphisms and human aging. PloS one, 7(11), p.e45232

439 440

Campbell, M.C. et al. 2012. Evolution of functionally diverse alleles associated with PTC bitter taste sensitivity in Africa. Molecular biology and evolution, 29(4), pp.1141–1153

441

Cann, HM. et al. 2002. A human genome diversity cell line panel. Science. 2002;296(5566):261-2

442 443

Clark, A.A. Dotson, C.D. Elson, A.E.T. Voigt, A. Boehm, U. Meyerhof, W. Steinle N.I. Munger, S.D. 2015. TAS2R bitter taste receptors regulate thyroid function. FASEB J. Jan;29(1):164-72

444 445

Colares-Bento FC et al. 2012. Implication of the G145C polymorphism (rs713598) of the TAS2r38 gene on food consumption by Brazilian older women. Arch Gerontol Geriatr. 54(2):e13-8

446 447

des Gachons, C.P. Beauchamp, G.K. and Breslin, P.A. 2009. The genetics of bitterness and pungency. Detection and its impact on phytonutrient evaluation. Ann N Y Acad Sci. Jul;1170:140-4

448 449

Dinehart ME, Hayes JE, Bartoshuk LM, Lanier SL, Duffy VB. 2006. Bitter taste markers explain variability in vegetable sweetness, bitterness, and intake. Physiol Behav. 28;87(2):304-13

450 451

Duffy, V.B. et al. 2004. Bitter receptor gene (TAS2R38), 6-n-propylthiouracil (PROP) bitterness and alcohol intake. Alcoholism, clinical and experimental research, 28(11), pp.1629–1637

AC C

EP

TE D

M AN U

SC

RI PT

420 421 422

16

ACCEPTED MANUSCRIPT

Duffy et al. 2010. Vegetable Intake in College-Aged Adults Is Explained by Oral Sensory Phenotypes and TAS2R38 Genotype. Chemosens Percept. 1;3(3-4):137-148

454 455

Excoffier L, Laval G, Schneider S. 2007. Arlequin (version 3.0): an integrated software package for population genetics data analysis. Evol Bioinform Online 23(1):47–50

456 457

Feeney E, O'Brien S, Scannell A, Markey A, Gibney ER. 2011. Genetic variation in taste perception: does it have a role in healthy eating? Proc Nutr Soc. 70(1):135-43

458 459

Feeney, E.L. and Hayes, J.E. 2014. Exploring associations between taste perception, oral anatomy and polymorphisms in the carbonic anhydrase (gustin) gene CA6. Physiology & behavior, 128, pp.148–154

460

Finger T.E. and Kinnamon S.C. 2011. Taste isn’t just for taste buds anymore. F1000 Biol Rep. 3: 20

461 462

Fischer, A. et al. 2005. Evolution of bitter taste receptors in humans and apes. Molecular biology and evolution, 22(3), pp.432–436

463 464

Fischer, M.E. et al. 2014. The associations between 6-n-propylthiouracil (PROP) intensity and taste intensities differ by TAS2R38 haplotype. Journal of nutrigenetics and nutrigenomics, 7(3), pp.143–152

465 466 467 468 469

Franceschi, C. and Campisi, J. 2014. Chronic inflammation (inflammaging) and its potential contribution to age-associated diseases. J Gerontol A Biol Sci Med Sci. Jun;69 Suppl 1:S4-9

470

Glendinning JI. 1994. Physiol Behav. Dec;56(6):1217-27

471 472

Green BG, Dalton P, Cowart B, Shaffer G, Rankin K, Higgins J. 1996. Evaluating the 'Labeled Magnitude Scale' for measuring sensations of taste and smell. Chem Senses. 21(3):323-34

473 474 475

Hayes JE, Wallace MR, Knopik VS, Herbstman DM, Bartoshuk LM, Duffy VB. 2011. Allelic variation in TAS2R bitter receptor genes associates with variation in sensations from and ingestive behaviors toward common bitter beverages in adults. Chem Senses. 36(3):311-9

476 477

Jerzsa-Latta, M. Krondl, M. and Coleman, P. 1990. Use and perceived attributes of cruciferous vegetables in terms of genetically-mediated taste sensitivity. Appetite, 15(2), pp.127–134

478 479 480

Kabera, J.N. Semana, E. Mussa, A.R. and He, X. 2014. Plant secondary metabolites: biosynthesis, classification, function and pharmacological properties. Journal of Pharmacy and Pharmacology 2, pp.377-392

481 482

Ketterer, C. et al. 2011. Genetic variation within the TRPM5 locus associates with prediabetic phenotypes in subjects at increased risk for type 2 diabetes. Metabolism. 2011;60(9):1325-33

483 484

Kim, U. et al. 2003. Positional cloning of the human quantitative trait locus underlying taste sensitivity to phenylthiocarbamide. Science (New York, N.Y.), 299(5610), pp.1221–1225

M AN U

SC

RI PT

452 453

AC C

EP

TE D

Gabriel S, Ziaugra L, Tabbaa D. 2009. SNP genotyping using the Sequenom MassARRAY iPLEX platform. Curr Protoc Hum Genet. doi:10.1002/0471142905.hg0212s60

17

ACCEPTED MANUSCRIPT

Kobayashi, C. and Kennedy, L.M. 2002. Experience-induced changes in taste identification of monosodium glutamate. Physiology & behavior, 75(1–2), pp.57–63

487 488

Koch, W. et al. 2011. Variations of specific non-candidate genes and risk of myocardial infarction: a replication study. International journal of cardiology, 147(1), pp.38–41

489 490

Kumanyika, S.K. 2008. Environmental influences on childhood obesity: ethnic and cultural influences in context. Physiology & behavior, 94(1), pp.61–70

491 492

Lundgren, B et al., 1978. Taste discrimination vs hedonic response to sucrose in coffee beverage. An interlaboratory study. Chemical Senses, 3 (3): 249-265

493 494

Mennella J.A. 2014. Ontogeny of taste preferences: basic biology and implications for health. Am J Clin Nutr 99(suppl), pp.704S–711S

495 496

Mennella, J.A. Bobowski, N.K. 2015. The sweetness and bitterness of childhood: Insights from basic research on taste preferences. Physiol Behav. Dec 1;152(Pt B):502-7

497 498

Meyerhof, W. et al. 2010. The molecular receptive ranges of human TAS2R bitter taste receptors. Chemical senses, 35(2), pp.157–170

499 500

Niewind, A. Krondl, M. and Shrott, M. 1988. Genetic influences on the selection of brassica vegetables by elderly individuals. Nutrition Research, 8(1), pp.13–20

501 502

Nolden AA, McGeary JE, Hayes JE. Differential bitterness in capsaicin, piperine, and ethanol associates with polymorphisms in multiple bitter taste receptor genes. Physiol Behav. 2016 Mar 15;156:117-27

503 504 505

Padiglia et al. 2010. Sensitivity to 6-n-propylthiouracil is associated with gustin (carbonic anhydrase VI) gene polymorphism, salivary zinc, and body mass index in humans. Am J Clin Nutr. 2010 Sep;92(3):539-45

506 507

Pepino, M.Y. and Mennella, J.A. 2005. Factors contributing to individual differences in sucrose preference. Chemical senses, 30 Suppl 1, pp.i319-320

508 509

Perry, G.H. et al. 2015. Insights into hominin phenotypic and dietary evolution from ancient DNA sequence data. Journal of human evolution, 79, pp.55–63

510 511

Petrovska, B. B. 2012. Historical Review of Medicinal Plants’ Usage.” Pharmacognosy Reviews 6 (11), pp.1-5

512 513

Pieroni, A. Pawera L. and Shah G.M. 2016. Gastronomic Ethnobiology. In: Albuquerque, U.P. Alves, R.R.N. (Eds.) Introduction to Ethnobiology, 2016, Springer, New York, ISBN 978-3-319-28153-7, pp. 53-62

514 515 516

Pirastu, N. et al. 2015. Genome-wide association analysis on five isolated populations identifies variants of the HLA-DOA gene associated with white wine liking. European journal of human genetics: EJHG, 23(12), pp.1717–1722

517 518

Purcell, S. et al. 2007. PLINK: a tool set for whole-genome association and population-based linkage analyses. American journal of human genetics, 81(3), pp.559–575

AC C

EP

TE D

M AN U

SC

RI PT

485 486

18

ACCEPTED MANUSCRIPT

Risso, D. Tofanelli, S. et al. 2014a. Genetic variation in taste receptor pseudogenes provides evidence for a dynamic role in human evolution. BMC evolutionary biology, 14, p.198

521 522

Risso, D. Morini, G. et al. 2014b. Genetic signature of differential sensitivity to stevioside in the Italian population. Genes & Nutrition, 9(3)

523 524

Risso, D. et al. 2016. Global diversity in the TAS2R38 bitter taste receptor: revisiting a classic evolutionary PROPosal. Sci Rep. 6: 25506

525 526

Robino, A. et al. 2014. A population-based approach to study the impact of PROP perception on food liking in populations along the Silk Road. PloS one, 9(3), p.e91716

527 528

Roudnitzky, N. et al. 2011. Genomic, genetic and functional dissection of bitter taste responses to artificial sweeteners. Human molecular genetics, 20(17), pp.3437–3449

529 530

Roudnitzky, N. et al. 2015. Receptor Polymorphism and Genomic Structure Interact to Shape Bitter Taste Perception T. Gojobori, ed. PLOS Genetics, 11(9), p.e1005530

531 532

Roudnitzky N, et al. 2016. Copy Number Variation in TAS2R Bitter Taste Receptor Genes: Structure, Origin, and Population Genetics. Chem Senses. pii: bjw067

533 534 535

Rozin P. 1982. Human food selection: the interaction of biology, culture and individual experience, in L.M. Barker (Ed) The Psychobiology of Human Food Selection, AVI Publishing, Westport, CT, pp.225269

536 537

Santa-Cruz Calvo, S. and Egan, J. M. 2015. Endocrinology of taste. Nat. Rev. Endocrinol. 2015 Apr;11(4):213-27)

538 539 540

Shaik FA, Singh N, Arakawa M, Duan K, Bhullar RP, Chelikani P. 2016. Bitter taste receptors: extraoral roles in pathophysiology. The International Journal of Biochemistry & Cell Biology doi:10.1016/j.biocel.2016.03.011

541 542

Schembre, S.M. et al. 2013. Variations in bitter-taste receptor genes, dietary intake, and colorectal adenoma risk. Nutrition and cancer, 65(7), pp.982–990

543

Shepherd G. Neurogastronomy. New York: Columbia University Press; 2012

544 545

Shiffman, D. et al. 2008. Association of gene variants with incident myocardial infarction in the Cardiovascular Health Study. Arteriosclerosis, thrombosis, and vascular biology, 28(1), pp.173–179

546 547

Singh, P.B. Schuster, B. and Seo, H.-S. 2010. Variation in umami taste perception in the German and Norwegian population. European journal of clinical nutrition, 64(10), pp.1248–1250

548 549 550

Straub, R.H. 2014. TRPV1, TRPA1, and TRPM8 channels in inflammation, energy redirection, and water retention: role in chronic inflammatory diseases with an evolutionary perspective. J Mol Med, 92, pp.925–937

551 552

Tepper BJ. 6-n-Propylthiouracil: a genetic marker for taste, with implications for food preference and dietary habits. Am J Hum Genet. 1998 Nov;63(5):1271-6

AC C

EP

TE D

M AN U

SC

RI PT

519 520

19

ACCEPTED MANUSCRIPT

Tepper, B.J. 2008. Nutritional implications of genetic taste variation: the role of PROP sensitivity and other taste phenotypes. Annual review of nutrition, 28, pp.367–388

555 556 557

Turnbull, B. and Matisoo-Smith, E. 2002. Taste sensitivity to 6-n-propylthiouracil predicts acceptance of bitter-tasting spinach in 3-6-y-old children. The American journal of clinical nutrition, 76(5), pp.1101– 1105

558 559

Turner, B.L. Thompson, A.L. 2013. Beyond the Paleolithic prescription: incorporating diversity and flexibility in the study of human diet evolution. Nutrition Reviews Vol. 71(8):501–510

560 561 562

Szarc vel Szic, K. Declerck, K. Melita Vidaković, M. and Vanden Berghe, W. 2015. From inflammaging to healthy aging by dietary lifestyle choices: is epigenetics the key to personalized nutrition? Clinical Epigenetics, 7:3

563 564

Williams, J.A. et al. 2016. Exploring Ethnic Differences in Taste Perception. Chemical senses Jun;41(5):449-56

565 566

Wink, M. 2015. Modes of action of herbal medicines and plant secondary metabolites. Medicines 2, pp. 251-286; doi:10.3390/medicines2030251

567 568

Wooding, S. et al. 2004. Natural selection and molecular evolution in PTC, a bitter-taste receptor gene. American journal of human genetics, 74(4), pp.637–646

569 570

Zhao L, Kirkmeyer S, Tepper B. 2003. A paper screening test to assess genetic taste sensitivity to 6-n propylthiouracil. Physiol Behav 78:625–633

SC

M AN U

TE D EP AC C

571

RI PT

553 554

20