Exploration of Vanilla Pompona From The Peruvian Amazon As A Potential Source of Vanilla Essence. Quantifaction of Phenolics by HPLC-DAD
Exploration of Vanilla Pompona From The Peruvian Amazon As A Potential Source of Vanilla Essence. Quantifaction of Phenolics by HPLC-DAD
Exploration of Vanilla Pompona From The Peruvian Amazon As A Potential Source of Vanilla Essence. Quantifaction of Phenolics by HPLC-DAD
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Ponticia Universidad Catolica del Peru, Departamento de Ciencias Quimica, Av. Universitaria 1801, Lima 32, Peru
Botanical Research Institute of Texas (BRIT), 1700 University Drive, Fort Worth, TX 76107-3400, USA
a r t i c l e
i n f o
Article history:
Received 7 April 2012
Received in revised form 27 August 2012
Accepted 1 October 2012
Available online 8 November 2012
Keywords:
Vanilla pompona
HPLC-DAD quantication
Glucovanillin
Phenolic prole
Peruvian wetlands
a b s t r a c t
This study provides the rst chemical investigation of wild-harvested fruits of Vanilla pompona ssp. grandiora (Lindl.) Soto-Arenas developed in their natural habitat in the Peruvian Amazon. Flowers were
hand-pollinated and the resulting fruits were analysed at different developmental stages using an
HPLC-DAD method validated for the quantication of glucovanillin and seven other compounds. The
method showed satisfactory linearity (r2 > 0.9969), precision (coefcient of variation <2%), recoveries
(70100%), limit of detection (0.0080.212 lg/ml), and limit of quantication (0.0270.707 lg/ml). The
evaluation of crude and enzyme-hydrolyzed Soxhlet-extracted samples conrmed the leading role of glucosides in fruit development. LCESI-MS studies corroborated the identities of four glucosides and seven
aglycones, among them vanillin (5.7/100 g), 4-hydroxybenzyl alcohol (3.6/100 g), and anisyl alcohol (7.1/
100 g) were found in high concentrations. The attractive avor/aroma prole exhibited by wild V. pompona fruits supports studies focused on the development of this species as a specialty crop.
2012 Elsevier Ltd. All rights reserved.
1. Introduction
Natural vanilla, a unique spice of great commercial value on the
international market, is borne from the fruit of Vanilla planifolia Andrews (syn. Vanilla fragans) (Orchidaceae), Vanilla tahitensis J.W.
Moore (commonly referred to as Java), or Vanilla pompona Schiede.
These vanilloid orchids produce large owers that are ephemeral
and, in commercial settings, require pollination by hand. Subsequent development of the fruit into a mature and harvestable fruit
takes about seven to nine months. Upon harvest, the fruits undergo
a long curing process (58 months) to nally become the avorful
and scented product. All of these laborious steps contribute to the
high price of vanilla, known to have reached about $500 per kg in
the last decade (Brownell, 2011). Each year approximately 2000
3000 tons are used commercially by the cosmetic, avoring, and
food industries. Madagascar, despite the crises experienced during
the last two decades, is still the main producer (Brownell, 2011),
while, France, Germany, Japan, and the United States are the main
consumers.
Vanilla fruits are chemically complex. More than 300 compounds have been identied from cured Vanilla fruit extracts (Toth,
Lee, Havkin-Frenkel, Belanger, & Hartman, 2011), and that number
continues to rise. Twenty-ve phenolic compounds with concen Corresponding author. Tel.: +51 1 6262000x4226; fax: +51 1 626 2853.
E-mail address: [email protected] (H. Maruenda).
0308-8146/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodchem.2012.10.037
162
results indicate that vanillyl (13) and benzyl (56) derivatives are
key components of cured V. planifolia fruits. Vanillin is found in
higher concentrations (between 2 and 2.5/100 g dry weight) than
the other aglycones 26 (Brodelius, 1994; Cicchetti & Chaintreau,
2009; Gassenmeier, Riesen, & Magyar, 2008; Odoux, 2000; Odoux,
2011; Sinha et al., 2007; Toth et al., 2011). The importance of compounds 1, 3, 5 and 6 in dening the avor and aroma of vanilla essence is apparent from the role they play as markers in the
authenticity and quality control of natural V. planifolia products
(Gassenmeier et al., 2008; Jhon & Jamin, 2004). In contrast to the
chemical prole exhibited by V. planifolia fruits, in cured fruits of
V. tahitensis, the sweet sensory impact of compounds 13 is
blended with oral notes exerted by the anisyl derivatives (79;
Ranadive, 2011). Three compounds are found in considerable
amounts: vanillin (1.31.5/100 g dry weight), anisyl alcohol (1.4/
100 g dry weight), and anisic acid (0.8/100 g dry weight)
(Leper-Andrzejewski, Brunschwig, Collard, & Dron, 2011).
While the composition of the fruits of V. planifolia and
V. tahitensis is well documented, the chemical prole of V. pompona
fruits has received scant attention. This could be due to the inaccessibility of V. pompona fruits or to the fact that they have been
described as carrying signicantly less vanillin than the fruits of
the other two species (Ehlers & Pster, 1997; Ranadive, 2011; Toth
et al., 2011). However, most of the information available is based
on cured fruits of V. pompona of unknown origin.
Wild V. pompona ssp. grandiora has recently been reported as
the most abundant of six Vanilla species present in wetland ecosystems of Madre de Dios, a region within the southern Peruvian Amazon (Householder et al., 2010). Early eld observations established
that the large fruits of wild populations of V. pompona ssp. grandiora exhibited strong aromatic and avorful properties. This suggested the hypothesis that V. pompona ssp. grandiora of the
Peruvian Amazon could represent a new potential source of vanilla
essence. The complete lack of knowledge regarding the chemistry
of Amazonian Vanilla species in comparison with those from other
regions of the world was surprising. As a consequence, wild populations of V. pompona ssp. grandiora were selected from their natural wetland habitat to serve as sources of properly identied fruits
ideally suited for chemical analysis. The objective at this early
stage of the program was to determine if these wild fruits had
any value in terms of aromatic chemical content, and if so, to identify the month at which the chemical prole yield was optimal.
0.45 lm syringe lter prior to injection (10 ll) into the HPLC system. Each extract was evaluated in triplicate.
2.4. HPLC analysis
The samples were analysed on an Agilent 1200 series DAD
equipped with a binary pump unit (G1312A), UVVIS detector
(G1315D), degasser (G13798), column oven (G1316A), a 20 ll loop
manual injector (G1328B) controlled by Chemstation software
LC3D (Agilent Technologies, Inc.). The column used was a Cromolith RP 18e (100 4.6 mm, Merck-Darmstadt), and the mobile
phase a mixture of two solvents: A (1 103 M K PO4, pH 3.1)
and B (MeOH). Elution was achieved at 30 C with a gradient of
37% B in 2 min (1.0 ml/min), 79% B in 10 min (1.02.0 ml/min),
and 919% B in 7 min (2.0 ml/min). The compounds were monitored at: 230 nm for GV, vanillin, and alcohols 2, 4, and 7;
254 nm for vanillic acid and 4-hydroxybenzoic acid; and 280 nm
for 4-hydroxybenzaldehyde. The injection volume was 10 ll, and
all samples were ltered through a PTFE 0.45 lm syringe lter
(Millipore, Germany) prior to analysis. The LCESI-MS experiments
were performed on a Bruker Daltonics Esquire 6000 controlled by
Compass 1.3 for Esquire/HCT software (Bruker Daltonik GmbH).
The HPLC system was as described for the HPLC-DAD. The gradient
separation was carried out at 35 C using a volatile mobile phase
composed of solvents A (4.2 103 M Formic acid, pH 3.1) and B
(MeOH), following the same gradient as above but at a ow rate
of 1 ml/min. After reaching 19% B it was held isocratic for
16 min. The injection volume was 7 ll. The ESI-MS operating
conditions were: positive ionisation mode, drying gas (N2) ow,
12 l/min; nebulizer pressure, 65 psi; gas drying temperature,
350 C; capillary voltage 4000 V; scan mode m/z, 90800. ESI-MS
and UV-DAD spectra of authentic standards, 17 and GV, were recorded and used to corroborate identity. The semipreparative
HPLC purication was conducted with a Cromolith RP 18e
(100 10 mm, Merck Darmstadt) using the conditions established for the LCESI-MS run, but with a ow rate of 2.5 ml/min.
The lyophilized extract (75 mg), dissolved in 1.5 ml of 30% aqueous
ethanol, was injected through a 2 ml loop. Fraction collection was
guided by retention time and UV spectra of the eluate.
2.5. Method validation
The HPLC method was validated according to ICH requirements
(ICH, 1996). The external calibration curves were prepared with
solutions containing compounds 17 in concentrations between
0.167 and 992.8 lg/ml. In the case of GV the range was 0.864
1035 lg/ml. The linearity was checked by regression analysis of
at least six concentrations of each compound. Repeatability was
conrmed by performing injections on the same day (intra-day
precision, n = 5) at a given concentration of each compound. The
intermediate precision was calculated over ve different days
(inter-day) at an established concentration per substance. Highly
diluted solutions of each compound were used to achieve signalto-noise (S/N) ratios of 3 (limit of detection, LOD) and 10 (limit
of quantication, LOQ). The recovery of the complete analytical
protocol was evaluated by the enrichment of vanilla samples
(2 g) with 50 mg of compounds 1, 4, and 7, and 10 mg of 2, 3, 5,
and 6. The recovery was expressed as the percentage of the total
amount recovered. The process was repeated in triplicate. The
fruits used for this study were nine-month old.
2.6. Statistical analysis
Statistical analysis of the data was carried out using IBM SPSS
Statistics Version 19.0 (SPSS, Inc. Chicago, IL) and the multivariate
analysis was performed using MATLAB software Version 7.11
163
(Matworks, Inc., Natick, MA). The PCA analysis was carried out
using standardized variables.
3. Results and discussion
3.1. Humidity
The average humidity of intact V. pompona ssp. grandiora
fruits, three and a half to nine months old, is presented in Table 1.
The observed increment of dry weight through fruit maturation
has been noted earlier with V. planifolia fruits (Brodelius, 1994).
The size of the 24 fruits used to calculate humidity ranged from
10 to 25 cm in length and 24 cm in width.
3.2. HPLC-DAD validation
The HPLC method developed for this study achieved adequate
separation among the eight standards, GV and 17, in 19 min,
Fig. 2A. The linear regression parameters obtained for all eight calibration curves were optimal in the concentration range established for each compound (Table 2). The same could be argued
for the sensitivity of the method, with LOD values in the range of
0.0080.212 lg/ml and LOQ of 0.0270.707 lg/ml. In terms of linear range, sensitivity, and HPLC run-time, the results are comparable to those achieved by two other validated HPLC methods
available in the literature for the quantication of compounds 1
6 (Cicchetti & Chaintreau, 2009; Sinha et al., 2007). However, this
is the rst validated HPLC method to assess GV and anisyl alcohol
in the presence of aglycones 16. A good linearity was achieved for
both of these compounds, GV and compound 7, over a wide concentration range, 11000 lg/ml, Table 2. Precision (retention time
and peak area) was addressed through intra-day and inter-day
repetitive analysis (n = 5) at four different concentrations per compound. In general, the results were satisfactory, displaying coefcient of variation (CV) below 2%, as shown in Table 3 for a
particular concentration of each standard.
Accuracy of the proposed method was tested only for the free
aglycones 17, since the extraction protocol followed has already
been reported as optimal for the isolation of GV (Voisine et al.,
1992). The recoveries for the experiment were high (>97%) for almost all aglycones and, acceptable (>70%), for compounds 2 and
4 (Table 4), suggesting that, overall, the validated method is suitable for quantication of vanillin and related phenolic compounds.
3.3. Quantitative determination of GV and aglycones 17 present in
vanillons at different stages of development
The presence of GV and phenolic compounds 17 in extracts of
V. pompona ssp. grandiora was demonstrated through comparisons of their HPLC-retention times, UV-DAD proles, and ESI-MS
spectra against that of authentic samples. The alcohols were identied through the ion [M+HH2O]+, whereas all others showed the
[M+H]+ signal. In the case of GV, its structure was veried through
the presence of m/z 353, 337 and 153, corresponding to the ions
[M+K]+, [M+Na]+ and [M+1162]+, respectively. The latter ion fragment corresponds to loss of glucose.
The HPLC-quantitative study was performed on twenty-four
green intact fruits at seven different stages of maturation. The
quantities of GV and free and total aglycones 17, were calculated
based on the amount present on enzyme-treated and non-treated
extracts. The results, plotted against time after pollination, are
shown in Fig. 3. The data suggest that vanillin and the other six
phenolic compounds accumulate in the fruit as both glucosides
and free forms throughout maturation. This result coincides
with previous observations (Brodelius, 1994). In the case of
164
Table 1
Average humidity (H) of Vanilla pompona ssp. grandiora fruits at different stages of development.
Month
a
H(%) SD
a
b
c
91.3 0.5
86.3 1.6
83.6 1.3
82.8 0.8
79.4 1.6
80.6 0.4
76.7 2.0c
Humidity is expressed as percentages [(wwet wdry/wwet) 100]. The values are average of three fruits unless otherwise indicated.
SD, standard deviation.
The value is an average of six fruits.
165
Compound
4-Hydroxybenzyl alcohol
4-Hydroxybenzoic acid
4-Hydroxybenzaldehyde
Glucovanillin
Vanillin
Vanillyl alcohol
Vanillic acid
Anisyl alcohol
a
b
c
Linear regression
1.20287
0.17126
0.34255
0.861035
0.26501
0.56416
0.23176
1.06993
LODb (lg/ml)
LOQc (lg/ml)
0.212
0.008
0.020
0.196
0.048
0.157
0.073
0.053
0.707
0.027
0.065
0.654
0.159
0.524
0.244
0.177
2a
y = mx + b
2158.5x 3.7
5372.5x 10.6
4981.8x 25.9
3511.9x 120.5
3636.4x 8.3
2334.4x 8.5
2899.4x 12.8
2186.7x 3.3
0.9999
0.9999
0.9999
0.9969
1.0000
0.9999
0.9998
0.9998
Correlation coefcient.
LOD = limit of detection (S/N = 3).
LOQ = limit of quantication (S/N = 10).
Table 3
Precision of the retention time and peak area of analytes in the HPLC-DAD method used.
Compounda
4-Hydroxybenzyl alcohol
4-Hydroxybenzoic acid
4-Hydroxybenzaldehyde
Glucovanillin
Vanillin
Vanillyl alcohol
Vanillic acid
Anisyl alcohol
a
5.34
8.17
9.93
8.77
14.29
7.41
11.43
15.10
Retention time
Peak area
Retention time
Peak area
0.105
0.083
0.058
0.423
0.073
0.071
0.057
0.222
1.948
1.806
1.836
1.777
1.422
1.867
1.510
0.745
0.228
0.573
0.361
0.067
0.095
0.100
0.297
0.190
1.177
1.365
1.869
1.976
1.912
1.908
1.277
0.279
Concentration of GV was 1.65 mM, anisyl alcohol 0.76 mM; all other analytes in the range 0.140.44 mM.
Table 4
Recovery of the method used for the determination of compounds 17 in vanilla fruits
(n = 3).
Compound
Spike (mg)
Recovery (%)
RSDa (%)
4-Hydroxybenzyl alcohol
4-Hydroxybenzoic acid
4-Hydroxybenzaldehyde
Vanillin
Vanillyl alcohol
Vanillic acid
Anisyl alcohol
50
10
10
50
10
10
56
72.7
100.8
101.8
96.5
70.4
100.3
93.7
3.32
2.65
2.54
4.05
1.48
2.98
0.32
166
Fig. 3. Free () and total (-) aglycone content in fruits of V. pompona ssp. grandiora at different developmental stages. Error bars indicate standard error.
Table 5
Contents of GV and compounds 19 expressed in g/100 g dry weight present in
uncured V. planifolia, V. tahitensis and Peruvian V.pompona ssp. grandiora fruits.
Compound
Vanilla planifoliaa
Vanilla tahintensisb
Vanilla pomponac
GV
4
5
6
1
2
3
7
8
9
8.51
0.87
0.40
0.11
5.17
0.17
0.30
Not reported
0.21
0.60
2.02
0.06
0.02
2.11
0.01
0.79
5.869.88
1.733.55
0.010.03
0.010.02
2.665.71
0.220.61
0.050.14
4.057.13
NDd
ND
Acknowledgements
The authors thank Jason Wells, Javier Huinga, Angel Balarezo
and Manuel Huinga for assisting with the pollination of Vanilla
owers and harvest of the fruits. We owe much appreciation to
Renan Valega of BRIT Peru for support provided in the
administration of the project and to Alex Nieva from PUCP for his
assistance in the LCMS study. The Peruvian Instituto Nacional
de Recursos Naturales (INRENA) issued permits for research and
collection activities of this project. This research was nanced by
Programa de Ciencia y Tecnologia FINCYT (co-nanced by BID)
grant number PIBAP-2007-005. Field and herbarium research were
supported in part by the US National Science Foundation (NSF)
grant # 0717453 to BRIT, the Gordon Betty Moore Foundation,
and the Clayton Fund of Houston, Texas.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.foodchem.2012.
10.037.
References
Brodelius, P. E. (1994). Phenylpropanoid metabolism in Vanilla planifolia Andr. (V)
High performance liquid chromatographic analysis of phenolic glycosides and
aglycones in developing fruits. Phytochemical Analysis, 5, 2731.
Brownell, R. J. Jr., (2011). Fair trade The future of vanilla? In D. Havkin-Frenkel & F.
Belanger (Eds.), Handbook of vanilla science and technology (pp. 107116). West
Sussex, UK: Blackwell Publishing Ltd.
Cicchetti, E., & Chaintreau, A. (2009). Quantitation of the main constituents of
vanilla by reverse phase HPLC and ultra-high-pressure-liquid-chromatography
with UV detection: Method validation and performance comparison. Journal of
Separation Science, 32, 30433052.
Da Costa, N. C., & Pantini, M. (2006). The analysis of volatiles in Tahitian vanilla
(Vanilla tahitensis) including novel compounds. Developments in Food Science, 43,
161164.
Dignum, M. J. W., Kerler, J., & Verpoorte, R. (2002). Vanilla curing under laboratory
conditions. Food Chemistry, 79, 165171.
Dignum, M. J. W., van der Heijden, R., Kerler, J., Winkel, C., & Verpoorte, R. (2004).
Identication of glucosides in green beans of Vanilla planifolia Andrews and
kinetics of vanilla b-glucosidase. Food Chemistry, 85, 199205.
Ehlers, D., & Pster, M. (1997). Compounds of vanillons (Vanilla pompona Schiede).
Journal of Essential Oil Research, 9, 427431.
Gassenmeier, K., Riesen, B., & Magyar, B. (2008). Commercial quality and analytical
parameters of cured vanilla beans (Vanilla planifolia) from different origins from
the 20062007 crop. Flavour and Fragance Journal, 23, 194201.
Havkin-Frenkel, D., French, J. C., Graft, N. M., Pak, F. E., Frenkel, Ch., & Joel, D. M.
(2004). Interrelation of curing and botany in vanilla (Vanilla planifolia) Bean.
Acta Horticulturae (ISHS), 629, 93102.
Householder, E., Janovec, J., Balarezo, A., Huinga, J., Wells, J., Valega, R., et al. (2010).
Diversity, natural history, and conservation of vanilla (Orchidaceae) in
Amazonian Wetlands of Madre de Dios, Peru. Journal of the Botanical Research
Institute of Texas, 4(1), 227243.
International Conference on Harmonization (ICH). (1996) Guideline on the
validation of analysis procedures: Methodology Q2B.
167
Itokawa, H., Oshida, Y., Ikuta, A., Inatomi, H., & Adachi, T. (1982). Phenolic plant
growth inhibitors from the owers of Cucurbita pepo. Phytochemistry, 21,
19351937.
Jhon, T. V., & Jamin, E. (2004). Chemical investigation and authenticity of Indian
vanilla beans. Journal of Agricultural and Food Chemistry, 52, 76447650.
Kanisawa, T. (1993). Flavor development in Vanilla beans. Kouryou, 180(5),
113123.
Kitajima, J., Ishikawa, T., Tanaka, Y., Ono, M., Ito, Y., & Nohara, T. (1998). Watersoluble constituents of fennel. V. Glycosides of aromatic compounds. Chemical &
Pharmaceutical Bulletin, 46(10), 15871590.
Lapeyre-Montes, F., Conejero, G., Verdeil, J.-L., & Odoux, E. (2011). Anatomy and
biochemistry of vanilla bean development (Vanilla planifolia G. Jackson). In E.
Odoux & M. Grisoni (Eds.). Medicinal and aromatic plants-industrial proles:
Vanilla (Vol. 47). Boca Raton, FL: CRC Press, pp 164.
Leper-Andrzejewski, S., Brunschwig, Ch., Collard, F.-X., & Dron, M. (2011).
Morphological, chemical, sensory, and genetic specicities of tahitian vanilla.
In E. Odoux & M. Grisoni (Eds.). Medicinal and aromatic plants-industrial proles:
Vanilla (Vol. 47, pp. 205228). Boca Raton, FL: CRC Press.
Negishi, O., Sugiura, K., & Negishi, Y. (2009). Biosynthesis of vanillin via ferulic acid
in Vanilla planifolia. Journal of Agricultural and Food Chemistry, 57, 99569961.
Odoux, E. (2000). Changes in vanillin and glucovanillin concentrations during the
various stages of the process traditionally used for curing Vanilla fragans beans
in Reunion. Fruits, 55, 119125.
Odoux, E. (2011). Developing the aromatic quality of cured vanilla beans (Vanilla
planifolia G. Jackson). In E. Odoux & M. Grisoni (Eds.). Medicinal and aromatic
plants-industrial proles: Vanilla (Vol. 47). Boca Raton, FL: CRC Press, pp 189
204.
Palama, T. L., Khatib, A., Choi, Y. H., Payet, B., Fock, I., Verpoorte, R., et al. (2009).
Metabolic changes in different developmental stages of Vanilla planifolia pods.
Journal of Agricultural and Food Chemistry, 57, 76517658.
Ranadive, A. S. (2011). Quality control of vanilla beans and extracts. In D. HavkinFrenkel & F. Belanger (Eds.), Handbook of vanilla science and technology
(pp. 141160). West Sussex, UK: Blackwell Publishing Ltd.
Sharma, U. K., Sharma, N., Gupta, A. P., Kumar, V., & Sinha, A. K. (2007). RP HPTLC
densitometric determination and validation of vanillin and related phenolic
compounds in accelerated solvent extract of Vanilla planifolia. Journal of
Separation Science, 30, 31743180.
Sinha, A. K., Sharma, U. K., & Sharma, N. (2008). A comprehensive review on vanilla
avor: Extraction, isolation and quantication of vanillin and others
constituents. International Journal of Food Sciences and Nutrition, 59, 299
326.
Sinha, A. K., Verma, S. C., & Sharma, U. K. (2007). Development and validation of an
RP HPLC method for quantitative determination of vanillin and related
phenolic compounds in Vanilla planifolia. Journal of Separation Science, 30,
1520.
Sreedhar, R. V., Roohie, K., Venkatachalam, L., Narayan, M. S., & Bhagyalakshmi, N.
(2007). Specic treatments reduce curing period of vanilla (Vanilla planifolia)
beans. Journal of Agricultural and Food Chemistry, 55, 29472955.
Toth, S., Lee, K. J., Havkin-Frenkel, D., Belanger, F. C., & Hartman, G. (2011). Volatile
compounds in vanilla. In D. Havkin-Frenkel & F. Belanger (Eds.), Handbook of
vanilla science and technology (pp. 183218). West Sussex, UK: Blackwell
Publishing Ltd.
Voisine, R., Carmichael, L., Chalier, P., Cormier, F., & Morin, A. (1992). Determination
of glucovanillin and vanillin in cured vanilla pods. Journal of Agricultural and
Food Chemistry, 43, 26582661.