Food Chemistry 225 (2017) 154–161

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

A new natural source for obtainment of inulin and fructo-

oligosaccharides from industrial waste of Stevia rebaudiana Bertoni
Sheila Mara Sanches Lopes a, Gabriela Krausová b, José Walter Pedroza Carneiro c, José Eduardo Gonçalves d,
Regina Aparecida Correia Gonçalves a, Arildo José Braz de Oliveira a,⇑
a
Graduate Program in Pharmaceutical Sciences, Department of Pharmacy, State University of Maringá, Ave. Colombo 5790, 87.020-900 Maringá, Brazil
b
Department of Microbiology and Technology, Dairy Research Institute, Ke Dvoru 12a, 160 00 Prague, Czech Republic
c
Department of Agronomy, Iguatemi Research Farm, State University of Maringá, Av. Colombo 5790, 87.020-900 Maringá, Brazil
d
Program of Master in Health Promotion and Program of Master in Clean Technologies, University Center of Maringá, Av. Guedner, 1610, 87.050-900, Maringá, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Fructan-type inulin and fructo-oligosaccharides (FOS) are reserve polysaccharides that offer an interest-
Received 16 August 2016 ing combination of nutritional and technological properties for food industry. Stevia rebaudiana is used
Received in revised form 28 November 2016 commercially in the sweetener industry due to the high content of steviol glycosides in its leaves.
Accepted 31 December 2016
With the proposal of using industrial waste, the objective of the present study was to isolate, characterize
Available online 6 January 2017
and evaluate the prebiotic activity of inulin and FOS from S. rebaudiana stems. The chemical characteri-
zation of the samples by GC–MS, NMR and off-line ESI-MS showed that it was possible to obtain inulin
Chemical compounds studied in this article:
molecules from the S. rebaudiana stems with a degree of polymerization (DP) of 12, and FOS with a
Inulin from chicory (PubChem CID:
16219508)
DP < 6. The in vitro prebiotic assay of these molecules indicates a strain specificity in fermentation capac-
Fructose (PubChem CID: 5984) ity of fructans as substrate. FOS molecules with a low DP are preferably fermented by beneficial micro-
Glucose (PubChem CID: 5793) biota than inulin molecules with higher DP.
Myo-inositol (PubChem CID: 892) Ó 2017 Elsevier Ltd. All rights reserved.
Ethanol (PubChem CID: 702)
Trifluoroacetic acid (PubChem CID: 6422)
Hexamethyldisilazane (PubChem CID:
13838)
Chlorotrimethylsilane (PubChem CID: 6397)
Pyridine (PubChem CID: 1049).

Keywords:
Industrial waste
Stevia rebaudiana
Inulin
Fructo-oligosaccharides
Prebiotic activity

1. Introduction
Stevia rebaudiana Bertoni is a perennial herb of the Asteraceae
family and is native from South America. It has been used commer-
cially in the sweetener industry since the 1970s, when Japanese
researchers developed a process for the extraction and refinement
of stevioside and rebaudioside from its leaves (Serfaty et al., 2013;
Tavarini & Angelini, 2013). Fructans are another type of metabolite
⇑ Corresponding author at: Graduate Program in Pharmaceutical Science, Depart-
ment of Pharmacy, State University of Maringá, Ave. Colombo 5790, 87.020-900
Maringá, Brazil.
E-mail address: ajboliveira@uem.br (A.J.B. de Oliveira).
of commercial interest, and have been isolated from the roots of S.
rebaudiana (Lopes et al., 2016; Lopes, Krausová, Rada, Gonçalves,
Gonçalves, & De Oliveira, 2015; Oliveira et al., 2011).
Inulin and fructo-oligosaccharides (FOS) are fructan-type
polysaccharides that consisting of (2 ? 1)-linked b-D-
fructofuranosyl (b-D-Fruf) residues with a terminal glucose residue
linked by an a-D-(1 ? 2) bond. The general structure of fructans
can be described as GFn, where G and F represent glucose and fruc-
tose, respectively, and ‘n’ is the number of fructosyl units (Shoaib
et al., 2016; Singh, Singh, & Kennedy, 2016).
The fructans differ in chain length, or degree of polymerization
(DP). Fructo-oligosaccharides are smaller molecules with a DP < 9,
and inulin has DP ranging from 10 to 65 fructose units, but the

http://dx.doi.org/10.1016/j.foodchem.2016.12.100
0308-8146/Ó 2017 Elsevier Ltd. All rights reserved.
 S.M.S. Lopes et al. / Food Chemistry 225 (2017) 154–161
most common are molecules with DP between 12 and 15 (Singh
et al., 2016).
These compounds are considered a functional food ingredient
that can be used as a dietary fiber and prebiotic (Caleffi et al.,
2015; Moreno-Vilet et al., 2014). Regular consumption of such
non-digestible polysaccharides (i.e. prebiotics) has positive effects
on human health (Shoaib et al., 2016). These compounds can
reduce cardiometabolic risk by diminishing levels of triglycerides
and cholesterol, modulating hyperglycemia (Mitchell et al.,
2015), preventing colorectal cancer (Ambalam, Raman, Purama, &
Doble, 2016), increasing the intestinal absorption of minerals
(Lobo et al., 2011) and improving immune system efficiency
(Moreno-Vilet et al., 2014; Peshev & Van den Ende, 2014).
Fructans offer a unique combination of nutritional properties
and technological applications, they are used as a fat and sugar
substitute in low-calorie foods, while improving the taste and
mouthfeel of food items such as dairy products (Aravind, Sissons,
Fellows, Blazek, & Gilbert, 2012; Chi et al., 2011; Crispín-Isidro,
Lobato-Calleros, Espinosa-Andrews, Alvarez-Ramirez, & Vernon-
Carter, 2015).
Sustainable technological development aims to reduce environ-
mental impact using renewable resources such as raw materials
from industry (Misselbrook, Menzi, & Cordovil, 2012). Organic
wastes generated by agriculture and industry are an abundant nat-
ural biomass source, which can be recycled for the production of
compounds of commercial interest. Mainly, functional ingredients
with health promotion potential and prevention of diseases, that
meet consumer requirements for health foods (Castro, Soares,
Albernaz, & Sato, 2016; Salihu, Alam, Abdulkarim, & Salleh, 2012).
The objective of the present study was to isolate and character-
ize fructans (inulin and fructo-oligosaccharides) from S. rebaudiana
stems, as well as evaluate the prebiotic activity of these molecules
in vitro, with the propose of obtaining products of commercial
interest from waste generated by the sweetener industry.
2. Materials and methods
2.1. Materials
The S. rebaudiana stems were obtained from field cultivar of the
Iguatemi Research Station (located at 230200 SL, 520040 WL and alti-
tude of 506 m) of the State University of Maringá. The S. rebaudiana
voucher specimen (14301-HUEM) was deposited at the Herbarium
of the State University of Maringá, Brazil. The stems were dried in a
drying oven at 48 °C for 3 days and milled.
Inulin extracted from chicory (OraftiÒ GR) and fructo-
oligosaccharides (OraftiÒ P95) obtained by the hydrolysis of the
chicory inulin used as standard, were purchased from Beneo, Bel-
gium. The fructose, glucose and myo-inositol standards were from
Sigma-AldrichÒ. All other chemical reagents were analytical grade.
2.2. Fructan extraction
The S. rebaudiana stems (24.0 g) were extracted with deionized
water (300 mL) under reflux conditions, at 80 °C. Sequential
extraction was carried out at 4, 8 and 12 h. The crude extracts were
separately concentrated in a rotary evaporator and were then pre-
cipitated with ethanol at a 1:3 (v/v) ratio and maintained at 4 °C for
24 h under refrigerated conditions. The crude extracts were cen-
trifuged (6000  g at 4 °C for 20 min).
The inulin molecules with high molecular weight were obtained
from the precipitate fraction. The fructo-oligosaccharides (FOS)
were obtained in a soluble form from the ethanolic supernatant.
Both fractions were lyophilized to obtain the dry inulin and FOS
extracts.
155
2.3. Total sugar determination
The total carbohydrate content in the extracts was determined
by the phenol-sulfuric acid method with d-fructose as the standard
(Dubois, Gilles, Hamilton, Rebers, & Smith, 1956).
2.4. Monosaccharide composition
The samples of inulin and FOS (5.0 mg) from S. rebaudiana
stems were hydrolyzed with 0.5 M trifluoroacetic acid (TFA) at
60 °C for 1 h, followed by oxime derivatization with methoxya-
mine hydrochloric acid in pyridine (20 mg/mL) at 70 °C for 1 h.
Then, the oximes were trimethylsilylated with 200 lL of the fol-
lowing derivatization mixture: HMDS + TMCS + Pyridine, 3:1:9
(Lopes et al., 2015).
The silylated oxime derivatives (dissolved in 200 lL hexane)
were analyzed by Gas Chromatography Agilent 7890B coupled
with Mass Spectrometry Agilent 5977A MSD, using an HP5-MS
UI-Agilent with a fused silica capillary column
(30 m  0.25 mm  0.25 lm). Helium was used as the carrier gas,
with an oven temperature of 170–210 °C (2 °C/min). The injector
and interface were kept at 280 °C and 260 °C, respectively. The
injections were made in split mode with a split ratio of 1:40. The
mass spectrometer was operated in electron impact (EI) mode at
70 eV. The quadrupole and source temperatures were 150 °C and
230 °C, respectively.
The characterization of the compounds was performed using
the NIST Mass Spectral Library Database – NIST 11 and through
comparison with the mass spectrum and retention time of the
standard glucose and fructose analyzed under the same conditions.
The monosaccharide concentrations were obtained by analysis
of the relative area of the integration of the chromatographic
peaks, corrected by internal standard. The degree of polymeriza-
tion by GC–MS was calculated by the ratio between the fructose
and glucose concentrations (Lopes et al., 2015).
2.4.1. 1H and 13C NMR analyzes
The samples of inulin and FOS (20.0 mg) were dissolved in deu-
terium oxide (D2O) 99.9%, maintained at 45 °C for 24 h for hydro-
gen exchange and then lyophilized. The dried waste was
solubilized in 700 lL of D2O and then analyzed.
One- and two-dimensional NMR spectra were recorded at 298 K
using a Bruker Spectrometer (Avance III HD model) operating at
500.0 MHz for the 1H nucleus and 125.0 MHz for the 13C nucleus,
using the standard pulse sequences available in the Bruker soft-
ware. The chemical shifts (d) were expressed in parts per million
(ppm).
The spectra signals were assigned in comparison with the inulin
standard (OraftiÒ GR), FOS standard (OraftiÒ P95), and according to
literature data (Caleffi et al., 2015; Lopes et al., 2015; Oliveira et al.,
2011).
The DP by 1H NMR was calculated from the mean ratio between
the integral proton signal (H3-Fru and H4-Fru) and the integral of
the anomeric hydrogen glucose signal (H10 -Glc) (Caleffi et al.,
2015). The integrals of the protons were calculated by the Mestre
Nova 6.1 software package (Mestrelab Research S. L., 2010).
2.5. Electrospray ionization mass spectrometric analysis of FOS
The FOS molecules were dissolved in Milli-Q water (1 mg/mL),
and an aliquot of 250 lL of FOS was added to 250 lL of acetonitrile.
One drop of formic acid was added and the samples were cen-
trifuged at 6000  g for 5 min.
The supernatant was introduced into the Quattro MicroTM API
(Waters) mass spectrometer using a syringe pump (40 lL/min),
for off-line ESI-MS analysis. Mass spectra were obtained in the pos-
156 S.M.S. Lopes et al. / Food Chemistry 225 (2017) 154–161
itive ionization mode, setting the capillary voltage at 3500 V, the
cone voltage at 70 V, the source temperature at 120 °C and the des-
olvation temperature at 350 °C. Each spectrum was produced by
the accumulation of data over 1 min (Lopes et al., 2016; Oliveira
et al., 2011).
2.6. Prebiotic activity testing
Prebiotic activity was tested in seven strains of bifidobacteria
and lactobacilli originating from various sources including com-
mercially used dairy strains, as well as isolates from infant feces
and biopsy samples. The strains Bifidobacterium bifidum CCDM
559, Bifidobacterium breve CCDM 562 and Lactobacillus animalis
CCDM 382 were obtained from the Culture Collection of Dairy
Microorganisms Laktoflora (Prague, Czech Republic). The Lacto-
bacillus casei subsp. paracasei PE1 TB-P and Lactobacillus gasseri
PHM-7E1 strains were isolated from biopsy samples (Dairy
Research Institute, Tábor, Czech Republic) and the Lactobacillus fer-
mentum RL25 strain came from infant feces (Czech University of
Life Sciences, Prague, Czech Republic). The commercial strain Bifi-
dobacterium animalis subsp. lactis Bb12 was purchased from Chr.
Hansen (Hoersholm, Denmark).
Bacterial strains were grown in a basal medium containing
10.0 g of peptone, 10.0 g of tryptone, 5.0 g of yeast extract, 1 mL
of Tween 80Ò and 2.0 g of carbon source per 1 L distilled water.
The medium was autoclaved at 21 °C for 15 min. Inulin and
fructo-oligosaccharides from S. rebaudiana stems were used as a
carbon source. As a negative control, a medium without carbohy-
drate and an uninoculated medium were used. The Wilkins Chal-
gren broth (Oxoid, Basingstoke, UK) was used as a control of
bacterial growth and viability. The OraftiÒ P95, a commercially
available prebiotic (containing FOS from chicory) was used for
comparison with the FOS isolated from S. rebaudiana. All strains
were precultivated in Wilkins Chalgren broth (Oxoid, Basingstoke,
UK) overnight (37 °C for 16 h).
Bacterial suspensions (0.5 mL) of the respective strains in the
exponential growth phase were used to inoculate each tested
media. Prior to this, they were centrifuged (6000  g for 7 min)
and resuspended in saline (0.9%) at a concentration of 106 CFU/
mL. Subsequently, the media were incubated at 37 °C for 24 h in
anaerobic jars (Oxoid, Basingstoke, UK). All the strains were grown
in triplicate.
The optical density of the media was measured (densitometer
DEN-1, Dynex, Czech Republic) after inoculation (time zero), and
at the end of cultivation (24 h). The bacterial growth was evaluated
as the change in the absorbance (at the wavelength of 540 nm) of
the media during 24 h of fermentation. For statistical evaluation of
the data, StatgraphicsÒ Centurion XV Software (StatPoint, Inc.,
Warrenton, USA) and the multiple range comparison LSD test were
used. A difference was considered to be statistically significant at
the level of p < 0.05.
3. Results and discussion
3.1. Fructan extraction
The greatest extraction efficiency was observed during the first
4 h. The precipitate fraction identified as inulin exhibited a yield of
4.5% and the soluble fraction (FOS) presented a yield of 11%. Addi-
tional hours of extraction (8 and 12 h) did not significantly increase
the extraction yield of the polysaccharides. The carbohydrate con-
tent in the factions was 35% and 60% for precipitate and super-
natant fraction, respectively.
Hot-water extraction is a method that ensures a good yield, as
well as increasing the solubility of the polysaccharides and the dif-
fusion coefficient (Miao et al., 2011). Other studies consider an
ideal extraction time of 4 h, as beyond this time the polysaccharide
yield does not significantly increase (Miao et al., 2011). A similar
data was observed in the present study.
The total fructans content was about 15%, similar to the value
(15–20%) observed in species currently used for commercial pur-
poses, such as Jerusalem artichoke (Helianthus tuberosus) and chic-
ory (Cichorium intybus) (Chi et al., 2011).
3.2. Fractionation of the fructans crude extract
The fractionation of the fructan crude extract from S. rebaudiana
stems allowed to obtain two compounds with industrial interest,
inulin and fructo-oligosaccharides. The difference in the degree
of polymerization (DP) of these molecules plays a key role in the
functionalities of fructans, resulting in different technological and
nutritional properties (Karimi, Azizi, Ghasemlou, & Vaziri, 2015).
Inulin molecules are used as fat substitutes and texture agents
(Karimi et al., 2015), while FOS molecules are used as a low-
calorie sugar replacement (1–2 kcal g1) (Villegas, Tárrega,
Carbonell, & Costell, 2010). The use of blends of short- and long-
chain fructans, which are selectively metabolized in the proximal
and distal colon, respectively, enhances the fermentative and pre-
biotic effects of these molecules (Karimi et al., 2015; Tárrega,
Rocafull, & Costell, 2010).
The obtaining of two products (inulin and FOS) with different
industrial applications from the same source (S. rebaudiana stems)
is a low cost option for commercial extraction from industrial
waste.
3.3. Inulin characterization
In the monosaccharide composition analysis by GC–MS, the
peaks were identified according to retention time in comparison
with fructose (Fig. 1C) and glucose (Fig. 1D) standards, the frag-
mentation profile of the mass spectra of oxime-silylated deriva-
tives (Fig. S1) and literature data (Lopes et al., 2015).
The inulin chromatogram (Fig. 1A) showed peaks identified as
fructose units with retention times of 7.9, 8.0, 8.2, 9.6 and
9.9 min. Glucose units with retention times of 10.2 and 12.6 min
were observed. Retention times of 11.8, 15.3 and 15.9 min were
assigned to internal standard myo-inositol. Quantitative analysis
by GC–MS showed a higher content of fructose (32%) than glucose
(2.6%). The DP calculated for inulin obtained from S. rebaudiana
stems by GC–MS was 12.
In the NMR analysis of inulin molecules, the H10 signal of termi-
nal glucose was observed at low intensity in the anomeric region of
the 1H NMR spectrum (Fig. 2A) at d 5.45 (1H, d, J = 3.8 Hz). Charac-
teristic signals of fructose residues (?2-b-Fru) H3 and H4 were
observed at d 4.27 (11H, d, J = 8.0 Hz) and d 4.11 (14H, t,
J = 8.4 Hz), respectively.
The other fructosyl hydrogen signals (H1, H5 and H6) were
assigned between the d 3.70–3.95 region (Table 1). All chemical
shifts are consistent with the spectrum of the inulin standard
(OraftiÒ GR, Fig. S2) and literature data (Caleffi et al., 2015; Lopes
et al., 2015; Oliveira et al., 2011). The DP of inulin estimated
through 1H NMR analysis was 12, a similar value was observed
in the GC–MS analysis.
The 13C NMR spectrum (Fig. S3) showed six signals assigned to
fructose carbons. There was a peak of anomeric carbon at d 103,20
(C2-Fru), methylene group (CH2) peaks at d 60.74 (C1-Fru) and
62.26 (C6-Fru), and peaks assigned to the CAH group at d 74.39
(C4-Fru), 77.00 (C3-Fru) and 81.02 (C5-Fru). These data were con-
firmed using the DEPT 135 experiment (Caleffi et al., 2015; Lopes
et al., 2015).
 S.M.S. Lopes et al. / Food Chemistry 225 (2017) 154–161
Fig. 1. GC–MS Chromatogram (TIC) of oxime-silylated derivatives: (A) inulin and from S. rebaudiana stems, (B) fructo-oligosaccharides from S. rebaudiana stems, (C) fructose
standard, (D) glucose standard.
The correlation map HMBC (Fig. 3) showed the occurrence of a
cross-peak between H-1 and C-2, and the concomitant absence of
correlations between C-2 and H-6 suggests a b-(2?1) linkage
between fructosyl residues. The other correlations observed, H3-
Fru/C1-Fru, H4-Fru/C3-Fru, H3-Fru/C4-Fru, H4-Fru/C5-Fru, H6-
Fru/ C5-Fru and H4-Fru/C6-Fru, contribute to the proposed assign-
ments for the isolated inulin molecule (Caleffi et al., 2015; Lopes
et al., 2015).
From the degree of polymerization calculated by 1H NMR and
GC–MS was possible to estimated the molecular weight (Mw) of
inulin molecule using the following formula described by Suzuki,
Maeda, Grant, Grant, and Sporns (2013): [C6nH10n+2O5n+1], where
‘‘n” is the average value of DP of the inulin. Considering the
DP = 12 and the monoisotopic mass of atoms, the molecular weight
of the inulin molecules was 1947 g/mol. This value is in accordance
with the literature which shows that inulin molecules with DP
ranging from 2–13 have molecules with Mw in the range of
1400–2500 g/mol (Evans et al., 2016).
3.4. FOS characterization
In monosaccharide composition analysis, the FOS chro-
matogram (Fig. 1B) showed peaks with retention times of 7.7,
7.9, 8.1, 8.5, 9.5 and 9.7 min that were assigned to fructose units.
Signals relating to glucose units were observed at retention times
of 9.9, 10.1, 10.6, and 12.5 min. Similar retention times were
observed for fructose (Fig. 1C) and glucose (Fig. 1D) standards.
The mass spectra of the majority peaks of oxime-silylated deriva-
tives are shown in Fig. S1. The data obtained for the FOS molecules
is consistent with literature (Lopes et al., 2016). The quantitative
data showed 45% fructose and 11% glucose concentration. The
degree of polymerization by GC–MS showed molecules with a
DP = 4.
157
In the 1H NMR spectrum of FOS (Fig. 2B), characteristic signals
(?2-b-Fru) were observed at d 4.27 (3H, d, J = 8.0 Hz) and d 4.11
(6 H, t, J = 8.4 Hz), which corresponded to H3 and H4, respectively.
The correlation 1H-13C of these signals were observed in the HSQC
Correlation Map (Fig. 4) at d 4.27/76.93 (H3-Fru/C3-Fru) and
4.12/74.93 ppm (H4-Fru/C4-Fru). The H10 of the terminal glucose
was assigned in the anomeric region, with a low intensity signal
at d 5.45 (1H, d, J = 3.8 Hz) (Lopes et al., 2016).
In addition to the fructo-oligosaccharides, signals of mono- and
disaccharides were also observed in the extracts (Fig. 2B). These
signals were confirmed using 2D HSQC Correlation Map (Fig. 4).
The anomeric signals due to free glucose were observed at d
5.24/91.75 ppm for a-glucose and at d 4.65/95.81 ppm for b-
glucose (its H2 is present at d 3.29 ppm), these correlations agree
with data described by Matulová, Husárová, Capek, Sancelme,
and Delort (2011).
The sucrose chemical shifts are shown in Table 1 and the main
correlations for ?2)-b-Fruf residues were observed at d 4.21/76.50
(H3-Fru/C3-Fru), 4.04/71.78 ppm (H4-Fru/C4-Fru) in the HQSC
Correlation Map (Fig. 4). These molecules are commonly observed
in aqueous plant extracts, mainly in soluble fractions, as was
observed in the present study.
The other fructosyl (H1, H5 and H6) and glucosyl hydrogens
(H30 , H50 and H60 ) of the fructo-oligosaccharide and sucrose mole-
cules were found in the region between d 3.40–3.95 (Table 1). The
chemical shifts are consistent with FOS standard (OraftiÒ P95,
Fig. S2) and literature data (Lopes et al., 2016; Matulová et al.,
2011). The FOS molecules analyzed by 1H NMR showed an average
DP = 4.5.
The off-line ESI-MS spectrum (Fig. S4) showed fructo-
oligosaccharide peaks with potassium adducts [M+K]+. The peaks
at m/z 544 and 706 corresponded to FOS molecules with DP = 3
and DP = 4, that corresponding to 1-Kestose and 1-Nystose com-
158 S.M.S. Lopes et al. / Food Chemistry 225 (2017) 154–161

Fig. 2. 1H NMR spectra (500.0 MHz, D2O at 298 K) of the (A) inulin and (B) fructo-oligosaccharides from S. rebaudiana stems.

Table 1
1
H chemical shifts (d) of inulin standard (OraftiÒ GR), FOS (OraftiÒ P95) standard, inulin and fructo-oligosaccharides (FOS) from S. rebaudiana stems
1
Sample Sugar Unit H Chemical Shifts/d
1 2 3 4 5 6
OraftiÒ GR ?2)-b-Fruf 3.93–3.72 – 4.27 4.11 3.87 3.79–3.76
OraftiÒ P95 ?2)-b-Fruf 3.92–3.71 – 4.25 4.11 3.87 3.78–3.76
Stevia Inulin ?2)-b-Fruf 3.93–3.72 – 4.27 4.11 3.86 3.79–3.76
Stevia FOS ?2)-b-Fruf 3.91–3.70 – 4.26 4.11 3.88 3.78
Sucrose* a-Glcp-(1? 5.42 3.58 3.75 3.47 3.87 3.83
?2)-b-Fruf 3.68 – 4.22 4.04 3.87 3.80
*
Chemical shifts (d) of sucrose observed in the FOS from S. rebaudiana stems.

pounds, respectively. These are the two first products of the
biosynthesis of fructo-oligosaccharides molecules. The mass differ-
ence of 162 Da between the peaks corresponds to hexose units
(Evans et al., 2016; Lopes et al., 2016).
Some steviol glycosides were observed in the ESI-MS analysis of
the extract from the stems of S. rebaudiana, most notably the ste-
vioside and rebaudioside A in the adduct potassium form
(Fig. S4). Jackson et al., 2009, also report the presence of these gly-
cosides in small amounts in other plant parts besides the leaves of
S. rebaudiana.
3.5. Prebiotic activity testing
Inulin and fructo-oligosaccharides are prebiotic compounds
that are widely used in functional food formulations. They are
defined as non-digestible food ingredients which selectively stim-
ulate the growth of lactobacilli and bifidobacteria, which improve
human health (Karimi et al., 2015; Roberfroid et al., 2010).
The present study evaluated the prebiotic activity in vitro of
inulin and FOS previously isolated and characterized from S. rebau-
diana stems. The results were expressed as the differences in the
optical density of the growth media at inoculation (time zero)
and after 24 h of incubation. The resulting data is shown in Table 2.
The FOS isolated from S. rebaudiana stems were fermented by
all the strains tested, with a growth statistically (p < 0.05) higher
than bacterial growth in the basal medium (negative control).
The average increases in the growth of the bifidobacteria and lac-
tobacilli strains were 2.28-fold and 1.7-fold, respectively.
Among the tested bifidobacteria, the strain B. bifidum CCDM 559
showed better capacity to fermente FOS from S. rebaudiana and in
this case was detected even better growth in the medium contain-
ing S. rebaudiana FOS than in the medium with OraftiÒ P95, a com-
mercial prebiotic product containing FOS from chicory.
From lactobacilli group, the strain L. gasseri PHM-7E1 was eval-
uated to utilize FOS from S. rebaudiana stems most efficiently,
where the highest increase in optical density after 24 h incubation
 S.M.S. Lopes et al. / Food Chemistry 225 (2017) 154–161 159

OH
H 6'
5' O
O 4' H H
H H
O H
3' 2' 1'
OH
H

O
O H
6 H O O H
H 3
5 H 2
H H
4
C H
O H 1
n
O H O
6 H O O H
H 3
5 H 2
H
H H C H
4 1
OH
OH

Fig. 3. Heteronuclear Correlation Map Multiple-Bond (HMBC) of the inulin from S. rebaudiana stems.

was measured. The strain L. fermentum RL25 demonstrated a sim-
ilar ability to ferment S. rebaudiana FOS and chicory FOS (OraftiÒ
P95) as carbon sources. For the other tested strains the bacterial
growth in the medium containing S. rebaudiana FOS was significant
(p < 0.05) lower than the medium containing prebiotic control.
The inulin molecules isolated from S. rebaudiana stems was
evaluated to be not a suitable energy source neither for bifidobac-
teria nor for lactobacilli, because no statistically significant
increase (p < 0.05) in optical density of the medium containing this
sugar source compared to the basal medium (without sugar) was
detected (Table 2).
The prebiotic control (OraftiÒ P95) was the most suitable for
comparison with FOS extracts of S. rebaudiana stems, because both
are polysaccharides with similar chemical structure and degree of
polymerization that provide more similar fermentation parame-
ters, unlike the Wilkins Chalgren broth (WCB) where glucose mole-
cules are used as carbon source, this medium it is easily fermented
and it is not a selective medium. The WCB was used as a control of
bacterial growth and viability.
Strain specificity in fermentation capacity was observed and
each strain utilized FOS from S. rebaudiana with varying intensity.
This agrees with the general knowledge about strain and substrate
specificity (Gibson, Probert, Loo, Rastall, & Roberfroid, 2004;
Moreno-Vilet et al., 2014).
The human gastrointestinal tract represents a competitive envi-
ronment for bacterias. Lactobacilli and bifidobacteria, belong to a
group of saccharolytic bacteria able to utilize a wide range of
mono-, oligo- and polysaccharides as energy source (Vernazza,
Gibson, & Rastall, 2005). The capability of microorganisms to fer-
ment sugars depends on their enzymatic equipment, on the pres-
ence of hydrolases and transportes. As example, the ability to
ferment fructan molecules by certain bacterial strains is related
to the presence of the b-fructofuranosidase, an inulinase responsi-
ble for the hydrolysis of fructan chain in terminal position b-(2 ?
1) (Karimi et al., 2015; Moreno-Vilet et al., 2014).
In this study, FOS molecules with DP < 6 were better used as
substrate by probiotic strains than inulin molecules (DP = 12), both
isolated from the same source. It is also known, that carbohydrates
utilization as substrate by bacteria is influenced to a significant
extent by their chemical structure, such as molecule branching,
glycosidic bonding and chain length, i.e. degree of polymerization
(Caleffi et al., 2015). The latter parameter has great influence on
the rate of FOS and inulin fermentability by bacteria where their
growth increases with decreasing of the degree of polymerization
(Mclaughlin et al., 2015; Moreno-Vilet et al., 2014).
The fermentation process in the colon is the result of complex
microbial metabolic activity and in the ecosystem of the human
colon, many different species and genera of microorganisms live
together. It is known that comensal relationships between certain
microorganisms exist. As for example, where strains able to cleave
oligofructose (B. longum) can provide monosaccharides for other
strains (Anaerostipes caccae) this condition provide mutual growth
of bacterial strains (Pompei et al., 2008; Ward, Ninonuevo, Mills,
Lebrilla, & German, 2007). These kinds of interactions are obviously
not reflected in in vitro studies.
The FOS molecules isolated from S. rebaudiana stems, showed a
chemical profile and degree of polymerization similar to commer-
cial FOS (OraftiÒ P95). However, the carbohydrate content was sig-
nificantly different between the samples. In the total sugar
determination, a higher concentration of sugars was found for
160 S.M.S. Lopes et al. / Food Chemistry 225 (2017) 154–161

Fig. 4. Heteronuclear Single Quantum Correlation (HSQC) of the fructo-oligosaccharides from S. rebaudiana stems.

Table 2
Fermentability of inulin and fructo-oligosaccharides (FOS) from S. rebaudiana stems by bifidobacteria and lactobacilli strains.

Bacterial strains Inulin Bacterial density (change in A540)

FOS OraftiÒ P95 WCH BM
a d b c
B. bifidum CCDM 559 0,64 ± 0,21 3.08 ± 0.28 2.47 ± 0.16 7.19 ± 0.57 1.82 ± 0.53c
B. animalis subsp. lactis Bb12 0,69 ± 0,15a 1.33 ± 0.10ab 4.95 ± 0.18d 3.98 ± 0.13a 0.30 ± 0.08a
B. breve CCDM 562 0,73 ± 0,23a 1.57 ± 0.07b 2.42 ± 0.23b 6.83 ± 0.13c 0.50 ± 0.03a
Bifidobacterium average 0.68 ± 0.29¥ 1.99 ± 0.33– 3.28 ± 0.33£ 6.00 ± 0.33Ɛ 0.87 ± 0.33¥
L. fermentum RL25 0,95 ± 0,15ab 3.72 ± 0.29e 3.72 ± 0.11c 7.04 ± 0.12c 2.29 ± 0.20d
L. animalis CCDM 382 1,24 ± 0,23bc 2.52 ± 0.25c 5.47 ± 0.20e 5.79 ± 0.17b 1.00 ± 0.05b
L. casei subsp. paracasei PE1 TB-P 0,94 ± 0,21ab 1.11 ± 0.08a 1.38 ± 0.13a 5.84 ± 0.12b 1.02 ± 0.14b
L. gasseri PHM-7E1 1,49 ± 0,30c 4.03 ± 0.20e 6.93 ± 0.21f 6.84 ± 0.43c 2.34 ± 0.08d
Lactobacillus average 1.15 ± 0.29¥ 2.85 ± 0.33– 4.38 ± 0.33£ 6.38 ± 0.33Ɛ 1.67 ± 0.33¥

Data are expressed as increase in turbidity of the bacterial suspension estimated from the increase in absorbance (A540) during 24 h of incubation; values are means from
triplicate determination ± standard deviation (SD).
OraftiÒ P95 – commercial prebiotic (FOS positive control), WCH – Wilkins-Chalgren broth (positive control), BM – basal medium without sugar (negative control).
abcdef
Data in columns with different superscripts differ (p < 0.05).
Ɛ¥–£
Data in lines with different superscripts differ (p < 0.05).

the commercial FOS (100%) in comparison with FOS extract from S.
rebaudiana (60%). Therefore, the best prebiotic effect observed for
the commercial FOS can be justified by the industrial process of
this product that results in a higher purity and high carbohydrate
content.
The Lactobacilli and Bifidobacteria populations play a key role
in bowel function in health and well-being. Therefore, the ade-
quate knowledge of the metabolism of certain types of polysaccha-
rides as substrate by certain bacteria strains, may provide
important information for the rational development of selective
food ingredients for bacterial strains as well as in the production
of synbiotic products (Mclaughlin et al., 2015; Rossi et al., 2005).
4. Conclusions
The results of the present study suggested that S. rebaudiana
stems can be a new and promising source for obtaining inulin
and fructo-oligosaccharides. The in vitro prebiotic assay of these
molecules indicated the strain specific ability to use fructans as
carbon sources. FOS molecules with low DP are preferably fer-
 S.M.S. Lopes et al. / Food Chemistry 225 (2017) 154–161
mented in vitro by beneficial microbiota than inulin molecules
which higher DP.
S. rebaudiana stems can be further explored as an alternative
source of the extraction of bioactive compound with commercial
application from waste of the sweetener industry.
Acknowledgments
The authors thank the Brazilian agencies: Conselho Nacional de
Desenvolvimento Científico e Tecnológico (CNPq, process no.
481915/2013-3) as well as the Coordenação de Aperfeiçoamento
de Pessoal de Nível Superior (CAPES) and the Complexo de Centrais
de Apoio à Pesquisa (COMCAP) of the State University of Maringá.
Authors also thank the Ministry of Education, Youth and Sports of
the Czech Republic (COST LD 14123).
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.2016.
12.100.
References
Ambalam, P., Raman, M., Purama, R. K., & Doble, M. (2016). Probiotics, prebiotics
and colorectal cancer prevention. Best Practice & Research Clinical
Gastroenterology, 30(1), 119–131.
Aravind, N., Sissons, M. J., Fellows, C. M., Blazek, J., & Gilbert, E. P. (2012). Effect of
inulin soluble dietary fibre addition on technological, sensory, and structural
properties of durum wheat spaghetti. Food Chemistry, 132(2), 993–1002.
Caleffi, E. R., Krausová, G., Hyršlová, I., Paredes, L. L. R., dos Santos, M. M., Sassaki, G.
L., ... de Oliveira, A. J. B. (2015). Isolation and prebiotic activity of inulin-type
fructan extracted from Pfaffia glomerata (Spreng) Pedersen roots. International
Journal of Biological Macromolecules, 80, 392–399.
Castro, R. J. S., Soares, M. H., Albernaz, J. R. M., & Sato, H. H. (2016). Biochemical
characterization of solvent, salt, surfactant and oxidizing agent tolerant
proteases from Aspergillus niger produced in different agroindustrial wastes.
Biocatalysis and Agricultural Biotechnology, 5, 94–98.
Chi, Z. M., Zhang, T., Cao, T. S., Liu, X. Y., Cui, W., & Zhao, C. H. (2011).
Biotechnological potential of inulin for bioprocesses. Bioresource Technology,
102(6), 4295–4303.
Crispín-Isidro, G., Lobato-Calleros, C., Espinosa-Andrews, H., Alvarez-Ramirez, J., &
Vernon-Carter, E. J. (2015). Effect of inulin and agave fructans addition on the
rheological, microstructural and sensory properties of reduced-fat stirred
yogurt. LWT - Food Science and Technology, 62(1), 438–444.
Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A., & Smith, F. (1956). Colorimetric
method for determination of sugars and related substances. Analytical
Chemistry, 28, 350–356.
Evans, M., Gallagher, J. A., Ratcliffe, I., & Williams, P. A. (2016). Determination of the
degree of polymerisation of fructans from ryegrass and chicory using MALDI-
TOF mass spectrometry and gel permeation chromatography coupled to
multiangle laser light scattering. Food Hydrocolloids, 53, 155–162.
Gibson, G. R., Probert, H. M., Loo, J. V., Rastall, R. A., & Roberfroid, M. B. (2004).
Dietary modulation of the human colonic microbiota: Updating the concept of
prebiotics. Nutrition Research, 17, 259–275.
Jackson, A. U., Tata, A., Wu, C., Perry, R. H., Haas, G., West, L., & Cooks, R. G. (2009).
Direct analysis of Stevia leaves for diterpene glycosides by desorption
electrospray ionization mass spectrometry. The Analyst, 134(5), 867–874.
Karimi, R., Azizi, M. H., Ghasemlou, M., & Vaziri, M. (2015). Application of inulin in
cheese as prebiotic, fat replacer and texturizer : A review. Carbohydrate
Polymers, 119, 85–100.
Lobo, A. R., Cocato, M. L., Borelli, P., Gaievski, E. H. S., Crisma, A. R., Nakajima, K., ...
Colli, C. (2011). Iron bioavailability from ferric pyrophosphate in rats fed with
fructan-containing yacon (Smallanthus sonchifolius) flour. Food Chemistry, 126
(3), 885–891.
Lopes, S. M. S., Francisco, M. G., Higashi, B., Almeida, R. T. R., Krausová, G., Pilau, E. J.,
... Oliveira, A. J. B. (2016). Chemical characterization and prebiotic activity of
fructo-oligosaccharides from Stevia rebaudiana (Bertoni) roots and in vitro
adventitious root cultures. Carbohydrate Polymers, 152, 18–725.
161
Lopes, S. M. S., Krausová Rada, V., Gonçalves, J. E., Gonçalves, R. A. C., & De Oliveira,
A. J. B. (2015). Isolation and characterization of inulin with a high degree of
polymerization from roots of Stevia rebaudiana (Bert.) Bertoni. Carbohydrate
Research, 411, 15–21.
Matulová, M., Husárová, S., Capek, P., Sancelme, M., & Delort, A.-M. (2011). NMR
structural study of fructans produced by Bacillus sp. 3B6, bacterium isolated in
cloud water. Carbohydrate Research, 346(4), 501–507.
Mclaughlin, H. P., Motherway, M. O. C., Lakshminarayanan, B., Stanton, C., Ross, R. P.,
Brulc, J., ... O’toole, P. W., & Van Sinderen, D. (2015). Carbohydrate catabolic
diversity of bifidobacteria and lactobacilli of human origin. International Journal
of Food Microbiology, 203, 109–121.
Miao, Y. Z., Lin, Q., Cao, Y., He, G. H., Qiao, D. R., & Cao, Y. (2011). Extraction of water-
soluble polysaccharides (WSPS) from Chinese truffle and its application in
frozen yogurt. Carbohydrate Polymers, 86(2), 566–573.
Misselbrook, T. H., Menzi, H., & Cordovil, C. (2012). Preface – Recycling of organic
residues to agriculture : Agronomic and environmental impacts. Agriculture,
Ecosystems and Environment, 160, 1–2.
Mitchell, C. M., Davy, B. M., Halliday, T. M., Hulver, M. W., Neilson, A. P., Ponder, M.
A., & Davy, K. P. (2015). The effect of prebiotic supplementation with inulin on
cardiometabolic health: Rationale, design, and methods of a controlled feeding
efficacy trial in adults at risk of type 2 diabetes. Contemporary Clinical Trials, 45,
328–337.
Moreno-Vilet, L., Garcia-Hernandez, M. H., Delgado-Portales, R. E., Corral-
Fernandez, N. E., Cortez-Espinosa, N., Ruiz-Cabrera, M. a., & Portales-Perez, D.
P. (2014). In vitro assessment of agave fructans (Agave salmiana) as prebiotics
and immune system activators. International Journal of Biological
Macromolecules, 63, 181–187.
Oliveira, A. J. B., Gonçalves, R. A. C., Chierrito, T. P. C., dos Santos, M. M., de Souza, L.
M., Gorin, P. A. J., ... Iacomini, M. (2011). Structure and degree of polymerisation
of fructooligosaccharides present in roots and leaves of Stevia rebaudiana (Bert.)
Bertoni. Food Chemistry, 129(2), 305–311.
Peshev, D., & Van den Ende, W. (2014). Fructans: Prebiotics and
immunomodulators. Journal of Functional Foods, 8(1), 348–357.
Pompei, A., Cordisco, L., Raimondi, S., Amaretti, A., Pagnoni, U. M., Matteuzzi, D., &
Rossi, M. (2008). In vitro comparison of the prebiotic effects of two inulin-type
fructans. Anaerobe, 14(5), 280–286.
Roberfroid, M., Gibson, G. R., Hoyles, L., McCartney, A. L., Rastall, R., Rowland, I., et al.
(2010). Prebiotic effects: Metabolic and health benefits. British Journal of
Nutrition, 104, S1–S63.
Rossi, M., Corradini, C., Amaretti, A., Nicolini, M., Pompei, A., Zanoni, S., et al. (2005).
Fermentation of fructooligosaccharides and inulin by bifidobacteria: A
comparative study of pure and fecal cultures. Applied and Environmental
Microbiology, 71, 6150–6158.
Salihu, A., Alam, Z., Abdulkarim, I., & Salleh, H. M. (2012). Lipase production : An
insight in the utilization of renewable agricultural residues. Resources,
Conservation & Recycling, 58, 36–44.
Serfaty, M., Ibdah, M., Fischer, R., Chaimovitsh, D., Saranga, Y., & Dudai, N. (2013).
Dynamics of yield components and stevioside production in Stevia rebaudiana
grown under different planting times, plant stands and harvest regime.
Industrial Crops and Products, 50, 731–736.
Shoaib, M., Shehzad, A., Omar, M., Rakha, A., Raza, H., Sharif, H. R., ... Niazi, S. (2016).
Inulin: Properties, health benefits and food applications. Carbohydrate Polymers,
147, 444–454.
Singh, R. S., Singh, R. P., & Kennedy, J. F. (2016). Recent insights in enzymatic
synthesis of fructooligosaccharides from inulin. International Journal of
Biological Macromolecules, 85, 565–572.
Suzuki, T., Maeda, T., Grant, S., Grant, G., & Sporns, P. (2013). Confirmation of
fructans biosynthesized in vitro from [1–13C]glucose in asparagus tissues using
MALDI–TOF MS and ESI–MS. Journal of Plant Physiology, 170, 715–722.
Tárrega, A., Rocafull, A., & Costell, E. (2010). Effect of blends of short and long-chain
inulin on the rheological and sensory properties of prebiotic low-fat custards.
LWT - Food Science and Technology, 43(3), 556–562.
Tavarini, S., & Angelini, L. G. (2013). Stevia rebaudiana Bertoni as a source of
bioactive compounds: The effect of harvest time, experimental site and crop age
on steviol glycoside content and antioxidant properties. Journal of the Science of
Food and Agriculture, 93(9), 2121–2129.
Vernazza, C. L., Gibson, G. R., & Rastall, R. A. (2005). In vitro fermentation of chitosan
derivatives by mixed cultures of human faecal bacteria. Carbohydrate Polymers,
60, 539–545.
Villegas, B., Tárrega, A., Carbonell, I., & Costell, E. (2010). Optimising acceptability of
new prebiotic low-fat milk beverages. Food Quality and Preference, 21(2),
234–242.
Ward, R. E., Ninonuevo, M., Mills, D. A., Lebrilla, C. B., & German, J. B. (2007). In vitro
fermentability of human milk oligosaccharides by several strains of
bifidobacteria. Molecular Nutrition & Food Research, 51, 1398–1405.