Structural Characterization of Cranberry Arabinoxyloglucan Oligosaccharides

Article abstract of DOI:10.1021/acs.jnatprod.8b01044

Cranberry (Vaccinium macrocarpon) products are widely available in North American food, juice, and dietary supplement markets. The use of cranberry is popular for the prevention of urinary tract infections (UTIs) and other reported health benefits. Preliminary findings by our research group indicate that arabinoxyloglucan oligosaccharides are present in cranberry products and may contribute to the antiadhesion properties of urine produced after cranberry consumption, but relatively little is known regarding the oligosaccharide components of cranberry. This report describes the isolation from two cranberry sources and the complete structure elucidation of two arabinoxyloglucan oligosaccharides through the use of carbohydrate-specific NMR spectroscopic and chemical derivatization methods. These compounds were identified as the heptasaccharide β-d-glucopyranosyl-(1γ4)-[α-d-xylopyranosyl-(1γ6)]-β-d-glucopyranosyl-(1γ4)-β-d-glucopyranosyl-(1γ4)-[α-l-arabinofuranosyl-(1γ2)-α-d-xylopyranosyl-(1γ6)]-β-d-glucopyranose (1) and the octasaccharide β-d-glucopyranosyl-(1γ4)-[α-l-arabinofuranosyl-(1γ2)-α-d-xylopyranosyl-(1γ6)]-β-d-glucopyranosyl-(1γ4)-β-d-glucopyranosyl-(1γ4)-[α-l-arabinofuranosyl-(1γ2)-α-d-xylopyranosyl-(1γ6)]-β-d-glucopyranose (2). Selected fractions and the isolated compounds were subjected to antimicrobial, cell viability, and E. coli antiadhesion assays. Results indicated that enriched fractions and purified compounds lacked antimicrobial and cytotoxic effects, supporting the potential use of such compounds for disease prevention without the risk for resistance development. Preliminary antiadhesion results indicated that mixtures of oligosaccharides exhibited greater antiadhesion properties than purified fractions or pure compounds. The potential use of cranberry oligosaccharides for the prevention of UTIs warrants continued investigations of this complex compound series.

Full text of DOI:10.1021/acs.jnatprod.8b01044

Article
pubs.acs.org/jnp
Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX
Structural Characterization of Cranberry Arabinoxyloglucan
Oligosaccharides
Kimberly M. Auker,† Christina M. Coleman,*,† Mei Wang,‡ Bharathi Avula,‡ Susanna L. Bonnet,§
Lindsey L. Kimble,⊥ Bridget D. Mathison,⊥ Boon P. Chew,⊥ and Daneel Ferreira*,†
†Department of BioMolecular Sciences, Division of Pharmacognosy, and the Research Institute of Pharmaceutical Sciences, School
of Pharmacy, University of Mississippi, University, Mississippi 38677, United States
‡National Center for Natural Products Research and the Research Institute for Pharmaceutical Sciences, School of Pharmacy,
University of Mississippi, University, Mississippi 38677, United States
§Department of Chemistry, University of the Free State, 205 Nelson Mandela Drive, Bloemfontein, 9301, South Africa
⊥School of Food Science, Washington State University, Pullman, Washington 99164-6376, United States
S
* Supporting Information
ABSTRACT: Cranberry (Vaccinium macrocarpon) products
are widely available in North American food, juice, and dietary
supplement markets. The use of cranberry is popular for the
prevention of urinary tract infections (UTIs) and other
reported health benefits. Preliminary findings by our research
group indicate that arabinoxyloglucan oligosaccharides are
present in cranberry products and may contribute to the
antiadhesion properties of urine produced after cranberry
consumption, but relatively little is known regarding the
oligosaccharide components of cranberry. This report
describes the isolation from two cranberry sources and the
complete structure elucidation of two arabinoxyloglucan
oligosaccharides through the use of carbohydrate-specific
NMR spectroscopic and chemical derivatization methods. These compounds were identified as the heptasaccharide β-D-
glucopyranosyl-(1→4)-[α-D-xylopyranosyl-(1→6)]-β-D-glucopyranosyl-(1→4)-β-D-glucopyranosyl-(1→4)-[α-L-arabinofurano-
syl-(1→2)-α-D-xylopyranosyl-(1→6)]-β-D-glucopyranose (1) and the octasaccharide β-D-glucopyranosyl-(1→4)-[α-L-arabino-
furanosyl-(1→2)-α-D-xylopyranosyl-(1→6)]-β-D-glucopyranosyl-(1→4)-β-D-glucopyranosyl-(1→4)-[α-L-arabinofuranosyl-
(1→2)-α-D-xylopyranosyl-(1→6)]-β-D-glucopyranose (2). Selected fractions and the isolated compounds were subjected to
antimicrobial, cell viability, and E. coli antiadhesion assays. Results indicated that enriched fractions and purified compounds
lacked antimicrobial and cytotoxic effects, supporting the potential use of such compounds for disease prevention without the
risk for resistance development. Preliminary antiadhesion results indicated that mixtures of oligosaccharides exhibited greater
antiadhesion properties than purified fractions or pure compounds. The potential use of cranberry oligosaccharides for the
prevention of UTIs warrants continued investigations of this complex compound series.
formation,9 and anti-inflammatory effects on human immune
roducts prepared from the American cranberry [Vaccinium
components.11 To date, however, the exact composition of the
Pmacrocarpon Ait., (Ericaceae)] are widely available and of
NDM cranberry fractions used in these studies has not been
significant commercial interest in food, juice, and dietary
supplement markets.1 Cranberry products are popularly used
for a variety of reported health benefits, but most notably for
the promotion of urinary tract health and for the prevention of
urinary tract infections (UTIs).2,3 Compounds known to be
responsible for the aroma, flavor, color, processing stability,
and antioxidant properties of cranberry include terpenes,
aromatic acids, monosaccharides, simple phenolics, polyphe-
nols, and a wide variety of flavonoids.2,3 High molecular weight
nondialyzable materials (NDM) derived from cranberry
products have also been reported and have been found to
have various bioactivities including the inhibition of adhesion
by bacteria4−9 and viruses,10 inhibition of bacterial biofilm
fully characterized.4−11
The literature has been dominated since 1998 by the
hypothesis that polymeric flavan-3-ols known as proanthocya-
nidins (PACs) are the primary components responsible for the
antiadhesion properties of cranberry products and of urine
after cranberry consumption.12,13 Findings by our research
group, however, indicate that arabinoxyloglucan oligosacchar-
Special Issue: Special Issue in Honor of Drs. Rachel Mata and
Barbara Timmermann
Received: December 9, 2018
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
ides are an additional class of antiadhesive compounds found
in cranberry products and are likely to contribute to the
antiadhesion properties of urine produced after cranberry
consumption.14,15 This oligosaccharide hypothesis has been
supported by reports that cranberry fractions enriched in
xyloglucan oligosaccharides are able to prevent bacterial
adhesion and interfere with biofilm formation.16,17 Recent
analyses of cranberry NDM have indicated the presence of
xyloglucan oligosaccharides in addition to PACs and suggest
that relatively low molecular weight oligosaccharides may form
microaggregates in solution that are retained by the dialysis
membranes used in the preparation of such materials.9 These
findings shed new light on the original reports concerning
cranberry NDM4,5 that later led to studies that identified PACs
as the antiadhesive constituents of cranberry12,13 and make it
clear that more must be known about cranberry oligosacchar-
ides as they relate to the biological properties of cranberry
products.
viability assays and to a P-fimbriated Escherichia coli-
uroepithelial cell (EC-UEC) antiadhesion assay that was
under development at the time of this project.
RESULTS AND DISCUSSION
■
Isolation. The oligosaccharides 1 and 2 were isolated from
the aqueous extracts of two different cranberry materials, SN
and CJ (Figure S1, Supporting Information). These materials
were obtained from two separate suppliers and differed in that
the SN material was a cranberry fruit powder, while the CJ
material was a juice powder. The initial appearance of these
parent materials in aqueous solution differed markedly (Figure
S1, Supporting Information), but later fractions and sub-
fractions behaved similarly and appeared to contain similar
constituents as determined by NMR data and various chemical
analyses (data not shown). Both yielded oligosaccharide
1
mixtures with similar profiles as visible by HPLC and H
NMR spectroscopy (Figures 1 and 2; Figures S5, S6, S9, and
The carbohydrate composition of cranberry products has
not been well characterized, and nothing was known regarding
cranberry oligosaccharides when this project was initiated in
2011. The gel-forming characteristics of cranberries, as
exhibited by the classic cranberry sauce, led to early structural
investigations of complex carbohydrates such as pectins and
other cell-wall-associated insoluble polysaccharides related to
cellulose.18 Studies on soluble carbohydrates have been limited
to mono- and disaccharides with the use of standards for
comparisons.19 Metabolomics studies on cranberry chemical
composition with regard to carbohydrates have been limited by
the availability of known compound databases and by the
suitability of analysis and detection methods for complex
carbohydrates.20 Recent investigations of the oligosaccharide
profile of cranberry have used a pectinase treatment of
cranberry hulls followed by various separation methods and
chemical analyses of the resulting fractions with MALDI-TOF/
TOF MS/MS characterization of the xyloglucan portion of this
material.16,17,21,22 These reports propose arabinoxyloglucan
structures similar to that of the octasaccharide isolated from
porcine urine after cranberry consumption,14,15 but do not
provide full structural characterization of isolated compounds.
Our research group has found that the oligosaccharide
profile of cranberry materials is similar to that found in urine
after cranberry consumption, indicating that a portion of the
cranberry oligosaccharide series may be excreted intact into
urine and thereby contribute to the ability of cranberry
materials to prevent UTIs.14,15 Excretion of these compounds
by the fecal route is also likely, and beneficial gut microbes
have been identified that are able to ferment cranberry
xyloglucan mixtures.22 This class of compounds may therefore
also contribute to various gut-associated health benefits
attributed to cranberry materials such as systemic anti-
inflammatory effects.23 Full characterization of all components
of this complex mixture of cranberry oligosaccharides will
therefore be necessary to fully investigate the metabolism and
excretion profiles of these compounds in mammalian systems
and to further investigate their biological properties. Toward
this end, this report describes the isolation of two
arabinoxyloglucan oligosaccharides from two distinct sources
of cranberry material and provides details for the complete
structure elucidation of both compounds using extensive NMR
data analyses in combination with a series of optimized
chemical derivatization methods. Selected fractions and the
isolated compounds were subjected to antimicrobial and cell
Figure 1. 1H NMR spectra of cranberry oligosaccharide fractions
SNA-03 and CJA2-03 (D2O, 400 MHz). Resonances do not appear
outside of the region shown. See Figure S9 (Supporting Information)
for full spectra and expansions. Both of these materials are similar to
the CJA1-02 material discussed in the companion manuscript15
(Figure S7, Supporting Information).
S10, Supporting Information), and both parent materials
yielded the same two oligosaccharides via parallel purification
methods (Figures S10−S13, Supporting Information).
Compounds 1 and 2 were obtained via sequential
fractionation and purification steps, including Sephadex LH-
20 size exclusion chromatography and preparative HPLC with
C18 and polyamine stationary phases (Figures S2−S6, S11−
S13 and Tables S1 and S2, Supporting Information). Selected
Sephadex LH-20 fractions were determined to contain
1
oligosaccharides based on H NMR spectroscopic (Figure 1)
and chromatographic data and comparisons to previously
characterized material, including CJA1-02 and known mono-
saccharides (Figures S3−S10, Supporting Information).15
Fractions SNA-02 and CJA2-02 contained gylcans with
arabinosyl and glucosyl subunits, as determined by 1D and
HSQC NMR data (Figure S10, Supporting Information), but
material from these fractions could not be sufficiently purified
with the resources available for this study (Figures S5 and S6,
B
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Supporting Information). Fractions SNA-03 and CJA2-03 each
yielded two subfractions (Fr.4 and Fr.5 from each) that
1
contained similar profiles of components by H NMR and
HPLC-ELSD profiling (Figures S10−S13, Supporting In-
formation). Both Fr.4 subfractions, from both SNA-03 and
CJA2-03, contained 1, while both Fr.5 subfractions from the
two parent materials contained 2 (Figure 2).
Fractions SNA-03 and -04 and CJA2-03, -04, and -05
contained the primary oligosaccharide profile of interest
(Figures S5 and S6, Supporting Information) and, combined,
accounted for 15% and 29% of SN and CJ starting materials,
respectively. Although the SN fruit powder contained a lower
overall percentage of oligosaccharide content, fraction SNA-03
constitutes a larger portion of the parent aqueous extract
(44.3%) as compared to the parallel fraction CJA2-03 (22.6%).
This is likely attributable to the fact that ∼50% of the dry
weight of the original SN material was removed as pectins or
insoluble material through EtOH precipitation and subsequent
filtration. Fractions SNA-03 and CJA2-03 were used for further
purification steps, yielding compounds 1 and 2, which
composed ∼0.13% and 0.29% w/w of SN and ∼0.33% and
0.52% w/w of CJ, respectively. The combination of selected
subfractions from the two parent materials provided sufficient
material for full structure elucidation.
Figure 2. Selected portion of the HPLC-ELSD preparative separation
chromatograms for fractions CJA2-03 and SNA-03 (Atlantis dC18).
Compounds 1 and 2 were both obtained from the respective Fr.4
(∼23−24 min) and Fr.5 (∼24−25 min) fractions of both parent
materials (arrows). This chromatogram shows the relative resolution
and initial elution profiles for these fractions. See Figures S11 and S12
(Supporting Information) for representative ELSD and UV chromato-
grams for these separations.
MS Analysis. The presence of neutral glycosyl residues in
1
both 1 and 2 was indicated by initial H NMR spectroscopic
data and was confirmed using trimethylsilyl derivatization and
GC-MS analysis (data not shown). Consequently, ionization
using MSDTOF (data not shown) of both compounds in
native form was minimal, requiring permethylation of free
hydroxy groups to improve the intensity of indicative mass
peaks.24 Per-O-methylated 1 and 2 dissolved in 1 mM NaOH
(50% MeOH(aq)) were analyzed by UHPLC/QTOFMS and
yielded a high abundance of [M + NH4]+ molecular ions,
consistent with patterns observed in analyses of olive pulp and
other oligosaccharides.25,26 Per-O-methylated 1 eluted at 4.9
and 5.1 min and yielded [M + NH4]+ and [M + Na]+
molecular ions (m/z 1360.6652 and 1365.6263, respectively),
with [M + NH4]+ at 100% relative intensity in both eluted
peaks (Figure S14, Supporting Information). Per-O-methylated
2 eluted at 5.2 and 5.4 min and yielded the same types of
molecular ions as per-O-methylated 1 at m/z 1520.7398 and
1525.6988, respectively (Figure S14, Supporting Information).
The observation of two elution peaks is consistent with the
presence of α- and β-anomers of the methyl glycoside
derivative of the reducing glycosyl unit present in each native
compound. Ions were also detected at m/z 689.3450 and
769.3771, corresponding to doubly charged molecular ions [M
+ 2NH4]2+ as observed for xylan-type oligosaccharides.25
Fragmentation patterns resulting from tandem-MS analyses
can aid in determining the order of glycosyl units; however
Scheme 1. Alditol Acetate Derivatization of an Example Glucopyranosyl Unit for 1 or 2
C
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 1. Results from the Chemical Derivatization Analyses for Compounds 1 and 2
alditol acetate analysis
D/L configuration
glycosyl linkage analysis
distinguishing mass fragments
a
glycosyl
unit
glycosyl
RT
molar
ratio
glycosyl
RT
ring size and linkage
position
b
identity
(min)
stereoisomer
(min)
(m/z)
RRT
Compound 1
A
B
C
D
E
glucose
26.88
4
D-(+)-glucose
18.02
4,6-linked glcp
4-linked glcp
4,6-linked glcp
terminal glcp
2-linked xylp
terminal xylp
terminal araf
118, 261
118, 233
118, 261
118, 161, 162, 205
117, 190
0.776
0.654
0.776
0.497
0.569
0.438
0.394
xylose
21.73
19.82
27.06
2
1
2
D-(+)-xylose
18.76
18.62
18.02
F
G
117, 118, 161, 162
118, 161, 162
arabinose
L-(+)-arabinose
D-(+)-glucose
Compound 2
A
B
C
D
E
F
G
H
glucose
4,6-linked glcp
4-linked glcp
4,6-linked glcp
terminal glcp
2-linked xylp
2-linked xylp
terminal araf
terminal araf
118, 261
118, 233
118, 261
118, 161, 162, 205
117, 190
117, 190
118, 161, 162
118, 161, 163
0.776
0.655
0.776
0.497
0.569
0.569
0.394
0.394
xylose
21.8
19.9
1
1
D-(+)-xylose
18.76
18.62
arabinose
L-(+)-arabinose
a
b
RT, retention time. RRT, relative retention time, RT of sample relative to the internal standard myo-inositol.
Scheme 2. Partially Methylated Alditol Acetate Derivatization Involving a Partial Structure of 2
minimal fragmentation was observed in the QTOFMS spectra
for the permethylated derivatives of 1 and 2. This is consistent
with the branching pattern, as the level of fragmentation is
negatively correlated with the level of branching in
carbohydrates.27 The molecular formulas, degrees of polymer-
ization (DP), and numbers of pentose and hexose sugar
residues were calculated from the exact masses: 1 (C39H66O33),
DP = 7 with four hexose and three pentose moieties, and 2
(C44H74O37), DP = 8 with four hexose and four pentose
moieties.
Alditol Acetate Analysis. As both 1 and 2 were neutral
molecules, alditol acetate (AA) derivatization was appropriate
for the identification of individual glycosyl units and
subsequent molar ratio calculations. AA derivatives of 1 and
2, generated through successive hydrolysis, reduction with
NaBD4, and O-acetylation,28 yielded indicative fragment pairs
of deuterated pentose (m/z 290/291) and hexose (m/z 361/
362) AA derivatives. This process is shown for an example
glucopyranosyl unit in Scheme 1. The presence of glucose,
arabinose, and xylose was confirmed based on comparisons of
retention times and MS data to standards (Table 1; Figure S15
and Table S3, Supporting Information). Molar ratios,
calculated based on quantification of AA derivatives of the
monomers within each oligosaccharide, were applied to the DP
to establish the presence of one arabinosyl, two xylosyl, and
four glucosyl moieties in 1 and two arabinosyl, two xylosyl, and
four glucosyl moieties in 2. This result is consistent with the
numbers of pentosyl and hexosyl moieties calculated based on
the observed mass of each compound.
D/L Configuration. The absolute configurations of the
constituent monosaccharide moieties of 1 and 2 were
determined using the parent fraction SNA-03 in order to
preserve purified compounds. The sample and standards were
hydrolyzed and treated with L-cysteine methyl ester and phenyl
isothiocyanate to form optically active phenyl thiocarbamate
(PTC) diastereomers, which were analyzed by HPLC-UV.29
Comparisons of the retention times of standards to those of
the PTC derivatives of SNA-03 (Table 1; Table S4, Figures
S16−S19, and Section S4d, Supporting Information) indicated
the presence of D-(+)-glucose, L-(+)-arabinose, and D-
(+)-xylose, thus confirming the absolute configurations of the
monomeric constituent units of 1 and 2.
Glycosyl Linkage Analysis. Derivatives obtained through
per-O-methylation followed by AA derivatization identified the
ring size and linkage position of each monomeric unit upon
GC-MS analysis (Scheme 2). Key patterns of O-methyl and O-
acetyl substitutions resulted in distinguishing mass fragments
for the partially methylated alditol acetate (PMAA) derivatives
of 1 and 2 (Table 1). Relative retention times and mass
fragments were compared to standards to characterize the
carbohydrate monomers in each compound (Table S5 and
Figure S20, Supporting Information). Compound 2 differed
D
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
1
Figure 3. H NMR spectra for compounds 1 and 2, with anomeric and selected protons indicated (D2O, 600 MHz). See Figures S21−S23
(Supporting Information) for full spectra and expansions.
Figure 4. 13C DEPT NMR spectra for compounds 1 and 2, with anomeric and selected carbons indicated (D2O, 150 MHz). See Figures S24−28
(Supporting Information) for 13C NMR spectra and DEPT full spectra and expansions.
from 1 in the presence of a second terminal arabinofuranosyl
moiety and replacement of the terminal xylopyranosyl unit
with a 2-linked xylopyranosyl moiety.
Information), including two correlations pertaining to the α-
and β-anomers of the reducing Glc-A moiety.
The anomeric orientation of each glycosyl unit in 1 was
determined through examination of chemical shift and
coupling constant values for anomeric protons compared to
reference spectra (Table 2, Figure 3; Figure S8, Supporting
Information).30−32 H-1 and H-2 of Ara-G (δH 5.04 and 4.07,
respectively) gave 3J1,2 = 1.9 Hz, indicating an α-orientation for
NMR Data Analysis. The NMR-based structural elucida-
tion of each compound was initiated independently. Later
comparisons between the 1D and 2D NMR spectra and partial
structures for both compounds helped to assign chemical shift
values for the glucosyl backbones (Figures 5 and 6, Table
3
H-1, in agreement with published J1,2 values for α-L-
1
2). The H NMR spectrum of 1 (Figure 3; Figures S21−S23,
arabinofuranosides.31,32
Supporting Information) contained eight proton resonances in
the anomeric region (δH 4.3−5.1) that integrated to a total of
seven protons, thus indicating the presence of one reducing
glycosyl unit (Glc-A, Figure 5). The anomeric protons gave
HSQC correlations to eight carbon resonances in the δC 90−
110 region (Table 2, Figure 7; Figures S27−S29, Supporting
4
The C1 preferred conformations of the xylopyranose and
glucopyranose residues are similar, resulting in α-anomeric
protons with deshielded (e.g., δH > 4.70) chemical shifts and
coupling constants of 3J1,2 < 5 Hz, while those with β-anomeric
protons exhibit shielded (e.g., δH < 4.70) chemical shifts and
coupling constants of J1,2 > 5 Hz. Accordingly, both xylosyl
E
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 5. Key COSY, H2BC, HMBC, HSQC-TOCSY, and NOESY correlations for compound 1 with 13C and 1H chemical shifts indicated at each
position [δC (δH)].
1
Figure 6. Key COSY, H2BC, and HMBC correlations for compound 2 with 13C and H chemical shifts indicated at each position [δC (δH)].
3
residues (δH 4.97 and 4.83) were α-anomers, with J1,2 = 2.74
Information) in combination with additional 2D NMR
experiments (Figures S30−S37, Supporting Information).
Several methine carbons resonating in the δC 69.0−85.0
region were assigned using COSY, H2BC, and HMBC
correlations. The congestion of resonances in the δH 3.10−
3.85 region required the use of HSQC-TOCSY and phase-
sensitive COSY experiments to assign C-5 and C-6 of Glc-A
and C-6 of Glc-C (Figure 5; Figures S35 and S37, Supporting
Information). The HSQC-TOCSY experiment established the
assignment of the resonance at δC 79.0 to C-5 of Glc-A, but a
corresponding H-5 resonance could not be identified due to a
and 3.32 Hz, respectively, while Glc-B, Glc-C, and Glc-D (δH =
4.43, 4.38, and 4.42, respectively) were β-anomers with J1,2
3
=
7.6, 7.1, and 7.1 Hz, respectively. The reducing nature of Glc-A
resulted in duplicate resonances for the anomeric proton (δH =
5.09 and 4.53) and carbon (δC = 91.8 and 95.7, respectively)
corresponding to both the α- and β-anomers (3J1,2 = 3.6 and
7.6 Hz, respectively) (Table 2).
The remaining proton and carbon resonances (Figures 3 and
4; Figures S21−26, Supporting Information) were assigned
using the HSQC spectrum (Figure S29, Supporting
F
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
1
Table 2. H and 13C NMR Assignments for Compounds 1 and 2 (D2O, 600/150 MHz)
compound 1
compound 2
compound 1
compound 2
1
δ H (J in
a
1
1
1
C
δ
13C
δ H (J in Hz )
δ
13C
Hz)
C
δ
13C
δ H (J in Hz)
δ
13C
102.5 4.45 d (8.0)
δ H (J in Hz)
Glcp-A
β-anomer
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
95.7
73.8
74.3
69.9
79.0
66.3
91.8
71.2
73.2
69.9
4.53 d (7.6)
3.16 t (9.4)
3.52
95.6
73.5
74.3
69.9
79.2
66.1/66.3
91.7
71.2
4.58 d (8.0)
3.23
3.55
3.87
3.62
3.87
5.14 d (3.9)
3.51
3.23
3.87
Glcp-D
1
2
3
4
5
6
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
102.5 4.42 d (7.1)
72.7
74.0
69.4
79.4
60.6
98.6
78.8
71.9
69.4
61.2
98.8
71.5
73.0
69.4
61.5
3.28 t (3.8)
3.54
3.53
72.6
73.9
69.4
79.5
60.5
98.5
78.7
71.8
69.4
61.1
98.5
78.7
71.8
69.4
61.1
3.32
3.59
3.34
3.62
3.77, 3.84−5
5.01 d (3.8)
3.49
3.78
3.59
3.66, 3.76
5.01 d (3.8)
3.49
3.78
3.59
3.83
3.62
b
3.83
3.70, 3.79
4.97 d (2.7)
3.45
3.72
3.53
3.60−3.62
4.83 d (3.3)
3.41−3.43
3.59
Glcp-A
α-anomer
5.09 d (3.6)
3.41−3.43
3.71
Xylf-E
Xylf-F
Araf-G
Araf-H
73.5
69.9
3.83
66.3
3.83
Glcp-B
Glcp-C
102.5 4.43 d (7.6)
102.5
72.6
73.9
69.4
79.4
4.47 d (8.2)
3.32
3.59
3.59
3.62
3.80, 3.72,
3.87, 3.83
4.43 d (7.8)
3.24
3.42
3.40
72.7
74.0
69.4
79.2
60.0
3.28 t (8.8)
3.54
3.29 t (9.2)
3.55
3.43−3.44
3.66, 3.76
109.2 5.04 br. d (1.9) 109.1 5.09 br. d (2.0)
3.64, 3.75
59.8/60.0
81.1
76.4
83.8
61.2
NA
NA
NA
NA
NA
4.07
3.81
3.96 m
3.59, 3.73
81.0
76.3
83.7
61.1
4.12 dd (3.8, 1.6)
3.86
4.00 m
1
2
3
102.6 4.38 d (7.1)
102.5
73.0
75.4
75.9
69.4
78.9
73.1
75.5
3.18 t (7.9)
3.36
3.64, 3.76
c
109.1 5.09 br. d (2.0)
c
81.0
76.3
83.7
61.1
4.12 dd (3.8, 1.6)
3.86
4.00 m
4
5
6
69.4
79.0
66.1
3.29 t (9.2)
3.62
3.34
3.62
3.87 dd
b
3.78, 3.88 dd (11.0, 5.0) 66.1/66.3
3.64, 3.76
a
b
Only coupling constants that could be measured accurately are shown. Two resonances for C-6 of Glc-A and C in 2 were present at 66.1 and
c
66.3; however, the two could not be distinguished with the present data, and both are assigned as 66.1/66.3. Two carbon and proton chemical
shift pairs could be assigned for C-3 of Glc-C.
Ara-G and Xyl-F side chains, a phase-sensitive COSY spectrum
(Figure S37, Supporting Information) was used to differentiate
between the two unassigned oxymethylene carbons at δC 66.1
and 66.3 in 1. The objective of this experiment was to detect a
correlation with the well-resolved proton resonance observed
at δH 3.88 (dd) associated with the carbon resonating at δC
66.1 via HSQC data. The observed correlation between the δH
3.88 proton resonance and H-5 (δH 3.62) of Glc-C
unequivocally established the chemical shifts of C-6 and H-6
of Glc-C in 1 as δC 66.1 and δH 3.88, respectively, and, by
elimination, C-6 of Glc-A at δC 66.3 in 1. The nature of the β-
(1→4)-glucose backbone for 1 was established based on
comparisons to spectra for cellotetraose and compound 2
(Figures S38 and S39, Supporting Information) and the
presence of indicative NMR correlations. The NOESY
spectrum of 1 (Figure 5; Figure S36, Supporting Information)
contained correlations between H-3 (δH 3.52) of Glc-A and H-
1 (δH 4.43) of Glc-B and between H-3 (δH 3.54) of Glc-B and
H-1 (δH 4.38) of Glc-C. The presence of a (1→2)-linkage
between Ara-G and Xyl-E for 1 was indicated by an HMBC
correlation (Figure 5; Figure S44, Supporting Information)
Figure 7. Expansion of the anomeric region of the HSQC spectrum
for compound 1 with labeled correlations (D2O, 600/150 MHz). See
Figure S29 (Supporting Information) for full HSQC spectrum of
compound 1.
between H-1 (δH 5.04) of Ara-G and C-2 (δC 78.8) of Xyl-E.
The (1→6)-linkages between xylosyl and glucosyl residues
were indicated by HMBC correlations between H-1 (δH 4.97)
of Xyl-E and C-6 (δC 66.3) of Glc-A and between H-1 (δH
4.83) of Xyl-F and C-6 (δC 66.1) of Glc-C, thus locating the
E→G- and F- glycosyl moieties on alternating glucosyl
residues.
lack of correlation. As the assignments of C-6 for Glc-A and
Glc-C were integral for determining the locations of the Xyl-E-
G
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Based on NMR correlations and the derivatization results,
the structure of 1 was defined as the heptasaccharide β-D-
glucopyranosyl-(1→4)-[α-D-xylopyranosyl-(1→6)]-β-D-gluco-
pyranosyl-(1→4)-β-D-glucopyranosyl-(1→4)-[α-L-arabinofura-
nosyl-(1→2)-α-D-xylopyranosyl-(1→6)]-β-D-glucopyranose.
pyranosyl-(1→6)]-β-D-glucopyranosyl-(1→4)-β-D-glucopyra-
nosyl-(1→4)-[α-L-arabinofuranosyl-(1→2)-α-D-xylopyranosyl-
(1→6)]-β-D-glucopyranose.
Comparisons of 1 and 2 to Known Xyloglucans.
Naturally occurring xyloglucans are a common type of
hemicellulose present in the primary cell walls of higher
plants.33,34 The solubilization and enzymatic depolymerization
of hemicellulosic polysaccharides such as xyloglucans is a
feature of fruit ripening35 and contributes to the physical
softening of fruit.36 Xyloglucans are composed of repeating
units of a cellotetraose backbone containing β-D-(1→4)-
glucopyranosyl residues with α-D-(1→6)-xylopyranosyl sub-
stituents (side chains) attached to the glucosyl moieties.
Xylopyranosyl moieties are often substituted with galactose or
α-L-(1→2)-arabinofuranosyl moieties,37 increasing the solubil-
ity of these compounds when dissociated from cellulose.
Arabinoxyloglucan polymers have been identified in tomato,
potato, and tobacco (Solanaceae),37 and similar parent
polymers are likely to be the source of the oligosaccharides
found in cranberry products. As compounds 1 and 2 were both
identified in both source materials, it is possible that these
compounds originate from the natural process of cranberry
fruit ripening. A similar series of oligosaccharides has also been
detected in cranberry concentrate material (Ocean Spray
Cranberries, Inc., Lakeville-Middleboro, MA, USA) using
methods similar to those presented here.38 Further work will
be necessary to determine the effects of various commercial
processing and product preparation methods on cranberry
oligosaccharide content and profiles in different products.
Compounds 1 and 2 are previously uncharacterized
oligosaccharides that possess structural characteristics similar
to those reported elsewhere for cranberry oligosacchar-
ides.16,17,21,22 The repeating oligosaccharide units identified
from Solanaceous plants37 differ from those of 1 and 2 in that
the branch side chains of Solanaceous oligosaccharides are
located on adjacent glucose moieties. This observation may be
due, in part, to the methods used to generate the various
oligomers from the parent polymers being studied. Other
reports of cranberry xyloglucans provide MALDI-TOF/TOF
MS/MS fragmentation pattern evidence for both adjacent and
alternate placements of branch side chains.16,17 PTC
derivatives of the parent fraction SNA-03 yielded only four
monomeric units, D-(+)-glucose, D-(+)-xylose, L-(+)-arabinose,
and D-(+)-galactose, indicating that additional oligosaccharides
may be present in the cranberry series that contain galactosyl
moieties. From these data, it is likely that the overall cranberry
oligosaccharide profile contains both types of branching motifs
(alternate and adjacent), in conjunction with variations in side
chain composition.
1
The H NMR spectrum of 2 (Figure 3; Figures S21−S23,
Supporting Information) contained seven resonances in the
anomeric region that integrated to eight protons. The
anomeric proton resonances at δH 5.01 (3J1,2 = 3.8 Hz, 2H)
and 5.09 (3J1,2 = 2.0 Hz, 2H) were identified as those of α-
xylosyl and α-arabinosyl residues, respectively, while the
resonances at δH 5.14 (3J1,2 = 3.9 Hz) and 4.58 (3J1,2 = 8.0
Hz) corresponded to the respective α- and β-anomers of a
reducing glucosyl residue (Table 2 and Figure 6). The
anomeric protons gave HSQC correlations to five carbon
resonances in the δC 91−110 region (Table 2, Figure 4;
Figures S22−S28 and S40, Supporting Information), with the
two at δH 5.14 and 4.58 correlating with the α- and β-anomers
of Glc-A (δC 91.7 and 95.6, respectively), one (δH 5.01; 2H)
correlating with the two xylosyl moieties (δC 98.5; 2C; Xyl-E
and Xyl-F), and one (δH 5.09; 2H) correlating with the two
arabinosyl moieties (δC 109.1; 2C; Ara-G and Ara-H). The
3
three remaining anomeric protons (δH 4.47, J1,2 = 8.2 Hz;
3
3
4.43, J1,2 = 7.8 Hz; and 4.45, J1,2 = 8.0 Hz) were consistent
with nonreducing β-glucosyl residues (δC 102.5; 3C) and
corresponded to the Glc-B, Glc-C, and Glc-D moieties. The
remaining proton and carbon resonances were assigned using
the HSQC spectrum in conjunction with COSY, H2BC, and
HMBC data (Table 2, Figure 6; Figures S40−47, Supporting
Information).
HMBC correlations (Figure 6; Figure S44, Supporting
Information) between H-1 of the Ara moieties (δH 5.09) and
C-2 of the Xyl moieties (δC 78.7) confirmed the presence of
two arabinosyl-(1→2)-xylosyl linkages identical to those
present in 1. Correlations between H-1 of the Xyl moieties
(δH 5.01) and C-6 of Glc-A and C (δC 66.1/66.3) also
matched the branching pattern of 1. While neither NOESY nor
HMBC spectra provided specific correlations indicating the
linkages between the glucosyl moieties due to overlap of the
Glc-B, Glc-C, and Glc-D resonances, the same β-(1→4)-
glucosyl backbone was assigned due to the similarities
observed in the spectra of 1, 2, and cellotetraose (Figures
S38 and S39, Supporting Information).
In contrast with 1, compound 2 showed considerable
conformational flexibility. This was indicated by more than two
correlations between proton resonances and the C-6 oxy-
methylene carbons of Glc-B (δC 60.0) and Glc-D (δC 60.5,
Table 2). The δC 60.0 resonance of Glc-B is observed as two
small signals at δC 59.85 and 60.0, which were considered
together as a single resonance. C-3 of Glc-C (δC 75.4) also
gave correlations indicative of conformational flexibility and
labile hydrogen bonding. Hydrogen bonding between the
tetrahydropyran heteroatom and the 3-hydroxy of an adjacent
glucosyl moiety is a common feature of a cellulosic β-D-(1→4)-
glucose backbone.33 In the case of 2, C and H resonances
vicinal to C-3/H-3 of Glc-C showed dual correlations with two
separate C-3/H-3 pairs (δC 75.4/δH 3.42 and 75.9/δH 3.40) in
COSY, H2BC, and HMBC spectra (Figures S42−S44,
Supporting Information). Resonances at δ 75.4 and 75.9
were therefore both assigned to C-3 of Glc-C.
Bioactivity. Antimicrobial and Cell Viability Assays. All
cranberry materials tested were ineffective at inhibiting the
growth of the pathogenic microorganisms used in the NCNPR
antimicrobial screening assay (Table S6, Supporting Informa-
tion). Compounds 1 and 2 and oligosaccharide-enriched
fractions also did not exhibit cytotoxic effects against two
different breast cancer cell lines (Table S7, Supporting
Information). These preliminary data suggest that the reported
antiproliferative39,40 and antimicrobial41,42 effects of cranberry
extracts are due to compounds other than arabinoxyloglucan
oligosaccharides. In this case, the lack of activity for the tested
compounds and fractions is considered to be a positive result,
as it supports possible mechanisms of microbial disease
prevention or treatment that involve noncytotoxic effects. If
Based on NMR correlations and the derivatization results,
the structure of 2 was defined as the octasaccharide β-D-
glucopyranosyl-(1→4)-[α-L-arabinofuranosyl-(1→2)-α-D-xylo-
H
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
compounds such as cranberry oligosaccharides were to be
developed into products intended for the management or
prevention of diseases such as UTIs, resistance development
associated with long-term or repeated use would be less likely
to cause problems.
An apparent loss of bioactivity as fractionation proceeds is a
well-known phenomenon in the field of natural products
research, and multiple possible reasons for this occurrence
could be proposed. As several independent lines of evidence
support the hypothesis that oligosaccharides are involved in
the in vivo antiadhesive properties of cranberry materi-
als,9,14−17,38 it is proposed that such antiadhesion effects are
possibly the result of (1) mixtures of various oligosaccharides
within the series, (2) mixtures of oligosaccharides with other
classes of compounds, (3) higher overall concentrations of
oligosaccharides than could be tested in the present study, or
(4) some combination of possibilities 1−3. The inherent in
vitro nature of the bioassay further confounds the issue, as (1)
various in vivo processes may contribute to the urinary
antiadhesive effects of cranberry materials and (2) the
antiadherence assay used may not be effective at detecting
the mechanism by which cranberry oligosaccharides prevent E.
coli adhesion. For further discussion see Section S3
(Supporting Information).
Significance of Structural Features for Future
Studies. This is the first report describing the complete
structure elucidation of two arabinoxyloglucan oligosaccharides
from cranberry products. The new compounds 1 and 2 are
representative of a series of structurally related oligosaccharides
that possess chemical features that make these compounds
difficult to detect and characterize without the application of
specialized methods and techniques. These compounds are
neutral carbohydrates that do not readily ionize or produce
diagnostic fragments using standard MS-based detection
methods. They are UV-transparent and require specialized
detection via evaporative light scattering or refractive index
detectors. They are of relatively high molecular mass (>500
Da) and are highly polar, requiring that relatively large mass
quantities (>5 mg) of samples be solubilized in water or
DMSO for spectroscopic analyses and that samples be dried
via lyophilization. While methods are available for the
complete structural analysis of these compounds, such
methods are destructive, requiring fairly large quantities of
purified starting material for multiple sample derivatizations
and analyses. Some specialized resources are available for
studying xyloglucans such as those found in cranberry
materials, including NMR reference spectra and MS
fragmentation data,30,37,43 but the general applicability of
such resources is confounded by the structural complexities
inherent to carbohydrates.44
These compounds are difficult to isolate due to their polar
nature and the close structural similarities of the compounds in
the series. The use of Sephadex LH-20 as an initial enrichment
method was a serendipitous discovery,14 and subsequent
attempts by our research group to use carbohydrate-specific
isolation and purification methods, such as carbohydrate
extraction resins, diol chromatography, hydrophilic-interaction
chromatography, and various other sorbents and separation
techniques, met with limited success. This is possibly due to
the fact that the methods attempted are typically used for the
isolation of carbohydrate monomers, dimers, or trimers, and
therefore do not readily enrich for or resolve neutral
oligosaccharides with DP > 6. Carbohydrates are often
considered to be “nuisance compounds” in terms of natural
products isolation. Methods and sorbents pertaining to
carbohydrates are often intended to remove carbohydrates
from a sample rather than to intentionally enrich for specific
types of carbohydrates. The optimization of isolation or
Antiadhesion Properties. The EC-UEC antiadhesion assay
used in this study was under development during the time
period when samples were submitted. A series of unexpected
methodological and sample-management issues (Section S2,
Supporting Information) arose during testing, and the data
reported herein are therefore considered to be preliminary.
The relative activity of each test sample was based on
qualitative assessments of the dose−response curves obtained
at the sample concentrations tested relative to that of the
positive control, myricetin (Figure S48, Supporting Informa-
tion). Samples with an overall negative slope were considered
active (e.g., Figure S49 C and D, Supporting Information),
while samples with an overall positive slope were considered
inactive (e.g., Figure S50 A, Supporting Information). Samples
in which a negative slope appears at higher concentrations
(e.g., Figure S50 B and C, Supporting Information) were
considered potentially active, as these samples may have
resulted in increased antiadhesion effects if tested at higher
concentrations.
The aqueous SN extract (SNA) and subfractions SNA-03
and SNA-04 exhibited dose-dependent antiadhesive effects as
determined by the observation that the slopes of the dose−
response curves for these samples were negative, similar to that
of myricetin. This effect appears to be greater with SNA, as this
sample yielded a dose−response curve with an increased slope
and lower final cpm values relative to myricetin and
subfractions SNA-03 and SNA-04. The HPLC subfractions
that contained compounds 1 and 2 (SNA-03 Fr.4, Fr.5, and
Fr.6) and the purified compounds 1 (from SNA-03 Fr.4) and 2
(from SNA-03 Fr.5) showed a lack of antiadhesion activity
relative to the parent material SNA-03 and the positive control
(Figures S50 and S51, Supporting Information). Fractions
SNA-03 Fr.5, Fr.6, and Fr.9 (Figure S50 B, C and Figure S52
E, Supporting Information) resulted in dose−response curves
that displayed negative slopes above the test concentration of 7
mg/mL, suggesting that higher test concentrations may have
resulted in greater antiadhesion properties. The remaining
SNA-03 fractions, Fr.1, Fr.3, Fr.7, and Fr.8, showed a lack of
antiadhesion activity (Figure S52, Supporting Information).
The higher antiadhesion activity of the parent fraction SNA
relative to that of the Sephadex LH-20 fraction SNA-03 and
the subsequent RP-HPLC fractions of SNA-03 (e.g., Fr.4, Fr.5)
supports the theory that multiple compounds as mixtures may
exhibit greater antiadhesion activity than fractions or purified
components. Further studies on Sephadex LH-20 fractions
containing only oligosaccharide components obtained from
cranberry juice concentrate38 provide additional evidence that
mixtures of cranberry oligosaccharides prevent the adhesion of
uropathogenic E. coli adhesion to uroepithelial cells. The
relative contributions of individual cranberry oligosaccharide
components to overall antiadhesion activity, however, will
require further studies such as those that use proportional
recombinations of fractions or purified compounds to identify
additive or synergistic effects. Studies of this type and further
investigations of concentration effects were beyond the scope
of the present project due to limitations in both material and
available resources.
I
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
1
Figure 8. Comparison of the H NMR spectra for 2 (11.2 mg) and HF2-2P1t22A17 (∼2 mg), an enriched fraction from porcine urine, showing
similar composition (D2O, 400 MHz). The HF2-2P1t22A17 material was obtained from a fraction of porcine urine that exhibited antiadhesive
properties.15 NMR and analytical HPLC-ELSD data for HF2-2P1t22A17 indicate that it contains a single primary component that is structurally
related to an octasaccharide isolated from a related fraction. See the accompanying article15 for a detailed discussion of the oligosaccharide series
detected in porcine urine after cranberry consumption.
detection methods is made additionally difficult by the fact that
applicable standards for this class of compound are not readily
available. Knowledge of the complete structures of at least two
compounds from the cranberry oligosaccharide series may aid
future studies seeking to develop improved enrichment and
isolation methods.
Bacterial fermentation of cranberry xyloglucans involved the
hydrolysis of a terminal arabinosyl moiety from the reducing
end of the molecule (e.g., Ara-G for 1 and 2). The ability of
certain strains of beneficial bacteria to utilize specific substrates
is known to impact the overall composition of the gut
microbiota by multiple mechanisms.51
While the structural features of cranberry oligosaccharides
make them difficult to detect, isolate, and chemically
characterize, these same features may contribute to specific
biological properties for these compounds. Several structural
features support the hypothesis that these compounds could be
directly absorbed and excreted into mammalian urine. High-
MW, neutral oligosaccharides from other sources, such as
human milk,45,46 with as many as eight glycosyl units, have
been detected in the urine of human infants, and cranberry
oligosaccharides fall within this molecular size range.16 The
presence of Ara-Xyl-, Xyl-, and other types of glycosyl side
chains16 may contribute to the aqueous solubility of this class
of compounds, as has been observed for other plant-derived
polymeric xyloglucans37 and by molecular dynamics simu-
lations investigating xyloglucan intramolecular hydrogen
bonding and solvation.47,48 The possibility of direct absorption
for cranberry oligosaccharides early in the digestive process is
supported by the acid stability of the cellulose-type, β-(1→4)-
glucose backbone49,50 and by the detection of compounds with
similar elution profiles and structural features in porcine urine
(Figure 8) as discussed in the accompanying manuscript.15
The conformational flexibility observed by NMR for
compounds 1 and 2 in aqueous solution, combined with
intra- and intermolecular hydrogen bonding, may contribute to
the observed formation of microaggregates in solution.9 The
formation of such microaggregates may influence both the
absorption and excretion of these compounds and their
antiadhesion and other biological properties.
The gut is the primary reservoir for strains of E. coli that
cause community acquired UTIs.52 Cranberry compounds,
including oligosaccharides with particular structural features,
may help decrease the population of this pathogen in the gut
reservoir by influencing the binding of certain strains of E. coli
to mammalian epithelial cells and by providing a preferred,
plant-cell-wall-based substrate for bacteria, enabling fecal
clearance. Cranberry xyloglucan mixtures have been able to
inhibit the in vitro adhesion of several different strains of E. coli
to different types of epithelial cell lines.16 The investigated
strains of E. coli expressed various adhesins, including type 1,
type IV, P-fimbriae, and the E. coli common pilus, and were
affected differently by fractions of cranberry materials that
contained various profiles of oligosaccharides and other
compounds. A study investigating the adhesion of two
enteropathogenic E. coli strains (O157:H7 and O18:K1:H7)
to spinach leaves found that the E. coli common pilus allows
these organisms to adhere to plant cells by targeting arabinosyl
residues of pectin and other plant cell wall components.53 This
study used oligosaccharides with known structures and found
that various structural features, such as backbone and side
chain lengths, monosaccharide composition, branching pat-
terns, and overall DP, influenced E. coli binding and adhesion
processes. These types of specific oligosaccharide structural
features, in combination with the type of E. coli adhesin being
expressed, may therefore be of relevance for the various
biological properties of cranberry oligosaccharides, especially
with regard to the prevention of E. coli adhesion.
Structural characteristics of cranberry oligosaccharides, such
as the presence of terminal arabinosyl residues and character-
istic branching patterns, may influence the effects these
compounds have on fecal microbes and thereby contribute
to the gut health benefits of cranberry materials.23 Specific
strains of Bifidobacterium longum have been identified that can
ferment cranberry xyloglucans similar to those described in this
report at low rates, while other strains and organisms are not
capable of using these compounds as a carbon source.22
In conclusion, the series of structurally related oligosacchar-
ides represented by compounds 1 and 2 is likely to play
important roles in the bioactivities and health benefits
attributed to the American cranberry (V. macrocarpon).
Enriched oligosaccharide fractions and purified compounds
exhibited a lack of antimicrobial and cytotoxic effects,
supporting the potential use of such compounds for the
prevention of disease without the risk for resistance develop-
ment. Preliminary results from the antiadhesion assay used in
J
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
reduced pressure with a water bath temperature of ∼40 °C. The SN
aqueous fraction (250 mL) was further treated with EtOH (500 mL)
to precipitate proteins and pectins using sonication (20 min) and
centrifugation at 3000 rpm for 5 min. The resulting fractions were
concentrated and dried via lyophilization to give EtOAc fractions SNE
(0.31 g) and CJE (0.41 g) and aqueous SNA (14.04 g) and CJA (38.4
g).
this study indicated that mixtures of oligosaccharides exhibited
greater antiadhesion properties than more purified fractions or
pure compounds. The potential use of cranberry oligosacchar-
ides for the prevention of infection, especially UTIs, warrants
continued investigation into the structural characteristics of
this complex series of compounds.
SNA and CJA (13.7 and 25.3 g, respectively) were individually
fractionated using Sephadex LH-20 (356 g dry resin, 7 × 37 cm
column) equilibrated in 70% EtOH. Each sample was dissolved in
70% EtOH and centrifuged for 5 min at 2000 rpm to remove
precipitate, yielding final load volumes of 60 and 100 mL for SNA and
CJA, respectively. Both samples were fractionated using gravity-driven
isocratic elution with 70% EtOH, flow rates of 1−2 mL/min, and the
collection of 15−20 mL fractions. A time-lapsed color image of the
fractionation of CJA material on Sephadex LH-20 (separation
CJA1)15 is included in the Supporting Information (Figure S2).
Fractions were combined based on visual assessment and elution
volume to yield five major fractions from each starting material: SNA-
02 (1.14 g), SNA-03 (3.62 g), SNA-04 (2.44 g), SNA-05 (2.11 g),
SNA-06 (3.06 g) and CJA2-02 (2.09 g), CJA2-03 (3.98 g), CJA2-04
(1.74 g), CJA2-05 (1.81 g), and CJA2-06 (3.81 g). Oligosaccharide-
containing fractions of interest were identified using analytical HPLC-
PDA-ELSD (Figures S3−S6, Table S1, and Section S4, Supporting
Information). For the SNA column, fractions containing oligosac-
charides, SNA-03 and SNA-04, eluted between 320 and 620 min and
were dried to yield lavender-maroon powders. For the CJA2 column,
fractions containing oligosaccharides, CJA2-03, CJA2-04, and CJA2-
05, eluted between 350 and 810 min and were dried to yield pink-
lavender powders. Material eluting before 320 and 350 min from the
SNA and CJA2 columns, respectively, yielded fractions SNA-02 and
CJA2-02.
Fractions CJA2-03 (1.0 g) and SNA-03 (3.17 g) were dissolved in
H2O (4 and 9 mL, respectively), filtered (0.20 μm), and fractionated
by RP-HPLC using a Waters Atlantis dC18 (19 × 250 mm, 10 μm)
column. Injection volumes ranged from 400 to 600 μL, and fractions
were eluted using a gradient mobile phase of H2O to 30% MeOH
over 30 min at a flow rate of 25 mL/min. Analytes were detected
using a dual-wavelength UV detector (210 and 254 nm) connected in
series to an ELSD. A flow splitter in line between the UV detector and
the ELSD allowed for an ELSD inlet solvent flow rate of 1−2 mL/min
and parallel collection of the remaining eluent. The ELSD used N2 as
the carrier gas at a flow rate of 0.8 SLM, a nebulizer temperature of 50
°C, and an evaporator temperature of 100 °C. Multiple fractions were
collected and combined based on retention times and similar
retention profiles (Table S2, Figures S11 and S12, Supporting
Information). For the SNA-03 fractionation, analytes eluting at 23.8−
24.5 and 24.5−25.5 min were collected and pooled to yield fractions
SNA-03 Fr.4 (149 mg) and SNA-03 Fr.5 (212 mg), respectively
(Figure 2). For the CJA2-03 fractionation, analytes eluting between
23.4−24.3 and 24.3−25.3 min were pooled to yield CJA2-03 Fr.4 (56
mg) and CJA2-03 Fr.5 (97 mg), respectively (Figure 2). These
selected subfractions (SNA-03 Fr.4, SNA-03 Fr.5, CJA2-03 Fr.4, and
CJA2-03 Fr. 5) were analyzed by 1H NMR spectroscopy (Figure S10,
Supporting Information) and chosen for further purification.
SNA-03 Fr.4 (80.1 mg), SNA-03 Fr.5 (169.4 mg), and CJA2-03
Fr.5 (60 mg) were fractionated by NP-HPLC using a YMC
Polyamine II column (20 × 250 mm, 5 μm) and an isocratic mobile
phase of 63% MeCN(aq). Samples were dissolved in sufficient
volumes of 50% MeOH(aq) to give concentrations of 0.1 mg/mL,
and injection volumes ranged from 100 to 200 μL per run. SNA-03
Fr.4 yielded 1 (31.0 mg) with a retention time of 25 min, while SNA-
03 Fr.5 and CJA2-03 Fr.5 both yielded 2 (65 and 20 mg, respectively)
with a retention time of 28 min (Figure S13, Supporting
Information). Analytical NP-HPLC using a YMC Polyamine II
column and conditions similar to those used for the final preparative
separation showed that CJA2-03 Fr.4 contained a major component
that eluted at the same retention time as that of SNA-03 Fr.4.
However, the CJA2-03 Fr.4 material was lost during handling before
final preparative purification could be achieved.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Solvents and Chemicals.
Solvents were purchased from Fisher Chemicals (Thermo-Fisher
Scientific, Waltham, MA, USA) and were ACS certified or HPLC-
grade as appropriate unless otherwise stated. DMSO and MeOH were
dried with 3 Å molecular sieves. Chemical reagents included Tri-Sil
reagent (from Thermo-Fisher Scientific) and L-cysteine methyl ester
and phenyl isothiocyanate (from Sigma-Aldrich, Inc., St. Louis, MO,
USA). All carbohydrate standards and myo-inositol were purchased
from Sigma-Aldrich unless otherwise stated. Cellotetraose was
purchased from Dextra (Reading, UK), and Sephadex LH-20
(Lipophilic Sephadex LH-20, 25−100 μm bead size) was from
Sigma-Aldrich. High-purity NMR solvents DMSO-d6 (99.99%) and
D2O (99.99%) were obtained from Cambridge Isotope Laboratories,
Inc. (Andover, MA, USA).
Equipment. Analytical HPLC separations were performed using a
Waters 2695 separations module equipped with an autosampler and a
Waters 996 photodiode array (PDA) detector (Waters Corporation,
Milford, MA, USA) and connected in series to a Polymer Laboratories
PL-ELS 2100 evaporative light scattering detector (ELSD).
Preparative HPLC separations were performed using a Waters Delta
Prep 4000 HPLC equipped with a Waters 2487 dual wavelength
absorbance detector connected in series to a Polymer Laboratories
PL-ELS 1000 ELSD with a flow splitter for collection of eluent.
UHPLC-MS/MS analyses were performed using an Agilent LC-1200
Infinity Series (1290) UHPLC system coupled to a QTOFMS (model
#G6530A, Agilent Technologies, Palo Alto, CA, USA) with data
acquisition and analysis controlled by Agilent MassHunter Acquis-
ition software Ver. A.02.00 and processed by MassHunter Qualitative
Analysis software Ver. B.02.00. GC-MS was performed using an
Agilent Packard 6890 Series GC system coupled with a Hewlett-
Packard 5973 mass selective detector, and data acquisition and
analyses were performed using Agilent ChemStation and MestreNova
software (Mestrelab Research, Escondido, CA, USA). NMR spectra
were recorded on Bruker Avance III 400 MHz spectrometers
equipped with 3 mm probes and either Ultrashield or Ultrashield
Plus magnets. Additional NMR spectra for compounds 1 and 2 were
obtained on a Bruker Avance II 600 MHz spectrometer (Bruker
Biospin, Reinstetten, Germany).
Isolation. Two starting materials were used in this investigation:
Sundown Naturals brand cranberry fruit powder (100 × 475 mg
capsules; SN) (Rexall Sundown Inc., Boca Raton, FL, USA) and
spray-dried cranberry hull extract powder (CJ), provided by Ocean
Spray Cranberries, Inc. (Lakeville-Middleboro, MA, USA). Color
images of these materials and their labeling are included in the
Supporting Information (Figure S1). The CJ material was created by
spray-drying cranberry juice (cranberry hull extract) that had been
produced via proprietary methods. The CJ powder was fully soluble in
water and produced a dark burgundy solution that was similar in
appearance to commercially available cranberry juice products
prepared from cranberry juice concentrate. The SN material appeared
to be composed of ground, lyophilized cranberries mixed with
powdered silica, as determined by available label information, product
appearance, and behavior in solution.
The isolation methodology was based on methods developed and
used to isolate oligosaccharides from porcine urine produced after the
feeding of CJ cranberry powder.14,15 The SN and CJ powders (40.0 g
each) were dissolved separately in H2O (150 mL) and centrifuged for
5 min at 1800 rpm, and the precipitate was washed with H2O (1 ×
100 mL and 1 × 50 mL) to yield water-soluble supernatants (300 mL
each). The supernatants were extracted with EtOAc (4 × 150 mL),
and aqueous fractions were concentrated by evaporation under
K
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
NMR Analysis. Fractions were characterized by 1D and 2D NMR
spectroscopy using both D2O (99.99%) and DMSO-d6 (99.99%). Pu-
rified compounds were dissolved in a minimum amount of D2O
(99.99%), and the following spectra were recorded for both materials:
1H, 13C, DEPT (Figures 3 and 4; Figures S21−S28, Supporting
Information), COSY, HSQC, HMBC, H2BC, NOESY, TOCSY, and
HSQC-TOCSY (Figures S29−S36 and S40−S47, Supporting
Information). A phase-sensitive COSY experiment was additionally
carried out for 1 using a pulse program with the optimized 90 deg
pulse of 11 μs and a relaxation time (d1) of 1 s (Figure S37,
Supporting Information). Water suppression was applied to 1H NMR
spectra of pure compounds using standard procedures.
of the hydrolyzed glycosyl units was done in a solution of NaBD4 (10
mg/mL) in 1 M NH4OH (200 μL) at room temperature overnight
followed by neutralization with glacial HOAc in MeOH and drying.
Any remaining NaBD4 was removed by repeated addition and drying
of MeOH/glacial HOAc (9:1). O-Acetylation was performed with
Ac2O (250 μL) in TFA (99.9%, 230 μL) at 50 °C for 20 min. 2-
Propanol was added to the cooled sample, and the mixture was dried.
Na2CO3 (0.2 M, 2 mL) was added and alditol acetate derivatives were
extracted using CH2Cl2 (2 mL). The extract was washed with H2O (2
× 1 mL) and concentrated to ∼100 μL in preparation for GC-MS
analysis.
Each sample (3 × 5 uL/injection) was separated on a Restek Rtx-
2330 [30 m, 0.25 mm i.d., 0.2 μm film thickness (df)] column with an
injector temp of 250 °C, a split ratio of 20:1, helium as the carrier gas,
and a mass scan mode from m/z 50 to 500. An initial temperature of
80 °C was held for 2 min, with a first ramp at 30 °C/min to a final
temperature of 170 °C and a second ramp at 4 °C/min to a final
temperature of 240 °C, which was held for 20 min.
Quantification of Alditol Acetate Derivatives. In order to
determine the relative molar quantities of each monomer in the
oligosaccharides, a six-point calibration curve was created using the
averaged peak areas (PA) from three injections of the standard
(AAstd). The detector response factor (RF; slope of the linear
calibration curve) was used to calculate a relative response factor
(RRF) of each monosaccharide standard (S) relative to the internal
standard (IS), as follows: RRF = RFS/RFIS. The concentration (C) of
each monosaccharide moiety (S) in the oligosaccharide samples was
calculated using the corresponding standard monosaccharide RRF
(e.g., the glucose RRF was used to calculate the concentration of
Permethylation. Compounds 1 and 2 (200−500 μg of each)
were permethylated as described previously54 with modifications as
follows. Base was freshly prepared using 4 mL of dry DMSO added to
100 μL of 50% NaOH (w/w), dried with repeated rinsing with
DMSO, and dissolved in 2 mL of DMSO. Base (200 μL) was added
to each carbohydrate sample and stirred briefly, after which
iodomethane (100 μL) was added and the samples were stirred at
room temperature for 40 min. The addition of base and iodomethane
and stirring for 40 min was repeated and followed by addition of H2O
(2 mL) to quench the reaction, yielding a cloudy solution. The
remaining iodomethane was removed under a low flow of N2. The
dried per-O-methylated carbohydrates were purified by solid-phase
extraction (SPE; Sep-Pak C18 12 cc Vac Cartridge, 2 g). The SPE
cartridge was conditioned with sequential aliquots of 10 mL each of
MeOH and MeCN and 15 mL of H2O, in that order. The dried
extracts were loaded onto the cartridge with H2O and eluted with
MeCN. The samples were dried under N2 prior to UHPLC/QTOF-
ESIMS analysis using an Agilent Bonus RP18 (2.1 × 100 mm, 1.8
μm) column with a flow rate of 0.25 mL/min and a gradient of 3%
MeOH(aq) to 100% MeOH over 5 min followed by isocratic elution
with MeOH for 3 min. A 5 min wash with MeOH followed by an
equilibration period of 13 min was completed between each run. The
column temperature was set at 35 °C, and the injection volumes
ranged from 3 to 5 μL, with all samples prepared in 1 mM NaOH in
50% MeOH(aq). Total ion chromatogram peaks were assigned
according to the mass of the compounds. The conditions of the
QTOFMS were as follows: ESI source with Jet Stream technology
using a drying gas (N2) at 11.0 L/min and 250 °C; nebulizer, 35
PSIG, sheath gas at 10 mL/min and 325 °C; capillary, 3500 V;
skimmer 65 V; Oct RF V, 750 V; fragmentor voltage, 100 V. The
sample collision energy was set to 65 V. Samples were analyzed in ESI
positive mode, and spectra were acquired across the range of m/z
100−1700 with accurate mass measurements obtained through the
use of a reference ion correction using m/z 121.0509 (protonated
purine) and 922.0098 [protonated hexakis(1H,1H,3H-
tetrafluoropropoxy)phosphazine or HP-921]. The reference solution
was introduced into the ESI source using a T-junction with an Agilent
Series 1200 isocratic pump (Agilent Technologies, Santa Clara, CA,
USA) containing a 100:1 splitter with a flow rate of 20 μL/min.
Alditol Acetate Derivatization. Alditol acetate derivatives of 1
and 2 were formed via hydrolysis, reduction with NaBD4, and O-
acetylation. AA derivatives of the standards L-(+)-arabinose, D-
(+)-glucose, L-(−)-xylose, D-(+)-fucose, L-(+)-rhamnose, D-(+)-gal-
actose, and myo-inositol and of 1 and 2 were prepared individually
using a modified version of a reported procedure.55 Individual
monosaccharide standards [100 μg each of L-(+)-arabinose, D-
(+)-glucose, L-(−)-xylose, D-(+)-fucose, L-(+)-rhamnose, D-(+)-gal-
actose] were combined with 20 μg of the internal standard, myo-
inositol, in separate screw-cap tubes. A standard set of multiple
monosaccharides (AAstd) was used to create a GC-MS calibration
curve for quantification purposes and consisted of 50 μg each of L-
(+)-arabinose, D-(+)-glucose, and L-(−)-xylose and 20 μg of myo-
inositol. Compounds 1 and 2 (∼100 μg each) were combined with 20
μg of myo-inositol in separate screw-cap tubes and were hydrolyzed
via incubation in 2 M trifluoroacetic acid (TFA) at 121 °C for 2 h. 2-
Propanol (200 μL) was added to the cooled samples followed by
drying under N2 while sitting in a water bath at 40 °C. The addition
and subsequent drying of 2-propanol was repeated twice. Reduction
glucose in the oligosaccharide sample) with the formula CS = (CIS
×
PAS)/(RRFS × PAIS). The concentration of each monomeric moiety
in the oligosaccharide sample was used to calculate the relative molar
ratio of the monomers present. The concentration of the hexose
moiety (glucose) was divided by its approximate molar mass of 180,
and, likewise, the pentose moieties (arabinose and xylose) were
divided by 150. The resulting values were compared to determine the
relative molar ratio of each monomeric moiety.
Partially Methylated Alditol Acetate Derivatization. PMAA
derivatives of 1 and 2 (400 μg each) were prepared using the
permethylation procedure as described above, immediately followed
by the alditol acetate procedure. Derivatives were reduced to a volume
of ∼100 μL in CH2Cl2 and analyzed by GC-MS, as described above.
The ring size and linkage positions for the specific glycosyl moieties of
1 and 2 were determined using GC-MS via comparison of relative
retention times (RRT; relative to the internal standard) to those of
prepared PMAA monosaccharide standard derivatives (Table S5 and
Section S1, Supporting Information). RRT data were combined with
MS-fragmentation pattern information, and the identification of
distinguishing mass fragments56 was used to aid assignments due to
the potential overlap of RRTs.
Phenyl Thiocarbamate Derivatization. Phenyl thiocarbamate
derivatives of SNA-03 and the standards D-(+)-glucose, L-
(+)-arabinose, D-(−)-arabinose, L-(−)-xylose, D-(+)-xylose, L-
(+)-rhamnose, D-(+)-galactose, and D-(+)-mannose were formed as
follows. Each sample (1 mg) was hydrolyzed in 2 M HCl for 1 h with
stirring in an oil bath at 100 °C. Samples were neutralized by repeated
addition of H2O and evaporation, and subsequent steps were
performed as described previously.55 The samples were cooled, and
the retention times of the resulting optically active diastereoisomers
derived from 1 and 2 were compared to those of the standards and a
reagent blank by analytical scale HPLC using a Gemini C18 (4.6 ×
250 mm, 110 Å, 5 μm) column with a flow rate of 1 mL/min and a
gradient of H2O (0.1% formic acid)/MeCN (90:10) to 100% MeCN
over 45 min. A 5 min wash period with MeCN and a 15 min
equilibration period were applied between each injection. The UV
max plot and the UV absorbance at 210 nm were monitored, and each
sample was injected twice to verify the retention times for each peak,
with injection volumes of 2 μL for individual samples and standards
and 3 μL for samples spiked with standards. An isocratic method with
L
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
a mobile phase of 27% MeCN(aq) was employed for samples spiked
with L-(−)-xylose and L-(+)-arabinose.
control, solvent control, and two positive controls (myricetin and a
cranberry extract enriched for A-type proanthocyanidins; Figure S48,
Supporting Information). Inhibition curves were collected with four
concentration points, resulting in serial dilutions of the highest
concentration attainable. The highest concentration tested varied
depending on sample availability. The IC50 value was calculated for
positive controls but was unobtainable for cranberry samples (Section
S2, Supporting Information).
Antimicrobial Assays. These assays were performed as a part of
the antimicrobial screening program available at the National Center
for Natural Products Research (NCNPR), at the University of
Mississippi, according to previously reported methods57,58 modified
from the Clinical and Laboratory Standards Institute (CLSI) methods
[formerly the National Committee for Clinical and Laboratory
Standards (NCCLS)].59−62 Samples tested included CJ, CJE, CJA,
fraction SNA-03, and compound 2. All samples were tested against a
panel of 10 pathogenic organisms including five fungi (Candida
albicans ATCC 90028, C. glabrata ATCC 90030, C. krusei ATCC
6258, Cryptococcus neoformans ATCC 90113, and Aspergillus fumigates
ATCC 204305) and five bacteria (Staphylococcus aureus ATCC
29213, methicillin-resistant S. aureus ATCC 33591, E. coli ATCC
35218, Pseudomonas aeruginosa ATCC 27853, and Mycobacterium
intracellulare ATCC 23068). Samples of CJ, CJE, and CJA were
additionally tested against the protozoans Plasmodium falciparum (D6
clone) and Leishmania donovani. Amphotericin B served as the
positive control for the L. donovani assay (producing 100% inhibition
under the parameters used) and for the antifungal assay. Ciprofloxacin
served as the positive control for bacterial pathogens. Percent growth
inhibition was calculated relative to positive and negative controls,
and samples were considered to be inactive if they produced growth
inhibition of less than 50% (Table S6, Supporting Information). In
the antibacterial and antifungal assays, CJ, CJE, and CJA samples were
tested at a concentrations of 50 μg/mL; in the antiprotozoal assay
these samples were tested at a concentration of 15.9 μg/mL against P.
falciparum and at 80 μg/mL against L. donovani. SNA-03 was tested at
concentrations of 200, 40, and 8 μg/mL, and compound 2 was tested
at 20, 4, and 0.8 μg/mL.
Heptasaccharide (1). β-D-Glucopyranosyl-(1→4)-[α-D-xylopyra-
nosyl-(1→6)]-β-D-glucopyranosyl-(1→4)-β-D-glucopyranosyl-(1→
4)-[α-L-arabinofuranosyl-(1→2)-α-D-xylopyranosyl-(1→6)]-β-D-glu-
1
copyranose: white, amorphous powder; moderately hygroscopic; H
and 13C NMR data, Figures 3−5, Table 2, Figures S21, S22, S24, S25,
S27, and S28 (Supporting Information); 2D NMR data, Figure 7,
Figures S29−S37 (Supporting Information); QTOFESIMS of per-O-
methylated compound (positive ion mode) [M + Na]+ at m/z
1365.6545, indicative of a molecular formula of [C59H106O33Na]+
(Figure S14, Supporting Information).
Octasaccharide (2). β-D-Glucopyranosyl-(1→4)-[α-L-arabinofura-
nosyl-(1→2)-α-D-xylopyranosyl-(1→6)]-β-D-glucopyranosyl-(1→4)-
β-D-glucopyranosyl-(1→4)-[α-L-arabinofuranosyl-(1→2)-α-D-xylo-
pyranosyl-(1→6)]-β-D-glucopyranose: white, amorphous powder;
1
moderately hygroscopic; H and 13C NMR data, Figures 3, 4, and
6, Table 2, Figures S21, S23, S24, and S26−S28 (Supporting
Information); 2D NMR data, Figures S40−S47 (Supporting
Information); QTOFESIMS of per-O-methylated compound (pos-
itive ion mode) [M + Na]+ at m/z 1525.7012, indicating a molecular
formula of [C66H118O37Na]+ (Figure S14, Supporting Information).
ASSOCIATED CONTENT
* Supporting Information
■
S
Breast Cancer Cell Viability Assay. Breast cancer cells T47D
and MDA-MB-231 were seeded into a 96-well culture plate at 30 000
cells per well and incubated at 37 °C for 24 h in 200 μL of 10% fetal
bovine serum (FBS) DMEM/F12 medium supplemented with 50 μg/
mL penicillin/streptomycin. Compounds 1 and 2 (100 μL) were
added into separate wells at concentrations of 10, 30, and 100 μM.
Fraction SNA-03 was added at 30 and 100 ppm. Samples were tested
in triplicate. Chlorhexamide (CHX; 100 μL, 10 μM) was used as a
positive control in addition to a medium control. The cultures were
incubated at 37 °C for 48 h under normoxic conditions. The cells
were fixed with 10% trichloroacetic acid for 1 h at 4 °C, washed four
times with H2O, and allowed to air-dry. Sulforhodamine B (0.4%) in
1% HOAc was added, and samples were allowed to sit for 15 min at
room temperature. The culture wells were washed four times with 1%
HOAc and allowed to air-dry. Tris HCl buffer (200 μL, 10 mM) was
added, and the plates were placed on a rotary table and vortexed for 5
min. Plates were read at 490/690 nm on a microplate reader. Percent
inhibition was calculated as follows: % Inhibition = (1 − XS)/(Xctrl),
where XS is the average absorbance of the triplicate wells incubated
with a sample and Xctrl is the average absorbance of the triplicate
media control wells (Table S7, Supporting Information).
Escherichia coli-Uroepithelial Cell Antiadhesion Assay.
Samples were submitted to the EC-UEC antiadhesion assay while it
was under development. A report of this assay has since been
published.63 Briefly, the procedure used for the submitted cranberry-
derived samples was as follows: human uroepithelial cells (UEC;
HTB-4) were seeded at 1 × 105 cells per well in a 96-well culture plate
and rinsed with phosphate-buffered saline (PBS). P-fimbriated E. coli
(ATCC 700928) were subcultured in colony forming antigen (CFA)
agar to promote expression of P-fimbriae64 and incubated for 4 h at 35
°C with [3H]-uridine. The cells were rinsed with PBS and incubated
with cranberry samples for 30 min at 35 °C. [3H]-E. coli and UECs
were incubated together in a ratio of 400:1, respectively, for 1 h at 37
°C and washed with PBS. Radioactivity of the wells was measured by
liquid scintillation.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.8b01044.
Complete NMR spectra of 1 and 2; additional tables and
figures containing chromatographic, spectroscopic, and
carbohydrate composition data; color images of the
starting materials and an example Sephadex LH-20
separation with CJA material; further descriptions of the
bioassay methods and results; additional discussion
points (PDF)
AUTHOR INFORMATION
Corresponding Authors
*E-mail: cmcoleman4321@gmail.com.
*Tel: +1 662-915-7144. E-mail: dferreir@olemiss.edu.
■
ORCID
Christina M. Coleman: 0000-0001-5785-7425
Mei Wang: 0000-0003-3686-4292
Daneel Ferreira: 0000-0002-9375-7920
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This project was funded in part by Ocean Spray Cranberries,
Inc., Lakeville-Middleboro, MA, USA, for the isolation and
structure elucidation of cranberry oligosaccharides and
antiadhesion testing. Dr. C. Khoo, at Ocean Spray Cranberries,
Inc., provided administrative assistance and coordinated initial
contact between the University of Mississippi and Washington
State University for antiadhesion bioassay testing. Various
resources within the National Center for Natural Products
Research (NCNPR) at the University of Mississippi were
made available by Dr. I. Khan, including access to
Samples submitted included SNA, SNA-03, SNA-04, SNA-03
fractions Fr.1, Fr.3, Fr.4, Fr.5, Fr.6, Fr.7, Fr.8, Fr.9, and compounds 1
and 2. SNA-03-Fr.2 was not assayed due to sample loss during
handling. Each sample was tested in duplicate along with a media
M
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(20) Turi, C. E.; Finley, J.; Shipley, P. R.; Murch, S. J.; Brown, P. N.
J. Nat. Prod. 2015, 78, 953−966.
antimicrobial testing, performed by Ms. M. Wright, and to
breast cancer cell viability testing, performed by Ms. F. Mahdi.
Funding for antimicrobial testing was provided in part by
USDA Agricultural Research Service, Specific Cooperative
Agreement No. 58-6408-1-603 (antibacterial and antiprotozoal
assays) and in part by the NIH, NIAID, Division of AIDS grant
number AI 27094 (antifungal assay). Analytical and prepara-
tive HPLC-ELSD equipment was made available by Dr. J.
Williamson and Dr. S. J. Cutler in the Department of
Medicinal Chemistry, School of Pharmacy, University of
Mississippi. The Complex Carbohydrate Research Center,
University of Georgia, Athens, GA, provided training for
K.M.A., general information, and support for derivatization
techniques. Dr. J. Parcher, in the Department of Chemistry and
Biochemistry at the University of Mississippi, assisted with
GC-MS analytical procedures.
(21) Sun, J.; Liu, W.; Ma, H.; Marais, J. P. J.; Khoo, C.; Dain, J. A.;
Rowley, D. C.; Seeram, N. P. J. Berry Res. 2016, 6, 149−158.
̈
(22) Ozcan, E.; Sun, J.; Rowley, D. C.; Sela, D. A. Appl. Environ.
Microbiol. 2017, 83, e01097−17.
(23) Blumberg, J. B.; Basu, A.; Krueger, C. G.; Lila, M. A.; Neto, C.
C.; Novotny, J. A.; Reed, J. D.; Rodriguez-Mateos, A.; Toner, C. D.
Adv. Nutr. 2016, 7, 759S−770S.
(24) Oakley, E. T.; Magin, D. F.; Bokelman, G. H.; Ryan, W. S. J.
Carbohydr. Chem. 1985, 4, 53−66.
(25) Reis, A.; Domingues, M. R. M.; Domingues, P.; Ferrer-Correia,
A. J.; Coimbra, M. A. Carbohydr. Res. 2003, 338, 1497−1505.
(26) Fredriksson, S.-A.; Nilsson, B. Adv. Mass Spectrom. 1998, 14,
C025930/1.
(27) Cancilla, M. T.; Penn, S. G.; Carroll, J. A.; Lebrilla, C. B. J. Am.
Chem. Soc. 1996, 118, 6736−6745.
(28) Whiton, R. S.; Lau, P.; Morgan, S. L.; Gilbart, J.; Fox, A. J.
Chromatogr. 1985, 347, 109−20.
(29) Tanaka, T.; Nakashima, T.; Ueda, T.; Tomii, K.; Kouno, I.
Chem. Pharm. Bull. 2007, 55, 899−901.
DEDICATION
■
Dedicated to Dr. Rachel Mata, of the National Autonomous
University of Mexico, Mexico City, Mexico, and to Dr. Barbara
N. Timmermann, of the University of Kansas, for their
pioneering work on bioactive natural products.
(30) Duus, J. O.; Gotfredsen, C. H.; Bock, K. Chem. Rev. 2000, 100,
4589−4614.
(31) Islam, S. M.; Richards, M. R.; Taha, H. A.; Byrns, S. C.; Lowary,
T. L.; Roy, P. N. J. Chem. Theory Comput. 2011, 7, 2989−3000.
(32) Sudjaroen, Y.; Hull, W. E.; Erben, G.; Wuertele, G.;
Changbumrung, S.; Ulrich, C. M.; Owen, R. W. Phytochemistry
2012, 77, 226−37.
REFERENCES
■
(1) Smith, T.; Kawa, K.; Eckl, V.; Stredney, R. Herbalgram 2017,
115, 56−65.
(33) Wyman, C. E.; Decker, S. R.; Himmel, M. E.; Brady, J. W.;
Skopec, C. E.; Viikari, L. In Polysaccharides: Structural Diversity and
Functional Versatility, 2nd ed.; Dumitriu, S., Ed.; CRC Press/Marcel
Dekker: New York, 2005; pp 1023−1062.
(2) Blumberg, J. B.; Camesano, T. A.; Cassidy, A.; Kris-Etherton, P.;
Howell, A. B.; Manach, C.; Ostertag, L. M.; Sies, H.; Skulas-Ray, A.;
Vita, J. A. Adv. Nutr. 2013, 4, 618−632.
(34) Waldron, K. W.; Faulds, C. B. In Comprehensive Glycoscience,
from Chemistry to Systems Biology, Vol. 1; Introduction to Glycoscience;
Synthesis of Carbohydrates; Kamerling, J. P.; Boons, G. J.; Lee, Y. C.;
Suzuki, A.; Taniguchi, N.; Voragen, A. G. J., Eds.; Elsevier: Boston,
2007; pp 181−201.
(3) Pappas, E.; Schaich, K. M. Crit. Rev. Food Sci. Nutr. 2009, 49,
741−781.
(4) Zafriri, D.; Ofek, I.; Adar, R.; Pocino, M.; Sharon, N. Antimicrob.
Agents Chemother. 1989, 33, 92−98.
(5) Ofek, I.; Goldhar, J.; Zafriri, D.; Lis, H.; Adar, R.; Sharon, N. N.
Engl. J. Med. 1991, 324, 1599.
(35) Brummell, D. A.; Dal Cin, V.; Crisosto, C. H.; Labavitch, J. M.
J. Exp. Bot. 2004, 55, 2029−2039.
(6) Ofek, I.; Goldhar, J.; Sharon, N. Adv. Exp. Med. Biol. 1996, 408,
179−183.
(36) Goulao
19, 4−25.
̃
, L. F.; Oliveira, C. M. Trends Food Sci. Technol. 2008,
(7) Burger, O.; Ofek, I.; Tabak, M.; Weiss, E. I.; Sharon, N.;
Neeman, I. FEMS Immunol. Med. Microbiol. 2000, 29, 295−301.
(8) Rafsanjany, N.; Senker, J.; Brandt, S.; Dobrindt, U.; Hensel, A. J.
Agric. Food Chem. 2015, 63, 8804−8818.
(37) York, W. S.; Kumar Kolli, V S.; Orlando, R.; Albersheim, P.;
Darvill, A. G. Carbohydr. Res. 1996, 285, 99−128.
(38) Coleman, C. M.; Akgul, Y.; Kimble, L. L.; Mathison, B. D.;
Chew, B. P.; Ferreira, D. Phytochem. Anal., 2019, in preparation.
(39) Weh, K. M.; Clarke, J.; Kresty, L. A. Antioxidants 2016, 5, 27/
1−27/20.
(40) Vu, K. D.; Carlettini, H.; Bouvet, J.; Cote, J.; Doyon, G.;
Sylvain, J. F.; Lacroix, M. Food Chem. 2012, 132, 959−967.
(41) Cote, J.; Caillet, S.; Doyon, G.; Dussault, D.; Sylvain, J. F.;
Lacroix, M. Food Control 2011, 22, 1413−1418.
(42) Das, Q.; Islam, M. R.; Marcone, M. F.; Warriner, K.; Diarra, M.
S. Food Control 2017, 73, 650−662.
(43) Tuomivaara, S. T.; Yaoi, K.; O’Neill, M. A.; York, W. S.
Carbohydr. Res. 2015, 402, 56−66.
(44) Holck, J.; Hotchkiss, A. T.; Meyer, A. S.; Mikkelsen, J. D.;
Rastall, R. A. In Food Oligosaccharides: Production, Analysis and
Bioactivity; Moreno, F. J.; Luz Sanz, M., Eds.; John Wiley & Sons:
New York, 2014; pp 76−87.
(45) Bode, L. Early Hum. Dev. 2015, 91, 619−622.
(46) Rudloff, S.; Pohlentz, G.; Borsch, C.; Lentze, M. J.; Kunz, C. Br.
Br. J. Nutr. 2012, 107, 957−963.
(47) Umemura, M.; Yuguchi, Y. Cellulose 2009, 16, 361−371.
(48) Umemura, M.; Yuguchi, Y. Carbohydr. Res. 2005, 340, 2520−
2532.
(49) Haard, N. F.; Chism, G. W. In Food Chem.;Fennema, O. R.,
Ed.; Marcel Dekker, Inc.: New York, 1996; pp 943−1012.
(50) Robyt, J. F. Essentials of Carbohydrate Chemistry; Springer-
Verlag: New York, NY, 1998.
(9) Neto, C. C.; Penndorf, K. A.; Feldman, M.; Meron-Sudaim, S.;
Zakay-Rones, Z.; Steinberg, D.; Fridman, M.; Kashman, Y.; Ginsburg,
I.; Ofek, I.; Weiss, E. I. Food Funct. 2017, 8, 1955−1965.
(10) Weiss, E. I.; Houri-Haddad, Y.; Greenbaum, E.; Hochman, N.;
Ofek, I.; Zakay-Rones, Z. Antiviral Res. 2005, 66, 9−12.
(11) Tipton, D. A.; Christian, J.; Blumer, A. Arch. Oral Biol. 2016,
68, 88−96.
(12) Howell, A. B.; Der Marderosian, A.; Foo, L. Y. N. Engl. J. Med.
1998, 339, 1085−1086.
(13) Foo, L. Y.; Lu, Y.; Howell, A. B.; Vorsa, N. J. Nat. Prod. 2000,
63, 1225−1228.
(14) Coleman, C. M. Cranberry Metabolites With Urinary Anti-
Adhesion Activity. Ph.D. Dissertation, University of Mississippi,
University, MS, 2014.
(15) Coleman, C. M.; Auker, K. M.; Killday, K. B.; Azadi, P.; Black,
I.; Ferreira, D. J. Nat. Prod. 2019, 82, DOI: 10.1021/np8b01043.
(16) Hotchkiss, A. T.; Nunez, A.; Strahan, G. D.; Chau, H. K.;
White, A. K.; Marais, J. P. J.; Hom, K.; Vakkalanka, M. S.; Di, R.; Yam,
K. L.; Khoo, C. J. Agric. Food Chem. 2015, 63, 5622−5633.
(17) Sun, J.; Marais, J. P. J.; Khoo, C.; LaPlante, K.; Vejborg, R. M.;
Givskov, M.; Tolker-Nielsen, T.; Seeram, N. P.; Rowley, D. C. J.
Funct. Foods 2015, 17, 235−242.
(18) Holmes, A. B.; Rha, C. J. Food Sci. 1978, 43, 112−115.
(19) Marlett, J. A.; Vollendorf, N. W. Food Chem. 1994, 51, 39−44.
N
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(51) Wang, J.; Chen, C.; Yu, Z.; He, Y.; Yong, Q.; Newburg, D. S.
Eur. Food Res. Technol. 2017, 243, 133−146.
(52) Katouli, M. Iran J. Microbiol. 2010, 2, 59−72.
(53) Rossez, Y.; Holmes, A.; Lodberg-Pedersen, H.; Birse, L.;
Marshall, J.; Willats, W. G. T.; Toth, I. K.; Holden, N. J. J. Biol. Chem.
2014, 289, 34349−34365.
(54) Ciucanu, I.; Kerek, F. Carbohydr. Res. 1984, 131, 209−217.
(55) Whiton, R. S.; Lau, P.; Morgan, S. L.; Gilbart, J.; Fox, A. J.
Chromatogr. 1985, 347, 109−120.
̈
(56) Jansson, P. E.; Kenne, L.; Liedgren, H.; Lindberg, B.; Lonngren,
J. Chem. Commun. (Stockholm Univ.) 1976, 8, 1−75.
(57) Zhang, J. R.; Aziz, A.; Jain, S.; Jacob, M. R.; Khan, S. I.;
Tekwani, B. L.; Ilias, M. Res. Rep. Med. Chem. 2012, 2, 1−6.
(58) Rahman, A. A.; Samoylenko, V.; Jacob, M. R.; Sahu, R.; Jain, S.
K.; Khan, S. I.; Tekwani, B. L.; Muhammad, I. Planta Med. 2011, 77,
1639−1643.
(59) NCCLS. Reference Method for Broth Dilution Antifungal
Susceptibility Testing of Yeasts; Approved Standard M27-A2; National
Committee on Clinical Laboratory Standards, 2002.
(60) NCCLS. Methods for Dilution of Antimicrobial Susceptibility
Tests for Bacteria that Grow Aerobically, Approved Standard, Seventh
ed.; National Committee on Clinical Laboratory Standards, 2006.
(61) NCCLS. Susceptibility Testing of Mycobacteria, Nocardia, and
Other Aerobic Actinomycetes; Tentative Standard-Approved Standard,
M24-A; National Committee on Clinical Laboratory Standards, 2003.
(62) NCCLS. Reference Method for Broth Dilution Antifungal
Susceptibility Testing of Filamentous Fungi; Approved Standard, M38-
A; National Committee on Clinical Laboratory Standards, 2002.
(63) Mathison, B. D.; Kimble, L. L.; Kaspar, K. L.; Khoo, C.; Chew,
B. P. J. Nat. Prod. 2013, 76, 1605−1611.
(64) Evans, D. G.; Evans, D. J.; Tjoa, W. Infect. Immun. 1977, 18,
330−337.
O
DOI: 10.1021/acs.jnatprod.8b01044
J. Nat. Prod. XXXX, XXX, XXX−XXX