Characterization of molecular structural changes in pectin during juice

cloud destabilization in frozen concentrated orange juice

Ashley L. Galant, Wilbur W. Widmer, Gary A. Luzio, Randall G. Cameron
*
Citrus and Other Subtropical Products Research Unit, US Horticultural Research Laboratory, USDA-ARS, 2001 South Rock Road, Ft. Pierce, FL 34945, USA
a r t i c l e i n f o
Article history:
Received 25 September 2013
Accepted 12 March 2014
Available online 20 March 2014
Keywords:
Pectin
Juice cloud
Orange juice
Pectin methylesterase
Cloud loss
Citrus sinensis
a b s t r a c t
Pectin comprises one of the major components of cloud material in citrus juices. Juice cloud is a complex
mixture of polysaccharides, proteins and lower molecular weight compounds that are responsible for the
turbid appearance of citrus juices. The stability of juice cloud depends on a number of factors, including
pectin degree of methylation (DM) and the availability of sufciently-sized, charged demethylated
blocks, but detailed information on the precise relationship between cloud state and pectin architecture
is limited. To address this gap, we have systematically treated commercial frozen concentrated orange
juice (FCOJ) with pectin methylesterases extracted from orange pulp cells to mimic the aggregation of
pectin that naturally occurs in the presence of calcium in unpasteurized or under pasteurized juice. We
then assessed the pectin structural and functional properties, including juice optical density (OD) and
soluble sugar composition, DM, degree of blockiness (DB), degree of absolute blockiness (DB
abs
) and
rheological capacity from juice in various states of cloud loss. A strong positive correlation ([r] 0.916)
was observed between juice OD and DM, while a strong negative correlation ([r] 0.914) was observed
between OD and DB
abs
. Most critically, a reduction of average DM from 74.7% to 68.3% in the course of
cloud loss resulted in the formation of rheologically active pectin.
Published by Elsevier Ltd.
1. Introduction
In orange juice, cloud is responsible for imparting desirable
texture, color, and avor, as well as consumer-expected turbidity.
Cloud is a component of juice distinct from pulp; while pulp tends
to settle in juice, cloud forms a suspension of particles ranging from
0.4 to 5 mm in size (Klavons, Bennett, & Vannier, 1994). Smaller
particles, typically those below2 mm, formmore stable suspensions
than large particles, and if particle sizes become too large, cloud
precipitation, known generally in the literature as cloud loss, can
occur (Corredig, Kerr, & Wicker, 2001).
Orange juice cloud is comprised primarily of pectin, protein,
cellulose, hemicellulose, some lipids, and small quantities of
entangled bioavonoids. Of these components pectin and protein
are the most abundant, accounting for approximately one third and
one half of the insoluble material respectively (Baker & Bruemmer,
1969; Braddock, 1999; Gattuso, Barreca, Gargiulli, Leuzzi, & Caristi,
2007; Scott, Kew, & Veldhuis, 1965). Citrus pectin is a complex
molecule: it contains a homogalacturonan (HG) backbone of a-(1-
4)-linked galacturonic acid (GalA) residues e some portion of
which may be methylesteried e interspersed with regions con-
taining repeating rhamnogalacturonan dimers (RGI) (Coenen, Bakx,
Verhoef, Schols, & Voragen, 2007; Yapo, Lerouge, Thibault, & Ralet,
2007). The rhamnose moieties of RGI regions may be decorated
with variably-sized arabinans or galactans (Doco, Williams, Vidal, &
Pellerin, 1997). The degree to which pectin HG is methylated (de-
gree of methylation, DM), as well as the pattern in which methyl-
ation occurs, plays a dominant role in determining the functional
behavior of the molecule (Cameron, Luzio, Goodner, & Williams,
2008; Tanhatan-Nasseri, Crepeau, Thibault, & Ralet, 2011). Blocks
of six or more contiguous demethylated GalA residues are required
for formation of ionic sandwiches of pectin molecules and cal-
cium ions (Liners, Thibault, & Cutsem, 1992; Luzio & Cameron,
2008); commonly referred to as the egg-box model (Powell,
Morris, Gidley, & Rees, 1982). These demethylated blocks are
formed through the activity of pectin methylesterases (PMEs; EC
3.1.1.11), four of which can be found naturally in citrus juice
(Cameron, Baker, & Grohmann, 1998). When pectin networks form,
the suspended material present in the citrus juice separates from
the liquid phase, leaving behind a yellowish, unappetizing serum
and the resulting precipitate (Ackerley, Corredig, & Wicker, 2002;
Rouse, 1949). This precipitation represents the end-state of juice
* Corresponding author. Tel.: 1 772 462 5856; fax: 1 772 462 5986.
E-mail address: Randall.Cameron@ars.usda.gov (R.G. Cameron).
Contents lists available at ScienceDirect
Food Hydrocolloids
j ournal homepage: www. el sevi er. com/ l ocat e/ f oodhyd
http://dx.doi.org/10.1016/j.foodhyd.2014.03.013
0268-005X/Published by Elsevier Ltd.
Food Hydrocolloids 41 (2014) 10e18
clarication however the initial stages of cloud loss are not readily
visible to the unaided eye. For instance, Ellerbee and Wicker (2011)
and Croak and Corredig (2006) reported a shift in cloud particle size
in juice containing native PME activity from 2e50 mm to 2e100 mm
without a corresponding change in % transmittance. Likewise,
Croak and Corredig (2006) reported that the apparent diameter of
juice cloud particles began to increase almost immediately
following addition of PME, but remained below the 2 mm stability
threshold for all treatments during the 30 min observation period.
The longer the factors responsible for cloud loss are allowed to
persist in juice, the more readily apparent the effects become.
Cameron et al. (1998) observed an increase in settling pulp corre-
sponding to a reduction in juice cloud absorbance within three days
following addition of citrus PMEs isozymes to FCOJ samples; within
ten days the juice was completely claried. Krop (1974) likewise
observed clarication of reconstituted juice treated with PME
within two days, and the corresponding release of approximately
80% of the saponiable GalA methyl groups after a total of 12 days.
From the above referenced studies on the phenomenon of juice
cloud loss it is obvious that calciumis naturally present in sufcient
concentration in commercial orange juices to allow for the cross-
linking of sufciently sensitized pectin, though the minimal
structural modications and their relationship to calcium concen-
tration have not been determined. The United States Department of
Agriculture reports a value of 11 mg calcium per 100 g chilled or-
ange juice (includes from concentrate) and 140 mg per 100 g for
calcium fortied orange juice (U.S. Department of Agriculture,
2013).
From an industry perspective, issues of juice cloud and cloud
loss present unique challenges and difculties. Because cloud loss
eliminates juice marketability and thus has a negative economic
impact on juice producers and associated downstreamindustries, it
is important that the critical structural details of cloud and its
incorporated pectin be fully assessed.
From the antecedently illustrated studies and many others a
macroscopic image of juice cloud and cloud stability has been
documented; comparatively less is known however about the
nanostructural changes pectin undergoes during cloud loss. Previ-
ously, we characterized the composition, macromolecular proper-
ties, and architecture of pectin from Citrus sinensis FCOJ (Galant,
Luzio, Widmer, & Cameron, 2014). We now report on the changes
to pectin molecular structure, including DM, DB, and DB
abs
for
pectin isolated from the same FCOJ at various stages of cloud loss.
2. Materials and methods
2.1. Materials
Frozen concentrated orange juice was provided by Nestl Pro-
fessional Vitality (Tampa, FL). C. sinensis cultivar Pineapple pulp
cells were a generous gift from Don Gillette and Louis-Dreyfus
Citrus Inc. Polygalacturonases (EPG-M1, EPG-M2) were purchased
from Megazyme (Wicklow, Ireland). Pectinase (P2166) and pro-
teinase K were purchased fromSigma Aldrich (St. Louis, MO), while
Rapidase ADEX-P was purchased from DSM Food Specialties (Delft,
The Netherlands). Celluclast was purchased from Novozymes
(Hellerup, Denmark).
2.2. Methods
2.2.1. PME purication
Four liters of orange pulp cells were diluted into two volumes of
Buffer A (20 mMTris, pH8.0; 50 mMNaCl) and mixed for 1 h at 4

C
(Savary et al., 2010). The resulting slurry was ltered through four
layers of miracloth, and the solids were centrifuged at 12,000 g,
4

C for 30 m. The supernatant from both the ltration and
centrifugation were discarded after conrmation that it did not
contain PME activity (see below). The wash, ltration, and centri-
fugation were repeated a second time, and again the solids were
retained. Next, the pulp cell solids were diluted into 2 volumes of
Buffer B (20 mM Tris, pH 8.0; 0.5 M NaCl), and stirred overnight at
4

C to extract the PME. The slurry was ltered through 4 layers of
miracloth and then centrifuged at 12,000 g, 4

C for 30 m. The
solids were discarded once it was conrmed that most of the PME
activity had migrated to the supernatant fraction. An AkroPak 500
(Pall, Port Washington, NY) in line with a Tangential FlowFiltration
system (Pall, Port Washington, NY) containing a 30 kDa MWCO
cartridge was pre-equilibrated with Buffer B, and used to concen-
trate the supernatant to volume (approximately 2 L in this case).
The permeate was discarded once the absence of PME activity was
conrmed. The retentate was then diluted into 2.5 volumes of
Buffer C (20 mM Tris, pH 7.5; 0.2 M NaCl). Diethylaminoethanol
(DEAE, 320 g) was equilibrated with Buffer C, decanted, and added
to the retentate; the resulting slurry was allowed to stir overnight
at 4

C. The next morning, the slurry was decanted into a sintered
glass funnel, and the ow-through was collected using gravity
ltration. The DEAE was washed twice with Buffer C, and the ow-
through was pooled. Tangential owltration, after pre-incubation
with Buffer C, was used to concentrate the ow-through down to
1.25 L, and the PME activity in the retentate was determined
quantitatively by titrating 25 mL of retentate solution against 0.02 M
LiOH in the presence of 2% 94% DM pectin. PME activity or the
absence thereof was veried qualitatively throughout the puri-
cation by adding 30 mL of the fraction to be tested to 100 mL of in-
dicator (0.05% Bromothymol Blue, pH 8.0; 0.2 M NaCl; 0.1% 94% DM
pectin) and observing the color change after 1 min.
2.2.2. FCOJ demethylation
To the juice concentrate, three volumes of water, as well as
lithiumazide (LiN
3
) to 0.02% (w/v) and potassiummeta-bisulfate to
0.43% (w/v) were rst added. The solution was brought to 30

C in a
stirring water-jacketed vessel (Wilmad-LabGlass, Vineland, NJ) and
3104 Units of C. sinensis Pineapple PME activity was added per
0.5 L of pre-dilution juice concentrate. 10 mL aliquots of juice were
collected periodically (every 2 h initially and then every h until an
OD in the targeted range was reached) and the degree of cloud loss
was determined by centrifuging the solution for 10 min at 360 g
in a centrifuge (International Equipment Company, Universal
Model) with a swinging bucket rotor, and measuring the OD of the
resulting supernatant at 660 nm on a UV-2401 PC spectropho-
tometer (Shimadzu) using distilled water as a blank. Once the
desired OD was reached, the pH of the juice was adjusted to 1.8
with nitric acid, and the temperature was adjusted to 70

C for a
period of 3 h. At the conclusion of the incubation, the pHof the juice
was adjusted to 2.2 with potassiumhydroxide, and the solutionwas
drained into 3 volumes of chilled isopropyl alcohol (IPA).
A second, smaller aliquot of FCOJ was prepared in the same
manner for simultaneous demethylation and rheological analysis.
2.2.3. Pectin extraction
The procedure utilized for pectin extraction is outlined in Fig. 1.
After at least 16 h at 4

C, the juice in IPAwas centrifuged for 30 min

at 12,000 g, 4

C. The resulting supernatant was discarded, while
the pellets were resuspended in two volumes of water. Proteinase
K, 0.3 mg/mL, was added to the suspension, and it was allowed to
stir overnight at 37

C. After the incubation, the suspension was
centrifuged for 30 min at 12,000 g, 20

C, and the supernatant
(soluble fraction) and pellet (insoluble fraction) were separated.
The supernatant was dialyzed (Spectra/Por MWCO: 6000e
8000 Da) against two volumes of distilled water containing 0.02%
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 11
LiN
3
for 24 h with four solution changes throughout. At the
conclusion of dialysis, three volumes of chilled IPA were added to
the supernatant (soluble fraction). After 16 h at 4

C, the solution
was again centrifuged for 30 min at 12,000 g, 4

C and the su-
pernatant (soluble fraction) was discarded. The resulting pellet
(soluble fraction) was briey allowed to air dry, lyophilized over-
night with a Labconco Freezone Lyophilizer (Kansas City, MO), and
homogenized with a coffee grinder. The pellet (insoluble fraction)
was resuspended in water adjusted to pH 8.0 with potassium hy-
droxide, and allowed to stir overnight at 4

C. Following resus-
pension, the suspension was centrifuged for 30 min at 12,000 g,
20

C, and the pellet was discarded. The supernatant was dialyzed
(Spectra/Por MWCO: 6000e8000 Da) against two volumes of
distilled water containing 0.02% LiN
3
for 24 h with four solution
changes throughout. As above, at the conclusion of dialysis, three
volumes of chilled IPA were added to the supernatant, and after
16 h at 4

C, the solution was again centrifuged for 30 min at
12,000 g, 4

C. The supernatant was discarded, while the pellet
(insoluble fraction) was allowed to air dry, lyophilized, and ho-
mogenized with a coffee grinder.
2.2.4. Soluble sugar content
FCOJ pectins were prepared at a 1% concentration in 10 mM
lithium acetate, pH 4.5 containing 0.02% LiN
3
and digested for 16 h
at 45

C with 0.36 mL Celluclast, 1 mL Rapidase ADEX-P, and 1 mL
Pectinex per mg of pectin. The reactions were terminated by raising
the temperature to 80

C for 10 m. Analysis for soluble sugars was
carried out using HPAECePADas previously described (Galant et al.,
2014; Widmer, 2011).
2.2.5. Degree of methylation
DM was determined by titration according to a method modi-
ed from that found in the United States Pharmacopeia (2000). In
order to convert it to the acidic form, the pectin was rst stirred
(20 mL/1 g of pectin) in 37% hydrochloric acid:60% ethanol (1:20)
for 30 m, ltered and then further washed with successive portions
of 37% hydrochloric acid:60% ethanol (1:20), 60% ethanol, and 95%
ethanol. The ltrate was dried in an oven for 1 h at 65

C, resus-
pended to 0.5% in 1% sodium dodecyl sulfate and the solution
degassed for 15 min to remove dissolved carbon dioxide. The so-
lution was then titrated against sodium hydroxide of known
molarity using bromothymol blue as an indicator, and saponied
for 15 min at room temperature with an excess of base. After the
excess base was neutralized, the solutionwas titrated a second time
and the DM was calculated as in Equation (1).
DM

moles methyl esterified GalA

moles total GalA

100 (1)
2.2.6. Degree of blockiness
FCOJ pectins were prepared at a concentration of 1% in 10 mM
lithium acetate, pH 5.5 and digested for 16 h at 30

C with 2.1 Units
of EPG-M1 and 5 Units of EPG-M2 (both fromMegazyme, Wicklow,
Ireland) per mg of pectin. The reactions were terminated by raising
the temperature to 80

C for 10 min and the denatured enzymes
were separated from the pectin by centrifugation at 12,000 g for
2 m. The pectin solutions were diluted to 0.1% and analyzed by
HPAEC using a Dionex CarboPac PA-1 anion exchange column
(4 250 mm) coupled in line behind a CarboPac PA-1 guard col-
umn (4 50 mm) as previously described (Galant et al., 2014;
Guillotin et al., 2005). A linear gradient of ammonium formate
was used to produce calibration curves for 0.0126% monomeric,
0.0119% dimeric, 0.0109% trimeric GalA, and for sample analysis.
Eluents were detected with a Sedex 75 Evaporative Light Scattering
Detector (ELSD; Sedere, Alfortville, FR) connected to house air (0%
humidity, ltered to 0.01 mm) and pressurized to 4.5 bar; Gain was
set to 11 and drift tube temperature was set to 97

C. Analyte peaks
were observed and integrated using EZChrom Elite (Agilent Tech-
nologies, Santa Clara, CA). Three replicate injections were per-
formed, and the results averaged. DB and DB
abs
were calculated
based on the averages.
2.2.7. Rheology
The rheological properties of the pectic extracts and PME
treated reconstituted FCOJ were determined using an AR 2000
Rheometer (TA Analytical, Wilmington, DE) equipped with a
60 mm acrylic cone geometry with a 2

02
0
angle of deviation be-
tween the cone and the plate. FCOJ pectins were prepared at 0.2%
concentration in distilled water and adjusted to pH 7.0 with
ammonium hydroxide. Reconstituted FCOJ was also adjusted to pH
7.0 prior to the addition of PME. Pectin concentration in recon-
stituted juice ranges from 0.04% to 0.2% (Braddock, 1999).
FCOJ Extracon IPA Precipitaon Supernatant:
Discarded
Resuspend Pk digeson
Pellet Supernatant
IPA Precipitaon
Centrifugaon
Lyophilizaon
Soluble Fracon
Resuspend:
pH 8.0
Centrifugaon Pellet:
Discarded
IPA Precipitaon
Centrifugaon Lyophilizaon
Insoluble Fracon
Centrifugaon
Pellet Centrifugaon
Supernatant Dialysis
Dialysis
Fig. 1. Overview of the procedure utilized for pectin extraction.
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 12
Ca(H
2
PO
4
)
2
was then added to 0.1% (w/v) and allowed to dissolve
completely. The solutionwas then adjusted to pH5.5 with glucono-
d-lactone, which gradually hydrolyzes to gluconic acid and prompts
the uniform release of calcium ions. After pH adjustment, the gap
between the plate and the geometry was set to 56 microns, and
4 mL of sample was injected under the cone. A volume 500 mL of
water was added to the reservoir on top of the geometry, serving as
a hydration reservoir so as to maintain constant vapor pressure
throughout the experiment as per the manufacturers recommen-
dation. A manufacturer designed cover was placed over the ge-
ometry to prevent evaporation, and gel formation was monitored
over the course of 20 h at a frequency of 1.257 rad/s and at 20

C.
Frequency sweeps were preformed between 0.1 and 100 rad/s at
20

C, 1% strain to determine the storage modulus (G
0
) and loss
modulus (G
00
).
2.2.8. Oligomer degree of polymerization
FCOJ pectins were prepared at a concentration of 0.25% in
50 mM lithiumacetate, pH 5.5 and digested with 0.01 Units of EPG-
M2 per mg of pectin for 5e15 min at 30

C. To terminate the re-
action, 1 mL of solution was pipette into a beaker containing 7.5 mL
of concentrated HCl, microwaved for 4 s on high power (Kenmore
Model 721.62463200, 1100 W maximum, 2450 MHz), and then
heated to 100

C for 10 m. After a sample cooled to room temper-
ature, its pH was adjusted back to pH 5.5 with ammonium hy-
droxide and LiN
3
was added to 0.02%. The digested sample (200 mL)
was injected onto the CarboPac PA-1 anion exchange column, and
oligomers were separated using a 0e1 M ammonium formate
gradient (0e60 min, ammonium formate concentration from 0 to
60%; 60e180 min, 60 to 80%; 180e195 min, 80 to 100%) in deion-
ized water at a ow rate of 1 mL/min. Injections were performed in
triplicate and the results averaged. Masses for each oligomer were
calculated using a pooled calibration curve of GalA trimer, and GalA
octomer samples. Oligomer mass estimates were converted to
molar concentration as previously described (Cameron et al., 2008)
and then scaled to account for purity differences between the
extracted FCOJ pectins.
2.2.9. Molecular weight
FCOJ pectins were prepared at a 1% concentration in 1% SDS
containing 0.02% LiN
3
. Molecular weight determinations were
carried out as previously described (Galant et al., 2014; Luzio &
Cameron, 2010) with a few changes. First, the injection volume
was set to 100 mL instead of 50 mL. Additionally, instead of 1 mM
ammonium formate, 1 mM ammonium laurel sulfate (ALS) was
employed as the mobile phase.
3. Results and discussion
In this work, we characterized the structural prole and
compositional details of pectin extracted from FCOJ in progressive
states of cloud loss. Since the Units of PME activity required for
sufciently clarifying the juice samples was greater than that which
could be economically obtained from commercial sources, bulk
PME (comprising four PME isozymes) (Cameron et al., 1998) was
rst puried fromC. sinensis cultivar Pineapple pulp cells. From4 L
of pulp cells, we obtained 1.25 L of bulk extracted PME containing
243,000 Units of activity, or 0.194 mEquivalents/mL/min. This PME
extract was added to FCOJ that was diluted 1:3 in distilled water,
and its effects were monitored at 660 nm as described. Once an OD
within a desired range was reached, the pectin was extracted via a
procedure that is commonly employed commercially (70

C for 3 h
at low pH, in this case pH 1.8) and puried using the methodology
outlined in Fig. 1 (Joye & Luzio, 2000; May, 1990). Using this tech-
nique, we procured ve samples from three reactions of PME with
FCOJ which were terminated at OD 2.202, OD 1.595, and OD 0.109.
We also previously assessed uncompromised juice cloud which had
an OD of 3.220 at a 1:3 dilution, and have included some of those
results for comparison (Galant et al., 2014).
As previously stated the initial stages or lowest degrees of cloud
loss are not readily visible to the unaided eye. Ellerbee and Wicker
(2011) reported a shift in cloud particle size of juice containing
native PME activity from 2e50 mm to 2e100 mm without a corre-
sponding change in % transmittance. Likewise, Croak and Corredig
(2006) reported that the apparent diameter of juice cloud particles
began to increase almost immediately following addition of PME,
but remained below the 2 mm stability threshold for all treatments
during the 30 min observation period. Visually, the cloud from the
OD 2.202 sample that remained in suspension after centrifugation
was indistinguishable from that found in untreated juice; the OD
1.595 cloud was similar, but appeared noticeably thinner when the
two samples were compared side-by-side. The cloud from the OD
0.109 sample, on the other hand, consisted of a clear, yellow-hued
serum, the likes of which can be seen in the graphical abstract
accompanying this manuscript.
At some ODs, specically 0.109 and 3.220, the pectin in the juice
cloud was isolated from a single fraction. For the OD 0.109 sample,
this fraction was termed the insoluble fraction because it origi-
nated from the pellet portion of the extract following the second
round of centrifugation. In contrast, the pectin isolated from the
cloud at ODs 2.202 and 1.595 comprised both a soluble fraction
from the supernatant portion of the extract and an insoluble frac-
tion. Fromthe previously isolated OD3.220 sample, pectinwas only
isolated from a soluble fraction. Henceforth, in the text and tables,
samples originating from the insoluble or soluble fraction are
delineated with an i or s respectively following their OD. The
presence of pectic fractions of differing solubility was rst noted in
reconstituted commercial lemon juice concentrate, though the
differences between the fractions were not further investigated
(Klavons & Bennet, 1987). The yield of extracted solids from each
fraction is outlined in Table 1. The highest material yields are
associated with the samples originating from OD 2.202 and 0.109
clouds, the highest and lowest ODs respectively. Although extrac-
ted at a 1:1 dilution instead of 1:3, OD 3.220 cloud also yielded a
comparatively large quantity of material. Cloud particles are least
aggregated in the absence of PME or shortly after the beginning of
Table 1
Yield of extracted solids from starting juice concentrate.
Sample name Starting juice concentrate volume (L) Extracted solids (g) Solids concentration (mg/L) at single strength (11.8

Bx)
0.109i 0.5 2.00 715.6
1.595i 1.0 1.98 354.2
1.595s 1.0 0.43 76.9
2.202i 1.0 0.92 164.6
2.202s 1.0 7.04 1259.4
3.220s
a
2.5 20.7 1480.4
a
From Galant et al. (2014).
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 13
treatment (Croak & Corredig, 2006), though aggregation effects are
lessened by the presence of a sufcient quantity of free carboxyl
groups (Klavons, Bennett, & Vannier, 1992); it is unclear whether
these factors are causative or merely correlate with the yield of
pectin from juice cloud at different ODs.
3.1. Optical density (OD) and degree of methylation (DM)
Once extracted, citrus juice cloud can be very difcult to resol-
ubilize, even when placed under conditions similar to those found
in juice serum. For orange juice cloud in particular, Klavons noted
that 53e77% of the cloud material reprecipitated when returned to
native-like conditions (Klavons, Bennet, & Vannier, 1991). Likewise,
Baker and Bruemmer (1969) noted that while cloud could be
resuspended in water alone, the presence of salts, pectin, and sugar
led to serumclarication. To avoid cloud reprecipitation and ensure
full solubilization, we opted to resolubilize our extracts in 1% SDS
instead of a more serum-like liquid. In order to determine the
molecular weight of each resolubilized pectic extract, SEC-MALLS-
RI was performed using ALS as the mobile phase. We previously
reported the molecular weight of undigested pectin extracted from
FCOJ to be 1.5 10
6
, which suggested aggregation was occurring.
The molecular weight of pectin isolated from each of the ve FCOJ
cloud samples was consistent with this value (data not shown),
indicating that no widespread degradation of the HG backbone it-
self occurred during OD reduction.
For each pectic extract, titration and back-titration with base
yielded a measure of the quantity of methylesteried GalA and total
quantity of GalA present, the ratio of which was taken to determine
the DM of each sample. A positive correlation was observed be-
tween FCOJ OD after centrifugation and pectin DM, with the DM
dropping from an average of 74.7% in unadulterated cloud, to an
average of 19.6% in OD 0.109i cloud (Table 2). The DMs of the OD
2.202 and 1.595 samples fell between these two bounds, with
distinct DMs for insoluble and soluble fractions in each pair. In both
cases, the insoluble fractionpossessed the lower DMof the two. The
insoluble fraction was isolated by resuspension at pH8.0, instead of
at the pH 3.5e4.5 range typical of juice (Swift & Veldhuis, 1957);
accordingly it is likely that this elevated pH favored the solubili-
zation of the more demethylated insoluble fractions due to their
greater overall negative charge relative to the soluble fractions. At
OD 0.109, pectin was isolated only from the insoluble fraction; this
is in line with the lower DM, and hence greater overall negative
charge of this pectin relative to those isolated at higher ODs.
As described above, at OD 1.595 the juice cloud visually retained
much of the appearance of uncompromised juice. As the OD dipped
belowOD1.000 however, the cloud began to lose its opacity, and by
OD 0.109 only a clear yellow serum remained suspended post-
centrifugation. At OD 0.109, the DM of the juice pectin was deter-
mined to be 19.6%; well below the 50% threshold for a low DM
pectin. This DM is in line with those obtained by Rouse (1949) and
Baker (1979), who reported juice clarication when an average DM
of 38% or 14% respectively was reached. In the authors experience
it is not possible to reduce the serum OD below 0.09, although
demethylation may continue beyond this point pending the avail-
ability of PME binding sites (Fries, Ihrig, Brocklehurst, Shevchik, &
Pickersgill, 2007; Mercadente, Melton, Jameson, Williams, & De
Simone, 2013).
While the 0.109i sample represents the end stages of cloud loss
in juice, the 2.202i and 2.202s samples represent the beginning.
Although not visually apparent, cloud loss was already well un-
derway at this point, with the overall OD having been reduced
nearly a third. At OD 2.202, the juice cloud was able to be frac-
tionated for the rst time into soluble and insoluble samples, with
pectin DMs of 68.3% and 47.4% respectively. While the soluble
fraction represents a comparatively small DMdrop of 6.4% fromthe
uncompromised juice, the DM of the insoluble fraction fell by more
than 25%. Fortunately, based on the solids extracted from each
sample (Table 1), this low DM fraction represents only a minor
subset of the total pectin present in the overall OD 2.202 sample.
From the perspective of a juice producer, the presence of low DM
pectin alone is not so much a concern as whether that low DM
pectin, or indeed any pectin, is demethylated such that blocks large
enough for the formation of calcium crosslinks are present.
3.2. Degree of blockiness (DB), absolute degree of blockiness
(DB
abs
), and oligomer degree of polymerization (DP)
As described in the introduction, calcium crosslinking of pectin
molecules depends upon the availability of sufciently-sized
stretches of demethylated GalA, which is negatively charged,
along the length of the molecular backbone. One method by which
the blockiness of a given pectin population may be calculated is
through determination of degree of blockiness (DB) and absolute
degree of blockiness (DB
abs
). DB describes the presence of deme-
thylesteried blocks as a percentage of the total non-
methylesteried GalA, while DB
abs
describes the presence of
demethylesteried blocks as a percentage of the total GalA present
(Daas, Voragen, & Schols, 2000; Guillotin et al., 2005). Both metrics
rely on quantication of the GalA blocks cleaved frompectin via the
activity of EPG, which requires a stretch of four contiguous deme-
thylesteried GalA residues in order to cleave the pectic backbone
(Chen & Mort, 1996). Larger demethylated stretches will result in
more backbone breaks, and different EPG enzymes have different
specicities for cleaving the smaller fragments that may result.
Although DB and DB
abs
somewhat overestimate the blockiness of a
pectin molecule (a block of four or more required for EPG versus a
block of 6 or more demethylated GalA residues required for
crosslinking), the metrics are nonetheless a good indicator of
crosslinkability.
With our pectic extracts, we observed an inverse correlation of
blockiness to DM, with high DM pectins having a low DB and DB
abs
and vice versa (Table 3). Because DB
abs
is dependent upon the total
Table 2
Degree of methylation for pectins from intact and cloud-compromised juice
samples.
Sample name OD
660nm
DM (%)
0.109i 0.11 19.6 1.4
1.595i 1.6 39.9 1.5
1.595s 1.6 54.0 2.6
2.202i 2.2 47.4 0.8
2.202s 2.2 68.3 2.0
3.220s
a
3.220 74.7 0.5
a
From Galant et al. (2014).
Table 3
Degree of blockiness, degree of absolute blockiness and rheological properties of
pectic extracts.
Sample name DB DBabs G
0
G
00
0.109i 52.78 1.24 42.46 1.00 16.63
a
1.28
1.595i 55.02 3.16 33.10 1.90 18.03 2.07
1.595s 52.35 2.64 24.06 1.21 6.28 0.47
2.202i 48.67 0.87 25.63 0.46 0.46 0.19
2.202s 36.03 2.29 11.42 0.73 0.46 0.09
3.202s 15.92 0.42 4.03 0.11 DNG
b
DNG
a
G
0
and G
00
values were recorded at 3.1 rad/s during a frequency sweep of each
extract (Fig. S1).
b
Did not gel.
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 14
quantity of GalA present e a number which generally remains
constant e and the quantity of demethylated GalA present, it
typically displays a very strong correlation to DM. DB, on the other
hand, guarantees no such correlation. In this case, we observed that
the DB of the samples increased with decreasing DM initially, but
then plateaued once the DM of the cloud pectin reached approxi-
mately 50%. Since increasing quantities of demethylated GalA were
available as the DM of the cloud pectin decreased, and since
increasing quantities of monomeric, dimeric, and trimeric GalA
were detected as the DMdecreased, this suggests that EPG cleavage
sites were becoming available at a somewhat slower rate than that
at which GalA residues were being demethylated. Because EPG, in
addition to its minimum block-size requirements, is also restricted
to cleavage sites anked on each side by at least one non-esteried
GalA residue, PME activity may create demethylated blocks on the
pectic backbone which are difcult or impossible for EPG to access
(Limberg et al., 2000a, 2000b). Such a mode is consistent with the
activity of salt-independent citrus PME, which is the most abun-
dant PME present in citrus extracts, and which operates via a multi-
attack mechanismwith a lowdegree of processivity at pH4.5 when
starting-site availability is limited (Cameron et al., 2008). In this
instance, the formation of juice cloud aggregates, as evidenced
visually and by the large pectin MWvalues observed, may preclude
PME acting with a higher degree of processivity in some instances.
Oranges also contain three other pectin methylesterases
(Cameron et al., 2008; Cameron et al., 2003; Savary et al., 2010); one
has not yet been characterized, while the other has biochemical
properties similar to those recently reported for papaya PME (Kim
et al., 2013; Savary & Vasu, 2012). The nal pectin methylesterase,
thermally-tolerant PME, accounts for 5e10% of the total PME pre-
sent in peel and also operates via a multi-attack mechanism,
though with a degree of processivity equal to w10 at pH 4.5
(Cameron, Luzio, Vasu, Savary, & Williams, 2011). To assess the
presence of longer demethylated blocks in the extracted pectin, we
performed a time-limited digest of each extract using EPG-M2 and
separated the resulting oligomers via HPAECeELSD (Fig. 2). The
resultant oligomer degree of polymerization (DP) generally corre-
lated with OD and DM, with the 1.595i and 0.109i samples pos-
sessing the largest DP values of 52 and 62 respectively. The largest
oligomer detectable in the 2.202s sample had a DP of 5, while the
largest oligomer detectible in the 3.220s sample was trimer (data
not shown). Larger blocks still are likely present in each sample,
albeit at concentrations too low and for periods too brief to be
detected with the available methodology. Because calcium cross-
links rely on the presence of demethylated stretches of pectin
backbone in order to form, the presence of larger blocks correlates
with the increased occurrence of juice cloud collapse (as measured
by OD) and the likelihood of pectin network formation.
3.3. Rheology and correlation analysis
In order to further quantify susceptibility to calcium cross-
linking, the rheological properties, in particular G
0
and G
00
, of each
extract were also assessed. G
0
, the storage modulus, measures the
energy stored in a viscoelastic solid, and correlates with the
crosslink density (or blockiness in the case of pectin) of the
comprising molecules. G
00
, the loss modulus, represents the energy
dissipated as heat and the liquid component of the gel matrix.
Finally the ratio of G
00
/G
0
, also known as tan d, indicates the relative
uidity of the matrix; tan d < 1 suggests the presence of well-
ordered molecules, while tan d > 1 indicates that the particles are
less well ordered and that the matrix may appear liquid in form
(Audebrand, Kolb, & Axelos, 2006; Espinosa-Muoz, Renard,
Symoneaux, Biau, & Cuvelier, 2013). As expected, the undigested
parent pectin fromthe OD 3.220 juice did not gel in the presence of
0.1% Ca(H
2
PO
4
)
2
during the 20 h time sweep. Gel formation was
observed in all other samples, and moduli values were obtained by
frequency scan between 0.1 and 10.0 rad/s at 25

C (Table 3, Figs. S1
and S2). G
0
values were consistently larger than G
00
values, with
tan d < 0.5 in all cases, indicating the presence of a strong, well-
ordered gel. Gel formation generally correlated with DM, with G
0
and G
00
increasing as DM decreased. Given its relatively high DM
(68.3%) and comparatively low DB and DB
abs
values, gel formation
in the 2.202s sample was unexpected. The 2.202s sample
comprised the majority of pectin extracted at OD 2.202, but both
the soluble and insoluble fractions gelled and displayed similar G
0
values These results indicate that juice cloud susceptibility to gel
formation only requires a small decrease in pectin DM brought on
by a comparatively small increase in the availability of stretches of
demethylated GalA. The DM values as correlated to cloud state that
were obtained are valuable in that they indicate the availability of a
very small windowin which juice producers may safely fortify juice
without risking cloud collapse via calcium crosslinking.
As a separate control (Fig. S3), PME was added to reconstituted
FCOJ together with calcium phosphate at pH 5.5. This was done to
determine if time effects of PME modication of juice pectin could
be monitored directly with a rheometer and perhaps relate changes
to cloud loss. At time zero of PME addition to orange juice, G
0
was
approximately 0.1 Pa and G
00
was approximately 0.08 Pa which in-
dicates no gel formation as expected. After 48 h of exposure to PME
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
5 10 15 20 25 30 35 40 45 50 55 60
R
e
l
a

v
e

A
b
u
n
d
a
n
c
e
Degree of Polymerizaon
0.109i
1.595i
1.595s
2.202i
2.202s
Fig. 2. Demethylated oligomer degree of polymerization (DP) for pectic extracts isolated from cloud-compromised juice. Monomer through pentamer was present in all samples,
and off scale where not included.
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 15
there was no discernible change in G
0
or G
00
again indicating no gel
formation, but not necessarily pectin modication. Thus direct
measurement of modulus values of juice with pectins being
modied by PME in situ was not useful for relating to cloud loss.
These data indicate that the preferred method for monitoring PME
modication of juice pectins is to rst isolate the PME modied
pectins from the juice and follow up with rheology and block
analysis of said pectins.
Correlations between the major variables used to describe the
pectic extracts from destabilized FCOJ were veried by calculating
Pearsons correlation coefcient [r] for each pairwise comparison
(Table 4). When all six juice extracts were included in the analysis, a
highly signicant negative correlation (p < 0.01) was observed
between DM and DB
abs
, and a highly signicant positive was
observed between G
0
and G
00
. Strong positive correlations (p <0.05)
were further observed between OD and DM, DB and DB
abs
, and G
0
and DB
abs
. Strong negative correlations were observed between OD
and DB
abs
, and G
0
and DM. These relationships e particularly those
between OD, DM, DB
abs
, and G
0
e conrm the previously illustrated
synergy between cloud loss, pectin demethylation, and calcium
crosslink formation. The juice extracts were further broken apart by
solubility, and [r] was recalculated for each subset. Due to limited
sample size, signicance thresholds were much more difcult to
achieve, though overall trends remained consistent.
3.4. Polysaccharide composition
In order to determine polysaccharide composition, each FCOJ
extract was digested with a cocktail of Rapidase ADEX-P, Pectinex,
and Celluclast, and the resultant soluble sugars were analyzed via
HPAEC-PAD. Rapidase and Pectinex both contain a wide assortment
of enzyme activities, making them appropriate for use in total hy-
drolysis experiments (Galant et al., 2014; Grohmann & Baldwin,
1992; Wilkins, Widmer, Grohmann, & Cameron, 2007). Celluclast
specically targets 1,4-beta-D-glycosidic linkages and reduces cel-
lulose to cellobiose and glucose with assistance from b-glucosi-
dases found in the other enzyme compositions. For any given pectin
extract, glucose and cellobiose together comprised no more than
5.18% of the detected soluble sugars, indicating that the majority of
the insoluble cellulosic material was removed during the extraction
procedure (Table 5). Glucose is also present natively as a mono-
saccharide in orange juice and, along with fructose and sucrose,
contributes to juices sweet avor (Bai et al., 2009). No sucrose, and
only small quantities of fructose were detected in our extracts,
indicating that the majority of soluble sugars of non-pectic origin
were removed during the purication procedure. The remaining
sugars detected e arabinose (Ara), galactose (Gal), rhamnose (Rha),
xylose (Xyl), and GalA e are all known components of pectin, with
GalA comprising the HG rich regions and the other sugars as
components of the backbone (GalA, Rha) or sidechains (Ara, Gal,
Xyl) of the RGI region. Of the two sugars that make up the pectin
backbone, GalA represents approximately 60e80% of the total,
while Rha represents 1e2%. These values are in line with what has
been observed previously with other pectin, particularly other
citrus pectins (Grohmann, Cameron, Kim, Widmer, & Luzio, 2013;
Kravtchenko, Voragen, & Pilnik, 1992; Yapo et al., 2007). Together
Ara and Gal, the two major contributors to the RG1 sidechains,
make up 17e35% of the detected soluble sugars, with Gal contrib-
uting 11e24%. Interestingly, our previous extraction of pectin from
FCOJ that was not treated with PME yielded a material comparably
enriched in Gal (31%) and Ara (16%) and decient in GalA (48%)
(Galant et al., 2014). Our investigation into the source of the Gal and
Ara indicated that a large proportion was not conjugated directly to
the pectic backbone, but rather was present as sidechains on ara-
binogalactan proteins (AGPs). While proteinase K (Pk) was only
used as part of a molecular weight assessment in the prior study, in
this study we employed it during dialysis of the extracts to improve
overall pectin yields. Accordingly, much of the non-pectic Ara and
Gal content was lost, resulting in higher relative percentages of
GalA.
4. Conclusions
Pectin methylesterases were extracted from C. sinensis cultivar
Pineapple and used to systematically reduce the suspended cloud
present in commercial FCOJ. Pectin was acid extracted from the
resultant series of FCOJ samples, and its structural properties,
including DM, DB, and G
0
were assessed. Strong correlations were
observed between cloud OD, pectin DM, and pectin G
0
, indicating
pectin crosslinking proceeds, or has the potential to proceed
pending the availability of calcium, in a predictable manner. While
uncompromised juice cloud extracts were previously found to be
enriched in arabinose and galactose due to the presence of AGPs,
Table 4
Pearsons correlation coefcient [r] for variables describing the extracted pectins.
OD DM DB DBabs G
0
G
00
All
OD e **0.916 *0.772 **0.914 *0.782 0.629
DM e e *0.792 ***0.988 **0.818 0.725
DB e e e **0.879 0.662 0.650
DBabs e e e e **0.836 *0.763
G
0
e e e e e ***0.962
G
00
e e e e e e
Insoluble
OD e **0.9998 0.438 *0.985 0.672 0.370
DM e e 0.418 0.981 0.656 0.349
DB e e e 0.586 0.960 **0.997
DBabs e e e e 0.790 0.524
G
0
e e e e e 0.936
G
00
e e e e e e
Soluble
OD e 0.935 *0.996 0.957 0.824 0.885
DM e e 0.962 **0.998 0.971 *0.993
DB e e e 0.978 0.869 0.921
DBabs e e e e 0.953 0.982
G
0
e e e e e *0.993
G
00
e e e e e e
Asterisks indicate the signicance of the correlation coefcient; *p < 0.1, **p < 0.05,
***p < 0.01.
Table 5
Contribution of individual sugars (%wt) in supernatants of enzyme hydrolyzed pectin extracts.
Sample name Rhamnose Arabinose Galactose Glucose Xylose Fructose Sucrose Cellobiose GalA
0.109i 0.86 11.90 19.71 1.62 0.64
a
1.46 63.81
1.595i 0.76 5.40 11.82 0.85 0.50 0.16 80.49
1.595s 1.33 5.75 15.29 1.78 0.60 0.42 0.65 74.20
2.202i 2.38 10.99 21.53 3.29 0.97 0.26 1.89 58.69
2.202s 1.06 12.43 18.80 2.03 0.68 0.13 0.22 64.65
a
Empty cells indicate that no sugar was detected.
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 16
treatment of cloud-compromised extracts with Pk results in soluble
sugar abundances more in line with those seen in citrus peel. Most
importantly, a relatively low threshold for rheological activity,
corresponding to a reduction in pectin extract DM from 74.7% to
only 68.3%, was found. Although accompanied by a corresponding
drop in OD, this reduction in DM was not accompanied by any vi-
sual indication of cloud loss. The availability of such a narrow DM
window for safely fortifying juice with calcium has concerning
implications for juice producers and distributors whose anecdotal
reports suggest that previously stabilized juice cloud can become
unstable upon calcium fortication (140 mg calcium per 100 g
juice; U.S. Department of Agriculture, 2013). This prompts the need
for future work on the early detection of cloud loss potential, the
structural events involved in its initiation and the effect of calcium
addition to juice when only minimal amounts and sizes of deme-
thylated blocks have been introduced into the pectin component of
juice cloud.
Funding
This research was supported by USDA ARS CRIS 6621-41000-
016-00D and the USDA, ARS Postdoctoral Research Associate Pro-
gram: Class of 2011. These Funding Sources had no role in study
design; in the collection, analysis, and interpretation of data; in the
writing of the report; or in the decision to submit the paper for
publication.
Acknowledgments
We thank Elena Branca, Steven W. Kauffman, and Sandra Mat-
lack for technical assistance.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.foodhyd.2014.03.013.
References
Ackerley, J. L., Corredig, M., & Wicker, L. (2002). Clarication of citrus juice is
inuenced by specic activity of thermolabile pectinmethylesterase and inac-
tive PMEepectin complexes. Journal of Food Science, 67, 2529e2533.
Audebrand, M., Kolb, M., & Axelos, M. A. V. (2006). Combined rheological and ul-
trasonic study of alginate and pectin gels near the solegel transition. Bio-
macromolecules, 7(10), 2811e2817.
Bai, J., Baldwin, E., Plotto, A., Manthey, J. A., McCollum, G., Irey, M., et al. (2009).
Inuence of harvest time on quality of Valencia oranges and juice. Proceedings
of the Florida State Horticultural Society, 122, 308e315.
Baker, R. A. (1979). Clarifying properties of pectin fractions separated by ester
content. Journal of Agricultural and Food Chemistry, 27(6), 1387e1389.
Baker, R. A., & Bruemmer, J. H. (1969). Cloud stability in the absence of various
orange juice soluble components. Proceedings of the Florida State Horticultural
Society, 82, 215e220.
Braddock, R. J. (1999). Handbook of citrus by-products and processing technology (1st
ed.). New York: John Wiley and Sons, Inc.
Cameron, R. G., Baker, R. A., & Grohmann, K. (1998). Multiple forms of pectinme-
thylesterase from citrus peel and their effects on juice cloud stability. Journal of
Food Science, 63, 253e256.
Cameron, R. G., Luzio, G. A., Goodner, K., & Williams, M. A. K. (2008). Demethylation
of a model homogalacturonan with a salt-independent pectin methylesterase
from citrus: I. Effect of pH on demethylated block size, block number and
enzyme mode of action. Carbohydrate Polymers, 71(2), 287e299.
Cameron, R. G., Luzio, G. A., Vasu, P., Savary, B. J., & Williams, M. A. K. (2011).
Enzymatic modication of a model homogalacturonan with the thermally
tolerant pectin methylesterase from citrus: 1. Nanostructural characterization,
enzyme mode of action, and effect of pH. Journal of Agricultural and Food
Chemistry, 59(6), 2717e2724.
Cameron, R. G., Savary, B. J., Hotchkiss, A. T., Fishman, M. L., Chau, H. K., Baker, R. A.,
et al. (2003). Separation and characterization of a salt-dependent pectin
methylesterase from Citrus sinensis var. Valencia fruit tissue. Journal of Agri-
cultural and Food Chemistry, 51(7), 2070e2075.
Chen, E. M. W., & Mort, A. J. (1996). Nature of sites hydrolyzable by endopolyga-
lacturonase in partially-esteried homogalacturonans. Carbohydrate Polymers,
29(2), 129e136.
Coenen, G. J., Bakx, E. J., Verhoef, R. P., Schols, H. A., & Voragen, A. G. J. (2007).
Identication of the connecting linkage between homo- or xylogalacturonan
and rhamnogalacturonan type I. Carbohydrate Polymers, 70(2), 224e235.
Corredig, M., Kerr, W., & Wicker, L. (2001). Particle size distribution of orange juice
cloud after addition of sensitized pectin. Journal of Agricultural and Food
Chemistry, 49(5), 2523e2526.
Croak, S., & Corredig, M. (2006). The role of pectin in orange juice stabilization:
effect of pectin methylesterase and pectinase activity on the size of cloud
particles. Food Hydrocolloids, 20, 961e965.
Daas, P. J., Voragen, A. G., & Schols, H. A. (2000). Characterization of non-esteried
galacturonic acid sequences in pectin with endopolygalacturonase. Carbohy-
drate Research, 326(2), 120e129.
Doco, T., Williams, P., Vidal, S., & Pellerin, P. (1997). Rhamnogalacturonan II, a
dominant polysaccharide in juices produced by enzymatic liquefaction of fruits
and vegetables. Carbohydrate Research, 297, 181e186.
Ellerbee, L., & Wicker, L. (2011). Calcium and pH inuence on orange juice cloud
stability. Journal of the Science of Food and Agriculture, 91(1), 171e177.
Espinosa-Muoz, L., Renard, C. M. G. C., Symoneaux, R. B., Biau, N., & Cuvelier, G.
(2013). Structural parameters that determine the rheological properties of ap-
ple puree. Journal of Food Engineering, 119(3), 619e626.
Fries, M. F., Ihrig, J., Brocklehurst, K., Shevchik, V. E., & Pickersgill, R. W. (2007).
Molecular basis of the activity of the phytopathogen pectin methylesterase.
EMBO Journal, 26, 3879e3887.
Galant, A. L., Luzio, G. A., Widmer, W. W., & Cameron, R. G. (2014). Compositional
and structural characterization of pectic material from frozen concentrated
orange juice. Food Hydrocolloids, 35, 661e669.
Gattuso, G., Barreca, D., Gargiulli, C., Leuzzi, U., & Caristi, C. (2007). Flavonoid
composition of citrus juices. Molecules, 12(8), 1641e1673.
Grohmann, K., & Baldwin, E. A. (1992). Hydrolysis of orange peel with cellulase and
pectinase enzymes. Biotechnology Letters, 14, 1169e1174.
Grohmann, K., Cameron, R., Kim, Y., Widmer, W., & Luzio, G. (2013). Extraction and
recovery of pectic fragments from citrus processing waste for co-production
with ethanol. Journal of Chemical Technology and Biotechnology, 88(3), 395e
407.
Guillotin, S. E., Bakx, E. J., Boulenguer, P., Mazoyer, J., Schols, H. A., & Voragen, A. G. J.
(2005). Populations having different GalA blocks characteristics are present in
commercial pectins which are chemically similar but have different function-
alities. Carbohydrate Polymers, 60(3), 391e398.
Joye, D. D., & Luzio, G. A. (2000). Process for selective extraction of pectins from
plant material by differential pH. Carbohydrate Polymers, 43(4), 337e342.
Kim, Y., Williams, M. A. K., Galant, A., Luzio, G. A., Savary, B. J., Vasu, P., et al. (2013).
Nanostructural modication of a model homogalacturonan with a novel pectin
methylesterase: effects of pH on nanostructure, enzyme mode of action and
substrate functionality. Food Hydrocolloids, 33(1), 132e141.
Klavons, J. A., & Bennet, R. D. (1987). Nature of the pectin constituent of commercial
lemon juice cloud. Journal of Agricultural and Food Chemistry, 35, 159e161.
Klavons, J. A., Bennet, R. D., & Vannier, S. H. (1991). Nature of the protein constituent
of commercial orange juice cloud. Journal of Agricultural and Food Chemistry, 39,
1545e1548.
Klavons, J. A., Bennet, R. D., & Vannier, S. H. (1992). Stable clouding agent from
isolated soy protein. Journal of Food Science, 57(4), 945e947.
Klavons, J. A., Bennett, R. D., & Vannier, S. H. (1994). Physical/chemical nature of
pectin associated with commercial orange juice cloud. Journal of Food Science,
59(2), 399e401.
Kravtchenko, T. P., Voragen, A. G. J., & Pilnik, W. (1992). Analytical comparison of
three industrial pectin preparations. Carbohydrate Polymers, 18(1), 17e25.
Krop, J. (1974). The mechanism of cloud loss phenomena in orange juice. Wageningen,
Netherlands: Agricultural University.
Limberg, G., Korner, R., Buchholt, H. C., Christensen, T. M. I. E., Roepstorff, P., &
Mikkelsen, J. D. (2000a). Analysis of different de-esterication mechanisms for
pectin by enzymatic ngerprinting using endopectin lyase and endopolyga-
lacturonase II from A. niger. Carbohydrate Research, 327(3), 293e307.
Limberg, G., Korner, R., Buchholt, H. C., Christensen, T. M., Roepstorff, P., &
Mikkelsen, J. D. (2000b). Quantication of the amount of galacturonic acid
residues in blocksequences in pectin homogalacturonan by enzymatic nger-
printing with exo- and endo-polygalacturonase II from Aspergillus niger. Car-
bohydrate Research, 327(3), 321e322.
Liners, F., Thibault, J.-F., & Cutsem, P. V. (1992). Inuence of the degree of poly-
merization of oligogalacturonates and of esterication pattern of pectin on their
recognition by monoclonal antibodies. Plant Physiology, 99, 1099e1104.
Luzio, G. A., & Cameron, R. (2010). Contactless conductivity: an HPLC method to
analyze degree of methylation of pectin. Proceedings of the Florida State Horti-
cultural Society, 123, 213e216.
Luzio, G. A., & Cameron, R. G. (2008). Demethylation of a model homogalacturonan
with the salt-independent pectin methylesterase from citrus: II. Structure
function analysis. Carbohydrate Polymers, 71, 300e309.
May, C. D. (1990). Industrial pectins: sources, production and applications. Carbo-
hydrate Polymers, 12, 79e99.
Mercadente, D., Melton, L. D., Jameson, G. B., Williams, M. A. K., & De Simone, A.
(2013). Substrate dynamics in enzyme action: rotations of monosaccharide
subunits in the binding groove are essential for pectin methylesterase proc-
essivity. Biophysical Journal, 104, 1731e1739.
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 17
Powell, D. A., Morris, E. R., Gidley, M. J., & Rees, D. A. (1982). Conformations and
interactions of pectins. II. Inuences of residue sequence on chain association in
calcium pectate gels. Journal of Molecular Biology, 155(4), 517e531.
Rouse, A. H. (1949). Gel formation in frozen citrus concentrates thawed and stored
at 40

F. Proceedings of the Florida State Horticultural Society, 170e173.

Savary, B. J., Cameron, R. G., Luzio, G. A., McCollum, T. G., Vasu, P., & Nunez, A. (2010).
In U. S. P. Ofce (Ed.), Thermally-tolerant pectin methylesterase (p. 35). USA.
Savary, B. J., & Vasu, P. (2012). Routine identity conrmation of recombinant pro-
teins by MALDIeTOF mass spectrometry. In A. Lorence (Ed.), Recombinant gene
expression, reviews and protocols (3rd ed.). New York, NY: Humana Press.
Scott, W. C., Kew, T. J., & Veldhuis, M. K. (1965). Composition of orange juice cloud.
Journal of Food Science, 30, 833e837.
Swift, L. J., & Veldhuis, M. K. (1957). Composition of commercial, segment and peel
juices of Florida oranges. Journal of Agricultural and Food Chemistry, 5, 49e52.
Tanhatan-Nasseri, A., Crepeau, M. J., Thibault, J. F., & Ralet, M. C. (2011). Isolation and
characterization of model homogalacturonans of tailored methylesterication
patterns. Carbohydrate Polymers, 86(3), 1236e1243.
United States Pharmacopeia and National Formulary USP24-NF19. (2000). Ofcial
monographs (Vol. 2). Rockville, MD: United States Pharmacopeia Convention,
1265.
U.S. Department of Agriculture, Agricultural Research Service. (2013). USDA National
Nutrient Database for Standard Reference. Release 26. Nutrient Data Laboratory
Home Page http://www.ars.usda.gov/ba/bhnrc/ndl.
Widmer, W. (2011). Analysis of biomass sugars and galacturonic acid by gradient
anion exchange chromatography and pulsed amperometric detection without
post-column addition. Biotechnology Letters, 33, 365e368.
Wilkins, M. R., Widmer, W. W., Grohmann, K., & Cameron, R. G. (2007). Hydrolysis of
grapefruit peel waste with cellulase and pectinase enzymes. Bioresource Tech-
nology, 98(8), 1596e1601.
Yapo, B. M., Lerouge, P., Thibault, J.-F., & Ralet, M.-C. (2007). Pectins from citrus peel
cell walls contain homogalacturonans homogenous with respect to molar mass,
rhamnogalacturonan I and rhamnogalacturonan II. Carbohydrate Polymers,
69(3), 426e435.
A.L. Galant et al. / Food Hydrocolloids 41 (2014) 10e18 18