Effect of Pectin Methylesterase on Carrot (Daucus carota) Juice CloudStabilityAlison K. Schultz,†,§ Gordon E. Anthon,† Stephanie R. Dungan,‡ and Diane M. Barrett*,†

†Department of Food Science and Technology, University of California, Davis, California 95616, United States‡Department of Chemical Engineering and Materials Science, University of California, Davis, California 95616, United States

ABSTRACT: To determine the effect of residual enzyme activity on carrot juice cloud, 0 to 1 U/g pectin methylesterase (PME)was added to pasteurized carrot juice. Cloud stability and particle diameters were measured to quantify juice cloud stability andclarification for 56 days of storage. All levels of PME addition resulted in clarification; higher amounts had a modest effect incausing more rapid clarification, due to a faster increase in particle size. The cloud initially exhibited a trimodal distribution ofparticle sizes. For enzyme-containing samples, particles in the smallest-sized mode initially aggregated to merge with the secondpeak over 5−10 days. This larger population then continued to aggregate more slowly over longer times. This observation of amore rapid destabilization process initially, followed by slower subsequent changes in the cloud, was also manifested inmeasurements of sedimentation extent and in turbidity tests. Optical microscopy showed that aggregation created elongated,fractal particle structures over time.

KEYWORDS: pectin methylesterase, Daucus carota, carrot juice, particle size, relative sediment height, relative turbidity

■ INTRODUCTION

Many of the sensory characteristics associated with fruit andvegetable juices, such as color, flavor, texture, and appearance,are attributable to the presence of cloud particles suspended inthe liquid matrix of the juice.1,2 The destabilization, or“clarification,” of cloud is a serious problem for the juiceindustry. When cloud becomes unstable, the particles flocculateand settle out of the continuous phase, resulting in a clear,flavorless serum and a potentially unappealing precipitate at thebottom of the container. It is commonly thought that clouddestabilization occurs through the action of either the enzymepectin methylesterase (PME) or by acidification of juices belowtheir natural pH.The purpose of this study is to examine the effects of

different amounts of PME on cloud stability and how suchadditions affect particle size as well as the rate and amount ofclarification within the juice. From these results, mechanisms ofclarification can be proposed. This study seeks to improve ouroverall understanding of how PME affects cloud stability, a goalthat is of broad interest to scientific and commercialcommunities. To prevent cloud destabilization, the commonindustrial method is to inactivate PME entirely.3 During juiceprocessing, carrots undergo enzyme inactivation within one dayof peeling by several common blanching methods, includingsteaming or immersion in hot water or acid.3 If inactivation isnot accomplished within one day, loss of cloud stability andclarification will result.3 The attempt to completely inactivatePME using heat results in deleterious effects on juice color,flavor, nutrients and viscosity. The present study looks atresidual enzyme levels instead of complete inactivation, andcould be used to determine whether blanching operations couldbe reduced, saving energy and maintaining quality.There is considerable interest in minimally processed juices,

and so in order to improve retention of initial fruit andvegetable quality, advanced processing methods such as high

pressure and electric field processing are being explored. Theseprocesses achieve microbial inactivation using very little heat,and therefore, quality is superior; however, residual enzymeactivity of pectin methylesterase in particular has been shown tolimit shelf life. Very few studies have addressed the effects of aknown amount of residual enzyme activity on refrigerated juicesThe orange juice industry sets a goal of inactivating PME to

below 10% of its initial level,4,5 because it is assumed that evena small amount of residual PME can be deleterious to cloudstability. However, in pineapple juice residual PME levels of 4.1× 10−4 U/mL resulted in a product with acceptable cloudstability, while 7.1 × 10−3 U/mL caused complete clarification.6

For citrus fruit juices, it has been established that when PMEactivity levels are kept below 10−4 U/mL cloud loss isprevented.7 This residual level for citrus fruits has beenassumed to be the same for tomato products;8,9 however, thishas not been investigated. Some level of residual enzymeactivity may be acceptable in juice products, depending on theshelf life of the product.

Characterization of PME. In plants, PME is cell wall-bound and mainly found in the middle lamella between theindividual cells, where it is important in plant development andfruit ripening.10−12 PME catalyzes the hydrolysis of methylesters from the O6 position of galacturonic acid monomers ofwhich pectin is composed. This reaction increases the numberof free carboxylic acid groups within the pectin chain, andproduces methanol and hydrogen ions.8,11,13,14 PME actiontypically does not result in complete deesterification, as 20−30% of the methyl esters are not removed.15

Received: October 1, 2013Revised: January 2, 2014Accepted: January 8, 2014

Article

pubs.acs.org/JAFC

© XXXX American Chemical Society A dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Hydrolysis of methyl ester groups from the pectin backboneby PME can occur via three different mechanisms.16,17 The firstmechanism involves a multiple chain reaction, in which PMEremoves one methyl ester from the pectin chain and thendissociates from the substrate, resulting in random deester-ification of the chain. This is typically the mode of action ofacidic fungal PME.16,18 The second mechanism is single chain,or linear, deesterification, where PME follows along one chainand cleaves all the methoxyl groups, which results in blocks ofdemethylated pectin. This is generally the mode of deester-ification for higher plant PME.16,18 This would suggest thatcarrot PME also employs linear deesterification, although noresearch to date has confirmed this. The third mechanisminvolves multiple attacks of the enzyme on methyl esters, whereonly a limited number of methyl groups are removed from eachpectin substrate. The mechanism by which carrot PMEcatalyzes deesterification of carrot pectin has not yet beendetermined.Carrot Juice Cloud Stability. The rate of juice clarification

is governed by two mechanisms: flocculation and sedimenta-tion. Flocculation is the phenomenon in which particles clustertogether to form aggregates, increasing the effective particlesize. Sedimentation is the settling out of flocculated particlesfrom suspension to the bottom of the container. Typically,flocculation leads to faster sedimentation in accordance withStokes’ Law. Larger particles formed from flocculation increasethe rate of sedimentation and, hence, cause more rapid loss ofcloud stability.If particle size influences the rate of cloud loss and

clarification, then it is worthwhile exploring how particle sizeincreases within juices over storage time. In a suspension,particles remain separate from one another due to the presenceof an energy barrier between them, which reduces their rate ofcollision. The energy barrier is influenced by several differentcontributions, including attractive van der Waals interactions(which reduce the barrier) and repulsive electrostatic forces(which increases it).19−22 Factors that lower the energy barriermake it more likely that cloud particles will closely approachone another, at which point they can “stick” to each other andfloc as a result of short-range attraction.In juices with active enzymes, PME cleaves methoxyl groups

from the pectin backbone and additional negatively chargedresidues are exposed. These like-charged residues wouldnormally enhance the repulsive interparticle energy barrier;however, naturally present divalent cations such as calcium maybind to these sites, creating electrostatic attractive interactionsbetween particles. The calcium can then cross-link pectinchains on different particles together, thereby forming insolublecalcium pectate within aggregates that are large enough tosediment out according to Stokes’ Law.1,17,20,23

Little research has been conducted regarding the effect thatresidual PME levels have on cloud stability of carrot juice,which is a low-acid juice product. Therefore, the goal of thisstudy was to observe the effects of adding PME to pasteurizedcarrot juice and suggest a mechanism for cloud flocculation andsedimentation.

■ MATERIALS AND METHODSMaterial Sourcing. Sugarsnax variety carrots (Daucus carota L.

ssp. sativa cv. ‘Sugarsnax’) were used to make the PME extraction, andwere sourced from Bolthouse Farms in Bakersfield, CA. The averagePME activity in raw carrot juice made from these carrots was found tobe 0.43 U/g (range: 0.218−0.806 U/g), as measured by individually

juicing 30 Sugarsnax carrots on a benchtop juicer. Commercialpasteurized Bolthouse carrot juice (pH 6.20) was purchased from alocal grocery store. All containers of purchased juice were mixedtogether to ensure a uniform starting material. The PME activity of thecommercial juice was found to be less than 5% of the average carrotjuice PME.

Reagents. Sodium chloride, monobasic monohydrate sodiumphosphate, dibasic anhydrous sodium phosphate, citric acid mono-hydrate, sodium citrate, hydrochloric acid, sodium hydroxide, sulfamicacid, and ferric ammonium sulfate were obtained from Fisher Scientific(Fair Lawn, NJ). Trizma base, Trizma hydrochloride, citrus pectin,Bradford Reagent, lauryl sulfate, bovine serum albumin (BSA),Fluoral-P, alcohol oxidase, and 3-methyl-2-benzothiazolinone hydra-zine hydrochloride hydrate were obtained from Sigma Aldrich Co. (St.Louis, MO). Micro Bio-Spin desalting columns were obtained fromBio Rad (Hercules, CA), and membrane tubing was sourced fromSpectrum Medical Industries, Inc. (Houston, TX). Deionized distilledwater was used in all experiments.

PME Preparation and Characterization. PME Extraction. PMEadded to commercial carrot juice was prepared from an extraction ofraw carrots based on the method of Ly-Nguyen et al.24 A sample of600−700 g of washed carrots was homogenized with 600 mLdeionized water in a Vita-Mix blender set on high for 2 min (ModelVM0101D, Vita-Mix Co., Cleveland, OH). The homogenate wasfiltered through four layers of cheesecloth, and the filtrate wasdiscarded. The pellet was washed four times with 1 L deionized waterto ensure the removal of color and organic compounds. The pellet wasthen mixed with 1:4 (w/v) 0.2 M Tris and 1 M NaCl (pH 8.0) for 30min to extract PME into the supernatant, and the carrot slurry wasfiltered through four layers of cheesecloth. The supernatant wascollected and the pellet discarded.

The supernatant was stirred with 30% ammonium sulfate for 30 minas a purification step to precipitate large proteins that were not PME.This mixture was then centrifuged at 18 000g for 15 min at 20 °C in aSorvall centrifuge (Model RC5C, DuPont, Wilmington, DE). Aftercentrifugation, the pellet was discarded and the supernatant wasvacuum filtered (Whatman #4) to remove any remaining largeparticles.

To precipitate the PME enzyme, 80% ammonium sulfate was slowlymixed into the supernatant. The mixture was stirred for 30 min, andthen centrifuged at 18 000g for 15 min at 20 °C. The supernatant wasdiscarded, and the PME precipitate that collected on the side of thecentrifuge tubes was redissolved in 0.05 M citric acid/sodium citratebuffer (pH 6.5). The redissolved PME was then dialyzed in membranetubing (MWCO 3500) in the citric acid/sodium citrate buffer in twostages. In the first stage, PME was dialyzed for 1 h at 4 °C, duringwhich time highly concentrated molecules such as salts rapidly diffuseinto the surrounding buffer and slow further diffusion. Therefore, forthe second stage, fresh buffer solution was added to allow forcontinuous rapid diffusion, and the PME was dialyzed overnight at 4°C. Once dialysis was complete, the PME, suspended in buffer, wasstored in 1 mL aliquots at −20 °C.

The specific activity of three sample replicates from the extractionwas measured by the PME assay and Bradford protein determination,and these three activity values were averaged to obtain the specificactivity of that batch of PME extract. This concentrated PMEpreparation had an activity of 116 U/mL, and a specific activity of 123U/mg protein.

Bradford Protein Determination. The Bradford protein assay wasbased on the method described in the Technical Bulletin of theBradford Reagent by Sigma-Aldrich.25 The procedure was modified sothat only one-third of the assay volume was used, in order to cut downon hazardous waste. Therefore, 0.33 mL of samples and standards wasadded to 1 mL Bradford reagent, and the solution was incubated atroom temperature (25 °C) for 5−10 min and then measured at 595nm in a Shimadzu spectrophotometer (Columbia, MD).

PME Assay. PME activity was determined based on the assaydeveloped by Anthon and Barrett (2004), in which a crude extract ofPME from carrot juice was added to a standard amount of pectin.26

The methanol produced in this reaction was used to quantify the

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXB

amount of enzyme activity and, hence, the amount of enzyme presentin the sample. Pectin used in the assay was dialyzed overnight in water.Assays were started by adding 5 μL of a 10-fold dilution of the carrotjuice to 95 mL of the assay medium, which contained 3 mg/mL pectinand had a pH of 7.5. All samples were placed into a 30 °C water bath,and at four time points (40, 70, 100, and 130 min) after the start of theassay, 0.1 mL aliquots were removed and added to 0.1 mL of asolution containing 5 mg/mL of each ferric ammonium sulfate andsulfamic acid. After 30 min at room temperature, 0.8 mL of water wasadded and absorbance at 620 nm was determined using a Shimadzuspectrophotometer (Columbia, MD). Methanol concentrations weredetermined from a separate methanol standard curve. The amount ofmethanol formed was plotted and the activity was calculated from theslope of the plot.Cloud Stability during Juice Storage. PME Addition. The PME

extract, made as described above, was added to commercial carrot juicein 5 quantities, 0, 0.15, 0.25, 0.5, and 1 U/g, where a unit of enzymeactivity (U) is defined as the number of micromoles of methanolproduced in 1 min, and grams refers to the mass of carrot juice. Sincedifferent volumes of PME extract (enzyme suspended in 0.05 M citricacid/sodium citrate buffer) were added to the juice to achieve differentenzyme levels, it was necessary to compensate for the subsequentchanges in juice volume. The largest volume of extract added (foraddition of 1 U/g) was used as a baseline, and the other samples had acombination of enzyme extract and citric acid/sodium citrate bufferadded so that all samples had the same final volume. This was done toensure each juice was diluted the same small amount and any effects ofthe buffer were mitigated. Final juice pH after PME addition was 6.20.To prevent microbial growth during the storage study, 0.2% sodiumazide was added to the juice. Juices were stored for up to 56 days at 5°C.Relative Sediment Height. Relative sediment height, also known as

the sedimentation test, is a measure of the quantity of visible settlingthat occurs in the juice.21,27 It is the height (Hsediment) of the cloudsediment divided by the total height (Htotal) of the juice in a vial.Samples used for measuring the relative sediment height of the

sediment were placed in 7 mL glass vials with screw caps and allowedto remain undisturbed except for gentle transportation of the vialsfrom the refrigerator to the laboratory bench for measurements. Acaliper was used to make height measurements in triplicate.Relative Turbidity. Relative turbidity is a spectrophotometric

method to determine cloud stability by comparing the turbidity in ajuice sample before and after centrifugation.21,22,28−30 For the carrotjuices examined here, absorbance at 660 nm was used as the measureof juice turbidity. At this wavelength, absorbance is predominantlydetermined by the scattering due to the particles, rather than specificchemical absorption.31 Relative turbidities were determined bymeasuring the absorbance values for the whole juice and for thesupernatants obtained after centrifugation (4200g, 15 min, 20 °C).When absorbance values measured in standard 1 cm cuvettes wereunacceptably high, shorter path length cuvettes (0.5 mm for wholejuice and 5.0 mm for juice supernatants) were used. Turbidity (τ) wascalculated as the absorbance divided by the path length used formeasurement. Relative turbidities (expressed as %) were calculated asτserum/τtotal × 100.Particle Size Determination. Static light scattering was used to

measure average particle size.21,28,30,32−34 A laser is passed through thesample and, upon hitting a particle, diffracts and creates a light patternagainst the detector that depends on the particle size. Measurementswere performed using a Microtrac S3500 (Montgomeryville, PA).Results were obtained as the volume fraction of particles containedwithin a discrete size range (bin). The mean diameter was determinedas a volume average:

=∑∑

Dn dn d

( )( )

I I

I I4,3

4

3 (1)

Here D4,3 is the volume mean diameter, ni is the number of particles inbin i, and di is their diameter. Carrot juice was inverted gently 10 timesto ensure even distribution of particles and then added to the

instrument sample port, which was operated at a flow rate of 15% ofmaximum. Particles were assumed to be irregular in shape andabsorbing, and a refractive index value for water of 1.333 was used.Three replicates were measured for each sample.

■ RESULTSEffect of Added PME on Cloud Stability. It is generally

understood that PME impacts juice cloud stability, but the timeframe over which destabilization occurs and the impact of theamount of active enzyme is not known, nor is the mechanism ofcloud destabilization by PME well understood. Therefore, astorage study was conducted that examined the effects ofdifferent quantities of added enzyme on the cloud stability ofcommercial carrot juice. Commercial pasteurized carrot juicewas preferred as the starting material for these studies, ratherthan raw unpasteurized juice made in the laboratory, because ofthe former’s uniform particle content and minimal active PME(see Material Sourcing in Materials and Methods). Startingwith juice that lacked any PME, then adding known amounts ofa partially purified carrot PME, juices with different knownlevels of PME activity could be obtained. The levels of PMEadded (0.15, 0.25, 0.5, and 1.0 U/g) were based on the level ofendogenous PME measured in raw Sugarsnax carrot juice madewith a benchtop juicer. This level varied widely across the juicefrom individual carrots, even within carrots planted andharvested at the same time and grown in the same field. For30 individually measured carrots, the PME activity of the juiceranged from 0.218 to 0.806 U/g, with an average of 0.43 U/g.Cloud stability was measured via relative turbidity (τserum/

τtotal), which has been used previously by others as a methodsensitive to the extent of flocculation within the cloud.21,28,29

Flocculation creates larger particles within the suspension,which are removed more effectively by centrifugation, and leavebehind a serum with a lower particle concentration and lowerturbidity. Thus, smaller values of τserum/τtotal indicate that moreflocculation has occurred. As shown in Figure 1, at all levels of

added PME the cloud was less stable than in the control juicewith no added enzyme. The relative turbidity of the controlsample remained consistently high throughout storage, whilethose samples with added PME showed a large degree ofclarification after centrifugation (Figure 1).When the amount ofadded PME was at or above 0.25 U/g, there was an inverserelationship at the 90% confidence level between the level ofPME addition and the relative turbidity of the cloud at periodslonger than 15 days.

Figure 1. Relative turbidity (τserum/τtotal, %) of carrot samples with 0(□), 0.15 (◇), 0.25 (△), 0.5 (○), and 1.0 (∗) U/g of added PMEenzyme over the course of 56 days of storage.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXC

Loss of cloud stability and juice clarification was alsomonitored by measuring relative sediment height (Hsediment/Htotal). Consistent with the results for relative turbidity, muchless clarification occurred in the control than in the enzyme-treated samples (Figure 2). However, the relative sediment

height of the control juice increased slowly over the course ofthe storage period, indicating that there was natural settlingoccurring within the juice that is not attributable to PMEactivity.All samples containing active PME had an initial period (∼10

d) of rapid sedimentation, after which the sediment heightleveled off (Figure 2). The kinetics of this process can berepresented using the empirical formula

=+

∞HH

t Ht T

sediment

total

2

2o2

(2)

where t is time, H∞ is the steady-state dimensionless plateauheight, and To is a characteristic time of the initial, rapidsedimentation period. H∞ and To could thus be obtained by afit of eq 2 to the data in Figure 2; results for these two fitparameters are given in Table 1.

Carrot juice with greater amounts of added enzyme becameunstable more quickly during storage, as shown by the numberof days (To) it took for the initial rapid sedimentation period tooccur (Table 1). With added enzyme, just a few days wereneeded for the cloud to separate visually into a sediment layerwith approximately steady thickness. The thickness of thatsediment layer increased slightly but significantly (p ≥ 0.05) asmore enzyme was added into individual samples.

Effects of Added PME on Particle Size.Measurements ofrelative turbidity and sedimentation indicated that PMEaddition increased both flocculation and clarification rates inthe cloud. Rates of flocculation can be estimated more directlyby analyzing changes in particle sizes. As discussed above, theparticle size distribution of the entire juice sample wasmeasured by gently inverting the juice samples, thus ensuringan even distribution of particle sizes throughout each sample,and then using static light scattering to determine particlediameters. Inspection of the particle size distribution profilesfor each level of enzyme added showed that the carrot juicestarted with a trimodal distribution (Figure 3), i.e., threeoverlapping populations with different modes of approximately0.7, 1.7, and 13 μm. In the absence of enzyme, this distributiondid not change over time (Figure 3a). For the enzyme-containing systems, (Figure 3b−e), there was a rapid initialchange occurring over the first ∼9 days, during which thesecond modes shifted to larger diameters, leaving the size of theother two modes unaffected. As a result, the height of the firstpeak at smaller sizes shrank over time as that of the larger modegrew.Examination of all enzyme levels (Figure 4a) indicated that

the extent of this shift at earlier times did not strongly dependon the amount of enzyme; there was a similar increase inparticle size for mode 2 at all levels of added enzyme. The firstrapid shift was followed by a second, slower change in thedistribution, as the second peak diameter gradually increased tomerge with the third mode over ∼30−40 days. This secondgrowth process occurred more rapidly at higher enzymeconcentrations at longer times. Consequently, data at day 5exhibited little dependence on the amount of enzyme added, aslong as some active enzyme was present (Figure 4a). Sampleswith 0.15, 0.25, and 0.5 U/g added PME also had very similarparticle size distributions at 33 days of storage; it was only thesample with 1.0 U/g added PME that had a significantly greaterparticle size (Figure 4b). Finally, at day 43, the two highestenzyme concentrations both exhibited more significant particlegrowth than the other samples (Figure 4c).The increase in particle size is an indication that flocculation

was occurring within the cloud.30,35 This is qualitatively visiblein the light microscopy photos (Figure 5), as one candistinguish the individual particles that make up an aggregate.Clearly, the presence of active enzyme corresponded to theformation of flocs in the mixture, with the size of those flocslarger in the two samples with highest amounts of enzyme.Overall, the data in Figures 1−5 consistently indicates that asparticles flocculated and grew in diameter, they began to settleout, causing clarification. Therefore, it can be concluded thatthe amount of enzyme present had a direct effect on the rate offlocculation and sedimentation. In addition, measurements ofturbidity, sedimentation, and particle size distributions allindicate a process that occurred more rapidly over an initialperiod of ∼5−10 days. This period is followed by a time ofslower flocculation rates with little change in the observedsediment layer, although changes in the cloud structure werestill occurring.

■ DISCUSSIONAdding PME to carrot juice caused juice clarification, asmeasured by sediment height and relative turbidity. With moreadded enzyme, the juice clarified more rapidly and to a greaterextent. The clarification was accompanied by an increase inparticle size. Although no previous research has been found on

Figure 2. Relative sediment height (Hsediment/Htotal, %) of carrotsamples with 0 (□), 0.15 (◇), 0.25 (△), 0.5 (○) and 1.0 (∗) U/g ofadded PME over the course of 56 days of storage at 5 °C.

Table 1. Sediment Plateau Height and Time of Initial HeightIncrease From Eq 2, for the Four Levels of Added PMEEnzyme

quality of fit

enzyme added (U/g) H∞ plateau height (%) To (days) R2 χ2

0 - - - -0.15 11 4.9 0.99 1.30.25 11 3.9 0.98 0.800.50 13 3.2 0.98 1.21.00 13 2.1 0.92 2.1

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXD

the addition of different enzyme levels back into juice, theseresults are consistent with those reported for native residualenzyme activity in commercial juices. In citrus and tomatojuices, clarification occurred above a certain minimum thresholdof PME activity.7−9 In pineapple juice, residual PME activitiesof 4.1 × 10−4 U/mL did not cause complete clarification,whereas PME values 10 times greater resulted in completeclarification.6

Commercial juice was used instead of raw, unpasteurizedjuice for consistency and relevancy to the industry. However,there are important differences between commercial, pasteur-

ized juice and raw juice that could affect the results. Preliminarywork showed that juice made in the laboratory and filteredthrough cheesecloth had particle sizes 50 times greater thanthose of commercial carrot juice, with much greater variation inparticle size. Additionally, our laboratory juice had 1.25 timesmore solid matter than commercial juice. Interestingly, whenthe carrots were blanched before juicing in the laboratory, theamount of solid matter became equal to that of the commercialjuice, which would suggest that the heating of the carrotweakens the internal structure and allows particles to be brokendown into smaller sizes and removed. This could account in

Figure 3. Particle size distribution curves for commercial carrot juice with no added PME (a) and with the following amounts of enzyme added: 0.15(b), 0.25 (c), 0.5 (d), and 1.0 U/g (e). Though particle sizes were measured every three days for 56 days, data for only days 1, 5, 9, 26, and 43 wereused to represent the size changes that occurred.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXE

part for the smaller particle sizes in pasteurized commercialcarrot juice versus raw laboratory juice, and is supported by thework of Reiter et al. who found that pasteurized juice hadsmaller particle diameters than unpasteurized juice.30 Process-ing techniques have a significant impact on cloud particle size; astudy by Reiter et al.30 showed the wide range of particle sizes(0.1−700 μm) for carrot juices produced with decanter

technology from five different manufacturers. The commercialjuice used in this study had <1% of particles >10 μm.28 Thisindicates there is a wide range of processing techniques used inthe juice industry that can result in different particle sizes.The changes in sediment height and relative turbidity

observed here were similar to one another in that bothindicated an initially rapid destabilization process, followed byslower changes over time. Both also showed small butsignificant differences between juices with the three highestamounts of PME added at the later stages of the storage timecourse. Changes in particle size showed a similar trend. Overthe first 5−10 days of storage, there was a relatively rapidincrease in diameter of the population of smallest particles inthe mixture. At longer times, sizes in the population of largerparticles grew slowly, with the rate of this growth showing amodest dependence on enzyme concentration especially at thehighest levels of added PME. By the end of the storage timecourse, all three measures of juice clarification, sediment height,relative turbidity, and particle size distribution, showed similarsmall differences between the juices with different levels ofadded PME. If PME induced clarification is due to enzymaticde-esterification of the juice pectins, then one might expect thatclarification would occur more rapidly with more added enzymebut that all the juices would eventually reach the same endpoint.Why the final states of the juices differed with differing

amounts of added enzyme is not clear. One possibility is thatthe de-esterification of pectin by PME was only partial and that

Figure 4. Comparison of particle size distribution of carrot juices with different amounts of added PME after 5 (a), 33 (b), and (c) 43 days.

Figure 5. Commercial carrot juice with added enzyme viewed with20× light microscopy at 56 days of storage at 5 °C. From left to right,enzyme levels are 0, 0.15, 0.25, 0.5, and 1.0 U/g.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXF

the final degree of pectin esterification achieved depended onthe amount of enzyme added. It is known that in the absence ofhigh salt concentrations, PME binds strongly to de-esterifiedstretches of pectin, resulting in the formation of inactive PME−pectin complexes.36 As a result, PME typically only partially de-esterifies a pectin chain before becoming inactive. It is possiblethat at higher levels of added PME, a higher final level of de-esterification is obtained because more pectin can be de-esterified before all the PME enzyme forms inactive PME−pectin complexes and de-esterification stops. The differences inparticle size, aggregation, flocculation, and sedimentationbetween samples with different levels of added PME may allresult from this difference in this final degree of esterification. Itis even possible that all the PME catalyzed de-esterification inthe juice occurred very rapidly, within the first day or two of theexperiment. In that case, not only the different extents but alsothe different time courses for the particle sizes andsedimentation over the next 56 days could be due to thedifferences in final pectin esterification between the differentlevels of added PME, rather than being from different rates ofde-esterification.A second possibility is that particle aggregation and

clarification is partially due to direct protein binding effects.Proteins (including PME) present in the high salt extract of cellwall material that was used as the source of the added PMEmay bind to the particulate matter in the juice, possiblyaffecting particle size aggregation. This is supported by theobservations of Schmelter et al. who found that even thermallyinactivated PME was able to cause pectin gelation in thepresence of salts.37 This ability of PME to create gels evenwhen inactivated was taken as an indication that, in addition toits catalytic mechanism, the PME protein could participate indirect electrostatic and hydrophobic interactions with thepectin. Proteins also have the capability of interacting with eachother in high acid environments (pH 4.4), and it has beenargued that it is protein−protein interactions that influencecloud stability more than protein-pectin interactions.30

Our micrograph data and the overall kinetics in the systemmay indicate that particle growth is in the diffusion limitedaggregation (DLA) regime. In such a regime, once a nucleationsite is formed, additional particles may be added one at a timevia the “hit and stick,” describing a rapid, irreversibleflocculation, without time for rearrangement.38 Loose, fractalflocs are the result of such a mechanism,35,39 consistent withwhat is observed in Figure 5. As more PME is added,flocculation occurs at a greater rate, creating even looser andtherefore larger flocs. Such flocs have a greater effectivediameter, and so sediment out more with centrifugation (Figure1), yet they do not pack as tightly on the bottom because oftheir loose structure (Figure 2). Initially, in juice with a highlevel of added PME, it may be that particles do not have todiffuse far to encounter another particle with which they areable to flocculate. High amounts of PME create more particlesof pectin that have a higher lower degree of esterification, andthe PME itself can participate in flocculation. This wouldaccount for the relatively rapid increase in particle size at thebeginning of storage.

■ AUTHOR INFORMATIONCorresponding Author*Telephone, (530)752-4800; e-mail, dmbarrett@ucdavis.edu.Present Address§A.K.S.: 500 Linden Street, Fort Collins, CO, 80524.

FundingThe authors would like to thank the National ScienceFoundation Center for Advanced Processing and Packagingfor their generous funding of this research.

NotesThe authors declare no competing financial interest.

■ ABBREVIATIONS

DLA, diffusion limited aggregation; PME, pectin methylester-ase; U, unit of enzyme activity

■ REFERENCES(1) Hirsch, A. R.; Foerch, K.; Neidhart, S.; Wolf, G.; Carle, R. Effectsof thermal treatments and storage on pectin methylesterase andperoxidase activity in freshly squeezed orange juice. J. Agric. FoodChem. 2008, 56 (14), 5691−5699.(2) Anthon, G. E.; Barrett, D. M. Kinetic parameters for the thermalinactivation of quality-related enzymes in carrots and potatoes. J. Agric.Food Chem. 2002, 50 (14), 4119−4125.(3) Suzuki, Y.; Sugimoto, A.; Kakuda, T.; Ikegawa, Y. Manufacturingprocess of carrot juice. United States Patent US 6340489, Jan. 22,2002.(4) Sampedro, F.; Geveke, D. J.; Fan, X.; Zhang, H. Q. Effect of PEF,HHP and thermal treatment on PME inactivation and volatilecompounds concentration of an orange juice-milk based beverage.Innovative Food Sci. Emerging Technol. 2009, 10 (4), 463−469.(5) Carbonell, J. V.; Contreras, P.; Carbonell, L.; Navarro, J. L. Pectinmethylesterase activity in juices from mandarins, oranges and hybrids.Eur. Food Res. Technol. 2006, 222 (1−2), 83−87.(6) Castaldo, D.; Laratta, B.; Loiudice, R.; Giovane, A.; Quagliuolo,L.; Servillo, L. Presence of residual pectin methylesterase activity inthermally stabilized industrial fruit preparations. LWTFood Sci.Technol. 1997, 30 (5), 479−484.(7) Rothschild, G.; Vliet, C.; Karsenty, A. Pasteurization conditionsfor juice and comminuted products of Isralie citrus fruits. J. Food Sci.Technol. 1975, 10, 29−38.(8) Fachin, D.; Van Loey, A. M.; Nguyen, B. L.; Verlent, I.; Indrawati;Hendrickx, M. E. Comparative study of the inactivation kinetics ofpectinmethylesterase in tomato juice and purified form. Biotechnol.Prog. 2002, 18 (4), 739−744.(9) De Sio, F.; Dipollina, G.; Villari, G.; Loiudice, R.; Laratta, B.;Castaldo, D. Thermal resistance of pectin methylesterase in tomatojuice. Food Chem. 1995, 52 (2), 135−138.(10) Jolie, R. P.; Duvetter, T.; Houben, K.; Clynen, E.; Sila, D. N.;Van Loey, A. M.; Hendrickx, M. E. Carrot pectin methylesterase andits inhibitor from kiwi fruit: Study of activity, stability and inhibition.Innovative Food Sci. Emerging Technol. 2009, 10 (4), 601−609.(11) Walter, R. H. The Chemistry and Technology of Pectin; AcademicPress: San Diego, CA, 1991; p 276.(12) Brummell, D. A.; Harpster, M. H. Cell wall metabolism in fruitsoftening and quality and its manipulation in transgenic plants. PlantMol. Biol. 2001, 47 (1−2), 311−340.(13) Anthon, G. E.; Sekine, Y.; Watanabe, N.; Barrett, D. M. Thermalinactivation of pectin methylesterase, polygalacturonase, and perox-idase in tomato juice. J. Agric. Food Chem. 2002, 50 (21), 6153−6159.(14) Duvetter, T.; Sila, D. N.; Van Buggenhout, S.; Jolie, R.; VanLoey, A.; Hendrickx, M. Pectins in processed fruit and vegetables: PartIStability and catalytic activity of pectinases. Compr. Rev. Food Sci.Food Saf. 2009, 8 (2), 75−85.(15) van Alebeek, G.-J. W. M.; van Scherpenzeel, K.; Beldman, G.;Schols, H. A.; Voragen, A. G. J. Partially esterified oligogalacturonidesare the preferred substrates for pectin methylesterase of Aspergillusniger. Biochem. J. 2003, 372 (1), 211−218.(16) Catoire, L.; Pierron, M.; Morvan, C.; du Penhoat, C. H.;Goldberg, R. Investigation of the action patterns of pectinmethylester-ase isoforms through kinetic analyses and NMR spectroscopy

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXG

Implications in cell wall expansion. J. Biol. Chem. 1998, 273 (50),33150−33156.(17) Wicker, L.; Ackerley, J. L.; Hunter, J. L. Modification of pectinby pectinmethylesterase and the role in stability of juice beverages.Food Hydrocolloids 2003, 17 (6), 809−814.(18) Benen, J. A. E.; Alebeek, H.-J. W. M. V.; Voragen, A. G. J.;Visser, J. Handbook of Food Enzymology; Marcel Dekker, Inc.: NewYork, NY, 2003; Vol. 122.(19) Benítez, E. I.; Genovese, D. B.; Lozano, J. E. Effect of pH andionic strength on apple juice turbidity: Application of the extendedDLVO theory. Food Hydrocolloids 2007, 21 (1), 100−109.(20) Castaldo, D.; Servillo, L.; Loiudice, R.; Balesterieri, C.; Laratta,B.; Quagliuolo, L.; Giovane, A. The detection of residual pectinmethylesterase activity in pasteurized tomato juices. Int. J. Food Sci.Technol. 1996, 31 (4), 313−318.(21) Mensah-Wilson, M.; Reiter, M.; Bail, R.; Neidhart, S.; Carle, R.Cloud stabilizing potential of pectin on pulp-containing fruitbeverages. Fruit Process. 2000, 2, 47−54.(22) Mollov, P.; Mihalev, K.; Buleva, M.; Petkanchin, I. Cloudstability of apple juices in relation to their particle charge propertiesstudied by electro-optics. Food Res. Int. 2006, 39 (5), 519−524.(23) Cameron, R. G.; Baker, R. A.; Grohmann, K. Multiple Forms ofPectinmethylesterase from Citrus Peel and Their Effects on JuiceCloud Stability. J. Food Sci. 1998, 63 (2), 253−256.(24) Ly-Nguyen, B.; Loey, A. M. V.; Fachin, D.; Verlent, I.;Indrawati; Hendrickx, M. E. Partial purification, characterization, andthermal and high-pressure inactivation of pectin methylesterase fromcarrots (Daucus carrota L.). J. Agric. Food Chem. 2002, 50 (19), 5437−5444.(25) RC; JDS; MAM. Bradford Reagent Technical Bulletin. Sigma-Aldrich, Ed., 2011.(26) Anthon, G. E.; Barrett, D. M. Comparison of three colorimetricreagents in the determination of methanol with alcohol oxidase.Application to the assay of pectin methylesterase. J. Agric. Food Chem.2004, 52 (12), 3749−3753.(27) Silva, V. M.; Kawazoe Sato, A. C.; Barbosa, G.; Dacanal, G.;Ciro-Velasquez, H. J.; Cunha, R. L. The effect of homogenisation onthe stability of pineapple pulp. Int. J. Food Sci. Technol. 2010, 45 (10),2127−2133.(28) Reiter, M.; Neidhart, S.; Carle, R. Sedimentation behaviour andturbidity of carrot juices in relation to the characteristics of their cloudparticles. J. Sci. Food Agric. 2003, 83 (8), 745−751.(29) Ibrahim, G. E.; Hassan, I. M.; Abd-Elrashid, A. M.; El-Massry, K.F.; Eh-Ghorab, A. H.; Manal, M. R.; Osman, F. Effect of cloudingagents on the quality of apple juice during storage. Food Hydrocolloids2011, 25 (1), 91−97.(30) Reiter, M.; Stuparic, M.; Neidhart, S.; Carle, R. The role ofprocess technology in carrot juice cloud stability. Lebensm.-Wiss.Technol. 2003, 36 (2), 165−172.(31) Harrison, D.; Fisch, M. Turbidity Measurement [Online]; CRCPress LLC: Boca Raton, FA, 2000. http://www.engnetbase.com.(32) Croak, S.; Corredig, M. The role of pectin in orange juicestabilization: Effect of pectin methylesterase and pectinase activity onthe size of cloud particles. Food Hydrocolloids 2006, 20 (7), 961−965.(33) Ellerbee, L.; Wicker, L. Calcium and pH influence on orangejuice cloud stability. J. Sci. Food .. Agric. 2011, 91 (1), 171−177.(34) Wicker, L.; Ackerley, J. L.; Corredig, M. Clarification of juice bythermolabile valencia pectinmethylesterase is accelerated by cations. J.Agric. Food Chem. 2002, 50 (14), 4091−4095.(35) Benitez, E. I.; Lozano, J. E.; Genovese, D. B. Fractal dimensionand mechanism of aggregation of apple juice particles. Food Sci.Technol. Int. 2010, 16 (2), 179−186.(36) Lineweaver, H.; Ballou, G. A. The effects of cations on theactivity of alfalfa pectinesterase (pectase). Arch. Biochem. 1945, 6,373−387.(37) Schmelter, T.; Vreeker, R.; Klaffke, W. Characterisation of anovel gel system containing pectin, heat inactivated pectinmethylesterase and NaCl. Carbohydr. Polym. 2001, 45 (3), 277−284.

(38) Russel, W. B.; Saville, D. A.; Schowalter, W. R. ColloidalDispersions; Cambridge University Press: Cambridge, MA, 1989.(39) Beveridge, T. Letter to the Editor: Fractile images and applejuice haze. Food Res. Int. 1998, 31 (5), 411−414.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4043979 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXH