Author's personal copy

13 Pages·2012·0.53 MB·English

Author's personal copy
FoodResearchInternational46(2012)438–450
ContentslistsavailableatScienceDirect
Food Research International
journal homepage: www.elsevier.com/locate/foodres
Review
Carotenoids bioavailability from foods: From plant pigments to efficient
biological activities
Elisabet Fernández-García, Irene Carvajal-Lérida, Manuel Jarén-Galán, Juan Garrido-Fernández,
Antonio Pérez-Gálvez 1, Dámaso Hornero-Méndez 1,⁎
FoodBiotechnologyDepartment,InstitutodelaGrasa(CSIC),Av.PadreGarcíaTejero,4,41012Sevilla,Spain
a r t i c l e i n f o a b s t r a c t
Articlehistory: Carotenoid pigments are a group of bioactive compounds that are of interest to the food scientists,
Received28January2011 nutritionistsandfoodindustriesduetotheirpositiveimpactonhumanhealthandtheireconomicbenefits.
Accepted3June2011 Carotenoidsareresponsiblefortheattractivecolourofmanyplantfood(mainlyfruitandvegetables),which
is perhaps the first attribute that consumers assess when determining the quality and appearance of a
Keywords:
product,andthereforeconditionsitsacceptability.Inaddition,carotenoidshavediversebiologicalfunctions
Bioactivecompounds
andactivities,suchasthewellknownprovitaminAactivity,antioxidantcapacityandenhancementofthe
Bioaccessibility
Bioefficacy immune system. There are an extensive number of factors affecting the efficient incorporation of these
Carotenoids phytochemicalsfromthediet,althoughinmanycasesnobiologicalactivitywillbeputinactionwithinthe
Digestibility/absorptionefficiency consumer body (animal or human) without a first visual attraction. The term bioaccessibility is used to
ProvitaminA evaluate the amount of a nutrient that is released from a food during the digestion process. The
bioaccessibilityoflipophiliccompounds,suchascarotenoids,innaturalfoods(mainlyfruitsandvegetables)
isusuallyfairlylowandisconstrainedbyvariousfactors,particularlythedegreeoffoodprocessingandmatrix
composition.Thereareevidencesthathomogenisationandthermaltreatmenthavepositiveeffectsonthe
bioaccessibilityofthesecompounds,whereasthepresenceofdietaryfibrehasanegativeeffect.Thepresence
andco-ingestionoffatinthedietareakeyfactor,withaminimumquantityneededtofacilitatecarotenoid
absorption,andthisseemstobeoneoftheadvantagesoftheMediterraneandiet.Mostoftherelevantdataon
the bioaccessibility of carotenoids from natural or processed foods has been obtained in postprandial
absorptionstudiesandsupplementationstudies.Thisapproach,althoughhighlyvaluable,isinsufficientfora
detailedanalysisofthefoodmatrixcompositioneffects,andalsoitdoesnottakeintoconsiderationother
factors that may be involved in carotenoid absorption in each stage of the bioaccessibility process
(digestibilityandabsorption).Invitroexperimentalprocessesthatreproducethephysiologicalconditionsand
eventsthattakeplaceinthehumangastrointestinaltractduringdigestionhavebeendevelopedandfine-
tunedinrecentyears.Thesedigestionmodelsbecomeanexcellentanalyticalresourcetoestablishboththe
significanceandscopeofdiversefactorsintheefficiencyofdigestibilityofcarotenoidsallowingadetailed
analysisoftheinfluenceofthefoodmatrixcompositiononthedigestiveprocess.
©2011ElsevierLtd.Allrightsreserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
2. Carotenoidcontentinfood:TheMediterraneandiet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
3. CarotenoidsandprovitaminAactivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
3.1. MetabolismofcarotenoidswithprovitaminAactivity.Bioefficiency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
4. Carotenoidbioavailability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
4.1. Digestibility:Solubilisationandmicellisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
4.2. Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
Abbreviations:BBMV,brush-bordermembranevesicles;HDL,high-densitylipoproteins;LDL,low-densitylipoproteins;O/W,oilinwateremulsions;VLDL,verylow-density
lipoproteins.
⁎ Correspondingauthor.Tel.:[phone];fax:[phone].
E-mailaddress:[email protected](D.Hornero-Méndez).
1 Bothauthorscontributedequallytothiswork.
0963-9969/$–seefrontmatter©2011ElsevierLtd.Allrightsreserved.
doi:10.1016/j.foodres.2011.06.007
Author's personal copy
E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450 439
5. Carotenoidsbioaccessibilityfromhydrophilicmatrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
6. Invitrodigestionmodels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
7. Invitroabsorptionmodels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
1.Introduction wouldbeinteresting,andpotentiallyuseful,toincludecarotenoidsin
processed foods or new food matrices in order to increase their
The colour of food is perhaps the first attribute that consumers bioavailability. Food formulations that provide one alternative to
assesswhendeterminingthequalityandappearanceofaproduct,and natural foods, in terms of carotenoid contribution, are based on
therefore conditions its acceptability. Colour becomes a measure of hydrophilicmatricesthathavebeenenrichedwiththesecompounds
quality and also an indication of deterioration. In fruits and throughdispersion,emulsionorencapsulationwithothercompounds
vegetables, colour is due mainly to three pigment families: chloro- to modify fat-solubility if necessary (Pérez-Gálvez & Mínguez-
phylls,carotenoidsandanthocyanins,whichareresponsibleforgreen, Mosquera,2004;Richelle,2002).Thesetypesofformulationspresent
red-yellow, and blue-violet colourations, respectively. The first severaladvantagestotheconsumer,includingtheabilitytoselecta
function of these pigments that can be observed in plants is the combination of various functional compounds at adequate concen-
attractionofanimalstoactasvectorsforthedisseminationofseeds trations for their activity, and the application of additional design
andfruits,thusensuringreproductivesuccessandperpetuationofthe criteria that could further increase assimilation of the said
species (Bartley & Scolnik, 1995). Certainly the acquisition of these compounds.
strikingcolourshasbeenselectedforthroughoutevolutionandhas Whenclaimsaremaderegardingphytochemicals(theirpresence
helped to establish other characteristics among which are the infoodandthehealthbenefitthattheyprovidebystrengtheningthe
bioactiveproperties.Humansarenotindifferenttothisphenomenon poolofantioxidantsinourbodies),theassessmentoftruthofthese
ofattraction(Mínguez-Mosquera,Pérez-Gálvez,&Hornero-Méndez, claimsisnotonlyamatterofcompositionbutratherwhetherthose
2005). componentsareeffectivefortheclaimedfunction.Then,ithastobe
Alargebodyofepidemiologicalevidenceshasdemonstratedthat shownhowthesecompoundsareingested,digested,andeffectively
eating fruits and vegetables may reduce the risk of several assimilated,passingthepre-systemicmetabolismtoreachthetarget
degenerative processes (Fung, Willett, Stampfer, Manson, & Hu, tissuewheretheycarryouttheirfunction,stagesthatcanbeverified
2001; Gramenzi et al., 1990; Hertog, Feskens, Hollman, Katan, & through a bioavailability study. This fact is summarised in the
Kromhout, 1993; Hirvonen et al., 2001; Joshipura et al., 2001; Key, Regulation EC 1924/2006 on nutrition and health claims made on
Thorogood,Appleby,&Burr,1996;Knektetal.,1994;Law&Morris, foods. Specifically in section 15 it is established that “in order to
1998;Sasazuki,2001).Increasedconsumptionoffruitsandvegetables ensure that the claims made are truthful, it is necessary that the
leadstoanincreasedintakeofwater-andfat-solublevitamins,dietary substancethatisthesubjectoftheclaimispresentinthefinalproduct
fibreandotherphytochemicals(suchascarotenoids,glucosinolates, in quantities that are sufficient, or that the substance is absent or
phenols and chlorophylls), which have beneficial effects on health presentinsuitablyreducedquantities,toproducethenutritionalor
(Liu, 2004; Lock, Pomerleau, Causer, Altman, & McKee, 2005). This physiologicaleffectclaimed.Thesubstanceshouldalsobeavailableto
claimissupportedbystudiesthatrefertothebeneficialeffectofsome beusedbythebody.Inaddition,andwhereappropriate,asignificant
ofthosecompounds(Aggarwaletal.,2004;deJong,Plat,&Mensink, amount of the substance producing the claimed nutritional or
2003; Lambert & Yang, 2003; Verhoeven, Goldbohm, van Poppel, physiological effect should be provided by a quantity of the food
Verhagen,&vandenBrandt,1996;Ziegler,1989). thatcanreasonablybeexpectedtobeconsumed”.
Promotional campaigns have raised consumer awareness that Thesestatementsapplyparticularlyforcarotenoids.Bioavailability
healthyeatinghabitsarenowrecognisedtoimproveorincreaselife ofcarotenoidsfromdietarysourcesdependsonmultiplefactorssome
expectancy,thenpositivelyimpactingtheconsumptionoffruitsand ofthemrelatedtofoodmatrixcharacteristicsandthepropertiesof
vegetables. Encouraging consumption of these foods has steadily anyco-ingestedfood,otheronesrelatedtonutrientstatusandgenetic
increasedfromameanof2to3portionsadayinthe1970stothe profileofthehost(vanhetHof,West,Weststrate,&Hautvast,2000).
currentmeanof5.Therecommendationofincreasingourconsump- These factors condition the efficiency of carotenoid bioavailability
tion of fruits and vegetables also suggests that our intake of fromdifferentfoodsandconsequentlytheexertionofbenefitsthey
phytochemicals should increase, and one of the driving reasons coulddevelopinhumans.
could be the data from studies concerning their bioavailability. The
amount of phytochemicals that are efficiently absorbed from raw 2.Carotenoidcontentinfood:TheMediterraneandiet
fruitsandvegetablesislowincomparisonwithprocessedproducts
andsupplements.Thus,only25%ofthevegetablesconsumedprovide Although more than 700 carotenoids have been described in
enoughquantitiesofphytochemicalstoobtainfavourableeffects,and Nature,notallnaturalsourcesofthemarepresentinournormaldiet.
this proportion increases up to 50% in the case of fruits (Johnston, Itisestimatedthatweonlyhaveaccesstoabout40carotenoidsthat
Taylor,&Hampl,2000). canbeabsorbed,metabolised,and/orusedinourbodies.Thatnumber
In particular, an interest in increasing the consumption of is reducedto6 if weconsider thecarotenoidprofile that is usually
carotenoids has been evident since 1981, when Peto, Doll, Buckely, detected in human blood plasma. This group includes α- and β-
and Sporn (1981) suggested that β-carotene consumption reduces carotene,lycopene,β-cryptoxanthin,zeaxanthinandlutein,whichare
the incidence of some types of cancer, and further evidences were regularly present in the foods listed in Table 1 (for structures see
obtained in subsequent studies (Ziegler, Maine, & Swanson, 1996). Fig.1)(Britton&Khachik,2009).Thecarotenoidcontentofthefoods
However, the percentage of carotenoid ingested versus the amount listedinthisandothertablescanbefoundindatabasesthathavebeen
assimilated, reaches up to 10% when a natural food is consumed developedforthispurpose.ThedatabaseofMangels,Holden,Beecher,
(Boileau,Moore,&Erdman,1999).Inlightofthislowpercentage,it Forman, and Lanza (1993) only included fruits and vegetables as
Author's personal copy
440 E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450
Table1
CarotenoidcontentinfoodsrepresentativeoftheMediterraneandiet,expressedin
mg/100g.
Lycopene
α-Carotene β-Carotene β-Cryptoxanthin Luteinor Lycopene
Zeaxanthin
Greenvegetables
Lettuce – 1272 – 2635 –
Spinach – 5597 – 11938 –
-Carotene
Brussels 6 450 – 1590 –
sprouts
Vegetables/tubers
Beans 147 408 – – –
Broccoli 1 779 – 2 –
Pepper 59 2379 2205 – – -Carotene
Pumpkin 4795 6940 – – –
Potato – 6 – – –
Tomato 112 393 – 130 3025
Carrot 4649 8836 – – –
Onion 6 – – – – HO
Fruits -Cryptoxanthin
Pineapple 30 – – – – OH
Banana 5 21 – – –
Grape 5 603 12 13
Mango 17 445 11 – –
Melon 27 1595 – 40 – HO
Orange 16 51 122 187 – Zeaxanthin
Watermelon – 295 103 17 4868 OH
Pear 6 27 – – –
Cereals
Corn 33 30 – 884 –
Wheat – 100 – 35 – HO
Lutein
Vegetableoils
Olive – 219 30 5990 –
Palm 24 38 – – – Fig. 1. Chemical structures of the carotenoids most frequently found in fruits,
vegetablesandhumanbloodplasma.
carotenoid-containingfoods.Lateron,anewdatabasewasdeveloped, Canada (1997). However, the study published by the WCRF/AICR
publishedbyHoldenetal.(1999),thatalsoincludedotherfoodssuch (1997)raisestheaverageintakevalueto9–18mg/day.Forinterven-
as vegetable oils, butter, eggs, cheese, and other products made of tionstudiesconductedwithcontrolleddietarycarotenoidcontent,it
vegetables (pizzas, salads, etc.). This database evaluated up to 200 issuggestedthatanintakeof3to6mg/dayofcarotenoidsissufficient
references on the carotenoid content of 215 foods, tabulating the tomaintainplasmalevelsofthesecomponents(Micozzietal.,1992;
average content and standard deviation, as well as the number of Yongetal.,1994;Zino,Skeaff,Williams,&Mann,1997).
studiesconductedforeachfood.Itisavailableatthefollowingweb In particular, the Mediterranean diet offers perhaps the most
site: www.nal.usda.gov/fnic/foodcomp (last accessed 20/01/2011). diversityandamountofcarotenoidintakeduetoitshighcontentof
One of the main uses of this database is the estimation of the fruits and vegetables (fresh and/or processed) and vegetable oils
provitaminAcontributionofagivendietaryintake.Anotherdatabase (especiallyoliveoil)(Saura-Calixto&Goñi,2009).Bearinginmindthe
ontheaveragecontentofcarotenoidsinfoodsisofferedbytheU.S. 6mostrepresentativecarotenoidsmentionedabove,Table1shows
Department of Agriculture, Agricultural Research Service(2005). In the amounts found in foods of the Mediterranean diet. All green
thelistofnutrientsincludedinthisdatabase(www.nal.usda.gov/fnic/ vegetablescontainaconsiderableamountoflutein,β-carotene,and
foodcomp/Data/SR18/nutrlist/sr18list.html, last accessed 20/01/2011) β-cryptoxanthin, with the concentration varying greatly from one
thereisinformationonα-andβ-carotene,lycopene,β-cryptoxanthin, source to another. The best sources of α-carotene are carrots and
zeaxanthinandlutein. pumpkins, while β-carotene is found more widely in fruits and
Despite the correlation between high carotenoid content in vegetables such as carrots, red bell peppers, oranges, potatoes,
plasma, which comes exclusively from the intake of foods rich in broccoli,andgreenvegetables.β-Cryptoxanthinisfrequentlyfound
carotenoids, and lower risk of developing severe degenerative atlowconcentration,althoughinriperedpeppersandsometropical
processes, adequate intake levels of these components have not fruitslikepapayaisoneofthemajorpigments.Tomatoanditsderived
beenestablishedsincethepositivehealtheffectsmaybeduetoother products (pasta and sauces), together with watermelon and pink
constituents that are co-ingested with carotenoids. Neither the grapefruit, are the main sources of lycopene. Rich sources of lutein
health-promoting biological actions that these compounds may includegreenvegetablessuchasspinach,Brusselssprouts,broccoli,
have in our bodies (antioxidant capacity, immune enhancement, and peas, while zeaxanthin is found in high concentrations in egg
increased intercellular communication) nor the fact that some of yolksandcorn.
themexhibitprovitaminAactivityhasbeen,atthemoment,reasons
to establish a recommendation on the appropriate amount of 3.CarotenoidsandprovitaminAactivity
carotenoid intake. However, by using data from epidemiological
studiesontheconsumptionoffruitsandvegetablesandtheireffecton Inanimals,carotenoidpigmentshaveseveralimportantbiological
health,normalvaluesmaybesetforcarotenoidintake,whichmaybe activitiesfromnutritionalandphysiologicalstandpoints.Animalsand
associated with a lower risk of developing degenerative diseases humans cannot synthesise carotenoids de novo although they can
(cancer,cardiovasculardisease,etc.).Evenso,therearediscrepancies metabolisesomeofthemintovitaminA(retinol).Approximately10%
onmeanintakevaluesintheconsultedreferences.Thus,individuals of carotenoids meet the main structural requirement for acting as
whoeatadietrichinthesefoodsingestabout6mg/dayaccordingto vitaminAprecursors,i.e.,containaβ-typenon-substitutedring,being
studies by Lachance (1997) and published guidelines from Health β-carotene and β-cryptoxanthin the most representatives. Table 2
Author's personal copy
E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450 441
showstheprovitaminAactivity(relativetoβ-carotene)ofthemain since every molecule of pigment produces two of retinal, which is
carotenoids present in the diet. The only source of these retinol thenreducedtovitaminA(retinol).
precursorsisthedietand,inmostcases,fruitsandvegetablesarethe The metabolic process for conversion of carotenoids with
food sources in our diet that primarily provide carotenoids with provitamin A activity into retinol occurs via an oxidation process
provitamin A activity. The extensive presence and distribution of with central cleavage of the molecule, which is catalysed by the
carotenoids in Nature, mainly in fruits and vegetables (foods that enzyme β,β-carotene 15,15′-mono-oxygenase (EC 1.13.11.21). This
occupy or should occupy an important place in our diet), make enzyme,originallycalledβ,β-carotene-15,15′-dioxygenase,although
carotenoidswithprovitaminAactivitythemostimportantsourceof discovered 45years ago, (Goodman & Huang, 1965), had not been
retinol.Somegroupsofpeople,thevegetarians,evendependalmost purified until recently through the successful cloning of cDNAs in
exclusivelyonfruitsandvegetablesasasourceofretinolintheform different species, including humans (Yan et al., 2001). That key
of its precursors.In mammals,therefore, theunique and important processisfollowedbytheconversionofretinalintoretinylestersby
biologicalfunctionofcarotenoidswithretinolequivalenceistheirrole retinalreductaseandlecithin-retinolacyltransferase(Moore,1930).
as vitamin A precursors, which is necessary for vision, growth, cell Theproposedmechanism,firstlydescribedbyGoodman,Huang,and
differentiation,andotherphysiologicalprocesses(Olson,1996). Shiratori(1966),consistsofoxidationofthetwocentralcarbonatoms
Data published in the study “Global prevalence of vitamin A of β-carotene with molecular oxygen and subsequent homolytic
deficiencyinpopulationsatrisk1995–2005”publishedbytheWorld cleavage of the dioxyethane intermediate at the C15–C15′ linkage,
HealthOrganisationin2009,indicatethat190millionpreschool-age generatingtwomoleculesofretinal(Fig.2A).However,Leuenberger,
childrenand19.1millionpregnantwomenhadlevelsofserumretinol Engeloch-Jarret, and Woggon (2001) showed that the enzyme
lessthan0.7μmol/L,whichisthelowerlimitofnormal,andbelow involved in the retinal transformation is really a mono-oxygenase
which is considered a state of vitamin A deficiency. The deficient and that the mechanism involves epoxide formation at the 15–15′
populationisdistributedincountrieswhosegrossdomesticproduct bondandthesubsequentnon-selectiveringopeningwithwaterand
(GDP) is less than US$15,000 and in those with 92% of the world's cleavagetogeneratethecorrespondingaldehydes.Theexperimental
population. A priori, this situation may seem paradoxical since the strategyusedinthisworkiswellknownandconsistsofcleavingan
majorityofthatpopulationbasesitsdietonplant-derivedfoods,and asymmetric carotenoid substrate (α-carotene) and its enzymatic
thereforerepresentsthemaincontributionofprovitaminA.However, incubationinamediumenrichedin17O andH18O.Thepresenceof
2 2
and in light of thedata,that contributionis not enoughdue tothe 17O and 18O in the aldehyde obtained at the end of the enzymatic
equivalence or efficiency of the conversion of provitamin A into processrevealedthepreviouslydiscussedmechanism(Fig.2B).
retinol,asitwillbediscussedlater.Inordertoreachalevelhigher Thevalidityofonecleavagepathoranother,centraloreccentric,
than that which is considered deficient, it is necessary to consume has been the subject of profound debate for decades. Although
foods of animal origin that have been fortified. The Institute of evidenceexistedfortheoperationoftheeccentriccleavagepath,the
MedicineoftheNationalAcademyofSciences(USA)hasestablished alternativerouteofcentralcleavagewasacceptedfordecadesuntil
recommendeddailyallowancelevelsforthepopulationaccordingto thepublicationoftheworkofWang,Tang,Fox,Krinsky,andRussell
age,withanadditionaldistinctionforpregnantandlactatingwomen. (1991) that demonstrated the production of apocarotenals and
Foradultsbetween31and50yearsofage,forexample,theRDAvalue retinoicacidasthemainproductsofthebreakdownofβ-carotene.
is900mg/dayformenand700mg/dayforwomen. Both pathways can be considered to occur during the conversion
processofβ-carotenetoretinalwithapredominanceofoneorthe
otherdependingonthetypeoftissuewheretheprocessiscarriedout.
3.1.MetabolismofcarotenoidswithprovitaminAactivity.Bioefficiency Alloftheseevidencessupporttheexistenceofafamilyofoxygenase
enzymes that produce central or eccentric cleavage of carotenoids
As stated before, not all carotenoids have the structural re- withprovitaminAactivity,givingrisetometaboliteswithdifferent
quirementsforconversionintovitaminA.Onlythosewithatleastone structuresdependingonthelocationofepoxidationandcleavageof
β-type ring, without oxygenated functional groups, along with one thepolyenechain(Fig.3).Thisgroupofenzymesislocalisedinthe
polyene chain containing at least 11 carbon atoms are potential cytosolofcellsand,inmammalsitsactivityismostlylocatedinthe
precursorsofvitaminA.Amongthe700naturalcarotenoidsdescribed intestine.Infact,inhumansitisbelievedthat60–70%ofingestedβ-
uptonow,only10%showprovitaminAactivity.Themostimportant caroteneisconvertedtoretinal.Activityhasalsobeendetectedinthis
ones,asmuchfortheirhighactivitylevelasfortheiravailability,are familyofenzymesinothertissuessuchastheliver,lungs,kidneys,
α-andβ-carotene,andsomexanthophyllsincludingβ-cryptoxanthin andbrain.Thecharacteristicsoftheenzymaticprocessesinvolvedin
and some apocarotenoids (Mínguez-Mosquera & Hornero-Méndez, the conversion of carotenoids into retinol esters indicate that the
1997). Of these, β-carotene has the greatest provitamin A activity processesareregulatedwithinlipidandretinoidmetabolism.Factors
related to the individual (vitamin A status, malnutrition, intestinal
infection/inflammation) and those related to the food matrix
TCaabrolete2noidswithprovitaminAactivity.Provitaminvalueisrelativetoβ- (presence of other carotenoids, fat and dietary fibre, processed/
caroteneaccordingtoBauernfeind(1972). cooked) that affect carotenoid bioavailability, all determine the
bioefficiencylevel(Castenmiller&West,1998).
Carotenoid Percentactivity Bioefficiencyisdefinedasthefractionofingestedcarotenoidswith
trans-β-Carotene 100 provitaminAactivitythatisabsorbedandconvertedtoitsactiveform,
9-cis-β-Carotene 38
that is, retinol. In 1967 the WHO established 6 as the ratio value to
13-cis-β-Carotene 53
trans-α-Carotene 53 calculateβ-carotenebioefficacy.Inthecaseofothercarotenoidswith
9-cis-α-Carotene 13 provitaminAactivity,theratiowasdoubled.Thus,1mgofretinolwas
13-cis-α-Carotene 16 obtained from 6mg of β-carotene or 12mg of other proactive
trans-β-Cryptoxanthin 57 carotenoids.Theestablishmentoftheseparameterswasbasedondata
9-cis-β-Cryptoxanthin 27 showingthattheequivalencebetweenpureβ-caroteneinoilandretinol
15-cis-β-Cryptoxanthin 42
β-Carotene-5,6-epoxide 21 is2:1.Estimatingthatthebioavailabilityofcarotenoidsfromthediet
Mutatochrome 50 wasonethirdofthatofretinol,theratioof6wassetforβ-carotene
γ-Carotene 42–50 (3×2:1), and 12 for other carotenoids with retinol equivalence
β-Zeacarotene 20–40
(2×3×2:1). However, it was recognised that these ratios could
Author's personal copy
442 E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450
A
O2 O2
(1) (1) β,β-carotene9’,10’monoxygenase
, -carotene 15,15’ dioxygenase
O
O
O
(2)
(2)
2 H O
2
O OH
(3) OH
(3)
B
O
O2
(1) O
, -carotene 15,15’ monoxygenase (4)
(5)
β-oxidation
mechanism
O
(4) O
H2O
HO (6)
Fig.3.Reactionofβ-carotene(1)toretinal(6)viamonoxygenaseandmonoxyethane
OH (2), diol (3), β-ionone (4) and β-apo-10-carotenal (5) intermediates. Enzymatic
activityisperformedoveraneccentricpositionofβ-carotene.
(5)
2
meetthedailyvitaminArequirementishigherthanthatwhichhad
O been established for years, which had overestimated the intake of
vitaminAinthediet.Table3showstheunitsandconversionfactors
for International Units (IU), Retinol Equivalent (RE) and Retinol
(3)
ActivityEquivalent(RAE).ThislastunithasemergedtoreplacetheRE
Fig.2.Reactionofβ-carotene(1)toretinal(3)viadioxygenaseenzymeanddioxyethane andcorrecttheoverestimationofbioefficacy.Alltheseaspectshave
intermediate(2)(panelA)orviamonoxygenaseenzymeandmonoxyethane(4)and recentlybeenreviewedbyThurnham(2007).
diol(5)intermediates(panelB).Enzymaticactivityisperformedovercentralpositionof
β-carotene.
4.Carotenoidbioavailability
overestimateorunderestimatethebioavailabilityofdietarycarotenoids Sincehumanscannotsynthesisecarotenoidsandmusttherefore
whenconsideringotherfactorsthataltertheefficiencyofbioavailability, depend on diet to supply these compounds and their beneficial
such as carotenoid content in the food, type of food, degree of effects,thebioavailabilityofcarotenoidshasalwaysbeenofinterest.
processing,andsoon.Laterstudies(vanhetHofetal.,1999)conducted The concept of bioavailability comes from a fusion of the words
with stable isotopes have confirmed the fact that they had over- biologicalandavailability(Metzler&Huang,1983),anditsdefinition
estimatedtheefficiencyofbioavailability,whichisamorecrucialfactor hasevolvedwithtime.Thus,bioavailabilityincludesaccessibilityfor
thantheconversiontoretinol,indicatingthattheefficiencyofdietaryβ- absorption,absorption,metabolism,transportandtissuedistribution,
carotene absorption was one seventh of that corresponding to β- andbioactivity.Howeverthetermbioavailabilityisusuallydefinedas
caroteneadministeredinoil,thusleadingtoratiosof14(7×2:1)and28 thefractionofanoraldoseofaparentcompoundoractivemetabolite
(2×7×2:1)forβ-caroteneandothercarotenoidswithretinolequiv- that reaches the systemic circulation (Schumann et al., 1997), a
alence,respectively.TheInstituteofMedicineoftheNationalAcademy definition more related with the term bioaccessibility. In general,
of Sciences (USA) report from 2001 reduced the ratios to 12 for β- bioaccessibility can be described as the sum of digestibility and
caroteneand24forothercarotenoidswithretinolequivalence(Scott& assimilation. Although bioavailability and bioaccessibility are often
Rodriguez-Amaya,2000). used indistinctly, it is important to note that definition of bioavail-
The abovementioned change in ratios between β-carotene and ability includes metabolism, transport and tissue distribution, and
carotenoidswithprovitaminAactivityobtainedfromtheconsump- bioactivity.However,bioactivitygenerallyfadesintotheconceptof
tion of fruits and vegetables, prompted by bioavailability studies, bioavailabilitygiventhat,pharmacologicallyspeaking,oncetheactive
indicatesthattheamountofintakeofthesefoodsthatisrequiredto ingredient has reached its target organ, it is not inert, but rather
Author's personal copy
E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450 443
Table3 carotenoidsarelocated,thusimprovingtheirreleaseandsolubilisa-
UnitsusedtoexpressthevitaminAcontentinfoodsandtheequivalenceswithretinol, tion (Gärtner, Stahl, & Sies, 1997; Paetau et al., 1998; Stahl & Sies,
β-caroteneandothercarotenoidswithprovitaminAactivity.
1992).
μgretinol μgβ-carotene μgcarotenoida The micellisation phase involves reduction of the size of lipid
IUretinolb 0.3 3.6 7.2 particles produced in the previous phase to form molecular
IUβ-caroteneb – 0.6 1.2 aggregatesof3–10nm(micelles)inwhichthelipidmaterialwillbe
RE 1 6c 12 stored (Johnson & Gerwin, 2001). Micelles may not only contain
RAEd(μg) 1 12 24 carotenoids, but also acylglycerols, cholesterol and phospholipids.
a CarotenoidswithprovitaminAactivitywitha50%conversionvalue(seeTable2) Secretionsfromthegallbladderandpancreasarecriticalinthisphase,
withrespecttoβ-carotene(seeTable3). asthefirstsecretionsconsistmainlyofbiliarysaltsthatcontributeto
b 1IU(InternationalUnits)retinol=3IUβ-carotene.
reduce micelle size and stabilise these aggregates. These secretions
c Equivalencywithrespecttoβ-carotenefromdiet.1RE(RetinolEquivalent)=2μg
ofpureβ-caroteneinoil.Analternativeequivalencyratiocouldbe1RE=14μgβ- provide digestive enzymes such as the pancreatic lipases that
caroteneor28μgtootherprovitaminAcarotenoids. hydrolyse triacylglycerides, cholesterol esters and other esterified
d 1RAE(RetinolActivityEquivalent)=3.33IUretinol. compounds, such as carotenoids from fruits. The key factor in this
phase is the ingested fatty matter. Its presence stimulates biliary
carries out some action (Stahl et al., 2002). Digestibility refers secretions and pancreatic lipase levels, which in turn increases
specificallytothefractionoffoodcomponentsthataretransformed micellisation capacity. It is estimated that the consumption of 3 to
bydigestionintopotentiallyaccessiblematterthroughallphysical– 5g of fat notably increases the bioavailability of carotenoids
chemicalprocessesthattakeplaceinthelumen.Assimilationrefersto (Jayarajan, Reddy, & Mohanran, 1980; Roels, Trout, & Dujacquier,
theuptakeofbioaccessiblematerialthroughtheepitheliumbysome 1958).However,somecomparativestudiesindicatethatcertaintypes
mechanismoftransepithelialabsorption. offatsmaydisfavourthisprocess.Thebioaccessibilityofcarotenoids
Inthefieldofcarotenoidssomeworksthatmeasuredigestibility decreases when short/medium-chain triacylglycerides are ingested,
stage refer this term as bioaccessibility while others equate whilebioaccessibilityappearstobeincreasedbytheconsumptionof
measurementofassimilationwithbioavailability(Granado-Lorencio long chain-triacylglycerides (Borel et al., 1998; Hollander & Ruble,
et al., 2007; Hedrén, Diaz, & Svanberg, 2002; Hornero-Méndez & 1978).Duetotheestimationmethodapplied(assessmentinplasma
Mínguez-Mosquera,2007;Lemmens,VanBuggenhout,Oey,VanLoey, of the quantity of carotenoids accumulated during intake) it is not
&Hendrickx,2009;O'Sullivan,Jiwan,Daly,O'Brien,&Aherne,2010). possibletodifferentiatewhetherthetriacylglyceridechainlengthhas
Finally, digestibility plus assimilation are often referred to as aneffectonmicellisationorthestage(s)followingassimilation.
bioavailability (Chitchumroonchokchai, Schwartz, & Failla, 2004; Thefatsolubilityofindividualcarotenoidsalsoaffectsmicellisa-
Chitchumroonchokchai&Failla,2006). tion. Fat solubility depends on the structural particularities of each
Inmanycases,theinformationpublishedonthebioavailabilityof carotenoid. The general distinction of carotenoids as carotenes or
carotenoidsisobtainedbydeterminingthequantityofcarotenoidsin xanthophylls implies also differentiation with respect to fat and
plasmaaftertheintakeoffoodorsupplementeitherinasingledose solventsolubility.Comparativestudiesofbioaccessibilityindicatethat
(postprandialstudy)oroveracertainperiodoftime(supplementa- xanthophylls, such as lutein, are absorbed more readily than
tionstudy).Thus,bioavailabilityisestimatedoncethetotaldosehas carotenes (van het Hof et al., 1999). Therefore, the joint intake of
been administered, and the quantity of carotenoids accumulated in carotenoids from both groups somehow leads to an absorption
plasma is known. Furthermore, experiments may be designed such competition in which xanthophylls achieve better bioaccessibility
that relevant factors that modify either digestibility or assimilation (Fotouhi etal., 1996;Gärtner,Stahl,&Sies, 1996;Hendersonetal.,
maybedetermined.Forexample,thecarotenoidcompositionofthe 1989). However, these studies did not establish the step of the
ingested food or the type of food matrix may be changed or other carotenoidabsorptionprocessatwhichthiscompetitionarises.Itis
components that may interfere or complement absorption may be possible that these two types of carotenoids are incorporated
determined.Considerableamountofdatahasbeenobtainedwiththis differentlyintomicelles,duetotheirdifferentfatsolubility,butthe
strategy(Castenmiller&West,1998;O'Neill&Thurnham,1998;Furr absorption difference could also occur during the subsequent
&Clark,1997;Stahl&Sies,1992;vanVliet,Schreurs,&vandenBerg, assimilationstage(vandenBerg,1999).
1995).However,digestionandassimilationinvolveseveralsteps,and Anotherimportantfactorinthemicellisationphaseisthepresenceof
eachonecouldcauseaneffectonthenutrientorbioactivecompound soluble fibre. Dietary fibre consumption reduces the efficacy of this
so that a detailed picture is not obtainedwith an in vivo approach. processbecausethebiliarycontentbecomessolubleinthegelsformed
Therefore, application of in vitro digestion and assimilation pro- duringgastricdigestion,affectingtheefficacyofcarotenoidbioacces-
cedures is a suitable process to gain information about factors that sibility(Erdman,Fahey,&White,1986;Jones&Ntanios,1998;Rock&
modulate different steps of carotenoid bioaccessibility and Swendseid,1992;Westrate&Meijer,1998).Ithasbeendemonstrated
bioavailability. thatthepresenceofdietaryfatsubstitutes,suchassucrosepolyester,
notablyreducesthe bioaccessibilityofcarotenoids(Cooper,Webb,&
4.1.Digestibility:Solubilisationandmicellisation Peters, 1997). Pancreatic lipase inhibitors, which are used to treat
obesity,alsoreducetheefficacyofmicellisation,decreasingtheplasma
Carotenoids are lipophilic compounds and must first undergo concentration of carotenoids (James, Avenell, Broom, & Whitehead,
solubilisation from the food matrix followed by micellisation. This 1997;Zhi,Melia,Koss-Twardy,Arora,&Patel,1996).
refers to their incorporation into micelles, which are molecular Certain structural differences may alter fat solubility and,
aggregatesthat transport fat-soluble material,makingit potentially therefore,modifytheefficiencyofthemicellisation.Oneofthemis
accessible by the intestinal epithelium (Fig. 4). The degree of food the esterification of xanthophyll with fatty acids. This modification
processing is significant for micellisation efficiency as a high hasnotbeenwidelyconsideredinbioaccessibilitystudies,although
processing degree can maximise the amount of compound that is the xanthophylls we consume are usually esterified. Esterified
made soluble from the matrix. Mechanical homogenisation, the xanthophylls exhibit increased fat solubility relative to their corre-
application of a thermal treatment and the addition of fat during spondingfreexanthophyllsandevenagainstcarotenes(Fernández-
foodprocessingarealltechniquesthatincreasetheefficiencyofthis García, Mínguez-Mosquera, & Pérez-Gálvez, 2007). In addition,
phase (van het Hof, Gärtner, West, & Tijburg, 1998). The above- studiesofcarotenoidbioaccessibilityafterintakeoffoodrichinesters
mentioned techniques modify the subcellular structures where the show that only the corresponding free forms are detected in the
Author's personal copy
444 E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450
SOLUBILIZATION 1. Release from the food matrix.
2. Enzymatic hydrolisis (pancreatic
lipases, cholesterol esterase).
3. Interaction with biliary salts and
mixed micelles production.
MICELLIZATION
Mixed micelle
Transport to enterocytes
membrane
Simple diffusion Facilitate diffusion
UPTAKE
1. Monoacylglycerides re-acylation.
2. Cholesterol and retinol re-acylation.
3. Chylomicrons assembly.
Chylomicrons
Transport to liver vialymphatic
system
Fig.4.Representationofdifferentbioaccessibilityanduptakestagesoflipidiccompoundsinintestinalepithelium.
plasma, which implies that de-esterification has occurred at some unexpectedthathumanpancreaticlipasedoesnotactonxanthophyll
pointalongtheabsorption(Pérez-Gálvez,Martin,Sies,&Stahl,2003; esters,ithasalsobeenshowntolackactivitytowardsotheresterified
Roodenburg, Leenan, van het Hof, Westrate, & Tijburg, 2000). compounds (Rigtrup, McEwen, Said, & Ong, 1994). This study also
Therefore,itwaslogicaltoassumethatthisreactioncouldbecarried investigatestheactivityofcarboxyl-esterlipase(EC3.1.1.13),which
out by pancreatic lipases during the micellisation phase (Khachik, showedthexanthophyllesteraseefficiency.Carboxyl-esterlipaseisa
Beecher,Goli,Lusby,&Smith,1992;Olson,1984).Thesetwoaspects non-specificlipasethatisactiveinthepresenceofbiliarysaltsand
ofxanthophyllestermetabolism(higherfatsolubilityandsubjection hydrolyses cholesterol esters, tri-, di- and monoacylglycerides,
to the de-esterification process) also increase the requirements for phospholipids, lysophospholipids and ceramides (Hui & Howles,
biliary salts and digestive enzymes to guarantee xanthophyll ester 2002). This enzyme role complements the hydrolytic activity of
assimilation. In fact, the bioaccessibility studies that have been other lipases that act upon the previously mentioned substrates
performed indicate that a greater fat intake is required to achieve (Harrison, 1998) and it has even been suggested that its activity is
anefficientabsorptionofxanthophyllestersatlevelssimilartothose directlyinvolvedinvitaminAesterhydrolysis(Rigtrupetal.,1994;
of their corresponding free forms (Aman, Bayha, Carle, & Schieber, Rudd&Brockman,1984).
2004; Bowen, Herbst-Espinosa, Hussain, & Stacewicz-Sapuntzakis, Ifcarboxyl-esterlipaseisresponsibleforthehydrolysisofxanthophyll
2002;Roodenburgetal.,2000). estersthensignificanteffortsshouldbedirectedatdeterminingwhere
Specialmentionshouldbegiventothequestionofwhatenzyme/s thisenzymeactivityislocated.Thereareevidencesthatitsactivitytakes
is/areresponsibleforperformingthede-esterificationprocess.Some placeinthemembraneoftheintestinalepithelialcells(Gallo,Newbill,
havesuggestedthatpancreaticlipasescouldefficientlyperformester Hyun, Vahouney, & Treadwell, 1977; Lechene De la Porte, Aboukail,
hydrolysisandconvertxanthophyllesterstotheircorrespondingfree Lafont,&Lombardo,1987).Asaresult,onlylipidmaterialthatisfinally
forms.However,Breithaupt,Bamedi,andWirt(2002)demonstrated emulsifiedandmicellisedandreachesthemembraneoftheseepithelial
that human pancreatic lipase (EC 3.1.1.3) does not efficiently cellsissusceptibletohydrolysisbythisenzyme.Ithasbeenproventhat
hydrolyseesterifiedformsofxanthophyllssuchasβ-cryptoxanthin, xanthophyllestersareincorporatedintomicelles,whichitisnecessaryto
zeaxanthin,luteinandcapsanthin,whereasporcinepancreaticlipase reach the enterocyte membrane where they can be hydrolysed
and Candida rugosa lipase did posses this activity. Although it is (Chitchumroonchokchai&Failla,2006).
Author's personal copy
E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450 445
4.2.Absorption accumulate differentially in certain tissues depending on their
receptors density for LDLs or HDLs (Handelman, Snodderly, Adler,
Carotenoid uptake by enterocytes involves contact between the Russett,&Oratz,1992;Schmitz,Poor,Wellman,&Erdman,1991).
assimilable material of the micelles and the intestinal epithelium,
whichactsasa“selectivebarrier”fornutrients.Ourunderstandingof 5.Carotenoidsbioaccessibilityfromhydrophilicmatrices
theeventsthattakeplaceatthisstagehasbeensignificantlyenhanced
over the last years, demonstrating the existence of different Carotenoidbioaccessibilityis,asdiscussedpreviously,limitedand
absorption mechanisms from those initially considered. It was conditionedbydifferentfactors,mainlythedegreeoffoodprocessing
supposed that the assimilation of lipid micelle contents, including and matrix composition (Deming & Erdman, 1999). Fat (type and
carotenoids, took place in a passive diffusion process where the quantity)isarelevantfactorthatisrequiredinminimumquantitiesto
micelles collided with the cellular membrane and released their ensurebioaccessibility(Jalal,Nesheim,Agus,Sanjur,&Habicht,1998;
contents into the cytosol (El-Gorab, Underwood, & Loerch, 1975; van het Hof et al., 2000). However, it has been demonstrated that
Hollander & Ruble, 1978). Parker (1996) suggested that the carotenoid bioaccessibility from fatty foods (e.g. oily concentrates
concentration gradient between the micelle and the cytosol deter- such as tomato and paprika oleoresins) is not as high as expected,
minedthediffusionspeedandevenexplainedhowassimilationcould particularly when compared with the bioaccessibility provided by
besaturatedatcertaincarotenoidmicelleconcentrationsbyreducing hydrophilic matrices elaborated from these vegetables (Oshima,
the concentration gradient. However, several studies have signifi- Ojima,Sakamoto,Ishiguro,&Terao,1997;Pérez-Gálvezetal.,2003;
cantlyupdatedthemodelofcholesterol,carotenoidandotherlipid Richelle, 2002). Could hydrophilic matrices be considered as an
compounds absorption mechanisms (von Bergmann, Sudhop, & alternative to oily formulations to increase carotenoid bioaccessi-
Lütjohann, 2005; Werder et al., 2001). The biokinetic profile of bility? Lipophilic compound formulation in a hydrophilic matrix
absorption,obtainedinmodelsystemssuchasbrush-bordermembrane requiresanadequatesolubilisationtechnique(emulsion,encapsula-
vesicles(BBMV)andintestinalepitheliumcellcultures(Caco-2),shows tion)butinthiscaseitwouldbepossibletodiversifythetraditional
some characteristics of a facilitated diffusion (Compassi et al., 1997; sourcesofthesebioactivecompounds.Inthissense,applyinganyof
Thurnhofer & Hauser, 1990). Based on this observation, the type B theappropriatetechniquescouldbeanadequatestrategytoachieve
residual receptors (SR-BI), cluster of differentiation 36 (CD36) and stable,digestibleandwell-assimilatedbioactivecompounds.Tothis
Niemann–PickC1-Like1protein(NPC1L1)wereeventuallyidentifiedas end,liquidformulationsinwhichcarotenoidshavebeenincorporated
facilitatorsoftheabsorptionofcholesterolandcarotenoids,although asoilinwateremulsionshavebeendeveloped(Amar,Aserin,&Garti,
theirpreciseroleinthisprocessisstillunderdebateandamatterof 2003;Spernath,Yaghmur,Aserin,Hoffman,&Garti,2002).Intheory,
furtherinvestigations(vanBennekumetal.,2005). the adequate incorporation of a carotenoid into a hydrophilic
The model of facilitated diffusion has been supported by the environment emulsion should increase bioaccessibility, although
findingofmacromoleculessuchasα-andβ-peptidesorezetimibe(a this fact has scarcely been determined. Then, quality design of the
drugthatlowerscholesterolabsorption)thatinhibitslipidcompound emulsion formulation will be completed when the estimation of
absorption(Altmannetal.,2002;Boffellietal.,1997).Inaddition,this digestibility/assimilation from that matrix is incorporated as a
mechanism is consistent with the significant variability of lipid criterionduringtheoptimisationprocessoftheformulationcompo-
absorption efficacy that is observed (e.g. cholesterol) among in- sition(Fernández-García,Rincón,&Pérez-Gálvez,2008).
dividualswithinthesamepopulation,whichisattributedtogenetic
differencesexpressedintheintestinal epithelium (Wang,Paigen,& 6.Invitrodigestionmodels
Carey,2001).Asimilarvariabilityhasalsobeenobservedinstudiesof
postprandial carotenoid absorption (Paetau, Chen, Goh, & White, Over the last decades, the available information on carotenoid
1997;Pérez-Gálvezetal.,2005).Theseachievementsarelargelydue bioavailabilitywasobtainedwithpostprandialabsorptionandsupple-
tothedevelopmentofinvitroabsorptionmodels,basedonobtaining mentation studies. This approach, although highly valuable, is not
BBMVandcellularcultures.Thesemodelshavebeenusedasanalytical sufficientforadetailedanalysisoftheeffectsoffoodmatrixcomposition
tools in experiments aimed at determining the biokinetics of lipid orotherfactorsoneachstageintheprocessofabsorption.Therewerein
absorption in general and of the carotenoids in particular (During, vitro digestion models ready for the estimation of digestibility of
Hussain,Morel,&Harrison,2002;During,Dawson,&Harrison,2005; minerals such as iron and phosphorus (Liu, Ledoux, & Veum, 1998;
Moussaetal.,2008;Rebouletal.,2005; vanBennekumetal.,2005; Miller,Schricker,Rasmussen,&vanCampen,1981),cholesterol(Fouad,
Wilson,1990). Farrell, Marshall, & van de Voort, 1991) and hydrosoluble vitamins
Oncethelipidmaterialisinternalisedbythecells,itisthenpacked (Ekanayake&Nelson,1986).Thesearestaticgastrointestinalmodels,
intolipoproteins,calledchylomicrons.Thesemolecularaggregatesare simulatingthetransitthroughthehumandigestivetract,andreprodu-
assembledintheGolgiapparatus,reachasizeof75to1200nmand cing the physiological conditions of gastric and intestinal digestion.
arecharacterisedbytheirapoproteinB-48content,whichisessential Thesemodelsareeasytoapplytoalargenumberofsamplesandare
totheassemblyoftheseparticles.Chylomicronsareexcretedintothe thuswellsuitedforstudiesoftheeffectsofvariousdigestionconditions
lymphaticsystemandareeventuallytakenupbytheliver,wherethey or other factors linked to the food matrix on the bioaccessibility
are stored or re-excreted into circulation within very low-density (digestibilityandassimilation)ofanutrient.
lipoproteins(VLDLs).Somechylomicronsdegradebeforereachingthe The procedure consists of subjecting the food to the digestive
liverduetotheactivityoflipoproteinlipasethereforepartofthelipid processwhichisdividedintotwostages,gastricandintestinal,whose
content of these particles is released and absorbed into the standard conditions are generally based on those in Miller et al.
endothelialtissue.However,mostchylomicronsreachtheliverand (1981).Duringthegastricphase,aportionofthehomogenisedfoodis
are incorporated into VLDLs; once released into the circulatory acidified to pH 2 with 1M HCl and a porcine pepsin suspension is
system, they are transformed into low-density lipoproteins (LDLs) added. The homogenised portion is transferred to a test tube and
andfinallyintohigh-densitylipoproteins(HDLs).Apolarcarotenoids incubated at 37°C in a water bath with circular agitation. Then, to
suchasβ-caroteneandlycopeneareconcentratedatthecentresof simulatetheintestinalphase,thepHisincreasedto5.3withsodium
lipoproteic particles and are not frequently exchanged with other bicarbonatesolution.ThepHisagainincreasedto7.5andthemixture
lipoproteins;thus,LDLsexperienceanincreaseincarotenecontent, isincubatedfor2hunderthetemperatureandagitationconditions
whereas HDLs are comprised of xanthophylls (Paetau et al., 1998). mentioned above. When this digestion procedure is applied to
Thispointisrelevantasitindicatesthatcarotenesorxanthophyllswill estimate digestibility of a hydrosoluble nutrient, it is not necessary
Author's personal copy
446 E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450
toisolatethemicellisedcontentfromtherestofthenon-digestedor pancreatin enzymatic pool is replaced by pancreatic lipase and
partially emulsified food. However, these models were not wholly carboxyl-ester lipase. The use of the carboxyl-ester lipase enzyme,
applicable for digestibility studies of fat-soluble compounds, since togetherwithpancreatinlipase,intheintestinaldigestionphaseisthe
they did not include a micelle fraction separation step, and this mainmodificationtothemethodofGarrettetal.(1999)thathasbeen
separation is essential when the digestibility and/or assimilation of introduced in other studies, as this enzyme mixture seems to have a
fat-soluble materials are to be analysed. This procedure was not positiveeffectonmicellisationandsubsequentcarotenoidassimilation
developeduntil1990.Hernell,Staggers,andCarey(1990)conducted (Chitchumroonchokchai&Failla,2006).InthestudyofGranado-Lorencio
adetailedstudyofthephysical–chemicalpropertiesoftheduodenal et al. (2007), an initial mastication simulation phase incorporating a
contentsduringthedigestionoffoodrichintriglyceridesingestedby salivarysolutionwithorganicandinorganiccomponentsandα-amylase
volunteers.Theapplicationofanultracentrifugationstepallowedthe (EC 3.2.1.1) is applied before the two digestion phases, gastric and
separation of non-digested triglycerides, the micelle fraction and a intestinal.Invitrodigestionmodelandexperimentaldesignofmixtures
precipitate fraction. The combination of an in vitro digestion model wereusedasanalyticaltoolstomeasurecarotenoidbioaccessibilityand
with the micelle fraction separation method by ultracentrifugation tooptimisetheformulationofanO/Wemulsion,includingcarotenoidsas
yields a complete experimental tool for the analysis of fat-soluble functionalingredients(Fernández-Garcíaetal.,2008).Twoexperimental
componentdigestibility,becausethemicellelipidcontent,whichis stageswereapplied.First,ascreeningphasewascompletedtodetectthe
susceptibletoassimilationbyepithelialtissuecells,canbeseparated. critical factors that exerted a significant effect on the response
The experimental conditions of any digestion model should not (bioaccessibility).Duringthisphase,itwasobservedthattheresponse
differsignificantlyfromthosejustmentioned,asitcloselyreproduces was modified mainly by secondary effects such as synergies and
thephysiologicalstagesandconditionsoftheformermodel,butthese antagonisms of the emulsifying mixture. Due to the effect on the
consist of specific changes or modifications to the basicmodel that response, a group of four emulsifiers was selected at this phase to
dependontheparticularproblembeingtargetedortheapplicationto perform the second experimental stage, the optimisation phase. This
be developed. In vitro digestion models have multiple applications allowedobtainingthemixturethatproducedthemaximumcarotenoid
and can contribute to knowledge in diverse areas. Some uses have bioaccessibility.
includedtoxicologyanalysesofpossiblefoodcontaminants,thestudy
ofdietaryfactorsinvolvedinthebioaccessibilityofcertainnutrients 7.Invitroabsorptionmodels
orfoodcomponentsanddeterminationoftheeffectthatdifferentor
unusualhepatobiliaryconditionshaveonthebioaccessibilityoflipid Overall,theprocessestakingplaceduringassimilationmaybethe
compounds (Leo et al., 1995; Oomen, Tolls, Sips, & Groten, 2003). most complex to reproduce with in vitro models. Apart from the
Adaptationstothismodelhavealsobeenusedinthedesignoffood physical–chemical processes that take place during assimilation,
matrices, in which bioaccessibility is a matrix component selection biochemical events must also be included to adequately reproduce
criterion along with physical structure, stability or palatability intestinalabsorption.Theintestinalmucosaisnotonlyabarrierinthe
(Manzocco&Nicoli,2002).Bioaccessibilitymustbeacriteriontobe physicalsense,butalsoabiochemicalbarrier.Thecelllayercarriesout
consideredinthedesignoffunctionalfoods,giventhatanyclaimof complex metabolic processes that are part of the pre-systemic
health benefits establishedthroughtheintake of thesefoods isnot metabolism, which is separated from the nutrient/bioactive com-
validunlessitcanbeshownthatthebioactivecomponentisabsorbed poundtransportviasimpleorfacilitateddiffusionoractivetransport.
efficiently (Aggettetal., 2005). Therefore,in vitrodigestionmodels However,despitethiscomplexity,therearenumerousalternativesto
are an effective analytical tool that measures bioactive compound reproducingtheintestinalassimilationprocess.Therefore,itisworth
digestibility from different matrices and can be used as a starting mentioningtheexistenceofexvivo(Ungell,2002),insitu(Schanker,
point for later analysis of assimilation without resorting to in vivo Tocco, Brodie, & Hogben, 1958) and in vivo techniques (Poelma &
experimentationmodels. Tukker,1987)beforedescribinginvitroassimilationmodels.
Intheparticularcaseofcarotenoidpigments,invitrodigestionhas The various types of in vitro models include those based on a
beenappliedtomeasurethedigestibility ofthesecompoundsfrom physicalreproductionofthemembrane,thosebasedoncellcultures
naturalfoods.ThestudybyGarrett,Failla,andSarama(1999)isthe andthosethatuseisolatedorreconstitutedcellmembranes.Models
firstapplicationofaninvitrodigestionmodelincombinationwithan based on artificial membranes aim to specifically measure the
intestinal assimilation model, based on the Caco-2 cell line, to membrane permeability of the nutrient or bioactive compound
determine carotenoid bioavailability in food. Their experimental usuallyfollowingthemethodologyofKansy,Senner,andGubernator
procedure follows the stages described above: a gastric phase, an (1998). The membrane is reproduced artificially by integrating
intestinalphaseandseparationofthemicellefraction.Children'sfood phospholipids and other membrane constituents in a solution
basedonfruitsandvegetablesandtomatopastewasusedtoanalyse containingorganicsolvent.Thissolutionisplacedoveraporousfilter
various factors that control the digestive process. Therefore, it was thatservesasphysicalsupportbetweenthedonatingbuffersolution
determinedthatomittingthegastricdigestionphasedoesnotmodify below and the receiving buffer solution above. The forte of
theefficiencyofthemicellisationprocessintheintestinaldigestion membrane-typemodelsliesinthefactthatthelipidcompositionof
phase. The quantities of biliary salts and pancreatic enzyme the artificial membrane that is placed over the filter and can
suspension that provide the best transfer of carotenoids from food reproduce,asfaraspossible,themembranecompositionofintestinal
tomicelleswerethendefinedandwerefoundtocorrespondtothe epithelium cells. The greatest number of modifications has been
physiological concentrations estimated in earlier studies (Charman, introducedinthisparametertooptimisethemodel.Fromtheoriginal
Porter,Mithani,&Dressen,1997;Tso,1994). composition described by Kansy et al. (1998) to the modifications
Subsequently,themethoddevelopedbyGarrettetal.(1999)has made by Sugano et al. (2001) or by Avdeef, Nielsen, and Tsinman
beenappliedtoandadaptedforotherstudiesofdigestibilityand/or (2004),improvementshavemadeitpossibleformembraneperme-
assimilation of carotenoids, with slight modifications. Hedrén et al. abilitymodelstoreproducetheresultsobtainedwithinvivomodels.
(2002) studied the digestibility of carotenes in fresh and cooked Anotherrecurringthemeistheattempttoreproducethelumen/in-
carrots,applyingsimilarconditionsintermsofdigestiveenzymeand testinalepitheliumdynamicinthesemodels.Althoughtheconditions
biliarysaltsconcentrationsaswellasincubationtimes,butmodifying ofthesemodelsareessentiallystatic,thisapproachofmodellinghas
theconditionsofmicellefractionseparation.Inworkspublishedby greatpotential.TechnicalarticlesfromR&Dpharmaceuticalcompa-
Breithaupt et al. (2002) and Zorn, Breithaupt, Takenberg, Schwack, nieshavepredictedthattheywilldisplaceinvitromodelsbasedon
andBerger(2003)thegastricdigestionphaseiseliminatedandthe cell cultures, considering their lower costs and a more acceptable
Author's personal copy
E.Fernández-Garcíaetal./FoodResearchInternational46(2012)438–450 447
correlation with results obtained in vivo (Kansy et al., 1998). This oftheassimilationofglucose,biliarysalts,lipids,xenobiotics,cholesterol
techniqueisidealasahigh-throughputscreeningmodelfordrugor andcarotenoids(Alcorn,Simpson,Leahy,&Peters,1991;Burckhardt,
phytochemicalcompounddevelopment. Kramer, Kurz, & Wilson, 1983; Fernández-García, Carvajal-Lérida,
Modelsbasedoncellculturesareverypopularandhavebeenlargely Rincón, Ríos, & Pérez-Gálvez, 2010; Hopfer, Nelson, Perrotto, &
accepted by the scientific community as valid references to estimate Isselbacher, 1973; Ikeda, Tanaka, Sugano, Vahouny, & Gallo, 1988;
nutrient and bioactive compound assimilation. Given that primary Keelan,Burdick,Wirzba,&Thomson,1992;Kessleretal.,1978;Ling,Lee,
culturesofenterocytesfromexplantsofepithelialtissuedonotforman & Hollander, 1989;Moore,Gugger,& Erdman, 1996; Proulx, McNeil,
epithelial monolayer and therefore do not have apical and basolateral Biglez, & Williamson, 1982). Intestinal tissue from pigs, rabbits,
areasitisnecessarytoresorttocellularlines(usuallyfromtumours)that chickens, rats, mice, and humans has been used in these studies,
do have the structural characteristics of a cell monolayer. The cell although human BBMV studies are less frequent due to the limited
monolayerisculturedoverasupportingfiltersuchthattheapicalarea availabilityof tissuein the conditionsnecessary for thesetrials.This
contactsthedonatingbuffersolutionandthebasolateralareacontactsthe modelalsohaslimitations,mainlybecauseitisonlypredictiveforthe
receivingbuffersolution.Differentcelllineshavebeendescribed,suchas step of the absorption process that occurs on the interior of the
Madin–DarbyCanineKidneycells(MDCK),cellsfromratfoetusintestine membrane.Furthermore,thevesiclesareinsuspension,whichisthe
(2/4/A1),cellsfrompigkidneyepithelium(LLC-PK1)andthebestknown reasonwhydynamicconditionsareappliedtotheassay;however,this
cell line, Caco-2. This cell line comes from human adenocarcinoma prevents the differentiation of a basolateral area that represents
colonocytesandwhenculturedunderappropriateconditionspresentthe transportthroughthemembrane.
morphological characteristics of enterocytes and also produce a large
majorityofmembraneenzymesandproteins(Jumarie&Malo,1991). 8.Conclusions
Active transport of glucose, peptides, amino acids, biliary acids,
carotenoids and flavonoids has been characterised in this cell line, Carotenoidpigmentsareagroupofbioactivecompoundsthatareof
demonstrating they are a reliable in vitro model (Bleasby,Chauhan, & interesttothefoodscientists,nutritionistsandfoodindustriesdueto
Brown, 2000; Delie & Rubas, 1997; Garrett et al., 1999; Manach & their positive impact on human health and their economic benefits.
Donovan,2004).Miniaturisation,thedevelopmentofphysiologicalbuffer Carotenoidsareresponsiblefortheattractivecolourofmostfruitand
solutionsandstandardisationwithreferencepatternsubstancestoenable vegetables, having diverse biological functions and activities. An
comparisonofresultsbetweenlaboratories,areallbeingintroducedto extensive number of factors determine the efficient incorporation of
bringthesemodelsclosertotheinvivoreality(Ingels&Augustijns,2003). thesephytochemicalsfromthediet.So,ifwewanttotakeadvantageof
Asinthecasewithmodelsbasedonartificialmembranes,modelsbased allofthesebenefitswemustconsiderthebioaccessibilityeventasa
on cell cultures are excessively static and thus they are subject to necessarystageinthecomplexprocessofconvertingavisualattraction
successive improvements. Additional effort is being directed towards intobiologicalactions.Asitcanbeappreciatedfromthisreview,thereis
enabling cells to conveniently express the large variety of protein a wide range of experimental models that can be used to obtain
transportersthatisusuallyfoundinintestinalepithelialenterocytes.Itis informationabouttheevents thatoccurredduringeach stage of the
importanttobearinmindthatthislinecomesfromthecolon,notfromthe carotenoid digestive process, as well as the mechanisms involved in
small intestine, which implies a handicap in expressing the adequate every step. Each of these models focuses on a particular aspect
numberoftransporters,giventhatmorereceptorsareexpressedinthe (digestibilityorassimilationefficiency,structuralmodificationsdueto
smallintestinethaninthecolon(Seithel,Karlsson,Hilgendorf,Bjorquist,& the digestive process or the intestinal metabolism, mechanisms
Ungell,2006).Inordertoimprovethisaspect,Caco-2-derivedclonallines involvedintheassimilationorinhibitionofassimilation),thusproviding
suchasTC7havebeendeveloped(Carriereetal.,1994;Rebouletal.,2005; informationthatisrelevanttocarotenoidbioavailability.
Verma,Hansch,&Selassie,2007).Invitroassimilationmodelsbasedon
cell cultures consist of an experimental scanning technique (high-
Acknowledgements
resolutionthroughputscreening)thatislargelyusedinthepharmaceu-
tical industry to estimate drug assimilation, which uses specialised
ThisworkwassupportedbytheMinisteriodeEducaciónyCiencia
instruments to perform certain experiments many times in parallel.
(Spanish Government, Projects AGL2007-61146/ALI and AGL2007-
This model has provided excellent insights of the carotenoid cellular
60092/ALI)andConsejeríadeEconomía,InnovaciónyCiencia(Junta
uptakeandtransportincreasingthepossibilityofdatacorrelationwiththe
deAndalucía,ProjectsP07-AGR-03025andP08-AGR-03477).E.F.G.is
results from in vivo studies (Chitchumroonchokchai et al., 2004;
a research fellow of the Spanish National Research Council (I3P,
Chitchumroonchokchai&Failla,2006;Duringetal.,2002;Duringetal., predoctoralprogrammecofinancedbytheEuropeanSocialFund).
2005;Moussaetal.,2008;Rebouletal.,2005;vanBennekumetal.,2005).
I.C.L. is a research fellow of the Junta de Andalucía (Proyecto de
PreparationsofBBMVareaclassicmodelformechanisticstudiesof Excelencia, predoctoral programme cofinanced by the European
nutrient bioactive compounds or drug transport through the cell
SocialFund).
membrane. The vesicles, cell membranes in suspension, act as
receptorsastheypreservetheproteinictransportersandmembrane
References
enzymeswithwhichtheyinteracttogetherwiththenutrientsfrom
the donor solution in which they were contacted. The vesicles are Aggarwal,B.B.,Bhardwaj,A.,Aggarwal,R.S.,Seeram,N.P.,Shishodia,S.,&Takada,Y.
(2004).Roleofresveratrolinpreventionandtherapyofcancer:Preclinicaland
obtainedbyisolatingcellsfromtheintestinalmucosabycurettageor
clinicalstudies.AnticancerResearch,24(5A),2783–2840.
vibrationandthenremovinganycontaminatingtissueorbasolateral Aggett,P.J.,Antoine,J.M.,Asp,N.G.,Bellisle,F.,Contor,L.,Cummings,J.H.,etal.(2005).
membrane residues. The purified mucosal cells are fractioned by Passclaim:Consensusoncriteria.EuropeanJournalofNutrition,44(1),i5–i30.
centrifugation and precipitation into BBMVs separated from the Alcorn,C.J.,Simpson,R.J.,Leahy,D.,&Peters,T.J.(1991).Invitrostudiesofintestinal
drugabsorption.BiochemicalPharmacology,42,2259–2264.
nuclear and mitochondrial fraction. This technique, described by Altmann,S.W.,Davis,H.,Yao,X.,Laverty,M.,Compton,D.S.,Zhu,L.J.,etal.(2002).The
MillerandCrane(1961)hasbeensomewhatmodifiedtoimprovethe identificationofintestinalscavengerreceptorclassB,typeI(SR-BI)byexpression
purity,stabilityandyieldofthepreparation,butoverallthetechnique cloninganditsroleincholesterolabsorption.BiochimicaetBiophysicaActa,1580,
77–93.
has remained quite unaltered. ThepreparationofBBMVsinisotonic
Aman, R., Bayha, S., Carle, R., & Schieber, A. (2004). Determination of carotenoid
buffer is incubated with the donating solution or a suspension that stereoisomers in commercial dietary supplements by high-performance liquid
contains the nutrients of interest. After some incubation time, the chromatography.JournalofAgriculturalandFoodChemistry,52,6086–6090.
solutionisfilteredorcentrifugedtoseparatetheBBMVfractionandthe Amar,I.,Aserin,A.,&Garti,N.(2003).Solubilizationpatternsofluteinandluteinesters
infoodgradenonionicmicroemulsions.JournalofAgriculturalandFoodChemistry,
buffer.Thistechniquehasbeenusedsuccessfullyinmechanisticstudies 51,4775–4781.