Journal of the Science of Food and Agriculture J Sci Food Agric 86:1481–1488 (2006)

Antioxidant capacity and secondarymetabolites in four species of Andeantuber crops: native potato (Solanum sp.),mashua (Tropaeolum tuberosum Ruiz &Pavon), Oca (Oxalis tuberosa Molina) andulluco (Ullucus tuberosus Caldas)David Campos,1 Giuliana Noratto,2 Rosana Chirinos,1 Carlos Arbizu,3 Willian Roca3

and Luis Cisneros-Zevallos2∗1Instituto de Biotecnologıa, Universidad Nacional Agraria-La Molina, Lima 12, Peru2Department of Horticultural Sciences, Texas A&M University, College Station, TX 77843-2133, USA3International Potato Center, Lima 12, Peru

Abstract: Four species of edible tubers endemic to and domesticated in the Andes, native potato (Solanum sp.),mashua (Tropaeolum tuberosum Ruiz & Pavon), oca (Oxalis tuberosa Molina) and ulluco (Ullucus tuberosusCaldas), were studied for their antioxidant capacity and associated secondary metabolites. The antioxidantcapacity was measured using ABTS− radicals and total phenolics, carotenoids, anthocyanin, betaxanthin andbetacyanin content were also characterized. The antioxidant capacity found in the crops studied ranged from 483to 9800 µg trolox equiv. g−1, phenolics ranged from 0.41 to 3.37 mg chlorogenic acid equiv. g−1, anthocyanins rangedfrom 0.08 to 2.05 mg cyanidin 3-glucoside g−1 and carotenoids ranged from 1 to 25 µg β-carotene g−1. The content ofbioactive compounds was high and variable between crops and within the genotypes studied. In general, mashuatubers showed the highest antioxidant capacity and phenolic, anthocyanin and carotenoid content compared withother crops. Ulluco was the only crop that contained betalains in the acid form of betaxanthins (22–96 µg g−1) andbetacyanins (64 µg g−1) with no presence of carotenoids or anthocyanins. This is the first publication regarding theantioxidant capacity of and associated secondary metabolites in Andean tubers. This information can be usefulin the identification of Andean tubers species and genotypes with potential value as a novel dietary source ofantioxidants for food, and also for medicinal use. 2006 Society of Chemical Industry

Keywords: antioxidants; Andean tuber crops; potato; mashua; oca; ulluco

INTRODUCTIONThere is a continuous search for new plant com-pounds with antioxidant potential. Interest in natu-ral antioxidants has increased considerably in recentyears.1–4 Antioxidants, which can neutralize free rad-icals, may be of central importance in the preventionof cancer and cardio- and cerebrovascular diseases.5–9

Many natural antioxidants exhibit a wide range ofbiological effects, including antibacterial, antiviral,anti-inflammatory, antiallergic, antithrombotic andvasodilatory activity.10 The Andean region is wellknown for its great genetic diversity in different types ofcrops, which unfortunately have been poorly studied.Andean crops show variability in colors, forms, sizes,primary nutrient constituents and bioactive secondarymetabolites.

People from the Andean region have utilized nativepotato (Solanum sp.), mashua (Tropaeolum tuberosum

Ruiz & Pavon), oca (Oxalis tuberosa Molina) and ulluco(Ullucus tuberosus Caldas) tubers for their nutritionaland medicinal properties since ancient times. It isthought that the health-related properties of Andeantuber crops claimed from folklore use could be in partattributed to the antioxidants present in these crops.However, little is known about the chemical natureand identity of the bioactive compounds present.Hence there is a need for scientific data on theantioxidant content of these Andean crops in orderto increase our understanding of their role in thediet and in reducing chronic diseases. These initialsteps may lead to an increased antioxidant intakeby breeding Andean crops with higher antioxidantactivity.

There have been a few attempts at quantifyingthe nutritional value of Andean tubers,11,12 but thisstudy is the first attempt to measure accurately the

∗ Correspondence to: Luis Cisneros-Zevallos, Department of Horticultural Sciences, Texas A&M University, College Station, TX 77843-2133, USAE-mail: lcisnero@taexgw.tamu.edu(Received 17 February 2005; revised version received 26 September 2005; accepted 20 February 2006)Published online 15 June 2006; DOI: 10.1002/jsfa.2529

2006 Society of Chemical Industry. J Sci Food Agric 0022–5142/2006/$30.00

D Campos et al.

antioxidant capacity and related bioactive compoundsof a wide germplasm of different Andean tubers.

The objective of this study was to provide newdata regarding Andean tuber crops as a sourceof antioxidant compounds and compare them withother commercial crops. The information obtainedshould stimulate an increase in the evaluation andconservation of tuber diversity within the Andeanregion and in the identification of these crops as newsources of functional foods.

MATERIALS AND METHODSPlant materialFifteen native potato (Solanum sp.), 14 oca genotypes(Oxalis tuberosa Molina), 15 ulluco (Ullucus tuberosusCaldas) and 11 mashua (Tropaeolum tuberosum Ruiz &Pavon) genotypes were supplied by the InternationalPotato Center (CIP) (Lima, Peru). The tubers wereharvested during June 2002 from CIP’s experimentalfields in Huancayo (3700–3800 m), Peru. Strawberryand carrots were purchased from a local market inLima and used as reference crops. Tuber sampleswere analyzed for total phenolics (TPH), anthocyanins(ACY), carotenoids (CTC), betaxanthins (TBX),betacyanins (TBC), hydrophilic antioxidant capacity(HAC) and lipophilic antioxidant capacity (LAC).Strawberries were analyzed for ACY and TPH andcarrots for TCT; these two crops were used asreference crops for comparison purposes. All sampleswere frozen and stored at −20 ◦C until the time ofanalysis. Samples were taken from the middle sectionin slices which included skin and flesh in the analysis tokeep the original ratio between skin and flesh presentin the tuber. The analyses were run in triplicate foreach tuber using three tubers per genotype.

ChemicalsAll solvents and other chemicals were ofanalytical grade and obtained from Merck(Darmstadt, Germany), J.T. Baker (Phillipsburg,NJ, USA) and Mallinckrodt (Phillipsburg, NJ,USA). 2,2′-Azinobis(3-ethylbenzthiazoline-6-sulfonicacid (ABTS) and Folin–Ciocalteu reagent (Catalog#F9252) were purchased from Sigma Aldrich (StLouis, MO, USA).

Total phenolic compounds (TPH)Total phenolics were determined using theFolin–Ciocalteu reagent following the procedurereported by Swain and Hillis.13 A 5-g sample washomogenized with 20 mL of 95% ethanol to a uniformconsistency using an Ultra-Turrax homogenizer andleft at 4 ◦C for 24 h before filtration. An aliquot ofextract (0.5 mL) was diluted with 8 mL of water.Simultaneously, a blank sample was prepared with95% ethanol and treated in the same way as thesamples. The Folin–Ciocalteu reagent was dilutedwith water (1:7) and 0.5 mL were added to the dilutedextracts, vortex mixed and allowed to react for 3 min.

At 3 min, 1 mL of 0.5 mol L−1 Na2CO3 was addedand allowed to react for 10 min. The absorbance wasmeasured at 725 nm until a constant reading wasreached. Chlorogenic acid was used as standard forthe calibration curve and TPH were expressed as mgof chlorogenic acid equiv. g−1 fresh weight. A Gensys-5 UV–visible spectrophotometer (Milton Roy, NewYork, USA) was used for absorbance readings.

Total anthocyanins (ACY)The method reported by Fuleki and Francis14 wasused for ACY determination. A 5-g sample and15 mL of solvent (85:15 95% ethanol–1.5 mol L−1

HCl) were homogenized as described above and leftat 4 ◦C for 24 h before filtration. A 2-mL volumeof extract was transferred to a graduated cylinderand anthocyanin solvent was added to obtain a finalvolume of 100 mL. Hexane was added to removeany carotenoids present. Spectrophotometric readingsat 535 nm were taken, subtracting the absorbance at700 nm (due to turbidity). ACY were expressed as mgcyanidin 3-glucoside equiv. g−1 fresh weight, using amolar extinction coefficient15 of 25 965 L mol−1 cm−1

and a molecular weight of 449 g mol−1.

Total carotenoids (TCT)The method reported by Talcott and Howard16 wasused for measuring TCT. A 2-g sample with 20 mL ofacetone–ethanol (1:1) solution containing 200 mg L−1

BHT was homogenized as described above beforefiltration. The filtrate was transferred into a graduatedcylinder and solvent added to a final volume of100 mL. A 50-mL volume of hexane and 25 mLof H2O were added and shaken vigorously beforestanding for 30 min to allow separation of phasesto occur. The spectrophotometer was blanked withhexane and the absorbance of the hexane phasewas measured at 470 nm. β-Carotene was used asa standard for the calibration curve and TCT wasexpressed as µg β-carotene equiv. g−1 fresh weight.

Betaxanthins (TBX) and betacyanins (TBC)The method adapted from Cai and Corke17 wasused for measuring betalains. A 5-g sample and20 mL of solvent (McIllvaine buffer, pH 5.2) werehomogenized for 2 min as described above. Solventwas added to a final volume of 50 mL beforefiltration. The spectrophotometer was blanked withMcIllvaine buffer. The absorbance of the extractwas measured at 476 nm for betaxanthin content(orange–yellow pigment) and 536 nm for betacyanincontent (purple–red pigment). The content of totalbetaxanthins and betacyanins was calculated usingabsorptivity values, E1%

1cm, of 750 and 1120,respectively. Results were expressed as mg TBX orTBC g−1 fresh weight.

Determination of antioxidant capacityAntioxidant capacity was determined by the ABTSmethod adapted from Arnao et al.18 A 5-g sample and

1482 J Sci Food Agric 86:1481–1488 (2006)DOI: 10.1002/jsfa

Antioxidants in Andean tuber crops

25 mL of methanol were homogenized as describedabove and left at 4 ◦C for 24 h before filtration.An aliquot of this extract was taken for measuringthe hydrophilic antioxidant capacity (HAC). Thepellet was mixed with 25 mL of dichloromethaneand shaken for 15 min before filtration. An aliquotof the yellow–orange supernatant was taken formeasuring of lipophilic antioxidant capacity (LAC).A 150 µl volume of extract was mixed with 2.85 mL ofABTS solution (∼1.1 absorbance, 734 nm) preparedas described by Awika et al.19 This mixture wasallowed to react at 20 ◦C until a steady absorbancewas reached. Simultaneously, a 150-µl aliquot ofmethanol was treated in the same way as the sampleand used as a control. The spectrophotometer wasblanked with methanol and the decrease in absorbancedue to antioxidant activity was recorded at 734 nm.Antioxidant capacity was calculated from a calibrationcurve developed for Trolox, providing a relativeantioxidant capacity of the extracts compared withthis standard and expressed as µg TE (trolox equiv.)g−1 fresh weight.

Statistical analysisSPSS for Windows version 11.0 was used forstatistical analysis. Pearson correlation at the levelof α = 0.01 and analysis of variance (ANOVA) withmean comparisons using Tukey’s multiple range testat α = 0.05 were performed. Results are expressed asaverage ± standard deviation (SD).

RESULTS AND DISCUSSIONThe overall average moisture content of the genotypesstudied in native potatoes, mashua, oca and ullucotubers was ∼71.5 ± 3.4, 89 ± 2.2, 83 ± 1.9 and86.7 ± 1.7%, respectively, indicating a small variationof moisture content among the genotypes studied ineach crop. In general, results expressed on dry or wetbasis did not alter the overall ranking for each crop inrelation to TPH, ACY, TCT, HAC and LAC in thisstudy. We consider it meaningful to present the dataon a wet or fresh basis since that is the way in whichthese crops are consumed and the way in which mostreferences are available in the literature.

Total phenolic content (TPH)Our results indicate that the TPH content fornative potato ranged from 0.64 to 2.32 mg g−1

(Fig. 1). Genotypes 707 132, 704 463 and 705 841had the highest TPH contents with 2.32, 2.09and 1.9 mg g−1, respectively. These values are higherthan those reported previously for purple potatoes20

(0.76–1.27 mg g−1). In mashua tubers, the TPHranged from 0.92 to 3.37 mg g−1 (Fig. 2). GenotypesARB-5241, DP-02-24 and AGM-5109 had the highestTPH with 3.37, 3.05 and 2.75 mg g−1, respectively.Purple mashua genotypes presented higher TPHvalues, whereas lower values of TPH were found in

0

1

2

3

4

5

6

7044

81

7038

44

7038

25

7032

79

7029

61

7054

68

7043

53

7068

19

7032

86

7050

24

7039

85

7052

63

7071

32

7044

63

7058

41

TP

H o

r A

CY

(m

g g-1

)

0

1000

2000

3000

4000

5000

6000

HA

C (

ug T

E g

-1)

TPH ACY HAC

Figure 1. Total phenolic content (TPH), anthocyanins (ACY) andhydrophilic antioxidant capacity (HAC) of native potato tubers(Solanum sp.).

0

2

4

6

8

10

12

DP

-02-

03

DP

-02-

23

DP

-02-

07

AR

B-5

576

AV

M-5

562

AG

M-5

109

M6C

OL

2C

DP

-02-

15

AR

V-5

366

DP

-02-

24

AR

B-5

241

TP

H o

r A

CY

(m

g g-1

)

-500

1500

3500

5500

7500

9500

11500

HA

C (

ug T

E g

-1)

TPH ACY HAC

Figure 2. Total phenolic content (TPH), anthocyanins (ACY) andhydrophilic antioxidant capacity (HAC) of mashua tubers.(Tropaeolum tuberosum Ruiz & Pavon).

0

0.5

1

1.5

2

2.5

3

3.5

4

TP

H o

r A

CY

(m

g g

-1)

-500

500

1500

2500

3500

4500

5500

6500

HA

C (

ug

TE

g-1

)

TPH ACY HAC

AR

B-5

054

AJA

-527

0

AJA

-524

5

AJA

-525

2

AT

T-5

466

O-2

66-8

5

O-1

16-8

4

O-0

18-8

3

DP

-01-

46

CO

C-5

37

AM

M-5

223

AA

Q-5

477

MU

-31

PIC

A-H

I-92

OC

Figure 3. Total phenolic content (TPH), anthocyanins (ACY) andhydrophilic antioxidant capacity (HAC) of oca tubers (Oxalistuberosa Molina).

yellow mashua genotypes. The TPH content of ARB-5241 was comparable to that of strawberries, used asa reference (3.35 mg g−1). For oca tubers, the TPHranged from 0.71 to 1.32 mg g−1 (Fig. 3). GenotypesO-018-83, MU-31 and AAQ-5477 showed thehighest TPH values with 1.32, 1.31 and 1.26 mg g−1,

J Sci Food Agric 86:1481–1488 (2006) 1483DOI: 10.1002/jsfa

D Campos et al.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

-300

200

700

1200

1700

2200TPH TBX HAC

U-0

9-01

24

AM

M-5

142

CL

C-0

53

CL

C-0

04

AJA

-527

5

AQ

P-5

454

MH

-296

U-0

9-01

22

U-1

08-8

4

U-0

34-8

3

DP

A-0

3-43

DP

A-0

3-23

PR

OD

EK

ON

-14

PR

OD

EK

ON

-16

AR

V-5

337

TP

H o

r T

BX

(m

g g

-1)

HA

C (

ug

TE

g-1

)

Figure 4. Total phenolic content (TPH), betaxanthines (BTX) andhydrophilic antioxidant capacity (HAC) of ulluco tubers (Ullucustuberosus Caldas).

respectively. Oca genotypes with the highest TPHwere purple, whereas lower TPH values were foundin yellow genotypes. Ulluco tubers had the lowestTPH content compared with the other crops studied.The TPH of ulluco tubers ranged from 0.41 to0.77 mg g−1 (Fig. 4), with genotypes U-108–84, AJA-5275 and U-09-0122 showing the highest phenoliccontent with 0.77, 0.63 and 0.61 mg g−1, respectively.The phenolics identified previously in ulluco includerutin, narcissin and kaempferol 3-O-(2′,6′-di-O-α-L-rhamnopyranosyl)-β-D-glucopyranoside.21 In general,the TPH range values for the crops studied followedthe descending order mashua ≥ native potato ≥ oca >

ulluco.

Total anthocyanins (ACY)The ACY in the pigmented native potato tubers rangedfrom 0.08 to 0.8 mg g−1 and was only present in somegenotypes (Fig. 1). Genotypes 704 463 and 707 132showed higher levels of ACY with 0.8 and 0.59 mg g−1,respectively. The ACY values were comparable tothose reported previously20 for purple flesh potatoes,which ranged from 0.11 to 0.6 mg g−1, and higher thanthe ACY content in strawberries (0.4 ± 0.13 mg g−1,in this study). The high ACY content of nativepotatoes indicates that these red and purple flesh tubergenotypes may be a novel source of natural colorants.The ACY may constitute a significant fraction of theTPH. For example, the ACY/TPH ratio was highest(∼0.38) in genotype 704 463 and ranged between 0.09and 0.38 for all pigmented native potato samples.

The ACY in mashua pigmented genotypes rangedfrom 0.5 to 2.05 mg g−1 (Fig. 2). This ACY contentis higher than those reported22 for radish cv.Fuego (0.30 mg g−1), red cabbage (0.25 mg g−1),strawberry (0.15–0.3 mg g−1) and red raspberry(0.3–0.4 mg g−1). The highest amount was found ingenotype DP-02-24, which is in the range reported forblueberry,23 1.38–3.85 mg g−1. The ACY in mashuatubers seems to be a significant component of theTPH in the pigmented genotypes. For example, theACY/TPH fraction ranged between ∼0.3 and 0.67.

For oca pigmented tubers, the ACY ranged from0.14 to 1.3 mg g−1 (Fig. 3). The highest amount wasfound in MU-31 and seems to be concentrated mostlyin the tuber skin. For pigmented tubers, the ACY/TPHfraction ranged from 0.14 to 0.9. The ACY/TPHfraction was low (0.14) in genotype PICA-HI-92OC,owing to the relatively high TPH values, whereas forgenotype MU-31, the fraction was high (0.9) owingto the high ACY content. On the other hand, ourresults showed that ulluco tubers do not containanthocyanins. In general, the ACY range values ofthe crops studied followed the descending ordermashua ≥ oca ≥ native potato. The presence or not ofACY and the differing proportions of ACY observedindicate that the genotypes studied in each type ofcrop would have different qualitative and quantitativephenolic profiles.

Total carotenoids (TCT)The TCT in native potatoes ranged from 2 to5 µg g−1 (Table 1), these values being higher thanthose reported previously for commercial varieties,24,25

which are in the range 0.27–3.43 µg g−1. The705 468 genotype had the highest carotenoid contentwith 5 µg g−1. Interestingly, only one red pigmentedgenotype (705 024) presented carotenoids, phenolicsand anthocyanins in the same tuber. In general, tuberswith carotenoids tended visually to correlate withflesh yellow intensity. Violaxanthin is considered themain potato carotenoid reported in previous studies.followed by lutein, antheraxanthin and others.26

Lutein has attracted interest since high blood serumlevels are correlated with a reduced risk for age-relatedmacular degeneration in humans (AMD).25,27

In mashua tubers, the TCT ranged from 1 to25 µg β-carotene g−1 (Table 2). Genotypes ARB-5576, M6COL2C and DP-0207 had the highestcarotenoid contents with 25, 23 and 21 µg g−1,respectively. The carotenoid content of mashuatubers is relatively high compared with commercialpotatoes24,25 and those found in native potatoes in thisstudy, and also the carotenoid content in papaya28

(4.08 µg g−1). However, TCT in mashua are lowerthan those reported28 for tomato (56–210 µg g−1),mango (74.3 µg g−1) and carrot (90 ± 16 µg g−1 in thisstudy).

For oca tubers, the TCT ranged from 2 to 25 µgβ-carotene g−1 (Table 3). Genotypes AJA-5245 andAJA-5270 had the highest carotenoid content with25 and 17 µg β-carotene g−1, respectively. GenotypeAAQ-5477 showed high ACY, TCT and TPH contentamong all oca genotypes studied. In the case of ulluco,the results indicate that these tubers do not containcarotenoids in the pigmented yellow tissue. Accordingto the present results, the carotenoid content rangevalues for the crops studied followed the descendingorder mashua ≈ oca ≥ native potato.

Betaxanthins (TBX) and betacyanins (TBC)By using absorption spectra, the aqueous solubilityproperties and the response to alkaline (change to

1484 J Sci Food Agric 86:1481–1488 (2006)DOI: 10.1002/jsfa

Antioxidants in Andean tuber crops

Table 1. Total carotenoid content (TCT) and lipophilic antioxidant capacity (LAC) in 15 native potato (Solanum sp.) genotypes

GenotypeCommon

name SpeciesColor of

flesh/skinaMoisture

(%)bTCT

(µg g−1)bLAC

(µg TE g−1)b

707 132 Yana Shiri S. × curtilobum P/P 78 ± 3.9f ND 321 ± 15e,f

705 024 Pachacona S. tuberosum subsp. tuberosum P/P 76 ± 1.6e,f 4 ± 0.1d,e,f 211 ± 7c,d,e

705 841 Unknown S. tuberosum subsp andigena R/R 75 ± 1.0d,e,f ND 152 ± 35c,d

704 463 Gogon Magash S. tuberosum subsp andigena P/P 75 ± 1.2d,e,f ND 131 ± 26c,d

703 286 Yana Pituwayaka S. stenotomum P/Y 74 ± 0.7c,d,e,f ND 357 ± 30f

703 825 China Runtush S. goniocalyx Y/Y 72 ± 3.9c,d,e,f 4 ± 1.3d,e,f 133 ± 53c,d

705 263 Chaucha S. phureja Y/Y 72 ± 4.9c,d,e,f 4 ± 1d,e,f 239 ± 102c,d,e,f

703 985 Runtu S. goniocalyx Y/Y 70 ± 0.7c,d,e 4 ± 1e,f 155 ± 82c,d

704 353 Puna S. tuberosum subsp andigena Y/Y 70 ± 2.9c,d,e 2 ± 0c,d 199 ± 76c,d,e

702 961 Garhuash Pashon S. goniocalyx Y/Y 70 ± 0.5c,d,e 2 ± 1c,d,e 170 ± 40c,d

705 468 Huasa Amarilla S. goniocalyx Y/Y 69 ± 3.3c,d,e 5 ± 1f 249 ± 30c,d,e,f

706 819 Tablera Jubalena S. tuberosum subspp. andigena Y/Y 69 ± 1.7c,d,e 4 ± 0d,e,f 115 ± 28c

703 279 Seda Monilla S. goniocalyx Y/Y 69 ± 3.0c,d,e 3 ± 1c,d,e,f 125 ± 19.21c,d

704 481 Amarilla S. goniocalyx Y/Y 68 ± 1.5c,d 3 ± 0c,d,e,f 265 ± 28d,e,f

703 844 Anil Papa S. stenotomum Y/Y 66 ± 2.8c 4 ± 1d,e,f 361 ± 13f

a P, purple; Y, yellow; R, red.b Mean ± SD (n = 3). Means within a column with the same superscript letter are not significantly different (α = 0.05). ND, not detected.

Table 2. Total carotenoid content (TCT) and lipophilic antioxidant

capacity (LAC) in 11 mashua (Tropaeolum tuberosum Ruiz & Pavon)

genotypes

Genotype

Color offlesh/skina

Moisture(%)b

TCT(µg g−1)b

LAC(µg TE g−1)b

AVM-5562 Y/Y 92 ± 1.6c 14 ± 3d,e 217 ± 37.7e,f

DP-02-07 Y/Y 90 ± 1.0c,d 21 ± 3d,e 221 ± 13.0e,f

DP-02-15 Y/Y 90 ± 1.9c,d,e 17 ± 1d,e 186 ± 9.6d,e

M6COL2C Y/Y 90 ± 1.6c,d,e 23 ± 6e 279 ± 15.0f

DP-02-23 Y/Y 90 ± 1.6c,d,e 13 ± 6d,e 93 ± 22.85c

AGM-5109 P/Y 89 ± 1.2c,d,e 18 ± 0d,e 128 ± 56.1c,d

DP-02-03 Y/Y 87 ± 2.8c,d 21 ± 4d,e 224 ± 24.7e,f

DP-02-24 P/P 87 ± 1.3c,d ND NDARB-5576 Y/Y 87 ± 1.4c,d 25 ± 9e 218 ± 12.8e,f

ARB-5241 P/Y 85 ± 1.7c 10 ± 1c,d 202 ± 19.3d,e,f

ARV-5366 Y-P/P 86 ± 0.6c 1 ± 0c 225 ± 44e,f

a P, purple; Y, yellow; R, red.b Mean ± SD (n = 3). Means within a column with the same superscriptletter are not significantly different (α = 0.05). ND, not detected.

a stable yellow color) and acidic conditions (colordisappearance on boiling),29 it was found that thepigments in ulluco correspond to betalains in the baseform of betacyanins (red pigment, λmax 535–540 nm)and the acid form of betaxanthins (yellow pigment,λmax 453–475 nm). The results indicate that TBXin ulluco tubers ranged from 22 to 96 µg g−1 (Fig. 4).The highest amount was found in genotypes CLC-004and U-09-0122 with 96 and 83 µg g−1, respectively.In general, these pigments have been shown to bea potential source of both natural antioxidants andnatural colorants in Amaranthaceae.30 Only genotypeAQP-5454, which corresponds to a red skin ullucotuber- presented betacyanins (TBC = 64 µg g−1).

Hydrophilic antioxidant capacity (HAC)The HAC obtained for native potatoes ranged from860 to 3780 µg TE g−1(Fig. 1). Genotypes 705 841,

Table 3. Total carotenoid content (TCT) and lipophilic antioxidant

capacity (LAC) in 14 oca (Oxalis tuberosa Molina) genotypes

Genotype

Color offlesh/skina

Moisture(%)b

TCT(µg g−1)b

LAC(µg TE g−1)b

PICA-HI-92OC

Y-R/P 82 ± 1.4c,d,e 8 ± 0c,d 165 ± 23c,d,e

O-116-84 Y-P/P 80 ± 1.7c,d 7 ± 1c,d 126 ± 44c,d

O-266-85 Y/Y 82 ± 2.0c,d,e 13 ± 2d,e 164 ± 39c,d,e

ARB-5054 Y/Y 84 ± 1.5c,d,e 14 ± 1d,e 101 ± 4c,d

ATT-5466 Y-R/P 84 ± 0.7c,d,e 5 ± 1c,d 219 ± 33d,e,f

COC-537 Y-P/P 85 ± 2.6e 2 ± 1c 268 ± 41e,f

MU-31 Y-P/P 85 ± 0.8e 3 ± 1c 76 ± 24c

AMM-5223 Y-R/P 80 ± 1.0c 3 ± 1c 69 ± 58c

AJA-5245 Y/Y 84 ± 0.2c,d,e 25 ± 4f 127 ± 34c,d

AAQ-5477 Y-P/P 82 ± 0.8c,d,e 11 ± 9c,d,e 320 ± 59f

AJA-5270 Y/Y 84 ± 0.8c,d,e 17 ± 2e,f 90 ± 25c,d

O-018-83 P-Y/P 85 ± 1.3d,e 3 ± 1c 181 ± 30c,d,e

DP-01-46 Y-P/P 81 ± 1.3c,d,e 3 ± 1c 135 ± 79c,d,e

AJA-5252 Y/Y 84 ± 2.4d,e 10 ± 2c,d,e 107 ± 72c,d

a P, purple; Y, yellow; R, red.b Mean ± SD (n = 3). Means within a column with the same superscriptletter are not significantly different (α = 0.05).

704 463 and 707 132 showed the highest HAC valueswith 3780, 3369 and 2473 µg TE g−1, respectively.Potatoes are known to contain water-soluble phenolicantioxidants that act as radical scavengers such asquercetin and chlorogenic acid.31 We found thatincreasing TPH in tubers showed a trend of increasingHAC values (Fig. 1), despite the low correlationbetween HAC and TPH (r2 = 0.45, P = 0.00). Ahigh correlation between HAC and TPH for differentgenotypes in a certain type of crop would only bepossible if the genotypes have phenolic compoundswith similar antioxidant properties (e.g. similar proton-donor properties). Since the correlation obtainedis low, we hypothesize that the studied nativepotatoes had different phenolic profiles (qualitative,

J Sci Food Agric 86:1481–1488 (2006) 1485DOI: 10.1002/jsfa

D Campos et al.

quantitative or both) with different antioxidantproperties. Furthermore, when HAC is expressed ona phenolic basis (defined as the HAC/TPH ratio inthis study), this specific HAC ranged from 1066 to2294 mg TE g−1 chlorogenic acid equiv. This indicatesthat the phenolics present in some of the genotypesshowed a higher antioxidant capacity to stabilize agreater number of free radicals than others. Thespecific HAC values obtained are comparable to thosereported previously for blueberries (685–1966 mg TEg−1 phenolics) and higher than for blackberries32

(532 mg TE g−1 phenolics).For mashua tubers, the HAC ranged from 955 to

9800 µg TE g−1 (Fig. 2). Genotypes ARB-5241, DP-02-24 and ARV-5366 showed the highest HAC valueswith 9800, 9309 and 7867 µg TE g−1, respectively.Our results indicate that mashua genotype ARB-5241 is comparable to blueberry32 (cultivar premier,HAC = 9572 µg TE g−1), which is considered one ofthe fruits with the highest antioxidant capacity.33–35

Again, for mashua tubers the antioxidant capacitycould be related to the ACY and TPH contentaccording to Fig. 2. The low ACY correlation withHAC (r2 = 0.48, P = 0.11); and the higher correlationbetween TPH and HAC (r2 = 0.84, P = 0.00) aremost likely due to the presence of different phenolicprofiles in mashua tubers, as in the case of nativepotatoes. When the HAC is expressed on a phenolicbasis, the values obtained ranged from 729 to 3052 mgTE g−1 chlorogenic acid equiv., which could berelated to these different phenolic profiles. The highestvalues were found in genotypes DP-02-24 (3052 mgTE g−1 chlorogenic acid equiv.) followed by ARB-5241 (2908 mg TE g−1 chlorogenic acid equiv.). Thespecific HAC values for mashua are higher than thoseobserved for blueberries.32

In the case of oca tubers, HAC ranged from 1637to 4771 µg TE g−1 (Fig. 3). Genotypes O-018-83,AAQ-5477 and MU-31 had the highest HAC valueswith 4771, 4396 and 3994 µg TE g−1, respectively.Genotype O-018-83 had a HAC value of ∼70%of that in Rabbiteye blueberry cultivars with highHAC (average 6900 µg TE g−1).32 Once again, theantioxidant capacity could be related mainly to theTPH content (Fig. 3). For oca tubers, the correlationbetween HAC and TPH was high and significant(r2 = 0.75, P = 0.00), indicating in this case that thephenolics present in oca tubers could have similarantioxidant properties. When HAC is expressed ona phenolic basis, a range from 1935 to 3614 mgTE g−1 chlorogenic acid was found, which is higherthan those reported for blueberries and blackberries.32

These specific HAC values are comparable to thosefound in mashua, indicating the potential that thesecrops have as a novel source of phenolic compoundswith high antioxidant capacity.

Ulluco tubers presented a HAC range from 483 to1524 µg TE g−1 (Fig. 4). Genotypes U-108-84, AJA-5275 and U-034-83 showed the highest HAC valueswith 1524, 1372 and 1183 µg TE g−1, respectively.

When HAC is expressed on a phenolic basis, thehighest specific HAC among the ulluco genotypes wasfound in genotype DP-03-43 with 2322 mg TE g−1

chlorogenic acid equiv. A low correlation betweenHAC and TPH was observed (r2 = 0.64, P = 0.00),most likely due to different phenolic profiles amongulluco cultivars. In general, the trend observed wasan increased antioxidant capacity for genotypes withhigher phenolic content (Fig. 4). Among the phenolicspresent in ulluco,21 the flavonoid kaempferol has beenreported to exhibit strong antioxidant activity36,37 andto inhibit hemolysis, lipid peroxidation and superoxideradical generation.38 On the other hand, HAC andTBX were not correlated, suggesting that the TBXdoes not contribute to the HAC in ulluco. Accordingto the results obtained, the HAC range values forthe crops studied followed the descending ordermashua ≥ oca ≥ native potato ≥ ulluco.

Lipophilic antioxidant capacity (LAC)In native potatoes, the LAC ranged from and 115 to361 µg TE g−1 (Table 1). Genotypes 703 844, 703 286and 707 132 showed the highest LAC values with 361,357 and 321 µg TE g−1, respectively. Interestingly, twogenotypes lacking carotenoids showed LAC propertiessuggesting the presence of lipophilic compounds otherthan carotenoids with radical scavenging activity.Among the native potatoes, genotypes 705 841,704 463 and 707 132 showed the highest totalantioxidant capacity (lipophilic + hydrophilic) with3932, 3500 and 2794 µg TE g−1, respectively. Thelipophilic fraction of native potatoes contributed only3.7–24.5% of the total antioxidant capacity values.

For mashua tubers, the LAC ranged from 93 to279 µg TE g−1 (Table 2). Genotypes M6COL2C,ARV-5366 and DP-02-03 showed the highest LACvalues with 279, 225 and 224 µg TE g−1, respectively.In mashua, LAC was observed only in genotypes thathad carotenoids. Genotypes ARB-5241 and AGM-5109 presented LAC values (1346 and 1163 µg TE g−1

DW, respectively) comparable to tomato39 (1425 µgTE g−1 DW). The correlation between LAC andTCT was very low (r2 = 0.14, P = 0.058), suggestingthat the carotenoid profiles between genotypes weredifferent. In general, genotypes ARB-5241, DP-02-24 and ARV-5366 had the highest total antioxidantcapacity (lipophilic + hydrophilic) with 10 002, 9309and 8092 µg TE g−1, respectively. The lipophilicfraction contributed 2–19% to the total antioxidantcapacity values for mashua tubers.

The LAC of oca tubers ranged from 69 to 320 µg TEg−1, with genotypes AAQ-5477, COC-537 and ATT-5466 showing the highest LAC values with 320, 268and 219 µg TE g−1, respectively (Table 3). There is nocorrelation between LAC and TCT in oca tubers, mostlikely owing to different profiles of carotenoids presentin the different genotypes. Genotypes O-018–83,AAQ-5477 and MU-31 showed total antioxidantcapacities (lipophilic + hydrophilic) of 4952, 4716 and4070 µg TE g−1, respectively. The lipophilic fraction

1486 J Sci Food Agric 86:1481–1488 (2006)DOI: 10.1002/jsfa

Antioxidants in Andean tuber crops

A B

DC

Figure 5. Andean tubers with high antioxidant capacity: (A) native potato 707 132; (B) mashua ARB-5241; (C) oca O-018-83; (D) ulluco U-108-84.

contributed 1.9–10.2% to the total antioxidantcapacity values for oca tubers.

For ulluco tubers, our results indicate that they donot present LAC for any of the genotypes studied. Ingeneral, the LAC value ranges obtained for potatoes,mashua and oca tubers were similar. The totalantioxidant capacity of these crops is mainly attributedto the phenolic compounds, which are present inhigher amounts than carotenoids.

CONCLUSIONSThis survey of Andean tuber germplasm for antiox-idant compounds confirms that each type of cropstudied can be considered as an excellent source ofdietary phytochemicals (Fig. 5). The phytochemicalsresponsible for the antioxidant capacity are relatedto the phenolics and carotenoids present in eachtuber. The antioxidant values of these tuber cropsare higher than or comparable to known sources ofnatural antioxidants, such as blueberries, indicatingthat these crops have the potential to be consideredas important novel sources of nutraceuticals. TheseAndean tuber crops hold great promise for producingnew and unique healthy, functional products for thebenefit of producers and consumers.

ACKNOWLEDGEMENTSCarla Rios, Daniel Segura and Katherine Alejandroare thanked for technical assistance.

REFERENCES1 Loliger J, The use of antioxidants in foods, in Free Radicals and

Food Additives, ed. by Arouma OI and Halliwell B. Taylor andFrancis, London, pp. 121–150 (1991).

2 Liu RH, Health benefits of fruit and vegetables are from additiveand synergistic combinations of phytochemicals. Am J ClinNutr 78:517S–520S (2003).

3 Boyer JB and Liu RH, Apple phytochemicals and their healthbenefits. Nutr J 3:5 (2004). Available: http://www.nutritionj.com/content/3/l/5 [15 April 2005].

4 Arts ICW and Hollman PCH, Polyphenols and disease risk inepidemiologic studies. Am J Clin Nutr 81:317S–325S (2005).

5 Ames BM, Shigena MK and Hagen TM, Oxidants, antioxidantsand the degenerative diseases of aging. Proc Natl Acad Sci USA90:7915–7922 (1993).

6 Gey KF, The antioxidant hypothesis of cardiovascular dis-ease: epidemiology and mechanisms. Biochem Soc Trans18:1041–1045 (1990).

7 Steinberg D, Parthasarathy S, Carew TE, Khoo JC and Witz-tum JL, Beyond cholesterol: modification of low-densitylipoprotein that increased its atherogenicity. N Engl J Med320:915–924 (1989).

8 Steinberg D, Antioxidants and atherosclerosis: a currentassessment. Circulation 84:1420–1425 (1991).

9 Liu RH, Potential synergy of phytochemicals in cancerprevention: mechanism of action. J Nutr 134:3479S–3485S(2004).

10 Cook NC and Samman S, Flavonoids – chemistry, metabolism,cardioprotective effects and dietary sources. Nutr Biochem7:66–76 (1996).

11 King SR and Gershoff SN, Nutritional evaluation of threeunderexploited Andean tubers: Oxalis tuberosa (Oxalidaceae),Ullucus tuberosus (Basellaceae) and Tropaeolum tuberosum(Tropaeolaceae). Econ Bot 41:503–511 (1987).

12 Gross R, Koch F, Malaga I, de Miranda AF, Schoeneberger Hand Trugo LC, Chemical composition and protein qualityof some local Andean food sources. Food Chem 34:25–34(1989).

13 Swain T and Hillis W, The phenolic constituents of Prinusdomestica. I. The Quantitative analysis of phenolic con-stituents. J Sci Food Agric 10:63–68 (1959).

14 Fuleki T and Francis F, Quantitative methods for anthocyanins.1. Extraction and determination of total anthocyanin incranberries. J Food Sci 33:72–77 (1968).

15 Abdel-Aal ES, A rapid method for quantifying total antho-cyanins in blue aleurone and purple pericarp wheats. CerealChem 76:350–354 (1999).

J Sci Food Agric 86:1481–1488 (2006) 1487DOI: 10.1002/jsfa

D Campos et al.

16 Talcott S and Howard R, Phenolic autooxidation is responsiblefor color degradation in processed carrot puree. J Agric FoodChem 47:2109–2115 (1999).

17 Cai YZ and Corke H, Production and properties of spray-driedAmaranthus betacyanin pigments. J Food Sci. 65:1248–52(2000).

18 Arnao M, Cano A and Acosta M, The hydrophilic andlipophilic contribution to total antioxidant activity. Food Chem73:239–244 (2001).

19 Awika JM, Rooney LW, Wu X, Prior RL and Cisneros-Zevallos L, Screening methods to measure antioxidant activityof sorghum (Sorghum bicolor) and Sorghum products. J AgricFood Chem 51:6657–6662 (2003).

20 Reyes LF and Cisneros-Zevallos L, Wounding stress increasesthe phenolic content and antioxidant capacity of purple-flesh potatoes (Solanum tuberosum L.). J Agric Food Chem51:5296–5300 (2003).

21 Dini A, Rastrelli L, Saturnino P and Schettino O, Minorcomponents in food plants – Note I. Flavonol glycosides fromUllucus tuberosus. Boll Soc Ital Biol Sper 67:1053–1058 (1991).

22 Giusti M and Wrolstad R, Radish anthocyanin extract as anatural red colorant for maraschino cherries. J Food Sci61:688–694 (1996).

23 Cevallos-Casals BA and Cisneros-Zevallos L, Stoichiometricand kinetic studies of phenolic antioxidants from andeanpurple corn and red-fleshed sweetpotato. J Agric Food Chem51:3313–3319 (2003).

24 Lu W, Haynes K, Wiley E and Clevidence B, Carotenoidcontent and color in diploid potatoes. J Am Soc Hortic Sci126:722–726 (2001).

25 Iwanzik W, Tevini M and Hilbert R, Carotinoidgehalt und-zusammensetzung verschiedener deutscher Kartoffelsortenund deren Bedeutung fur die Fleischfarbe der Knolle. PotatoRes 26:149–162 (1983).

26 Muller H, Determination of the carotenoid content in selectedvegetables and fruits by HPLC and photodiode arraydetection. Z Lebensm Unters Forsch 204:88–94 (1997).

27 Olmedilla B, Granado F, Blanco I, Vaquero M and Cajigal C,Lutein in patients with cataracts and age-related maculardegeneration: a long-term supplementation study. J Sci FoodAgric 81:904–909 (2001).

28 Gross J, Pigments in Fruits. Academic Press, London (1987).29 Nielson LR and Harley SM, Chemataxonomy: simple test for

distinguishing between anthocyanins and betacyanins. J BiolEduc 30:88–90 (1996).

30 Cai Y, Sun M and Corke H, Antioxidant activity of betalainsfrom plants of the Amaranthaceae. J Agric Food Chem51:2288–2294 (2003).

31 Al-Saikhan MS, Howard LR and Miller JC, Antioxidant capac-ity and total phenolics in different genotypes of potato(Solanum tuberosum, L.). J Food Sci 60:341–347 (1995).

32 Sellappan S, Akoh CC and Krewer G, Phenolic compoundsand antioxidant capacity of Georgia-grown blueberries andblackberries. J Agric Food Chem 50:2432–2438 (2002).

33 Prior RL, Cao G, Martin A, Sofic E, McEwen J, O’Brien C,et al, Antioxidant capacity as influenced by total phenolic andanthocyanin content and variety of Vaccinium species. J AgricFood Chem 46:2686–2693 (1998).

34 Wang H, Cao GH and Prior RL, Total antioxidant capacity offruits. J Agric Food Chem 44:248–251 (1996).

35 Cao G, Sofic E and Prior RL, Antioxidant capacity of teaand common vegetables. J Agric Food Chem 44:3426–3431(1996).

36 Jung MJ, Chung HY, Choi JH and Choi JS, Antioxidantprinciples from the needles of red pine, Pinus densiflora.Phytother Res 17:1064–1068 (2003).

37 Cho EJ, Yokozawa T, Rhyu DY, Kim SC, Shibahara Nand Park JC, Study on the inhibitory effects of Koreanmedicinal plants and their main compounds on the 1,1-diphenyl-2-picrylhydrazyl radical. Phytomedicine 10:544–551(2003).

38 Ng TB, Liu F, Lu Y, Cheng CH and Wang Z, Antioxidantactivity of compounds from the medicinal herb aster tataricus.Comp Biochem Physiol C Toxicol Pharmacol 136:109–115(2003).

39 Pellegrini N, Re R, Yang M and Rice-Evans C, Screeningof dietary carotenoids and carotenoid rich fruit extractsfor antioxidant activities applying ABTS radical cationdecolorization assay. Methods Enzymol 291:379–389 (1998).

1488 J Sci Food Agric 86:1481–1488 (2006)DOI: 10.1002/jsfa