Some varieties have been studied and are appreciated for their nutritional properties because they are rich in carbohydrates, lipids and proteins (2-5). Variation in carbohydrate composition (starch content and free sugar) of the fruit of the soft and firm varieties of Artocarpus heterophyllus Lam have been reported (5). Free sugars and fatty acid of different parts of jack fruit (A.. heterophyllus) have been isolated (6). Functional properties (A.. heteropyllus) seed flour and the baking properties of A.. altilis flours have been studied (1-7). Studies on the chemical composition for the seedless and seeded varieties have shown a protein content of 15.10 and 1.70 g /100g, fat 29.0 and 0.30 g /100 g and moisture 20.20 and 70.80 g/100g respectively (2, 7).

Bread prepared with 30 % of A.. altilis flour did not show any significant difference with bread made with 100 % wheat flour, and a substitution of 10 % gave the best pasting characteristics (1). Nutritional value of composite flours and of the dried meal were evaluated (8), and some of the starch properties obtained from immature fruits have been studied (3).

Although the major commercial sources of starch are cereal grains seeds (maize, wheat, rice), tubers (potato) and certain roots (sweet potato, cassava), several potential non conventional sources have been reported (9-11).One of the potential alternative sources of commercial starch could be the seedless fruit of A.. altilis, due to its high yield.

Literature survey reveals little work on the physical characteristics, chemical properties and digestibility in vitro of A.. altilis starch. Since the chemical composition and physical characteristics of a given starch are typical of its biological origin (12), the purpose of the present study was to characterize this starch and provide information on its composition, morphology, selected physico-chemical characteristics, and assessment of its pasting properties and in vitro digestibility.

MATERIALS AND METHODS
Materials
Fifty two samples of immature fruits of Artocarpus, seedless variety, were collected in Margarita Island, Venezuela. All analyses were performed with analytical grade reagents, and results expressed as mean ± standard deviation (SD) of n = 3. 

Morphology, chemical composition and physical characteristics of fruits samples
Fruits of A. altilis were analyzed for their physical attributes: morphology, size and weight. Size was evaluated on a representative sample of fruits (100), by measuring length and width using a measuring tape. Fruits were peeled manually and the peeled material (edible portion) was weighed. Yield was calculated from the following equation:

Yield = (wt of edible portion)/wt of whole fruit) x 100

Peel fraction = 100 – Yield

Peeled fruits were sliced and dried in a Labline Imperial oven at 45°C for 24 h, followed by milling in a Braun food processor to obtain a flour or meal which was sifted through a 60 mesh sieve and stored at room temperature in plastic bags.

Chemical composition of the flour was determined by the AOAC official methods(13). Soluble and insoluble fiber by Mañas and Saura-Calixto method (14).

Starch isolation and chemical composition
Starch extraction was based on that described by Rincón et al (10). Briefly, the seedless fruits of Artocarpus were washed, peeled and pulp suspended in large quantities of distilled water, blended for 1 min in a Oster blender, at medium speed. The homogenate was filtered through a muslin bag. The filtrate was centrifuged and the supernatant was decanted. The starch was washed several times with distilled water and dried at 45°C for 24 h, and stored at room temperature in plastic bags. Quantitative evaluation of moisture, ash and nitrogen were performed by the standard methods of the AACC (15). Total lipids by Schoch (16) with previous hydrolysis. The pH and acidity by Smith (17).

Apparent amylose content
Quantitative estimation of apparent amylose was determined by the method of Ratnayake et al (18) slightly modified. Starch (20 mg, db) was dissolved in 90% dimethylsulfoxide (3 ml) in 10 ml screw-cap reaction vials. The content of the vials was vigorously mixed for 2 min and then heated in a water bath (with intermittent shaking) at 85°C for 15 min. The vials were then cooled to ambient temperature, and the content diluted with water to 25 ml in a volumetric flask and sonicated for 10 min. 1.0 ml of the diluted solution was mixed with 40 ml of water and 5 ml I₂/KI solution (0.0025 M I₂ and 0.0065M KI) and then adjusted to a final volume of 50 ml. The contents were allowed to stand for 15 min at ambient temperature, before absorbance measurements at 600 nm. The amylopectin content was obtained by difference.

Scanning electron microscopy
Granule morphology was determined by scanning electron microscopy (SEM); using a Hitachi electron microscope, model S 2400 at 20 kV. Samples were mounted, and the sample holder was sealed with silver paint and coated with gold/palladium at 8-10mA for 10 min, under low pressure less than 10 torr.

Swelling power, solubility and water absorption
Swelling, water absorption and solubility determinations were carried out at a temperature range of 60° to 90°C according to Rincón and Pérez methods(19). An appropriate amount of the starch (2 g d.b) was accurately weighed and quantitatively transferred with distilled water into a dry, 500 ml Pyrex flat bottom flask with three angle necks and t-joints. Sufficient distilled water was added to give 200 ml total suspension volume. A magnetic stir bar was placed inside the flask. The central neck was connected to a reflux condenser, which in turn was attached to a support stand by a single clamp. In the right neck a thermometer was placed attached to a rubber stopper and the left neck was covered with a rubber stopper. The system was placed on a hot, porcelain top, magnetic stirrer and heated from 65ºC up to 95ºC at an uniform rate (1.5ºC/min increments) under constant stirring (75 rpm). At each 10ºC temperature increment, an aliquot of 10 ml was taken and placed into a tared centrifuge tube, and centrifuged 15 minutes at 2200 rpm. Using a glass tube attached to a clean dry suction flask, all supernatant was removed from settled paste. Each solution was transferred to a tared, 50-ml nickel evaporating dish and evaporated to dryness on a steam bath. The dishes were dried for 4 hours in a vacuum oven at 65ºC, cooled in a desiccator, and weighed. The centrifuge tubes with settled paste were weighed. The following calculations were used:

W₁= [ Starch weight (dw)/ Starch weight (fw) + 200] x 100

W₂= [ A x W₁] / 100

W₃= W₂ - b

%SS= Soluble Solids = [b / W₂] x 100

WA = Water Absorption ( g water/ g starch) = [a - W₃] / W₃

SP = Swelling Power = [a x 100] / W₂ x (100 - % S.S.)]

W₁= % Starch (dw) in suspension

W₂= Starch content in each aliquot

W₃= Residual starch in each aliquot (settled paste)

A= Weight of aliquot (g)

a= Weight of settled paste

b= Weight of dry residue

dw= Dry weight

fw= fresh weight

Results used for calculations were means of triplicate measurements.

Brabender viscography
Brabender viscographic amylogram of a 6.0 g/100 mL slurry was obtained on a Brabender Micro visco-amylograph following the procedure and condition described by the manufacturer. The aqueous starch suspension was heated from 50°C to 95°C, kept at this temperature for 5 min and then cooled to 50°C and held at this temperature for 1 min. Speed rotor was fixed at 75 rpm and the rate of heating and cooling was 7.5°C.min^-1 throughout the range of gelatinization, holding and cooling steps. Amylogram parameters and determination of breakdown and setback points calculated by viscograph program for Windows N° 72300 (software) and analyzed according to Rasper (20). The peak of viscosity (P), hot paste viscosity (H, viscosity after 5 min stirring at 95°C) and cold paste viscosity (C, viscosity after cooling to 50°C) were recorded. Breakdown (BD) is the viscosity difference between the maximum viscosity (P) and the hot paste viscosity (H). Setback (SB) is the viscosity difference between viscosity at 50°C (C) and viscosity at 95°C (P). Additionally, in order to monitor disintegration behavior upon controlled energy input and accompanied tendencies to reconstitute super molecular glucan structures, Brabender viscosity was determined for 10 % (w/v) aqueous starch suspensions, according to Praznik (21), with some modification as follows: a heating program from 30 to 90°C, kept at this temperature for 5 min and then the cooling period started and maintained till the end of the study (22min). 

Light transmittance of starch pastes
The light transmittance of starch pastes was carried out according Hoover et al method (22). Starch pastes were prepared at 1% (w/v) by weighing 50mg (db) and suspended in 5ml of water in screw cap tubes and the pH is adjusted by addition of 0.1N HCl or NaOH as required between pH 2 and 14. The tubes were heated in a boiling water bath, with occasional shaking, for 30 min. After cooling to ambient temperature, the percentage transmittance (%T) at 650 nm was determined against a water blank in a Genesis 2 spectrophotometer.

In vitro of starch digestibility
Estimation of in vitro digestibility by pancreatic a -amylase of A altilis starch was assessed by a colorimetric assay method (23). Percentage of starch hydrolysis was calculated as follows:

% starch hydrolysis = mg glucose x 9,90 x 100

                                   mg starch (drybasis)

Results and discussion
Morphology and physical characteristics of fruits
Breadfruit (A.. altilis) fruits are round with a smooth surface and a diameter in the range of 12 to 35 cm Fig 1. The whitish to pale yellow pulp presented a weight range of 233.7- 656.6 g. The edible portion was 88.64 g/100 g, and a waste (peel) of 11.36g/100 g.

FIGURE 1
Morphological appearance of Artocarpus altilis, seedless variety

Proximal composition
The results are shows in Table 1. The moisture content of the fresh fruits, and the chemical composition of the flour. Carbohydrate content as well as its composition in starch, and fiber: insoluble (lignin, resistant starch, and resistant protein), and soluble (as neutral sugars and uronic acid) are comparable to the results reported for A.. heterophyllus (5). Protein content (15.09 g/100 g) is similar to one reported for A.. heterophyllus (7), but much higher than the one reported for the same specie(1), and for A. communis (8,24). The soluble and insoluble dietary fiber as well as the starch content were comparable to values reported in different varieties of A..heterophyllus (5). The low starch content is probably due to the maturity stage, variety, and different climatic and agronomic conditions. Results for protein and soluble and insoluble dietary fiber are likely to make this fruit of nutritional and physiological importance.

TABLE 1
Chemical composition (g/100 g) and some physical characteristics
of seedless variety of Artocarpus altilis fruit, flour, and starch ¹

Scanning electron microscopy (SEM)
Scanning electron microscopy of A.. altilis starch granules is presented in Figure 2 (a, b). The starch granules have a rounded irregular shape, with some surface indentations, and a width and length diameter range of 4.24 – 6.67 m m and 3.03 – 7.88 m m, respectively. Indentations might be caused by compression of small starch granules or protein bodies during the early development stages in the amyloplast (25). Fissures suggest that the extraction and purification procedure were somewhat drastic.

FIGURE 2
Artocarpus altilis starch granules photomicrograhs

Starch chemical composition
Proximate analysis and yield are shown in Table 1, as well as some of the physical properties of the breadfruit starch. 18.50 g/100 g (dw) yield is somewhat low, nevertheless one has to consider that during the extraction and purification procedure there is some loss. The purity of the starch was judged on the basis of composition and microscopic observation. According to this the starch has a purity of 98.86 g/100. A low protein content of 0.06 g/100 g corroborates this, which is similar values reported (3). Amylose content was higher than the results presented in the literature (3) for two different varieties of Artocarpus (16.4 and 18.2).

Swelling, water absorption and solubility
The results are presented in Table 2. Swelling of starch granules is the first stage in the initiation changes in hydration related properties. The swelling power of the A. altilis starch increased with temperature. This increase was more pronounced within 70 – 80⁰C, and values are higher than the ones reported (3). The swelling power has been related to the associative binding within the starch granule, and apparently, the strength and character of the micellar network is related to the amylose content of the starch, low amylose content, produces high swelling power (26). However, these results are much higher when compared with the ones reported for amaranth (6.57 g/100 g amylose) and corn (17.00 g/100 g amylose) (27). As a result of swelling there is an increment in the solubility, showing the highest value at 90⁰C. These results are also similar to the ones reported for African breadfruit (4). Increase solubility could be attributed to the amylose content, since the solubilized amylose molecules leach from the swelled starch granules.

TABLE 2
Physicochemical properties of native
starch from A. altilis (seedless variety)

Water absorption capacity and gelatinization temperature are specific for each type of starch, and they depend on several factors such as size of granules, amylose/amylopectin ratio, and intra and inter molecular forces. The breadfruit starch presented a lower absorption of water below 70⁰C, this can be explained by the slow dispersion that polymers show when they are added to a solvent and the separation and absorption of moisture by particles do not take place before the swelling occurs. This is due to the degree of intermolecular association within the starch granule (28). Water absorption is also dependent on the size of the starch granule, the smaller the size of the granules, the higher the absorption capacity.

Viscoelastic behavior
In the presence of water and heat, starch granules swell by imbibing water. As the temperature is increased the gelatinization temperature is reached after which a paste is formed. Pasting properties of Artocarpus starch are given in Figure 3 and Table 3. According to Brabender viscoamylograph, the gelatinization temperature of a 6.0 g/100 g of Arthocarpus starch paste was found to be 73.3 °C. This value is higher than that of potato starch (61.6 °C) but much lower than that of maize starch (83.3 °C) (27). The gelatinization temperature depends on the size of the starch granule; small granules are more resistant to rupture and loss of molecular order, so this might explain the relative high gelatinization temperature. After gelatinization, the viscosity of each starch increases markedly, mainly because of lack of water, which acts as a lubricant between the swollen granules. The peak viscosity (P) at any concentration is an important distinguishing feature of a given starch from other species of starch. Artocarpus starch shows a peak viscosity value of 790 BU. Peak viscosity (P) of Artocarpus starch is similar that of Dioscorea starch (781 BU) (29) but higher than that of maize starch (302 BU). Viscosities of maize and potato starches decrease during the isothermal holding (27), while for artocarpus starch viscosity increases during this period. The hot paste viscosity (H) has been attributed to the mixed effect of swollen starch granules, granule fragments, colloidal dispersed starch molecules, molecularly dispersed starch molecules, rate of amylose exudation, and competition between exuded amylose and remaining granules for free water (29). Artocarpus starch maintains its structural integrity under shear and heat as the breakdown in hot paste viscosity is only 4 BU, compared to other starches (27,29). Artocarpus starch could be used in food products that will require continuous heating like those for children and the elderly. When hot starch pasted is cooled, the extent of viscosity increase is governed by the starch retrogradation tendency. This behavior is largely determined by the affinity of hydroxyl groups of one molecule for another. Amylose molecules being randomly dispersed can orient themselves in parallel fashion to form aggregates of low solubility, leading to gel formation (29,30). The cold paste viscosity (C: 1091 BU) of Artocarpus starch was greater that its corresponding peak of viscosity (P:790 BU). Hence it may be assumed that on cooling the viscosity of Artocarpus starch rises due to the high retrogradation tendency of the amylose fraction.

TABLE 3
Viscographic characteristics of 6.0 g/100 ml
A. altilis (seedless variety) native starch. n = 3

FIGURE 3
Brabender pasting curves for A. altilis starch
(seedless variety) at 6.0 % aqueous suspensions

Brabender viscosity was also determined for a 10% suspension even though such a high concentration is atypical, but it was applied because for wheat a reasonable response could be only achieved at such concentration and it was a way to compare Artocarpus starch behavior. This test was introduced to monitor disintegration behavior upon controlled energy input and accompanied tendencies to re-constitute super molecular structures.

The amylograph (Figure 4) shows that maize starch has a similar disintegration in the initial temperature raising period than wheat. At similar temperatures maize and wheat disintegrate in a similar way. Further increase in temperature reduced viscosity for wheat by increasing the mobility of disintegrated components. In the cooling period, mobility of glucans is reduced and thus, viscosity smoothly rises again for wheat and with less intensity for maize.

FIGURE 4
Brabender pasting curves for A. altilis starch (seedless variety)
[A], maize [B], and wheat [C] at 10% aqueous suspensions

Artocarpus starch disintegrates much more strongly which causes an enormous increase in viscosity. Further increase in temperature produces higher viscosity and disintegration takes place more slowly. Super molecular structures of Artocarpus starch are seriously disintegrated, but exhibit a high glucan/glucan interaction potential. In the cooling period viscosity increases indicating a tendency to re-constitute super molecular structures. Similar characteristics were observed for wheat but at lower temperatures.

When comparing these results with those of Praznik et al .(21), performance of Artocarpus starch could be similar to buckwheat starch, which is categorized on the group of mixed short chain branched/long chain branched glucan starches. When comparison is made taking in consideration [BU] ratio T_30ºC/T_90ºC (Table 4) then artocarpus starch (1.39) is similar to quinoa (Chenopodium quinoa)(1.5) that is considered by Praznik et al. (21) a short chain branched glucan starch. Further studies should be performed in order to elucidate the real structure of artocarpus starch glucans.

TABLE 4
Brabender viscograph characteristics of 10g/100 mL
aqueous starch suspensions of wheat, maize and breadfruit

Light transmittance of starch pastes
The results of transmittance (%T) measurement differed at all pH levels. Figure 5. Artocarpus starch showed a lower %T at neutral pH compared to the alkaline pH. The starch chains increase their negative charges and association with increasing pH, this effect occurred rapidly between pH 10 and 14. This result is similar to the one reported Craig et al. (31) When a beam of light passes through native starch granules, a large proportion of the light is reflected back and the starch appears white and opaque due to the surface of the granule being larger than the wavelength of light (31). These authors proposed that the separation of starch chains during gelatinization decreases the reflection ability of starch granules and thus, increases the %T of starch paste. They also showed that amylose-lipid inclusion complexes decrease the %T of starch paste and that %T increases with the degree of swelling. These results suggest that A.. Altilis starch might have lipid complexed amylose since it shows a low %T.

FIGURE 5
Effect of pH on light transmittance of A. altilis starch

In vitro digestibility studies
Digestion of starch samples was performed with a -amylase and the results are shown in Figure 6. Digestibility of Artocarpus and wheat starches, represented as percentage of hydrolysis at 60 min were 93 and 74% respectively. Digestibility of Artocarpus starch was higher than wheat starch. Literature shows that amylose rich starch is difficult to swell or to gelatinize, and it is digested slowly because of higher crystallinity in the structure due to extensive hydrogen bonding (32). Other studies reported that retrogradation of amylose in starch generally suppresses the reaction with amylolitic enzymes (33,34). However, it is believed that the amylose content is not the only factor influencing digestibility and that digestion is a complicated procedure affected by many factors such as amylopectin structure, gelatinizing temperature and phosphorus content besides, the amylose content. A plausible reason for the high digestibility is that the amylopectin in Artocarpus starch might contain a larger amount of longer branched chains, which was suggested by the results of the amylo-viscographic study.

FIGURE 6
Artocarpus and wheat starch digestibility

CONCLUSIONS AND RECOMMENDATIONS
From these results it can be concluded that the A.. altilis (Breadfruit) starch showed a high degree of purity, and has a tendency to retrogradation. The physicochemical and rheological characteristics suggest that this starch could be useful in products that require long heating processes. The excellent digestibility of Artocarpus starch might be advantageous for medical and food use.

Acknowledgements
This research was supported by a grant of the Consejo de Desarrollo Científico y Humanístico (CDCH) [PI 06.30.4358.99], and the Instituto de Investigaciones Farmacéuticas (IIF) of the Faculty of Pharmacy, Universidad Central de Venezuela.

References

1. Eusoso K, Bamiro F. Studies on the baking properties of non wheat flours I. Breadfruit (Artocarpus altilis), Int J Food Sci and Nutr, 1995; 46: 267-273.
   
   

2. Duke J, Alan A. Handbook of proximate analysis tables of higher plants. CRC Press Inc. Boca de Raton, 1986.p. 21
   
   

3. Tumaalii F, Wooton R. Properties of starch isolated from Western Samoan breadfruit using a traditional method. Starch/Starke, 1988; 40: 7-10
   
   

4. Akubor PI. Proximate composition and selected functional properties of African breadfruit and sweet potato flour blends. Plant Food for Human Nutr, 1997; 51:53-60
   
   

5. Rahman MA, Nahar N, Jabbar Mian A, Mosihuzzaman M. Variation of carbohydrate composition of two forms of fruit from jack tree (Arthocarpus heterophyllus L.) with different maturity and climatic conditions. Food Chem, 1999; 65: 91-97
   
   

6. Chowdhury FA, Raman Md. A, Mian A.J. Distribution of free sugars and fatty acids in jackfruit (Artocarpus heterophyllus). Food Chem, 1997; 60: 25-28
   
   

7. Singh A, Kumar S, Singh IS. Functional properties of Jack fruit seed flour. Lebensm Wiss u Technol, 1991; 24: 373-374
   
   

8. Nochera C, Caldwell M. Nutritional evaluation of breadfruit containing composite flours products. J Food Sci, 1991; 57: 1420-1451
   
   

9. Gebre-Mariam T, Ababa A. Some Physicochemical Properties of Dioscorea starch from Ethiopia. Starch/Starke, 1998; 60: 241-246.
   
   

10. Rincón AM, Padilla FC, Araujo C, Tillett, S. Myrosma cannifolia, chemical composition and physicochemical properties of the extracted starch. J Sci Food and Agric, 1999a; 79: 532-536.
    
    

11. Rincón AM, Pérez E, González Z, Rodríguez P. Microstructural changes of Canavalia ensiformis starch modified by thermal methods. Food Sci Tech Int. 1999b; 5:31-40
    
    

12. Zobel H, Stephen A. Starch: structure, analysis, and application. In: Food Polysaccharides and their applications. Ed by A. M. Stephen. Marcel Dekker, Inc. New York, 1995. pp.19-66.
    
    

13. AOAC. Official Methods of Analysis, 15 Ed. Association of Official Analytical Chemists, Washington,DC. 1990
    
    

14. Mañas E, Saura-Calixto F. Ethanolic precipitation: a source of error in dietary fibre determination. Food Chem, 1993; 47: 351-355
    
    

15. AACC. Approved Methods of the AACC. American Association of Cereal Chemists, St. Paul, MN, USA. 1993.
    
    

16. Schoch TJ. Fatty substances in starch. In: Methods in Carbohydrates Chemistry. Ed by Whistler, R.L, Academy Press. New York, 1964a. pp.56-61.
    
    

17. Smith R. Characterization and analysis of starch. In: Starch/Chemistry and Technology. Ed by Whistler, R.L. and Paschall, E.F.,. Academic Press. New York, 1967. pp.569-635
    
    

18. Ratnayake WS, Hoover R, Shahidi F, Perera C, Jane J. Composition, molecular structure, and physicochemical properties of starches from four field pea (Pisum sativum L.) cultivars. Food Chem, 2001; 74: 189-202
    
    

19. Rasper V. Theoretical aspects of amylographology. In The amylograph Handbook . Ed by Shuey W.C, & K.H. Tipples. The American Association of Cereal Chemists. St Paul, 1980. pp 1-6
    
    

20. Praznik W, Mundigler N, Kogler A., Pelzl B, and Huber A. Molecular background of technological properties of selected starches. Starch/Starke 1999; 51:197-211
    
    

21. Hoover R, Swamidas G, Kok LS, and Vasanthan T. Composition and physicochemical properties of starch from pearl millet grains. Food Chem. 1996; 56:355-367
    
    

22. Holn J, Björck I, Asp, NG, Sjöberg, LB, Lundquist, I. Starch availibity in vitro and in vivo after flaking, steam-cooking and popping of wheat. J Cereal Sci, 1985; 3: 193-206
    
    

23. Calzetta AN, Tolaba MP, and Suarez C. Some physical and thermal characcteristics of amaranth starch. Food Sci and Techn Intern 2000;. 6: 371-378
    
    

24. Gebre-Mariam T and Schmidt PC. Isolation and physicochemical properties of enset starch. Starch/Starke, 1996; 48 :208-214
    
    

25. Gebre-Mariam T, Ababa A, and Schmidt PC. Some Physico-chemical Properties of Dioscorea Starch from Etiopia. Starch/Starke, 1998; 50:241-246
    
    

26. Craig SAS, Maningat CC, Seib PA, & Hoseney RC. Starch paste clarity. Cereal Chem. 1989; 68:173-182
    
    

27. Panlasigui LN, Thompson BO, Juliano CM, Perez C.M, Yiu SH, and Greenberg GR. Rice varieties with similar amylose content differ in starch digestibility and glycemic response in humans. A J Clin Nutr. 1991; 54:871-877
    
    

28. Berry CS, Anson KI, Miles MJ, Morris VJ, Russel P.L. Physicochemical characterization of resistant starches from wheat. J Cereal Sci, 1988; 8:203-206.