Symposium

Starch Digestion and Absorption in Nonruminants123 GARY M. GRAY The Gastroenterology Stanford

ABSTRACT

University

Starch

Division and The Digestive Disease Center, School

digestion

of Medicine,

and absorption

Stanfoi4

IMPORTANCE OF THE PHYSICAL STATE OF STARCH

Is aug

mented appreciabRyby physical processing of grain or legume and by heating to 100°Cfor several minutes before fts Ingestion.

Starch,

a potysacchaiide

CA 94305

When starches

composed

are well penetrated

by a polar so

of cil,4-linked glucose units (amylose) and al,4-

lution contiiining salivary and pancreatic a-amylases,

1,6-linked branched stmcture (amylopectin), is cleaved in the duodenal cavity by secreted pancreatic ci-amylase to a disaccharide (maltose), trlsaccharlde (maltotriose), and branched ci-dextrins. These final oligosaccharides

they are cleaved extremely efficiently in the intes ti.nal lumen. Because the food starches are always present in grains and legumes in association with

proteins, many of which are relatively hydrophobic, these polysaccharides tend to be misintained in the

are hydrolyzed efficiently by compllmentaiy action of three hit@

brush boider enzymes at the intestinal

surface: glucoamylase (maltase-gincoamylase,am34oglucosidase), dextrinase

sucrase

(maltase-sucrase)

(isomaltase).

The

final

and

interior of the ingested

a-

monosaccharide

glucose product Is then cotransported Into the en terocyte 75-kDa

along with Na transport

luminal and membrane digestion followed by glucose transport, starch Is assimilated in a veiy efficient manner in nonniminants. J. Nutr. 122: 172-177, 1992.

process

processes,

• atnylopcdlln

•oltgosacchartdase •dlsaccharkiase •a-amylase

• glucose

facilitating

starch

availability

for

water

penetration and consequent a-amylase action are physical processing (2) of the grain or legume and cooking by heating to 100C for several minutes (3, 4). Assimilation of starches is enhanced by a series of

INDEXINGKEY WORDS: ainylose

from water.

access to a-amylase in the intestinal lumen unless they have been physically altered. The principal

step for

overallstarchassimilation.By virtueof this sequential

.

protected

fluids, but even cereals such as wheat may not have

by a specific bnish border

protein In the rate-limiting

particle,

Starches in tubers and legumes are particularly well protected from the polar environment of hmiinal

inducing

cracking

the grains,

converting

to

course flour and milling to fine flour. Cooking wheat

transport

or potatoes alters the starch by converting it from a crystalline

to a gel structure,

cient entry into the luminal

action with the a-amylases

which

promotes

polar solution

effi

for inter

(5). Chilling after cooking

re-alters the polysaccharide's physical state suffi ciently to reduce the digestibility (5). The extent of overall assimilation of polysaccharides from grains or Dietary

carbohydrate,

the least expensive

source

legumes also depends upon the amount of non-starch saccharides such as cellulose, hemicellulose and

of

energy for hum@ins, consists of 60—70%starch of two

principal types (1). Amylose is composed almost corn pletely of al,4-linked glucosyl linear chpins of an average molecular weight of 100 kDa (—600glucose residues), amylopectin, a branched starch (1000 kDa, -.6000 glucose residues), has both the elongated a1,4@-linked glucosyl straight chains and a-1,6-linked

branching points approximately residues

along the chain.

structure

of amylopectin

‘Presentedat the 31st Annual Rnmiiiant Nu@rifi@Conference, entitled “Starch Thgestion Understanding and Potential for hn provement,―at the Annual Meeting of the Federation of American Societies for Experimental Biology, April 1, 1990, Washington, DC. 2Guest editor for this symposium was C. B. Theurer@

every 20 glucose

Figure 1 diagrams the

and the action

Department of M@iiyii1Sciences, University of Arizona, Tuceon, AZ 85721. @Workof the author is eupported by grants from the US. Public Health Service, National Institutes of Health for reeearch @DK 11270)and forthe Digestive DiseaseGenter(DX 38707).

of a-amyl.ase

to produce oligosacchaxide breakdown products, the details

are considered

below.

0022-3166/92 $3.00 C 1992 American Institute of Nutrition. Received 26 June 1991. Accepted 16 July 1991. 172 Downloaded from https://academic.oup.com/jn/article-abstract/122/1/172/4754868 by guest on 08 March 2018

SYMPOSIUM: STARCH DIGESTION

173

the terminal reducing glucose unit to specific cats lytic subsites of the a-amylase, followed by deavage

AMYLOPECTIN

between

the second and third a-l,4.-linked

glucosyl

residues (8-10). The final products from amylose di gestion are principally the disaccharide maltose and

the trisaccharide maltotriose. Although substantial amounts

@-omylase@@@

of free glucose may eventually

be released

after prolonged incubation of starch with a-amylase in vitro (11), little glucose is formed by a-amylase action under physiological

conditions within the in

testinal lumen. Notably, the a-amylaaes have less specificity MALTOTRIOSE cg- LIMIT

MALTOSE

DEXTRINS

for

smaller

glucosyl

oligosaccharides,

which can bind nonproductively to two or three sub sites of the catalytic site by failing to span the scission site. a-Amylase has no specificity for the a-

fiGURE 1 Action of ot-amylaseon amylopectin. The

1,6 branch linkage in amylopectin, and its capacity to

partial structure of the branched 8tarch, amylopectin, is shown. Each circle represents a glucose residue linked

break a-l,4 links adjacent to the branching point is

either ct-l,4 (horizontally) or ccl,6 (vertically). The chains are shown beginning with the non-reducing end on the left and, for the hydrolysis products, the terminal reducing residue at the right. Because salivary and pancreatic a-

amylases have specificity for the a-l,4 links and are in hibited in the area of the ci-l,6 branching point, the final products are linear oligosaccharides

of 2-3 glucose units and

the brancheda-dextrins.

pectin,

which

are connected

of starch

the a-linked amylopectin,

principally

availability

most

of

of the branched

is actually

Based on analysis

of intestinal

amylopectin

are the a-dextrins,

branched oligosaccha

rides having one or more a-1,6 link (average molecular weight of 800-1000, Fig. 1). Most dietary amylopectin

is hydrolyzed

to its final oligosaccharide

by the time ingested

distal duodenum

of hwmins.

MEMBRANE

SURFACE

carbohydrates

reach the

by the mdi

by cooking,

glycan consists which

hindered.

contents removed from normal human subjects (12), nearly one-third of the final breakdown products of

products

gestible f@-glycanlinkage. Also, despite the major en hancement

sterically

rendered

less

DIGESTION

OF

THE RELEASED OUGOSACCHARIDES

di Because there are no carbohydrases

gestible by virtue of the heating process (5).

himinal

fluid except

a-amylases

in the duodenal

secreted

from

saliva

and the pancreatic duct, the final glucosyl oligosac

ACTION OF a-AMYLASE ON STARCHES The vast majority of starches are digested within the intestinal lumen by a-amylase secreted via the pancreatic duct. Although salivary a-amylase can be rapidly degraded in the acidic environment @

encoun

tered in the stomach (6) and hence may play a very minor role in hydrolysis of ingested starch, recent studies

(7) have

shown

considerable

protection

for

salivary a.mylase when starch or its a-amylase breakdown products are present at low pH. There is at least a 70% reduction in the degradation rate of sal ivary a-amylase

at pH 3.0 when it is accompanied

by

its starch substrate or the oligosaccharide breakdown products (7). The substrate, by interacting with the active

hydrolytic

site,

seems

to

ni@intain

amylase in a more favorable conformation allow

salvage

of the active

enzyme

the

a-

and may

as it passes

into

charide products must be handled by another mecha

rnsm. Notably, there is no integral transport process in the intestinal enterocyte that can accommodate anything

larger than free glucose.

Oligosaccharides

are hydrolyzed by surface oligosaccharidases, large glycoprotein components of the mtestinal surface brush border membrane. These carbohydrases are syn thesized within the depths of the enterocyte and transferred to the brush border surface, where they

retmiin anchored by a short t&minal hydrophobic segment

of

majority

of their domains

their

protein

ch@iins,

allowing

the

vast

(including the active cats

lytic sites) to be free at the lumen-cell interface (13-18). The characteristics of these a-glucosyl sac charidases

are given in Table 1. Although

the term

maltase can be used for these as a group,4 it is

may occur in the

4Nomenclature for this group of enzymes varies eomewhat Maltase is often used as a prefix hyphenated to glucoamylase, suaase or isomaltase, but the Principal physiological specific sub

neonatal period even though pancreatic a-amylase secretions are not yet developed. Arnylose and a.mylopectin are hydrolyzed by virtue

have nii@rim@lactivity are ul,4-Iinked oligoglycosidee (inal tooligossccharides), sucrose and the al,4-1,6-linked branched oligoglycosides (the ci-dcxtrina).The terms that identify the specific

the duodenum in premature infants. Hence signif icant

himinal

digestion

of starch

of binding of five of their glucose residues adjacent to Downloaded from https://academic.oup.com/jn/article-abstract/122/1/172/4754868 by guest on 08 March 2018

strata

in the intestinal

lumen for which the oligosaccharidaaee

physiologicalsubstrate have been used in this paper.

174

GRAY

TABLE 1 Intestinal no.znmol/Lsec@1Sucrase EnzymePrincipal

@ace membnme CI-gIOcOSIdaSeS

substrateKmK@tRef.

3

a-1,4 Glucosyloligoeaccharides

a-Dcxtrina:

a-Dcxtrin&se

2-4 1

(a-1,4 Links) (ci-1,6 Links) GlucoamylaseSucrose

1-4

a-Dcxtrins (ci-1,4 links only)18

1120

specific substrate. Glucoamylase (amyloglucosidase, maltase-glucoamylase) is capable of removing single glucose residues sequentially from the nonreducing end of the a-1,4 chain, but (analogous to the situation with a-aniylase) is blocked when an a-1,6--linked glucose becomes located at the terminiti end of the sacchaxide. Sucrase-a-dextrinase, (commonly called sucrase-isomaltase), a hybrid carbohydrase initially

synthesized as a single glycoprotein chain within the enterocyte's interior, is divided after its insertion into the brush border membrane by pancreatic proteases

into sucrase and a-dextrinase units, which then reas non-covalently

20-40

at the intestinal

surface.

Su

crase is a highly efficient a-l,4 glucosidase that corn

13, 16 13

120 50-65

a-1,4 Glucoeyloligosaccharidcs

preferable to name them according to their primary or

sociate

13—16

110

16-18

30-4013-15

16-18

plernents the specificity of glucoamylase by preferring shorter a-1,4--linked oligosaccharides, particularly maltotriose and maltose. a-Dextrinase is the sole sac charidase that has the capacity to deave the non reducing terminal a-1,6 link once it becomes uncov ered. In the sequential action of the intestinal

membrane glycosyloligosaccharidases on a typical adextrinase occurs

(Fig. 2), deavage from the nonreducing

by sequential

orderly

removal

end

of individual

glucose residues, with glucoamylase being the most efficient initially, the a-dextrinase being essential for

the deavage of the a-1,6 branching link and sucrase being the preferred a-glucosidase for finishing up the final deavage of the chain after it has been reduced to

two or three glucose residues. Transport across the intestinal cussed

membrane

is restricted

to glucose, as die

below.

TRANSPORT OF THE RELEASED GLUCOSE @

GLUCOSE

The final glucose product can then be transported by the specific glucose carrier or transporter, a 75-kDa integral brush border glycoprotein expressed

only in the small intestine (and possibly in kidney tubules), which has high affinity for the monosac chañde (19). Kinetic

studies

of brush border mem

brane vesides support the presence of both the high GLUCOSE FIGURE

2

Sequential

oligosaccharidaaes

to

action

cleave

an

of

intestinal

a-dextrin.

The

surface

model

depicts the a-dextrin hcxasaccharide (of. Pig. 1 legend for

affinity transporter (low glucose concentration re quirement or K@)throughout the small intestine and, in the jejunum, a low iiffinity, high capacity trans porter

(low ïç and high ni@rimel

transport

rate or VJ.

designation of glucose residues and linkages) initially at tached by g1UCOaII1yIaSCor a-dcxtrinase in the sequential

Both transporters depend on the presence of Na@in the htniinal glucose solution to facilitate efficient

removal

transport (20). Two molecules

of single

glucose

residues

from

the non-reducing

end (upper left). When the 1,6-linked glucose residue be comes termin@1, only the a.dextrinase is capable of cleavage. The final conversion of the triuCCharide (mel totriose) and the disaccharide (maltose) is most efficiently

achieved by the action of either sucrase or glucoamylase

(basedon data from ref. 13—18).

Downloaded from https://academic.oup.com/jn/article-abstract/122/1/172/4754868 by guest on 08 March 2018

of sodium bind at a site

on the transporter (21) separate from glucose and (along

with

the

glucose)

are

internalized

and

die

charged to the interior of the cell. The actual driving force for uphill transport of glucose into the en terocyte

is

provided

by

the

sodium-potassium

175

SYMPOSIUM: STARCH DIGESTION

E1HYDROLYSIS

SURFACE MEMBRANE

@

GLUCOSE PRODUCT ABSORPTION

1.UMEN

PANCREATIC INSUFFICIENCy

fiGURE 3 Mechanism of glucose-Na@ cotransportin the intestine. Glucose (circle) interacts with the 75-kDa mem at the brush border surface and

is joinedby Na@at a Na@/glucose molarratioof 2.0, which greatly facilitates basolateral centration

glucose entry. The ATPase in the

membrane enables Na@extrusion against a con gradient and constitutes the driving force to

overall coupled transport. The percentagesare estimates of fractional poo1s of glucose at transport equilibrium. The

local intracellular and interstitial concentrations of Na@are given

within

brackets.

transport

membrane.

Na@across the

The elements

in the enterocyte

are shown

of glucose

in Figure 3. The

coupled nature of sodium and glucose and their co transport

through

maltose

in human

intestine.

Comparison

of hydrolysis

(total column height) and absorption of the glucose product (hatchedarea)from 1%amylopectinor maltose in normal healthy subjects and in those lacking duodenal a-amylase because

of pancreatic

insufficiency.

strate the reproducibility 30-cm intubated

Disappearance

segment

Normal

I and 11 demon

of the in vivo hydrolysis

in a

beg@nning at the distal duodenum.

of the test saccharide

was taken as its hy

drolysis, and absorption calculated by subtracting the re sidual saccharide products remaining in the intestinal

lumen from that hydrolyzed.Notably, hydrolysis exceeded absorption and was at least as great from amylopectin as from maltose. In pancreatic insufficiency, the absence of aamylase reduced the amylopectin hydrolysis and subse

ATPase, which pumps the intracellular

basolateral

PANCREATIC INSUFFICIENCY

FIGURE4 Hydrolysis and absorption of starch vs.

- Nci'1140m.ci/L1

brane carrier or transporter

NORMAL

the membrane

surface

quent glucose absorption appreciably but had no effect on maltose

hydrolysis

and absorption,

because the intestinal

oligosaccharidase activities are normal in this condition (from ref. 12, with permission of the American Physio logical Society).

are essential

for the efficient final step in the assimilation of di etary starch. Glucose itself probably diffuses from the

tose, as shown in Figure 4. After the cleavage to

basolateral

oligosaccharides in the duodenal lumen, the surface hydrolysis under the influence of appropriate oligosaccharidases seems to be even more efficient. Glucose is usually produced in excess and absorption

(serosal)

surface

to the capillaries

of the

villous core, but this may also involve interaction with a second carrier protein at the basolateral surface (22) (Fig. 4).

is at least equal

(23) or superior

(24) to that

achieved

by an equivalent quantity of free glucose. Such effi

RATE-UMITING

STEPS

IN

ciency

may

be

STARCH ASSIMILATION

tective unstirred Certainly starch

the most

assimilation

is

important

aspect

preparation

of

possible

because

oligosaccharide

deavage at the lumen-cell interface within the pro of overall

the

starch

layer of water and mucoproteins

(25)

may allow the released monosaccharides to be ‘m'@n tained at relatively high local concentrations, thereby

containitig food before ingestion, in a way that favors its solubility in the polar limiinal solution to promote

favoring

efficient interaction with a-amylase. For humans, this is usually a process such as grinding and milling and

final glucose transport into the depths of the entero cyte, rather than intrairiminal or surface digestion of

relatively

thorough

cooking,

for

tnimiil@,

this

in

efficient

transport

into

the

enterocyte.

Overall, it seems that the rate-limiting step is the starch and its oligosaccharide products (1, 12, 23).

volves some industrial processing to promote gelatini zation at the interior of the crystalline

granule of raw

starch. After this has been accomplished, hydrolysis in the duodenal lumen is most often extremely effi

@ent.Indeed, in human digestion-absorption studies (12), overall assimilation from a perfused solution of 1% starch was found to be equal to that from mel Downloaded from https://academic.oup.com/jn/article-abstract/122/1/172/4754868 by guest on 08 March 2018

DEGREE OF STARCH ASSIMILATION IN HUMAN BEINGS Despite

the apparently

amylase in the duodenal

efficient

interaction

with a-

lumen with the processed

176

GRAY quent digestion at the lumen-brush border interface. More than sufficient

2

glucose

monosaccharide

is then

released in the vicinity of the brush border glucose

starcb@Food AssimilationTABLE of carbohydrate from foodsanzkhed In

transporters to insure saturation. By virtue of the sequence of these three integrated processes, starch is assimilated in a very efficient ni@nner in nonrumi

sourceAbsorbedMalabsorbedg %gOatsStarch57

Earns.

(2)Non-starch0.4

(98)0.9 (6)6.3

(94)CornflakceStarch71 (5)Non-starch00.7 (100)BreadStarch61

UTERATURE CITED

(95)3.4

(2)Non-starch02.3

1. Gray, C. M. (1970) Carbohydrate digestion and absorption. Gastrocnterology 58: 96—107. 2. Heaton, K. W., Marcus, S. N., Emmett@ P. M. & Bolton, C. IL

(98)1.1 (100)

‘Datacalculated from the ileostomy effluent studies of Englyst and Cummings (27). Starch and non-starch carbohydrate compo nents are listed separately.

(1988)Particle size of wheat, maize, and oat test meals: effects on plasma glucose and insulin responses and on the rate of starch digestion in vitro. Am J. Clin. Nut. 47: 675-682. 3. Hoim, J., Hagander, B., BjOrck,L, Eliuson, A-C. & LUndqUiIt, L (1989) The effect of various thermal processes on the

glycemic response to whole grain wheat products in human. and rats. J. Nutr. 119: 1631—1638. 4. Bornet,F. R., Foiitvieille,A. M., RizkalIa, S., Colonna,P.,

amylose and amylopectin,

there is considerable cvi

dence that a significant proportion of some starches passes all the way through the intestine to be mets bolized by bacteria in the colon. For the first two to four weeks of life, newborns, especially pre-term in fants, have a relatively reduced capacity to assimilate carbohydrate because of reduction in pancreatic aamylase. In adults, assimilation of polysaccharides

(including starch) is dependent upon the ingested flu trient.

Carbohydrate

maldigestion

ranges from 1 % for

rice to 7% for white wheat, 10% for whole oats and 19% for baked beans (26). This may be due to the

presence of amylase inhibitors in the food, to the relatively slow process of wetting of the starch corn ponent from the depths of the grain or legume, or to the presence of non-starch carbohydrate (so-called di etary fiber). In patients

with ileostomies,

whose ter

minal ileal effluent is available for direct qualitative and quantitative analysis of unabsorbed carbohydrate, Englyst and Cummings (27) found the assimilation of starch from grains to be nearly complete (Table 2). Malabsorbed carbohydrate consisted mainly of non starch

glycans

that

are

cleaved by a-amylase

@3-linked and hence

are not

(27).

In the aggregate, starch assimilation in non nimin@nts is apparently a highly efficient process that involves the following: 1) duodenal intraluminal di

gestion by a-arnylase from saliva and pancreas, 2) surface

hydrolysis

of

the

starch

oligosaccharide

breakdown products by integral oligosaccharidases of the brush border, and 3) transport of the final released glucose by the specific glucose transporters of the enterocyte. Provided that the food containing has been processed and cooked to facilitate

bifity to the intraduodenal

water-soluble

starch ava.ila

a-amylases,

the initial luminal digestion is very efficient and pro vides sufficient oligosaccharide products for subse Downloaded from https://academic.oup.com/jn/article-abstract/122/1/172/4754868 by guest on 08 March 2018

BI*yo, A., Mercier, C. & Slama, C. (1989) Insulin and glycemic responses in healthy humans to native starches processed in

different ways: correlation with in vitro alpha-amylsse hydrol ysis. Am. J. Clin. Nutr. 5@ 315-323. 5. Englyst, H. N. & Ciinlmingk, j. H. (1987) Dig@rion of polysac

cb.arides of potato in the small intestine of man. Am. J. Clin. Nutr. 45: 423-431.

6. Pried,M., Abrameon,S. & Meyer, J. H. (1987) Passageof salivary amylase through the stomach in human.. Dig. Die. Sri.32: 1097—1103. 7. Roeenblum,J.L, Irwin, C. L. &Alpere, D. H. (1988)Starchand glucose oligoeaccharides protect salivary-type amylase activity

at acid pIL Am. J. Physiol.254: G775-G780. 8. Robyt, J. F. & French, D. (1970) The action pattan of porcine pancreatic aipha-amylase in relationship to thc substrate

binding site of the

@zyme. J. BioL @heni. 245: 3917-3927.

9. chan, Y., Braun, P. J., French, D. & Robyt, J. P. (1984)Porcine

pancreatic alpha-amylaee hydrolysis of hydzvxyethylated amyloee and specificity

of eubsite binding. Biochemistry

23:

5795—5800.

10. Desseauz@V., Seigner,C., Pierron,Y., Grieoni, K L & Marchie-Mouren,C. (1988)Porcine pancreatic alpha-amylaae a modelforstructure-function studiesof hOmOdepOlyIneraeeS. Biochimie (Paris) 70: 1163—1170. 11. Kaczmarek, M. J. & Rosenmund, H. (1977) The action of

human pancreatic and salivary ieoamylaees on starch and glycogen. din. @hem.Acta 79: 69—73. 12. Fogel, M. R. & Gray, G. M. (1973) Starch hydrolysis in man an

intralnniinal process not requiring membrane digestion. J.

AppL Physiol. 35: 263—267. 13. Gray,C. M., LaDy,B. C. & Conklin,K. A. (1979)Actionof

intestinal sucraee-ieomaltaee and Its free monomers on an alpha-limit dextrin. J. BioL [email protected]: 6038-6048. 14. Rodriguez@L it, Taravel, F. R. & Whelan, W. J. (1984) @harac terization and function of pig intestinal sueraae-ieomakaac and Its separate subunits. Eur. J. Biochein. 143: 575-582. 15. Lorenzsonn, V., Korsmo, H. & Olsen, W. A. (1987) Localization

of suorase-ieomaltaee in the rat enterocyte.Caetroaiterology 92@98—105. 16. Taravel, F. R., Datema, R., Wo1oszczuk@ W., Marshall, J. J. & Whelan, W. J. (1983) Purification and characterization of a pig

intestinal alpha-limit dextrinaee. Eur. J. Biochem. 13th 147—153. 17. Lee, L. &

Forstner,G. (1984)Rat intestinal maltese.

SYMPOSRJM@@ STARCH DIGESTION glucoaniylaae.

Purification

of

the

detergent-eolubilized

enzyme in the presence of protesee inhibitors: properties and

identificationof a protease-eeneithesubunit. Can. 7.Biochem. CelL Biol 62: 36-43. 18. Naim, H. Y., Sterchi, E@E. & Lentze@M. J. (1988) Structure@ biosynthesis, and glycosylation of human small intestinal maltaee-glucoamylase. 7. BiOL @hem.263: 19709-19717. 19. Hediger, M. A., Coady, M. J., Ikeda, T. S. & Wright@ E. M.

(1987) Expression doning and eDNA sequencing of the N&/ glucose co-transporter.Nature (Loud.)330: 379-381. 20. Hang, I. M., Barry, J. A., Rajendran, V. M., Soergel, L IL & Ramsawamy, K. (1989) D-Glucoee and L-leucine transport by

human intestinal brush-border membrane vesicles. Am. J. PhysioL 256: G618-G62.3.

2.1. Kimmich,G. A. & Randles,J. (1984)Sodium-sugarcoupling stoichiometry

in chick intestinal

cells. Am. J@Physiol. 247:

C74-C82. 22. Maenz@D. D. & Clieceeman, C. L (1987) The Na@-indepesdent D-glucose transportc in the enterocyte basolateral membrane

Downloaded from https://academic.oup.com/jn/article-abstract/122/1/172/4754868 by guest on 08 March 2018

177

orientation and cytochalasin B binding characteristics. J. Membr. BioL 97: 259-266. 23. Gray, G. 1st & Santiago, N. A. (1966) Diuccharide absorption in normal and diseased human intestine. Gastroenterology 51:

489-498. 24. Jones, B. J., Brown, B. E., Loran, J. S., Edgerton, D., Kennedy, J. F., Stead, I. A. & Silk, D. B. (1983) Glucose absorption from starch hydrolysates in the human jejunum. Gut 24: 1152—1160. 25. SmithsOn, L W., Millar, D. B., Jacobs, L. R. & Cray, G. M. (1981) Intestinal diffusion barrier@unetirred wat@ layer or

membrane surface mucous coat? Science @Wubingtcn, DC)

214: 1241—1244. 26. Levitt, M. D., Thigh, P., Fetzer, C. k, Sheehan, M. & Levine, A. S. (1987) 112 exaction after ingestion of complex carbohy drates. Cutroenterology 92 383—389. 27. Englyst, IL N. & Cununinr, J. H. (1985) Digestion of the

polysaccharides of some cereal foods in the human small intestine.

Am. J. Cm.

Nut.

42 778-787.