Children When consumption is normalized to body weight, young children are the highest consumers of fructose (Marriott et al. 2009) (Fig. 1). Total fructose consumption ranged from 30 g day−1 (∼3 g kg−1 day−1) in 1 year olds to 50 g day−1 (∼2 g kg−1 day−1) in 10-year-old children. This consumption rate is relatively high, and equivalent to adults consuming four or more cans of soda sweetened with high-fructose corn syrup (∼50% fructose) each day. It comes as no surprise that there is a greater incidence of positive breath hydrogen tests (a symptom of carbohydrate malabsorption), or much higher breath hydrogen peaks, in young children fed diets or juices containing high fructose levels (Cole et al. 1999; Duro et al. 2002). Incidence, severity of symptoms or peak breath hydrogen seemed to increase with decreasing age (Nobigrot et al. 1997; Tsampalieros et al. 2008) and with increasing fructose level in the diet (Gomara et al. 2008). Recently, when children were challenged with 0.5 g fructose (kg body weight)−1, a level lower than average consumption rates, in a 5 year study involving 1100 subjects, frequency of positive breath hydrogen was inversely related to age, so that ∼80% of infants <1 year old tested positive but only ∼25% of children >10 years old (Jones et al. 2011a) (Fig. 2A). It is expected that the per cent of children with symptoms of fructose malabsorption will decrease by ∼5% per year of age. Thus, in humans, fructose malabsorption is prevalent in infants, toddlers and young children compared with that in adults, suggesting a pattern inversely proportional to that of GLUT5 expression in rodent intestine (Fig. 2B). Developmental increases in GLUT2 are much more modest and not specific for fructose (Cui et al. 2003).
Figure 2. Low or modest expression of the fructose transporter GLUT5 may cause intestinal fructose malabsorption in humans A, fructose malabsorption in humans measured by breath hydrogen (data from Jones et al. 2011a). Subjects received either 0.5 g (kg body weight)−1 of fructose (maximum of 10 g) or 2 g kg−1 of lactose (maximum of 20 g), and were tested for 2.5 h. Patient age had a remarkable effect on the proportion of subjects that tested positive, so that the odds of testing positive for fructose malabsorption in patients 15 years or younger decreased by a factor of 0.82 for each year of increasing age. B, relative GLUT5 (left red bars) and GLUT2 (right blue bars) mRNA expression in the small intestine of rats as a function of age, from suckling to adults (V. Douard and R. P. Ferraris, reanalysis of archived materials from Douard et al. 2010, 2012). Bars are means ± SEM. Weaning and adult rats were fed a high-glucose or high-fructose diet; the ‘diets’ of suckling (<14 days old) rats reflected those of their mothers. All mRNA levels were normalized to GLUT5 expression (arbitrarily set as 1.0) in rats fed high glucose. Intestines of rats 1–2 days of age represented those of humans in the last trimester of gestation, 10–13 days of neonatal humans, 19–22 days of weaning, 45–50 days of teenagers, and >90 days of adult humans. GLUT5 expression increases dramatically while that of GLUT2 increases modestly with age. GLUT5 expression is enhanced specifically by a high-fructose diet when compared with a high glucose or any fructose-free diet. GLUT2 expression increases modestly with both glucose and fructose when compared with protein diets (not shown). Since GLUT5 expression levels determine rates of rat fructose transport rates (Jiang & Ferraris, 2001), fructose malabsorption in young children may be caused by lack of GLUT5.
Download figure to PowerPoint
The mechanisms underlying fructose malabsorption in children may be due to the facts that both intestinal GLUT5 expression, and the ability to regulate GLUT5 expression, are tightly constrained by development in mammals as established in rats, mice and rabbits (Ferraris, 2001; Douard & Ferraris, 2008; Boudry et al. 2010). Unlike SGLT1 (Slc5a1, sodium-dependent glucose transporter 1) whose expression levels are relatively substantial even in the fetal intestine and increases only gradually during postnatal development, baseline intestinal expression and activity of GLUT5 are low throughout the suckling and weaning stages, but increases markedly in postweaning. However, precocious and dramatic increases in GLUT5 expression during weaning, but not during suckling, can be induced by increases in intestinal luminal fructose concentrations (Douard et al. 2008a). The effect of fructose is specific, as no other sugar leads to GLUT5 upregulation, and seems dependent on metabolism, as 3-O-methylfructose fails to induce GLUT5 (Jiang & Ferraris, 2001). In contrast, intestinal GLUT2 expression during weaning and in adults can be enhanced by either fructose or glucose in either the lumen or blood (Cui et al. 2003). Perfusion of either glucose or fructose, but not amino acids, induces GLUT2 transcription.
The ontogenetic timing mechanism of enhancing GLUT5 expression and activity is rather complex. Fructose cannot induce GLUT5 during the suckling stage, but if suckling (<14 days old) rats were injected (primed) with the corticosterone analogue dexamethasone prior to introduction of dietary fructose, intestinal GLUT5 expression can increase dramatically (Douard et al. 2008b), along with expression of other genes involved in fructose transport and metabolism (Cui et al. 2004). Coincidentally, endogenous levels of corticosterone increases markedly at ∼14 days of age, also prior to allowing GLUT5 regulation by luminal fructose during weaning (Monteiro & Ferraris, 1997). Fructose induces de novo transcription and translation of GLUT5 and involves glucocorticoid receptor translocation to the nucleus followed by fructose-induced histone H3 acetylation and RNA polymerase II binding to the GLUT5 promoter (Suzuki et al. 2011). Since even very low levels (40 ng (kg body weight)−1 or ∼10% of typical human dose) of luminal or intraperitoneal dexamethasone allows fructose to induce GLUT5 in suckling rats (E. S. David and R. P. Ferraris, unpublished), symptoms of fructose malabsorption in young infants may be relieved by low levels of glucocorticoid analogues taken orally. Endogenous thyroxine levels increase following increases in corticosterone, and thyroxine has also been shown to regulate gut development. Thyroxine may have no effect on GLUT5 regulation since the magnitude of enhancement of intestinal fructose transport by dietary fructose in hypothyroid rat pups was similar to that observed in euthyroid pups (Monteiro et al. 1999). Surprisingly, thyroxine injections increase expression of intestinal GLUT5, GLUT2 and SGLT1 during weaning by increasing transcription (Mochizuki et al. 2007), although it was not clear whether thyroxine injections also altered corticosterone levels, which could explain their findings.
Isolated fructose malabsorption, a rare pediatric disease resolved by a fructose-free diet, does not arise from expression of mutant GLUT5 (Wasserman et al. 1996), suggesting that developmental limitations of GLUT5 expression is a potential explanation. It seems that omnivorous mammals have evolved in the wild so as not synthesize GLUT5 until after weaning is completed or after foraging begins and access to fructose-bearing fruits is feasible. In prehistoric humans who selectively preferred fruits, fructose was probably made available to infants when they were weaned at 3 to 4 years of age (Clayton et al. 2006). One can speculate that the human intestine, like that of the rat, rabbit and mouse, may not be fully capable of fructose absorption until completion of weaning.
Adults Intestinal fructose malabsorption also occurs in adult humans who, unlike infants, should have ample GLUT5 and GLUT2. Fructose malabsorption is not associated with GLUT2 or GLUT5 mutations (Wasserman et al. 1996). The essential role of GLUT2 for intestinal sugar absorption remains unclear since a GLUT2-deficient patient displayed normal breath hydrogen levels after glucose and sucrose tolerance tests (Santer et al. 2003) and GLUT2-null mice transported glucose but not the non-metabolizable 3-O-methylglucose, at rates similar to wild type (Stumpel et al. 2001). This indicates that another transepithelial, phosphorylation-dependent glucose transport system may exist. These findings can now be evaluated better with similar studies using GLUT5−/− and KHK−/− models.
Fructose malabsorption by human adults, as evaluated by breath hydrogen, increases as a function of dietary fructose concentration, so that for every 10 g increase in fructose dose, the number of positive breath hydrogen tests increases by 15% (Jones et al. 2011b). At 50 g, a dose well below the average daily fructose intake in the USA, about 60–80% of adults experience some form of malabsorption (Truswell et al. 1988; Ladas et al. 2000; Rao et al. 2007), suggesting that intestinal absorptive capacity for fructose in some people is insufficient at current rates of fructose intake. Unlike that of isolated fructose malabsorption, the mechanism underlying adult malabsorption is not clear. There is a consensus that rat or mouse intestinal absorptive capacity for glucose exceeds total intake by 20% or more (Ferraris et al. 1990; Diamond, 2002) so that glucose even from high-carbohydrate diets is completely absorbed. However, GLUT5 is facilitative and, in rats and mice, absorptive Vmax for fructose is ∼3–4 times less compared with that for glucose, a difference in activity that parallels the 8-fold differences in mRNA expression between SGLT1 and GLUT5 (Ferraris & Vinnakota, 1995; Shu et al. 1998). The total absorptive capacity of a mouse intestine for fructose is about 1.5 μmol min−1 on a low-sucrose diet, and 4.0 μmol min−1 on a high-sucrose diet (Ferraris et al. 1990). In contrast, that for glucose is about 7 and 12 μmol min−1, respectively.
Regulation by diet In experiments utilizing surgical rat models with Thiry Vella loops, fructose intake did not induce GLUT5 upregulation in the bypassed section but only in the anastomosed (reconnected) intestine, suggesting that GLUT5 requires interaction with its substrate for upregulation (Shu et al. 1998). GLUT2 expression is upregulated by glucose and fructose from the lumen or the blood (Cui et al. 2003). In support of these findings, baseline expression and function of GLUT5 and GLUT2 do not change after complete intestinal denervation and are thus independent of intrinsic or extrinsic neural connections to the jejunoileum (Iqbal et al. 2009).
In summary, in some human adults potentially unable to sufficiently upregulate GLUT5 expression, absorptive capacity for fructose may be limited, and when dietary concentrations of fructose is well above normal and exceeds that of glucose, absorptive capacity may be exceeded, and malabsorption occurs. In infants, abnormal breath hydrogen levels caused by dietary fructose are probably due to low levels of intestinal GLUT5 expression.