What Is Known
- Starchy foods are a major source of complementary food energy.
- There are 4 different duodenal mucosal activities assayed as maltase.
- Developmental α-amylase insufficiency is countered by mucosal maltase starch digestion.
What Is New
- The maltase associated with sucrase activity is the dominant activity in duodenal biopsy assays.
- Non-sucrase maltase can be approximated by the Rule of Two: total maltase activity − 2.24 times sucrase activity.
- A novel non-sucrase maltase deficiency is identified by the Rule of Two for further investigation.
Discovery of small intestinal sucrase activity, then termed invertase, was reported by Claude Bernard, considered the father of physiology, in 1856 (1). He found this activity in many mammalian species. In 1880, H.T. Brown discovered mucosal maltase activity and differentiated it from the pancreatic enzyme diastasis, now called amylase, His view was that pancreatic amylase hydrolyzed starch into maltoside fractions, which were then hydrolyzed to glucose by maltase activity in the small intestine (2).
Advances in protein chemistry in the 1960s allowed Dahlqvist (3) and Semenza (4) to fractionate and characterize small intestinal maltase activities. Both groups demonstrated that there were 4 major fractions of the maltase activity that were intrinsic to 2 different peptide structures. One maltase peptide was also active in hydrolyzing sucrose (Ib) and isomaltose (Ia) substrates, and was called sucrase-isomaltase (SI). The other maltase peptide did not have added substrate specificities (II, III) and was named maltase-glucoamylase (MGAM). Dahlqvist reported an assay for enzyme activities from biopsy homogenates in 1970, which has greatly influenced pediatric gastroenterology clinical practice (5). Maltase enzyme activity results are reported by many clinical laboratories using modified Dahlqvist assays to search for disaccharidase deficiencies from clinically indicated duodenal biopsies. Congenital and acquired deficiencies of lactase and sucrase activities have been characterized with well-recognized clinical presentations and effective management. In contrast, duodenal maltase activity deficiencies are less well understood and remain difficult to interpret. This review will test Dahlqvist's hypothesis (3), that is, hereditary maltase intolerances will never be found because maltase activities from mucosal tissues of all studied species can be separated into multiple fractions. Dahlqvist's related hypothesis will be confirmed, that is, maltase intolerance, which demands the simultaneous absence of 4 separate enzymes, will probably always be accompanied by intolerance for sucrose and isomaltose (3).
Fifty years later we enter the genome age and now can test, and perhaps extend, Dahlqvist's biochemical observations and understandings about these maltase activity fractions. Cloning and sequencing of the mucosal starch hydrolase enzymes (6,7) in the last decades of the 20th century confirmed that there are 2 maltase activities expressed by maltase-glucoamylase and 2 by sucrase-isomaltase. Recent research has provided detailed in vitro enzyme substrate selectivity, activity modulation, and structural specificity of the 4 α-glucosidases summarized in this workshop. Except for lactase deficiency (8) and rare individuals with congenital sucrase-isomaltase deficiency (CSID, (9)); however, clinical relevance of many biopsy activities remains obscure. For this reason, translational in vivo animal models of mucosal maltase deficiencies have been developed.
ANIMAL MODELS OF MALTASE DEFICIENCIES
A model of CSID (maltase Ib deficiency) has been investigated in the Asiatic shrew (Suncus marinus). The strain studied has a gene mutation after the membrane-binding domain in the isomaltase moiety of SI (Nt-Si, maltase Ia), which blocks transcription of the sucrase (Ct-Si, maltase Ib) exons. As a result, the mucosal sucrase activity levels were virtually absent. The mutant shrews developed severe diarrhea and dehydration when fed sucrose, but feeding of the starch hydrolytic product, maltodextrin, was tolerated. The hypothesis for this study was that a quantitative reduction in small intestinal starch digestion occurs in sucrase-deficient animals (10,11). When fed, blood-glucose and glucose-enrichment values from labeled starch were reduced in the mutant shrews, compared to controls, lending support to the above hypothesis. Confirming evidence was obtained when poor starch digestion, caused by sucrase deficiency, could be corrected with an oral supplement of recombinant glucoamylase (maltase II). The response to enriched glucose accounted for about 20% of the prandial rise in total blood glucose. In summary, mucosal sucrase deficiency reduced starch digestion to glucose without detectable intolerance.
A model of maltase-glucoamylase (maltases II and III) deficiency was studied in mice (12–14). Null mice were produced by ablation of the membrane form of MGAM. When the mucosal maltase activity was measured, the null mice had a 40% reduction in activity and a compensatory increase in sucrase activity. The hypothesis for this study was that residual sucrase-isomaltase (maltases Ia and Ib) activity can maintain starch digestion. A 40% reduction in labeled breath 13CO2 was shown after the null mice were fed 13C-labeled starch. This 40% reduction was confirmed in a gastric infusion study measuring 13C-maltodextrin digestion into 13C-glucose in circulating blood. Despite reduced blood 13C-glucose enrichment, no differences in total blood glucose concentration could be detected between the null and wild type (WT) genotypes. Total glucose turnover was the same in both genotypes, although blood insulin responses were significantly diminished in the infused null mice. These results supported the hypothesis that maltase-glucoamylase-(maltases II and III)-deficiency had 40% reduced intestinal starch glucogenic capacity, and that MGAM plays a significant role in insulin response to starch feeding. The apparent paradox between exogenous glucogenesis from starch and total glucose turnover led to an investigation of the role of MGAM in prandial control of endogenous gluconeogenesis. Because SI is a more abundant maltase than MGAM in the small intestine, we hypothesized that MGAM is important for efficient glucogenesis during limited intakes and that SI activities dominate gluconeogenesis during unlimited feeding. This was investigated in the MGAM null and WT mice using computer-controlled feedings of a 13C-starch diet. In addition to 13C-glucose enrichment from fed starch, gluconeogenesis was measured by 2H-glucose enrichment from body water. Again, there was no difference between genotypes in total blood glucose response to feedings, and 13C-glucose enrichment from the timed-feeding was the same as that on the unlimited feeding. Blood glucose response, however, was 40% lower in null mice on limited feeding. At 4 hours after the feeds, the null mice had about 40% more undigested starch and less free glucose in the small intestine. Post-prandial gluconeogenesis was increased in the MGAM null mice compensating for the reduced glucogenesis from starch. In summary, a 40% reduction of in vivo glucogenesis from starch was observed only in limited feeding of MGAM-deficient mice, indicating that MGAM maltase activity provides a high-efficiency glucogenesis, whereas on unlimited starch intakes, slower SI activities play a dominant role in prandial glucogenesis. Internal gluconeogenesis compensated for the reduction of starch glucogenesis in the null mice.
HUMAN MALTASE DEFICIENCIES
The standard human sample used for the Dahlqvist assay is a single <10 mg tissue fragment obtained by duodenal mucosal biopsy. The sampling of solids presents a universal statistical challenge, whereas the sampling of the complex villus structured duodenal mucosal surface increases complexity. Reporting activities as a ratio to biopsy protein (eg, maltase of 125 U/g protein) is only a partial solution. This sampling problem places a limitation on the precision of diagnostic boundaries of Dahlqvist assay activity values. Although the variation of activities of replicate assays from a single human biopsy homogenate are in the range of 7%, sampling variations of duodenal activities measured from replicate biopsies or repeated endoscopies of the same patient have a variation of around 20% ((15), unpublished data). This variability of replicate activity has limited confidant diagnostic applications to the most severe levels of deficiency.
For the present review, we used the biochemical database that was reported in 2012 (16) and has been further analyzed (Supplementary Digital Content 1, http://links.lww.com/MPG/B302). Briefly, 27,847 complete sets of disaccharidase assays were identified from 30,334 (30K) duodenal biopsies collected for clinical indications between January 1, 2006, and July 29, 2011. The remainders were only assayed for lactase and/or sucrase activity. All assays were according to the methods of Dahlqvist (5) and reported as U/g protein. All assay values of 0.00 U/g protein were treated as 0.01 for statistical purposes. Statistical deficiency was defined as below a lower limit (LL) of mean –1 standard deviation or using the laboratory's clinical definitions of LL for lactase, <10; sucrase, <25; maltase, <100; and palatinase <5 μmol−1·min−1·g−1 protein (15). The clinical definitions of sucrase and maltase LL agree with a reduction of variability in values below this point (Fig. 1). In an unpublished study of duodenal biopsies, the glucogenic activity from maltose was compared to that of maltodextrin substrate, often called glucoamylase activity. Glucoamylase activity can be converted to maltase activity using the regression equation (±SE) of:
Lactase deficiency was classified according to the L/S ratio (17,18) where non-persistent deficiency was recognized with a ratio <0.2 and congenital lactase deficiency as <0.06.
Patterns of disaccharidase activities below LL
Of the 30K biopsies assayed, 55% had all substrate activities above the statistical LL of activity and were considered normal. The most common deficiency was that of lactase with 32% of all biopsy activities below LL.
Pandisaccharidase deficiencies (PDDs) were next in frequency with 9% of all assays below LL. In PDD, all 4 assayed substrates fell below LL. Isolated, primary maltase-deficiencies (PMD) were next, with (2%) and without (1%) lactase deficiency. Isolated sucrase-deficiencies followed with normal lactase (CSID) (0.5%) and several variations (<0.2%). When the statistical frequency of LL values was plotted against age, the α-glucosidase (sucrase-, maltase-, and palatinase-) deficiencies dropped in the second year to about 10% with advancing age. This is in contrast to β-glucosidase-(lactase)-deficiency which increased in frequency to about 40% in adults (Fig. 2).
Dominance of sucrase in total maltase assayed by the Dahlqvist clinical assays
Of the 4 duodenal mucosal maltase activities, that associated with sucrase activity is dominant in the Dahlqvist assays (3). When maltase activities are plotted against sucrase activity in the total 28,793 biopsy series (30 K, Fig. 1), the 2 show high correlation with a slope of 2.24 and an intercept of 39 U/mg of maltase, which followed the equation:
This degree of fractionation of the Ib from Ia, II, and III, maltase activity is confirmed by protein isolation and denaturation experiments (15). The intercept of the equation suggests that an average of only 40 units of maltase-activity are not contributed by the sucrase-Ib-activity, but correspond to the contributions of Ia, II, and III maltase-activities. This residual non-sucrase maltase-activity will be discussed below.
Maltase activity patterns below 100 U/g protein
Of the 30K biopsies with all activities assayed, 3832 maltase activities (13.3%) fell below the statistical LL of 100 U. Four patterns of low maltase activities can be recognized (Supplemental Table S1, Supplementary Digital Content 2, http://links.lww.com/MPG/B303). The best studied maltase-deficiency pattern is CSID (maltase Ib and/or Ia deficiency). Sucrase (maltase Ib), as described previously, correlated with maltase activity and was below the LL of 25 U/g protein in 2936 (10% of total biopsies). With the criteria requiring lactase activity >10, we found that 576 biopsies had the CSID pattern, which corresponded to 20% of all sucrase deficiencies (50, or 0.1% of all biopsies). The CSID patterns frequently have elevated lactase activities. The remaining sucrose-deficient biopsies were either PDD with lactase <10 (2853, or 10% of all biopsies) or α-glucosidase–deficient (AMD) (872, or 3% of all biopsies), where all but lactase were below the LL. The fourth maltase deficiency is a novel isolated primary maltase deficiency (PMD), which involved the sum of maltase activities Ia, II, III (53, or 0.1% of all biopsies), but without sucrase deficiency. PMD frequently had elevated lactase, sucrase, and palatinase activities. This maltase-non-sucrase-deficiency is identified by an equation derived from equation 3, which we call “The Rule of Two”:
Deficiency is recognized by a negative PMD and the replicate variability of biopsy sampling is considered in a revised LL of -15. The above classification of maltase and sucrase deficiencies depends on the levels of lactase activity. Recent molecular insights into human lactase deficiency (17,18) identify 2 forms, congenital and non-persistent forms. The non-persistent lactase deficiency is age-related with onset in late childhood or adolescence (Fig. 2).
The strength of Dahlqvist assays is their capacity to report 4 different substrate activities, all expressed by 3 different genes, and measured from a single biopsy specimen. It has become an axiom for pattern recognition of both lactase and sucrase deficiencies, that lactase activity be compared with that of sucrase activity; the deficiency of either activity requires normal activity of the other. For the recognition of CSID, it has become accepted as a requirement that lactase activity be greater than the LL; conversely, sucrase activity must be above the LL. The genomic era calls these criteria into question. The recognition that non-persistent lactase deficiencies increase after weaning as a result of a polymorphism in chromosome 2, and that CSID exists as congenital mutations in chromosome 3, suggests that the molecular basis for these defects is different or independent, one from the other. Also, the fact that frequency of non-persistent lactase-deficiency increases with age (Fig. 2) and is widespread in the normal population, suggests that there are CSID patients who develop concurrent non-persistent lactase deficiencies. Tests for the polymorphism of C/T at nucleotide –13,910 found in non-persistent lactase-deficiency, are commercially available and may confirm this hypothesis.
Another widely accepted axiom is that PDD, when lactase, sucrase, maltase, and palatinase are all below the LL, is an artifact of processing of the biopsy. This assumption should be reconsidered, especially when the accompanying duodenal histology is normal. Although sucrase was named for its specific substrate, as demonstrated by animal studies above, the maltase activity of this enzyme plays a major role in mucosal starch digestion. As mentioned above, maltase and sucrase activities show a high direct correlation with a slope constant of 2.24. Thus, we suggest that the “Rule of Two” can be used as a rough estimate of the non-sucrase fraction of maltase, by the subtraction of 2.24 times sucrase activity from the total maltase activity. The control by sucrase on maltase activities is strongest in the MGAM KO mice, mentioned above, and clearly differentiates null from heterozygous or WT homogenates by changes in the maltase-2.24 sucrase (M-2.24S) subtraction. Although somewhat empirical, the rationale for the M-2.24S activity subtraction is visually evident in our 30 K series of duodenal biopsies (Fig. 1).
CLASSES OF MALTASE DEFICIENCY
If the statistical LL of 100 maltase activity is applied, 4 groups of maltase-deficiency can be identified; the major 2 groups have sucrase and maltase activities which follow the 2.24 rule, some with (PDD) and others without (AMD) lactase-deficiency. The PDD are older than either CSID or AMD (Supplemental Table S1, Supplementary Digital Content 2, http://links.lww.com/MPG/B303). When the M-2.24S subtractions are applied to biopsy activities with various grades of mucosal histologic abnormalities, the general pattern of “Rule of Two” is conserved (unpublished).
The application of “Rule of Two” also reveals a novel group of severe maltase deficiencies with normal sucrase, palatinase, and lactase activities. The frequency is approximately the same as classic CSID. Because these were identified from unselected clinical biopsies, no clinical information is available about the indications for endoscopic investigations.
PANDISACCHARIDASE DEFICIENCY AND α-GLUCOSIDASE–DEFICIENT MAJOR GROUPS
In the present analyses, the conventional criteria for CSID enzyme recognition include lactase activity >10 U/g protein (19,20). For purposes of analysis, the PDD biopsies were divided by those with lactase activities below 10 U/g protein; those above this clinical LL were reclassified as AMD (Fig. 3). The mean class activity comparisons are reported in Supplemental Table S1 (Supplementary Digital Content 2, http://links.lww.com/MPG/B303), where all enzyme activities are significant lower in lactase. With lactase <10, the PDD regression equation becomes:
In patients with PDD, the correlation of maltase and sucrase decreases, displaying a slope of almost 3 (2.91 ± 0.06) and, notably, the intercept is substantially lower reflecting the reduced overall contribution of maltases Ia, II, and III. A review of reported CSID cases reveals that about 20% have lactase below the LL (unpublished). In Sander's report, in which all patients were >9 years (21), half had low lactase-deficiencies combined with sucrase-deficiency and documented exon mutations of the SI gene. This is clear evidence that lactase-deficiency can coexist with CSID. Our case PDD 2 suggests that, like Sander's patients, lactase-deficiency can be of the late-onset non-persistent class. A new hypothesis arises that some or many of PDD-pattern patients could be undiagnosed cases of CSID with lactase-deficiency. These observations suggest the need of genotyping AMD patients to confirm CSID and the search for C/T –13,910 polymorphism in PDD.
With the hypothesis that CSID can co-exist with lactase-deficiencies, the PDD and AMD data sets were reviewed without reference to lactase activity. The relationship between maltase- and sucrase-activities has been reviewed above and was tested as a lactase-independent indicator for the detection of sucrase-deficiency. As previously described, the PDD data set was divided by lactase >10 U/g protein to produce lactase normal AMD. The calibrating M/S ratio from published CSID activities was 21.64 ± 19.17 (M ± SD) with a range from 0.00 to 93.67 (n = 27, LL: 2.5). The upper limits (UL, M + 1 SD) of AMD and PPD are 5.3 and 5.5, respectively. A conservative M/S cutoff of >6 was used to identify a possible sucrase-isomaltase-deficient subset from the AMD and PDD data sets (Table S3). The putative deficient subsets were named α-glucosidase sucrase-isomaltase-deficiency (AMSID) and pandisaccharidase sucrase-isomaltase deficiency (PSID), respectively. The lactase values were the same in AMD and AMSID classes and slightly lower in PSID versus PDD classes. The sucrase activities were the same in CSID versus AMSID and vs PSID classes, suggesting that similar sucrase-deficiencies exist independently of lactase activities. As expected, palatinase-deficiencies agreed with this sucrase equivalences between these classes. The age patterns reveal that CSID is recognized at 6.5 years and AMD and PDD at 10 and 11 years, respectively. The ages at which AMSID and PSID were recognized were 1 year younger than the larger groups. Reviewed in this context, this evidence of additional SID deficiencies within PDD (at least 13% PSID) and AMD (at least 4% AMSID) support the above hypothesis that CSID can exist in PDD patients with non-persistent lactase deficiency (Table 1).
DEVELOPMENTAL STARCH DIGESTION HYPOTHESIS
The human mucosal maltases Ia, Ib, II, III, and lactase are fully developed at birth (22). The maturation of human-secreted α-amylases is delayed until weaning in later infancy (23). In the more mature child, α-amylase amplifies the activity of mucosal Ib (24), but amylase hydrolyzed starch products apply a “brake” on the more active glucoamylase (maltase II, 15). In early infancy, starch is digested by active maltase II and III (MGAM), but with α-amylase maturity, it shifts to less-active maltase Ia and Ib (SI). A hypothesis arises, that is, transition to slower mucosal starch digestion reflects the onset of non-persistent lactase deficiency, and a longer period of infant weaning to complimentary food starches in ancient societies.
Dahlqvist's hypothesis (3), that hereditary maltase intolerances will never be found, because maltase activities from mucosal tissues of all studied species can be separated in to multiple fractions, has been refuted by identification of PMD-deficiency with elevated sucrase and palatinase activities. His prediction, that maltase intolerance will probably always be accompanied by intolerance for sucrose and isomaltose, is partly refuted (3).
The availability of molecular biology has begun to amplify Dahlqvist's biochemical understanding of mucosal maltase activities and their capacity to digest starch into absorbable glucose. Complementary in vitro and in vivo experiments confirm the dominant role of the maltase activity associated with sucrase activity (Ib) on prandial glucogenesis after a major meal, and the modulating role of glucoamylase (II) in snacking. Two specific patterns of maltase- deficiency can be recognized; CSID and PMD. The largest group of maltase-deficiencies remains poorly understood; PDD could include unrecognized CSID deficiencies with coincidental non-persistent lactase deficiency, and AMD may include a large number of homozygous and heterozygous CSID-deficiencies. The increasing availability of sequencing technology for clinical diagnosis has enhanced our understanding of the relationship between genotype and phenotype in CSID and similar advances can be predicted for PMD. The molecular approach to diagnosis probably can identify undiagnosed portions of PDD and AMD.
1. Bernard C. De l’assimilation du sucre de canne (digestion of sugar by dogs). Comptes rendus des séances de la Société de biologie
2. Brown HT, Heron J. Some observations upon the hydrolytic ferments of the pancreas and small intestine. Proc R Soc Lond
3. Dahlqvist A. Specificity of the human intestinal disaccharidases and implications for hereditary disaccharidase intolerance. J Clin Invest
4. Aurrichio S, Semenza G, Rubino A. Multiplicity of human intestinal disaccharidases II. Characterization of the individual maltases. Biochim Biophys Acta
5. Dahlqvist A. Assay of intestinal disaccharidases. Scand J Clin Lab Invest
6. Chantret I, Lacasa M, Chevalier G, et al. Sequence of the complete cDNA and the 5’ structure of the human sucrase-isomaltase gene. Possible homology with a yeast glucoamylase. Biochem J
7. Nichols BL, Eldering J, Avery S, et al. Human small intestinal maltase
-glucoamylase cDNA cloning. Homology to sucrase-isomaltase. J Biol Chem
8. Robayo-Torres CC, Nichols BL. Molecular differentiation of congenital lactase
deficiency from adult-type hypolactasia. Nutr Rev
9. Nichols BL Jr, Auricchio S. 50 years of progress since congenital sucrase-isomaltase deficiency recognition. J Pediatr Gastroenterol Nutr
2012; 55 (suppl 2):S2–S7.
10. Nichols BL, Avery SE, Quezada-Calvillo R, et al. Improved starch digestion of sucrase-deficient shrews treated with oral glucoamylase enzyme supplements. J Pediatr Gastroenterol Nutr
11. Robayo-Torres CC, Baker SS, Chumpitazi BP, et al. Poor starch digestion in children with CSID and recurrent abdominal pain. J Pediatr Gastroenterol Nutr
2012; 55 (suppl 2):S32–S34.
12. Nichols BL, Quezada-Calvillo R, Robayo-Torres CC, et al. Mucosal maltase
-glucoamylase plays a crucial role in starch digestion and prandial glucose homeostasis of mice. J Nutr
13. Quezada-Calvillo R, Robayo-Torres CC, Opekun AR, et al. Contribution of mucosal maltase
-glucoamylase activities to mouse small intestinal starch alpha-glucogenesis. J Nutr
14. Diaz-Sotomayor M, Quezada-Calvillo R, Avery SE, et al. Maltase
-glucoamylase modulates gluconeogenesis and sucrase-isomaltase dominates starch digestion glucogenesis. J Pediatr Gastroenterol Nutr
15. Quezada-Calvillo R, Robayo-Torres CC, Ao Z, et al. Luminal substrate “brake” on mucosal maltase
-glucoamylase activity regulates total rate of starch digestion to glucose. J Pediatr Gastroenterol Nutr
16. Nichols BL Jr, Adams B, Roach CM, et al. Frequency of sucrase deficiency in mucosal biopsies. J Pediatr Gastroenterol Nutr
2012; 55 (suppl 2):S28–S30.
17. Rasinperä H, Savilahti E, Enattah NS, et al. A genetic test which can be used to diagnose adult-type hypolactasia in children. Gut
18. Kuokkanen M, Kokkonen J, Enattah NS, et al. Mutations in the translated region of the lactase
gene (LCT) underlie congenital lactase
deficiency. Am J Hum Genet
19. Treem WR. Clinical Aspects and treatment of congenital sucrase-isomaltase deficiency. J Pediatr Gastroenterol Nutr
2012; 55 (suppl 2):S7–S13.
20. Nichols BL, Avery SE, Karnsakul W, et al. Congenital maltase
-glucoamylase deficiency associated with lactase
and sucrase deficiencies. J Pediatr Gastroenterol Nutr
21. Sander P, Alfalah M, Keiser M, et al. Novel mutations in the human sucrase-isomaltase gene (SI) that cause congenital carbohydrate malabsorption. Hum Mutat
22. Auricchio S, Rubino A, Muerset G. Intestinal glucosidase activities in the human embryo, fetus, and newborn. Pediatrics
23. McClean P, Weaver LT. Ontogeny of human pancreatic exocrine function. Arch Dis Child
24. Lin AH, Nichols BL, Quezada-Calvillo R, et al. Unexpected high digestion rate of cooked starch by the Ct-maltase
-glucoamylase small intestine mucosal α-glucosidase subunit. PLoS One