Main

Although most humans down-regulate intestinal lactase (EC 3.2.1.23-62) during postnatal development, a significant minority does not. Studies of proximal intestinal mucosal biopsies have demonstrated high lactase levels until after the age of 3 y when two groups become apparent: a group in which enzyme activity remains high and a group in which levels fall to about 15% of earlier values(1, 2). Although the mechanism for lactase down-regulation has not been studied in children, the heterogeneity of lactase expression has been extensively studied in adults. High levels of lactase in adults (lactase persistence) have been shown to be a dominant genetic trait(3) which is characterized in general by higher rates of synthesis of precursor LPH than found in individuals with lactase nonpersistence(4–6). Although initially controversial(7), most studies have found a correlation between lactase enzymatic activity and LPH mRNA levels, suggesting regulation at the level of mRNA abundance(6, 8–10). Recent evidence from analysis of polymorphisms within LPH exons in heterozyotes suggests that the persistence/nonpersistence polymorphism involves a cis-acting element with presumably a mutation in this element accounting for differences in LPH expression in adults(11). In addition, a few adults have been identified with hypolactasia which appears to be completely(4) or partially(5) the result of alterations of intracellular processing.

It occurred to us that transcriptional regulation might not be the sole physiologic determinant of lactase activity in children and that other mechanisms, most likely posttranslational, might account for differences in enzymatic activity, especially in view of the changes in diet and pancreatic function which occur in childhood. This possibility seemed particularly attractive in view of the evidence obtained in other mammalian species in which developmental down-regulation of lactase is universal. A number of studies, particularly in rats and rabbits, suggest that more than one mechanism is involved. Transcriptional regulation almost certainly contributes. Many, although not all, studies in rats and rabbits have shown a correlation between LPH mRNA abundance and LPH synthesis or lactase enzyme activity(12–14) during postnatal development, and the recent study by Krasinski et al.(13) has demonstrated an apparent decrease in LPH gene transcription after weaning. Transcriptional control is also suggested by the findings of Troelsen et al.(15) who used transgenic mice to show that a 1-kb segment of the LPH gene upstream of the transcriptional start site results in the down-regulation of a reporter gene at the appropriate developmental time. On the other hand, other investigators have reported that LPH mRNA does not fall during development(16) or at least the changes did not appear sufficient to completely account for the pattern of fall in enzymatic activity(7, 14, 17, 18). Evidence supporting additional mechanisms contributing to developmental lactase down-regulation has been found by a number of investigators. These mechanisms include changes in intracellular processing of precursor LPH(19, 20) and in degradation of brush border LPH(21). Because pancreatic proteases are responsible for surface removal of brush border hydrolases in mature animals(22), the changes in LPH degradation could reflect the development of exocrine pancreatic function.

The purpose of our study was to determine whether the heterogeneity of lactase expression in human children is solely the consequence of differences in mRNA abundance or whether additional mechanisms contribute.

METHODS

Children undergoing upper gastrointestinal endoscopy were studied. Duodenal mucosal biopsies were obtained from children undergoing endoscopy at Columbus Children's Hospital for clinical diagnostic purposes. Endoscopic indications included abdominal pain, vomiting or gastroesophageal reflux, or growth delay. The Olympus GIF-100 (Olympus America, Inc., Melville, NY) endoscope with a 2.8-mm Microvasive disposable biopsy forceps was used to obtain biopsies from the second and third portions of duodenum. In two subjects pancreatic exocrine function was tested before biopsy with duodenal fluid collected through the endoscope for 15-20 min after cholecystokinin (0.02 μg/kg) and secretin (1 clinical unit/kg). The two subjects were included in the final analyses, because both sucrase and lactase enzymatic activities from both fell within normal limits. Children with clinical, radiographic, or endoscopic small intestinal disease, nutritional deficiency, or histologic abnormalities were excluded from the study. Biopsies from the remaining 39 subjects were immediately placed in a cryovial, frozen, stored in liquid nitrogen, and shipped in two batches on dry ice to Madison via Federal Express. They were then stored in liquid nitrogen until processed.

For enzyme activity, biopsies were homogenized with a Polytron PT1200C(Brinkman, Westbury, NY) in 200 μL of 0.01 M sodium potassium phosphate, 0.15 M NaCl, pH 8.0, for 25 s on ice. Disaccharidase activities were measured according to Dahlqvist(23) with lactose and sucrose as substrates. Twenty microliters of homogenate were incubated with 20 μL of substrate in duplicate for 60 min at 37°C, and the reaction was terminated by the addition of 0.6 mL of 0.5 M Tris-HCl, pH 7.0, containing 22 mM sodiump-hydroxybenzoate, 0.5 mM 4-aminoantipyrine, 7 U/mL glucose oxidase, 0.5 U/mL peroxidase, and the color was allowed to develop for 10 min at 37°C. OD at 500 nm was read using a DU-64 spectrophotometer (Beckman, Fullerton, CA) with microcuvettes. Blanks, enzyme blanks, and glucose standards were treated similarly. Protein was measured using the Bio-Rad(Richmond, CA) protein dye reagent, and results were expressed as units per g of protein. Lactase activity was also normalized to sucrase enzymatic activity and expressed as a sucrase/lactase ratio(1) to account for possible differences in sampling site.

RNA was extracted by homogenizing biopsies in 150 μL of 4 M guanidinium thiocyanate, 25 mM sodium citrate, pH 7.0, 0.5% sarcosyl containing 7.2μL/mL β-mercaptoethanol and processed according to Promega (Madison, WI) technical bulletin no. 064. The final RNA pellet was suspended in 25 μL of 0.5% SDS and stored at -80°C. For Northern analysis 5 μg of total RNA was electrophoresed on a 1.2% agarose gel according to Kroczek(24). After photographing the ethidium bromide-stained RNA, the gel was soaked successively in 0.05 M NaOH for 30 min, 0.5 M Tris-HCl, pH 7.0, for 30 min, and 1 × SSC for 30 min before capillary transfer to a Hybond-N membrane (Amersham Corp., Arlington Heights, IL) overnight in 20 × SSC. After transfer the membrane was soaked for 10 min in 3 ×SSC, allowed to air dry for 30 min and then cross-linked using a UV Stratalinker 1800 (Stratagene, La Jolla, CA) on setting auto cross-link. Membranes were first probed for LPH mRNA using pHlac5, a generous gift of Drs. N. Mantei and G. Semenza(25), as described previously(6). Membranes were stripped with boiling 0.1% SDS with gentle agitation while the solution cooled to room temperature for approximately 40 min. A β-actin probe(26) was then used to probe the membrane for actin mRNA. Densitometry of bands on the autoradiograms was performed with a Zeineh Video Densitometer (Biomed Instruments, Inc., Fullerton, CA), and adjusted LPH mRNA density was calculated by dividing LPH mRNA densitometric units by actin mRNA densitometric units. Although in the past there was some concern about normalization to levels of actin mRNA because of possible developmental increases in actin expression(7), more extensive recent studies have shown no increase in actin mRNA in postnatal intestinal development(13, 18).

Each batch of biopsies was studied as described above, but to ensure comparable hybridization conditions, the membranes were stripped, and RNA from all biopsies was reprobed together for LPH and actin mRNA and presented in this report. No significant differences were noted in relative signal strengths between the first and second analysis.

RESULTS

Lactase and sucrase enzymatic activities were measured in all 39 biopsies. We excluded three biopsies from further study because low sucrase-specific activities (11, 17, and 14 units/g of protein) suggested the possibility of subtle mucosal injury. In addition we excluded 13 more biopsies because of concern about the integrity of the RNA after examining the ethidium bromide-stained gels after electrophoresis-either the ribosomal bands were not sharp or there was not an approximately two to one density ratio of the 28 S to the 18 S band. Thus our report concerns the remaining 23 biopsies.

Figure 1 demonstrates the relationship of lactase enzymatic activity to age. Subjects with lactase activities of less than 13 U/g of protein are considered to represent the lactase nonpersistence phenotype(1) and are plotted in the shaded area of the figure. Because a sucrase/lactase activity ratio of more than 4 may be a better indicator of hypolactasia(1), such individuals are indicated by the square symbols. We identified five individuals with hypolactasia, four of whom had sucrase/lactase ratios of 4 or greater. None of these individuals was less than 7 y of age, consistent with other reports that primary hypolactasia does not appear before the age of 3 y. Our failure to identify more individuals with hypolactasia than we did no doubt reflects the racial background of these children, only one of whom is black and the great majority Caucasian. We found no evidence of an age-related decline in lactase activity affecting all individuals.

Figure 1
figure 1

Relationship of lactase enzymatic activity to subject age. Values in the shaded area represent lactase nonpersistence. The square symbols indicate values from biopsies where the sucrase/lactase activity ratio was greater than 4.

Full size image

To determine the role of mRNA abundance in the regulation of lactase activity in these children we studied the correlation between LPH mRNA and enzyme activity. We used a single step technique to isolate RNA because of the limited amount of tissue available. Nevertheless the mean 260/280 absorbance ratio of RNA isolated from the biopsies was 1.7 ± 0.03 (SEM), indicating a high degree of purification. In general the quality of RNA was excellent, but we excluded 13 samples as noted above. After transfer to a Hybond-N membrane (Amersham Corp., Arlington Heights, IL), the RNA was probed with a radiolabeled LPH cDNA. To compensate for variations in the amount of total RNA probed, the membranes were also probed with an actin cDNA, both signals quantitated by densitometry, and an adjusted LPH mRNA level calculated by dividing the LPH signal by the actin signal. The ethidium bromide-stained gels together with Northern analysis are shown in Figure 2. The relationship between enzymatic activity and adjusted mRNA level is shown in Figure 3. As can be seen, we were able to find no relationship between LPH mRNA abundance and lactase enzymatic activity(r = 0.33, p > 0.1). A similar lack of relationship was found when we plotted sucrase/lactase activity ratios against adjusted LPH mRNA levels.

Figure 2
figure 2

Northern analysis of total RNA isolated from biopsies.Panel A demonstrates the ethidium bromide stained gels, panel B the signal obtained with the LPH cDNA probe, panel C the signal obtained with the actin probe, and panel D lactase specific activity(units/g of protein).

Full size image
Figure 3
figure 3

The relationship between lactase enzymatic activity and adjusted LPH mRNA abundance (calculated as described in“Methods”). Values in the shaded area represent lactase nonpersistence. The square symbols indicate values from biopsies where the sucrase/lactase activity ratio was greater than 4. Two virtually identical points with lactase specific activities of 14 and adjusted LPH mRNA levels of 0.50 and 0.46 are nearly superimposed.

Full size image

DISCUSSION

Until the 1960s it was assumed that humans, unlike other mammals, retained high levels of intestinal lactase throughout life. Hypolactasic individuals were identified, however(27, 28), but considered rare. Not until population studies beginning with the work of Cuatrecasaset al.(29) was it realized that the lactase nonpersistence phenotype is in fact the most common one except in certain populations in which a high frequency of an autosomal dominant trait leads to lactase persistence(3). Thus, most human children undergo a developmental fall in lactase activity which corresponds to the decline in other mammalian species.

Despite considerable study, the mechanisms responsible for the developmental regulation of lactase are not completely understood. As summarized in the introduction section, it is evident that both transcriptional and posttranscriptional mechanisms contribute in nonhuman mammals. It is also likely that the dominant mechanism differs between different regions along the longitudinal axis of the small intestine, between regions along the crypt-villus axis, and even from enterocyte to enterocyte of a single villus(13, 17, 18, 30, 31). It is even possible that the major mechanisms differ between species as some have suggested. The nature of the reported posttranslational mechanisms remains undefined although the evidence of increased rates of LPH degradation(21) after weaning suggests the possibility that pancreatic protease secretion may be responsible. Pancreatic proteases appear to be responsible for the surface removal and rapid turnover of brush border hydrolases in general(22), and, in fact, human adults with exocrine pancreatic insufficiency are known to have higher levels of disaccharidases than normal controls(32). Because the development of pancreatic secretion in rats is temporally related to weaning(33), pancreatic proteases may be important in lactase regulation in at least some species.

Although there are no prior studies of the mechanism of lactase regulation in human children to our knowledge, there are a number of studies which address the nature of the persistence/nonpersistence polymorphism in adults. Although recent studies have demonstrated a complex distributional pattern of gene expression in duodenal biopsies in lactase nonpersistence with a patchy distribution of both LPH(34, 35) and LPH mRNA(36) even on a single villus, most of the evidence, summarized in the introduction section, indicates a good correlation between LPH mRNA and enzymatic activity at least in mucosal biopsy samples of the duodenum. Thus LPH mRNA abundance probably has central importance in lactase regulation in adults, a conclusion which is also consistent with recent evidence indicating that the persistence/nonpersistence polymorphism in adults is controlled by a cis-acting element(11).

To what extent does the evidence indicating the importance of mRNA in regulating LPH in adults bear on the mechanism of lactase regulation during human development? More than one mechanism may regulate lactase, and we hypothesized that differences in LPH gene transcription which might explain the persistence/nonpersistence polymorphism in adults might be independent of posttranslational mechanisms operative in children which could relate to changes in diet or exocrine pancreatic function. The studies implicating posttranslational mechanisms during weaning in nonhuman mammals made this hypothesis particularly attractive to us.

We therefore studied duodenal biopsies in children undergoing diagnostic upper gastrointestinal endoscopy. It would not have been possible for ethical considerations to biopsy asymptomatic children, and all of our subjects had gastrointestinal symptoms. It is therefore conceivable that some of our subjects had unrecognized intestinal disease. The possibility seems unlikely, however, because we excluded subjects with clinical, radiographic, endoscopic, or histologic evidence of intestinal disease or evidence of significant nutritional problems. We also excluded from further analysis three additional subjects in whom low sucrase levels suggested the possibility of a secondary disaccharidase deficiency. We found a marked variation in lactase enzymatic activity from subject to subject with no evidence of a general down-regulation of the enzyme with age. A group of five individuals had lactase activities of less than 13 U/g of protein, thus meeting the usual criterion of the lactase nonpersistence phenotype; four of these met the more strict criterion of a sucrase/lactase activity ratio of 4 or greater. None of our hypolactasic subjects was less than 7 y of age. These findings are consistent with other reports of lactase activity in mucosal biopsies which suggest that the nonpersistence phenotype does not appear before the age of 3 y(1, 2), at least in the populations of largely Caucasian subjects studied. When the relationship of enzymatic activity to LPH mRNA was examined we found, in striking contrast to the findings in adults of our own group and others, little evidence that enzymatic activity was regulated by message level. The five subjects with hypolactasia were particularly interesting. Although LPH mRNA levels were in fact somewhat lower than in many of the subjects with lactase persistence, we found many subjects with similar lower message levels who expressed high levels of enzymatic activity.

It is difficult to place our findings in context with our current understanding of the regulation of human lactase. Although we have been convinced that lactase regulation in adults, at least in the duodenum, is primarily at the level of mRNA abundance, these findings suggest regulation at least in part by posttranslational mechanisms. It is possible that some of the differences reflect mechanisms operative in somewhat more distal intestine biopsied in children than ordinarily biopsied in adults. Whatever the explanation, our findings emphasize the complexity of lactase regulation and suggest that in children the enzyme is regulated at least in part by posttranslational mechanisms.