In order to study the effect of glucose on the differentiation of cultured human colon cancer cells, a subpopulation of HT-29 cells was selected for its capacity to grow in the total absence of sugar. These cells (Glc-cells) exhibit, after confluency, an enterocytic differentiation, in contrast to cells grown with glucose (Glc+ cells), which always remain undifferentiated. The differentiation is characterized by a polarization of the cell layer with apical brush borders and tight junctions, and by the presence of sucrase-isomaltase. The differentiation of Glc- cells is reversible: the addition of glucose to postconfluent cultures of Glc- cells results in an inhibiting effect on the expression of sucrase-isomaltase; switching growing cultures of Glc- cells to the Glc+ medium for several passages results in a progressive reversion to the undifferentiated state, which is completed after seven passages. The dedifferentiation process is associated with a parallel, passage-related, increase in the rates of glucose consumption and lactic acid production, and decreases of intracellular glycogen content, which return to the values of the undifferentiated original Glc+ cells. The values of these metabolic parameters are correlated, at each passage, with the degree of dedifferentiation of the cells. When these dedifferentiated cells, after having been cultured in Glc+ medium for 20 passages, are switched back to the Glc- medium, they readily grow without mortality, and reexpress the same enterocytic differentiation as the parent Glc- cells. These results show that the capacity of this subpopulation to grow and differentiate in the absence of sugar is a stable characteristic. They further suggest that glucose metabolism interferes with the program of differentiation of HT-29 cells.

The presence of sucrase-isomaltase (SI), a glycoprotein hydrolase normally restricted to the brush border membrane of the enterocytes of the small intestine, was investigated in tumours which developed in nude mice inoculated with six human colon carcinoma cell lines (HT-29, Caco-2, HRT-18, HCT-8R, SW-480, and CO-115). Foetal and normal adult human small intestines and colons were used as controls. SI was studied by (1) immunofluorescence with rabbit antibodies raised against purified human small intestine SI; (2) polyacrylamide gel electrophoresis and immunoblotting; and (3) determination of the enzyme activity. SI was antigenically present, and enzymatically active, in all the tumours derived from Caco-2 and HT-29 cells. The presence of the enzyme was associated with that of typical brush borders at transmission electron microscopy examination. SI was absent from the tumours developed with the other four cell lines, as well as from the normal adult colon mucosa. SI was also present and active in the colons of mid-gestation foetuses, ranging in ages between 20 and 28 weeks; it was absent from the colons of late-gestation foetuses. The presence of SI in tumours derived from two cell lines suggests that this enzyme is a marker, so far unsuspected, of certain human colon cancers, and that the differentiation pattern of these particular cancers closely resembles that of the foetal colon.

D-glucose transport across the intestinal brush-border membrane involves two transport systems designated here as systems 1 and 2. Kinetic properties for both D-glucose and methyl a-D-glucopyranoside transport were measured at 350C by using brush-border membrane vesicles prepared from either control, fasted (48 hr), or semistarved (10 days) animals. The results show the following: (i) The sugar influx rate by simple diffusion was identical under either altered condition. (ii) Semistarvation stimulated D-glucose uptake by system 2 (both its V__ and Km increased), whereas system 1 was untouched. (iii) Fasting increased the capacity of system 1 without affecting either Km of system 1 or V..: and Km of system 2. The effect of fasting on Vx,,,, of system 1 cannot be attributed to indirect effects from changes in ionic permeability because the kinetic difference between control and fasted animals persisted when the membrane potential was short-circuited with equilibrated K+ and valinomycin. This work provides further evidence for the existence of two distinct sodium-activated D-glucose transport systems in the intestinal brush-border membrane, which adapt independently to either semistarvation or fasting.D-glucose transport in the small intestine responds to a variety of pathophysiological conditions that include qualitative and quantitative modifications of the diet, small-bowel resection, and diabetes. Often, however, published results are contradictory. For instance, under apparently similar conditions total sugar absorption has been described as either increasing, decreasing, or exhibiting no change (for reviews, see refs. 1-3). No satisfactory explanation of these discrepancies thus far exists, but many variables that have escaped appropriate control may be involved. Furthermore, the procedures of analysis and data expression differ widely and complicate the situation (3-7).An additional source of confusion is the heterogeneity in intestinal transport. Although the suggestion that two Dglucose transport systems occur in the apical border of the intestine has existed since the work of Honegger and Semenza (for review, see ref. 8), D-glucose absorption is usually treated as involving a single homogeneous transport system, the D-glucose/Na+ cotransporter identified in the sixties (for reviews, see refs. 8-11). Consequently, work on intestinal adaptation has been concerned with overall transport function of the intestine, providing no answers to whether individual transport systems were being selectively affected. In this paper we describe experiments permitting such a diagnosis.We demonstrated recently that D-glucose transport across the intestinal brush-border membrane involves at least two distinct, sodium-activated transport agencies (12-14). Although we identified the first, system 1 (S-1), as being identical with the classical D-glucose/Na' cotransporter (12), the exact nature of system 2 (S-2) remains to be established.[In this paper we use our own classification (12-14); although two distinct D-glucos...