Fructose-responsive genes in the small intestine of neonatal rats

Xue-Lin Cui1, Patricia Soteropoulos2,3, Peter Tolias2 and Ronaldo P. Ferraris1

1 Department of Pharmacology and Physiology, University of Medicine and Dentistry of New Jersey (UMDNJ)-New Jersey Medical School, Newark 07103-2714
2 Center for Applied Genomics, Public Health Research Institute, Newark 07103-3506
3 Department of Microbiology and Molecular Genetics, UMDNJ-New Jersey Medical School, Newark, New Jersey 07103-2714


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
The intestinal brush border fructose transporter GLUT5 (SLC2A5) typically appears in rats after weaning is completed. However, precocious consumption of dietary fructose or in vivo perfusion for 4 h of the small intestine with high fructose (HF) specifically stimulates de novo synthesis of GLUT5 mRNA and protein before weaning is completed. Intermediary signals linking the substrate, fructose, to GLUT5 transcription are not known but should also respond to fructose perfusion. Hence, we used microarray hybridization and RT-PCR to identify genes whose expression levels change during HF relative to high-glucose (HG) perfusion. Expression of GLUT5 and NaPi2b, the intestinal Na+-dependent phosphate transporter, dramatically increased and decreased, respectively, with HF perfusion for 4 h. Expression of >20 genes, including two key gluconeogenic enzymes, glucose-6-phosphatase (G6P) and fructose-1,6-bisphosphatase, also increased markedly, along with fructose-2,6-bisphosphatase, an enzyme unique to fructose metabolism and regulating fructose-1,6-bisphosphatase activity. GLUT5 and G6P mRNA abundance, which increased dramatically with HF relative to HG, {alpha}-methylglucose, and normal Ringer perfusion, may be tightly and specifically linked to changes in intestinal luminal fructose but not glucose concentrations. G6P but not GLUT5 mRNA abundance increased after just 20 min of HF perfusion. This cluster of gluconeogenic enzymes and their common metabolic intermediate fructose-6-phosphate may regulate fructose metabolism and GLUT5 expression in the small intestine.

gluconeogenesis; glucose-6-phosphatase; GLUT5; regulation; glucose-6-phosphate translocase


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
THE INTESTINAL BRUSH BORDER fructose transporter GLUT5 (SLC2A5) is apparently underexpressed in rats, rabbits, and humans (3) during early development, potentially causing malabsorption of fructose. In fact, consumption of fruit juice containing high levels of fructose results in infantile colic (10) and carbohydrate malabsorption, particularly in newborns (27). In rat pups, mRNA expression and transport activity of GLUT5 appear late during development, typically after weaning is completed (4, 9, 28, 31). However, precocious consumption of dietary fructose by weaning rats or in vivo perfusion of the small intestine for 4 h with high-fructose (HF) solutions can dramatically enhance GLUT5 expression and activity well before weaning is completed (8, 25, 28). The effect of fructose on GLUT5 is highly specific: only luminal fructose increases fructose transport, and only brush border fructose transport increases with fructose perfusion. Activity and expression of the intestinal glucose transporter SGLT1 are also greatest during the weaning stage (22) but are not affected by fructose perfusion. Fructose induction of GLUT5 requires de novo synthesis of GLUT5 mRNA and protein (20). GLUT5 is then a gene that responds to the luminal presence of its substrate, fructose, if that substrate is made available earlier in development. Luminal fructose, however, must be linked to GLUT5 by factors that also respond to fructose perfusion. In this study, we used microarray hybridization to identify those genes whose expression changes during fructose perfusion. Glucose perfusion was used as control, because it did not affect GLUT5 expression and activity in weaning rats (8, 20, 29).

There have been only five microarray studies on mammalian small intestine, and there is as yet no microarray study on the response of the small intestine to a nutrient signal. This is surprising, since the small intestine constitutes the first barrier to nutrient metabolism in all mammals and is therefore strategically positioned to provide the major regulatory mechanism(s) that may be altered by nutrient availability. There has been one study on the effect of protein quantity and quality on rat hepatic gene expression, but the RNA from several rats was pooled so that results were based on only a single microarray (12). In this study, we found many genes along with GLUT5 to change in expression after a 4-h perfusion with fructose, indicating that many signaling and metabolic pathways in the small intestine are responsive to the presence of fructose in the lumen. To determine potential signals that may increase shortly after initiation of fructose perfusion, we also determined gene expression in rat intestines perfused for only 20 min. Here, we found a smaller number of genes whose expression changes with fructose perfusion.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Experimental Design
Rats, 20–22 days old, were divided into four groups and perfused for different periods of time and with different sugar solutions: 1) 4-h perfusion with high glucose (HG) (100 mM in Ringer solution); 2) 4-h perfusion with HF (100 mM); 3) 20-min perfusion with HG; and 4) 20-min perfusion with HF. The perfusion conditions were based on previous studies from this laboratory and were physiological. For example, the time course and magnitude of increases in GLUT5 mRNA abundance and activity were similar between free-living rats fed with 60% fructose pellets and anesthetized rats whose intestines were perfused with 100 mM fructose (20). When measured during the day, the average intestinal luminal concentration of fructose in rats fed HF pellets was 30 mM. However, rats are nocturnal, so peak luminal sugar concentrations at 8 PM can average 100 mM (15). We decided to use the highest average luminal sugar concentration, because GLUT5 expression is directly proportional to intestinal luminal fructose concentration (28).

For the microarray study, we chose to perfuse fructose, because it was the only nutrient known to stimulate GLUT5, and it did not affect the mRNA expression of other brush border sugar transporters. Glucose perfusion was an excellent control, because it did not affect GLUT5 expression and activity.

In a subsequent experiment to determine the regulation of glucose-6-phosphatase (G6P) expression, rat intestines were perfused with HF, HG, 100 mM {alpha}-methylglucose (a nonmetabolizable, SGLT1-transportable glucose analog; Ref. 11) in Ringer and normal Ringer alone. The osmolalities of all perfusion solutions were the same (290 mosmol/kgH2O), with fructose, glucose, and {alpha}-methylglucose replacing NaCl (see section on perfusion model for Ringer composition).

Although GLUT5 expression is known to peak at 4 h after HF perfusion, other intestinal genes may respond at a shorter time intervals. The choice of 20-min perfusion was based on a previous study indicating that the mRNA and protein expression of the immediate-early genes c-fos and c-jun in epithelial cells peaked at this duration when the intestine was perfused with either glucose or fructose (21).

An everted sleeve method was employed for determining glucose and fructose uptake rates. Relative quantitative reverse transcription-polymerase chain reaction (RT-PCR) was initially used to examine GLUT5 mRNA abundance in perfused intestinal tissues, to ensure that tissues used for subsequent (and labor-intensive) microarray experiments responded in a manner similar to tissues in previous work. An oligonucleotide microarray was used for screening the genes that were affected by fructose and glucose perfusion. Finally, some of the up- and downregulated genes were confirmed by semiquantitative RT-PCR.

Animals
Adult male and female Sprague-Dawley rats weighing ~200 g were purchased from Taconic (Germantown, NY) and bred. Rats were housed in a temperature-controlled room (22–24°C) with a 12:12-h light-dark cycle in the research animal facility, and allowed access to water and food ad libitum (Purina Mills, Richmond, IN). After the female rats became pregnant, they were separated from the male rats and carefully monitored until the pups were born. The time and date of birth were recorded; age at birth was considered day 0. Midweaning rat pups were used in this study. The pups were kept with their dams until they were 20–22 days old, and then they were randomly selected according to experimental needs. All the procedures conducted in this study were approved by the Institutional Animal Care and Use Committee, UMDNJ-New Jersey Medical School.

Perfusion Model
Rat intestinal perfusion was conducted following the method of Jiang and Ferraris (20). Briefly, rat pups (20–22 days old, not starved) were anesthetized by intraperitoneal injection of urethane (1.4 g/kg body wt). After its contents were flushed, the small intestine was continuously perfused with solution (100 mM glucose or fructose in Ringer) at a rate of 60 ml/h at 37°C using a peristaltic pump. The composition of the perfusion solution was (in mM) 78 NaCl, 4.7 KCl, 2.5 CaCl2·H2O, 1.2 MgSO4, 19 NaHCO3, 2.2 KH2PO4, and 100 sugar. Perfusion for the microarray experiment was performed in sets of two (one perfused with HG, another with HF). Perfusion for the G6P experiment (see Fig. 5) was done in sets of four (normal Ringer, {alpha}-methylglucose, fructose, glucose) parallel perfusions. Rats perfused within a set were littermates subjected to exactly the same conditions except for differences in perfusion solution.



View larger version (31K):
[in this window]
[in a new window]
 
Fig. 5. The expression of G6P in the small intestine of neonatal rats perfused with normal Ringer (KRB) and Ringer with 100 mM {alpha}-methylglucose (AM), glucose (HG), and fructose (HF) isosmotically replacing NaCl. Perfusion lasted for 4 h, and mRNA abundance was determined by Northern blot. 18S rRNA was used as loading and transfer control. The ratio of G6P mRNA abundance to 18S rRNA abundance in KRB-perfused intestine was designated as 100%, and the ratios in AM, HG, and HF were subsequently normalized to this value. A: representative blot. B: mean relative abundance ± SE (n = 5). HF perfusion dramatically increased G6P mRNA abundance, whereas HG perfusion had no effect.

 
Fructose and Glucose Uptake Measurements
After each experiment, the jejunum was immediately isolated and gently flushed with ice-cold Krebs-Ringer bicarbonate solution (20). For each pup, a series of four 1-cm jejunal segments was everted and mounted on a grooved steel rod (3-mm diameter) and preincubated at 37°C for 5 min in Ringer solution bubbled with 95% O2-5% CO2. The sleeves were then incubated in solutions containing either 10 µCi D-[14C]glucose (NEN, Boston, MA) for 1 min or 10 µCi of D-[14C]fructose for 2 min. We used 20 µCi of L-[3H]glucose to correct for adherent fluid and passive diffusion of glucose or fructose.

Total RNA Extraction
The total RNA was extracted from 100 mg tissue (stored in –80°C) using RNeasy Midi kits (Qiagen, Valencia, CA) according to the manufacturer’s instructions. The RNA sample was resuspended in RNase-free water and stored at –80°C in a freezer for the subsequent RT-PCR, Northern blot, and microarray analysis.

Microarray
The design, analysis, and interpretation of this microarray experiment followed in general the suggestions of other microarray studies (24, 33, 37). The detailed protocols and primary data from this study are available at the Gene Expression Omnibus database (GEO; http:/www.ncbi.nlm.nih.gov/geo), under accession number GSE996.

Since spotted microarrays allowed only two treatments to be compared directly, we were constrained to use glucose as control not only because its perfusion (20) and consumption (8) had no effect on the gene of interest GLUT5, but also because of the following reasons. Unperfused controls were not useful since intestinal perfusion even by a nonmetabolizable, nonabsorbable sugar such as mannitol triggered marked responses by many genes in the small intestine (21). Perfusion with a nonmetabolizable sugar as control also had a problem with distinguishing fructose-response genes from numerous metabolic-response genes. In fact, many genes were nonspecifically regulated by starvation and sugar metabolism (39).

Because our microarray did not contain oligonucleotides representing the entire rat genome, the list of genes was far from complete but nonetheless identified for the first time genes responsive to luminal fructose. Expression of these genes must be specifically and directly or indirectly enhanced by luminal fructose because we used as control intestines perfused with glucose, eliminating from consideration the genes that responded nonspecifically to luminal sugars. Conversely, by using fructose as the reference, readers interested in glucose metabolism can interpret the downregulated genes as genes that were upregulated with glucose perfusion.

Array design description.
We used a rat 5,000-oligonucleotide array [Rat 5K Oligo Array (NGEL 2.0.1); Center for Applied Genomics, PHRI, Newark, NJ]. A detailed description of array design elements is available from the following web sites: http://www.cag.icph.org and http://www.labonweb.com. The array is built from 4,854 oligonucleotides representing 4,803 independent genes. All oligonucleotides are 60–70 nucleotides in length. Oligonucleotides were printed in 3x SSC at a concentration of 40 µM on poly-L-lysine-coated glass slides (http://www.microarrays.org) using a GeneMachines OmniGrid 100 microarrayer (Genomic Solutions, Ann Arbor, MI). Arrays were stored at room temperature in a desiccator until use.

Labeling and hybridization procedures and parameters.
The 3DNA Submicro Oligo Expression Array Detection Kit (Genisphere, Hatfield, PA) was employed to synthesize and label the complementary DNA (cDNA) with fluorescent dyes Cy3 or Cy5. The indirect labeling method was chosen because it is more sensitive and consistent than the direct labeling method or other indirect techniques (38). In the preliminary experiments, 3, 4, and 5 µg of total RNA was tested for labeling and hybridization conditions. In the actual experiments, 4 µg of total RNA from each sample was used for labeling and hybridization, because at this amount of total RNA, the hybridization signal-to-noise ratio was high, signals from almost all modestly expressed genes were detectable, and signals from highly expressed genes were not fully saturated.

Experiment description.
In the first experiment, rat intestines were perfused with HF (n = 5) or HG (n = 5) for 4 h. RNA from fructose-perfused intestines was labeled with Cy5, and RNA from glucose-perfused intestines with Cy3. To eliminate dye-bias, one randomly selected fructose-perfused intestine was labeled in the opposite manner with Cy3, as was the corresponding glucose-perfused intestine with Cy5. Results from this dye-flip were used to correct results as described later. In the 4-h perfusion experiment, we decided to label one-way only and eliminated dye bias with a dye-flip experiment because we had an odd number of paired samples, and the number of Cy3-labeled fructose-perfused intestines would never be the same as that of Cy3-labeled glucose-perfused intestines.

In the second experiment, rat intestines were perfused with HF (n = 4) or HG (n = 4) for 20 min. RNA from two HF-perfused intestines was labeled with Cy5, and their corresponding HG-perfused intestines were labeled with Cy3. Conversely, RNA from another two HF intestines was labeled with Cy3, and their corresponding HG intestines were labeled with Cy5.

To identify additional potentially false-positive genes, three randomly chosen samples from HF- and HG-perfused intestines were used for "self-against-self" experiments, whereby RNA from each sample was labeled with both dyes. Since the dye-labeled samples were identical, these hybridizations should ideally produce similar intensities for every spot for both Cy3 and Cy5 channels. Under these conditions, no gene should be over- or underexpressed, and spots exhibiting significant differences in intensity between the two channels were eliminated from consideration (see Data processing below). Finally, two "capture-primer" experiments were performed where Cy3 and Cy5 capture sequences with no sample RNA were loaded onto the array. Since the capture sequences should not correspond to any of the oligonucleotide sequences on the array, ideally, there should be no fluorescence for each microarray spot for both Cy3 and Cy5 channels.

Data processing.
Slides were scanned for Cy3 and Cy5 fluorescence using an Axon GenePix 4000A microarray reader (Axon instruments, Union City, CA). The density of the spots was detected using GenePix 4.0 (Axon Instruments). When measuring the density of the spots, the ratio of Cy3 and Cy5 dyes was adjusted to 1 as a global normalization for each sample slide. The results (median pixel density of each array spot was used) were normalized by LOWESS (locally weighed scatter smoother) method so that the results can be compared among the experiments performed at different times (37).

The following rules were made to eliminate genes that were considered false positives and to choose genes that might change significantly with fructose perfusion. First, genes in array spots with a density over twofold that of background in the 3DNA capture primer experiment were eliminated, because these genes should not fluoresce above background, as no sample RNA was added. Second, in the actual array experiments, genes whose expression was so low that their average median density was less than twofold that of background density were also eliminated (23). Third, genes whose expression changed by >50% in any one of three self-self hybridization experiments were eliminated, because these genes should not change in expression when Cy3 and Cy5 labeled the same sample. Fourth, nutrient-induced changes in gene expression should not be altered when a different dye was used to label the same sample. To minimize dye bias, any gene whose expression changed from a positive value (in this study, defined as gene expression was greater in HF- over HG-perfused intestine) to a negative value when the dye was switched was eliminated. Conversely, any genes whose expression changed from a negative to a positive value (gene expression greater in HG- over HF-perfused intestine) when the dye was switched were eliminated. Fifth, only genes that changed by ≥50% in at least three of five comparisons in the 4-h perfusion experiment, and in two of four comparisons in the 20-min perfusion experiment, were considered. Although genes whose expression changed markedly (>>50%) with sugar perfusion might initially be more interesting, we chose a 50% change as guide, because we also wanted to examine genes whose expression might change moderately, but consistently, with HF (or HG) perfusion.

RT-PCR
A OneStep RT-PCR kit (Qiagen) was used in this experiment to verify changes in expression as detected by microarray analysis. Briefly, the reaction was performed on a thermal cycler (Techne, Duxford, Cambridge, UK) for 30 min at 50°C, 15 min at 95°C for reverse transcription reaction and activation of Taq DNA polymerase. The DNA amplification procedure was set at 45 s, 94°C (denature); 45 s, 60–65°C (annealing); and 90 s, 72°C (extension) for a total number of cycles optimal (see next paragraph) for PCR amplification of different genes. Pictures of the gels were taken using a CCD camera for determination of the density of the bands by a densitometer (FluorChem 8800; Alpha Innotech, San Leandro, CA). The relative amount of mRNA in the samples for each target gene measured was normalized by a housekeeping gene, GAPDH, and presented as a ratio to GAPDH (16, 26, 35).

Determination of Optimal PCR Cycles for Each Gene
The number of PCR cycles optimal for determining relative changes in mRNA abundance varied. An exponential relationship between RT-PCR products (represented by ethidium bromide density) and DNA amplification cycles was demonstrated between 21 and 33 cycles for GLUT5, 21 and 27 cycles for NaPi2, and 18 and 27 cycles for GAPDH. The optimal number was considered to be the midpoint of the exponential phase, or 27 cycles for GLUT5, 24 for NaPi2, and 23 for GAPDH. We measured the optimal number of cycles for 20 candidate genes whose mRNA abundance were confirmed using semiquantitative RT-PCR (Table 1).


View this table:
[in this window]
[in a new window]
 
Table 1. Sequence of primers used for the determination of optimal number of PCR cycles for measuring gene expression

 
Northern Blot Analysis
At the end of each perfusion, about 10 cm of jejunum 15 cm from the ligament of Treitz was quickly frozen in liquid nitrogen and then stored at –80°C for later G6P and 18S rRNA (loading and transfer control) abundance measurement by Northern blotting following protocols previously described (7). The G6P probe was obtained from the 907-bp RT-PCR product (Table 1).

Statistical Analysis
Data are presented as means ± SE. Paired student t-test was used to determine perfusion-related differences in fructose and glucose uptake, and in the relative amounts of gene expression in RT-PCR experiments using log10 of the ratio of the target gene to the housekeeping gene, GAPDH. A one-sample t-test was used to determine significance of changes in gene expression in the microarray results. The values tested were derived from the GenePix algorithm M = log2(R/G), where R = red channel signal = F635; G = green channel signal = B532, and the hypothetical mean was 0 (meaning no change between HF and HG). For consistency with the microarray results, the uptake rates and gene expression results from RT-PCR were converted to ratios of HF to HG. Concordance between the results of RT-PCR and the microarrays was evaluated by correlation analysis. P < 0.05 was considered statistically significant. Statistical analysis was conducted using the StatView program (Abacus Concepts, Berkeley, CA).


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Fructose Perfusion for 4 h but not 20 min Increased Fructose Uptake Rate
Fructose uptake rate increased by approximately twofold (P < 0.01) in the intestines of 20-day-old pups perfused with HF solutions for 4 h (Fig. 1A). In contrast, fructose perfusion for 4 h had no effect on glucose uptake rate (P > 0.05). When perfusion duration was reduced to 20 min, HF and HG solutions did not (P > 0.05) enhance fructose uptake rate (Fig. 1B).



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 1. The relative rates of fructose and glucose uptake in the small intestine of neonatal rats perfused with high-glucose (HG) and high-fructose (HF) solutions (100 mM) for 4 h (A) and for 20 min (B). Bars are means ± SE (n = 5). Uptake in HG-perfused pups was designated as 100%, and the uptake in HF-perfused pups was normalized to this value. *Significant difference between HG and HF groups. Fructose perfusion for 4 h increased rates of fructose, but not glucose, uptake.

 
Fructose Perfusion for 4 h Increased GLUT5 mRNA Abundance
The expression of GLUT5 mRNA increased by almost fivefold (P < 0.01) in the intestines of 20-day-old pups perfused with HF solutions for 4 h (Fig. 2, A and C). These results confirmed findings for GLUT5 by microarray analysis of RNA from the same tissues (Tables 2 and 4). When perfusion duration was reduced to 20 min, the HF solution failed to enhance GLUT5 mRNA abundance as measured by RT-PCR (P > 0.05) (Fig. 2, B and C), also confirming the microarray findings (Table 4C). The magnitude of fructose-induced increases in GLUT5 mRNA abundance measured by microarray and RT-PCR (~3- to 5-fold) was similar to that of fructose-induced increases measured by Northern blots and in situ hybridization (19).



View larger version (45K):
[in this window]
[in a new window]
 
Fig. 2. The expression of GLUT5 mRNA in the small intestine of neonatal rats perfused with HG and HF solutions as determined by RT-PCR. Gel pictures of 4-h (A) and 20-min (B) perfusion. M, DNA size marker. C: mean relative abundance ± SE (n = 5). The expression of GLUT5 (represented as the ratio of GLUT5 to GAPDH) in HG-perfused pups was designated as 100%, and then its expression in HF-perfused pups was normalized to this value. *Significant difference between HG and HF groups. Fructose perfusion for 4 h increased GLUT5 mRNA abundance.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Genes whose mRNA abundance, as estimated by microarray, upregulated with intestinal fructose perfusion for 4 h

 

View this table:
[in this window]
[in a new window]
 
Table 4. Ratio of mRNA abundance as measured by RT-PCR and microarray for selected genes

 
Differentially Expressed Genes in 20-day-old Rat Small Intestines Perfused with HG or HF for 4 h
Upregulated genes.
Using HG-perfused intestines as control, we identified 24 candidate genes that were upregulated when intestines of 20-day-old rats were perfused with HF for 4 h (Table 2). We found five genes that consistently increased with HF perfusions by ≥50% for each of five experiments, eight genes that increased ≥50% for four of five experiments, and 13 genes that increased for three of five experiments. We then selected five of the most upregulated genes [GLUT5, G6P, fructose-1,6-bisphosphatase, activating transcription factor 4, and cyclic nucleotide-gated channel (specifically, CNG3)] to determine whether the upregulation indicated by microarray would be confirmed by semiquantitative RT-PCR (Fig. 3A). Except for CGN3, the average HF/HG ratio of expression of other genes determined by microarray was similar to that determined by RT-PCR (Table 4A). The P values of the RT-PCR results for increases in GLUT5, G6P, fructose-1,6-bisphosphatase, and activating transcription factor 4 were P < 0.01, P < 0.05, P < 0.05, and P < 0.001, respectively, and were comparable to the P values of the microarray results (Table 2). In contrast, there was a disconcordance between microarray (statistically significant, Table 2) and RT-PCR (insignificant, P > 0.05) results for CNG3.



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 3. The expression as determined by RT-PCR of glucose 6-phosphatase (G6P), fructose-1,6-bisphosphatase (FBP), activating transcription factor 4 (ATF4), sodium-phosphate cotransporter NaPi2b, testicular ecto-ATPase (ECTO), and T clone 15 (TC15) mRNA in the small intestine of neonatal rats perfused with HG and HF solutions for 4 h. A: representative upregulated genes. B: representative downregulated genes. Values are mean relative abundance ± SE (n = 5). The expression of each gene (represented as the ratio of each gene to GAPDH) in HG-perfused pups was designated as 100%, and then the expression of each gene in HF-perfused pups was normalized to this value. *Significant difference between HG and HF groups. Fructose perfusion for 4 h increased G6P, FBP, and ATF4 mRNA abundance, and decreased NaPi2, ECTO, and TC15 mRNA abundance.

 
Downregulated genes.
Twenty-six genes were potentially downregulated in HF-perfused intestine of 20-day-old rats, which included three genes underexpressed by 50% for five of five experiments, as well as five genes consistently underexpressed by 50% in four of five experiments (Table 3). We selected three highly downregulated (as measured by microarray) genes in HF-perfused intestines, NaPi2, ecto-ATPase (ECTO), and T clone 15, then redetermined their expression by RT-PCR. Expression of these three genes clearly decreased markedly with HF as shown by RT-PCR (Fig. 3B, Table 4B): NaPi2 (P < 0.01), ECTO (P < 0.01), and T clone 15 (P < 0.05). Hence, the microarray and RT-PCR results were similar.


View this table:
[in this window]
[in a new window]
 
Table 3. Genes whose mRNA abundance, as estimated by microarray, downregulated with intestinal fructose perfusion for 4 h

 
Differentially Expressed Genes in 20-day-old Rat Small Intestines Perfused with HG and HF for 20 min
From the microarray results, eight candidate genes seemed upregulated (≥50% in at least two of four comparisons) in rat intestine perfused with HF for 20 min, but only two (G6P and cyclin D1) reached statistical significance (Table 5A). The degree of overexpression of G6P, heat shock protein 70, putative alkaline phosphatase, phototransducing protein, and peroxisomal enoyl hydratase-like protein genes as determined by microarray were then confirmed by RT-PCR. Only the expression of G6P was clearly matched between the microarray (Table 5A) and the RT-PCR results (P < 0.05) (Fig. 4A). HF-induced differences in heat shock protein 70.2 expression, not statistically significant in the microarray, were significant in the RT-PCR (P < 0.05, Table 4A). Putative alkaline phosphatase expression also tended to increase in the RT-PCR assay (statistically borderline, P < 0.1). Although there were no significant difference for phototransducing protein and peroxisomal enoyl hydratase-like protein genes (P values of HF/HG ratios of RT-PCR measurements were < 0.1), these tended to increase in HF-perfused intestines (not shown). RT-PCR attempts to confirm microarray-detected differences in cyclin D1 were unsuccessful.


View this table:
[in this window]
[in a new window]
 
Table 5. Genes whose mRNA abundance as estimated by microarray changes with fructose perfusion for 20 min

 


View larger version (19K):
[in this window]
[in a new window]
 
Fig. 4. The expression as determined by RT-PCR of G6P, heat shock protein (HSP70), putative alkaline phosphatase (PAP), butyrate response factor (BRF1), transforming growth factor (TGFß), and intrinsic factor B12 receptor (IFR) mRNA in the small intestine of neonatal rats perfused with HG and HF solutions for 20 min. A: representative upregulated genes. B: representative downregulated genes. Values are mean relative abundance ± SE (n = 5). The expression of each gene (represented as the ratio of each gene to GAPDH) in HG-perfused pups was designated as 100%, and then the expression of each gene in HF-perfused pups was normalized to this value. *Significant difference between HG and HF groups. +Borderline significant. Fructose perfusion for 20 min increased G6P and HSP70 mRNA abundance and decreased TGFß mRNA abundance.

 
Nine candidate genes were apparently downregulated (≤50% in at least two of four comparisons) in HF-perfused intestine of 20-d old rats for 20 min (Table 5B). Butyrate response factor 1, cardiovascular heat shock protein, transforming growth factor-ß (TGFß), intrinsic factor B12 receptor (IFR), and chromaffin granule amine transporter were selected to be confirmed by RT-PCR. However, only TGFß (P < 0.05) and IFR (borderline, P < 0.1) seemed to match between microarray and RT-PCR (Fig. 4B, Table 4B).

Comparison of Microarray and RT-PCR Results
There were 24 genes whose mRNA abundance was reanalyzed by RT-PCR, and in 80% of these genes, the results from RT-PCR matched those from the microarray experiments (Table 4). The slope of the line between microarray and RT-PCR results was not significantly different from unity, indicating that after appropriate correction for false positives as detected by capture primer and dye-flip experiments, changes in mRNA abundance as detected by microarray will likely be similar to those detected by RT-PCR. Based on a correlation analysis, r2 was 0.843 (P < 0.0001) indicating the general reliability of the results obtained from the microarray. Genes that did not change in expression by microarray were also found to be unchanged when evaluated by RT-PCR (Table 4C).

G6P Expression Increases with HF
Since microarray results were expressed in ratios of RNA abundance in HF- relative to HG-perfused intestine, we determined whether high ratios of G6P expression increased with HF and/or decreased with HG perfusion by comparing mRNA abundance among HG-, HF-, {alpha}-methylglucose-, and normal Ringer-perfused intestines. G6P mRNA abundance was markedly greater in HF-perfused intestine compared with normal Ringer-perfused (P < 0.01) and {alpha}-methylglucose-perfused (P < 0.001) intestines (Fig. 5), paralleling results obtained with GLUT5 (7). G6P mRNA abundance was the same (P > 0.05) among HG-, normal Ringer-, and {alpha}-methylglucose-perfused intestines.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Since dietary fructose stimulates the precocious expression of its transporter GLUT5, here we list genes whose mRNA expression changes in parallel with that of GLUT5. Products ofsome of these genes, particularly those belonging to the gluconeogenic pathway, may act as intermediary signals linking fructose to its transporter.

Genes Upregulated with Luminal Fructose
Genes involved in sugar metabolism.
Ten of 24 upregulated genes at 4 h of fructose perfusion are either directly involved in enzymatic metabolic reactions (Table 2) or regulate those reactions. Of these 10 upregulated genes, 6 metabolize sugars, and 5 of the 6 add or remove phosphate groups. These six genes participate in fructose and mannose metabolism, glycolysis/gluconeogenesis, pentose phosphate pathway, and galactose metabolism. One gene, 6-phosphofructo 1-kinase, catalyzes the formation of fructose-1,6-bisphosphate from fructose-6-phosphate with hydrolysis of ATP and participates directly in all four of these metabolic pathways (Kyoto Encyclopedia of Genes and Genomes, http://www.genome.ad.jp/kegg/). Another gene, fructose-1,6-bisphosphatase catalyzes the opposite reaction (hydrolysis of fructose-1,6-bisphosphate to produce fructose-6-phosphate and inorganic phosphate) and is thought to participate in fructose and mannose, gluconeogenic, and pentose phosphate metabolic pathways.

The one gene that is upregulated and is unique to the fructose and mannose metabolic pathway is fructose-2,6-bisphosphatase (Table 2), which hydrolyzes fructose-2,6-bisphosphate to fructose-6-phosphate and orthophosphate. This same enzyme, when modulated by a cAMP-dependent protein kinase, becomes 6-phosphofructo 2-kinase, which catalyzes the opposite reaction (synthesis of fructose-2,6-bisphosphate from fructose-6-phosphate and ATP). It is interesting to note that dibutryl-cAMP, a permeable nonmetabolizable analog of cAMP, can significantly alter intestinal fructose transport without changing GLUT5 mRNA abundance (Cui et al., unpublished observations). GLUT5 is a facilitative transporter, and whenever free fructose concentration is decreased in the cytosol, it will result in increased downhill gradient for fructose transport into the cell. Hence, conditions that favor intracellular fructose metabolism may alter fructose uptake rate.

Fructose-2,6-bisphosphate is an allosteric inhibitor of fructose-1,6-bisphosphatase (Fig. 6). Since fructose-2,6-bisphosphatase is overexpressed during fructose perfusion, removal of fructose-2,6-bisphosphate may in turn remove the inhibition of fructose-1,6-bisphosphatase, thereby enhancing gluconeogenesis. Significant gluconeogenesis is now known to occur in the small intestinal epithelial cells, which can supply up to a quarter of total glucose production during insulinopenia (6).



View larger version (24K):
[in this window]
[in a new window]
 
Fig. 6. Model showing potential effects of fructose perfusion on the gluconeogenic pathway of neonatal rat intestinal cells. Expression of G6P, GLUT5, and fructose-2,6-bisphosphatase/6-phosphofructokinase-2 (F2,6BP/6PFK2) increases solely with fructose perfusion. Expression of fructokinase (FK), glucose-6-phosphate isomerase (G6PI), fructose-1,6-bisphosphatase (F1,6BP) and 6-phosphofructo 1-kinase (6PFK1) may increase with fructose or decrease with glucose perfusion. Note that G6P is a membrane-bound, multimeric enzyme associated with the endoplasmic reticulum, and hydrolysis of glucose-6-phosphate occurs in the reticular lumen (34). Glucose-6-phosphate is transported from the cytosol by the translocase component of the G6P into the lumen of the endoplasmic reticulum. This translocase is thought to be the same as the putative glycogen storage disease type 1b protein (34), which is upregulated with fructose perfusion (Table 2); – = inhibits; + = stimulates. Fructose-2,6-bisphosphate inhibits F1,6BP and stimulates 6PFK1. In turn, cAMP inhibits 6PFK2, which synthesizes fructose-2,6-bisphosphate from fructose-6-phosphate.

 
The common metabolite linking 6-phosphofructo 1-kinase, fructose-1,6-bisphosphatase, and 6-phosphofructo 2-kinase (or fructose-2,6-bisphosphatase) is fructose-6-phosphate. This fructose metabolic intermediate is also the link between fructose metabolism and glycolysis or gluconeogenesis (Fig. 6). Hence, this cluster of genes and metabolites may represent reactions that control fructose metabolism in the enterocyte and may therefore also regulate fructose transport. In contrast, these do not regulate intestinal glucose transport, because glucose transport in weaning rats is not regulated by diet or luminal substrates (13).

The two other genes in the glycolytic/gluconeogenic pathway clearly increasing with fructose perfusion are G6P catalyzing the hydrolysis of glucose-6-phosphate to glucose and orthophosphate as well as phosphoglycerate mutase, an isomerase which converts 2-phospho-D-glycerate, a precursor of phosphoenolpyruvate, to 3-phospho-D-glycerate (Table 2). Changes in phosphoenolpyruvate concentrations may affect cellular levels of pyruvate, which is a major metabolite whose concentration is pivotal in regulating many metabolic pathways. Expression of pyruvate dehydrogenase kinase 2, which inactivates pyruvate dehydrogenase (enzyme complex converting pyruvate to acetyl-CoA) also increases with fructose perfusion. This means that fructose perfusion may inhibit the conversion of pyruvate to acetyl-CoA, an inhibition consistent with the increase in expression of gluconeogenic enzymes. Acetyl-CoA is an important intermediate in lipid metabolism and in oxidative phosphorylation.

Of all the metabolic enzymes discussed above, only the expression of G6P increases after just 20 min of HF perfusion (Table 5A), well before GLUT5 mRNA abundance increases. G6P mRNA abundance increases from about twofold greater than HG perfusion at 20 min to fourfold greater than HG perfusion after 4 h. This indicates that luminal fructose may trigger transcription of G6P even more rapidly than it triggers GLUT5 transcription. It is also intriguing to note that the mRNA abundance of G6P seems highly sensitive to luminal fructose and not to luminal glucose (Fig. 5). This pattern is much like that of GLUT5 but not that of SGLT1 (28) and GLUT2 (7). The equilibrium concentrations of fructose-6-phosphate:glucose-6-phosphate lie sevenfold toward glucose-6-phosphate, and may suggest why glucose loading has little effect on fructose-responsive enzymes.

G6P is part of a system consisting of four subunits found in the membrane of the endoplasmic reticulum (34). Not only does the expression of G6P (the hydrolase or catalytic subunit) increase with fructose perfusion, but also increased is the expression of a second subunit, glycogen storage disease type 1b, also known as the G6P translocase (Table 2). These parallel increases highlight the importance of the entire G6P system to intestinal fructose metabolism, and suggest a linkage to GLUT5 regulation.

Membrane proteins.
Expression of GLUT5 and GLUT4 (Table 2), the facilitative glucose transporter in skeletal and smooth muscle and in adipose tissue, also increases. Either GLUT4 mRNA is also found in intestinal epithelial cells, or GLUT4 expression increased in smooth muscle tissue found in the submucosal and muscularis layers of the small intestine. The role of other membrane transporters or proteins that regulate those transporters is not clear.

Development, transcription factor, and other genes.
Whereas luminal fructose can stimulate GLUT5 expression in weaning (≥15 days) rats, it has no effect on GLUT5 expression in younger rats (19). Hence, the ability of fructose to stimulate GLUT5 transcription is modulated by development. There are two fructose-upregulated genes that are also strongly involved in regulating development: glypican 3 and activin receptor-like kinase (Table 2). Glypican 3 is involved in controlling the growth of embryonic mesodermal tissues, and in regulating the shape of intestinal epithelial cells during morphogenesis (5). In contrast, activin receptor kinase is not well characterized but belongs to a novel subclass of cell-surface receptors with predicted serine/threonine kinase activity. These receptors regulate activins, which belong to the TGFß family, and are involved in the patterning of the mesoderm like glypican 3 (2).

Activating transcription factor 4 and zinc finger protein are factors likely involved in upregulating transcription of genes turned on by fructose perfusion (Table 2). In hepatocytes of aged rats, GADD153 (also known as C/EBP homologous protein, or CHOP) expression is induced by epidermal growth factor (EGF) (18). Perfusion of EGF in neonatal rat intestine, specifically enhances fructose transport (Cerquiera, Cui, and Ferraris, unpublished observations). The role of other remaining genes is not known.

Genes Downregulated with Luminal Fructose
A relatively large number of genes is significantly downregulated with fructose perfusion, but it is not clear how these are specifically related to fructose induction of GLUT5. Fructose perfusion inhibits the expression of cytochrome-c oxidase, an electron carrier in the respiratory chain (Table 3). Inhibition of a key component of oxidative phosphorylation is consistent with increased expression of gluconeogenic enzymes discussed previously. Since acetyl-CoA is not catabolized when oxidative phosphorylation is inhibited, its concentration increases. Increases in acetyl-CoA stimulate the initial steps of gluconeogenesis. The gluconeogenic pathway should dramatically increase intracellular inorganic phosphate concentrations, which in turn, should downregulate transporters that take in inorganic phosphates from the intestinal lumen or blood. Indeed, fructose perfusion consistently and markedly downregulates the expression of NaPi2b (Table 3), the sodium-dependent phosphate cotransporter that absorbs inorganic phosphates from the intestinal lumen into the cytosol. This suggests that NaPi2b, known to be regulated by dietary phosphate levels (30), may also be modulated by intracellular phosphate concentrations.

The role of the sodium-myo-inositol transporter SMIT (Table 3) is not clear, because it has not previously been reported in the intestine. SMIT is a member of SLC5, the gene family of sodium-dependent glucose cotransporters (36), and SGLT1 is known to increase with luminal glucose in adult rodent intestines (14). GLUT2 is the intestinal basolateral glucose and fructose transporter (13), and its expression may increase with glucose perfusion. GLUT2 expression is regulated by both luminal fructose and glucose (7), but glucose is often reported as a more important regulator (32).

The endothelin receptor (Table 3) is involved in development, and a mutation in this gene results in intestinal hypertrophy (17). Likewise, the catenins are abundant cellular proteins, and their spatial distribution is critical in establishing tissue patterns during embryonic development.

TGFß (Table 5B) typically inhibits cell proliferation and is found in high levels in both rat milk and neonatal rat small intestine, and endogenous production in the intestine increases during the weaning period (40). Inhibiting the expression of this factor may stimulate proliferation.

Future Studies
Changes in gene expression may arise not only from various cell types in the submucosal, muscle, and serosal layers but also from other cell types in the mucosal layer itself; hence, it should be established that these changes occur in the enterocytes where GLUT5 is found. However, changes in gene expression are likely to be events occurring mainly in the enterocytes because only the mucosal epithelium was exposed to and can absorb luminal fructose and glucose. The mucosal epithelium consists mainly of absorptive enterocytes (1), and we showed by in situ hybridization (21) as well as immunocytochemistry (21) that mRNA and protein expression of GLUT5 and the transcription factors c-fos and c-jun increase only in enterocytes of perfused intestines. Ongoing studies also suggest that phosphatidylinositol-3,4,5-triphosphate, a second messenger in the PI3-kinase system, concentrations also increase only in enterocytes whose GLUT5 mRNA abundance increases with HF perfusion.

The biggest challenge is to distinguish which of the genes specifically modulate the effect of luminal fructose on its transporter GLUT5 and which of the genes are responding secondarily to fructose perfusion. Fortunately, a cluster of genes involved in glycolysis and gluconeogenesis stand out as candidate genes regulating GLUT5 expression, and studies using gluconeogenic inhibitors are warranted, to assess the role of some gluconeogenic enzymes such as G6P on GLUT5 regulation.


    GRANTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
This work was supported by National Science Foundation Grants IBN-998808 and 0235011, as well as by USDA Grant 2004-35206-14154.


    ACKNOWLEDGMENTS
 
We thank Carla Cerquiera and Tong-sheng Wang for excellent technical help, as well as Dr. Bin Tian for valuable suggestions.

Present address of P. Tolias: Ortho-clinical Diagnostics-a Johnson and Johnson Co., 1001 US Hwy 202, Raritan, NJ 08869.


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: R. P. Ferraris, Dept. of Pharmacology and Physiology, UMDNJ-New Jersey Medical School, Newark, NJ 07103-2714 (E-mail: ferraris{at}umdnj.edu).

10.1152/physiolgenomics.00056.2004


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Altmann GG and Enesco M. Cell number as a measure of distribution and renewal of epithelial cells in the small intestine of growing and adult rats. Am J Anat 121: 319–336, 1967.[ISI][Medline]
  2. Bourillot PY, Garrett N, and Gurdon JB. A changing morphogen gradient is interpreted by continuous transduction flow. Development 129: 2167–2180, 2002.[Abstract/Free Full Text]
  3. Buddington RK and Diamond JM. Ontogenetic development of intestinal nutrient transporters. Annu Rev Physiol 51: 601–619, 1989.[CrossRef][ISI][Medline]
  4. Castello A, Guma A, Sevilla L, Furriols M, Testar X, Palacin M, and Zorzano A. Regulation of GLUT5 gene expression in rat intestinal mucosa: regional distribution, circadian rhythm, perinatal development and effect of diabetes. Biochem J 309: 271–277, 1995.[ISI][Medline]
  5. Choo B, Li M, Carey DJ, Cano-Gauci DF, and Buick RN. Expression of OCI-5/glypican 3 during intestinal morphogenesis: regulation by cell shape in intestinal epithelial cells. Br J Cancer 77: 890–896, 1998.[ISI][Medline]
  6. Croset M, Zitoun C, Montano S, Mithieux G, and Rajas F. The glucose-6 phosphatase gene is expressed in human and rat small intestine: regulation of expression in fasted and diabetic rats. Diabetes 49: 1165–1168, 2000.[Abstract]
  7. Cui X, Jiang L, and Ferraris RP. Regulation of rat intestinal GLUT2 mRNA abundance by luminal and systemic factors. Biochim Biophys Acta 1612: 178–185, 2003.[ISI][Medline]
  8. David ES, Cingari DS, and Ferraris RP. Dietary induction of intestinal fructose absorption in weaning rats. Pediatr Res 37: 777–782, 1995.[Abstract]
  9. Davidson NO, Hausman AM, Ifkovits CA, Buse JB, Gould GW, Burant CF, and Bell GI. Human intestinal glucose transporter expression and localization of GLUT5. Am J Physiol Cell Physiol 262: C795–C800, 1992.[Abstract/Free Full Text]
  10. Duro D, Rising R, Cedillo M, and Lifshitz F. Association between infantile colic and carbohydrate malabsorption from fruit juices in infancy. Pediatrics 109: 797–805, 2002.[Abstract/Free Full Text]
  11. Dyer J, Vayro S, King TP, and Shirazi-Beechey SP. Glucose sensing in the intestinal epithelium. Eur J Biochem 270: 3377–3388, 2003.[Abstract/Free Full Text]
  12. Endo Y, Fu Z, Abe K, Arai S, and Kato H. Dietary protein quantity and quality affect rat hepatic gene expression. J Nutr 132: 3632–3637, 2002.[Abstract/Free Full Text]
  13. Ferraris RP. Dietary and developmental regulation of intestinal sugar transport. Biochem J 360: 265–276, 2001.[CrossRef][ISI][Medline]
  14. Ferraris RP and Diamond J. Regulation of intestinal sugar transport. Physiol Rev 77: 257–302, 1997.[Abstract/Free Full Text]
  15. Ferraris RP, Yasharpour S, Lloyd KC, Mirzayan R, and Diamond JM. Luminal glucose concentrations in the gut under normal conditions. Am J Physiol Gastrointest Liver Physiol 259: G822–G837, 1990.[Abstract/Free Full Text]
  16. Fuster G, Vicente R, Coma M, Grande M, Felipe A, and Pinto Garcia V. One-step reverse transcription polymerase chain reaction for semiquantitative analysis of mRNA expression. Methods Find Exp Clin Pharmacol 24: 253–259, 2002.[ISI][Medline]
  17. Gilbert SF. Developmental Biology. Sunderland, MA: Sinauer Associates, 2003.
  18. Ikeyama S, Wang XT, Li J, Podlutsky A, Martindale JL, Kokkonen G, van Huizen R, Gorospe M, and Holbrook NJ. Expression of the pro-apoptotic gene gadd153/chop is elevated in liver with aging and sensitizes cells to oxidant injury. J Biol Chem 278: 16726–16731, 2003.[Abstract/Free Full Text]
  19. Jiang L, David ES, Espina N, and Ferraris RP. GLUT-5 expression in neonatal rats: crypt-villus location and age- dependent regulation. Am J Physiol Gastrointest Liver Physiol 281: G666–G674, 2001.[Abstract/Free Full Text]
  20. Jiang L and Ferraris RP. Developmental reprogramming of rat GLUT-5 requires de novo mRNA and protein synthesis. Am J Physiol Gastrointest Liver Physiol 280: G113–G120, 2001.[Abstract/Free Full Text]
  21. Jiang L, Lawsky H, Coloso RM, Dudley MA, and Ferraris RP. Intestinal perfusion induces rapid activation of immediate-early genes in weaning rats. Am J Physiol Regul Integr Comp Physiol 281: R1274–R1282, 2001.[Abstract/Free Full Text]
  22. Khan JM, Wingertzahn MA, Teichberg S, Vancurova I, Harper RG, and Wapnir RA. Development of the intestinal SGLT1 transporter in rats. Mol Genet Metab 69: 233–239, 2000.[CrossRef][ISI][Medline]
  23. Lee HM, He Q, Englander EW, Uchida T, and Greeley GH Jr. Age-associated changes in gene expression patterns in the duodenum and colon of rats. Exp Biol Med (Maywood) 226: 692–700, 2001.[Abstract/Free Full Text]
  24. Miller RA, Galecki A, and Shmookler-Reis RJ. Interpretation, design, and analysis of gene array expression experiments. J Gerontol A Biol Sci Med Sci 56: B52–57., 2001.
  25. Monteiro IM and Ferraris RP. Precocious enhancement of intestinal fructose uptake by diet in adrenalectomized rat pups. Pediatr Res 41: 353–358, 1997.[Abstract]
  26. Nakayama H, Yokoi H, Fujita J, and Watanabe H. Quantification of mRNA by non-radioactive RT-PCR and CCD imaging system. Ovulation defect and its restoration by bone marrow transplantation in osteopetrotic mutant mice of Mitf(mi)/Mitf(mi) genotype. Nucleic Acids Res 20: 4939, 1992.[ISI][Medline]
  27. Nobigrot T, Chasalow FI, and Lifshitz F. Carbohydrate absorption from one serving of fruit juice in young children: age and carbohydrate composition effects. J Am Coll Nutr 16: 152–158., 1997.
  28. Shu R, David ES, and Ferraris RP. Dietary fructose enhances intestinal fructose transport and GLUT5 expression in weaning rats. Am J Physiol Gastrointest Liver Physiol 272: G446–G453, 1997.[Abstract/Free Full Text]
  29. Shu R, David ES, and Ferraris RP. Luminal fructose modulates fructose transport and GLUT-5 expression in small intestine of weaning rats. Am J Physiol Gastrointest Liver Physiol 274: G232–G239, 1998.[Abstract/Free Full Text]
  30. Sugiura SH, McDaniel NK, and Ferraris RP. In vivo fractional Pi absorption and NaPi-II mRNA expression in rainbow trout are upregulated by dietary P restriction. Am J Physiol Regul Integr Comp Physiol 285: R770–R781, 2003. First published June 19, 2003; 10.1152/ajpregu.00127.2003.[Abstract/Free Full Text]
  31. Toloza EM and Diamond J. Ontogenetic development of nutrient transporters in rat intestine. Am J Physiol Gastrointest Liver Physiol 263: G593–G604, 1992.[Abstract/Free Full Text]
  32. Tsang R, Ao Z, and Cheeseman C. Influence of vascular and luminal hexoses on rat intestinal basolateral glucose transport. Can J Physiol Pharmacol 72: 317–326, 1994.[ISI][Medline]
  33. Tseng GC, Oh MK, Rohlin L, Liao JC, and Wong WH. Issues in cDNA microarray analysis: quality filtering, channel normalization, models of variations and assessment of gene effects. Nucleic Acids Res 29: 2549–2557, 2001.[Abstract/Free Full Text]
  34. van Schaftingen E and Gerin I. The glucose-6-phosphatase system. Biochem J 362: 513–532, 2002.[CrossRef][ISI][Medline]
  35. Wada K, Honma H, Tanaka A, Kaneko T, Sakakibara S, Ohsumi J, Serizawa N, Fujiwara T, Horikoshi H, Fujita T, and Horikoshi T. Quantification of relative mRNA expression in the rat brain using simple RT-PCR and ethidium bromide staining. J Med Chem 43: 3052–3066, 2000.[CrossRef][ISI][Medline]
  36. Wright EM and Turk E. The sodium/glucose cotransport family SLC5. Pflügers Arch 447: 510–518, 2004.[CrossRef][ISI][Medline]
  37. Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J, and Speed TP. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res 30: e15, 2002.[Abstract/Free Full Text]
  38. Yu J, Othman MI, Farjo R, Zareparsi S, MacNee SP, Yoshida S, and Swaroop A. Evaluation and optimization of procedures for target labeling and hybridization of cDNA microarrays. Mol Vis 8: 130–137, 2002.[ISI][Medline]
  39. Zinke I, Schutz CS, Katzenberger JD, Bauer M, and Pankratz MJ. Nutrient control of gene expression in Drosophila: microarray analysis of starvation and sugar-dependent response. EMBO J 21: 6162–6173, 2002.[Abstract/Free Full Text]
  40. Zola H, Read L, Penttila I, and Zhang MF. Localization of transforming growth factor-beta receptor types I, II, and III in the postnatal rat small intestine. Immunol Cell Biol 79: 291–297, 2001.[CrossRef][ISI][Medline]