Hypoxia differentially regulates nutrient transport in rat jejunum regardless of luminal nutrient present K. A. Kles1 and K. A. Tappenden1,2

1 Division of Nutritional Sciences and 2 Department of Food Science and Human Nutrition, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801

    ABSTRACTAggressive enteral nutrition and poor intestinal perfusion are hypothesized to play an important pathogenic role in nonocclusive

small bowel necrosis. This study tests the hypothesis that glucose

and glutamine transport are differentially regulated during hypoxia

regardless of the luminal nutrient present. Sprague-Dawley rats

(247 ± 3 g; n = 16) were randomized to receive 1 h of intestinal hypoxia or serve as normoxic controls. During this hour, jejunal loops were randomized to receive in situ perfusions of mannitol, glucose, or glutamine. When compared with normoxic groups, glucose but not glutamine transport was impaired (P < 0.001) during hypoxia. Messenger RNA abundance of the sodium glucose cotransporter sodium-dependent glucose cotransporter-1 (SGLT-1) and neutral basic amino acid transporter Bo did not differ with hypoxia or nutrient perfused. Jejunal brush-border SGLT-1 abundance was decreased (P = 0.039) with hypoxia; however, total cellular SGLT-1 protein abundance did not differ among treatment

groups. These data indicate that SGLT-1 activity is regulated during hypoxia at the posttranslational level. Additional information

regarding the mechanisms regulating nutrient transport in the

hypoperfused intestine is critical for optimizing the composition of enteral nutrient formulas.

sodium-dependent glucose cotransporter-1; nutrient absorption; small intestine

TOP ABSTRACT INTRODUC

TION METHODS RESULTS DISCUSSIO

N REFERENC

ES

    INTRODUCTIONTHE INCIDENCE OF NONOCCLUSIVE small bowel necrosis is significant in traumatically injured patients, despite adequate systemic

resuscitation. Currently, the causative mechanism remains unknown; however, enteral nutrition has been provided in ~90% of cases (21), suggesting that the inappropriate administration of specific nutrients into a poorly perfused small bowel may play a pathogenic role (12, 21, 23, 25). During states of low blood flow to the intestine, enteral nutrients may increase oxygen demand beyond that available, potentially increasing intestinal hypoxia and impairing intestinal function. Gastrointestinal function has been shown to be impaired during hypoperfusion or decreased oxygenation of the intestine (28). Previous in vivo rat studies from our lab indicate that, during hypoperfusion, impaired barrier function, increased lactate concentration, decreased ATP concentration, and altered nutrient transport are characteristic (15). The detrimental effects of hypoperfusion may be minimized with an

enteral formula composed of nutrients easily processed by the

hypoperfused intestine.

The purpose of the current study was to determine how nutrient perfusion alters sodium-dependent glucose and glutamine transport

activity and their regulation at the cellular level during hypoperfusion. Glutamine transport occurs via system Bo, which is part of a family of transporters for neutral amino acids (14). Glucose transport occurs via the sodium-glucose cotransporter [sodium-dependent glucose cotransporter-1 (SGLT-1)]. We previously reported that, during hypoxia, brush-border glucose transport is impaired, but glutamine transport remains unaltered following luminal perfusion of hexoses (15). It is possible that glutamine transport was maintained in this previous study, because the jejunum had not been exposed to luminal glutamine. Therefore, the current study was designed to examine the effect of luminal glucose or glutamine on jejunal nutrient transport during hypoxia. We hypothesize that glucose and glutamine transport are differentially regulated at the cellular level during hypoxia regardless of the luminal nutrient present.

TOP ABSTRACT INTRODUC

TION METHODS RESULTS DISCUSSIO

N REFERENC

ES

    METHODSAnimals. Sixteen male Sprague-Dawley rats (Harlan Sprague Dawley, Indianapolis, IN) weighing 247 ± 3 g were acclimatized and housed in a temperature- and humidity-controlled facility with a 12:12-h light-dark cycle in individual cages. Animals were maintained on a standard rodent diet and were given free access to water. The University of Illinois at Urbana-Champaign Laboratory Animal Care Advisory Committee approved all procedures described in accordance with the Guide for the Care and Use of Laboratory Animals.

Experimental model. After a 24-h food restriction, surgical plane was achieved using ketamine (87 mg/kg im) and xylazine (13 mg/kg im) anesthesia, and a laparotomy was performed. Procedures were performed at the same time each day to account for diurnal variations of SGLT-1 activity (24). Two centimeters distal to the ligament of Treitz, three 8-cm in situ jejunal loops were cannulated with flexible Silastic tubing (3.2 mm OD × 2.0 mm ID; Fisher Scientific, Itasca, IL) and secured with 4-0 silk suture (Ethicon, Somerville, NJ). After the preparation of jejunal loops, rats were randomized to one of two groups for 1 h: 1) hypoxia (created by clamping the superior mesenteric artery with a microbulldog clip (Roboz Surgical, Rockville, MD) or 2) normoxia control (with unrestricted small bowel perfusion). Sterile instruments and aseptic technique were used at all times.

Within each animal, in situ jejunal loops were randomized to receive luminal perfusions with one of three nutrients: 1) mannitol (an osmotic control), 2) glucose (absorbed by active transport via SGLT-1), or 3) glutamine (actively transported via Bo). All of the nutrients were perfused for 60 min at a concentration of 120 mM in a modified Krebs solution [(in mM) 141 Na, 117.6 Cl, 26 HCO3, 1.2 Mg, 1.2 Ca, 5.2 K, 2.4 HPO4, and 0.4 H2PO4, pH 7.4] maintained at 38°C throughout the experiment. Fresh nutrient infusion solutions were continuously perfused during the 60 min. Body temperature was maintained at 38°C with a heating element. At the end of the experimental protocol, anesthetized animals were euthanized by cardiac puncture.

TOP ABSTRACT INTRODUC

TION METHODS RESULTS DISCUSSIO

N REFERENC

ES

Tissue preparation. After 60 min of luminal nutrient perfusions, the perfused segments were rapidly excised and the mesentery was removed. Two 2-cm sections were removed for electrophysiological analysis in modified Ussing chambers. A 1-cm section was snap-frozen and stored at 80°C for RNA isolation and determination of relative mRNA abundance of SGLT-1 and Bo. A 1-cm section was snap frozen and stored at 20°C for determination of ATP, lactate, pyruvate, and protein concentration. A 1.5-cm section was frozen for quantification of total cellular and brush-border SGLT-1 protein

abundance.

Ion and nutrient transport. Techniques assessing gastrointestinal function through measurement of ion flux in modified Ussing chambers have been previously described (3, 7, 14, 19). Jejunal sections were cut longitudinally along the mesentery and mounted in modified Ussing chambers (Physiologic Instruments, San Diego, CA) to expose 0.5 cm2 of the brush border and serosal sides. The tissue was bathed in 8 ml of oxygenated (95% O2-5% CO2) modified Krebs buffer solution maintained at 37°C with a circulating water bath (Fisher Scientific). Basal transmural short-circuit current, resistance, and potential difference were measured using established techniques after a 20- to 30-min equilibration period. Sodium-dependent nutrient transport was determined by measuring the changes in short-circuit current induced by addition of either 10 mM glucose or glutamine to the medium in the mucosal reservoir. The modified Ussing chambers were connected to dual-channel voltage/current clamps (VCC MC2, Physiologic Instruments) with a computer interface allowing for real-time data acquisition and analysis (Acquire & Analyze software, Physiologic Instruments).

Jejunal ATP/ADP ratio. To ensure that jejunal mucosa exhibited metabolic indicators of oxygen deprivation, jejunal ATP/ADP ratio was measured. Frozen tissue sections were homogenized (model PCU-11, Brinkman Instruments, Westbury, NY) and sonicated for 30 s at setting 3 (model 450, Branson Sonifier, Danbury, CT) in a modified Krebs buffer for determination of jejunal ATP levels using the luciferin/luciferase method (30). One hundred microliters (20 mg/ml) of luciferin/luciferase (Sigma Chemical, St. Louis, MO) were added to 100 µl of sample and counted for 30 s on a scintillation counter (Beckman LS 6500, Fullerton, CA) using chemiluminescence parameters. Sample ATP concentrations were then calculated based on an ATP standard curve (Sigma Chemical). Additionally, ADP concentration was measured on a fluorometer (F-2000; Hitachi Instruments, Chicago, IL) by reduction of NADH and

read at excitation of 340 nm and emission of 460 nm. Samples and standards were placed in a buffer of NADH (20 µM) and phosphoenolpyruvate (40 µM), and ADP reduction occurred following addition of pyruvate kinase (0.30 U/ml) (20). The sample ADP concentration was calculated based on an ADP standard curve.

Jejunal lactate concentration. As a marker of anaerobic metabolism, jejunal lactate concentration was measured in the prepared tissue homogenate using the Sigma Diagnostics Lactate kit (procedure no. 735) at 540 nm using an Elx800 plate reader (Bio-Tek, Winooski, VT). Sample lactate concentrations were determined using a lactate standard curve (Sigma Chemical). To normalize for differences in glycolytic substrate available between treatment groups, data were expressed as (micromoles lactate/micromoles pyruvate)/milligram protein. From the prepared homogenate, jejunal pyruvate concentration was quantified on a fluorometer (F-2000) based on a pyruvate standard curve. Ten microliters of each sample were placed in 1 ml NADH buffer (0.95 µM NADH; 0.5 M NaH2PO4; 0.5 M K2HPO4) read at excitation of 340 nm and emission of 460 nm. The sample was then reduced following addition of 5 U of lactate dehydrogenase (LDH) and read at excitation of 340 nm and emission of 460 nm. The differences in sample emission readings (initial reading minus reading following LDH addition) were determined and compared with a pyruvate standard curve (Sigma Chemical) (20, 30).

Relative RT-PCR mRNA abundance of nutrient transporters. Total cellular RNA was isolated from the snap-frozen jejunal samples using the guanidium isothiocyanate phenol-chloroform method of Chomczynski (8). Total RNA was quantified by using a spectrophotometer (U-2000, Hitachi) at OD260. RT-PCR was performed in a Gene Mate Genius thermocycler (ISC Biosexpress, Kaysville, UT) in a 20-µl total volume containing 3 µg RNA, 2.5 µM random decamers, 0.5 mM deoxynucleotide triphosphate (dNTP) mix of the four dNTPs, 1× first strand buffer, 10 mM DTT, and 200 U Superscript II RT (GIBCO-BRL, Rockville, MD; Ambion, Austin, Texas). Samples were then stored at 20°C.

The oligonucleotide primers used for the detection of cDNA specific to rat SGLT-1 and Bo mRNA were synthesized by GIBCO-BRL. Primer sequences for Bo determined using Primer3 (http://www-genome.wi.mit.edu/cgi-bin/primer/ primer3_www.cgi) and GeneBlast (http://www.ncbi.nlm.nih. gov/BLAST/) were: (forward) 5'-ATG GCT CTG GGA GAC AGA GA-3' and (reverse) 5'-GGA GAA ATG GAC TGG

GTG TG-3'. SGLT-1 mRNA primers were based on a report by Scholtka et al. (26). PCR mixtures for amplification of cDNA were performed in a 50-µl total volume containing 0.2 mM dNTPs, 1× PCR buffer without MgCl2, 1.5 mM MgCl2, template cDNA, forward and reverse primers, 2.5 U Taq polymerase (GIBCO-BRL), and primers and competimers for 18S (4:6 ratio; Ambion). The SGLT-1 reaction mixture underwent 25 cycles of denaturing at 94°C for 45 s, annealing at 48°C for 30 s, and extension at 72°C for 50 s, followed by 72°C for 10 min. The Bo reaction mixture underwent 33 cycles of denaturing at 94°C for 45 s, annealing at 55°C for 30 s, and extension at 72°C for 50 s, followed by 72°C for 10 min. Tris-borate ethylenediaminetetraacetic acid (TBE) agarose gels (1.2%) were stained with ethidium bromide and photographed using the FOTO/Analyst image-analysis system (Fotodyne, Hartland, WI). Densitometry of the 18S ribosomal product and the nutrient transporter of interest was performed using Collage image-analysis software 4.0 (Hartland, WI). Preliminary trials determined the appropriate level of 18S competimers/primer ratio and temperature cycles necessary to ensure both the gene of interest and 18S bands produced were well within the linear range of ethidium bromide

detection.

SGLT-1 protein abundance. SGLT-1 protein abundance was determined in tissue homogenate in modified Krebs solution and brush border isolated by the procedure described by Tappenden and McBurney (29) and confirmed by ensuring sucrase-isomaltase enrichment with the brush-border fraction. Total and brush-border protein concentrations were determined using the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA) with a BSA standard (1). Next, the protein was denatured by boiling for 4 min, and proteins were separated by size using 12.5% SDS-PAGE and were transferred to polyvinylidene difluoride (PVDF) membranes (Bio-Rad) using a semidry transfer apparatus (Bio-Rad). Before analyzing experimental samples, a linear range from 0 to 30 µg was established as the appropriate amount of protein per well to ensure that the bands produced fell within the linear range of the colorimetric detection system. Western blot analysis for SGLT-1 protein was performed using polyclonal antibody with known rat reactivity (Chemicon, Temecula, CA). In addition, actin monoclonal antibody with known rat reactivity (Chemicon) was added to the membrane to allow for SGLT-1 normalization to a constitutively expressed protein. PVDF membranes were developed using the Opti-4CN kit (Bio-Rad). According to the method established by Tappenden and McBurney (29), PVDF membranes were blocked using 5% nonfat dry milk in

Tris-buffered saline (20 mM Tris, 137 mM NaCl, pH 7.6, 0.1% Tween) for 3 h at room temperature on a metabolic shaker. The membrane was washed for 5 min in phosphate-buffered saline with Tween 20 (PBST; 4.3 mM Na2PO4, 1.4 mM KH2PO4, 2.7 mM KCl, 137 mM NaCl, 0.1% Tween 20, pH 7.3) at room temperature. The SGLT-1

primary and actin primary antibodies were diluted 1:5,000 and

1:2,000, respectively, in PBST and 0.01% BSA. The primary antibodies were coincubated for 4.5 h at room temperature and washed at room temperature (3 × 10 min). Secondary antibodies with horseradish peroxidase conjugated to goat anti-rabbit IgG (for the SGLT-1 primary antibody) and goat anti-mouse IgG (for the actin primary antibody) were added at room temperature for 30 min (1:10,000 dilution in PBST and 0.01% BSA; Bio-Rad) and washed with PBST (3 × 10 min). Colorimetric detection followed addition of Opti-4CN diluent solution. Photographs of gels were taken using the FOTO/Analyst image-analysis system. Densitometry of SGLT-1 protein abundance was performed using Collage image-analysis software 4.0.

Statistical analysis. The effects of hypoxia and nutrient perfusions on the outcome parameters were determined using a two-way ANOVA. The sources of variation were hypoxia (h = 2), nutrient perfused (p = 3), and hypoxia interacted with nutrient perfused. For analysis of relative transporter mRNA and protein abundance, the effect of gel was calculated using a randomized block ANOVA and was significant due to expected experimental variation. Therefore, a completely randomized block ANOVA was performed for these analyses (block = gel; main effects = hypoxia and nutrient perfused; interaction = hypoxia × nutrient perfused). When a significant effect existed, comparisons were completed using Tukey's post hoc analysis. Computations were performed using SAS (Version 8.1, SAS Institute, Cary, NC). Statistical significance was defined as P   0.05.

    RESULTSIndicators of jejunal hypoxia. A pilot trial established the presence of hypoxia by determining a reduction of the partial pressure of oxygen in the mesenteric

venous effluent using the protocol described herein [normoxia (n = 3; PO2 = 75.6 ± 1.5 mmHg; %O2 saturation = 96.0 ± 1.7); hypoxia rats (n = 3;

TOP ABSTRACT INTRODUC

TION METHODS RESULTS DISCUSSIO

N REFERENC

ES

PO2 = 44.0 ± 10.9 mmHg; %O2 saturation = 79.6 ± 7.6)]. In addition, indirect physiological responses to hypoxia were evaluated (basal transmural resistance, lactate concentration, and ATP/ADP). Consistent with previous reports (19), basal transmural resistance was significantly lower (P = 0.005) in the hypoxia groups, indicating increased permeability (Fig. 1). Additionally, glucose perfusion resulted in significantly higher resistance than glutamine perfusion (P = 0.002), regardless of hypoxia. However, jejunal lactate concentration was higher (P < 0.001) during hypoxia with glucose perfusion (Fig. 2). This outcome was normalized to jejunal pyruvate concentration to account for flux through the glycolytic pathway and therefore indicated that luminal glucose alone increases the level of anaerobic metabolism. The ATP/ADP ratio was measured as a metabolic indicator of energy stores. ATP/ADP ratio was significantly lower (P < 0.001) in the hypoxia groups compared with the normoxia control groups (Fig. 3). Furthermore, glucose perfusion significantly increased ATP/ADP ratio compared with mannitol (P = 0.01) and glutamine (P = 0.055) both in normoxia and hypoxia groups.

View larger version (13K):

[in this window] [in a new window]

 

Fig. 1.   Effect of jejunal nutrient perfusion during hypoxia on basal transmural resistance in rats. As expected, during oxygen deprivation, transmural resistance was significantly lower (P = 0.005) compared with control normoxia. Additionally, glucose perfusion resulted in significantly higher resistance than glutamine or mannitol perfusion (P = 0.002), irrespective of oxygenation status. Data are reported as pooled means ± SE. Means with different letters are significantly different from each other.

View larger version (11K):

[in this window] [in a new window]

 

Fig. 2.   Effect of luminal perfusion of mannitol, glucose, or glutamine on jejunal lactate concentration following hypoxia in rats. Jejunal lactate concentration was significantly higher (P < 0.001) in the hypoxia group perfused with glucose than all other groups. Data are reported as means ± SE. Means with different letters are significantly different from each other.

View larger version (10K):

[in this window] [in a new window]

 

Fig. 3.   Effect of luminal perfusion of mannitol, glucose, or glutamine on jejunal ATP/ADP ratio following hypoxia in rats. ATP concentration was significantly lower (P < 0.001) in the hypoxia groups compared with the normoxia control groups, irrespective of nutrient perfused. In addition, glucose perfusion significantly increased (P = 0.006) ATP/ADP ratio compared with mannitol and glutamine. Data are reported as means ± SE. Means with different letters are significantly different from each other.

Jejunal sodium-dependent nutrient transport. The effect of hypoxia on sodium-dependent glucose and glutamine transport activities was assessed. Compared with normoxia controls, glucose transport was significantly impaired (P < 0.001) in the hypoxia groups (Fig. 4). However, glucose transport activity did not vary among nutrient perfused (Fig. 4). In contrast, glutamine transport activity was not

affected by hypoxia (Fig. 5) and was significantly higher following luminal perfusion of glutamine compared with glucose or mannitol (P = 0.02), regardless of oxygenation state (Fig. 5).

View larger version (11K):

[in this window] [in a new window]

 

Fig. 4.   Effect of luminal perfusion of mannitol, glucose, or glutamine on jejunal sodium-dependent glucose transport (SGLT 1) following hypoxia in rats. SGLT 1 was significantly lower (P < 0.001) in the hypoxia groups than the normoxia controls, irrespective of nutrient pe