2002 Harry M. Vars Research Award

0148-6071/02/2606-0333$03.00/0 JOURNAL OF PARENTERAL AND ENTERAL NUTRITION Copyright © 2002 by the American Society for Parenteral and Enteral Nutrition

Vol. 26, No. 6 Printed in U.S.A.

2002 Harry M. Vars Research Award Keratinocyte Growth Factor Stimulates the Recovery of Epithelial Structure and Function in a Mouse Model of Total Parenteral Nutrition Hua Yang, MD, PhD; Barbara Wildhaber, MD; Yuko Tazuke, MD; and Daniel H. Teitelbaum, MD From the Section of Pediatric Surgery, Department of Surgery, University of Michigan Medical School and C. S. Mott Children’s Hospital, Ann Arbor

ABSTRACT. Background: Keratinocyte growth factor (KGF) increases intestinal growth and is expressed by intestinal intraepithelial lymphocytes (IEL). Because total parenteral nutrition (TPN) leads to villus atrophy and a loss of epithelial function, we hypothesized that KGF administration could reverse these changes. Methods: Mice were randomized into three groups: oral feeding (Control); TPN; or TPN with recombinant human KGF. Mice were killed at 7 days, and the small bowel was harvested for histology, DNA, and protein content analysis. Epithelial cell proliferation was studied by 5-bromo-2-deoxyuridine (BrdU) incorporation, and apoptosis was detected by flow cytometry with Annexin V staining. Epithelial ion transport function was studied by Ussing chambers. Epithelial barrier function was assessed with transepithelial resistance and transmural passage of 3 H-mannitol. Epithelial KGF receptors expression was studied by using reverse transcriptase-polymerase chain reaction

(RT-PCR) and Western blot. Results: TPN decreased intestinal DNA, protein content, villus height, and crypt cell proliferation. TPN also resulted in an increase in epithelial cell apoptosis. KGF administration significantly stimulated the recovery of mucosal structures including intestinal protein and DNA content, villus height, and crypt cell proliferation, and decreased epithelial apoptosis. KGF also up-regulate the epithelial KGF receptor expression. Moreover, KGF attenuated the TPN-induced increase in ion transport and increased the epithelial barrier function. Conclusions: KGF administration reversed many of the adverse epithelial changes associated with TPN administration. Additionally, KGF up-regulated epithelial KGF receptor expression. It is possible that KGF may have a therapeutic efficacy in patients who are receiving TPN. ( Journal of Parenteral and Enteral Nutrition 26:333–341, 2002)

Intestinal mucosal homeostasis depends on a balance between cell proliferation and cell death. A major consequence where cell death exceeds cell proliferation is the development of villus atrophy. Use of total parenteral nutrition (TPN) with the absence of enteral nutrition results in such villus atrophy.1,2 Because the combination of TPN and total bowel rest is widely used in the treatment of a variety of reversible gastrointestinal and surgical disorders, it is important to maintain gut structure and function to near normal capacity to aid recovery in an already nutritionally compromised critically ill patient. Maintenance of gut function may also have a benefit in preventing bacterial endotoxins from entering the host from the intestinal lumen. The regulation of gastrointestinal cell growth and differentiation is complex and influenced by many factors.3 A number of growth factors support the intestinal epithelium including epidermal growth factor,4 insulin-like growth factor 1,5 glucagon-like growth fac-

tor,6 and keratinocyte growth factor (KGF).7 KGF is a known mitogenic growth factor, which is expressed in the mucosal layer by T-cell receptor ␥␦⫹ intestinal intraepithelial lymphocytes (IEL).8 Because KGF receptors (KGFR) have been detected in high numbers in the gastrointestinal tract,9 KGF seems to play a critical role in intestinal epithelial growth and maintenance.10,11 Recent studies have found that KGF administration can have a number of beneficial effects. Administration of exogenous KGF prevents mucositis during the administration of chemotherapy and radiation to the intestine12 and ameliorates mucosal injury in an experimental model of colitis in rats.13 Additionally, KGF given to rodents in a starvation model has been shown to prevent villus atrophy.14 More recently, KGF administration can enhance early gut adaptation in a rat model of short bowel syndrome.15 These studies might suggest a relative deficiency of KGF in many of these disorders. In this current study, we hypothesized that KGF administration can prevent mucosal atrophy and attenuate the abnormality of epithelial function in a TPN mouse model. Additionally, we hypothesized that KGF administration would increase the expression of KGFR in the intestine, as a means of augmenting mucosal proliferation. The purpose of this study was to determine the effect of KGF on the recovery of epithelial structure and function in a mouse model of

Received for publication, February 11, 2002. Accepted for publication, June 12, 2002. Correspondence: Daniel H. Teitelbaum, MD, Section of Pediatric Surgery, University of Michigan Hospitals, Mott F3970, Box 0245, Ann Arbor, MI 48109. Electronic mail may be sent to [email protected]. Presented at the A.S.P.E.N. 26th Clinical Congress, February 23–27, 2002, San Diego, CA.

333

334

YANG ET AL

TPN. The study also to investigated the effect of KGF administration on intestinal mucosal KGFR expression. METHODS

Animals Male, specific pathogen-free, adult C57BL/6J mice (Jackson Laboratories, Bar Harbor, ME) were maintained in temperature-, humidity-, and light-controlled conditions. During administration of IV solutions, the mice were housed in metabolic cages. TPN Model Administration of TPN was performed as previously described.16 Mice were infused with crystalloid solution (dextrose 5% in 0.45 NS with 20 mEq KCl/L) at 4 mL/d. After 24 hours, mice were randomized into 3 groups (n ⫽ 6 per group). The Control group received the same IV solution at 7 mL/24 h, in addition to standard laboratory mouse chow and water ad libitum. The TPN group received a standard TPN solution IV at 7 mL/24 h with no oral intake. The TPN solution (prepared by the hospital pharmacy) contained a balanced mixture of amino acids, lipids, and dextrose, in addition to electrolytes and vitamins.16 Caloric delivery was based on estimates of caloric intake by the Control group and from previous investigators so that caloric delivery was essentially the same in the 3 groups. KGF administration to TPN mice (TPN-KGF group) was given daily by IV injection (5 mg/kg per day), after the first day of TPN, and continued for 6 days. Recombinant human KGF (rHuKGF) was a gift from Amgen Inc (Thousand Oaks, CA). All animals were killed on day 7 using CO2. Biochemical and Histologic Analysis At the time of death, 5 cm of jejunum was excised, the luminal contents were removed, and this segmental of intestine was weighed. The proximal 1-cm section was fixed with 10% formalin and used for histology and 5-bromo-2-deoxyuridine (BrdU) incorporation study, and the remaining 4-cm section was used for the measurement of intestinal DNA and protein content. DNA content was determined using a standard diphenylamine procedure.17 Protein determination was performed by using a Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA). A 0.5 cm section of jejunum was fixed in 10% formaldehyde for histologic sectioning (5 ␮m thickness). Tissues were then dehydrated with ethanol and embedded in paraffin. Sections were cut and stained with hematoxylin-eosin. The villus height was measured using a calibrated micrometer. Each measurement for villus height consists of the mean of 7 different fields. Epithelial Cells Proliferation Assay Mice were injected intraperitoneally with BrdU (50 mg/kg; Roche Diagnostic Corp, Indianapolis, IN) 1 hour before mice were killed. Paraffin-embedded sections of

Vol. 26, No. 6

5 ␮m thickness were deparaffinized with xylene. Immunohistochemistry was done by using a BrdU InSite detection Kit according to the manufacturer’s guidelines (PharMingen, San Diego, CA). Briefly, endogenous peroxidase was quenched with 3% H2O2. Slides were incubated with biotinylated anti-BrdU antibody in a 1:10 dilution and then with StreptavidinHPR. Slides were exposed to diaminobenzidine (DBA) substrate. Finally, slides were counter-stained with hematoxylin. An index of the crypt cell proliferation rate was calculated by the ratio of the number of crypt cells incorporating BrdU to the total number of crypt cells. The total number of proliferating cells per crypt was defined as a mean of proliferating cells in 10 crypts (counted at 45⫻ magnification). Epithelial Cells Apoptosis and KGFR Expression Mucosal cell isolation. Isolation was performed after the protocol described by Mosley and Klein.18 Briefly, after removing the small intestine, it was placed in tissue culture media (RPMI 1640, with 10% fetal calf serum). The intestinal segment was cut into 5-mm pieces, washed in an extraction buffer, and incubated in the same buffer with continuous brisk stirring at 37° for 25 minutes. The supernatant was then filtered rapidly through a glass wool column. After centrifugation, the pellets were purified in 40% isotonic Percoll (Upjohn & Pharmacia, Uppsala, Sweden) and reconstituted in tissue culture media. Viability exceeded 95% using trypan blue exclusion staining. Epithelial cells apoptosis study. Apoptosis was determined with flow cytometry by determining the cell surface expression of phosphatidylserine (Annexin V staining). Briefly, Annexin V assay was performed with an Apoptosis Kit (PharMingen) according to the company protocol. Cells were double stained with fluorescein isothiocyanate (FITC)-conjugated Annexin V and propidium iodide. Cells (10,000 per sample) were analyzed by a FASCalibur (Beckman Dickinson Co, Franklin Lake, NJ). Epithelial cells KGFR mRNA expression (reverse transcriptase-polymerase chain reactions [RT-PCR]). A guanidine isothiocyanate/chloroform RNA extraction method was used with Trizol (Gibco BRL, Gaithersburg, MD) following the manufacturer’s guidelines. RNA was then reversed transcribed into cDNA by adding 50 ␮g/mL of total RNA to the following mixture: nucleotides (adenosin triphosphate [ATP], cytosin triphosphate [CTP], thymidin triphosphate [TTP], and guanidin triphosphate [GTP], each at 1 mmol/L; Boehringer Mannheim, Mannheim, Germany), Moloney Murine Leukemia Virus (MMLV) (reverse transcript) (8 U/␮L; Gibco BRL), Oligo dT (2.5 ␮mol/L; New England Biolabs, Beverly, MA), and RNAase inhibitor (2 U/␮L; Boehringer Mannheim). Diethylene pyrocarbonate (DEPC)-treated H2O was added to yield an appropriate final concentration. Samples were incubated at 39°C for 1 hour, and the reaction was stopped by incubating at 95°C for 5 minutes. RT product (5 ␮L) was added to forward and reverse specific oligomers (5 mmol/L); PCR buffer (with 10 mmol/L Tris and 50 mmol/L KCl), MgCl2 (2.5 mmol/L), and Taq polymerase (0.4

November–December 2002

KGF STIMULATES RECOVERY OF EPITHELIAL STRUCTURE

U/sample; Perkin Elmer, Foster City, CA), with sufficient DEPC-treated H2O to allow for appropriate concentrations. Specific oligomers were designed using an optimization program (OLIGO 4.1; National Biosciences, Plymouth, MN). For KGFR, the forward primer was 5⬘-CGC AAA TGG ATA CTG ACA CG-3⬘ and the reverse primer was 5⬘-GGG CTG GAA CAG TTC ACA CT-3⬘. ␤-Actin primers were previously described. The forward primer was GAG GGA AAT CGT GCG TGA CAT and the reverse primer was AGA AGG AAG GCT GGA AAA GAG.16 The following thermal cycler (PTC-100; MJ Research, Inc, Watertown, MA) settings were used: 94°C for 2 minutes; followed by 35 cycles of 94°C ⫻ 15 seconds; 55°C ⫻ 15 seconds, and 72°C ⫻ 30 seconds; followed by a 5-minute extension time at 72°C. The PCR products were run out on a 2% agarose gel containing ethidium bromide for 1 hour at 170 V and then visualized under ultraviolet light. Quantification of cDNA product used Kodak 1D Image software (Kodak Co, Rochester, NY), and a semiquantification of the amount of target PCR product was done by normalizing each sample to the production of ␤-actin. Immunoblot analysis for epithelial cell KGFR expression. Briefly, isolated epithelial cells were homogenized on ice in lysing buffer.19 Protein determination was performed by using a Bio-Rad protein assay kit (BioRad Laboratories). Approximately 80 ␮g of total protein in loading buffer was loaded per lane and was separated on sodium dadecyl sulfate (SDS)-polyacrylamide-gel electrophoresis (8%). After electrophoresis, proteins were transferred to a polyvinoylidene difluoride (PVDF) membrane (Bio-Rad Laboratories). The membrane was first incubated with a blocking solution (Zymed Laboratories Inc, San Francisco, CA) and then probed with rabbit anti-KGFR antibody Bek (C-17) (purified polyclonal antibody, 1:200 in blocking solution; Santa Cruz Biotechnology, Inc, Santa Cruz, CA) for 1 hour. Bound antibodies were exposed to a peroxidase-conjugated secondary antibody, goat antirabbit IgG 1: 8000 (Zymed Laboratories Inc), detected on X-ray film. Blots were then stripped and reprobed with monoclonal mouse anti ␤-actin antibody (Ac-15) (1:8000 in blocking solution; Sigma, St Louis, MO) to confirm equal loading of protein. The peroxidase-conjugated second antibody is goat antimouse IgG (1:8000 in blocking solution) from Zymed Laboratories Inc.

included 10 mmol/L glucose as an energy source and was osmotically balanced by 10 mmol/L mannitol in the mucosal buffer. Transepithelial potential difference (PD) and short circuit current (Isc), as an indication of the ion transport state of the tissue, were measured after a 30-minute equilibration period. The transmembrane resistance was determined using Ohm’s law. To assess the effect of TPN on intestinal ion transport, an absorptive agent D-glucose (stimulated Na⫹ absorption) and a secretory agonist, carbachol (stimulated Cl⫺ secretion) were used in this study.20 Basal electrical parameter measurements were taken after the equilibration period. The changes in Isc induced by mucosal addition of 10 mmol/L glucose or serosal addition of 10 ␮mol/L carbachol in Ussing chambers were registered on a chart recorder. The change in Isc was measured by subtracting the basal current (before the addition of the agonist) from the peak current after the addition of the agonist. Intestinal permeability experiments. The permeability of the small intestine was also assessed with 3H-mannitol. After a 30-minute equilibration period in Ussing chambers, 3H-mannitol (3 ␮Ci/mL; Sigma) was added to the mucosal compartment. One-milliliter samples were taken every 10 minutes from the serosal compartment for analysis of 3H-mannitol and replaced with 1 mL fresh Krebs buffer. Incubations were carried out for 90 minutes after equilibration. Radioactivity of 3 H-mannitol was measured in a scintillation counter (Beckman LS-1801; Beckman Instruments Inc, Irvine, CA). Permeability of isotopes was assessed by measuring the appearance of the marker on the serosal side during the experiments. The apparent permeability coefficient (Papp) was calculated by standard equation.21,22 Statistics All data are expressed as mean ⫾ SD. Kruskal-Wallis one-factor analysis of variance (ANOVA) was used for statistical analysis. Differences were considered significant at the p ⬍ .05 level. Ethics This study was approved by the Animal Ethics Committee of the Faculty of Health Science, University of Michigan. RESULTS

Epithelial Function Studies Ion transport experiment. Intestinal segments were taken 5 cm proximal to the Treitz ligament and were mounted in modified Ussing chambers (Physiologic Instruments, Inc, San Diego, CA), with an exposed tissue area of 0.3 cm2. Each half-cell (mucosal and serosal) was filled with 5 mL of preheated 37°C Krebs buffer. The Krebs buffer contained NaCl (110.0 mmol/L), CaCl2 (3.0 mmol/L), KCl (5.5 mmol/L), KH2PO4 (1.4 mmol/L), NaHCO3 (29.0 mmol/L), and MgCl2 (1.2 mmol/L), was adjusted to a pH of 7.4, and was continuously oxygenated with O2/CO2 (95/5%) and stirred by gas flow in the chambers. The serosal buffer

335

Mucosal Parameters Intestinal wet weight, DNA, and protein. Intestinal wet weight, DNA, and protein content were reduced in the jejunum of mice receiving TPN. Intestinal wet weight, protein, and DNA decreased by 18.9%, 27.6%, and 44.9%, respectively, after 7 days of TPN. KGF administration led to a significant change in intestinal wet weight, DNA, and protein contents. The intestinal wet weight from TPN mice supplemented with KGF significantly increased by 23%, protein content increased by 23%, and DNA content increased by 48.1% when compared with TPN mice (Table I).

336

YANG ET AL

Vol. 26, No. 6

TABLE I Intestinal wet weight, protein, and DNA contents in the jejunum Control

TPN

TPN-KGF

Protein (mg/cm) 2.50 ⫾ 0.39* 1.81 ⫾ 0.24 2.23 ⫾ 0.12* DNA (␮g/cm) 29.8 ⫾ 2.2* 16.4 ⫾ 2.4 24.3 ⫾ 3.5* Intestinal weight (mg/cm) 116.2 ⫾ 6.2* 94.2 ⫾ 5.4 107.5 ⫾ 4.8* Data are expressed as mean ⫾ SD. *Compared with the TPN group, p ⬍ .05. TPN, total parenteral nutrition; KGF, keratinocyte growth factor.

Intestinal histology. TPN led to significant atrophy of the jejunum. Villus height was 210 ⫾ 45 ␮m in the TPN group compared with 310 ⫾ 42 ␮m in controls. KGF administration significantly reversed the loss of villus height (Fig. 1). Villus height in the TPN-KGF group was 273 ⫾ 39 ␮m. The increase in villus height with KGF, however, did not fully reach the values of the control group. Intestinal epithelial proliferation rates of crypt cells. There was no difference in labeled cell position between groups of mice injected with BrdU. BrdU-positive cells were all distributed in the crypts of Lieberkuhn of the small intestine. TPN significantly (p ⬍ .01) decreased BrdU positive cells to 13.5 ⫾ 2.4% compared with 24.4 ⫾ 4.2% in controls. KGF administration significantly (p ⬍ .01) attenuated the decrease in crypt cells proliferation caused by TPN (19.8 ⫾ 1.9%; Fig. 2). Intestinal epithelial cells apoptosis. We found that morphologic changes also correlated with an increase in intestinal epithelial apoptosis. The epithelial cell apoptotic rate in mice receiving TPN was significantly (p ⬍ .01) higher (21.1 ⫾ 3.2%) compared with controls (7.8 ⫾ 0.7%). KGF administration significantly decreased the epithelial apoptosis to 15.1 ⫾ 4.3%. The difference between each group was significant (p ⬍ .05; Fig. 3). Epithelial Transport Function Epithelial basal ion transport. Isc is an indicator of active ion transport. TPN led to significant Isc transport abnormalities. The baseline Isc significantly (p ⬍ .01) increased in intestinal tissues from mice receiving TPN (56.5 ⫾ 8.3 ␮A/cm2) compared with controls

FIG. 1. Change in the jejunal villus height after 7 days of total parenteral nutrition (TPN), or TPN supplemented with keratinocyte growth factor (KGF), compared with control mice. The TPN group received a standard TPN solution IV with no oral intake. Controls received physiologic saline in addition to standard laboratory mouse chow and water ad libitum. *p ⬍ .05 compared with the TPN group. Data are expressed as mean ⫾ SD.

FIG. 2. 5-Bromo-2-deoxyuridine (BrdU) positive cells in mouse jejunum with immunohistochemical staining. Mice were injected with BrdU, 50 mg/kg body weight intraperitoneally 1 hour before death. (A) Control mice; (B) mice receiving total parenteral nutrition (TPN). (C). TPN mice supplemented with keratinocyte growth factor (KGF). Magnification, 45⫻.

(27.4 ⫾ 4.8 ␮A/cm2). KGF administration in the TPNKGF group significantly (p ⬍ .01) attenuated the increase in Isc caused by TPN. Basic Isc in the TPNKGF group was 35.6 ⫾ 4.9 ␮A/cm2 (Fig. 4); however, levels remained elevated above control values. Epithelial-stimulated ion transport. Carbachol was added to the serosal side to determine secretagogueinduced changes in Cl⫺ secretion. Carbachol-induced elevation in Isc was significantly (p ⬍ .01) higher in the TPN group compared with controls. This increase in Isc with TPN suggests an increased flow of electrolytes into the intestinal lumen that could lead to fluid losses.

FIG. 3. Flow cytometry histogram of apoptosis (Annexin V staining) in small intestinal epithelial cells for control (A), total parenteral nutrition (TPN) (B), and TPN-keratinocyte growth factor (KGF) (C) groups. The apoptotic index is expressed as percentage of apoptotic cells per 100 epithelial cells determined by flow cytometry. TPN caused a significant increase in epithelial cell apoptosis compared with controls. KGF administration significantly (p ⬍ .05) decreased the rate of epithelial apoptosis.


337


changes in Isc with glucose absorption. Glucose evoked a significantly (p ⬍ .01) greater increase in Isc in intestinal segments from TPN mice. KGF supplement did not significantly reduce this response to levels, which approximated the control group (Fig. 5B). Epithelial Barrier Function

FIG. 4. Baseline short circuit current (Isc) (␮A/cm2) at 0 minutes after mounting mouse intestine from different groups. Total parenteral nutrition (TPN) significantly increased the intestinal baseline Isc. Keratinocyte growth factor (KGF) administration attenuated the increase in Isc induced by TPN. *p ⬍ .01 compared with the TPN group. Values are means ⫾ SD. Time 0 represents time at which all tissues were mounted in Ussing chambers after a 30-minute equilibration period to achieve steady-state conditions.

The decline in this TPN-induced Isc elevation with KGF may result in a conservation of electrolytes. KGF administration (p ⬍ .01) significantly lowered the carbachol-induced Isc response to levels that approximated controls (Fig. 5A). Glucose uptake into enterocytes is linked to sodium absorption, thus leading to

Transepithelial resistance. Transepithelial resistance (TER) is a measure of intestinal epithelial integrity and tissue viability.23 Baseline TER (⍀䡠cm2) at 0 minutes after mounting mouse intestine was 87.1 ⫾ 6.8 ⍀䡠cm2 and 52.1 ⫾ 5.3 ⍀䡠cm2 in control and TPN groups, respectively. TPN resulted in a significant decrease in TER (40%) when compared with tissue from controls. KGF administration partially attenuated the decrease of TER caused by TPN (26.3%), although TER remained lower than controls (Table II). Epithelial permeability. Mannitol was selected because it is reported to diffuse across the epithelium through intracellular and paracellular pathways. It has also been shown that mannitol passes through tight junctions at both the level of the villus and crypt.24 After the equilibration period, there was a constant permeation of 3H-mannitol in the TPN and control groups in the small intestine during the 90-minute incubation period. The Papp for 3H-mannitol in the small intestine can be seen in Table II. TPN significantly increased the permeability values of 3H-mannitol when compared with controls. Similarly, KGF administration significantly decreased the TPN-induced loss of intestinal epithelial barrier function. Changes in Epithelial Cell KGFR Expression Because of the observed increase in epithelial cell loss with TPN, we hypothesized that intestinal epithelial cell KGFR expression would be decreased with TPN administration, and KGF administration would increase this KGFR. RT-PCR results showed that TPN resulted in a slight increase in KGFR mRNA expression. KGF administration up-regulated epithelial KGFR mRNA expression to a much greater extent (threefold higher) to levels significantly higher than controls or the TPN group (Fig. 6). Western blot studies of KGFR protein expression results confirmed these changes. Figure 7 shows a composite Western blot from control, TPN, and TPN supplemented with KGF mice. KGF administration increased epithelial KGFR expression. DISCUSSION

In this study, the effects of KGF administration on gut growth and epithelial function in a mouse TPN TABLE II Changes in intestinal permeability and TER in mice after TPN FIG. 5. Changes in intestinal short circuit current (Isc) induced by serosal addition of 10 ␮mol/L carbachol (A) or mucosal addition of 10 mmol/L glucose (B) in tissues from mice after total parenteral nutrition (TPN), TPN-keratinocyte growth factor (KGF), and control groups. Values are means ⫾ SD. ⌬Isc was measured by subtracting the basal current before the addition of the agonist from the peak current after the addition of the agonist, which was then normalized to the serosal area exposed in flux chambers. *p ⬍ .01 compared with TPN.

Papp of mannitol TER (⍀cm2)

Control

TPN

TPN-KGF

7.4 ⫾ 2.8* 87.1 ⫾ 6.8*

18.7 ⫾ 4.1 52.1 ⫾ 5.3

12.0 ⫾ 4.7* 65.8 ⫾ 8.8*

Papp is expressed in 10⫺6 ⫻ cm 䡠 s⫺1. All values are expressed as means ⫾ SD. *p ⬍ .05 compared with the TPN group. TER, transepithelial resistance; TPN, total parenteral nutrition; KGF, keratinocyte growth factor.

338

YANG ET AL

FIG. 6. (A) Composite gel demonstrating epithelial keratinocyte growth factor receptor (KGFR) mRNA expression from a representative specimen in control, total parenteral nutrition (TPN), and TPN-keratinocyte growth factor (KGF) groups. (B) Changes in KGFR mRNA expression in small intestinal epithelial cells in the TPN, TPN-KGF, and control groups. *p ⬍ .05, comparing expression with control groups. M represents a 123 base pair ladder.

model were investigated. The results of this study demonstrate that KGF administration had significantly beneficial effects in preventing the loss of epithelial structure and function with TPN. KGF supplementation improved the intestinal DNA, protein content, increased mucosal crypt proliferation cells, and villus height. At the same time, KGF decreased epithelial cell apoptosis. Additionally, KGF administration attenuated TPN-induced epithelial barrier and ion transport dysfunction. Further KGF led to an up-regulation in small intestine mucosal KGFR expression, which may play an important role in epithelial proliferation. A number of alterations occur in the intestinal mucosa during the administration of TPN, including a decline in mucosal epithelial growth and function. In several animal models of TPN, the intestinal epithelium shows a loss of villus height and a decline in epithelial growth.1,25 This onset of hypoplasia and hypofunction could occur as early as after 3 days of TPN in rats and 1 to 2 weeks in mice and humans.26 In our studies, atrophy, as shown by a loss of villus height, occurred in mice within 7 days of TPN. This atrophy seems to be caused by both a decline in epithelial cell proliferation and an increase in epithelial cell apoptosis. KGF is a known mitogenic growth factor and is a fibroblast-derived member of the fibroblast growth factor family. It has been shown to stimulate proliferation of a variety of epithelial cell lines.27 Intestinal T-cell receptor ␥␦⫹ IEL has been found to express KGF.8 The

Vol. 26, No. 6

close proximity of IEL and epithelial cells suggests that this source of KGF plays an important role in regulating growth of small intestinal mucosa. Administration of exogenous KGF can prevent mucositis during the administration of chemotherapy and radiation to the intestine.12 KGF can also ameliorate mucosal injury in an experimental model of colitis in rats.13 Additionally, KGF has been shown to be able to prevent mucosal atrophy when given in a rat model of starvation.14 All these results suggest a relative deficiency of KGF in many of these disorders. We considered that intestinal villus atrophy may be related to lack of KGF production. In fact, a preliminary report by our group showed a decrease in intraepithelial lymphocyte KGF mRNA expression in mice receiving TPN.28 Based on this, we hypothesized that KGF administration might up-regulate KGFR expression. We further hypothesized that this up-regulation in KGFR may in turn stimulate the recovery of epithelial structure and function in a mouse model of TPN. Epithelial cell proliferation takes place in the crypt region, before movement and differentiation of cells along the villi toward the villus tip.29 BrdU becomes incorporated into the proliferating cells of tissues. We found that BrdU positive cells in small intestine of mice treated with TPN alone was the lowest among all groups, but proliferation markedly increased with the addition of KGF. Our results are supported by a study from Playford et al,30 which showed that recombinant KGF administration led to a significant dose-related increase in the weight of the stomach, small intestine, cecum, and colon in a rat TPN model. They also found that the increase in the weight of the stomach and small intestine were associated with increased epithelial proliferation, as assessed by a 2-hour metaphase determination. More recently, Johnson et al15 found that KGF supplementation increased mucosal wet weight, DNA, and protein content in rats 3 days after an 85% small bowel resection. KGF administration

FIG. 7. Composite western blot of keratinocyte growth factor receptor (KGFR) expression in control, total parenteral nutrition (TPN), and TPN-keratinocyte growth factor (KGF) mice. Protein from epithelial cells were subjected to Western blot analysis with polyclonal KGFR antibody. ␤-actin was included as a loading control. Two KGFR proteins (120 and 110 kDa) were detected. The major size of KGFR was 120 kDa. Lane A, control; lane B, TPN group; lane C, TPN-KGF group. KGF administration increased epithelial KGF receptor expression.



also significantly increased the mucosal villus height and crypt depth. Our study showed that this increase in weight was caused by an increase in mRNA and protein formation. We further showed that KGF not only led to an increase in epithelial cell proliferation but a decline in apoptosis. Apoptosis is an active process of gene-directed cell nuclear self-destruction. Apoptosis occurs in many tissues, including the gastrointestinal tract.31 There is normally a steady low level of apoptosis in the small intestine.32 It is suggested that apoptosis plays an important role in the small intestine in terms of its homeostasis. In this study, we found that the apoptotic index in the control group was the lowest. It increased greatly in mice treated with TPN. KGF administration partially reversed the increase in epithelial apoptotic index caused by TPN. This might suggest that aside from the inhibition of proliferation, atrophy of the small intestine was also caused by an augmentation of apoptosis. It also suggested that the role of KGF in gut adaptation is not only by an increase in epithelial cell proliferation, but also by a decrease in epithelial cell loss (apoptosis). KGFR has been found to be ubiquitously express in the mucosal epithelium of all gastrointestinal tract segments. Using immunofluorescent staining of KGFR, Chailler et al9 found that KGFR is located at the basolateral membrane and cytoplasm in differentiated enterocytes and in crypt cells and has an increasing gradient of expression toward the apex of the crypt-villus axis. This baso-lateral position allows the receptor to lie in close proximity to IEL. Our study found that KGF administration significantly increased KGFR expression in a TPN mouse model. Our data are supported by a study from Estivariz et al33 They found that administration of recombinant KGF significantly increased small intestinal mucosa KGFR mRNA expression in rats subjected to 3 days of starvation or provision of a low level of luminal nutrition (25% of control group intake) for 3 days after 3 days starvation. Interestingly, this increase in receptor expression was noted in both the ileum and the colon. A parallel increase in KGFR mRNA and protein abundance was demonstrated in hyperplastic epidermal keratinocytes from patients with psoriasis.34 The mechanisms of KGF-induced gut growth are unknown. The increased expression of KGFR correlates with epithelial cell proliferation and supports the hypothesis that KGFR signaling is crucial for the mitogenic stimulus for gut adaptation. A study from Jonas et al35 found that KGF markedly enhanced glutathione (GSH) levels and GSH antioxidant capacity in small intestinal mucosa in 25% refed rats compared with saline-treated controls. Their results demonstrate that KGF may up-regulate tissue GSH by an endocrine mechanism, which may represent another mechanism for KGF-induced mucosal growth.14 Transport capacity is the essential functional measurement that determines the nutritional status of an animal. Extremes of electrogenic ion transport can result in malabsorption. We therefore investigated whether KGF administration is associated with improved epithelial transport and barrier function.

339

Our results showed that intestinal fluid and ion transport is different in orally fed animals compared with those given TPN. Baseline Isc was determined as an indication of the ion transport state of the tissue.36 We found that TPN resulted in a significant increase in basal Isc. Similarly, Peterson et al37 found that TPN caused a twofold rise in Isc in a TPN rat model when compared with chow-fed rats. Carbachol, which is reported to evoke active chloride secretion by crypt glands,38 – 40 increased Isc in TPN mice. Glucose uptake by the apical membrane of enterocytes is linked to sodium and increases Isc by stimulating mucosal to serosal flux (ie, stimulating transepithelial sodium absorption). Glucose evoked a significantly increased Isc in tissue from TPN mice. However, KGF treatment significantly attenuated all of these ion transport changes induced by TPN except the increase in glucose stimulated sodium absorption. The barrier function of the gut permits the absorption of nutrients while preventing systemic contamination by luminal toxins and microbial products.41 TER is considered to reflect tissue tight junction integrity. We found that TER significantly decreased in mice receiving TPN. These results are supported by the findings of Peterson et al20,37 who showed that TPN administration in a rat model led to an increased intestinal conductance. We also found that, concomitant with our observed decrease in TER, there was an increase in intestinal permeability to mannitol. KGF administration partly attenuated the increase of permeability caused by TPN. The reason why KGF administration increases the intestinal integrity function is unclear. Future studies on the changes in intestinal permeability in TPN supplemented with KGF may help to answer whether this is because of the effect of KGF on epithelial growth, or if KGF has a direct role on epithelial tight junction (independent of epithelial structure changes). In conclusion, this study demonstrated that administration of TPN in a mouse model caused profound changes in small bowel epithelial morphology and physiology. These changes include mucosal atrophy and alteration in ion transport and epithelial barrier function. KGF administration reversed most of these TPN-induced changes. KGR administration also upregulates the epithelial KGFR expression. It is possible that KGF may have therapeutic efficacy in patients who are receiving TPN. ACKNOWLEDGMENTS

This research was supported by National Institutes of Health Grant AI44076-01. REFERENCES 1. Levine GM, Deren JJ, Steiger E, et al: Role of oral intake in maintenance of gut mass and disaccharide activity. Gastroenterology 67:975–982, 1974 2. Buchman A, Moukarzel AA, Bhuta S, et al: Parenteral nutrition is associated with intestinal morphologic and functional changes in humans. JPEN 19:453– 460, 1995 3. Klein RM, McKenzie JC: The role of cell renewal in the ontogeny of the intestine. II. Regulation of cell proliferation in adult, fetal, and neonatal intestine. J Pediatr Gastroenterol Nutr 2:204 –228, 1983

340

YANG ET AL

4. Chowdhury A, Fukuda R, Fukumoto S: Growth factor mRNA expression in normal colorectal mucosa and in uninvolved mucosa from ulcerative colitis patients. J Gastroenterol 31:353– 360, 1996 5. Lo HC, Hinton PS, Peterson CA, et al: Simultaneous treatment with IGF-I and GH additively increases anabolism in parenterally fed rats. Am J Physiol 269(2 Pt 1):E368 –E376, 1995 6. Tsai C, Hill M, Drucker K: Biological determinants of intestinotrophic properties of GLP-2 in vivo. Am J Physiol 272:G662– G668, 1997 7. Dignass A, Lynch-Devaney K, Kindon H, et al: Trefoil peptides promote epithelial migration through a transforming growth factor beta-independent pathway. J Clin Invest 94:376 –383, 1994 8. Boismenu R, Havran WL: Modulation of epithelial cell growth by intraepithelial gamma delta T cells. Science 266:253–255, 1994 9. Chailler P, Basque JR, Corriveau L, et al: Functional characterization of the keratinocyte growth factor system in human fetal gastrointestinal tract. Pediatr Res 48:504 –510, 2000 10. Simmons J, Pucilowska J, Lund P: Autocrine and paracrine actions of intestinal fibroblast-derived insulin-like growth factors. Am J Physiol 39:G817–G827, 1999 11. Housley R, Morris CF, Boyle W, et al: Keratinocyte growth factor induces proliferation of hepatocytes and epithelial cells throughout the rat gastrointestinal tract. J Clin Invest 94:1764 –1777, 1994 12. Farrell CL, Bready JV, Rex KL, et al: Keratinocyte growth factor protects mice from chemotherapy and radiation-induced gastrointestinal injury and mortality. Cancer Res 58:933–939, 1998 13. Zeeh JM, Procaccino F, Hoffman P, et al: Keratinocyte growth factor ameliorates mucosal injury in an experimental model of colitis in rats. Gastroenterology 110:1077–1083, 1996 14. Estivariz CF, Jonas CR, Gu LH, et al: Gut-trophic effects of keratinocyte growth factor in rat small intestine and colon during enteral refeeding. JPEN 22:259 –267, 1998 15. Johnson WF, DiPalma CR, Ziegler TR, et al: Keratinocyte growth factor enhances early gut adaptation in a rat model of short bowel syndrome. Vet Surg 29:17–27, 2000 16. Kiristioglu I, Teitelbaum DH: Alteration of the intestinal intraepithelial lymphocytes during TPN. J Surg Res 79:91–96, 1998 17. Burton K: A study of the conditions and mechanisms of the diphenylamine reaction for the colormetric estimation of deoxyribonucleic acid. Biochem J 62:315–323, 1956 18. Mosley RL, Klein JR: A rapid method for isolating murine intestine intraepithelial lymphocytes with high yield and purity. J Immunol Methods 156:19 –26, 1992 19. Witkowski J, Miller R: Increased function of P-glycoprotein in T lymphocyte subsets of aging mice. J Immunol 150:1296 –1306, 1993 20. Peterson CA, Carey HV, Hinton PL, et al: GH elevates serum IGF-I levels but does not alter mucosal atrophy in parenterally fed rats. Am J Physiol 272(5 Pt 1):G1100 –G1108, 1997 21. Madara JL, Trier JS: Structure and permeability of goblet cell tight junctions in rat small intestine. J Membrane Biol 66:145– 157, 1982

Vol. 26, No. 6

22. Grass G, Sweetana S: In vitro measurement of gastrointestinal tissue permeability using a new diffusion cell. Pharm Res 6:372– 376, 1988 23. Smith PL: Methods for evaluating intestinal permeability and metabolism in vitro. Pharm Biotechnol 8:13–34, 1996 24. Bjarnason I, MacPherson A, Hollander D: Intestinal permeability: An overview. Gastroenterology 108:1566 –1581, 1995 25. Johnson LR, Copeland EM, Dudrick SJ, et al: Structural and hormonal alterations in the gastrointestinal tract of parenterally fed rats. Gastroenterology 68(5 Pt 1):1177–1183, 1975 26. Gavrieli Y, Sherman Y, Ben-Sasson SA: Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol 119:493–501, 1992 27. Rubin J, Osada H, Finch P: Purification and characterization of a newly identified growth factor specific for epithelial cells. Proc Natl Acad Sci USA 806:802–906, 1989 28. Antony P, Yang H, Teitelbaum DH, et al: Altered expression of intraepithelial lymphocyte (IEL) keratinocyte growth factor (KGF) mRNA in the mouse. Gastroenterology 118(Suppl 2):A658, 2000 29. Johnson LR: Regulation of gastrointestinal mucosal growth. Physiol Rev 68:456 –502, 1988 30. Playford RJ, Marchbank T, Mandir N, et al: Effects of keratinocyte growth factor (KGF) on gut growth and repair. J Pathol 184:316 –322, 1998 31. Cummings MC, Winterford CM, Walker NI: Apoptosis. Am J Surg Pathol 21:88 –101, 1997 32. Hall PA, Coates PJ, Ansari B, et al: Regulation of cell number in the mammalian gastrointestinal tract: The importance of apoptosis. J Cell Sci 107(Pt 12):3569 –3577, 1994 33. Estivariz CF, Gu LH, Scully S, et al: Regulation of keratinocyte growth factor (KGF) and KGF receptor mRNAs by nutrient intake and KGF administration in rat intestine. Dig Dis Sci 45:736 –743, 2000 34. Finch PW, Murphy F, Cardinale I, et al: Altered expression of keratinocyte growth factor and its receptor in psoriasis. Am J Pathol 151:1619 –1628, 1997 35. Jonas CR, Estivariz CF, Jones DP, et al: Keratinocyte growth factor enhances glutathione redox state in rat intestinal mucosa during nutritional repletion. J Nutr 129:1278 –1284, 1999 36. Saunders PR, Kosecka U, McKay DM, et al: Acute stressors stimulate ion secretion and increase epithelial permeability in rat intestine. Am J Physiol 267(5 Pt 1):G794 –G799, 1994 37. Peterson CA, Ney DM, Hinton PS, et al: Beneficial effects of insulin-like growth factor i on epithelial structure and function in parenterally fed rat jejunum. Gastroenterology 111:1501– 1508, 1996 38. Carey HV, Cooke HJ: Effect of hibernation and jejunal bypass on mucosal structure and function. Am J Physiol 261(1 Pt 1):G37– G44, 1991 39. Carey HV, Cooke HJ, Gerthoffer, WT, et al: Intestinal transport in megacolonic mice. Alterations in sugar absorption. Dig Dis Sci 34:185–192, 1989 40. Cooke HJ: Neurobiology of the intestinal mucosa. Gastroenterology 90:1057–1081, 1986 41. Aranow JS, Fink MP: Determinants of intestinal barrier failure in critical illness. Br J Anaesth 77:71– 81, 1996

Discussant Dr DeLegge: This was a very elegant study and based on sound scientific data. The recent work with growth factors and the small intestine is exciting. Previous work has been done on growth factors and other organ systems; however, the effects of growth factor on the intestine is now a growing field. The hypothesis in this study was very sound. We know from other work that TPN does cause difficulty with small bowel atrophy, which may lead to increased mucosal permeability and an increased rate of sepsis secondary to bacterial translocation. There is also decreased absorption.

Most of the previous work has been done on villus height, DNA content, and the overall weight of the small intestine. There has not been much work reported on actual small intestinal function. Also, there has not been much work reported regarding the RNA expression within the small intestine. There is some controversy regarding exactly where KGF is produced. The authors here have some previous data regarding it’s production on the intraepithelial lymphocyte. Other investigators have not been successful at delineating the IEL as the site of production of KGF.



The statistical analysis for this paper was particularly appropriate. The conclusions were soundly based on the results. One of the most exciting issues for me was the transport of intestinal iron and intestinal permeability to iron with regard to KGF. An earlier work looked at hypocaloric feedings on small intestinal KGF expression from the mucosa. Within that work, intestinal KGF production was not seen in the jejunum, but was seen in the ileum and in the colon. It was

341

also seen in the duodenum. This raises the question regarding the production site of KGF within the intestine. This same study begs the question of whether KGF production is affected by hypocaloric feedings, or in response to the use of TPN. In conclusion, I would say this was a well thought out paper. Its statistical analysis was done very elegantly, and the work on iron transport and membrane permeability is very exciting.

Response Dr Yang: Let me first answer your questions one by one. The different changes in KGFR expression are not caused by an increase in villus height. In fact, the expression is based on identical amounts of either mRNA or protein as isolated from control, TPN, and KGF ⫹ TPN mice. Thus, an increased

number of cells in the KGF-treated group would not interfere with our results. And the second question, the KGF dose used is based on the experience of two additional groups who used 5 mg/kg per day and several different doses from 0.1 to 3 mg/kg (the highest dose was the most effective).