A8 REPEATED SUBMERGENCE OF AIR-LIQUID INTERFACE COLONOID CULTURES IMPAIRS INFLAMMATORY AND REGENERATIVE RESPONSES

Background

Damage in the intestinal epithelium is repaired via de-differentiation of mature intestinal epithelial cells to a stem-like state. Indeed, literature has primarily focused on acute forms of intestinal damage, but there is a lack of models to study how intestinal stem cells function after chronic injury, such as in inflammatory bowel disease (IBD). A previous report found that growth of mouse intestinal organoids in air-liquid interface (ALI) follows by submergence caused differentiation and reversible injury, but this has not been demonstrated in human cells or with repeated cycles of injury. Understanding how chronic damage alters human intestinal stem cell fate and function is imperative to developing novel therapies that repair the epithelium in people with IBD

Aims

To develop a robust in vitro model to differentiate and damage human intestinal epithelial cells, with or without the addition of bacterial flagellin to mimic pathogen exposure.

Methods

Human colonoid monolayers were seeded on Transwell inserts for 10 days until fully confluent and then differentiated by removing the apical media to create ALI growth conditions for 7 days. To induce damage, media was added to the apical side of the Transwell, with or without the addition of flagellin in the basolateral compartment. Following submergence induced damage, the apical media was removed and collected for chemokine analysis, and the cells were grown back in ALI for 3 days to recover them from injury. This cycle was repeated 5 times to induce chronic damage. Cells were collected for qPCR analysis, immunofluorescence imaging, RNA sequencing and DNA methylation analysis

Results

Repeated rounds of damage impaired the ability of intestinal epithelial cells (IECs) to respond to TLR stimulation (a decrease in basolateral IL-8 with each round), likely due to a decrease in TLR signaling pathways, as demonstrated by GSEA and qPCR. Chronic submergence damage led to an increase in differentiation of cells expressing MUC2, SLC26a3 and CHGA, and a decrease in stemness as shown by qPCR for BMI1, HOPX, and LGR5. After several rounds of damage, colonoid monolayers were unable to regrow as monolayers after passaging, likely due to a decrease in YAP signaling. We also identified mRNA expression and DNA methylation changes in genes associated with IBD and colon cancer.

Conclusions

We have developed a novel chronic damage model of recurrent IEC injury, which possibly mimics pathologies seen in people with inflammatory bowel disease. This model can be used to understand how chronic damage alters the ability of IECs to respond to pathogens and regenerate to repair and protect the epithelium from further damage.

Funding Agencies

CCC

A202 DYSREGULATED ENDOPLASMIC RETICULUM STRESS PATHWAYS IN COLON-DERIVED ENTEROIDS FROM INFLAMMATORY BOWEL DISEASE PATIENTS DRIVE DC MATURATION LEADING TO A PRO-INFLAMMATORY PHENOTYPE

Background

Intestinal epithelial cells (IECs) rely on danger signals such as adenosine triphosphate (ATP) and endoplasmic reticulum (ER) stress to form an appropriate, coordinated immune response. Moreover, ER stress in IECs contributes to the pathogenesis of inflammatory bowel disease (IBD), and an increase in extracellular ATP concentrations is a risk factor for Crohn’s disease (CD).

Aims

We hypothesized that cells exposed to ER stress or ATP modulate innate immune responses, creating a pro-inflammatory environment that drives dendritic cell (DC) maturation.

Methods

Caco-2 cells and human colon-derived enteroid monolayers were exposed to ATP or the ER stress inducer thapsigargin, stimulated with E. coli FliC flagellin, and ER stress markers CHOP, GRP78, and XBP1 s/u were measured via qPCR and western blot, and cytokine release was measured by ELISA. Next, monocyte-derived dendritic cells (moDCs) were cultured in Caco-2 or enteroid conditioned supernatants and their activation status measured via flow cytometry, and cytokine analysis was performed using Luminex platform. We also assessed ER stress markers, TLR5 expression, and cytokine expression differences between IBD and healthy controls (HC).

Results

We found that ER stress amplified FliC-induced IL-8 and decreased CCL20 in Caco-2 cells. Moreover, in IBD subjects, we found an increase FliC-induced IL-8 response, and decreased TNFa and CCL20. moDCs cultured with conditioned media from Caco-2 or enteroid monolayers showed a proinflammatroy phenotype, with an increase in CD80, CD86, MHCII, and a decrease in CD103. Moreover, moDCs cultured in stressed Caco-2 supernatants increased release of IL-6, TNFa, and IL-12p70, and decreased IL-10, suggesting potential to induce inflammatory Th1 and/or Th17 cells.. DC activation correlated with the amount of FliC-induced IL-8. Interestingly, there were distinct differences in cytokine expression and basal ER stress between IBD and HC enteroid monolayers, suggesting a dysregulated ER stress pathway in IBD-derived enteroids.

Conclusions

ER stress in Caco-2 cells and colon-derived enteroid monolayers enhances FliC-induced TLR5 responses, leading to a pro-inflammatory environment that drives DC maturation, which may link epithelial ER stress and immune cell activation in IBD. Furthermore, the cytokine and ER stress pathway differences between IBD and HC-derived enteroids suggests that prolonged periods of stress in IBD patients may rewire the IEC stem cell compartment, further perpetuating inflammation and disease.

Funding Agencies

CCC, CIHR

The developmental origins of health and disease: current theories and epigenetic mechanisms

The retrospective cohort studies of David Barker and colleagues during the late 1980s established the principle that the incidence of certain adult diseases such as stroke, type 2 diabetes and dyslipidaemia may be linked to in utero development. Later termed the "Developmental Origins of Health and Disease (DOHaD)" hypothesis, there have been several more recent attempts to explain this phenomenon. Although a general conceptual framework has been established to explain how mechanisms may have evolved to facilitate rapid adaptations to changing ecological conditions, it doesn't identify the actual mechanisms responsible for such effects. Extensive covalent modifications to DNA and related proteins occur from the earliest stages of mammalian development. These determine lineage-specific patterns of gene expression and so represent the most plausible mechanisms by which environmental factors can influence development during the life course. In providing a contemporary overview of chromatin modifications during early mammalian development, this review highlights both the complexity and our current lack of understanding of how epigenetic alterations may contribute to in utero programming. It concludes by providing some thoughts to future research endeavours where the emphasis should be on bettering our understanding of epigenesis and devising more thoughtful experimental approaches that focus on specific environmental factors in appropriate animal and cellular models.

DNA damaging agents increase gadd153 (CHOP-10) messenger RNA levels in bovine preimplantation embryos cultured in vitro

DNA damage and other forms of stress are believed to be important factors in reducing the efficiency of in vitro embryo transfer techniques in farm animals. The expression of mRNAs from stress-responsive genes such as gadd153 (CHOP-10, ddit3) may provide a means of assessing the quality of embryos produced in vitro. Treatment of bovine granulosa cell cultures with the DNA-damaging agents, methyl methane-sulphonate (MMS) or sodium arsenite, induced the expression of an mRNA, which hybridized with the hamster gadd153 cDNA. Part of the corresponding bovine cDNA was amplified by nested polymerase chain reaction (PCR), cloned, and sequenced. Using a sensitive reverse transcriptase-PCR assay we have investigated the expression of gadd153 and beta-actin in blastocyst-stage bovine embryos treated with MMS or sodium arsenite. Both agents produced an increase in the ratio of gadd153 mRNA relative to beta-actin. These results show that there are changes in gene expression in blastocyst-stage bovine embryos in response to genotoxic stress, suggesting that an increase in gadd153 mRNA is a useful marker of DNA damage and metabolic stress in preimplantation embryos.

Maternal protein deficiency causes hypermethylation of DNA in the livers of rat fetuses

Maternal protein deficiency during pregnancy is associated with changes in glucose tolerance and hypertension in the offspring of rats. In this study the growth of rat fetuses was examined when the dams were fed diets containing 18% casein, 9% casein or 8% casein supplemented with threonine. The extra threonine was added to reverse the decrease in circulating threonine concentrations that occurs when pregnant rats are fed protein-deficient diets. The fetuses of the group fed the low protein diet supplemented with threonine were significantly smaller than those of the control group and not significantly different from those fed low protein. Homogenates prepared from the livers of dams fed the diet containing 9% casein oxidized threonine at approximately twice the rate of homogenates prepared from dams fed the diet containing 18% casein. We conclude that circulating levels of threonine fall as a consequence of an increase in the activity of the pathway that metabolizes homocysteine produced by the transulfuration of methionine. Serum homocysteine was unaffected in the dams fed low protein diets compared with controls, but was significantly greater in dams fed the low protein diet supplemented with threonine. Elevated levels of homocysteine are associated with changes in the methylation of DNA. The endogenous methylation of DNA was greater than that of controls in the livers of fetuses from dams fed the 9% protein diets and increased further when the diet was supplemented with threonine. Our results suggest that changes in methionine metabolism increase homocysteine production, which leads to changes in DNA methylation in the fetus. An increase in maternal homocysteine may compromise fetal development, leading to the onset of glucose intolerance and hypertension in adult life.

Upregulation of CHOP-10 (gadd153) expression in the mouse blastocyst as a response to stress

CHOP-10 (also known as gadd153 or Ddit3) is one of the genes overexpressed by mammalian cells exposed to cytotoxic agents or to nutrient stress. The response of this gene to stress was studied in the mouse blastocyst and in F9 embryonal carcinoma cells. When mouse blastocysts were exposed to the alkylating agent MMS, the metabolic inhibitor sodium arsenite or an inhibitor of protein glycosylation tunicamycin, levels of the CHOP-10 mRNA were increased by two- to threefold relative to the mRNA for beta-actin. There was no increase in gene expression when blastocysts were treated with the inhibitor of nucleotide synthesis PALA. These results show that the response of CHOP-10 is dependent on the type of stress applied to the embryo. When F9 embryonal carcinoma cells were treated with MMS or sodium arsenite, CHOP-10 expression was induced by fourfold within 4 hr of treatment. The induction following tunicamycin treatment was slower requiring at least 24 hr. The response to tunicamycin was greater in cells treated with retinoic acid to induce differentiation. The results suggest that there is a link between the extent of glycoprotein synthesis and the sensitivity of CHOP-10 to tunicamycin. The inhibitor PALA did not change CHOP-10 expression in the presence or absence of retinoic acid. In F9 cells an increase in the expression of CHOP-10 was followed by cell death due to apoptosis. The overexpression of CHOP-10 may be a marker for one of the pathways that lead to apoptosis in the blastocyst. These results suggest that there is more than one control system regulating growth arrest in the blastocyst and the fetal outcome may differ depending on the type of stress encountered in culture.

Expression of the growth arrest genes (GAS and GADD) changes during organogenesis in the rat fetus

Mammalian cells mount an active response to nutrient limitation by overexpressing the growth arrest specific (GAS) and the growth arrest and DNA damage (GADD) genes. During embryogenesis in rats, there are quantitative and temporal differences in GAS and GADD gene expression during the development of the placenta, heart and kidney. Genes associated with the inhibition of DNA synthesis (p53 and GAS1) were predominantly expressed during the early stages of development, whereas those genes associated with inhibition of protein synthesis [GADD153 (also known as CHOP-10 or Ddit3) and C/EBP-beta] were more highly expressed during the later stages. The GADD45 gene was expressed throughout development. There were distinct periods of GAS3 and GAS6 gene expression during the development of the placenta, heart and kidneys, which is consistent with the proposed roles of these genes in cell interactions. These results show that there is a change in the expression of genes associated with the negative regulation of growth as the fetus develops.

The effects of maternal protein restriction on the growth of the rat fetus and its amino acid supply

Maternal protein deficiency causes fetal growth retardation which has been associated with the programming of adult disease. The growth of the rat fetus was examined when the mothers were fed on diets containing 180, 90 and 60 g protein/kg. The numbers of fetuses were similar in animals fed on the 180 and 90 g protein/kg diets but the number was significantly reduced in the animals fed on the 60 g protein/kg diet. The fetuses carried by the mothers fed on the 90 g protein/kg diet were 7.5% heavier than those of mothers fed on 180 g protein/kg diet on day 19 of gestation, but by day 21 the situation was reversed and the fetuses in the protein-deficient mothers were 14% smaller. Analysis of the free amino acids in the maternal serum showed that on day 19 the diets containing 90 and 60 g protein/kg led to threonine concentrations that were reduced to 46 and 20% of those found in animals fed on the control (180 g/kg) diet. The other essential amino acids were unchanged, except for a small decrease in the branched-chain amino acids in animals fed on the 60 g protein/kg diet. Both low-protein diets significantly increased the concentrations of glutamic acid+glutamine and glycine in the maternal serum. On day 21 the maternal serum threonine levels were still reduced by about one third in the group fed on the 90 g protein/kg diet. Dietary protein content had no effect on serum threonine concentrations in nonpregnant animals. Analysis of the total free amino acids in the fetuses on day 19 showed that feeding the mother on a low-protein diet did not change amino acid concentrations apart from a decrease in threonine concentrations to 45 and 26% of the control values at 90 and 60 g protein/ kg respectively. The results suggest that threonine is of particular importance to the protein-deficient mother and her fetuses. Possible mechanisms for the decrease in free threonine in both mother and fetuses and the consequences of the change in amino acid metabolism are discussed.

Effects of maternal vitamin A status on fetal heart and lung: changes in expression of key developmental genes

Vitamin A is required during pregnancy for fetal lung development. These experiments monitored fetal lung morphology in normal and vitamin A-deficient rats. The expression of elastin and the growth arrest-specific gene 6 (gas6) in fetal and neonatal hearts and lungs was assessed by Northern blotting. In normal-fed rats, elastin and gas6 were expressed in the fetal lung and heart from day 19 of gestation up to day 2 postnatally. Maternal vitamin A deficiency altered fetal lung development. On day 20, the bronchial passageways were less developed and showed reduced staining for elastic fibers, and in the neonates, the relative air space and the size of the sacculi were reduced. In the fetal lung, the mRNAs for elastin and gas6 were reduced to 56 and 68% of the control values, respectively. In the fetal heart, the mRNA for elastin was reduced to 64% of the control value, whereas gas6 was increased twofold. In the neonate, there was no change in elastin expression in the lung or heart, but gas6 expression in the heart was increased twofold. These results suggest that, in the pregnant rat, vitamin A deficiency may retard fetal lung development or influence the differentiation of critical cell lines. The changes in elastin and gas6 expression may be used to identify the cell types affected.

Gastrointestinal tolerability of meloxicam and piroxicam: a double-blind placebo-controlled study

AIMS

The aim of the study was to compare the effects of meloxicam and piroxicam on the gastroduodenal mucosa in healthy adults.

METHODS

Forty-four healthy volunteers were given a 28 day course of either meloxicam 15 mg, piroxicam 20 mg or placebo. Damage to the oesophageal, gastric and duodenal mucosa was assessed, mucosal blood flow (MBF) measured at endoscopy and biopsies taken for prostaglandin content and microscopic assessment of damage before NSAID administration and during days 1, 7 and 28 of continued intake.

RESULTS

Maximal macroscopic gastric mucosal damage (median grade+IQR) occurred within 24 h of piroxicam administration, the damage score increasing from 0 to 2.5 (0-3) (P=0.02) at day 1 before falling to 2.0 (0-2) at day 7 and 0 (0-1) at day 28 with resolution of damage observed in six out of the seven subjects who sustained acute injury. No significant macroscopic gastric damage occurred in either of the two other groups although some minor damage was observed in seven subjects taking placebo and five taking meloxicam. There was a trend towards piroxicam causing more acute gastric damage than meloxicam (P=0.06). Baseline antral, body and duodenal MBF were similar in all three groups. No significant changes occurred in any of the groups on any of the visits. There were also no changes in gastric mucosal prostaglandin content in any group.

CONCLUSIONS

These observations suggest that meloxicam causes little acute damage to the upper gastrointestinal tract and piroxicam causes some acute gastric injury but such damage resolves in most subjects by 28 days.

Effects of nutrient deprivation and differentiation on the expression of growth-arrest genes (gas and gadd) in F9 embryonal carcinoma cells

The growth-arrest genes (gas and gadd) are widely expressed during mammalian embryogenesis and may be useful as markers of nutritional stress in the embryo. F9 embryonal carcinoma cells have been used to characterize the effect of serum or amino acid deficiency on growth-arrest gene expression in a differentiating embryonic cell. The differentiation markers, homeobox B2 (HoxB2), collagen type IV and laminin B2, were not induced by growth arrest. Treatment with all-trans retinoic acid (RA) produced a dose-dependent increase in alkaline phosphatase activity, which was unchanged in lysine-deficient medium and reduced in low-serum medium. Low-serum medium also reduced HoxB2 expression. There was a transient 2-6-fold increase in mRNAs for C/EBP-beta, gadd153/CHOP-10 and gas5 genes 24 h after transfer to amino-acid-deficient media. The mRNAs for the gas2 and gas6 genes began to rise slowly by 5-10-fold after a delay of approx. 24 h. The transient increases did not occur in low-serum medium where there was a much smaller and slower increase. Differentiation caused 1-2-fold increases in gas2, gas3 and gas6 mRNA levels. The transient overexpression of gas5, gadd153/CHOP-10 and CCAAT-enhancer-binding protein-beta, and the later expression of gas6 mRNAs in response to amino acid deficiency, were not affected by differentiation. RA treatment increased the expression of gas3 and caused gas2 to be transiently overexpressed in amino-acid-deficient medium. Differentiation in serum-deficient medium did not significantly alter the levels of the growth-arrest gene mRNAs. These results show that in F9 cells the growth-arrest genes are expressed sequentially as a result of nutrient stress.

The growth arrest genes gas5, gas6, and CHOP-10 (gadd153) are expressed in the mouse preimplantation embryo

The gas and gadd family of genes, known collectively as the growth arrest genes, are associated with the negative control of mammalian cell growth. The steady-state levels of their mRNAs are increased by three to fivefold when exponentially multiplying cells are exposed to a variety of stresses including inadequate nutrition or the removal of serum. Reverse transcription-polymerase chain reaction (RT-PCR) has been used to analyze growth arrest gene expression in the preimplantation mouse embryo. The gas5, gas6, and CHOP-10 (gadd153, Ddit3) genes were expressed from the eight-cell stage onward. The gas2 and gas3 genes associated with apoptosis were not expressed. Embryos were cultured in kSOM medium and a semiquantitative RT-PCR method was used to measure the relative gene expression using beta-actin mRNA as a reference. The ratio of gas5 to beta-actin mRNA was high at the eight-cell stage and fell three to fivefold during development. The decline in the gas5:beta-actin ratio corresponded to the activation of true cell growth (cytokinesis). The gas6:beta-actin ratio was low at the eight-cell stage and increased by twofold as the blastocyst formed. CHOP-10 was expressed at a constant level throughout development. Embryos that had developed in vivo were compared with the equivalent blastocyst-stage embryos cultured in kSOM medium. There were no significant differences in the ratio of CHOP-10, gas5, or gas6 mRNAs relative to beta-actin. These results suggest that these genes are expressed as part of normal early embryonic development. The potential roles of the growth arrest genes are discussed.

The influence of age, gender, Helicobacter pylori and smoking on gastric mucosal adaptation to non-steroidal anti-inflammatory drugs

INTRODUCTION

Oral NSAIDs cause acute gastric injury that resolves, despite continued administration, by a process known as adaptation. Little is known about the factors that influence this process.

METHODS

Sixty-two healthy volunteers were given a 28-day course of either etodolac 300 mg b.d. (13 subjects), naproxen 500 mg b.d. (23), enteric-coated diclofenac (10) or effervescent diclofenac 50 mg b.d. (16). All subjects were gastroscoped before and on days 1, 7 and 28 during drug administration, to assess gastric mucosal damage using a modified Lanza scale. Subjects were then divided into three categories: those who adapted completely, those who adapted incompletely and those who showed no adaptation. The proportion of subjects in each group was compared with respect to age, gender, smoking, the presence of Helicobacter pylori, and the NSAID prescribed.

RESULTS

Fifty-nine subjects (median age 25.0 years, range 18-70) developed initial gastric injury to NSAIDs of whom 42 adapted completely, 13 adapted incompletely and four showed no evidence of adaptation. The mean age of subjects was lower in those who adapted (26.8 +/- 9.8 years) than those who adapted incompletely (32.5 +/- 10.3 years) and those who did not adapt (42.0 +/- 15.7 years, P = 0.01). There was no evidence of gender influencing adaptation. Of 17 H. pylori-positive subjects, a higher proportion had incomplete adaptation, with only nine subjects adapting completely (53% vs. 81%, P = 0.04). Sixteen subjects were smokers, of whom a greater proportion showed no evidence of adaptation (19% vs. 2%, P = 0.03). A smaller proportion of those who took naproxen (48%) adapted completely than those who took enteric-coated diclofenac (89%), effervescent diclofenac (75%) or etodolac (91%, P = 0.03).

CONCLUSION

Some adaptation occurred in over 90% of subjects after 4 weeks dosing with an NSAID, but adaptation was less frequent in older subjects and in smokers. Complete adaptation occurred less frequently in H. pylori-positive subjects and in those who were given naproxen.

The use of percutaneous endoscopic gastrostomy (PEG) feeding tubes in patients with neurological disease

Percutaneous endoscopic gastrostomy (PEG) is being used increasingly in the treatment of patients with neurogenic dysphagia to improve nutrition and prevent choking and aspiration pneumonia. PEG is used in a wide range of general medical conditions, but its role in clinical neurology is sometimes controversial. This paper reviews the place of PEG in the management of 32 patients with a variety of chronic and progressive neurological disorders. All the patients found it to be an effective and acceptable method of feeding that prevented weight loss, reduced chest infections, facilitated nursing care and improved their quality of life. PEG has an important role in neurological rehabilitation.

ob gene expression and secretion of leptin following differentiation of rat preadipocytes to adipocytes in primary culture

Expression of the ob gene and production of leptin have been examined on differentiation of rat fibroblastic preadipocytes to adipocytes in primary culture. Preadipocytes were obtained from the inguinal fat pad of suckling rats, and following differentiation the cells contained lipid droplets and the mRNAs for both lipoprotein lipase and adipsin were detected by Northern blotting. ob mRNA was not, however, detected on Northern blots, but analysis by RT-PCR indicated that the ob gene was expressed, particularly after differentiation. Measurement of leptin in the culture medium by ELISA showed that the ob gene product was secreted by adipocytes from approximately 4 days after the induction of differentiation. Leptin production was sustained over a 2-week period with a peak at 8-10 days post-induction. Dexamethasone stimulated leptin production, while an inhibition was observed with the beta-adrenoceptor agonist isoprenaline. These results demonstrate that following the differentiation of fibroblastic preadipocytes to adipocytes in primary culture, leptin is secreted with the cells responding to stimuli which regulate production of the hormone.

Influence of Helicobacter pylori on gastric mucosal adaptation to naproxen in man

Our objective was to determine whether H. pylori influences gastric mucosal injury and adaptation caused by naproxen. Twenty-four healthy volunteers, 12 H. pylori-positive and 12 H. pylori-negative, were given a 28-day course of naproxen 500 mg twice a day. They were each gastroscoped to assess gastric mucosal damage and mucosal blood flow before and at 1, 7, and 28 days during treatment. Maximal gastric mucosal damage (median grade + IQR) occurred during the first 24 hr in both groups and was of similar magnitude (H. pylori-positive: 2.5, 2.0-3.0 P < 0.01; H. pylori-negative: 2.0, 1.0-3.0 P < 0.01). This damage was associated with a fall in antral but not corpus mucosal blood flow. With continued NSAID administration, gastric damage resolved confirming adaptation (H. pylori-positive 1.0, 0-2.0, H. pylori-negative: 1.0, 0-1.0) and antral mucosal blood flow returned to baseline in both groups by day 28. These observations suggest that initial gastric mucosal injury is not influenced by H. pylori colonization and adaptation occurs regardless of its presence.

Gastric mucosal injury and adaptation to oral and rectal administration of naproxen

INTRODUCTION

Oral nonsteroidal anti-inflammatory drugs (NSAIDs) cause acute gastric mucosal injury but the relative importance of systemic and topical effect of NSAIDs to overall gastric damage remains uncertain.

METHODS

Twenty-four healthy volunteers were allocated either oral or rectal naproxen 500 mg b.d. and gastroscoped before and during days 1, 7 and 28 of dosing. Macroscopic gastric damage was assessed using a modified Lanza score, mucosal blood flow recorded using laser Doppler flowmetry and prostaglandin E2 (PGE2) measured in antral mucosal biopsies.

RESULTS

Maximal gastric damage occurred during the first 24 h in the oral naproxen group and was associated with a fall in antral mucosal blood flow (mean +/- S.E.M.) from 58.2 +/- 3.3 to 46.6 +/- 4.1 arbitrary units (a.u.) (P < 0.05). With continued administration of oral naproxen, gastric damage resolved and antral mucosal blood flow returned to baseline (54.2 +/- 3.7 a.u.). No macroscopic damage or significant changes in mucosal blood flow were observed during rectal administration. There was no significant difference between mucosal PGE2 concentrations in those receiving oral or rectal naproxen, falling from an initial level of 335 +/- 29 to 155 +/- 49 pg/mg at day 1 (P = 0.06) in those receiving oral naproxen and from 235 +/- 55 to 107 +/- 31 pg/mg at day 1 (P = 0.1) in those receiving rectal naproxen, and remaining suppressed throughout the study in both groups.

CONCLUSIONS

These observations suggest that acute mucosal damage and changes in mucosal blood flow are caused by the topical rather than systemic actions of naproxen.

Gastric mucosal adaptation to etodolac and naproxen

BACKGROUND

Nonsteroidal anti-inflammatory drugs (NSAIDs) frequently cause damage to the gastroduodenal mucosa, principally by suppressing mucosal prostaglandin synthesis. However, such acute mucosal injury usually resolves, despite continued NSAID administration, by a process known as adaptation. Newer NSAIDs, such as etodolac, have been developed to minimize effects on prostaglandin synthesis.

AIM

To determine whether etodolac causes less acute damage than naproxen, and whether the damage produced resolves with continued NSAID administration.

METHODS

Twenty-four healthy volunteers were given a 28-day course of either etodolac 300 mg b.d. or naproxen 500 mg b.d. Gastroduodenal damage was assessed using a modified Lanza scoring system and mucosal blood flow with laser doppler flowmetry at endoscopy before NSAID administration and during days 1, 7 and 28 of continued intake.

RESULTS

Maximum gastric damage (median grade and interquartile range, IQR) occurred during the first 24 h of administration, being greater with naproxen (2.0, IQR 1.0-3.0) than etodolac (1.0, IQR 1.0-1.5; P = 0.03). Such damage was associated with a fall in antral blood flow in the naproxen group (mean +/- S.E.M.) from 54.5 +/- 3.4 to 43.8 +/- 3.4 arbitrary units (P = 0.07) and a slight increase in mucosal blood flow in the etodolac group from 43.5 +/- 2.24 to 49.5 +/- 3.6 arbitrary units. With continued intake this damage resolved in all subjects taking etodolac and in eight of 14 subjects on naproxen. Resolution in the naproxen group was associated with a return to normal of antral blood flow.

CONCLUSIONS

These observations suggest that etodolac causes less mucosal damage than naproxen and that adaptation occurs to both.

Neural influences on human esophageal and salivary alkali secretion

Esophageal secretion HCO3- ions occurs in opossum and man and may contribute to mucosal defense. Using a perfusion technique, neuroregulatory influences on esophageal and salivary HCO3- secretion were investigated in 24 healthy human subjects. The sight and smell of food increased median salivary HCO3- output from 424 to 573 mumol/15 min (P = 0.014), without significantly altering esophageal HCO3- secretion (74-105 mumol/15 min, P = 0.24). Atropine reduced both salivary (610 to 68, 17, 10, and 3 mumol/15 min in successive periods; P < 0.028) and esophageal HCO3- output (108 to 78, 35, 18, and 7 mumol/10 cm/15 min; P < 0.028, respectively. Following atropinization, cholinergic stimulation failed to increase salivary secretion but did "unmask" a small rise in esophageal alkali output (7 to 27 mumol/10 cm/15 min, P = 0.036), implicating a noncholinergic mechanism. Cold-induced pain activated sympathetic reflexes and reduced esophageal HCO3- output (91 to 64 mumol/10 cm/15 min, P = 0.041) without influencing salivary secretion. These observations support a role for the autonomic nervous system in modulating human esophageal and salivary HCO3- secretion.

The biosynthesis of threonine by mammalian cells: expression of a complete bacterial biosynthetic pathway in an animal cell

The coding regions for the Escherichia coli gene for aspartokinase I/homoserine dehydrogenase I (thrA) and the Corynebacterium glutamicum gene for aspartic semialdehyde dehydrogenase (asd) have been subcloned into a Simian Virus 40 (SV40)-based mammalian expression vector. Both enzyme activities are expressed in mouse 3T3 cells after transfer of the corresponding chimaeric gene. The kinetic parameters are similar to those of the native bacterial enzymes, and aspartokinase I/homoserine dehydrogenase I retains its allosteric regulation by threonine. An extract of the cells expressing aspartokinase I/homoserine dehydrogenase I, mixed with one from cells expressing aspartic semialdehyde dehydrogenase, produced homoserine when the mixture was incubated with aspartic acid, ATP and NADPH. The thrA and asd expression cassettes were combined into a single plasmid which, when transfected into 3T3 cells, enabled them to produce homoserine from aspartic acid. Homoserine-producing 3T3 cells were transfected with the plasmid pSVthrB/C (homoserine kinase and threonine synthase) and selected for growth on homoserine. Cell lines isolated from these cells expressed the complete bacterial threonine pathway, were independent of threonine for growth and could be maintained in medium which contained no free threonine. The threonine in the proteins of these cells became enriched in 15N when the culture medium contained [15N]aspartic acid. The production of homoserine and the growth of cells was at a maximum when there was more than 2.5 mM aspartate in the medium. Below this concentration the high Km of aspartokinase limited the flux through the pathway. In the presence of additional aspartic acid the new pathway could sustain a cell cycle time close to that of the same cells cultured in threonine-containing medium.

Review article: factors protecting the oesophagus against acid-mediated injury

Reflux of gastric acid and pepsins into the lower oesophagus causes symptoms such as heartburn and nausea, and tissue injury leading to erosive oesophagitis and stricture formation. This article reviews the mechanisms involved in protecting the oesophagus against acid-mediated injury, including the role of the lower oesophageal sphincter, secondary oesophageal peristalsis and swallowed saliva. The oesophageal mucosa has inherent abilities to resist acid damage, and recent data from three laboratories suggest a secretory function with local production of bicarbonate and mucus responsive to local acidification. The evidence for these putative oesophageal defence mechanisms is discussed.

Effect of topical oesophageal acidification on human salivary and oesophageal alkali secretion

Recent human studies suggest that oesophageal HCO3- secretion, in conjunction with salivary HCO3- secretion and secondary oesophageal peristalsis, is important for the protection of oesophageal mucosa from refluxed gastric contents. This study evaluated simultaneously the responsiveness of oesophageal and salivary HCO3- secretion to oesophageal acidification in eight healthy subjects. A 10 cm segment of oesophagus was perfused at a constant rate of 5 ml/min with a specially designed tube assembly. Saline was used initially, and then 10 mM and 100 mM HCl. The perfusates contained 3H-polyethylene glycol (PEG) as a concentration marker to determine volumes. Corrections were applied for a small degree of contamination by swallowed saliva and refluxed gastric alkali. Oesophageal perfusion with 10 mM HCl did not cause symptoms (nausea and heartburn), but tripled the oesophageal HCO3- output from a baseline of 51 mumol/10 cm/10 min (p = 0.021), while doubling the rate of salivary HCO3- secretion from a median basal value of 140 mumol/10 min (p = 0.021). Oesophageal perfusion with 100 mM HCl was associated with symptoms of nausea and heartburn in all subjects. The median oesophageal HCO3- output increased 32 fold to 1659 mumol/10 cm/10 min (interquartile range 569 to 3373; p = 0.036), and salivary HCO3- secretion approximately tripled from basal values (p = 0.036). In conclusion, oesophageal acidification stimulates both salivary and oesophageal HCO3- secretion, responses which may be protective to the oesophageal epithelium.

The expression of Escherichia coli diaminopimelate decarboxylase in mouse 3T3 cells

We have subcloned the coding sequence for the Escherichia coli lysA gene coding for diaminopimelic acid decarboxylase (DAP decarboxylase) into a eukaryotic expression vector based on the SV40 early promoter. The activities of a series of constructs with different lengths of non-coding DNA at the 5' and 3' ends of the coding region have been compared by measuring the synthesis of lysine from diaminopimelic acid (DAP) in mouse 3T3 cells. A short non-coding sequence at the 3' end reduced the expression of enzyme activity. Stable lines of 3T3 cells have been produced by co-transfection of the chimeric gene with a plasmid coding for G-418 resistance. Cells were grown in medium containing G-418 and resistant clones were screened for an ability to synthesise lysine from DAP. [3H]Lysine produced from [3H]DAP was incorporated into cell proteins. An enzyme extract from a cell line which had incorporated two copies of the gene synthesised 0.082 nmol of lysine/min per mg protein. In the intact cell the rate of lysine synthesis is limited by the uptake of DAP which is taken up at only 5% of the rate of lysine. lysA has a potential as a reporter gene in studies of gene expression in mammalian cells.

Threonine synthesis from homoserine as a selectable marker in mammalian cells

The plasmid pSVthrBC expresses the Escherichia coli thrB (homoserine kinase) and thrC (threonine synthase) genes in mouse cells and enables them to synthesize threonine from homoserine. After transfection with pSVthrBC and culture in medium containing homoserine, only cells that have incorporated pSVthrBC survive. Homoserine at concentrations greater than 1 mM is toxic to mammalian cells. Mouse cells selected from medium containing 5 mM homoserine had incorporated 20-100 copies of the plasmid per cell and had homoserine kinase activities of 0.001-0.012 nmol/min per mg of protein per copy. Cells selected from medium containing 10 mM homoserine had incorporated one or two copies of the plasmid per cell and had homoserine kinase activities of 0.06-0.39 nmol/min per mg of protein per copy. By using high concentrations of homoserine, it is possible to use pSVthrBC to select and isolate cell lines that have one or two copies of the plasmid incorporated into an active region of chromatin. CHO and HeLa cells have also been successfully transfected with pSVthrBC. COS-7 cells are naturally resistant to homoserine as they are able to metabolize homoserine.

Measurement of bicarbonate output from the intact human oesophagus

Injury of the oesophageal mucosa can result from exposure to refluxed gastric acid and pepsin. Competence of the lower oesophageal sphincter and peristaltic activity serve to reduce contract time between luminal acid and oesophageal mucosa, but intraluminal neutralisation of residual acid by bicarbonate may also be important in preventing oesophageal mucosal injury. Whereas swallowed saliva contains bicarbonate, recent experiments have also demonstrated alkali secretion from the mammalian oesophagus. Bicarbonate secretion from the human oesophagus was therefore examined with an intubation technique and perfusion of the oesophagus with a non-absorbable marker. Saliva, gastric, and oesophageal aspirates were collected and bicarbonate concentrations determined by measurements of pH and pCO2 or by back titration. In 32 normal subjects (17 women, 15 men) median basal oesophageal bicarbonate secretion determined by a pH/pCO2 method was 416 (range 139-1050) mumol/hour/10 cm. In a subgroup of 15 experiments median oesophageal bicarbonate output was 489 (range 157-1033) mumol/hour/10 cm (pH/pCO2 method) compared with a median alkali output of 563 (range 135-799) mumol/hour/10 cm as determined by back titration. The difference was not significant. Salivary contamination of the oesophagus accounted for 25% of all bicarbonate measured within the oesophagus and refluxed gastric bicarbonate accounted for 2.5%. Bicarbonate secretion from the normal human oesophagus may, in combination with swallowed salivary bicarbonate, play a part in preventing oesophageal mucosal damage due to refluxed gastric acid and pepsin.

The expression of Escherichia coli threonine synthase and the production of threonine from homoserine in mouse 3T3 cells

We have subcloned the coding sequence for the Escherichia coli threonine synthase gene into a eukaryotic expression vector based on the simian-virus-40 early promoter. When mouse 3T3 cells which already expressed homoserine kinase were transfected with the new plasmid, the cells were able to incorporate radioactivity from [14C]homoserine into their cell proteins. Stable cell lines were established by co-transfecting 3T3 cells with the plasmid coding for threonine synthase and another coding for homoserine kinase and G-418 (Geneticin) resistance. Cells were selected for G-418 resistance and then screened for an ability to synthesize threonine from homoserine and incorporate it into the cell protein. A cell line which expressed both the homoserine kinase and threonine synthase genes was capable of growth in a threonine-deficient medium containing homoserine.

Expression of Escherichia coli homoserine kinase in mouse 3T3 cells

The Escherichia coli gene for homoserine kinase (thrB) has been cloned into a simian-virus-40-based eukaryotic expression vector which also includes a neomycin-resistance gene. Mouse 3T3 cells transfected with this plasmid were selected for resistance and screened for homoserine kinase activity. It has thus been possible to isolate clones which are capable of accumulating homoserine O-phosphate when supplied with homoserine. In broken-cell preparations the kinetic constants for the production of homoserine O-phosphate were similar to those of the wild-type E. coli enzyme. These experiments demonstrate that E. coli homoserine kinase can be expressed in an animal cell and that it can successfully phosphorylate L-homoserine in the intact cell utilizing endogenous ATP.

Mucosal adaptation to indomethacin induced gastric damage in man--studies on morphology, blood flow, and prostaglandin E2 metabolism

The effect of 28 days' continuous administration of oral indomethacin on gastroduodenal morphology, gastric mucosal blood flow, and gastric mucosal prostaglandin E2 (PGE2) metabolism in man was studied to define further the mechanisms of mucosal injury induced by indomethacin. Indomethacin caused acute gastroduodenal damage in all cases, which was maximal at 24 hours of administration. With continued intake, mucosal adaptation occurs resulting in resolution of endoscopic mucosal damage. At the time of maximal mucosal damage, gastric mucosal blood flow was significantly reduced compared with values before treatment (p less than 0.001 in fundus and p less than 0.002 in antrum), with good correlation between the severity of damage and the magnitude of the reduction in blood flow (r = 0.76). Mucosal recovery was associated with a return of the blood flow to normal. PGE2 in mucosal homogenate was significantly reduced by indomethacin in both the fundus (p less than 0.01) and antrum (p less than 0.01) after 24 hours but there was no correlation between the magnitude of this reduction and the severity of mucosal damage (r = -0.34). Despite mucosal recovery by 28 days, PGE2 values remained significantly below those before treatment in both the fundus (p less than 0.01) and antrum (p less than 0.01). The PGE2 degradation capacity was not influenced by indomethacin. In conclusion, mucosal adaptation to acute damage by indomethacin occurs in man and seems independent of local PGE2 metabolism.

Prevention of peptic ulcer relapse by sucralfate: mechanisms of action

Sucralfate has a complex effect on the luminal and mucosal environment of the stomach and duodenum. Some of the actions are important in ulcer healing whilst others are important in preventing subsequent ulcer relapse. Although sucralfate has little direct effect on acid secretion, there is evidence that after ulcer healing with this drug, parietal cell responsiveness is reduced. This may in part be mediated by increased somatostatin release from gastric D cells and may be important in reducing ulcer relapse. Sucralfate has been shown to increase mucosal resistance to damaging agents, such as ethanol and aspirin. Studies have shown that this protective action may be related to the drugs effect on various protective zones such as the 'mucous-bicarbonate' barrier, mucosal hydrophobicity, epithelial cell function and morphology, and mucosal blood flow. These complex actions of sucralfate are in part related to direct interaction between the drug or its components and gastroduodenal tissues, and in part related to effects on various mediators of tissue injury and repair.

Mechanisms of gastroduodenal protection by sucralfate

Over the past 5-10 years, a number of studies have shown that topical sucralfate enhances a number of gastric and duodenal mechanisms, e.g., the "mucus-bicarbonate barrier," mucosal hydrophobicity, mucosal blood flow, cell viability, and local production of prostaglandins, as well as endogenous mediators of tissue injury and repair. It seems likely that the complex actions of sucralfate are in part related to direct interaction between the drug or its components (aluminum, sucrose, and sulfate) and gastric mucosal tissues, and in part related to effects of the drug on the various mucosal mediators of tissue injury and repair. Local actions may play a role in accelerating healing of ulcer-damaged mucosa, but this does not explain the protective actions of sucralfate on normal mucosa. Thus sucralfate appears to enhance the protective function of the "mucus-bicarbonate" barrier by actions on both components. This may depend in part on an interaction with the unstirred layer overlying gastric epithelium. Sucralfate has also been shown to increase the hydrophobicity of mucus gel. There is little doubt that sucralfate increases local production and release of protective prostaglandins (PGs), but the precise role played by these agents in mediating mucosal protection and in chronic ulcer healing remains uncertain. Currently, the mechanism of action of sucralfate on vascular integrity remains unknown and the role of PGs in this protective function is unclear. There is little evidence that epidermal growth factor plays any role in mediating mucosal protection by sucralfate, but it may be important in its ulcer-healing action. Sucralfate has been shown to be truly "cytoprotective" in that it protects isolated epithelial cells from damage by noxious agents. In animals treated with sucralfate, the surface epithelial cells were disrupted, but necrotic lesions in the deep proliferative zone were virtually absent. It seems likely that investigations of the actions of sucralfate and its components will move ever closer to defining the target cells, the intracellular events, and the mediators that bring about its protective and ulcer-healing activity.

Sarcosine kinetics in pigs by infusion of [1-14C]sarcosine: use for refining estimates of glycine and threonine kinetics

To investigate in vivo the interconversion between glycine (Gly) and its N-methyl product sarcosine (Sar), [1-13C]Gly and [1-14C]Sar were infused into hourly fed pigs receiving diets with low- and high-threonine levels. An open two-pool model was developed to calculate Sar demethylation (DM) and Gly methylation (GM). During [1-14C]Sar infusion, intracellular Gly specific radioactivities (SA) in the liver and kidney were higher than plasma Gly SA, suggesting that demethylation of Sar occurred in those tissues. DM estimated by using hippuric acid (HA) as the production pool had a mean value of 1.55 mumol.kg-1.h-1, similar to the Sar production rate (mean 1.85 mumol.kg-1.h-1). GM was undetectable (less than 0.5 mumol.kg-1.h-1). These results suggest that, in fed pigs, Sar is produced mainly from choline catabolism and is degraded only to Gly in liver and kidney. On the assumption that Sar degradation gave rise only to Gly, the production rate of Gly (Gly PR) was calculated from [1-13C]Gly and [1-14C]Sar infusions using either the primary pools (plasma Gly and HA, respectively) or the secondary pools (HA and plasma Gly, respectively). The results were explained by a liver-plasma Gly exchange model. The whole body Gly irreversible loss, i.e., direct loss from plasma and liver, was calculated from this model to be 832 +/- 58 mumol.kg-1.h-1, showing that the estimation of Gly PR with [1-13C]Gly infusion and plasma Gly enrichment (599 +/- 56 mumol.kg-1.h-1) was a significant underestimate of the true value.(ABSTRACT TRUNCATED AT 250 WORDS)

A molecular biological approach to reducing dietary amino acid needs

Rapid developments in transgenic animal technology make it possible to consider introducing new metabolic capabilities into animals, using genes from other species. Lysine and threonine are both essential amino acids in mammals, and are commonly the first and second limiting amino acids, respectively, for protein accretion in pigs and poultry fed cereal based diets. Here we consider the potential for transgenic animals with microbial biosynthetic pathways for these amino acids.

Quantitative partition of threonine oxidation in pigs: effect of dietary threonine

Kinetic aspects of threonine (Thr) metabolism were examined in eight pigs fed hourly with a diet containing either 0.68% (LT group) or 0.81% (HT group) of Thr (wt/wt), corresponding to 10 and 30% Thr excess, respectively, compared with an "ideal" diet. Primary production (PR) and disposal (DR) rates were obtained for Thr, glycine (Gly), and 2-keto-butyrate (KB) after a 12-h continuous infusion of L-[U-14C]-Thr together with [1-13C]Gly and a 6-h continuous infusion of [1-14C]KB. Transfer of Thr into secondary pools was also monitored, and from these the rates of Thr oxidation through the catabolic pathways of L-Thr 3-dehydrogenase (DR(Thr-Gly)) and threonine dehydratase (DR(Thr-KB)) were estimated. For the LT group the results were (mumol.kg-1.h-1) PR(Thr) 314 +/- 3, PR(Gly) 551 +/- 24, PR(KB) 41 +/- 3, DR(Thr-Gly) 22 +/- 2, and DR(Thr-KB) 7 +/- 1. For the HT group they were PR(Thr) 301 +/- 23, PR(Gly) 598 +/- 55, PR(KB) 39 +/- 4, DR(Thr-Gly) 32 +/- 2, and DR(Thr-KB) 8 +/- 1. The increase in Thr intake (14 mumol.kg-1.h-1, P less than 0.01) induced a commensurate increase in the sum of DR(Thr-Gly) and DR(Thr-KB) (14 mumol.kg-1.h-1, P less than 0.001) when liver was used as the precursor pool. This was mainly due to the increased DR(Thr-Gly) (13 mumol.kg-1.h-1, P less than 0.01); the change in DR(Thr-KB) was not statistically significant. By comparison of intracellular-to-plasma ratios of specific activities (or enrichments) for different tissues with each type of infusion, liver was shown to be the major site of production of Gly and KB from Thr. These data suggest that in fed growing pigs a 30% excess of Thr in the diet does not alter the partition of Thr oxidation, since 80% of Thr oxidation occurs through the L-Thr 3-dehydrogenase pathway for both LT and HT groups.

The effects of indomethacin on gastroduodenal morphology and mucosal pH gradient in the healthy human stomach

To define further the injury and the mechanisms of mucosal injury induced by indomethacin, the effect of 28-day continuous administration of oral indomethacin on gastroduodenal morphology, gastric histology, and the protective mucus-bicarbonate barrier overlying gastroduodenal mucosa in humans was studied. In the studies, indomethacin caused acute gastroduodenal damage in 100% of cases, with maximal damage at 24 hours of administration. With continued intake this damage resolves, although a minority (two study subjects) progressed to discrete ulceration. Why these two subjects failed to adapt is unknown. Biopsy specimens taken during the studies showed no significant changes in inflammatory or regenerative features, and thus failed to shed any light on this process of adaptation to damage. Mucosal pH gradient studies showed a significant increase in juxtamucosal pH at the time of maximal damage (24 hours); this is thought to represent passive diffusion of alkali from damaged mucosa. In conclusion, mucosal adaptation to acute damage by indomethacin occurs in humans. The mechanisms through which the mucosa adapts in this intriguing way remain unknown.