sign in sign up
<<
Home , Mesalazine

Gut 1995; 36: 55-59 55 Selective inhibition of fatty acid oxidation in colonocytes by : a cause of ? Gut: first published as 10.1136/gut.36.1.55 on 1 January 1995. Downloaded from W E W Roediger, S Millard

Abstract '4-16 and in the defunctioned Ibuprofen is associated with initiation or colon low concentrations of n-butyrate result exacerbation of ulcerative colitis. As in diversion colitis'7 which is remedied by ibuprofen selectively inhibited fatty acid re-introduction of luminal short chain fatty oxidation in the liver or caused mito- acids (SCFAs).18 chondrial damage in intestinal cells, its Ibuprofen inhibits SCFA oxidation in effect on substrate oxidation by isolated isolated mitochondria of mouse liver,19 an colonocytes of man and rat was examined. action which in whole animal studies'9 20 was Ibuprofen dose dependently (2.0-7.5 confirmed to be selective for SCFA and long mmol/l) and selectively inhibited 14C02 chain fatty acid oxidation, without affecting production from labelled n-butyrate in glucose oxidation. As NSAIDs may uncouple colonocytes from the proximal and distal oxidative phosphorylation in mitochondria and human colon (n=12, p=<0.001). Glucose reduce adenosine triphosphate (ATP) produc- oxidation was either unaltered or in- tion21-24 it remained unclear whether the effect creased. Because short chain fatty acid of ibuprofen on the colonic mucosa was due to oxidation is the main source ofacetyl-CoA a selective action on fatty acid oxidation or to for long chain fatty acid synthesis, the inhi- generalised mitochondrial damage. Because bition of synthesis by ibuprofen produces colitis' 25-28 independent ibuprofen in the colonic mucosa could also of bacterial activity,29 the aim of the present occur at this level. Because the concentra- study was to establish the action of ibuprofen tions of ibuprofen that can be attained in on fatty acid and glucose oxidation in isolated the human colon are not known, conclu- colonocytes of healthy rat and man. The effect sions drawn from current dosages are of ibuprofen was studied with a view to com- tentative. The inhibition of fatty acid oxi- paring changes observed in healthy colono- dation by ibuprofen may be biochemically cytes with those observations made in implicated in the initiation and exacerba- colonocytes affected ulcerative by colitis.14'6 http://gut.bmj.com/ tion of ulcerative colitis, manifestation of which would depend on the ibuprofen concentrations reached in the colon. Methods (Gut 1995; 36: 55-59) TISSUE COLLECTION AND CELL PREPARATION Keywords: ibuprofen, colitis, colonocytes, fatty acid Colonocytes were prepared from female, oxidation. rats30 between Sprague-Dawley weighing 130 on October 2, 2021 by guest. Protected copyright. and 200 g. Animals were bred in the animal houses of the University of Adelaide, kept on a Non steroidal anti-inflammatory drugs timed cycle, and used in the fed state, as this (NSAIDs) can have deleterious effects on the produced the optimal yield of cells. Animals intestinal mucosa. In the colon, these drugs were killed by stunning/cervical fracture and can produce colitis de novo or cause exacerba- the colon was removed and flushed clear of tion of quiescent ulcerative colitis.'-3 NSAIDs luminal contents with water at 22°C. produce permeability changes in the colonic Human colonocytes were prepared from mucosa independent of colitis,4 or cause unaffected segments of colon obtained from colonic bleeding5 with or without stricturing surgical specimens of patients undergoing and diaphragm formation.6 7 Even though the colectomy for colorectal cancer. The warm inhibition of cyclo-oxygenases and therefore ischaemia time of the colectomy specimens Cell Physiology prostaglandin production has been invoked as was less than three minutes. The colonic Laboratory ofthe a causative factor of damage,' 3 there also was Departnent of specimen washed clear of debris with Surgery, The Queen seems a lack of correlation between reduced water, cut open, and pinned to assist removal Elizabeth Hospital and mucosal synthesis of and the of mucosal strips from the muscularis propria The University of degree of mucosal injury.8 with scissors. Strips were taken at least 10 cm Adelaide, Adelaide, Australia The epithelial barrier of the colonic mucosa from tumour tissue. W E W Roediger in health is maintained by nutrition from the Isolated colonocytes were prepared and S Millard lumen through n-butyrate9-11 derived from tested for viability as previously described.30 Correspondence to: bacterial fermentation of starches. Fatty acid Aliquots of 1 ml of cells were used for experi- Associate Professor W E W oxidation of n-butyrate in colonocytes governs mental procedures. Roediger, Department of Surgery, The Queen the supply of acetyl-CoA for oxidative and Elizabeth Hospital, synthetic functions of epithelial cells, such as Woodville SA 501 1, Australia. lipogenesis,12 mucus synthesis, and processes REAGENTS AND PREPARATION Accepted for publication of detoxification.'3 Inhibition of n-butyrate Sodium n-butyrate and D-glucose were 29 April 1994 oxidation is associated with formation of obtained from BDH (Melbourne, Australia), 56 Roediger, Millard

2.0- isolated cells were reduced to 2.0-7.5 mmol/l as isolated cells are more sensitive to drug dosages a) than whole organs. The luminal concentrations 4- of ibuprofen that can be attained in the human cn colon have not been determined and would

E >. depend, among other things, on absorption in Gut: first published as 10.1136/gut.36.1.55 on 1 January 1995. Downloaded from CO the small intestine and transit time. A dose com- parison was made with 5-ASA even though a

..- 1.0- .-o E 2 0-7.5 mmolIl concentration of ibuprofen may - 0 -r * only be achieved at the mucosal level in the V0 , E upper . 0.= Q-) **: t** CELL INCUBATIONS ~t** Cell suspensions, 1 ml representing 5-14 mg dry weight of epithelial cells, were incubated I | - for 40 minutes in conical flasks equipped with 2-0 5-0 7-5 a glass centre well and stoppered with Suba- Drug concentration (mmol/1) seals (William Freeman, UK). The gas phase Figure 1: 14C02 prod&wction by rat colonocytesfrom was 02 and CO2 (19:1, vol/vol). Incubations [1-14Cl butyrate (AA) and [6-14C] glucose OO) in the were performed at 370C in 1 ml of physio- presence of5-ASA (A.0) or ibuprofen (AO) at doses logical saline containing 20/o-5% (wt/vol) indicated. The mean ('SEM) ofseven experiments each bovine serum albumin, dithiothreitol 1 mmol/l, comprising cellsfrom fiour rats are given. Student's paired t test: *=p<0.005 an d **=p

CALCULATION AND STATISTICAL ANALYSIS TABLE I Acetoacetate and lactate production by isolaited rat colonocytes in the presence of Results of metabolic products were expressed glucose or n-butyrate with 5-ASA or ibuprofen. (Values, mean (SEM); number of colons Rslso eaol rdcswr xrse in brackets) per g dry weight, which was obtained by drying 1 ml of cell suspension to constancy at 90°C Metaboliiteformation (p.mol/minlg/dry wt) and corrected for the dry weight ofthe albumin Metabolite Drug and dose (mmoll) n-butyra,rte(5 mmokll) D-glucose (5 mmol/l) contained in the medium. 14C02 generation Acetoacetate Control 2-72 (0:'29) (6) - from fatty acid glucose was calculated from the Ibuprofen specific activities and trapped 14C02 in NaOH. 2-0 1-93 (0230) (6) - 50 0-61 (0:21) (6)* Observations obtained from the same tissue or 7-5 0-28 (0 10) (5)** - animals on which parallel experiments were Lactate Control 6-33 (058) (6) performed were subject to Student's paired Ibuprofen 2-0 - 6-43 (0.68) (6) t test. The 0.05 probability level was taken to 5.0 _ 500 (054) (6) indicate a difference in observa- 7.5 - 3.00 (0.47) (6)*** significant 5-ASA tostions. 2-0 - 5-38 (0.43) (6) 5.0 _ 5-20 (0.13) (6) 7.5 - 4-52 (0.40) (6)**** Results *Student's paired t test, p<0-01 compared with control accetoacetate;**p<0 001 compared with Initial experiments were undertaken with control acetoacetate; ***p<0.005 compared with control lactate; ****p<0.05 compared with control lactate. isolated colonocytes from the whole colon of Selective inhibition offatty acid oxidation in colonocytes by ibuprofen: a cause ofcolitis? 57

00) 0 'a Gut: first published as 10.1136/gut.36.1.55 on 1 January 1995. Downloaded from

" c 0- '_ ro E 0. Qi Uo=es 'c=1

0.0 ,o20 02 o2h< # 020 #0t0s

Control Ibuprofen 5-ASA Control Ibuprofen 5-ASA Figure 2: 14C02 production by human colonocytes from [6-'4C] glucose in the presence of2-0 and 7 5 mmoNl of5-ASA and ibuprofen in the proximal and distal human colons. Observations are mean (SEM) ofsix colons at each site, with parallel experiments from the same batch ofcells. Student's paired t test: *=p<0. 05 compared with control; **=p<0. 05, 2-0 mmolll ibuprofen compared with 7-5 mmol/l ibuprofen; ***=p<0.025 and ****=p<0. 01 compared with controL

the rat to establish effects of dosage changes a slight reduction in both butyrate and glucose with ibuprofen and 5-ASA. Ibuprofen dose oxidation, a reduction that at 7.5 mmol/l was dependently decreased both n-butyrate and significantly different from that produced by D-glucose oxidation to CO2. Reduction of ibuprofen (p=<0 01 for glucose and 14C02 production from 1.68 (0.12) to 0.36 p=

79% reduction of fatty acid oxidation, while exposure comprising cells from the same batch http://gut.bmj.com/ reduction of glucose at that concentration of of cell isolates. Unlike findings with rat colono- drug was 60% from 0.60 (0.04) to 0.24 cytes, 5-ASA did not reduce either n-butyrate (0.06) ,umol/min/g (dry weight). Ketogenesis or glucose oxidation in the proximal or distal and lactogenesis were reduced in parallel with human colon, but ibuprofen at 7.5 mmol/l the reduction of 14C02 (Table I). 5-ASA led to selectively diminished fatty acid oxidation on October 2, 2021 by guest. Protected copyright.

1.6 I- 1.4 *

L-1.2

c) 1.0 J2. 0.8

0. 0 ~0.4. 0.2

0.0

Control Ibuprofen 5-ASA Control Ibuprofen 5-ASA Figure 3: 14CO2 production by human colonocytesfrom [1-14C] n-butyrate in the presence of2-0 and 7-5 mmol/l of 5-ASA and ibuprofen in the proximal and distal human colon. Observations are mean (SEM) ofsix colons at each site with paralel experimentsfrom the same batch ofcels. Student's paired t test: *=p<0.01 and **=NS compared with control; ***=p<0.001, 2-0 mmolJI ibuprofen compared with 7-5 mmolA; ****=p<0.025 and *****NS compared with control. 58 Roediger, Millard

TABLE II Acetoacetate and lactate production by isolated human colonocytes in the presence ofglucose or n-butyrate with 5-ASA or ibuprofen. (Values, mean (SEM); number ofcolons in brackets) Metabolite formation (pumollminlg/dry wt) Proximal colon Distal colon

Drug and dose n-butyrate D-glucose n-butyrate D-glucose Gut: first published as 10.1136/gut.36.1.55 on 1 January 1995. Downloaded from Metabolite (mmol/l) (5 mmol/l) (5 mmoL/l) (5 mmolJI) (5 mmol/l)

Acetoacetate Control 2-57 (0.48) (6) - 2-17 (0.53) (5) - Ibuprofen 2-0 2.14 (0.47) (6) 1-40 (0.22) (5) - 7-5 0-27 (0.08) (6)* - 0-13 (0.03) (5)* - Lactate Control - 3-12 (0.55) (6) - 3-54 (0.21) (6) Ibuprofen 2-0 - 3-35 (0.58) (6) - 3-70 (03) (6) 7-5 - 2-71 (0.51) (6) - 2-62 (0.24) (6)** 5-ASA 2-0 - 3.07 (0.67) (5) - 3.39 (0.27) (6) 7.5 - 2-69 (0-41) (5) - 2-80 (0.23) (6)** *Student's paired t test, p<0001 compared with acetoacetate control; **p<0.01 compared with lactate control.

(Figs 2 and 3) and ketogenesis (Table II). It is oxidation for lipid and prostaglandin synthesis. noteworthy that ibuprofen stimulated glucose Mucosal damage, therefore, may not always be oxidation, particularly in the distal colon through loss of cytoprotection of reduced (Fig 2). Lactogenesis was significantly reduced prostaglandins but perhaps via metabolic in the presence of 5-ASA (Table II), a feature damage through changes in fatty acid oxida- that was not parallelled by a reduction of tion. A precise relationship between drug 14C02 from the respective substrates. dosages, mucosal injury, impaired fatty acid oxidation, and reduced mucosal prostaglandin synthesis in colonocytes needs to be estab- Discussion lished by further experimentation as was done The detrimental change in fatty acid but not in gastric mucosal injuries.38 glucose oxidation induced by ibuprofen in The reduction of butyrate oxidation by human colonocytes suggests maintenance of 79% by ibuprofen in human colonocytes the Krebs cycle and therefore mitochondrial reflects an overall decrease of 55% of cell function in cells. 5-ASA did not change fatty energy as butyrate normally supplies 70% of acid oxidation of colonocytes in line with cellular oxidation9-1' in colonocytes. Because observations made by Ireland and Jewell.15 both lipogenesis12 and efficient absorption are

The effect of ibuprofen on fatty acid oxidation dependent upon butyrate-oxidation, any http://gut.bmj.com/ in human colonocytes was analogous to that impairment of fatty acid oxidation can explain observed by Freneaux et al 19 with isolated liver changes of membrane permeability4 observed mitochondria and in intact animals. As both with ibuprofen. The mechanism proposed ketone body and CO2 production were equally for ibuprofen damage in the colon is in line reduced in colonocytes, a change within the with the evolution of ulcerative colitis that P-oxidation pathway rather than the Lynen is related to impairment of fatty acid oxida-

(ketone body) pathway or Krebs cycle must be tion.14-6 A reduction in fatty acid oxidation on October 2, 2021 by guest. Protected copyright. involved. Damage to oxidative phosphoryla- suggests a mechanism whereby ibuprofen tion could, however, have occurred in rat can both exacerbate and cause ulcerative colonocytes as both glucose and fatty acid colitis. oxidation were impaired by ibuprofen. Such Ibuprofen reduces the growth of colonic damage usually occurs in experimental animals polyps and tumours.39 40 This effect may at higher doses of drugs.23 depend upon the inhibition of butyrate oxida- The biochemical mechanisms whereby tion because n-butyrate normally promotes NSAIDs cause mucosal damage have received the growth of colonic tumour cells to a more considerable review.3 33 34 Aspirin35 and differentiated form.41 The proposed explana- ibuprofen36 at dosages much lower than those tion would relate to metabolic changes in currently used inhibit cyclo-oxygenase activity mitochondria and cytoplasm of colonocytes which reduces the synthesis of prostaglandins rather than nucleic acids, implying that neo- that are generally considered to be pro-inflam- plastic lesions may be diminished by change in matory. Prostaglandins and 12 lipoxygenase fatty acid oxidation but not eradicated. This activity in ulcerative colitis are increased in is in keeping with the effects of ibuprofen colonic epithelial cells.37 With regard to the noted in the formation of experimental colon mucosal toxicity of NSAIDs, Whittle and cancer.39 Vane33 proposed a two-step damaging process The observations now made with ibuprofen - an initial step of topical irritation possibly on fatty acid oxidation cannot be extended to related to uncoupling of oxidative phos- other NSAIDs until these have been individu- phorylation21-24 followed by a step in which ally investigated. Nevertheless changes in fatty loss of prostaglandins represented a loss of acid oxidation with ibuprofen may explain a cytoprotection and therefore an additive effect number of clinical and experimental observa- to mucosal damage. Present observations with tions with regard to colonic mucosal disease, fatty acid oxidation in human colonocytes especially ulcerative colitis. The dose depen- indicate that very high dosages of ibuprofen dent metabolic changes observed in the limit the supply of acetyl-CoA from fatty acid colonocytes indicate that continued use of Selective inhibition offatty acid oxidation in colonocytes by ibuprofen: a cause ofcolitis? 59

in colonic disease but at a 21 Glanborg-Jorgensen T, Weis-Fough US, Nielsen HH, ibuprofen is possible Olsen HP. Salicylate and -induced uncoupling of lower dosage range. oxidative phosphorylation in mitochondria isolated from the mucosal membrane of the stomach. Scan Jf Clin Invest 1976; 36: 649-53. With sincere thanks to Mrs E Mazel and Mrs S Ireland for word 22 Spenney JG, Bhown M. Effect of acetylsalicylic acid on processing the manuscript. gastric mucosa. II Mucosal ATP and phosphocreatine content and salicylic effects on mitochondrial metabolism. 1 Rampton DS, Sladen GE. Relapse of ulcerative procto- Gastroenterology 1977; 73: 995-9. Gut: first published as 10.1136/gut.36.1.55 on 1 January 1995. Downloaded from colitis during treatment with non-steroidal anti-inflamma- 23 Ohe K, Hayashi K, Shirakawa T, Yamada K, Kawasaki T, tory drugs. Postgrad MedJ 1981; 57: 297-99. Miyoshi A. Aspirin - and taurocholate-induced 2 Gibson GR, Whitacre EB, Ricotti CA. Colitis induced by metabolic damage in mammalian gastric mucosa in vitro. non-steroidal anti-inflammatory drugs. Arch Intern Med AmJ7Physiol 1982; 239: 624-35. 1992; 152: 625-32. 24 Somasundaram S, Machperson AJ, Hayllar J, Saratchandra 3 Bjarmason I, Hayllar J, Macpherson AJ, Russell AS. Side P, Bjarnason I. Enterocyte mitochondrial damage due to effects of non-steroidal anti-inflammatory drugs on the NSAID in the rat. Gut 1992; 33 (suppl): S5. small and large intestine in humans. Gastroenterology 25 Schwartz HA. Lower gastrointestinal side effects of non- 1993; 104: 1832-47. steroidal anti-inflammatory drugs. J Rheumatol 1981; 8: 4 Jenkins AP, Trew DR, Crump BJ, Nukajam WS, Foley JA, 952-4. Menzies IS, et al. Do non-steroidal anti-inflammatory 26 Walt RP, Hawkey CJ, Langman MJS. Colitis associated drugs increase colonic permeability? Gut 1991; 32: 66-9. with non-steroidal anti-inflammatory drugs. BMJ 1984; 5 Holt S, Rigoglioso V, Sidhu M, Irshad M, Howden CW, 288: 238. Maneiro M. Nonsteroidal anti-inflammatory drugs and 27 Charuzi I, Ovnat A, Zirkin H, Peiser J, Sukenik S. lower gastrointestinal bleeding. Dig Dis Sci 1993; 38: Ibuprofen and benign cecal ulcer. Y Rheumatol 1985; 12: 1619-23. 188-9. 6 Huber T, Ruchti C, Halter F. Nonsteroidal anti-inflamma- 28 Ravi S, Keat AC, Keat ECB. Colitis caused by non- tory drug-induced colonic strictures: a case report. steroidal anti-inflammatory drugs. Postgrad Med Jf 1986; Gastroenterology 1991; 100: 1 1 19-22. 62: 773-6. 7 Halter F, Weber B, Huber T, Eigenmann F, Frey MP, 29 Melarange R, Moore G, Blower PR, Coates ME, Ward FW, Ruchti C. Diaphragm disease of the ascending colon. Ronaasen V. A comparison of indomethacin with Association with sustained release . J Clin ibuprofen on gastrointestinal mucosal integrity in conven- Gastroenterol 1993; 16: 74-80. tional and germ-free rats. Aliment Pharmacol Ther 1992; 6: 8 Whittle BJR. Temporal relationship between cyclo- 67-77. oxygenase inhibition as measured by bio- 30 Roediger WEW, Truelove SC. Method of preparing synthesis and gastrointestinal damage induced by isolated colonic epithelial cells (colonocytes) for metabolic indomethacin in the rat. Gastroenterology 1981; 80: studies. Gut 1979; 20: 484-8. 94-8. 31 Staerk Laursen L, Stokholm M, Bukhave K, Rask-Madsen 9 Roediger WEW. The role of anaerobic bacteria in the meta- J, Lauritsen K. Disposition of 5-aminosalicylic acid by bolic welfare of the colonic mucosa in man. Gut 1980; 21: and three mesalazine preparations in patients 793-8. with ulcerative colitis: comparison of intra-luminal 10 Ardawi MSM, Newsholme EA. Fuel utilization in colono- colonic concentrations, serum values and urinary cytes of the rat. BiochemJ_ 1985; 231: 713-9. excretion. Gut 1990; 31: 1271-6. 11 Fleming SE, Fitch MD, Devries S, Liu ML, Kight C. 32 Bergmeyer HU, Bergmeyer J, Grass M. Methods of Nutrient utilization by cells isolated from rat jejunum, enzymatic analysis. 3rd ed. Weinheim: Verlag Chemie cecum and colon.JNutr 1991; 121: 869-78. 1984, vol VI, 582-8, vol VIII, 61-9. 12 Roediger WEW, Kapaniris 0, Millard S. Lipogenesis from 33 Whittle BJR, Vane JR. A biochemical basis for the gastro- n-butyrate in colonocytes. Action of reducing agent and intestinal toxicity of non-steroid antirheumatoid drugs. 5-aminosalicylic acid with relevance to ulcerative colitis. Arch Toxicol 1984; 7 (suppl): 315-22. Moll Cell Biochem 1992; 118: 113-8. 34 Williams KM, Day RO, Breit SN. Biochemical actions 13 Ramakrishna BS, Roberts-Thomson IC, Pannall PR, and clinical pharmacology of anti-inflammatory drugs. Roediger WEW. Impaired sulphation of phenol by the Adv Drug Res 1993; 24: 121-98. colonic mucosa in quiescent and active ulcerative colitis. 35 Vane JR. Inhibition of prostaglandin synthesis as a Gut 1991; 32: 46-9. mechanism of action of aspirin-like drugs. Nature (New 14 Roediger WEW. The colonic epithelium in ulcerative colitis 231: 232-35. Biol) 1971; http://gut.bmj.com/ - an energy deficiency disease? Lancet 1980; ii: 712-5. 36 Rome LH, Lands WEM. Structural requirements for time- 15 Ireland A, Jewell DP. 5-aminosalicylic acid (5-ASA) has no dependent inhibition of prostaglandin biosynthesis by effect on butyrate metabolism in human colonic epithelial anti-inflammatory drugs. Proc Nat Acad Sci USA 1975; cells. Gastroenterology 1990; 98: A176. 75: 4863-5. 16 Chapman MAS, Grahn ME, Hutton M, Rogers J, Williams 37 Shannon VR, Stenson WF, Holtzmann NJ. Induction of NS. Failure of colonic mucosa to oxidise butyrate in epithelial arachidonate 12-lipoxygenase at active sites of ulcerative colitis. Gut 1991; 33: 540 (abstract). inflammatory bowel disease. Am J Physiol 1993; 264: 17 Glotzer DJ, Glick ME, Goldman H. and colitis G104-11. following diversion of the fecal stream. Gastroenterology 38 Levine RA, Petokas S, Nandi J, Enthoven D. Effects ofnon- 1981; 80: 438-41. steroidal anti-inflammatory drugs on gastrointestinal 18 Harig JM, Soergel KH, Komorowski RA, Wood CM. injury and generation in healthy volunteers. on October 2, 2021 by guest. Protected copyright. Treatment of diversion colitis with short-chain-fatty acid Dig Dis Sci 1988; 33: 660-6. irrigation. New Engl Jf Med 1989; 320: 23-8. 39 Reddy BS, Tokumo K, Kulkami N, Aligia C, Keloff G. 19 Freneaux E, Fromenty B, Berson A, Labbe G, DeGott C, Inhibition of colon carcinogenesis by prostaglandin Letteron P, et al. Stereoselective and non-stereoselective synthesis inhibitors and related compounds. effects of ibuprofen enantiomers on mitochondrial Carcinogenesis 1992; 13: 1019-23. ,-oxidation of fatty acids. J Pharmacol Exp Therap 1990; 40 Boone CW, Steele VE, Kelloff GJ. Screening for chemo- 255: 529-35. preventative (anticarcinogenic) compounds in rodents. 20 Zhao B, Geisslinger G, Hall I, Day RO, Williams KM. The Mutat Res 1992; 267: 251-5. effect of the enantiomers of ibuprofen and on 41 Whitehead RH, Young GP, Bhathal PS. Effect of short the 3-oxidation of palmitate in the rat. Chirality 1991; 4: chain fatty acids on a new human colon carcinoma cell 137-41. line (LIM 1215). Gut 1986; 27: 1457-61.