Nutrición Hospitalaria

ISSN: 0212-1611

info@nutriciónhospitalaria.com

Grupo Aula Médica

España

Singh Kanwar, S.; Vaish, V.; Nath Sanyal, S.

Altered membrane lipid dynamics and chemoprevention by non-steroidal anti inflammatory drugs

during colon carcinogenesis

Nutrición Hospitalaria, vol. 26, núm. 5, septiembre-octubre, 2011, pp. 1141-1154

Grupo Aula Médica

Madrid, España

Available in: http://www.redalyc.org/articulo.oa?id=309228876034

How to cite

Complete issue

More information about this article

Journal's homepage in redalyc.org

Scientific Information System

Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal

Non-profit academic project, developed under the open access initiative

Nutr Hosp. 2011:26(5):1141-1154ISSN 0212-1611 • CODEN NUHOEQ

S.V.R.318

Original

Altered membrane lipid dynamics and chemoprevention by non-steroidalanti inflammatory drugs during colon carcinogenesisS. Singh Kanwar', V.Vaish' and S. Nath Sanyal-

'Prcscntly at School ofMedicine. Department ofInternal Medicine. Wayne State University. Detroit. Michigan. USA.-Deportment ofBiophysics. Panjab University. Chandigarh. India.

Abstract

The present work focuses on the anti-neoplastic role ofnon steroidal anti-inflammatory drugs (NSAIDs) inmodulating the biophysical parameters of the colonicmembranes in 1,2-dimethylhydrazine dihydrochloride(DMH) induced carcinogenesis. The steady-state fluores­cence polarization technique was applied to assessmembrane fluidity, membrane polarity and Iipid phasestates. The decline in cholesterol content, biosynthesisand cholesterol: phospholipids ratio with DMH treat­ment indicates more fluidity associated with carcinogen­esis. The DMH group had shown lower order parameterindicating more fluidity whereas NSAIDs resulted inincreasing the membrane Iipid order. The convergingeffects of these changes were more in membrane phaseseparations and membrane phase state. In DMH treat­ment membrane shows lesser phase separation or highpolarity, and more Iiquid crystalline state while forNSAID groups membranes ha ve higher phase separa­tions or low polarity, and more ofthe gel phase. Further,NSAIDs induced anti-proliferative effects were evidentlyobserved by apoptosis in the colonocytes by using acri­dine orange-ethidium bromide fluorescent staining andTerminal de-oxynucleotidyl transferase dUTP nick endlabeling (TUNEL) assay. The results suggest that NSAIDsinduced alteration in the membrane biophysical parame­ters may be an important initiating event for the chemo­preventive action.

(Nutr Hosp. 2011;26:1141-1154)

DO 1:1O.3305/nh.2011.26.5.5295Key words: Lipid phase separations. Membrane polarity.

Membranejluidity. Apoptosis. NSAIDs.

Correspondence: Sankar Nath Sanya!.Proíessor.Department ofBiophysics.Panjab University.160 014 Chandigarh. India.E-mail: sanyalpu@gmai!.com

Recibido: 2-V-20l1.Aceptado: 9-V-20l1.

ALTERACIÓN DE LA DINÁMICA DE LOS LÍPIDOSDE MEMBRANA Y QUIMIOPREVENCIÓN

MEDIANTE LOS FÁRMACOSANTIINFLAMATORIOS NO ESTEROIDEOS EN LA

CARCINOGÉNESIS DE COLON

Resumen

Este trabajo se centra en el papel antineoplásico delos fármacos antiinflamatorios no esteroideos (AINE)en la modulación de los parámetros biofísicos de lasmembranas colónicas en la carcinogénesis inducida por1,2-dihidrocloruro de dimetilhidracina (DMH). Se aplicóla técnica de polarización de la fluorescencia en estado deequilibrio para evaluar la fluidez de la membrana, supolaridad y los estados de fase Iipídica. El declive delcontenido de colesterol, la biosíntesis y el cociente coles­terol: fosfolípidos con el tratamiento con DMH indicamás fluidez asociada con la carcinogénesis. El grupoDMH mostraba un parámetro de menor orden, lo queindica más fluidez, mientras que los AINE produjeron unaumento del orden de lípidos de membrana. Los efectosconvergentes de estos cambios fueron más notables en lasseparaciones de la fase de membrana y en el estado de fasede membrana. Con el tratamiento con DMH, la membranamuestra menor separación de fase o polaridad elevada, yun estado cristalino más líquido o polaridad elevadamientras que los grupos de AINE tienen mayores separa­ciones de membrana o polaridad baja y más fase enestado gel. Además, los efectos antiproliferativos indu­cidos por los AINE se observaron de forma evidente utili­zando tinción fluorescente con naranja de acridina­bromuro de etidio y el ensayo de marcado final de ladUTP transferasa desoxinucleotidil terminal (TUNEL).Los resultados sugieren que los AINE inducían una alte­ración de los parámetros biofísicos de la membrana, locual podría ser un acontecimiento inicial importante parala acción quimiopreventiva.

(Nutr Hosp. 2011;26:1141-1154)

DO 1:1O.3305/nh.2011.26.5.5295Palabras clave: Separaciones en fase lipidica. Polaridad de

la membrana. Fluidez de la membrana. Apoptosis. AINE.

1141

Introduction

In recent years. modifications of physical characteris­tics of the memb rane lipid bilayer. such as the fluid itychanges alon g with mem brane phase state and lipidphase separations, have proved important to clarifybiological mechanisms and elucidate a variety of patho­logical conditions. such as, chronic inflammatory condi­tions and carcinogenesis.' Many pathological casesincluding cancer are manifested in abnormal lipidcomposition of cell memb ranes which are perhaps theleast understood phenornena.' Indeed, a relationship hasbeen suggested between the abilityof the drugs to changemembrane fluidity and their therapeutic acti on.' Nonsteroidal anti-inflammatory drugs (NSAIDs) are the mostimportant agents used in the treatment of inflammation,pain and fever and currently being seen as one of the besttherapeutics for colon cancer chemoprevention.' Likeother drugs , the therapeutic effects of NSAIDs arestrongly influenced by their relative lipid affinity whichcan have several consequences. including an inducedperturbation of the physicochemical state of themembrane. ' Earlier, we have presented the evidence ofpathogenesis of colon cancer and chemoprevention withNSAID s along with the relationship of membrane lipidcompo sition, fluid ity and antioxidant defen se in themorph ogene sis of chemically induced neoplasm of ratintestine.??AIso it is shown that during colon carcinogen­esis the NSAIDs induced chemoprevention is mediatedthrough the alteration of membrane bound sodium­proton exchanger and low intracellualr pH, followed bymitochondrial dysfunction ing by Ca" ATPase inhibitonand an enhanced intracellular Ca'rflux which may furtherlead to activate a strong apoptotic signal."

In the present study the chemopreventive actions oftbree NSAIDs were analyzed for lipid composition andfunctions of the colonic plasma membranes as a possibleinitiating event for the anti-proliferative role . The threeNSAIDs chosen were Aspirin, Celecoxib and Etoricoxibfor their inhibitory action against the two isoforms ofCycloxygenase (COX ), COX-I and COX -2, withdifferent selectivity. Aspirin, a non selective COXinhibitor, is a benchma rk NSAID and represents the classof traditional-NSAIDs (tNSAIDs) whereas Celecox iband Etoricoxib are from "coxib" family representing thesecond-generation NSAID s. Although. both the coxibsselectively inhibit COX-2, Etoricoxib has the highestselectivity ratio of any cox ib for inhibition of COX-2relative to COX-l." These newly developed coxibs areconsidered to have tremendous potential in cancerchemopreventi on as demon strated in epidemiologicalstudies, animal experimentation. cell culture and in clin­ical studies." Althou gh evidences strongly suggest thatNSAIDs induce their anti-proliferative effect by trig­gering programm ed ce ll death (apoptosis)" whichprimarily follow s through the inhibition of the COX-2pathway," the molecular mechanism of such inhibition isfar from clea r. Therefore . in order to have an under­standing of the outcomes of altered membrane lipid

components and phase changes. the status of apoptosis inthe colonocytes was looked for the possible pathway ofcancer preventing mechanism by the NSAIDs.

Materia l and methods

Animal care and drug treatm ent

Male Sprague-Dawley rats weighing 140-1S0 gwere procured from the Central Animal House ofPanjab University and hous ed in polypropylene cagesunder hygienic conditions, ambient light and tempera­ture (18 -22'C). They were fed with pellet diet anddrinking water ad libitum and assorted into thefollowing groups:

Group 1- Control (vehic le treated);Group 2- DMH treated ;Group 3- DMH + Aspirin treated;Group 4- DMH + Celecoxib treated : and

- Group S- DMH + Etoricoxib treated.

Animals in the groups 2 to S were given DMHweekly at a dose of 30 mg/k g body weight subcuta­neously (s.c .) for 6 weeks . In addition. the animals ingro up 3 to S received daily dose of the re spectiveNSAIDs (v iz., Aspirin- 60 mg, Celecoxib- 6 mg andEtoricoxib- 0.6 mg) orally per kg body weight.Animals were acclimati zed for a peri od of one weekprior to the experimental work .

1,2-dimethylhydrazine dih ydrochloride (DMH)(Sigma, USA) was prepared fresh immediately beforeinjection in 1 mM EDTA-saline. pll being adjusted to7.0 using 100 mM NaOH. Aspirin, Celecoxib andEtoricoxib were generous gift from Ranbaxy ResearchLab (Gurgaon, Indi a). Their reported anti -inflamma­tory doses were prepared fresh every day in O.S %sodium carboxymethyl cellulose (CMC). The vehicletreated animals were admini stered with O.S% CMCeveryday per oral and lmM EDTA saline solution oncewee kly by s.c. injection. The bodywei ght of the ratswas recorded weekly till the termination of the experi­ments. They were sacrificed in overnight fasted condi­tions 6 weeks after beginning of the study by keepingthem under an overdo se of ether anesthes ia and sacri­ficed aro und 8 AM to avoid diurnal variatio n in theparameters . Animal pro cedures were approv ed byPanjab Universit y Ethical Conunittee on ExperimentalAnimals for Biomedical Research . The guidelineswere prepared according to th e principies of animalcare laid down by the Nationallnstitute of Health , USA(NlH publication No. 23-8S. revised in 1985).

Isolation 01colonic brush border membrane (BBM)

The BBM of rat colon was isolated using the method ofBrasitus and Keresztes15 and had been variously described

1142 Nutr Hosp , 2011;26(5) :1141-1154 S. Singh Kanwar el al.

in earlier publications.'" The final membrane soobtained was analyzed for the membrane specificenzyme, cysteine sensitive alkaline phosphatase.'

lsolation ofcolonocytes

Colonocytes were obtained from the freshlydissected colons by the method of Mouille et al. 16 Theintegrity ofthe colonocytes was assessed by the abilityof the cells to exclude vital dyes. The number ofcolonocytes re-suspended in DMEM buffered withMOPS (pH 7.5, 25 mM) were counted on a hemocy­tometer and it was determined that the isolation proce­dure led to the recovery of the viable colonocytes, tothe extent of at least 97%. In order to assess the effi­ciency of colonic epithelial cell retrieval, pieces ofmucosa were fixed in 10% formalin saline both beforeand after cell separation. The tissue was processed forparaffin embedding and then the sections stained withhematoxylin and eosin.

Quantitative analysis ofLipids

Extraction of Lipids from the colonic mucosa wasdone by the method of Folch et al and as described vari­ously earlier." IR The lipid extracts obtained were storedat -20'C and used as such for various lipid estimations.The pooled upper aqueous phase (also called the Folchupper phase) was used for ganglioside estimation. Totallipids were estimated by the method of Fringes andDunn, phospholipid by the method of Ames, cholesterolby the method of Zlatkis et al and ganglioside-sialic acidby the method of Warren, respectively .17 IR

Dynamics oflipid metabolism

14C incorporation ofPalmitic acid-1-C-14 (BhabhaAtomic Research Centre, Mumbai, India) in themembrane lipids was studied by incubating the colonictissues in the radio-labelled substrate and then isolatingthe lipid by thin layer chromatography and quantitatingby liquid scintillation spectrometry following themethod of Butler and as described earlier."

Liposome preparation

Liposomes were prepared from the extracted lipids bythe method described earlier." A known volume of lipidextract (in chloroform: methanol) was evaporated todryness at room temp and suspended in sodium maleatebuffer to a final canco of approx 2.5 mg/rnl. The mixturewas sonicated for 1 min with brief pause in between, 2-3times using a probe type sonicator (Sonics, USA) in anice cold environment. The suspensions were centrifugedfor 10 min at 50,000 g in a Beckman Coulter ultracen-

trifuge. The supernatant liposomes were tested with thevarious fluorescent probes.

Fatty acid analysis

Fatty acid methyl esters were prepared as describedearlier. IR and quantitated in a gas liquid chromatog­raphy (GLC) set up. Appropriate standards were alsorun simultaneously. Conditions for GLC were such thatthe injection temp was set at 230'C, column temp at220'C, detector temp at 240'C, and the carrier gas wasNitrogen. Flame ionizing detector was used and thecolumn was of FFAP from Machery Nagel GmbH,Duren, Germany. Retention times of different fattyacids were recorded and the results expressed in termofweight%.

Evaluation oflateral phase separationby N-NBD-PEfiuorescence quenching

Membrane lipid phase separations were monitoredin the liposomes, BBM and colonocytes by the methodas described by Hoekstra" using the self-quenching ofN-NBD-PE fluorescence. The method is based on theself quenching of the NBD fluorophore that occurswhen the local concentration of NBD lipid in thebilayer increases during the segregation of themembrane lipids into the discrete domains in the planeof the bilayer. 2ml vol containing approx 150 ug ofBBM/O.25 mg liposomes or colonocytes (2 x l O-cells)were labeled with N-NBD-PE (final concentration 0.6flM) (Sigma, USA) by incubating the tubes in dark for15 min at 37'C. The samples of colonocytes and BBMwere loaded with N-NBD-PE using the same buffer aswas used for their original isolation procedure while incase of liposomes, the small unilamellar vesicles weregenerated in 0.1 M NaCl 10.01 M Hepes, pH 7.4, byultrasonication (Sonics, USA) in cold conditions andultracentrifugation (at 150,000 g for 1 h). Fluorescenceassay was done by measuring the fluorescence intensi­ties at A"= 475 nm and A,m= 530 nm before and after thesolubilization of vesicles or cells as described byNicholas & Pagano." BBM vesicles/liposomes orcolonocytes were solubilized by the addition of TritonX-lOO (100 ul). The percentage of quenched fluores­cence (% Q) was calculated from the relative fluores­cence of unquenched N-NBD-PE (Fu) in the presenceof detergent and quenched molecules (FQ) in vesiclesby the relation:

Evaluation ofmembrane phase state

Membrane phase determination was done to evaluatethe membrane polarity in terms of generalized polariza-

NSAIDs and membrane lipid dynamicsin colon cancer

Nutr Hosp, 2011;26(5):1141-1154 1143

tion (GP) using fluorescence probe, Laurdan (6-dode­ca noyl-2 -d imethy laminonaphta lene) fo11owing themeth od as described by Amb ro sini et al." Laurd an isparticul arl y sens itive to the phase state of themembrane . The basis of its spectral sensitiv ities lies inits ability to sense the polarity change of the memb raneand is therefore. a powerful technique to detectph ysico-chemi cal modifi ca tions . The enviro nme nt­sensi tive spectroscopic prop ertie s of Laurd an havebe en compreh en sivel y described by a parametertermed generalized polarization (GP), originally intro­duced by Gratton and co-workers." GP values fromexcitation and emission spec tra are characteristic of themembrane phase stat e. In order to quantify the GPparameter in biolog ical membra nes the relative valuesof exGp340and exGp40'" (measured on exGP at 340 and400nm) and emGp44°and cmGP'')() (measured on emGPat 440 and 490 nm) are considered important.

Calculation of genera lized polarization(GP) parameter

150 flg of BB M/ 0.25 mg liposomesl2 x IO- ccllpopulati on of co lo nocytes were equilibr ated for 20min at 3TC in the dark to be labeled with Laurdan(stock prepared in metha nol) (Sig ma, USA) at lA flMfina l con c . Th e buffers for BBM, lip osom es andco lonocytes were ke pt thc sa me as that of thei r or ig­inal preparation s for labcling the probe . Steady-stateLaurdan excit ation spec tra were obtained in the rangeof 300 to 420 nm , using bot h 440 and 490 emi ssionwaveleng ths . Simi lar ly , laurdan em ission spectrawere recorded in the range from 420 to 550 nm, usingboth 340 and 410 nm exci tatio n wavelengths . Blankspectra were obt ain ed with unlabe11ed samples andsubtracted from th e spec tra of the labeled samples.Readings were mad e at a fixed temp of 3TC for a11 thethree sample syst ems . From the spectroscopic data,Laurdan excitation generalized polarization (exGP)spectra were derived by ca lculating the GP value foreach excitation wavelength using Eq. (l): exGP = (1440- 14')()/(1440+ 1

4')() where 1440and 14')() are the intensities

measured, at each exci tation wavelength (from 320 to4 20 nm), on the flu orescence excita tion spectraobta ined by fixed emiss ion waveleng th of 440 and490 nrn , respecti vel y. Laurd an emission genera lize dpolarizati on (e mGP) spectra were derived by ca lcu­lating the GP va lue for eac h emission waveleng thusing Eq . (2) : emGP =(1410 - 1

340) / (1410+ 1

340) where 1410

and 1340

are the intens ities measured at each emi ssionwaveleng th, from 42 0 to 520 nm . These va lues areobta ined by the fluorescence emiss ion spectra record­ings using fixed exc itatio n wavele ngth of 41Oand 340nm, respectively. The choice of 4 10,340, 440 and 490nm for GP calculation was based on the charac teristicexcit ation and emiss ion waveleng ths of pure ge l andliquid-crystall ine lipid phases, according to Parasassiet al ."

Fluidity and orderparameter assay using DPH

The flu ore scent probe, 1, 6- diphenyl-I , 3, 5-hexa­trien e (DPH) (Sigma. USA) was used in the fluiditystudies (ro tational diffu sion ). A stock solution of 1 mMprob e in tetr ah ydrofuran (T HF) was prep ared andstored in dark colored bottle at room temp. A small vol(10 pl) of DPH solution in THF was injected with rapidstirring into 1,000 vol of Tr is/HCl buffer (10 mi) atroom temp . The suspension was stirred for at least 2 hafter which little or no odor of TI-IFwas detected andthe suspension showed neglig iblc fl uorescence . TheBBM were incubated in 3 mi of the aboye suspensioncon taining 3 fll DPH for 2-4 hr at 3TC. Thereafter, esti­mati ons of fluorescence intensity (F), fluorescencepolarization (P), and fluorescen ce anisotropy (r) weremade with an excitation wavelength of 365 nm andemissio n wavelength of 430 nm using a Perkin Elme rLu minescence Spectrometer LS-5 5 (Beaco nsf ield,United Kingdom). Order paramctcr (S) was calculatedus ing the formula. S' = (4/3r - 0.1)/rJ25] taking rova lue for DPH as 0.362 and flu idity was measured interm of reciprocal of flu orescen ce po lariza tion (l/P )values and as described by us earlier.">

Assessment ofapop tosis in colonocytes

Acridine orange-ethidium bromid c co-staining

Acr idine ora nge is a fluorescent age nt. which canintercalate (slip in) between base pairs (bp) in the centralstack of DNA heli x. Thi s staining procedure wasperformed according to the details provided by Baker etal." Briefly, the cells were suspended in PBS (pH 704,I-liMedia, Mumbai) containin g acridine orange (1 ug/ml)and RNaseA (Sigma, USA) as well as ethidium bromide(Sigma, USA) in the same concentration . The cells werewashed and examined under fluorcscence microscope(x400) (Axioplan, Zeiss, Germany). For quantification ofapoptotic cells, a total of 100 cells from four differentslides were observed and percentages of apoptotic cellscalculated for the individual animal.

Terminal de-oxynucl eotidyl transferasedUTP nick end labelin g (TU NEL) assay

TUNEL is a spec ific meth od for detecting DN Afragmentation (damage) that results from the apoptoticsignaling casc ades. The assay relies on the presence ofnicks in the DNA which ca n be identifi ed by terminaldeoxynucleotidyl transferase . an enzy me that willcatalyze the additio n of dUTP s that are seco ndarilylabeled with a marker. TUNEL assay was performedaccording to the prot ocol providcd by Trevigen (usingTACS-XL® in situ apoptosis detection kit from Trevigen(USA) which is based on the incorporation of bromod­eoxyuridine (BrdU) at the 3' OH ends of the DNA frag-

1144 Nutr Hosp, 2011;26(5):1141 -1154 S. Singh Kanwaret al.

Table 1Effect of different treatments on the contents of total lipid (TL), cholesterol (Ch), phospholipids (PL), cholesterol:

phospholipid ratio (Ch:PL) and ganglioside sialic acid (GSA) in the lipids extractedfrom colonic mucosa

TL Ch PI Ch/PI GSA(mg/gtissue) (mmole/g tissue¡ tmmole/g tissue¡ (mole/mole) tnmoles/g tissuewt)

Control 16.93± 1.55 11.724±0.171 2.603 ±0.049 4.504±0.078 3.570±0.108

DMH 15.17±0.45 7.068± 0.564' 6.721 ± 0.393' 1.057± 0.145' 5.870± 0.206'

DMH+Aspirin 18.58±2.23Y 7.529± 0.359' 4.619± O.13"Y 1.629± 0.040"Y 5.627± 0.107'

DMH+Celecoxib 19.33±2.14Y 8.010± 0.517' 3.747± 0.19"' 2.172± 0.254"Y 5.256± 0.090"'

DMH+Etoricoxib 15.87± 1.42 12.701 ± 1.294' 4.091± 0.13"' 3.100±0.243"' 5.267± 0.152"'

Thevaluesaremean± S.D.of8 animals. Significantly different from control at'P" 0.001 andDMHat'p" 0.01,'P" 0.001 byonewayANOVA.

ments that are formed during apoptosis). For quantifica­tion of apoptotic cells, a total of 100 cells from fourdifferent slides were observed and percentage of apop­totic cells calculated for individual animal.

Statistical analysis

Statistical examination ofthe data was performed bythe analysis ofvariances (one way ANO VA) followingPost- Hoc test using the least significant difference(LSD) method with the help of SPSS v 14 computersoftware (SPSS Inc. Chicago, IL, USA).

Results

Quantitative analysis oflipids

Table 1depicts the profiles of the totallipid, choles­terol, phospholipid, cholesterol: phospholipid (Chol:PL)and ganglioside sialic acid (GSA) content present in thecolonic mucosa. It showed totallipid and cholesterolvalue to decrease in case of DMH treatment which wasrecovered in the NSAIDs groups. The phospholipidcontent was found to be increased with DMH treatmentand similarly corrected in the NSAIDs groups. Amongthe NSAIDs tested, the values in the Etoricoxib + DMHreached close to the control. Chol:PL ratio was found tobe reduced remarkably with the DMH administration,however, in comparison to DMH group a significantimprovement was observed with the NSAIDs treatmentand the maximum recovery of the Chol:PL ratio wasobserved in DMH + Etoricoxib. DMH group has alsoresulted in an increase in the PL which was lowered in theDMH + NSAID groups. Levels of GSA were found to beelevated in the DMH group as compared to the controlwhereas, DMH+NSAIDs treatment caused the substan­tial decrease in comparison to the DMH only group.

Dynamics oflipid metaholism

Results of membrane lipid and phospholipids biosyn­thesis studies using 1-14C-labelled palmitic acid incorpo-

ration are shown in the table Il. A level of the radioac­tivity incorporated into the cholesterol was found to bedecreased with DMH as compared to the control. Co­administration of Aspirin and Celecoxib proved helpfulin recovering the cholesterol biosynthesis in comparisonto DMH while DMH + Etoricoxib group recorded thehighest increase in the same. Free fatty acids (FFA) wereobserved to be unaltered in the DMH while in NSAIDstreated groups, FFA was significantly low in comparisonto DMH as well as the control group. Triglycerides (TG)registered a significant increase in DMH group ascompared to the controls while NSAIDs treatmentreduced its biosynthesis when compared to the DMHgroup. The lower level of triglycerides was recorded inthe DMH + Celecoxib and DMH + Etoricoxib groups.Similarly, the fraction corresponding to the monoglyc­erides + diglycerides was observed to be lower in theDMH + NSAID groups when compared to the DMHalone. The DMH alone registered a slight increase overthe control. Cholesteryl ester was observed to be unal­tered in DMH group as compared to the control whilehigher levels were recorded in the DMH + NSAIDgroups. S, PC and PE were observed to be enhanced inthe DMH group whereas with NSAIDs these fractionsrecorded lower concentrations in comparison to DMHgroup. Fraction corresponding to PI + PS content showedan increment in the DMH + NSAIDs groups with respecttotheDMH.

Fatty acid analysis

Table Ill shows the quantitative values (wt %) asderived from the elution profile of various fractions (wt%) offatty acid methyl ester derivatives. Carbon 16 fattyacids, palmitic acid (16:0) and palmitoleic acid (16:1)contents were observed to decrease in DMH group whileNSAID showed an increase. Increment of the carbon 18fatty acids, stearic acid (18:0) increased in DMH groupwhile NSAID caused the decrease. Higher unsaturatedfatty acids such as the linoleic acid (18:2) and arachi­donic acid (20:4) showed an increase in the DMH groupas expected, which declined subsequently in theNSAIDs groups, Etoricoxib in particular. Overall results

NSAIDs and membrane lipid dynamicsin colon cancer

Nutr Hosp, 2011;26(5):1141-1154 1145

Table 11(a) Ejfect ofdifferent treatments on the biosynthesis ofmembrane neutral lipid fructions: freefatty acids, cholesterol,triglycerides, mono- + di-glycerides and cholesteryl ester as measured by "Cslobellcd palmitic acid incorporution.

(b) Ejfect ofdifferent treatments on the biosynthesis ofmembrane phospholipid fractions: spliyngomyelin (S),

phosphatidylcholine (PC), phosphatidvlethanolamine (PE) and phosphatidylinositol (PI) + phosphatidylserine (PS)as measured by "Cslobellcd palmitic acid incorporation

(a)Biosyntllesis oimembrane nelltrallipids./l'GC,tion (b)Biosynthesis ot'membrane pIIOSpllOlipidsfi'act/ontumole (~fradi()activit}¡g tissue lVt) tumo!« (~fradi()activit}¡g tissue lVt)

PllOsphatidJI-

CllOlesterolFree

TriglyceridesMOlwglycerides+ CllOlestervl

Sph}'llgomyelillPllOsphaidJI- PllOsphatidvl- inositol+

Fattdcids Diglycerides Ester choline ethG/wlamilie PllOsphatidJI-serine

Control 46.94±2412 46.84± 1.87 61.286± 3.22 6148±5.1O 10.52 ±0.53 1O.26±0.31 10.53 ±0.68 23.99 ±0.33 10.25 ±0.52DMH 23.52± 141' 47.32±2.32 83.91 ±4.62' 69.81 ±2.87 10.56 ±0.94 13.51 ±1.04c 12.01 ±0.12' 40.91 ±1.51' 942±0.01DMH +Aspirin 31.21 ±1.53'" 27.14±0.98'" 70.64 ±3.99'" 55.72 ±2.05" 1842 ±0.84'" 9.70± 0.52 9.95 ±0.27' 13.18 ±0.72'" 10.51 ±045'DMH +Celecoxib 39.61 ±2.97'·' 22.53± 1.15'" 54.90 ±4.34' 48.27 ±3.26'" 14.89±0.59'·' 949±0.97' 9.11 ±0.29'·' 13.82± 1.35'" 13.21 ±0.91'"DMH +Etoricoxib 51.62 ±1.76" 3745 ±3.15'" 56.44 ±4.17' 44.02±3.31'" 16.61 ±0.68'" 8.73 ±0.57' 7.50±0.39'·' 13.39± 0.78'" 13.91 ±o.n'

Thevalues aremean ±S.D. of8 animals. Significantly differenl from control al"p" 0.05, "p" 0.01, 'p" 0.001 andDMH al'p" 0.05, 'p" 0.001 byoneway ANOVA.

Table 111Quantitative anaiysis ojfatty acid (weight %) as obtained from the elution profilesjrom GLC ojfatty acid

methyl ester derivatives

Myristie acid Palmitie acid Palmitoleieacid Stearie acid Oleieacid Linoleie Araehidonie acid(14.0) (16.0) (16.1) (18.0) (18.1) (18.2) (20.4)

Control 2.0±0.05 32.1 ±0.038 7.2±0.28 7.3±0.04 37.2 ±1.9 11.9±0.531 2.3±0.075

DMH 1.8±0.02'" 28.8 ±0.04I' 4.4±0.16' 9.2±0.13' 39.1 ±2.3 12.9±0.332 3.7±0.l3'

DMH +Aspirin 1.6±0.02'" 29.5 ±0.027'" 7.5±0.D9' 7.3±0.098' 39.0 ±1.6 12.5 ±0.84 2.6±0.05'"

DMH +Celecoxib 1.6±0.02'" 32.1 ±0.08' 8.1±0.24'" 7.0±0.92' 41.8 ±1.38' 8.3±0.46'" l.l ±0.08'"

DMH +Eloricoxib 1.7 ±0.013'" 33.0±0.51'" 5.70 ±0.67'" 8.1±0.51' 37.9 ±1.46 12.1 ±0.48 1.5±0.07'"

Thevalues aremean ±S.D. of8 animals. Significantly differenl from control al"p" 0.01, 'p" 0.001 andDMH al'p" 0.01, 'p" 0.001 byoneway ANOVA.

showed no significant change in the amount of freeunsaturated fatty acids in the colonic mucosal lipidswhile saturated fatty acids were found to be lowered inthe DMH group and increased in DMH + Etoricoxib(table IV). Percentage with respect to control valuescalculated for arachidonic and palmitic acid content isshown in the table V. Arachidonic acid (AA) was foundto be significantly increased in the DMH group, (60%with respect to control) and significantly reduced in the

Table IVSaturation (wt%) and unsaturation (wt%) calculatedfrom the quantitative data obtained from the elution

profiles ojfatty acid methyl ester derivativesjrom GLC

NSAID treated groups. Content of palmitic acid wasfound to be lower for DMH (89%) while in DMH +NSAIDs groups (especially Celecoxib and Etoricoxib)the values are greatly recovered. Overall, in comparisonto the DMH group it was noted that AA levels weredecreased and palmitate levels increased very signifi­cantly with NSAIDs co-administrations.

Table VPercentage of arachidonic acid (20:4) and palmitic acid(16:0) prcsent in the lipids extract calculatedfrom the

quantitative data (wt%) obtainedfrom the elutionprofiles ojfatty acid methyl ester derivativesjrom GLC

Arachidonic acid Palmitic acidSatllration(%) Unsaturation (%) (20:4) (16:0)

Control 41.4±0.12 58.6±2.78 Control 100±3.47 100±O.l2

DMH 39.8±0.19' 60.1±2.26 DMH 160.86± 5.65' 89.72± 0,12'

DMH+Aspirin 38.4±O.II 61.6±2.58 DMH+Aspirin 113.04± 2.17"' 98.13± 0.08"'

DMH + Celecoxib 40.7 ± 1.02 59.3±2.51 DMH + Celecoxib 47.83± 3.47" 100±0.24'

DMH +Etoricoxib 42.8 ± 10.3" 57.2±2.07 DMH +Etoricoxib 65.22± 3.04" 102.80± 1.58"

Thevalues aremean ± SDof8 animals. Signiticantly differenl from control al"p" 0.05 andDMHal'p" 0.001 byoneway ANOVA.

Thevalues aremean ± SDof8 animals. Signiticantly differenl from control al"P" 0.01, 'P" 0.001 andDMHal'p" 0.001 byoneway ANOVA.

1146 Nutr Hosp, 2011;26(5):1141-1154 S. Singh Kanwar el al.

o Control O DMH O DMH + Aspirin O DMH + Celecoxib DI DMH + Etoricoxib

60

BBM

z

e/z j:e/z

el-.J.. .5:- ely

~f.L1 u

-..L

50

40

:;? 30

20

10

oLiposomes Colonoeytes

Fig. 1.-Effects of various treat­ments on the lipid phase separa­tions as measured in terms ofper­centage ofjluorescence quenching(%Q) using NBD-PE (L-a-Phos­phatidylethanolamine, Dipalmitoyl.N-NBD) as the probe in BBM vesi­desea), liposomes (b) and isolatedviable colonocytes (c). The valuesare mean ± S.D. of six animalseach. Significantly different fromcontrol at"p s 0.05, hp s 0.01, 'p s0.001 and DMH at 'p s 0.01, 'p s0.001 by one way ANOVA.

Evaluation oflateral phase separationby N-NBD-PEfiuorescence quenching

The effects ofDMH and DMH + NSAIDs treatmenton the physical state of BBM vesicles, liposomes andcolonocyte membranes on the NBD fluorescence areshown in the figure 1. Results showed a decrease in %Q values in all the membrane systems for DMH groupwhen compared with the control, which was in generalrising in the DMH+NSAIDs. The effect was mostprominent in case of the Etoricoxib group.

Evaluation ofmembrane phase state

Laurdan exGP and emGP spectra were measuredfrom 320 to 400 nm and from 440 to 490 nm, respec­tively, to characterize the lipid phase of three differentmembrane bilayer systems, viz, liposomes, BBM andcolonocytes. In order to quantify the slope changes, therelative values of exGp340 were measured on exGPspectra (at 340 nm) (fig. 2a) and emGp490 measured onemGP spectra (at 490 nm) (fig. 2b). The exGp340 valuewas found to be decreased in DMH group for all themembrane systems, while a significant improvementwas observed for the DMH + NSAIDs and in particu­lar, DMH + Etoricoxib group. ExGP340 is an importantparameter of lipid phase state and membrane polarityas this value was shown to be directly related to theincreasing cholesterol content of a membranc.>

Fluidity and order parameter assay using DPH

Fluorescence polarization was expressed as the fluo­rescence anisotropy (r), and as the anisotropy parameter[(r/r)-l]', which are probe dependent phenomena andreflect the Oyeran motional freedom of the fluorescent

molecule without distinguishing the specific mecha­nisms affecting its motions. The specific mechanismsthat affect the motion include alteration of the rate of rota­tion of the probe owing to viscous drag of the environ­ment, anisotropic rotations, and hindered motions owingto structural factors. Thus, within certain limitations,anisotropy and anisotropy parameters are accepted asproportional to the fluidity. The results were obtainedaccording to the modified Perrin relationship as given byEq. (3): r = r

w+ (ro-rJ[T/(Te+T

f) ] where ro is the maximal

limiting anisotropy, taken as 0.365 for DPH, roo is thelimiting hindered anisotropy, Te is the correlation timeand T, is the mean lifetime of the excited state. It shouldbe noted, however, that with the present instrument(Perkin Elmer LS 55), r values were obtained directlyfrom the instrument and anisotropy parameter calculatedusing ro= 0.365. The static component of the membranefluidity was assessed by an order parameter (S) accordingtoEq. (4)." S = [(4/3r-0. l)/ro] 112.

Results of order parameter as calculated from theobserved anisotropy values are given in figure 3a andshow the lowest S value and therefore, increased state oflipid fluidity in the DMH group while with the NSAIDtreatments the situation was reversed with higher Svalues and lesser fluidity. Similarly, the fluidity changesas calculated from the inverse of polarization values (lIP)show an enhanced state of fluidity in DMH group whichwas corrected in the NSAIDs groups (fig. 3b). The lipidfluidity was observed minimum in DMH + Etoricoxibgroup as also shown by maximum ordered state of lipids(higher S value).

Morphological evaluation ofapoptosisin colonocytes

Apoptosis in the colonocytes was studied byassessing the nuclear morphology using DNA

NSAIDs and membrane lipid dynamicsin colon cancer

Nutr Hosp, 2011;26(5):1141-1154 1147

o Control O DMH O DMH + Aspirin [] DMH + Celecoxib m DMH + Etoricoxib

.l..z

y --=- z

b,..L....r.... e/x e e

f...i- e e I ~--Í..~r-i-

e

----"'-- b/z

e/z

e/z

Fig. 3.--{a) Effect ofdifferent treat­ments on the membrane order pa­rameter as measured in the lipo­somes p rep ared from the lip idsextractedfrom colonic mucosa. Thevalues for order parameter (S) cal­culated from the anisotropy valuesrecorded from the fluorescence ofmembrane permeable fluorophore.Diphenylhexairiene (DPH). (b) Ef­fects ofdifferent treatments on themembrane [luidity (JlP) as mea­su red from the DP H fluorescencepolariration values (P). The valuesare mean ± S.D. of six animalseach. Significantly different fromcontrol at hp s 0.01, '/J s 0.001 andDMH at ir: s 0.001 by one wayANOVA.

Fig. 2.-Effects of different treat­ments on the relative values ofexGY'40 (a) and emGP49°(b) asquantiiatedfrom the emGP spectraofLaurdan in liposomes, BBM vesi­eles and isolated colonocytes. Thevalues are mean ± S.D. of six ani­mals each. Significantlv differentfrom control at"ps 0.05, hp S 0.01,'/Js 0.001 and DMH at fJ s 0.05, '/J s0.01, 'p s 0.001 by one way ANOVA.

Colonoeytes

DMH +Celecoxib DMH +Etoricoxib

DMH +Celecoxib DMH +Etoricoxib

a

BBM Colonoeytes

aly 1a

BBM

DMH+Aspirin

DMH+Aspirin

xD DMH + Aspirin D DMH + Celecoxib DMH + Etoricoxib

DMH

DMH

alz

DDMH

Control

Control

y y

Liposomes

D Control

Liposomes

[1

.u:.sz:

e/z.....--=-

~

e

B3.83.7

3.6

§'3.5

304

O 3.3'i3

3.2Bp. 3.1

3

2.9

2.8

A

0.7

0.6

0.5

a. 004O" 0.3'"

0.2

0.1

O

B

O

-0.1

-0.2

iO -0.3S'" -004

-0.5

-0.6

A0.78

0.76

~ 0.74

:u 0.72tiS

0.7"Elo- 0.68:u-eO 0.66

0.64

0.62

binding dyes acridine orange-ethidium bromide co­staining as well as by the TUNEL assay in theparaffin sections of the colonic tissue. Using fluores-

cent DNA binding dyes apoptotic cells were indi­cated by the appearance of brightly labelled nuclei(bright orange coloured). Figure 4a-e shows the

1148 Nutr Hosp, 2011;26(5):1141-1154 S. Singh Kanwaret al.

Fig. 4.-Morphologieal [eatures ofeolonoeytes from different treat­ments showing the induction ofapoptosis as visualized hy acridineorange & ethidium bromide eo­staining: (a) Control, (h) DMH, (e)DMH + aspirin; (d) DMH + eele­coxib and (e) DMH + etoricoxib.Arrows indicate apoptotic eells.

photomicrographs of the appearance of normal andapoptotic cells as observed in the various groupsusing fluorescence microscopy while figure Sa-eshows the results of TUNEL assay. To quantify theextent of apoptosis from these studies percentage ofapoptotic colonocytes (as observed using fluores­cence microscopy) as well as apoptotic nuclei (asobserved under light microscope using TUNELassay) were calculated and presented in figure 6.Results from both the studies showed similar trends.It was found that only a minimum number of apop­totic cells were present in the DMH group while anincreased percentage of apoptotic cells observed inthe DMH + NSA1Ds groups. Maximum occurrenceof apoptotic cells were recorded in the DMH + Eotri­coxib group.

Discussion

Cell membrane is an integral part of the alive cell andplays an essential role in life processes; it makes the cellan isolated system and determines its specific propertiesunder a well controlled equilibria comprising ofproteins, phospholipids, neutrallipids and fatty acids.A cancer transformation results in the appearance of anew cellline in the organism whose malignant activityis transmitted from one cell generation to another and isconnected with the changes in the cell membranes."Cell proliferation and tumor growth occur due to anincreased rate of de novo synthesis of its constituentcomponents. Also, cancer transformations occur at thecost of ccll's immune capacities and changes inmembrane composition would affect growth, interac-

Fig. 5.-TUNEL assay for deteetingthe apoptotic nuclei (dark hrown)in the colonic segments from differ­ent treatment groups. (a) Controlsection showing the occurrence ofmoderate apoptosis, (h) sectionfrom DMH showing the negligibleamount of apoptosis , (e) DMH +aspirin showing the moderate levelsof apoptotis, (d) DMH + celecoxiband (e) DMH + etoricoxib showingthe increased presenee of apoptoticeells.

NSAIDs and membrane lipid dynamicsin colon cancer

Nutr Hosp, 2011;26(5):1141-1154 1149

b/z

e/z

Agp 40

-2

-~ 35

~l30

.S::-O 25

¡J 20

'5~ 15~ ~c.,

10s~ 51 o

Control

B

60

50

40.~ ¿(j

~~30

E

~t:c 20

c.,

10

OControl

DMH

DMH

alz

DMH+Aspirin

DMH+Aspirin

DMH +Celecoxib DMH +Etoricoxib

e/ze/z

DMH +Celecoxib DMH +Etoricoxib

Fig, 6.-Percent ap optoticce lis. (a) Fluo rescence mi­croscopy using acridine or­ange- ethidium bromide co­staining, (h) Light microscopyofParaffi n emhedded sectionsusing TUNEL assay. The val­ues are mean ± S.D. offour tosix animals. Significantly dif­ferent from control ai-p s 0.05,hp s 0.01, '/Js 0.001 and DMHat ip « 0.001 hy ANOVA.

tion with other cells (irnmune system), and the functionof proteins and other components in the membrane.

Alterations in the synthesis, uptake, and membranecontent of cholesterol have been observed in a varietyof experimental tumor models as well as in humanneoplasms." Cholesterol content and Chol:PL ratio areimportant from malignancy point of view as also indi­cated by many previous reports." In the present studythe decrease in the cholesterol content and its synthesiswas observed in the DMH group which was found to beassociated with the decrease in Chol:PL ratio alongwith increased triglyceride (TG) biosynthesis. Thesechanges were found to be reversed with NSAIDs treat­ments. It has been found that cancer cell membranesha ve low cholesterol content and low Chol: Proteinratio. Chol:PL ratio is also closely related to themembrane fluidity. Addition of cholesterol is known tocause a transition from gel to liquid-crystalline phaseresulting in alteration in the nature and extent of phos­pholipid-protein interaction and a decrease in ampli­tude of motion of the chain axes and reduced fluidity ofthe system." Similarly, TG contents have also beeninvestigated in previous works to be involved incolorectal cancer. A study of Mckeown-Eyssen" hasshowed the elevated levels of TG in the serum to bedirectly involved with colorectal cancer. An increasedrisk of colorectal adenomas was found among menwith the highest level of serum TG in a Japanesestudy," therefore, an increased synthesis for the TG asobserved in the DMH group seems to be an importantfinding.

Further, increments were also observed in thebiosynthesis of major phospholipids studied i.e., S, PCand PE (except for PI + PS fractions) in the DMHgroup while lower values were observed in NSAIDgroups. An increment in biosynthesis of phospholipidsmay be required for cell proliferation, especially duringcarcinogcncsis," in particular, PC biosynthesis hasbeen implicated in the process, while an inhibition ofPC biosynthesis in tumor celllines has been associatedwith an increased event of apoptosis. The increase inPC values in DMH group therefore, indicates the moreproliferative capability while a decreased synthesisindicates the inhibition of PC biosynthesis in NSAIDgroups which is consistent with the enhanced role ofthe NSAIDs in apoptosis. Further, in NSAIDs groupsincreased PI was noted. An increased production of PIwas observed to be associated with disruption ofmembrane integrity of both mitochondrion and endo­plasmic reticulum. Agents, such as etoposide, camp­tothecin, farnesol, and chelerythrine, and ether lipidswhich are known inducers of apoptosis have also beenshown to inhibit PC biosynthesis, although the effectsof these agents on PI biosynthesis ha ve not beendescribed."

Gangliosides are important membrane constituentsthat may influence growth and cell-to-cell interactionsand therefore, playa vital role in the development ofmalignancy."-" Thus in the present study gangliosidewas assayed in the lipids derived from the colonicmucosa to look into any possible NSAIDs mediatedregression in colon carcinogenesis. The present results

uso Nutr Hosp, 2011;26(5):1141-1154 S. Singh Kanwaret al.

showed an elevated level of gangliosides in the colonof DMH treated animals whereas significant regressionwas observed in the coxib treated groups. Previously,Lu et al." ha ve shown that ganglioside contents intumor tissues as well as the plasma sialic acid levels tobe significantly increased in experimental tumorigen­esis. Earlier studies have also demonstrated the alter­ations in the levels of ganglioside sialic acid to be asso­ciated with the process of carcinogenesis in varioustissues.":"

Cancer causation is a process that require a constantsupply of fatty acids from the organism's metabolicapparatus and it is well established that lipid composi­tion of the immune and tumor cell membranes areinfluenced by the fatty acid constituents." Fatty acids,for example: palmitoleic 16:1, oleic 18:1, linoleic 18:2,along with stearic 18:0, palmitic 16:0, myristic 14:0and arachidonic acid 20:4 are known to potentiallyinfluence these interactions.

Fatty acids like arachidonic acid influences thesynthesis of eicosanoids (prostaglandins, leukotrienesand thromboxanes) and in addition to their role in theregulation of immune and inflammatory responses,eicosanoids may also be needed to sustain growth oftumor cells. Palmitic acid on the other hand has beenreported to have a key role in the mitochondrialmembrane potential and cytochrome e release, thereby,inducing apoptosis in the cells. In view of this, lipidsextracted from the colonic mucosa were analyzed forfatty acid composition including most importantly AAand palmitic acid. An increased level of AA was foundin the colonic tissues obtained from the DMH group ascompared with the control or NSAID treated groups.Previously, Nicholson et al." showed consistently highvalues of cell membrane AA in an experimentalcolorectal tumor tissue. A recent study and earlier clin­ical reports on the colorectal tissues of human cancerhave also showed an increase in AA.",41 These studiescontributed to the idea that the increase in AA is afeature of the rapidly growing cells. On the other hand,the increased level of palmitate in NSAIDs groups thussignifies their role as inducers of apoptosis. Inductionof apoptosis by palmitate has also been observed inother cell types, including cardiomyocytes, hematopoi­etic cells, pancreatic f:\-cells, and astrocytes, but formost of these cell types, the mechanism by whichpalmitate induces the cell deathremains elusive." It hasbeen proposed that excess palmitate could induce celldeath through increased intracellular concentration ofceramide, a metabolite exclusively produced from thesaturated FFAs. Alternatively, other workers" ha vesuggested that apoptosis induced by palmitate couldoccur through the generation of the reactive oxygenspecies (ROS). In the present studies, apoptosis wasevidently observed to occur in the NSAID groups asobserved through fluorescent staining of colonocytesand TUNEL assay. These results were found to be inagreement to those ofLiu et al.," Kim et al." and Xu etal.," where apoptosis was found to be induced after

NSAIDs administration. These observations takentogether therefore strongly indicate the role of palmi­tate in preventing the DMH induced carcinogenesispossibly through apoptosis.

The segregation of membrane lipids in domains,called lateral phase separation assumes importance dueto their regulatory role in cellular functions such asmembrane sorting and signal transduction." Fluores­cent lipid probes have been proved very useful due totheir ability to monitor lipid molecules by a variety ofphysicochemical approaches at increasing spatiotem­poral resolution. A widely used extrinsic fluorophorein the studies of membrane is the NBD (7-nitroben­zenz-2-oxa-l,3-diazol-4-yl) group which is veryweakly fluorescent in water and upon transfer to ahydrophobic medium, fluoresces brightly in the visiblerange and also exhibits a high degree of environmentalsensitivity. Various fluorescent analogues of nativelipids labeled with NBD have been used to study avariety of biological processes such as cell growth andproliferation, apoptosis and signal transduction."Originally, Hoekstra" presented a method for moni­toring lipid phase separation by using small amount offluorescent phospholipids (NBD labelled) analoguesincorporated into the vesicle bilayer, which becomeself quenched during the process of phase separation.Separation of lipid phase leads to an increase in localconcentrations of the NBD-lipid in the bilayer, andhence to self quenching of NBD fluorescence. The rateand extent of NBD self quenching provides themeasure of the rate and extent of phase separation inmembrane bilayers.

The increase in %Q was found to be to a larger extentin case of DMH + Etoricoxib group, both in compar­ison to control as well as DMH group. This indicatesthat under present experimental conditions the DMHtreatment promotes the gel state of the membrane asmore separations in the lipid phases of the membraneoccurs in the bilayers of BBM vesicles, liposomes andplasma membranes of the colonic epithelial cells.Previous studies have demonstrated that the presenceof phase separations in mixed phospholipicl/cholesterolvesicles" in which, a increase in initial fractional %Qwas found to be associated with increasing cholesterolconcentration. Since, in the present study also, choles­terol content in the colonic lipids and the BBM vesicleswere found to be decreased more in the DMH groupwhile it was increased in NSAIDs group, it can be spec­ulated that the NSAIDs may induce the intrinsic lipidphase separation by perhaps regulating the cholesterolconcentration in the membrane systems.

To study the polarity of the lipid interfaces, Laurdan(6-dodecanoyl-2-dimethylaminonaphthalene) was used,which is an amphiphilic fluorescent membrane probesynthesized by Prof G. Weber to study the dipolarrelaxation processes." This probe has the advantage ofdisplaying spectral sensitivity to the phospholipidphase state. Its steady state excitation and emissionspectra can be resolved into two spectral components

NSAIDs and membrane lipid dynamicsin colon cancer

Nutr Hosp, 2011;26(5):1141-1154 1151

that correspond to the two phases. The time evolutionof the Laurdan emission spectrum has been measuredto detect the coexistence of different phospholipidphases in mixed vesicles and to determine the intercon­version rate between the phases." Also, Laurdan exci­tation spectrum is different in the two phospholipidphases. Quantitization ofthe relative amount of gel andliquid crystalline phases has been described by steady­state generalized polarization (GP) which is defined as:

Where Ig

and 11are the fluorescence intensities at the

maximum emission in phospholipids in the gel and inthe liquid crystalline phase, respectively (excitation GPor exGP). In addition, the GP value can be obtainedusing the fluorescence intensities at the maximum exci­tation in phospholipids in the gel and in the liquid-crys­talline phase (emission GP or emGP). The particularwavelength behavior of the GP value gives informationon the coexistence of different phases and on their inter­conversion. The GP values are limited between +1 and­1. The characteristic GP values in the pure gel (ordered)and in the pure fluid liquid-crystalline (disordered)membranes have previously been deterrnined." Ourresult showed lower values of exGp340 for DMH treatedgroup and increasing values with coxib treatrnents. Theobserved increase in exGp340 values in membranes of"coxib" groups indicated the presence of more lipids ingel phase as in the gel phase dipolar relaxation of thephospholipids is strongly reduced either because of theabsence of water or because the phospholipid tightpacking preventing a rapid relaxation of water mole­cules." Further, Parasassi et al.," has shown that theLaurdan in different phospholipid vesieles has a welldefined value for the gel phase (GP = 0.6) and for theliquid crystalline phase (GP = -0.2), respectively. In thepresent study taking the exGp340 (fig. 3a) and emGP490(fig. 3b) values into considerations it can be elearly seenthat the control values are eloser to the gel phasewhereas for the DMH group values are eloser to theliquid crystalline phase. In the coxib groups (Celecoxiband Etoricoxib) there is a shift in GP values towards thegel phase. These trends are better followed in case of theliposomal vesieles than in the BBM or colonocyte vesi­eles probably because of more complexities involved inthe nature and composition of the later membranes. 23

Further, in the present studies, steady-state fluores­cence polarization of lipid soluble probe was applied toassess membrane state of fluidity in terms of rotationaldiffusion as assessed by order parameter (using DPH).The particular usefulness of these methods stems fromthe fact that the polarization of the fluorescence of amolecule depends upon its rate of rotation." Hencc,changes in rotation rate due to interactions of the probewith other species in the environment are readilyobserved and quantified. Since the rotation rate dependson the resistance offered by the microenvironment to themotion of the probe, fluorescence polarization provides

an estimate of the environmental resistance which isinterpretable as an apparent "microviscosity". From theDPH fluorescence measurements of order parameter, itcan be seen that DMH group has lower order parameterindicating lesser phospholipid packing in themembrane bilayer. NSAIDs on the other hand haveresulted in increasing the membrane lipid order andtherefore, are responsible for creating more of themembrane into gel phase as also found with theLaurdan studies.

In the biological membranes phospholipids have acomplex composition, ineluding differences in lengthand unsaturation of their acyl chain residues and in thetype of polar heads. Gel phase is assumed to be "lesspolar" with respect to the liquid crystalline phasc" andtherefore, provides a good basis for the observedinduction in lipid phase separations (as evident fromthe NBD-PE studies) in the NSAID groups. Theconverging effects of these changes in the membranephysical state was observed more evidently in terms offluidity (fig. 4b), which was found to be more in DMHgroup (also indicated by less phase separations or highpolarity, and liquid crystalline phase state) and lesserfor NSAID groups (as al so indicated by higher phaseseparations or low polarity, and more of the gel phase).In addition, cholesterol and cholesterol: PL ratio is animportant chemical index of membrane fluidity. Thedecrease in the chol: PL ratio as observed in the DMHgroup correlates with an increase in local polarity,fluidity and molecular disordering (as seen in terms oflesser phase separation and lower order parameter).From these results, it seems that DMH induces signifi­cant alterations at the membrane structural level bymodulating its composition and alters its physicaldynamics so as to make the conditions more favorablefor the process of carcinogenesis, whereas, NSAIDsmay cause their membrane modulating effects byaffecting the lipid phase separations and therebycausing the decrease in membrane polarity.

The concomitant increase in the number of apoptoticcells in NSAIDs treated groups in comparison to DMHgroup substantiates the anti-proliferative effects ofNSAIDs. Indeed, in our previous study, within thecolonic cell membranes NSAIDs have been found toaffect many membrane associated processes (such assodium proton exchange activity, intracellular pH,calcium homeostasis, mitochondrial dysfunction) andalso responsible for the inhibition of membrane associ­ated enzymes like cyclooxygenasc, thereby, creatingthe microenvironment favorable for the inhibition ofcellular progression of the colon cancer. 10 The inhibi­tion of COX was believed to be the sole explanation forthe chemopreventive actions of NSAIDs. However, itis now believed that there should be additional mecha­nisms to explain the overall chemopreventive pharma­cological behavior of NSAIDs. Since membranedynamics are responsible for the functioning ofmembrane bound enzymes and permeability of the ions,the present study provides a possible explanation for the

1152 Nutr Hosp, 2011;26(5):1141-1154 S. Singh Kanwaret al.

ability of NSAIDs to influence these biophysical para­rneters contributing towards the induction of apoptosisduring colon carcinogenesis. Results presented in thisstudy can be higly relevant in providing sorne insightinto the usefulness ofNSAlDs as a potential chernopre­ventive agents.

Acknowledgernents

We thank lndian Council of Medical Research(lCMR), New Delhi for providing the financial assis­tance to Shailender Singh Kanwar as Senior ResearchFellowship. Financial support for this work was alsoprovided by University Grants Cornrnision (UGC)­Special Assistance Programme, New Delhi.

References

1. Stern RG, B.N. Milestone BN, Gatenby RA. Carcinogenesisand the plasma membrane. Med Hypotheses 1999; 52: 367-372.

2. Bogdanov M, Mileykovskaya E, Dowhan W. Lipids in theassembly 01'membrane proteins and organization 01'proteinsupercomplexes: implications for lipid-linked disorders.Subcell Biochem 2008; 49: 197-239.

3. Sousa C, Nunes C, Lúcio M, Ferreira H, Lima JL, Tavares J,Cordeiro-da-Silva A, Reis S. Ellect 01' Nonsteroidal anti­inflammatory drugs on the cellular membrane fluidity. J PharmSci 2008; 97: 3195-3206.

4. Cho NL, Redston M, Zauber AG, Carothers AM, Hornick J,Wilton A, Sontag S et al. Aberrant crypt Ioc¡ in the adenomaprevention with celecoxib trial. Cancer Prev Res (Phila Pa)2008; 1: 21-31.

5. Lucio M, Nunes C, Gaspar D, Golebska K, Wisniewski M,Lima JLFC, Brezesinski G, Reis S. Ellect 01' anti -intlammatorydrugs in phosphatidylcholine membranes: A fluorescence andcalorimeteric study. Chem Phys Lett 2009; 471: 300-309.

6. Kanwar SS, Vaiphei K, Nehru B, Sanyal SN. Chemopreventiveeff'ects 01' non steroidal anti-inflammatory drugs on 1, 2­dimethylhydrazine-induced colon carcinogenesis in rats.ToxicoIMechMethods2007; 17: 197-204.

7. Kanwar SS, Vaiphei K, Nehru B, Sanyal SN. Chemopreventiveeff'ects 01' nonsteroidal anti-inflammatory drugs in themembrane lipid composition and fluidity parameters 01'the 1,2­dimethylhydrazine-induced colon carcinogenesis in rats. DrugChem ToxicoI2007; 30: 293-309.

8. Kanwar SS, Vaiphei K, Nehru B, Sanyal SN. Antioxidativeeffects ofnonsteroidal anti-inflammatory drugs during the initi­ation stages 01' experimental colon carcinogenesis in rats.J Environ Pathol Toxicol Onco12008; 27: 89-100.

9. Behal N, Kanwar SS, Kaur J, Sanyal SN. Cyclooxygenase-2inhibition by etoricoxib modulates plasma membrane fluidityin rat colon. J Environ Pathol Toxicol Oncol 2008; 27: 295­302.

10. Kanwar SS, Roy KR, Nehru B, Reddanna P, Sanyal SN. Na'stimulated N a'/H' exchange and an unfavorable Ca" home­ostasis initiate the Cycloxygenase-2 inhibitors induced apop­totic signals in the colonic epithelial cells during the early stage01'colon carcinogenesis. Oncol Res 2009; 18: 243-257.

11. Furst DE, Ulrich RW. Nonsteroidal Anti-inflammatory drugs,discasc-modifying antirheumatic drugs, nonopoid analgesicand drug used in gout. Mac Graw Hill Lange, Lange, 2006.

12. Shi S, Klotz U. Clinical use and pharmacological properties 01'selective COX-2 inhibitors. Eur J Clin Pharmacol2008; 64:233-252.

13. Chung-Faye GA, Kerr DJ. ABC 01'colorectal cancer: Innova­tive treatment íor colon cancer. BMJ2000; 321: 1397-1399.

14. Di Pace RF, Massa S, Ribeiro OG, Cabrera WH, De Franco M,Starobinas N, Seman M, Ibañez Oc. Inverse genetic predispo­sitian to colon versus lung carcinogenesis in mouse linesselected based on acute inflammatory responsiveness. Corcino­genesis2006;27: 1517-1525.

15. Brasitus TA, Keresztes RS. Protein-lipid interactions inantipodal plasma membranes 01' rat colonocytes. BiochimBiophysActa 1984; 773: 290-300.

16. Mouille B, Robert V, Blachier F. Adaptative increase 01'ornithine production and decrease 01'ammonia metabolism inrat colonocytes after hyperproteic diet ingestion. Am J PhysiolGastrointest Liver Physio112004; 287: G344-G35l.

17. Sanyal S, Agarwal N, Subrahmanyam D. Ellect 01' acutesublethal and chronic administration 01'DDT (chloropheno­tane) on brain lipid metabolism 01'rhesus monkeys. Toxicol Lett1986; 34: 47-54.

18. Sadana T, Dhall K, Sanyal SN, Wali A, Minocha R, MajumdarS. Isolation and chemical composition 01'surface-active mate­rial from human lung lavage. Lipids 1988; 23: 551-558.

19. Sarwal R, Sanyal SN, Khera S. Lipid metabolism in Trichurisglobulosa (Nematoda). J Helminthol1989; 63: 287-297.

20. Gumbhir K, Sanyal SN, Minocha R, Wali A, Majumdar S.Glucose-6-phosphate phosphohydrolase activity in guinea pigliver microsomes is influenced by phosphatidylcholine. Inter­action with cholesterol-enriched membranes. Biochim BiophysActa 1989; 981: 77-84.

21. Hoekstra D. Fluorescence method for measuring the kinetics 01'Ca-t-induced phase separations in phosphatidylserine-containinglipid vesicles. Biochemistry 1982; 21: 1055-1061.

22. Nichols JW, Pagano RE. Kinetics 01'soluble lipid monomerdiffusion between vesicles. Biochemistry 1981; 20: 2783-2789.

23. Ambrosini A, Zolese G, Wozniak M, Genga D, Boscaro M,Mantero F, Balercia G. Idiopathic infcrtility: susceptibility 01'spermatozoa to in-vitro capacitation, in the presence and theabsence 01' palmitylethanolamide (a homologue 01' anan­damide), is strongly correlated with membrane polarity studiedby Laurdan fluorescence. Mol Hum Reprod2003; 9: 381-388.

24. Parasassi T, De Stasio G, Ravagnan G, Rusch RM, Gratton E.Quantitation 01'lipid phases in phospholipid vesicles by thegeneralized polarization 01'Laurdan fluorescence. Biophys J1991; 60: 179-189.

25. Pottel H, Vandermeer W, Herreman W. Correlation between theorder parameter and the steady-state fluorescence anisotropy 01'I ,6-diphenyl-1 ,3,5-hexatriene and an evaluation 01'membranetluidity. Biochim BiophysActa 1983; 730: 181-186.

26. Nair P, Kanwar SS, Sanyal SN. Ellects ofnon steroidal anti-intlam­matory drugs on the antioxidant defense system and the membranefunctions in therat intestine. Nutr Hosp 2006; 21: 638-649.

27. Baker AJ, Mooney A, Hughes J, Lombardi D, Johnson RI, SavillJ. Mesengial cell apoptosis: The major mechanism for resolution01'glomerular hypercellularity in experimental mesengial prolif­erativenephritis. .l Clin Invest 1994; 94: 2105-2116.

28. Dobrzynska 1,Szachowicz-PetelskaB, Sulkowski S, FigaszewskiZ. Changes in electric charge and phospholipids composition inhuman colorectal cancer cells. Mol Cell Biochem 2005; 276:113-119.

29. Yachnin S, Larson RA, West EJ. Rates 01'cholesterol biosyn­thesis are related to early differentiation in acute non-Iympho­cytic leukaemia cells. Br J Haematol1983; 54: 459-466.

30. Yamada K, Araki S, Tamura M, Sakai 1, Takahashi Y,Kashihara H, Kono S. Relation 01' serum total cholesterol,serum triglycerides and fasting plasma glucose to colorectalcarcinoma in situ. Int J Epidemiol1998; 27: 794-798.

31. Hubbell WL, McConnell HM. Molecular motion in spin­labeled phospholipids and membranes. J Am Chem Soc 1971;93: 314-326.

32. McKeown-Eyssen G. Epidemiology 01'colorectal cancer revis­ited: are serum triglycerides and/or plasma glucose associatedwithrisk? Cancer Epidemiol Biomarkers Prev 1994; 3: 687-695.

33. Kono S, Imanishi K, Shinchi K, Yanai F. Yanai, Serum lipidsand left-sided adenomas 01'the large bowel: an extended study01'self-defense ofticials in Japan. Cancer Causes Control 1993;4: 117-121.

NSAIDs and membrane lipid dynamics

in colon cancer

Nutr Hosp, 2011;26(5):1141-1154 1153

34. Jackowski S. Cel! cycle regulation of membrane phospholipidmetabolismo J Blol Chem 1996; 271: 20219-20222.

35. Finney RE. Nudelman E. White T. Bursten S. Klein p. Leer LL.Wang N et al. Pharmacological inhibition of phosphatidyl­choline biosynthesis is associated with induction of phos­phatidylinositol accumulation and cytolysis of neoplastic cel!lines. Caneer Res 2000; 60: 5204-5213.

36. Lu CQ. Lu J. Wang BL. Cui YZ. Changes in gangliosidecontents, plasma sialic acid and cAMP levels in experimentalhepatoma in mice. Mol Cell Biochem 2000; 207: 29-33.

37. Uemura S. Feng F. Kume M. Yamada K. Kabayama K.Nishimura S. Igarashi Y. Inokuchi J. Cel! growth arrest by sialicacid clusters in ganglioside GM3 mimetic polymers. Glycobio­logy 2007; 17: 568-577.

38. Stefenel!i N. Klotz H. Engel A. Bauer P. Serum sialic acid inmalignant tumors, bacterial inf'ections, and chronic liverdiseases. J Caneer Res cu« Oneo11985; 109: 55-59.

39. Field CJ. Thomson CA. Van Aerde JE. Parrott A. Euler A. LienE. Clandinin MT. Lower proportion of CD45RO+ cel!s anddeficient interleukin-IO production by formula-fed infants,compared with human-fed, is corrected with supplementationof long-chain polyunsaturated íatty acids. J Pediatr Gastroen­terol Nutr 2000; 31: 291-299.

40. Nicholson ML. Neoptolemos JP. Clayton HA. Talbot IC. Bel!PRo Increased cel! membrane arachidonic acid in experimentalcolorectal tumours. Gut 1991; 32: 413-418.

41. Neoptolemos JP. Husband D. Imray C. Rowley S. Lawson N.Arachidonic acid and docosahexaenoic acid are increased inhuman colorectal cancer, Gut 1991; 32: 278-281.

42. Blázquez C. Galve-Roperh l. Guzmán M. De novo-synthesizedceramide signals apoptosis in astrocytes via extracel!ularsignal-regulatedkinase. FASEB J2000; 14: 2315-2322.

43. Listenberger LL. Ory DS. Schaffer JE. Palmitate-inducedapoptosis can occur through a ceramide-independent pathway.J Blol Chem 2001; 276: 14890-14895.

44. Liu X. Yue p. Zhou Z. Khuri FR. Sun SY. Death receptor regu­lation and celecoxib-induced apoptosis in human lung cancercel!s. J Natl Caneer Inst 2004; 96: 1769-1780.

45. Kim SH. Song SH. Kim SG. Chun KS. Lim SY. Na HK. KimJW et al. Celecoxib induces apoptosis in cervical cancer cel!sindependent of cyclooxygenase using NF-kappaB as a possibletarget, J Caneer Res cu« Oneo12004; 130: 551-560.

46. Xu Y. Zhao YM. Huang H. Celecoxib-induced apoptosis inacute promyelocytic leukemia cel!line MR2 and its mecha­nism. Zhejiang Da Xue Xue Bao YI Xue Ban 2007; 36: 319­324.

47. Kusumi A. Nakada C. Ritchie K. Murase K. Suzuki K.Murakoshi H. Kasai RS. Kondo J. Fujiwara T. Paradigm shiftof the plasma membrane concept frorn the two-dimensionalcontinuum fluid to the partitioned fluid: high-speed single­molecule tracking of membrane molecules. Annu Rev BiophysBiomol Struet2005; 34: 351-378.

48. Raghuraman H. Shrivastava S. Chattopadhyay A. Monitoringthe looping up of acyl chain labeled NBD lipids in membranesas a function of membrane phase state. Biochim Biophys Acta2007; 1768: 1258-1267.

49. Gebhardt C. Gruler H. Sackmann E. On domain structure andlocal curvature in lipid bilayers and biological membranes. ZNaturforsch. C 1977; 32: 581-596.

50. Weber G. Farris FJ. Synthesis and spectral properties of ahydrophobic fluorescent probe: 6-propionyl-2-(dimethylamino)naphthalene. Biochemistry 1979; 18: 3075-3078.

51. Parasassi T. Di Stefano M. Loiero M. Ravagnan G. Gratton E.Influence of cholesterol on phospholipid bilayers phasedomains as detected by Laurdan fluorescence. Biophys J 1994;66: 120-132.

52. Schachter D. Shinitzky M. Fluorescence polarization studies ofrat intestinal microvil!us membranes. J Clin Inves 1977; 59:536-548.

1154 Nutr Hosp, 2011;26(5):1141-1154 S. Singh Kanwar el al.