‘. i ,1 i 2 7 g 5 V l-Tv .rfil- ~yl;rf IlNa-».. “"4”“5‘ .IX‘HH...‘ 3 A. ; ) v!..| to" -A r - a “ 33.7 ..~ . .., . .3: Aft: 5:. - x < .r a. .. 17.5,.» . yt—F-f—d‘ , .. uv 5%. JA 4‘ - ri X. "1“ .7... " ' ,‘tfitéfi-fifi 014 ‘ g . ‘2‘ t; 1 .3 ‘ ~ ‘33 «’4‘? ‘ ~~ ~ If.“ ta: . ' fiwfii‘fl‘m’ ping?» ,7 -, 3.1%,.‘3'1; . .. .11}. , h .1 - I“: In , . ‘33“th 5.24. “E. ‘ L‘I‘K‘J‘iflunfifi '1' {Ls 3‘ ‘3‘ - ~ "BEA. i.g:.,\;.€;‘ 3. sne- m-.;.; ”\M tI.--VZ'.'I: 3' J52 u'w""“" 53x. Mun u.“ as . film an ‘ dwa- (i. .5331 3‘1 '3u3v a,“ . -- w ‘5‘.“ "- 125‘. 41. Hxnr - .nv 1"?“ a ' T: . V‘ .- 5.1.-..“ ug,u‘- M 8“ V L \'\ “39 a. . ‘3 u. 43.5.; amp .- .n- -:4 q." :- 5mm: . h. ‘ u '4 1 .. m-..—..' M, .woM _. ._. ,_ -v-m. .0. ...~- v . _, »- “v.1“.w v... v—o.-s‘. M...” 0“. u ‘ v~A-‘l. u ..‘- \- I - I, ‘ ‘ 3““:4' :I‘I. ‘ I . m ‘0 1 1“- . War“ .1? '4‘ l‘q?§:d~g ‘L ,3 . ’ln32i1:}t[q=ul-u ,L G b A: ‘1“. . - _'..l . fift‘fi.‘ .. - u .33 ngiw... -.£';\§t:“'.z ,¢— I“! VS < 1 u‘UIS-RF‘ *a‘géswi‘; 3‘": .sfirguivm :33; Chan‘s" 232.31“ “wk ~ t‘a‘: "w “-33%?" 1‘21.— .g..«: 4- k... '“m'Z-K' 1.‘ .um. ‘ \. v.\‘ y... x . ‘u... x,_... \ H’s‘ ~. u MIC CHlG CAN STATE !!!!3!!!! !!!2!! ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 300912 1603 This is to certify that the dissertation entitled Cellular Aspects of Aluminum Toxicity : Aluminum uptake by Neuroblastana Cells and Inhibition of Inositol Phosphate Formation by Aluminum in Neuroblastana Cells presented by Biao Shi has been accepted towards fulfillment of the requirements for Ph.D degree in Microbiology A. WV a' professor Date3O Oc‘!« (CM! MS U it an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY ‘ Michigan State University PLACE ill RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before duo duo. DATE DUE DATE DUE DATE DUE MSU is An Affirmative Action/Equal Opportunity Institution cmmS-nt CIIIUIARlASPECTS OFWAIUMINUM'TOXICITY: ALUMINUMTUPTAKE BY NEURDBLASTUMA.CELLS AND INHIBITION OF INDSIToliPHDSPHETE FORMATION BY.AUUMINUM.IN’NEURDEIASTCMAACELL Biao Shi A.DISSERENEHIW smtndxted.to MidtiganStateUniversity inpartialfulfillmentofthemiratents forthedegreeof DDCTDR:OF’PHIIDSOPH! Department of Micmbiology and Public Health 1991 CEIIUIARASPEXII’SOFAHDENIM‘IDHCITY: WWBYWWCEHSANDDHIBITIWOF MEDLWWWBYWDJNHHDWCEIIB Biao Shi Aluminm isamaraltlyinvolvedinabroadspectnmof mysiological disorders, e.g., in certain neurodegenerative diseases ‘of hunts; saneofttmedisorders may originate from almilm'sprimxyinteractim with a key target in cells.(me possibility is that alumilun interferes with phosphoinositide signal transdutim where an intracellular secmd messenger, inositol LLB-twists (IP3) is generated which, in blrn, signals intracellular (22+ release. uploying mine naminlastana cells, labelled with [331-111570- inositol, acpez'inierrtal reallts denonstrate‘ that alumimlm amlimtim drastically reduce inositol phosphate prochction stimlated bprproteinactivatoxs, GTP[S] or fluoride, or the recqltor agcnist bradyldnin. 'me inhibitim principally affects fonnatial of IP3 , fran hydrolysis of phosphatidylincsitol 4,5-bimcsghate (PIPZ), rather than downstream or upstream‘ reactions distal to 1P3 formation along the signal transduction. Maltmirmispredlelatedwifll agents inpermeabletotheplasma radar'ane, aluminnn-related inhibition is almst cmpletely reversed in intact cells. the application of great excess of 1492*, GI‘P[S] ard GTPreduces mly partially or mt all the inhibition of incsitol phosphate formation by aluminum. In addition to m2+/cp protein-mediated PIPZ hydrolysis, phospholipase C reaction can also be activated directly by applying increasing concentrations of Ca2+ in neuroblastana cells. IP3 production in both pathways is sensitive to alumirum, whereas (22+-triggered 1P2 production is not affected by alumimm applicatim. 'Ihese findings suggest that almirm idlibits incsitol prostrate producticn through its putative interactims with the Gp protein arri Midlipase C. Employing atomic absorpticn spectroscqu, alumirum uptake by nalrdllastana cells was studied. At physiological pH, cells can protectthalselves franincorporatl'ngtcndcallmimm. Asthemedium pHdecreases, cells accumulate large anumts of aluminum against the cmcentratim gradient. At neutral pi, transferrin facilitates alumimm uptake, presunably via marbrane receptors. Iowmcleollar weight alminnn-dlelating metabolites like citrate always act to mu'bit alumimm internalization. <22”, alleged to ameliorate alumirum toxicity, does not measurably inhibit aluminnn uptake by nalrcblastana cells and alumilun birding onto the cellular surface. iii 'Ibmyparents,mywifearrimydaughter. iv I would like to thank the following timbers of my dissertation ccmnittee for their advice, help and guidance: Dr. William Frantz, Dr. Alfred Haug, Dr. Patrick Oriel, Dr. Rcmald Patterson, and Dr. James Tiedje. Ialsowishtoaclorlwledge Dr. Karen Chou, inAnimalScience Department, for her concern and surport. And thanks to Li Glen, Harrie Kcenraadt, Dr. Christopher Weis ardnr. ShiximYuanfortheirfrierrishipandhelp. 'IEBIEOFCIIH'EIH‘S page LISTOF FIGJRES.. ............................................. viii LISTOF m............. ..................................... xi m1. mm"... ................... . ............ l Alumimm toxicityinanimals............ ...... . ............. 2 Chemical prcperties of almninnn........... .................. 4 Primary cellular lesials of aluminum taxicfity ............... 7 1. Interactim of alumimnn with chranatin ................. 8 2 . Effect of alumirum an intracellular calcium metabolism........... .......... .................... 10 3. Substitutim of alumirum for M32+ ............. . ....... 13 4. Perulr‘batial of almnimmmtheplasmanmbrane ....... 14 Alminmuptakeaniintracellular distributiwn...” ........ 17 Phcsfiloincsitide signal trarsductlml9 Activaticn of picsphoincsitide signal transduction by flucrcalumlnate ..... ....... ...... 22 ijectives: possible inpact of aluminum m phcsgi'loincsitide signal tramductimandcellularalmninnnuptake.... ..... 24 Listof References... ...... .. ........ . ............. . ....... 27 vi CHAPTER.II. AIUMENUM INTERFERES WITH INDSITOLIPHDSPHAIE mats!WM.................... ....... 34 i. Wu”... .................. . ................... 35 ii. Introductimm... ................................... 36 iii. Materialsanduethods ............ . ....... ....... 38 iv. Resultsw... .......................... . ........ 44 v. Discussim......... .................................. 84 vi. Listof References ..... ...94 CHAPTERIII. WMMWCEHSHHHHHH.99 i. Abstract....... ............ . ................. . ...... 100 ii. Introductim ....... ....101 iii. Materials andHethods... ............. . .............. 103 iv. mm.............. ....... . ............ . ...... ....108 v. Discussim ............... .......129 vi. Listof references........... ......... ..............138 cm IV. WWW... ............... . ....... 142 APPENDIX A. loss of inositol mosplclipids in epididymal sperm following exposure of mice to long-term feeding of WWOOOOOOOI......OOOOOOOOOOOOOOOOOOOI....150 B. Bicdlanical basis of aluminum tolerance in plant mIOOOOOOOOOOC......OOOOOOIO00......0.00.00.00.00I.152 C. Charges in intracellular calcium of porcine sperm durirginvitroilnlbatimwithsanl’nalplaanaarri acapacitationmedium164 D. Uptakeofalmnimmbylipidvesicles..................171 E. mannel-closing activity of porin fran M ganinbilayerlipidmenbrams ........... ............184 vii LISTOF FIGURES Page Gtapter II 1. Elution profiles of incsitcl phosphates by column chrmiatcgraphy ......................................... 42 2. Incorporatim of [3H]myc—inositol into inositol phosphates in nalrcblastana cells ...................... 45 3. Effects of fluoride (:1 [32P1phosrhcincsitide biz-hover in naJroblastana cells ....... . ......................... 48 4 . Dose-deperdence of fluoride-induced inositcl phosphate formtiminneuroblastcna cells...... ..... . ........... 50 5. Almirum—irduced inhibiticn of fluoride-related total inositol Mats and IP3 formtion ................... 52 6. Almirun—iniuced inhibitim of GI'P[S]-triggered total inositol plosphate and 1P3 formatim ................... 57 7. GI'P[S]-irduced [31!]incsitcl frustrate formation in the presalce of nan-radioactive IP359 8. Effect of alminnn cm the dumatograplic assay of inositol Mate in GI'P[S]-stinulated permeabilized QBOOOOOOOOOOOOOOOO ................................. 60 9. GI'P[S] dose-deperdence of incsitol phosphate formation inpermeabilized cellsm... .......................... 62 10. Irhibitim of aluminnn m inositcl phosphate formation in the presence of emgenom GI'P ..... . ................. 63 11. Time cause of GI'P[S]-irduced inositol phcsriiate formationintheabsenceorpresenceof almnirun. ...... 66 Effects of aluminnn-dlelatcrs cm alumimm-l'nduced inhibitim of inositol phosphate formation ............. 69 viii 13. Effect of aluminnn m GI'P[S]-induced plospilcl‘ncsitide 14 . Effect of alumirum on Mgz+-, Ca2+- or an+- 15. 16. 17. mediated inositol phosphate formation in the presence or absence of GI'P[S] ........... . ....................... 75 Respcnse of mztmediated inositol phosphate to almnirumstress ...... 78 11:2" dose-deperrience of GI'P[S]-irduced incsitol [insulate formatlm ....... 80 Idlibitim of aluminnn m bradykinin-trriggered incsitcl [insulate formation and intracellular Ca2+ MWOOOOOOOOOCIO ............. ....- OOOOOOOOOOOOOOOOOO 83 Chapter III 1. 9. Ebrperimentalprocedurecf alumirumuptake ....... ......105 pt! Dependence of aluminumuptake by neurcblastara x118.......OOOOOOOOOOOOOOOOI......OOOOOIOOOOIOOO ..... 111 Time course of superficial birding of aluminum on and incorporatim of aluminum in nalroblastana cells. ..... 114 Dose-effect of alumimm uptake by neuroblastana @180...0.000000000000,00.........OOOOOOOOOOOOOOOOO0.0117 Effect of transferrin on aluminm uptake..............120 Effect of linoleic acid (:1 aluminnn tptake............122 Effects of low mlecular weight chelatirr; agents on alumirun uptake124 Effect of energy metabolisn inhibitor m alminum MOOOOOOOOOOOO......OOOOOOOOO0.00.0.0...0.0.0.....126 Effects of inorganic ions (:1 aluminum uptake” ........ 128 APPENDIX C 1. Depenience of the intracellular free (22+ calcentratim in ejaculated porcine sperm on incubatlaltlme .......... 166 2. Respcnse of the intracellular free (22+ canentraticn in ejaculated porcine sperm, incubated in a physiological medium, to the application of a (22+- specific icnophcre A2318? .............................. 167 APPENDIX D APPENDIX 1. 2. Aluminnn coprecipitatim with [NFC liposaues atvariwspivalues......... .......................... 175 Alumimm coprecipitition with phosphctidylserine- curtainin; liposcmes (80% DIPC + 20% phosphatidylserine) atvariouspHvalues” ..... 175 Time course of alumimm coprecipitation with [MPG intheabsenceorpresenceofcitrate............. ..... 176 Time course of aluminum «precipitation with DIPC liposcmes attwodifferenttarperamres........... ..... 178 E Stqwiseamrtdaangesacrossawbranecaposedof PC/oxidized cholesterol (2:1) inthepresence of mtg/mlofameroteininabafllirgsolutimof l.onNaCl (pH 55)186 Distrihltimofthesizeparameterformprrotein addedtoabilayerlipidmenbraneomprisedof PC/oxidized cholesterol (2:1) at different membrane potartials ........ . ................. 187 'Ihedistrihrtialofthesizeparameterofonly the closing event fran Fig.2 at different transmitter-ans WOOOOOOOOOOOOOI00.0.0.0...00.000.000.000...00.187 Distributimofthesizeparameterfcrthecnpr‘protein in a bilayer lipid medal-ans cmprised of PE/oxidized cholesterol (21)188 Voltagedmadernecfporindmlelclosingfrunmpr‘ proteinsuspendedatpiss in0.1MNaCl oratpH3.5 in 0.1 n MOOOOOOOOOIOOOOOOOOOOOOOI... ........ 00......188 LISTOF'DABIES page clapter II 1. Irhibitim of GI'P[S]-induced inositol phosphate formationbyalumlmlm .............. 54 2 . ugztdcse dqlerdence of aluminum-related inhibition of GI'P[S]-stinulated inositol phosphate production ..... 81 OlapterIII 1. Viability of neurdalastana cells following incubation with alumirum for several hourle? 2. Recovery of alminum by washing procedures............1o9 3 . Intracellular subdistribution of internalized aluminum in nalrdalastana cellslls APPENDIXA 1. Charges in 32p-lanellimg of inositcl pnosmolipids in epididymal sperm frcm marths old mice.............150 APIFNDIXD 1. Altmirum :cfprca'egipitatim with DIPC liposanes in the presel'ceo ..... ..... .........l79 2. Alumirum recovery in the presence of Ca2+.............179 APPENDIXE 1. Sizeofsingle—dxamreleventsatpiSj formpF treasured at differentnatbrane potentials............186 CEAPI‘ERI WW Alminlm Toxicity in Animals Aluminlm toxicity in plants was famd first thiscentury (Hartwell and Barber, 1918). In recent decades aluminum toxicity has drawn enhamed attention, since it creates serious problems in agricultural production and environmental protection. In vast subtrqlic ard trcpic areas, accounting for 40% of arable soils in the world, alumirum toxicity in plants cultivated on acidic soils constitutes a formidable barrier to food and biomass production (Osmond et al., 1980). The ecological inpact of aluminum toxicity hasbeenaggravated by enviramental pollution. Broughtaboutby acid rain causing acidification osoil and water bodies, the incmasingly Mile alumirum icns are in part responsible for forest decline in ampean countries (Stiortle and anith, 1988) and the lossoffishpopulatiminlakesandstreansinCanadaand the northeastern U.S. (Godbold et a1., 1988). As an ecological corsequence,alumimm taken up by plants and aquaticorganisms finallyreadlesthefooddlainforlnmananianimals. Until recently, the significance of aluminlm toxicity in humans had been largely overlooked. Cmtinual an} unavoidable exposure to alumirum, slow accumulation and chronic toxicity by the metal have cartrihltedtothis ilfliffererne (Ganrot, 1986). In the past decade a cmsiderable body of evidence has accunulated inplicating alumirum in various diseases, particularly neurological disordersof tumorsAlumirum toxicity is being recognizedwith increasing frequelcy in patients withrenalfailure, leading to osteanalacia, anemiaorernqilalopathy (Ievine et al., 1990). In these patients, tissue aluminum levels correlated directly with left ventriallar mass arri inversely sith the velocity of cimmfera'rtial fiber shortening (Iadcn at al., 1989). Several studies identified aluminlm as a potential causative factor in the pathology of various types of dementia, especially Alzheimer's syrdrcme, which is a major health problanamongthe elderly (Martynetal., 1989). InBritain, there are anestimated 600,000 pecple afflicted with these diseases, accounting for 5% of people over 65 yearsandZO%ofthcseover80%. Almninumhasbeen found in highcmcentratims on hippocanpal neurons containing smile plaques and neurofibrillary tangles, which occur in the brain of subjects with Alzheimer's syndrane (Roskams modular, 1990). In cell alltmre, 70% of aluminlm-treated l'nman nalroblastcma cellsreactedpositivelywith antibodytotaoproteinandtopaired helical filament, i.e., the Ganges resable those seen in Alzheimer disease brain specimens (Guy at al., 1990). Amyotrtpic lateral sclerosis (Arm-WW cmplex mcuamhasbeenspeallated to originate franagenetiCdisorder. miter, Wisniewski et al. (1980) noticed that acamulation of almintminnalrofilamentsinperikarya and proximal axonssnared sane features with the nanolesions found in patients. later, Hiranoetal. (1984) cuifiruedthatthisabnormalstructure closely reserbled mural alterations in the early stage of ALS. Recently, employing x-ray microanalysis, Hirsch et al., (1991) identified increased alumirum accumulation in the substantia nigra of patients with Parkinsal's disease. Ecperinaltally, aluminum intoxication in nalrological disorders hasbeenwell establishedinanimalnndels. Intracr‘anial administratim of aluminum to animals produced a progressive encepialqaathy with neurofibrillary degeneration of intermediate filaments (Selkoe et al., 1979). Abnormal nalronal axonal transport of nalrofilament proteirs has been reported in almnimm—intmricated ratioits ('rr'aicoso et al., 1985). This inpaired transport is caused by allmlinum-induced formation of protease-resistant high molecular weight oatpleices fran neurofilamant protein (Nixon, et al., 1990) . Since cartact and absorption of aluminiml is unvoidable, everyday exposure to aluminnn unlikely leads to pathological events ofhlnanneirodisorders. Itispostulatedthat the onset ofthese narrodisorders may involve, besides aluminum itself, intrinsic factors like genetic lesims, aging and biological agents (Liss et al., 1989). Alumirum narrotoodcity is rarely expressed in a normal organism itooalrsonly when thenormalprotectivemechanismsare inpeired or altered by intrinsic or extrinsic factors. Cranial Prqlerties of Aluminum Asthehardesttrivalentmetal elementinthenatnlre, aluminum has high imic diarge and snall crystalline radius. Its high ratio of diarge-ovwradils (zz/r = 43.6C2 111'1 x 1028) yields a reactivity urnatdled by other soluble metals (Parker et al., 1989) . 'Iherefore alumimm does not exist as a positive aquo ion, rather, the nmhydrated aluminum ionhas a great tendency to polarize adjacent atans, in particular ligands which contain snall, hard negative oxygen donor groups (Martell and Motekaitis, 1987) . In the aqueals solutial, aluminlm ion strongly polarizes 0-H bonds of the water rnolecule, resulting in the dissociation of a proton fran water mlecule. With a small ionic volume and coordiration number of 6,thehydrated aluminum ion is coordinated in its primary hydration shell by six water molecules in an octahedral ccnfiguratial (Nordstran and May, 1989) , represented by A1050) 63+. Because of the high positive charge of aluminum im, these water rolecules- form a tightly bound primary hydration shell. 'lhe solvation of the aluminum ion largely depends on its ca'centratimandthesolutim pa. Free 1113+ is thepredaninant species at low w (< 5.2) solutim. As the solution pH increases, the initially hydrated 1113+ urdergoes stepwise hydrolysis, progressively losing its hydration shell protm to water molecules to mintain dissociatial equilibrium, thus doubly and singly charged munlclear species are formed: (Bees and Hester, 1976) M0120)63+ + H20 == Al(l'120)5((1i)2+ + 1130’r 111(1120)?’2+ + 1120 = 1.10120) 4(on) 2* + 1130+ 1110120) 42+ 4- H20 = Ala-120)3(a-I)3 + 1130+ Further [it increase in solution will lead totheformationof naltral and negatively charged species. Alumimm dunistry is further calplicated by a variety of calplexatim reactions. It tends to form electrostatic bonds preferentially with oxygen donor ligands. In biological systems, carboxylate ard phosphate groups, inorganic phosphate, nucleotides, ard polyrucleotides meet this remirenent (Martin, 1986). Therefore, mrboxyl ootyanions of proteins and phosphate oxyanions of lipidsprovidetargets for aluminum birding. Sane polyhydroxy acid metabolites in biological systans like malate, tartrate and particularly citrate, are tell-known alumimm chelators (thrtell andMotekaitis, 1987). Incitrate, carboxylategruipsarearrarged inamannerfavorablefor very strong chelation of thealuminum im. An inportant feature of alumimm iansistheslowrate of ligard embargo in and out ofthecoordination sphere. Ligand mange rate for 113+ is lOS-fold slower than that for 192+ (Martin, 1986). 'lhis feature takes at special inportance for aluminum touticity,because the low ligard exchange rate makes aluminlmuseless as a metalatactivesitesonproteinsardother cellularompments.'lhis low rate ispartiallyduetothestrict coordinate umber 6 foralmninnn.'nleligard exchange generally occurs by a dissociative mechanism (alrgess, 1978), and the fanatic) of a dissociative intermediate often requires a coordination umber higher than 6. With the cmplexity of solutim and coordination chemistry, relatimships between aluminum stress and biological responses have mtyetbeenwell established, e.g., the identity of aluminum's toxic species. Anong various aluminum species in solution, free ion Al3+ is generally believed to be the active form causing many aluminlm effects (Zhang and Oolmbili, 1989), but aluminum toxicity sanetimes can be ascribed to Al(OH)2+, nuanz", or the sum of all maniclear species activities (Parker et al., 1989). Under certain caditiore, polynlclear hydroxy-alumimnn is highly toxic (Wagatsma and Kaneko, 1987). Furthermore, thepossibility that certain alumimm-carplexing ligarris may be responsible for alumirum’s adverse effects cannot be excluded. An exanple is given by alumirum-fluoride caplet, which interferes with cellular signal tramductial (Chabre, 1990). Primary Cellular lesions for Aluminum Toxicity For plants, alminlm toxicity is a soilborne problen, and the majortoxic reactim appears (:1 therootsurface (Akesonetal., 1989), unrealmninlm is takenup. Since the syuptms of aluminum stress in plants are those characteristic of deficiencies of several physiologically inportant ions including calcium, magnesium and phosphorus , it is generally agreed thatalmnirnnninjures plantsnainlybyinterferingwiththemetabolian ofthoseessential elanaits (Pay at al., 1978). In lumen and animals, the mechanism ofaluminum toxicity ranaire nuch nore obscure. Atthecellular level, a variety of organelles like chromatin (Walker et al., 1988), cytoskeleton (Oteizaetal., 1989), mitochondria (Dilletal.,1987) andplasma netbrane (Weis arri Hang, 198 )havebeen listed as potential primary injury site(s) . At the molecular level, nucleic acids (Karlik et al., 1980), filospholipids (Deleers et al., 1986), polysaccharides (Moreno et al., 1985), and proteins are reported to be vulnerable to alumimm intoxication. In in vitro experiments, aluminum inhibits various enzymes. In these cases, alumirummay iiilibit enzyme through its substitution for physiological factors such as 1192+, or tlmcugh its allosteric interaction with enzyme protein. (Macdonald and Martin, 1988). The firsthnownexanpleishemokimsetthebrainismerismre sensitive to alumimm relative to nuscle isaners. ‘Ihe aluminum- -induced inpairment of cerebral glucose utilizaticn affects particularly the metabolism of acetyldloline and other narrotransmitters, which is typical of diseases associated clinically with dementia (Iai and Bless, 1984) . In the following sections, we will briefly review sane current theories regardirg the primary cellular lesions of aluminum. 1. Alumirum Interaction with Chrmatin Aluminlmisreportedtohave high affinity to nucleic acid polymers, 111A ard RNA. Besides phosphate oxygen, heterocyclic nitrogenarrieimcycliccarbalylmpmineardpyrimidine basesalso provide potential binding sites, despite lower affinity for aluminlmtmrlik et al., 1980). Investigations show chranatin being aeofthecellstructmres most vulnerable to almninum's action. In experimental aluminum encephalopathies, aluminum acomnllates rapidly upon INA containing structmes in the nucleus. and Al/mA-P ratio in dnronatin of the animalwasfoundashighasl.05% by weight (Crapper et al., 1980). 'Iheinteractionofaluminlm with genetic machinery has not been investigated in depth. According to very limited information, at the replication level, [NA synthesis in osteoblast-like cells is substantially ixhibited by micronolar almninum, which in turnn disturbs bone cell proliferation and differentiation (Kasai et .al. , 1991,). ' At the transcription level, the aluminum-induced codensation and aggregation of chronatin may prevent the formation ardmaintenance of a transcriptionallycoipetentopenstructures, an inportant mechanism controlling gene expression, which allows INA polymerase access to coding region of 114A. Walker et a1. (1989) examined divalent ard trivalent cations in terns of their ability tocausethecoipactionof dnronatin fronratbrainandliver, and found that aluminum, with its large ionic index zz/r and high covalent irriex, was the most reactive. Quotatin treated with alumirum in vivoandin vitro is less sersitive to mass II digestion (Matamnto, 1988) , and ooincidently Alzheimer chronatin wasfoundtobelessaccessibletodigestionbymicrocoocal nuclease. Ehploying a quantitative dot blot method, results stunned that aluminmitdeedhadspecificdepressiononscnemessenger RNA levels, which might belinkedtothedirecteffectofaluminumon the transcription of genes. In altnnimm-treated rabbit brain, level ofcalnodulinnE'NAwasreduced by 60-70%, ardthesamenagnioxie 10 of reduction of these messengers was found in Alzheimer's rabbit brain (Cramer Macladnlan, 1989). line actual birding sites of almninum onchronatinarenot known. One assunption is that aluminummightbind between the linker histones ard INA, basedonthe fact that dinncleosones releasedfr'onaluminm-treated cerebral cortexcontainedtwicethe content oflinkerhistonemthanthatfrm the coitrol (Cramer mladflan, 1989). A potential site for aluminum’s bridge may be provided by coordination at amino acid asp-98 and glu-99 of H]. and [laymen-late. ByanchorinnglinkerhistoneonlNA, aluminum may prevent the gene expressionwhidnonlyooolrsonopaneudnrmatin regions where H]. linker histones are depleted. Alternatively, almninlmuayresideon a site mummnidnvmldbedismptiveto the birdirg of associatirg cationic proteins (Record et al., 1978). In addition, alumimm may act as counterion to finysiologiczl cations, such as Mg”, Ca2+ or Zn”, required in gene expression, diqnlacing them fron their normal binding sites on dnmtin (Garnot, 1986). 2. Effect of Aluminum on Cellular Calcium Metabolism calcium serves as a second messengerinbioregulation via variols intracellular calcium trigger proteins. In response toa large variety of external stimli, intracellular calcium transients are generated. Within the lifetime of these transients, (22+ binds to trigger proteins (signal input), causing conformational 11 dnanges. 'nnese changes play a key role in signal anplification annd transmission (signal output) fron the trigger protein to respective effector enzymes annd structural elements. When (22+ coupled signnal transduction is interrupted, severe repercussions on biological annd pnysiologiml processes are expected to occur. As an inportant biological messennger, Ca2+ is subjected to nultiple cellular controls. Most cells have the capacity of regulating intracellular (22+ over a broad rannge by using diverse transport systems in endoplasmic reticulum, plasma membrane annd mitochoriria, aswellasthrmghcaz” binding proteins (Patneyet al., 1989). Free (32+ concentration incytosolthusreflects a balance between influx, efflux, and intracellular exchange and redistribution (rogin et al., 1987). It has beenn known thataluminum stress on plannts always results in an interference with cellular (22+ metabolism (Zhao et al., 1987). Inrecentyears, increasingevidencehas emerged that (22+ regulation may alsobeatargetforaluminmnintoxicationin animl cells. Application of aluminum to csteoblast-like cells inhibits (22+ accumulation in the cell matrix, which may underl ie the develqzment of aluminum-induced osteonalacia in certain patients (Ikeda et al., 1986). In laboratory rats, aluminum overload decreases sacrrplasmic reticulum (22+ transport (Ievine et al., 1990). Perturbation offreecn2+ transients by aluminum was also reported in pnenylefinrine-stinnlated hepatccytes (Sonofl et al., 1990). One possible way by which aluminum nanipulates intracellular (22+ metabolism is involved in the interaction of aluminum with intracellular Ca2+ regulator proteins. A well knncwn example is calmdulin, which mediates a nultitude of Ca2+-depehdent bicchonicnl processes (Siegel anniHaug, 1983). Calmdulin has a profomdto'denncy to bird 4 (22+ in specific locikncwnasEF hands. As first two on2+ bind to the high-affinity birding sites III and IV,theprotein undergoes conformational changes which exposeahydrophobicregionservingasaninnterfaceforthe interaction between calmcdulin and target pmteins.Almninmmis able to binnd to calmnddin stoichionetrically at a mlar ratio of 3:1. then hand on the protein, aluminum triggersahelix-coil transition conconitant with an increase in tcpograrhic surface (Yuan annd Haug, 1988). ‘Iheinduced geonetric rearrangement of calmdulin severely antagonizes its activity to stimlate effector proteinslikecalmcdulin-dependenntprotein kinase which, inturn, controls Ca2+ channels onsarcoplasmic reticulum (Gasseret al., 1988). In a newly-proposed approach, aluminum is suggested to domregulate intracellular (22” through pbsooimsitide signal transduction (Birdnall and dnappell, 1988), in which second unssenyer molecule Ins(l,4,5)P3, I133, is generated to signal (32" release fron intracellular stores (Berridge, 1983) . Moreover, IP3 , together with its phosphorylation product Ins(1,3,4,5)P4, presumably controls the entry of external 13 Ca2+ throlgh seco'd nnessengerhcperated dnannnels on the plasma nenbrane (Berridge and Irvine, 1989). This nednanism will be discussed in detail in later sections. 3. Substitution of aluminum for MQZ+ A prevailing hypothesis regarding aluminum intmniontion anfinasizes the element's substitution for magnesium bound on crucial cellular coxponennts (Kraal et al., 1990). As a regulator in various biodnemical processes, divalent cation Hg” is required forthefunstiosofnmnermsenzymesandthemainntenanceof dnronatin conformation (Wadrer 1980). A13+/ug2+ substitution theory has its solid enenical mnderpimning. Itiskncmthat size similarity, rather thancharge identity, plays a key role inpermittingmetal ion substitution (Garnet, 1986) . 'Ihe ionic radii of A13+ most closely resemble thoseof n32". Insixfold coordinnation, the radius isO.54Afor m3+ and 0.72 A for 1432+ aecoonnald and Martin, 1988), respectively. With a slightly smaller ionic volume but nuch stronger ionic index, free A13+ ion has onhanced association constants with many ligands, and is able to conpete effectively for 1132+ on binnding sites in biological systole, even at M32" molar concentrations 10"8 fold higher than A13+ (Miller et a1. , 1989). In nanny cases of aluminum-mediated enzyme inhibition, replacanennt of aluminum for M32“ is inplicated (MacDonald and 14 Martin, 1988). 'Be lgztmediated enzymes vulnerable to aluminum innclude those involved in energy netabolism like hexokinnase, glucose-o-ptncsphate dehydrogenase (Ono and Joshi, 1989), those involved in pnospnate transfer reactios like 3',5’-cyclic nnucleotide rincsphcdiesterase, acidic and alkaline pncsphatases adenylate cyclase, and several carboxyl acid esterases like acetylcholine esterase (Maoionald and Martin, 1988). 'Ihe fundamental bicchonical lesion effected by Al3+/M32+ substitution is illustrated by alumimnm’s interaction with tubulin (Macdonald et al, 1987). Hf", as tne physiological mediator for assonblyofulbulin, isthcughttobindat theexchangeableguanine nucleotide (61‘? or GDP) binding site (B site) of tie protein. After the polymerization of MEN-GT? bound tubulin m innto microtubules, tre bound (:1? ishydrolyzedtoGDPanddissociated fronthetubulin, leadingtothe nextcycleofmicrotubule assonbly. Aluminum ion Al3+ at suonanonolar concentration conpetes effectively for E site with ugz“ at l mu. Because of the extremely slow ligand exchange rate of Al3+, the hydrolysis and dissociation of are is inhibited on Al3+-bound GI‘P-tubulin, leading to aberrant microtubule assonbly and disassonbly. 4. Perturbation of Aluminumon PlasmaMenbrane Plasnna mennbranm manifest aluminum toxicity in two ways: servirgasaprimary lesionsiteoras acontroloftleaccessof 15 aluminum to intracellular targets. In eitler case, interaction of alnminlmwithplasnamennbranenmldbean initial stage inaluminum Manyconponentsofplasnamanbranenay serve as targets for aluminum, but polar phosfinolipics are likely the prine candidates. 'Ihe negatively charged noseblipids, prosphatidylserine (p5) and [inspetidylincsitol (PI) especially have high affinities for aluminum binding, but they aremainly located on the cytoplasmic side of the plasma nenbrane. an the extracellular side, polar head regios of zwitterionic peelelipids, e.g. , phosphatidylcholine (PC) and rincqnhatidylethanolamine (PE) , are attractive ligands for alnminm binding. Such binding my cause drastical dnanges in mbreesurfacednargedesityandtransnenbrane potential (Akeson et al., 1989). With this electrostatic interaction, alnminnm was reportedto ininihitvoltagegatingofVAIxzdnannelonmitodnodrial outer Mbrane by neutralizing the channel's sesor responsible for voltage dependence (Dill et al., 1987) . Alnminnm crosslinking of the polar regions at the memorane surfacewasfonndtobetranslateddeeplyintotneinntemal nopolar regios, inducing a membrane phase separation, aggregation andnanbrane flsion (Deleers et al., 1985, Deleere et al., 1986). In runan erythrocytes, tte association of alnminm with plasma manhranesresults in an inncreasedlipidorderparameterandphase transition temperature, indicative of more rigid lipid packing (WeisandHaLg, 1989). anch inpcsedonanges inmenbranestructure and physical prcperties will influence mmrane functios like 16 permeability. Zhaoetal.,(l987) demonstrated thattheperturbation of lulk lipid matrix, upon aluminum application, led to enlnannced non-electrolyte transport across plasma membranes. In addition, the activity of sone membrane-bound proteins might be also modified by the aluminum-induced dnanges in a lipid environment, particularly intheboundarylipid area. For instance, inhibition of nnenbrane- bondK“-A'1‘Pasebyalnmin1mwasfoundtocorrelatewiththe alnminnnm-induced decrease in mennbrane fluidity (Suhayda andHaug, 1986). Aninterestirgresearoharea is the effect of aluminum on menbrane lipid peroxidation, which is believed to be a culprit (reusing cellular aging. In vivo, tre production of 2-thiobarbituric acidreactive substances('I'BARS)wasenhancedinbrainandliverof miceafterdietary aluminm intoxication (Fragaetal., 1990). In in in vitro experiments, alnminm was sham to facilitate iron-dependent peocidationinerythrocytemenmraneandliver microsones (minnlan et al., 1988). me to its electronic configuration, alnminnm isnot able to innteract directly with oxidative free radicals, therefore the observed acceleration of alnminm on lipid peroddation more likely results fron its interaction with lipid substrates. One interpretation is that bindingofalnminmonmenbranemayeausearearrangonentof membrane piospnolipid molecules, which reders lipids more accessibletotheattackof free radicals (Fragaet al., 1990). 17 Aluminum Uptake and Intracellular Distribution Since aluminum amarently does not have physiological importance, it is unlikely that there are specific cellular transport device for its entry. A major criticism of tie aluminum hypothesis in neurodegenerative disorders is the postulated relative inaccessibility of the elenent. War, abnormally high accumlationofaluminumin the CNS argues against thisnotion. Because CNS nneural cells are terminally differentiated, the altminumtransported to these cells will be accunulatedunless specific systems are available to remove then. This is different fronothertissueswhichhaveasetturncverrate,andthus altminmaccunulation overtimewouldbelessprofound (Roslamaand Ch‘nncr, 1990).. In posumrten brain sanples from patients with dialysis encqhalqnathy, alnminumconcentration ranges fron 100 - 800m (Garnet, 1986). In ecperimental intoxication, total brain aluminum concentrationaverage abontloomandnayreadnthOuM. Becauseof its nonuniform accumulation in different nalros, in vivo levels of alnminuminscneneuronalpcpulatios maygreatlyexceedthelOOuM average (Nixon etal., 1990). Ilumanneurdnlastcna cells (DIR-32) wreshowntoaccunulate 10-20 malnminumagainstloomalnminum in the medinm (Guy et al., 1990). 'me inntracellular aluminum concentratios withinthisrange were found in tangle-bearing 18 nenros of alananian patients with amyotropic lateral sclerosis or Parkinson’s disease (Perl and Good, 1987). Raw does altminm enter manmalian cells? With limited experimental data, it is geerally believed that mononuclear forms of alnminm are reqnosible for its transport across the plasma menbrane.'nehydroearboninnteriorof tte lipid medarane is not permeable to these alnminum ios, buttte ioncphor'etic capacity of Melipids may provide a means to translccate alnminum, e.g., by forming non-bilayer configuratios like reversed vehicles or “II“ no illustrate, nil3+ adsorption by nospniatidyldioline was reportedlyinvolved in alnminnm uptake intocytoplasm (Akesonet al., 1939). In living cells, alnminum ions especially A13+ is likely able to take advantage of co—transport or nonspecific transport. machineries for metal catios. In seardn of possible carriers for cellular uptake of alnminum, investigatioshavebeenfconssedontwo typesof biological molecules: Wrting protein, i.e., transferrin, and low molecular weight metabolites with capacity of delating alnminum. Based on dnenicnl study, the snall alnminum-delating metabolites like citrate lnave been suggested to provide an effective means for aluminum transport innto cells (Martin, 1986). W, in most experiments, these little alnminm-delating agentsvirtuallyprotectedanimls and their cells bypreventing alnminm internalization (Doningo et al., 1988, Guy et al., 1990). On the other bands, transferrin has been inplicated in facilitating aluminmuptake by various malian cells includingluman 19 nelrdJlastona cells (Harris et al., 1987) and cells in CNS (Rosskam and Connor, 1990), presnmably through a transferrin receptor-mediated process. Since it probably cannnnot be used physiologically, intracellular alnminum will possibly be bound on sites that are stronger than cytosolic pool chelators. At slightly acidic inntracellular pi, at which cationic alnminum species are present, the prine candidates are menbrane lipcphosninates and phosphorylated proteins (Birdnall and Chappell, 1988). 'He existence ofsudn stro'g intracellular delating pools for' aluminum can partially explain how mannnnalian cells are able to establish high aluminum contents against low extracellular aluminum concenntration. ansphoinesitide Signal Transduction Hesnnoinncsitide signal transduction is a ubiquitous secod messe'ger systeninenkaryoticcellstoregulatealarge array of cellular processes incltding metabolism, secretion, contraction, nnenral activity and cell proliferation (Berridge and Irvine, 1989) . In analogy with cAMP secod nnnessenger systempncsninoincsitide signalling pathway cosists of three conponents, viz., receptors, guanine nucleotide binding protein (Gp protein) and phospholipase C, residingintteplaenamenbrane. In neuronalcells, receptorson the cell surface detectextracellular stimuli like hornnones or nelrotransmitters,and transfer the signals to Gp protein.'1he activated G protein in turn triggers effector enzyme [incspholipase 20 c, and the later cleaves nospniatidylinositol 4,5-biphosphate (PtdIns(4,5)P2) into inncsitol 1,4,5 ~triphosoiate (1P3) and diacylglycerol , which nobilizes intracellular ca?” or activates protein kinase C, respectively (Berridge, 1987). Inthelastdecade, a number of signal-nediatinngroteins have been identified by the combination of Classical bicdnenistry andINA reconbinationtedmiqte.'nneyarealll‘eterotriueric molecules, cosistingofadistinetatsuhmitandidentialp and “Y subunits (Bimbaumer et al., 1990). 0L Subunnit contains a high-affinity guanine nucleotide binding site and at least oe high-affinity Mg?“ binding site. So far nine genes encoding a subunnits havebeenidentified, andlZpolypeptideproductsofthese geesareknewn to be inplicatedinntheactivities (heissnnnthet al., 1989). 'nemlecular process of G protein activation is brieflyasfollowing: after receiving tte signal fron tl'ecell surface receptor, the G protein mdergoesconformationaldnange which facilitates the exchange of bond GDP for 61?, followed by dissociation of on submit fron $1 subunits. Afterthis dissociation, tie Md (1 sutunitbecones functional, its catalyzing center on tl'e sutunit activates Melipase C. Cn'eonitantly, GI'Pase activity of anbmit hydrolyses the bound GTP, leadingto association of a and 51 submits and inactivation of Gp protein (Freissnuth et al., 1989). Gproteinwasoriginally believedtobearasoncogeeproduct (Wakelam et al., 1986), sinnce enpression of pZIN’raS in 3T3 cells ‘rosqodingtostimilusbygrowthfactors ledtoalargeincrease in 21 incsitol phosphate formation. Now it is known thatthereisa subset of "little" runner GI'P binding proteins with molecular masses 20-25 km. 'nney inncludemanyfactors controlling protein synthesis like the elongation factor, EF-Tu, and ras, rho, ral and racgeeproducts (Nozawaetal., 1991). There is a nultiplicity of phcqinolipases C (PIC) in mannmalian tissues (Rhee et al., 1989). FolrtypesofPICisoners, assigned as a, p, TandJ, have beenidentified. Incultured neuroblastoma cells, tie najor isoner is PICn, accounting for 99% of total PIC activity. CloningofthefcurtypesofPICisoenzymes has revealed a surprisingly low degree of similarity in treir primary structure, suggesting different roles and regulatory properties (Vicenti and attaneo, 1991). 'neactivities of these PLCisonersaredistinct in treir respective comartnentation, substrate specificity and ligand regiireneit. nhe soluble phospholipa'se c apparently uses onlyPIassubstrateandisnotabletoreadntlehome- sesitive pespnoinositide pool residing in plasma membrane, while the Q protein-mediatedmenbrane—bondPICactivityisseeningly nore specific to agonist-sesitive PIP and PIP2 (Codtcroft, 1987). line PLC activity can also be activated directly by intracellular (22+ of increasing concentratios, bypassing receptorstimlation and Gp protein mediation (CradleandCrevs, 1990). Resphoinositide hydrolysis is followed by the extremely conplicated inncsitol pnioeniate netabolism. In a bewildering array of Merylation and dephcsnrmylation reactios, dozes of 22 isoners of incsitol nono-, bi-, tri-, tetraki-, penta-, Winesaswellascyclic derivatives are produced (Putney et al., 1989). Among then, only Ins(l,4,5)P3 is firmly established as the primary signal to inntracellular Ca2+ nebilization. It innteracts with receptors on the edoplasmic retionllm and sarccplasmic reticulum, and triggers the opening of ca2+ channels on theseorganellarmedoranes, resulting in ea2+ efflux fron these intracellular stores. 'Ihe depletion of IP3-sesitive intracellular Ca2+ pools, in turn, signals Ca2+ entry mednanisun, and extracellular <22" isallowedtoenter to refill tie depleted Ca2+ pools (Berridge and Irvine, 1989) . Activation of Fluoroalnminate on Phosphoinositide Signal Transduction Fluoride anios (F') havelog been know to stinulate G protein-calmed transmenbrane signalling. 'Ihis activation process was later revealed tobedepedentontraceaneunts of alnminum (Stermveis andGilman, 1982). In fluoride solutios, alnminum is able to form various soluble ionic conplexes AlFxx'B (wlnere x = 1-6) whose stoichionetry depeds ontheexcessconcentrationof fluoride. 'nnese alnminum-fluoride conplees, nest likely A1F4', areprcposedtostinnulate G protein byactingasanalcguesofthe terminal pesphate of are (Fain et al., 1988). 23 On the atouic and molecular bases, there are close similarities between allminofluoride conplex AlF4' and phosphate group 1303': lere fluorine conpares to oxygen, and aluminum to desphorus. With tie sane size and valence orbitals as oxygen, fluorine has also the very strog electroegativity and this agreatcapacityofforminghydrcgen bonds. Fluorine in an ionicconplex teds to bind to ahydrogenboddonorgroupona protein. Moreover, anO—H..F isjust slightly loger thanO-H..O, and N-H..F and N-H..0 areexactlyinthesamebodlegth. Like phosphorus, alnminumhascoordinate number of 1 - 6,dletothe possible hybridization of its olter shell 3p electrons with 3d orbitals. mrtlernere, Al-Fbodinalnmineflucride has the sane legthasaP-Obo'dinpncsphate. Because of these resenblances, Q protein may erroneously take alumineflucride as {resonate group. Accordingto Chabre’s nedel (Chabre, 1990), inaG'I‘P-binding protein whose nucleotide site already contains a GDP, presumably AlF4' is tetrahedrally bound tofolrflneridesndnionarehydrogen-bonndontleprotein at ornear tie nucleotide site, thenn AlF4' exchanges oe ofits fluorides by binding ionicallytotheterminaloxygenof -phcsphateoftlebound GIP. Homver, unnlike r-pncsphateofGIP, tie alnminefluoride bound to tie GDP can not form the pentacoordinatedbipyramidal structure which is requiredforGTP hydrolysis, thusit locksthecatalyticcenteroprrotein in tre activated tetrahedral configuration. Inotterwords, uponbinding of allminofluoride, G proteins are blocked at tie active state as 24 with bound nun-hydrolyzable GI‘P analogue GI‘P[S]. The similar structure changes in G proteins modified by allminofluoride or by GTP[S] were recasidered (Higashijima et al., 1987) Objectives: Possible Inpact of Aluminum on Phosphoinositide Signal Transductian and Cellular uptake of Aluminum. Takentogetler, although knowledge has been builtregarding tiedevelqmentofalnmimmtmdcitysynflraneinlmnan andanimals, scant informatian is available an cellular and mlecnlar nechanisnns associated with altminm inntoodcationn. Fmdannenntal questions that nust be resolved are those of alnminm—caused primary cellular lesim(s) anti altminm uptake across the plasnna manbrane. The possible impact of aluminm an ginospnoinositide signnal transduction might berelated to alnminm-caused neurodiorders. Furthermore, basedin partanfinndinqs(ShiandHaug,1988)ino\mlaboratorywe aretherefore advancing tre hypothesis that altminnmiuptakeby intactcellsisdepenflentanneditmwandmayalsobe interriorizedbycertain types of carrier—nediatednednanisns.'ro testtlesehypotlesestmposetopursetlefollmringobjectives: OBJECTIVE I: Determine tte effect of alnminnm on inositol phosphate fonatian in Meinositide signal transduction pathway. 25 OBJECTIVE II: Determine alnminnm uptake by viable neuroblastoma cells, by varying medinm {fl and enploying potential carriers. Allexperinentswillbecarriedoutanmn-differenrtial nurine neurdnlastana cells, C1300, clue neuro—ZA. RATIONAL as to cbjective I: 1). Alnminm toxicity has been inplicated in a broad spectrum of finysiological disorders. 'Ihis led to a proposal that aluminnm’s toxicity is a nultigene-cantrolled syndrcme. The ability of alnminm to bind raspecifically to different cellular canponents offers tie element omorumities to affect various cellular reactias (Ganrot, 1986). I‘m, it is also possible that a group of these disorders originates frrm a single primary cellular lesian: alnminminrteracts with a keytargetmidniscritical in cellular netabolisn. Wet, at present, amear to be pleiotropic altmimmeffects innanlnaliancells, mayintteendprovetoanerge frcm a basic, underlying molecular nedlanism. Phospnoinositide signaltransductian could be aeofsuchtargets, becauseitisa ubiquitalssystaninnalkaryoticcells to regulate nanny inportant cellular processes. 2). A nejor consequence of alnminm intoxication is the interference with cellular (22+ metabolism, which is kncm to be partially caholled by ghosphoinositide signal transduction. Altminm generally permrbs intracellular (22+ netabolisnn in an 26 inhibitory nanner. This .fact suggests that, apart frun the activation node in tte presence of fluoride, aluminum may dcunregulate intracellular <22" through signal transduction: with a distinctnechanisn. 3). Ebcperimenntally, altminnmhasbeenreportedto inhibit a nunber of signal-mediating G proteins like transducin in retinal rod cuter segments (Miller et al., 1989), and Gs/Giin cAMP signalling pathway (mnsour et al., 1983, Johnson, 1988,). 4). A principalnednanismtcxicity expressionisbelieved to be substitution by aluminum for ugz“ at critical cellular site(s), and ng+ is a physiological ligand in GP protein- coupled rinospnoinositide signnalling process. 'Ite replacement of n1g2+hy altminm at the nucleotide bindingcenteronGpprotein may result in tle inactivation of Gp protein. 5). Alnminm is expected tobebounrion vicinnal phosphate grams an phosfinoinositide in particular PIP2 (Birchall and enamel, 1989) , leading to a depletion of hydrolyzable substrate pools. Alternately, alnminnm may binnd to I133 (Schofl et al., 1990), thereby altering IP3 netabolic fate and/or 1P3 binding affinity to its receptor. 6). Finally, all key cmporents of phospnoincsitide signnal transductian, i.e., receptors, Gpprotein and ninospholipase C, areresidinginplasmamenbr'ane.'nerefore, tl‘e perhnrbanceoftle nenbrarepropertiesbyalnminnmnayinriirectlyaffect tle phospho- inositide metabolism 27 LIST'OFIREFERENCES Akeean M.A., and Munns D.N. (1989) Lipid bilayer perneation by neutral citrate and by three hydroxy carboxylic acids. Biochim. Biophys. Acta. 984, 200-206. Akesm 14.15., Minus D.N. and nnrau R.G. (1989) Adsorptian of Al3+ to Metidyldnolile vesicles. Biochilu. Biophys. Acta. 986, 33-40. Baes C.F. annd Mesner R.E. (1976) 'Be Hydrolysis of Cations, pp. 112-113. J. Wiley & Sas, New York. Berridge M.J., mwsanR.M.C., Domes C.P. and Irvine R.F. (1983) Changesin tlelevelsofinositolrhosmatesafteragonist -depadent hydrolysis of meter-are phosrhoinositides. Biochem. J. 212, 473-482. Ber-ridge H.J. (1987) Inesitol triphosphate annd diacylglycerol: two interacting secad messengers. Annu. Rev. Biochem. 56, 159-193. Berridge NJ. and Irvine R.F. (1989) Inositol ninosphates annd cell 8mm. mm. 341, 197.205. Birdnall J.D and dnappell J.S. (1988) Altminium, dnemical physiology, andAlzheiner’s disease. 'ne Iancet, Oct. 29, ii: 1008-1010. Birnbauner L., Abranowitz J. and Brown Add. (1990) Receptor- effector coupling by G protein. Biodnim. Biophys. Acta. 1031, 163-224. alrgess, (1978) Chapter II, Kinetics and mednanisns: solvent exchange. In Metal ICI'IS in Solutian, pp 310-348 Chabre M. (1990) Aluminefluoride and beryllofluoride canplexes: nan phoqinate analogs in enzymology. TIE. 15, 6-10 Ono S.W. andJoshi J.G. (1989) Tine-dependent inactivation of gltncose-G-unosphate dehydrogenase from yeast by alunninum. 'onicol. ILett. 47, 215-219. Chchran 11., (hates J. and Nedn S. (1884) 'Be cmpetitive equilibrinm between aluminum and ferric ice for tie binnding sites of transferrin. FEB Lett. 176, 129-132. 28 Cockcroft S. (1987) Polyphosdnoinositide pnosphodiesterase: regulationby a novel guanine nucleotide binding protein, Gp. Trends Biodnan. Sci. 2, 75-78. Cottan F.A. and Wilkinsan G. (1962) In Advanced Innorganic Chanistry. John Wiley & Sans. Cradle L.J. and Crews F.T. (1990) Calcinm- versus Gprotein- uediated rinosrhoinnositide hydrolysis in rat cerebral cortical synaptcreurosanes. J. Neurochem. 55, 1022-1030. Crapper mladnlan D.R. (1989) Allminum neurotoxicity: criteria for assigning a role in Alzneiner's disease, in Environmental Chanistry and Toxicology of Aluminum (Lewis T.E. ed), pp 299-315. Levis Publ., Boca Raton, FL. Crapper D.R., Quittkat S., Krishman S.S., Dalton A.J. and Debcni U. (1980) Intranuclear altminumcancert inAlzheiner’s disease, dialysis encefinalopathy, and experinental alnminum encepnalopathy. Acta Neuropathol. 50, 19-24. Deleers M., Servais J-P. and Wulfert E. (1985) Micrauolar cancentratian of Al3+ induced phase separation, aggregation and dye release in [inosphatidylserine-conrtaining vesicles. Biochim. Biqinys.Acta.813, 195-200. Deleers M., Servais J-P. and Wulfert E. (1986) Nelrotoxic cations induce manbrane rigidificatian and mendorane fnsian at micruuolar cancenntraticrs. Biochim. Biqnhys. Acta. 855, 271-276. Dill E.T., Holden M.J. and Colanbini M. (1987) Voltage gating in vmc is mrkedly inhibited by micronolar quantities of alminum. J. Meier-are Biol. 99, 187-196. Daningo J.L., Gates M., and Liobet J.M. (1988) Citric, malic and succinic acids as possible alternatives to deferrounamine in altminum toxicity. Clin. 'onicol. 26, 67-79. Fain J.N., Wallace M.A. andWojcikiepnicz R.J.H. (1988) ENidence for involvenent of guanine nucleotide-binding regulatory proteins in tl'e activatian of nhosdnolipases by homes. FASEBJ. 2, 2569-2574. C.G., Oteiza P.I., Golub M.S., Gershwin M.E. and Keen C.L. (1990) Effects of alnminum on brain lipid peroxidationn. Toxicol. Iett. 51, 213-219. Freissnlth 14., Casey, P.J. and Gilman A.G. (1989) G protein control diverse pathway of transmenbrane signaling. FASEB J. 3, 2125-2131. 29 Pay C.C., Chaney R.L. and finite M.C. (1978) line finysiology of netal toxicity in plants. Anun. Rev. Plant Physiol. 29, 511-566. Ganrot P.O. (1986) Metabolism and possible health effects of allminum. Envircxun. Health Perspect. 65, 363-441. Gasser J., Paganetti P. Carafoli E. and dniesi. (1988) Heterogeneas distribntion of calmdulin- and clam-dependent regulatian of Ca + uptake in cardiac sarooplaenic reticulum anbfractias. Eur. J. Biochen. 176, 535. Godbold D.L., Fritz E. and I-hnttermann A. (1988) Alrminum toxicity and forest decline. Proc. Natl. Acad. Sci. USA. 85, 3888-3892. any S.P. Seabright P. and Itzhaki R.F. (1990) Uptake of altminlm by l'uman nnenroblastaua cells. J. Trace Elan. Electrolytes Health Dis. 4, 183-187. Barbell B.L. and Father F.R. (1918) 'Ihe presence of alnminum as a reasan for tte differece in tie effect of so-called acid soil on barley and rye. Soil Sci. 6, 259-279. I-ligashijima T., Ferguson K.M., Sternweis P.C., RossE.M., anigel M.D. and Gilman A.G. (1987) 'De effect of activating ligands on tre intrinsic fluorescence of guanine nucleotide-binding regulatory proteins. J. Biol. Gm. 262, 752-756. Hiram. A., Danenfield H., Sasaki. S. and Nakanno I. (1984) Fine structural observations of nelrofilanenrtons dnanges in amyotrophic lateral sclerosis. J. Neuropathol. mp. Nelrol. 43, 461-470. HirschE.C., BrandelJ.-P., Galle F., Javoy-Agid F. andAgid Y. (1991) Iran and aluminum increase in tie substantianigra of patients with Parkinsan's disease: an X-ray microanalysis. J. Nelrochem. 56, 446-451. Ikeda et al., (1986) Calif. Tissue Internet. 39, 319-323 Jdnnsan G.V.W. (1988) 'Be effects of alnminum an agonist-induced alteraticu's in cyclic AMP and cyclic (MP concentrations in rat brain regias in vivo. Toxicol. 51, 299-308. Karlie S.J., Eidnhorn G.l. and Crager Mclachlan D.R. (1988) Moleallar innteractias of aluminum with DNA. Nelrotoocicol. 1, 83-88. Kasai K., I-iori, FLT. and Goodman W.G. (1991) Transferrin enhances tte anntiproliferative effect of aluminum on osteoblast-like cells. Am. J. Bnysiol. 260, E537-F543. 3O Klanzo I., Wisinievski H. andStreider E. (1965) Dcperinenntal productim of nelrofibrillary degeneration. J. Neuropathol. Exp. Nelrol. 24, 187-199 Kraal B., de Graaf J.M., Masters J.B., van Hoof P.M., Jacquet E. and Parmeggiani A. (1990) Fluoroaltminates do not affect the guanine-nucleotidebinding centre ofttepeptidechainelongation factor EF-Tu. Eur. J. Biodnen. 192, 305-309. Lai J.C.K. and Blass J.B. (1984) Inhibition of brain glycolysis by alnminum. J. Neurochen. 42, 438-446. Levine S. N., SannierG. B. andAbreo K. (1990) Effect of diabetes nellitts and altminum toxicity an myocardial calcilm transport. 'beicol. 65,137-148. Liss L., Lag J. and 'nnornntonnD. (1989) Aluminum nelrotoocicity, Alzheiner disease, and alcoholic ecerhalqnathy. In Environmental aneuistr'y and Toxicology of Altminum (Levis T. E., ed) pp 317-325. Levis Publ., Boca Ratonn, FL. Laden 6.11., do Vernejoul M.C., Fabiani F., Mardnais 8., Guerin A., Metivier F., Camus P. and Llsch F. (1989) Association between alminm accunulatian and cardiac hypertrophy in henodialyzed patients. Am. J. Kidney Dis. 13, 75 Login I.S., Jnldd A.M. and Incleod M. (1987) Activationn of calcium channel by maitotoxin. Heth. Enzymol. 141, 63-79. Macianald T.L. and Martin R.B. (1988) Altminum ion in biological systen. Trends Biodenn. Sci. 13, 15-19. mcdmald T.L., W W.G. and mrtin R.B. (1987) prcmotion of unbulin assednly by altminum ice in vitro. Sciece. 236, 183-186. Marsalr J.M., Ehrlich A. and Mansour T.E. (1983) 'Ite dual effects of aluminum as activator and inhibitor of adenylate cyclase in tne liver fluke @191; Mg. Biodnen. Bionhys. Res. Conn. 112, 11-918. Nartell A.E. and mtekaitis R.J. (1989) Coordination dnenistry and speciation of A1(III) in aqneons soluticun, in Environmental Gmistry and Toxicology of Alnminum (Levis T. E., ed), pp. 3-17. Levis Rlb1., Boca mtm, FL. )hrtin R.B. (1986) The chenistry of alnminum as related to biology and medicine. Clin. Chen. 32, 1797-1806. mtsmoto H. (1988) (hangesofttestructureofpeadnrdnatin by altminum. Plant mll Physiol. 29 , 281-287. 31 Mildevin A.S. and Loeb L.A. (1981) In Metal Ions in Genetic Information Transfer. (Eichhorn G.L. and Marzilli L.G. eds) pp. 103-123, Elsevier/North-Holland, Nev York. Miller J.L., Hubbard C.M., Litman B.J. and Macdonald T.L. (1989) Inhibition of transducin activation and guannosine triphosnhatase activity by alminum ion. J. Biol. Chen. 264, 243-250. Moreno C., Lifely 14.12. and Esdaile J. (1985) Effect of aluminum ios on denial and immunological properties of meningococcal grolp B polysaccharide. Infection and Immunity. 49, 587-592. mrimnras. andHatsmcto H. (1978) Effect af alnminum on sore es and template activity of purified pea 111A. Plant Cell Physiol. 19, 429-436. Morris C.H., CandyJ.M., CourtJ.A., mitford C.A. and Edwardson J.A. (1987) 'Be role of transferrin in tl‘e uptake of altmimm and manganese by tte DIR-32 neuroblastoua cell line. TIES. 15, 498. Martyn C.N., Barker D.J.P., Oenod C., Harris E.C., Edwardson J.A. and Lacey R.F. (1989) Geographical relation between Alzheiner’s disease and altminum in drinking water. Lancet. i, 59. Nixon R.A., Clarke J.F., Logvinenko R.B., Tan M.K.H., Hoult H. and W F. (1990) Altminum inhibits calpain-nediated proteolysis and inducedlumnnenrofilamentproteinstoformprotease-resistant high ncleollar weight conpleaes. J. Neurodneu. 55, 1950-1959. Nordstroen D.K and Hay H.H. (1989) Chapter 2, Aqueos equilibritm data for monoulclear alnminum species. In Environ. Chen. of Altminum (Sposito G. ed), [p 29-53. CRC Press, Boca Baton, FL. Nozawa Y., Halashima S. and Nagata K. (1991) Phospholipid-nediated signaling in receptor activation of luman platelets. Biochinn. Biqhys. Acta. 1082, 219-238. Osmod C.B., Bjorkman 0. and Anderson DJ. (1980) In Physiological Processes in Plannt Ecology. Springer-Verlag, Berlin. Oteiza P.I., Golub 14.3., Gershwin 14.13., Donald SLR. and Keen C.L. (1989) 'Be influence of high dietary alnminum on brain microtubule polymerization in mice. Toxicol. Lett. 47, 279-285. Parker D. R., Zelany L.W. and Kinraide T. B. (1989) Chenical speciation and plant toxicity of aqneos a11minum.In Environmental Onenistry and Toxicology of Aluminum (Levis T. E., ed), pp. 299-315. Levis Publ., Boca Raton, FL. Parl D.P. and Good P.F. (1987) 'Be association of aluminum. Alzheiner's disease, and neurofibrillary tangles. J. Neurol. Transm. 24, 205-211. 32 Putney J.W., Takenura B., Hughes. A.R., Horstman D.A. and 'Ihastrup 0. (1989) How do innositol phosphates regulate calcinm signaling? FESEB J. 3, 1899-1905. Quinlan G.J., Halliwei B., Moorhouse C.P. and Gutteridge J.M.C. (1988) Action of 1ead(II) and allminnm(III) ios—stinullated lipid peroxidationinliposones, erythrocytes and rat liver microsonal fractions. Biodiim. Biqhys. Acta. 962, 196-200. Rtee S.G., Kim H., Suh P.G. and Choi W.C. (1991) Multiple forms of phosplcincsitide-specific rhoqholipase C and different modes of activation. Biochen. Soc. Trans. 19, 337-341. Record H.T.Jr., Anderson C.F. and Lehman T.M. (1978) Q. Rev. Bionhys. 2, 103-178. Rockams A.J. and Connor J.R. (1990) Aluminum access to tie brain: A role for transferrin and its receptor. Proc. Natl. Acad. Sci. USA. 37, 9024-9027 Savoryet al., 1985, Clin. Endocrincl. Metab. 14, 681-702. Schofl C., Sanctez-Bneuo A., Dixon G.J., Woods N.M., IeeJ.A.C., CuthberwonSJL, Cthold P.H. and Birchall J.D. (1990) Aluminum perturbs oscillatory pcsphoincsitide-nediated calcitm signalling in lnoruoe-stinnlated hepatocytes. Biochen. J. 269, 547-550. Selkoe D.J., Lian R.K.H., Yen S.-H. ardShelansld M.L. (1979) Biological and imnclogial characterization of neurofibrillary degeneration by altminum. Brain Res. 163, 235-252. ShearsS.B., mwson., Loonis-Htsselbee J.W. andolllen P.J. (1990) ‘Be perturbation, by altminum of receptor-generated calcinm transients in hepatocytes is nctduetoeffects of Ins(1,4,5)P- -stinullated (22” release or Ins(1,4,5)P metabolism by tPne S-phosphatase and 3-kinase. J. Biol. Chan. 70, 837. Shi B. and Hang A. (1988) Uptake of aluminum by lipid vesicles. Toxicol. Environ. Chem. 17, 337-349. Snortle W.C. and Smith K.T. (1988) Aluminum-induced calcium deficiency syndrone in declining red spruce. Science. 240, 1017-1018. Siegel N. and Hang A. (1983) Aluminum interaction with calmodulin: evideceforalteredstructureandfunctionfronopticaland enzymtic studies. Biochim. Biophys. Acta. 744, 36-45. Sternuveis P.C., and Gilman A.G (1982) Aluminum: a requireuent for activation of tle regulatory conponent of adenylate cyclase by Ilium. Proc. Natl. Acad. Sci. USA. 79, 4888-4891. 33 anhayda C.G. and Hang. A. (1986) Organic acids reduce aluminum toxicity in maize root menhrane. Physiol. Plantartm. 68, 189-195. Terry R.D. and me. (1965) anerinental production of neuro- -fibrillary degeneration. 2 . Electmmicroscopy, phosphatase histodnenistry, andelectronprobeanalysis. J. Neuropathol. Exp. Nenro. 24, 200-210. Trocoso J.C., Hoffman P.N., Griffin J.W., Hess-Koslov K.M. and Price D.L. (1985) Alnminum neurotoxicity: a disorder of nelrofilanent transport in motor neuros. Brain Res” 342, 172-175 Vicentini L.H. and Cattaneo H.G. ( 1991) Are tie nultiple rhosnholipases C related by ncre than one nechanism? Pharmacological Res. 24, 1-4. Wagatsuma T. and Kaneko H. (1987) High toxicity of hydrocy-aluminum polyner ios to plant roots. Soil Sci. Plant Nutr,. 33, 57-67 Walker P.R. LeBlanc J. and Sikorska M. (1989) Effect of altminum andotlercatiosonttestructure of brain and liverchrouatin. Biodm. 28, 3911- 3915. Weis C. and Hang A. (1989) Alnminum-altered menbrane dynamics in lumen red blood cell white ghosts. Thronbosis Res. 54, 141-149. Winnievski H.H. Sturman J.A. and Shek J.W. (1980) alnminum chroride- induced nenrofibrillary changes in tie develcping rabbit: a dnronic animal nodal. (1980) Ann. Nequnathol. 8, 479-490. Woolfitt A.R., Kellett G.L. and Hoggett J. G. (1988) Synergistic binding of glucose and alnminum-ATP to tencokinase fron W mi. Biochim Bicwys Acta 955. 346-351- mlller L. and Wilrelm H. (1987) Uptake and distribution of aluminum inrattepatocytes and its effect on enzymeleakageandlactate femti . Toxicol. 44, 203-212. Ynlan S. and HaugA. (1988) Frictional resistance to notions of bimane-labelled spinach calnnodulin in respoce to ligand binding. FEE Lett. 234, 218-223. Zhang D.W. and Cblonhini H. (1989) Inhibition by allml'num hydroxide of the voltage-depedent closure of ttemitochodrial channel, VDAC. Biochim. Biqhys. Acta. 991, 68-78. Zhao X.J., Sucoff E. and Stadelmannn J. (1987) A13+ and Ca2+ alteration of nenbrane permeability of onerous rubra root cortex cells. Plant Physiol. 83,159-162 . GIAPI'ERII WWWI'IHIMBHULH‘IBHMEWCN BYMJRINENHJIWCELIS 35 The effects of aluminum on incsitol phosphate formation were examined in nurine nelrcblastona NzA cells labelled with [3H1myo-inositol. In intact cells, aluminum reduced fluoride- induced incsitol phosphate formation in a dose-dependent manner. In digitonin-perneabilized cells, GI‘P[S]-stinullated incsitol phosnhate formation was inhibited with increasing aluminum doses in a birhasic manner with an ICSO value of 20 uM. At 50 uM aluminum, theincsitol MetelevelmsrednedbyabontZ.5-3fold.‘nne inhibitoryeffect of aluminum (50 uM) could nctbereversedby increasing GI'P[S] cocentrations up to 500 uM. Application of aluminum lovered tle accumulation of incsitol [hosnhates mainly by inhibiting 1P3 generation rather than by interfering with metabolic reactios after 1P3 formation. Pre-cl'elation of aluminum with citrate or ESTA conpletely abolisted tle inhibition of fluoride-induced incsitol nhosphate production by aluminum in intact cells, bit had little effect on us inhibition of GI‘P[S]-induced incsitol phosnhate production in permeabilized cells. Applying aluminum prior to GTP[S] stimulation, turnoverof PIPzandPIPbecznearpreciably slower (30 - 45%). Inaddition of ugZVGp protein stimulation, phosphoincsitide hydrolysis. can be also evoked by increasing the intracellular (22+ concentration. When modulated by varios divalent cations, incsitol phosphate femtion resqnoded to aluminum stress differently. When ng+ was enployed, formation of IP3, IP2 and IP was inhibited. 36 In C32+-nnediated production, howver, only 1P3 release was appreciably depressed under aluminum stress; IPZ level remained unaffected. Dcposure of cells to altminnm also reduced bradykinin -triggered :13 production and intracellular (22+ release. mete findings suggest that a primary lesion of aluminum toxicity may be related to the inhibition of innositol phosphate production through the netal's interaction with the phoqnhoinnositide signal tram- duction pathway, presumably at Gp protein and pnospholipase C. Altminmhasbeeninplicnted asatoxicagentinvarionsneuro— dagenerativedisorders, e.g., inoertain types of seniledemerrtia (Ganrot, 1986, Crappernaclacnlan, 1989). At this time, no single mednanismlnas been identified as cansingaprimarylesioninthe altminm toxicity syrdrone. Alminmisapparortlyinvolvedinabroadqnectrumof physiological disorders. It is possible that a group of these disorders originnatnsfronthemetal's interaction(s) with a key target in basic metabolic patlway(s). A major cosequence of alnminmstross is lanown to result inadisunr'banoeofoellular calcinm metabolism (Marque, 1989) whidn, in turn, is partially interrelated with signal transduction, involving polypnospnc- inositide hydrolysis linked to a guanine nucleotide binding protein, med op protein (Berridge et al., 1983). Via this pathway, nanral cells, in response to extracellular stinnnli, 37 generate the intracellular messenger incsitol-1,4,5-triphosphate (193) to mobilize intracellular <22” (Bansal and Majerus, 1990). In' many types of cells, this signal transduction could be activated by fluoride in the presence of very low concentrations of alnminm (Gilman, 1987). Such activation is prestmably acconplished by binding of fluoroaluminate conplexes to the nucleotide center of Gpprctein.0onsequently,Gpprcteinislocked innGI‘P-bonnd activation conformation, thus preventing the effector enzyme fronn being switched off, leading to an elevated 1P3 fornation .(Chabre 1990). I-Icwever, alnminm stress generally interferes with cellular (22+ metabolism in an inhibitory mannnner (Levine et al., 1990), suggesting that alnminm may dcwnregulate cellular (22" levels via interference with onosphoinositide signal transduction by a mednanism different fron the activation node. Alnmimm application indeed hasbeenshowntocauseprofonndnegative effects on Ca2+ signalling in various types of cells like pancreatic acinar cells (Wakui et al., 1990). Application of alnminnnm to rat cortical slices reportedly depressed the release of innositol {inosphate following stimlation of carbachol (Johnson annd Jone, 1986) or fluoride (Jope, 1987) . unrecver, alnminnm was found to inhibit PIPZ hydrolysis by phospholipase C fron bovine heart (McDonald and Mamrack, 1988). 'nnere are other possibilities for alnminm to interfere with Ila/Caz" second messenger system, e.g., direct binnding toIP3 (Bironall annd Gnamell, 1988). In addition, key elements, namely 38 receptors, Gp protein and Linospholipase C, of the signal transduction pathwayare residing in the plasma monbrarne, and interaction of aluminum with membrane constituents may indirectly impact the regulation of the pathway. Taken together, alnminum—induced changes in ghosonoinositide signal trannsdnnction may provide a basis for understanding the mechanisnn whereby the toxic netal exerts itsprinnnaryeffecton nonral cells. Hennce, he decided to innvestigatethe effect of alnminum on phosphoinositide hydrolysis, employing neuroblastona cells. Our results demonstrate that application of aluminum inhibits innositol [hosphate formation, presumably by interactions ofthemetalwithGpproteinandnhospholipaseC. WWW mas Alltissue onlture applieswerepnrdnasedfronGIBCDOoJGrand Island, NY). myo-[2-3minositol (15.6 Ci/nlnol) and [329]- orthophosfinoric acidwereobtainnedfronNewEnglandNuclear (Boston, MA). crp[31 was bought frou Bodnringer Mannheim (Indianapolis, m). Bradykininwas obtained fron Sign Omical Co. (St. louis, in). All dnemical reagents usedwere of high quality. Cleaning of plastic ware and the preparation of incubation buffers and solutions were performed as described (Shi and flag, 1990) . 39 ELLE—um Onlturos of C1300 nncuse neuroblastona cells, clone nneuro-ZA (American Type Onlture Collection, Rodwille, 140.), were grown as mentioned recently (Shi and Hang, 1990). 'Ihe cells were used at confluence, 7-10 days after our first passage. W of Imeiml. M __Fomation Neuroblastona cells in 6-well nunltidishos were prelabelled for 24-30 h with 1.0 (for inntact cells) or 2.0 uCi/ml (for permea- bilized cell) of myo-(3m-inosito1 (15.6 uci/mol) in 2.0 ml Dnlbecco Vogt's modified Eagle medinm, um. After labelling unntil eqnilibrinm, the radioactive medinm was ronncved, and cells of the monolayer (abort 2.5 - 5 X IDS/well) were washed twice with the incubation medinmconposed of 140 um NaCl, 5 11M KCl, 10 11M ram, 30 on glucose, 5 mM MgClz; the meditmwasbufferedto thedesiredpfivalueswith 101MTris, neposorPipos. Cellsofnnonolayersweresub‘jectedtoalnminum challengeinl.0 ml imlbatimnfiilmmininglowmm, 2.5mm and 1.0 11M 2,3-dinhospho-D-glycerate (Sigma, St. Louis, m), prior to stimlation. 'Ihe reactionwas terminated by an addition of an ice-cold solution of tridnloroacetic acid (M) , finnal concentration 10%.. Cellswereextractedonice for 10minandthen scrapedoff. 'nneextracts mrecentrifngedathOngorSmin, and 40 supernatants and pellets were kept for the assay of innositol phosphates. 'Ihe extraction procedure was carried out at 0 - 4°C. A 1.0 ml volume of the supernatant solution was extracted with 2 ml solution of 1, 1,2-tridnloro-1,2,2-trif1uoroethanne and tri-n-octylamine (3 : 1) to removem (Challissetal., 1988). Afterronncvalof'ICA, theugnerrhase solution was neutralizedto pH 7.0 - 8.0 with 0.2 N mp1. Innositol phosphate products in the solution were separated on a column containing AG 1-x8 (formate form, zoo-400 mash, Bio-Rad, Rockville centre, NY), and [3H]- innositol mono, bis-, and triphosrhatos on the columnn were eluted stepwise (Figure 1) by using 0.1 M formic acid solution containing an increasing annuoninm formate gradient (Berridgeetal., 1983). line radioactivity of the fractions was counnted by liquid scinntillation spectronnetry. Members The perneabilization procedure followed basically that reported by Wojcikiewicz and Fain (1988): a monolayer of cells was cooled on ice for 10 minandumionlturemedinmwasrowved. Afterwashing with innonbation medium once, cells were inncubated in an ice-cold medium rosenbling inntracellular milieu (140 um KCl, 20 nu NaCl, 10 in! glucose, 5 null )gClz and w 7.4 bufferswith 10 um Hepos) containing 15 ng/ml digitonin for 10 min. Microscopic examination denonstrated that over 95% of cellsfailtoexchdetrypanblue 41 Figure 1. Elution profiles of innositol phosphates by colunnnnn chrouatography. the water-soluble extracts of [3H]inositol -1abelled nnerroblastona cells, treated with lomNaF/lo uM A1C13 (.)orwithnnoNaFandA1Cl3 (O),were applied to PG 1-XB column and eluted with: (A) water; (B) 5 M sodium borate/60 uM sodium fonnnate: (C) 0.1 M formic acid/0.2 M amninm for-mate: (D) 0.1 M formic acid/ 0.4 M anmoninm formats; (E) 0.1 M formic acid/1.0 M anunoninm foruate. Five peaks representthemetabolits eluted intheorder (fr'onnleft): free incsitol, glyceroplnosphoimsitol, IP, 1P2 and IP3. 'nne volunne of each fraction was 1.0 m1. 42 AHE\~noH x Eo0.. auq>wuou0acom rm mfl . o 1 m n m Liquid:oeuauuunuuuuununnu'ummmmnnflu ”wn......“ ........................ Q t 1 Eh ooooooooooo o ooooooo .0 .I'lu Oofloeoooomilwln = p p P L p n 25 30 35 40 45 20 15 1O Elute fraction number Fflgnezl.Enuthxnpnmfilecnfinosfiufl.phoqdn¢es hyczflnmnniuounbgnuhy. 43 after digitonin poration. Subsequently the permeabilized cells were mshedoncennoreininnonbationmedinm forSmin. “W 2: mitol MEG—S Eccqnt for minnor modifications, lipids were contracted and separatedasdescribedaiorwitzandperlnan, 1987). Inthecaseof [32P]labelled cells, susperdai cells, harvested fronn monolayers, were prelabelled with [32P]or'thq.hos;horic acid for 2 h in medium in 37°C water bath. [3H]inositcl (8.0 uCi/ml)-or [32P]ortho- ghospnoric acid (5.0 uCi/m1)-labelledcellpellets were dissolved in 1.0 ml of ice-cold solvent conposed of chloroform /methanol /concenntrated I-lCl (200/100/0.75, v/v/v). The mixture was allowed to stand inicefor10min,thennwaswarmedtoroontoqneraunre.After an addition of 0.2 ml of 0.6M HCl, the solution was centrifuged at GSngorSminnandtheugnerpnaseandmidileflur-rymre discarded.’lhe ranaining lower finasewaswashedWicewithO.5ml of an "upper rinse-like solvent” dnlorofom/methanol/O.6 N HCl (3 : 48 : 47). line washed lower phasewasconpletelydriedunder a nitrogen stream, and the residuewas dissolved in onloroform/ methanol/150 (75:25:2) syston. Innositol phospnolipids were separated by thin layer dnronatograrhy with a developing syston of chloroform/ methanol/ H20/mim hydroxide (48:40:75). The labelled lipids were visualized by radioautograrhy afterincubation at -70° C, then scraped andenctractedinGmyMIO-I/OJ M HCl (10 :20 : 8). 44 Afterthe addition of 0.5 ml CHC13 and 0.5 mleo, thelcwer [hase was dried and quantified by liquid scintillation counting. W of We; Mm! 'Ihe inntracellular free calcinm concentration was determinned as described (Zhon et a1, 1990, 'ijyo et a1. 1991). Briefly, suspended cells (1.5 x 105/1111) were loaded with monhrane—permeant, esterase—hydrolyzable acetoxy-nnethyl ester of fura-Z ( 2 uM/ml) for 1hinEMEM, inacnz incubatorat37°C.After washing twice, the fura-Z/AM-loaded cells were resuspended in inncubation buffer, at 5 x 105 cells/ml. Fluorescence measurement was performed at excitation wavelength 339 rm and emission wavelength 500 nun. Intracellular free ca2+ concentration was calculated fron the relation: [(22+) = 19 (F - Fm-,,n/(F,m.,,x - F) where 1% =224nfl (thedissociationconstantof Ca2+ binding to furs-2). For our neuroblastonna cells, equilibritm labelling was verified by mnitoring the incorporation of [3H]inositol into the phosrholipid pool. Under onrexperimental coditions, 24 to 30 h wereneededtoattain equilibrium asdennonstratedbystable incsitol onospnate acomulation levels (Figure 2) . Changes in 45 m 3" 1' g . 1’ m 2.5 - é - F, - 3 *‘ 2 2': as $3 E; 1.5 gas a?“ Q: 'a 0.5 4.! :9. L L, L L n 1 O 6 12 18 24 30 36 Labelling time (hours) Figure 2. Incorporation of [3H1myo-imsitcl into incsitol phoatames intimaupbhnnbma crabs. care; of unnchnmmsnune grownindithcontainingznuci/ml [3H]inositcl for the times indicated. [33] innositol phosphates in sanples were extracted and counted as described in MaterialsandMethods.‘Dne data grunts, repreozming the annuvitnss in tacit bmxfitcl phcqjames, are nmans 1: SEM of tzdplflonzn saunas :un a repncaauzmive<0.025 vs. corresponding control (nno NaF adiition) by Student's test. 48 A 40 )- T . 7’ 1 1 PI . O X E On 3 20 - m 0.) 3 +1 10 n 0H U) 0 C '8 1 .c 0 20.0 3‘ 4 .2 B Q: C 'H 3 >1 4.: .... > 3 0 2 fl 0 ou-n '8 . m 1 ' X, m i ”i it 0 L..r1 1 L 4 n I' 0 0.1 0.3 1.0 3.0 10.0 20.0 NaF Concentration (mM) Figure 3. Effect of fluoride on [.32P1phcsphoinnositide turnover in nanrcblastoa cells. 49 Figure 4 . Dose-dependence of fluoride-induced incsitol phosphate fornaticn in neuroblastoma cells: 1p (0), 1P2 (A) andrp3 ( I ). [3H1innositol-labelled cells of monolayers were treated with 1011M LiCl for 15min, thenincubatedwithNaFatvarions concentrations plus 10 AlCl3, 30 min, in pH6.8 incubation medinm, in a37cm2incubator. Datapointswith * represent theproduction in sanples treated with 100nMNaCl. Dataare meanns 1 sun of duplicate sarples. *p < 0.05, **p < 0.025 and - *“P < 0.01 vs. correspoding control (nno NaF addition). [3HIIP Production (cpm x 10-3 ) 25 20 15 10 50 - “ 12.5 - ‘ 10 _ - 7.5 - - S I ' .,*“15’ .\$ mats-trims" , , l .1 1 I 1 1 1 L L 0 t 0 0.1 0.2 0.5 1 ' 2 s 10 20 so 100 NaF concentration (mM) Ffignrn4.Ixseeigenizcennffbmmddeehihceiinosfixfl phcqjame2flmmatflzninnrnuubhnnrma<0.025ennd"""“'p<0.01bypan.in:edtett. 81 Table 2 . ugztdose dependence of aluminum-relatedinhibiticn of GI'P[S]-stinulated innositol phosphate production Total innositol phosphate production (cpn/well) 0192*] (I!!!) [IR-31.151 [IPS]+A1 [DS]+A1/[IPS].A1 0 7370 i 170. 5640 i 290 0.76 1.0 19260 i 680 8600 i 30 0.45 2.0 30620 i 1100 10110 i 1550 0.33 4.0 38470 i 1700 13010 i 290 0.34 5.0 36890 i 1670 14400 i 0 0.39 6.0 38420 i 2180 17450 i 750 0.45 8.0 37380 i’ 230 18650 i 1550 0.50 10.0 33510 i 280 17780 i 1030 0.53 mta are derived fron Fig.16. innositol [3%er represents total finsphateproductionintheabsonceofaluminnm; [IPT]+A1 ranesents that in the presence of 50uMaluminum. Nunbers in column 4 are the calollated ratios of values in column Bevercorreqnondingonesincolmmz. 82 Effect gf gmnimm pg mun-mediated l—PB Production and Intraggllular Q3+ gease Stinmlation of intact cells with bradykinin, a nuscarinic receptor agonist, resulted in immediate increases in 1P3 formation annd intracellular calcium release, in accord with previous findings (lakemra annd antnney, 1989, Mouillac et al., 1989). Application of 20 uM bradykinin brought the IP3 level to itsmaxinralvalue of 4040cpnconparedwitha basal level of 1090 on within 15 s (Figure 17A). 'nnetimingofincreaseininntra- cellular Caz" concentration (Figure 17B) resanbled that of 1P3 leveldnange: peak value of1421-28nfl (n=5) Ca2+wasobtained inls-zos (basallevel59i13nM). 'nne cell's ability to respond to bradykinin stimulation was howeverinpaired by pro-exposure toaluminum. Inaluminum-treated cells (Fig 17A and 17C), bradykinin-induced IP3 level was lowered to 2170 mandintracellular[Ca2+]to9Oj_-18nfl (basallevel 64 i 8 m, n = 5). After permeabilizition, cells became innsennsitive to bradykininstinulation, presumably because the receptors onthe cell surface were injured by digitonin treatment. 83 3r {to}? A“ b 'c I II. n—JGO a: i 5 "1.13 _ ’-47 p- 8 3 i " o. ‘490 a‘.‘ ‘.150 8 .37 I: ‘59 ,,. t a. t: of rim (5) Figure 17. Inhibition of aluminum on bradykinnin—triggered inositol phosphate formation (A) and intracellular Ca2+ release (B,C) inintactcells. A: Gellswere treatedwichOOuM alnminum (0|)crwithnoa1uminum(OD), 30 min, followed by stimulation with 20 m4 bradykinin. IP3 (.0) aniIPz (ID) prediction are shown. *p < 0.05and**P < 0.01 by paired test.BandC:I-ura-2nethylester~loaded cells were incubated with 200 m alnminum, 30 min, thenwashed. Forfluorescence studiesat500nm(excitation at 339 nun), bradykinin of 10114 finnal concentration was added into 3mlcellsuqnennsion(5x 1105 cells/ ml). Stimlation and calibration are indicated by arrows: (1): 20 ll! bradykininn, (2): 12 ug/mldigitonin, (3): 3.3wm+20wn‘ris. 84 DISQISSICN Previous work had deuctratedthattraceannmtsofalmnimm inndeed activate G protein-mediated incsitol phosnhate production in the presence of fluoride (Gilman, 1987). ‘Ihe results can this scriydemmstrate for the first time atthecellular level that application of aluminum at higher concentrations reduces incsitol phosphate formation. Our data snggest that the observed inhibition of incsitol phosphate formation results from aluminum’ s interference with {hosnhoinositide signalling pathway, presnmably atfileprimlytargetsGPPmteinarfimlipaseC- Incurstudies, aluminumconcenntratioc varied from 0 tozoo 014, within arangefonndinthebrainafterdnronicadministration to animals (Ganrot, 1986). In experimental intoxication, total brain alnminum concentrations average about 100114, anndinsone nenronal pqunlationsalnminumlevelsmayreadnwo 114 because of its noumiform accumulation in differennt neurons. (Nixon et al., 1990). Innoursysten, ATPwasusedasaghcsphatescuroe as'well as chelator to buffer allminum. Inthepresenceof 2.5mMATP, the total free A13+ in the incubation medinim was 10"10 - 10‘12 M based on model calculation (Martin, 1986). Kinetics of aluminum exchange between various intracellular chelators is not clear. For instance, alnminum in the soluble hydrated form crotherbonnd forms may be able to dissociate quickly enough to be available to interact with the target(s) in the signal transduction pathway. 85 'Ihereforewearehereusingasparameter the total alnminum dose added in the extracellular milieu. Fluoride forms a series of alnminum-fluoride chelates (Marten anxiibtekaitis, 1989). Amog the conplexes, A1F4' isthoughtto be involved in activation of Gp protein (Gilman, 1987), leading to enhanced incsitol nhosphate formation. In the presence of milimolar fluoride and micrnouolar allminum, A1F4" is. apparently a major fluoroaluminate species. On first thought, the deserved inhibition of incsitol nhosphate formation with increasing alnminum doses mightbeattributed to changes in qneciation of fluoroaluminate conplexes. 011‘ results dispute anchanexplanation because the inhibitionpersisted desqnitetherencval of external aluminum in the medium by washes with dnelator-containing buffer, priorto fluoride addition (Figure 128) . More convincingly, in the absence of added fluoride, accumulation of incsitol [hosphates triggered by GI'P[S] (Figure 6) or bradykinin (Figure 17) administration has also depressed in aluminum-stressed cells. mile lessening incsitol phosphate production aluminum application slowed GI'P[S]-stinnlated PIPZ turnover in nenroblastona cells (Figure 13). therefore, aluminumappearsto block a site (8) responsible for PIPZ hydrolysis alog the signalling pathway, ratherthanupstreamordownstreamsitesdistal toPIchydrolysis.'nnisnctionisfurthersugportedbythe following chservations. Firstly, alnminum reduces the total incsitol nhosphate level principally by inhibiting 1P3 generation from PIPZ hydrolysis. 'me observed lower IPandIleevels in 86 thepresenceofalnminmseonedtooriginatefron the reductionof 1P3 production rather than from depression of degradation of respective [hospiclipid precursors, PI and PIP. this was suggested by the temporal develqmennt of individual incsitol phosphate prodlcts (Figure 11) and by the concomitant increasein [3HJIP3 level and decrease in [3mm level following addition of "cold" IP3 (Figure 7). Hand PIP may serve ashydrolysis substrates if Mg.ATP is not available. MnenATP is addedto mainntain incsitol lipi® in the phosphorylated state, PIP2 is reportedly the main substrate for hydrolysis (Cockcroft and Taylor, 1987) . Secodly, alnminum failed to measurably interfere with ant- or Ca2+- stimunlatedIPandIPzrelease,whicharenortcoupledtoGp protein-mediated phosphoinositide hydrolysis (Figure 14) . 'nne fact that alnminum reduces either fluoride-induced incsitol [hosghate formation in intact cells (Figure 5) orGI'P[S]-induced incsitol nhosnhate formation in permeabilized cells (Figure 6) suggests that the inhibition is amarently taking place at site(s) distaltoreceptorsatthecellularsurface.‘1hisnotionis consistent with the finding that the alnminum-related inhibition of incsitol nhospnate formation was dependent on the metal’s interiorization. Firstly, binding of aluminum onto the cellular surface is conpleted within a fan: minuntes following aluminum treatment of cells (Shi and Hang, 1989). If alnminum acts on receptors at the cellular surface, its effect on innositol phosphate prodnction would be seen immediately, similartothcse of receptor 87 agonistsoranntagonists including metal cations (Smith et al., 1989): moreover, the inhibition slnonld not bedependent onthe preinncubation time of cells with aluminum as log as surface-bound aluminum is removed. Pullover, in onr experiments, appreciably loger preiranbationtimesnerereqniredforalnminumtobecome an effective inhibitor, and the extort of inhibition increased as the preinmbationtimewaslogthoned up to 60min (curveeinFigure 118). Secodly, inhibition of inositol phosphate formation by aluminum colld be averted if cells had been preirncubated with alnminum dnelatedtoEGIAorcitrate inmedia, thereby preventing surface binding and interiorization (Figure 128, bar 3 and 4). Inhibition lnowever persisted if cells were first treatedwith aluminum, then washed with a chelator-containning medinm, in which only surface—bound alnminum was removable (Figure 128, bar 1 and 2) . Dbreover, when permeabilized cells were employed, the reversion of alnminum—related inhibition by citrate— or mm-dclation was hardlycrnno logercbserved (Figure 12A). Apltativetargetforalnminum interactionismanhrane—boundcp protein coupled to phosgholipase C, whose activation is thought to be the rate-limiting step in phosphoinositide signal transduction (Gilman, 1987). Within our knowledge there has been no direct evidencefor inactivationoprproteinbyalnminum,buta similarmechanisn has beenn reported on othersigal-mediatingG proteins like transducin (Miller et al., 1989), and small molecular vweigntGI'Pbindingproteinslikembllin (Macdoaldetal, 1987). 88 The molecular nature regarding inactivation of G protein by alnminumisnotknown. Ourdatasnggestthatallminumdoesnotbind directly to GI'P[S], because addition of G'I‘P[S] in tenfold excess over aluminum couldnot reducethe aluminum-related reduction of incsitol phospnate release (Figure 9). Compared with phoqnnate, the sulfur atom ofthephosphorounioate is more mncleqohilic towards ”soft” ligands (Eckstein, 1985). 'Iherefore, the "hard" alnminum aquo ion (Hartley et al., 1980) is probably only weakly bond (or nnotat all) tothesubstituentinthegamnmapositionofGI'P[S]. In addition, the probability of aluminum binding to GI'P was drastically reducedsinceonreuperimennts were conducted in the presenceofa great excessofATP.lbreover,whenGerasaddedin 20-100timeshigherconcentratios than altminum, the alnminumn -induced inhibition could nnot be reversed at all (Figure 10). 'Iherefore, alnminum inhibits Q protein-related incsitol phosphate fornmation apparently not by forming a unusable Al-GTP conplex, which would lead to depletionofthe GI'Ppoolreguired for Q protein function. More likely, alnminum occupies a site at or near the nucleotide binding center onGpprotein, the formation of aluminum-ligandederoteinhinders binding of Mg” or/and crp on Q protein, or prevents activationoprproteininrespose to 1432+ or/and crp binding. As a physiological ligand for the formation of the functional ore-bonus Gp protein,hgz+iscruciallyinvolvedin the regulatory cycle of Gp protein-coupled signal transduction (Freissnnnthetal., 1989). Having similar ionic radii (Hangand 89 Weis, 1986) aluminum may conpete for thebinding site of 192+- Because the ligand exchangerateofhydrated alnminum is abort 105-told lower than that of hydrated n32“ (Haug and Weis, 1986), kinnetic features of (5p protein regulation are expected to be altered dramatically if nag?+ is displaced by aluminum ion. 'Ihis is illustrated by findings that aluminum inhibited the intrinsic GTPase activity of tubulin throgh Al3+/lg2+ exchange at the E site of the protein (Maoionaldet al., 1987). line inactivation of transducinbyalnminum was also reportedlydueto alnminum's conpetition with no” on the metal-free transducin- guaninne nucleotide conplex (Milleretal. 1989). In onr system, moms nu) was added together with nigz+ (5mM)inexcess over alnminum (50 in), the alnminum-related inhibition conld not be reversed (Figure 10). 'Iherefore, the observed inhibition is apparently nctduetotheconpetitionofbdometalsforGI'Pbutfor the binding site on GP protein. The verification of the putative alnminum/Q protein interaction requires investigatios on purified Qproteinmidnhasnotbeenavailableinanysystem. Inhibition of incsitol phosghate formation by alnminum was only partially abolished with increasing "J24. dose (Figure 16 and Table 2). 'Ihis daservation might beenplainedbythefact that ”J24- amlication up to 10 ml! was perhaps inefficient in displacing aluminum bond on high affinity sites on the metal-Q protein-nucleotide conplex. 'nnis notion is in accord withdata that the association costannt for aluminum in metal-GI'P-tinbnlin conplex is amroximately 107 times higher than that of Mgz+ 90 (Maodonald etal., 1987). Anotherpossibility is that aluminum mayhave,besides Gp protein, an additional target alongthe Windsitide signal pathway. This multitarget hypothesis is further supported by our findings that aluminum also interferes with C22+-induced incsitol phosphate production. Two separate pathways have been implicated in stimulating phosphoinositide hydrolysis (Chandler and Crows, 1990). They apparently anploydistinctgnosfinolipase C isozym having different substrate specificity (e.g., PIPZ vs. PIP) and (22+ sensitivity (Rhee et al. 1939) . Besides the big/GP protein-mediated classical pathway, phosphoinnositide hydrolysis canbealsostinulated directly by an increase in intracellular Ca2+ concentration, thus bypassing receptor and Gp protein involvement (Eberhard and H012, 1991). (Mr experiments indeed corrcboratetheexistomof,thesetwopattmaysin nanrrblastoma Nzln cells. According to our data (Figure 14 and Figure 15), the substrate preference of phospholipase c activity in the (22+- mediated pathways is seaminglydependantonintraoellular free (22+ concentrations: PIP hydrolysis become more pronomcedat high (22+ concentration (micromeles/l). It is not known whether both PIPZ- and PHI-hydrolysis activities reside in a single enzyme protein, or unetherthereareactuallydifferentisoters of phosnhcalipasecinvolved. 'Ihe dependence of PIPz and PIP hydrolytic rates on ca2+ concentration may have its physio- logical significance. At high Caz" concentration, cells om 91 employ PIP-specific phosrholipase C activity to gennerate DAG, a messenger signalling protein kinase C activation, without formation of unnecessary 1P3 and expenditure of anadditioal ATP. The inhibitory action of alnminum apparentlyisinclined towards PIPZ hydrolysis over PIP hydrolysis; at micromolar free Ca2+, PIP hydrolysis ronaina either unaffected or even slightly enhanced in the presence of alnminum. Nevertheless, in onrneuroblastoma cells, PIPZ hydrolysis in both m2+/Gp protein- and Caz+-mnediated pathways are susceptible to aluminum inhibition. Therefore the PIPz-specific phosonolipaseCreactionappearstobealikelytarget for aluminum. We postulate that the biphasic inhibition curve of GTP[S]-mediated incsitol phosplate production (Figure 6) may ranesent the events occurring on these twosites.Presumably, inhibitionatlcwalnminumorncenntrations (<50 ml!) is caused by inactivation oftherpr'oteinmidnisseominglyamore vulnerable, and at high concentrations by direct innhibition of phospnolipase Creactionbyalnminum.flerewedonnotruleoutother nears by which alnminum disturbs C22+-mnediated inositol phosphate release. For innstance, alnminum may impair Ca2+-calmndulin- dependantpcqholipase C activity (Ievine et a1. 1990, ) by inflcingstrucunral changes inalmcdulin (YuanandHaug, 1988). M putative manipulation of alnminum on phospholipase C reaction may cperate a different mechanism. At substrate level, theoretical evido'ncepredictsanavidbirdingofalnminum to PIPZ on 4’- and 5'1hosp'ate groups (Birchall and Chapell, 1988), 92 This alnminum—liganded PIP2 might be less vulnerable for enzyme digestion. Such alnminum—nhosphoincsitide interaction was reportedly related to alnminum-caused inhibition of PIPZ hydrolysis by purified phospholipase C from bovine heart (McDonald andManmrack, 1988). Atthe enzymatic protein level, alnminmm may interact directly with phospholipase C, or indirectlythrongh its perturbation of menbranne lipid environment. The hydrolyzing efficacy of pnnosrrnolipasos is known tobeinpartdependent on rhysico-dnemical properties of their microenviromental lipid milieu (Dennis, 1983). line presence of alnminumuptoSOOuMshowednoeffectonthe incsitol pcspate assay in column dnronatcgraphy (Figure 8). 'Ihis dcesnctnecessarilymeanaluminumdcesnctbindtothese metabolites, particularly IP3 , in vivo. Some investigators predictedthatsnxinbindingwoildalterthekineticsofIP3 metabolismandthe binding of IP3toitsrechtors. Forinnstance, a prologed half-life of alnminum—bound 1P3 may potentiate the ability of 1P3 to release (22" from the stores (Schofl etal., 1990). antinonrsudy,a1nminumstressvirtually diminished the bradykinin-irrmced intracellular Ca2+ release while reducing the br'adykininr-triggered 1P3 production (Figure 17). Also, the inhibition of the 3-kinase or S—pncsfiatase reaction of 1P3 by alnminum wasnctobservedinonrneurdalastomacellsandottercell systems (Shears etal., 1990). 93 Summing up, regarding adverse effects ofaluminmminneuro— blastona cells, on: data demonstrate tlat application of aluminmm reduces incsitol [hosptate formation possibly through interactions with elements of phosphoincsitide signal transduction. 'Ihis alnminum-triggered malfunction of the universal signalling pathway in enkaryotic cells may be related to primary manifestations of alnminum toxicity by impairing the cell's ability to properly respod to diverse stimuli. 94 LISTOF WCES Bansal V.S. and Majerus P.W. (1990) Phosptatidylincsitol-derived precursors and signals. Annnun. Rev. Cell Biol. 6, 41-67. Berridge M.J., Dawson R.M.C., Downes C.P., Heslop J.F., and Irvine R.F.(1983) Ganges in the levels of incsitol phosptates after agonist-depedent hydrolysis of membrane fincsphoincsitides. Biochem. J. 212, 473-482. Berridge M.J. (1987) Incsitol triphcsrhate and diacyl glycerol: two interacting secod messegers. Annu. Rev. Biochem. 56, 159-193. Birdnall J.D. and damell J.S. (1988) Alnmininm, chemical physiology, and Alzheimer's disease. Iancet, Oct. 29, ii:1008-1010. BramnerJ.B., Hajimohamumadreza I., Sardiwal S and WeaverK. Is incsitol bisfinosptate tie products of A2318? and carbachol -mediated polyrhcspnoincsitide breakdown in synaptosomes? (1988) J. Neurodnem. 51, 514-521. dabre M. (1990) Aleinofluoride and beryllofluoride conplexes: new nhcspate analogs in enzymology. TE 15, 6-10. Challiss R.A.J., Batty 1.11., and Nahorsky 8.12. (1988) mes measure- ments of incsitol (1,4,5)-trinhosoate in rat cerebral cortex slices using a radioreceptor assay: Effects of nenrotransmitters and depolarization. Biochem. Biqnhys. Res. Comm. 157, 684-691. Chandler L.J. and Crews F.T. (1990) calcium-versus G protein- -mediated fincspnoinositide hydrolysis in rat cerebral cortical synaptonenrosones. J. Neurodnem. 55, 1022-1030. Codmcrof‘t S. and Taylor J.A. (1987) Flucroalnminates mimic guannosine 5’-[gamuna-thio] triphcspate in activating tie poly- fincsphoincsitide Linosflnodiesterase of lnepatccyte membranes. Biochem. J. 241, 409-414. Cramer McIadnlan D. R. (1989) Alnminum nnenrotoxicity: criteria for assignningaroleinAlzteimer'sdisease, inEnvirommental and Toxicology of Alnminum (Lewis T. B., ed), 1;). 299-315. Lewis Publ., Boca Baton, FL. Dean NM. and Heaven M.A. (1989) Metncds for the analysis of incsitol phosfiates. Anal. Biodnem. 183, 199-209. 95 Dennnis R.A., Rhee S.G., Billah, 14.14., and Hannmm, Y.A. (1991) Role of phcepholipases in generating lipid secod messegers in signal transduction. FASEB J. 5, 2068-2077. Dennnis E.A. (1983) Bncsdnolipases, in 'ne Enzymes, vol. 16 (Boyer P.D., ed), pp. 307-353. Academic Press, New York. Bernard A and Holz R.W. (1991) calcium promotes the accumulation of polypncsphoinncsitide in intact and permeabilized bovine adrenal dnronaffin cells. Cellular and Molecular Nenrchiol. 11, 357-379. Eckstein F. (1985) Nucleoside nhosfinorothioates. Annu. Rev. Biochem. 54, 367-402. Freissmuth H., (hsey P.J., and Gilman A.G. (1989) Gproteins control diverse pathways of transmembrane signaling. FASEB J. 3, 2125-2131. Ganrot P.G. (1986) Metabolism and possible health effects of alnminum. Environ. Health Parspect. 65, 363-441. Gilman A.G. (1987) G proteins: Transducers of receptor-generated signals. Annnun. Rev. Biochem. 56, 615-649. any S.P., Seabright P., Day J.P., and Itztaki R.F. (1990) Uptake of aleinum by tuman nneirdalastona cells. J. Trace Elem. Electrolytes Haalth Dis. 4, 183-187. Hartley F.R., angess C., and Alccck R.H. (1980) Solution Equilibria, 1;). 249-274. Johnn Wiley & Sos, New York, NY. Hang A., and Weis C. (1986) Alnminum-induced danges in calmedulin, in Molecular and Cellular Aspects of Calcinm in Plant Development (Trewavas A.J., ed), pp. 19-26. Pleum Press, New York. fickle W.R. and Connn P.H. (1987) Use of lithinm ion in measurement of stimulated pituitary incsitol plnospholipid turnover. Heth. Ehzymol. 141, 149-155. Horwitz J. and Penman R.L. (1987) Measurement of incsitol rhospholipid metabolism in PC12 ptecdnromocytona cells. Meth. Enzymol. 141, 169-175. Johnson G.V.w. and Jcpe R.S. (1986) Alnminum impairs glucose utilization and cholinergic activity in rat brain in vitro. Toxicol. 40, 93-102. Jcpe R.S. (1988) lbdulation of pasfinoincsitide hydrolysis by NaF and aluminum in rat cortical slices. J. Neurochem. 51, 1731-1736. 96 Lee E. and Yupsa 8.11. (1991) Aluminum fluoride stimulates incsitol phosphate metabolism and inhibits expression of differentiation markers in mouse keratinnocytes. J. Cell. Physiol. 148, 106-115. Levine S.N., Sonnnier 6.8. and Abroe K. (1990) Effects of diabetes mellitnus and aluminum toxicity on myocardial calcinm transport. 'beicol. 65, 137-148. Li P.P., Sibony D. andWarsh J.J. (1990) Ganesine 5'-o-thio- triphcsphate and sodinm fluoride activate polyphcsphoinncsitide hydrolysis inrat cortical menbranes by distinct mechanisms. J. Neurodnan. 54, 1426-1432. Icomis-Husselbee J.W., Cullen P.J., Irvine R.F. and Dawson A.P. (1991) Electrqnoration can cause artefacts due to solubilization of cations from electrode plates. Biochem. J. 277, 883-885. Maoioald T.L., Humfinreys W.G. and mrtin R.B. (1987) Pronotion of tubulin assenbly by aluminum ion. Sciece 236, 183-186. Marquis J .K. (1989) Neurotnxicity of alnminum, in Environmental Chemistry and Toxicology of Aluminum (Lewis T.E., ed), 1;). 289-298. Lewis Publ., Boa Raton, FL. Martell A.E. and mtekaitis R.J. (1989) Coordination chemistry and speciation of A1(III) in aqueos solution, in Environmental Chemistry and Toxicology of Aluminum (1M8 T.E., ed), pp. 3-17. levis Publ., Boca Raton, FL. lartin R.B. (1986) ‘Be chemistry of alnminum as related to biology and medicine. Clin. Chem. 32, 1797-1806. McDoald J.J. and ranmrack R.D. (1988) Alnminum affectsnhcspho- incsitide hydrolysis by phosonoincsitidase C. Bicdniun. Biqhys. Rs. (Inn. 155, 203-208. Miller J.L., Hubbard C.M., Litman B.J. and Macdonald T.L. (1989) Inhibition of transducin activation and guancsine triphcsrhatase activity by aluminum ion. J. Biol. Chem. 264, 243-250. Mouillac B., Balestre M.N. and Guillon G. (1989) Transient incsitol (1,4,5)tri;hos;hate accumulation udervascpressin stimulation in WRK cells: correlation with intracellular calcinm mobilization. Bi . Biqhys. Res. Comm. 159, 953-960. Nixon R.A., Clarke J.F., Icgvinenko R.B., Tan M.K.H., Hoult M. and Grynspan F. (1990) Alnminum inhibits calpain-mediated proteolysis andindnmsl'mmanneurofilamentproteins toformprotease-resistannt high molecular weight conplexes. J. Neurochem. 55, 1950-1959. 97 Rice S.G., Kim H., Suh P.G. and Choi W.C. (1991) Multiple forms of Mcinositide—specific phosrholipase C and different modes of activation. Biodnem. Soc. Trans. 19, 337-341. 53::er D.D. (1985) Manganese stimulates the incorporation of [ ] incsitol innto a pool of ocsplatidylinositol in brain tlat is nnot coupled to agonist-induced hydrolysis. J. Neurochem. 45, 1481-1486. Schofl C., Sandez-aueno A., Dixon C.J., Woods N.H., Iee J.A.C., Cuthbertson K.S.R., authold P.H. and Birchall J.D. (1990) Aluminum perturbs oscillatory pesocinositide-mediated calcinm signalling in tarmac-stimulated hepatoctes. Bicdnem. J. 269, 547-550. Shi B. and Hang A. (1990) Alnminnm uptake by neuroblastoma cells. J. Neurochem. 55, 551-558. Sinusson P., Rodriguez P.D., Ioman N. and Alling C. (1991) G protein coupled to pcsfinolipase C: molecular targets of log-term etlannol exposure. J. Ncnrochem. 56, 2018-2026. smith J.B., myer 3.0. and smith L., (1989) cumminm evokes incsitol polyphosp'ate formation and calcium mobilization. J. Biol . Chem. 264, 7115-7118. Takenura H. and Rutrey J.W. (1989) capacitive calcinmentry in parotid acinar cells. Biochem. J. 258, 409-412. 'ijyo Y., 'I‘animunaA.,Hatsul S., Matsnmoto Y., SLgiya H. and Rmnyamas. (1991) NaF-induced amylase release f5om ratparotid cells is mediated by p1 breanodcmleadirgtoca+mobilizatim. Am. J. HnyBiol. 260, C194-(200. Weis C. andHangA. (1989)A1uminum-a1tered manhrane dynamics in human red blood cell white ghosts. 'nnronbosis Res. 54, 141-149. Wojcikiewicz J.H. and Fain J.N. (1988) Polyamines inhibit ptnosrho- lipase Catalyzed polyphosphoinositide hydrolysis. Biodnem. J. 255, 1015-1021. Woods N.H., Dimn C.J., cunthbertson K.S.R., and Cobbold P.H. (1990) fluoroaluminate mimics agonist application in single rat hepato- cytes. Biochem. J. 265, 613-615. Waknui H., Itaya R., Birctall D. andPetersenC.H. (1990) Intra- ular alnminum inhibits acetyldnoline- and mffeine-evoked m mobilization. FEE 1.3111. 267, 301-304 98 Yuan S. and HangA. (1988) Frictioal resistance to motios of bimane-labelled spinach calmnodulin in response to ligand binding. FEE Iettr. 234, 218-223. thu R., Shi B., Chou K.C.K., Cswalt 11.0., and Hang A. (1990) Ganges in intracellular calciLm of porcine sperm during in vitro inncubation with seminal plasnna and a capacitating medinm. Biochem. Biqhys. Res. Comm. 172, 47-53. CHAPTER III AIIMINUMUPIRKEBYMJRDIENEIJKDBIASKMACEIIS Biao Shi and Alfred Haug J. Neurochem. 55, 551-558 1990 100 Aluminum uptake in viable nneuroblastoma cells was largely depedent on tl'e medium pH. At physiological PH: cells were apparently uable to incorporate detectable amounnts of aluminum in theabsenceofprquer mediators. Aluminum uptakeinncreasedasthe #1 decreased, attaining a plateau ataboutpH 6.0. Performing experiments on 2 x lo6 cells/ml, pH 6.0, and 25 uM aluminum in the medium, aluminum incorporation readed satuuration at 5 nmol aluminum/ mg cellular protein, accounnting for 60-70% of aluminum added. At pH6.0, cellsshowedalarge capacity of accumuulating aluminum: about 70% of inntracellular aluminum was associated with tie postumitoonodrial fraction. At neutral pH, application of transferrin seemed to facilitate aluminum translocation into cells via transferrin -transferrinreceptorrontes. Fatty acids were also capable of mediating aluminum uptake at neutral pH, probably by forming aluminum-fatty acid conplenes. low molecular weight aluminum delators, like citrate, inhibited aluminum uptake with different efficiency, whereas on?” failed to alleviate aluminum intoxication by inhibiting sitter aluminum incorporation innto cells or aluminum superficial binding onto the cellular suurface. Treatment of cells with eergy metabolism inhibitors had virtually no influence on aluminum uptake, indicative of passive medanisms. 'nneresultssuggesttlataluminumuptakeoccurs viadifferentmodes depedentongrowthcoditiossudnasmediumpl—I. 101 Aluminum toxicity inplantshasbeenkncwnfor a log tinme, whileitstoxiceffectson humans and animals Iaveonlyrecently major targets of toxic aluminum, and high level accumulation of the metalinthesetissueshasbeenimplicated in a number ofneunal diseases like dialysis encephalopathy, Alzheimer syndrone and Parkinsonsyndrome (Ganrot, 1986: Crapper HcIachlanandDeBoni, 1980). Althoughaluminumisamarentlyinvolvedin abroadspectrumof physiological disorders, meoanisms of its toxicity renain largely unloown.'1wogeeralhypotheseshavebeenadvancedintermsof primaryinjury sites. Acmrdingtotteextracellular lesion model, the plasna mneubrane is believed to play aroleastheprimary target. As to tie intracellular lesion model, different organelles and macronolecules, e.g., chromatin (Walker et al., 1989), cytcskeleton (Macounald etal., 1987), enzymatic proteins like heumkinase (IaiandBlass, 1984), and regulatory proteins like calmnodulin (Siegel and Haug, 1983) are potential targets for aluminum intoxiation. In this case tte plasma menbrane plays a role in the manifestation of aluminum toxicity by controlling aluminumentry.Ineittercase,tle interaction of aluminum with tl'eplasmamenbranerepresentsttefirststageofaluminum cytotoxicity. 'nerefore, irrespective of the target site, detailed 102 information about aluminum uptake by living cells is important for uderstanding meoanisns of aluminum toxicity. Aluminum uptake in plantshasbeenextensively studied, and tterehavebeenanuumberof reports on aluminum accumulation in human and animal model (Slanina, et al., 1986, Domingoetal., 1988). But until recentlyfewattenptshave been made regarding aluminum uptake at the cellular level partly because of technique difficulties, this only scant data are available on cellular and molecular processes of aluminumuptakeandtransport, e.g., the translocatable aluminum species, potential mrriers, cell ' s capacity of accumuulating aluminum and intracellular aluminum conpartmentation. In this article, we are therefore reporting results, dutained by atomic absorption spectroscopy, onaluminum uptake bymuurineneuroblastoma cells. Neuroblastona cells were selected as the biological model because elevated aluminum is found in neural tissues of patients with neurological disorders (Forrester and Yokel, 1985), and because neuroblastoma cells express many of the daracteristics found in normal differentiated nneurons (de Iaat et al., 1984). 103 MATERIAISANDMEII-DIB Chemicals All tissue culturesngplieswereobtainedfrom Gibco Co.(Grand Island, NY). Chemical reagents used were all ofthehighest quality available. Plastic ware, washed with diluted nitric acid and then with redistilled water, was enployed to prevent aluminum contamination from tie usage of glassware. All buffers and solutions were preparedfromredistilledwaterwhidnhadbeenpassedthrough a delex-loo column to remove residual aluminum. 991.122.1213 C1300 mnouse neuroblastona cells, clone Neuro-ZA, were obtained from the American-Type Culture Collection (Rockville, MD) . Cells were cultured in Dulbecco—Vogt's Lbdified Eagle Medium (1145M) supplerented with 5% (v/v) fetal bovine serum. Cellswere grown in a rumidified atnosphere of 10% (Dz/90% air at 37°C. lbnolayers of neuroolastoma cells were wasted with Spinner salt solution and subsequently treated with a 0.2% trypsin solution. The trypsinized cells were renoved by aspiration of inncubation buuffer, followed by centrifugation at 100 g for 5 min. 'Ine sedimennted cells were resuspeded in and wasted witlntheinncubation buuffer three times, and were then available for experimental use. 104 Cell viability was examined by using a vital stain, viz., trypan blue. 'Ite initial viability of cells was about 95%. Prior to aluminum uptake experiments, tte effect of a given experimental condition (e.g., pH, fatty acids, eergy inhibitors) on cell viability was evaluated. Coditios were selected whereby viability of treated cells resenbled that for untreated cells. W of Aim—m m ‘Ihe following standard procedure was applied for measuring (Figure 1). Employingan incubation medium cosisting of 140 mu Nac1,5mMI 20). Between 1 and 5 ill of porin insolutionAatZOOpg/mlwasaddedtoa7ml solution of 0.1 M NaCl or of 0.1 M KCl (pH 3.5 or pH 5.5) for 30 to 120 minutes prior to analysis in a bilayer lipid membrane. 311a”! Epid membrane Theprincipleandtechniquesusedtostudy bilayer lipid membranes have been described pre- viously [20]. The lipids used to form the mem- brane included oxidized cholesterol. soybean phosphatidylcholine (PC. Sigma Chemical Co.). and too-phosphatidylethanolamine (PE. type V. from E. coli. Sigma Chemical Co.). Oxidation of cholesterol (Eastman Chemicals) was performed by the method of Tien et al. [21]. The membrane was formed by applying a 1.5% lipid solution in a-decane to a hole in a teflon chamber submersed in a bathing solution at pH 5.5 or 3.5. The lifetime of the bilayer lipid membranes was > 0.5 h and the resistance was > 10‘I 0 - cm: in the absence of porin. The pH was adjusted using HO and mea- sured with a pH meter. The membrane was moni- tored with a microscope until it turned black. indicating that a bimolecular leaflet had formed. Porin was added either before or after the mem- brane beeame black. Constant voltage was main- tained during all experiments. and the current across the membrane was measured using a Keith- ley model 610 electrometcr. 186 The stepwise conductancechanges. A. across thememhraneweremeasuredanddividedhythe specific conductance. e. of the solution. Assuming thattheporinchannelisahollowwatcr-filled qlindcr. then A/a-arz/l where r is the inner radiusoftbeehannelatitsnarrowestsectionandl isthechannellengthlnthissnsdxmostconduc- tanccchangesarereportedas thesizeparameter A/o since this parameter is proportional to the cross-sectional area of the channel and corrects for the changes in conductance at different pH valuesandindifferentsaltsolutions. Results Ahruptstepwiseincreasesanddecreasesinthe current across the bilayer lipid membrane in the presence of porin were recorded from the elec- trometcr. as shown in Fig. I. We assume that the CIE represent opening of porin channels. while the CDE represent closing events (arrows. Fig. 1). The size parameter. A/e. represents the conduc- tance change of the membrane (A) divided by the specific conductance of the bathing solution (a) in“ .; Fig. I.Stepwiaemeh-gaaerttuamembranecowrised ofPC/oaidiaedeholesseroltzzll'mlhepracnceofdoglml ofomprrotciniaabadu'qaoltrtioooflth’aCltpHSJl. mmmmuwwmmmh aolutionhfollowedbyprcinabation'nthebatlungsoluuon for30to‘0rniaatroomreweratare1'hecurtcntdecrementa. htdicated by arrows. pres-stably represent channel ddng eventsofpon'n trimersandhavethesamesizeasthsmq‘onry ofincrententevcnnl'heeurrentehartgeiadicatsdbyhthma magmnrdeone-thirdofthe-ajorityofeventsandisasned bindiantheopeiqdaaingkchndmvoltageaemas themembranewaSOIV.Thshueliaeofthetraeiqwa omitted. and is praumed to be proportional to the cross sectional area of the channel at its narrowest point (see Material and Methods). At low voltages (< 100 call}. the majority of conductance changes had size parameter values of approx. 3.1 A. While afew smaller current jumps. indicated by M (Fig. l). were detected which had size values approxi- mately one-third of the main conductance change. The size parameters of bath CIE and C05 were recorded at different transmembrane poten- tials. In these studies the bilayer lipid membrane. comprised of a 2:1 mixture of PC/oxidized cholesterol (by weight). was bathed in 0.1 M NaCl (pH 5.5). and the tnnsmembranc potential main- tained at values between 25 and 150 mV. At each voltage. approximately 200 individual events were measured. and histograms of the relative number (P) of C18 and CDE events were planed against the A/o values (Fig. 2). One can see that at all voltages studied. the main channel size was ap- prox. 3 A (Table 1). However. with increasing voltage. smaller channel events became prominent (Fig. 2. Table l). At a transrnernbrane patential of 125 mV. three distinct populations of channels wereevidencthemainchannelofsize3.lA.and msrnallerchanneIaOSandIJAinsizeThese smaller channels have cross-sectional areas ap- proximately one~third and two-thirds the size of the main conductance channel. Above 75 mV. the number of large conductance jumps (A/o > 6 A) increased. which may reflect prorein aggregation TABLE l SIZE (A/e) OF SINGLE-CHANNEL EVENTS AT pH 5.5 FOR ompF MEASURED AT DIFFERENT MEMBRANE POTENTIALS Voltage GE 9 C05 ('05 ""v’ Maia part . n ' 17'; ' (Ar tAi tin 3 3.1 2., Id. ‘ so at 2.9 a.d. 75 3.I 2.1 ML 1m 3.0 2.9 2.6 125 3.1 2.4 2.! ‘ AvcageuadalClEaadCDEsmallathanJLasshown 'mFig.‘. .AveragesiaeofaICDEasahowniaf-‘igl ' “Loam 187 induccdhythehighpotcntialsAMtheuniform- ityinthcsizcoftheconductanachannelsde. medatthehighpmcntiakashashcenre- FglbisuibuuaaofthshmA/e.b-F proteinaddaduabilaychid-snbr-smafw oaidiaadeholasaarolC:I)mifI-Ilmpalao tials. A is firemen-pads ‘uthemde onduetancsofthebatfithucaf.ilbiuary—V.h therelativenumberofev-tswiththeivaiasm “WWW“hmah dauihadforthLaxcaptthatthspae'm-d'mhw hissegramflSmWwasaddadmaoamauau’naofflq/d andth’sprota‘nwasaddadmhabhycbid-dmm withoupreuaauneaalothtafiudddqmma 'mclrrdadiathehistcgram ported in studic of mitochondrial porin [ll]. FromtheertperimenudescribcdinFiglthe CDE menurements alone were plotted as separate histogramandareshowninf-‘tgl'f'henumher of CDE was relatively sttnll when the transmern- htane potential was under 75 mV. but increased above this voltage. At 1m and 12.5 luv. the aver- age sizes of the CDEs were significantly smaller than the size of the tatal evcms (Table 1). Further- more. the average size of the CDE shifted to amallervalueswhenthepotentialwasraised from IN to 125 mV (Table I). In preliminary studies. we have found that the size distn'but'mn of the ompF porin channels was essentially identical when the bathing solution sur- rounding the bilayer lipid membrane was changed from 0.1 M NaCl to 0.1 M KCl (pH 5.5) (unpub- ' lished data). However. when the pH of the bathing solution was lowered to 3.5. a treatment which reportedly decreases subunit interaction within the porin u-imers [16]. the size distribution of porin channel conductance is dramatically altered. The size distribution histograms of ompF porin chan- nels inserted 'mto bilayer lipid membranes bathed inOJ MKC'l(pl-i3.5)ateshowninl-'tg.4.When the mhrane potential was maintained at 'L . A - - - 7w .1 _ - A—m - too—v 125-w s t a a Ale-tit F‘l'l'heisuihatioaofmesiaeparameterofonlythe “av-chuml’tglatifferenttraasmembranepoten- fiThsespui-catalmfilioaswcrathesamsasforfigl 188 D 25mv M'ML h g a 3 Alt ta) p 75!!” r t a a w (A) mama-mum“; forthcoan lflflflmmflwadiatbbatlingnluuondull KOtpHJkaorOJtozhatroor-tunpuamrepriorto additioatothebilayerlipidmemhranelfieverttsmaasuradat that-ovoltagetwere 25 kW. the majority of channels (both CIE and CDE) had a size of 1.6A.while there werea significant number of channels with one-third and two-thirds the size of the major peak. Further- more. when the voltage across the membrane was raised to 75 mV. the channel size distribution at pH 3.5 shifted to even smaller values (Fig. 4. 75 mV).and themajofityofchannelanhadasize of 0.6 A. approximately one third the msceo tional area of the channel at 25 mV. Not only did the size of the channels decrease at low pH values. but the number of closing evenuaboincreasedFigSshowsthepcrcentof total events that were CDEL At pH 5.5. using transmernbrane potentials below 100 mV. the C055 were less than 5% of the torn! events. while a. "",—.... ‘ r .I i" ’ o D U 3 3 C b 'I i I I. O a- . II” 8 fl 7! a B U Wit-IV) Fr; 5. Voltage of porin channel closing from omprroteinsuspendedatpHSJinOJMNaCHowratpl-l J.SinOJMKCIIILThehilayerlipidmembranesystemsused forthcpH5JstudieswereasducnbedlorthsJand3. and forthepH JJstudiaasdracrihedforthd. at pH 3.5. the CDEs were > 25% of the total. and approached 50% of the total at 75 mV. The results indicate that lowering the pH and increasing the transmembrane potential both decrasc the size distribution of ompF porin channels and increase the probability of channel closing for ompF porin inserted into a bilayer lipid membrane. Further- more. when the transmembrane potential was de- creased from 125 mV to 25 mV. the number of closing events decreased and the channel size in- creased. indicating that changes induced at high voltage were reversible (data not shown). Discussion The changes in conductance across a bilayer lipid membrane containing porin protein are of two basic types; increases in conductance (CIE) and decreases in conductance (CDE). In most bilayer lipid membrane studies of porins the majority of_evertts have been reported to he C15 and are interpreted to represent the opening of porin channels [8.13]. Others have noted infre- quent CDE and suggest that. under the conditions used in bilayer lipid membrane studies. porin channels close relatively infrequently [16]. 189 Inourstudie.wehave natedtwotypeof CDEThevastmajority of CDEoceurveryfast (<0.3 ms) and are of a magnitude escntially idenuealtotheCIEOnrareoccasionsatlowpli orinhighionicnretgthwedetectedveryslow closingeventswhichoccurredovcrapcriodof seconds to minute. Thee slow events showed. random fluctuations in the conductance during this time. We propose that the fast CDE represent a reversible conformational change in the porin proteins from an 'open' to a ‘closed‘ state. The slowprocesseenonrareoccasionisthoughtto reflect an irreversible denaturation of an open porin channel. In analyzing both the CDE and ClE at pH 5.5. the peak value of the channel size at all voltage was3.1A.Assumingachannel length of7.5nm {22]. we calculated that this size correponds to an interior diameter of 1.7 nm. consistent with Other bilayer lipid membrane results [8.16]. Below 1m mV. there was little change in the size distribution histogram of conductance change measured at pH 5.5. Thus. the structure of the ompF complex was unaltered by these change in the electric fiddllfithinthismngeOhm’slawhasbeenshown to be valid for porin single-channel conductance mrements [16]. However. above 100 mV. the channel size distribution histogram changed dra- matically (Fig. 2). In addition to the main channel of3.1 A. twoadditional peaksat approx. 1 andZ A were evident. We propose that the main conductance change withasizeparametervalucof3.lAreprescntsthe opening of a porin complex that contains three separate channels which open cooperatively. At . high membrane potential. the cooperativity be tween the units within the complex appears to be weakened and the channels more readily open idcpendently. This would account for conduc- unce change corresponding to channel cross-sec- tional areas of one-third and two thirds that of the strain complex. For the single channel with a size parameter of 0. 9 A. the calculated diameter would be 0.9 nm. consistent with the size exclusion linit oftheoronporindetermincd byliposomeswcll- irtg assays [19.23]. Our results. however. cannot determine whether. in this model. the single chan- nel with a 0.9 nm diameter repreents a protein trimerwithasinglefuscdchannclorasingle protein subunit. Channel conductance studies of mitochondrialporinsalsohaveshown thatthesize of the conductance channels decrease with in- creasing membrane potentials [IO-12]. In thee studie. distinct size populations of channels were also detected which changed in levels with mem- brane pctential [11.12]. Thee results. however. were interpreted in terms of multiple conforma- tions of the porin channel with different channel size. We do not believe that the three populations of channel size that we see for the ompF pretein repreent three conformations with three size of channels. lfthisweethecasethcprcdominant conformation with a conductance size of 3.1 A would have a diameter of 1.7 run. much larger than the exclusion limit of the ompF protein. ' In addition to the change in the size distribu- tion of the channels. high voltage also affected the stability of the open channel conformation. At low voltage the number of CDE was very small. and below 100 mV. the size distribution of the CDE was difficult to asses. However. above 100 mV. the stability of the closed state increased with increasing voltage (Fig. 5). The size of the CDE was distinctly smaller than'that of the total events (Table I). suggeting that. for our model. the clas- ingeventshowslescooperativitybctweenthe subunits than the opening event. A sinu'lar in- crease in number of CDE with increasing paten- tial has been nored for mitochondrial porins. and CDE of the mitochondrial porins are also reported tobesmallerinsizethantheopeningevents [IO-12]. Procese similar to those seen at high voltage were also detected at low voltage when the pl‘l of the bathing solution was decreased to 3.5. As shown in Fig. 4. at 25 mV. channels of one-third and two-thirds the size of the main channel were detected in sizable amounts. Furthermore. when the voltage was increased to 75 mV. the mono- meric channels were the major channels detected. Thus. even at low pH. increased voltage further reduced either the aggregation state or the cooper- ativity in the opening of the porin channels. The smaller size of the porin trimer at pH 3.5 (A/o - 1.6 A) compared to pH 5.5 (Ah-3.1 A) is thought to reflect a medic phi-dependent alter- ation in protein structure that significantly affects the apparent channel size. The change in confor- 19o mation that alters the diameter of the channel at low pH is probably not the cause of the increase in channel closing events. High voltage caused thee later change. but did not affect the size of the main channel event. Furthermore. increasing the voltage at low pH to 75 mV further enhanced the level of monomeric units and CDE. but did not alter trimer or monomer channel size. At pH 5.5. there appears to be a threhold potential. above which the porin subunits undergo a reversible. voltage-dependent change in confor- mation. This second. voltage-induced structural change triggered above 75 mV and enhanced at low pH induce two detectable change in porin function in the bilayer lipid membrane. The first is an increase in the number of smaller subunit channels (C15 and CDE). and the second is an increase in the probability of channel closing. Thee two change are likely caused by the same structural alteration in the protein. We propose that. bath at low pH and at high transmernbrane potential. the made within the trimeric unit become less tightly associated. In the loosened complexe. the separate channels open and close independent of the other subunits. and the closed conformation become more favorable. This model is consistent with the studie of Markovic-I-iousley and Garavito [24] and Schindler and Rosenbusch [31. who show that. below pH 4.5. structural change are detected which are at least partially reversible and are dependent on the detergent used to solubilize the sample. Thee phi-depen- dent structural change are thought to wake inter-subunit contacts and change the porin struc- ture [24). and thus. may decrease the cooperativity of the CIE/CDE of the subunits and increase the stability of the closed state. Furthermore. this les tightly associated complex is stabilized in a bilayer lipid membrane by elevated membrane potentials. Whether transmernbrane potentials control porinactivityontheintaetcellhasbeenacon- troversial quetion. It has been ealculated that the Donnan potential across the outer membrane usuo ally doe n0t exceed approximately 30 ml! [25], Furthermore. in studie where the Donnan paten- tial was elevated toashighas 80mv.therateof cephaloridine diffusion through the porin chan- nels was unaffected [26]. However. we found that such a potential may have been near the threshold b value needed to induce an alteration in porin structure. In addition lowering the pH may dramatically drop the threhold potential required to induce channel closing events. Thus. several environmental factors. such as pH and salt con- centration. may dramatically affect porin-channel activity and allow for voltage-dependent control of diffusion across the channel. We found that we could detect distinct size papulations of porin subunits and CDEs when the porin sample were pretreated by repeated freeze-thaw cycle. fol- lowed by preincubation in 0.1 M NaCl. In the absence of such pretreatment. the histograms were very broad and almost no CDEs were detected. even above 75 mV (data not shown). We believe that the pretreatment did not denature the pro- tein. but allowed for dissociation of porin aggre- gate to trimeric units and perhaps weakened the interactions within the trimers. Acknowledgemene This work was supported in part by Public Service Grants GMl4971 and GM31202 (I-LT.T.). Reference 1 h L (19851CRC Irv. Iiochem. 19. 145-!” 1 MLKnrner.C..$cllnidrnayr.W-Chea-Schmeiaae. U.“ Honing. UJIMWLMJM 3 Sci-die. M. std am”. am) FEISLett 172. 8549 a Garavito. R..\Llenkins.l..lamortius.l.l~l.larlsanul. all leenbusch. LP. (1983) 1. Mol. Biol. 1“. 313-327 5 Wei. 8.. Garavito. KM. Baumeister. W. ltd loser butch. LP. (198” EMIO l. 4. 1589—1592 OOangC. Min-shinnindGlaeseLLMflflfllio- pigs. 1. 47. 629—639 7 Dorset. D.LMLMAandWJJ. (ti-8t Biophys. J. 45. 128-119 8 Sebdler. H. and Mllflflllhneflatlkad. 53' USA 78. 2302-1306 9 MC. and Rosetta-sch. J.P.tlflfl EM”). (1593-1597 lOFIIitagl-Lls'euperLW.andleln.LlIIZlEur.J. Iiochem. 123. 629-636 11 MN-“La~IIrdieka.D.llMliochimlio- pbya Acta 686. lot-21‘ IZDeFinro.V-Tommasino.M.leuLlandhls-'eri.fi (1‘5) Biochim hiophys. Aeta 813. 230-202 13 We. H. and W1). (1973)ch Natl. Acad. 5d. USA 75. ”SI-3755 telemLJankn.K.loos.W.artduuger.P.tlm am am Aeta stt. Jos-Jts 15 I‘LmnandLI-nznmwlio- ”mamas-w 16 Lahy.lJLWattI.LP.ndL-.LLA.(1I5)IioeHm IinphynAcaafl7.zl-116 17MLandthAfl’7’lCaeLwn 423-427 1IWLT.T~.SdHaGn-ty.um 19 Nahaa.T.(1!76)Iio¢-.IinphyaI-C_.7Lm ”Tami-LTJIIflFmgS-Lidltm-m 1'91 11 1" I-LT.(197£)Iiay¢LipidMa-brua(lLM):Ther-y aadfiaeticaHedDahharJ‘uYort 22 Wuwnumsnm); maximum 3 manumwne—m 14 mm.znammum)m licphyaAcsal‘tlSi-I‘IO 8W1LIa-Alndlmsnflblliol. “317850-781 ‘ 26 Nihaidel-LadV-ehtnmmeobidlathl-n