1111 E115Ew

coCelUlar effecu of eratlotxin in vivat mobility o( '-Hxn ate iý

o memtbrane 4f heptoc-tes wid onewobuztomn ceflh0) Laurie Kilpatrick-Smnith ", Grztgorz Manisa 1. hate M. Vandeakooi " and

* ~ ~ ~d. Maria Erminska "

IPwpsftIa Pha*,O* PA 19.1E (V S-AJ

K~md uo4: R*daqcws buadis LAumi &ffuwm, Looi - w .2 an... Swmeemww

Upooyacud labeled with rhkl, camm Wthiocyauiw (FITC4LPS) onwas s to examm hiterilmcobwea odotoxia ond pLiasmai membrane lo boised rat heutaocytes miid unwo * sevub1*)a%t'me NIURIA3

eeRLs A:th seine eadomxida to adl mndo. bepatm-yta bond umv te= than dA aet mbimfoto cdlls At adose of 12 pg/mg dry wt. a bomd mobile frncdo. of beewn 0 #ad 75% of MT C-LPS wan fewi "bq, cw om p ) Wa 5C with A 1W hmdi~ulemffUSMCldeau11 (D) Of 44 10@-*CUs/s. 1a amwoblASIom cCQ themob&l frwtdoi mm laWe M3-90%). with D 1.0 10- 1OC' a/s D VU WO kteipeatM-e ýd W et Me, 10mid 37C mid Wa~esme from LS * 10 -9 is 1.80.10 8 csri~/ls is ie-puiocytes ai4 tr%9 9A. - 1 -' to 1.9 IS 'em'/ hi wme ~woiisiom ees Is both Mmpe of c4l souviahi (ceft wbich did am ou xltd Tr pma bkw) w

10pee Vlo h CibeN cmiShow4d difforemi *Mel) poMis OWd 100% ot the -O WfttjCme WeM 0064CThee tmis mgem uc ~) mdosza blaifi g to waommalho evils eoraniis of" 1.ia*omboanis with

din"ees moblites; (2) b~ndig atfd the obat fractions *adp adela ceiharw Woepliy d (3) thesifference La Wlading., latmW mobiity, and sin, of U.. Wmmobi fraction is bepuawca mW sd ewaiAwftom

eels may be doe to vwvadom hi membane compoodtio mwd/o umbe of bindin sIk. r ~v~

Inuodoctiog ~~used fluorescein isothioryaat leedpoly-'-.

Sccazdide to examne dirft-ty the mechanis OfU~Lttle is known about the mechanism of action toxin satichmnent to mammaL~An cad] memýbranes.

of endotoxins at the celluar level although their Farlier investigtions have shown that endotoxmseffects in clinical situations and experimental hAve the ability to bind to and interact with almostanimals have been characterized exteansvely. Previ- a&H types of a( la cell 14-9. Whether thisous studies from this laboratory 11-31 have used interaction oemnr throgh nonsecific lipid-lipidisolated cels as an in vitro model rystern and coutwtcs, which cause membrane detbdiuitionexpiord early ctlula response (changes = and/at charges to fluidity, or through binding tomtetabolites, ion contents and enzyme activities) to a specfic membrane component. which initiatesain.otoxin eqxpocar In the pteet work. we have celular perturbations. remains to be determined. I

In vivo stdahawe shown that the liver is animportant sate for clearance o( eadotcoin and that

Towain FA-qmmi *AA IND Sd*4111 pareachymal cells as wel] as Lzpfter cell areAbeeuwkmim F=T.bmn h c Lm. Np. involved in this p s 19.141 It is also known

- ~~that so alD cell typo bind endotoemn with the

Oi4S.4559,1S/M pk 3~.* (bsnjr DhjWM)

UTIUTION T EM A04

Ap~edforr~ mlm 7 12 2:3 8a09 OCT IYURm

same affmity or are equally susceptible to its toxic lated from perfused rat livers according to theefiects [4,7.71 In particular, transformed cells are method of Berry and Friend [161 with modifica-far more resistant to the toxic effects of endotoxin tions described by Krebs et al. [17]. Male Ithan are primary cells [5,11]. Therefore, in the Sprague-Dawley rats (200-300 g) were starvzd forpresent study, we examined the binding charactcr- 48 h then anesthetized with sodium pentobarbitalistics and lateral mobility of endotoxin in two cell (Nembutal, Abbott Laboratories, Chicago. IL) bytypes, primary cells (hepatocytes isolated from intraperitoneal injection at a dose of 5 mg/100 gperfused rat livers) and transformed cells (mouse body wt. Livers were perfused with Krebs-neuroblastoma NB41A3), using the method of flu- Henseleit saline (pH 7.4) equilibrated idth.reLb.e-orescence recovery after photobleaching. This is a 614- sal;,, 4pH '1) -quilibrated with 95%1relatively new technique which can be used to O2/5% CO2 and containing 0.4 mg collagenase/mlstudy lateral mobility of membrane proteins, gly- (Bochringer-Mannheim, Indianapolis, IN) and 1%coproteins, and lipids (for reviews, see Refs. bovine serum albumin. Perfusions were carried out12-14). The lateral mobility of these membrane with flow rates of 40 ml/min and at 370C. Aftercomponents is important in controlling cellular isolation, the cells were suspended in krebs- L/responses to the external environment. Because Henseleit buffer containing 2% dialyzed bovinemembrane proteins and lipids have different diffu- serum albumin to a cell concentration of 5 mg drysion properties on the cell surface, it should be wt./ml. Cell suspensions were maintained in anpossible, using the fluorescence recovery after pho- atmosphere of 5% CO2 in 0, and continuouslytobleaching technique, to determine whether endo- shaken at 25C. Dry weights of the cells in suspen-toxin is binding to either a protein or lipid compo- sion and of the medium were determined for eachnent of the plasma membrane. This knowledge cell preparation. .should, in turn. not only allow us to gain some Experimenia. design. Primary and transformedinsight into the mechanism of endotoxin attach- cells were-isolated and suspended as decribedment but also to test the possibility that the initial above and incubated with various concentrationsinteraction of the toxin with the plasma membrane (0.2-50 pg/mg dry wt.) of fluorescein iso-influences cellular susceptibility Zo its action. thiocyanate-lipopolysaccharide conjugate (FITC-

LPS) (List Biological Laboratories, Campbell, CA).Materials and Methods In this preparation, the lipopolysaccharide was

isolated from Eicherichia coU 055: B5 and theCell culture. Mouse neuroblastoma cells FITC probe was attached onto the hpid A portion

(NB41A3) were grown in RPM] 1640 medium of the molecule (24.1 mol FITC/mol 'PS. At "supplemented with 10% fetal calf serum and peni- various incubation times, two aliquots of cells werecillin/streptomycin. The cultures were maintained removed. One aliquot wassd without any further A2.

in a humidified atmosphere of 5% CO0 in air at manipulation and the second was washed by dilut-370C for 2-3 days. The cells were detached mech- ing 10-fold in endotoxin-free medium to ap.-anically from their solid support, washed in Hanks' proximately the same cell cocentration (5 mg drymedium [15] containing 5-10 mM glucose and wt./ml). In both samples (washed and unwashed),0.2% bovine serum albumin, and resuspended in fluorescence intertsity was measured using a Per-the same medium to a final concentration of 4-S kin-Elmer 650-IOS fluorimeter with 465 nm andmg dry weight cells/ml ((8-10). 105 cells/ml). The 520 nm as exctation and emission wavelengths,cell suspension was passed through a wire mesh respectively. Bound and free concentrations ofio • I(60-80 p&m) to ensure homogeneity. Incubations FITC-LPS were determined from the fluorescence1 -•were carried out at 25"C in a Dubnoff metabolic intensity. The washed cell suspension destined forshaker, in open flasks (to ensure adequate the fluorescence recovery after phoobleaching, 0oxygenation). The dry weights of cells in suspen- measurements was mixed with Trypan blue (10:1) 1sion and of the medium were determined for each to differewtiate viable from damaged cells. The ___preparation. Trypan blue-cell suspension (5-10 #d) was placed

Hepatocyte preparation. Hepatocytes were iso- on a glass microscope slide and secured with a-A -v.111abilltY Codes3OAvail a'nd/o--

COWC Spela,

-ii i III10S

cmverssp for phctobleacbol vwetrv expermwuu. sad AD>P - twratuo owe wisaevoW evu-mutPtotttz Cooct"TATUOS uft" drewrmumo. for lasbe~d c~ally b7 tbe stibcds of Lwnrm#f.i "~4and unuo~ed ceU ~swioc&w accrdmg to Lhe Trajucbw~d 1= W jawdJettie uin 1: 1 relrft-me-,Sod of Lowry el aI. I IS. Uvev'v It *as kfow tbAt a dcvof d 1'64 og

F.'wCroki,~ pw ~ rtotvv% &-te Dtoimhol W FITTC-L1FS/,w#4m4~ wi. prad,4c. a *41 ft.ztAfi"soo coef cfut (D) and mobile fwacocs (frac- IATj:T*?VmP ra wn im~a I b uwwmun ,wtr;

U)(1w of th.* flioro~,bcm wti"c we Waerayi Ow. - .72 :t -2 aid ?TTCLPS ~raad -6 61N~e in the mpenutwou umt-,cAk) wm e- 4.19 ILV~lis wt the momn *- S E rot thre "~: s-vewwwaww byreactaoee aOwvwy Owie 06OWo e~mnwatt In wmoaid beporwu. I og MTC-blaching. The sq~aratt was dewped &as am - LFS/ag *y wL peoduce a EWt aasae toI#vmeuoea bans [19.2101 A det&Wa deaatpumo of ba m wt w rumr~ withas 60the eqwpme i and dau walyhis en be fmind maw me e vwrvWa (eaovi - 0 44 ±003 sandteF'erobere ['All Diffusa'e axffloens of the fluo. FITC LPS uweted - O1'1 ± 0, *-m &,mrecrnt probe wese detv~riied from the time de. aseutabots we measue sciorLit to V-.ZzPCZdeMCV Of fluorescence recovery fteco'e we et &L JU.cuar'a wet mmutnely weurted for 5 mo Wo~OW-mn# a I a bleach Laerd wb&~nr of Owboim bowki & soes kea,?rwe,

ROinds Lateral inoboty of FITC-LS wa deterlmwedis hepwcvmi (WLO"ma a 30 MSn sub*nUO" with

Frwwnord wn Aw toxtror tf fTC-LPS 12 sig MlC4-Lf/=g dry a. a 231C (T~g. I)The loxaty of Oluorucem labelletf mdotaia (noi oavrpaftm to a sro &wde of eadotona~

was ausummed both id4 t"i by weuuria aher- of apprtia 3- 5 majIOO & The m -Amaaa *as basedacsuma isno mwOSUeubobc Parametm Is mouse an the c6Ajoij a(ieo Knesg &L 11 ~71 thot 65% a(DeuobLAstosna calls the effeact of FITtC.LIS ca rat liver a eomposed of PWAivmw Cag ceIssd omceiluLa IATF14ADFI ratio was dewcsed. ATP the reqxn of Z2~Iaszyk and Moos N) that Wo~o-

'Up-

L & ww smaif o wsos* swo m k"b *%ww o Ww Oesrw WA&OM IyaS.Mwwa ýat( e" N Mwbwaa

og M45/g y L ndr * snsonows Awhe i (4 ibme w%ý s ,0, -0av/$*i

ftav ftny of761L---

ing intravenous injection of endotoxin. 80% is Effect of temperature on lateral mobility of endo-recovered in the liver.) The fractional recovery of toxinfluorescence after the first bleach was uced as a Lateral mobility of FITC-LPS in both primarymeasure of the mobile fraction of the fluoro- and transformed cells was measured at 10. 25 andphores. It was found that 65% of the FITC-LPS 37 0c-to determine whether diffusion was tempera-bound to hepatocytes was mobile in the plasma ture dependent. Cells were incubated for 30 minimembrane with a diffusion coefficient of 4.5.10-9 with 12 jig FITC-LPS/mg dry wt. at the threecnu/s. After the second and further bleaches of temperatures, then washed and samples preparedthe same area the curves returned to their original as described in Methods. Measurements of fluores-pre-bleach levels, which demonstrates that the re- cence recovery after photobleaching were carriedcoveries were 100% complete within the limits of out using a thermostatically controlled stage soexperimental error (see Fig. 2). Moreover. there that the incubation temperature could be main-was no significant difference between diffusion taned Constant. The results of these experimentscoefficients derived from the first and successive are shown in Table L In both primary and trans-bleaches at the same membrane location. This formed cells, lateral mobility of FITC-LPS in-behavior indicates that the laser beam had no creased as the temperature was raised from 10Cadverse effect on the cellular pLsma membrane, to 37*C. In hepatocyt the diffusion coefficientwhich is in agreenent with numerous studies of increased 6-fold. from 1.78- 10-9 cm/s at 100Cother investigators [12-14,24,29,31). to 1.00. 10-' cu/s at 37*C. with no significant

Fig. 1lb shows a representative fluorescence re- change in the size of the mobile fraction over thiscovery after photobleaching profile for a neuro- temperature range. In neuroblastoma cells thereblastoma cell incubated with FITC-LPS under the was a 2-fold increase in the diffusion coefficient,same experimental conditions as described for from94.10- /cmý/sat10OCto1.97.10"'cm2 /shepatocytes. In neuroblastoma cells, lateral mobil- at 37"C, wile the mobile fractioc appeared to beity of FITC-LPS was greater. with an average somewhat smaller (by about 10%) at 37*C.diffusion coefficient of 1 10-8 cm/s. There wasalso an increase in fractional recovery of fluores- Effect of intuiation time on binding and lateralcence to 80% as compared to 65% in the hepato- mobility of endotoxincytes. The effect of length of incubation time on bi-

-;50.500

- 40,400

30.300--

!10.100

L.. 0 f k,

360 20d

Time (we)

Fig. 2. Fltaaacm id lWyO t Y 2a for.a tt teasm.tiv bl• esod FITC.LP'5&beWldmaou H mew m a ) ;atd

with 12 pg FITC-LS/mg dry w. at 2S0C for 30 ma. Diffusiom eaetci~em is 1.12-1Om/ aw rmmmmm rmwu. (1) i(2) 155 OWd (3) 19%.

TASIN I

trICT OF ~TIM flkATV~t ON LATTXAL ?d(4tLrr Of CLi.w"awd *Ms~" M~d 6srLc .bh cs~(iu ""U d*he IPA at C4 "s7 am b w - ý *No 1"'* veo"

WO& 2c S~~r L,' 0oC t"~1 hilN Vi "WMMMJ SIN n0JWMMMW '"OmrMbrum M( d saidMM -MR awr

we 1 u 14 'C wta.t I IJmr h P -1 0 01C Si %ot sos WA %wbUw& boob D M4 % %M 0

Pa 0Mt. hi I o.uad P *@a M T!C- . so ft~fafhinI wod P.OUJ (1 M"auhi owm o

M asew fewr, m)s0 wýi1 sA sw

md~ng and lateral uobibry of FMT4abelld endo. binge 1lL)hk~sb hep~txvis bound 3..3.ironaw .weusou wa ezamznW eds both primary &ad trans. rT~C-LPS thas dad miwoblsoaa oog&towaas cela and the mreslts awt shown in Tabie ILThet ceLls were omcbsatd for 10. 45. sed 90 mai at Effeaof oft.~x & ow~Mne so~r bw&d*25*C with 12 og FTrC-LS/nig dry wt. It was EqwL.bnas bumhcg stvud a differest endca.founad that beither lateral MobdiNt .r an of the wtsr mcunumtm vae perlamd as broiino&bi fractios (data sot shwe) was dependent bep.tocnes sod asoroblainam cells Udescvi)w.on wnubatios rne as e.ba type of cell (bepato. a Mateah& sod Mttboa am4 the reskaP4 wre

cvtes. D -3.63 -10`0 cn:/s seurobtatowna shows in F'. 3. Ceils were incuatowd ior 3o mut~ ' .NU4A3. D - .0O* 10O'cu?/s) at 250C wizb bervses I and Mp g FTTC4JPS/,wj

'do~mcca fucrwtic euqws wo thes ftO wenoe Ovw tQua moceu'sta Age~,.wwd(mwMet&) o 4tam wb the tW i- that owa so evidence of aniwtuos at bwakoI

admS a( FTTC-LPS was time IepuszJcw a other stas a sther aGltya Hawevgr. it sw observsd.primary at ransfortanl aMLs It was famed that &t *adh ~sedtain euuurothat bepatacylegwhes cells were incubated with 12 pL/mg d&I vn bound between 1.5-3-tim. awe eD"oaos/wqat 25*C for 10-90 uen the ascmot a( labele proten tbasm sowstoen ceaL&toxia bound to atLh dad soc vary nmpdicastiy with Scatchad asalysis of the bwada of FTC-LPS

TARLI IIZN~tNO A"D LATKLAL WWSLMT (W FT7V4A3U.L= V47DOTOMN IN MX"4AY A^tt ThANVOWV D CV.LL

W0440 bsIn md I'~m ah va mwbmd hr do ft new wi* 46 * FMT-LM/= ad ftmwm a

6wD (CMI/s Ie~ S 4.mw/10 (3.J1 * 0,3). 10" (101 11.14 *1.41(7 M4L4)CS SA 2 *41 (4)*

1443*123 (M 1tU30tcW *U 0.33**031O'- (6) 1&47 * IJUC M99 C S

lated to be approx. I. 10' per/cell, a value similar

so to that reported by Larsen and Sullivan [25] for

:10 human monocytes.It is worth mentioning that in the range of

endotoxin concentration of 0.1-40 ug/ml the im-20 mobile fraction on the hepatocyte bec-ime greater.

0 . . Although it is tempting to speculate that the im-t_ mobile fraction represents specific binding of

10 50 100200 250 lipopolysaccharides. interference by backgroundS-L fluorescence in these measurements does not allow

Rig. 3. bhadin of FrTC-LPS to bepawtcvt (W) and anevo- us. at the present time to draw a definite conclu-blastmm mis (A) a fianctm of •oam co u-tmuo. sionHepawocym and moblastoma odl (5 mS d&y wL/ml) wea

bcubated nuth FITC.LPS at the indiwsd coocarauous for La-tOJ n&obilizy of endotoxin bound to viable and30 man at 25SC. Bindn was dermmned fluoromeutrly.Values an the mesa + S.E. for four differrat cell preparauons. nonviai

Further information on the nature of endotoxininteraction with the plasma membrane was ob-

to hepatocytes is shown in Fig. 4. The binding tained by comparing the fluorescence recoverycurve was found to be non-linear, which suggests profiles of viable and nonviable cells. The diffu-the existence of two types of FITC-LPS binding sion coefficent for FITC-LPS bound to viable

,,. J , site: one with a high affinity which constitutes a hepatocytes was (4.58 ± 1.15 -10-9 cm2/s and thesma/] fraction of total sites and one with much immobile fraction was about 35% (see Table I). Inlower affinity composed of a relatively large popu- a nonviable hepatocyte (i.e., that which did notlation of sites. The number of FITC-LPS high-af- exclude Tr.ian blue) the rate of lateral mobility offinity binding sites on hepatocyte cells was calcu- FITC-LPS was higher (D (2.63 ± 1.04.10-1

cm2/s) and the fluorescence intensity curve showed100% recovery. (Values are means ± S.E for nine

0.05 separate experiments.) Fluorescence recovery of100% was also obtained when nonviable neuro-blasaoma cells were bleached (data not shown).

0.04.

• .0.03. An area of particular interest in understandingthe medhaism of action of endowoxin at the cellu-lar level is the made of attachment of endotoxin toS• 02 •the sur'face of -- .2 cells, It is reasonable to

"z assume that befor: endotoxin can elicit cellular.perturbatios, it must firt come in contact with

0.0I. and/or interact with the plasma membrfne. Stud-ies with radioactive labeled endotoxin showed that

0 rit binding of emdotoxin to a variety of rmn-0 02 0.4 0.6 02 1.0 1.2 marian cells was rapid (within minutes) and in-

00 FITC-LPS/ 10 6 depedet of temperature 4,5,8,25). It has alsobeen shown that pretreatment of jells with un-

F% 4L 3Cabd ly of bsu-lng of FrTC-4S to * lablded eadotoxi decreased the binding of labeledtywL Coaenruaoum of fhu and bomd FrTC-LPS w" do-

am, - g m ,we,, of h , toxin 18,26). However, if cells were r•t labeledbe I -t0'., PWas 0." VaUmw from fi, dfout bqaoyw with "C-labeled than exposed to unlabeled LPS,pqmuaua• very little displacement was observed, which may

¾~\ 7- $ U * 1

suggest that endotoxin binds to cell membranes It is unldy that the immobile fraction repre-with high affinity [8]. This finding seems, however, sents the binding of a mixed population of endo-to contradict the reports that endotoxin binding to toxin molecules (i.e.. aggregates of different sizes),a number of different cell types does not follow or aggregation of endotoxin molecules on the cellsaturation kinetics [8.27.281. Hence, it is difficult surface either spontaneously or due to chemicalto ascertain whether interaction of endotoxin with crosslinking produced by the photobleaching. Itmammalian cels is accomplished through direct has be= shown that successive bleaches of theattachment to a specific membrane component same arem of the cell surface result in recoverywhich generates a trnsmembrane signal or through profrO.s which approach 100%. If the immobilenonspecific lipid-lipid contacts which produce fraction were a result of croulinking of the endo-membrane destabilization and/or alteration in toxin molecule• then successve bleaches of themembrane fluidity. Either type of endotoxn- same spot would continue to produce an "immo.plasma membrane association could conceivably bile' fraction of fluorophores. In addition. eitherproduce the cellular perturbations that have been binding of various sizes of aggregates or formationobserved. of aggregates of endotoxin molecules should be

The results of this study have shown that endo- observed in both viable and nonviable cells. Wetoxin bound to mammalian cells consists of two did not observe an immobile fraction in nonviablefractions with different mobilities. A larger. mobile cells, indicating that cellular integrity rather thanfraction has a diffusion coefficient of between aggregation of endotoxin molecules is important1 • 10-1 (neuroblastoma) to 1 10-9 cm/s for the presence of an immobile fraction on the

' ~ , (hepatocytes, at 25"C which is dependent on cell surface.Stemaperature in the range 10-371C but indepen- It is more difficult to ascertain whether the/ dent of length of incubation time between 10 and immobile fraction represents specific binding to a

90 min. These values for diffusion coefficients are membrane component or internalization of a frac-simila to those reported for the lateral mobility of tion of th% endotoxin molecules. However, inter-membrane lipids [29-311 and. most likely, repre- n-li-ation, which most likely involves changes insent nonspecific binding of endotoxin to these both the protein and lipid componenu of themembrane constituent. plasma membrane, is known to be inhibited at low

The remaining fraction of endotoxin molecules temperatures [33,341 and increases with looger in-bound to cells is relatively immobile with a diffu- cubation times [331. By contrast, our studies showsion coefficient of len than 10-1 ciM/. The size that the immobile frwtion was temperature- andof this immobile fraction is independent of tem- incuba-ton -independent. Hence, it can beperature between 10 and 371C and independent of suggested that this smaller fraction of bound endo-length of incubation between 10 and 90 min. The toxin molecules which is relatively immobile onimmobile fraction is absent from nonviable cells in the cell surface, represents a population of siteswhich all of the endotoxin bound to cells was which interact with a specific membrane compo.mobile with a diffusion coefficient of approx. 10" nent, most likely a protein, since the latter iscm/s. Hence, the immobile fraction must repre- known to diffuse in the plane of the membrane atsnt a type of binding, the existence of which a slower rate [12,13,20,29,31.depends on ability of the cell to maintain its vital In comparing binding of endotoxin to hepato-function, i.e., to produce energy. cy" and neuroblastoma cells, we found the fol-

The nature of the immobile fraction of endo. lowing differences: (1) hepatocytes bournd moretoxin molecules is of great interest. At least three e no /mg protein than eurmoblastoma cells;possibilities can be entertaine: (1) apegatioa of (2) the lateral mobility of endotoxin on the cellendotoxin molecules on the cell surface; (2) inter- srface was lower in hepatocytes; and (3) the size"naization of A subpopulation of endotoxin mole- of the immobile fraction was larger in bepatocytescuas; or (3) binding of endooxin to a specific The asm n ax these differences are not cl.,, atmembrane component which is anchored on the preseUL One possibility is that the presence ofcell surface and not freely mobile. nanopam quantities of endotoxin in fetal calf

serum a component of the culture medium. causes 11 McCiiv-y.. A. and Bradley. S.G. (1979) . Ruculoeadothel.

transformed cells to become selectively resistant to Soc. 26. 307-316

the toxin. An alternative is that various cells ex- 12 Peters. R. (19811 Cell Biol 10taftaL Reports 5, 733-76013 Jacobso. IL (19M) wn Wase in Biology and Mediaine

hibit intrinsic differences in both the lipid and O3J cnk. F. F19ae IL .id Saco , C.9- eds.) pp.

protein composition of their plasma membrane 271-2H. Ponum. New York

and hence bind diff.rent amounts of endotoxin. 14 McClokey. M. and Poo, M-M. (1984) InaaCnaL Rev. Cytol

This means that the number or binding 'sites' for $7. 19-1I

lipopolysaccharide on plasma membranes of van- 15 H-nks J. and Wall&= R. (1949) Proc. Soc. Exp. BioL Mea.- ~71.1%W•-200ous cells may be different. Since in our hands 16 lny. I.N. s Friend. D.S. (1969) J. Cell Biol. 43.

hepatocytes appeared to be more sensitive to en- 4.06-520

dotoxin than neuroblastoma cells it is tempting to 17 Kwbs. HA. ornel. NW.. Lund. p. and Hems. L. (1974)

specUate that the amount of lipopolysaccharide i Alr Bcx= Symposium VI (lunquist. A.P. ed.). pp.7n1-743. M•ukspard. Cop a ""

bound per cell and/or the size of the immobile 18 Lowy. O.H. o NJ. Fr. Al. and Randall.fraction is correlated positively with ceUular sus- Rt. (1"1)3. 1 i r 193, 265-2.275ceptibility to the toxin. 19 Axeizdw D. LoppeL D.. Schlesuinger, ., Es. E and

Webb. W.W. (1976) Biophby. . 16.1055-1069

Acknowledgments 20 Jabsomn. K. Derko. 7- Wu. E-S. Hou% Y. and Poste. G.(1976). SupMramol. Saume 5.5365-575

Ths work was supported under th' Office of 21 VCaedkol 3.M. Mainar 0. and EIecisa. 4 (19433. 7.Naval Research Couutact No. NO00014-84-K-0514 22 Lampercch. W. sad Traautad . L (1974, in Methods of

with funds provided by the Naval Medical Re- Eacymi Analysis(Djmeyer. LU. ad.) pp. 21G1-2110,

search and Development Command and by Academic Pre. New YorkUSPHS grant No. HL1 708 from the National 2 • D. iber. W. and 3runeya, FLU. (194) in

Muuetods o Enrzyu•m Analyi (Uagmeyer. KU. ad.) pp.Institutes of Health. 23V-27 3L Academi PrI. New York

2 W'ilisma. D...M. Mlmby, J. -,d Kebe HA. (1962)

References 3iac . 3. 32. 90-9

I Klajtric.k-Smith. L. Ereciank . and Silve•. IA. (91) USA 31. 3"1-3495Sshock 8, 5$5-4w 26 D1A U M. Swart-TuL D.E.S. and Jackson. DAU (1973)

2 .lpamck-Smih.S L sad Eruzask. U. (1913) OC Soek BiUocim. Bioq . A:a SK0. 260-27611. 25-" 27 Zamm amm. D.HL. Gn•yr, I md Kam. 1. (1977) J.

3 K lauick-Smiak. L. DO&. 3.. Emnsnka. U. sad Silvw. htmaimo 119. 1013-10

IA. (1"83) Cire. Sbock 11. 101-111 23 Haufta-Cavl. N, Cbsby, R,. Cavailloa 3.AC and 14 Morrmo. D.C. and Rudbac JA (19•1) Cmtmp. Top. Subo. L (19• 1. Im oL 32A. 190-19• 4

o In• unoL -1,7- 29 S,1ainmuinw. 3J. Aidiud. D, KppeL. DX, Webb, W.W.

S Salver. LA. (1911) an Pazhopbysiok*Wca Effects of Endo- and Elsom. EL. (1977) Scismos 195. 307-309tomn at the Celular LAve (Majde. JA.. ad.). Mp 81-95. 30 SdchminW. J., SebchaW. Y- Casacasas. P. Wminsa

Alan I. Liss. New Yor U.C sd Pasa.. L (197) Proe. MAL. Acad. Sc. USA 75,

6 ahlry. LJ. (1977) CRC Ciit. Rev. TolcoL. , 239-313 3353-3357

7 Davies. U. sad Sewacrt-TuIL D.LS. (1931) Iehiom Di- 31 daLsat. &W, van der S"q. P.T1. Elso.. i-. sod Sdd-

pb-!. Acts 643, 17-29 8F J. (1"0) Proe. NaiL Aced. S&. USA 77. 1526-152

3 Prapni. R. Pona s, U.T. md Music.o, AM. (19•1) FEBS 32 Schkamaw J. Scbw Y. W'dfliansam. UC and Psu•.n

LeJt 131.103-107 L (1973) Pme. Na:l. Aced. Sd. USA 75, 2659-2663

9 Zlydauyk. .•C and Moo. 3.3. (1976) Intect. Imuna. 14. 33 reaclher. M. (1914) Sciece 224, 681-666ICO-105 34 Ku•=:bm TJI, K-lacimd MJ., Rocu S.

10 Ruador. 0. Hopf. U. and Meyer zam BDasm elede. LI. sod Lag Y.C. (193) Nochemiaay 23. 6437-"444

(1979) Aces Hepw.at-istamroma~L 26.C h-7

u VER SI Trr of PENNS YL VA NAPHILADELPHIA 19174

School of Medicine

DEPARTMENT OF PHARMAcoLoGy G3

Navy Contract NOOO14-84-K-0514

FINAL REPORT

Two types of studies were carried out under the above Navy research contract.

1. Interactions between endotoxin and plasma membrane were examined in isolatedhepatocytes and neuroblastoma cells. The results suggested that: i) endotoxinbinding to mammalian cells consists of two subpopulations with different mobi-lities; ii) binding of the immobile fraction is dependent on cellular integri-ty; iii) differences in binding, lateral mobility and size of the immobilefraction in hepatocytes and neuroblastoma cells may be due to variation inmembrane composition and/or number of binding sites.

Summarized in: "Cellular effects of endotoxin in vitro: mobility of endotoxinin the plasma membrane of hepatrcytes and neuroblastoma cells" L. Kilpatrick-Smith, G. Maniara, J.M. Vanderkooi and M. Erecinska. Biochim. Biophys. Acta847, 177-184, 1985.

2. Effects of intraperitoneal administration of a sublethal dose of E. coli endo-toxin were evaluated in fasted rats. The results demonstrated that there wasan early impairement of nitrogen metabolism in endotoxemia which may be ofsignificant clinical importance.

Summarized in: "Endotoxin-induced changes in nitrogen metabolism" L. Kilpa-trick-Smith, M.C. Yoder, R.A. Polin, S.D. Douglas and M. Erecinska. Res.Commun. Chem. Pathol. Pharmacol. 53, 381-398, 1986.

ENDOTOXIN INDUCED CHANCES IN NITROGEN METABOLISI

*12 2Laurie Kilpatrick-Smith *Mervin C. Yoder *Richard A. Polin

Steven D. Douglas and Maria Erecifiska'

D:epartment of Pharmacology

Department of Pediatrics'iUniversity of Pennsylvania Medical SchoolPhiladelphia, PA 19104

Running Title: Endotoxin Induced Changes in Nitrogen Metabolism

*Corresponding Author: Dr. Laurie Kilpatrick-Smzith, Department ofPharmacology, University of Pennsylvania Medical Schoo.,'Philadelphia, PA 19104 (215) 898-3661

ABSTRACT

IntraDeritoneal administration of a sublethal dose of F. coli endotoxin

to fasted rats produced within 12 hr a rise in plasma ammonia, urea, glutamine,

glutamate and aspartate as compared to controls. At this time, hepatic urea

levels also increased as did glutamine, glutamate, aspartate, alanine and

asparagine. Hepatic [ATP]/[ADP] ratios declined in endotoxin-treated animals

and there was an oxidaticn of mitochondrial pyridine nucleotides anrd a reduction

of c)tosolic pyridine nucleotides. The rise in blood ammonia and amino

nitrogen concomitant with increased ureagenesis suggests that endotoxin

administration markedly enhances protein catabolism.

KEY WORDS

hyperammonemia

ureagenesis

proteolysis

amino acids

Sprague-Dawley rats

INTRODUCTION

Studies utilizing radiolabelled endotoxin have shown that following

an IV injection, the toxin accumulates in the liver (17,19,29) where it

binds both to parenchymal and Kupffer cells (20,24,29). These observations

indicate that the liver plays an important role in tLe clearance and

detoxification of endotoxin.

The liver is also the primary site for gluconeogenesis, ketogenesis

and ureagenesis and serves a vital role in supplying the body's energy

needs. While the effect of endotoxin on hepatic gluconeogenesis and

ketogenesis has been studied in some detail (For reviews, see

1,2,9,18,23), its action on nitrogen metabolism is less well understood

(1,28). Preliminary studies from our laboratory have shown that a

sublethal dose of E. coli endotoxin increases plasma ammonia levels

(27) which suggests that hepatic clearance of this compound may have been

impaired. Therefore, the object of this investigation was to examine, in

more detail, the effect of endotoxin on hepatic function and in particular

on nitrogen metabolism.

METHODS

Sample Preparation

Male Sprague-Dawley rats (250-300 gr) were fasted for 12 hours then

injected intraperitoneally (IP) with either a sublethal dose (0.2 mg/kg)

of E coli 0111:B4 endotoxin (Difco Laboratories, Detroit MI) or

placebo (5% dextrose in water). With the particular batch of endotoxin

used, a dose of 0.2 mg/kg produced no parenchymal inflammation or cellular

necrosis within 12 hr (27). There were no changes in serum SCOT or SGPT

levels and the livers of the animals were essentially normal 48 hrs post

endotoxin. Other histological studies (27) showed no evidence of

inflammation or cellular necrosis in the brain, uor was there any

.*.ndications of vascular damage. It does not appear that this level of

endotoxin produced intestinal damage since the animals did not have

diarrhea or show any other evidence of gastrointestinal disfunction. In

addition, at autopsy there was no evidence of gross intestinal damage such

as apparent hemorrhages or necrotic areas.

Plasma and liver samples were taken at 1, 4 or 12 hrs post-treatment.

At sampling times, animals were anesthetized with sodium pentobarbital (35

mg/kg) by IP injection. In one group of control and endotoxin-treated

animals, a 3 cm incision was made .in the left axilla of the recumbant

(supine) animals to lacerate the axillary artery. Two ml of blood were

aspirated from the axillary fossa, placed in a heparinized tube and

centrifuged. The plasma was removed and quickly mixed with cold

perchloric acid (final concentration of 4%), centrifuged and the clear

supernatant was neutralized (pH 6.8-7.2) with 2N K 2CO 3-0.58 M

triethanolamine. Aliquots of the extracts were used for the assay of

metabolites.

The second group of animals (control and endotoxin treated) was

anesthetized as described above. A 3 cu. incision was made in the right

upper quadrant of the abdomen to expose the liver. The animals were

quickly turned 180 to allow the liver to protrude vertically from the

incision and a section of the liver was freeze-clamped using aluminum

tongs which had been precooled in liquid N2. (Care was taken during

this procedure not to obstruct blood flow and introduce anoxic conditions

to the lobe of the liver to be removed.) The liver samples were stored in

liquid N2 until .;etabolites were extracted according to the method of

Williamson and Corkey (26) and neutralized as described above. Wet

weights of the liver samples were de:ermined by weighing the freezed

clamped tissue prior to extraction with perchloric acid.

A 12 hr urine collection was obtained from 5 control and 5 endotoxin

treated rats. Animals were housed in hanging rodent metabolic cages (Acme,

Cincinnati, Ohio). Urine was collected for 12 hrs post treatment,

measured, and a I ml sample was mixed with perchloric acid (final

concentration 4Z). Samples were neutralized as described previously and

aliquots were utilized for determination of urine ammonia and urea

concentrations.

Measurements of Metabolites

The concentrations of ATP and ADP were measured enzymatically using

the methods of Lamprecht and Trautschold (15) and Jaworek et al (I0),

respectively. Ammonia and urea were determined by the method of Gutmann

and Bergmeyer (7), acetoacetate and 3-OH-butyrate according to Williamson

et al (25), lactate according to Gawehn and Bergmeyer (6) and pyruvate by

the procedure of Czok and Lamprecht (5).

Measurement of amino acids

The concentrations of various amino acids in both plasma and liier

samples were determined by high pressure liquid chromatography (HPLC) of

their O-phthaldialdehyde 2-mercaptoethanol derivatives and fluorescence

detection (8). Sample preparation was carried out as described previously

(11).

Reagents

Enzymes were obtained from Boehringur-Mannheim or Sigma Chemical Co.

All other reagents were of the highest purity commercially available.

4.

RESULTS

Effect of endotoxin on plasma [amino acids)

The concentrations of amino acids in the plasma of control and

endotoxin-treated animals were measured at 1, 4 and 12 hr post treatment.

Figure 1 shows a profile of selected amino acids involved in nitrogen

metabolism over this time. At I and 4 hours, there were no significant

changes in the levels of aspartate, alanine, glutamate, arginine,

glutamine or asparagine in endotoxin-treated as compared to control

animals. However at 12 hr after endotoxin treatment, several changes were

observed: both aspartate and glutamate levels increased by 33Z, while

arginine rose by 24% as compared to controls. The greatest change was in

glutamine levels which increased by 65% in endotoxin-treated animals.

Effect of endotoxin on plasma (ureaJ and [NH,-1

The concentrations of urea and NH4+ were also determined in

plasma obtained from control and endotoxin-treated animals at 1, 4 and 12

hr post treatment (Table I, Part A). Plasma NH 4÷n control animals

remained stable at approximately 60 uM. In endotoxin treated animals, the

[NH 4+] was within the same range as controls at 1 and 4 hr but after

12 hr, had risen by 50%, i.e. to 99 uM. Urea concentrations were also

relatively stable in control animals over 12 hr with an average [ureal of

7 mM. In endotoxin-treated animals, there was a slight elevation in urea

to 8.02 and 8.41 mM at 1 and 4 hrs, respectively but by 12 hr, a

significant 41% increase to 10.05 mM was observed.

Effect of endotoxin on urine (NH,-]. and (urea]

In order to ascertain whether the rise in plasma NH 4+ and urea in

endotoxin-treated animals was due to an alteration in hepatic nitrogea

5.

metabolism or a change in renal clearance of ammonia, urine samples were

collected from both control and treated animals as described in Methods.

It was found that there was no statistically significant difference in the

volume of urine excreted by the two groups. The urine [NH 4 ] was

368+2 uM in control and 356+20 uM in endotoxin-treated animals (values

are the mean + SD for 5 animals in each group). In controls, the urine

urea content was 103.6+6.2 mg/12 hr while in endotoxin-treated animals

there was a significant (P < 0.01) 38% increase to 138.7+18.5 (as

determined using a t-test for 2 means with an a of 5 for each group).

Effect of endotoxin on hepatic (amino acids]

The profile of liver amino acids was determined in control and

endotoxin-treated animals at 1, 4 and 12 hr post treatment (Figure 2).

One hr after treatment there were no statistically signwficavtt differences

between endotoxin-treated animals and controls. At 4 hr, glutamate levels

increased by 53% and those of glutamine by 41% as compared to controls.

At 12 hr, there were statistically significant increases in the following

amino acid concentrations; aspartate 42%, glutamate 72%, glutamine 76%,

asparagine 16% and alanine 96%.

Effect of endotoxin on hepatic (urea] and [NH+]

Hepatic ammonia and urea concentrations are shown in Table I, Part

B. There was no change in urea concentration in endotoxin-treated animals

at I hr as compared to controls. At 4 hr there was a slight increase

(24%) and at 12 hr the urea concentration had risen by 67%. In control

animals, the hepatic ammonia concentration remained stable at 0.7 umol/gr

wet wt over 12 hr. Endotoxin-treated animals had similar NH4 levels

at 1 and 4 hr, but at 12 hr post treatment there was a 24% increase to

0.88 umol/gr wet wt.

• 'i, ,

Effect oi endotoxin on henaric *nergy production.

Several parameters of hepatic energy production are shown in Table

II. (ATPI/[ADPI ratios were similar in the two groups of animals at 1 and

4 hrs. However at 12 hrs, endotoxin-treated animals showed a 25% decrease

in [ATP]/[ADPj as compared to controls (3.42 anid 2.47, respectively). The

redox state of the mitochondrial pyridine nucleotides was determined from

the [3-OH-butyratee/[acetoacetate] ratios. It was found~ that there was no

difference at 1 and 4 hr post treatment but at 12 hr, the

[3-OH-butyratel/[acecciacetatel in endotoxin-treated animals had declined

by 53%. This indicates that there was oxidation of the mitochondrial

pyridine nucleotides at that time. It should be pointed out that even

though the (3-OH-butyratel/[acetoacetatej ratio declined in these animals,

there was no change in the concentration of total ketones.

[Lactatej/[pyruvatel ratios were also determine4 at 1, 4 and 12 hr post

treatment. At 12 hr there vas a 54% increase in the tlactatel/[pyruvatej

in endotoxin-treated animals as compared to controls. This rise was due

to an increase in hepatic lactate concentration, while the pyruvate

concentration remained unaltered. The (lactatel]/[pyruatel is a reflection

of the redox state of the cytosolic pyridine nucleotides and the. increase

indicates their reduction.

DISCUSSION

In this study, we have found that a single sublethal dose of

endotoxin given to fasted rats produced alterations in plasma and hepatic

concentrations of several metabolites within 4 hr. Plasma levels of urea

began to increase at 4 hr and by 12 hr both plasma urea and aumonia levels

were significantly elevated as compared to controls. There were also

increases in plasma levels of aspartat., glutanine, glutamate and arginine

* 7.

12 hr after endotoxin. In the liver, glutamate, glutamine and urca were

increased by 4 hr after treatment and by 12 hrs, glutamate, Slutamine,

aspartate, alanine, asparagine, urea and lactate levels had risen

significantly. In addition, there was oxidation of mitochondrial pyridine

nucleotides, a reduction if cytosolic pyridine nucleotides and a decline

in the [ATP]/[ADPI ratios.

One of the most striking alterations produced by a sublethal dose of

endotoxin was the large increase in plasma ammonia levels, 12 hr post

treatment. A similar rise in blood ammonia has been reported by Yoshino

(28) in rabbits that were injected intravenously with Shigella

flexneri endotoxin (1 mg/kg). Annonia is extremely toxic to animals for

reasons that are not yet clear (4) and its blood level is normally

maintained within narrow limits at very low concentrations (•* 60-70 uM

in fasted rats (16)) through conversion to urea.. Hyperammonemia can occur

in an animal for several reasons; 1) increased absorption from production

by intestinal flora; 2) decreased excretion due to renal failure; or 3)

increased protein catabolism and/or decreased ability of the liver to

metabolize ammonia. Because we used a preparation of purified E. coli

endotoxin aud the animal showed no signs of intestinal damage (see

Methods), it does not seem likely that the increased blood NH is due4

to increased bacterial activity in or enhanced absorption from the gut.

We can also rule out altered renal function because endotoxin-treated

animals did not have decreased urine output or ammonia excretion as

compared to controls. In addition, we found that the urea concentration in

urine from endotoxin- treated animals was elevated as compared to .

controls.

8.

The two remaining possible explanations are that endotoxin either

inhibits hepatic ureagenesis or increases protein catabolism for which

hepatic nitrogen metabolism can not compensate. The former seems unlikely

because we found increased urea levels in both the plasma and urine of

endotoxin-treated animals. Furthermore, the hepatic concentration of this

compound was increased by 24% at 4 hr and by 67% at 12 hr post endotoxin.

These observations indicate that endotoxin enhances rather than inhibits

urea synthesis.

It has been observed that in both sepsis and after.endotoxin

administration amino acids are mobilized in peripheral tissues for

synthesis of proteins and preservation of liver and other central organs

(3,21,22). Our results are consistent with these findings in that they

show increases in-the levels of several amino acids-in-both the plasma and

liver of endotoxin-treated animals. Hence increased protein catabolism

seems to be the most likely explanation for increased blood ammonia

levels, although the protein source and the site of the lesion remain to

be established.

It is worth pointing out that synthesis of urea in the liver is not

the only pathway for detoxification of ammonia. There are two other

biosynthetic reactions which are important in the regulation of ammonia

levels (16); formation of glutamate (from NH 4 + O(-ketoglutarate)

+and glutamine (from glutamate + NH• ). Our results show that there are

large increases in the hepatic concentrations of both glutamate and

glutamine which suggest that these amino acids may also be products of

hepatic ammonia detoxification and not solely of protein catabolism. This

suggestion is consistent with studies using N labeled ammonium

compounds in vivo which have shown that the label appeared not only in

9.

urea but also in meveral amino acids, specifically glutamate, aspartate,

glutamine and asparagine (16).

The question that remains to be answered is why is there still a rise

in plasma ammonia level when the liver appears to be responding to an

increased load appropriately by stimulation of detaxification mechanisms?

A possible explanation is that glutamine synthesis and urea production are

not rapid enough to deal with the overload. It is well established that

both of these processes require ATP. Our measurements of hepatic energy

parameters showed that there is an alteration in the energy level in

endotoxin-treated animals. The oxidation of intramitochondrial pyridine

rucleotides and concomitant decline in the [ATP]/[ADP] suggest that due to

the decrease in the reducing power of the intramitochondrial metabolites,

the liver cannot control the increased amino acid load. Consequently,

blood ammonia rises in spite of increased urea synthesis and stimulation

of glutamine and glutamate formation. These findings support our in

vitro studies (1a,lt) which showed that one Gf the earliesL

perturbations produced by endotoxin was a suppresion of the aerobic

metabolic capacity and thereby of the cellular energy levels.

A frequent observation following the administration of endotoxin is

depressed hepatic gluconeogenesis (2,9,23). Assuming that our observation

of iuczeased ureagenesis is a general feature of endotoxemia, then this

decrease in gluconeogenesis must occur simultaneously with increased

nitrogen metabolism. Ureagenesis and gluconeogenesis are interrelated on 2

levels, by sharing intermediate steps and through competition for ATP

(14). It is tempting to speculate that endotoxin produces hepatic

conditions which favor urea synthesis over gluconeogenesis and that the

" *" :N1' I " ' I I 1 " i I'

10.

decreased gluconeogensis often observed following ondotoxin administration

is a secondary effect resulting from increased nitrogen metabolism.

ACkNOWLEDGQENTS

This work was supported under the Office of Naval Research Contract

#N00014-84-K-0514 with funds provided by the Naval Medical Research

and Development Command and by NIH PO1-NS-17752-01.

12.

References

1. Beisel, W.R. and R.W. Wannemacher, Jr. Gluconeogenesis, ureagenesis and

ketogenesis during sepsis. JPEN 4:277-285, 1980.

2. Berry, L.J. Bacterial toxins. CRC Crit. Rev. Toxicol. 5:239-318, 1977.

3. Clowes, Jr., G.H.A., H.T. Randall and C.J. Cha Amino acid and energy

metabolism in septic and traumatized patients. JPEN 4:195-203, 1980.

4. Colombo, J.P. Amnonia mtabolism. Monographs in Paediatrics 1:1-27,

1971.

5. Czok, R. and W. Lamprecht. Pyruvate, phosphoenolpyruv~te and

D-glycerate-2-phosphate. in: Methods of Enzymatic Analysis, edited by

L.A. Bergmeyer. New York: Academic Press, 1974, p. 1446-1451.

6. Gawehn, K. and H.U. Bergmeyer, Lactate. In: Methods of Enzymatic

Analysis, edited by H.U. Bergumyer. New York:'Academit Press, 1974, p.

1492-1493.

7. Gutmann, I. and H.U. Bergmeyer. Urea. In: Methods of Enzymatic Analysis,

edited by H.U. Bergmeyer. New York: Academic Press, 1974, p. 1791-1798.

8. Hill, D.W., F.H. Walters, T.D. Wilson and J.D. Stuart. High performance

liquid chromatographic determination of amino acids in the picomole

range. Anal. Chem. 51:1339-1341, 1979.

9. Hinshaw, L.B. Concise review: The role of glucose in endotoxin shock.

Circ. Shock 3:1-10, 1976.

10. Javorek, D., W. Gruber, and H.U. Bergmeyer. Adenosine-5'-diphosphate and

adenosine-5'-monophosphate. In: Methods of Enzymatic Analysis, edited b-,

H. U. Bergmeyer. New York: Academic Presi, 1974, p. 2127-2131.

11. Kashiwagura, T., M. Erecinska and D.F. Wilson. pH dependence of hormonal

regulation of gluconeogenesis and urea synthesis from glutamine in

suspensions of hepatocytes. J. Biol. Chem. 260:407-414, 1985.

V m: '

13.

12. Kilpatrick-Smith, L. and M. Ereciiska. Cellular effects of endotoxin in

vitro . I. Effect of endotoxin on mitochondrial substrate metabolism

and intracellular calcium. Ci-c. Shock 11:85-99, 1983.

13. Kilpatrick-Smith, L., M. Erecifiska and I.A. Silver. Eerly cellular

responses in vitro to endotoxin administration. Circ. Shock 8:

585-600, 1981.

14. Krebs, H.A., P. Lund and M. Stubbs. Interrelatiuns between

gluconeogenesis and urea synthesis. In: Gluconeogenesis: It's Regulation

in Mammalian Species, edited by R.W. Hanson and M.A. Mehlman. New York:

John Wiley & Sons, 1976, p. 269-291.

15. Lamprecht. W. and I. Trautschold. Adenosine-5 -tr;?hosphate. In:

Methods of Enzymatic Analysis, edited by H.U. Bergmeyer. New York:

Academic Press, 1974, p. 2101-2110.

16. Lund, P., J.T. Brosnan and L.V. Eggleston. Tke regulation of ammonia

metabolism in mammalian tissues. In: Essays in Cell Metabolism, edited by

W. Bartley, H.L. Kornberg and J.R. Quayle, New York: Wiley-Interscience,

1970, p. 167-188.

17. McCallum, R.E. and R.D. Stith. Endotoxin-induced inhibition of steroid

binding by mouse liver cytosol. Circ. Shock 9:357-367, 1982.

18. Neufeld, H.A., J.A. Pace and F.E. White. The effect of bacterial

infections on ketone concentrations in rat liver and blood and on free

fatty acid concentrations in rat blood. Metabolism 25:877-884, 1976.

19. Nowotny, A. Molecular aspects of endotoxic reactions. Bact. Revs. 33:

72-98, 1969.

20. Rmadori, G., U. Hopf and K.H. Meyer sum Buschenfelde, Binding sitess for

endotoxin lipopolysaccharide on the plasma membrane of isolated rat

hepatocytes. Acta Hepato-Gastroenterol. 26:368-374, 1979.

""

21. Romanosky, A.J., G.J. Iagby, E.L. Bockman and J.J. Spitzer. Increased

muscle glucose uptake and lactate release following endotoxin

admin~stration. Am. J. Physiol. 239:311-316, 1980.

22. Rosenblatt, S., G.H.A. Clowes, Jr., B.C. George, E. Hirsch, and B.

Lindberg. Exchange of amino acids by muscle and liver in sepsis. Arch.

Surg. 118:167-174, 19,d3.

23. Schumer, W. ai~d ?. R. Erve. Cellular metabolism in shock. Circ. Shock 2:

109-127, 1975.

24. Wilberson, J.T., R.L. Trelstad, T. Zincus, S.B. Levy and S.M. Wolff,

Subcell~ular localization of Salmonella anteritidis endotoxin in liver

and spleen of mice and rats. Infect. Iomzun. 1:440-445, 1970.

25. Williamson, D.H., J. Mellanby and H.A. Krebs. Enzymic determination of

D(-)-B-hydrcxybutyric acid and acetoacetic acid in blood. Biochem. J. 82:

90-96, 1962.

26. Williamson, J.R. and B.E. Coracey. Assays of intermediates of the citric

acid cycle and related compounds by fluorometric enzymic methods. In:

Methods in Eazymplogy, edited by J.M. ColovL.,c. New York: Academic Presc,

1969, p. 434-514.

27. Yoder, M.C., J.M. Egler, M. Yudkoff, J. Chatten, S.D. Douglas and R.A.

Polin. Metabolic and mitochondrial morphological changes that mimic Reye

syndrome after endotoxin administration to rats. Infect. Immun. 47:

329-331, 1985.

28. Yoshino, M. Effect of Shigella flexneri endotoxin on ureagenesis and

liver ultrastructure in rabbits. Exp. Mol. Path. 32:253-263, 1980.

29. Zlydaszyk, J.C. and R.J. Moon, Fate of 51Cr-labeled lipopolysaccharide

in tissue culture cells and livers of nt.rmal mice. Infect. Inanun. 14:

100-105, 1976.

4, 02 03 +.L0+I

0X CA Cn C . co

C *

r4, C_

%C~~If *l1%m c

4)4

- 241

ca IS

cl . . . 4)EC caC Cr w CID 4J

ic CO0I ' 4IS

00 '0 InLn'

"0 *

".44

4a 0

*4) m lc0 1)to -4

-4)

Cu CuZ '~ ' In 4)41 42 Cc r- 4

In CIS 41 0 0 '0 4 0 M CIS II

C.. Cu U5)

jjjý 11

Table II lt.

Effect of Endotoxin on Hepatic (ATP]/[ADP],

Cvtosolic and Ntitochondrial Redox States

1 HIr 4 Hr 12 Hr

(ATPI/ [ADP]

Control 4.16 + 1.19 4.12 1.13 3.42 + 0.73

Endotoxin-Treated 4.17 *1.48 3.80 + 0.91 2.47 + 0.14*

13-OH- Butyrat eLAcetoacetatel

Control 2.08 + 0.73 2.36 +. 0.52 2.60 + 0.84

Endotoxin-Treated 2.26 + 1.51 2.61 + 0.35 1.38 + 0.43"*

[Lactate ]! Pyruvate]

Control 26.12 + 5.49 35.37 + 6.01 28.96 + 4.63(4:) () (S~)

Endotoxin-Treated 2S.35 + 9.89 34.39 + 10.32 44.79 + 9.85'

Metabolite conc-.ent rations were determined as described in Methods.

[ATP + ADP] -2.36 - 2.74 urnol/gr wet wt and [3-OH-butyrate * acetoacetate]

0.83 - 1.14 uimol/gr wet wt in both control and endotoxin-:treated animals.

In contr-ils, (lactate] = 1.28 - 1.39 Uimol/gr wet wt and in endotoxin-treated

animals 1.25 at 1 hr and 2.15 at 12 hr.

Values are means + SD for the number of experiments in parenthesis.-

*P<0.0l, **P<0.001 as determined using t-test for 2 means.

Figure Legend

Figure 1

Effect of Endotoxin on Plasma Amino Acids

The levels of plasma amino acids were determined by HPLC in control and

endotoxin-treated animals at 1, 4 • 12 hrs post treatment. Control - liiEndotoxin-treated - (1 hr), (4 hr) and (12 hr)

Values are means + SE. (N-5).

Figure 2

Effect of Endotoxin on Hepatic Amino Acids

The concentrations of hepatic amino acids were determined by HPLC

in control and endotoxin-treated animals at 1,4 and 12 hr post treatment.

Control u Endotoxin-treated = (1 hr), (4 hr) and

S. (12 hr). Values are means + SDI for the number of experiments in

parenthesis.

4 '.

........... A AIOiL Gin O.P'O0025 ~700I600-

10 Soo- 500-*P0025

8- 400- 400-

30 300

* 00- tOO-

IVr 4* 42I INt 46 12h IN 4% 10J

Arg Giu Asn

140- 40 2

1200

801.40

0.0 0306

0.- 0040 20-,

0.20 002