÷ Ill)MAN SERI.M ALBUMIN

] .~:"

llr GO1LDoN L. NORD]~Y* ^~ ,1. MURI{AY LUC~

:;~ (From the D~partm~nl of Chemistry, Sla~tford University, Stanf~ California)

(Received for publication August I, I9~) , For sm, eral yeura Perflum’o(mt.anoie acid (1) (CP~-(CF~)~,COOH) }ms

.:;~ ~en ,sed’in this laboratory as a proteiu precipitant (2). Q~alita~ively it

.~Thas been observed ~,o bind so completely to proteins that little o~ the pre- cipitant remains in the tilt, rate; this is an obvious advantaga in chromatog- ~phy s.nd in vm’ious analygleal procedures.

The purpose of the present study is to define the conditions under which oic acid wil! effect a precipitation of hu~m ~erum albumin

to describe in de~uil the in~raetion between these two reactam~s. Kuman serum albnmin was used because of its unusuaf properties in bind-

~g a variety of ions.

M at~r~al~

A solut.ion of human ~erum albumin (decanol procedure) wa,s electro- at 4~ vol~s ~r era. against conductivity water and clarified by filtra6on ~rough a sterilizing’filter pad. The product was lyo-

stored at 3~; appropriate amounts o~ the powder were removed for the preparation of albumin solutions.

Sodium Per~uorooctanoate (PFt) aolubio~s were prepared by carefully a 0.5 per cent aqueous solution of the acid with a minimal

of sodium hydroxide.

Procedure

Precipitalio~--A number of buffered 0.3 per cent. solutions of serum al- in PF8 were prepared. Each solution was distinctive with respect

to i~s PF8 coneent~ration or to i~s pH, ~e later beir~g determined 0,I M Mcllvaine buffers (3). The pH of each solution was between

and 5.25. The PF8 concentration of each solution was su~ that ~he ratio (i%’8 to MbumJu) ranged between 0 and 200? The solutions

thoroughly mixed, allowed ~o s~and for 15 minutes, and centrifuged a ehnica ce tr fuge at room temperature for 20 rninutes. For ~onditions

gt.m protein in a given solution was only partially prec]pi~eed,

Supernatant solu~ionu were sometimes formed; in all other c~es the Public l/ea~th Service Research ~ltow of the National Heart

of 61 ,~ was .sod for ~hc molecular weight of human serum albumin

399 -~ g~

Exhibit 3013

State of Minnesota v. 3M Co.,

Court File No. 27-CV-10-28862

3013.0001 3MA00631310

4O0

supcrnat~nt solutions were clear. An nliquot of cach ~upernatant liquidI or solution i~ no prec~pit.~te existed, xwv titan :mMyzed for lm~cin b~ ~ L,~wry et aL mvditication of the Folin t.es~ (5)- PF8 d~s no~ intcrfe~

with t, his analysis. The percen~ag~ of albumin ~hat was precipitated fr0~ each solutio~ under ~he s~a~d condR.ions of pH and 1’1~ c~,ccntra%~

w~ cMcu~d. The results appear h~ Fig. 1. ’~3i Ani~ Bi~i~A number of 0.3 ~r cent albumin ~olu~ions were p~

pared in radioactive PFS. The ratio (PF8 to albumin) h~ each solu6~

w~ between 80 and 2O0. The solu~bns were tllea 6t.rated at 25° to abo~. pH 3 with hydrochloric acid. In the coume of the titration between

4 and 3, the pro[eiii precipitated complexly; severai 0.005 ml. were removed from eacl~ supernatsant liquid, placed on aluminum and immediately dried under an infra-red tamp. The disks were placed on an automatic sample changer which operated into a g~ counter (6). The time required for an arbitrary number of eomzb

recorded for the sampb on each disk. M~r making sni~able for efficiency and background interfermme ht the cmm~ing, the relative: tizzies of the s~nlples were de~rmined and compared with the activigic~ of ~heir respective soNtions before the tit, ra~io~ was

The molar ratio (bound PF8 to albumin) in eac~ sdufion w~ the~ N~d as a -fun0~ion d the molar ratio (PEg ~ albumin). The data:

given in Fig. 2. Hydrog~ !o~ Bindin~A number of 0.3 per cent albumin

prepared. Half ~-hieh the mol’,’~r r c¢l:tnined an eqn:

BOUN

5O

Fro÷ 2. Curve A

~f the number of P ~zlar ratio (album

l:m. 3, The ~f ~!~a pH of ~he s

s~@t~m PF8: Cur~’

was then tltrate The amount it, wa.~ calculate’

"01ution and r~:~t~eted by :vas calcula~

3013.0002 3,MA00631311

ha) in each

.-,.ted at 25¢ to

~,ration between

al 0.005 mL aliq~(

¯ on aluminum The disks were rated i,ato a gas umber of counts .g suit, sble nt, ing, the

,red with the ~i~ration was lation was ~hen

stain). The

40l G, lJ. NoRDBY AND J- M. LUCK

prepared. Half of the soiu~ion~ cont~ained ~hat cottcenbraI, ion of PF8 for ~chicb the molar ratio (PF8 ~o ~lbumin) was 196; the remaining solutions

contained an equivalent concentration of sodium chloride. Each solution ¯

~oo .02

200 ..?! ....

.,50 ~01~-:" .... 150

Fro. 2. Curve A, t, he molar r~tia (PF8 ~e albumin) in a system determlne~ the number of P~ pinions that bind ~o aach Mbumin molecule, Curve B, thv recip~ocM

o~ the number of F)~ anlo~ bound ~o e~ch albumb~ molecule is plotted agaio~ the

molar r~fio (Mbumia to [r~e PFS).

-5O

IOO

I

B

3013.0003 3,MA00631312

402

Pr*clpita~iun--Alt~ough Fig. ! indicates a ce~’tain critical mng~ of ~ad PF8 concentraLi0~ which mus~ prevail for the precipitation of hu~na~

serum albumin, ~t also indicates in a qualitative manner t~he under xvhic]l most other protein~ would be ~xpeeted to preclpit~e. Thr~.

rather striking features are noted for atMnticn. (i) The ccnditio~s pH and ionic strength under wttich compleM albumin precipitation ocm are very ~nild. (2) The precipi~tion of albumin is completely with respect, to pH. An albumin precipita~ formed in a PF8 solution

be completely di~olved by making the solu6on somewhat alkaline ~pH ~ 7) to the pH at Which preclpi~agion occurred. The minimal pH which the solution must be raised wilt depend, of course, upon the PF~

concentration of the solution. (3) In acidic solutions, the precipitation albumin is irreversible with r~pect to the PF8 concentration. Provided

*ha~ an albumin precipitate i~ diabmed against an appropria#~iy acidic solution, it wil! not di~olve appreciably ~ the concentration of PF8 in

equ~fibrium with the precipitate decr~ses. However, the P~ precipitator can rcadiIy be dialyzed from albumin in neutra! or slightly alkaline sol,-

dons.

Anion Bi~Min~The scope of Curve A in Fig. 2 ]s restrict~ed at the upper limi~ by the ra~her low solubility of PF8 in aqueous solution and at the lower linait by the concentration of PF8 which wotHd adeqtmtcly precipi- t.a~ albumin under t, he conditions of the experiments. Rowever, it is c]~r from t,!~e graph that a considerable m~mber of PF~ anions biud to each

albumin molecule. The molar ra6o (bound PF8 to albumins) i~crea~e~- qu~te rapidly ~ the PF8 concentration is increased. In contrast, ~he mola~ ratio (bound P~ to albumin) decre~es very slowly ~ the molar ratio (PF8 to albumin) i~ decreased below about 120. bi soIutionz conLa.i~:iag

the minimal concentration of PF8 Which is effective in con~Vletety preo?i- tat~ng a given c~neentratior~ of albumin, the PF8 is almost; completely r~

moved from the solution as par~ of the precipit, afh~g complex. The dat, a of Curve A arc tr@tcd by an exprcssion derived hv ~;lo~z (7~;

the results appear as Curve B of Fi~. 2.

~ ~ I 1

In the above equation, r i.~ the molar ratio (bound P~8 to albumin),

ghe to!!at ratio (maxima] bound PF8 ~ albumin), c is the mo]n.r raHo PF8 to albumin) in the solut.iou, and’K ~s a v~nstn.nt proportiomd to the

eqvi~ibrium vonst.ant for the PF8 n~ion-Hnd;ng reaction. The

cvidc~lcc lltat iJic equation does apply t~ l’F8 anion binding, h. can be

~ed, by virtue ~ apply to the

conc.~.ntrati mo!ecu le

Additions

some less

Curw number of P]

be calcula" acid

~36 ccta~mate

isotof bound 5o

~ydrog~m. Ion

a]bumln at app]ic~ble

is readily appa~

mel¢cu]~ by Stcir~ha~

6-trinitroresor means nf tht

of bound ~nm]3er agre~

~e albumin of

not

dotted line ~ function of

pH 4.25 at in the prcse~

~ extra hydro

beund tu ~,]w.

A and B evidcn~:e in alkalin

3013.0004 3,MA00631313

all|

4O8

, by virtue of the linearity of a portion of Curve B, tb~A, theequation

apply to the linear portion. Tim il~dicated extrapolation to infinite co~ccntration,revea,~ tha~ a m~ximumof 63 .P~?8 anions bind to an

molecule by reactions having nearly the same equilibrium c.on- Additional anions are bound ~o albumin a~ high PF8 conccn~r~tlons

some less e~sily d~scribed seri~ of reactions. At very high PY-8 con- Cu~a,e B of Fig, 2 becomes nearly vertical; ~hel~fore the max~

of PF8 anions th~ ca.n possibly bind to an albumin molecule

be calculated by the evidence available. In con~r~L, a study of

acid anion binding by Teresi and Luck (8) reveals that a to~I

oetanoat.¢ ions binds to two types of sites on the albumin molecule.

5minary i~pe dilu~.i~n stau~es indicate th~ the PF8 gnio~ ~e re- bound to the albmnin molecule.

1o~ Bi~i~g~Two hydrogen ion binding curves [or humal~ albumin are depicted in Fig. 3. These binding cu~ are com-

app]icgble Lo rever~ble ti~rations over the pH ~nge illustrated.

~ ~adily appareu~ from these curves that P~ strongly influences the molecule charge io ~cld solution. A similar effec~ has been ob-

by StehibardL (9) for the tiLratlon of wool prot,ein in ~he preserl~ of 54rinitroresorcinol, picric acid, or flavianic ~cid. means of *he equation described above, the ex~apolation ~ Curves

B to infinite hydrogen ion concentration indica~s ~h~t the maximal

of bound hydrogen ions per albumin molecule is 107 in each ~ase. number ~grces with tha~ found by Tanford (10). Thus, the charge

albumil) molecule in extremely acid s~lutioas in the presence of that of PF8 represented by a molar r~tio (PF8 ~o albumin) of

exceed 28; 79 FF8 anions ~re bound under fl~eae col~ditions. dotMd line in Fig. 3 represents the diffsren~ ~twecn Our~¢c, A and a function of pH. The maximal difference be*wren the two curves

pH 4.25 at which an extra 64 hydrogen ions are bound hy the al- fl~c !~resence of PFS. The significance of the coincidence between

extra hydrogen ions and the 63 strongly bound PF8 atdons whh’h

to the tdbumin is not known at this thne; but ~he coincidence is ~haL iL is brought t.o a.t.teut~i(m. Finally, its. is gppgren~ tha~

A and ~ coincide above pH 8~ This is additional, although incon- evidence that thc PF8 comph~ly di~ociaLes from the albumin

in alkaline solutions.

3013.0005 3,MA00631314

~n molecute binds both hydrogen ions and per~uor~.

oct~maff i~ ~orm, under ~hc conditions o~ the ezperimen~ ~ h~ s nc~ char~ of almost. ’zero.

Acknowledgment is made to the Cu~ter Laborat, ori~ for a sample

Fraction V (lt), dccanol human serum Mbumin, and to the

Mining and Manufacturing Company for a ~mple of ~rflu0roovtanoi~

acid. CarboxyMabeled (O0 perSuorooc~noic acid w~ obtained

the Atomic Energy Commission from the Minn~o{,a Mining and

tacturing Company,

Addc~dura--l~levens and E[lenbogen have recently published ’~heir research on lk~ van dee WaM~ association of bovine serum s]bumin in the presence of perflcoro seid:

(12).

BIBLIOGRAPHY

I. Kauck, E. A., and Diesslin, A. R., Ind. a~d Rn~. Chem., d3, 2~2 (1951). 2. Luck, J. M., Gri~n, A. C., Boer, G, and Wil~on, M., J. Biol. Ch,~.,

(1954).

~. Charhvood, P, A., B~ochcm. ,t., 51, ii~ (i952). & Lowry, O. I’I., P~sebrovgh, N. J., Fare, A, L., a~d ~.aoda]l, I~. $., ./. Bill

193, 26~ (t95I). 6. Ny~, w. N+, ~nd Teresi, J. D+, Arml. Ch~m., ~, ~3 (!951).

7. Klo~z, I. M., Arch. Biochem_, 9, l~ (1946)~ 8. Teresi, J. D., a nd Luck, J. M., J. Bi~l. Chem., !94, g23 (I952). 9. Steinhardq J., Ann. New ]rerk Acid. Sc., ~I; 287 (19~1).

!0. T&nford, C., J. Am. Chem. Sac., 7~, ~1 11. Cobs, E. J., Gurd, F. R. N., gurg0nor, D. M., B~rnes~ B. A., Browa~ R.

Deroua.ux, G., Giltespio, J. M., Xabnt, F. W., Lever, W. F., Liu, C. H., Mittd- man, D., Mouton, R. F.~ Sehmid, K., and Uroma, E., ,L Am. Chem. Soc.,?L

12. Klevens, H. B., and Ellenbogen, E., Discus¢ion, Fara~ay Sot., 1~, 277

UARL M. STEVENS,

(Received

fi.db~gs of Arnstein

(3-5) have estab] added to penicilli~

eate~sively incorporate that the entire cy.~

htilized, while the ~ported have failed

amino acids ioto seemed po~ibtc tha~ 1

of penicillin by valise undec different ,

oa the rel~tion~

procedures for eull, m

and the methods for

of radios

the present experimm

at. varying times after

penicillin was

irlocultllT~ wR~

were somewhat higher

J~-cysfine was o

Illinoiz, and

Company, Inc.,

re~olved hy the genarM

~esolution of labeled

~e and 350 mg

af ho~ water and crys

work wau support,ed in

~ndianaoolis, ~odlana, of Washington Init

~cp;trLinew

3013.0006 3MA00631315