126

Report of Dr. Avery (::ssisted by Drs. Eotchkiss, McC~rtq and Taylor

a

       Stl*dizs rin tke trnnsfornation of pneur.lococcus (Avery, LlcCarty and
           -__---_-----_~-.--       --

T~yZor). Tk siz;;ific-?nce of the phenomenon of trcnsfornation lies Ln the

fact t.'ra t hwit-blr 2nd spxific alterations in structui-e nnd function of

unencnpul.kted (k) variants :i :I be selectively bro:lght about by highly

pwified d esox:{ribonucleic acid fractions isolnter.?. i'rom encnpsulated strnitx

of different :.psci.fic typr;. T!le ckiIgeS induced a-:: not rcndom but are

p?sr?ictabln 2nd pre~?~t~r!nil:sd F.ccording to t.lle spec<.fic ty-re of Pneumococcus

used as source of the t.rc;nffxnin,~ cubstxnce.  frVKifGrmatiC1~ imcJlV$s the

synthesis of '7. nc:'I cz;3sulcr polwrJL. c"cchr:ricle chemically cni! serologically

different frcm that !.:nor-n to be I)roduced by the encapsulated ~211s from

which the v;riant strain was derived.  Once transform::tion h:.s occurred

the newly i cquired ckxctcr; nre therc;:ft,cr transmitted to daughter cells

through innumerable tmnsfcrs in culture r.lsdia nithout further addition of

the transforming aubctanco.  From the transformed cells, c?. nucleic acid

fraction having identic;J chemical properties XC? biologica ;.cti-rib] can

again be recovered in amounts greatly in excess of thzt used t;> initiate

the reaction.

     It is evident that not only is the capsular polysaccharide re-

produced but the inducing substance which determines the synthesis and

specificity of the capsule is also reduplicated in successive generations.

     Various interpretations have been advanced as to the nature of

this phenomenon.  However, those of us actively engc.gsd in the work have

for the most 1)x-t left matters of intsrpretction to othzx and have chosen

rather to devote our time ::;ld thought to experimental analysis of the

factors involved in the reaction. This is not to say thzit we are


.

127

indifferent 2nd have not among  OW`SF: lves indulged in speculction r;nd di.S-

cussion of the relation of ti-le problem to other similar phencmenz in re-

lnte6. fields of biolo,T.  i.lor*eover , considerzSle Doubt i=s been expressed ns

to t2.e chetnicnl identity of i,lie tr2nsfcrming scbsknce.  Frcm th? beginiling

we oursi:lvcs ixivc been kecnl-r
                               ./ alert t3 the ;ws:;i?x 1
                             'I' Ly: U-at t!;n pesence of


som2  suidst:.rse ztksr tlr n t?;t: dEzso~~~ribomlclent~ in our preparations may

be respc,;ls;ibiO for >iolo~iccl. rrctivity. Thus far, the eccur.wlate< evic'ence

strongly si;sPortc the :Aief originally ,zx?.?essed that Ifa nuclric ncic! of

tk,c desoy{ribose type ic: the fund~mentc.1 unit of t!lc trznsforning sub-

stance" .  Ceverthelcss, this question is of ;~~ch inport,;ilCe that it is

hoped tlL'!t it rn%y h-3 made pcssible  tc continue the LQL-~: on this and other

chcmicA aspects of the problem.

      BioCh%llC:!l. r;t.utiiC:; ol ~2vi~om-?nt.:~l fcctors essential in trans-
                  _--_-- -_- ._- ---.-A -.E  .-`-.--                   --

formntios (Hotci-kiss 2nd Tr.yl.or). One of tl:e major problems under investi-

gEtion at, present concorns the chemic,-,l n;:ture and biological activity of

certain coaLlonentS in serum :7hich :Ire esssntinl. in the transforming reaction.

In the ,lreceding report, Cr. ilcCkrty 2escribeci ,2 method in which the

specific enzyme, dcsoxyribonuclease, is used as i: biochemical tool for

determining the time required for the transforming principle to bo taken

up or "fixed" by com2etcnt R cells during gror?th in the presence of serum.
Using this method it can be readily shorrn that the addition of the enzyme

to the reaction system at any time up to 3-4 hours after inoculation inter-

fers cith the reaction so that transformation does not occur.  It appears

thet throughout this initinl period.the transforming substance has not yet

been taken up but is still free in the mdiun r-?ncl hence subj.zct to the

destructive action of the enzyme. On t!;c other knd, if the enpJrne is

added letor than 4-5 hours after inoculation, it has no effect on the


128

subsequent developncnt 0-r trsnsformation.  The results indicate thtlt c,

critical ?i*oyiih period of from 3-5 hours in ser-m medium is required


bCfO?2 the specific desoxJribonucleate is t&en up by the R cells end
ccnssqu.ently FroLected from Cestruciion by the enzyme.

      'These ubssrvctions hxve b,- *on conf`i~~~~d kv,r the resliitn of

corolliLly experiments in 71:ich ti cells arc first gro\-in 4-5 ho:lr*s in sorum

broth alone and then traxferre2 to meoium coritr\ining the Lransforming

;ubstznco.  Under these conditions it ~LX beon found that the serum trer.ted

cells x-e now capable of taking up the transforming substclnce zithin

5 minutes after inoculction in contrast to the f,-5 hours normally required

r~~hen untrsntcd Cell:; ;Ire wed 3s iilOCUlUIil.  It therefox appears that in

trcnsformction the primary action of the serum is on the R cells as a

result of which they become capz.blc of intcrcctir,.,c vith thz transforming

princi$e.  Although ti:2 I:,xtinisrn of this action is as yet unknonn, its

offect.is manifest by the  1 *et:diness nith v:hich "sensitizodtt cells take up

the specific Cesoyrribcnuclxte. _
                 For purposes of convenience and without

:'
`i
\!

any implication as to its nature, this preparatoc: process is rofcrrad to      i/j
                                                                           i Ii
                                                          ; !~
                                                                           p I'
as "sonsitixation" ?nd the potency of whole serum or a ::erum fraction can
be defined in terms of its "sensitizing activity." It is thus apparent

                                                                           I
that, the interaction betneon competent R cells and ths specific transform-

ing substance is mdiated bjr serum, the function of which is to propare R
cells for the rapid uptake of th:! transforming principle.  As rrill be

discussed below, it has been found recently that a crystalline protoin          II
                                                                           1.1
                                 ;.i

isolated from human anC! bovine sorz poss,~, ncs sensitizing activity comparable           I

to that of the action of whole serum.

In the last report it `ylas tentatively proposed that human serous

fluids provided at lxst three important components:  a) hgglutinins for I?


129

pneumococci vihich serve the purpose of cLusing essentkl colonial aggrega-


tion.  These aggregates of agglutinated cells in turn are thought to creata

i:: their environment the 3.0~2 conditions , possibly reducing in character,


tl-!:..t are required for tr&nsfornatiJn; b) A non dialyxabl2 factor, i-+ire-

sumabl,; protein in chartct%, which hr?C: net zt that til.le been sufficiently

defined to establish its c'he:.;icni i.dsntit,y.  The fact that serum can be

ccmA21etei-- in:;ctiw. ts! by prolonged dizlysil; sugGested the wistence of

(4  a dialyzeble component.  A primary concern of this laboratory during

the past yea: has been to clarify ths nature of the last two components.

      As previously reported, a serw inactivated by prolonged dialysis

regains its activity on incubation with inorganic phosphate.  It has since

been found that reactiviticn of dialyzed :;erm can be immediately brought

about without incubation ti/ the addition of sodi:ln pyro;hosphate.  A few

organic phosphates h~v, 0 bo?n tested but thus far none has been found to

have this effect.

     As progress has been made in the fractionation of serum, pyro-

phospkte has l>rovod to be a useful reagent for eliciting the maximal

sensitizing activity of serum fractions.  A further useful step has been

the development of a semi-quantitative method of assay for the serum factor.

Briefly, the procedure is as follow:  Graded amounts of serum or protein

fractions are tested in an enriched.broth containing optimal quantities of

.pyrophosphate, purified R antibodies, and transforming substance. Under

these standard conditions, the amount of serum component added becomes the

limiting factor, and the rate and extent of occurrence of pneumococcal

transformation serve as a satisfn.ctory m ensure of the sensitizing activity

of this component.

The protein fractions tested by this nsthod were prepared either


                                                        130

bf the usual salting out Trocedurcr; or by the newer alcohol fractionation

method devised bar Professor Cohn of !hrvard Nedical School.  The evidence
        #J

clearly indiwtes that the albumin fraction of normal serum of human and

bovine origin, trn6 of $,I- P*x+:.l flui.Ls :!re zxotent sources of the non

dizlyza~o 13 .
      ;7xtoln factor.  As little 1s 3c;o micro;;rz.ms of purified serum

alburlin per cubic centimeter cz.n be the decisive: fnc%or determining prompt

::iid typic,=1 trrcsformat?on.

     That the sensitizing sctivitiy just described nay re:iso;lc;bly be

attributed to one or other o.? tk serum nlbwins iti indicxtcd 'by the

following olserv~tions.  Four pr'pnrations of crystcliine albumin isolnted

forum bovin? s-?rum by thz Coiln procedure,  and considered l,o be electro-

phoretically homsgcneous, have ~wcvzd to be a.:: active :IS Cre the crude

albumin fractions.  In ;:tidition 2 a s::mpl^ of human serum zlbumin :,rhich had

been recrystallized four time:; and found to be olectrophoretically

homogeneous was likewise active in supportirx; tr~nsfornation.

      A further indicrtion that szcsitizing activity ia a property

of serum albumin is c.ffordcd Ly t!!e following experiments.  Uhen a solution

of albumin is heated under various condiLions of hydrogen ion concentration

and temperature, its sensitizing activity dininishes progressively as more

and more of the protein becomes altered in solubility.  In very acid

solutions (pH 2) albumin is remay!:cbl.y stable. At this acid reaction

solutions of albumin m&y b e hoatsd in boiling water for 15 minutes, cooled,

and reneutralized without apparent change in solubility or sensitizing

activity.  On the other hnnd, a precipitate is formed vrhen a solution of

albumin is heated at 62OC. in the range of its isoelectric point.  The

supernatant fluid has little or no sensitizing activity.  Qhcn, however,
the arecitate is dissolved in hydrochloric acid ($I 2) and the solution


131

is then neutralized the protein regains  it s original solubility and the

sensitizing activity is restored.  Thus,  there is a striking pai~~l1elism

betweer the native, solubility of bovine serum albumin and it:; sensitizing

activity in tliC2 transfarmins system.

      ijefore testing human serum or serous fluids for their ability to

support transformation, it has b$en customary to subject them to heating

(6OOC. for :$O minutes) at their normal slightly : llr~li.ne reaction. This



WL :: done in or?er to inactivate the serum enzyme, dezoxyribonuclcase, v:hich

zould otherwise depolymnriee and thus destroy the trcnsforming act.iy;ity

of the specific nucleic acid pre sent in the transforming system.   It is not

necessary to heat pwified serum albumin for this lwpose, but when it is

heated 2under these conditions, the protein becomes more readily precipitable

at lower concentrations of salt.  Accordingly, it appears not unlikely that

the rctivity previously observed in globuiin fractions of the serum may

have been due to such aitercd albumin.

.Qthoxh alblunin is by far the most active of the serum proteins,
                                                                    *

it is not possible to str;t.e with certainty that it bears the total

sensitizing activity of whole serum. Vhile in many instances the activity

level of whole serum or a serum fraction appears to be commensurate with its

albumin content, there is now experimental evidence that sensitization by

albumin known to be active can in some day be masked or blocked by com-

ponents associated with the globulin constituents. This blocking action

may appear when an active albumin is mixed v:ith the globulin fractions of

normal serum.  It is known that normal sera, which naturally contain their

full complement of albumin, vary markedly in their sensitizing activity.

It is possible tkt their action is masked as a rcsuit of the interactions

mentioned above,  This wbloc!:ing" phcnomencn is under investigation at


132

i

i
3

present and it may PurniFh an ex$Ac.
                    'lnaticn of the variable results

proviowly obszv~3 with certain sera and serum fractions.
     Indeed, the reactivation of a dialyzed swum by inorganic ghos-

phate or pyrc:7hcsphate may, as previously suggested, be regarded as tile
restoration, directly or indirectly, of a diak;zablo cofactor.  Xltsrnative-

ly, i+, may also b;+ Looked apon as the removal 0;' 31: inhibiting substance

th'e presence of nhich is indicated by the experiments men$ioncd above. An
inhibitor i&lay combine ;;iIt.h tile albumin during tho redistribution of
constituents t!;rt follow:, the loss of electrolytes !7hen such a complex

system as serum is di:Ayzed.  Tho observaticn that albumin more or less


free of other serum constituents does not appear to be inactivated by

dialysis but is nevsrtheless in ssrnc: instances brought to higher ssnsitis-

ing activitij by thL: Ltdditioll of p;ropi;osphate, is in harmony with the

second view.

    In summary, t!lc results of analysis of the role of serum in the
transforming reaction indicate that albumin is ths principal and effective

component of sera, y!!tich, in conjunction with R cell agglutinins, supports
pneumococcal transformation.

     Transformation studies on several spontaneously appearing

'variants of Pneumococcus (Taylor).  Pncumococci exhibit marked variability

in colonial fo-rms covering a wide range of differences in contour,

morphology and surface topograpkry.  The colonial variants dis;)lay all

degrees of smoothness and roughness from the glistening mucoid to the

extremoly granular and pebbled surfaced colony.  Those p&sical features

are in many instances refl.U, +tod in recognized differences in biochemical

and physiological properties of the cells.  Considerabl.2 information is
available with regard to alterations in tb antigmic structure,


                                                      133

serological behavior and virulent, 0 associated nith these colonial char:;;ee.

It seemed possi ble that the principles and techniques of transformation

migllt afford a specific means of investigating-the hereditary basis of

this phenozrenon.

      Indeed, by ';t,t! application 2nd extension of th-? methods used in

transformation it has been possibie to determine ni.th reasonable assurance

that the intermediate smooth organisms arising from mucoid forms of Type III

represent cells in which the SIII transforming determinant itself has
become modified. These intermedinte smooth variants are reminiscent of

rllelic series ir: higher organisms.

      The property of extron, 0 roughness (91) as opposed to moderate

roughness (MR) is apparently due to the lack of a factor or factors which

can be experimentally introduced into the cells by the techniques of

transformation.  Tho idR factor is present in the desoxyribonucleic acid

fraction of the modera tely rough variant and of Type III organisms as well.
The factor (MR) which determines the change from extreme to moderate
roughness can be introduced independently of the SIII factor.  Indeed, the
ER forms cannot be transformed to encapsulated Type III organisms until
they have first been converted into MR cells; that is, the MH factor must
be introduced first, (EEL->MR-->SIII). These observations furnish
experimental evidence in favor of the view previously expressed that the
desoxyribonucleic acid fraction derived from 'QJpe III pneumococci although

chemically homogeneous contains molecules of biologically diverse

specificities.

       This form of analysis is being pursued with three objectives in

mind.  First, it is hoped that some information nay be gained on the

nature of "competence", t'hat property of cells vihich determines their


.

134

capacity to respond to a [{jven specific stimulus or factor in transformc-


tion.  Seconi!, r;uoii es?erimznts may m?ke possible further analysis of the

natur., of the s,xcificity of the transformins substance.  Third, the
demon;tratiV-n that t\-io fr.ciors, one for moderate roughness (W), the

other f'o- synthesis 0; capsular pol/sacch;ride (XII), each coexisting in

the same pr%pcration of' nucleic acid, c;in be selectively taken up by

pneumococcnl cells and independently bring about development of the

corresponding troita, justifin,c the hope that the pneumococcal cell may

prove a deiicats tool for the study oi' many hereditary characters at a

tiological level horn-.oCor *c unrealized by geneticists.

Crystellizc.tion of the C-rcactiv-? orotein (McCarty). In the

past, several studies ha-Je been rsported from the laboratories of the

Department of ZespiraLorJ Diueasns on the occurrence during acute in-

fections of a protein not normally present in the blood. This protein

is designated as the C-reactive protein since it is precipitated from

acute phase human Serb by the addition of Lhs somatic C polysaccharide of

Pneumococcus.  Although the precipitation reaction botwaen the C-reactive

protein and the C polysaccharide is superfifially analagous to an antigen-

antibody reaction, it was shown to differ from the latter in several

important respects.  Briefl;r stated, thcsc differonccs am as follows:

(1) The occurrence of the protein is non-specific with respect to tho

inciting agent of the disease; (2) The protein is present early in the

acute phase and disappears rapidly with the onset of convalescence;

(3) 1 n contrast to antibodies,  it occurs in the albumin rather than the

globulin fraction of serum; and (4) The prosxce of calcium ion is

required in the precipitation reaction 'bctwecn the protAn and C poly-

saccharidc.  The precipitation appears to be a phonomenon brscd on a


                                                      135

chance complementary relationshis between the moieculer configuration of

the Tao reactin;: substances. It is felt that the occurrence cf the

grote!n ii: the blood may be in some FXY related to the host reaction to

infection.

      The present studies doal with th2 isolation of the C-reactive

proteirl in x-pta11ine form.  HLI;~~~ serous f Luids, i.e. ,  chest and

;:bdomi!:;,l fluids accumulptin& in t:,e cour~3 of acute infectious processes,

have been employed as source material.  ?,fter initial fr22tionnticn r:ith
                                              _

ammonium sulfate, tile C-yrotain is seijarated from the bulk of serum

proteins by pecipitntion q;Iith t1io C polysaccharide.  Cissociation of the

polysacch;ritie-protoin ::ori!pfex nnd crystallization of the protein depend

on the use of concentrated sodium culfatc ;ciutions.  Recrystallization

of the material in thl: for::; of s:mm:t:*ical i-hornbold pintos occurs readily

in the presence of sodium sillfat;. at 0.7's saturation.  The concentration

of C-protein in sorum is relatively 10~1, and less than 100 mg. of re-

crystallized ;irotein was obtained from 1.5 liter:: of chest fluid.
     Prc~~nrations of the crystalline protein contain none of the

C-polysaccharid o omplpJed in the isolation, and indeed solutions of'ths

material giv, 0 the characteristic precipitLo.tion reaction when mixed with

the polysaccharido in the precance of traces  o f calcium ion. The

crystalline protein is insoluble in distilled water, but dissolves

readily in physiological salt solution.

      Evidence regarding the purity of the crystalline protein has

been obtained by the use of immunological tochniqucs. The protein is

highly antigenic for rabbits, and antisera react in high titer aith the

purified antigen as vie11 as v[ith acute phas-e human ccra which contain the

C-reactive protein.  On tn.2 other hand, the antisora give no rsaction


136

whatsoever with normal humm sex-a.  This finding indicates that no

appreciable amoz~t. of normal serum protein was present in the crystalline

C-protein used for immunizzt*ion of the cnimais. The antiserum provides

a ustlfu2. tool for further. studi.-s, sinca it serves as a sensitive acd

selective reagent for detecting the  pr esxce of C-reactive protein in


minuC,e amounts.

Publication

IWk%y, 5.  Chemicni nature and biological specificity of the
      substance inducixi; trcnsformnticn of pnsumococcel types.