affects the &a&n of events which fba33y rem3ta ha the prodwtPon? of pfmnt and its

BeposUxAon in the hair.  The "one gem-ore cmzm" fommlatfo~ has led ti the

gimes wasgicturedas some sortofmld-and-c%& oystem,%nd so Sthasprovedto be.
But fop 8 "te?&%te" to detetintt3 the stmtW of some product f'uadamnt83..Jy dissbU.ar
                                                _ .
frcma#?lf reqtire6%lnoTe ehhoT8te pipgar%tus.  PTotew are 3QRg, lxihT%Rched ch8%lzs,

made of twenty d$fferent kids of Unks - a&m acids.  !Eheaxact 0rderofhundmAs of

these 35.x&s %ppe%lpf; to be $en&b?ally detemxbed %r e%ch of f&ousaMs of dffferent


EmA corl-t%w nucleo%iaes com2sponding to/IBU, wIith t?bymine

messe~s   goes go thepro~in-prodW~asgane3les IntJx3

the prodw%%on of 8pecfnc pMx?fur Itxoticties, acco-~~ %o

nu&?o%%des.  It does not do th-is directly, but v&z a rmnber

of ad%p%er Ilmlecules, called "%mfes l-lm'sW*  Preembly each aamino acid has one or

nare tmsfer REUI whech specifically attaches to it.  Bhnot~rpar;lPeoftktEIBnsfermA

attaches to its spec%ffc ntrclf33tide kdop1" (tziplet?) whexvmr 3% may occur 33.3 the

polyribonucleotfde chain, (2).  !Ehw e&3~hmt%n08CPd 5s h6i.d i.nitf3 prrrperp3iaeevhi&3


                                &@ 2.
                                  R.V*Eck

%bat da.-&, a.nd predfc% certati o%hem.  The tiscolmy of %d&tlon%l t;rfplets h%s
then required %'hat many of %hese hypo%heses be abandoned or much modtfied.  Judgi;ag
bythe ra%eat~chrtewo;nes havelseendiacovesed ~tapp~ars%ha%near3yall, ifno%
all. 4 x 4 x 4 = 64 mastic possible ttiplets wSll ti%ima%ely be fdentff3ed a~

COdOIlS. uu %&Se 8-Y' %O be 8 Ch80%&? jMBlblf3, 08" W%!ih &@ltZ?y f&il i&O Somc Xk?ar

p%ttemAt


kll 64 possible pe~%iono are usea*  Of %be 49 repox%ai triplets, four were discax&&.:

grea%ly reduced, bu% alI the tissing %riplets cm be accomted for ti%h only one

%x-u.e pattelm.  X% would be poW88 %o cqute a 'pmbabXW,y" that 1% cotid 'have

occurred "by cti"> bi?cau&? the d%ta obviout3ay fall 3.rzto s(pFJE? sort of pattern - they


%ppCt%XS that %hik? mi3qy d8t% could fit only ia%0 8 tm p%%%eti.  l?rmiOur;ly, w-hen

there were too few data it was possible ix devzlse a eJ.most adless am&a= of


Page 7
R. V.Eck

occur ia two dfffemrb psirs.  However, if the alanSnes at C...G were exchange8 ~281

serine and arginbe at 0.. . C, there wotid thsn be no codons of sJan2ne and serbe

ba~twolett8rsinco35on.  SincealLadne-3etieisthemst frequently-occurr%ng
"allele" pafr i(7), the a nmgement as shaam fs much pr&erred. S~mUarly, if valise

and cystew were exchanged, the nmrou3 allele pah-3 -JlGal%, v&l-ilu, val-leu, arId

others would not be producible by single-ink "&nterchanges".  @h&, with the

p~ti~ly-~y-rme31rffarPed~~t~onof~ sequence vaulx-em, constit*s sqpport

fox th%s procedure.)  m conpark~g each possM.e alternative U t&Is way tith a 39st of

   There are tiuxise OtheP unc@tstna~ aewls in table 3.  rnffexent %resao acads tight

hayebeendiscarded. Fo~e~le, @WatMm%@tbs~beenreta~d,an8
arg%nine at A?& discarded (table 2).  l!hirbkmxe, glycine at GM IT&$& exchange plscss
tith glutemine, if ft were retained at AGG.  ~emaM,however, onlya few such

alterniat~ves, and at each cbo%ce them seemed some good clue to tb& selection- Another

source of uncerkafnty is the possibility of exgerW&al error.  Of the 32 patrs, Uve
consist of an unreported [predfctedgtfiplet and a t-piplet repo*d fron only QltE!

laboratory. Anyofthe3emQIltplWe to be ZLzkePror.

  Even if there WBT~ no basfs for cho5c,g 5.n the altemat5ve pos%tlom -W&&d in

table 3, 1% would contain much infomttin &out sequence.  For each trsplet %n table 2

there are one, three, or six possZble seQv.ences.  !iT%ble 3 recaUC@S these to two

alternatives at most.

  l!hiis pattern predicts all ths amino ac%ds coded by the ram- 19 ordered

ttiplets, suggest;bragtI~ttkremaybe no '?mseme" combinations. ThQi is not a

strong i.&e3ace, howwe~.  33 one of the repopwizii ttiplets 53 erroneou3 S&s asslgned

pa2r might xepresent "non5ense".  OP, in a few ase3 the pairs m%ght be sub-dftided$

one of the two triplet3 be&q "ncmser~ewo



isrRnEfer mai of leucti 3?espom%3 ta mw ana ta wls (2).


                                   w?B 10
                                          R,V.Eck

  In retrospect, 1t se- that thla SinIple pattern could hRve lm?a d%3cc5vered with

fewer clues, and we may voIEbba3r why lt was not found earlier.  In the l,tiusttwo years


be mducedbysom aysWtPe constraTr&on%be sequences ofa&noacBds. Phis c&d
be caused try %he SaLBe xucleo%~des servfng 1pzmQre %hRn ORe trip&St sM.ly (I).


thmzeatat~,are resolvea Uto thi*-mpa32m.  The emxmdm&rofesch~frie

ldentbsl with the fir&, except that 2x1 one ponStiou a ~IW-&ZC is replawd by the othep

pe or a py2?5dm by the other pylridm.  Almost all of the r%?pom tr5pkts

fit into thb p&&m, and St pred%cts wh%ch &no acids wiXI. be found to correspmd to


5.

6.

7.

8.

9"

10.

Pmt. Nati. Acad. Scf. 48, a683 (lp62).
w--a-

R. v. Eck, 2. !i!heort?t.       . 2, 139 qlp62).

b$. chSZb;rrlti and P. &3ag, -0C. %t. &x&d. &if. b8s 81. @9;2); E. Qeassa,
                              m--m-
E. Btitsti, and 23. Oriwa, ibfd. 48, k-25 (1962).
so BPeRn%T, me. Rat. AcaZ$. 43> 6&f (3.957).
             _I---
B. E. C. Ctick, E. Bamx?t4z+ 3. l&x?nne~, and R. 3. Watts-Tobh, Xattzre, 19,

ll* R. wall, x%mxra 193, I.268 [lp62).