Professional Documents
Culture Documents
24413-24419,1991
(0 1991 by The American Society for Biochemistry andMolecular Biology, Inc. Printed in U.S. A.
Susan L. Knock*, Brian T. Miller§, James E. Blankenship*§, Gregg T. NagleSQ,John S. Smithll, and
Alexander Kuroskyll 11
From the $Marine Biomedical Institute, the §Department of Anatomy and Neurosciences,and the VDepartment of Human
Biological Chemistry and Genetics, the University of Texas Medical Branch, Galveston, Texas 77550
Chemicalmodification of the egg-laying hormone abdominal ganglion of several species of the marine mollusc
(ELH) of Aplysia by reaction with the N-hydroxysuc- Aplysia (Chiu et al., 1979; Nambu and Scheller, 1986; Nagle
cinimide ester of biotin, which contained 6-aminohex- et al., 1988a). ELH is one of several peptides encoded by the
anoic acidas spacer, yielded seven distinct derivatives ELH gene that isprocessed from a37-kDa hormoneprecursor
that were readily separated by reversed-phase high (Scheller et al., 1983; Mahon et al., 1985; Newcomb and
performance liquid chromatography. The derivatives Scheller, 1987; Newcomb et al., 1988). When the bagcells
were chemically characterized by amino acid compo- undergo a prolongedafterdischarge of intense repetitivefiring,
sitional analysis, sequence analysis, and mass spec- ELH and other peptides coordinately
are released to influence
trometry. The seven derivatives resulted fromcombi- specific targets (Kupfermann and Kandel,1970; Stuart et al.,
nations of differential modification of the three amino 1980; Sigvardt et al., 1986). ELH acts asa nonsynaptic trans-
24413
24414 Biotinylation of Aplysia Egg-laying Hormone
TABLE I
Bioactivity and stoichiometryof biotinylated E L H derivatives
Derivative mol biotin/ Egg-laying Residue
fraction" mol ELHb bioactivity modified'
1 0 + None
2 1.01 k 0.02 (1) +
-
Lys36
3 1.01 ? 0.01 (1) LysX
4 1.94 -t 0.13 (2) - Lys', Lys'G
5 1.09 ? 0.05 (1) - a-NH2
6 2.06 k 0.07 (2) - a-NH2,Lys'"
7 2.26 k 0.10 (2) NDd o(-NH2,Lys8
8 3.01 2 0.09 (3) - a-NH2, Lys', and Lysj6
8' 2.73 ? 0.20 (3) -
From Fig. 1.
*Established by compositional analysis; n = 3 for all analyses,
except fraction 7, which was a duplicate; fraction 2was also confirmed
by FAB/MS to contain one biotinyl group and fraction 1 to contain
none.
Deduced from compositional and sequence analyses of biotinyl-
ated derivatives (see Table 11) and from compositional analysis of
tryptic peptides (see Fig. 3, Table 111).
ND, not determined.
Fraction 8 was reacted with hydroxylamine and repurified by
HPLC.
EXPERIMENTAL PROCEDURES AND RESULTS Biorinylorion Reocrionr - For most of the expcrmcnts described. E L H was chemically modified with
NHS-rh-biotin. On O E C ~ S I O NHS-biotin
~, and rulfo-NHS-.Ah-biotin wcre also employed as dircurrd 10
the text. Sulfo-NHS-Ahx-biotin wasweighed and added to the reaction solution ar a dry powder whcreai
Mrrrerinlc - Tritluuroaeetie aad (HPLC Grade), sulfo-NHSt&-biotm, NHS-Ax-hiotin. NHS-biotin,
NHS-Aw-blotin and NHS-biotin were dissolved in Nfl-dimcthyllormamide just prior to addition to the
and avidiwagarorewere purchasedfrom Pierce ChemicalCo. Acetonitrile (HPLC Grads) and hydroxylamine
reaclion mixture. Reactions were lypically carried out in 0.05 M NaHCO,, p H 8.2, a1 ZS'C for 30 minor ar
hydrochloride(Certified Cradc) were obtained from Fisher Scientific. Vydae C18 (5 ym particle sire; 300A
specified. Rsaetion~were stopped by lowering the p H to 6.0 with the addition of 1 ml of 20% acstonilrile
pore diameter) reversed-phasecolumm (4.6 mm x 25 cm; phase 218Tp) and direct mnncct guard mlumm
in 0.1% TFA. T h e pH-adjusted reaetmn mixtures were subsequently injected directly onto the C18 HPLL
containmg Vydac CIS (5 pm particic size) eartrldges were purchased from Alltech Applied Science Labs
column equilibrated a1 23% acetonitrile in 0.1% TFA.
N,N-dimethyiformamidc.acctyicholinechloridc.and biatm were obtainedfromSigmaChemica1 Co.['*C]biotin A separatc set of experiments war conducted to observe the time-course of reaaion of E L H with NHS-
(40-60 mCi/mmol) was a product of Amersham. Reagents for amino acid mmporitional and wquenec
Ahx-biotin. The molar ratio of reagsnt:psplide was 401. Other reaction mnditionr were the Same as
analyses. includmg HPLC of Plh-amino acids. were from Applied Biosystems. [ne. r-Bulyloxyc'arbonyl amino
staled above.
acids and Other chemicals for peptide synthesis were purchased from Biosearch. Lyophilized trypsin war a
HPLC - Rearlion mixtures containing biotinylatcd E L H were applied to a Vydac C18 reversed-phue
product of Bwhringer Mannheim. Aptpie califomica (200-300 g) were obtained from Alacrily Marine
HPLC mlnmn and eluted with a linear gradient of ~olvcntA (0.1% 1"fluoroacctie acid) andsolvent B (100%
Biological S e ~ e e s(Redondo Beach. CA).
acetonitrile containing 0.1% trifluoroacetic acld) at a flow rate of 1.0 ml/min. The standard linear gradient
ELHSynrherir -Synthetic E L H was prepared in the Biosynthesis Laboratory a1 the Univcrslly of Teras
smploysd to fractionate E L H and 11s biotinylatcd derivatives was 2343% wkent B i n 100 min at Z'C. The
Medical Branch by solid-phase synthesis (Mernfield. 1963) using a fully automated synthesizer (Bioseareh
column eluate was monilored at 215 nm and 0.5-minfractions wsre collected. Fractions were pooled hared
Model 9600)and employing r-bulyloryearbonylchemistry. Reactive amino acld side chains were protected
on absorbance and were rvhjcctcd to amino acid compositional analysis. automated microJequsncs analysis.
ar follows: Asp and Glu with 0-benzyl ester; Scr and Thr wilh benzyl ester; Tyr with 2-
FAB/MS. and bioassays. A l l fractions were dried by vacuum centrifugationand stored at -2OC.
hromobenrylorycarbonylester; k g with 4-ralucnesulfonylester; andLys with 2-ehlorobcnzyloxyearbonyl
Amino Acrd Analysi$ . Sampler were applied directly to an Applied Bioryrtems
420H
errcr. T h e pepfrdc was r~multaneourlydeprorccted and cleaved from the resin by treamcm with H F (Stewart
(derivatizer/hydroly2er) amino acid analyzerwilh on-lme actd hydrolysisand pre-columnphcnylthmcarbamyi
and Young. 1984). Synthetic E L H was subsequently puriflsd on a Vydac C18 reversed-phase HPLC mlumn
derivatiration. T h e number of biotin groups ~ o v a l ~ n tbound
ly to pcptidcs war determined by amino acid
as de5crihcdbelow usmg a linear gradlent of 0.1% trifiuoroacctic aeldand aceLonitrile containing 0.1%
analysis of the 6-ammohexanaic acid spacer arm portion of the sulfonated and non-suifonated N H S - t h -
trifluoroacetic acid. E L H authenticity was confirmed by amino acid compar~tianal analysis
and
btotin rellgentr (Smith ef nl.. 1991a).
mmrorequencing (rcsuIts not shown) on an Applled Binsystems Modci 475A sequencer.
Biotinylation of Aplysia Egg-laying Hormone 24417
Ar,omurr.d Psptrde S q ~ r e r # c eAndyrir - Pumrry wucturdl mnlyr~swa\ drtermcncd urmg an Applied M~~ Speom,,wnc Anvlyru. HPLC-purifwl synthetic ELH and hiutinylatcd ELH (fracrlans I and 2.
Eio\y$rcms47% protcin/peptide mirrorequenrcr with an o n - l m M d d 120A mirrohore phenylthwhydantoin 8)were anslyzedby FABfMS. The d8aina molecular ion clusters observed in these
re\pectively: Fig. 1
amino a c d analyzer and a Model 900A drla prucerror as previously descrihed (Nagle ef ol., lY88hl. p - analyses embitshed rhe nurnher uf hmtin moieties present in these Two fractions (Fig. 18). The spectNm uf
,Ah*-hiotinyl-Lyr war ldentihed and quantified as described hy Smith el ul. (1991h). syntheuc ELH indicated a protonated mole~uleat m/z 4385.1. consonant wth the SlNClUrC of ELH (results
Mars Sprcrrom~rry- Mars rpertromctrie analyses were performed at the Mass Spcctromclry Facil~ly. not shown) Similarly, the ipcctrum of fraction 2 indicated a major ion at m/r 4724.5which w u in agreement
Department of Chemistry.Mawachusettr Institute of Teechnolugy. Cambridge.Massachusetts. FABfMS wth the calculated value for monobiotinylated ELH.
analysis of ELH and its hiottnylnted derivative was carred out in the first of two mass spectrometers of a Ammo Acld Comporltionol and sequence Andyscs. T h e stoichiometry ofhmtinyl groups bound per mol
tandem high-resololution mass spectrometer (JEOL HXIIOfHXIIO) a$ prev~ourlydescrihed (Barher cf ul.. ELH dsriwttvc was quantitated hy analysis of 6-ammoheranoic acidin fractions 1-8 (Fig. LB:Table 1) using
1981) pre-column denvatizauon ammo and analysis (Smith el "1.. 1P91a). Compos~ttonalanalysis gave identical
Hyhqlamme Rpocfion .Hydroxylamme hydrochloride ~ o l ~ t l o n
(0.5
s M in 0.2 M trimethylamine) were results as that obtained from F4B/MS for fractions 1 and 2 which provided solid evidenrr that fraction 2
adjusted to pH 9.0 with NaOH. HPLC-purified and lyophillzed. hiolmylated ELH derivatives were d ~ o l v c d eontamed monobiotinylated ELH (Table I).Microsequence analyss demonstrated that this derivative was
in 1M pl of 20% acetonitiil~and 50 +Iof thk m m u n was added 10 0.45 ml of buffered hydroxylamine monohiotinylatcd at Lyi" (Table 11). Tho peptide was not blocked at the amino ternnus and N6-Ahx-
T h. e reaction was mired and allowcd to proceed for 4 h at 25°C. The p H was tested at the end of
~Iut~m biolinyl Pth-Lyb wa5 not detected at cycle 8 hut was detected at qcle 36 (Smith el d..1991b).
!hecncuhation period to verify that i t had rcmamcd unchanged. Trifluoroacetir acod (500 PI of 0.1%) was Curnpor~tionalannlyr~rof the hialinylatcd ELH derivative in fraction 3 (Ftg. Is) indicated that this
then added to the reac~ionmixture prior TO >nlectlon onlo a C18 revendphase HPLC wlumn. For a fraction wa5 a l m monohlotinylated(Tahlc I). Microsequence analysisdcmonslrared that this derlvative was
comparative control. !he remaming Sop1 of each biotmylatcd ELH derwative war chromatographcd withoul bmtinylrted at Lyr'(Tablc 11). Similar analys~sof the hiotinylated derivative I" fraction 4 (Fig. 18)indicated
hydrnxylamme treatment. that this fraamn was dihminylated (Table I).
Microsequence analysis revealed that
it was unblocked and wa-
Tppyp,ic Hydrotym - ELH and hiotinylated-ELH dcrtvativer were hydrrblyzed with trypsin at an enzyme hiut~nylatedI I Lyse
and Lys" (Table ll). The hiaunylaled derivative I" fraction 5 (Fcg. IS) w&%
to ruhitnle ratio of 1:lO (w/w) for 8 h at 3 7 T ~n2SU PI of 0.05 M trls-HCI buffer, pH X.1. After reaction munuhiotinylated at the amino terminus since microsequence analysis of 550 pmol indicated only 8.7 pmol
the sampler were acidified to pH-3 with lr!fluoroaeetic acid and appled directly 10 P C18 reuerxd.phase uf I l e confirming that lhls derivative war hlocked. Comporitional analysis of fraction 5 milcared one hiotln
HPLC column. HPLC conditions were a$ descrihed above and included a h e a r gradient of 8 4 5 % solvent momty (Table I).Similarly. the dihiotinylated derivatives m fractmns 6 and 7 (Fig. LB: Table I) were also
B o v a 135 mm followed hy 45.100% 101vent E for IO min Pooled fractions were lyophilized and subjected blocked ai ihc ammo lcrmmus (II2 pmol and 393 pmnl \cqucnred, rerpectiuely; 2 8 pmol and 23 pmol lie
to amino acid analysis. recovered, respectively). S~ncewe were unable to unamhiguuusly distinguish whichof thc lysyl residues w a
Avidin Bhdlng Arrq - To establish whether or not hiotinylated ELH derivatives would hmd avidin. rpeciflcally hmtinylalcd m fractions 6 and 7 by sequence analysis, lheac two derivatives were ruhscquently
we employed a mmpetitive binding assay wich ['%]biotin smilac Is that deswbed hy Vskshmamurti el el subjected 10 trypuc hydrolyss and fraaionation by HPLC (Fig. 3). Compositional analyrls of the wlatcd
(1974). Avidin-eoatedagarose berdr W C ~ Fadded 10 0.01 M phosphate bufferedsaline. p H 7.5, and mcuhated peptides cunflrmed fhal fraction 6 war dihmmylated on llc' and Lys" whereas fraelwn 7 was dlhrutmylated
with EU1-Lyr-3kAhr-biotin. unmodified ELH. or excess hioun for 5 min at 2YC. ['Tlhiotin (-1.4x I(P on Ile' and Lys'(Tah1e Ill) Finally. compositional analysis ofthe derivative tn fraction 8 (Fig 18) Indicated
dpm) WPI then added and incubated for IO min. Thhe beads were mllscted by vacuum filtration onto glasz that this der8Yatwe war rribiulinylated (Table I).Mtcrosequence analysis confirmed that it was blocked due
filten and washed er.tensively with phosphate buffered saline. After a final wash with 100% ethanol, howd 10 hiotinylation of the amino tarmmus (136 pmol sequenced,4.7 pmol Ile rccmcrcd).
radiomtivily was determined m a Tracor Analytic Scintdlalian Counter. Significanl differences in bindmg Hydrolylrrmlne Tr.?mmenr of Blotinylufed ELH Detivoriver. To verify thatthe hiotinyl moieties were
between groups were determined by analysis of variance. attached v u amide linkagesrather than by ester linkages to hydroxyamino acids. as shorn for gonadotropin-
blade sliver and the exposed ganglion was superfused at 2S'C with artificial seawater at a rate of ahout 1 derivatives 2. 3 and 5 (Rg. SA). At later time points, the recovery of these dewatives dccreavd somewhat
ml/min. The raft war often slwed or stopped for apprormatsly 9-15 mi" at the time Of peptide application as they were converted to the dihiotinylated derivatives 4.6.and 7. and finally lo the tribiotinylated derivative
and then resumed. No protanare inhibitors were utilized. Neuron RIS and for one of the LB or LC clu~tcr x (Fig. 58). The tom1 HPLC recovery of 811 modtfied and unmodofied ELH spccier decreasedsomewhat
neurons ~n the left Iowcr quandrant of the ganglion were penetrated for intrarrllular recording using (-7.5%) with lnercarlng reaction time (Fig. SA). Ths reduction I" recovery after HPLC was likely due IO
the increase ~n peptide hydrophobicity P I more hiotinyl groups were added to the ELH moieculc. The
procedures prcviourly described (Rock el u l , 1986). L B / K neurons were identified by their topography, size.
color. rpantancaur aelivily, and m some cases. response IO applied acetylcholine. ELH sampler in artificial combined recovery of hiutmylated ELH modified at Ly? Lyra,or the amino Icrminus is Shawn in Figure SC
seawater (25 to 250 4 ) were popelled into the hathlng medium directly ahhove the ganglion. giving a finalbath and demonhtrateda relatirely rapid hiotinylamn of Lys" and to a lesser ~ x t e n tLyra. The bootmylationof the
concentration of IOb M 10 10' M. u s rurprwngly low and was ahout27% that of Lyr' after 30 mi"
ammo f ~ r m ~ n was Thhe Iummalion rccoverlec
for biotinyl Lys-36 and biotinyl Lyr-8 shown in Figure 5C did not include fractions 6 and 7: however. as
RESULTS lndicatcd m F~gure58, the 0vcraII contnhution of these fiaFtionS was relatively small (<5%1 Derivatives
m fractions 6 and 7 were moslly converted Io the trib~otmylatcdform after 15-20 min of reactlon.
Biolinylorwn of ELH - T h e lime-course reaction of ELH with NHS-ch-biotm was evaluated by HPLC Avidin 8~nd;ngRrsq. The ahilily of hiotinyl Lyr-36-ELH to bind to avidin and compete with biotin for
as shown in Figurc 1. Multoplc peptide derivatives were wmitiltsntly observed at the elution poritiom binding to avldin is shown in Figure 6 . Excess cold biotin and biotinyl Lyr.36.ELH rignifieantly reduced the
illustrated durlng the chromatography of many different hiatinylatmn reaction mklurer carried out under ahmy of ["Clhmtin IO hind to awlmagarose beads (p<O.OS), whereas unmodified ELH did not inhibit
valmus reaetmn condniom. T h e relative amount of each derivative formed.however. was dependent on ["C]biotin binding.
biotinylation reaaion mnditionr. Amino acid wmposilional analysis indicated that eachpeak contained a Egg-Loymg Bioauq. Synthetic ELH was found to he equipotent to native ELH in an egg-laying
pcptide identical in campition to the origlnal E L H Compasitionalanalyses of fractions 2 through 8 gavc bioassay. Biatinylatsd derivatives 2-6 and 8 obtained by HPLC purification of products of the reaction of
positive wnfirmalion for the presence of bmtm whereas fracnon I. whwh eluted with the Same retention time NHS-Ahx-biotin with ELH (Fig. ZB)were tested for egg-layingaclivlly by inlection of 1.5 m o l into recipient
as the unmodified synthetic ELH. did not Contain biotin (Table I). Within the first 5 min of the reaction, animals. The E L H derivative in fraction 2 consistently elicited egg laying (n =3) whereas all other derivatives
denvativcr in franions 2, 3, and 4 occurred on relatively large amounts (Fig. LB). By I5 mi& derivatives in were inactive at the same dow (n=3). Subsequently. fraction 2 andsynthettc ELH were mmparativcly
fractions 4. 6, and 8 had Increased and there was a concomitant decrease in fractions 2 and 3 (Fig. IC). bioawaycd using 0 5. 1.0. and I 5 n m d of peptide Both produced tg laying at 1 0 and 1.5 nmol but no1 a!
A comparative series of reactions of ELH with NHS-&-biotin at dtfercnt molar ratios of 0 5 "mol indlcaringthat derivative 2 wa equipotent lo synthetic ELH. Two days later. ind~ndual
animals that
reagcn1:pcptide (+I, 8 1 . and 401) resulted in the HPLC derivative distribution rho- in Figure 2. The were injected with fraction 2 were tested with the same dose of synthetic ELH and thore mnjected with
reanion of E L H with NHS-biotin (without <Ah*) under ldsntical condmons as for NHS-dmbiatin (Fig. 2.4) synthetic ELH were thentested with fraction 2. Under these mnditions. fraction 2 and synthetic ELH
resulted in a greater degree of hiorinylatwn as evidenced by the increase in relatwe yields of later eluting produced an egg-laymng rcrponse at the same dose (1.0 nmol). A l l peptide concentrations were determined
ELH reaction products (Fig. W ) . The predominantderivatives obtained were dihiotinylared (Lys'and Ly?; hy ammo acid analysis.
fraction 4) and tribtalinylalcd (e-NH, Lyra, and Ly?: fraction 8). There results Indicated that the NHS- Electroplyslolo~colEffects on Identified Neumnr. The comparative hioacuvily of hiotinyl Lys-36-ELH
hmin reagent was more reactive than the NHS-Ahx-biotin reagent andfor less rusceptihlc to hydrolysis by and ELH was also lcrled eleelrophyriologically.Peptideswere applied ID the bath surrounding theahdommal
ganglion whlle recording from identified neurons whose response to ELH has heen well documented (Mayeri
water. On the other hand, biotinylation prof~lesusing sulfo-NHS-Ah-blotin were very similar to those using
the "an-sulfonated reagent. NHS-rAhx-hiotc (rcsullr nut ihown).
el d 1979a,h, 19x5). Peptide\. soluhllmcd in 1U% ethanol and diluted with artiIiCia1seawnl~r,were applied
24418 Biotinylation Egg-laying
ofHormone
Aplysia
ID the bath by automalcc plpeucr; concentrationsgiven were final for thebath. Vehicle alone (artificial
seaya~erand ethanol) had no effecl on the neurons. Blotmy1 Lys-36-ELH had effects on cell RIS and LB
and LC cluster neurons indistinguishablefrom those causedby a bag cell afterdischargeor by ELH. As show
in Flgure 7. biolinylatcd ELH (1 r M ) produced more inlenle firing of Fel RIS and induced finng in an LC
c l u s w neuron. The number of spikes per hunt and the interbunt hyperpolarizationincreased in RLS. and
the previously silent LC cell began to fire steadily. There c f f e a psrrirtcd for aver I h. Figure 8 illustrates
results from another experiment i n which a comparison was made of the effects of ELH and biolinyl Lyr-
36.ELH on cell R1S In this case. IO ELH and biotmyl Lyr.36-ELH eachcaused comparableand strong
madhcations in the firing of lhir cell. AI there concenlratms each peplade caused a greatly accentuated
intorburst hyperpolarinatton and the cffsetr perrizted for prolonged pciiodr. Cell L8. Icprcscntative of the
LB cluster. was a150 equivalently rtmulatcd by equal concentratiem of either bmtinyl Lyr-36-ELH or E L H
(Fig. 9). In this experiment neuron LE displayed moderately hbgh sponlancour. irregular firing that w s often
inlcrruplcd by interneuron activiry. After several minuter m the prcsencc of 2 rM biotinyl Lyr-36-ELH or
synthetic ELH. Ihe ftring pnltern was changed to prolonged b u m of interne firing that continued for aver
1 h after washout Thus, comparable results with ELH and biotinyl Lys-36-ELHon LB/LC nCUrON and R1S
were seen in three Of three experiments
klr""' cleavage sitcs arc indicated by vertical arrows and theoretical resulting peptides are designated "a"through
"g". Actual tryptic peptidcr obtained from hydrolyzates of HPLC fractions I. 6, 7, and 8 are as indicated.
Dipeptides corresponding to regions "c", '"d".and "g" were not idenltfied and likely eluted early with (he salt
fraction. HPLC fractions 1 (uobiotinyhtcd) and 8 (sibiotinylatcd) served as controls to establish reternion
times and mnfirm susccptibllily to trypsin hydrolysis. B) Comporlre of observed retention times of tryptlc
,. . . . .
0 60 65 70 75 80 85
Elutlon time(min)
F@ 2 R~rullrof blollnyl~llm
ofELH at vwlaur reagent:pcplldc ntlos. HPLC analyses of the reaetion
(30 min. 292) of NHS-rh-biotin with ELH at molar ratios of: A) 4:); 8 ) SI; C) 401: and D) HPLC
analysis of thc rcactlon (30 min; 25°C) of NHS-biotin with ELH at a 4:l molar ratio yielded primarily two
pcptidc derivatives as shown.
Biotinylation Egg-laying
ofHormone
Aplysia 24419
Fraction 4
bzz
1v Ile I26 211 162
Illy I19 148
75 YY 86
Ill3 120 160
X8 146 11Y
I4 I24 150
S2 153 131
Y3 135 ill
71 135 123
2 62 95 84
66 ~ox 97
54 874 103
0 51 Y4 74
;-11 : : : : : : 3s 71 91
0 5 10 20 15 25 30 IJ 20 17
Amlno a c d 64 6b 6e 6g 7ai7h 7g
0 11.1
0.9 ( I ) 1.1 (I)
0 0.1
nl 0.2
n 0
11 0
0 0. I
O 0.1
n 0
n n
0 0
Methionme 0 0 0
6-aminohexanoicrcd' 1.0 ( I ) 1.0 ( I ) 0.1
lP"le"Clne 1.9 (2) 0 0. I
Leucine 1.1 (I) 2.0 (2) 2.1 ( 2 )
Phenylalanine 0 0 0
Lysine 1.0 (I) 1.0 (1) 1.0 ( I )