Download Eight bacterial proteins, including UDP-N

FEMS MicrobiologyLetters97 (1992)249-254
© 1992Federationof EuropeanMicrobiologica!S~:ietiest)3"/g-109"//q2/$t$.(g;
Publishedby Elsevier
249
FEMSLE 05(195
Eight bacterial proteins, including UDP-N-acetylglucosamine
acyltransferase (LpxA) and three other transferases
of Escherichia coli, consist of a six-residue periodicity theme
Martti Vaara
De.partmem of Bactcriolo,gs"and Immtmoh)gy, Unirersity of tlelsb~ki, th.l.smki. Finland
Received 17July 1992
Accepted 22 July 1992
Key words: Ipx.A Gene: ssc Gene: Lipid A biosynthesis: Hexapeptide repeat theme
1. SUMMARY
Only a few prokauotie or eukaryotic enzymes
are known to consist of a tandem-repeat structure. This report describes a common hexapept/de-repeat theme in four Escherichia colt transferases and in four less-characterized bacterial
proteins The proteins are the Ssc protein of
Salmonella typhimurium (25), UDP-N-acetylglucosamine acyltransferase of E. coli (24), the hypothetical proteins Tins of Bacillus subtilis (23)
and Yglm of E. colt (22), succinyidiaminopimelate aminotransferase of E. colt (14), serine
acetyltransferase of E. colt (13), NodL of Rhizobium leguminasorum (13), and thiogalactoside
acetyltransferase of E. colt (8) (number of repeats
indicated in parentheses). In UDP-N-acetyiglucosamine acyitransferase, the repeats constitute
55% of the total protein. Each hexapeptide re-
Correspondence to: M. Vaara. Department of Bacteriology
and Immunology,Universityof Helsinki.Haartmaninkatu3.
00290 Helsinki,Finland.
peat of the eight proteins starts with lie, Lcu, or
Val. Position b is occupied by Gly, position d by
Gly, Asn, or Asp, and position e by Val or Ala in
52%, 54%, and 56% of the hexapeptide repeats,
respectively.
2. INTRODUCTION
Tripeptide GXY periodicity in collagens, the
PGV and PGVGV periodicity in elastins, and the
Gin-, Pro-, and/or Leu-rich oligopeptide repeats
of cereal prolamines are well-known examples of
the occurrence of tandem repeats in structural
and storage proteins. Others include the imperfect a-helical heptapeptide repeats in coiled-coil
proteins [I,2], the glycine-rich nonapeptide repeats in cytolytic RTX toxins of Gram-negative
bacteria [3] and the flexible hinge sequences in
immunoglobulins [4]. Zinc-finger and Lcu-zipper
proteins as well as Ca2+-binding proteins with
EF-hands contain functionally specialized domains with oligopeptide repeats. Repeats of
longer oligopeptides (often longer than 10-15
residues) are a common feature in proteins involved in recognition processes. Enzymes of either
prokaryotic or eukaryotic origin which are known
to consist of a tandem repeat s,.ructure are extremely scarce. They include certain carbohydrate-binding streptococcal enzymes [5], the bacterial transfer RNA synthases with the coiled-coil
structure [2], and RNA polymerase II.
This study shows that five bacterial transferases, including an E. coli acyltransferase
(UDP-N-acetylglucosamine acyltransferase; ref.
6), two E. coil acetyltransferases (thiogalactoside
aeetyltransferase, serine acetyitransferase; refs. 7,
8), a probable acetyltransferase (the nodulation
protein NOdL of Rhizobium leguminasorum; ref.
9), and an E. coli aminotransferase (succinyldiamino-pimelate aminotransferase; ref. 10), contain multiple successive imperfect-hexapeptiderepeat units. In UDP-N-acetylglucosamine acyltransferase, these hexapeptide repeats cover 55%
of the total length of the protein. A large number
of hexapeptidc repeats bearing this theme also
occur in a recently discovered [11,12] enterobacterial protein Ssc, as well as in two hypothetical
bacterial proteins.
After this work had been completed, another
report [13] described 26 hexapeptide repeats in
the E. coli analogue of ssc.
3. MATERIALS AND METHODS
FASTA searches [14] were made by using the
University of Wisconsin GCG program package
[15] (version 6.2) and the National Biomedical
Research Foundation/Protein Identification Resource ( N B R F / P I R ) protein sequence database (release 26). When testing the theme (IGXXXX)61G, ktup 2 was used. The amino acid
sequence of UDP-N-acetylglucosamine acyltransferase and the Ssc protein were deduced from
GenBank entries M19334 [6] and M35193 [11],
respectively.
All protein homology alignments were made
using the H1BIO PROSIS program package (Hitachi Software Engineering, Yokohama,
Japan, ktup 1) and the SWISS-PROT database
(release 14) included in this package.
The secondary structures of the hexapeptide
repeats were calculated by the method of Chou
and Fasman [16] by determining the parameters
P,, Po, P,, and p,.
4. RESULTS
4.1. Six.residue periodicity
The Ssc protein of Salmonella typhimurium is
a new bacterial protein, recently discovered and
sequenced in this laboratory [11]. The sequence
of its analogue in E. coli, the FirA protein, is also
now known [17]; it is 96% identical to Ssc. The
function of Ssc/FirA has not yet been elucidated
but the protein has significant homology to E.
coil UDP-N-aeetylglucosamine acyltransferase
(the LpxA protein; ref. 18), which catalyzes the
first fatty acid transfer step in the biosynthesis of
the lipid A component of the outer membrane
[19]. Furthermore, both Ssc and LpxA have been
shown [19] to have sequence homology to the
'bacterial acetyltransferase family' [20,21] consisting of E. coil thiogalactoside acetyltransferase
(LacA) and E. coil serine acetyltransferase
(CysE), as well as of the NodL protein of R.
leguminasorum.
It was now realized that Ssc has regions notably rich in isoleucine residues and, more strikingly, that these residues were often regularly
spaced. The distance between successive lies in
Ssc was six residues in ten out of 33 occasions
(Fig. la). Intriguingly, a similar six-residue
periodicity was found in the isoleucine residues
of NodL (Fig. lb).
The isoleucine periodicity, although less pronounced, was also observed in LpxA, LacA, and
CysE (data not shown). In those proteins, as well
as in Ssc and NodL, the periodic lie was often
followed by Gly, and was occasionally replaced by
Leu or Val. An [I,V,L]G dipeptide residue was
found altogether 54 times in these five proteins,
(16, II, 11, 9, and 7 times in Ssc, LpxA, NodL,
CysE, and LacA, respectively), and its occurrence
conformed the six-residue periodicity pattern (Fig.
lc). In only 26%, the distance between the successive [I,V,L]Gs was not proportional to six (10%,
27%, 30%, 33%, and 38%, for LpxA, Ssc, N o d e
LacA, and CysE, respectively).
251
"ol
t
l
o
RESIDUE DISTANCE
10
011
1
•
•
RESIDUED~TANCE
,:tl
0..
I..li
,....
~a
c1
RES~(JE OI~TANCE
Fig. I. (A) Periodic occurrence of lie residues in the Ssc
protein of S. o'phimurium. The distance between each successive lie residue in Ssc was calculated. The figure shows that in
ten occasions, the distance was six residues (i.e. lie residues
were located in relative positions + I and + 7). (B) Isoleucine
periodicity in the nodulation protein NOdLof R. leguminasorum. (C) Periodic occurrence of dipeptide residues IV, VG.
and LG in S~, NodL, and the E. coil proteins thiogalactoside
acetyltransferase (LacA), ~rine acetyltransferase (CysE), and
UDP-N-acetylglucosamine acyltransferase (LpxA). The
[i,V,L]-[I,V,L] distance between successive [IV,L]G residues
in each of the five proteins were calculated, and the data
combined.
The hexad repeat arrangement of the five proteins is illustrated in Fig. 2a-b. Twenty-five more
or less identical hexapeptide units, each starting
with lie, Leu, or Val, were found in Ssc. Characteristically, t h e ~ units were clustered to form
long stretches: they covered 44% of the total
length of S ~ . The LpxA protein, aligned by
PROSIS to reveal homology with Ssc, carried 24
analogous and similarly clustered hexads (covering 55% of LpxA). By identical criteria, NodL
had 13, CysE 13, and LacA 8 hexad units.
Other proteins which have been shown [19] to
have homology with Ssc and LpxA include the
hypothetical protein Yglm of E. coil [22] and its
Bacillus subtilis analogue Tins [23]. Yglm is encoded by the flanking region of elmS (glucosamine phosphate isomerase) gene. Both Yglm
and Tins carried 21 hexad units (Fig. 2c). Finally,
limited homology has also been found [19] with
succinyldiaminopimelate aminotransfcrase
(DapD) of E. coil [10]. it had 14 hexads (Fig. 2d).
The
theme
(1GXXXX),,
used
as
(IGXXXX)61G, was then employed to screen all
N B R F / P I R database proteins by the FASTA
program. This theme was able to depict from all
database proteins Ssc (10.0), NodL (10.0), LpxA
(9.8), LacA (9.5), and Tins (8.2) as the best
matches (values in parentheses indicate the deviation [in SDs] of the optimized score from the
mean initial ,score of all database proteins). These
were iollowed by bovine collagen a-2 (8.0).
Altogether, 139 hexad units were found in the
proteins shown in Fig. 2.
4.2. Amino acid occurrence in the hexads
The occurrence of amino acid residues in different positions (a through f ) of the hexads is
shown in Table 1. Besides the specificity of a and
b, a clearcut feature was the predominance of
Gly, Asn, and Asp in position d and that of Val
and Ala in e. These findings were consistent for
the hexads from each of the eight proteins (for
details, see the footnotes of Table 1). Eighty-six
percent (107/125) of the glycine residues occurred in b or d, and only 3% in positions e or f.
Only 7% (8/110) of Val residues occurred in b,
c, or d. Eighty percent (16/20) of Pro occupied c
252
A
SSC:
~LPSIRLAD]J~OL~AZLHGD~D I V I TGVASMOSATTGHITFMVNPK¥ REHLGLCQASAV1/M'DODDLP FA~Sk,ALV1/KNPYLTYA~ I L D ~ P ~
L~XA:
SSC:
JIDATATI~ISNVS ~VIV~#rkNAV
~ E , , ~ I~DNVV .LGAGCF ~KNSK I~GSR ~
~PSAV
(105)
M ~D KS~.F
(?)
,IYHDI0.L~ENCL ~QSSTV,J~N)GFGYJ~ND~GNNVKI~LGRVI
(20;*)
ooo.o..,..,o.,,.o.o..,..,,o,,...,,o.
$SC: ~RP'[TEgGVYS 5G1pLOPN1On~KTARLVHNZDDHSKRL~; = ' ~ O
(341)
UPXA: ~AOOV1PPWIAOG~HATPFG1/NIEGL~RRGFSRE&I~AI RNAYKLI YRSGKTLDZVKPE1A£ I J ~ T Y P E V I ~ F ~ F F ~ S T ~ L I R
(262)
B
U~C&:
NCX)L:
C'~5E:
MNY~HTERI P,N3K LF~DRCEGLPEKRLRGKT~*J~
m" R-~KZ ~ML&GE/4YNArDPEI OREL L LTGANLK
HSCEEL£IVI~NNI KA£ARTIJ~DCEPHLASFYH/~TLLKHENLGSAL5Y,v~/~NKLSSPIMPAI AZ R£VVEEAYAADPE M I A S A A C O Z Q & ~ P A ~ K ¥ S
(34)
434)
(100)
1J~CA: EFNHSHPSEV~KRESLI KEMFAT ~ E I ~ N V£ PPWFS TGSNIH VGRN~y ANFNLT IVDDYW'~ Z~GDNVL~JUPN'VT13I'SV"/GHPVHHELRKNGEH ¥SFPIT
(133)
NODL: R1~)TLGDSAERNNGLL
~ERLGA~LGAV ~R PPFHCDYGFNIR~I~AWVY~N~'NCV~LDVAAVTI~DGTA~PAVO ~¥TADHPHDPEQROAGLQ
~RPVS
(135)
CYSE:
TPLLYLKGFHJ~,OAYR IGHNLI~1~GRRRL~ IF ~NOV$ VTFQVD ~HPAAK ~IRGIH •I*Df~TGIV ~ £ T A V .~£NDVS 11I~OSV~mLGGTGKSGCORHPK
4195)
IJ~C~:
~.~k~VW ~ 5 h ' V l ~NI~'I~ ~.~DNSV J ~ G S I ~KDIPPNVVAAGVPCRVIREINDRDKHYYFRDYKVESSV
(203)
NODL: I~GRHAN~.~GGAI ~LPGVT J..GDI~V I~A!35V mVTRDVpAGSTM4GNPARVK~GGRLPKS
(ZgO)
C~5£:
(2?3)
~R£~
~k~/kCd~K~LG~IE ~RG/kK ~m~/~SV ~q.OPVPPH~AAGVPARIVGKPDSDKPSMDHDOHFNGINHTF£¥GDG!
C
~:
TGLM:
P~)KRFA~P.m(SKLYKV[JiPVCGKPHVEHV VOE/tIJ( LSLSKL
MLNN/kHSWIIJ~GKG~HYSDLPKVLHTLJEGKAHV~HV ~O~NE ~ & J ~ V
(49J
(51)
TMS: VTIVG~.,&EEVKKQ-LGOKSE¥
RVQN<Q~TJ~A ~/KQ~PFI.~DEKGVTIVICGDTPLLTAETt~QMLKEHTQREN~RTILTAVABDPTGyGRIIRSENG&V~KIV
TGI21: HLVYGHGGDLLMQALI~)DNLNW
VLOAEQ~TGHA~AAPFFADDEDI-IJILYGDVPLISVETL-QRLRD-AKPOGGIGLLTVKLDDPTGYGRITR-ENGKVTGZV
(lb4)
(153)
T~tS: EHKDASEEERLVTEINTGTYCFDNEALFRAIDQVS~NAQGEYYLPDVIEZLKNEGETVAAyQTGNFOET
~VNDR VALSOAEOFMKERINXRI~NGVTLIDPHNTY
(262)
YGM4:
b~tKDJLTDEO~I(~EINTGILIANG~MKPddIJ~;VTN~QGEY¥ITDII/~L~yQ£GREIVAVHPQRLSEV
EGVNNR LQLSRLERVYQ$EQAEKLLLAG'~ILRDPARFD
"n,~s: .l.sPo^v .L~sD'T'{ 2.~rPG'rv .;.KG~VQ ]GEDTZ ..~P,TE ;raSP, ZC~niV ;KQ-S mVVN,SK ',~NDW ZGPrP,H .;SeDSV ;GNEVK ]..GNrVt ;KK'tQrG
•
• ,,*,,,,,.,,,,o.
lttS:
DP.SK~HLSWGDAJE ~ V N
,o,,,oo, ........
am
•
mm
Im
ms
•
(261)
O~S)
,o,,
~j~CGSTCVNYDGKNKYLTK;EDG&F T_GGCNSNLV/~PV~
~EGAY
_VNtGST BVTEDVPGKN~IAR/~QVNKDDYVKNIHK~
lq3I~q: ~GSK~HLT~'D~E ~.~DNI~ ~.~&GI'I~CN~DG~NKFK'~ ~'DDVF ~$DCO I~tVh.PV'i"~KGkT ~AJ~TF mV33NVGENAIJ~ISRVPOl~KEGNRRPVKKX
(456)
(456)
D
DAPD:
H~t~N~TAFERR~E~TPANN7Fv~R~DN~V~AL~J)5G~kLRVA£K~DC~NV~H~LKK~VLL~FRINDN~V1~GA£~R~V~D~K~GVR~P~V~
,,,,,,,,,,*,o,.o..,.,,,..,,,,,
~J~oD: ~YI)RETG¢IHYGm/PJU3SVVVSGNLPSKDGX¥SLYCAVIVKKVD~KTRGK
~INEL ~RTID
(114)
,,.,,.,
(274)
253
Table I
Percentage occurrence of amino acid residues in each position
of the hexapeptide ~
Residue
Position
Tip
a
58.3 b
25.2
16.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Cys
0.0
lie
Val
Leu
Met
Gly
Ala
Ser
Thr
Pro
Ash
Asp
Gin
Glu
Arg
Lys
His
Phe
Tyr
b
c
d
e
f
0.7
3,6
5.0
0.7
51.8 0
7.2
2.2
1.4
0.0
3.6
6.5
1.4
5.8
2.9
1.4
2.2
0.0
2.2
0.7
2.2
0.7
0.7
! .4
10.1
15.1
7.9
4.3
I 1.5
8.6
12.2
2.9
10.1
4.3
2.9
2.9
0.7
0.7
0.0
0.0
1.4
0.7
1.4
25.2 c
6.5
4.3
2.2
2.9
18.7 c
I0.1 c
2.2
1.4
5.0
0.7
7.9
5.0
0.7
1.4
2.2
32.4 c
0.0
0.7
1.4
23.7 c
10.8
12.2
(1.0
2.2
1.4
0.0
1.4
0.0
0.0
2.2
0.0
0.0
0.0
9.4
15.8
2.9
4.3
1.4
5.0
5.0
13.7
0.0
3.6
0.7
5.8
8.6
4.3
5.8
3.6
3.6
4.3
2.2
2.2
9.4
0.7
0.7
0.0
a The tandem hexapeptide repeats (n = 139, shown in Fig. 2
and filling the criteria in the legend of Fig. 2) were included
in the analysis.
b 72%, (Ssc); 46%, (LpxA); 88%, (LacA): 62%, (CysE): 77%
(NodL); 43% (Yglm); 62% (Tins); 36% (DapD).
c Val+Ala: 48%, (Ssc); 54%, (LpxA); 76%, (LacA); 69%,
(CysEP.69% (NOdL); 57% (Yglm); 48% (Tmsk 50% (DapD).
d 52%, (Ssc); 42%, (LpxA); 75%, (LacA); 38%, (CysE); 69%
(NodL); 52% (Yglm); 62% (Tms); 36% (DapD).
= Gly+Asn+Asp: 56%, (Ssc): 29%, (LpxA): 88%, (LacA):
54%, (CysE); 46% (NodL); 62% (Yglm); 57% (Tmsk 64%
(DapD).
a n d 72% ( 1 3 / 1 8 ) o f Cys in e. T h e majority o f
A s n (75%, 3 8 / 5 1 ) a n d A s p (72%, 3 1 / 4 3 ) w e r e in
cord.
T h e m e t h o d o f C h o u a n d F a s m a n [16] p r e dicted a major p a r t o f the hexads to contain a
//-turn. This was e x p e c t e d on the basis o f the
a b u n d a n c y o f residues with h i g h - b e n d potential
(such as Gly). Forty-nine p e r c e n t ( 6 8 / 1 3 9 ) o f the
t e t r a p e p t i d e s bcde (i.e. p e p t i d e s with s u b s e q u e n t
residues b, c, d, and e o f t h e hexad) had p, >
0.75 × 10 -4 (a value indicating a p r o b a b l e b e n d
[at b]) as ",veil as ( P , ) > 1.00 a n d ( P , , ) < (P~) >
<P0)" F o r the bcde p e p t i d e s from individual p r o teins, t h e c o r r e s p o n d i n g p e r c e n t a g e was: 75%
(LacA), 62% (Tins), 50% ( D a p D ) , 50% (LpxA),
48% (Ssc), 48% (Yglm), 31% (CysE), and 31%
(NodL).
5. D I S C U S S I O N
T h e p r e s e n t finding indicates that e n z y m e s
such as t r a n s f e r a s e s can b e m a d e from hexad
blocks. This might h e l p b o t h in u n d e r s t a n d i n g
m o l e c u l a r evolution o f e n z y m e s a n d in p l a n n i n g
genetically e n g i n e e r e d enzymes. It would b e interesting to evaluate w h e t h e r o t h e r e n z y m e s show
any signs o f periodicity (with a d i f f e r e n t pattern).
LpxA, LacA, a n d D a p D are all E. coil transferases but have very d i f f e r e n t t r a n s f e r a s e functions. L p x A transfers an (R)-3-hydroxymyristoyl
moiety from (R)-3-hydroxymyristoyl-acyl carrier
p r o t e i n to the 3 - O H o f U D P - G I c N a c and has
e x t r e m e specificity for acyl-chain length and an
(R)-3-hydroxy function [18,24]. L a c A transfers the
acetyl g r o u p o f acetyl-coenzyme A to t h e 6 - O H o f
thiogalactosides a n d certain galactosides and has
b e e n s u g g e s t e d to be a detoxifying e n z y m e [25].
D a p D transfers an a m i n o moiety from g l u t a m a t e
to N-succinyl-~-keto-a-aminopimelate [26]. It utilizes pyridoxal p h o s p h a t e as a coenzyme, as d o
Fig. 2. The imperfect tandem hexapeptide repeat structure of eight bacterial proteins. The hexapeptide arrangement was revealed
by locating those IG, LG, and VG dipeptide residues which are followed or preceded by lie, Leu, or Val at an [I,V,LHI,V,L]
distance of six residues. Such dipeptides as well as each lie, Leu, and Val continuing this six-residue periodicity pattern is
underlined (by a double line). (A) The Ssc protein of S. typhimurhan and UDP-N-acetylglucosamine acyltransferase (LpxA) of E.
coil The overlapping regions of these proteins are 22.4% identical [19]. (B) Thiogalactoside acetyltransferas¢ (LacA) and serine
acetyltransferase (CysE) of E. coll. and the nodulation protein NodL of R. leguminasorum. These proteins have been suggested to
form a single 'bacterial acetyltransferase' family [20]. (C) The hypothetical proteins Yglm of E. coil and Tins of B. subtilis. These
proteins are 43% identical [23]. (D) Succinyldiaminopimelate aminotransferase (DapD) of E. coll. The alignments were made by
using HIBIO PROS1S. The 31-residue long region of highest homology between all the eight proteins [19] is marked with asterisks.
Deletions are marked with ( - ).
254
the other aminotransferases. Accordingly, it is
difficult to find any common functional denominator specific enough to give any clue why these
proteins all resemble each other and have the
long hexapeptide repeat regions with the common theme. No three-dimensional structures have
yet been determined for any of these proteins.
Besides LpxA, LacA, and DapD, five other
proteins were found to have the hexapeptide
repeat theme. While no conclusions can be made
regarding to the hypothetical proteins (Yglm of
E. coli and its Bacillus analogue Tms), CysE and
probably NodL are acetyltransferases as is LacA,
and S ~ / F i r A could well have a function related
to LpxA. In addition to LpxA, at least three other
acyltransferases participate in lipid A biosynthesis [18,27], but none of them have been characterized at the protein or DNA level yet. Studies in
this laboratory arc in progress to define the role
o f Ssc.
Thc hcxapcptide repeat theme now found
m i g h t f a v o u r t h e f o r m a t i o n o f /3-turns, as d o
s o m e o f t h e previously d e s c r i b e d r e p e a t motifs.
These include the [L,I,F]XGGXG[N,D]DX motif
f o u n d in t h e p o r e - f o r m i n g R T X cytolysins o f
G r a m - n e g a t i v e b a c t e r i a [3,28] a n d t h e Gly-, A s n - ,
a n d A s p - r i c h m o t i f s in several c a r b o h y d r a t e - b i n d ing p r o t e i n s o f G r a m - n e g a t i v e b a c t e r i a [5]. A l s o
t h e Pro-rich t a n d e m p e n t a - to n o n a p e p t i d e m o tifs in p r o l a m i n s , R N A - p o l y m e r a s e 11, r h o d o p s i n ,
s y n a p t o p h y s i n , a n d synexin f o r m r e p e a t e d / 3 - t u r n s
[29,30]. It h a s b e e n s u g g e s t e d t h a t c o n s e c u t i v e
p - t u r n s m i g h t be o n e a d v a n t a g e n o u s way to b i n d
c e r t a i n ions, ligands, o r s u b s t r a t e c l a s s e s [3,5].
During evolution, intragenic recombination
b e t w e e n t h e r e p e a t s e q u e n c e s in t h e g e n e a n d
t h e r e s u l t a n t v a r i a t i o n in t h e n u m b e r a n d arr a n g e m e n t o f r e p e a t s m i g h t p r o f o u n d l y alter t h e
s u b s t r a t e specificity. R e g a r d i n g to possible evolut i o n a r y relations, it s h o u l d be n o t e d t h a t t h e
genes encoding Ssc/FirA, LpxA, and DapD are
all l o c a t e d at 4 m i n in E. coli c h r o m o s o m e , a n d
very close to e a c h o t h e r [6,12,17,31].
ACKNOWLEDGEMENTS
T h i s w o r k w a s s u p p o r t e d by G r a n t 1011749
f r o m t h e A c a d e m y o f F i n l a n d a n d by t h e Sigrid
Juselius Foundation.
REFERENCES
[I] Cohen. C. and Parry, D.A.D. (199(I) Proteins 7, 1-15.
{2] Lupas. A., Van Dyke, M. and Stock, J. (1991) Science
252, 1162-1164.
[3] Welch, R.A. (1991) Mol. Microbiol. 5. 521-528.
[4] Tan. L.K.. Shopes. R.J.. Oi, V.T. and Morrison. S.L.
(19911) Proc. Nail. Acad. Sci. USA 87, 162-166.
[5] Wren, B.W. (1991) Mol. Microbiol. 5, 797-803.
[6] Coleman, J. and Ractz, C.R.H. (1988) J. Bacteriol. 170,
1268-1274.
[7] Hedlger, M.A.. Johnson, D.F., Nierlich, D.P. and Zabin,
F. (19851 Proc. Nail. Acad. Sci. USA 82, 6414-6418.
[8] Denk, D. and Bock. A. (19871 J. Gen. Microbiol. 133,
515-525.
[9] Surin, B,P. and Downie. J.A. (1988) Mol. Microbiol. 2,
173-183.
lll)l Richaud. C.. Richaud, F., Martin. S.. Ilaziza, C. and
Patte, J.C. (19841 J. Biol. Chem. 259, 14824-14828.
[11] Hirvas, L.0 Koski, P. and Vaara, M. (19911) Biochem.
Biophys. Res. Commun. 173, 53-59.
[121 Hirvas, L., Koski. P. and Vaara, M. (19911EMBO J. I0.
1017-11123.
113] Dicker. !. and Seetharam, S. (19921 Mol. Microbiol. 6,
817-823.
[141 Pear,n, W.R. and Lipman, D.J. (19881 Proc. Natl. Acad.
Sci. USA 85, 2444-2448.
[15] Devereaux, J.. Haeberli, P. and Smithies, O. (19841 Nucleic Acids Res. 12. 389-391.
1161 Chou, P.Y. and Fasman, G.D. (19781 Annu. Rev.
Biochem. 47, 251-276.
[17] Dicker, i. and Seetharam, S. (19911 J. Bacteriol. 173,
334-344.
1181 Raetz, C.R.H. (19911)Annu. Rev. Biochem. 59, 129-1711.
[19] Vuorio. R., Hirvas. L. and Vaara. M. 119911 FEBS Len.
292, 911-94.
[201 Downie, J.A. (1989) Mol. Microbiol. 3, 1649-1651.
[21] Bairoch, A. 09901 A dictionary of protein sites and
patterns, the PROSlTE database. University of Geneva.
122] Walker, J.E., Gay. N.J.. Sarast¢. M. and Eberle. A.N.
11984) Biochem. J. 224, 799-815.
123] Nilsson, D., Hove-Jensen, B. and Arnvig, K. (19891 Mol.
Gen. Genet. 218, 565-571.
[24] Anderson, M.S. and Raetz, C.R.it. (19881 J. Biol. Chem.
262, 5159-5169.
125] Andrews. K.J. and Lin. E.C.C. (19761 J. Bacteriol. 128,
510-513.
[26} Peterkofsky, B. and Gilvarg, C. (19611. J. Biol. Chem.
236, 1432-1438.
[27] Brozek, K.A. and Raetz, C.R.H. (19901 J. Biol. Chem.
265. 15410-15417.
[281 Felmlee, T. and Welch. R.A. (19881 Proc. Natl. Acad.
Sci. USA 85, 5269-5273.
[29] Shewry, P.R. and Tatham, A.S. (19901 Biochem. J. 267,
1-12.
[30] Matsushima, N., Creutz, C.E. and Kretsinger, R.H. (19901
Proteins 7, 125-155.
[31] Bachmann. B.J. (19911)Mierobiol. Rev. 54, 130-197.