Download Comparisons between the Primary Structure of the Coat Proteins of

J. gen. Virol. (1985), 66, 2571-2579. Printed in Great Britain
2571
Key words: tymovirus/eoat proteins/structure
Comparisons between the Primary Structure of the Coat Proteins of Turnip
Yellow Mosaic Virus and Eggplant Mosaic Virus
By A N D R I ~ D U P I N , *
DANIEL
COLLOT,
RI~MY PETER
AND J E A N
WITZ 1
Laboratoire de Physiologic VOg(tale, Institut de Botanique, 28 rue Goethe, 67083 Strasbourg
Cedex and 1DOpartement de Virologic, Institut de Biologic MolOculaire et Cellulaire du
C.N.R.S., 15 rue Descartes, 67084 Strasbourg Cedex, France
(Accepted 28 August 1985)
SUMMARY
Comparison of the primary structures of eggplant mosaic virus (EMV) and turnip
yellow mosaic virus (TYMV) coat proteins shows that 3 2 ~ of their amino acids are
conserved. Alignment of the two sequences requires only one deletion near the N
terminus and two insertions at the C terminus of T Y M V coat protein. Although the
coat protein of EMV is on average less hydrophobic than that of T Y M V , structural
predictions yield fairly similar conformations for the two proteins, a p a r t from the N
terminus. Neither coat protein possesses an accumulation of basic residues able to form
the strong ionic R N A - p r o t e i n interactions observed in several other isometric viruses.
The nature of the amino acid exchanges seems to be different from that seen in families
of homologous proteins. The highly conserved regions encompass a (probably weak)
potential R N A - p r o t e i n interaction site. Implications for the structure and stability of
small isometric viruses are discussed.
INTRODUCTION
Tymoviruses are small isometric plant viruses stabilized by strong p r o t e i n - p r o t e i n
interactions. Their properties have been reviewed recently by Koenig & Lesemann (1979, 1981).
The type m e m b e r of the group is turnip yellow mosaic virus (TYMV). The primary structure of
its coat protein (Peter et al., 1972) as well as a large part of the nucleotide sequence of its R N A
are known (see for instance Guilley & Briand, 1978, and references therein). N o high resolution
details on its three-dimensional organization are available, but chemical cross-linking has shed
some light on the R N A - p r o t e i n interaction sites (Ehresmann et al., 1980) and several antigenic
determinants have been characterized (Pratt et al., 1980; Quesniaux et al., 1983a, b).
The primary structure of the coat protein of a second tymovirus, eggplant mosaic virus
(EMV), has been determined recently (Dupin et al., 1984). Since E M V is serologically not
directly related to T Y M V (Koenig, 1976) and since the stabilities of its virions and capsids differ
from those of T Y M V (Bouley et al., 1977), we compare here the primary structures of the two
coat proteins and discuss the implications for various properties of the virions.
METHODS
The primary structures of the coat proteins ofTYMV (P-TY) and EMV (P-E) were taken from Peter et al. (1972)
and Dupin et al. (1984). The amino acid sequences were established by automatic sequencing and mass
spectroscopic studies of the products of cleavage of the purified proteins by cyanogen bromide, BNPS-skatole,
trypsin, chymotrypsin and thermolysin.
A computer program written by G. de Marcillac for a Univac 1100 computer was used to align the protein
sequences. Sliding one sequence along the other, it indicated in a two-dimensional plot whether identical amino
acids face each other; alignments show up as lines of dots parallel to a diagonal [this program is similar in
principle to the simplest program of Staden (1982); it does not make use of a scoring matrix].
Assuming that EMV and TYMV derive directly from each other, expected amino acid substitution frequencies
between the primary structures of EMV and TYMV coat proteins were calculated the following way: both aligned
sequences (as in Fig. 1) encompass N amino acids, n~~,n,i, n~i and n2j are the numbers of amino acids i andj in the
0000-6621 © 1985 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
2572
A. D U P I N AND OTHERS
first and second sequence, respectively (1 ~< i, j ~< 21 if deletions and insertions are taken into account). If
substitutions are random, the number of possible substitutions involving residues i andj :~ i is nl~ n2j + n~j nzi, and
the number of possible conservative exchanges is n~, n_,~.The total number of possible exchanges (conservative or
not) is N 2, of which N actually occur. The expected numbers of random substitutions and conservations are (nli nzj
+ n~j n2~)/N and nl~ n_,jN, respectively. If the pattern of substitutions follows the rules established for a large
sample of protein families, these figures should be multiplied by the coefficients of the corresponding log-odds
matrix (Fig. 85 in Schwartz & Dayhoff, 1978).
Hydrophilicity plots were computed according to Hopp & Woods (1981).
Secondary structures were predicted according to Chou & Fasman (1978), taking into account the changes of
frequencies of the amino acids close to the N and C termini of helices and ~-sheets, or in their interior (Chou &
Fasman, 1974).
R E S U L T S AND D I S C U S S I O N
Alignment o f the amino acid sequences
Visual inspection of the sequences of P-E and P-TY showed that superposition of several
tripeptides and one tetrapeptide could be achieved by introducing in P-TY one single deletion,
facing Asp 3 (P-E), and two insertions at the C terminus to compensate for the longer
polypeptide chain (Fig. 1). The computer search showed that no major homologies could be
found other than those shown in Fig. 1.
Assuming that P-E and P-TY derive from a common ancestor, 3 2 ~ of the residues (including
one tetrapeptide and five tripeptides) are strictly conserved in P-E and P-TY. About two-thirds
of the exchanges (89 of 127) require only one nucleotide change in the messenger R N A and none
requires all three nucleotides of the corresponding codon to be changed. Conserved regions are
indicated by boxes in Fig. 1. These are approximately evenly distributed along the sequence and
the number of boxes of each size corresponds well with that expected for a 3 2 ~ homology
distributed at random along 189 residues, i.e. 27.5 single amino acids, 9.6 dipeptides, 3.4
tripeptides and 0.8 tetrapeptides compared with the observed numbers of 21, 10, 5 and 1,
respectively. This situation is very similar to that observed for the isometric phages Qfl and
MS2, although their aligned sequences contain many deletions and insertions (Konigsberg et al.,
1970; alignment 58 in Dayhoff, 1972). In contrast, for tobacco mosaic virus, a block of eight
amino acids in helix R R and another of ten amino acids in helix L R are conserved in all
sequenced strains. In this virus the polynucleotide chain and the protein subunits are arranged
along the same basic helix, and the conserved regions correspond to the two s-helical segments
that interact with three nucleotides in each subunit (Stubbs et al., 1977). In no isometric virus yet
investigated by single-crystal X-ray diffraction could the R N A or the N terminus of the coat
protein be seen in the electron density maps, although in each case the R N A is rigidly linked to
the capsid (Harrison, 1983; Rossmann et al., 1983). The icosahedral symmetry of the surface
domains of the shell does not hold for the interior of the virion, and it is possible that R N A protein interactions differ from one protein subunit to another, even if these belong to the same
group of 60 related by strict icosahedral symmetry at their surface domains.
Although the number of boxes in Fig. 1 follows the distribution expected for a random
distribution of the conserved residues in P-E and P-TY, their location is unlikely to be defined by
chance only. Indeed, we also determined the primary structure of three fragments of the coat
protein of another, as yet unidentified, plant virus which we presume to be also a tymovirus. The
fragments could be aligned with regions 63 to 68, 108 to 117 and 140 to 166 of Fig. 1. This virus
was about as close to EMV as it was to TYMV : out of 44 residues, 23 matched with P-E and 20
with P-TY. But it is remarkable that in these regions all but one (position 160) amino acids that
are common to P-TE and P-TY also occurred in the third sequence (A. Dupin & R. Peter,
unpublished). The fragment corresponding to positions 140 to 166 possesses several remarkable
properties: it contains the unique common tetrapeptide, together with a neighbouring
tripeptide. The region 144 to 153 corresponds to an antigenic determinant that was found in
dissociated TYMV protein only (see below). Peptide 133 to 153 could also be crosslinked in situ
to TYMV RNA (Ehresmann et al., 1980) and contains one of the two short sites (positions 142
and 143) that were found to remain bound to viral RNA after dissociation and drastic hydrolyais
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
2573
Tymovirus coat protein sequences
i0
1
20
40
TYMV
{Ac M El. I D K E L AlP 0ID R T V T V A T VIL PIA v P O P S P L TIIiK 0IP F OIsIEIv L F
I x
ol
{
I
I
Io
o
x
[--}
ol i
}
I_I
l-I
I--I I--I
70
{-I
{xx oo oox{
i
I-I
0IF{EIA
i i {x
}
I-{
I-I
AIAI
x{ ioxoo
i I
AIOIT K D A EIA{
{_I
I_{
T T FIG{A T S T
I}{
80
90
l-}
i00
I-I }--{
l--
I--I
{ooo{
{xx{ {xx
x{
{x{
{x { {x o
{ { {
{ o
}
{
{
S L T I{A{N I D S V S TiLiT T FiY R H{A S{LiE S L W V{T IiH{P TiL O{AiP T F{PiTiT V{G V CiW UIP A{N
{_I
i01
I
ii0
I
{_I
I
120
I {
I
I_I
I_I I
130
{
I
{
140
I
150
--I
}-I I-I
I--I
I--I
I-I
I--I
I-I
{-I
{--{
I-I I-I
S{T A K{P{T EII{L N V F{G O{S S Y T F{G O AIL{N{A T K{P LiT I P L{P{M N S V{N{e M L{K D S[V L{Y{T{D{
I xxl loo{ l o ol Ix oxol
lol {o x{ l o
l {oo o{ l
ol
{o [ { l {
TYMV
SIP V TIP{A O{IIT K T Y{G GI0 I F C IIG G A[IINIT L SIP LI[ V K CIPIL E M MIN[P R VIK D SiI OIYILIDI
{
i_l
151
i_l
I
}
I
160
I 1_I
}
170
I-{
I--I
{-I
I
I_I
}_}
180
I-I
l-IN
}
{
}_I I_}
190
i--I
C{P K L L{A Y S A{A{P S S P S K T P{T{A T{IIQ I H{GIK{L{R L S{S P L{L O A N . .
x{
TYMV
i
60
I_I
EMV
I--I
O I S LIAIS A N A I T KILIA S L{Y R HIV RILIT O C A AIT IITIP TIA AIA{I A NIPIL{T VIN I VIW V{S D{N
xooo{
EMV
oxo
{
{-{
TYMV
I-{
T A I I R SiP OIP S I N A P G F HIL P{P T D S 0 O S S A I I I E LiP F
51
EMV
{--I
50
I--I
IA= M EID
l
}
I
l--I
30
EMV
S{PK
i
Ixx
o{ {
oo
L L{I S I T { A { O P T A P
i
}_[
o
} {
P AS}T{C
}_}
{ {xox{
Ix{ {xox{
{ox
x
I{I{TV S{G{T{L{S M H{S P L{I T D T S T
}_I
} { } }
i
{
Fig. 1. Alignment of the sequences of the coat proteins of EMV and TYMV. Common residues are
indicated by boxes and conservative differences by circles. Differences that conserve only the
hydrophobic or hydrophilic nature of the amino acids, without change of the sign of the charge, are
indicated by crosses. The identified antigenic determinants of TYMV protein are underlined. Regions
of P-TY that could be crosslinked in situ to TYMV RNA are marked with an interrupted line.
of the crosslinked virions. Carbon-13 nuclear magnetic resonance studies have shown that in
another tymovirus, belladonna mottle virus, glutamic and aspartic residues are in contact with
the RNA (Virudachalam et al., 1983): two aspartic acid residues are conserved in all three
sequences, at positions 144 and 150. This makes this region a good candidate for a protein-RNA
interaction site in the tymoviruses.
Fig. 1 also shows the regions of approximate homology in P-E and P-TY. Circles indicate
amino acid substitutions frequently observed in closely related proteins. They often correspond
to exchanges between hydrophobic side-chains of similar size or to exchanges between certain
charged and uncharged hydrophilic residues (Fig. 85 in Schwartz & Dayhoff, 1978; Rice et al.,
1984). We also indicate by crosses substitutions that conserve the non-polar or polar (without
change of the sign of the charge) character of the residues. It is striking that only one large region
of low homology exists in the sequence. It corresponds to the N-terminal part of the chain (45
approximate homology for residues 1 to 25, and 50~ for residues 1 to 50). Apart from a short
stretch (residues 160 to 170) close to the C terminus, the score was above 60~o everywhere else. It
reached 100~ for the region 51 to 74.
Coat proteins of both EMV and TYMV contain an exceptionally large number of prolines,
reflecting the high cytosine content of the viral RNAs. Neither of the two sequences contains a
region rich in basic residues able to undergo strong ionic interactions with the RNA, and the
amino acid composition of other tymovirus coat proteins (Paul et al., 1980) makes it likely to be a
general property of the family. Such regions, deeply embedded in the RNA, have been found in
several small viruses stabilized by strong RNA-protein interactions (for reviews, see Rossmann
et al., 1983; Harrison, 1983). Their absence in the coat proteins of tymoviruses and poliovirus
(Racaniello & Baltimore, 1981) is probably a prerequisite condition for the self-assembly in the
cell of capsids that, at least for TYMV, are isomorphous with the virions.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
2574
A, DUPIN AND OTHERS
I
A 13.ol
I--I--I~-~[
I n Io.51o.31
I--I--I--I
3---[
I N It.41o.21o.31
I--I~--I--Ii---I
2-~[
i D It.310.210.810.4t
I--Ii'--I--I--I--I
I C 10.510.tlO.110.tlt.O[
1
I--Ii--I--I
I
I
12 [
I Q 11.51o.410.71o.71o.11o.81
I--I¥--I--Ii-Ii---I--Ii--I
2
I E 11.31o.210.611.olo.tlo.81o.4l
1--17l--li---I
I
I
I
6 I
a 12,210.210.610,slo.llo.slo.510.91
--1
I
I
I
I
I
It I
H 10.510.210.310.2t0.t l o . s l o . 2 1 0 . t 10.2I
--12--1
I
I
I
I1
I
I~
I
I
I
I
11
I
I1
]
12.810.310,vI0,slo.310,810.slo.610.a13.t I
I~IY--I
I
2 12 I
I
14 18 I
12.slo.alo.vIo.410.21~.OlO.Slo.610.414.916,71
I
I1
12
K It.110.610,710.410.~ I0.V10.410.410,210.710.710.81
I
I
It I
I
I
I
I
I
I
2 I
It
M 10.610.110.210.t 1 0 . 0 1 0 . 2 i o . l l o . t 10.110.91 1.6]o.31o.31
I
II t l
I1 I
I1 I
1 I
I
i
11
v 10.510.110.210.zlo.110.210.110.110.t I i . t l 1.61o.11o.211.3l
12 I
I
lO I
[2 I
I
]2
1 12 I
I
I
It
P 15.210.71 z.310.alo.011.610.911.310.311.81 2.011.110.410.317.0]
5 13---]1--12--13
It I
1 14 I
I1 12 12 1 I1 I1 I1
i s 14.Blo.611.511.210.711.411.zlz.910.512.01 1.81Z.310.510.514.412.51
[2
1 14 14
I
Ito [
I1 I
1 12 11 I
I1 14 5 14 I
i T i6.6i0.711.911.310.5il.611.2]1.8[0.513.611.0[1.710.710.714.816.0i5.11
--I~1--1--1--
--I
I
I
I
L
I
I
I
I
I
I
I
I
I
I
I
12
I]---I
I
I
I
I1--Ig--12
I
It
I
I
I
I
I
I
I--I--I--I--I--I--I--I
I
I
I
I--I--I--I--I--I--I--I--I~l--
I--I--I--I--I--I--I--I
I
I--I--I--I
I--I
1----i
w 10.z10.110.~10.010.010.010.010.010.010.~10.210.~10.010.1i0.110.210.I10.51
--I--I--I
I
I
12
Y 10.310.1t0.z I0.110.110.110.1 I0.110.110.410.510.110.tl z.010.210.310.510.110.71
--I~--I--Ii--I--I1
~
I
i--1~l--1--12---r
v 12.s10.210.610.310.210.610.410.610.213.613.210.5 0.610.51z.511.612.710.410.211.41
i--i
i
I
i
i
i
i
I
i
i
I--I--I--I--I--I--I-_IA
i n_l._ i ~ 1 c
Q I E I G I n I x I L I K M I ~ I P I S_I T_T_Iw._~_IJ!_IVY_
I
I
I
I
I
I
I
I
I
I
In(i)! 351 6 I 111 9 t 7 l tnl 9 t 9 I 5 I 251 2BI 121 6 I 101 27l 331 42
IN(i)l 301 5 I B 1 7 I 7 I 121 7 I 3 I 4 I 2tl 201 tOI 5 I 9
IA'(i~)l 4___£07___1 1__~411 ~ 1 7 1
I
171 301 3e
I--I--I--I--I--I--
I
I
I--I
2 I 5 I i gl
t I 3 I 171
1.._~611_!11~1 6_&_l291 3__661
t_A417___1 t.~ll 3__zl 3__661
4___~61_~SI L l
2__!1
Fig. 2. Number of observed amino acid exchanges between the coat proteins of EMV and TYMV, and
comparison with the numbers expected for random sequences of the same length and amino acid
composition, but following the rules established for homologous proteins (see text). Figures along the
diagonal refer to the conserved residues of each kind observed in Fig. 1. Observed numbers of
exchanges are at the top and expected ones at the bottom of each square, n(i), N(i) and N(i) represent
the number of positions at which amino acid i appears in Fig. 1, the total number of exchanges
involving amino acid i, and the total number of residues i, respectively (insertions and deletions are not
taken into account).
The exchanges observed in Fig. 1 are tabulated in Fig. 2. It is striking that glycine residues are
highly conserved whereas all seven cysteine positions are exchanged. The 'mutability' of several
other amino acids, normalized to that of alanine, also departs from the values observed in
families of homologous proteins (Dayhoff et al., 1978). To analyse such differences in more
detail, we calculated the number of exchanges for random sequences possessing the amino acid
composition and structural similarity of P-TY and P-E (see Methods). These figures were
weighted using the log-odds matrix corresponding to 250 point accepted mutations ( P A M s ; Fig.
85 in Schwartz & Dayhoff, 1978). This evolutionary distance corresponds to an 80% difference
in amino acid positions instead of the 68 % observed in Fig. 1 (corresponding to 150 PAMs), but
proved well adapted to the analysis of distantly related proteins, although it overestimates
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
Tymovirus coat protein sequences
2575
somewhat the exchange rates in the present case. Fig. 2 shows that the exchangeability of all
charged and amino-amide amino acids is much lower than expected ifP-E and P-TY behaved as
homologous proteins (12 residues are conserved, instead of an expected number of three or four):
this high conservation rate may at least in part compensate for the small number of these amino
acids in tymovirus coat proteins. As already mentioned, glycine residues are highly conserved,
especially when they occur in closely spaced doublets (positions 113 and 114, and 120 and 121,
Fig. 1). Several exchanges occur at a low rate (A P, A-S, T-P), but several others are unusually
frequent (S-R, S-Q, T - L and T - K , for instance). If this observation is confirmed by analysis of
the primary structures of more coat proteins of distantly related viruses, it would imply that
standard amino acid exchange tables such as those of Dayhoff (1972) cannot be used reliably to
establish phylogenetic trees of virus families and/or to predict possible primary structures of
putative common ancestors, as has been done for instance for the MS2/Q/~ family (alignment 78
in Dayhoff, 1972). It is indeed striking that the foldings of the polypeptide chains of the surface
domains of tomato bushy stunt virus, southern bean mosaic virus and satellite tobacco necrosis
virus are so similar that it has been suggested that these viruses derive from a common ancestor,
although there is no obvious similarity in the primary structures of the three coat proteins
(Rossmann et al., 1983). Even if EMV and TYMV derived from a common ancestor, the
variability of their stability (Bouley et aL, 1977) would allow amino acid exchanges in their coat
proteins that are different from those occurring in homologous proteins, mainly enzymes, that
must preserve the stereochemistry of an active site, or immunoglobulins.
Antigenic determinants
No data are available on the nature of the antigenic determinants of EMV but a detailed
immunological investigation of TYMV is in progress (Pratt et al., 1980; Quesniaux et al., 1983a,
b). Three antigenic regions have been identified with antibodies raised against virions. They
correspond to residues 1 to 13, 58 to 65 and 184 to 190 respectively, using the numbering
convention of Fig. 1. Two other regions (34 to 46 and 144 to 153) have been found in dissociated
protein only.
Fig. 1 clearly shows why EMV and TYMV do not possess common antigenic determinants
(Koenig, 1976). At best, four residues in 12 are conserved in peptide 1 to 13, five in 13 in peptide
34 to 46 and six in 10 in peptide 144 to 153. Only one residue in seven or eight is conserved in the
two other sites.
Only one determinant (positions 1 to 13 in Fig. 1) corresponds to a highly hydrophilic region of
TYMV coat protein (Fig. 3). The second one (positions 34 to 46) cannot be assigned with
accuracy. The fourth determinant (positions 144 to 153) corresponds to a moderate oscillation of
hydrophilicity and the third site (positions 58 to 65) clearly corresponds to the descending part, if
not the trough, of one of the deepest minima of the hydrophilicity plot. Similar observations
have also been made for tobacco mosaic virus, myoglobin and lysozyme where a better
correlation has been found between antigenicity and segmental mobility than between
antigenicity and hydrophilicity (Westhoff et al., 1984).
Structure predictions
Fig. 3 and 4 show the hydrophilicity plots (Hopp & Woods, 1981) and the predicted secondary
structures (Chou & Fasman, 1974, 1978) of P-E and P-TY.
Secondary structure predictions proved very ambiguous at several points. The number of
regions possessing a high probability of being in a helical conformation is very low. We included
in Fig. 4 segments possessing at least five residues, although Chou & Fasman (1974) argued that
a helix containing less than six residues may not be stable on its own. The existence of the
predicted helix from P81 to A87 in P-E is questionable. Although it has a high calculated
probability of being c~-helical (P~), it contains four alanines and one isoleucine in seven amino
acids, rendering it very hydrophobic. In P-TY two overlapping turns are predicted in the region
from position 139 to 145. We chose the turn that was also predicted in P-E.
Peaks and troughs occur at similar positions in the two average hydrophilicity plots of Fig. 3,
although their amplitudes may be very different, being often larger in P-TY; several regions that
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
2576
h.
I
I
I
l
DUPIN
I
I
AND
I
N
I
OTHERS
I
I
I
I
I
I
I
I
i
I
I
I
maD
1.5
i'l
1
ti
IA
II
I
i,d,,,,
0.5
t'I i
-
i
0
,>
17
1
-0.5
,I,!,I ,,
.
1
! ~,.7 i
!
';u
~ I
'. I~
.~
7
i:
l
t t
,,
',
, i
[
Li ~'
--1
.
I~
f '
I
,
,
,7
i I
' i
<
", i(
:1
"~
;
,
,
7" ,
I [1"- (;
,
li i
i :
,'
~LI" ~'
:
btl
I~;,
'A'
1.5
U
a
1.5
1
i
0.5-
i
iil't
N
i ',
i't
0
(
l' L,
h
I
,
it.,
--0.5
II P ,,'-' ~ ~ ,,
0
I ,r'l/I
i
I
I
I
I
50
[
I
I
I
I
I
i
', II
I
i
~
'~
,1J t,!''I, I If fJ
I ~'!q t,'
h~ t .....
.
~,I,E !/
'<,
!i
v
.
I,,'
/ .,
t!
I'
rr r
(b)
I
!'ti'i
I I
I I ! ~',
't4
-I
-1.5
',
i
~i #'~
'
i
,~..
i
I
I
100
Residue number
I
I
I
I
1
I
I
150
Fig. 3. Hydrophilicity plots of (a) EMV and (b) TYMV coat proteins. The hydrophilicities were
averaged over six contiguous residues. The position of the known antigenic determinants of TYMV and
dissociated TYMV protein is indicated by the horizontal bars.
are hydrophobic in P-TY contain one hydrophilic residue in P-E, rendering P-TY more
hydrophobic than P-E, on average. P-TY contains three clusters of very hydrophilic amino
acids, giving rise to the three highest peaks in Fig. 3. These include the N terminus of P-TY,
much more hydrophilic than that of P-E. The region 75 to 95 is very hydrophobic in both
proteins, containing no hydrophilic amino acid. It is the longest stretch of this kind found in the
published primary structures of coat proteins of simple isometric viruses.
A similar rather approximate similarity can also be seen in the predicted secondary structures
(Fig. 4). Both proteins contain a few rather short a-helices ( 2 0 ~ of the residues of P-TY; 22 to
29% for P-E) but many/~-sheets and turns, a feature c o m m o n to several viral proteins (Argos,
1981). Both C and N termini possess different predicted structures in the two proteins and at
several places the helical segments do not face each other in Fig. 4 (P105 to 1108 of P-E has a
high Pc(, but is too short to form a stable helix).
The N-terminal arms of P-E and P-TY are among the less conserved parts of the primary
structure (Fig. 1 and 4). This difference results from many exchanges between hydrophobic and
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
2577
Tymovirus coat protein sequences
~ A A A A A A A A A A A A - - _ - - _ - - _ - - ~-= ~.~.---------= ~------ ~---=
• wv
:====:========== -
-
-
I
£ ~.1 1 1 1 £ £ £ £ £ A A A A A A A A£ £ £ ~ , ~ . ~ I I / , £ £ £
AcMEDTAI I R S P Q P S I N A P G F H L P P T D S Q Q S S A I ELPFQFEATTFGATETAA
10
20
30
40
50
rrMv AcME, I D K E L A P Q D R T V T V A T V L P A V P G P S P L T I K Q P F Q S E V L F A S T K D A E A
ikhhAhAhAhAhA
t£££££~,£~.£££.~£~,£~,L££££££££££££££AAAAAAAAAAAAAAAAAAAAAAAAAAAA
££££££££££££~£zzz zz ~Z~AAAAAAAAAAAA~Z~
EMV
Q I S L A S A N A I T K L A S L Y R H V R L T Q C A A T I T P T A A A I A ~ P L T V N IVWVSD~
60
70
100
90
80
S L T I A N I D S V S T L T T F Y R H A S L E S L W V T I HPTLQAPTFPTTVGVCWVPAN
~,11R
--
_=E =-~_==-=-=-AAAAAAAAAAAAAAAAAAAA
t £ 11 I £ £ ~.£ ££~AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=~ ~-=-~~-
.AAAAAAAAAAAA--- =---------- z---~ z ~-------------
z~-z . . . .
--AAAAAflAAAAAA
.....
~-=--------
=. . . . .
"= -=m-=~-~-
~V
STAKPTE I LNVFGGSS YTFGGA L,NATKPLT I PLPMNSVN CMLKDSVLYTD
T~V
SPVTPAQITKTYGGQ I FC I GGAIi~TLSPLIVKCPLEMMNPRVKDS I QYLD
110
_=--
120
130
140
~,11 ~,~.~,~.~.i£ i_--=-~ =-_--~=_AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAA~ . 1 ~ . ~ I ~
~-~-£ ~ 9.1£ ~ .~£ £ 9.1 ,~l £ ~ ~,£ £
.......
~AAAAAAAAAAAAAAAA.
~
........
AAAAAAAAAA
-££111~.19.£111
ENV
CPKLLAYSAAPSSPSKT PTAT I Ol HGKLRLSSP LLQAI~
TIrMV
SPKLLIS ITAQPTAPPASTCI ITVSGTLS~HSPLITDTST
460
150
180
170
_--:_=_--=-=-=---AAAAAAAAAAAA_=_=
~ ~ ~._=_==-_=_=_==-=-:_-- ~ ^ A A A A A A A A A ~ . A A A A A A A A~A~ ~ ~ ~ ~ ~ ~ ^ A A A A A A A A A A A A A A ^
Fig. 4. Predicted secondary structures of EMV and TYMV coat proteins. ~,, c(-helix; A, fl-sheet; =-, flturn ;
, disordered. Where two equally probable possibilities occur, both structures are indicated.
E2
E2
E
.
.
.
.
~
K6
.
I 4 ~
D3
...... / A M
1
R8
I7 ~
T4
L8 ~"-"-'~ ~
D5
Fig. 5. Helical wheels (Schiffer & Edmundson, 1967) for the N terminus of the coat proteins of EMV
and TYMV. The orientation and length of each vector correspond to the hydrophilicity of the amino
acid, according to Hopp & Woods (1981). The numbering of the residues corresponds to that of Fig. 1;
position 3 in TYMV coat protein is a deletion.
hydrophilic side chains. It is also seen in the helical wheel diagrams (Schiffer & Edmundson,
1967) of the first nine (P-TY) or 10 (P-E) residues preceding the first proline : in P-E each of the
three hydrophilic amino acids is surrounded by hydrophobic residues, an unfavourable
distribution for a stable c(-helix (Fig. 5). In the present state of our knowledge of the architecture
of tymoviruses, it is not possible to say if the difference between the N-terminal arms of P-E and
P-TY contributes to the observed difference in the stabilities of T Y M V and EMV (Bouley et al.,
1977). The localization of the N-terminal arm of P-TY in the capsid has been controversial.
Peptides 1 to 13 (numbering of Fig. 1) could be successfully crosslinked in situ to the viral R N A
either by irradiation at p H 4.8 (but not at pH 7.0) or by bisulphite treatment at neutral p H
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
2578
A. D U P I N
AND OTHERS
(Ehresmann et al., 1980). Immunochemical studies (Quesniaux et al., 1983a), on the other hand,
led to the suggestion that it corresponds to an antigenic determinant of the virion, i.e. that it is
located at the outer surface of the capsid. These studies were based on tests of the inhibitory
capacity of peptides in reaction with antisera collected from rabbits immunized by injections of
virions. It is known, however, that about 4 0 ~ of the virions release their R N A and partially
dissociate when TYMV is incubated for 2 h at 37 °C in a neutral buffer of physiological ionic
strength (Lyttleton & Matthews, 1958). It is likely that antisera raised against virions also
contain antibodies against dissociated protein subunits; these antibodies may be detectable by
sensitive modern immunological methods. A definite immunological proof that peptides 1 to 13
are at the surface of the virion would require raising antibodies against this purified peptide and
testing their activity against the virions.
The core of the region that could be crosslinked in situ to T Y M V R N A (Ehresmann et al.,
1980) is the couple Lys-Asp (positions 143 to 144) located in a predicted fl-turn: it could make
specific contacts with the R N A by a combination of hydrogen bonds and charge neutralization
(H61~ne & Lancelot, 1982). According to the secondary structure predictions, none of the regions
of P-TY that could be crosslinked in situ to T Y M V R N A is able to form the e-helix-turn-e-helix
super-secondary structure, a suggested common feature of repressor and other DNA-binding
proteins (Sauer et al., 1982).
Conclusion
The coat proteins of the distantly related tymoviruses EMV and TYMV possess similar
structures although the observed amino acid differences and similarities do not correspond to
those found in homologous proteins such as enzyme families. One or two more tymovirus coat
proteins should be sequenced in order to ascertain which regions of the polypeptide chain are
common to all viruses of the family. However, many of the similarities observed between the
structures of P-E and P-TY are unlikely to have occurred by chance.
Analysis of the primary structure of P-E confirms several observations already made on the
structure of P-TY and helps understanding of why tymoviruses are stabilized by weak R N A
protein and strong protein-protein interactions (Kaper, 1975) with little interpenetration of
TYMV R N A and protein in the virion (Jacrot et al., 1978). The possibility of crystallizing
virions as well as empty protein shells possessing the same surface structure, combined with the
existence of short conserved potential R N A protein interaction sites, makes this group of
viruses a very interesting candidate both for high resolution crystallographic studies and for
studies on the relative stabilities of virions and capsids possessing R N A s and coat proteins
modified by directed mutagenesis.
We thank Professor R. E. F. Matthews for stimulating discussions, Mr Daney de Marcillac for computing Fig. 3
and 4, and Mrs C. Peter for her help in the preparation of this manuscript.
REFERENCES
ARGOS,P. ([981). Secondary structure prediction of plant virus coat proteins. Virology ll0, 55-62.
BOULEY,J. P., BRIAND,J. P. &WITZ,J. (I977). The stability of eggplant mosaic virus: action of urea and alkaline pH
on top and bottom components. Virology 78, 425 432.
CHOU,P. V. & FASMAN,G. O. (1974). Prediction of protein conformation. Biochemistry 13, 222-245.
CHOU,P. v. & FASMAN,G. D. (1978). Empirical predictions of protein conformation. Annual Review oJBiochemistry
47, 251 276.
DAYHOFF,M. O. (editor) (1972). Atlas q/' Protein Sequence and Structure. Washington, D.C. : NBR Foundation,
Georgetown University.
DAYHOFF, M. O., SCHWARTZ,R. M. & ORCUTT, B. C. (1978). A model of evolutionary change in proteins. In Atlas of
Protein Sequence antl Structure, vol. 5, supplement 3, pp. 345 352. Edited by M. O. Dayhoff. Washington,
D.C.: N B R Foundation, Georgetown University.
DUPIN, A., PETER, R., COLLOT, D., DAS, B. C., PETER, C., BOUILLON,P. & DURANTON,H. (1984). The primary structure
of the eggplant mosaic virus (EM V) coat protein. Comptes rendus hebdomadaires des skances de l'Acadkmie des
sciences, s~rie C 298, 219-221.
EHRESMANN, B., BRIAND,J. P., REINBOLT, J. & WITZ, J. (1980). Identification of binding sites of turnip yellow mosaic
virus protein and R N A by crosslinks induced in situ. European Journal of Biochemistry 108, 123 129.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27
Tymovirus coat protein sequences
2579
GUILLEY, H. & BRIAND,J. P. (1978). Nucleotide sequence of turnip yellow mosaic virus coat protein m R N A . Cell 15,
113-122.
HARRISON, S. C. (1983). Virus structure: high resolution perspectives. Advances in Virus Research 28, 175-240.
H/~LENE, C. & LANCELOT, G. (1982). Interactions between functional groups in protein nucleic acid associations.
Progress in Biophysics and Molecular Biology 39, 1 68.
HOPP, T. P. & WOODS, K. R. (1981). Prediction of protein antigenic determinants from amino acid sequences.
Proceedings of the National Academy of Sciences, U.S.A. 78, 3824-3828.
JACROT, B., CHAUVIN, C. & WITZ, J. (1978). Comparative neutron small angle scattering study of small spherical
RNA viruses. Nature, London 266, 417-421.
KAPER, J. M. (1975). The Chemical Basis of Virus Structure, Dissociation and Reassembly. Frontiers of Biology series.
Edited by A. Neuberger & E. L. Tatum. Amsterdam: North-Holland.
KOENIG, R. (1976). A loop-structure in the serological classification system of tymoviruses. Virology 72, 1-5.
KOENI6, g. & LESEMANN,D. E. (1979). Tymovirus group. Commonwealth Mycological Institute/Association of Applied
Biologists Descriptions of Plant Viruses, no. 214.
KOENIG, R. & LESEMANN, D. E. (1981). Tymoviruses. In Handbook of Plant Virus Injections and Comparative
Diagnosis, pp. 33-60. Edited by E. Kurstak. Amsterdam: Elsevier/North-Holland.
KONIGSBERG, w., MATTA,T., KATZE, J. & WEBER, K. (1970). Amino acid sequence of the Qfl coat protein. Nature,
London 227, 271-273.
LVTrLETON, J. W. & MATTHEWS,R. E. F. (1958). Release of nucleic acid from turnip yellow mosaic virus under mild
conditions. Virology 6, 460 471.
PAUL, H. L., GIBBS, A. & WITTMAN-LIEBOLD, B. (1980). The relationship of certain tymoviruses assessed from the
amino acid composition of their coat protiens. Intervirology 13, 99 109.
PETER, R., STEHELIN, D., REINBOLT, J., COLLOT, D. & DURANTON, H. (1972). Primary structure of turnip yellow mosaic
virus coat protein. Virology 49, 615 617.
PRATT, D., BRIAND, J. P. & VAN REGENMORTEL, M. H. V. (1980). Immunochemical studies of turnip yellow mosaic
virus. I. Localization of four antigenic regions in the protein subunit. Molecular Immunology 17, 1167 1171.
QUESNIAUX, V., BRIAND, J. P. & VAN REGENMORTEL, M. H. V. (1983a). Immunochemical studies of turnip yellow
mosaic virus. II. Localization of a viral epitope in the N-terminal residues of the coat protein. Molecular
Immunology 20, 179-185.
QUESNIAUX, V., JAEGLE, M. & VAN REGENMORTEL, M. H. V. (1983 b). Immunochemical studies of turnip yellow mosaic
virl~. III. Localization of two viral epitopes in residues 57 64 and 183 189 of the coat protein. Biochimica et
~iophysica acta 743, 226 231,
RACANIELLO,V. & BALTIMORE,D. (1981). Molecular cloning of poliovirus cDN A and determination of the complete
nucleotide sequence of the viral genome. Proceedings oJthe National Academy of Sciences, U.S.A. 78, 48874891.
RICE, D. M., SCHULZ, G. E. & GUEST, J. R. (1984). Structural relationship between glutathione reductase and
lipoamide dehydrogenase. Journal of Molecular Biology 174, 483-496.
ROSSMANN, M. G., ABAD-ZAPATERO, C., MURTHY, M. R., LILJAS, L., JONES, A. & STRANDBERG, B. (1983). Structural
comparisons of some small spherical viruses. Journal of Molecular Biology 165, 711 736.
SAWER, R. T., YOCUM, R. R., DOOLITTLE, R. F., LEWIS, M. & FABO, C. O. (1982). Homology among DNA-binding
proteins suggests use of a conserved super-secondary structure. Nature, London 298, 447-451.
SCHIFFER, M. & EDMUNDSON,A. B. (1967). Use of helical wheels to represent the structures of proteins and to identify
segments with helical potential. Biophysical Journal 7, 121 135.
SCHWARTZ,R. M. & DAYHOFF,M. O. (1978). Matrices for detecting distant relationships. In Atlas oJProtein Sequence
and Structure, vol. 5, supplement 3, pp. 353 359. Edited by M O. Dayhoff. Washington, D.C.: NBR
Foundation, Georgetown University.
STADEN, R. (1982). An interactive graphics program for comparing and aligning nucleic acid and amino acid
sequences. Nucleic Acids Research 10, 2951-2961.
STUBBS, G., WARREN, S- & HOLMES,K. C. (1977). Structure of R N A and R N A binding site in tobacco mosaic virus
from 4 ]~ map calculated from X-ray fibre diagrams. Nature, London 267, 216 221.
VIRUDACHALAM, R., SITARAMAN, K., HEUSS, K. L., ARGOS, P. & MARKLE¥, J. L. (1983). Carbon-13 and proton nuclear
magnetic resonance spectroscopy of plant viruses: evidence for protein-nucleic acid interactions in
belladonna mottle virus and detection of polyamines in turnip yellow mosaic virus. Virology 130, 360-371.
WESTHOFF, E., ALTSCHUH, D., MORAS, D., BLOOMER, A. C., MONDRAGON, A., KLUG, A. & VAN REGENMORTEL, M. H. V.
(1984). Correlation between segmental mobility and the location of antigenic determinants in proteins.
Nature, London 311, 123 126.
(Received 22 March 1985)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 03 May 2017 01:31:27