Download 669 Salmonella typhimurium - Journal of General Virology

669
J. gen. ViroL 0978), 4o, 669-673
Printed in Great Britain
A Model for the Adsorption of Phage P22 to
Salmonella typhimurium
(Accepted 3 April 1978)
SUMMARY
A new model for the adsorption of bacteriophage P22 to its host Salmonella
typhimurium is proposed. The main feature of this model is that only three of the
six tail proteins found on the mature phage function during adsorption. This model
explains why there is a difference in the specific endoglycosidase activity of the
tail protein of mature virions as opposed to unattached tail protein. It also
accounts for the cubic relationship between p.f.u, and tail protein concentration
in in vitro assembly experiments.
Phage P22, a temperate phage of Salmonella typhimurium, is an example of a Type C
phage in the Bradley classification (Bradley, 1967). The tail (or base plate) of this phage
consists of six spindle shaped proteins, each about I8 nm long, set around a short neck
(Israel et al. I967). Each tail protein, or tail part, is a dimer of the gene 9 product (P.
Berget & A. Poteete, personal communication), which is a polypeptide of 76ooo daltons
(Botstein et al. I973)- The neck, or central core, to which the tail proteins are bound is
about IO nm long and is attached to the icosohedral head (Israel et al. 1967; Botstein et al.
1973). A slender fibre, or spike of about 2 × 2o nm is attached to the tip of the neck (Yamamoto & Anderson, 196I ; Botstein et al. 1973).
The group of six tail proteins is the organ of initial contact with the host. Heads which
are complete except for the gene 9 product are unable to adsorb to cells. To such heads tail
protein binds readily and irreversibly in vitro, (Israel et al. I967). Free tail protein can also
bind the O antigen on the cell surface but here the binding is not irreversible. This is because
tail protein has an endoglycosidase activity which cleaves the O antigen at rhamnosylgalactose linkages (Israel et al. 1972; Iwashita & Kanegasaki, 1973). Thus, the tail proteins
destroy the sites they bind to. Tail proteins which are part of the mature virion still show
the endoglycosidase activity, but their specific activity is about four times lower than that of
unattached tail protein (Israel, 1976).
It has been known for some time that when P22 is assembled in vitro from heads and
free tail parts, the number of plaque forming units (p.f.u.) made depends on the tail concentration in the following way:
T3.3
p.f.u. - H2.3,
or
log (p.f.u.) = 3"3 log T - 2 . 3 log H,
where T equals tail protein in phage equivalents, H equals heads (constant), and T < H
(Israel et al. 1967). Thus the number of p.f.u, is proportional to approximately the third
power of T. Recently we showed that ejection of the phage DNA is also proportional
to T 3"3, but that initial adsorption of phage to cells is close to a linear function of T (Israel,
I976).
Any model of P22 adsorption must account for the following facts: (i) phage bind sensitive bacteria irreversibly, but unattached tail proteins do not; (ii) tail proteins which are
free are more active enzymatically than the tail proteins of mature virions, and (iii) the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 11:43:28
670
Short communications
plating efficiency of Pz2 is ext.remely sensitive to the number of tail proteins per particle. In
this note we present a model which we believe accounts for these findings and we also discuss
some discrepancies between our results and those of others.
It has been shown that the lipopolysaccharide (LPS) molecule of S. typhimurium consists
of three subunits, each of which can in turn be subdivided into three parts: lipid A, core
polysaccharide, and O antigen. There are thus three O antigen chains for each LPS molecule
(Rothfield & Romeo, I97i ). We believe that the initial event in Pz2 adsorption is the attachment of only one of the six tail parts to one of the O antigen chains. This is based on geometrical considerations. The O chains are of variable size consisting on the average of
about 30 hexose units. They are thus long and flexible, so the chance of any two ends being
close enough to be spanned by the width of the tail assembly seems remote. This idea that
the initial event involves one O chain and one tail protein is consistent with our observation
that heads with only one tail spike still show the endoglycosidase activity (Israel, I976, and
unpublished observations).
Immediately after binding, the tail proteins begin to cleave the O antigen. As the rhamnosylgalactose linkages are cleaved, the proteins apparently migrate down the chain. This is
based on the fact that when unattached tail protein is bound to ceils, it is not found free
in the medium until some time has passed, the extent of which depends on the multiplicity
of tails added (Israel et al. ~972). The matter is perhaps best put in terms of probability.
After each cleavage a tail protein could (i) continue down the chain; (ii) dissociate and
immediately bind to the same chain or one nearby, or (iii) dissociate and become free of the
cell. It seems clear that at the beginning of the reaction, the third possibility is the least
likely, but it is impossible to say which of the other two is more likely.
As the phage with one of its tail proteins bound to O antigen moves down the chain, the
chances increase that one of the other tail parts will bind to a sister chain. Ultimately each
of the three O antigen chains becomes bound to (and cleaved by) a separate tail protein.
This, we believe, is the critical moment in the adsorption process. Once the phage particle is
bound by three chains it becomes correctly oriented to the cell wall. This enables the long
slender fibre (Botstein et al. ~973) to react properly with some still hypothetical target site.
For the puposes of this model the site need only be small, inaccessible, and bounded by the
three subunits of the LPS. The O antigen thus acts as a kind of three sided funnel to bring
the fibre to the target site. When the phage is bound by only one or two chains, the chances
of hitting the site are drastically decreased. This explains why the number of p.f.u, formed
from heads and tail parts is dependent on the third power of the tail concentration. The
fact that no more than three of the six tail proteins are bound at one time to the O antigen
also explains why the specific activity of free tail protein is higher than that of virion
bound tail protein.
Presumably it is the interaction of the fibre with thc target site which triggers the beginning of
the DNA ejection process. It seems likely that the fibre-target interaction is also responsible
for the tight binding which is shown by Pz2 (Israel, ~977). This tight binding is probably
not the result of the action of the tail protein. In our original demonstration of the site
destroying activity of the tail protein, we showed clearly that free tail protein does not bind
irreversibly to the cell (Israel et al. I972). Nonetheless, there has been no direct evidence to
rule out the idea that irreversible binding is impossible for unattached tail protein but
possible for protein attached to heads. This could be explained by saying that attachment
to heads causes the tail proteins to undergo a conformational change which alters their
properties. In support of this idea one could point to the reduction in endoglycosidase activity of tails after they have become bound to heads.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 11:43:28
67I
Short communications
30
(a)
t ' ; '
?
x
1 '
30
'#
(b)
C
20
>2
2o
I
I
I
20,
40
60
10
!
I
i
20
40
60
Time (min)
Fig. I. Approximately ]o 9 phage equivalents of tail protein per ml were allowed to react overnight
with (i) IO~ heads per ml; (ii) 2'5 x io 9 heads per ml; or (iii) no heads. Each was assayed for endoglycosidase activity as previously described (Israel, 1976) as shown in (a). After 60 rain the cells
were removed by low speed centrifugation and the supernatant fluid was diluted twofold into
fresh labelled cells. The endoglycosidase reaction was then followed for another 60 rain as shown in
(b). Curve A, complete phage; curve B, tail deficient phage; curve C, free tail spikes. Endoglycosidase activity is expressed as the amount of filterable radioactivity which was released from
the labelled cells. Background values, obtained from unreacted labelled cells, were subtracted in
all cases. Heads and tails were assayed as previously described (Israel et al. 1967). The tails were
assayed with respect to the heads used in the experiment.
W e n o w have d a t a which show t h a t b i n d i n g tail p r o t e i n s to heads does n o t significantly
alter their O antigen b i n d i n g properties. I n the e x p e r i m e n t shown in Fig. I cells with r a d i o actively labelled O antigen were a l l o w e d t o react with m a t u r e phage, free tail protein, a n d
h e a d s which possessed a n average o f 2"4 tail p r o t e i n s each. The a m o u n t o f tail p r o t e i n was
equal in each case. A f t e r I h at 37 °C, the cells were r e m o v e d b y centrifugation a n d fresh
labelled cells were added. T h e d a t a show t h a t essentially all o f the p h a g e were r e m o v e d
with the cells b u t t h a t the b u l k o f the free tail p r o t e i n a n d the heads with two tail p r o t e i n s
r e m a i n e d behind, as j u d g e d b y their c o n t i n u e d ability to digest O antigen. This e x p e r i m e n t
m a k e s it difficult to argue t h a t the a t t a c h m e n t o f tail p r o t e i n to heads changes the p r o p e r t i e s
o f the p r o t e i n in a n y f u n d a m e n t a l way.
There are p u b l i s h e d d a t a which a p p e a r to conflict with the d a t a we have used to c o n s t r u c t
o u r m o d e l . E r i k s s o n & L i n d b e r g (I977) r e p o r t e d t h a t the hydrolysis rates o f free tail
p r o t e i n a n d w h o l e p h a g e were equal only when the c o n c e n t r a t i o n o f p h a g e was Iooo times
greater than the c o n c e n t r a t i o n o f free tail protein. W e have shown a difference in rates o f
a b o u t f o u r f o l d in earlier w o r k (Israel, I976) a n d a similar difference is also o b s e r v e d in
Fig. I. E r i k s s o n a n d L i n d b e r g used f o r m a l d e h y d e inactivated cells as substrate while
we used w h o l e cells a n d this m a y a c c o u n t for p a r t o f the difference. F o r example, the form a l d e h y d e inactivated cells m a y be m o r e efficient in causing the p h a g e to a t t a c h irreversibly
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 11:43:28
672
Short communications
i
|
I
i
i
10s
107
/
106
l0 s
~n
/
10*
I
1.65
I
I
I
1
2
4
I
I
810
Relative number of tails added
Fig. 2. Approxamately lO9 heads were reacted with varying amounts of tail protein and incubated at
25 °C. Each tube was assayed for p.f.u, at I5 rain, and 6 h after the incubation was begun.
and eject their DNA. Ejection has recently been shown to prevent further hydrolysis (A.
Wright, personal communication).
It is also possible that Eriksson & Lindberg (I977) underestimated their tail titres. They
indicated that they assayed their tails by using a formula like that given above, but they
obtained a slope of I'5 to 2.0. As shown in Fig. 2, when the reaction has not been allowed to
go to completion, the observed slope is less than the reported value of 3"3. Underestimating
the slope can cause a substantial error in estimating the tail titre. Eriksson and Lindberg
may also have underestimated their head titres, since they made their heads by thermally
inducing a lysogen carrying a prophage with a mutation in the tail gene. We have shown that
high temperature following induction results in the formation of many inactive heads
which are capable of binding tail protein (Israel, I967). Underestimation of the head
titre would cause a corresponding underestimation of the tail titre.
Bayer 0968) has proposed a model of phage adsorption in which the phage first attach to
the outer envelope, then move to points of adhesion between the inner and outer membrane.
Eriksson & Lindberg (I977) suggest that this is the case for P22, and Tomita et al. (I976)
have evidence which supports the Bayer model for phage eI5, which is very similar to
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 11:43:28
Short communications
673
P22. Although our model and Bayer's model are quite different, they may not be irreconcilable
I thank Myron Levine, Sherwood Casjens, and Andrew Wright for their critical reading of
the manuscript. This work was supported in part by grant number GB38236 from the
National Science Foundation of the U.S.A.
Department of Cell and Molecular Biology
Medical College of Georgia
Augusta, Georgia 3o9o2, U.S.A.
VANCE ISRAEL
REFERENCES
BAYER, M. E. (I968). Adsorption of bacteriophages to adhesions between wall and membrane of E. coll.
Journal of Virology 2, 346-356.
BOTSTEIN, D., WADDELL, C. & KING, S. J. (I973). Mechanism of head assembly and D N A encapsulation in
Salmonella phage P22. I. Genes, proteins, structures, and D N A maturation. Journal of Molecular
Biology $o, 669.
BRADLEY, D. (I967). Ultrastructure of bacteriophages and bacteriocins. Bacteriological Reviews 3x, 23o-314.
ERIKSSON, U. & LINDBERG, A. A. (I977). Adsorption of phage P22 to Salmonella typhimurium. Journal of
General Virology 34, 2o7-22I.
ISRAEL, V. (I967). The production of inactive phage P22 particles following induction. Virology 33, 317-322.
ISRAEL, V. (X976). Role of the bacteriophage P22 tail in the early stages of infection. Journalof Virology xfl,
361-364.
ISRAEL, v. (I977). E proteins of bacteriophage P22. I. Identification and ejection from mutant and wild type
particles. Journal of Virology 23, 91-97.
ISRAEL, V., ANDERSON, T. F. & LEVINE,M. (t967). The in vitro morphogensis of phage P22 from heads and base
plate parts. Proceedings of the National Academy of Sciences of the United States of America 57, 284-291.
ISRAEL, V., ROSEN, H. & LEVrNE,M. 0972). Binding of bacteriophage P22 tail parts to cells. Journalof Virology
xo, II52-I 158.
IWASmTA, S. & KANEGASAKI, S. (I973). Smooth specific phage adsorption: endorhamnosidase activity of
parts of P22. Biochemical and Biophysical Research Communications 55, 403-409.
ROTHFIELD, L. & ROMEO, D. (I971). Role of lipids in the biosynthesis of the bacterial cell envelope. Bacteriological Reviews 35, t4-38.
TOMITA, r., IWASmTA, S. & KANEGASAKI,S. (I976). Role of cell surface mobility on bacteriophage infection:
translocation of Salmonella phages to membrane adhesions. Biochemical and Biophysical Research
Communications 73, 8o7-813.
YAMAMOTO,N. & ANDERSON, T. (I96I). Genome masking and recombination between serologically unrelated
phages P22 and P22I. Virology I4, 430.
(Received 3I October I977)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 11:43:28