. an. .. I 3.1.... .. ......Hm
. .. . Lfifirfi. :flwufimé... «mammkvua .....-
. . Flflvunwuflm. 31.3%.”... Iflnnfir.m.3r....fi........u.wu.fi
. . . . . 4...-.. . 2 -..... I. . rI. ..u..I.I..
. . . . . . .. . . . . u-.......fih.l.a.u. M”. I .m. ...... ...... 4....
i . .... -. «...—.216“... [... ...- ...
I .. . . _. . . . .. . .. .r..,....r,.hI....u.......Nn.IrHNuv . ...... ..rtsxfinufi.
. ... _. .. . ..I.....Iv,..ltm.....u ...... IvIanvflI... Hflflflflnf ...
. . um.» . .. . .. . . . . , #1....1.........WIUII. - .-...vfi...f.--,I.1.. ..
. . . . . . . ......I -..... “fitness--. mama- .
_ 1 n, . ... .. . . . . . .. .... I “I. -
. . .... 5%.... . . .. ... .. [...l
I . m “L ..m V . e .. .. . .. . . I .gmnmulr
.... ... , ..NS . .. . .. .. .. -... in}. .. ..
. . . .I... . . . .. .....I INII .12 .. .
, . . . a 9% ...... a...%......§.+fid. ...".m. 3 . , . _ ., .. . . ......fimt. WWI... - 1.53;..-»
. .. . ..j.............. .53». 5.52.96.34.57 . . . . . . . ”EVILI .l. . .. . ---...
‘ ~31} u.* a or I _ “$.73;- if Iknl.u1- . . .. . . .. . .QltIII'... .’ o ‘ v. I I’Mfi
. A... . ...g r- . 9.1 3 3.1;? . . .. . ,. . . . . ..II...I.M..I.a_I.. I .... n...
. . . . - . . . a.» 5 .H ”m. . 3.9....3483. .. t . .. .. . . . . ... . . . _. . ....av..--......I n.....I -....-
. . .5 .c . . 3; .....rr; ... ... .I....._..h9 . . , . . . . . . . . . 1.1.... LI”... 2
v . . . I..Inu...c algal . 4.113: .... . l... t. . . . . .. . . . . . .... . . II.II.
,. . . Juan... ... :kau WEBM. flammnflgcmg . . .. . . .. , . I
Y I 3" h .7" I I . a .- I . v
. . .. . ...... . . ..J?. t.....«.....wu.r A..." .
. .. “um... ...... . Exam...» .. 3...... ......fiMdfimmmw .
. .41.... I .93.. 531...?” . ......J... ..
. 3. .. .....-T. . . --.... .... r ..
. - ... V pl ...-...!“ ‘1
.

    

   

.. r 5...»? c .. .9”. .wnMdhme .
. . Hug”. . ......r... . .9..%:. ...... 3 .
. . LEA... .. . .hfigmhhw ....fi....W.m.§r .
. . -....amvh ....nmw .31."... $39.11. ... . ...
. .... an... . v”: 5%.: “......IEuamdmam. +1... l..rm...u..~...fl.$....3th.§ ..I

"I n—N. 1 O 1" u ‘ h h in I

mmfimm ......» I a?

WP... 51".... ..I .

  

     

o I: ...}.kl...

 

. ...? .I . ...J.» ...ma.33sr.7.$umm.mvmu}f .....mwm. ...
. .3- ...nc. -. .-- . ..-%.:.n.. . fin... ...... ..mnqham.-w.§h-
... 4.1 I .5 ...iJGhflIIuZF 1:91.! v.» . .L. Iii}.-. I y... 35.
. .....-nui. $1.3... Ilunmhw... fig”... .. .....n......n...nm.# fi. . .
W ... .fl iI‘O-Vchi 64. A . It.” ‘ . Elma“... I .Iviundduanr: : $.- .7”.
......lwa. u... ....W...£Eonfin .. . Ida .mmfi , I.-.........wa.um._ - ..
.-. - «ISM ...! . 5mm... P
infinJRmK . .mWfiWaln. ”......firutu I WWWMMI .§th in...“ m..... _
“rill-FIE. finmdLKvachftfiflllr 1 . vi. ,. r 78.. ‘ e .o .
.lnufifidn ... "M1,...thjnlf 53;...
.- . I . ‘5»...

‘l

     

 

                            

 

......e...£c ... .Wm... .
-.. ...- ..flc.;....b....hI.u.m.w....kéumwmmmmwmwn .n...
s -..- . ......L.......ffi
. ......“wfifibfim... -fihu ...rw.........w..uu.mtu.h.w.a5mwm. ...fiwwmw. ....HJVV.....§B.I..¢-::
1.7 ., I}. A

_ v: .2 .... 3......31. I....I
....lJ..3.........L.nmm.r:.. . . ..
I I... III ...»- ..r . 6...... ...“...Ig. . 5.1.... . . .
...Wup<l.nflrlnluw.aumliuunlum.....k .... I-.. . .n... -.

 

7 7...
.

   
  

Ark!
- -.. .I
. . -
. -... .7)":
......I..- .. . I....:
. 3.0:)..- ......I {I
..- . . - 2.31..
-....J.......Ill
. -. ...- .
. . 4:31:42...
. .I. .r.”-|!.I“IU. CF... '0 ul
. 57.. .-.-..L: .
. . 1.1.2:..-

      

 

 

 

 

               

 

                                         

                                          

v
I. III-1.,
‘IA .4“.

.. .33....n: an?

   
  

   

      

 

    

       

             

  

. ...

. . x. . . I...“

. ..‘rio .IM'O gt: 1211.2hbdhfl‘l-r‘ .. .

. .- - I~I.H.I)J.I..mntmmm.~.:nr..ޤst mythiw. . 3...... 1.5%: . . u .

....n... . IIWM ..ymVLwIflul.......w5.f£..a.x..mmo.v.raru.. LhfiJer.lwnwurqv.mHsz¢zmun...»wm..fiuxuuwwnn ....L: .. I...

. -..--- - 2 I u . ..XI . v I . . .. . -... .. . - .. II.-.

.. . . Q .32.... 2 - ....no - $.93“. ...... . .1..J....::..%... 9...... - 3...... ...... -... -....-I ...-

.. 2:31.”... -wn... ...... ...- £53. ..h...-...§alI..... ......»h. «1%.... ....T unnm ......8. . heart... .
, 71.41.)..9. ...II II. Iv! £11-.. I?! ULIIcfgIVII.‘ 11..<Q¢|AV‘C 150”" $9 In... 1 4| 'Iallrut.
. 4 a I ICIEII: .. oI. !.I.- JV. IV! lay-315090;.» vulv.‘?..~1~ .. u I .....v 1.2:}...ir

..UJ. .v. ... 2...”. I: .....n..v....u.....|. Job.“ .... «...... ..xflhanonruflfl. ...-.... . .. . ....-- I-

. I31... .JI. I....I...IUfi.I.I...I, I..II Iv ...!n. -rsmru...I.{.. . l ......mx;s.r .

....I.. ....c ....r..u....u. tuflllwfiqu... he... . I. .mfllux.N.I.IB- L... #Hhuuwfi. I...

. .... ....H. dwfiwfld p.83 Ih

.c INANIU .IIA-...H... .

       

                                                 

V?’ n
. -...-.I..?

. I? I. .. . ..
. . I u 5. Iflhhmfififib. Ln... -.n. 1n....Mfl.I.vmx.fl...nq.ufluInnUnA£
... ......u-.P.....mN.u....i...IL...w........ figmqrmfiufiizwlfi. .
-.... .... ... .. «Fly... 6mm?- .: .. ... J.
. . . ... I... . 3.....-B -.....ij I... ...Iiime .Cbunuamm.
. . -».I..1fi- . Ithrawhfx %:_I.nw.§...-3z, ..
. . . .. ...‘I" 4". DII‘ 2‘ng ‘i .I. v v .r ...» I?
. . . .. I... l .....d......h.1 .....9...n..~....r..1..5......§rwfi. .-.-
. . Immunwv .....ubuunanmfiuI - , ..I. .1... ...... . ...

0£1eoi .

ES...

-...
“59-... -- .

  

. I'd:

         
 

'1 .
I . 3......» .... prep hIImWfiWWb
. o . ..
.31..-.1- --. - I... .-.. -
. .. . . n. avg...“ wflIeflgdfiflnnéII .
-....III. ...n. (.11... ...V...I.....I...HI .ILIltufl M... ...}... ......34. .....u... . .
.I‘IslkIh .I.0'I .‘JIOIIIIvtIKIJ I YIIV‘I vaihvu‘fiv - II. .11.}:
. . ..u I...I .... ....I.........-I 9... ...?I....-t-
I... -II.....uI.h.,.....I..-IIr.2I:-II ......3...
I ... {III}... I--.I..I.I:. .....r. .... fl .
. . .I......| n «a... . 1 II II........!.... .....u....n|o........r ..
6.11 IIIIgI. 4%.“.--1 5.....II III.-.
..- ...I -..! ... I. ... i412... ...: ... -..a
. In It .30 . Ln. hull...) a: o'ulc!....v.l .fil!
I IIIIu..1.u-F.I.”II. .. ... .
. ...».-IIIIE .I-~..-§I%. .
...... . v. .- ......1. II
I IV... ......

I!
Q . II I IIIV

.
. . . .. .
I I J I:

.4 - . b. , .uu... .. -... .I. “but?!
...m...1...t.......n Human“? ....I... . .....n. ..

5

  

Haul 6.1.». .
t

    

 

   

 

                      

  

    

   

  

 

1...... t
I ... u I r 13....
I}!!! u v I u. r .98!
..II...I..I.IIJI.I .. IL..-X.!..I...)..
It...a.v......II.nV._...I.I....... nu..I........I.-. .................... -.. u .I.-.....-,. . I. . . .
2.1.. .-..u .3. .....4 . .-.II...vI.:5 .. yl...va....:.. ..\ 1.1.90»..le I...
I. .-.LV...I.I.....MV.|..I.\ ...-nv... ..I.......L....~....-.v.l.--.. .-.. $97-11...
.-. -..! II. . .I-..| It! I. III. ...... .... ......
I-I.V.I.II (I IIIIII ..IIhII... ...w»......o...3!.9...sb . v v
.II... . ... III... «31!-..
II.‘ 4. .Whli ...III..--I.4...9..I.... ..I
. ‘5..I..IIV1. . . “In ... .1315..-
.. .hIsI - _ . ......l.u.. .. .
. InfilfIhBI-.. .. ...Amwrhflflvu ....
nu...".....Ia.I.~. ...... . . a...» .........r....4.t. .
. III...uu........o......h- . . . .q .-.I-
.I.I...II.II
0". o'l‘ z. I! .
-.-- . I.I
-.....I-I III
.I‘U" J“-
I... .....
‘UI'I v

   

. .
.I. ..... .... -I. ...
.9h............Is....... in...
.....J. ......ufi.fi.«w1...3. .
.- . I. .....Iurlo. Aw..-.r........l..t. -
... ...n. .- . I. $.41“)...
. . I :I $93153...
. ......JJSI.
”HM-rad-

I 4‘. v.

 

‘I
I.
I.
KI
:1
:3

LL

 

.2:
:3'
q

o I;

   

-.
'i

.nI.)

I

.. In ..- u. “...th
. .. . . .. I... “IuIInV.
. .- ,I.I.qu rd... ...I..WP..

. . gunk”. Wu.mu..fl.n.,uvh....m.nhI. ...I. .... I ..
.40.“.1‘ --.”- In“... .II LI I

  

I'D-7':

‘1:

n I...

         

 

u. . .....u... .....Iorm .....n...

            

 

I.
I!

. . - Inga
'. ' ‘
. III III. II
. . . . .. IJIIJJ: . IIIIIIIFI
. . . -....tiyr .-II....II«I..II.¢PIM. IIIII-IBHIII
. .. z s-.. ...-I..- ..I I--. -.I I.IIH..,-II
.-....uaL. -..!tflufv... III. . 2...- «WI... ..... ..Iwu ...... ..-)..urrl .-.... .. I... .
II . . . . - II..- - I}..- . ....-I.- --I.-- .. .
. I -I It... 1.1.1.5..- - --J. --..-r... - c. .I . -.
. . I...I........o u-....IIc...I ...-.nIPlJfil -I II. .... in. . ..
.I.-. .... ...IJ! II--2.\I..-5..5-II.I If! .1. . ..Jil. .
- I- . .3 ......i ......I....... ..
--.... . 51.3 ....I .... . .....- . .,
.- .I- .. k1... I.I -.. . 1 . I
n.kw....u...x....uIII -...II. -- I - - .

III:
II I.
III. IfiIII

. I..-.IIIIII.I~I»--.. l.

I III .

 

0 ._
m.“
L I lvhxulll‘lui I.
I . . . ".v| I I

 

2” IllllllllllllllllllllmlllllllllltllllllllllIllllllllllllll

 

LIBRARY
Michigan State

L Unlverslty

 

 

 

This is to certify that the

thesis entitled

Genomic localization of the 5'—linked protein (VPg)
of Pea Enation Mosaic Enamovirus

 

presented by

Christiane Elisabeth Wobus

has been accepted towards fulfillment
of the requirements for

Master's degree in Want Pathology

WM;

M!

Date 3/5/97

 

0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution

 

PLACE ll RETURN BOX to remove this checkout from your rocord.
TO AVOID FINES Mum on or baton dd. duo.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

F—T

 

 

 

 

 

 

 

l:[__J
l l-
l I-
II I |

MSU I. An Namath/o ActloNEquol Opportmlty Institulon
W

 

 

 

 

 

 

 

 

 

 

ulna-9.1

Genomic localization of the 5’-linked protein (VPg)

of Pea Enation Mosaic Engmgvirus
By

Christiane Elisabeth Wobus

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Department of Botany and Plant Pathology

1 997

ABSTRACT

Qenomic localization of the 5’-linked protein (VPg)
of Pea Enation Mosaic Enamgvirys
By

Christiane Elisabeth Wobus

This study reports that the protein sequence of Pea Enation Mosaic Virus
(PEMV) genome-linked, viral protein (VPg) is encoded within the 84K ORF 1 of
RNA 1. It is composed of 28 amino acids with a molecular mass (M,) of 3157
Da. Position and size of the VP9 suggest that ORF 1 is expressed as a
polyprotein. This is the first evidence of proteolytic cleavage of a protein
encoded by PEMV RNA 1 or by Luteoviruses. Computer-generated VPg
sequence comparisons did not reveal sequence similarity to other viruses
including Luteoviruses. Anti-VPg antibodies were produced and used in the
identification and characterization of the VP9. A 15 kDa protein doublet was
detected in protein extracts from RNA 1 and RNA 2 infected plants but not in
extracts of plants infected with RNA 2 alone. Biotinylated anti-VPg antibodies

detected a similar size protein in RNase A-treated preparations of viral RNA.

ACKNOWLEDGEMENTS

I gratefully acknowledge my professor Dr. Gustaaf de Zoeten for his
support in making this thesis possible. I also thank Dr. Richard Allison and Dr.
Rebecca Grumet for their help on my graduate committee. I thank Dr. Jihad
Skaf and Debby Rucker for their assistance in the laboratory. Thank you to Joe
Leykam and the staff of the MSU macromolecular structure facility for protein
sequencing and scientific assistance. Thanks to the staff of the University
Laboratory Animal Resources (ULAR) for the care of the rabbit, to Kurt Stepnitz
and Marlene Cameron for their photographic expertise, and to the department
office staff for their help. Lastly, I like to thank my husband Erik and my family
for their love and support, without whose help this work would not have been

possible.

iii

TABLE OF CONTENTS

LIST OF TABLES .............v
LIST OF FIGURES........... VI
CHAPTER 1: Literature revrew .......1

Pee Enation Mosaic Virus (PEMV)
1. genomic organization...

2. interaction of RNAs and symptoms

3. economic importance, host range, environment, control
4. aphid transmission and vector-virus relationship... ..
5. cytopathology, replication and translation.................................12
Relationship ofPEMVtoothervirusgroups..................................16
1. evolutionarylinks... ....16
2. similarities and dissimilarities between PEMV. and Luteovrruses ..18
3. What type of complex does PEMV represent? 21
VPg (viralprotein, genome-Iinked)..............................................22
1. general characteristics..........................................................23
2. poliovirus VPg .24
3. VPgs of plant VIruses .........27
4. rationale for prolect .........34

CHAPTER 2: genomic localization of the 5’ linked protein (VPglm of Pea
Enation Mosaic Enampvirgs... ...35

basis

Introduction... 35
Materials and Methods ........38
Results ....49

CHAPTER3: DIscussmn55
APPENDIX... 67
Appendix A: mutageneSIs of the 33K ORF of RNA 2 ..................... 67

LIST OF REFERENCES... ....90

iv

LIST OF TABLES

Table 1: Northern blot analysis of plants infected with wild-type and/or
UORF infectious clones... ............87

Figure 1:

Figure 2:

Figure 3:

Figure 4:

Figure 5:

LIST OF FIGURES

Localization of PEMV RNA1 and RNA2 open reading frames ...... 3

Western blot analysis of total protein extracts isolated from pea
leaves ....50

Differentially treated VPg attached to PEMV RNA probed with

biotin labeled rabbit anti-VPg lgG... .. ......52
PEMV-VPHg protein sequence and its localization within ORF 1 of
Mutated region of ORF1 (33kDa) of PEMV RNA2... ..........84

Chapter 1

Literature review

Pea Enation Mosaic Virus (PEMV)
1. Genomic organization

Pea Enation Mosaic Virus is the sole member of the Enamovirus genus.
The two linear single-stranded (ss), plus-sense (+) RNAs are encapsidated in
two icosahedrons of different sizes. The 5706 nucleotides (nt) of RNA 1 are
encapsidated in a polyhedron of 180 subunits (T=3) with the diameter of 28nm.
While the 25nm diameter particle consisting of 150 subunits contains the RNA 2
genome of 4253 nt. In viral preparations the particles were found to sediment at
different rates, leading to the naming of bottom (B) and top (T) component, for
RNA 1 and RNA 2 respectively (for review see Demler et al., 1996a). The RNAs
could not be aminoacylated and contain no poly (A) sequence at their 3’ end
(German et al., 1978). The 5’ end of the viral genome is linked covalently to a
protein (VPg) of 17.5kDa (Reisman & de Zoeten, 1982). Some strains also
contain a third RNA of 717nt, which is a satellite RNA of RNA 2 (for review see
Demler et al., 1996a).

RNA 1 contains five open reading frames (ORF) and encodes all
structural proteins (the numbering system is based on Miller et al., 1995) (Fig.
1). ORF 0 starts after a 183 nt non-coding region and has the potential to

encode a 34kDa protein. The C-terminus of the putative protein is acidic while a

small region towards the N-terminus is characteristic of transmembrane
proteins (Demler & de Zoeten, 1991 ). However, no significant sequence
homology was found in database searches leaving the function of the putative
protein unknown. ORF 1 (84K) is initiated out of frame from ORF O at a second
AUG at nt 265. The putative 84kDa protein contains several hydrophobic
stretches at its N-terminus characteristic for membrane spanning domains. A
helicase lV-like motif and a 3C-like protease motif were identified towards the
center of the ORF. The sequence at the C-terminus did not resemble any known
protein sequence (Demler & de Zoeten, 1991). ORF 2 encodes the viral
polymerase and extensively overlaps at its 5’ end with ORF 1. The occurrence
of a frameshift-motif led to the hypothesis that the product ORF 2 is expressed
as a frameshift fusion protein with ORF 1 (Demler & de Zoeten, 1991). A
different mechanism is used to express the 3’ half of the genome encoding the
structural proteins. The intergenic region between ORFs 1, 2 and ORFs 3, 4, 5
resembles a signal for subgenomic RNA production (Demler 8. de Zoeten, 1991).
ORF 3 encodes the major capsid protein (cp) of 21 kDa (Demler & de Zoeten,
1991), previously identified as a 22 kDa protein (Shepherd et al., 1968; Hull &
Lane, 1973; Hill & Shepherd, 1972). In addition to the 21 kDa cp, aphid
transmissible isolates also contain a minor capsid protein of 54 kDa. The latter
is the result of read-through (RT) of the 21 kDa cp (ORF 3) and a 33kDa protein
(ORF 5) and is involved in aphid transmission. At the N-terminus of the 33kDa
protein a 4kDa proline rich region connects both proteins; this might form a

hinge to allow for better interaction of the aphid transmission protein with the

RNA1 PEMV 5706 nucleotides
ORF1 84K

 

 

 

 

 

Jhydrophoblc protease I “K @
[ ORFO | [ ORF2 | '21K(cp) 33K (An'
34K 67K [oars] oars |

 

(Polymerase) ®

RNA 2 PEMV 4253 nucleotides

 

 

 

27K
33K
[ ORF1 | f oars
* | ORF2 J 26K
’ 65K
(Polymerase)

Figure 1: Localization of PEMV RNA1 and RNA2 open reading frames (ORFs).
The localization of RNA1 and RNA2 ORFs is based on Demler & de Zoeten
(1991) and Demler et al. (1993), respectively. The numbering system of RNA1
ORFs is based on Miller et al. (1995). The hydrophobic region and the protease
motif, identified by Demler 8. de Zoeten (1991) are indicated. Abbreviations:
coat protein (Cp), aphid transmission factor (AT), subgenomic RNA signal (SG),

read-through (RT), K=kDa

aphid receptors (Demler & de Zoeten, 1991). The most convincing evidence
for the expression of the 54 kDa fusion protein via RT was obtained when N-
tenninal protein sequencing of the 21kDa and 54kDa proteins resulted in
identical sequences (Demler at al., 1996a). The involvement of the 54kDa
protein in aphid transmission was shown by the use of infectious cDNA clones.
These clones of the 3’ portion of RNA 1, including ORF 3 and ORF 5, of an
aphid-transmissible (AT‘), aphid-nontransmissible (AT) and a naturally
occurring, nontransmissible deletion strain (ATA’) were used to generate cDNA
clones with various exchanges of portions of the parental clones. When these
infectious clones were inoculated together with RNA 2, it was determined that
the 110nt near the 3’ end of the 33K ORF are critical for aphid transmission
(Demler et al., 1996a). The 3’ end of RNA 1 ends in a 227 nt non-coding region.

RNA 2 has four confirmed ORFs (Fig. 1) and a potential 15 K 3’ terminal

ORF that lacks a termination codon is dispensable in infections and expression
of this protein is unlikely (Demler et al., 1993). ORF 1 is initiated after a short
5’terminal non-coding region of 21 nt. This ORF can potentially encode a 33
kDa protein but no function is assigned due to lack of statistically significant
sequence homology to known sequences (Demler et al., 1993). It is believed
that the protein product of this ORF migrates anomalously and corresponds to a
45 kDa protein detected after in vitro translation of viral RNA (Gabriel & de
Zoeten, 1984; Demler et al., 1993). The second ORF encodes a 65kDa protein,
the second viral polymerase. ORF 2 overlaps slightly with ORF 1, however, in a

different reading frame. The occurrence of a minor translation product

corresponding in size to a fusion protein derived from ORF 1 and 2 suggests
that expression occurs via frameshifting (Demler at al., 1993). An intergenic
region thought to encode a subgenomic promoter separates ORF 2 from ORFs 3
and 4 which have the potential to encode a 26kDa and a 27kDa protein,
respectively. The reading frames are superimposed upon each other but are in
different reading frames (Demler et al., 1993). Preliminary results provide
evidence that these proteins are involved in movement (J. Skaf, unpublished
results).

Some strains also contain a satellite RNA of 717m (Demler & deZoeten,

1989). Like other satellite RNAs, this RNA is characterized by a lack of
sequence homology with the host plant or viral genome, it is non-essential in
virus infection and depends on the helper virus for replication and encapsidation
(Roossinck et al., 1992). No extensive sequence homology was found between
host RNA, RNA 1 or RNA 2 and the satellite (Demler & de Zoeten, 1989) or to
other satellites (Demler et al., 1994b). It is non-infectious by itself and has no
influence on symptom expression, aphid transmission or particle morphology in
peas when inoculated together with RNA 1 and 2 (Demler 8. de Zoeten, 1989).
RNA 1 provides the structural and aphid transmission functions while RNA 2 is
solely responsible for replication of the satellite. Very limited sequence
homology exists between the satellite RNA and RNA 2, only 12 out of 14 nt at
the 5’ end and 7 of 8 nt at the 3’ terminus are identical. This is thought to be
required by the polymerase of RNA 2 in order to recognize the satellite RNA for

replication (Demler et al., 1994b). No multimeric or circular forms were found

suggesting a replication mechanism involving linear, monomeric
intermediates. Computer analysis of the satellite sequence identified several
ORFs, up to 4.3K on the plus-sense and up to 7.2 K ORFs on the minus-sense,
but in vitro translation of purified satellite or transcript RNA showed no protein

products (Demler et al., 1994b).

2. Interaction of RNAs and symptoms

PEMV historically has been characterized as a bipartite virus since both
particles are required for infectivity. However, after the elucidation of both
genomic sequences, the interdependencies of RNA 1 and 2 became more
obvious. The production of infectious transcripts and with it the possibility of
infecting plants and protoplasts with individual RNAs or with a combination of
RNAs brought clarification to the functions provided by each RNA. RNA 1 is
solely responsible for encapsidation and aphid transmission. Evidence shows
that genomic and subgenomic RNAs of RNA 1 were formed in the absence of
RNA 2 and that RNA 1 is independently capable of cp expression and virion
assembly (Demler et al., 1994a). RNA 2, on the other hand, is needed for
mechanical inoculation and phloem evasion. Protoplasts inoculated with RNA 1
or RNA 2 alone, showed replication of the respective genomes, indicating the
replicative independence of each RNA. However, inoculation of plants with RNA
2 resulted in a systematic, although almost symptomless, infection (Demler et
al., 1994a). Only after prolonged infection (14-28 days post-inoculation [dpi]) did

the RNA 2 infected plants show chlorotic yellowing and mosaic patterns or mild

mottling of upper leaves (Demler et al., 1994a). The development of
symptoms correlates with electron microscope (EM) findings that only RNA 1
infection causes the formation of double-membrane bound replication complexes

in the cytoplasm of infected protoplasts (see below) (Demler et al., 1994a).

3. Economic importance, host range, environment, control

PEMV occurs worldwide and causes significant losses in pea (Pisum
sativum), chickpea (Cicer arietinum), lentils (Lens culinan’s) and field bean (Vicia
faba) in the United States as well as in Europe (for review see Hagedom, 1996).

Its natural host range is restricted to leguminous plants. Alfalfa
(Medicago sativa) is the most important and widespread host, while some
members from the Vicia , Lathyrus, Medicago, Pisum, Melilotus, Lupinus,
Tn'folium and Lens genera have also been identified as natural hosts. However,
some non-leguminous plant species are local lesion hosts, mainly members of
the genera Chenopodium and Nicotiana (for review see Hagedom, 1996).

In addition to the plant species, environmental factors also influence
occurrence and severity of symptoms. Due to the aphid-transmissibility of the
virus, factors influencing the life cycle of aphids become especially important.
For instance, peas sown later in the growing season always display higher
incidence of infection by PEMV due to an increase in aphid populations. The
weather also has a profound effect. A high rate of transmission is observed on
warm, humid and stormless days as many alatae (winged) aphids develop and

move with the wind to uninfected peas. However, cold or hot, dry weather and

rainy days reduce aphid multiplication and the development of winged
individuals, which results in low PEMV occurrence (for review see Hagedom,
1996). Most efficient aphid transmission occurs around 20°C. In addition to the
aphid, the outcome of a PEMV infection is also determined by the temperatures
following inoculation. Generally, lower temperatures (12-14°C) reduce and
higher temperatures (28-30°C) increase the manifestation and severity of
symptoms in plants (for review see Hagedom, 1996).

The different parameters influencing incidence and spread of PEMV can
be controlled to reduce crop losses. Today, the development of host plants
tolerant to PEMV infections is used commercially. Genetic resistance of varying
degrees is determined by a single, dominant locus En. Introduction of this gene
into susceptible cultivars resulted in excellent resistance. In recent years, the
marker locus Adh-1 (coding for alcohol dehydrogenase) has been identified, that
is tightly linked to En, permitting the screening for PEMV resistance by Adh-1
signals (for review see Hagedom, 1996). The following control measures
influencing different parameters are also possible. Aphid transmission of this
virus is in a persistent manner and the use of insecticides can reduce the vector
population and limit virus spread. However, due to the erratic occurrence of
severe outbreaks, this practice is not widespread (for review see Hagedom,
1996). PEMV has to overwinter in leguminous perennials in order for aphids to
carry PEMV to new plants in the spring. A solution would be the development of
aphid-resistant perennial legumes but this is not an active control effort at this

time (for review see Hagedom, 1996).

4. Aphid transmission and virus-vector relationship

As described earlier, the aphid vector plays an important role in the
dissemination of naturally occurring PEMV infections. Therefore, a considerable
amount of research has focused on the interaction of the virus and its vector.

Aphid transmissible (AT') and nontransmissible (AT) strains are known
and naturally occurring PEMV infections contain a mixture of AT' and AT strains
(Demler et al., 1996a). Both strains in these mixtures can be transmitted by the
aphid due to a process called phenotypic mixing, where the AT strain is
transported within a virion containing cp determinants of the AT’ strain
necessary for aphid transmission during encapsidation (Adam, 1978). In the
laboratory, an AT' strain is often converted into an AT strain after repeated
mechanical passage. This is manifested upon purification of the virus in a
higher yield of aphid-nontransmissible compared to aphid-transmissible strains
(reviewed in Demler et al., 1996a).

PEMV is transmitted in a circulative, nonpropagative manner. Circulative,
also called persistent, transmission is characterized by the following: (I) The
virus is acquired through the food canal and subsequently moves to the aphid’s
intestines. (ll) After the virus has been taken up by the aphid through feeding,
termed the acquisition feed, a latent period follows during which the virus passes
through the hindgut wall into the hemocoel and subsequently into the salivary
gland. (lll) Finally, the virus is ejected with the saliva from the maxillary salivary

canal and can infect another plant (Matthews, 1991). Support for this type of

10

virus movement through the aphid comes from many EM observations
(Harris & Bath, 1972; Harris et al., 1975; Harris, 1979; Harris, 1990). Virus was
found concentrated after a 24 hr. acquisition period in the stomach, the intestine
and the hindgut lamina but not in the digestive tract or foregut (reviewed by
Demler et al., 1996a). It is thought that, in the case of PEMV, crossing the
barrier between hindgut and hemocoel is accomplished by simple penetration of
the gut epithelial cells, since no evidence could be obtained indicating an
interaction of the virus with the plasma membrane of these cells as characteristic
for phagocytosis. Once in the hemocoel the virus circulates until membrane
mediated endocytosis allows the virus to cross a second barrier to reach the
salivary system. AT‘ and AT virions have been observed crossing the gut-
hemocoel barrier but only AT’ virions could be detected in the salivary system.
Therefore, aphid transmissibility seems to be determined by a successful
interaction with the accessory salivary gland membranes (reviewed in Demler et
al., 1996a).

Circulative transmission is divided into propagative and nonpropagative
transmission, depending on virus replication in the aphid vector. Only a few
aphid species transmit viruses circulatively; and these viruses commonly
produce yellowing and leaf-rolling symptoms (Matthews, 1991). Myzus persicae
and Acyrthosiphon pisum are the two most important vectors of PEMV which
produces mosaic patterns (yellow spots) in the leaves as well as rolling of the

infected leaves (see below).

ll

PEMV is a circulative, nonpropagative virus. Although, its short
acquisition and inoculation feeding periods as well as the ability to be inoculated
during brief aphid test probes of non-phloem tissue are characteristic of non-
circulative viruses. However, a latent period prior to successful aphid
inoculation, typical for circulative transmission has been observed (reviewed in
Demler et al., 1996a). Other evidence for circulative transmission is that
prolonged acquisition feeding leads to an increased transmission efficiency
which results in a higher incidence of infected plants (Fargette et al., 1982).
Support for the nonpropagative nature of PEMV, eg. no virus replication in the
aphid, comes from the following observations: One, transmission efficiency
decreases with time in aphids denied subsequent acquisition feedings. Two,
aphids lose inoculativity (the ability of an aphid or insect to deliver infectious
virus into a healthy plant (Matthews, 1991]) after serial passage of hemolymph
between aphids. Third, transmission efficiency correlates with the amount of
virus acquired by the aphid (reviewed in Demler et al., 1996a). Adam (1978)
hypothesized that PEMV is a nonpropagative virus that is stored but not
replicated in fat bodies of certain cells of the mesoderrn. Electron-microscopy
showed virus inclusions in the cytoplasm of stomach and midgut cells, as well as
in connective tissue, fat and basophilic mesodennal cells. Phagocytosis causes
the inclusions to be incorporated into endocytotic vacuoles and other parts of the
Iysosomal apparatus. However, no virus breakdown was observed (reviewed in
Demler et al., 1996a). Heterologous encapsidation during infections of plants

with AT and AT’ mixtures was observed. Transencapsidation might take place

12

during replication in an aphid. However, injections of aphids with a mixture
of strains does not yield heterologous virions (Adam, 1978). Furthermore, no
virus multiplication was detected by enzyme-linked immunosorbent assays of

aphids (Fargette et al., 1982).

5. Cytopatholog y, replication and translation

Changes in symptoms of PEMV infected plants are accompanied by
distinct changes in ultrastructure. No symptoms and no ultrastructural changes
are visible up to 4 days post-inoculation (dpi). Plants infected for four to nine
days show signs of necrosis, a downward curling of upper leaves and an
increase of vesiculation in the sieve elements (Demler et al., 1994a; de Zoeten &
Gaard, 1983). At the same time electron-dense structures appear in the
vascular bundles and PEMV virions are detected in the cytoplasm of all cell
types as well as in the nuclei of these cells especially in the region of the
nucleolus (de Zoeten & Gaard, 1983). However, it is currently not known which
of these sites, nucleus or cytoplasm, is the place for virion assembly. Virions in
the cytoplasm appear randomly scattered in loose aggregates or are contained
in electron dense, dagger-like tubules in the plasmodesmata. During this time, 7-
10 dpi, visible symptoms like vein clearing and yellow mosaic patterns appear
while stunting and epinasty of aerial parts is seen in combination with an
amplification of the foliar symptoms 7 - 14 dpi (Demler et al., 1994a). Elongated
feather-like crystalline inclusions are found in the perinuclear space and in

epidermal cells starting 14 days after inoculation (reviewed in Demler et al.,

l3

1996a). Enations on veins on the lower leaf surface do not become evident
until late in the infection (starting at 20 dpi) (Demler et al., 1994a).

The typical symptom of PEMV infected tissue at the ultrastructural level of
resolution is the presence of double membrane bound vesicles in phloem related
tissues and other parenchymatic cell types. These vesicles originate at the
inner envelope of the nucleus and migrate into the perinuclear space. On their
way into the cytoplasm the vesicles gain their second membrane when crossing
through the outer envelope of the nucleus. Each of these single-membrane
bound vesicle contains 2-30 individual single-bound vesicles that contain
aggregated granular material and DNase sensitive fibrils (de Zoeten et al.,
1972). These vesicular structures often appear near plasmodesmata and are
especially numerous in phloem tissue and sieve elements, maybe indicating a
role in cell-to-cell or long distance movement (Demler et al., 1996a). The extent
of vesiculation and the amount of nuclei containing virus particles increase with
time after inoculation (de Zoeten et al., 1972). This and the following
observations lead to the conclusion that replication is linked to the nucleus and
the vesicles originating from it.

Positive strand viruses produce (-) strands during RNA replication and
subsequently more (+) strands via the formation of replicative forms (RF) and
replicative intermediates (RI). The RF is a structure of full-length (+) and (-)
strand basepaired to each other while RI only contains partially basepaired
structures (Matthews, 1991). De Zoeten et al. (1976) used in situ hybridization

and ferritin labeled antibodies (Abs) to determine the site of viral RNA synthesis

14

by localizing double-stranded (ds) viral RNA to the nuclei of infected cells.
However, the applied techniques were not able to establish whether RF and/or
RI were present. Both forms were previously shown to be present in PEMV
infected plants (German & de Zoeten, 1975). The involvement of vesicles in
replication in addition to the nucleus was shown by the association of negative
strand RNA with these structures (Powell et al., 1977). The hybridization studies
were accompanied by experiments showing incorporation of radiolabeled
precursors into nuclei and vesicles in the presence of actinomycin D (AMD) (de
Zoeten et al., 1976; Powell & de Zoeten, 1977). AMD binds tightly and
specifically to ds DNA but not to ss or ds RNA, therefore, preventing RNA
synthesis of the host without affecting viral RNA synthesis (Stryer, 1988). It
could also be shown that the viral RNA polymerase was localized to nuclei and
vesicles and required the presence of all four nucleotide-triphosphates and Mg”
for maximum efficiency (Powell et al., 1977; Powell & de Zoeten, 1977).
Extended experiments used isolated nuclei to study PEMV replication (Powell &
de Zoeten, 1977). It was determined that 60% of virus-specific RNA is (+) $8
and 40% ds and that H ss RNA exists only in ds form or bound by protein but
not as free RNA. Intact progeny virions were not observed indicating that only
early stages of the virus life cycle occur in isolated nuclei. There is an absolute
requirement for uncoating of virions prior to replication in this nuclear system.
Although the viral polymerase is AMD insensitive, host protein synthesis is still
required for full activity, pointing to the involvement of host gene products in viral

replication. The accumulation of double stranded (ds) RNA followed the

15

infectivity pattern (an increase in infectivity above the initially infected
leaves) and coincided with the appearance of vesicles in the phloem (de Zoeten
8 Gaard, 1983).

Viral RNA was translated using rabbit reticulocyte lysate and wheat germ
extract (Gabriel & de Zoeten, 1984). RNA 1 supported the translation of three
major proteins: vp1 (147kDa), vp2 (88kDa) and vp4 (36kDa) while translation of
RNA 2 resulted in only one protein, vp3 (45kDa). After the genomic sequence
became available, ORFs could be assigned to the protein products. Vp1
corresponded to the fusion protein of ORF1 and 2, vp2 to ORF1 and vp4 to
ORFO. No protein product encoded by ORFs of RNA 2 with the size of vp3 was
found but additional translation experiments with deletion clones demonstrated

that the 45kDa protein corresponded to ORF1 (Demler et al., 1993).

Relationship of PEMV to other virus groups
1. Evolutionary links

Since no fossil records exist, the study of virus evolution became possible
only after the advent of comparative sequence analysis. Positive strand RNA
virus evolution is generally believed to be influenced by two trends, 1)
conservation of distinct gene combinations involved in viral replication and
expression and 2) recombination and reshuffling of conserved gene blocks
(Koonin & Dolja, 1993; Gorbalenya & Koonin, 1993). In an exhaustive review,
Koonin and Dolja (1993) analyzed the origin of all positive strand virus groups

known to that date. Since RNA dependent RNA polymerases (RdRps) contain

16

the only universal sequence motifs, “RdRp is the only domain of positive-
strand RNA viruses allowing an all-inclusive phylogenetic analysis...” (Dolja &
Koonin, 1993). Based on RdRps, three supergroups were delineated. PEMV
RNA 1 has a polymerase similar to Luteovirus subgroup ll viruses (Demler & de
Zoeten, 1991) and therefore was classified in supergroup 1 as part of the
sobemo-lineage (Dolja 8. Koonin, 1993), while the polymerase of PEMV RNA 2 is
related to viruses of the Carmo-, Necro—, Diantho—, Tombusvirus group and
Luteovirus subgroup l (Demler et al., 1993) which are all part of the canno-
lineage in supergroup 2 (Dolja & Koonin, 1993). It therefore appears that PEMV
with its genomic origin from two separate supergroups is providing a live
example for the hypothesis of gene block shuffling. The different origins of the
two PEMV RNAs is also supported by the order of genes involved in replication.
PEMV RNA 1 in addition to a viral polymerase, related to the supergroup 1
RdRp, also possesses a protease while only a polymerase, related to
supergroup 2 RdRp, could be identified as part of the RNA 2 replicase. The
latter is typical for other members of the newly proposed order of Carmovirales
(Dolja & Koonin, 1993). The putative protease of PEMV RNA 1 (Demler & de
Zoeten, 1991) is part of a 3C-like protease family, a chymotrypsin-related
protease (Gorbalenya & Koonin, 1993). Other proteases in this family belong to
the Picoma-, Como-, Nepo-, Poty-, Sobemo- and Luteoviruses subgroup ll
(Gorbalenya & Koonin, 1993). There is no direct experimental evidence for
protease activity of Sobemo-, Luteo- and Enamoviruses. However, all these

proteases contain a conserved catalytic triad which prompted the authors to

17

suggest a similar cleavage specificity for Sobemo-, Luteo- and
Enamoviruses as is found in proteases of Picoma-, Como-, Nepo-, Potyviruses
(Gorbalenya & Koonin, 1993). In all these virus groups the protease is located
upstream of the polymerase. Picorna-, Como-, Nepo-, Potyviruses also contain
a VPg and a helicase as part of their replicase unit, resulting in a domain order
of helicase-VPg-protease-polymerase. Sobemo- and Luteoviruses lack the
helicase and the location of their VPg gene is unknown. Dolja and Koonin
(1993), therefore, predicted that the VP9 domain in the latter viral groups is
upstream of the protease. Demler et al. (1996a) came to the same conclusion

when proposing the location of PEMV-VPg upstream of the protease.

2. Similarities and dissimilanties between PEMV and Luteoviruses

As mentioned above, PEMV is a virus that contains two genome parts
derived from different supergroups (Dolja & Koonin, 1993). The same
phenomenon is seen in Luteoviruses, where Luteovirus subgroup II has a
polymerase as found in RdRp of supergroup 1 and is part of the order
Sobemcvirales in the Picomavirata class while Luteovirus subgroup I has a
supergroup 2 polymerase and belongs to the order Carmovirales in the
Flavivirata class (Dolja & Koonin, 1993). Similarities between PEMV (RNA 1 in
particular) and Luteoviruses exist especially in respect to genomic organization,
cytopathology, and aphid transmission. The following paragraphs are based on
a summary by Demler et al. (1996a) and the reader is referred to that review for

individual references.

18

The general characteristics of the virus provide a good example of
both similarities and dissimilarities that provided the basis for creating a new
virus group for PEMV (Enamovirus). Luteoviruses are monopartite viruses that
have isometric particles containing a ss (+) RNA with a 3’ OH group and a 5’
VPg (e.g. Randles & Rathjen, 1995). PEMV, on the other hand, has isometric
particles each encapsidating a different ss (+) RNA with a 3’ OH and a 5’ VPg
(e.g. de Zoeten 8. Demler, 1995).

Sequencing of the viral genome of PEMV revealed strong similarities
between RNA 1 and Luteoviruses subgroup ll. Statistically significant sequence
homology was found after comparisons of the viral polymerase, the protease
domain, the cp and the op readthrough (RT). The main differences between
PEMV RNA 1 and Luteovirus subgroup II occurred in the 3’ region of the
genomic RNAs. The Luteovirus ORF 4 (17-19K) is missing from the PEMV
genome. For PLRV, the protein product of ORF 4 was shown to bind to single-
stranded nucleic acid (Tacke et al., 1991) and the authors later suggested that
this protein may act as the phloem-specific movement protein (Tacke et al.,
1993). Furthermore, the RT of PEMV is reduced by about 20 kDa and the
conserved 12 nucleotide sequence surrounding the stop codon in Luteovirsues
is not conserved in PEMV.

The differences in cytopathology can be related to different virus titers
and infected tissues of Enamo- and Luteovirus groups. PEMV accumulates to
higher concentration in the plant compared to Luteoviruses because it infects

cells of all cell types and is not limited to the phloem as is the latter. However,

19

despite a broader range of infected tissue types, PEMV preferentially
accumulates in phloem cells and sieve elements as well as in plasmodesmata
connecting these cells. PEMV and Luteoviruses induce similar double
membrane bound vesicles during the infection that contain the viral RNA.
Luteovirus induced vesicles can fuse with the nuclear envelope and particles are
sometimes seen in the nucleus. However, it was not determined whether
Luteovirus replication is associated with the nucleus. As for PEMV, virus
replication in the nucleus was unequivocally proven (see above).

Aphid transmission of both virus groups is circulative, nonpropagative, but
the dynamics of these processes differ. Luteoviruses have a longer acquisition
time and inoculation threshold and inoculation of plants only occurs during
feeding of the aphid in the phloem. Inoculation with PEMV already occurs
during probing of the aphid of non-phloem tissue. The fate of these viruses in
the aphid vector is very similar. They are acquired by their respective vector
and move through the hindgut into the hemocoel and later to the accessory
salivary glands. The site of specificity in these glands, however, differs. BYDV
aphid uptake is determined by the hindgut and by the plasmalemma of the
salivary accessory gland, where the latter is responsible for the high level of
vector specificity (Gildow 8 Rochow, 1980). In PEMV, the basal lamina, not the
plasmalemma, of the accessory salivary gland is responsible for the exclusion of
AT but not AT’ strains whereas the hindgut barrier can be crossed by both
strains. The uptake mechanism of both viruses also differs; Luteoviruses use

receptor mediated endocytosis to move across the membrane barriers of the

20

hindgut and the accessory salivary gland, while PEMV crosses the former by
simple penetration. It is not known which mechanism mediates movement of
PEMV into the salivary gland system. Luteoviruses can only be transmitted by
aphids, whereas PEMV is also mechanically transmissible. However, when
plants are mechanically inoculated with RNA 1 alone, the RNA more closely
related to Luteoviruses, no systemic infection can be established. This is viewed
as an evolutionary relict of Luteovirus ancestry of PEMV, that was overcome by

the acquisition of RNA 2.

3. What type of complex does PEMV represent?

Many different terms have been used to describe the complex of RNA 1
and RNA 2 of PEMV. Although often described as a bipartite virus because of
its two particles, PEMV is not truly bipartite. Bipartite viruses have a single
polymerase while PEMV possesses two. In addition, PEMV shows no sequence
homology at the 5' and 3’ ends of its genomic RNAs while in bipartite viruses
similar sequences at the 3’ and 5’ ends of the two RNAs are thought necessary
for the single viral polymerase to recognize and replicate the genome. It is not a
mixed infection either which would require two independent viruses. However,
RNA 1 can not systemically move in the plant nor can it be mechanically
inoculated without RNA 2 which in turn can not be encapsidated and transmitted
by aphids without RNA 1. Therefore, one could describe this relationship as

mutualistic symbiosis between two defective viruses (Demler et al., 1996a).

21

The relationship of the two RNAs of PEMV also shows some
resemblance to the helper-dependent complexes observed between some
Luteo- and Umbraviruses. Due to transencapsidation, a mechanically
transmitted virus gains aphid transmissibility when both viruses are acquired
simultaneously (Falk & Duffus, 1981 ). The Umbravirus replicates independently
and is capable of systemic infections but lacks a cp. The Luteovirus replicates
also independently but is aphid transmissible because of the ability to express a
cp. PEMV RNA 1 represents the luteoviral part and PEMV RNA 2 the umbraviral
part of the complex. However, since RNA 1 is not an independent virus and it is
not phloem limitated in the complex this relationship does not fit the accepted
definition of a helper-dependent complex. PEMV, therefore, can not be
classified in the existing categories without broadening their definitions.

Evolutionarily PEMV seems most closely related to the groundnut rosette
complex, made up of groundnut rosette assistor virus (GRAV), a subgroup ll
Luteovirus, groundnut rosette virus (GRV), an umbravirus, and a satellite virus.
The satellite depends on GRV for multiplication and both, GRV and the satellite,
depend on GRAV for aphid transmission (Demler et al., 1996a). RNA 1, related
to subgroup ll Luteoviruses, resembles GRAV while RNA 2, related to
Umbraviruses, is similar to GRV. Like the satellite of the groundnut rosette
complex, the satellite of PEMV also depends on the umbraviral part for
replication while the satellite and RNA 2 require the luteoviral part for
encapsidation and aphid transmission. The evolutionary relationship between

PEMV and the groundnut disease complex was further strengthened when

22

Demler et al. (1996b) showed sequence similarities between the satellites of

GRV and PEMV and that GRV and PEMV could replicate each others satellite.

VPg (viral protein, genome-linked)
1. General characteristics

VPgs are small proteins covalently attached to the 5’ end of many single-
stranded, positive sense, viral RNAs. To date two animal virus families, Picoma-
and Caliciviruses, and six plant virus groups, Enamo-, Luteo—, Como-, Nepo-,
Poty- and Sobemoviruses possess VPgs (for review see Vartepetian 8
Bogdanov, 1987). This would be expected since all of the above mentioned
virus groups belong to the Picomavirata with Luteovirus and PEMV representing
a link to the Flavivirata (Dolja 8 Koonin, 1993). It was reasoned that protein-
primed RNA replication mechanisms (see below for possible VPg functions) only
arose once, early in evolution (Dolja 8 Koonin, 1993). As mentioned earlier the
domain localization of VPg, upstream of a 3C-like protease, is identical for
Picoma-, Como-, Nepo- and Potyviruses (Dolja 8 Koonin, 1993; Gorbalenya 8
Koonin, 1993). Despite sequence similarities within a group, no common
sequence elements were found, hampering efforts to predict the VP9 localization
of Schema- and Luteoviruses (Gorbalenya 8 Koonin, 1993).

VPgs often run anomalously on SDS-PAGE gels due to their cationic
character (Vartepetian 8 Bogdanov, 1987). The electrophoretically derived
molecular weights, therefore, frequently predict much larger proteins than

protein sequencing reveals. The sizes of the VPgs range primarily between 20

23

and 30 amino acids (aa) (e.g. Wimmer, 1982; Mayo 8 Fritsch, 1994; Jaegle
et al., 1987; Drygin et al., 1987). VPgs are not only found on viral genomes but
also on at least two satellite RNA (Meyer et al., 1984; Koenig 8 Fritsch, 1982;
Liu et al., 1991). The aa involved in the covalent linkage of the protein with the
RNA can be either tyrosine, serine or threonine due to the free hydroxyl group in
their sidechains. Tyrosine is involved in Picomaviruses (for reviews see
Wimmer, 1982; Vartepetian 8 Bogdanov, 1987) and Potyviruses (Murphy et al.,
1991; Jayaram et al., 1992; Vance at al., 1992) while serine has been implicated
for viruses belonging to Como- (Jaegle et al., 1987; Drygin et al., 1987) and
Nepoviruses (Pinck et al., 1991; Margis et al., 1993; Zalluoa et al., 1996).
Threonine, although theoretically possible, has not yet been found as an aa
involved in RNA linkage. The function of VPgs remains obscure despite
intensive research especially with poliovirus. Possible functions include: primer
of RNA synthesis, signal for virion encapsidation, and in some instances an
involvement in infectivity (e.g. Vartepetian 8 Bogdanov, 1987). The VPg is
needed for infectivity of Sobemo—, Nepo-, Poty- and Caliciviruses but removal of
the protein from the genomes of Picoma-, Como-, Enamo-, Luteoviruses has no
effect on infectivity (for review see Vartepetian 8 Bogdanov, 1987). Since most
research on the function of VPg has involved poliovirus, I will provide an
overview of poliovirus replication and the involvement of its VPg in the viral life

cycle before summarizing data obtained with the VPgs of plant viruses.

2. Poliovirus VPg

24

Poliovirus VPg is a basic, acid-soluble, 22 aa protein (for review see
WImmer, 1982). The basic amino acids are located in the center and C-terminal
portion of the protein (Vartepetian 8 Bogdanov, 1987). Tyrosine at position 3
from the N-terminus is linked via its O‘-hydroxyl to the terminal phosphate of
uracil, the most 5’ terminal nucleotide of the RNA (for review see Wimmer,
1 982).

The presence and absence of VPg influences several stages of the
poliovirus life cycle. Poliovirus VPg is not needed during very early stages (e.g.
entry and uncoating) since removal of the protein has no influence on infectivity
of the RNA (Flanegan et al., 1977). The protein is found on (+) and H ss RNA,
on replicative forms and replicative intermediates as well as on nascent strands
suggesting a role in viral replication (Wimmer, 1982). It is not necessary for
translation and a 27kDa unlinking enzyme of the host is involved in the removal
of VPg from RNA to become mRNA (Vartepetian 8 Bogdanov, 1987). However,
it is again needed during virion formation. Only RNA linked to a VPg is found in
virions indicating that it might serve as a signal determining encapsidation of
RNA (Nomoto et al., 1977a).

Two different approaches have been used to elucidate the mechanisms
underlying poliovirus replication. The first one uses isolated crude membranous
RNA replication complexes from virus infected HeLa cells. This is mainly used
to study the synthesis of (+) strand RNA from (-) strands (Ehrenfeld 8 Richards,
1989). A second approach uses in vitro studies to elucidate primarily (-) strand

from (+) strand RNA synthesis. Components involved in poliovirus replication

25

have been isolated and purified and individual steps were reconstituted
(Ehrenfeld 8 Richards, 1989). The ultimate goal to construct a complete in vitro
replication system has not been achieved and caution should be exercised when
analyzing in vitro results (Hey et al., 1987).

Different poliovirus replication models have been proposed but each

takes the strict template and primer dependence of the viral polymerase 30"
into account (F lanegan 8 Baltimore, 1979; van Dyke 8 Flanegan, 1980). Also,
the formation of membranous vesicles and their involvement in viral replication is
well documented (for review see Mirzayan 8 Wimmer, 1994). One model states
that positive strand RNA synthesis occurs in a membranous environment where
uridylylated VPg serves as primer for RNA elongation by the viral polymerase
3D"°' (for review see Mirzayan 8 Wimmer, 1994). Evidence showing the
involvement of VPg-pU as a primer is: l) VPg is attached to the 5’ end of
genomic, (+), H and nascent RNA but not to mRNA (Nomoto et al., 1977a;
Pettersson et al., 1978; Flanegan et al., 1977). ll) Antibodies (Abs) against VPg
(anti-VPg Abs) immunoprecipitate VPg and VPg-pUpU (Crawford 8 Baltimore,
1983). III) Anti-VPg Abs specifically inhibit initiation of viral RNA synthesis in
vitro (Baron 8 Baltimore, 1982). Data indicating the importance of membranes
during replication led to the discovery of a VPg (3B) precursor protein 3A8, a
12kDa protein with a hydrophobic stretch of 22 aa in the 3A portion. This
hydrophobic stretch is involved in attaching 3A8 to membranes of the replication
complex and delivering VPg to its site of function. The viral proteinase 3Cpro

cleaves 3A8 during or after (+) RNA elongation. This releases the VP9 and the

26

associated viral RNA from the membrane (for review see Mirzayan 8
Wimmer, 1994).

Negative strand synthesis involves the uridylylation at the 3’ poly A of the
(+) RNA by host factor (HF), resulting in the formation of a polyA - polyU hairpin
that serves as a primer. The attachment of VPg to the H strand takes place by
a self-catalyzed transesterification reaction that concomitantly cleaves the
hairpin and separates both strands (Mirzayan 8 Wimmer, 1994). Two main
observations support this model. One, negative strand covalently linked to (+)
strand RNA was seen after in vitro synthesis in the presence of viral polymerase
and HF (Young et al., 1986; Young et al., 1985). Two, HF has a terminal
uridylyltransferase activity that stimulates initiation of RNA synthesis (Andrews 8
Baltimore, 1986; Andrews et al., 1985).

The evidence obtained for poliovirus replication led people to postulate
the following two functions of VPg. First, VPg or uridylylated VPg serve as
primer for (+) strand RNA synthesis. Second, VPg has a nucleolytic activity that
cleaves the RNA hairpin during (-) strand synthesis and a linkage activity to
attach itself to the 5’ end of the (-) RNA. The strongest evidence against the
essential role of VPg during RNA replication comes from mutational experiments
in which a VPg mutation was replication positive but nonviable. This suggests
that an essential function other than viral RNA replication was affected. As
indicated earlier, the VP9 could also serve as a signal for encapsidation
although no direct experimental evidence for this is available. None of these

functions could be positively verified and other functions are possible as well.

27

3. VPgs of plant viruses

The plant viruses closely related to poliovirus and the Picornaviruses are
Como-, Nepo- and Potyviruses which are combined in the superfamily of
picoma-Iike plant viruses (Matthews, 1991). A comparison of genome
organization and nucleotide sequence led to the following common features of a
(+) ssRNA genome, a 5’ VPg and a 3’ poly (A) tail. The viral expression strategy
is via a genome containing a long single ORF that codes for a polyprotein which
is processed by viral proteases. Several nonstructural proteins of similar
function also share significant sequence similarity and these conserved proteins
are arranged in the order of membrane binding region - VPg - protease - RdRp
(for review see Matthews, 1991). The following paragraphs will provide a
discussion on the knowledge of VPgs from each group of the picoma-like
viruses.

VPgs have been identified on the Comoviruses CPMV, cowpea severe
mosaic virus (CPSMV) and red clover mottle virus (RCMV) (for review see Mayo
8 Fritsch, 1994), radish mosaic virus (RaMV) (Drygin et al., 1987), squash
mosaic virus (SqMV) and Echtes Ackerbohnenmosaik Virus (EAMV) (Daubert 8
Bruening, 1979). Protein sequencing of VPgs from CPMV, CPSMV and RCMV
revealed basic, hydrophilic proteins of 28 aa with a relatively similar sequence
(Mayo 8 Fritsch, 1994). The linkage to the viral RNA in the case of the 28 aa
CPMV-VPg and the 36 aa RaMV-VPg was a B-OH (serine) linked via a

phosphodiester bond to the 5’ terminal phosphate of the last nucleotide of the

28

viral RNA (Jaegle et al., 1987; Drygin et al., 1987, respectively). CPMV, the
type member of Comoviruses, was the first plant virus for which the VP9 was
identified (Daubert et al., 1978). However, data obtained from poliovirus VPg
nevertheless guide the experimental analysis of all VPg functions.

CPMV-VPg has a positive charge of +5. If VPg functions as a primer
during CPMV RNA replication as was proposed for poliovirus, the positive
charge could enable the protein to interact with the negatively charged viral RNA
(Jaegle et al., 1987). Not all 5’ ends of RF are bound to VPg (Lomonossoff et
al., 1985). It was hypothesized that if the VP9 is initially present and later
removed, those (+) strands could serve as mRNA, in analogy to poliovirus where
the mRNA lacks a VPg. Also, like Picornaviruses, the VP9 was not required for
infectivity or in vitro translation (Stanley et al., 1978). Early research with anti-
VPg Abs indicated that CPMV-VPg was part of a membrane bound 60kDa
precursor molecule (label at al., 1982; Goldbach et al., 1982; Wellink et al.,
1986) that was responsible for bringing the VP9 to the replication complex.

More recent evidence indicates a 112kDa VPg precursor protein is present
during the initiation of replication (Peters et al., 1995). It is unknown whether the
precursor is directly involved, or whether involvement is via a 26kDa protein
(VPg + 24kDa protease) derived from this precursor. Free VPg is rapidly
degraded or undetectable (de Varennes et al., 1986; Peters et al., 1995; Peters
et al., 1992) but viral RNA of as few as 17 nucleotides linked to VPg provided

protection from proteolytic degradation (de Varennes et al., 1986).

29

Nepoviruses are closely related to Comoviruses and are classified
within the picoma-like plant virus superfamily as Comoviridae. The Comoviridae
have isometric particles with a bipartite genome. Like Comoviruses, the VP9 is
encoded by the larger RNA (RNA 1) (Mayo et al., 1982a), and it plays no role in
translation (Koenig 8 Fritsch, 1982; Chu et al., 1981), and covalent linkage of.
the VP9 occurs via the serine [I-OH to the 5’ terminal phosphate of the viral RNA
(Pinck et al., 1991; Margis et al., 1993; Zalluoa et al., 1996). However, unlike
Comoviruses, infectivity is greatly reduced or eliminated by protease treatment
(Mayo et al., 1982a; Harrison 8 Barker, 1978; Chu et al., 1981). The size of six
different nepoviral VPgs is around 4kDa on SDS-PAGE (Mayo et al., 1982a).
Microsequencing revealed a 24aa VPg for grapevine fanleaf virus (GFLV) (Pinck
et al., 1991) and 27aa VPg in case of tomato black ring virus (TBRV) (Mayo 8
Fritsch, 1994). A 26 aa VPg sequence for tomato ringspot virus (TomRSV)
(Mayo 8 Fritsch, 1994) and the 29 aa sequence of grapevine chrome mosaic
virus (GCMV) were inferred based on strong sequence similarities between the
RNA 1 polyproteins. Despite this, nepoviral VPgs show relative sequence
diversity and only the insertion of gaps led to a consensus (Mayo 8 F ritsch
1994). Mutations in the 5’ end of satellite RNA from cDNA clones of arabis
mosaic virus (ArMV) were corrected after infection of plants providing indirect
evidence that VPg - 5' terminal RNA bases might serve as a primer in (+) strand
RNA synthesis (Liu et al., 1991). A previously unknown possible VPg function
emerged from translation studies of the GFLV polyprotein. Cleavage by the viral

encoded 24kDa protease (Pro) is two times more efficient after cleavage of the

30

VPg-Pro polyprotein and release of the VP9. It was therefore postulated
that the VP9 could modulate the cleavage activity of the 24kDa Pro and
ultimately regulate protein expression (Margis et al., 1994).

Potyviruses like Como- and Nepoviruses belong to the same superfamily
but unlike Como— and Nepoviruses they are filamentous with a monopartite ss
(+) RNA genome (Matthews, 1991). The VPg has been identified and localized
to the N-terminus of the small nuclear inclusion protein Nla for tobacco etch
virus (TEV) (Dougherty 8 Parks, 1991), tobacco vein mottling virus (TVMV)
(Shahabuddin et al., 1988) and turnip mosaic virus (TuMV) (Laliberte et al.,
1992). The C-terminus of this protein is a serine-like 3C protease that can utilize
a suboptimal internal cleavage site of Nla to release the protease from the VP9
(Laliberte et al., 1992; Carrington et al., 1993). The VPg is linked via a
phosphodiester bond of a tyrosine hydroxyl group to the viral RNA (Murphy et
al., 1991; Jayaram et al., 1992; Vance at al., 1992) and mutations of the
conserved tyrosine results in a lethal phenotype (Murphy et al., 1996). Like
Nepoviruses the infectivity of the RNA decreases with the removal of the protein
(Riechmann et al., 1989). The size of the VP9 is between 22kDa and 24kDa on
SDS-PAGE (Riechmann et al., 1989; Laliberte et al., 1992; Murphy et al., 1990;
Shahabuddin et al., 1988). A 21 kDa (189 aa) VPg for tobacco etch virus (TEV)
was deduced from computer sequence analysis (Daugherty 8 Parks, 1991).
Potyvirus VPgs are much larger than any other identified VPgs. This might be
due to the multifunctional activities of the Nla precursor protein. In addition to

the VP9, a nuclear localization signal has been identified within the N-terminus

31

of Nla (Carrington et al., 1993; Schaad et al., 1996). Both the 49kDa Nla
proteinase and a 24kDa protein, the N-terminal part of the proteinase, were
found attached to the viral RNA (Murphy et al., 1990) indicating that cleavage is
not required for VPg activity (Carrington et al., 1993). However, attachment of
either protein to the RNA might not be sufficient for VPg function as it is believed
that the functional VPg form includes an additional 6kDa protein that serves as a
membrane anchor like the 3A protein of poliovirus (Restrepo-Hartwig et al.,
1 994).

Sobemoviruses are related to Luteoviruses (Koonin, 1991) and both
groups belong with the picoma-like plant viruses to the Picornavirata (Dolja 8
Koonin, 1993). Members of the two groups have isometric particles with a
monopartite, ss (+) RNA genome that contains a 5’ VPg but is not
polyadenylated at the 3’ end (Matthews, 1991). Based on electrophoretic
mobility, the type member of the Sobemoviruses, southern bean mosaic virus
(SBMV), has a 12kDa VPg (Ghosh et al., 1979). Sequencing of the genomes of
rice yellow mottle virus (RYMV) (Yassi et al., 1994) and cocksfoot mottle virus
(CfMV) (Makinen et al., 1995), led to the discovery of a conserved VPg motif that
is located upstream of the protease and polymerase in the second ORF. The aa
involved in the protein-RNA linkage is unknown but Yassi et al. (1994) proposed
the involvement of a conserved tyrosine residue in analogy with poliovirus data.

Luteoviruses are divided into two subgroups which are distinguished
based on the genomic organization and sequence similarity (e.g. Martin, 1990).

They contain viruses related not only to the Picornavirata (subgroup II) but also

32

to the Flavivirata (subgroup l) (Dolja 8 Koonin, 1993). VPgs have been
found at the 5’ end of RNAs of subgroup ll viruses. The RPV isolate of barley
yellow dwarf (BYDV) and potato leafroll virus (PLRV) have a 17kDa and 7kDa
VPg, respectively, attached to the 5' end of their genome (Murphy et al., 1989;
Mayo et al., 1982b, respectively). No VPg has been found attached to the RNA
of a subgroup l Luteovirus. This is not unexpected since no virus of the
Flavivirata class contains this protein. Like poliovirus, Como- and Potyviruses,
the protein is not required for infectivity of the RNA (Mayo et al., 1982b). No
protein sequence has been determined and the aa involved in the VP9 linkage
to the RNA as well as the precise genomic localization of this protein are
unknown. However, hypotheses regarding the genomic localization of the VP9
have been proposed. The first complete aa sequence of a Luteovirus was
determined for subgroup l BYDV-PAV (Miller et al., 1988). Since ORF 4 has the
potential to encode a 17kDa protein and the M, of BYDV-RPV VPg was of the
same size, Miller (1994) proposed that the VP9 is encoded in ORF 4. A 17K
ORF 4 was also found after sequencing of PLRV. The authors suggested
proteolytic cleavage was the most logical explanation to derive the 7kDa PLRV-
VPg (van der Wilk et al., 1989). However, more recent evidence points to ORF
1 as the VP9 coding region (Miller et al., 1995, Mayo 8 Ziegler-Graft, 1996).
Frameshift mutations of BWYV-ORFs were introduced and RNA transcripts used
to infect protoplasts. Only mutations in ORF 1 and 2 are lethal (Reutenauer et
al., 1993). The essential nature of the two ORFs points to their involvement in

RNA replication, a function also proposed for VPg. Also, subgroup ll

33

Luteoviruses show a protease motif in ORF 1 and a polymerase motif in
ORF 2 and are part of the picomavirus-like superfamily which possesses a
replicase with the domain order of VPg-protease-polymerase (Dolja 8

Carrington, 1992; Koonin 8 Dolja, 1993) placing the VP9 in ORF 1.

4. Rationale for project

As mentioned earlier, Enamoviruses contain a divided genome with RNA
1 related to subgroup ll Luteoviruses and Picornavirata and RNA 2 to the
Carmovirales in the Flavivirata (Dolja 8 Koonin, 1993). PEMV-VPg was
identified on SDS-PAGE gel by iodination as a 17.5kDa protein which was
covalently attached to the 5’ end of the viral genome (Reisman 8 de Zoeten,
1982). It is unknown whether the protein is attached to both RNAs. The actual
size of the protein is most likely smaller as other VPgs have turned out to be
smaller than the size predicted by SDS-PAGE (for review see Vartepetian 8
Bogdanov, 1987). No ORF of 17K/ 18K or smaller was identified within the
PEMV genomic sequence (Demler 8 de Zoeten, 1991). The protein is not
required for infectivity since in vitro RNA transcripts were successfully used for
mechanical inoculations (Demler et al., 1996a). It was speculated that the VP9 is
localized upstream of the protease motif of RNA 1 as part of the conserved
domain arrangement of membrane binding - VPg - protease - polymerase in
analogy with Picoma- and picorna-Iike viruses (Demler et al., 1996a). However,
the actual size and the genomic localization of the VP9 and the aa involved in

the protein-RNA linkage is not known.

34

The goal of my project was to isolate and characterize the PEMV-VPg
and to obtain protein sequence to determine the genomic localization of the VP9

gene.

Chapter 2

Genomic localization of the 5’-linked protein (VPg)

of Pea Enation Mosaic Ensmovims

Introduction

Pea enation mosaic virus (PEMV) is the sole member of the Enamovirus
genus (de Zoeten 8 Demler, 1995). Its genome consists of two taxonomically
distinct, single-stranded, (+) sense RNAs. RNA 1 (5706 nucleotides) shares
significant homologies with subgroup ll Luteoviruses in the arrangement and
sequences of RNA polymerase, putative protease, coat protein (cp), and coat
protein readthrough domain (Demler 8 de Zoeten, 1991; Demler et al., 1996a).
RNA 2 (4253 nucleotides) encodes a polymerase, distinct from the polymerase
of RNA 1, that reveals a closer relationship to members of the Luteovirus
subgroup l as well as the Umbra-, Carmo-, Tombus-, Diantho- and Necroviruses
(Demler et al., 1993). Despite the independent replication of RNA 1 and RNA 2,
both are required for wild-type infections. RNA 1 provides for encapsidation and
vector transmission while RNA 2 facilitates systemic invasion and mechanical
transmission (Demler et al., 1994).

Many single-stranded viral RNAs, including PEMV (Reisman 8 de Zoeten,
1982), encode, and are covalently linked to a small protein (VPg) at the 5’

terminus (for reviews see Wimmer, 1982; Vartapetian 8 Bogdanov, 1987 and

35

36

references therein). The function of VPgs remains under investigation but
experimental data from poliovirus has led to several hypothesis. Poliovirus VPg
was proposed to act as a primer for RNA synthesis (e.g. Nomoto et al., 1977a;
Mirzayan 8 Wimmer, 1994). Poliovirus RNA polymerase is strictly primer-
dependent (Van Dyke 8 Flanegan, 1980) and antibodies (Abs) raised against
the poliovirus VPg inhibit RNA replication (Baron 8 Baltimore, 1982; Barton et
al., 1995). In addition, poliovirus VPg may also play a role in virus encapsidation
since only poliovirus-RNA that is covalently linked to the VP9 is packaged
(Nomoto et al., 1977a; Nomoto et al., 1977b).

The nucleotide sequence of the PEMV genome (Demler 8 de Zoeten,

1991) did not reveal a discrete open reading frame (ORF) corresponding to the
17.5 kDa PEMV-VPg reported by Reisman 8 de Zoeten (1982). However, the
predicted amino acid sequence of the 84K protein of RNA 1 ORF 1 revealed a
hydrophobic region near the N-terminus of the protein that is characteristic of
membrane spanning proteins. The central region of this 84K protein is similar
to the catalytic site of a serine-based member of the 3C-like proteinases
although no proteinase activity had been reported for PEMV or any Luteovirus.
The N-terrninus and the central part are separated by a very basic region, a
characteristic common to VPgs (Demler 8 de Zoeten, 1991). This linear
arrangement of transmembrane region - proteinase - polymerase domains in
PEMV was similar to Picoma-like viruses (membrane binding domain - VPg -
proteinase - polymerase) (Demler 8 de Zoeten, 1991; Domier et al., 1987,

respectively) which utilize proteinases in protein processing. Based on this

37

analogy, Demler et al. (1996a) suggested that the PEMV-VPg gene is
encoded upstream of the protease motif.

Several members of the Luteovirus subgroup ll possess a VPg (Mayo et
al., 1982; Murphy et al., 1989). However, attempts to determine ORFs encoding
these VPgs led to conflicting and inconclusive results. Miller et al. (1988) and
Miller (1994) proposed that the VP9 may be encoded by the 17K ORF 4, since a
17 kDa VPg was reported for the RPV isolate of barley yellow dwarf virus
(BYDV-RPV) (Murphy et al., 1989). Similarly, van der WIlk et al. (1989) inferred
that the 7kDa VPg of potato leafroll virus (PLRV) is produced via proteolytic
processing of the PLRV—ORF 4 product. Experimental proof for both
conclusions is lacking and subsequent data prompted researchers to propose
the involvement of ORF 1 (Mayo 8 Ziegler-Graft, 1996; Miller et al., 1995) in the
production of VPg. The similarities of PEMV RNA 1 to subgroup ll Luteovirus
RNAs and the relatively high yield of purified virus from peas make PEMV an
attractive system to address questions surrounding the genomic origin of the
VP9.

In this study we report for the first time the amino acid (aa) composition of
PEMV VPg. Subsequent computer sequence alignments mapped the VP9 to the
84K ORF of RNA 1, downstream of the protease motif. This provides the first

indication for proteolytic processing as part of the PEMV replication process.

Materials and Methods

38

Virus propagation and purification:

The naturally occurring, aphid-nontransmissible “deletion" strain PEMV-
AT-A (Demler et al., 1996a) was propagated in Pisum sativum L. cv. 8221 by
mechanical inoculation of the lower two sets of leaves of 9 to 11 day old
seedlings. Pea plants (Pisum sativum) were grown from seeds in a growth
chamber with a 16 hour light regime, daytime temperature of 22°C and night
temperature of 18°C. PEMV infected fresh leaves (29) or dried leaves (0.59)
were ground in 10ml of 0.1M phosphate buffer pH 6.5 with celite (diatomaceous
earth) in a chilled mortar and pestle and used for mechanical inoculation.

Infected plants were harvested ten to twelve days post-inoculation and
PEMV virions were purified according to a protocol from German 8 de Zoeten
(1975). All extraction equipment was placed at 4°C prior to the virus isolation
and solutions were made from a stock solution of 2M sodium acetate (NaOAc)
buffer, pH 6.0 (total volume of 2 liters: 230 ml of glacial acetic acid [17.4M] were
added to 1600 ml of deo in a beaker on ice, 1309 of NaOH were slowly added
under constant stirring and the pH was adjusted with 10N NaOH). Plants were
harvested by cutting them at the soil level with a razor blade and the tissue
weight was determined. A volume of extraction buffer (0.2 M NaOAc, pH 6.0)
equal to 1.5 times the weight of the tissue was added to the harvested tissue in a
commercially available Waring blender and well homogenized. One volume of
chloroform, based on the weight of the tissue, was added and the tissue was
homogenized for one minute at high speed. The homogenate was poured into

250ml centrifuge bottles(Beckman Instruments Inc., Palo Alto, CA), placed in a

39

JA 14 rotor (Beckman) and centrifuged at 4°C for 20 minutes at 7000rpm
(7K) in a Beckman J2-21 centrifuge. The upper layer was then decanted into a
1l flask on ice and dialyzed overnight in 10 l of 0.2M NaOAc in the cold room
(4°C). On the following day, the solution was centrifuged in 250ml centrifuge
bottles in a JA-14 rotor at 4°C and 7K for 20 minutes. The supernatant was then
centrifuged in a Beckman L-70 ultracentrifuge in 26.3ml polycarbonate tubes
with 3 piece cap assembly (Beckman, No. 355616) in a 50.2Ti rotor (Beckman)
at 4°C and 30K for 3 hours. The resulting supernatant was decanted and
discarded. Two milliliters of 5% sucrose in 0.1M NaOAc buffer were added to
the pellet to resuspend it overnight on a shaker (Gyrotory shaker model 62, New
Brunswick Scientific Co., Inc., Edison, NJ) in the cold room. The resuspended
pellets were combined the next day in 30ml glass Corex tubes (Corning
Incorporated, Corning, NY) and spun in a JA-20 rotor (Beckman) at 4°C and 7K
for 20 minutes in a Beckman J2-21 centrifuge. The high speed centrifugation
and the resuspension of the pellets were repeated. The virus concentration was
calculated after determining the absorbency of a 1:100 dilution of an virus
aliquot with a Beckman DU-64 spectrophotometer at 260nm, using an extinction
coefficient of 7.5mg/ml. Linear sucrose gradients of 10% to 40% were prepared
from stock solutions of 10% and 40% sucrose in 0.1M NaOAc buffer in Ultra—
clear centrifuge tubes (25x89mm, Beckman) and 2-20 mg of virus were loaded.
The gradients were spun in a SW 28 rotor (Beckman) at 4°C and 25K for 3 hours
in a Beckman L-70 ultracentrifuge and fractionated afterwards using an ISCO

density gradient fractionator, model 185, (ISCO lnc., Lincoln Nebraska) and the

40

ISCO UA-5 absorbance/fluorescence detector. PEMV peaks were collected,
combined and centrifuged overnight in a 50.2Ti rotor (Beckman) at 4°C and 50K
rpm. The resulting pellets were resuspended in diethyl pyrocarbonate (DEPC) -

treated 20mM Tris-HCI, pH 7.2 and used for viral RNA isolation.

Viral RNA isolation and anti-VPg antibody production:

Viral RNA was purified according to a protocol from Reisman 8 de Zoeten
(1982). Special precautions were taken to prevent contamination with RNases.
All steps were carried out in 15ml Corex tubes (Corning) that were autoclaved
and baked for 8 hours at 180°C in an oven (lsotemp oven, model 6556, Fisher
Scientific, Pittsburgh, PA). The solution containing the resuspended pellet was
brought to 2% SDS, 20mM EDTA, mixed and placed at 68°C for 5 min. Two
extractions with an equal volume of 10mM Tris-HCI pH 7.2 saturated phenol
each followed. Each time the samples were vortexed vigorously and centrifuged
for 10 min at 7K rpm in a JA—20 rotor (Beckman). To increase RNA yield, the
phenol phase of the first extraction was back-extracted with 10mM Tris-HCI pH
7.2. The aqueous phase was collected from each extraction, pooled, extracted
twice with chloroform - isoamyl alcohol (24:1) and later brought to 0.2M NaOAC.
To each sample, 2.5 volumes of -20°C, concentrated ethanol were added and
the RNA was precipitated at -80°C overnight. The precipitate was pelleted in a
JA-20 rotor (Beckman) at 4°C and 7K for 30 min. The RNA was washed by
adding 2ml of chilled, 70% ethanol and centrifuging for 10 min. The pellet was

resuspended in 500m of DEPC-treated deo. Three additional phenol-

41

chloroform extractions followed by ethanol precipitation were performed to
remove noncovalently attached proteins. This extensivly purified RNA was then
used for antibody production.

To obtain anti-VPg antibodies (Abs), approximately 250ug of precipitated,
extensivly purified RNA was resuspended in 400m of DEPC - treated deO and
used for subcutaneous injections of a rabbit. The RNA was emulsified in an
equal volume of Freunds complete adjuvant (Sigma Chemical Company, St.
Louis, M0) for the first injection and Freunds incomplete adjuvant (Sigma) for
booster injections. The care of the rabbit, administration of injections and
drawing of blood was performed by the personnel of the University Laboratory
Animal Resources (ULAR) at Michigan State University (MSU), East Lansing,
MI. Booster injections were administered 28, 46 and 56 days after the first
injection. Immune serum was collected before each booster injection and Abs
were concentrated following standard protocols (Harlow 8 Lane, 1988). The
obtained blood was placed at 37°C for 1 hour prior to an overnight incubation at
4°C and the serum was separated from the clotted blood the next day by
centrifugation in a table top centrifuge (lEC HN-Sll centrifuge, lntemational
Equipment company, Needham, MA) for 10 min at 1800 x g. The serum was
removed and placed into eppendorf tubes in 250 pl aliquots and stored at -20°C.
The Abs were tested in Western blots of total protein extracts from infected
plants to determine the specificity and strength of the Abs. Proteins were
separated on 15% sodium dodecyl sulfate - polyacrylamide gels (SDS-PAGE)

and electroblotted onto lmobilon-P membranes (Millipore, Maryland) prior to the

42

Western blotting protocol (see below). Only anti-VPg Abs collected 56 days

after the first injection were used in the described experiments.

Production of RNA transcripts and transcript inoculation of plants:

Pea plants used for total protein isolation and characterization of the
putative anti-VPg Abs were infected with transcripts of RNA 1 and RNA 2, RNA 2
alone or mock inoculated. Transcripts of RNA 1 and RNA 2 were prepared from
the cDNA clones PEMV-AT’ (Demler et al., 1996a) and pPER2d (Demler et al.,
1994a), respectively. Ten ug DNA of a clone was linearized for each
transcription reaction, with Pst l (Boehringer Mannheim, Indianapolis, IN) for
RNA 1 or Sma l (Boehringer Mannheim) for RNA 2, during a 60 min incubation
at 37°C. Care was taken during the reactions to prevent contamination with
RNases. For each transcription reaction 165 pl of deo, 100 pl of 5x optimized
transcription buffer (Promega Corporation, Madison, WI), 50 pl of 0.1 M
dithiothreitol (DTT), 50ul of nucleotide triphosphates (4mM ATP, CTP, UTP,
each and 0.75mM GTP), 125 pl 2mM 7-methylguanosine (CAP) and 1ul (80ng)
of T7 polymerase (Promega) were added to linearized, phenol-chloroform
extracted and ethanol precipitated DNA. Following a 45 minute incubation at
37°C, 4 ul of 10mM GTP and 1 pl (80ng) of 17 polymerase (Promega) were
added and the incubation continued for another 45 min. The DNA template was
destroyed by addition of 5 units (U) R01 DNase (Promega) during a 15 min
incubation at 37°C. The RNA was then extracted twice with phenol-chloroform

and precipitated at -20°C for at least one hour with 50 ul of 3M NaOAc and 1ml

43

chilled ethanol. The precipitated RNA was pelleted at 14000 (14K) rpm in
an eppendorf centrifuge 5415C (Brinkmann Instruments, Inc, Westbury, NY) for
20 min, washed with chilled, 70% ethanol and afterwards resuspended in 50 pl
DEPC-treated deo. The size and integrity of RNA transcripts was assessed on
non-denaturing 1% agarose gels (for detailed protocol see appendixA).

Seven day old pea seedlings were inoculated with various combinations
of infectious transcripts or mock inoculated as previously described (Demler et
al., 1993). Each inoculum contained 5 to 10 pg of viral RNA transcripts
suspended in an equal volume of buffer (10mM Tris-HCI, pH 8.0, 1mM EDTA,
1% wlv celite, 1% wlv bentonite). The seedlings were dusted with 600 mesh
carborundum (Fisher Scientific) prior to rub-inoculation. Plants were harvested

10 to 14 days post-inoculation for total protein isolation.

Total protein isolation:

Five leaves from individual pea plants were ground in 750pl of
homogenization buffer (700mM sucrose, 500mM Tris-HCI pH8.0, 50mM EDTA,
100mM KCI, 2% B-mercaptoethanol, 2mM phenylmethylsulfonyl fluoride [PMSF]
in isopropanol). The homogenate was centrifuged in an eppendorf centrifuge
(Brinkman Instruments, Inc.) for 5 minutes at 14K rpm. The supernatant was
collected and the protein concentration was determined using the Bio-Rad
Protein Assay Kit I (Bio-Rad Laboratories, Hercules, CA). Based on the
protocol of the manufacture, dilutions of 0 - 20 pg of bovine gamma globulin

(BGG) in 0.8ml of dHZO were prepared to create a standard curve. Diluted

44

samples were prepared by adding 1pl and 10pl of each sample to dHZO to
obtain a final volume of 0.8ml. Dye reagent concentrate (0.2ml) (Bio-Rad) was
added to all samples and standards and solutions were mixed by gentle
inversion of each test tube. The OD“ was measured after 5 min in a
spectrophotometer (Beckman). A standard curve was created by plotting
absorbency versus concentration of standards. The concentration of each

sample was then determined from this curve.

Purification and biotinylation of anti VPg-Ig Gs:

In order to increase the detection limit, anti-VPg lgGs were isolated and
biotinylated. Anti-VPg lgGs were purified using protein A low salt columns
(Harlow 8 Lane, 1988). A column was made by filling 200mg protein A beads
(Pharmacia Biotech Inc., Piscataway, NJ) dissolved in 100mM Tris-HCI pH 8.0
into a disposable chromatography column (BioRad) and washing the column
extensively with the 100mM Tris-buffer. The pH of the serum was adjusted to
8.0 by adding 1/10‘” volume of 1.0M Tris-HCI pH 8.0. After application of 1ml
serum, pH 8.0, to the column, washes with 10 column volumes of 100mM Tris-
HCI, pH 8.0, and with 10 column volumes of 10mM Tris-HCI, pH 8.0, followed.
lgGs were eluted from the column by applying 100mM glycine pH 3.0 in 500pl
aliquots. The eluate was collected in eppendorf tubes containing 50 pl of 1M
Tris-HCI pH 8.0 to adjust the pH to neutral. The lgG containing fractions were
identified by absorbance at 280nm (1 OD=approximately 0.8 mg/ml) in a

spectrophotometer (Beckman).

45

Anti-VPg lgGs were biotinylated in solution using the Amersham
LifeScience ECL-protein biotinylation module. The provided 20x bicarbonate
buffer, pH 8.6, was diluted 1:20 by adding 2.5pl of buffer to 50pl (55pg) of anti-
VPg 1965. After addition of 3pl biotinylation reagent the solution was mixed well
and incubated for one hour at room temperature (RT) under constant agitation
(eppendorf mixer 5432, Brinkmann Instruments, Inc.).

Unbound biotin was removed first by incubating the reaction mixture with
1l10°‘ volume of 1M Tris, pH 8.0, at room temperature for 15 min, followed by
purification of labeled Abs on a Sephadex G-25 spin column. The Sephadex G-
25 resin (medium grade, Sigma) was equilibrated overnight at 4°C in an excess
volume of phosphate-buffered saline pH 7.5 (PBS; 80mM di-sodium hydrogen
orthophosphate anhydrous, 20mM sodium dihydrogen orthophosphate, 100mM
sodium chloride) containing 0.01% Thimerosal (mercury-[(o-carboxyphenyl)thio]-
ethyl sodium salt). For each biotinylation reaction, a 1ml plastic syringe was
stoppered with glass wool and filled with preswelled resin. Each column was
placed into a 15ml orange cap tube (Corning Incorporated, Corning, NY) and
centrifuged upright in a IEC HN-Sll centrifuge. Centrifugations were performed
by going up to % speed and down again. The eluate was discarded, the column
refilled with preswelled resin and respun until the resin was within 1cm of the top
of the syringe. The column was calibrated immediately prior to use by adding a
volume of PBS, identical to the reaction volume, to the top of the column and
centrifuging at % speed for 2 min. The steps were repeated until the volume of

the eluate equaled the volume added. After calibration of the column, labeled

46

anti-VPg lgGs were added. Again, the centrifugation was performed at ’/4
speed for 2 min and the purified, biotinylated protein eluate was transferred to
an eppendorf tube.
The biotinylated anti-VPg lgGs (primary antibody preparation) were
diluted 1:600 (vlv) in PBS + 0.1% Tween 20, pH 7.5 (PBS-T) containing 2%
dried milk and used in western blot analysis as described below (ECL Western

blotting detection, Amersham LifeScience).

Gel electrophoresis and western blotting:

In order to test unlabeled anti-VPg Abs, total protein from healthy and
infected plants was mixed with Laemmli buffer (Laemmli, 1970) and individual
proteins were separated on 15% SDS-PAGE (Sambrook et al., 1989). The
protein gel and lmmobilon-P membrane (Millipore) were incubated in transfer
buffer (a 1:5 dilution of methanol containing 25mM Tris, 192mM glycine, 0.005%
SDS) for 5 min prior to assembly of the electroblot sandwich. After completion of
the transfer in the cold room at 30V overnight or for 2 hours at 80V but prior to
the antibody incubation, the membrane was placed in a plastic tray and washed
for 5 min in Tris-buffered Saline + 0.1% Tween 20 (TBS-T; 20mM Tris-HCI pH
7.4, 150mM NaCl). Blots were probed for 45 min with primary Abs diluted TBS-T
containing 2% dried milk at a concentration of 1:1000 (vlv) for preimmune serum,
1:5000 (vlv) for anti-PEMV Abs, or 1:1000 (vlv) for anti-VPg Abs. Three washes
for 3 min in fresh changes of TBS-T were performed between antibody

incubations. Phosphatase-labeled, affinity purified goat anti-rabbit lgGs [diluted

47

122000 (vlv)] (Kirkegaard8Perry Laboratories Inc., Maryland) were used as
secondary Abs for another 45 min incubation. Afterwards membranes were
washed as previously indicated and proteins were visualized by the use of
chromogenic substrates with alkaline phosphatase-coupled antibodies
(Sambrook et al., 1989). Bromo chloro indolyl phosphate (BClP, 40pl of
75mglml in 100% dimethyl formamide [DMF]) and nitroblue tetrazolium (NBT,
40pl of 150mglml in 70% DMF) were added to 10ml of color buffer, pH 8.5 (0.1M
Tris, 0.1 M NaCl, 5mM MgClz). This buffer was used to incubate the blot with
agitation until the bands developed to the desired intensity. To remove all
reagent, the blot was extensively washed in tap water, once in TBS-T, five times
in deo and then placed into 0.02% sodium azide (NaNa) overnight in the dark

before drying. Dry blots were stored in the dark to prevent signal fading.

Western detection with biotinylated anti- VPg lgGs:

Extensively purified viral RNA (55pg), untreated or treated with 10pg
RNase A or 125 p9 Protease K for 90 min at 37°C, separated electrophoretically
and electroblotted as described above, were probed with biotinylated anti-VPg
lgGs according to the manufacturer's protocol (ECL-protein biotinylation module,
Amersham LifeScience). The membrane was blocked during a one hour
incubation in a plastic tray containing 5% blocking reagent after electroblotting.
After rinsing the blot twice with PBS-T, it was incubated for another hour, this
time with biotinylated anti-VPg lgG, primary antibody solution. Prior to the one

hour incubation with streptavidin-horseradish peroxidase (S-HRP; diluted 1:5000

48

in PBS-T), the membrane was rinsed twice in PBS-T, washed once for 15
min and twice for 5 min in fresh changes of PBS-T. The S-HRP was discarded,
the membrane rinsed twice and washed three times for 15 min in PBS-T.

Proteins were detected using ECL reagents and following the
manufacturer's protocol (Amersham LifeScience). Equal volumes of ECL
detection solution 1 and 2 were mixed (0.125mllcm2 membrane) at the end of the
last wash. The blot was placed on SaranWrap (DowBrand, Inc.), protein side
up, and detection reagents were added for 1 min without agitation. Excess
detection reagents were drained and the membrane was wrapped in SaranWrap,
smoothing out all air bubbles and placing it , protein side up in a Kodak x-ray
exposure holder (20.3x25.4cm, Eastman Kodak Company, Rochester, NY).
Autoradiography film (Kodak X-OMAT AR, 20.3x25.4cm) was placed on top and
the film was initially exposed for one minute. Film was developed in an
automatic developer (Kodak RP X-OMAT Processor, model M78) and additional
pieces of film were exposed for varying length of time based on the strength of

the signal.

Sequencing of VPg:

Fivehundred pg of extensivly purified viral RNA was incubated with three
units of T1 RNase (Sigma) for ten hours (based on the protocol described by
Zalloua et al., 1996). The preparation was given to the Biochemistry
Macromolecular Structure Facility, MSU for protein sequencing. They applied

the sample to a ProSorb membrane (Applied Biosystems) and subjected it to

49

Edman degradation in a Protein Sequencer, Procise model 494 (Applied

Biosystems).

Sequence analysis of VPg:

The obtained protein sequence was compared to other sequences in the
database using the BLAST program (Altschul et al., 1990). Protein secondary
structure was predicted using the University of Wisconsin (UW) GCG -
Peptidestructure program (Devereux at al., 1984) and the PHD (Profile fed
neural network systems from Heidelberg) protein mail server at EMBL
Heidelberg, Germany (Rost 8 Sander, 1993; Rost 8 Sander, 1994a; Rost 8
Sander, 1994b; Rost, 1996). Additional analysis was performed using the
default parameters of individual functions of the UW GCG program (Devereux et
al., 1984): Motifs to determine possible protein modification sites, BestFit for
protein sequence alignments of PEMV-VPg and other Luteoviruses , lsoelectric
to calculate the charge of the VP9 peptide at neutral pH , and FindPattems to

identify possible cleavage sites in the vicinity of the VP9.

Results

To characterize PEMV-VPg, antibodies (Abs) against the VP9 were
produced. The viral RNA purified from pea plants infected with the deletion
strain of PEMV (AT- A) was subjected to extensive phenol-chloroform extractions

to minimize the amount of contaminating cp and injected for antibody production.

50

4-cp

,4—VPg

 

 

I I I
PI or-PEMV or-VPg

Figure 2: Western blot analysis of total protein extracts isolated from pea
leaves. Protein extracts (40pg per lane) were separated by 15% SDS-PAGE
(5% stacking gel) and electroblotted onto membranes. These were analyzed
using: lane 1, 121000 (vlv) rabbit preimmune serum (Pl); lanes 2-4, 1:5000 (vlv)
rabbit anti-PEMV serum (or-PEMV); lanes 5-7, 1:1000 (vlv) rabbit anti-VPg serum
(a—VPg) in a double antibody sandwich with goat anti-rabbit alkaline
phosphatase conjugate. Total proteins isolated from leaves of plants inoculated
with: RNA 1 and RNA 2 infectious transcripts, lanes 1, 3, 6, respectively; RNA 2
infectious transcripts, lanes 4 and 7; lanes 2 and 5, mock-inoculated plants. The
positions of M, standards are indicated on the left as well as the position of the
gel front (arrow). The positions of PEMV coat protein (cp) and VPg are indicated
on the right.

51

Comparisons between total protein extractions from mock inoculated and
infected plants probed with anti-VPg Abs indicated a lightly staining band at
approximately 18kDa and a heavily staining doublet around 15kDa in RNA 1 +
RNA 2 infected plants (Fig. 2, lane 6) that were not detected with preimmune
serum (Fig. 2, lane 1). In the same protein extractions, Abs raised against virion
PEMV detect four major bands: the op at 21 kDa, a band at 18kDa and the
doublet at 15kDa (Fig. 2, lane 3). The reduction in cp signal between the two
Abs suggests anti-VPg Abs were raised mainly against the 15kDa doublet. A
reduction in cp signal but no change in the strength of the VP9 signal was
observed with sequential phenol-chloroform extractions (data not shown). We
concluded from these experiments that the minor reactivity of the anti-VPg Abs
against the op (Fig. 2, lane 6) was due to a slight contamination of the injected
sample with cp. None of the above mentioned bands were observed when total
protein extractions from RNA 2 infected plants were probed with anti-PEMV Abs
or anti-VPg Abs (Fig. 2, lanes 4 and 7, respectively) or in protein extractions of
mock-inoculated plants probed with either antibody (Fig. 2, lanes 2 and 5,
respectively). Since the immunoreactive 18kDa band (Fig. 2, lane 3 and 6)
appeared erratically in different experiments we concluded that the 15kDa
protein doublet most likely represents the VP9.

Biotinylated VPg-lgGs were used to detect the VP9 isolated from viral
RNA. No protein was detected in the untreated, viral RNA lane (Fig. 3, lane 1).

Both, the 18kDa and 15kDa proteins, were detected after RNase A-treated viral

52

 

Figure 3: Differentially treated VPg attached to PEMV RNA (55pg per lane)
probed with biotin labeled rabbit anti-VPg lgG. Treatments were: lane 1,
untreated RNA; lane 2, RNA treated with RNase A; lane 3, RNA treated with
Proteinase K The positions of M, standards are to the left of this figure while the

position of the PEMV-VPg is indicated by an arrow.

53

RNA was probed with biotinylated anti-VPg lgGs (Fig. 3, lane 2). Protease K
treatment resulted in the degradation of the proteins as evident from the
increased electrophoretic mobility (Fig. 3, lane 3). This indicates that anti-VPg
lgGs recognized a protein that was released from the viral RNA by RNase A
treatment.

Purified viral RNA was treated with RNase T1 to release the VP9.
Microsequencing of the protein revealed a 28 amino acid (aa) sequence with a
M' of 3157 Da. Using computer sequence alignments, the sequence encoding
this protein mapped to nucleotides 1811 to 1894 within ORF 1 of RNA 1,
downstream of nucleotides encoding the proteinase (Fig. 4). The small size and
location of the VP9 suggested a polyprotein expression strategy of ORF1 of
RNA1. Secondary structure prediction programs were used to determine a
possible folding pattern of the hypothetical 84K protein. In addition, the
protease accessibility of the putative VPg cleavage sites was analyzed in order
to identify possible cleavage sites. Both cleavage sites, Glu-Ala at the N-
terminus and Gln-Ser at the C-terminus of VPg, were solvent accessible and
loops were found at or near these sites. In addition, these dipeptide sequences
are cleavage sites known to be utilized by other viral 3C-like proteases (Wellink
8 van Kammen, 1988). These results are consistent with a total length of 28 aa

for the VP9.

54

PEMV-RNA 1 (5706 nucleotides)

 

 

  

 

ORF1-84K
h d h bl t E
I y r2310: 6 Pro eas1 I
| ORFO | ORF2 ] 21K(CP)33K(AT)
34K 67K (Pol) |0RI=3| ORF5 |

1811 1894

o o o o 0 0°
FEIGCE YSEAETAYE MADNFsocEEIufilch SYRTF

Figure 4: PEMV-VPg protein sequence and its location within the ORF 1 of
RNA1. Arrangements of the ORFs of RNA1, the protease, and the N-terminal
region containing hydrophobic patches, are based on Demler 8 de Zoeten
(1991). The shaded box indicates the position of the coding region of VPg. The
enlarged area shows the nucleotide coordinates and protein sequence of VPg
and putative cleavage sites ([1). The boxed amino acids indicate the P4 position
respective to the putative cleavage sites. Abbreviations: coat protein (CP),
aphid transmission factor (AT), RNA polymerase (Pol), basic amino acid (*),

acidic amino acid (0), K = kDa

Chapter 3

Qiecu_s_s_i.cm

In a previous study, the 5’ end of the viral RNA of PEMV was shown to
possess a covalently linked VPg with an apparent Mr of 17.5 kDa as determined
by electrophoresis (Reisman 8 de Zoeten, 1982). Despite speculations about
the coding region of this protein, no independent ORF corresponding to this
protein size was detected in the viral genome (Demler 8 de Zoeten, 1991).
Therefore, the goals of this study were a) to produce polyclonal Abs that could
verify the presence of the VP9 and b) to obtain and analyze the aa sequence of
the VP9, in order to deduce its placement on the viral genome.

The use of extensively purified viral RNA as immunogen was a simple
and effective way to obtain polyclonal Abs against the PEMV-VPg. These Abs,
either unlabeled or biotinylated, were used to characterize the VP9 by western
blot analysis. Several lines of evidence support our conclusion that the 15kDa
immunoreactive bands correspond to the genome-linked VPg. First, western
blot experiments that utilized anti-PEMV Abs and anti-VPg Abs to analyze
infected plant protein extracts (Fig. 2, lanes 3 and 6) showed that the VP9 has
different antigenic sites from those of the cp. Thus, anti-VPg Abs were raised
against a protein different from the 21 kDa cp. Second, experiments analyzing

purified viral RNA with biotinylated anti-VPg 1965 (Fig. 3) and infected plant

55

56

protein extracts with anti-VPg Abs (Fig. 2) both labeled a 15kDa doublet as
the predominant signal, suggesting that the same protein was recognized when
isolated by two different sample extraction procedures. Third, the protein
identified in this study as the VP9 bound to the RNA. Noncovalently bound
proteins were drastically reduced by the extensive viral RNA purification
procedure as reflected by the increasingly faint cp—derived signal (Fig. 2,
compare lanes 3 and 6). In addition, when viral RNA was treated with RNase A,
the 15kDa protein was preserved (Fig. 3, lane 2) but when the same viral RNA
was treated with Protease K (Fig. 3, lane 3) the protein was degraded. Thus, the
proteinaceous nature of the signal was confirmed. Vartepatian 8 Bogdanov
(1987) suggested that combined nuclease and protease treatment are the best
way to identify proteins covalently linked to RNA. However, we were unable to
detect a signal when untreated viral RNA was analyzed (Fig. 3, lane 1). This
was probably due to the inability of the large RNA-protein complex to enter the
15% polyacrylamide gel.

The detection of a double band in protein extracts (Fig. 2), as opposed to
the expected single band, may be due to several reason: 1) Protein degradation
could have generated the 15kDa protein doublet. Although a proteinase
inhibitor (PMSF) was added during the extraction of total protein, it is possible
that not all proteinase activity was inhibited and such activity could have
generated proteins containing VPg antigenic sites. 2) The anti-VPg Abs might
have recognized a VPg precursor protein in addition to the VP9. One protein of

the 15kDa doublet could be a VPg precursor. Poliovirus (Baron 8 Baltimore,

57

1982) and cowpea mosaic virus (CPMV) (label at al., 1982) have been
shown to posses precursor proteins to the VP9 that are recognized by Abs
raised against the VP9. 3) A differing number of RNase resistant bases might
be left attached to the protein. One residual RNA nucleotide linked to some VPg
molecules would be sufficient to change the M, and charge of the protein.

Protein modifications are also theoretically possible. The UW GCG -
Motifs program (Devereux et al., 1984) identified a possible N-glycosylation site
at aa 14 (Asn) and two possible casein kinase II phosphorylation sites at aa 7
and 16 (Thr and Ser). However, protein sequencing did not indicate any
modifications as ea signal peaks would have been altered.

Interestingly, a similar strength of the VP9 and cp signal was found in
protein extracts from infected plants probed with anti-PEMV Abs (Fig. 2, lane 3).
The reason for this is currently not understood. The amount of each protein
needed during the viral life cycle differs widely. Only one molecule of VPg is
thought to be attached to each RNA molecule while 180 or 150 subunits of cp
are required for formation of the RNA1 and RNA2 capsid, respectively (for
review see Demler et al., 1996a). Therefore, it is important to note that western
blotting techniques are not quantitative. The relative band intensities of VPg
and cp do not necessarily reflect that these proteins are expressed in equal
amounts. Hence, these proteins may differ in their antigenicity.

Protein sequencing data and subsequent computer analysis showed that
the PEMV-VPg gene was located downstream of the proteinase motif within the

84K ORF of RNA 1 (Fig. 4). This result differs from an earlier proposal by

58

Demler et al. (1996a) that predicted the coding region upstream of the
proteinase motif based on similarities of motif arrangements in Picorna-Iike
viruses and PEMV (Demler 8 de Zoeten, 1991). Proteolytic cleavage is a
strategy used by the Picornaviruses (Daugherty 8 Semler, 1993). Demler 8 de
Zoeten (1991) identified a potential proteinase sequence motif in the 84K ORF,
suggesting the processing of expressed viral proteins in PEMV infected tissues.
Size and location of the VP9 protein would argue for an involvement of
proteolytic activity in the release of the VP9 from its precursor protein. This is
the first direct indication for polyprotein processing in the expression strategy of
Enamo- and Luteoviruses.

The deletion strain of PEMV (AT-A) was cloned from a naturally occurring
isolate of PEMV characterized by the lack of 727 nucleotides from the 33K ORF
5 of RNA 1 (Demler, et al., 1996a). This isolate consistently produced three to
five times the yield of the aphid-transmissible [AT’] strains. This high yield made
it well suited for the isolation of large quantities of RNA needed for this study
(Demler et al., 1996a). Protein sequencing determined that the first aa of the
PEMV-AT-A-VPg is alanine. The corresponding aa in the published sequence of
RNA 1 derived from the WSG strain is threonine (Demler 8 de Zoeten, 1991).
The single base substitution from ACC to GCC causes this change from alanine
in the protein sequence of PEMV-AT-A -VPg to threonine in the AT-WSG-VPg.
Unpublished results (Skaf et al.) showed that the same ACC to GCC substitution
was found in the aphid transmissible PEMV-AT’. Hence, the discrepancy is

likely due to strain differences.

59

PEMV VPg consists of 28 aa. This is supported by secondary
structure predictions of the hypothetical protein encoded by ORF 1 and by the
proteinase accessibility of its putative cleavage. Accessible cleavage sites are
most likely exposed as surface loops (e. g. Daugherty 8 Semler, 1993; Pallai et
al., 1989; Arnold et al., 1987). In an analysis with different computer programs
(data not shown), the solvent accessibility at the cleavage sites surrounding the
VP9 was shown to be favorable and loops were found to be at or near these
sites.

Additional support comes from the likelihood that the putative VPg
cleavage sites are those used preferentially by 3C-like proteases. The majority
of proteolytic cleavages of the picomaviral polyprotein are carried out by 3C-like
proteases (Wellink 8 van Kammen, 1988). Most cleavages occur between Gln
or Glu and Gly or Ser, less frequently Thr/AlalMeWal, amino acid pairs (Wellink
8 van Kammen, 1988). Named after the 3C protease (30 "'°) of Picornaviruses,
the family of 3C-like proteases possess a conserved catalytic triad and belong in
the chymotrypsin -related protease superfamily (Gorbalenya 8 Koonin, 1993). In
addition to Picornaviruses, PEMV (Demler 8 de Zoeten, 1991) and Como-, Poty-
, and Nepoviruses (Daugherty 8 Semler, 1993) possess 3C-like protease motifs.
Hence, the region of the protein surrounding the PEMV-VPg was searched for
the presence of cleavage sites similar to those utilized by known 3C-like
proteinases (cleavage site sequences were taken from Daugherty 8 Semler,
1993; Wellink 8 van Kammen, 1988). This search indicated that the N-terminus

of PEMV-AT-A-VPg could be generated by a cleavage between glutamic acid

60

and alanine (Fig. 4), as has been previously observed for Picorna— and
Nepoviruses (for review see Wellink 8 van Kammen, 1988). In the published
sequence of the aphid-nontransmissible AT-WSG strain of PEMV, threonine is
encoded instead of alanine at the N-terminus (Demler 8 de Zoeten, 1991). This
Glu-Thr cleavage is also observed in Picornaviruses (for review see Wellink 8
van Kammen, 1988). Therefore, it appears likely that Ala, the first amino acid
obtained by protein sequencing, was actually the N-terminus of the VP9. The C-
terminus of the PEMV-VPg is believed to result from a Gln-Ser cleavage (Fig. 4).
Such a cleavage site is common for Picornaviruses as well as Como- and
Potyviruses (for review see Wellink 8 van Kammen, 1988). In addition,
comparative analysis of sequences at tentative protein domain boundaries in
polyproteins of Luteo- and Sobemoviruses previously predicted cleavage sites at
Gln or Glu and Gly or Ser or Ala (Gorbalenya 8 Koonin, 1993). We, therefore,
reason that the VP9 protein of PEMV is 28 aa long, with a calculated molecular
mass of 3157 Da.

The discrepancy between the calculated (3157 Da) and observed (15 kDa
or 17.5kDa) M, of PEMV-VPg may be due to residual RNA nucleotides that
remained attached to the VP9 thus contributing to the overestimated Mr of the
protein when run on SDS-PAGE. Other VPgs run also anomalously in SDS-
PAGE due to their cationic character (Daubert 8 Bruening, 1984). Interestingly,
the net charge of the PEMV-VPg is negative (-5.02) at neutral pH, suggesting an

anionic character, with one fourth of the aa being acidic (7aa) and only 7% being

61

basic (2aa) (Fig. 4). At present, It is not clear what factors contribute to the
aberrant migration of PEMV-VPg in SDS-PAGE.

The basic amino acids found in the C-terminal half of the poliovirus VPg
are thought to be involved in electrostatic interactions of VPg and RNA while the
N-terminal portion of the protein is involved in the protein-RNA linkage (for
review see Vartepatian 8 Bogdanov, 1987). Interestingly, the only two basic aa
found in PEMV-VPg are close to the C-terminus (see Figure 4). Two residues of
tyrosine and serine, amino acids shown previously to be involved in the RNA-
protein linkage of other viruses (see Chapter 1 for references), can be found
within the PEMV-VPg sequence. The tyrosine and the serine residues are each
found towards the N-terminus and in the center of the VP9. Based on the
position of the linking ea in the N-terminal half of the VP9 in several viruses
(Vartepatian 8 Bogdanov, 1987; Murphy et al., 1991; Jaegle et al., 1987; Pinck
et al., 1991; Zalluoa et al., 1996), we predict that the tyrosine at aa 2 or the
serine at aa 3 in PEMV-VPg are involved in the covalent linkage of the protein
and the RNA. However, at the moment experimental evidence for this
hypothesis is lacking.

No significant sequence similarity was found between the protein
sequences of PEMV-VPg and Luteoviruses or any other virus. This is not
surprising, since to date no common VPg sequence elements have been
identified and attempts to predict VPg domains on Luteo-, Sobemo- and
Calicivirus genomes through comparative sequence analysis have been

inconclusive (Gorbalenya 8 Koonin, 1993). Based on other similarities between

62

PEMV and Luteoviruses, information concerning the PEMV-VPg might prove
useful in evaluating the VPgs of Luteoviruses.

The cleavage site sequences Glu-Ala (or Glu-Thr) and Gln-Ser utilized by
the viral proteinase to release the VP9 were found at other sites within the 84K
polyprotein. However, not all possible dipeptide sequences within a respective
genome are cleaved. For example, only 9 out of 13 Gln-Gly pairs in the
poliovirus polyprotein are cleaved (Dewalt 8 Semler, 1989; Ypma-Wong et al.,
1988). It was proposed that in poliovirus 3C°'° specifically recognizes Gln-Gly
pairs while residues flanking this dipeptide together with structural determinants
mediate cleavage efficiency (Mirzayan 8 Wimmer, 1994).

Threonine was found at P4 in both PEMV-VPg cleavage sequences. The
cleavage site nomenclature is based on Berger 8 Schlechter (1970) where P1
corresponds to the newly generated C-terminal residue and P1' to the new N-
terrninal aa. Studies from Picoma-, Como- and Potyviruses determined aa
requirements in the vicinity of cleavage sites and their effect on processing
dynamics (for review see Wellink 8 van Kammen, 1988). The importance of the
P4 aa in addition to the dipeptide cleavage sequence is well documented (e. 9.
Blair 8 Semler, 1991; Cordingley et al., 1990; Pallai et al., 1989; Harris et al.,
1992). In the cases studied, the P4 position is preferentially occupied by a small
subset of amino acids with hydrophobic groups, preferentially Ala and less often
Val, Leu, Ile, Pro, Phe (for review see Wellink 8 van Kammen, 1988). In
poliovirus the P4 position regulates the cleavage efficiency (Mirzayan 8

Wimmer, 1994). The P5, P3 and P2 aa appear less critical but are still part of a

63

minimum length peptide sequence required for cleavage (Long et al., 1989;
Cordingley et al., 1990). This data is supported by X-ray crystallography of
another Picornavirus protease, the human rhinovirus 3C protease (Matthews at
al., 1994). A variety of amino acids are tolerated in the P5 and P3 position as aa
sidechains point into the solution while a spacious cavity in 3C"'° is found
opposite of P2. However, only small hydrophobic sidechains can be tolerated in
the P4 position due to steric constrains. In addition to the favored aa residues,
threonine can also be found at P4. Like PEMV-VPg, the poliovirus 3CD protein
carries an unfavored threonine in the P4 position of its 3C*D cleavage site
(Harris et al., 1992; Pallai et al., 1989). The polyprotein 3CD processing
intermediate is the proteolytically active precursor of the 3C protease and the
viral polymerase 3D (Mirzayan 8 Wimmer, 1994). Interestingly, this unfavorable
cleavage site still allows for sufficient processing to yield the necessary amounts
of cleavage products (Harris et al., 1992). Pallai et al. (1989) proposed that the
slow processing and the long 3CD half-life were a regulatory mechanism by the
virus to govern accessibility and availability of proteins during its life cycle. Also,
Semler et al. (1988) speculated that the initiation of RNA synthesis in vivo may
be controlled by proteolytic processing of VPg precursor polypeptides. In
PEMV, the viral polymerase of RNA 1 is expressed as a fusion protein with the
84K polyprotein containing the protease and the VP9. It is, therefore, possible
that like in poliovirus, different cleavage activities determine the availability of

proteins and that PEMV- VPg, whose release from the polyprotein is delayed by

64

suboptimal cleavage sites, may be activated by proteolysis at a specific
point during the viral life cycle, 99. RNA synthesis.

A similar concept, that proteolysis might temporarily and functionally be
connected with the RNA replication process, was proposed by Riechman et al.
(1993) for Potyviruses. A conserved heptapeptide recognition sequence on
substrate proteins is absolutely required for proteolysis by potyviral proteases
(Daugherty 8 Parks, 1989). The P6, P3 and P1 aa are strictly conserved in all
cleavage sites of a particular potyvirus. They define an active cleavage site
while the amino acids at P5, P4, and P2 vary and determine the rate and extent
of processing (Daugherty 8 Parks, 1989; Daugherty et al., 1989). Like other
viruses of the Picornavirata, the VP9 in Potyviruses lies upstream of the 3C
protease (Riechman et al., 1993). The nuclear inclusion protein a (Nla) of
Potyviruses consists of a N-terminal VPg domain and a C-terminal protease
domain (Riechman et al., 1993). In tobacco etch virus (TEV), the Nla protease
catalyzes a suboptimal internal cleavage site, cleaving some but not all Nla
molecules at a slow rate. PEMV-VPg is adjacent to the protease in the PEMV
genome, but in this case is located downstream rather than upstream of the
putative protease motif (see Figure 4). However, based on the appearance of a
suboptimal aa at P4 and the adjacent arrangement of VPg and protease, the
possibility exists that suboptimal cleavage sites are responsible for the release
of PEMV-VPg.

Based on the observations presented above, we suggest that the 84K

ORF of PEMV encodes a polyprotein. Therefore, the aa sequence of the 84 kDa

65

protein was searched for other possible cleavage sites of the utilized
dipeptide sequences, Glu-Ala (or Glu-Thr) and Gln-Ser. In addition to the
previously identified cleavage sequences, one Glu-Ala, four Glu-Thr and six Gln-
Sar dipeptide sequences were found. From those, only two Gln-Ser sites (at aa
600 and 720) downstream of the VP9 in the C-terminus of the 84K protein, seem
likely to be utilized by the viral protease. These sites have a favorable
secondary structure of solvent accessible loops and the appearance of a small
hydrophobic aa at P4. Both of these putative cleavage sites are located within
the region identified by Demler 8 de Zoeten (1991) for use as frameshifting site
in the expression of the RNA 1 polymerase. Maybe, one of these sites is used to
release the polymerase from the 84K-67K fusion protein (see Figure 1). No
consensus sequence was found comparing the minimum required peptide
cleavage sequences (P6 to P1 ’) of these putative sites with the VP9 cleavage
sites, however, no consensus flanking sequence could be identified in Picorna-
and Comoviruses either (Wellink 8 van Kammen, 1988).

The different evolutionary origins of PEMV RNA 1 and RNA 2 were
discussed previously (see Chapter 1). The polymerase of RNA 1 is closely
related to members of the Luteovirus subgroup Il (Demler 8 de Zoeten, 1991)
and belongs to the picornavirus-like supergroup (Koonin 8 Dolja, 1993).
Members of this supergroup possess a VPg and are thought to have evolved
from a common, VPg containing ancestor (Koonin 8 Dolja, 1993). Therefore, the
polymerase of the ancestor virus and polymerases derived from it, including

PEMV RNA 1 polymerase, may have become functionally dependent on a VPg.

66

On the other hand, PEMV RNA 2 polymerase is related to the carrnovirus-
ralated plant viruses (Demler at al., 1994a) that evolved from a different ancestor
which did not possess a VPg nor did any of the virus groups evolutionary
derived from this ancestor ever acquire a VPg (Koonin 8 Dolja, 1993).
Subsequently, these polymerases may have evolved RNA replication
mechanisms independent of a VPg. Based on these considerations, we believe
it unlikely that the RNA 2 polymerase depends on a VPg for replication of the
RNA 2 genome. No experimental evidence is available but based on these
evolutionary considerations, we pose that PEMV-VPg is only attached to RNA 1.

Future research on PEMV-VPg will answer unresolved questions such as:
Which aa is involved in the protein-RNA linkage? Is PEMV-VPg covalently
linked only to the RNA 1 molecule or does it also attach to RNA 2? What is the
function of the VPg? Other topics related to the PEMV-VPg also remain
unresolved, such as the unknown cleavage requirements of the viral protease or
the relationship between the protease and VPg in the viral life cycle.
Mutagenesis experiments could be used to study the function of the VP9 and
determine the aa involved in the protein-RNA linkage.

In conclusion, we have provided evidence that the VP9 of PEMV is
encoded in the ORF1 of PEMV RNA1 in a polyprotein arrangement of membrane
anchor-protease - VPg - polymerase. Post-translational proteolysis is the most

logical explanation for the expression of mature VPg.

Appendix

Appendix A

APPENDIX A

mutagenesis of the 33K ORF of RNA 2

Introduction

Pea Enation Mosaic Virus (PEMV), the sole member in the Enamovirus
genus (for review see Demler at al., 1996a), is aphid-transmissible and causes
significant losses in the pea-growing regions of the United States and in Europe
(for review see Hagedom, 1996). Two, isometric particles, each encoding a
different RNA, make up the infectious virus. RNA 1 is responsible for
encapsidation and aphid transmission, while RNA 2 determines mechanical
transmissibility and systemic movement (for review see Demler at al., 1996a).
Genomic sequencing of RNA 1 and RNA 2 revealed five and four open reading
frames (ORFs), respectively (Demler 8 de Zoeten, 1991; Demler at al., 1994a).
The functions of the gene products of individual ORFs was elucidated by
comparative sequence and/or experimental analysis. However, the functions of
the 34K and 84K ORFs of RNA 1 and of the 33K ORF of RNA 2 (see Figure 1)
remain obscure.

In order to learn molecular cloning techniques, I decided to introduce, in a
side-project, a nonsense mutation into the 33K ORF of RNA 2. We hoped that
this would possibly provide preliminary information pointing to a function of the

33K protein.

67

68

The strategy involved selecting a restriction site near the beginning of
the 33K ORF and modifying this site for early translation termination. These
modified complete cDNA clones were used for production of infectious
transcripts and plant inoculation. Infected plants were tested for viral RNA and
proteins and symptoms were observed. However, preliminary data are recorded
here and additional experiments are suggested to determine a functional role for

the RNA2 ORF1 product.

Materials and Methods

Creation of a frameshift mutation in the 33K ORF:

Five pg of pPER2d cDNA (Demler, et al., 1993) were linearized with 4
units (U) of BsaR1 (New England Biolabs lnc., Beverly, MA) in 1x NEB 2 buffer
(New England Biolabs) during a 60 min incubation at 37°C. One U of Klenow
enzyme (Boehringer Mannheim, Indianapolis, IN) and 1pl of 0.5mM of each
dNTP (dATP, dCTP, dGTP, dTl‘ P) were added to create blunt ends and the
incubation continued for another 15 min. The reaction was stopped with 1pl of
0.5M EDTA and the extent of DNA linearization evaluated on 1% non-denaturing
agarose gel (see below).

The linearized DNA was isolated using the GENECLEAN ll kit from BIO
101 (Vista, CA), following the manufacturer's recommendations. Initially, three

volumes of sodium iodine (Nal) stock solution were added to the reaction and

69

mixed before 5pl of “Glassmilk”, a silica matrix that binds DNA, were added.
The tube was vortexed horizontally for approximately 1 minute and placed for 5
min at 55°C while mixing every one to two minutes. DNA bound to Glassmilk
was pelleted by centrifugation for 5 seconds in an Eppendorf centrifuge at high
speed of 14000 (14K) rpm (Brinkmann Instruments, Inc., Westbury, NY), the
supernatant was removed and the pellet was resuspended and washed three
times with “New Wash”. Each wash involved resuspending the pellet in 10 to 50
volumes of “New Wash” and centrifuging for five seconds at high speed with
disposal of the supernatant. After the last wash, the “Glassmilk” was removed
from the DNA by resuspending the pellet in a volume of dH20 at least equal to
the volume of “Glassmilk” (5pl) previously added, and incubating at 50°C for
three minutes. The mixture was then spun for 30 sec. at 14K rpm and the
supernatant, containing the DNA, removed. The resuspension and incubation of
the “Glassmilk” pellet was repeated once to increase the efficiency of DNA
recovery. The DNA was ready for ligation and transformation reactions.

The ligation was performed by adding 5x ligation buffer (Boehringer
Mannheim) and 1 U of T4 DNA Ligase (Boehringer Mannheim) to the DNA
followed by incubation of the mixture at room temperature (RT) overnight. The
ligated DNA was transformed into competent JM101 Escherichia coli cells
(Sambrook at al., 1989). Competent cells were thawed in ice-water before 100pl
of cells were mixed gently with 50 - 200ng of ligated plasmid. The mixture was
placed on ice for 30 min, heated for 1 min at 42°C, and moved back on ice for

another 2 min before plating out the cells. Pre-poured LB plates (1% wlv

70

tryptone, 0.5% wlv yeast extract, 0.55% wlv NaCl, 0.12% v/v 4N NaOH and
1.5% wlv Bacto Agar) containing 40pglml ampicillin were warmed to 37°C prior
to plating the transformation reactions. One microliter of each transformation
reaction was mixed in 50pl of dH20 and plated out. The remainder of each
transformation reaction (roughly 99pl) was plated onto a separate plate. This
ensured varying concentrations of colonies growing on the different plates. All
plates were incubated overnight at 37°C in a incubator (Precision Gravity
Convection Incubator) and individual colonies were picked using sterile
toothpicks the next day.

Each selected bacterial colony was grown in 2ml of liquid 2xYT media

(1.6% wlv tryptone, 1% wlv yeast extract, 0.5% wlv NaCl) containing 40pglml
ampicillin ovamight at 37°C in an Orbit Environ-Shaker in preparation for
plasmid DNA isolation (Lee 8 Rasheed, 1990). Cultures were centrifuged in
1.5ml microcentrifuge tubes at 14K rpm for 1 min to pellet the cells and the
supernatant was discarded. Pelleted cells were resuspended in 100pl lysis
buffer (50mM glucose, 25mM Tris-HCI pH 8.0, 10mM EDTA) vortexed and
incubated for 5 min at RT. Two hundred pl of freshly prepared alkaline solution
(0.2N NaOH pH 5.5, 1% vlv SDS) and later 150pl ice-cold 7.5M ammonium
acetate, pH 5.5 were added. Each time the tubes were gently mixed by
inversion, followed by a five minute incubation in an ice-water bath. The mixture
was centrifuged at 14K rpm for 5 min. The clear supernatant was transferred to
a new microcentrifuge tube, containing 0.6 volumes of isopropanol, and the

reaction was incubated at RT for a minimum of 10min. A centrifugation at 14K

71

rpm for 10min followed and 100 pl of 2M ammonium acetate (pH 7.4) were
used to resuspend the DNA pellet. After a 5 min incubation in an ice-water bath,
the mixture was centrifuged for 5 min at 14K rpm, the supernatant transferred to
a new microcentrifuge tube containing 100pl of isopropanol and incubated and
centrifuged as described for the previous isopropanol step. The resulting pellet
was washed with 70% ethanol, centrifuged, vacuum-dried and resuspended in
Soul of water.

The success of plasmid isolations was evaluated on 1% non-denaturing
agarose gels containing 0.1p9lml of athidium bromide (EtBr). A 5pl aliquot of
plasmid DNA was mixed with 1pl of 6x loading dye (0.25% wlv bromophenol
blue, 0.25% wlv xylene cyanol FF, 40% wlv sucrose in water), electrophoresed
in 1xTBE (1.08% wlv Trizma base, 0.55% wlv boric acid, 0.037% wlv EDTA)
running buffer at 100V (BioRad minisub DNA call gel system, BioRad power
supply model 500l200) and the gel was photographed under UV light (312nm

Transilluminator FBTI 614, Fisher Scientific; Polaroid MP-4 Land Camera).

DNA sequencing and sequence analysis:

To confirm the frameshift mutation, the manipulated region was
sequenced in both directions using the dideoxynucleotide chain-termination
method of Sanger at al. (1977). DNA sequencing (Sequenase Version 2.0,
United States Biochemical, Cleveland, OH) was based on the manufacturer’s
protocol. The DNA was denatured by adding 0.2pl of 10N NaOH for each 10pl

of DNA (1-2pg). After a 5 min incubation at RT, 4pl of 5N NH4OAC and three

72

volumes of ethanol were added. The DNA was mixed wall and placed at -
80°C for at least 15 min. Precipitated DNA was centrifuged at 14K rpm for 10
min in an Eppendorf centrifuge and the resulting pellet was resuspended in 7pl
of sterile dH20 for each 10pl of starting DNA.

A short region of RNA 2, 50-100 bp upstream of the manipulated region in
either direction, was chosen as sites for primer annealing. Each primer was
complementary to the respective part of RNA 2. The17 nucleotide (nt) forward
primer, d263, spans nts 269 - 285 of RNA 2 (5' TGT sac TCT AGT GGA cc 3').
The 23nt reverse primer, dz64, hybridizes to nts 363 - 385 (5' TI'C AGC CAA
CAG GCG ACG ATG GA 3'). Primers were synthesized by the Biochemistry
Macromolecular Structure Facility, Michigan State University, East Lansing, MI.

One pl of either primer (25nglpl) and 2pl of 5X Sequencing buffer (0.2M
Tris-HCI pH 7.5, 0.1M M9012, 0.25M NaCl) were combined with the denatured
DNA, incubated for 2 min at 65°C and cooled to RT by placing it on the bench for
at least 15 min. DNA with the annealed primer was added to the labeling
reaction containing 1 pl of 0.1M D'l'l', 2pl of a dGTP,dTT P, dCTP mix (diluted 5-
fold in dHZO), 1pl of 3‘58 ATP (diluted 1:1 in dHZO) and 2pl of Sequenase
(diluted 1:8 in ice cold TE buffer [1% vlv 1M Tris-HCI pH 8.0, 0.2% vlv 0.5M
EDTA pH 8.0]). The reaction was mixed and placed at RT for 10min.

In preparation of the termination reaction, V-bottom microliter plates (96
well, 86x128mm, Fisher Scientific, Pittsburgh, PA) containing 2.5pl of the
respective didaoxynucleotide triphosphates(ddNTP’s: ddCTP, ddGTP, dd‘l'l'P,

ddATP) were pre-wanned at 42°C for at least 1 min. After completion of the

73

labeling reaction, 3.5pl of each reaction were added to each ddNTP and
incubated at 42°C for 5 min. Reactions were terminated by addition of 4pl stop
solution (95% formamide, 20mM EDTA, 0.05% bromophenol blue, 0.05% xylene
cyanol) to each well, and samples were heated in boiling water for 2 min prior to
loading of the electrophoresis gel.

Electrophoresis was performed on vertical 8% acrylamida gels. Gels
were poured at least 45 min prior to electrophoresis to ensure adequate
polymerization. The 8% acrylamide gel mixture consisted of 60ml of 20%
acrylamide stock solution (96.59 acrylamida, 3.359 N,N-methylene-bis—
acrylamida, 233.59 ultra-pure urea in 500ml 1xTBE), 90ml urea mix (233.59 urea
in 500ml 1xTBE), 150pl 25% ammonium persulfate and 150p| N,N,N’,N’-
tetramethyl-ethylene-diamine (TEMED). Half the volume of each sequencing
reaction was loaded between teeth of a shark tooth comb and electrophoresed
at a constant gel temperature of 50°C in 1xTBE until xylene cyanol reached the
bottom of the gel (Sequi Gen sequencing cell, BioRad; electrophoresis constant
power supply ECPS 30001150, Pharrnacia, Piscataway, NJ).

The gel was fixed in a solution of 12% methanol and 10% glacial acetic
acid for 20min and transferred to 3mm thick Whatman chromatography paper
(Whatman lntemational Ltd., Maidstone, England). The fixed gel was placed on
a gel dryer (Gel Dryer model 583, BioRad), covered with Saran wrap and dried
under vacuum for 1 hour at 80°C (Speed Vac Water Jet SW J120, Savant
Instrument lnc., Holbrook, NY). The dried sequencing gel was exposed

ovamight on autoradiography film (Kodak Scientific Imaging Film X-OMAT AR

74

35cm x 43cm, Eastman Kodak Company, Rochester, NY) in a Kodak X-ray
exposure holder (35.6cm x 43.2 cm) and the film was developed in an automatic
film developer (Kodak RP X-OMAT Processor, model M78) and analyzed.
The obtained sequences were compared to the original PEMV RNA 2
sequence using the default parameters of the University of Wisconsin (UW)

GCG - GAP program (Devereux at al., 1984).

Large scale CSC/z plasmid isolations:

After confirmation of the mutations, large quantities of selected plasmids
were purified on cesium chloride (CSCiz) gradients for RNA transcription
reactions (Sambrook at al., 1989).

A 1| flask containing 250ml liquid LB with 40pglml ampicillin was
inoculated with bacterial cultures harboring the mutation and grown ovamight in
a shaker incubator at 37°C (Orbit Environ-shaker). Each culture was divided in
half and poured into 250ml centrifuge tubes. The bacterial cells were pelleted in
a JA-14 rotor (Beckman) at 6000 rpm (5520 x g) for 15 min in a Beckman J2-21
centrifuge. The supernatant was aspirated and pellets were transferred to 40ml
centrifuge tubes (Beckman) using a spatula. Four ml of ice-cold solution I (
2.5% vlv 1M Tris-HCI pH 8.0, 10% vlv 100mM EDTA, 5% vlv1M sterile sucrose)
were added to each 250ml centrifuge tube to resuspend the remaining cells,
which were subsequently transferred to 40ml centrifuge tubes and vortexed until
the pellets were completely resuspended. Eight ml of freshly made solution ll

(2% vlv 1ON NaOH, 10% vlv 10% SDS) were added, mixed by inverting the

75

tubes and then incubated on ice for 10 min. Another 10 min incubation on
ice with occasional shaking followed after addition of six ml of solution III. This
solution was made by dissolving 29.4% (wlv) KOAc in a small volume of H20
and adjusting the pH with glacial acetic acid to 4.8 before bringing the volume to
500ml. At this point, the bacterial mixture was clear but with a white coagulant.
After centrifugation in a JA-20 rotor (Beckman) at 15K rpm (27,200 x g) for 20
min, the clear supernatant, was transferred (avoiding debris) to new tubes
containing 10ml isopropanol. The now cloudy solution was mixed well,
incubated for 5 min at RT and centrifuged at 10K rpm (12,100 x g) for 10 min.
The supernatant was discarded and pellets were washed with 70% ethanol,
vortexed briefly and spun at 10K rpm for 5 min. The resulting pellets were semi-
driad under vacuum for 5 min and completely dissolved in 4ml of 25mM Tris-HCI
pH 8.0 before being transferred to 14ml Falcon tubes (Becton Dickinson and
Company, Lincoln Park, NJ) containing 4.559 of CsClz. After the CsClz was
dissolved, 667pl 10mglml ethidium bromide (EtBr) were added; samples were
vortexed and incubated in the dark for 20 min. Solutions were spun in a JA-20
rotor (Beckman) at 10K rpm for 10 min and the clear supernatant was transferred
to Quick-Seal tubes (Beckman) for the WI 65.2 rotors (Beckman). The volume
usually filled the tubes but 1mglml CsClz in TE was sometimes used to top off the
tubes. All tubes were sealed with a Beckman tuba sealer and then centrifuged
at 45K rpm for 16 hours at 20°C.

The resulting gradients were handled with great care and appropriate

safety equipment was worn for the remainder of the isolation procedure. To

76

extract the DNA band from each gradient, a tube was securely clamped in
front of a hand-held long-wave UV lamp (Mineralight lamp model UVGL-25,
UVP, Inc., San Gabriel, CA), above a, EtBr waste beaker. The top of each tube
was pierced with a 256 needle to allow air to enter as the sample was
withdrawn. A fluorescing horizontal band in the middle of the tube represented
plasmid DNA and a vertical dark band the RNA pellet. A 3cc syringe with an
186 needle was carefully inserted, beveled side up and avoiding the pellet, just
under the plasmid band which was then slowly drawn into the syringe by
sweeping back and forth while filling the syringe. In case the needle became
plugged, another needle was inserted. Once the plasmid was collected, the
needle was removed and the sample dispensed into 14ml Falcon tubes. After
plasmid DNA was collected from all tubes, EtBr was removed by three
extractions with 10ml of NaCl-saturatad isopropanol each. The plasmid DNA
remained in the bottom fraction while the top fraction retaining EtBr was
removed. Afterwards, each sample was adjusted to a total volume of 5ml with
deo, mixed, transferred to 15ml Corex tubes (Corning Incorporated, Corning,
NY) and precipitated with 10ml of EtOH at -20°C for 1 hour. The precipitated
DNA was pelleted by centrifugation in a JA-20 rotor for 30min at 8K rpm, washed
with 70% EtOH and dried under vacuum. Dried pellets were resuspended in
100pl TE, the concentrations determined by UV-absorption (1 ODm=50pglml
DNA) and the remaining DNA stored at -20°C.

The DNA was then used in RNA transcription reactions and infectious

transcripts were used to inoculate pea plants as described in chapter 2.

77

Northern and western blot analysis of infected plants:

Pea plants inoculated with infectious transcripts containing the mutation
were analyzed by northern and western blot analysis for the presence of viral
RNA and proteins, respectively.

Total RNA was extracted from infected pea plants based on a protocol by
Demler 8 de Zoeten (1989). Care was taken to prevent any contamination with
RNases during the procedure and preparation of solutions. Freshly cut plant
tissue (0.6 - 1.09 for each sample) was placed at 4°C for approximately 10 min
prior to the isolation. The tissue was first powdered by freezing in liquid nitrogen
and grinding with a mortar and pestle. The pulverized tissue was transferred to
30ml Corex tubes (Corning) using a spatula. One sample at a time, 5ml of
chilled, freshly made grinding buffer (4% wlv paraamino salicylic acid Na salt
(PAS, Sigma], 1% triisopropylnaphtalene sulfonic acid [T NS, Kodak], 10mM
DTT, 10mM Na meta bisulfite in DEPC-treated 50mM Tris-HCI pH 8.0) were
added and later 4ml of Tris-saturated phenol while vortexing the sample
between additions. All samples were placed on ice during processing of
remaining samples and then centrifuged in a JA-20 rotor (Beckman) for 15 min
at 9K rpm and 4°C in a Beckman J2-21 centrifuge. Each supernatant was
transferred to another tube and proteins were removed by extraction with 2ml
chloroform and centrifugation as indicated above. Again, the supernatant was
removed and transferred to a 15ml Corex tube for Lithium Cloride (LiCI)

precipitation overnight at —80°C. For each 5ml of supernatant 2.2ml of 10M LiCl

78

were added. The precipitated RNA was centrifuged in a JA20 rotor for
20min at 9K rpm and 4°C. Using a glass pipette, the supernatant was aspirated
and discarded. The pellets were washed with 2ml of 2M LiCI, respun and dried
under vacuum. Each pellet was resuspended in 300pl deo, before transfer to
an Eppendorf tube. Unsuspended pieces were removed by centrifugation for
5min at 14K rpm in an Eppendorf centrifuge and the supernatant was transferred
to a new tube. An ethanol precipitation at -20°C for at least one hour followed
after adding 30pl 3M NaOAc and 825pl ethanol. Precipitated RNA was
centrifuged at 14K rpm for 20min, the supernatant aspirated and the remaining
pellet washed with 70% EtOH, respun, dried under vacuum and resuspended in
200pl dH20. The RNA concentration was determined by UV absorbance at
260nm. Each RNA sample was separated on 1% non-denaturing agarose gels
as described above.

Electrophoretically separated, total RNA was transferred to a nylon
membrane (Micron Seprarations, Inc. (MSI), Westboro, MA) as part of the
northern blotting procedure. For this purpose a “sandwich”, described below,
was assembled. Interfering bubbles between bread slices were removed by
rolling a pipette firmly on the surface as each new layer was added. Four pieces
of Whatman 3mm paper were cut to double-gel size, saturated in 10xSSC (1.5M
NaCl, 0.15M Na citrate) and stacked on an oversized piece of Saran wrap. The
gel was placed in the center of the Whatman paper after trimming off excess
pieces. The oversized Saran wrap was then used to cover the exposed parts of

the Whatman paper before the nylon membrane (MSI) was placed on top of the

79

gel. The membrane was cut to the same size of the gel, wetted in dH20 and
saturated in 10xSSC prior to the application. Three gel-size pieces of Whatman
paper, saturated in 10xSSC, constituted the next layer. A stack of paper towels,
cut to the approximate gel-size, and a weight centered on top of a glass plate
completed the “sandwich”. The next day, the membrane was removed, air dried
and the RNA fixed by baking for 2 hours in a 68°C oven (lsotemp oven model
6556, Fisher Scientific). At this point, the membrane was either stored or used
immediatelly.

The Genius 3 system “Nonradioactive Nucleic Acid Detection Kit" from
Boehringer Mannheim was used to prepare RNA hybridization probes and
perform the hybridization, washing and chemiluminescent detection of labeled
RNA-RNA hybrids. To make digoxigenin (DIG) - labeled RNA probes, the DNA
template was linearized at a restriction site. For the RNA 1 specific probe, 10 pg
of DNA were linearized with 2pl Hindlll (Boehringer Mannheim), while the same
amount was linearized with 2pl Pstl (Boehringer Mannheim) for the RNA 2
specific probe. Each restriction enzyme digest was carried out at 37°C for one
hour in the optimal buffer as specified by the manufacturer. The DNA template
was purified by phenol/chloroform extraction and ethanol precipitation.

During probe preparation great care was taken to prevent contamination
of the samples with RNases. RNA was labeled during a transcription reaction
for which the following reagents, in the indicated order, were added to 2pg DNA
template in a microfuge tube on ice: 4pl 0.1M DTI', 4pl NTP labeling mixture, 8pl

5x transcription buffer (200mM Tris-HCI pH 8.0, 30mM MgCIz, 50mM DTT, 10mM

80

sperrnldine, 50mM NaCl), 1pl T7 RNA polymerase and DEPC-dH20 to adjust
the total volume to 40pl. The solution was mixed, incubated at 37°C for 2 hours,
and the reaction was stopped with 2pl EDTA. One pl glycogen was added to
each reaction tube prior to RNA precipitation with 0.1 volume of 3M NaOAc and
3.0 volumes of chilled EtOH. After completion of a 30 min incubation at -80°C,
the reactions were centrifuged for 15 min at 14K rpm in a microcentrifuge, the
EtOH decanted, the pellet washed with 70% EtOH and centrifuged again at the
same speed for 5min. The EtOH was removed prior to vacuum drying the pellet
and resuspension in 100pl dHZO. The resuspension was carried out at 37°C for
10 min under frequent vortexing. Labeled probes were stored at -80°C.

The hybridization procedure followed the labeling of probes and
preparation of the northern blots. Duplicate blots were always prepared in order
to probe each blot with a separate probe. Each blot was placed into a
hybridization bag (Gibco BRL Life Technologies, Inc., Gaithersburg, MD)
containing 20ml of northern prehybridization (prehyb) solution (50% formamide,
0.02% SDS, 0.1% N-lauroylsarcosine, 2% [wlv] blocking reagent for nucleic acid
hybridization, 20mM sodium maleate pH 7.5 in 5xSSC pH 7.0 [0.75M NaCl,
75mM sodium citrate]) per 1000m2 of membrane surface. The bags were sealed
(Impulse Sealer, TEW Electric Heating Equipment Co., Ltd.), incubated for at
least 2 hours in an oven at 68°C and the solution was discarded. Previously
made, DIG-labeled probe was added to 25ml of northern prehyb solution to
make the northern hybridization solution which was added to the membrane for

an overnight incubation at 68°C. The next day, the hybridization solution

81

containing unannealed DIG-labeled probe was transferred into orange
capped tubes (Coming) and stored at -80°C for future experiments. (For reuse,
probes were thawed and heated for 10min to 68°C prior to use.) After the
annealing step, blots were removed from the bag and placed into plastic trays for
washes. Membranes were washed twice for 5 min in 1xSSC containing 0.1%
SDS at RT and twice for 15min in 0.1xSSC + 0.1%SDS at 68°C.

As part of the chemiluminescent detection procedure, membranes were
first washed in 50ml of maleate buffer (100mM maleic acid pH 7.5, 150mM NaCl)
for 1 min prior to a 30 to 60min incubation in 50ml of northern blocking solution
(10% blocking reagent stock solution [10% wlv blocking reagent for nucleic acid
hybridization, 0.1M sodium maleate pH 7.5] diluted 1:5 in maleate buffer). All
remaining steps were performed at RT. Next, anti-DlG-alkaline phosphatase
(Boehringer Mannheim) diluted 1:5000 in blocking solution was used to replace
the northern blocking solution and the membranes were incubated for 30min.
Unbound antibody was discarded; the membrane was moved to a new plastic
tray and washed twice with 200ml maleate buffer for 15min. Blots were
equilibrated in 50ml of Genius 3 buffer (100mM Tris-HCI and 100mM NaCI
adjusted to pH 9.5, 50mM MgClz) and placed between plastic page protectors.
Afterwards, 0.5ml per 100cm2 of Lumi-Phos 530 (Boehringerr Mannheim) was
scattered and agitated gently for 1 min to distribute the reagent over the entire
surface. The excess solution was then drained and the top sheet of plastic was
lowered. All air bubbles were removed prior to covering with Kodak x-ray film

(Kodak X—OMAT AR, 20.3 x 25.4 cm) in a film cassette holder (Kodak x-ray

82

exposure holder, 20.3 x 25.4 cm). Exposure times varied depending on
signal strength. After adequate exposure, the film was developed.
Total protein isolation from leaves of infected pea plants, SDS-PAGE,
electroblotting and wastem blot analysis were carried out as described above
(see chapter 2). Anti PEMV Abs (diluted 1:5000 vlv in TBST containing 2%

dried milk) were used as primary Abs preparations.

Results and Discussion

In order to introduce a nonsense mutation into the 33K ORF of RNA 2, the
restriction enzyme BsaR1 was chosen. The enzyme has only one restriction site
in the cDNA clone at nucleotide (nt) 342 of RNA 2, one-third into the 33K ORF
(nt 21-1001), and none in the plasmid. The following sequence is recognized by
BsaR1:

5' 6A66A6(N)10 3'

3' crccrcm). 5'
This results in a two base pair 3’ overhang that was filled by the Klenow unit of
DNA polymerase I to create blunt ends. Successful cloning was verified by
genomic sequencing in the target region due to the unusual restriction
characteristics of the enzyme, which has a recognition sequence different from
the cleavage site. Two clones, UORF1-12 and UORF1-13, containing an
identical two bp deletion (nt 341 and 342), were identified and used for

subsequent experiments. Figure 4 shows the sequence in the target region and

83

the aa sequences before and after mutagenesis. The early termination of
the protein results in a putative protein of 110aa instead of the 326 aa full-length
protein.

Both mutant clones were then purified in large scale purifications and
infectious transcripts were produced for inoculation of young pea plants. Visual
comparison of symptoms between plants infected with wild-type or mutant
clones, northern and wastem blotting techniques to analyze RNA replication and
expression of viral proteins were used, respectively, to study the effect of the
mutation in plants.

Four different sets of pea seedlings were Inoculated with different
combinations of wild-type (wt) and mutant clones: mock inoculated, wt RNA 1
and RNA 2, RNA 2 alone, RNA 1 and UORF1-12, RNA 1 and UORF1-13,
UORF1-12 alone, UORF1-13 alone. Each time, one pot with four to five plants
was used for mock or wt inoculations and two pots (8-10 plants) for each mutant
inoculation. The majority of plants in each treatment group showed typical PEMV
symptoms after ten days post-inoculation, excluding the mock-inoculated plants
which had no visible, virally induced symptoms. No difference in severity of the
symptoms could be distinguished between wt and mutant inoculated plants.
Symptom onset appeared in some mutant infected plants one to two days
delayed. However, due to the subjective nature of these observations no
definitive statement can be made.

Total RNA and total protein extractions were obtained from a symptomatic

plant of each treatment. Northern and wastem blot analysis showed differing

84

s' (327) (359) 3'
RNA2 GAG GAG T66 1361' GCG GAG AT'I' SAT 661'
E E S C K s A E I D P

5' (327) (357) 3'
UORF GAG one was was no ace car; car 'rca mac '1'
a a s c . n s c o *

Figure 5: Mutated region of ORF1 (33kDa) of PEMV RNA2. A stop codon was
introduced in the ORF1 at nucleotides 351 to 353 that resulted in an early
termination of the 33kDa protein after 110 amino acids (see Materials and
Methods in Appendix A for details). The nucleotide and amino acids sequence
of RNA2 and mutant UORF are shown. The location of the nucleotides in RNA2
is indicated above the nucleotide sequence in brackets as well as the orientation
of the RNA. The BsaR1 recognition site is underlined and the cleavage site is
indicated by the arrow in the RNA2 sequence. The boxed nucleotides were
deleted during the mutagenesis, causing the frameshift. The stop codon is

indicated by the star (*).

85

results for each isolation. However, taken all the data together, there appears to
be no obvious difference in the RNA or protein pattern between wt and mutant
infected plants. Due to the limited number of samples and technical problems, it
is difficult to draw general conclusions and the results presented below should
be viewed with caution.

Northern blotting procedures used RNA 1 and RNA 2 specific probes
which hybridized to the viral polymerase of each genomic RNA. Therefore, a
positive signal indicated replication of the respective RNA. Table 1 shows a
graphical representation of the obtained results indicating the detection of a
positive (+) or negative (-) signal. Total RNA extractions from each of the four
sets of inoculations (A-D) were tested. Two of the four RNA extractions were
tested twice (A and B) and the others (C and D) only once. As expected, total
RNA of mock inoculated plants showed no signal when tested with either RNA 1
or RNA 2 specific probes in all cases, indicating that these plants were not
infected with PEMV. Also, no signal was seen when total RNA of RNA 2,
UORF1-12 and UORF1-13 inoculated plants were tested with the RNA 1 specific
probe. The absence of a RNA 1 specific signal was expected in these samples
since the plants were only infected with infectious clones of wt and mutant RNA
2 but not with RNA 1. Therefore no contamination with RNA 1 occurred in these
samples. Total RNA of RNA1 + UORF1-12 or UORF1-13 inoculated plants
showed the same infectivity pattern (see table 1) as total RNA of RNA1 + RNA2

infected plants. Some but not all samples of the three treatments gave positive

86

results. The result, that the wt infections (RNA1 + RNA2) were not always

successful, suggest that the negative results obtained for some mutant
treatments are likely due to technical problems, like bad probes or unsuccessful
inoculations, rather than a property that can be attributed to the mutated 33K
protein. When UORF1-12 and UORF1-13 where inoculated alone, the isolated
total RNA showed fewer positive signals than total RNA from plants infected with
RNA2 alone, after probing with the RNA2 specific probe. The reason for this is
not known. It could be due to either experimental inconsistencies, or to the
mutation in the 33K ORF of RNA2. If the latter was the case, it might suggest
that the presence of RNA 1 has a positive effect that overcame the mutation.

Total protein was isolated from three sets of inoculations (B-D). Proteins
were isolated, when possible, from leaves of plants also used for total RNA
extractions. Two of the three sets of inoculations (C and D) showed virus-
specific signals in the positive control (total protein extracts from RNA1 + RNA2
infected plants) when probed with anti-PEMV Abs. One of the wastem blots did
not show the expected signals in the positive control and was, therefore,
excluded from the discussion. A typical total protein extraction from RNA1 and
RNA2 infected plants probed with anti-PEMV Abs show three to four specific
bands: the 21 kDa coat protein, a 15kDa doublet representing the VP9, and
sometimes an 18kDa protein that appears erratically in different experiments
(see Figure 2, lane 3). However, total protein extracts of RNA 2 infected plants
do not show these PEMV-specific bands (see Figure 2, lane 4). Therefore as

expected, these bands were not seen when total protein extract of mock-

87

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A A B B C D
probe RNA1 RNA2 RNA1 RNA2 RNA1 RNA2 RNA1 RNA2 RNA1 RNA2 RNA1 RNA2
—
treatment
mOCk - u u I u I I I u c I -
RNA2 . + b - If“ - QI. I +b. I +u I +h
1+12 . - . . . +. - +5. - +8 - 4'1).
1+1: . + b - +u - +‘ -9 o 9 I I I +u

 

 

 

Table 1: Northern blotanalysis of plants infected with wild-type and/or UORF

infectious clones. Treatments are indicated in the left column while the different

sets of inoculations (A-D) are indicated in the top row. The RNA1 or RNA2

specific probe used during Northern blot analysis is indicated for each treatment.

Abbreviations: positive signal (+), negative signal (-), b = clear bands, bs = clear

bands but associated with a smear, s = smear but no clear band, 9 = RNA was

degraded in the agarose gel before transfer and no signal is expected.

88

inoculated, RNA 2, UORF1-12 or UORF1-13 infected plants were tested in
this study. However, when mutant or wt RNA 2 were infected together with RNA
1 and total proteins were isolated and tested, the results differed between the
two wastem blots. In one wastem blot, total protein extracts of RNA 1 and RNA
2 infected plants reacted specifically with the 21kDa cp, the 18kDa protein and
the 15kDa VPg doublet. Only total protein extract of RNA1 + UORF1-12 but not
RNA1 + UORF1-13 infected plants reacted with the cp and the 15kDa protein
doublet. No conclusion can be drawn about the 18kDa protein signal because of
its erratic nature. It is not clear why the total protein extraction of the RNA1 +
UORF1-13 inoculated plant from set C did not show any virus-specific signals
when probed with anti-PEMV Abs. In the other wastem blot from set D, the
15kDa protein doublet was detected in total protein extracts of RNA1 + 2, RNA1
+ UORF1-12 and RNA1 + UORF1-13 infected plants, but neither the cp nor the
18kDa protein. It is not clear, why no cp signal was detected. More repetitions
of this experiment are needed to allow a statistical conclusion. However, it
appears that no distinctive difference in the virus-specific protein signals was
detected in total protein extracts of RNA1 + UORF versus RNA1 + 2 infected
plants. The cp and the 15kDa protein were detected in at least one case in the
mutant infected plant. This could indicate that these proteins were not affected
by the mutation in the 33K ORF of RNA 2. Therefore, the 33K gene product is
unlikely to regulate the expression of either protein.

As already indicated, this data is preliminary and more repetitions of

northern and western blots are needed to derive any statistical data. The 33K

89

gene product does not appear to be absolutely essential. Infectious clones
carrying the mutation remained infectious either in conjunction with RNA 1 or
alone. Analyzing viral RNA from mutant infected plants by PCR and genomic
sequencing is one of the several things necessary to come to a definitive
conclusion.

In conclusion, an infectious clone containing a mutation in the 33K ORF
was constructed. However, we were unable to obtain data indicating a possible
function of this ORF gene product. As already indicated, more repetitions of the
described experiments are needed to make this preliminary data statistically
relevant. In addition, supplemental experiments are needed. This would include
PCR and sequence analysis of viral RNA from mutant infected plants but also
additional constructs. The described mutant construct results in a protein one
third of its original length. It will also be necessary to completely eliminate the
33K ORF to determine which function resides in the N-terminal portion of the

protein.

List of References

List of References

Adam, 6. (1978). Die gemeinsame Ubertragung eines vektoriellen und
avektoriellen Stammas des Pea Enation Mosaik Virus durch die Blattlaus
Acyrthisophon pisum. Zeitschrift far Pflanzenkrankheiten und Pflanzenschutz 85,
586-593.

Altschul, S. F., Gish, W., Miller, W., Myers, E. W. 8 Llpman, D. J. (1990).
Basic local alignment search tool. Journal of Molecular Biology 21 5, 403-410.

Andrews, N. C. 8 Baltimore, D. (1986). Purification of a terminal uridylyl
transferase that acts as host factor in the in vitro poliovirus replicase reaction.
Proceedings of the National Academy of Sciences of the United States of
America 83, 221-225.

Andrews, N. C., Levin, D. 8 Baltimore, D. (1985). Poliovirus replicase
stimulation by terminal uridylyl transferase. The Journal of Biological Chemistry
260, 7628-7635.

Arnold, E., Luo, M., Vriend, 6., Rossmann, M. 6., Palmenberg, A.C., Parks,
6. D., Nicklin, M. J. H. 8 Wimmer, E. (1987). Implications of the picomavirus
capsid structure for polyprotein processing. Proceedings of the National
Academy of Sciences of the United States of America 84, 21 -25.

Baron, M. H. 8 Baltimore, D. (1982). Anti-VPg antibody inhibition of the
poliovirus replicase reaction and production of covalent complexes of VPg-
related proteins and RNA. Cell 30, 745-752.

Barton, D. J., Black, E. P. 8 Flanegan, J. B. (1995). Complete replication of
poliovirus in vitro: preinitiation of RNA replication complexes require soluble
cellular factors for the synthesis of VPg-linked RNA. Journal of Virology 69,
5516-5527.

Berger, A. 8 Schlechter, l. (1977). Mapping the active site of papain with the
aid of peptide substrates and inhibitors. Philosophical Transactions of the Royal
Society of London, Series B 257, 249-264.

Blair, W. S. 8 Semler, B. L. (1991). Role for the P4 amino acid residue in
substrate utilization by the poliovirus SCD proteinase. Journal of Virology 65,
6111-2123.

91

Carrington, J. C., Haldeman, R. Dolja, V. V. 8 Retsrepo-Hartwig, M. A.
(1993). lntemal cleavage and trans—proteolytic activities of the VPg-proteinase
(Nla) of tobacco etch potyvirus in vivo. Journal of Virology 67, 6995-7000.

Chu, P. W. 6., Boccardo, 6. 8 Franckl, R. I. B. (1981). Requirement of a
genome-associated protein of tobacco ringspot virus for infectivity but not for in
vitro translation. Virology 109, 428-430.

Cordingley, M. 6., Callahan, P. L., Sardana, V. V., Garskky, V. M. 8 Colonno,
R. J. (1990). Substrate requirements of human rhinovirus 3C protease for
peptide cleavage in vitro. The Journal of Biological Chemistry 265, 9062-9065.

Crawford, N. M. 8 Baltimore, D. (1983). Genome-linked protein VPg of
poliovirus is present as free VPg and VPg-pUpU in poliovirus-infected cells.
Proceedings of the National Academy of Sciences of the United States of
America 80, 7452-7455.

Daubert, S. D. 8 Bruening, 6. (1979). Genome-associated proteins of
Comoviruses. Virology 98, 246-250.

Daubert, S. D. 8 Bruening, 6. (1984). Detection of genome-linked proteins of
plant and animal viruses. Methods in Virology 8, 347-379.

Daubert, S. D., Bruening, 6. 8 Najarian, R. C. (1978). Protein bound to the
genomic RNAs of cowpea mosaic virus. European Journal of Biochemistry 92,
45-51.

Demler, S. A. 8 de Zoeten, 6. A. (1989). Characterization of a satellite RNA
associated with pea enation mosaic virus. Journal of General Virology 10, 1075-
1 084.

Demler, S. A. 8 de Zoeten, G. A. (1991). The nucleotide sequence and
luteovirus-like nature of RNA 1 of an aphid non-transmissible strain of pea
enation mosaic virus. Journal of Virology 72, 1819-1834.

Demler, S. A., Rucker, D. 6. 8 de Zoeten, 6. A. (1993). The chimeric nature of
the genome of pea enation mosaic virus: the independent replication of RNA 2.
Journal of Virology 74, 1-14.

Demler, S. A., Borkhsenious, O. N., Rucker, D. 6. 8 de Zoeten, 6. A.
(1994a). Assessment of the autonomy of replicative and structural functions
encoded by the Iuteo-phase of pea enation mosaic virus. Journal of General
Virology 75, 997-1007.

92

Demler, S. A., Rucker, D. G. 8 de Zoeten, G. A. (1994b). Replication of the
satellite RNA of pea enation mosaic virus is controlled by RNA 2-encoded
functions. Journal of General Virology 75, 1399-1406.

Demler, S. A., de Zoeten, G. A., Adam, 6. 8 Harris, K. F. (1996a). Pea

Enation Mosaic Enamovirus: properties and aphid transmission. In The plant
viruses. Volume 5: Polyhedral virions and bipartite RNA genomes, 1" edn, pp.
303-344. Edited by B. D. Harrison 8 A F. Murrant. New York: Plenum Press.

Demler, S. A., Rucker, D. G., de Zoeten, G. A., Ziegler, A., Robinson, D. J. 8
Murant, A. F. (1996b). The satellite RNAs associated with the groundnut rosette
disease complex and pea enation mosaic virus: sequence similarities and ability
of each other’s helper virus to support their replication. Journal of General
Virology 77, 2847-2855.

Devereux, J., Haeberll, P. 8 Smithies, O. (1984). A comprehensive set of
sequence analysis programs for the VAX. Nucleic Acids Research 12, 387-395.

de Varennes, A., Lomonossoff, G. P., Shanks, M. 8 Maule, A. J. (1986). The
stability of cowpea mosaic virus VPg in reticulocyte lysates. Journal of General
Virology 67, 2347-2354.

Dewalt, P. G. 8 Semler, B. L. (1989). Molecular biology and genetics of
poliovirus protein processing. In Molecular Aspects of Picornavirus Infection and
Detection. Edited by B. L. Semler 8 E. Ehrenfeld, pp. 73-93, Washington, D. 6:
American Society for Microbiology.

de Zoeten, 6. A., Gaard, G. 8 Diez, F. B. (1972). Nuclear vesiculation
associated with pea enation mosaic virus-infected plant tissue. Virology 48, 638-
647.

de Zoeten, G. A., Powell, C. A., Gaard, G. 8 German, T. L. (1976). In situ
localization of pea enation mosaic virus double-stranded ribonucleic acid.
Virology 70, 459-469.

de Zoeten, G. A. 8 Gaard, 6. (1983). Mechanisms underlying systemic
invasion of pea plants by pea enation mosaic virus. Intervirology 19, 85-94.

de Zoeten, G. A. 8 Demler, S. A. (1995). Genus Enamovirus. In Virus
Taxonomy - Classification and Nomenclature of viruses. Sixth Report of the
lntemational Committee of Taxonomy of Viruses, 1" edn, pp. 384-387. Edited by
F. A. Murphy, C. M. Fauquet, D. H. L. Bishop, 8. A. G. Jarvis, G. P. Martelli, M.
A. Mayo 8 M. D. Summer. Vienna: Springer Verlag.

93

Dolja, V. V. 8 Carrington, J. C. (1992). Evolution of positive-strand RNA
viruses. Seminars in Virology 3, 315-326.

Domier, L. L., Shaw, J. 6. 8 Rhoads, R. E. (1987). Potyviral proteins share
amino acid sequence homology with picoma-, coma- and caulimoviral proteins.
Virology 158, 20-27.

Daugherty, W. 6., Cary, S. M. 8 Parks, T. D. (1989). Molecular and genetic
analysis of a plant virus polyprotein cleavage site: a model. Virology 171, 356-
364.

Dougherty, W. G. 8 Parks, T. D. (1989). Molecular genetic and biochemical
evidence for the involvement of the heptapeptide cleavage sequence in
determining the reaction profile at two tobacco etch virus cleavage sites innn
cell-free assays. Virology 172, 145-155.

Daugherty, W. 6. 8 Parks, T. D. (1991). Post-translational processing of the
tobacco etch virus 49-kDa small nuclear inclusion polyprotein: identification of
an lntemal cleavage site and delimitation of VPg and proteinase domains.
Virology 183, 449-456.

Daugherty, W. G. 8 Semler, B. L. (1993). Expression of virus-encoded
proteinases: functional and structural similarities with cellular enzymes.
Microbiological Reviews 57, 781 -822.

Drygin, Y. F., Sapotsky, M. V. 8 Bogdanov, A. A. (1987). Radish mosaic virus
VPg. FEBS Letters 215, 247-251.

Ehrenfeld, E. 8 Richards, 0. C. (1989). Studies of poliovirus RNA polymerase
expressed in Escherichia coli: attempts to understand viral RNA replication. In
Molecular Aspects of Picornavirus Infection and Detection. Edited by B. L.
Semler 8 E. Ehrenfeld, pp. 95-105, Washington, D. 6: American Society for
Microbiology.

Falk, B. W. 8 Duffus, J. E. (1981). Epidemiology of helper-dependent
persistent aphid transmitted virus complexes. In Plant Diseases and Vectors:
Ecology and Epidemiology. Edited by K. Maramorosh 8 K H. Harris, pp. 161-
179, New York: Academic Press.

Fargette, D., Jenniskens, M.-J. 8 Peters, D. (1982). Acquisition and
transmission of pea enation mosaic virus by the individual pea aphid.
Phytopathology 72, 1386-1390.

94

Flanegan, J. B. 8 Baltimore, D. (1979). Poliovirus polyuridylic acid
polymerase and RNA replicase have the same viral polypeptide. Journal of
Virology 29, 352-360.

Flanegan, J. 3., Pettersson, R. F., Ambros, V., Hewlett, M. J. 8 Baltimore, D.
(1977). Covalent linkage of a protein to a defined nucleotide sequence at the 5’-
terrninus of virion and replicative intermediate RNAs of poliovirus. Proceedings
of the National Academy of Sciences of the United States of America 74, 961 -
965.

Gabriel, C. J. (1983). Investigation into the in vitro translation and aphid
transmission of pea enation mosaic virus. Ph.D. thesis, University of Wisconsin,
Madison, WI.

Gabriel, C. J. 8 de Zoeten, G. A. (1984). The in vitro translation of pea enation
mosaic virus RNA. Virology 139, 223—230.

German, T. L. 8 de Zoeten, G. A. (1975). Purification and properties of the
replicative forms and replicative intermediates of Pea Enation Mosaic Virus.
Virology 66, 1 72-1 84.

German, T. L., de Zoeten, G. A. 8 Hall, T. C. (1978). Pea Enation mosaic virus
RNA contains no polyadenylate sequences and can not be aminoacylated.
lntervirolog y 9, 226-230.

Gildaw, F. E. 8 Rochow, W. F. (1980). Role of accessory salivary glands in
aphid transmission of barley yellow dwarf virus. Virology 104, 97-108.

Goldbach, R., Rezelman, G., Zabel, P. 8 van Kammen, A. (1982). Expression
of the bottom-component RNA of cowpea mosaic virus: evidence that the 60-
kilodalton VPg precursor is cleaved into single VPg and a 58-kilodalton
polypeptide. Journal of Virology 42, 630-635.

Gorbalenya, A. E. 8 Koonin, E. V. (1993). Comparative analysis of amino-acid

sequences of key enzymes of replication and expression of positive-strand RNA

viruses: validity of approach and functional and evolutionary implications. Soviet
Scientific Review, Section D: Physicochemical Biology 1 1 , 1-84.

Gash, A., Dasgupta, R., Salemo-Rife, T., Rutgers, T. 8 Kaesberg, P. (1979).
Southern bean mosaic virus RNA has a 5’-linked protein but lacks 3-terminal
poly(A). Nucleic Acids Research 7, 2137-2146.

Hagedom, D. J. (1996). Pea enation mosaic virus: ecology and control. In The
plant viruses. Volume 5: Polyhedral virions and bipartite RNA genomes, 1" edn,

95

pp. 345-356. Edited by B. D. Harrison 8 A. F. Murrant. New York: Plenum
Press.

Harlow, E. 8 D. Lane. (1988). Antibodies: a laboratory manual. 1". edn. Cold
Spring Harbor, N. Y.: Cold Spring Harbor Press.

Harris, K. F. 8 Bath, J. E. (1972). The fate of pea enation mosaic virus in its
pea aphid vector, Acyrthosiphon pisum (Harris). Virology 50, 778-790.

Harris, K. F., Bath, J. E., Thottappilly, G. 8 Hooper, G. R. (1975). Fate of pea
enation mosaic virus in PEMV-injected pea aphids. Virology 65, 148-162.

Harris, K. F. (1979). Leafhoppers and aphids as biological vectors: vector-virus
relationships. In Leafhopper Vectors and Plant Disease Agents, pp 217-308.
Edited by K Maramorosch 8 K F. Harris. New York: Academic Press.

Harris, K. F. (1990). Aphid transmission of plant viruses. In Plant Viruses, Vol.
2, Pathology, pp. 177-204. Edited by C. L. Mandahar. Boca Raton, FL: CRC
Press.

Harris, K. s.. Reddigarl, S. R., Nicklin, M. J. H., Hammerle, T. 8 Wimmer, E.
(1992). Purification and characterization of poliovirus polypeptide 300, a
proteinase and precursor for RNA polymerase. Journal of Virology 66, 7481-
7489.

Harrison, B. D. 8 Barker, H. (1978). Protease-sensitive structure needed for
infectivity of nepovirus RNA. Journal of General Virology 40, 711-715.

Hey, T. D., Richards, 0. C. 8 Ehrenfeld, E. (1987). Host factor-induced
template modification during synthesis of poliovirus RNA in vitro. Journal of
Virology 61, 802-81 1.

Hill, J. H. 8 Shepherd, R. J. (1972). Molecular weights of plant virus coat
proteins by polyacrylamide gel electrophoresis. Virology 47, 817-822.

Hull, R. 8 Lane, L. C. (1973). The unusual nature of the components of a strain
of pea enation mosaic virus. Virology 55, 1-13.

Jaegle, M., Wellink. J. 8 Goldbach, R. (1987). The genome-linked protein of
cowpea mosaic virus is bound to the 5’ terminus of virus RNA by a
phosphodiester linkage to serine. Journal of General Virology 68, 627-632.

Jayaram, C., Hill, J. H. 8 Miller, W. A. (1992). Complete nucleotide sequences
of two soybean mosaic virus strains differentiated by response of soybean
containing the Rsv resistance gene. Journal of General Virology 73, 2067-2077.

96

Koenig, l. 8 Fritsch. C. (1982). A protein linked at the 5’ end of satellite
and genomic tomato black ring virus RNAs: Study of in vitro translation after
protease treatment. Journal of General Virology 60, 343-353.

Koonin, E. V. (1991). The phylogeny of RNA-dependent RNA polymerases of
positive-strand RNA viruses. Journal of General Virology 72, 2197-2206.

Koonin, E. V. 8 Dolja, V. V. (1993). Evolution and taxonomy of positive-strand
RNA viruses: implications of comparative analysis of amino acid sequences.
Critical Reviews in Biochemistry and Molecular Biology 28, 375-430.

Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature 227, 680-685.

Laliberte, J.-F., Nicolas, D., Chatel, H., Lazure, C. 8 Morosoli, R. (1992).
Release of a 22-kDa protein derived from the amino-terminal domain of the 49-
kDa Nla of Turnip Mosaic Potyvirus in Escherichia coli. Virology 190, 510-514.

Lee, S 8 Rasheed, S. (1990). A simple procedure for maximum yield of high-
quality plasmid DNA. Biotechniques 9, 676-679.

Lomanossoff, G. P., Shanks, M. 8 Evans, D. (1985). The structure of cowpea
mosaic virus replicative form RNA. Virology 144, 351-362.

Long, A. C., Orr, D. C., Cameron, J. M., Dunn, B. M. and Kay, J. (1989). A
consensus sequence for substrate hydrolysis by rhinovirus 3C proteinase.
FEBS Letters 258, 75-78.

Liu, Y. Y., Cooper, J. l., Coates, D. 8 Bauer, G. (1991). Biologically active
transcripts of a large satellite RNA from arabis mosaic nepovirus and the
importance of 5’ end sequences for its replication. Journal of General Virology
72, 2867-2874.

Makinen, K., Tamm, T., Naess, V., Truve, E., Puurand, 0., Munthe, T.
Saanna, M. (1995). Characterization of cocksfoot mottle sobemovirus genomic

RNA and sequence comparison with related viruses. Journal of General
Virology 76, 2817-2825.

Margis, R., Hans, F. 8 Pinck, L. (1993). VPg northem-immunoblots as a
means for detection of viral RNAs in protoplasts or plants infected with
grapevine fanleaf virus. Archives of Virology 131, 225-232.

97

Margis, R., Viry, M., Pinck, M., Bardonnet, N. Pinck, L. (1994).
Differential proteolytic activities of precursor and mature forms of the 24K
proteinase of grapevine fanleaf nepovirus. Virology 200, 79-86.

Martin, R. R., Keese, P. K., Young, M. J., Waterhouse, P. M. 8 Gerlach, W. L.
(1990). Evolution and molecular biology of luteoviruses. Annual Reviews in
Phytopathology 28, 341-363.

Matthews, R. E. F. (1991). Plant Virology. 3"' ed. Academic Press, Inc. San
Diego, CA.

Matthews, D. A., Smith, W. W., Ferre, R. A., Candan, B., Budahazi, G.,
Sissan, W., Villafranca, J. E., Janson, C. A., McElroy, H. E., Gribskav, C. L.
8 Worland, S. (1994). Structure of human rhinovirus 3C protease reveals a

trypsin-like polypeptide fold, RNA binding site, and means for cleaving precursor
polyprotein. Cell 77, 761 -771.

Maya, M. A., Barker, H. 8 Harrison, B. D. (1982a). Specificity and properties of
the genome-linked proteins of nepoviruses. Journal of General Virology 59, 149-
162.

Mayo, M. A., Barker, H., Robinson, D. J., Tamada, T. 8 Harrison, B. D.
(1982). Evidence that potato leafroll virus RNA is positive-stranded, is linked to
a small protein and does not contain polyadenylate. Journal of General Virology
59, 163-167.

Maya, M. A. 8 Fritsch, C. (1994). A possible consensus for VPg of viruses in
the family Comoviridae. FEBS Letters 354, 129-130.

Mayo, M. A. 8 Ziegler-Graft, V. (1996). Molecular biology of luteoviruses.
Advances in Virus Research 46, 413-460.

Miller, W. A., Waterhouse, P. M. 8 Gerlach, W. L. (1988). Sequence and
organization of barley yellow dwarf virus genomic RNA. Nucleic Acids Research
16, 6097-611 1.

Miller, W. A. (1994). Luteoviruses. ln Ecyclopedia of Virology, 1" edn, pp. 792-
798. Edited by R. G. Webster 8 A. Granoff. London: Academic Press.

Miller, W. A., Dinesh-Kumar, S. P. 8 Paul, S. P. (1995). Luteovirus gene
expression. Critical Reviews in Plant Sciences 41 , 179-21 1.

98

Mirzayan, C. 8 Wimmer, E. (1994). Polioviruses: Molecular Biology. In
Encyclopedia of Virology, 1" edn, pp. 1119-1132. Edited by G. A. Webster 8 A.
Ganoff. London: Academic Press.

Meyer, M., Hammer, 0. 8 Fritsch, C. (1984). Complete nucleotide sequence of
a satellite RNA of tomato black ring virus. Journal of General Virology 65, 1575-
1 583.

Murphy, J. F., D’Arcy, C. J. 8 Clark, J. M. JR. (1989). Barley yellow dwarf
virus RNA has a 5' -terminal genome-linked protein. Journal of General Virology
70, 2253-2256.

Murphy, J. F., Rhoads, R. E., Hunt, A. G. 8 Shaw, J. G. (1990). The VPg of
tobacco etch virus RNA is the 49-kDa proteinase or the N-tenninal 24-kDa part
of the proteinase. Virology 178, 285-288.

Murphy, J. F., Rychlik, W., Rhoads, R. E., Hunt, A. G. 8 Shaw, J. G. (1991).
A tyrosine residue in the small nuclear inclusion protein of tobacco vein mottling
virus links the VP9 to the viral RNA. Journal of Virology 65, 5511-513.

Murphy, J. F., Klein. P. 6., Hunt, A. G. 8 Shaw, J. G. (1996). Replacement of
the tyrosine residue that links a potyviral VPg to the viral RNA is lethal. Virology
220, 535-538.

Nomoto, A., Detjen, B., Pozzatti, R. 8 Wimmer, E. (1977a). The location of
the polio genome protein in viral RNAs and its implication for RNA synthesis.
Nature 268, 208-213.

Nomoto, A., Kitamura, N., Gallni, F. 8 Wimmer, E. (1977b). The 5’ -terminal
structures of poliovirion RNA and poliovirus mRNA differ only in the genome-
linked protein VPg. Proceedings of the National Academy of Sciences of the
United States of America 74, 5345-5349.

Pallai, P. V., Burkhardt, F., Skoog, M., Schrelner, K., Bax, P., Cohen, K. A.,
Hansen, 6., Palladina, D. E. H., Harris, K. S., Nicklin, M. J. 8 Wimmer, E.
(1989). Cleavage of synthetic peptides by purified poliovirus 3C proteinase.
Journal of Biological Chemistry 264, 9738-9741.

Peters, 8. A., Voarharst, W. G. B., Wellink, J. 8 van Kammen, A. (1992).
Processing of VP-containing polyproteins encoded by the B-RNA from cow pea
mosaic virus. Virology 191, 90-97.

Peters, S. A., Mesnard, J.-M., Kaoter, l. M., Verver, J., Wellink, J. 8 van
Kammen, A. (1995). The cowpea mosaic virus RNA 1-encoded 112kDa protein

99

may function as a VPg precursor in viva. Journal of General Virology 76,
1 807-1 81 3.

Pettersson, R. F., Ambras, V. 8 Baltimore, D. (1978). Identification of a
protein linked to nascent poliovirus RNA and to the polyuridylic acid of negative-
strand RNA. Journal of Virology 27, 357-365.

Pinck, M., Reinbolt, J., Loudes, A. M., Le Ret, M. 8 L. Pinck. (1991). Primary
structure and localization of the genome-linked protein (VPg) of grapevine
fanleaf virus. FEBS Letters 284, 117-119.

Powell, C. A., de Zoeten, G. A. 8 Gaard, G. (1977). The localization of pea
enation mosaic virus-induced RNA-dependent RNA polymerase in infected
peas. Virology 78, 135-143.

Powell, C. A. 8 de Zoeten, G. A. (1977). Replication of pea enation mosaic
virus RNA in isolated pea nuclei. Proceedings of the National Academy of
Sciences of the United States of America 74, 2919-2922.

Randles, J. W. 8 Rathjen, J. P. (1995). Genus Luteovirus. In Virus Taxonomy
- Classification and Nomenclature of viruses. Sixth Report of the International
Committee of Taxonomy of Viruses, 1" edn, pp. 379-383. Edited by F. A.
Murphy, C. M. Fauquet, D. H. L. Bishop, 8. A. G. Jarvis, G. P. Martelli, M. A.
Mayo 8 M. D. Summer. Vienna: Springer Verlag.

Reisman, D. 8 de Zoeten, G. A. (1982). A covalently linked protein at the 5’
end of the genomic RNAs of Pea Enation Mosaic Virus. Journal of General
Virology 62, 187-190.

Restrepo-Hartwig, M. A. 8 Carrington, J. C. (1994). The tobacco etch
potyvirus 6-kilodalton protein is membrane associated and involved in viral
replication. Journal of Virology 68, 2388-2397.

Reutenauer, A., Ziegler-Graft, V., Lot, H., Scheidecker, D., Guilley, H.,
Richards, K. 8 Janard, J. (1993). Identification of beet western yellows
luteovirus genes implicated in viral replication and particle morphogenesis.
Virology 195, 692-699.

Riechmann, J. L., Lain, S. 8 Garcia, J. A. (1989). The genome-linked protein
and 55’ end RNA sequence of plum pox potyvirus. Journal of General Virology
70, 2785-2789.

Riechmann, J. L., Lain, S. 8 Garcia, J. A. (1993). Highlights and prospects of
potyvirus molecular biology. Journal of General Virology 73, 1-16.

100

Raoslnck, M. J., Sleat, D. 8 Palukaitis, P. (1992). Satellite RNAs of plant
viruses: structures and biological effects. Microbiological Reviews 56, 265-279.

Rost, B. 8 Sander, C. (1993). Prediction of protein structure at better than 70%
accuracy. Journal of Molecular Biology 232, 584-599.

Rost, B. 8 Sander, C. (1994a). Combining evolutionary information and neural
networks to predict protein secondary structure. Proteins 19, 55-72.

Rost, B. 8 Sander, C. (1994b). Conservation and prediction of solvent
accessibility in protein families. Proteins 20, 216-226.

Rost, B. (1996). PHD: predicting one-dimensional protein structures by profile
based neural networks. Methods in Enzymology 266, 525-539.

Sambraok, J., Fritsch, E. F. 8 Maniatis, T. (1989). Molecular Cloning: a
laboratory manual, 2"" edn, pp. 18.1 - 18.88. Cold Spring Harbor, N. Y.: Cold
Spring Harbor Laboratory Press.

Sanger, F., Nicklen, S 8 Caulsan, A. R. (1977). DNA sequencing with chain-
terminating inhibitors. Proceedings of the National Academy of Sciences of the
United States of America 74, 5463-5467.

Schaad, M. C., Haldeman-Cahill, R., Cronin, S. 8 Carrington, J. C. (1996).
Analysis of the VPg-proteinase (Nla) encoded by tobacco etch potyvirus: Effects
of mutations on subcellular transport, proteolytic processing, and genome
amplification. Journal of Virology 70, 7039-7048.

Semler, B. L., Kuhn, R. J. 8 Wimmer, E. (1988). Replication of the poliovirus
genome. ln RNA Genetics, pp. 23-48. Edited by E. Domingo, J. J. Holland 8 P.
Alquist. Boca Raton, Fl: CRC Press.

Shahabuddin, M., Shaw, R. G. 8 Rhoads, R. E. (1988). Mapping of the
tobacco vein mottling virus VPg cistron. Virology 163, 635-637.

Shepherd, R. J., Wakeman, R. J. 8 Ghabrial, S. A. (1968). Preparation and
properties of the protein and nucleic acid components of pea enation mosaic
virus. Virology 35, 255-267.

Stanley, J., Rattler, P., Davies, J. W., Zabel, P. 8 van Kammen, A. (1978). A
protein linked to the 5’ termini of both RNA components of the cowpea mosaic
virus genome. Nucleic Acids Research 5, 45-5-4522.

101

Stryer, L. (1988). Biochemistry. 3"I ed. W. H. Freeman and Company, New
York.

Tacke, E., Prtlfer, D., Schmitz, J. 8 Rahde, W. (1991). The potato leafroll
luteovirus 17K protein is a single-stranded nucleic acid-binding protein. Journal
of General Virology 72, 2035-2038.

Tacke, E., Schmltz, J., Prllfer, D. 8 Rahde W. (1993). Mutational analysis of
the nucleic-acid binding 17kDa phosphoprotein of potato leafroll luteovirus

identifies an amphipathic a—helix as the domain for protein/protein interactions.
Virology 197, 274-282.

Vance, V. B., Moore, D., Turpen, T. H., Bracker, A. 8 Hallowell, V. C. (1992).
The complete nucleotide sequence of pepper mottle virus genomic RNA:
Comparison of the encoded polyprotein with those of other sequenced
potyviruses. Virology 191, 19-30.

van der Wilk, F., Huisman, M. J., Comelissen, B. J. C., Huttinga, H. 8
Goldbach, R. (1989). Nucleotide sequence and organization of potato leafroll
virus genomic RNA. FEBS Letters 245, 51-56.

van Dyke, T. A. 8 Flanegan, J. B. (1980). Identification of poliovirus
polypeptide P63 as a soluble RNA - dependent RNA polymerase. Journal of
Virology 35, 732-740.

Vartapetian, A. B. 8 Bogdanov, A. A. (1987). Proteins covalently linked to viral
genomes. Progress in Nucleic Acid Research and Molecular Biology 34, 209-
251.

Wellink, J., Rezelman, G., Goldbach, R. 8 Beyreuther, K. (1986).
Determination of the proteolytic processing sites in the polyprotein encoded by
the bottom-component RNA of cowpea mosaic virus. Journal of Virology 59, 50-
58.

Wellink, J. 8 van Kammen, A. (1988). Proteases involved in the processing of
viral polyproteins. Archives of Virology 98, 1-26.

Wimmer, E. (1982). Genome-linked proteins of viruses. Cell 28, 199-201.

Yassi, M. N. A., Ritzenthaler, C., Brugidou, C., Fauquet, C. 8 Beachy, R. N.
(1994). Nucleotide sequence and genome characterization of rice yellow mottle
virus RNA. Journal of General Virology 75, 249-257.

102

Young, D. C., Tuschall, D. M. 8 Flanegan, J. B. (1985). Poliovirus RNA-
dependent RNA polymerase and host cell protein synthesize product RNA twice
the size of poliovirus RNA in vitro. Journal of Virology 54, 256-264.

Young, D. C., Dunn, B. M., Tobln, G. J. 8 Flanegan, J. B. (1986). Anti-VPg
antibody precipitation of product RNA synthesized in vitro by the poliovirus
polymerase and host factor is mediated by VPg on the poliovirus RNA template.
Journal of Virology 58, 715-723.

Ympa-Wong, M. F., Filman, D. J., Hagle, J. M. 8 Semler, B. L. (1988).
Structural domains of the poliovirus polyprotein are major determinants for
proteolytic cleavage at Gln-Gly pairs. The Journal of Biological Chemistry 263,
1 7846-1 7856.

Zabel, P., Moennan, M., van Straaten, F., Goldbach, R. 8 van Kammen, A.
(1982). Antibodies against the genome-linked protein VPg of Cowpea mosaic
virus recognize a 60,000-Dalton Precursor polypeptide. Journal of Virology 41,
1 083-1088.

Zalloua, P. A., Buzayan, J. M. 8 Bruening, G. (1996). Chemical cleavage of
5’-linked protein from tobacco ringspot virus genomic RNAs and characterization
of the protein-RNA linkage. Virology 219, 1-8.

"illllllllllllllllllllllll