Download Geminivirus Replication Origins Have a Modular

The Plant Cell, Vol. 6, 405-416, March 1994 O 1994 American Society of Plant Physiologists
Geminivirus Replication Origins Have a Modular
Organization
Elizabeth P. B. Fontes,ai‘ Heather J. Gladfelter,a Reenah L. Schaffer,bi2 lan T. D. Petty,b and
Linda Hanley-Bowdoin
a
Department of Biochemistry, P.O. Box 7622, North Carolina State University, Raleigh, North Carolina 27695-7622
Department of Microbiology, North Carolina State University, Raleigh, North Carolina 27695-7615
Tomato golden mosaic virus (TGMV) and bean golden mosaic virus (BGMV) are closely related geminiviruses with bipartite genomes. The A and B DNA components of each virus have cis-acting sequences necessary for replication, and
their A components encode trans-acting factors required for this process. We showed that virus-specific interactions
between the cis- and trans-acting functions are required for TGMV and BGMV replication in tobacco protoplasts. We
also demonstrated that, similar to the essential TGMV AL1 replication protein, BGMV AL1 binds specifically to its origin
in vitro and that neither TGMV nor BGMV AL1 proteins bind to the heterologous origin. The in vitro AL1 binding specificities of the B components were exchanged by site-directed mutagenesis, but the resulting mutants were not replicated
by either A component. These results showed that the high-affinity AL1 binding site is necessary but not sufficient for
virus-specific origin activity in vivo. Geminivirus genomes also contain a stem-loop sequence that is required for origin
function. A BGMV B mutant with the TGMV stem-loop sequence was replicated by BGMV A, indicating that BGMV AL1
does not discriminate between the two sequences. A BGMV B double mutant, with the TGMV AL1 binding site and stemloop sequences, was not replicated by either A component, indicating that an additional element in the TGMV origin
is required for productive interaction with TGMV AL1. These results suggested that geminivirus replication origins are
composed of at least three functional modules: (1) a putative stem-loop structure that is required for replication but does
not contribute to virus-specific recognition of the origin, (2) a specific high-affinity binding site for the AL1 protein, and
(3) at least one additional element that contributes to specific origin recognition by viral trans-acting factors.
INTRODUCTION
The geminiviruses are a large and diverse family of plantinfecting viruses that share a unique particle structure of fused
incomplete icosahedra and have genomes comprised of circular, single-stranded DNA (reviewed by Stanley, 1991;
Lazarowitz, 1992).Their genomes replicatevia double-stranded
DNA intermediates that are transcribed in the nuclei of infected
plant cells. Geminiviruses encode only a few proteins necessary for their replication and transcription and depend largely
on host factors to mediate these processes. Consequently,
geminiviruses are excellent model systems for investigating
both DNA replication and transcription in plant cells.
The geminivirus family can be divided into four subgroups
defined by their biological properties (Harrison, 1985) and molecular genetics. One of the largest subgroups comprises the
whitefly-transmitted geminiviruses that infect dicotyledonous
plants and have bipartite genomes (Haber et al., 1981). Their
Current address: Department of Biochemistry, BIOAGRO-CTQ,
Universidade Federal de Vicosa, 36570.000 Vicosa MG, Brazil.
2 Current address: Department of Forestry, NOrth Carolina State
University, Raleigh, NC 27695-8008.
To whorn correspondence should be addressed.
genomes are composed of two -2.6-kb DNA components,
designated A and 6,that are dissimilar in sequence except
for a highly conserved common region of approximately 200
nucleotides. The common region is located in a 5’ intergenic
region on both DNA components and includes replication
(Revington et al., 1989; Lazarowitz et al., 1992) and transcription (Townsend et al., 1985; Hanley-Bowdoin et al., 1988; Sunter
et al., 1989) signals. The common region also contains a GCrich inverted repeat that could form a cruciform or stem-loop
structure. (In this study, “stem-loop” refers to the sequence and
potential secondary structure, whose existence has not yet
been demonstrated.) An invariant AT-rich sequence, 5‘-TAATATTAC, in the loop is found in all geminivirus genomes. Mutations that either disrupt the invariant sequence or partially
delete the GC-rich stem sequence prevent replication in vivo
(Revington et al., 1989; Lazarowitz et al., 1992).
The A component of bipartite genome geminiviruses encodes all of the viral proteins required for replication and
encapsidation (Rogers et ai,, 1986; Townsend et
1986;
Sunter et
1987), whereas the component contributes functions necessary for the spread of infection in plants (Brough
et al., 1988; Etessami et al., 1988). Two genes, AL7 and AL3,
406
The Plant Cell
on the A component encode viral proteins involved in replication. The only viral protein essential for replication is AL1 (Elmer
et al., 1988; Hayes and Buck, 1989; Hanley-Bowdoin et al.,
1990), but AL3 is necessary for efficient accumulation of viral
DNA in plants (Elmer et al., 1988) and in protoplasts (Sunter
et al., 1990). There is strong amino acid sequence and functional conservation among the AL1 and AL3 proteins encoded
by geminiviruses with bipartite genomes (Howarth and
Vandemark, 1989; Etessami et al., 1991; Lazarowitz et al.,
1992). This conservation also extends to the AL1 homologs
encoded by geminiviruses with a single genome component
(Mullineaux et al., 1985; Accotto et al., 1989; Lazarowitz et al.,
1989; Schalk et al., 1989; Stanley et al., 1992).
Three independent lines of evidence show that geminiviruses reproduce their circular, single-stranded DNA genomes
by a rolling circle replication mechanism analogous to some
prokaryotic viruses and plasmids. First, the nucleotide sequences of beet curly top virus genomes generated from
recombinant heterodimers are consistent with rolling circle
replication that initiatesherminates (+) strand DNA synthesis
within the stem-loop sequence (Stenger et al., 1991). This possiblity is further supported by studies of African cassava mosaic
virus (ACMV) and wheat dwarf virus (Etessami et al., 1989;
Heyraud et al., 1993). Second, the intracellular DNAforms that
accumulate during ACMV infection correspond to rolling circle replication intermediates (Saunders et al., 1991). Third,
geminivirus AL1 proteins contain three amino acid sequence
motifs that are related to consensus motifs found in the replication initiator proteins of two bacterial plasmid families (Koonin
and Ilyina, 1992) and in the gene A protein of bacteriophage
(pX174 (Ilyina and Koonin, 1992). These prokaryotic proteins
introduce a sequence-specific nick in the (+) strand to initiatekerminate rolling circle replication. The AL1 protein may
serve the same function in geminivirus replication. AL3 does
not show homology to any proteins involved in rolling circle
replication, and the mechanism whereby AL3 acts as a
geminivirus replication accessory factor is unknown.
In spite of the strong functional conservation between the
trans-acting replication factors encoded by bipartite genome
geminiviruses, these proteins can show specificity for replication of their cognate genomes. The A and B genome
components are usually only infectious on plants when both
are derived from the same geminivirus. Studies carried out
with squash leaf curl virus (SqLCV) and tomato golden mosaic virus (TGMV) revealed that incompatibility between the
B component replication origin and frans-acting replication
factors encoded by the A component can be responsible
for the failure of heterologous DNA components to produce a
viable infection (Lazarowitz et al., 1992). These studies also
showed that an approximately 90-nucleotide fragment, which
includes the stem-loop sequence and 60 nucleotides from
the AL7-proximal (left) side of the SqLCV common region,
contains the replication origin and the determinant(s) required
for specific recognition of the origin by AL1 in vivo. An analogous fragment of the TGMV common region contains the
minimal replication origin for this virus (H.J. Gladfelter and
L. Hanley-Bowdoin, manuscript in preparation). Thus, SqLCV
and TGMV display exclusive origin recognition, but the region(s) of the replication origins that mediates specific
recognition and the mechanisms by which specificity is
achieved have not been elucidated. Other studies that established that TGMV AL1 binds specifically to a directly repeated
sequence on the left side of the TGMVcommon region in vitro
(Fontes et al., 1992, 1994) suggest a possible mechanism for
virus-specific, replication origin recognition.
In this study, we showed that TGMV and another closely
related geminivirus, bean golden mosaic virus (BGMV), also
exhibit strict specificity in the interactions between the cis- and
trans-acting factors required for their replication in vivo. Using
this system, we examined the role that AL1 plays in specific
recognition of the origin, both in vitro and in vivo. The results
of this investigation suggested that geminivirus replication origins consist of at least three functional modules.
RESULTS
TGMV and BGMV DNA B Components Are Replicated
Selectively by Their Homologous A Components in
Tobacco Cells
To establish an experimental system for investigating the determinants of geminivirus replication specificity, it was first
necessary to identify two closely related viruses that show
selective replication of their homologous origins. Comparison
of the nucleotide sequences of TGMV and BGMV revealed
that these viruses are closely related (Howarth et al., 1985).
However, it was not known whether TGMV and BGMV display
selectivity in the interaction between the cis- and trans-acting
factors required for their replication. It was also not known if
BGMV can replicate in the tobacco cell lines used in TGMV
transfection systems (Sunter et al., 1990; Brough et al.,
1992; H.J. Gladfelter and L. Hanley-Bowdoin, manuscript in
preparation).
To address these questions, partial tandem copies of TGMV
A and B or BGMVA and B were introduced by electroporation
into protoplasts derived from tobacco suspension cells and
assayed for their capacity to replicate autonomously (A components) or in combination with the homologous component
(6 components), as shown in Figure 1. Recombinant DNAs
containing partial tandem repeatsof geminivirus genomes support the release and replication of unit-length, circular viral
DNA in plant cells. Total DNA was isolated after a 48-hr culture period and analyzed by DNA gel blot hybridization by
sequentially probing with virus-specific radiolabeled DNAs under conditions that prevented cross-hybridization between
TGMV and BGMV DNA sequences. Cross-hybridization between the A and B components of the same virus was
prevented by excluding common region sequences from the
probes. Prior to DNA gel blot analysis, replicon DNA was
linearized, and the samples were treated with Dpnl, which
Geminivirus Replication Origins
TGMVA
TGMVB
BGMVA
+ - +
+
- + + - - - -
- - - - - +
+ - + +
BGMVB
-
-
-
-
+
+
+
-
TGMVA
ds-
I
ss-
TGMVB
dsss-
BGMVA
ds-
407
only digests dam-methylated, input plasmid DNA. Newly
synthesized, double-stranded DNA was identified by its resistance to Dpnl digestion and by length (2.6 versus 6 kb for input
DNA). The identity of newly synthesized, single-stranded DNA
was confirmed by its sensitivity to mung bean nuclease digestion (data not shown). Double- and single-stranded forms of
replicated viral DNA were readily detected for TGMV A (Figure 1, lanes 1, 3, and 4) and BGMV A (lanes 5, 7, and 8) either
in the presence or the absence of the homologous B component. In contrast, TGMV B only replicated in the presence of
TGMV A (Figure 1, cf. lanes 2 and 3), while BGMV B only replicated in the presence of BGMV A (cf. lanes 6 and 7). These
results established that BGMV is able to replicate efficiently
in tobacco cells and that, similar to TGMV A (Rogers et al.,
1986; Hayes and Buck, 1989), BGMV A provides all of the viral frans-acting factors required for replication.
The capacity of TGMV and BGMV A components to replicate the heterologous B components was examined in
cotransfection experiments shown in Figure 1. No replication
of BGMV B was detected in the presence of TGMV A (lane
4) or of TGMV B in the presence of BGMV A (lane 8). Replication of the heterologous A components in the same samples
(Figure 1, lanes 4 and 8) and of the B components in the presence of their homologous A component (lanes 3 and 7)
indicated that the protoplasts were competent for viral replication. Thus, the TGMV- and BGMV-encoded replication factors
displayed a high degree of specificity for their respective origins, making these two closely related viruses an ideal system
for investigating the molecular basis of replication specificity
in common host cells.
TGMV and BGMV AL1 Proteins Bind Specifically to
Their Cognate Origins of Replication in Vitro
ss-
We demonstrated previously that TGMV AL1 specifically binds
to a DNA sequence on the left side of the TGMV common region (Fontes et al., 1992) and more recently that this high-affinity
binding site is essential for TGMV replication in tobacco cells
(Fontes et al., 1994). The TGMV and BGMV AL1 proteins are
functionally equivalent, suggesting that BGMV AL1 may specifically bind to sequences in the BGMV common region. This
hypothesis was tested using an in vitro immunoprecipitation
DNA binding assay devised for TGMV AL1 (Fontesetal., 1992).
BGMVB
dsss1
2
3
4
5
6
7
8
Figure 1. Virus-Specific Interactions between the c/s- and frans-Acting
Functions Required for TGMV and BGMV Replication.
Plasmids containing partial tandem copies of TGMV A, TGMV B, BGMV
A, and BGMV B were electroporated into tobacco protoplasts in the
combinations indicated at the top. Transfection of a given DNA is
indicated by a (+), whereas the absence of a given DNA is indicated
by a (-). The transfected DNAs (20 ng of each) were pTG1.3A (lane
1), pTG1.4B(lane2), pTG1.3A + pTG1.46 (lane 3), pTG1.3A + pGA1.2B
(lane 4), pGA1.2A (lane 5), pGA1.2B (lane 6), pGA1.2A + pGA1.2B
(lane 7), and pGA1.2A + pTG1.4B (lane 8). Total DNA was isolated 2
days post-transfection, digested with Dpnl and Xhol (lane 1), Dpnl and
Bglll (lanes 2, 5 to 8), or Dpnl, Xhol, and Bglll (lanes 3 and 4). The
digested DNA was resolved on an agarose gel and analyzed by DNA
hybridization using 32P-labeled probes specific for the viral DNAs indicated on the left. Double-stranded (ds) and single-stranded (ss) forms
of the viral DNAs are marked.
408
The Plant Cell
1
2
3
-43
AL1 -
-30
B
Competitor DMAs
TGMV
BGMV
TGMV
BGMV
B
B
B-B1
B-T1
-
-
AL1 protein
TGMV
Prior to the binding studies, it was necessary to confirm that
a monoclonal antibody raised against TGMV AL1 was able to
cross-react efficiently with BGMV ALL Protein extracts from
leaves of healthy and BGMV-infected bean plants were
immunoprecipitated using a TGMV AL1-specific monoclonal
antibody. The precipitated proteins were resolved by SDSPAGE and detected by immunoblotting using a polyclonal antiserum raised against TGMV AL1, as shown in Figure 2A. Both
the monoclonal and polyclonal antibodies cross-reacted with
a protein from BGMV-infected plants (Figure 2A, lane 3). Three
lines of evidence indicated that this protein is BGMV AL1. First,
the protein is approximately 40 kD, which is the size predicted
by the BGMV AL1 open reading frame. Second, the BGMV
protein comigrated with the TGMV AL1 protein produced in
recombinant baculovirus-infected insect cells (Figure 2A, lane
1). Third, no cross-reacting material was detected in protein
extracts from healthy beans (lane 2), indicating that the protein is specific to BGMV-infected plants.
The TGMV AL1 monoclonal antibody was then used to immunoprecipitate TGMV AL1 from recombinant baculovirusinfected insect cell extracts and BGMV AL1 from infected bean
extracts. The immunocomplexes were incubated with 3' radiolabeled fragments containing either the TGMV B or the BGMV
B common region (Figure 2B, lane 1), and the bound DNA was
eluted and resolved on denaturing gels. TGMV AL1 bound to
the TGMV common region probe (TGMV; Figure 2B, lane 2),
BGMV
1 2 3 4 5 6 7
Figure 2. In Vitro DMA Binding by TGMV and BGMV AL1 Proteins.
(A) AL1 proteins were immunoprecipitated using a monoclonal antibody raised against TGMV AL1 and rabbit anti-mouse IgG beads. The
immunoprecipitated proteins were analyzed by SDS-PAGE followed
by immunoblotting using a rabbit polyclonal antibody raised against
TGMV ALL Immunoprecipitates from total soluble protein extracts of
recombinant baculovirus-infected Sf9 cells expressing TGMV AL1 (25
ng, lane 1), healthy beans (0.5 mg, lane 2), or BGMV-infected beans
(0.5 mg, lane 3) are shown. Molecular mass markers are given at right
in kilodaltons.
(B) Immunoprecipitation assays were used to examine the binding of
TGMV or BGMV AL1 proteins to their respective common regions, as
indicated on the left, in the presence of 50-fold molar excesses of the
competitor DMAs indicated by (+) at the top. A (-) indicates the absence of a competitor DMA. TGMV AL1 from recombinant
baculovirus-infected insect cells or BGMV AL1 from infected beans
was included in the binding reactions in lanes 2 to 6. The TGMV B
common region probe was a 3' end-labeled, 370-bp Ndel-Clal fragment from pTG0.4B (TGMV, lane 1). The BGMV B probe was a 3'
end-labeled, 489-bp BamHI-Hindlll fragment from pGA0.2B (BGMV,
lane 1). The probe DMAs were not immunoprecipitated prior to electrophoresis in lane 1 (TGMV and BGMV). In lane 2, the binding reactions
contained AL1 protein and its homologous probe DNA in the absence
of competitor DNA. The competitor DMAs in lanes 3 to 6 were a 370bp Ndel-Clal fragment from pTG0.4B (lane 3), a 489-bp BamHI-Hindlll
fragment from pGA0.2B (lane 4), a 368-bp Ndel-Clal fragment of pTGBB1 (lane 5), or a 491-bp BamHI-Hindlll fragment from pGAB-T1 (lane
6). Control extracts were incubated with probe DNA and electrophoresed
and binding was prevented by competition with a 50-fold molar excess of the unlabeled homologous probe (TGMV; lane
3). Similarly, BGMV AL1 bound to the BGMV common region
probe (BGMV; Figure 2B, lane 2), and binding was prevented
by competition with a 50-fold molar excess of the unlabeled
homologous probe (BGMV; lane 4). In contrast, neither TGMV
AL1 binding to the TGMV common region (TGMV; lane 4) nor
BGMV AL1 binding to the BGMV common region (BGMV; lane
3) was prevented by competition with a 50-fold molar excess
of the heterologous common region DNA. Control extracts lacking the AL1 proteins from recombinant baculovirus-infected
insect cells expressing p-galactosidase (TGMV; Figure 2B, lane
7) or from healthy beans (BGMV; lane 7) did not bind to the
common region probes. These results demonstrated that the
two AL1 proteins bind specifically to their respective common
regions and are unable to interact with the heterologous common region in vitro. The conditions used to wash the
immunocomplexes after DNA binding were stringent (Fontes
et al., 1992), indicating that both AL1 proteins bind to their recognition sequences with high affinity.
To define the BGMV AL1 recognition sequence, site-directed
mutagenesis was used to exchange the known TGMV AL1 binding site with the BGMV AL1 binding site predicted by nucleotide
sequence alignment and vice versa, as shown in Figure 3A.
In the mutant TGMV B-B1, the TGMV B common region was
in lanes 7. The TGMV control extract was from recombinant baculovirusinfected insect cells expressing p-galactosidase, and the BGMV control extract was from healthy bean leaves.
Geminivirus Replication Origins
- -
A
TGMV B-B1
TGMV B
409
...................CACGTGGCGGCCATCC~~MTATTACCGGATGGCCGCGCGAT
T A T G . M T T ~ T M ~ C T C T . T A T A T A T T A G A A G T T C C T A A G G G...................
G
CACGTGGCGGCtATCCmrrMTATTACCGGATGGCCGCGC~T
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.........................................
.........................................
TATC.M~ac~CTCT.TATATATTAGAAGTTCCTAAGGGG
BCMV B
TATCCAA~~ACAATATATACTAGTA.CCCTCAATCTCGTCAATTATCAGA~CACACACGTGGCGGCCATCCGATATMTATTACCGGATGGCCGCGCGCC
BCMV B - T 1
BCMV B-T4
BCMV B-T14
TATGCAATT~tao~ACAATATATACTAGTA.CCCTCAATCTCGT~TTAT~~TTCACACAC~GGCGGCCATCCGATATMTATTACCG~TGGCCGCGCGCC
TATGCM-CACMTATATACTAGTA
CCCTCAATCTCGTGAATTATCAGATTCACACACGTGGCGGC~TCCGtTt~MTATTACCGGATGGCCGCGCGCC
TATGCM'TT~too~ACAATATATACTAGTA.CCCTCAATCTCGTGAATTATATCAGATTCACACACGTGGCGGCCATCCGtTtTAATATTACCGGATGGCCGCGCGCC
.
I
B
I
-
TGMV
ACAAAAGTTA. . T A T G . A A T T m . T M a C T C T T A T A T A T T
-
BGMV
TCAAACTCTCGCTATGCAAW. .ACTBACAATATATACT
- 40
AbMV
TCAAACTTGCTC.ATGTA~T.ATTGGACGTCTllATATACT
BOMV
TCAAAACTATCTCATTCTm.mTACllATATACr
-
23
31
30
- 31
- 28
TCAACmCTCATATT.. G C T C T A T A T A T A C C - 41
T C A A A C l l C . C T C A T T C A A l l ~ T T A ~ C T T A T A T A T A- 4 1
BGMV-BZ TCAAAACTCr.CTATG.AATCGGIGT.AAT~C.CAATAAATAGT
WMV
SqLCV
ToMoV
TCAAAACTCT. . T A T G . A A w n C l T A T A T A G T
Figure 3. Comparison of the TGMV and BGMV AL1 Binding Sites with Other New World Geminivirus DNA Sequences
(A) The DNA sequences from TGMV 6, BGMV B, or their mutant derivatives are shown. The ALI binding site, putative stem-loop, and intervening
sequence of TGMV B are compared to the equivalent region of BGMV 6. The colons mark nucleotides that are conserved between the TGMV
B (nucleotide positions 64 to 157) and BGMV B (nucleotide positions 77 to 188) sequences. The repeated elements of the TGMV ALI binding
site and the equivalent sequences for BGMV are underlined. The inverted repeats that form the putative stem-loop structure are marked by bars
above the sequences. Mutations that alter the TGMV AL1 binding site to match BGMV (Bl) or the predicted BGMV ALI binding site to match
TGMV (TI) are shown by lowercase letters or by a dash for deleted bases. Mutations that alter the BGMV loop sequence to the TGMV loop sequence (T4) are also shown in lowercase letters. In BGMV BT14, the mutations of the BGMV ALI binding site and loop sequence to TGMV sequences
are combined. Gaps introduced to maximize the sequence alignments are marked by dots.
(B) Sequences from the TGMV and BGMV B common regions were compared to sequences on the left of the common regions of other bipartite
geminiviruses from the New World. The sequences were aligned using the predicted TATA box sequences (long bar) for transcription to the left.
Directly repeated elements are underlined, and conserved GG motifs are marked by short bars. The numbers on the right are the distance in
nucleotides to the stem-loop sequence for each virus. The viruses and their GenBank accession numbers are TGMV (K02030), BGMV (M91605),
abutilon mosaic virus (AbMV; X15984), bean dwarf mosaic virus (BDMV; M88180), BGMV-Brazil (-BZ; M88687), potato yellow mosaic virus (PYMV;
D00941), SqLCV (M63158), and tomato mottle virus (ToMoV; M90494).
modified at five positions to change the TGMV AL1 binding
site to the predicted B G M V A L l binding site. In mutant BGMV
BT1, the BGMV B common region was also altered at five positions to change the predicted B G M V A L l binding site to that
for TGMV ALI. Restriction fragments containing these mutated
common regions were used in 50-fold molar excess as competitors for AL1 binding to wild-type common region probes
in vitro (Figure 28). The TGMV B-B1 mutant did not compete
for TGMV AL1 binding to the wild-type TGMV common region
(TGMV; Figure 28, lane 5), while BGMV BT1 DNA was an efficient competitor for TGMV AL1 binding (TGMV; lane 6).
Conversely, the BGMV BT1 mutant was unable to compete
for BGMV AL1 binding to the wild-type BGMV common region
(BGMV; Figure 2B, lane 6), while TGMV B-B1 DNA was an efficient competitor for BGMV AL1 binding (BGMV; lane 5). Thus,
the introduced mutations altered the in vitro binding specificities of the common region fragments such that TGMV B-B1
is bound by BGMV AL1 and BGMV BT1 is bound by TGMV
AL1. These data demonstrated that the recognition sequence
required for high-affinity binding of BGMV AL1 to the BGMV
common region contains the repeated motif 5'-TGGAGACTGGAG, which is analogous to the TGMV AL1 binding site
5I-GGTAGTAAGGTAG (Fontes et al., 1994). It is possible that
the BGMV AL1 recognition sequence includes additional 5'
sequences, because the BGMV and TGMV B common regions
are identical at seven of eight positions immediately upstream
of the repeated motif (Figure 3A).
The HighAffinity AL1 Binding Site 1s Necessary but
Not Sufficient for Specific Geminivirus Origin
Recognition
The in vitro DNA binding experiments established that the
TGMV and BGMV AL1 proteins recognize related, but distinct,
high-affinity binding sites. The failure of either AL1 protein to
support the replication of a heterologous B component may
be due to the inability of AL1 to bind to the heterologous
410
The Plant Cell
common region, resulting in the replication specificity observed
in vivo. To test this hypothesis, we took advantage of the switch
in AL1 binding specificity exhibited by the TGMV B-B1 and
BGMV B-T1 mutants. Plasmids containing partial tandem copies of the mutant B components were cotransfected with each
of the wild-type A components into tobacco protoplasts and
assayed for replication by DNA hybridization, as shown in Figure 4. Consistent with their altered AL1 binding specificities
in vitro, TGMV B-B1 failed to replicate in the presence TGMV
A (Figure 4, lane 3; TGMV B probe), and no replication of
BGMV B-T1 was detected in the presence of BGMV A (lane
8; BGMV B probe). However, TGMV B-B1 also failed to replicate in the presence of BGMV A (lane 7; TGMV B probe), even
though BGMV AL1 binds to TGMV B-B1 in vitro. Similarly,
BGMV B-T1 did not replicate in the presence of TGMV A (Figure 4, lane 4; BGMV B probe), even though TGMV AL1 binds
to BGMV B-T1 in vitro. Both A components retained the capacity to replicate their homologous, wild-type B components
(Figure 4, lane 1, TGMV B probe; lane 6, BGMV B probe). In
addition, accumulation of TGMV A (lanes 1 to 4) or BGMV A
(lanes 5 to 8) was detected in all samples, indicating that the
protoplasts were competent for geminivirus replication.
These results showed that mutation of the high-affinity AL1
binding site prevented replication by the homologous AL1 protein, confirming the importance of this site. However, transfer
of the in vitro AL1 binding specificity was not sufficient to create a functional replication origin for the nonhomologous AL1
protein. These results established that specific binding of AL1
to its high-affinity site in the common region is necessary for
the replication of two different geminiviruses and, thus, contributes to the specificity of geminivirus origin recognition.
Nevertheless, because transfer of the high-affinity AL1 binding site alone does not result in productive interaction between
AL1 and the mutant origin, additional determinants of specificity must be required for geminivirus origin function.
Nonconserved Nucleotides in the Putative Stem-Loop
Do Not Mediate the Specific Interaction between
frans-Acting Replication Factors and the Origin
Lazarowitz et al. (1992) suggested that nonconserved nucleotides in the loop sequence of the stem-loop might be
responsible for the specificity of interaction between AL1 and
its cognate replication origin. The stem sequences of TGMV
and BGMV are identical, whereas their loop sequences differ
at only two nucleotide positions (Figure 3A). To directly assess
the role of these variant nucleotides in replication, site-directed
mutagenesis was used to alter the loop sequence of wild-type
BGMV B to that found in TGMV B, resulting in a mutant designated BGMV B-T4. The loop mutations were also introduced
into the BGMV B-T1 mutant, which contains the high-affinity
binding site for TGMV AL1, to generate a double mutant designated BGMV B-T14. Plasmids containing partial tandem copies
of these mutant B components were cotransfected into tobacco
cells with each of the wild-type A components and assayed
TGMV A
TGMVB
TGMV B-B1
BGMV A
BGMVB
BGMV B-T1
TGMV A
fflHM^^I»
TGMVB
BGMV A
BGMVB
1
2
3
4
5
6
7
8
Figure 4. Replication Analysis of the AL1 Binding Site Mutants.
Plasmids containing partial tandem copies of TGMV A, TGMV B, TGMV
B-B1 (binding site mutant), BGMV A, BGMV B, and BGMV B-T1 (binding site mutant) were electroporated into tobacco protoplasts as
indicated at the top. Transfection of a given DNA is indicated by a (+),
while the absence of a given DNA is indicated by a (-). In lanes 1
to 4, the transfected DMAs (20 ng of each) were pTG1.3A plus either
pTG1.4B (lane 1), pGA1.2B (lane 2), pTG1.3B-B1 (lane 3), or pGA1.2BT1 (lane 4). In lanes 5 to 8, the transfected DMAs were pGA1.2A plus
pTG1.4B (lane 5), pGA1.2B (lane 6), pTG1.4B-B1 (lane 7), or pGA1.2BT1 (lane 8). Total DNA was isolated 2 days post-transfection and digested
with Dpnl, Xhol, and Bglll (lanes 1 to 4) or Dpnl and Bglll (lanes 5
to 8). The digested DNA was resolved on an agarose gel and analyzed by DNA hybridization using 32P-labeled probes specific for the
viral DNAs indicated on the left. Double-stranded forms of replicated
viral DNA are shown.
for replication by DNA gel blot analysis, as shown in Figure
5. The loop mutant BGMV B-T4 was replicated by BGMV A (Figure 5, lane 7; BGMV B probe), but not by TGMV A (lane 3;
BGMV B probe), as was wild-type BGMV B (lanes 2 and 6;
BGMV B probe). In contrast, no replication of the double mutant BGMV B-T14 was detected in the presence of either TGMV
A (Figure 5, lane 4) or BGMV A (lane 8). DNA gel blot analysis
of TGMV A (Figure 5, lanes 1 to 4; TGMV A probe), BGMV
A (lanes 5 to 8; BGMV A probe), and TGMV B (lanes 1 and
4; TGMV B probe) verified that the protoplasts were competent for geminivirus replication. These results demonstrated
Geminivirus Replication Origins
that the loop sequence does not contribute significantly to the
ability of BGMV AL1 to specifically recognize the BGMV origin of replication. In addition, because the double mutant
BGMV B-T14 does not replicate in the presence of TGMV AL1,
the combination of the cognate high-affinity AL1 binding site
and the wild-type TGMV loop sequence was not sufficient to
confer TGMV replication specificity on the BGMV origin.
DISCUSSION
Although the geminiviruses TGMV and BGMV are closely
related, we showed that they exhibit strict specificity in the interactions between the c/s- and frans-acting factors necessary
TGMV A
+
TGMVB
+ . . - + - - -
+ + + -
-
-
-
BGMVA
BGMV B
BGMVB-T4
BGMVB-T14
. . . . + + + +
. + . - - + - - - + - - - + - - - + - - - +
TGMV A
TGMVB
BGMV A
BGMVB
1
2 3 4 5 6 7 8
Figure 5. Replication Analysis of the BGMV Loop Sequence Mutants.
Plasmids containing partial tandem copies of TGMV A, TGMV B, BGMV
A, BGMV B, BGMV B-T4 (loop mutant), and BGMV B-T14 (binding site
and loop mutant) were electroporated into tobacco protoplasts as indicated at the top. Transfection of a given DMA is indicated by a (+),
while the absence of a given DMA is indicated by a (-). In lanes 1
to 4, the transfected DMAs (20 ng of each) were pTG1.3A plus pTG1.4B
(lane 1), pGA1.2B (lane 2), pGA1.2B-T4 (lane 3), or pGA1.2B-T14 (lane
4). In lanes 5 to 8, the transfected DMAs were pGA1.2A plus pTG1.4B
(lane 5), pGA1.2B (lane 6), pGA1.2B-T4 (lane 7), or pGA1.2B-T14 (lane
8). Total DNA was isolated 2 days post-transfection and digested with
Dpnl, Xhol, and Bglll (lanes 1 to 4) or Dpnl and Bglll (lanes 5 to 8).
The digested DNA was resolved on an agarose gel and analyzed by
DNA hybridization using 32P-labeled probes specific for the viral DNAs
indicated on the left. Double-stranded forms of replicated viral DNA
are shown.
411
for their replication in vivo. One mechanism that might mediate this specificity is sequence-specific binding of AL1 to its
origin. This type of interaction governs origin recognition of
the rolling circle replicons in the pT181 plasmid family (Wang
et al., 1993) and probably the filamentous bacteriophages M13
and IKe (Peelers et al., 1986). We demonstrated previously that
TGMV AL1 binds with high affinity to a site in the TGMV common region (Fontes et al., 1992) and recently located this
binding site on a directly repeated motif near the left edge of
the minimal origin of replication (Fontes et al., 1994). In this
study, we showed that BGMV AL1 also binds specifically to
its common region in vitro. Furthermore, BGMV AL1 did not
bind to the TGMV common region, and TGMV AL1 did not
recognize the BGMV common region.
Alignment of the common region sequences from various
bipartite genome geminiviruses revealed that they frequently
contain a directly repeated motif with conserved GG dinucleotides in an analogous position to the TGMV AL1 high-affinity
binding site (see Figure 3B). We used this alignment to identify the likely binding site for BGMV AL1 in the BGMV common
region and constructed two B component mutants in which
the binding site sequences were exchanged between BGMV
and TGMV. The BGMV B-T1 mutant contains the TGMV AL1
high-affinity binding site in a wild-type BGMV common region
background, and the TGMV B-B1 mutant contains the predicted
BGMV AL1 high-affinity binding site in a wild-type TGMV common region background. We found that the in vitro AL1 binding
specificities of the mutant common regions were reversed relative to their wild-type parents, i.e., BGMV B-T1 bound
specifically to TGMV AL1, and TGMV B-B1 bound specifically
to BGMV ALL These results confirmed that the TGMV sequence 5'-GGTAGTAAGGTAG and the analogous BGMV
sequence 5'-TGGAGACTGGAG determine the binding specificity for the cognate AL1 protein.
The success of our nucleotide sequence alignment in
predicting the AL1 recognition motif for BGMV AL1 suggested
that the related, directly repeated sequences observed in other
geminivirus genomes also constitute high-affinity binding sites
for their respective AL1 proteins (Figure 3B). The sequence
heterogeneity of the predicted binding sites in different
geminiviruses makes it likely that AL1 proteins recognize their
origins in a highly specific manner. This suggestion is consistent with recent findings that the TGMV and abutilon mosaic
virus A components are also unable to support replication of
the heterologous B component (Frischmuth et al., 1993). Similar results were also obtained with TGMV and the Old World
geminivirus ACMV (Frischmuth et al., 1993).
We have shown previously by site-directed mutagenesis that
the high-affinity AL1 binding site forms an essential element
of the TGMV replication origin (Fontes et al., 1994). This observation raised the possibility that the AL1 recognition motif
is the sole determinant of origin recognition specificity. To test
this hypothesis, B component mutants, which have their in vitro
AL1 binding specificities exchanged (TGMV B-B1 and BGMV
B-T1), were analyzed for their ability to undergo DNA replication in tobacco protoplasts. Neither of the mutant B components
412
The Plant Cell
was replicated in the presence of either TGMV A or BGMV
A. This lack of replication cannot be attributed to a nonspecific
loss of replication competence by the mutant B components,
because both mutants can replicate in the presence of chimeric AL1 proteins (E.P.B. Fontes, H.J. Gladfelter, I.T.D. Petty,
and L. Hanley-Bowdoin, manuscript in preparation). Thus, as
previously shown for TGMV, the BGMV high-affinity AL1 binding site is an essential element of its replication origin (BGMV
BT1 is not replicated by BGMV A). However, because neither
of the mutant origins was replicated by the ALI protein to which
it bound efficiently in vitro, additional elements of the replication origin must be involved in specific interaction(s) with viral
trans-acting factors in vivo.
Lazarowitz et al. (1992) suggestedthat nonconsetved nucleotides in the loop of the stem-loop sequence in the common
region may be responsible for discrimination of the TGMV and
SqLCV origins by their respective AL1 proteins. The loop sequences of the TGMV and BGMV B components are also
distinct (Figure 3A) and therefore might have a role in origin
recognition specificity in vivo. To test this hypothesis, we used
site-directed mutagenesis to substitute the TGMV loop sequence into a wild-type BGMV common region background.
The resulting mutant, BGMV BT4, was able to replicate in the
presence of BGMV A, but not in the presence of TGMV A. Thus,
variation in the putative loop sequence did not contribute to
the ability of BGMV AL1 to discriminate between the BGMV
and TGMV replication origins. The TGMV loop sequence was
also substituted into the mutant BGMV BT1, which already
contains the TGMV AL1 high-affinity binding site in the BGMV
common region. The resulting double mutant, BGMV 6714,
had the same phenotype as its BGMV BT1 parent; i.e., it was
unable to replicate in the presence of either the BGMV or TGMV
A component. Because the high-affinity AL1 binding site and
stem-loop essentially delimit the TGMV minimal origin of replication (H.J. Gladfelter and L. Hanley-Bowdoin, manuscript in
preparation), it is likely that additional determinants required
for replication of the origin by TGMV-encoded, trans-acting factors are located between these two elements.
The replication assays with BGMV, TGMV, and their mutant
derivativeswere all performed in tobacco cells so that the specificity of origin recognition could not be mediated by host factors.
However, because the source of viral frans-acting factors in
these experiments was always a wild-type A component, a possible contribution of the accessory replication protein AL3 to
origin recognition specificity cannot formally be excluded.
Because the mutants with reversed in vitro ALI binding specificities cannot be replicated by either of the A components in
vivo, we can infer that if AL3 has any role in specific origin
recognition, it is unable to act in the absence of the appropriate AL1 binding site. Although AL3 is not essential for TGMV
replication, viral DNA accumulation is reduced m50-fold in its
absence (Sunter et al., 1990). Consequently, if a specific interaction between AL3 and the origin is disrupted in the
mutants, viral DNA accumulation might fall below detectable
levels. Two other factors could exacerbate this problem. First,
TGMV AL1 negatively regulates its own synthesis (Sunter et
al., 1993), and DNA replication may be limited by the supply
of AL1. Second, the replication origin of the A component that
supplies the trans-acting factors is wild type in character, and
it is in direct competition for AL1 with the mutant B component
origin being assayed. To address these potential limitations
of the replication assay, we are currently pursuing experiments
in which the viral frans-acting factors are supplied from nonreplicating expression cassettes.
The results presented here extend our understanding of the
structure and function of geminivirus replication origins and
represent an important step toward using these viruses to characterizethe nuclear DNA replication processesof higher plants.
By combining our results with those of previous studies, we
can now propose that the replication origins of TGMV, BGMV,
and related geminiviruses are modular in structure. Modular
organization is also associated with the (+) strand origins of
bacterial plasmids (de Ia Campa et al., 1990) and parvoviruses
(Snyder et al., 1993; Tam and Astell, 1993) and with the chromosomal origins of yeast (Marahrens and Stillman, 1992). The
geminivirus minimal origin contains at least three functional
elements, and, by analogy with other viral and plasmid systems, there may also be as yet unidentifiedreplicationenhancer
elements that lie outside the minimal origin (DePamphilis, 1988;
Gennaro, 1993). The three elements of the geminivirus minimal origin, shown schematically in Figure 6, are (1) a high-
invariant
sequence
a ATA
A T -n
r i i p
spacer
10000
AL1
binding
second specificity
determinant
,nickingr
virus-specific recognition
Figure 6. Modular Organizationof a GeminivirusOrigin of Replication.
A schematic drawing of the functional domains of a geminivirus origin of replication is shown. The relative locations of the ALI binding
site (hatched box) and the stem-loopmotif in the origin are indicated.
The invariant sequence and the AT-rich spacer motif in the loop are
marked. The limits of the DNA sequence that contains the probable
nick site for initiation of rolling circle replication mapped by Etessami
et al. (1989), Stenger et al. (1991), and Heyraud et al. (1993) are shown
(A).Other sites in the origin that may be involved in additional interactions with viral replicationproteins are illustratedby the open boxes.
Sites that may function in a sequence-specific manner are marked
by the large open rectangle, whereas specific interactions that may
be mediated by differential spacing are indicated by the small open
boxes.
Geminivirus Replication Origins
affinity binding site for the AL1 protein that is located on the
left side of the origin, (2) the putative stem-loop structure that
delimits the right side of the origin and contains the invariant
sequence 5'-TAATATTAC, and (3) an intervening sequence that
contains at least one element that must interact specifically
with viral frans-acting factors for replication to occur in vivo.
The high-affinity AL1 binding sites are essential for function of both the TGMV and BGMV replication origins, probably
because they promote the recruitment of AL1 to the origin.
Sequence-specific interaction between AL1 and the origin also
contributes to the ability of geminivirus trans-acting factors to
discriminate between the origins of closely related viruses. The
integrity of the putative stem-loop is essential for replication,
but the variant AT residues preceding the invariant 5'-TAATATTAC sequence in the loop apparently do not contribute to
the specificity of origin recognition. This arrangement is
reminiscent of the AT-rich spacer adjacent to the nick site in
the (+) strand origin of bacteriophage (pX174 in which nucleotide substitutions can be tolerated without compromising
replication (Heidekamp et al., 1981).The existence of an additional element that contributes to origin recognitionwas inferred
from the inability of the TGMV A component to replicate a
BGMV mutant that carries both the high-affinity AL1 binding
site and the stem-loop sequence of TGMV. The sequences of
the two origins differ in the region between these elements,
and the BGMV origin contains 19 additional nucleotides compared to that of TGMV (Figure 3A). Consequently, the additional
interactions with viral proteins required for origin function may
be mediated by sequence specificity, differential spacing, or
some combination of both, as has been observed for the (+)
strand origins of the pT181 plasmid family (Wang et al., 1993).
Experiments designed to distinguish between these possibilities for the geminivirus origins are currently in progress.
METHODS
Plasmid Constructs and Site-Dlrected Mutagenesis
The position numbers used to describe the following clones refer to
the nucleotide coordinates of the tomato golden mosaic virus (TGMV)
sequence determined by Hamilton et al. (1984), as corrected by
MacDowell et al. (1986), von Arnim and Stanley (1992), and Schaffer
(1992). The corrected size of full-length TGMV DNA B is 2525 bp. The
coordinates for bean golden mosaic virus (BGMV) refer to the sequence
of the Guatemalan isolate (BGMV-GA) determined by J.C. Faria, R.L.
Gilbertson, F.J. Morales, ?
! Ahlquist, and D.P. Maxwell (personal communication, University of Wisconsin, Madison; GenBank accession
number M91605). In these numbering schemes, the common regions
of both viruses are delimited by positions 1 to 210.
The construction of recombinant plasmids containing partial tandem repeats of the A or B components of TGMV and BGMV has been
described previously by Schaffer (1992). Briefly, plasmids pBH401
(Bisaro et al., 1982) and pBH604 (Hamilton et al., 1983), which contain single copies of TGMV A or B DNA, respectively, were used to
subclone fragments that included the TGMV A or B common region
413
into pZfl9U (Mead et al., 1986). The plasmid pTG0.3A contains the 736bp EcoRI-Xhol fragment from pBH401, while pTGOAB contains the 939bp Bglll-Clal fragment from pBH604. Partial tandem copies of TGMV
A and B were generated by insertion of the 2588-bp EcoRl fragment
from pBH401 into EcoRI-digested pTG0.3A and of the 2525-bp Clal
fragment from pBH604 into Clal-digested pTG0.4B. The resulting plasmids pTGl.3A and pTG1.48 contained 1.3 copies of TGMV DNA A or
1.4 copies of TGMV DNA 8, respectively, and had duplicated common
regions.
The plasmids pGAAl and pGAB1, which contain a single copy of
the A or B genome components of BGMV-GA, respectively, have been
described previously by Gilbertson et al. (1991).The 634-bp Bglll-Hindlll
fragment from pGAA1 was subcloned into pZfl9U to give pGAO.PA,
which contains the BGMV A common region and flanking sequences.
Similarly, the 489-bp BamHI-Hindlll fragment from pGABl was subcloned into pZfl9U to give pGAO.PB, which contains the BGMV B
common region and flanking sequences. A partial tandem copy of
BGMV A was constructed by releasing the 2647-bp EcoRl fragment
from pGAA1, circularizing it in vitro, followed by linearizationwith Spel,
and inserting it into the Spel site of pGAO.2A. A partial tandem copy
of BGMV B was generated by inserting the 2596-bp BamHl fragment
of pGABl into BamHI-digested pGAO.2B. The resulting plasmids
pGAl.2A and pGA1.28 contained 1.2 copies of BGMV DNA A or 1.2
copies of BGMV DNA B, respectively, and had duplicated common
regions.
Mutations were introduced into the common regions of plasmids
containing a single copy of TGMV B (pTGB1) or BGMV B (pGAB1) using
dU-containing, single-stranded DNA templates (Kunkel, 1985), as described previously (Petty et al., 1989). The plasmid pTGBl was made
by inserting the 2525-bp Clal fragment from pBH604 into the Clal site
of pBluescript II KS+ (Stratagene). The TGMV B common region was
modified at TGMV B positions 74, 77 to 79, and 82 using the primer
WGTTATATGAATTGGAGACTGGAGCTCTTATATATTAG-3' to create
pTGB-B1 (TGMV B-B1, Figure 3A). The BGMV B common region was
modified at positions 87, 90 to 92, and 94 using the primer 5'CGCTATGCAATTGGrAGlAAGCZACXACAATATATAC-3'
to create pGABT1 (BGMV BT1, Figure 3A), and at positions 161 and 163 using the
primer 5'-CACGTGGCGGCCATCCCZTTTAATATTACCGG-3' to create
pGABT4(BGMV BT4, Figure 3A). The T4 mutation was subsequently
introduced into pGABT1 to generate the double mutant pGABT14. All
mutations were verified by DNA sequencing. A subclone of pTGB-B1,
which contains the common region, was created by digestion with
BamHI. The resulting plasmid, pTGO.3B-B1, contained the 848-bp
BamHI-Clal fragment of TGMV B-81. A partial tandem copy of TGMV
B-B1was generated by inserting the 2523-bp Clal fragment of pTGBB1 into Clal-digested pTGO.3B-Bl. Partial tandem copies of BGMV
BT1, BGMV BT4, and BGMV BT14 were constructed from pGABT1,
pGABT4, and pGABT14, respectively, as described above for wildtype BGMV DNA B. The resulting plasmids, pTG1.38-81, pGA1.2BT1,
pGA1.2BT4, and pGA1.2BT14, contained duplicated, mutant common
regions.
lsolation and lmmunoprecipitationof AL1 Protein
Bean plants (Pbaseolus vulgaris cv Top Crop) were infectedwith BGMVGA bysap transmission. Systemicallyinfected leaf tissue or the equivalent from uninoculated control plants was harvested 12 days
postinfection and used to prepare whole-cell protein extracts, as described previously by Hanley-Bowdoinet al. (1990). Whole-cell protein
extracts were also prepared from healthy beans and from recombinant
414
The Plant Cell
baculovirus-infectedSpodopfem frugiperda Sf9 cells expressing TGMV
AL1 or b-galactosidase (Fontes et al., 1992).
AL1 protein was immunoprecipitatedfrom 5 mg of plant cell extract
or 25 pg of insect cell extract using 20 pg of an AL1-specific monoclona1 antibody (Fontes et al., 1992) and 20 pg of rabbit anti-mouse IgG
coupled to Sepharose beads (Harlow and Lane, 1988). The immunocomplexes were fractionated by SDS-PAGE and analyzed by
immunoblotting using a rabbit polyclonal antibody raised against ALI
(Hanley-Bowdoinet al., 1990) and the enhanced chemiluminescence
protein detection system (Amersham). Alternatively, the immunoprecipitatedAL1 protein was resuspended in binding buffer for DNA binding
assays.
DNA Binding Assays
DNA binding assays were performed as described by Fontes et al.
(1992). The immunocomplexes were incubated with 1 nM of radiolabeled probe DNA (-60,000 cpm) in the absence or presence of 50
nM competitor DNA. Bound DNA was extractedfrom the DNA/ALl immunocomplexes using phenol-CHC13-isoamylalcohol and visualized
on denaturing, 6% polyacrylamidegels. Probe and competitor DNAs
were prepared by restriction enzyme digestion followed by fractionation on agarose gels and purification by glass adhesion (Vogelstein
and Gillespie, 1979). A 370-bp Ndel-Clalfragment from pTG0.46, which
includesthe wild-type TGMV B common region and flanking sequences,
and a489-bp BamHI-Hindlll fragment from pGA0.2B, which includes
the wild-type BGMV B common region and flanking sequences, were
radiolabeled using a-=P-dATP and the Klenow fragment of DNA polymerase Ifrom Escherichiaco/i(Sambrooket al., 1989); these fragments
were used as probes in DNA binding experiments. Competitor DNAs
were isolated by Ndel-Clal digestion of pTG0.46 (wild-type TGMV B)
and pTGB-B1 (mutant TGMV 8-61) and by BamHI-Hindlll digestion of
pGAO.26 (wild-type BGMV B) and pGABT1 (mutant BGMV BT1).
Replication Assays
Protoplasts were prepared from the Nicotiana tabacum suspension
cell line NT-1 and cultured in NT-1 medium supplemented with 0.2
mg/mL 2,4-dichlorophenoxyaceticacid (An, 1985). The protoplasts were
isolated by digestion with 1% cellulase, 0.1% pectolyase Y30.4 M mannitol, 20 mM 2-(N-morpholino)ethanesulfonic acid, pH 5.5; washed twice
with 0.4 M mannitol, pH 5.5; and resuspended in 0.8% NaCI, 0.020/0
KCI, 0.02% KH2P04,0.11Vo Na2HP04,0.4 M mannitol, pH 6.5. Replication assays were performed by electroporation (250 V, 500 pF) of
20 pg of each replicon DNA and 30 pg of sheared salmon sperm DNA
into 5.0 x 106 protoplasts. Protoplasts were diluted into 8 mL of NT-1
medium supplementedwith 0.2 mg/mL 2,4-dichlorophenoxyaceticacid
and 0.4 M mannitol. Total DNA was isolated from the protoplasts 48
hr after transfection(Junghansand Metzlaff, 1990), digestedwith XholDpnl (TGMV A) or Bglll-Dpnl (TGMV B, BGMV A, and BGMV B),
resolved on 1% agarose gels, and vacuum transferred and UV crosslinked to nylon membranes (MagnaGraph; MSI, Westboro. MA). The
blots were analyzed by DNA gel blot hybridizationaccording the protocol
of Thomashow et al. (1980), except that 0.2% sodium pyrophosphate
was included in all solutions. Virus- and component-specificprobes
were randomly labeled using U-~'P-~ATP
and the Klenow fragment
(Sambrook et al., 1989). The probes were a 1.8-kb Xhol-EcoRI fragment from TGMV A, a 2.1-kb Pstl fragment from TGMV B, a 1.4-kb
Scal-Bglll fragment from BGMVA, and a 1.6-kb Hindlll fragment from
BGMV B.
ACKNOWLEDGMENTS
We thank Doug Maxwell and his colleagues (University of Wisconsin,
Madison) for generously providing the plasmids pGAAl and pGAB1
and complete nucleotide sequence data for BGMV-GA prior to publication. We also thank Verne Luckow (Monsanto Company, St. Louis,
MO) for production of TGMV ALI in recombinant baculovirus-infected
insect cells and Keith Everett for oligonucleotide synthesis. We thank
Ron Sederoff (Forestry, North Carolina State University) for critical reading of the manuscript. BGMV-infectedbeans were maintained in the
South Eastern Plant Environment Laboratory (Phytotron) at North Carolina State University. We are indebted to Daniel Bowdoin for the
photography. This work was supported in part by National Science
Foundation Grant No. DMB-9104150 to L.H.43. and in part by Public
Health Service Grant No. GM-48067 to I.T.D.P. E.P.B.F. was supported by a Conselho Nacional de Desenvolvimento Cientifico e
Technoldgicopostdoctoral fellowship from the Brazilian Government.
Received October 19, 1993; accepted January 11, 1994.
REFERENCES
Accotto, G.P., Donson, J., and Mullineaux, P.M. (1989). Mapping
of Digitaria streak virus transcripts reveals different RNA species
from the same transcription unit. EMBO J. 8, 1033-1039.
An, G. (1985). High efficiency transformationof cultured tobacco cells.
Plant Physiol. 79, 568-570.
Bisaro, D.M., Hamilton, W.D.O., Coutts, R.H.A., and Buck, K.W.
(1982). Molecular cloning and characterizationof the two DNA components of tomato golden mosaic virus. Nucl. Acids Res. 10,
4913-4922.
Brough,C.L.,Hayes, R.J., Morgan,A.J.,Coutts, R.H.A.,andBuck,
K.W. (1988). Effects of mutagenesis in vitro on the ability of cloned
tomato golden mosaic virus DNA to infect Nicotiana benthamiana
plants. J. Gen. Virol. 69, 503-514.
Brough, C.L., Sunter, G., Gardiner, W.E., and Bisaro, D.M. (1992).
Kinetics of tomato golden mosaic virus DNA replication and coat
protein promoter activity in Nicotianatabacum protoplasts.Virology
187, 1-9.
de Ia Campa, A.G., de1 Solar, G.H., and Espinosa, M. (1990). Initiation of replication of plasmid pLS1: The initiator protein RepB acts
on two distant DNA regions. J. MOI. Biol. 213, 247-262.
DePamphilis, M.L. (1988). Transcriptional elements as components
of eukaryotic origins of DNA replication. Cell 52, 635-638.
Elmer, J.S., Brand, L., Sunter, G., Gardiner, W.E.,'Bisaro, D.M., and
Rogers, S.G. (1988). Genetic analysis of the tomato golden mosaic
virus. 11. The product of the AL1 coding sequence is required for
replication. Nucl. Acids Res. 16, 7043-7060.
Etessami, P., Callis, R., Ellwood, S., and Stanley, J. (1988). Delimitation of essential genes of cassava latent virus DNA 2. Nucl. Acids
Res. 16, 4811-4829.
Etessami, P., Watts, J., and Stanley, J. (1989). Size reversion of African cassava mosaic virus coat protein gene deletion mutantsduring
infection of Nicotiana benthamiana. J. Gen. Virol. 70, 277-289.
Geminivirus Replication Origins
Etessami, P., Saunders, K., Watts, J., and Stanley, J. (1991). Mutational analysis of complementary-sense genes of African cassava
mosaic virus DNA A. J. Gen. Virol. 72, 1005-1012.
Fontes, E.P.B., Luckow, V.A., and Hanley-Bowdoin, L. (1992). A
geminivirus replication protein is a sequence-specific DNA binding
protein. Plant Cell 4, 597-608.
Fontes, E.P.B., Eagle, P.A., Sipe, P., Luckow, V.A., and HanleyBowdoin, L. (1994). lnteraction between a geminivirus replication
proteinand origin DNA is essentialfor viral replication.J. Biol. Chem.
269, in press.
Frischmuth, T., Roberts, S., von Arnlm, A., and Stanley, J. (1993).
Specificity of bipartite geminivirus movement proteins. Virology 196,
666-673.
Gennaro, M.L. (1993). Genetic evidence for replication enhancement
from a distance. Proc. Natl. Acad. Sci. USA 90, 5529-5533.
Gilbertson, R.L., Faria, J.C., Hanson, S.F., Morales, F.J., Ahlquist,
P., Maxwell, D.P., and Russell, D.R. (1991). Cloning of the complete DNA genomes of four bean-infecting geminiviruses and
determiningtheir infectivity by electric discharge particle acceleration. Phytopathology 81, 980-985.
Haber, S., Ikegami, M., Bajet, N.B., and Goodman, R.M. (1981). Evidente for a divided genome in bean golden mosaic virus, a
geminivirus. Nature 289, 324-326.
Hamilton, W.D.O., Bisaro, D.M., Coutts, R.H.A., and Buck, K.W.
(1983). Demonstration of the bipartite nature of the genome of a
single-stranded DNA plant virus by infection with the cloned DNA
components. Nucl. Acids Res. li,7387-7396.
Hamilton, W.D.O., Stein, V.E., Coutts, R.H.A., and Buck, K.W. (1984).
Complete nucleotide sequence of the infectious cloned DNA components of tomato golden mosaic virus: Potential coding regions
and regulatory sequences. EMBO J. 3, 2197-2205.
Hanley-Bowdoin, L., Elmer, J.S., and Rogers, S.G. (1988). Transient
expression of heterologous RNAs using tomato golden mosaic virus. Nucl. Acids Res. 16, 10511-10528.
Hanley-Bowdoin, L., Elmer, J.S., and Rogers, S.G. (1990). Expression of functional replicationproteinfrom tomato golden mosaic virus
in transgenic tobacco plants. Proc. Natl. Acad. Sci. USA 87,
1446-1450.
Harlow, E., and Lane, D. (1988). Antibodies, a Laboratory Manual.
(Cold Spring Harbor, NY Cold Spring Harbor Laboratory Press).
Harrison, B.D. (1985). Advances in geminivirus research. Annu. Rev.
Phytopathol. 23, 55-82.
Hayes, R.J., and Buck, K.W. (1989). Replicationof tomatogolden mosaic virus DNA B in transgenic plants expressing open reading
frames (ORFs) of DNA A: Requirementfor ORF AL2 for production
of single-stranded DNA. Nucl. Acids Res. 17, 10213-10222.
Heidekamp, F., Baas, P.D., van Boom, J.H., Veeneman, G.H.,
Zipursky, S.L., and Jansz, H.S. (1981). Construction and characterization of recombinant plasmid DNAs containing sequences of
the origin of bacteriophage X174 DNA replication. Nucl. Acids Res.
9, 3335-3354.
Heyraud, F., Maitzeit, V., Kammann, M., Schaefer, S., Schell, J.,
and Gronenborn, B. (1993) ldentification of the initiation sequence
for viral-strand DNA synthesis of wheat dwarf virus. EMBO J. 12,
4445-4452.
Howarth, A.J., and Vandemark, G. (1989). Phylogeny ofgeminiviruses.
J. Gen. Virol. 70, 2717-2727.
Howarth, A.J., Caton, J., Bossert, M.,and Goodman, R.M. (1985).
Nucleotidesequence of bean golden mosaic virus and a model for
415
gene regulation in geminiviruses. Proc. Natl. Acad. Sci. USA 82,
3572-3576.
Ilyina, T.V., and Koonin, E.V. (1992). Conserved sequence motifs in
the initiator proteins for rolling circle DNA replication encoded by
diverse replicons from eubacteria, eucaryotes and archasbacteria.
Nucl. Acids Res. 20, 3279-3285.
Junghans, H., and Metzlaff, M. (1990). A simple and rapid method
for the preparation of total plant DNA. BioTechniques 8, 176.
Koonin, E.V., and Ilyina, T.V. (1992). Geminivirus replication proteins
are related to prokaryotic plasmid rolling circle DNA replication initiator proteins. J. Gen. Virol. 73, 2763-2766.
Kunkel, T.A. (1985). Rapid and efficientsite-specific mutagenesiswithout phenotypic selection. Proc. Natl. Acad. Sci. USA 82,482-492.
Lazarowitz, S. (1992). Geminiviruses: Genome structure and gene
function. Crit. Rev. Plant Sci. 11, 327-349.
Lazarowitz, S.G., Pinder, A.J., Damsteegt, V.D., and Rogers, S.G.
(1989). Maize streak virus genes essential for systemic spread and
symptom development. EMBO J. 8, 1023-1032.
Lazarowitz, S.G., Wu, L.C., Rogers, S.G., and Elmer, J.S. (1992).
Sequence-specific interaction with the viral ALI protein identifies
a geminivirus DNA replication origin. Plant Cell 4, 799-809.
MacDowell,S.W., Coutts, R.H.A., and Buck, K.W. (1986). Molecular
characterization of subgenomic single-stranded and double-stranded
DNAforms isolated from plants infectedwith tomato golden mosaic
virus. Nucl. Acids Res. 14, 7967-7984.
Marahrens, Y., and Stillman, B. (1992). A yeast chromosomal origin
of DNA replication defined by multiple functional elements.Science
255, 817-823.
Mead, D.A., Szczesna-Skorupa, E., and Kemper, B. (1986). Singlestranded DNA 'blue' T7 promoter plasmids: A versatile tandem promoter system for cloning and protein engineering. Protein Eng. l ,
67-74.
Mullineaux, P.M., Donson, J., Stanley, J., Boulton, M.I., MorrisKrisinich, B.A.M., Markham, P.G., and Davies, J.W. (1985). Computer analysis identifies sequence homologies between potential
gene products of maize streak virus and those of cassava latent virus and tomato golden mosaic virus. Plant MOI. Biol. 5, 125-131.
Peeters, B.P.H., Schoenmakers, J.G.G., and Konings, R.N.H. (1986).
The gene II proteins of the filamentous phages IKe and Ff (M13, fd,
and fl) are not functionally interchangeableduring viral strand replication. Nucl. Acids Res. 14, 5067-5079.
Petty, I.T.D., Hunter, B.G., Wei, N., and Jackson, A.O. (1989). Infectious barley stripe mosaic virus RNA transcribed in vitro from
full-length genomic cDNA clones. Virology I7l, 342-349.
Revington, G.N., Sunter, G., and Bisam, D.M. (1989). DNAsequences
essential for replication of the B genome component of tomato golden
mosaic virus. Plant Cell 1, 985-992.
Rogers, S.G., Bisaro, D.M., Horsch, R.B., Fraley, R.T., Hoffman,
N.L., Brand, L., Elmer, J.S., and Lloyd, A.M. (1986). Tomatogolden
mosaic virus A component DNA replicates autonomously in transgenic plants. Cell 45, 593-600.
Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual. (Cold Spring Harbor, NY Cold Spring
Harbor Laboratory Press).
Saunders, K., Lucy, A., and Stanley, J. (1991). DNA forms of the
geminivirus African cassava mosaic virus consistent with a rolling
circle mechanism of replication. Nucl. Acids Res. 19, 2325-2330.
Schaffer, R.L. (1992). Host range determination of geminiviruses. M.S.
Thesis (Raleigh: North Carolina State University).
416
The Plant Cell
Schalk, H.J., Matzeit, V., Schiller, B., Schell, J., and Gronenborn,
B. (1989). Wheat dwarf virus, a geminivirus of graminaceous plants,
needs splicing for replication. EMBO J. 8, 359-364.
Sunter, G., Hartitz, M.D., and Bisaro, D.M. (1993). Tomato golden
mosaic virus leftward gene expression: Autoregulation of geminivirus replication protein. Virology 195, 275-280.
Snyder, R.O., Im, D A . , Ni, T., Xiao, X., Samulski, R.J., and
Muzyczka, N. (1993). Features of the adeno-associated virus origin
involved in substrate recognition by the viral Rep protein. J. Virol.
67, 6096-6104.
Tam, P., and Astell, C.R. (1993). Replication of minute virus of mice
minigenomes: Nove1 replication elements required for MVM DNA
replication. Virology 193, 812-824.
Stanley, J. (1991) The molecular determinam of geminivirus pathogenesis. Seminars Virol. 2, 139-149.
Thomashow, M.F., Nutter, R., Montoya, A.L., Gordon, M.P., and
Nester, E.W. (1980). lntegration and organization of Ti plasmid sequences in crown gall tumors. Cell 19, 729-739.
Stanley, J., Latham, J.R., Pinner, M.S., Bedford, I.,and Markham,
P.G. (1992). Mutational analysis of the monopartite geminivirus beet
curly top virus. Virology 191, 396-405.
Townsend, R., Stanley, J., Curson, S.J., and Short; M.N. (1985).
Major polyadenylated transcripts of cassava latent virus and location of the gene encoding coat protein. EMBO J. 4, 33-37.
Stenger, D.C., Revington, G.N., Stevenson, M.C., and Bisaro, D.M.
(1991). Replicational release of geminivirus genomes from tandemly
repeated copies: Evidence for rolling-circlereplication of a plant viral DNA. Proc. Natl. Acad. Sci. USA 88, 8029-8033.
Townsend, R., Watts, J., and Stanley, J. (1986). Synthesis of viral
DNA forms in Nicotianaplumbaginifolia protoplasts inoculated with
cassava latent virus (CLV): Evidence for the independent replication of one component of the CLV genome. Nucl. Acids Res. 14,
1253-1265.
Sunter, G., Gardiner, W.E., Rushing, A.E., Rogers, S.G., and Bisaro,
D.M. (1987). lndependent encapsidation of tomato golden mosaic
virus A component DNA in transgenic plants. Plant MOI. Biol. 8,
477-404.
Vogelstein, B., and Gillespie, D. (1979). Preparative and analytical
purification of DNA from agarose. Proc. Natl. Acad. Sci. USA 76,
615-619.
Sunter, G., Gardiner, W.E., and Bisaro, D.M. (1989). ldentification
of tomato golden mosaic virus-specific RNAs in infected plants.VirolOgy 170, 243-250.
von Arnim, A., and Stanley, J. (1992). Determinants of tomato golden
mosaic virus symptom developmentlocated on DNA 6.Virology 186,
286-293.
Sunter, G., Hartitz, M.D., Hormuzdi, S.G., Brough, C.L., and Bisaro,
D.M. (1990). Genetic analysis of tomato golden mosaic virus: ORF
AL2 is required for coat protein accumulationwhile ORFAL3 is necessary for efficient DNA replication. Virology 179, 69-77.
Wang, P.Z., Projan, S.J., Henriquez, V., and Novick, R.P. (1993). Origir! recognition specificity in pTl8l plasmids is determined by a
functionally asymmetric palindromic DNA element. EMBO J. 12,
45-52.
Geminivirus replication origins have a modular organization.
E P Fontes, H J Gladfelter, R L Schaffer, I T Petty and L Hanley-Bowdoin
Plant Cell 1994;6;405-416
DOI 10.1105/tpc.6.3.405
This information is current as of August 3, 2017
Permissions
https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532
298X
eTOCs
Sign up for eTOCs at:
http://www.plantcell.org/cgi/alerts/ctmain
CiteTrack Alerts
Sign up for CiteTrack Alerts at:
http://www.plantcell.org/cgi/alerts/ctmain
Subscription Information
Subscription Information for The Plant Cell and Plant Physiology is available at:
http://www.aspb.org/publications/subscriptions.cfm
© American Society of Plant Biologists
ADVANCING THE SCIENCE OF PLANT BIOLOGY