Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Reprogramming plant gene expression: a prerequisite to geminivirus DNA replication

Reprogramming plant gene expression: a prerequisite to geminivirus DNA replication INTRODUCTION The Geminiviridae family is classified into four genera, the mastreviruses, begomoviruses, curtoviruses and topocuviruses, based on genome structure, insect vector and host. Together, the four genera infect a broad range of plants and cause significant crop losses worldwide ( Mansoor ., 2003 ; Morales and Anderson, 2001 ; Rybicki and Pietersen, 1999 ). All geminiviruses are characterized by twin icosahedral capsids and single‐stranded DNA genomes that replicate through double‐stranded DNA intermediates in infected cells ( Hanley‐Bowdoin ., 1999 ). They contribute only a few factors for their replication and depend on the nuclear DNA polymerases of their plant hosts. This review focuses on recent studies addressing how geminiviruses modify host gene expression to enable efficient viral DNA replication in infected plant cells. GEMINIVIRUS REPLICATION: THE PLAYERS Geminiviruses amplify their small DNA genomes using a combination of rolling circle and recombination‐mediated replication ( Gutierrez, 1999 ; Jeske ., 2001 ). They provide the proteins required for initiation of replication and recruitment of host replication machinery. Members of the begomovirus, curtovirus and topocuvirus genera encode two proteins required for efficient viral replication ( Hanley‐Bowdoin ., 1999 ). AL1 (also designated AC1, C1 and Rep) is the initiation factor that mediates origin recognition and DNA cleavage/ligation to begin and end rolling circle replication ( Fontes ., 1994 ; Laufs ., 1995 ). AL3 (also designated AC3, C3 and REn) facilitates the accumulation of high levels of viral DNA ( Sunter ., 1990 ), possibly by modifying the activity of AL1 and/or aiding in the recruitment of the host replication enzymes ( Castillo ., 2003 ; Settlage ., 1996 ). Mastreviruses specify two replication‐associated proteins, RepA and Rep, which are translated from mRNA splicing variants. The Rep protein is the functional homologue of AL1 ( Heyraud‐Nitschke ., 1995 ). The RepA and AL3 proteins are not related and may serve different functions during infection. Several host factors have been implicated in geminivirus replication. Viral DNA accumulation is blocked by aphidicolin ( Nagar ., 2002 ), an inhibitor of eukaryotic DNA polymerases α and δ. Geminivirus replication proteins interact with two factors associated with Pol δ: the processivity factor, proliferating cell nuclear antigen (PCNA), and its clamp loader, RFC ( Castillo ., 2003 ; Luque ., 2002 ). These interactions probably represent early steps in the assembly of a DNA replication complex on the geminivirus origin. Geminivirus replication is also blocked by hydroxyurea ( Nagar ., 2002 ), which inhibits ribonucleotide reductase, a host enzyme required for deoxyribonucleotide synthesis and maintenance of dNTP precursor pools. THE DILEMMA: THE SOURCE OF HOST REPLICATION MACHINERY In mature plants, DNA replication and the corresponding enzymes are confined to meristematic and endoreduplicating tissues. Geminiviruses are excluded from meristems and do not have access to the host replication machinery in these dividing cell populations ( Peele ., 2001 ). They can infect young tissues undergoing endoreduplication but are also found in plant tissues at later stages of development. Some geminiviruses are restricted to vascular tissue ( Horns and Jeske, 1991 ; Morra and Petty, 2000 ; Sanderfoot and Lazarowitz, 1996 ) and may replicate in vascular and bundle sheath parenchyma cells using pre‐existing plant machinery. Other geminiviruses are not limited to vascular tissue and, instead, are found in mature cells throughout leaves, stems and roots ( Lucy ., 1996 ; Nagar ., 1995 ; Rushing ., 1987 ). These cells have exited the cell division cycle and no longer contain detectable levels of plant DNA replication enzymes necessary for geminivirus replication. By contrast, in cultured cells, geminiviruses preferentially replicate during S phase when host replication machinery is abundant ( Accotto ., 1993 ). The begomovirus, Tomato Golden Mosaic Virus (TGMV), is not restricted to vascular‐associated cells. In situ and immunohistochemical analyses showed that TGMV DNA and the viral replication proteins, AL1 and AL3, occur in differentiated mesophyll, epidermal and vascular cells of leaves ( Nagar ., 1995, 2002 ). In vivo labelling experiments using the thymidine analogue, 5‐bromo‐2‐deoxyuridine (BrdU), demonstrated directly that TGMV replicates in these cell types ( Nagar ., 2002 ). High levels of BrdU are incorporated into both viral and host chromosomal DNA of infected cells, indicating that they have acquired the ability to support efficient DNA replication characteristic of S phase ( Fig. 1A ). However, there is no evidence of cell proliferation during TGMV infection, suggesting that a true cell division cycle state is not established in infected cells. DNA replication in TGMV‐infected cells may reflect entry into an endocycle in which DNA is amplified in the absence of mitosis ( Bass ., 2000 ). Interestingly, hyperplasia and enations are observed in other geminivirus/host combinations ( Briddon, 2003 ; Latham ., 1997 ). Hence, some geminiviruses may induce a cell division cycle whereas others may trigger an endocycle ( Fig. 1B ). In both cases, gene expression is reprogrammed in differentiated plant cells to induce the accumulation of host DNA replication machinery. 1 The retinoblastoma‐related protein pRBR and geminivirus infection. (A) A TGMV‐infected nucleus showing BrdU incorporation into both plant and viral DNA ( Nagar ., 2002 ). Chromosomal and viral DNA was stained with DAPI (blue, left). BrdU was detected using anti‐BrdU antibodies and Alexa 488‐conjugated secondary antibodies (green, middle). TGMV DNA was detected using a Texas Red‐labelled oligonucleotide probe (red, right). (B) pRBR (pRB in animal systems) regulation during plant cell division, development and geminivirus infection. In healthy cells, the ability of pRBR to block cell cycle progression and promote differentiation is inhibited by phosphorylation from G1‐associated cyclin‐dependent kinases. During geminivirus infection, viral proteins interact with and inhibit pRBR to establish S phase and DNA replication competency of a cell division cycle or an endocycle. (C) The pRBR binding motifs of plant viral proteins. The begomovirus replication proteins, TGMV AL1, TYLCV C1 and CaLCuV AL1, interact with pRBR through the Helix 4 motif whereas the mastrevirus BeYDV, MSV and WDV RepA proteins bind to pRBR via an LXCXE sequence. The LXCXE motif of the nanovirus FBNYV CLINK protein is also shown. Conserved Helix 4 residues are shown in blue, whereas conserved LXCXE residues are in red. The sequence of the TGMV AL3 pRBR binding domain is given. GEMINIVIRUSES AND RBR Geminiviruses resemble mammalian DNA tumour viruses in their reliance on host replication machinery and their ability to replicate in differentiated cells. Polyoma, papilloma and adeno viruses encode proteins that modify animal cell cycle controls by binding to the retinoblastoma protein (pRB) or the related family members, p107 and p130 ( Herwig and Strauss, 1997 ). pRB family members negatively regulate cell cycle progression and promote differentiation, in part, through their interactions with E2F transcription factors ( Sidle ., 1996 ). These interactions repress transcription of genes encoding proteins required for entry into S phase and DNA replication ( Lavia and Jansen‐Durr, 1999 ) and facilitate maintenance of a quiescent state ( Harbour and Dean, 2000 ). In late G1, phosphorylation of pRB by cyclin‐dependent kinases (CDKs) disrupts its association with E2F and allows expression of genes required for S phase ( Fig. 1B ; Kaelin, 1999 ). DNA tumour viruses bypass this phosphorylation requirement by binding directly to pRB and disrupting its interaction with E2F ( Fig. 1B ; Jansen‐Durr, 1996 ). Functional pRB is also required to prevent endoreduplication in animal cells, and its inactivation can uncouple entry into S phase and passage through mitosis, leading to uncontrolled DNA replication ( Niculescu ., 1998 ). The pRB/E2F regulatory network is conserved in plants ( Dewitte and Murray, 2003 ). R etino b lastoma‐ r elated (RBR) and E2F proteins have been identified in a variety of plant species ( Durfee ., 2000 ; Gutierrez ., 2002 ). pRBR is a substrate of G1 and S‐phase CDKs ( Boniotti and Gutierrez, 2001 ; Nakagami ., 1999 ) and its levels are high in differentiated plant tissues ( Huntley ., 1998 ), consistent with an involvement in repression of genes required for cell proliferation. E2F consensus sites have been found in the promoters of a number of plant genes ( Menges ., 2002 ; Ramirez‐Parra ., 2003 ) that are transcriptionally modulated during the cell division cycle and development ( Castellano ., 2001 ; Chaboute ., 2002 ; de Jager ., 2001 ; Egelkrout ., 2002 ; Kosugi and Ohashi, 2002 ; Stevens ., 2002 ). Analogous to animal DNA viruses, geminiviruses encode proteins that interact with RBR proteins in plants ( Fig. 1C ). Binding activity has been demonstrated for both mastrevirus and begomovirus proteins, and it is anticipated that curtoviruses and topocuviruses encode pRBR binding proteins. Nanoviruses, the other family of plant single‐stranded DNA viruses that replicate via double‐stranded DNA intermediates, also encode a pRBR binding protein, CLINK ( Aronson ., 2000 ). Among the mastreviruses, pRBR binding has been shown for the RepA proteins of Maize Streak Virus (MSV: Grafi ., 1996 ; Horvath ., 1998 ), Wheat Dwarf Virus (WDV: Xie ., 1996 ) and Bean Yellow Dwarf Virus (BeYDV: Liu ., 1999 ). MSV RepA stimulates cell division and callus growth of maize cultures when expressed from a transgene, consistent with its capacity to bind to pRBR ( Gordon‐Kamm ., 2002 ). The RepA proteins bind to pRBR through an LXCXE motif. The CLINK protein of Favabean Necrotic Yellow Virus (FBYNV) also binds to pRBR via an LXCXE motif ( Aronson ., 2000 ). The mammalian DNA tumour virus proteins, SV40 large T‐antigen, Adenovirus E1A and Human Papillomavirus E7, interact with Rb through the same LXCXE sequence ( Lee ., 1998 ), indicating that this binding motif is of ancient origin. However, the mastrevirus Rep protein, which includes the LXCXE sequence, is unable to interact with pRBR ( Horvath ., 1998 ), demonstrating that the motif is not sufficient to confer binding activity. Both begomovirus replication proteins interact with pRBR. Binding activity has been detected for TGMV AL1 and AL3 ( Ach ., 1997a ; Settlage ., 2001 ), the AL1 protein of Cabbage Leaf Curl Virus (CaLCuV) and the C1 protein of Tomato Yellow Leaf Curl Virus (TYLCV) ( Arguello‐Astorga ., 2004 ). None of these viral proteins contains an intact LXCXE motif, indicating that they bind to pRBR via different mechanisms. TGMV AL1 requires a longer pRBR pocket domain than LXCXE‐containing proteins for efficient binding, further supporting a different binding mechanism ( Kong ., 2000 ). Some plant proteins, including E2F transcription factors ( de Jager ., 2001 ; del Pozo ., 2002 ) and chromatin remodelling proteins ( Ach ., 1997b ; Rossi ., 2003 ), also interact with pRBR through sequences that lack the LXCXE motif. The pRBR binding domain of TGMV AL1 has been located between amino acids 101 and 180 to a region that contains a conserved sequence designated as Helix 4 ( Fig. 1C ; Kong ., 2000 ). Alanine substitutions in Helix 4 significantly reduce the capacities of TGMV and CaLCuV AL1 proteins to bind to pRBR ( Arguello‐Astorga ., 2004 ; Kong ., 2000 ), indicating that this motif is part of the begomovirus pRBR binding interface. The pRBR binding domain of TGMV AL3 has been mapped to the first 36 amino acids ( Fig. 1C ; Settlage ., 2001 ), but has not been characterized further. GEMINIVIRUSES AND E2F Extensive analysis of the PCNA gene suggested that geminivirus infection activates host transcription in mature leaves by relieving pRBR/E2F repression. The first indication that PCNA expression is altered by geminivirus infection came from immunohistochemical studies that detected PCNA protein in mature leaves of Nicotiana benthamiana plants infected with TGMV, but not in equivalent leaves of healthy plants ( Fig. 2A ; Nagar , 1995 ). Uninfected cells immediately adjacent to infected cells did not accumulate PCNA antigen, indicating that induction is cell‐autonomous with the presence of TGMV infection (as determined by the presence of AL1 or viral DNA). PCNA, albeit at lower levels, was also detected in differentiated cells of transgenic plants expressing AL1. These results established that TGMV infection specifically induces PCNA accumulation in differentiated cells and that the AL1 protein is sufficient for induction. 2 Induction of PCNA expression during geminivirus infection. (A) PCNA protein accumulates in nuclei of TGMV‐infected cells (top) but not in equivalent cells of mock‐inoculated plants (bottom) ( Nagar ., 1995 ). PCNA antigen (top, arrows) was visualized using an anti‐human PCNA antibody and peroxidase detection ( Kong ., 2000 ). (B) PCNA promoter regulation in infected (top) and healthy (bottom) mature leaves. In geminivirus‐infected cells, AL1 and AL3 interact with pRBR to induce the release of a repressor complex bound to the E2F sites and reassembly of an active promoter. Immunohistochemical studies also provided evidence that PCNA protein accumulation depends upon AL1/pRBR interactions. N. benthamiana plants infected with TGMV variants mutated in the pRBR binding Helix 4 motif, which have about 15% wild‐type binding activity, display delayed and attenuated symptoms ( Arguello‐Astorga ., 2004 ; Kong ., 2000 ). In these plants, TGMV DNA and AL1 protein are confined to vascular cells, which are also the only cells that accumulate PCNA. The change in tissue specificity may reflect a smaller, more readily inactivated pool of RBR protein in vascular cells. Alternatively, vascular cells, which express the meristematic marker, CDC2, may be predisposed to return to the cell division cycle ( Boucheron ., 2002 ; Hemerly ., 1993 ; Martinez ., 1992 ). Plants encode a diverse array of CDKs and novel cyclins that integrate hormones, environmental signals and growth ( Dewitte and Murray, 2003 ). This complexity also provides a mechanism for tissue‐specific responses to cell cycle inducers like geminiviruses. Accumulation of the PCNA protein in geminivirus‐infected cells reflects changes in host transcription ( Egelkrout ., 2001, 2002 ). PCNA mRNA levels are high in young N. benthamiana leaves (smaller than 2 cm) but cannot be detected in mature leaves of healthy plants. By contrast, TGMV‐infected plants contain detectable levels of PCNA mRNA in both young and mature leaves, indicating that geminivirus infection alters its developmental expression profile. Analysis of PCNA promoter‐luciferase transgene expression showed that TGMV infection activates the PCNA promoter in mature leaves. The N. benthamiana PCNA promoter contains two E2F consensus elements, designated E2F1 and E2F2 ( Fig. 2B ; Egelkrout ., 2001 ). Gel shift assays using plant nuclear extracts or purified Arabidopsis E2F/DP proteins showed that different complexes bind to the two sites ( Egelkrout ., 2002 ). Analysis of mutant PCNA promoters in transgenic plants established that the two elements have different functional roles, consistent with their different binding specificities. The E2F1 element contributes to full promoter activity in young tissues by recruiting a strong activator and/or countering the activity of a repressor bound to the E2F2 site. Both elements contribute to repression in mature tissues, with the E2F2 site being the stronger negative regulatory element. Unlike the wild‐type PCNA promoter, TGMV infection has no detectable effect on the activities of the PCNA promoters carrying E2F binding site mutations in mature leaves. Together, these results demonstrated that TGMV infection induces the accumulation of an essential host replication factor by activating transcription of its gene in mature tissues, most likely by overcoming E2F‐mediated repression. Recent experiments have shown that CaLCuV also activates the PCNA promoter in mature leaves ( Egelkrout ., 2002 ). Hence, it is likely that induction of host gene expression is a common feature of geminivirus infection. REPROGRAMMING HOST TRANSCRIPTION: A MODEL AND QUESTIONS The pRBR binding activity of AL1 and the ability of TGMV infection to overcome E2F‐mediated repression of the PCNA promoter support a model whereby geminivirus replication proteins modulate host gene expression through the pRBR/E2F pathway ( Fig. 2B ). According to this model, in mature plant cells, E2F binds to the PCNA promoter and recruits pRBR, which in turn recruits chromatin remodelling activities like histone deactylases and SWI/SNF‐like enzymes to create a repressor complex ( Zhang and Dean, 2001 ). During infection, a geminivirus replication protein binds to pRBR and disrupts its interaction with E2F, thereby relieving transcriptional repression through release of a pRBR/corepressor complex and/or unmasking an E2F activation domain. In either case, host gene expression is activated, leading to the production of the requisite host DNA replication machinery. Several parts of the model are not yet supported by experimental evidence. To date, AL1/pRBR interactions have not been verified in infected plants because the AL1 protein is expressed at low levels and is not extracted efficiently using native conditions required to maintain protein complexes. In addition, it is not known if AL1 disrupts interactions between pRBR and E2F. It may be possible to test this idea in in vitro binding experiments or yeast three‐hybrid assays ( Tirode ., 1997 ) but the ultimate proof will depend on being able to detect interactions between the different proteins in plant cells. There are also no data showing that pRBR modulates the activity of E2F transcription factors in plants or that AL1/pRBR binding activity is required to relieve transcriptional repression. A first step toward addressing these questions will be a comparison of the abilities of wild‐type AL1 and a pRBR binding mutant to activate the PCNA promoter in the absence of infection. Another question is the role of AL3/pRBR binding in the induction process. PCNA is induced in transgenic plants that express AL1 in the absence of other viral proteins, but PCNA levels are lower than in infected plants ( Nagar ., 1995 ). PCNA accumulation has not been detected in mature leaves of transgenic plants expressing AL3, indicating that unlike AL1, AL3 is not sufficient to induce PCNA expression (S. B. Settlage and L. Hanley‐Bowdoin, unpublished observation). However, it is possible that AL3 enhances host induction through its interactions with both pRBR and AL1 ( Settlage ., 1996, 2001 ). All geminiviruses encode proteins that contain either LXCXE or Helix 4 motifs, and pRBR binding activity has been shown for seven geminivirus proteins ( Fig. 1C ). These observations strongly suggest that pRBR binding and modulation of the pRBR/E2F network are conserved features of geminivirus infection. However, the ability to overcome E2F‐mediated transcriptional repression has only been documented for TGMV and CaLCuV in N. benthamiana ( Egelkrout ., 2001, 2002 ). TGMV is the only geminivirus for which a pRBR binding mutation has been shown to impact symptoms and host gene expression ( Kong ., 2000 ). Unlike TGMV, mutation of the LXCXE motif of the BeYDV RepA protein has no detectable effect on symptoms in N. benthamiana and common bean ( Liu ., 1999 ). However, a reduction in pRBR binding activity may not visibly alter symptoms if BeYDV is a phloem‐associated geminivirus. The different results with TGMV and BeYDV emphasize the importance of examining the role of pRBR binding in a variety of geminivirus/host combinations. MORE QUESTIONS The model proposed in Fig. 2B does not address how the invading single‐stranded viral DNA is converted to double‐stranded DNA prior to the synthesis of the geminivirus replication proteins. In primary infection sites, this early replication step may be catalysed by the host DNA repair machinery. Mastreviruses package an oligonucleotide that could be used to prime DNA synthesis by the repair machinery ( Donson ., 1984 ). Similar primers have not been detected for other genera, and it is not clear how early DNA synthesis would be primed for these geminiviruses. For these viruses, accumulation of the host Pol α/primase complex (and other components of the DNA replication apparatus) may be induced in primary inoculated cells by a wound response triggered by insect feeding or laboratory inoculation procedures ( van de Ven ., 2000 ). Wounding activates the promoters of the cdc2a and PCNA genes ( Hemerly ., 1993 ; E. M. Egelkrout and L. Hanley‐Bowdoin, unpublished data) and is likely to induce other genes required for cell cycle re‐entry and DNA replication. In secondary infection sites, the early events might reflect translocation of viral mRNA specifying the replication proteins into adjacent cells or through the phloem, possibly via the action of geminivirus movement proteins, which bind to RNA as well as to DNA ( Pascal ., 1994 ). Translation of the viral mRNA in the recipient cell would lead to the production of the geminivirus pRBR binding proteins and reprogramming of host gene expression. PCNA protein can accumulate in plant cells containing the AL1 protein but undetectable levels of viral DNA ( Nagar ., 1995 ). These cells may represent early stages in the infection process prior to amplification of viral DNA and/or cells that received viral mRNA but no viral DNA. However, if geminivirus DNA moves as doubled‐stranded molecule instead of a single‐stranded form, it could serve as a transcription template for AL1 production prior to host induction and viral replication. The preferential double‐stranded DNA binding properties of the geminivirus movement proteins ( Rojas ., 1998 ) lend support to this mechanism. Another question is the extent to which geminivirus infection reprograms host gene expression. Microarray analysis showed that RNA virus infection alters the expression of genes encoding proteins involved in disease, metabolism, transport and signal transduction ( Golem and Culver, 2003 ; Whitham ., 2003 ). Although geminiviruses are likely to impact some of the same genes, they may also alter the activities of genes encoding proteins involved in cell division and development. Microarray analysis of synchronized cells established that a large number of plant genes display differential expression during the cell division cycle—many of them at the G1/S boundary and, hence, potential targets of geminivirus induction ( Breyne and Zabeau, 2001 ; Menges ., 2002 ). Preliminary gene profiling data indicated that about 10% of the Arabidopsis transcriptome is differentially regulated in geminivirus‐infected vs. mock‐inoculated leaves (J. T. Ascencio‐Ibañez and L. Hanley‐Bowdoin, unpublished results). Although many of these genes will be regulated through the E2F/pRBR pathway, others will be controlled by other transcription factors that modulate differentiation, metabolism and the disease response. Two protein kinases ( Hao ., 2003 ; Kong and Hanley‐Bowdoin, 2002 ), an acetyl transferase ( McGarry ., 2003 ) and an NAC transcription factor ( Xie ., 1999 ), have been identified as geminivirus protein partners. These plant proteins may modulate host gene expression for functions not directly related to DNA replication. Future experiments will address the identities and functions of the differentially regulated host genes and the transcriptional regulatory networks that are accessed by geminiviruses to establish successful infections. ACKNOWLEDGEMENTS We thank Hank Bass (University of Florida) and Steve Nagar (NCSU) for their contributions to the research described here. This review was supported by grants to L.H.B. from the National Science Foundation (MCB‐0110536) and USDA National Research Initiative (NRI‐2001‐02619). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Molecular Plant Pathology Wiley

Reprogramming plant gene expression: a prerequisite to geminivirus DNA replication

Loading next page...
 
/lp/wiley/reprogramming-plant-gene-expression-a-prerequisite-to-geminivirus-dna-S2SDB5nXrN

References (87)

Publisher
Wiley
Copyright
Copyright © 2004 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1464-6722
eISSN
1364-3703
DOI
10.1111/j.1364-3703.2004.00214.x
pmid
20565592
Publisher site
See Article on Publisher Site

Abstract

INTRODUCTION The Geminiviridae family is classified into four genera, the mastreviruses, begomoviruses, curtoviruses and topocuviruses, based on genome structure, insect vector and host. Together, the four genera infect a broad range of plants and cause significant crop losses worldwide ( Mansoor ., 2003 ; Morales and Anderson, 2001 ; Rybicki and Pietersen, 1999 ). All geminiviruses are characterized by twin icosahedral capsids and single‐stranded DNA genomes that replicate through double‐stranded DNA intermediates in infected cells ( Hanley‐Bowdoin ., 1999 ). They contribute only a few factors for their replication and depend on the nuclear DNA polymerases of their plant hosts. This review focuses on recent studies addressing how geminiviruses modify host gene expression to enable efficient viral DNA replication in infected plant cells. GEMINIVIRUS REPLICATION: THE PLAYERS Geminiviruses amplify their small DNA genomes using a combination of rolling circle and recombination‐mediated replication ( Gutierrez, 1999 ; Jeske ., 2001 ). They provide the proteins required for initiation of replication and recruitment of host replication machinery. Members of the begomovirus, curtovirus and topocuvirus genera encode two proteins required for efficient viral replication ( Hanley‐Bowdoin ., 1999 ). AL1 (also designated AC1, C1 and Rep) is the initiation factor that mediates origin recognition and DNA cleavage/ligation to begin and end rolling circle replication ( Fontes ., 1994 ; Laufs ., 1995 ). AL3 (also designated AC3, C3 and REn) facilitates the accumulation of high levels of viral DNA ( Sunter ., 1990 ), possibly by modifying the activity of AL1 and/or aiding in the recruitment of the host replication enzymes ( Castillo ., 2003 ; Settlage ., 1996 ). Mastreviruses specify two replication‐associated proteins, RepA and Rep, which are translated from mRNA splicing variants. The Rep protein is the functional homologue of AL1 ( Heyraud‐Nitschke ., 1995 ). The RepA and AL3 proteins are not related and may serve different functions during infection. Several host factors have been implicated in geminivirus replication. Viral DNA accumulation is blocked by aphidicolin ( Nagar ., 2002 ), an inhibitor of eukaryotic DNA polymerases α and δ. Geminivirus replication proteins interact with two factors associated with Pol δ: the processivity factor, proliferating cell nuclear antigen (PCNA), and its clamp loader, RFC ( Castillo ., 2003 ; Luque ., 2002 ). These interactions probably represent early steps in the assembly of a DNA replication complex on the geminivirus origin. Geminivirus replication is also blocked by hydroxyurea ( Nagar ., 2002 ), which inhibits ribonucleotide reductase, a host enzyme required for deoxyribonucleotide synthesis and maintenance of dNTP precursor pools. THE DILEMMA: THE SOURCE OF HOST REPLICATION MACHINERY In mature plants, DNA replication and the corresponding enzymes are confined to meristematic and endoreduplicating tissues. Geminiviruses are excluded from meristems and do not have access to the host replication machinery in these dividing cell populations ( Peele ., 2001 ). They can infect young tissues undergoing endoreduplication but are also found in plant tissues at later stages of development. Some geminiviruses are restricted to vascular tissue ( Horns and Jeske, 1991 ; Morra and Petty, 2000 ; Sanderfoot and Lazarowitz, 1996 ) and may replicate in vascular and bundle sheath parenchyma cells using pre‐existing plant machinery. Other geminiviruses are not limited to vascular tissue and, instead, are found in mature cells throughout leaves, stems and roots ( Lucy ., 1996 ; Nagar ., 1995 ; Rushing ., 1987 ). These cells have exited the cell division cycle and no longer contain detectable levels of plant DNA replication enzymes necessary for geminivirus replication. By contrast, in cultured cells, geminiviruses preferentially replicate during S phase when host replication machinery is abundant ( Accotto ., 1993 ). The begomovirus, Tomato Golden Mosaic Virus (TGMV), is not restricted to vascular‐associated cells. In situ and immunohistochemical analyses showed that TGMV DNA and the viral replication proteins, AL1 and AL3, occur in differentiated mesophyll, epidermal and vascular cells of leaves ( Nagar ., 1995, 2002 ). In vivo labelling experiments using the thymidine analogue, 5‐bromo‐2‐deoxyuridine (BrdU), demonstrated directly that TGMV replicates in these cell types ( Nagar ., 2002 ). High levels of BrdU are incorporated into both viral and host chromosomal DNA of infected cells, indicating that they have acquired the ability to support efficient DNA replication characteristic of S phase ( Fig. 1A ). However, there is no evidence of cell proliferation during TGMV infection, suggesting that a true cell division cycle state is not established in infected cells. DNA replication in TGMV‐infected cells may reflect entry into an endocycle in which DNA is amplified in the absence of mitosis ( Bass ., 2000 ). Interestingly, hyperplasia and enations are observed in other geminivirus/host combinations ( Briddon, 2003 ; Latham ., 1997 ). Hence, some geminiviruses may induce a cell division cycle whereas others may trigger an endocycle ( Fig. 1B ). In both cases, gene expression is reprogrammed in differentiated plant cells to induce the accumulation of host DNA replication machinery. 1 The retinoblastoma‐related protein pRBR and geminivirus infection. (A) A TGMV‐infected nucleus showing BrdU incorporation into both plant and viral DNA ( Nagar ., 2002 ). Chromosomal and viral DNA was stained with DAPI (blue, left). BrdU was detected using anti‐BrdU antibodies and Alexa 488‐conjugated secondary antibodies (green, middle). TGMV DNA was detected using a Texas Red‐labelled oligonucleotide probe (red, right). (B) pRBR (pRB in animal systems) regulation during plant cell division, development and geminivirus infection. In healthy cells, the ability of pRBR to block cell cycle progression and promote differentiation is inhibited by phosphorylation from G1‐associated cyclin‐dependent kinases. During geminivirus infection, viral proteins interact with and inhibit pRBR to establish S phase and DNA replication competency of a cell division cycle or an endocycle. (C) The pRBR binding motifs of plant viral proteins. The begomovirus replication proteins, TGMV AL1, TYLCV C1 and CaLCuV AL1, interact with pRBR through the Helix 4 motif whereas the mastrevirus BeYDV, MSV and WDV RepA proteins bind to pRBR via an LXCXE sequence. The LXCXE motif of the nanovirus FBNYV CLINK protein is also shown. Conserved Helix 4 residues are shown in blue, whereas conserved LXCXE residues are in red. The sequence of the TGMV AL3 pRBR binding domain is given. GEMINIVIRUSES AND RBR Geminiviruses resemble mammalian DNA tumour viruses in their reliance on host replication machinery and their ability to replicate in differentiated cells. Polyoma, papilloma and adeno viruses encode proteins that modify animal cell cycle controls by binding to the retinoblastoma protein (pRB) or the related family members, p107 and p130 ( Herwig and Strauss, 1997 ). pRB family members negatively regulate cell cycle progression and promote differentiation, in part, through their interactions with E2F transcription factors ( Sidle ., 1996 ). These interactions repress transcription of genes encoding proteins required for entry into S phase and DNA replication ( Lavia and Jansen‐Durr, 1999 ) and facilitate maintenance of a quiescent state ( Harbour and Dean, 2000 ). In late G1, phosphorylation of pRB by cyclin‐dependent kinases (CDKs) disrupts its association with E2F and allows expression of genes required for S phase ( Fig. 1B ; Kaelin, 1999 ). DNA tumour viruses bypass this phosphorylation requirement by binding directly to pRB and disrupting its interaction with E2F ( Fig. 1B ; Jansen‐Durr, 1996 ). Functional pRB is also required to prevent endoreduplication in animal cells, and its inactivation can uncouple entry into S phase and passage through mitosis, leading to uncontrolled DNA replication ( Niculescu ., 1998 ). The pRB/E2F regulatory network is conserved in plants ( Dewitte and Murray, 2003 ). R etino b lastoma‐ r elated (RBR) and E2F proteins have been identified in a variety of plant species ( Durfee ., 2000 ; Gutierrez ., 2002 ). pRBR is a substrate of G1 and S‐phase CDKs ( Boniotti and Gutierrez, 2001 ; Nakagami ., 1999 ) and its levels are high in differentiated plant tissues ( Huntley ., 1998 ), consistent with an involvement in repression of genes required for cell proliferation. E2F consensus sites have been found in the promoters of a number of plant genes ( Menges ., 2002 ; Ramirez‐Parra ., 2003 ) that are transcriptionally modulated during the cell division cycle and development ( Castellano ., 2001 ; Chaboute ., 2002 ; de Jager ., 2001 ; Egelkrout ., 2002 ; Kosugi and Ohashi, 2002 ; Stevens ., 2002 ). Analogous to animal DNA viruses, geminiviruses encode proteins that interact with RBR proteins in plants ( Fig. 1C ). Binding activity has been demonstrated for both mastrevirus and begomovirus proteins, and it is anticipated that curtoviruses and topocuviruses encode pRBR binding proteins. Nanoviruses, the other family of plant single‐stranded DNA viruses that replicate via double‐stranded DNA intermediates, also encode a pRBR binding protein, CLINK ( Aronson ., 2000 ). Among the mastreviruses, pRBR binding has been shown for the RepA proteins of Maize Streak Virus (MSV: Grafi ., 1996 ; Horvath ., 1998 ), Wheat Dwarf Virus (WDV: Xie ., 1996 ) and Bean Yellow Dwarf Virus (BeYDV: Liu ., 1999 ). MSV RepA stimulates cell division and callus growth of maize cultures when expressed from a transgene, consistent with its capacity to bind to pRBR ( Gordon‐Kamm ., 2002 ). The RepA proteins bind to pRBR through an LXCXE motif. The CLINK protein of Favabean Necrotic Yellow Virus (FBYNV) also binds to pRBR via an LXCXE motif ( Aronson ., 2000 ). The mammalian DNA tumour virus proteins, SV40 large T‐antigen, Adenovirus E1A and Human Papillomavirus E7, interact with Rb through the same LXCXE sequence ( Lee ., 1998 ), indicating that this binding motif is of ancient origin. However, the mastrevirus Rep protein, which includes the LXCXE sequence, is unable to interact with pRBR ( Horvath ., 1998 ), demonstrating that the motif is not sufficient to confer binding activity. Both begomovirus replication proteins interact with pRBR. Binding activity has been detected for TGMV AL1 and AL3 ( Ach ., 1997a ; Settlage ., 2001 ), the AL1 protein of Cabbage Leaf Curl Virus (CaLCuV) and the C1 protein of Tomato Yellow Leaf Curl Virus (TYLCV) ( Arguello‐Astorga ., 2004 ). None of these viral proteins contains an intact LXCXE motif, indicating that they bind to pRBR via different mechanisms. TGMV AL1 requires a longer pRBR pocket domain than LXCXE‐containing proteins for efficient binding, further supporting a different binding mechanism ( Kong ., 2000 ). Some plant proteins, including E2F transcription factors ( de Jager ., 2001 ; del Pozo ., 2002 ) and chromatin remodelling proteins ( Ach ., 1997b ; Rossi ., 2003 ), also interact with pRBR through sequences that lack the LXCXE motif. The pRBR binding domain of TGMV AL1 has been located between amino acids 101 and 180 to a region that contains a conserved sequence designated as Helix 4 ( Fig. 1C ; Kong ., 2000 ). Alanine substitutions in Helix 4 significantly reduce the capacities of TGMV and CaLCuV AL1 proteins to bind to pRBR ( Arguello‐Astorga ., 2004 ; Kong ., 2000 ), indicating that this motif is part of the begomovirus pRBR binding interface. The pRBR binding domain of TGMV AL3 has been mapped to the first 36 amino acids ( Fig. 1C ; Settlage ., 2001 ), but has not been characterized further. GEMINIVIRUSES AND E2F Extensive analysis of the PCNA gene suggested that geminivirus infection activates host transcription in mature leaves by relieving pRBR/E2F repression. The first indication that PCNA expression is altered by geminivirus infection came from immunohistochemical studies that detected PCNA protein in mature leaves of Nicotiana benthamiana plants infected with TGMV, but not in equivalent leaves of healthy plants ( Fig. 2A ; Nagar , 1995 ). Uninfected cells immediately adjacent to infected cells did not accumulate PCNA antigen, indicating that induction is cell‐autonomous with the presence of TGMV infection (as determined by the presence of AL1 or viral DNA). PCNA, albeit at lower levels, was also detected in differentiated cells of transgenic plants expressing AL1. These results established that TGMV infection specifically induces PCNA accumulation in differentiated cells and that the AL1 protein is sufficient for induction. 2 Induction of PCNA expression during geminivirus infection. (A) PCNA protein accumulates in nuclei of TGMV‐infected cells (top) but not in equivalent cells of mock‐inoculated plants (bottom) ( Nagar ., 1995 ). PCNA antigen (top, arrows) was visualized using an anti‐human PCNA antibody and peroxidase detection ( Kong ., 2000 ). (B) PCNA promoter regulation in infected (top) and healthy (bottom) mature leaves. In geminivirus‐infected cells, AL1 and AL3 interact with pRBR to induce the release of a repressor complex bound to the E2F sites and reassembly of an active promoter. Immunohistochemical studies also provided evidence that PCNA protein accumulation depends upon AL1/pRBR interactions. N. benthamiana plants infected with TGMV variants mutated in the pRBR binding Helix 4 motif, which have about 15% wild‐type binding activity, display delayed and attenuated symptoms ( Arguello‐Astorga ., 2004 ; Kong ., 2000 ). In these plants, TGMV DNA and AL1 protein are confined to vascular cells, which are also the only cells that accumulate PCNA. The change in tissue specificity may reflect a smaller, more readily inactivated pool of RBR protein in vascular cells. Alternatively, vascular cells, which express the meristematic marker, CDC2, may be predisposed to return to the cell division cycle ( Boucheron ., 2002 ; Hemerly ., 1993 ; Martinez ., 1992 ). Plants encode a diverse array of CDKs and novel cyclins that integrate hormones, environmental signals and growth ( Dewitte and Murray, 2003 ). This complexity also provides a mechanism for tissue‐specific responses to cell cycle inducers like geminiviruses. Accumulation of the PCNA protein in geminivirus‐infected cells reflects changes in host transcription ( Egelkrout ., 2001, 2002 ). PCNA mRNA levels are high in young N. benthamiana leaves (smaller than 2 cm) but cannot be detected in mature leaves of healthy plants. By contrast, TGMV‐infected plants contain detectable levels of PCNA mRNA in both young and mature leaves, indicating that geminivirus infection alters its developmental expression profile. Analysis of PCNA promoter‐luciferase transgene expression showed that TGMV infection activates the PCNA promoter in mature leaves. The N. benthamiana PCNA promoter contains two E2F consensus elements, designated E2F1 and E2F2 ( Fig. 2B ; Egelkrout ., 2001 ). Gel shift assays using plant nuclear extracts or purified Arabidopsis E2F/DP proteins showed that different complexes bind to the two sites ( Egelkrout ., 2002 ). Analysis of mutant PCNA promoters in transgenic plants established that the two elements have different functional roles, consistent with their different binding specificities. The E2F1 element contributes to full promoter activity in young tissues by recruiting a strong activator and/or countering the activity of a repressor bound to the E2F2 site. Both elements contribute to repression in mature tissues, with the E2F2 site being the stronger negative regulatory element. Unlike the wild‐type PCNA promoter, TGMV infection has no detectable effect on the activities of the PCNA promoters carrying E2F binding site mutations in mature leaves. Together, these results demonstrated that TGMV infection induces the accumulation of an essential host replication factor by activating transcription of its gene in mature tissues, most likely by overcoming E2F‐mediated repression. Recent experiments have shown that CaLCuV also activates the PCNA promoter in mature leaves ( Egelkrout ., 2002 ). Hence, it is likely that induction of host gene expression is a common feature of geminivirus infection. REPROGRAMMING HOST TRANSCRIPTION: A MODEL AND QUESTIONS The pRBR binding activity of AL1 and the ability of TGMV infection to overcome E2F‐mediated repression of the PCNA promoter support a model whereby geminivirus replication proteins modulate host gene expression through the pRBR/E2F pathway ( Fig. 2B ). According to this model, in mature plant cells, E2F binds to the PCNA promoter and recruits pRBR, which in turn recruits chromatin remodelling activities like histone deactylases and SWI/SNF‐like enzymes to create a repressor complex ( Zhang and Dean, 2001 ). During infection, a geminivirus replication protein binds to pRBR and disrupts its interaction with E2F, thereby relieving transcriptional repression through release of a pRBR/corepressor complex and/or unmasking an E2F activation domain. In either case, host gene expression is activated, leading to the production of the requisite host DNA replication machinery. Several parts of the model are not yet supported by experimental evidence. To date, AL1/pRBR interactions have not been verified in infected plants because the AL1 protein is expressed at low levels and is not extracted efficiently using native conditions required to maintain protein complexes. In addition, it is not known if AL1 disrupts interactions between pRBR and E2F. It may be possible to test this idea in in vitro binding experiments or yeast three‐hybrid assays ( Tirode ., 1997 ) but the ultimate proof will depend on being able to detect interactions between the different proteins in plant cells. There are also no data showing that pRBR modulates the activity of E2F transcription factors in plants or that AL1/pRBR binding activity is required to relieve transcriptional repression. A first step toward addressing these questions will be a comparison of the abilities of wild‐type AL1 and a pRBR binding mutant to activate the PCNA promoter in the absence of infection. Another question is the role of AL3/pRBR binding in the induction process. PCNA is induced in transgenic plants that express AL1 in the absence of other viral proteins, but PCNA levels are lower than in infected plants ( Nagar ., 1995 ). PCNA accumulation has not been detected in mature leaves of transgenic plants expressing AL3, indicating that unlike AL1, AL3 is not sufficient to induce PCNA expression (S. B. Settlage and L. Hanley‐Bowdoin, unpublished observation). However, it is possible that AL3 enhances host induction through its interactions with both pRBR and AL1 ( Settlage ., 1996, 2001 ). All geminiviruses encode proteins that contain either LXCXE or Helix 4 motifs, and pRBR binding activity has been shown for seven geminivirus proteins ( Fig. 1C ). These observations strongly suggest that pRBR binding and modulation of the pRBR/E2F network are conserved features of geminivirus infection. However, the ability to overcome E2F‐mediated transcriptional repression has only been documented for TGMV and CaLCuV in N. benthamiana ( Egelkrout ., 2001, 2002 ). TGMV is the only geminivirus for which a pRBR binding mutation has been shown to impact symptoms and host gene expression ( Kong ., 2000 ). Unlike TGMV, mutation of the LXCXE motif of the BeYDV RepA protein has no detectable effect on symptoms in N. benthamiana and common bean ( Liu ., 1999 ). However, a reduction in pRBR binding activity may not visibly alter symptoms if BeYDV is a phloem‐associated geminivirus. The different results with TGMV and BeYDV emphasize the importance of examining the role of pRBR binding in a variety of geminivirus/host combinations. MORE QUESTIONS The model proposed in Fig. 2B does not address how the invading single‐stranded viral DNA is converted to double‐stranded DNA prior to the synthesis of the geminivirus replication proteins. In primary infection sites, this early replication step may be catalysed by the host DNA repair machinery. Mastreviruses package an oligonucleotide that could be used to prime DNA synthesis by the repair machinery ( Donson ., 1984 ). Similar primers have not been detected for other genera, and it is not clear how early DNA synthesis would be primed for these geminiviruses. For these viruses, accumulation of the host Pol α/primase complex (and other components of the DNA replication apparatus) may be induced in primary inoculated cells by a wound response triggered by insect feeding or laboratory inoculation procedures ( van de Ven ., 2000 ). Wounding activates the promoters of the cdc2a and PCNA genes ( Hemerly ., 1993 ; E. M. Egelkrout and L. Hanley‐Bowdoin, unpublished data) and is likely to induce other genes required for cell cycle re‐entry and DNA replication. In secondary infection sites, the early events might reflect translocation of viral mRNA specifying the replication proteins into adjacent cells or through the phloem, possibly via the action of geminivirus movement proteins, which bind to RNA as well as to DNA ( Pascal ., 1994 ). Translation of the viral mRNA in the recipient cell would lead to the production of the geminivirus pRBR binding proteins and reprogramming of host gene expression. PCNA protein can accumulate in plant cells containing the AL1 protein but undetectable levels of viral DNA ( Nagar ., 1995 ). These cells may represent early stages in the infection process prior to amplification of viral DNA and/or cells that received viral mRNA but no viral DNA. However, if geminivirus DNA moves as doubled‐stranded molecule instead of a single‐stranded form, it could serve as a transcription template for AL1 production prior to host induction and viral replication. The preferential double‐stranded DNA binding properties of the geminivirus movement proteins ( Rojas ., 1998 ) lend support to this mechanism. Another question is the extent to which geminivirus infection reprograms host gene expression. Microarray analysis showed that RNA virus infection alters the expression of genes encoding proteins involved in disease, metabolism, transport and signal transduction ( Golem and Culver, 2003 ; Whitham ., 2003 ). Although geminiviruses are likely to impact some of the same genes, they may also alter the activities of genes encoding proteins involved in cell division and development. Microarray analysis of synchronized cells established that a large number of plant genes display differential expression during the cell division cycle—many of them at the G1/S boundary and, hence, potential targets of geminivirus induction ( Breyne and Zabeau, 2001 ; Menges ., 2002 ). Preliminary gene profiling data indicated that about 10% of the Arabidopsis transcriptome is differentially regulated in geminivirus‐infected vs. mock‐inoculated leaves (J. T. Ascencio‐Ibañez and L. Hanley‐Bowdoin, unpublished results). Although many of these genes will be regulated through the E2F/pRBR pathway, others will be controlled by other transcription factors that modulate differentiation, metabolism and the disease response. Two protein kinases ( Hao ., 2003 ; Kong and Hanley‐Bowdoin, 2002 ), an acetyl transferase ( McGarry ., 2003 ) and an NAC transcription factor ( Xie ., 1999 ), have been identified as geminivirus protein partners. These plant proteins may modulate host gene expression for functions not directly related to DNA replication. Future experiments will address the identities and functions of the differentially regulated host genes and the transcriptional regulatory networks that are accessed by geminiviruses to establish successful infections. ACKNOWLEDGEMENTS We thank Hank Bass (University of Florida) and Steve Nagar (NCSU) for their contributions to the research described here. This review was supported by grants to L.H.B. from the National Science Foundation (MCB‐0110536) and USDA National Research Initiative (NRI‐2001‐02619).

Journal

Molecular Plant PathologyWiley

Published: Mar 1, 2004

There are no references for this article.