Posts Tagged ‘plants’

The double life of a geminivirus: Bean yellow dwarf virus

29 May, 2015

Every now and then I solicit contributions for this site – but this one came without coercion, or even prompting!  I thank Romana for being so enthusiastic B-)

Romana Yanez

Biopharming Research Unit, UCT

All material copyright Romana Yanez and UCT


I want to tell the story of a geminivirus called Bean yellow dwarf virus that has two very distinct “lives”: one as a crop pest, infecting bean plants in South Africa and the other as a powerful molecular tool as a viral vector for recombinant protein expression in plants. As if each one of the “heads” of the twinned capsids had a life of its own. The dark side and the bright side. The yin and yang…


Geminiviruses are small, single-stranded, circular DNA plant viruses, so called because each particle is composed of two partially assembled icosahedra joined to form a twinned capsid [1], [2]. They infect plants and are carried by insect vectors such as leafhoppers and whiteflies [2]. They are divided into seven genera: Mastrevirus, Bogomovirus, Topocuvirus, Curtovirus, Becurtovirus, Eragrovirus and Turncurtovirus; according to their genomic organization, the hosts they infect, the insect vectors by which they are transmitted and by genome-wide pairwise sequence identities [3].

Geminiviruses belonging to the genus Mastrevirus are all monopartite viruses with genome sizes between 2.5 and 3.0 kb. They have as vectors different species of leafhoppers. They infect mostly monocotyledonous plants: Maize streak virus (MSV) causes devastating crop losses in African countries, Wheat dwarf virus (WDV), but also infect dicotyledonous plants: Bean yellow dwarf virus (BeYDV), Tobacco yellow dwarf virus (TYDV) and Chickpea chlorosis virus (CpCV) [2], [4], [5].

In 1997 the production of French beans (Phaseolus vulgaris cv. Bonus) was severely reduced  in South Africa, mainly in the Northern Province and Mpumalanga District [4]. Plants presented symptoms similar to a TYDV infection, which at the time was the only mastrevirus described to infect dicotyledonous plants. These symptoms included stunted growth, brittle and leathery leaves, and leaf curling. Investigating the aetiology of this disease were Liu and co-workers. They determined it was a geminiviral infection by identifying virus-like particles (VLPs) with the characteristic twinned morphology. They then sequenced DNA samples of the virus and found it to be most closely related to TYDV, both, in nucleotide sequence (65% identity) and in genomic organization. It was similar enough to be placed under the genus Mastrevirus but distinct enough to be considered a different virus. They Then called it Bean yellow dwarf virus – BeYDV [4].

In 2007 a mild strain of BeYDV (BeYDV-m) was described by Halley-Stott et al. also isolated from P. vulgaris (cv. Top Crop). It was phylogenetically similar to BeYDV having 97% nucleotide sequence identity, but presented sufficient phenotypic differences to be a different BeYDV strain. It contained 81 nucleotide differences compared to the BeYDV type, most of which (63 changes) were found in regions of the genome that directly influenced its replication. Thus, BeYDV-m produced typical symptoms that were less severe and temporally delayed when compared to BeYDV type. The authors also suggested that P. vulgaris is not the BeYDV natural host since it is a non-indigenous plant in South Africa. Furthermore, since both strains of BeYDV were isolated in the same region, it was also suggested that their natural host may show very mild or no symptoms upon infection [6]. Subsequently BeYDV-m was renamed as Chickpea chlorotic dwarf virus (CpCDV) [3]. [However, we like BeYDV, so we’re going to keep called it that – Ed]

Molecular Characteristics and Life Cycle of BeYDV

The genome of BeYDV (Figure 1) is 2,561 nucleotides long with an organization similar to that of other mastreviruses and that replicates by rolling circle mechanism [4], [7]. Its genome is bidirectional, consisting of virion-sense open reading frames (ORFs) V1 and V2, and complementary-sense ORFs C1, C2, C3 and C4. Of these, only C3 and C4 are non-functional and non-conserved between the mastreviruses; although they are also present in TYDV [4], [8]. Within the complementary sense ORFs C1 and C2 an intron is found which is also conserved in other mastreviruses. Virion sense and complementary sense ORFs are separated by a long intergenic region (LIR) and a small intergenic region (SIR) [4]. Liu and co-workers described the functions of each component of the BeYDV genome by mutational analysis.

Figure 1. Genomic organization of Bean yellow dwarf virus. CP, capsid protein. LIR, long intergenic region. MP, movement protein. Rep, replication associated protein. SIR, short intergenic region. [9]

Figure 1. Genomic organization of Bean yellow dwarf virus. CP, capsid protein. LIR, long intergenic region. MP, movement protein. Rep, replication associated protein. SIR, short intergenic region. [9]

The LIR contains a bidirectional promoter to which host factors can bind and a stem-loop structure within the origin of replication (ori) which is required for initiation of rolling circle replication. A binding site and nicking site for the replication associated protein (Rep) are also found in this region. The SIR in turn contains a primer binding region for initiation of complementary strand synthesis as well as transcription termination elements [8]. These are the only two cis-acting elements required for BeYDV replication [8], [10].

The V1 ORF encodes for the movement protein (MP) which is associated with plasmodesmata and is important for systemic spread of the virus. It was found to be a symptom inducer as transgenic plants expressing V1 developed wild type-like infection symptoms. The putative pathogen associated molecular pattern recognized by the host plant may be within the first 17 N-terminus amino acids as plants infected with a mutant  developed wild type-like symptoms as well [8]. 

The V2 ORF encodes for the capsid protein (CP) which is important for viral movement as well and therefore for systemic infection. Thus, intracellular movement or trafficking of the viral DNA may require encapsidation. This was suggested since V2 mutants did not infect plants systemically and also, a basic domain on the N-terminal of the CP was identified which putatively binds to DNA or is involved in nuclear localization [8], [11].

From the genome of BeYDV, the complementary sense ORFs C1 and C2 are the most interesting for me. These encode two regulatory proteins involved in the replication of the virus: Rep and RepA. Their expression is regulated by alternate splicing, where spliced C1 and C2 (C1C2) mRNA is translated into Rep and unspliced C1 mRNA is translated into RepA [8]. 

Rep is responsible for initiating rolling circle replication by nicking the stem-loop structure at the ori, and for releasing nascent virion sense single stranded DNA and later ligating it to form circular ssDNA molecules [8], [12]. Rep is the only protein required for BeYDV replication, but in the presence of RepA the replication is more efficient [8], [10], [11].

RepA is a multi-regulatory protein only found in mastreviruses [2]. Even though both Rep and RepA, have a retinoblastoma related protein (RBR)-binding motif, LeuXCysXGlu, in BeYDV only RepA is able to bind to RBR proteins [10]. In mammalian cells, the retinoblastoma protein is a tumor repressor that binds to and inactivates the transcription factor E2F. By binding to RBR proteins, RepA is thought to disrupt this interaction and force the plant cell cycle into the S-phase – where DNA is replicated just before cell division. RepA is thus acting like other viruses’ oncogenic proteins, such as the human papillomavirus E7 protein and the adenovirus E1A protein. Thus, keeping conditions favorable for enhanced viral replication and proliferation [10], [11]. This could be seen when Hefferon and Dugdale mutated the RBR binding-motif of Rep and RepA to LeuXCysXGln. Only the RepA mutant showed significantly decreased replication. While the Rep mutant showed wild type-like replication [11].

Having in mind what I just described, one can picture the life cycle of BeYDV as follows:

A leafhopper (which has not been identified yet) carrying the virus infects a host plant – this will be a dicotyledonous plant such as P. vulgaris, from which it was originally isolated. The virus releases its ssDNA genome into the cytoplasm. The ssDNA enters the nucleus where host’s replication machinery synthesizes the complementary strand from the primer located in the SIR region, generating a replicative double stranded circular DNA intermediate. At this point the dsDNA serves as template for gene expression, from which Rep and RepA are expressed. RepA transactivates virion-sense gene expression and interferes with plant cell’s life cycle to produce S-phase conditions. Rep nicks the stem-loop structure located at the ori and binds to the 5’ end of the nicked strand. The 3’ end acts as a primer for the synthesis of a new virion-sense strand displacing the previous virion-sense strand. When this new strand is complete, the ori is regenerated and Rep nicks it again. Subsequent release and recircularization of the nascent virion-sense strand is also mediated by Rep. The process continues on the new circular ssDNA molecules as well. Only later, when the amount of CP is high enough, ssDNA molecules are encapsidated. The CP and MP then mediate systemic spread of the viral genome [2], [8]–[12]. When another leafhopper visits the infected plant, the virus is transferred to other plants and all starts again (Figure 2).

gv fig 2

Figure 2. The life cycle of BeYDV. Black circle, BeYDV ssDNA with the stem-loop structure. Black and green circle, BeYDV dsDNA replicative intermediate. Orange spheres, plant host’s replication machinery. Yellow spheres, Rep protein. Black line, nascent ssDNA during rolling circle replication. Purple sphere, RepA. Green sphere, plant retinoblastoma-related protein. Red spheres, BeYDV movement protein. Geminal structures, BeYDV capsid proteins. Modified from [13], [14].

Liu et al. (1997) and Halley-Stott et al. (2007) showed that BeYDV is able to infect other dicotyledonous plants besides P. vulgaris, such as: Nicotiana tabacum, N. benthamiana, Datura stramonium and Arabidopsis thaliana [4], [6]. It has also been isolated from chickpeas in Pakistan [15]. It was noted by Liu and co-workers that the intron of BeYDV (and TYDV) is not as AU-rich as intron sequences present in dicotyledonous plants, which suggested that these viruses had evolved from monocotyledonous-infecting ancestors [8]. Other thing that suggests that BeYDV (and TYDV) evolved from monocotyledonous-infecting mastreviruses is that they encode for two variants of the Rep protein while other geminiviruses infecting dicotyledonous plants encode for only one Rep protein from a continuous ORF [11].

BeYDV as a Powerful Molecular Tool

I have talked about the relatively dark side of BeYDV as a crop pest and plant cell cycle manipulator. Now I would like to introduce you to the other face of this geminivirus.

The importance of recombinant proteins in pharmaceutical, medical and research fields makes them highly demanded, which in turn requires the use efficient production systems [16], [17]. Plants provide a cheaper, faster, more efficient and highly scalable platform for the production of proteins compared to other methods [18], [19]. Vectors based on DNA viruses can be used to express complex proteins without the limitations and complexity faced by RNA viruses such as the need to use more than one virus construct, size constraint imposed on the insert and genomic instability [2], [20]. BeYDV and other geminiviruses have small and simple DNA genomes which can be rapidly amplified to very high copy numbers using mainly host factors and that can be easily manipulated. These features make them attractive viruses for the design of plant vectors for the expression of recombinant proteins [21]. BeYDV has been extensively explored as a molecular tool for the expression of mainly pharmaceutically relevant proteins, such as vaccines, antibodies and enzymes [9], [21]. And recently it has also been used as a means to deliver reagents into plant cells to genetically engineer them [22].

Hefferon and co-workers were one of the first to design a vector derived from BeYDV. They expressed a synthetic version of Staphylococcus enterotoxin B (SEB) in tobacco NT-1 cells. The synthetic SEB sequence was placed under the control of a Cauliflower mosaic virus (CaMV) 35S constitutive promoter and flanked by the cis acting BeYDV LIR and SIR. The Rep encoding gene was provided in trans from a separate construct and also constitutively expressed from the CaMV 35S promoter. Constructs were co-delivered into NT-1 cells by bombardment [11]. They obtained expression levels of ≈0.025 mg SEB / kg of NT-1 cells. They showed that expression of SEB could be enhanced by 20 times by supplying Rep in trans compared to when no Rep was supplied. Overall they showed that BeYDV-based replicon systems promised enhancement of recombinant protein expression in plants [23].

In a more deconstructed approach, Mor et al. (2003) designed a replicon system similar to that of Hefferon and Dugdale (2003) in which the BeYDV MP and CP genes were replaced by the gene of interest (GUS), controlled by CaMV 35S promoter and flanked by the LIR and SIR sequences [24]. Since the CP can sequestrate viral ssDNA, preventing dsDNA to be formed [8], by removing the CP from the viral vector, expression levels can be increased. Removing non-essential features of the virus also gives more room for larger inserts and channels energy and building blocks that would be used to synthesize these proteins into expressing the recombinant protein [20]. Mor et al. obtained expression levels 40 times higher when supplying Rep as well as RepA than when no Rep/RepA was supplied. Showing that RepA also enhances expression levels, probably by making the cell environment more favorable for replication [24]. 

Regnard et al. (2010) designed a replicon vector, pRIC, based on the mild strain of BeYDV that contained the Rep/RepA coding regions in cis rather than in trans. This allowed the vector to autonomously replicate and thus generate high levels of gene copy number and in turn enhanced protein expression. They used N. benthamiana plants and Agrobacterium tumefaciens-mediated gene delivery. They obtained higher expression levels than previously described of three unrelated proteins: enhanced GFP, Human Papillomavirus type 16 major CP, L1, and a HIV-1 p24 antigen. Yields were higher when using the replicative vector than when compared to expression from a non-replicative A. tumefaciens expression vector: 550 mg ⁄ kg fresh leaf weight (FLW) vs. 337 mg L1 ⁄ kg FLW for L1 and 3.23 mg p24 ⁄ kg FLW vs. 0.95 mg p24 ⁄ kg FLW for p24. This study showed that autonomous replication of BeYDV-based vectors dramatically increases gene expression levels [25].

Huang et al. (2009) designed a three-component replicon system that consisted of a construct derived from a deconstructed version BeYDV similar to that described by Mor et al. (2003) containing the gene of interest expression cassette, a construct encoding for the Rep/RepA under CaMV 35S promoter control and a construct expressing the posttranscriptional gene silencing suppressor protein P19. They obtained 0.34 g of Norwalk virus CP (NVCP) / kg FLW and 0.8 g of hepatitis B core antigen (HBc) / kg FLW, which were able to form VLPs. In order to simplify the replicon system, they included the Rep/RepA sequences in cis. They obtained similar expression levels when using the simplified replicon, with or without P19 supplementation as when the three-component system was used [26].  Later they designed a single vector containing multiple replicon cassettes each flanked by a LIR and a SIR. The vector also contained the Rep/RepA sequences under LIR control. Co-delivering the single-vector replicon and a P19 expression vector, they expressed the light and heavy chain of an Ebola virus-targeting monoclonal antibody (mAB), 6D8. They obtained ≈0.5 g of 6D8 mAB / kg FLW which had been assembled correctly and could bind its antigen specifically. Expression levels were comparable to those obtained by Giritch et al. (2006) [27] using two vectors based on two non-competing RNA viruses. They speculated that using this single-vector multireplicon system, even four proteins could be expressed simultaneously using two vectors or placing expression cassettes in tandem  [28].

More recently, Moon et al. (2014) were able to express Brome mosaic virus (BMV) and Cucumber mosaic virus (CMV) VLPs at 0.5 and 1.0 g / kg FLW respectively, using a BeYDV-derived single-vector replicon system. This vector included the P19 coding sequence, the gene of interest as well as the Rep/RepA coding sequences in the same backbone. In this way enhanced expression of VLPs that can be used as carriers for nano-platforms with applications in material sciences and medicine was possible with only one agroinfiltration [29].

Finally, Baltes et al. (2014) demonstrated that BeYDV-based replicon system can be also used for plant genome engineering. They were able to deliver various nucleases (TALENs and CRISP/Cas system) as well as repair templates into tobacco cells and to regenerate plantlets with the desired DNA changes within 6 weeks. This highlighted the potential of vectors derived from BeYDV and other geminiviruses to be applied in the engineering of plants for, for example,  improvement of crop characteristics, crop resistance or in fundamental biology studies [22].

In conclusion, BeYDV is a small, dicotyledonous plant-infecting mastrevirus with apparently unlimited possible molecular applications.


[1] W. Zhang, N. H. Olson, T. S. Baker, L. Faulkner, M. Agbandje-McKenna, M. I. Boulton, J. W. Davies, and R. McKenna, “Structure of the Maize streak virus geminate particle.,” Virology, vol. 279, pp. 471–477, 2001.

[2] E. P. Rybicki and D. P. Martin, “Virus-derived ssDNA vectors for the expression of foreign proteins in plants,” Current Topics in Microbiology and Immunology, vol. 375, pp. 19–45, 2011.

[3] A. Varsani, J. Navas-Castillo, E. Moriones, C. Hernández-Zepeda, A. Idris, J. K. Brown, F. Murilo Zerbini, and D. P. Martin, “Establishment of three new genera in the family Geminiviridae: Becurtovirus, Eragrovirus and Turncurtovirus,” Archives of Virology, vol. 159, pp. 2193–2203, 2014.

[4] L. Liu, T. Van Tonder, G. Pietersen, J. W. Davies, and J. Stanley, “Molecular characterization of a subgroup I geminivirus from a legume in South Africa,” Journal of General Virology, vol. 78, pp. 2113–2117, 1997.

[5] J. Hadfield, J. E. Thomas, M. W. Schwinghamer, S. Kraberger, D. Stainton, A. Dayaram, J. N. Parry, D. Pande, D. P. Martin, and A. Varsani, “Molecular characterisation of dicot-infecting mastreviruses from Australia,” Virus Research, vol. 166, no. 1–2, pp. 13–22, 2012.

[6] R. P. Halley-Stott, F. Tanzer, D. P. Martin, and E. P. Rybicki, “The complete nucleotide sequence of a mild strain of Bean yellow dwarf virus,” Archives of Virology, vol. 152, pp. 1237–1240, 2007.

[7] K. E. Palmer and E. P. Rybicki, “The molecular biology of mastreviruses.,” Advances in virus research, vol. 50, pp. 183–234, 1998.

[8] L. Liu, J. W. Davies, and J. Stanley, “Mutational analysis of bean yellow dwarf virus, a geminivirus of the genus Mastrevirus that is adapted to dicotyledonous plants,” Journal of General Virology, vol. 79, pp. 2265–2274, 1998.

[9] Q. Chen, J. He, W. Phoolcharoen, and H. S. Mason, “Geminiviral vectors based on bean yellow dwarf virus for production of vaccine antigens and monoclonal antibodies in plants,” Human Vaccines, vol. 7, no. 3, pp. 331–338, Mar. 2011.

[10] L. Liu, K. Saunders, C. arole L. Thomas, J. W. Davies, and J. Stanley, “Bean yellow dwarf virus RepA, but not rep, binds to maize retinoblastoma protein, and the virus tolerates mutations in the consensus binding motif.,” Virology, vol. 256, pp. 270–279, 1999.

[11] K. L. Hefferon and B. Dugdale, “Independent expression of Rep and RepA and their roles in regulating bean yellow dwarf virus replication,” Journal of General Virology, vol. 84, pp. 3465–3472, 2003.

[12] C. Gutierrez, “Geminivirus DNA replication,” Cellular and Molecular Life Sciences, vol. 56. pp. 313–329, 1999.

[13] “Adult drawing grape leafhopper,” Koppert Biological Systems 9103. [Online]. Available: [Accessed: 09-Feb-2015].

[14] “Phaseolus vulgaris,” Belgium, Prelude – Royal Museum for Central Africa – Tervuren. [Online]. Available: [Accessed: 09-Feb-2015].

[15] N. Nahid, I. Amin, S. Mansoor, E. P. Rybicki, E. Van Der Walt, and R. W. Briddon, “Two dicot-infecting mastreviruses (family Geminiviridae) occur in Pakistan,” Archives of Virology, vol. 153, pp. 1441–1451, 2008.

[16] G. Pogue and S. Holzberg, “Transient Virus Expression Systems for Recombinant Protein Expression in Dicot-and Monocotyledonous Plants,” in Plant Science, N. K. Dhal and S. C. Sahu, Eds. InTech, 2012, pp. 191–216.

[17] F. Sainsbury, P.-O. Lavoie, M.-A. D’Aoust, L.-P. Vézina, and G. P. Lomonossoff, “Expression of multiple proteins using full-length and deleted versions of cowpea mosaic virus RNA-2.,” Plant biotechnology journal, vol. 6, no. 1, pp. 82–92, Jan. 2008.

[18] E. P. Rybicki, “Plant-produced vaccines: promise and reality.,” Drug Discovery Today, vol. 14, no. 1–2, pp. 16–24, Jan. 2009.

[19] V. Yusibov, S. Rabindran, U. Commandeur, R. M. Twyman, and R. Fischer, “The potential of plant virus vectors for vaccine production.,” Drugs in R&D, vol. 7, no. 4, pp. 203–17, Jan. 2006.

[20] Y. Gleba, S. Marillonnet, and V. Klimyuk, “Plant Virus Vectors: Gene Expression Systems,” Encyclopedia of Virology, vol. 4, pp. 229–237, Apr. 2008.

[21] K. L. Hefferon, “DNA Virus Vectors for Vaccine Production in Plants: Spotlight on Geminiviruses,” Vaccines, vol. 2, no. 3, pp. 642–653, Aug. 2014.

[22] N. J. Baltes, J. Gil-Humanes, T. Cermak, P. a Atkins, and D. F. Voytas, “DNA replicons for plant genome engineering.,” The Plant cell, vol. 26, no. January, pp. 151–63, 2014.

[23] K. L. Hefferon and Y. Fan, “Expression of a vaccine protein in a plant cell line using a geminivirus-based replicon system,” Vaccine, vol. 23, pp. 404–410, 2004.

[24] T. S. Mor, Y.-S. Moon, K. E. Palmer, and H. S. Mason, “Geminivirus vectors for high-level expression of foreign proteins in plant cells.,” Biotechnology and bioengineering, vol. 81, pp. 430–437, 2003.

[25] G. L. Regnard, R. P. Halley-Stott, F. L. Tanzer, I. I. Hitzeroth, and E. P. Rybicki, “High level protein expression in plants through the use of a novel autonomously replicating geminivirus shuttle vector.,” Plant biotechnology journal, vol. 8, no. 1, pp. 38–46, Jan. 2010.

[26] Z. Huang, Q. Chen, B. Hjelm, C. Arntzen, and H. Mason, “A DNA replicon system for rapid high-level production of virus-like particles in plants.,” Biotechnology and bioengineering, vol. 103, no. 4, pp. 706–14, Jul. 2009.

[27] A. Giritch, S. Marillonnet, C. Engler, G. van Eldik, J. Botterman, V. Klimyuk, and Y. Gleba, “Rapid high-yield expression of full-size IgG antibodies in plants coinfected with noncompeting viral vectors.,” Proceedings of the National Academy of Sciences of the United States of America, vol. 103, no. 40, pp. 14701–6, Oct. 2006.

[28] Z. Huang, W. Phoolcharoen, H. Lai, K. Piensook, G. Cardineau, L. Zeitlin, K. J. Whaley, C. J. Arntzen, H. S. Mason, and Q. Chen, “High-level rapid production of full-size monoclonal antibodies in plants by a single-vector DNA replicon system.,” Biotechnology and bioengineering, vol. 106, no. 1, pp. 9–17, May 2010.

[29] K. Moon, J. Lee, S. Kang, M. Kim, H. S. Mason, J. Jeon, and H. Kim, “Overexpression and self-assembly of virus-like particles in Nicotiana benthamiana by a single-vector DNA replicon system.,” Applied microbiology and biotechnology, vol. 98, pp. 8281–90, 2014.

Recombinant Bluetongue virus vaccines – or some, anyway

1 May, 2014

General model of reo-like viruses. Copyright Russell Kightley Media

I picked up yesterday – via @MicrobeTweets’ Twitter feed – on a very useful list of papers in a “Virtual Special Issue” of Elsevier’s recent coverage of vaccines – for “World Immunization Week”. Great stuff, I thought to myself, as I browsed the list – and downloaded at least those that were Open Access, or which I can get via our Libraries’ IP range.

“Even better!”, I thought, as I saw a review entitled “Recombinant vaccines against bluetongue virus?”  A meaty, well-sourced review, I thought; good reading for me and my students / coworkers, and good meat for upcoming Introductions for papers yet to be written.  Indeed, it promised the following:

“The multiple outbreaks of BTV in Mediterranean Europe in the last two decades and the incursion of BTV-8 in Northern Europe in 2008 has re-stimulated the interest to develop improved vaccination strategies against BTV. In particular, safer, cross-reactive, more efficacious vaccines with differential diagnostic capability have been pursued by multiple BTV research groups and vaccine manufacturers. A wide variety of recombinant BTV vaccine prototypes have been investigated, ranging from baculovirus-expressed sub-unit vaccines to the use of live viral vectors. This article gives a brief overview of all these modern approaches to develop vaccines against BTV including some recent unpublished data.”

So, I parked the conveniently Open Access-ible window away on the side of my desktop, to be got back to with every expectation of delight.

Until I read it, that is: well-sourced it may be; excellent in its coverage, it is NOT.  In fact, apart from a brief discursion on subunit vaccines – concentrating almost exclusively on baculovirus / insect cell-produced proteins – it is almost exclusively concerned with live viral vectors for bluetongue proteins, and of poxviruses in particular.  Now, this is all very well, if that is what they work on – but to dismiss one of the potentially most exciting developments in recent Bluetongue vaccinology like this:

“VLPs of BTV have been also produced in plants recently using the cowpea mosaic virus and their use in a vaccination study produced no clinical manifestations in sheep after homologous challenge, although viremia was no [sic] evaluated (Thuenemann et al., 2013).”

– boggles the mind somewhat.  Really?  That’s all they have, compared to the screed immediately before it on baculovirus-produced antigens?  They get the expression system wrong – it is an Agrobacterium tumefaciens-mediated transient expression system in Nicotiana benthamiana involving a Cowpea mosaic virus-derived enhanced translation vector – and neglect to mention that the VLPs produced are as good as anything produced in insect cells; will be FAR cheaper to produce, and WORKED AS WELL AS THE CONVENTIONAL ATTENUATED LIVE VIRUS VACCINE IN A CHALLENGE EXPERIMENT IN SHEEP.  True!

This is a big deal, folks, really: successful production of significant amounts of VLPs requiring simultaneous expression of 4 structural proteins of BTV-8 in plants AND their subsequent assembly, AND performing as well as the standard vaccine in an animal trial.  But no – not good enough for our review’s authors….

I must declare vested interests up front here: first, we work on plant-made recombinant Bluetongue vaccines; second, I and others in my group are co-authors of the paper whose lack of coverage I am aggrieved about.

But that’s not the point: what IS the point is that this review is a slipshod piece of work that damns our collective endeavour with faint praise, in community that might otherwise have been alerted to an alternative to the far-too-expensive-for-animal-use baculovirus expression technology.

Ah, well.  I suppose that’s what blogs are for B-)

PBVAB 5 Part 2

28 June, 2013

Session: Vaccines 1.

This session produced some of the most interesting talks of the conference, so I will go into some detail in describing them.

Charlie Arntzen (ASU, Tempe, AZ) gave a typically excellent presentation on their latest work on norovirus vaccine formulation for stability and oral delivery – using lyophilized aloe gel-derived nanoparticles.

Norovirus outbreaks are tracked by a CDC lab continuously; every 2 years or so new strains circulate, meaning vaccines will have to keep up.  Ligocyte makes VLPs in insect cells currently; however, plant expression has been shown to be able to respond quickly to strain changes, via Icon vectors used at KBP, with the possibility of very quickly making a lot of product.  Downstream formulation has been a problem, however, as the processing throws away lots of antigen protein downstream.

Ligocyte use MPLA and chitosan (an irritant) for nasal immunisation: this has 50% efficacy.  The FDA does not like adjuvants for nasal dosing – so they went for no adjuvant, and chose the nasal route as one gets a more uniform response for 5x less Ag than with oral administration.  The formulation is basically of VLP preparations with lyophilized and milled pectin content from aloe gel: the uniformly-sized nanoparticles absorb water, and stick to each other and to the nasal mucosa for 3 hrs+.

Charlie commented that “This is the one time I recommend putting white powder in your nose!”.

They have tried mixing VLPs of different virus types – and found that with 50 ug of each, you get same immune response as to one.  Apparently this virus is unusual as you can do virus challenge experiments quite easily: these cost $15K/patient, which is a bargain.  The group is looking at annual or biannual dosing for maximum protection, and is also formulating VLPs for oral vaccination.  Interestingly, the aloe gel also works for intramuscular vaccination – possibly as a result of a depot effect?


Yuri Gleba (Nomad / Icon Genetics, Halle, Germany) was supposed to speak on “Technology progress in PMP (=plant-made pharma) research” – but basically said “Transient technologies are the future!”, and then went on to demonstrate it.  He noted that KBP can process 1.2 tonnes biomass/day for agroinfiltrating plants, using a robot from a car factory and a converted industrial autoclave – and consequently have to grow plants in trays for infiltration. Nomad had therefore started investigating how spraying Agrobacterium onto plants might work – with a biosafe Agrobacterium as a prime requirement.  They also took the bold approach of doing transient agronomic trait engineering – for traits such as flowering control, drought tolerance, yield suites, cellulases and anti-microbials – and sold the idea to Bayer Crop Science.

Their technique uses an engineered Agrobacterium that is 100-1000x more efficient at gene delivery than standard strains, with surfactants that allow easy penetration of the leaf tissues.  In combination with the use of replicating vectors that spread cell-to-cell, they could get 100% of standard infiltration yields, by a far easier and much more scalable technique.  They found that spraying worked for most dicots and even for maize, albeit inefficiently, and that they could repeatedly dose plants for the same trait with no apparent harm.  Transient expression of cry1ab and cry2ab Bt toxins delivery worked well, as did delivery of the Cold Shock protein from B subtilis, which also works for drought tolerance.

Their technique does away with need for seed – they can do somatic trait addition / subtraction, they can use the technology outdoors, and there is no trait transmission and so no escape, as the genes do not get into seed.  It means they can produce proteins in  plants at commodity agricultural prices – which considerably broadens the scope of “biofarming” in terms of what products can viably be made!!

One good example was cellulases for bioethanol production: one needs 1-3% w/w relative to cellulose mass, meaning production must be high volume and cheap.  Yuri noted it was possible to store biomass as silage or possibly by vacuum-packing at room temperature for months and that the silage process also eliminated Agrobacterium.  He mused that it might be possible to make a sauerkraut-type oral formulation for recombinant protein delivery…B-)

As for antimicrobials in plants, he said organic crops have more microbes than standard, eg: the recent fatal infections caused by E coli in bean sprouts in Europe – and that a solution would be to make eg colicin E1 in the plants, to kill the bacteria in situ.  One can apply for GRAS status which is MUCH faster than for other routes of approval.  They were currently doing this for phage lysins, bacteriocins, and thaumatin, among others.  Yuri said transient expression tech was like flash drives vs old-style PCs: a versatile set of tools vs a one-trick pony.  He also mused that the technology could lead to reinvention of the old ideal of use of biofarming in undeveloped communities – presumably for low-cost remedies as well as for therapeutics, etc.

To a question on what was the shelf-life of recombinant Agro he replied that there was already field use of live bacteria to combat pathogenic strains; that one can take a Petri dish and dilute in 100l water and spray, and then keep the suspension for two days…it was a very robust bug!  An interesting regulatory point that came up was that the USDA thinks a plant is a GMO even if it is transiently sprayed.


Andres Wigdorovitz (INTA, Buenos Aires, Argentina) spoke on their experience of a decade’s worth of work on plant-made veterinary vaccines.  He opened by noting that he has a major problem of getting money from companies in Argentina – partly die to what a “product” is defined as, because what happens is that a “researcher has an egg, whereas the company wants a butterfly”.

They made the decision to work in platforms – to make diagnostic kits initially, which teaches one how to make recombinant proteins.  They use baculovirus/insect cells and plants – in the form of transgenic alfalfa or transplastomic tobacco – to make the same proteins for comparison purposes, and as products, depending on which was more suitable.  While they had had considerable experience with FMDV vaccines made in plants, which had been protective, their current work focused on making novel vaccines and products.  An example was camellid-derived VHH nanobodies – and the fact that fusing the E2 protein of Bovine viral diarrhoea virus (BVDV) with a anti-E2 VHH gave a better alfalfa-produced immunogen for something that was already protective.  Their experimental vaccine was better than the commercial vaccine from the Ab response – and they could get total protection with 3 ug vaccine, and even better efficacy if they made an E2-HLA fusion.  He believed they will have a commercial vaccine in less than 2 yrs as they were engaged in getting regulatory approval now.

In other work, a FMDV VP1 peptide-GUS fusion expressed 10x better in transplastomic tobacco than in transgenic alfalfa.  A rotavirus VP8* fusion protein was also 10x increased in chloroplasts, and dry leaves preserved the protein very well.  They were also making VHH nanobodies against human rotavirus as vaccine coverage of local strains was not good – and VHH against the conserved VP6 could penetrate the outer capsid and bind and neutralize infectivity whereas larger proteins did not work.  They got 3% TSP expression in tobacco chloroplasts.  They were also making VHH to other rotavirus proteins, and to human noro- and influenza viruses.  All in all, it was a very heartening demonstration of a good business model, and that developing countries too can lead the field in some respects.


The remainder of the session was taken up with two talks from our group: these were given by Drs Ann Meyers and Inga Hitzeroth, on the parrot-infecting Beak and feather disease virus CP-elastin fusion protein production, and Human rotavirus CP and VLP production in N benthamiana via agroinfiltration, respectively.

The BFDV work has just been published with MSc student Lucian Duvenage as first author – from PubMed, then:

J Virol Methods. 2013 Jul;191(1):55-62. doi: 10.1016/j.jviromet.2013.03.028. Epub 2013 Apr 9.
Expression in tobacco and purification of beak and feather disease virus capsid protein fused to elastin-like polypeptides.
Duvenage L, Hitzeroth II, Meyers AE, Rybicki EP.
Department of Molecular and Cell Biology, University of Cape Town, Rondebosch 7700, South Africa.


Psittacine beak and feather disease, caused by beak and feather disease virus (BFDV), is a threat to endangered psittacine species. There is currently no vaccine against BFDV, which necessitates the development of safe and affordable vaccine candidates. A subunit vaccine based on BFDV capsid protein (CP), the major antigenic determinant, expressed in the inexpensive and highly scalable plant expression system could satisfy these requirements. Full-length CP and a truncated CP (ΔN40 CP) were transiently expressed in tobacco (Nicotiana benthamiana) as fusions to elastin-like polypeptide (ELP). These two proteins were fused to ELPs of different lengths in order to increase expression levels and to provide a simple means of purification. The ELP fusion proteins were purified by inverse transition cycling (ITC) and it was found that a membrane filtration-based ITC method improved the recovery of ΔN40 CP-ELP51 fusion protein relative to a centrifugation-based method.

Essentially, Lucian managed in some very elegant work to show that BFDV CP fused to a 51-mer ELP allowed production and subsequent simple purification of quite high yields of fusion protein.  It remains to be seen how immunogenic or protective this is – however, it is a breakthrough, as expression of the CP alone has been VERY problematic, in everything from insect cells through E coli, to plants.

Inga spoke on our recent MSc student David Mutepfa’s work on expression in plants of the CPs of a South African rotavirus that is not well matched to current live attenuated vaccines.  The short story is that he succeeded very well indeed in expressing three of the four proteins.  From a recent publication from me and Nunzia Scotti on plant-made VLPs, then:

Current studies in the Rybicki laboratory have focused on expression of capsid proteins of a local isolate of human rotavirus (G9 P[6]) that is not well matched to available commercial vaccines.  Expression of VP2, VP4 and VP6 in N. benthamiana was targeted via co-agroinfiltration to the cytosol, endoplasmic reticulum, apoplast and chloroplast. Electron microscopy showed that co-expressed VP2/6 and VP2/6/4 produced virus-like particles in the cytosol, with yields as high as 1.1 g/kg of plant material, for batches of 100 g.

…with a picture to prove VLPs are made:

rota pic


Rotavirus VP2/6/4 co-expression in N benthamiana: protein ex- tract partially purified by sucrose gradient centrifugation, particles captured onto electron microscope grids with mouse-anti VP6 antibody. Bar = 200 nm

Influenza vaccines from plants??

22 April, 2008

I should have known Alan Cann would find this one; it’s just too good to miss – so I am going to add to what he said, as a way of further exploring what they could/should have done, as a result of discussions in our Journal Club this morning.

Alan wrote:

Influenza vaccines from plants

Posted by ajcann on April 16, 2008

 Our major defense against infection with influenza viruses is immunization of individuals with an annually updated vaccine that is currently produced in chicken eggs, with a global annual capacity of about 400 million doses, a scale of production insufficient to combat a pandemic. Furthermore, at least six months is required between the identification of new virus strains to be included in the vaccine formulation and the manufacture of bulk quantities. Uncertainties over the robustness of egg-based vaccine production are intensified even further by the emergence of H5N1 strains that are highly virulent to both chickens and eggs. There is a need to develop alternative vaccine production systems capable of rapid turnaround and high capacity. Recombinant subunit vaccines should circumvent some of the concerns regarding our current dependence on egg-based production.This paper reports on the production and evaluation of domains of influenza haemagglutinin (HA) and neuraminidase (NA) fused to the thermostable enzyme lichenase. All vaccine targets were produced using a plant-based transient expression system (Nicotiana). When tested in ferrets, vaccine candidates containing these engineered plant-produced influenza HA and NA antigens were highly immunogenic, and were protective against infection following challenge with homologous influenza virus. This plant-based production system offers safety and capacity advantages, which taken together with the protective efficacy data reported, demonstrates the promise of this approach for subunit influenza vaccine development.

A plant-produced influenza subunit vaccine protects ferrets against virus challenge
Influenza and Other Respiratory Viruses 2008 2: 33–40

There are a couple of interesting features of this paper, chief among them being the complete obscurity of the reasons why they use lichenase fusions, and what exactly their “launch vector” – which is what they use to express their proteins transiently – is.  Because the reference they give is incorrect – it is to a journal they erroneously call “Influenza”, which is not listed by PubMed, and turns out to be Influenza and Other Respiratory Viruses in fact – and is unavailable at our institution.  I am assuming, given the system uses a CaMV 35S promoter to drive RNA production, and they talk of “viral replication and target sequence expression from the [TMV] CP subgenomic mRNA promoter”, that the vector is a TMV-based replicon.  I was alerted by colleagues at the Journal Club to the fact that the same group used the same system – pBID4 “launch vector”, fusions to lichenase – for production of a HPV E7 vaccine in plants.  And referred to the same paper as this one does, for the vector and constructs.   Aargh!  I still don’t know why lichenase fusions are such a good idea!! 

A hint is given in the E7 paper: they say that “…these LicKM fusion proteins alone are able to activate both innate and adaptive antigen-specific immune responses”.  But they found in the paper under discussion here that alum was needed to get the best response…and they got the best yield AND immunogenicity out of their NA protein, which was expressed as a (presumably) soluble truncated native protein.  So the reason is still obscure.

The purification section of this paper is also woefully inadequate: saying “…recombinant antigens were enriched by ammonium sulphate precipitation followed by immobilised metal affinity chromatography and anion exchange chromatography, with dialysis after each step, to at least 80% purity” is NOT a method!  It is an anecdote, fit for a 1-minute talk maybe, but NOT for the Methods section of a paper.  Naughty, naughty!

Another interesting thing is the complexity of the vaccine constructs – again, exactly the same type of constructs as made for HPV E7; assembly-line vaccine producers, these guys!  These consist of the Gene of Choice (GoC) with a poly-His tag AND a KDEL (ER retention) tag at the C-terminus, AND the signal sequence of Nicotiana tabacum PR1a protein at their N-terminus.  This means (a) proteins get into the ER lumen, (b) get retained in the ER, (c) can be purified by Ni or other metal affinity column.  In addition to being fused to LickM.  Granted, the PR1 signal sequence is lost and the His tags can be removed – but the proteins still have significant “other” constituents – which is rather frowned on in a vaccine intended for humans.

I am also interested that they did not do the standard thing with their plant-produced HA GD protein and test for haemagglutination / RBC binding: this was in any case superseded by the fact that the vaccines were protective and antisera elicited by them worked in HI [haemagglutination-inhibition] assays, but it has long been regarded as a necessary first step.  I like these guys’ approach: forget the biochemistry; let’s see if it works!

All in all, a good paper despite our criticisms, which points up the very distinct possibility of being able to use plant production of influenza virus antigens for the rapid production of effective vaccines.

But I wish they’d included some more details….