I was privileged to attend – and give the first presentation in – this workshop, on April 12-13 this year, at Monash University’s Clayton Campus in Melbourne. While it was small – about 25 people all told – I think what was discussed was very interesting. Even more interesting was the discussion of where the technology should be taken in Australia, given the science environment there has many similarities to the South African situation.
The workshop included some 20-odd people, mainly from the School of Biological Sciences at Monash in Melbourne, but with representation from the University of Cape Town, Southern Cross University (NSW), the Agri-Science in Queensland, the University of Queensland, and –importantly for the implications of this kind of work – the Australian Department of Defence.
I kicked off proceedings with a Keynote Address on “Viruses, vaccines and plants: The Cape Town experience”: this was essentially an account of my laboratory’s work on plant-made vaccines and other pharmaceutically-important molecules over the last 15 years, focussing on optimisation of expression of relevant molecules, and our work on Human papillomavirus (HPV), HIV and influenza virus vaccines in particular. I emphasised that optimisation requires one to look at molecule, gene, vector, expression host, expression modality and intracellular localisation – and that there was no substitute for an empirical determination of optimisation parameters.
In the first session – Plant-Made Vaccines – chaired by Diane Webster of Monash, Sadia Deen of Monash described her work on making an experimental mouse vaccine against the causative agent of fowl cholera, Pasteurella multocida. Expression of the OmpX protein via transient agroinfiltration of N benthamiana was successful, as was protection of mice against lethal challenge by injection of freeze-dried leaf powder with alum as adjuvant. Interesting features for me were that the vaccine was apparently better than the E coli-produced version, and that it was apoplast-targetted – meaning purification could be very simple.
Assunta Pelosi of Monash then described an investigation of the in vivo fate in mice of the B subunit of the E coli enterotoxin (LTB) that had been produced in plants and then and formulated in different ways. Hairy root culture-produced protein was most protected and released antigen latest regardless of formulation; otherwise, LTB produced in N benthamiana leaves or tomato fruit was released in stomach or duodenum if formulated in aqueous (=apple juice + honey) media, and in the duodenum and ileum if formulated in lipid (=peanut butter) media. There seemed to be no difference in how much LTB went all the way through the animals regardless of origin or formulation. I was interested in the possibility of an enhanced adjuvant effect with protein produced in N benthamiana compared to other routes of production, as this has been documented for other proteins – and this is apparently being investigated, using protein produced in different Nicotiana spp.
Session 2 – Plants as Production Systems – was chaired by me, and featured a fascinating mix of production of an exciting new product, a chemical engineering view of production of antibodies via plant tissue culture, and an introduction to the very significant industrial potential of sugarcane and/or sugarcane processing infrastructure for biopharming.
Diane Webster of Monash described her group’s work on expression of soluble human-derived RAGE, which has potential both as a diabetes therapy to lessen uptake by the ordinarily membrane-linked receptor of advanced glycation end-products (AGE), and as a reagent for advanced assay techniques. She described how the Ig-like protein was very difficult to make in insect cells, and how the E coli or yeast-made versions were useless because of lack of or inappropriate glycosylation. sRAGE could be made successfully in plants; addition of a KDEL ER retention signal adversely affected yield, while a His-tag did not. While yield was only ~0.6% of total soluble protein (TSP), they could get 70% recovery of a 90% pure protein – which was identical to mammalian RAGE in terms of –S-S- bridges and functionality as assessed by surface plasmon resonance. It was interesting that use of ICON vectors did not appear to help increase yield.
Pauline Doran, of the Dept Chemical Engineering and the School of Biological Sciences at Monash, gave a fascinating account of her experiences with plant tissue culture as a production vehicle for biopharming. Her first point was that there are important trade-offs with bioreactor vs. whole plant production: for example, production of biomass was more expensive for the former, but purification of final product was probably cheaper. It was also much easier and more reproducible to control a wide variety of environmental and growth conditions for tissue cultures. She described years worth of experience of making anti-S mutans monoclonal Abs in transgenic suspension cultured tobacco cells, shooty teratomas derived from transgenic plants using Agrobacterium tumefaciens, and hairy root cultures produced via A rhizogenes transformation. Production over years was most consistent for root cultures, while levels of mAb production were similar initially for roots and suspension cultured cells. An important consideration in how to make the Abs was that suspension cultured cells produced significant amounts of proteases via the apoplast, so that secreted Abs could be degraded – which is why it was better to retain them in the ER. Rhizosecretion form root cultures was a useful means of production; however, adsorption to vessel walls was a major problem even though it could be blocked using proteins or PVP. The talk was valuable for laboratory-end scientists because it showed up practical problems – and solutions – not often dealt with during the research and development phase of biopharming research.
Peter Twine of the Queensland CRCSIIB recounted how they had been given Aus $28 million over 7 years to improve value in sugar in Australia. He noted that Australia could handle 40 million tonnes of cane in some 40 processing stations annually – so very little further investment would be needed to get much more product than the bagasse, sugar and ethanol that the crop was presently used to produce. In terms of biomass, a cane field could produce 50-100 tonnes / ha / yr, which could be significantly improved by strategic backcrossing. Value enhancements which had already been made included production of Barrecote – a biodegradable waterproof coating for recyclable paper – as well as a GI-lowering agent, and PHB (for production of plastics) that was 3-4 times cheaper than E coli-produced material. He said what was wanted for sugarcane was an Agrobacterium transformation protocol, as well as chloroplast transformation: this would allow significant expression enhancement as well as reliable transgenesis. The talk was valuable both as an informational window into a well established industrial-scale biotechnology, and as a window into possibilities for use of the established biomass processing infrastructure for biopharming purposes: for example, discussion after the talk ranged around the significant potential for use of even only a small fraction of that infrastructure – for example, one processing mill – to produce potentially very large amounts of lower-end pharmaceuticals (high volume, low cost), such as enzymes or nutraceuticals.
The last Monday session was a Student Forum, chaired by Rob Shepherd from Monash. Giorgio De Guzman (Monash) described how their team had systematically investigated production of LTB in hairy root cultures of tobacco, tomato and petunia: he mentioned that a distinct advantage of hairy root cultures was that they grew rapidly, and required only simple media with no hormones. They showed that tobacco and petunia cultures gave the best yields (60-70 ng/g root), and that while there was some negative correlation of growth with expression level, this was least for petunia. Optimum growth of cultures was reached after 22 days, and rhizosecretion peaked at 20 days. What was most interesting to me was that it was possible to regenerate plants from root cultures: this would allow seeding and preservation of good producer lines as seed rather than as root cultures, with the option or re-establishing root cultures again as needed. An added bonus was that such cultures seemed to grow better than the original cultures, with the same recombinant protein yield.
Huai-Yian Ling (Monash) gave a talk on a topic close to my heart: this was the use of plants to produce candidate pandemic influenza virus vaccines. She specifically addressed the potential problem for oral vaccine development of the high alkaloid content of the tobaccos that are otherwise very well suited for pharming, due to their ease of propagation, high biomass per plant, and prolific seed production, by crossing transgenic H5 haemagglutinin (HA)-producing N tabacum with a N glauca line in which alkaloid production had been silenced by siRNA. She determined that a plant codon-optimised HA gene expressed best, and gave 13 ug/g leaf tissue in the line used for crossing. The hybrid progeny had no alkaloid, but also only produced HA at levels <1 ug/g. Lyophilisation allowed this to be concentrated to 64 ug/g, which would be sufficient for oral immunisation experiments. She planned to see whether a LEA protein from a resurrection plant would protect the HA.
The final talk – and of only two in the whole workshop which did not involve transgenic plants or recombinant protein expression – was by Sylvia Malory (Southern Cross Univ). She had studied the possibility of “mining” rice domestication genes in the rice relative Micolaena stipoides: this grass is a drought-, frost- and shade tolerant perennial which produces rice-like seed. Her approach was to exploit the extensive sequence and map databases for rice and other grasses so as to rapidly extract relevant genetic information for homologues of important rice genes from whole-genome shotgun sequencing runs of M stipoides DNA. Traits of interest included reduced plant height, non-shattering seeds, increased yield, controlled seed dormancy, photoperiod insensitivity, drought tolerance and disease and insect tolerance. For me the interest in this talk was that the potential for crossover between the biopharming and crop improvement spheres seems to get closer and closer with improvements in technology such as pyrosequencing, in that crop targets such as secondary metabolite engineering can also be exploited for pharming.
The final session of the Workshop on the Tuesday morning was left to me to summarise what the strategic directions of the existing and future “Molecular Farming Network” in Australia could be. I approached the topic by summarising where I thought the biopharming field was presently in terms of direction and priorities – which was that the latter had shifted from the largely “vaccines in edible transgenic plants” hype of early years, to a more mature appreciation that transiently-expressed proteins in non-food plants were the way to go, and that the preference for products should be enzymes, reagents, diagnostics, and then therapeutics, and animal vaccines – with human vaccines as a long-term goal. This notwithstanding, I highlighted the fact that transiently-expressed plant-made candidate vaccines for H5N1 and H1N1 influenza and for noroviruses were probably the quickest emergency response vaccines ever made, which was a very important niche for the technology.
My impression from the workshop, and from discussions with various folk, was that this particular biotechnology in Australia is presently under-developed, despite the presence in Australia (and apparently, mainly at Monash University) of considerable expertise and potential products. I caution against the single-minded pursuit, for example, of “edible” or even of oral vaccines, where injectable versions may well work better and be cheaper to produce, simply because much less active ingredient is needed. I also urge that the production of various reagent- or diagnostic-related products be pursued more vigorously, given that there is either no regulatory barrier to overcome, or – for animal products in particular – a far lower one than exists for products made for use in humans.
And I thank Amanda Walmsley, John Hamill, Rob Shepherd and Assunta Pelosi for looking after me so well.