Regulatory agencies and the public have traditionally entrusted the scientists developing new technologies to conduct rigorous evaluation and research of their products to ensure safety to human health and the environment. But one biotech CEO bluntly said, "we not in the business of making problems for ourselves". Therefore, critique by scientists outside industry will continue to provide useful scientific information on the risks or safety of GMOs for the public and policymakers. Dr. David Quist offers us just the rigorous evaluation missing in the rush to insert genes into chloroplasts.
Chloroplast genetic engineering (cpGE) is the new horizon for GM technology [1]. It is rapidly becoming the transformation method of choice for the next wave of transgenic products in crop plants, plant-made pharmaceuticals. Instead of inserting transgenes into the plant genome, they are inserted into a chloroplast's genome. A chloroplast is an organelle in green plants containing the light harvesting pigment chlorophyll. In flowering plants, there could be thousands of copies of chloroplast DNA per cell, for each cell contains a large number of chloroplasts (10-100), and each chloroplast has multiple copies (20-50) of its genome that may be the same or different from the others.
This alone raises questions on gene expression, regulation and cell functioning that we currently have no answers to.
Developers claim that cpGE is a stable, precise, highly expressing system that's environmentally friendly. These claims rely mostly on the notorious "absence of evidence as evidence", and some are true only for a few plant families. Contrary reports are often dismissed as "rare" events and hence insignificant. And definitive experimental evidence is lacking that these claims are valid under realistic or ecologically relevant contexts. I shall deal with each of the claims below.
The integrity of chloroplast transgene expression and the integrity of transgenic sequences in successive generations or crosses have simply not been tested.
Reliable transgene expression is important to ensure a high and predictable level of agronomic performance of the GM crop, otherwise, the projected benefits will fail to be realized. Unfortunately, the claim of stable transgene expression is based on the absence of relevant evidence: None of the published studies looked at transgene expression beyond the first (T1) generation after transformation. The patterns (but not levels) of expression reported are no different from those observed in unstable nuclear transgenes at this stage, which are subject to spontaneous mutation, rearrangements, and inherent host defence mechanisms designed to identify and inactivate alien genes. Analysis of chloroplast DNA in wheat turned up a number of structural alterations, mediated by a suite of mechanisms, including illegitimate recombination [2]. Studies on the long-term integrity of chloroplast transgene sequences and their products have not yet been conducted to support claims of DNA or transgene expression stability.
The supposed chloroplast transgene resilience to gene silencing mechanisms, including post-transcriptional gene silencing, has not been experimentally tested.
Ensuring stable transgene expression has been a chronic problem due to a number of mechanisms in the host plant that inactivate foreign DNA inserted into a plant's genome. These natural host defence mechanisms act at genetic, transcriptional and post-transcriptional levels [3]. The claim that cpGE avoids gene silencing [4] is misleading. One might expect that if the transgenes can be targeted to particular locations in the chloroplast genome, for example, to avoid heterochromatin (regions of the genome that are inactive), that would help prevent silencing. But other forms of transgene silencing, including various post-transcriptional gene silencing mechanisms (PTGS) are not so "avoidable". Furthermore, there is evidence that a high homology of construct sequence with host genome sequence - necessary for regulating transgene expression in the chloroplast genome - actually promotes methylation (a chemical modification of the DNA) that leads to inactivation of intrusive DNA [5]. The extent to which silencing effects are silenced in cpGE plants, particularly with PTGS, remains unknown.
High levels of transgene expression is one of the most hyped features of cpGE plants, making it a most attractive technology for the production of pharmaceuticals. The "hyperexpressivity" of chloroplast transgenes is largely attributed to the thousands of copies of the chloroplast genome in each cell. Most studies report a 30-100 times increase in expression of transgenes in cpGE compared with nuclear-transformed single-copy transgenic plants. Very high levels of chloroplast transgene expression have been reported, with alien protein products, as for example, Bt endotoxin, accounting for as much as 45% of the total soluble proteins within the cell [6]. While this overabundance of product is seen as advantageous for the industry, it does not consider the toxic effects of these proteins on the plant or on non-target organisms in the environment.
The over-expression of transgenes is of particular concern with marker genes, where a marker protein (e.g. conferring antibiotic resistance) could constitute up to 10% of a cell's total soluble protein [7]. In the environment, this would intensify the development of antibiotic resistance of bacteria. A new non-antibiotic marker system for selection of transformed plants has been proposed [8] but is not in current use, possibly because of the time, cost and effort required in developing it.
Chloroplast gene expression is known to involve complex RNA-editing, a process in which the RNA transcript is chemically modified so that its base sequence, and hence the protein translated from it, is completely different from that encoded in the gene (See "What's wrong with GMOs?", Science in Society 16, out now). RNA editing of chloroplast transcripts is known to be critical for the health of the host plant. Translation of unedited foreign messenger RNAs (from chloroplast transgenes) or the expression of transcripts under foreign processing controls can result in functionally defective proteins [9,10]. These deviant proteins could have toxic effects on the host plants, humans or other non-targets, or negate the intended efficacy of the protein as vaccines or drugs.
It should come as no surprise that the latest strategy to gain public acceptance of biotechnology is by marketing cpGE as an "environmentally friendly" technology [11]. The transmission of chloroplast genes to offspring is thought to be strictly maternal, avoiding the problem of transgene escape via the paternally-produced pollen. But while most angiosperms exhibit maternal inheritance of chloroplast genes, it is not always the case (see "Piffalls of transgene containment in chloroplast" by Joe Cummins, Science in Society 16, now out). "Trace levels" of paternal chloroplast inheritance is observed in what was once thought, strictly uniparental-maternal chloroplast inheritance in flowering plants [12]. In some species, paternal or mixed inheritance of chloroplast is observed under some conditions [13]. These instances have been considered too "rare" for consideration, despite the fact that they turned up in the very few investigations actually carried out.
Methods for introducing DNA into chloroplasts may also transform nuclear DNA. Although transgene constructs can be tailored for homologous recombination in the chloroplast genome, shooting transgenic DNA by 'gene gun' into the plant cells [14] does not, and cannot, specifically target transgenic DNA to the chloroplast genome. Much of the DNA ends up in the cytoplasm or the nucleus, where it may be available for integration into the nuclear genome. Another transgene delivery system relies on the uptake of the transgenic DNA into the cytoplasm in the presence of polyethylene glycol [15], which is then taken up, by an unknown mechanism, into chloroplasts. Again, the transgenic DNA in the cytoplasm could also be taken up into the nucleus.
Clearly the best way to determine the frequency of concomitant nuclear integration is through molecular analysis. The absence of signal in Southern blots has been cited as evidence that nuclear integration has not also taken place in some cases [8]. But the method used was not sensitive enough to detect low-copy number integration into the nuclear genome. 'Event-specific' characterization of the genetic structure of the transgenic insert is still the most definitive means to determine its location within the cell [16], whether the insert is in the chloroplast or the nuclear genome in the first instance.
One of the greatest challenges for GM technology has been the reliable performance of the transgenes over time. The maintenance of transgenic-homoplasmic plants (where all the chloroplast genomes are transformed) remains a clear problem. Experimental data documenting the stability of chloroplast transgenes are nonexistent.
Well-designed and transparent scientific studies on the stability and ecology of transgenes are needed before considering large-scale release into the environment, and especially so for transgenes inserted into the chloroplast genome. It requires an empirical burden of proof, not just the absence of evidence that is currently cited as proof. The inherent hyper-expressivity of the transgenes and the complexity of the regulation of chloroplast gene expression greatly increase the risks to health and the environment. Current research on cpGE is inadequate in validating its claims of being "environmentally friendly".
Article first published 21/10/02
Got something to say about this page? Comment