Modeling Risk Frames for Genetically Modified Foods

by Morgan Beltz

The regulation of genetically modified organisms (GMOs) is a constant battle of how to balance science with the consumers’ perception of risk. Scientists argue that there is no evidence to prove GM foods cause harm, but opponents reply there is no way to look at the potential future risks involved. Lisa Clark (2013) provides a solution to the governance framework by outlining three potential risk frames to use when regulating GMOs; 1) proof of harm, 2) precautious, and 3) precaution through experience. Currently, the regulation of GMOs and biotechnology is split between proof of harm and precautious risk frames, causing tensions between world governments and the correct way to govern the risks associated with GMOs. Clark concludes that individually the proof of harm and precautious risks frames are good, but focus on different aspects of the regulatory process. Bringing the two frames together to create a precaution through experience risk frame constructs the most solid foundation for a concise regulatory framework across different government bodies. Continue reading

Global Perceptions of Genetically Modified Foods

by Morgan Beltz

Public perceptions of genetically modified foods are not generally the same in different regions of the globe and can help dictate the availability of GM products. Frewer et al. (2013) conduct a systematic meta-analysis of 70 journal articles published all over the world, between the years of 1994 and 2010, to compare risk and benefit perceptions of different global regions. The authors focused on papers including agriculture genetic modification. The papers then went through a coding process to detect the levels of risks and benefits presented. The continent results were compared to the mean values of European participants in 2008. The authors found that North America and Asia have a lower risk perception of GM foods than Europe. North America also has a higher benefit perception of GM foods than Europe, but Asia has a lower benefit perception. Continue reading

Decrease GM Crop Research to Preserve Agriculture Biodiversity

The questions of whether to increase research on genetically modified crops or on increasing agricultural biodiversity as solutions to sustainable agriculture are currently topics of intense debate in the food industry. Jacobsen et al. (2013) collected studies from all over the world to assess the present pros and cons of continuing GM crop research and increasing agricultural biodiversity. ­­­­After compiling the studies, the authors found there are two obstacles to having sustainable agriculture 1) the claim that GM crops are vital to secure food production, and 2) the shortage of research funds for agriculture biodiversity in comparison to research funds for GM crops. Evaluating these two obstacles in regard to the pros, cons, and economics of GMOs, the authors conclude that research funding currently available to GM technologies would be better spent financing more efficient breeding techniques to cultivate agriculture biodiversity.—Morgan Beltz

Jacobsen, S., Sorensen, M., Pederson, S., Weiner, J., 2013. Feeding the world: genetically modified crops versus agricultural biodiversity. Agronomy for Sustainable Development 33(4), 651—662.

                  Jacobsen et al. claimed that there is burden on the environment from the current agriculture practices and increase in monoculture production, in particular the reduction of biodiversity through soil degradation. The authors recognize that GM crops can help solve this problem through technological advances in gene sequencing to improve adaptive capabilities in the crops. However, the authors argue that GM crops are financially impossible in many developing countries because large manufacturers have a monopoly on the industry and drive up seed cost to a point that is not affordable for small-scale farmers. For example, the authors have found a correlation between suicides of Indian farmers and the prices of Monsanto seeds which have turned out to be overly expensive and have not had the high yields expected by the farmers, leaving them in debt. The economic burden is high for developing countries because the seeds have lowered the labor needed, increased the price, but have not increased the yield.
                  In larger more regulated countries, such as Denmark, the authors found that GM crops can be beneficial in lowering production costs; sugar beets came out 80 Euros/hectare on top, potatoes 108, and maize was the only negative crop at -5 Euros/hectare. These results show that in a regulated established industry GM crops can save a lot of money in labor and production. However, it does not guarantee biodiversity and a sustainable way to produce food.
                   The authors argue that the only way to increase biodiversity is to study the factors that influence it; genotype, environment, and management. They noted that a study on Australian yields over the past 100 years shows that management was responsible for 50% of yield increases, genotype for 35%, and environment for 15%. The study also showed that farms with more agricultural diversity produced higher yields than monoculture farms thus, they argue,  funding should continue to go to research these areas of crop.

                  The authors conclude that all the evidence shows a need for a greater emphasis on agriculture biodiversity instead of GM crops and technologies. GM crops will add to the biodiversity loss instead of helping it and will reduce the nutritional value of the soil and hinder the yield. The authors believe that GM crop research should be a basic foundation to learn from for future applications, but not in the short term to increase the world food production. Instead, research should go towards improved agricultural practices, agricultural biodiversity though breeding techniques, and sustainable production. Developing countries are the target of increased food production and practices, and if they cannot afford the GM seeds and have a profitable yield, than the focus needs to change to other sustainable methods.

Using PCR testing to Detect Foods With Unlabeled GM Ingredients

Before 2009 Turkey had minimal regulations of genetically modified foods. In a study containing 26 processed saybean, maize, cotton, and canola products, Arun et al. (2013) found that 42.3% were positive for GM ingredients. After regulations went into place, the authors reevaluated the amount of GM ingredients in foods using a polymerase chain reaction (PCR) test. The authors found that of 100 samples only 25 contained GM ingredients, a significant decrease since the 2009 study.  After evaluating results before and after GM regulations went into place, the authors conclude the regulations have been sufficient in decreasing the amount of unlabeled GM ingredients in food. These results show the effectiveness of regulating GMOs during importation and that PCR is a sufficient way to test for GMOs in food.—Morgan Beltz
Arun, O., Yilmaz, F., Muratoglu, K., 2013. PCR detection of genetically modified maize and soy in mildly and highly processed foods. Food Control 32, 525-531.

                  Arun et al. collected 100 different processed food samples that contained maize, soy, or both. The authors used certified reference materials consisting of soybean powder and maize powder as negative and positive controls for comparison with the samples. For the mildly processed samples and controls the authors extracted and purified DNA using the Promega Wizard DNA isolation kit. Highly processed samples had DNA extracted using the cetyltrimethyl ammonium bromide method and then purified with the Promega Wizard DNA isolation kit. The extracted DNA from each sample was primed in a PCR for an amplification reaction. This process allowed the authors to target a specific strand of DNA in each sample to study it. Each amplified DNA strand was separated and stained through electrophoresis in gel that had ethidium bromide in order to view the macromolecules in the DNA strands. To account for false negatives the authors analyzed each negative sample further for presence of soy lectin and maize zein sequences which are present in all GM foods. The authors also put a cauliflower mosaic virus (CaMV) (a virus that almost all GM food contains) 35S promoter in control samples to compare to false negative results related to PCR inhibitors.
                  The authors found that 14 of 43 maize samples and 11 of 57 soy samples tested positive for GM ingredients, 25 percent overall. Sixty-three negative samples were confirmed as true negatives with specific amplification of the lectin and zein sequences. The other 12 negatives were negative for both lectin and zein, suggesting that they had little or no soy or maize DNA present.The authors also found that three of the 14 positive samples contained two different GM strands, not uncommon, as other studies have found. Arun et al.did have difficulty extracting DNA from 5 negative lectin soy samples, as have other researchers, but the samples were evaluated again to make sure the negatives were not false.
These results are similar to those found in other studies conducted after new regulations are imposed. Other researchers reported having similar problems with DNA extraction with high processed foods, having so many negatives, and having negatives that did not have any detectable maize and soy DNA. This leads the authors to believe that the extraction process can be improved. However, the consistency with other studies shows the accuracy of PCR testing and effectiveness for tracing GM foods through to importation.

Although it is positive that the regulations have led to a decrease in unlabeled GM foods, some remain in the market, leading the authors to call for additional regulations and stricter enforcements. The success of this PCR study shows the authors that it is an effective way to monitor the GM content in unlabeled foods.

Improving the Use of Phosphate in Alfalfa

Phosphate (P) has always been one of the most limiting factors in agricultural soil, and will become more of a constraint as world population grows and environmental concerns gain precedence. It is therefore imperative to improve how efficiently and carefully P fertilizers are applied, as well as how efficiently crop plants use those fertilizers. To explore this, Ma et al. (2012) caused alfalfa plants to express genes that are found in a model legume organism, Medicago truncatula, in the hopes that these genes would improve the use of organic phosphate (Po) by alfalfa plants. Po is generally derived from animal manure, and takes the form of phytate in soil, which is difficult for plants to use. The genes MtPHY1and MtPAP1 produce enzymes that can assist in the digestion of Po, and therefore were selected to express in the roots of transgenic alfalfa. Ultimately, it was concluded that transgenic lines of alfalfa demonstrated significantly better use of Po than control lines when grown in either lab conditions or active farm soil.—Chad Redman
                  Ma, X.F., Tudor, S., Butler, T., Ge, Y., Xi, Y., Bouton, J., Harrison, M., Wang, Z.Y. 2012. Transgenic Expression of Phytase and Acid Phosphatase Genes in Alfalfa (Medicago sativa) Leads to Improved Phosphate Uptake in Natural Soils. Mol Breeding 377, 377–391.
                  Ma et al. inserted genes from Medicago truncatula into the roots of alfalfa

plants in order to promote the use of Po. They chose to use genes that had demonstrated significant ability to increase P uptake in other plants previously, and focused on the use of transgenic plants in the real-world conditions of active farm soil. Specifically, the authors selected phytase (MtPHY1) and purple acid phosphatase (MtPAP1) genes, which both produce phytases, or enzymes that break down phytate into usable P. They also identified two different promoters, or regions of DNA that initiate the process of expressing genes, which could be used for both MtPAP1 and MtPHY1 and tested them independently for each of the genes. These promoters are identified as CaMV35S and MtPT1. Thus, Ma et al.created four different transgenic alfalfa lines, combining one gene with one promoter in all four combinations.

                  Researchers began with two different lab tests, the first growing plants in a medium without any P and the second in a medium supplied with Po. In the first condition, each of the transgenic plant lines along with a control line were grown in sand and fertilized with a solution that did not contain any form of P. After two weeks of growth under P-stressed conditions, the plants were harvested for their roots in order to isolate RNA and measure enzyme activity.
                  The second growth condition also utilized sand, but in this case plants were grown in the presence of fertilizer. The fertilizer composition was standard, except that it contained no inorganic phosphate (Pi), only Po. After six weeks of growth plants were harvested and all above-ground parts of the plant were dehydrated using an oven over the course of one week. The dried biomass of each plant group was recorded to compare use of Po.
                  After these lab-generated medium tests, plants were grown in pots containing soils from active farm ground. Two different soils were tested, one from Texas (soil 1) and the other from Oklahoma (soil 2). Soil 1 was generally less nutrient rich, and also more acidic. Soil 2 had a significantly higher concentration of usable P and a fairly neutral pH. In each soil, all four transgenic plant lines, along with control plants, were grown for three weeks without adding any nutrients to the soil. After three weeks, fertilizer without any form of P was applied in order to isolate the effects of P-stress on the plants. In order to determine biomass, two cuttings of the alfalfa plants were performed, the first after eight weeks and the second after another four weeks. These cuttings were dried and weighed. Additionally, the second cuttings of these plants were used to measure total P contained in the plants.
                  Ma et al. present some intriguing findings. From the RNA and enzyme analysis of plants grown in sand without any added P fertilizer, they find that transgenic plants on the whole produce far greater levels of APase, an enzyme that breaks down forms of phosphorous that plants cannot use into forms that may be digested. This observed difference was highly statistically significant. Moreover, there was an observed significant difference between the enzyme activities induced by the two different promoters. MtPT1 controlled genes had higher levels of APase than CaMV35S controlled genes. However, there was no observed difference between the two genes themselves. These results indicate that the transgenic alfalfa is, in fact, superior to the wild type alfalfa. Furthermore, the MtPT1 promoter is more effective in promoting efficient P use than the CaMV35S promoter.
                  Examination of plant roots revealed that transgenic plants showed higher phytase activity than control plants. Phytase is another enzyme that breaks down P into forms that may be easily used by plants, but phytase is specific to braking down phytate. Phytate is the most abundant source of P in soils, so phytase production is essential to effective use of P fertilizers. Interestingly, MtPHY1 plants exhibited higher phytate activity than MtPAP1 plants, which suggests that MtPHY1 transgene alfalfa is preferable.
                  When it comes to growth performance, transgenic plants consistently outperformed wild type alfalfa. As the above enzyme analyses would suggest, the plants with the MtPT1-MtPHY1 construct had the most vigorous growth and the highest dry biomass when grown in sand with Po readily available. Also, there was a strong correlation between enzyme activity and biomass, with transgenic lines clearly using the Po much more efficiently than the non-transgenic line.
                  In natural soils, those taken from active farm ground, similar patterns were observed. Especially in soil 1, transgenic alfalfa grew as expected, much larger and healthier than the control plants. In soil 2, the transgenes MtPHY1 and MtPAP1 were less effective because the pH of soil 2 was close to neutral, which is too basic for the enzymes to work at optimal levels. Moreover, because soil 2 contained a higher concentration of P to begin with, there was less P stress on wild type plants.
                  What Ma et al. have shown is that transgenic alfalfa can be used to alleviate the economic and environmental costs of applying large amounts of P fertilizer. With more efficient P use, particularly in alfalfa expressing the MtPT1-MtPHY1construct, pasture grounds will require less application of P, resulting in fewer weed control issues and also less runoff into aquatic ecosystems.

Toxic Effects of Glyphosate (Roundup)

Glyphosate is a popular and broadly used herbicide that is effective against weeds, especially in association with transgenic glyphosate-resistant crop systems.  Although glyphosate is considered a low-toxic herbicide, recent studies, such as one conducted by Romano et al., have revealed toxic effects resulting from low-dose commercial formulations.  The aim of the study was to investigate the effect of gestational maternal glyphosate exposure on the reproductive development of male offspring.  Sixty-day-old male rat offspring were evaluated for sexual behavior and partner preference; serum testosterone concentrations, estradiol, follicle-stimulating hormone (FSH) and luteinizing hormone (LH); the mRNA and protein content of LH and FSH; sperm production and the morphology of the seminiferous epithelium; and the weight of the testes, epididymis and seminal vesicles.  The study found that there was an increase in sexual partner preference scores and the latency time to the first mount testosterone and estradiol serum concentrations; the mRNA expression and protein content in the pituitary gland and the serum concentration of LH; sperm production and reserves; and the height of the germinal epithelium of seminiferous tubules.  These results suggest that masculinization processes, behavior, histological processes, and endocrine processes can all be negatively impacted by maternal exposure to glyphosate.
Romano, Marco Aurelio, et al. (2013) “Glyphosate impairs male offspring reproductive development by disrupting gonadotropin expression.” Archives of toxicology 86.4: 663-673. [GSSS romano gonadotropin glyphosate]

                  Sexual differentiation in the brain takes place from late gestation to the early postnatal days.  This process is dependent on the conversion of circulating testosterone into estradiol by the enzyme aromatase.  This process will eventually determine the gener-specific reproductive endrocrinology and behavior in adults.  A reduction in aromatase activity was observed in placental and embryonic human cells treated with low concentrations of a commercial formulation of glyphosate (Benachour et al. 2007).  From previous studies, the authors suspect that the herbicide glyphosate may be characterized as a potential endocrine chemical disruptor.  Endrocrine disruptors are defined as exogenous agents that interfere with the production, release, transport, metabolism, binding, action or elimination of natural hormones responsible for the maintenance of homeostasis and the regulation of developmental processes (Kavlock et al. 1996).
                  Romano et al.investigated the effect of gestational maternal glyphosate exposure on the reproductive development of male offspring.  Sexual differentiation in the brain occurs during the late gestational and the early postnatal days. Multiple factors can influence sexual expression.  First, sexual behavior is influenced by hormones, so the serum concentrations of testosterone, estradiol, FSH and LH were measured.  The pituitary expression of mRNA and protein content of LH and FSH was also analyzed to assess the possible glyphosate-mediated interference with their production.  Sex hormone serum concentrations may also affect sperm production and the morphology of the seminiferous epithelium, which were evaluated by testicular and epididymal sperm counts and the morphometric analysis of histological sections.  Lastly, the weight of the testes, epididymides and the seminal vesicle, the growth of the animals, and the weight and age at puberty were recorded to evaluate the effect of the treatment on these conditions.
First, to verify sexual partner preferences, the males were exposed at PND60 to a sexually mature male and a female in estrous.  The males from dams treated with glyphosate spent significantly more time in contact with female rats than control animals, suggesting a preference for the female gender.  Next, the glyphosate treatment led to an increase in the latency to first mount, latency to first intromission and latency to mount after first ejaculation.  Furthermore, the levels of both testosterone and estradiol were different between the control group and the glyphosate treated group.  Specifically, the group treated with glyphosate showed higher levels of testosterone and estradiol compared to control animals.  Next, the analysis of LH mRNA expression showed increased levels in treated animals, which was accompanied by higher amounts of LH protein in the pituitary and the serum.  In addition to this, the FSH mRNA expression was increased in treated animals, but this was not associated with a rise in the FSH protein in the pituitary or the serum. 
The changes in hormones observed may influence spermatogenesis.  Thus, the researchers monitored total sperm production, daily sperm production, sperm reserves and sperm transit at PND60.  Glyphosate exposure during the perinatal period increased the total and daily sperm production.  Also, an altered morphometry of the seminiferous epithelium was observed in treated animals.  This alteration caused an increase in epithelial height and a reduction in luminal diameter without changes in the tubular diameter.  In addition, the weight of the testes was found to be similar between groups.  The weight of the undrained seminal vesicle was not altered, but the drained seminal vesicle was heavier than the control group.  This finding suggests that this structure contained a smaller amount of fluid.  Interestingly, the glyphosate rats were observed to initiate puberty at an earlier age.  This change was accompanied by a reduction in their body weight.  However, the weights of the animals at the same age were not different, indicating that observed lower weight is merely a function of the younger age at puberty onset.

This experiment indicated changes in most of the parameters evaluated.  The results suggest that maternal exposure to glyphosate disturbed the masculinization processes.  The authors conclude that glyphosate exposure promotes behavioral changes and histological and endrocrine problems.  Further study is suggested to evaluate whether the effects of maternal exposure to glyphosate are dose-dependent.

Works Cited
Benachour, Nora, et al.“Time-and dose-dependent effects of roundup on human embryonic and placental cells.” Archives of Environmental Contamination and Toxicology 53.1 (2007): 126-133.
Kavlock, Robert J., et al. “Research needs for the risk assessment of health and environmental effects of endocrine disruptors: a report of the US EPA-sponsored workshop.” Environmental health perspectives 104.Suppl 4 (1996): 715.

Safety of GM Crops

Genetically modified biotechnology is the fastest-adopted technology in the history of modern agriculture.  In 1996 there were 1.7 million hectares of GM crops; since then that number has increased to 148 million hectares—an 87-fold increase.  But, there is some concern surrounding the potential side effects of randomly inserting exogenous genes in plant genomes.  Mainly, an insertion of exogenous genes could produce modified biochemical processes, new proteins, or other secondary pleiotropic effects.  Evaluating the substantial equivalence of GM groups to traditional crops is therefore essential to guarantee the safe use of GM crops and alleviate the fears consumers have about GM food.  Here, the researchers evaluated the effects of transgenes on rice seed proteomes by 2-D differential in-gel electrophoresis (2D-DIGE) combined with mass spectrometry (MS). The study found that GM events do not substantially alter proteome profiles as compared with conventional genetic breeding and natural genetic variation.  Specifically, mass spectrometry revealed 234 proteins differentially expressed in the 6 materials (BAR68-1, D68, 2036-la, MH86, MH63, ZH10), which are involved in different cellular and metabolic processes.  This finding suggests that metabolism, protein synthesis and destination, and defense response in seeds are important in differentiating rice cultivars and varieties.
Gong, Chun Yan, et al. “Proteomics insight into the biological safety of transgenic modification of rice as compared with conventional genetic breeding and spontaneous genotypic variation.” Journal of Proteome Research 11.5 (2012): 3019-3029. [GSSS gong proteomics insight rice]

                  Studies of diverse plant species have demonstrated that changes in transcript levels are not fully followed by the same changes in protein levels.  Proteins are the key players in gene function and are directly involved in metabolism and cellular development or have roles as toxins, antrinutrients, or allergens.  Therefore, comparisons of GM proteomes and control lines are of necessary importance when trying to determine GM crop safety.  The results of these experiments can reveal molecular differences in varieties produced by conventional genetic breeding and natural genetic variation and help researchers better assess the safety of a GM crop. Here, the researchers evaluated the effects of transgenes on rice seed proteomes by 2-D differential in-gel electrophoresis (2D-DIGE) combined with mass spectrometry (MS).  Two sets of GM indica rice and controls were used: Bar68-1 transformed with herbicide resistant gene bar and its non transgenic control indica variety D68, and 2036-la transformed with insect-resistant genes cry1Ac/sckand its nontransgenic control indicavariety MingHui 86 (MH86).  In addition to these, the researchers used MH63, which is a parental line used for breeding MH86, and japonica rice Zhonghua 10 (ZH10).  In summary, the experimental design included GM rice and controls, different indica varieties (parental and filial), and indica and japonicacultivars.
                  First, the researchers confirmed that their transgenic lines were successfully transformed.  PCR with specific primers revealed BAR68-1 had one detectable DNA fragment with a size of 568 bp, which corresponded to the bar gene for herbicide resistance.  2036-la line had 2 detectable DNA fragments of 1709 bp and 358 bp corresponding to cry1Ac and sck genes, respectively (coding for insect resistance).
                  Next, a 2D-DIGE with pH 4–7 strips analyzed seed proteomes from different rice lines and detected about 2250 protein sports in each image.  A principal component analysis (PCA) was conducted to investigate similarities in the proteomes of the 6 rice lines.   Much less variation was found in the proteomes between transgenic lines and their controls than between different indica varieties or between indica and japonica cultivars.  In addition to this, the researchers analyzed differentially expressed proteins (DEPs) from the seed proteomes of the 6 lines which revealed 423 (Student’s t test) and 443 (ANOVA) protein spots, respectively, with statistically significant differences in expression.  The largest differences of protein expression were found between indica (varieties NH63, D68, and MH86) and japonica (ZH10) cultivars.  A smaller difference was found between the 3 indicavarieties, an even smaller difference between NH63 and MH86, and the least between transgenic lines and controls (Bar68-1 vs D68; 2036-la vs MH86).  Compared with conventional genetic breeding and natural genetic variation, rice seed proteomes were largely unchanged with transgenic modification.
                  In addition to the principal component analysis, mass spectrometry was used to analyze 264 differentially expressed proteins (DEPs) selected on the basis of (1) 1.2-fold changes in expression and (2) significant difference by both Student’s ttest and ANOVA.  These proteins were classified into eight functional categories: metabolism, protein synthesis and destination, defense response, cell growth and division, pyruvate orthophosphate dikinases (PPDKs), signal transduction, transcription, and transporters.  Most of the DEPs were involved in metabolism (31.2%), protein synthesis and destination (25.2%), and defense response (22.4%).  In summary, proteins implicated in central carbon metabolism, starch synthesis, protein folding and modification, and defense response showed altered expression in response to natural genetic variation, conventional breeding, and transgene modification.
                  To examine the identified DEPs in more detail, the authors analyzed the expression patterns of 218 DEPs using GeneCluster 2.0.  The DEPs were grouped into 6 clusters: c0, c1, c2, c3, c4, c5. The 6 clusters were further grouped into three antagonistic pairs (clusters pairs: c1 vs c3, c1 vs c4, c2 vs c5).  Proteins in c0c3 sets were assumed to contribute to the variability between D68/MH63 and MH86/ZH10.  Proteins grouped into c1c4 sets possibily separated MingHui from D68/ZH10.  Furthermore, the c2c5 sets may be the main contributors to the separation between japonica and indica rice.  Although there were large changes in expression of these proteins among nontransgenic varieties, their expression was similar between transgenic lines and their respective controls in all clusters.
                  The authors further performed PCA for the 218 DEPs to estimate the contribution of DEPs to the total variability observed within rice lines and identify proteins responsible for the variability.  The results of the PCA confirmed the authors finidings from cluster analysis that proteins in 3 clusters pairs (c1c4, c2c5, and c0c3) may have a distinct contribution to the entire variability of the data set. 
                  Next, the authors evaluated the distribution on 218 DEPs involved in different function categories and subcategories in cluster pairs c0c3, c1c4, and c2c5.  The results of this analysis clearly showed different functional categories and subcategories-related proteins distributed heterogeneously in these cluster pairs.  For example, metabolism-related proteins were mainly in c1c4 and c2c5, but less so in c0c3.  Protein synthesis and destination-related proteins were more often in c0c3 and c2c5 than in c1c4. Most of the defense response-related proteins were in c0c3 and c1c4.  Thus, the different functional categories in these clusters confirmed differences in biological processes in all analyzed rice lines.
                  To investigate the changes in biological processes as a result of natural genetic variation, conventional breeding, and transgene modification, the authors performed expression profile analysis of protein groups associated with nine functional categories and subcategories showing significant contribution to the differences.  The expression patterns of proteins involved in glycolysis and starch synthesis were similar, with relatively high levels in 2 nontransgenic MinHui varieties and low levels in D68 and ZH10. The opposite relationship was found with proteins involved in the TCA pathway.  All in all, the expression of proteins in transgenic lines and their respective controls was changed, but the changes were similar to those observed between certain non-transgenic varieties.
                  All these data combined suggested that GM does not significantly alter the rice seed proteomes as compared with natural genetic variation and conventional genetic breeding.  Specifically, the integration in rice genomes and expression of bar or cry1Ac/sck do not change the proteome patterns as compared with natural genetic variation and conventional breeding.  Apart from the safety conclusions of this experiment, the results also show that the proteins differentially expressed in nontransgenic rice varieties had functions in central carbon metabolism, starch synthesis, protein folding and modification, and defense response.  For future experiments, these processes can be further investigated to differentiate and explore rice varieties.