5.1 Introduction to the GMOs

26. The applicant proposes to release up to 1241 GM banana lines, comprising of 1221 GM Cavendish (cultivar Dwarf Cavendish or Williams) and 20 GM Pome (cultivar Lady Finger) banana lines. The genes introduced into the GM banana lines proposed for release are listed in Table 1 and the promoters and other genetic elements used for expressing these genes are listed in Table 2. Details of the constructs used to generate the GM banana lines are provided in Table 3.

27. Up to 836 of the GM banana lines contain from one up to three genes that are expected to enhance β-carotene content in banana while 215 lines contain one or two genes to increase iron. Up to 160 lines engineered for enhancing both iron and β-carotene will contain at least one gene for each trait and up to three genes in total.

28. The remaining 30 GM banana lines contain the uidA reporter gene under the Exp1 promoter. These plants will be used to study tissue specificity of the Exp1 promoter function.

29. In addition, all of the GM banana lines contain the antibiotic resistance selectable marker gene neomycin phosphotransferase type II (nptII) from the common gut bacterium Escherichia coli. This gene, encoding the enzyme neomycin phosphotransferase, confers kanamycin or neomycin resistance on GM plant cells. The nptII gene was used during initial development of the GM plants in the laboratory to select plant cells containing the introduced genes.

30. Short regulatory sequences would be used to control expression of the introduced genes or to direct proteins to appropriate cellular compartments. These sequences are derived from plants (maize, banana, castor bean, pea and chrysanthemum), a soil bacterium (Agrobacterium tumefaciens) and the plant virus cauliflower mosaic virus (CaMV; see Table 2).

Table 1. The genes introduced into the GM banana lines proposed for release.

Gene Gene – full name Source organism Intended function
DXS Deoxy-xylulose-5-phosphate synthase Thale cress (Arabidopsis thaliana) pro-vitamin A biosynthesis
PsyB73 Phytoene synthase 1 Maize (Zea mays) inbred line B73 pro-vitamin A biosynthesis
Psy1Q60 Phytoene synthase 1 Maize (Z. mays) inbred line Q60 pro-vitamin A biosynthesis
APsy2a Phytoene synthase 2 Banana (Musa spp.) cultivar Asupina pro-vitamin A biosynthesis
Crtl Phytoene desaturase Erwinia uredovora pro-vitamin A biosynthesis
LYCB Lycopene β-cyclase Rice (Oryza sativa) pro-vitamin A biosynthesis
Ferritin Ferritin Wild soybean (Glycine soja) Fe-assimilation
NAS Nicotianamine synthase Rice (O. sativa) Fe-assimilation
FEA1 Fe-assimilation 1 Chlamydomonas reinhardtii Fe-assimilation
uidA β-glucuronidase gene Escherichia coli Reporter gene
nptII Neomycin phosphotransferase type II gene E. coli Selectable marker


Table 2. Genetic elements used in GM banana lines proposed for release.

Element Source gene Function Souce oganism
Nos Nopaline synthase Promoter Agrobacterium tumefaciens
35S Cauliflower mosaic virus 35S Promoter Cauliflower mosaic virus
Exp1 Expansin 1 Promoter Banana (Musa acuminata) cultivar Williams
Exp4 Expansin 4 Promoter Banana (Musa acuminata) cultivar Cavendish
Ext Extensin Promoter Banana (Musa acuminata) cultivar Cavendish
MT2A Metallothionein Promoter Banana (Musa acuminata) cultivar Cavendish
ACS 1-aminocyclopropane-1-carboxylate synthase Promoter Banana (Musa acuminata) cultivar Williams
ACO 1-aminocyclopropane-1-carboxylate oxidase Promoter Banana (Musa acuminata) cultivar Williams
Ubi Polyubiquitin Promoter Maize (Zea mays)
Nos Nopaline synthase Terminator Agrobacterium tumefaciens
35S Cauliflower mosaic virus 35S Terminator Cauliflower mosaic virus
Cat Catalase Intron Castor bean (Ricinus communis)
CMSSP Chrysanthemum morifolium small subunit protein Signal peptide Chrysanthemum (Chrysanthemum morifolium)
Rbcs-CTP Rubisco small subunit – chloroplast targeting peptide Signal peptide Pea (Pisum sativum)


Table 3. Gene constructs used to generate the GM banana lines proposed for release<sup>&#42</sup>

Lines previously released under DIR 076/2007
Cultivar Construct(s) Gene(s) Promoter(s) Additional genetic elements` Max. lines Replicates per line Total plants
Dwarf Cavendish pCAM-Exp1-APsy2a APsy2A Exp1 Nos 5 10 50
pCAM-ACO-APSy2a APsy2A ACO Nos 5 10 50
pCAM-Ubi-APsy2a APsy2A Ubi Nos 2 10 20
pCAM-Exp1-PsyB73 PsyB73 Exp1 Nos 5 10 50
pCAM-ACO- PsyB73 PsyB73 ACO Nos 5 10 50
pCAM-Ubi-PsyB73 PsyB73 Ubi Nos 4 10 40
pCAM-Ubi-PsyB73 PsyB73 Ubi Nos 4 10 40
pCAM-ACO-APSy2a
pBin-Exp1-Crt1
APsy2A
CrtI
ACO
Exp1
Nos
CMSSP, Nos
5 10 50
pCAM-ACO-PsyB73
pBin-Exp1-Crt1
PsyB73
CrtI
ACO
Exp1
Nos
Nos
5 10 50
pCAM-Exp1-Ferritin Ferritin Exp1 Nos 5 10 50
Lady Finger pCAM-Exp1-APsy2a APsy2A Exp1 Nos 5 10 50
pCAM-ACO-APSy2a APsy2A ACO Nos 5 10 50
pCAM-Exp1-PsyB73 PsyB73 Exp1 Nos 5 10 50
pCAM-ACO- PsyB73 PsyB73 ACO Nos 5 10 50
Lines newly developed for release
Williams pOPT-A Psy1Q60
CrtI
Exp1
Exp1
Nos
Rbcs-CTP, Nos
10 1 10
pOPT-B Psy1Q60
CrtI
Exp1
Exp4
Nos
Nos
10 1 10
pOPT-C Psy1Q60
CrtI
Ferritin
Exp1
Exp1
Exp1
Nos
Nos
Nos
10 1 10
pOPT-D APsy2A
CrtI
Exp1
Exp1
Nos
Nos
10 1 10
pOPT-F APsy2A Exp4 Nos 30 1 30
pOPT-G uidA-Cat Exp1 Cat, Nos 30 1 30
pOPT-H Psy1 Q60
Ferritin
ACS
Exp1
Nos
Nos
30 1 30
pOPT-I Psy1 Q60 ACS Nos 30 1 30
pOPT-J Psy1 Q60 Exp1 Nos 30 1 30
pOPT-K APsy2A Exp1 Nos 30 1 30
pOPT-L Psy1 Q60
Ferritin
Exp1
Exp1
Nos
Nos
30 1 30
pGen2-A Apsy2a Exp1 Nos 30 1 30
pGen2-A Apsy2a Exp1 Nos 30 1 30
pGen2-B Psy1 Q60 Exp1 Nos 30 1 30
pGen2-C Ferritin Exp1 Nos 30 1 30
pGen2-D Ferritin MT2a Nos 30 1 30
pGen2-E Ferritin ACO Nos 30 1 30
pGen2-F Apsy2a MT2a Nos 30 1 30
pGen2-G Apsy2a Exp1 Nos 30 1 30
pGen2-CA Ferritin
Apsy2a
Exp1
Exp1
Nos
Nos
30 1 30
pGen2-DA Ferritin
Apsy2a
MT2a
Exp1
Nos
Nos
30 1 30
pGen2-AF Apsy2a
Apsy2a
Exp1
MT2a
Nos
Nos
30 1 30
pGen2-CD Ferritin
Ferritin
Exp1
MT2a
Nos
Nos
30 1 30
pGen3-EA Ferritin
Apsy2a
ACO
Exp1
Nos
Nos
30 1 30
pGen3-H DXS Exp1 Nos 30 1 30
pGen2-I LYCB Exp1 Nos 30 1 30
PGen3-HI DXS
LYCB
Exp1
Exp1
Nos
Nos
30 1 30
pGen3-HA DXS
Apsy2a
Exp1
Exp1
Nos
Nos
30 1 30
pGen3-AI Apsy2a
LYCB
Exp1
Exp1
Nos
Nos
30 1 30
pGen3-HAI DXS
Apsy2a
LYCB
Exp1
Exp1
Exp1
Nos
Nos
Nos
30 1 30
pGen3-J DXS MT2a Nos 30 1 30
pGen3-K LYCB MT2a Nos 30 1 30
pGen-JK DXS
LYCB
MT2a
MT2a
Nos
Nos
30 1 30
pGen3-JF DXS
Apsy2a
MT2a
MT2a
Nos
Nos
30 1 30
pGen3-FK Apsy2a
LYCB
MT2a
MT2a
Nos
Nos
30 1 30
pGen3-JFK DXS
APsy2A
LYCB
MT2a
MT2a
MT2a
Nos
Nos
Nos
30 1 30
pGen2-L Apsy2a ACO Nos 30 1 30
pGen2-M Psy1Q60 ACO Nos 30 1 30
pGen2-OL APsy2a
CrtI
ACO
Exp1
Nos
Rbcs-CTP,Nos
30 1 30
pGen2-OM Psy1Q60
CrtI
ACO
Exp1
Nos
Rbcs-CTP,Nos
30 1 30
pGen4-PC Ferritin
NAS
Exp1
Ubi
Nos
Nos
30 1 30
pGen4-QC Ferritin
NAS
Exp1
Exp1
Nos
Nos
30 1 30
pGen4-RC Ferritin
FEA1
Exp1
Ubi
Nos
Nos
30 1 30
*All lines contain the nptII gene with promoter and terminator from either the CaMV 35S or the A. tumefaciens nos gene

5.2 Introduction to pro-vitamin A and its biosynthesis

31. Vitamin A is an essential nutrient needed in small amounts for normal functioning of the visual system, growth, differentiation, maturation, reproduction and immunity among others (FAO/WHO 2001). Vitamin A is derived from pro-vitamin A carotenoids which are synthesized by plants and many species of bacteria, fungi and archaea. Human and most animals cannot produce these carotenoids and their need for vitamin A must be met through dietary intake. Carotenoid biosynthesis genes have recently been discovered in aphids (Moran & Jarvik 2010).

32. Dietary needs of vitamin A are normally supplied by the consumption of pre-formed vitamin A (from foods derived from animals e.g. milk, meats, liver and egg yolks) and pro-vitamin A carotenoids (from foods derived from fruits and vegetables e.g. green leafy vegetables and yellow/orange fruits such as pumpkin, carrot, mango and papaya). The recommended daily intake (RDI) of vitamin A as Retinol Equivalents (RE) for adult (19+ years) men is 900 µg/day and for adult women is 700 µg/day (NH&MRC 2006).

33. Of the major pro-vitamin A carotenoids (α-carotene, β-carotene and β-cryptoxanthene), β-carotene is most efficiently converted to vitamin A (Hess et al. 2005). Some examples of the levels of β-carotene contained in commonly consumed plant (or plant-derived) foods are given in Table 4. The Fe’i banana (Musa troglodytorum L.) cultivar ‘Asupina’ that is consumed in Micronesia, and from which the APsy2a gene has been sourced, contains 1,412 µg β-carotene/100 g (259 RE) and meets the estimated vitamin A requirements for an adult woman within normal consumption patterns (Englberger et al. 2006a). Several other Fe’i cultivars also contain very high levels of β-carotene with ‘Utin Iap’ and ‘Utimwas’ containing approximately 8,000 µg β-carotene/100 g and ‘Karat’ containing approximately 2,000 µg β-carotene/100 g (Englberger et al. 2006b).

Table 4. Levels of β-carotene, retinol equivalents (RE) and iron in commonly consumed plant-based foods (value per 100g of fresh weight)*.

Food Source β-carotene (µg) RE (µg)1 Iron (mg)
Banana - Cavendish (peeled fruit) 64 11 0.5
Blackberry - fresh 150 53 0.4
Bok Choy (stir-fried) 692 116 2.1
Breakfast cereal (mixed grain flakes), iron-fortified 3 0 8.1
Broccoli (boiled) 273 46 0.8
Canola Oil 0 0 0
Capsicum (red, fresh) 282 215 0.3
Carrot (fresh, mature) 5,996 1,316 0.3
Cashew 6 1 5.0
Kidney bean (canned) 0 0 2.1
Lettuce - Iceberg 120 23 0.6
Mango (fresh) 2,195 366 0.3
Olive Oil 13 2 0
Pasta (fresh, white flour) 0 0 0.7
Soybean (canned) 0 0 1.8
Spinach – English (boiled) 2,201 376 3.5
Walnut 21 4 2.5
* data taken from NUTTAB 2006 (FSANZ 2006)

1Vitamin A intakes or requirements are expressed as Retinol Equivalents (RE) (NH&MRC 2006)

34. In higher plants carotenoids are synthesised in the plastids from five carbon isopentyl di-phosphate (IPP) units (Giuliano et al. 2008; Figure 2). In plastids IPP is generated via the methyl erythritol phosphate (MEP) pathway (Rodriguez-Concepcion & Boronat 2002; Cunnigham 2002). The first step of this pathway is condensation of pyruvic acid and glyceraldehyde-3-phosphate to form 1-deoxy-xylulose-5-phosphate (DOXP). This is the rate limiting step of this pathway and is catalysed by 1-deoxy-xylulose-5-phosphate synthase (Estevez et al. 2001; Cazzonelli & Pogson 2010). DOXP is then converted to MEP, the first pathway specific compound, through function of DOXP reductoisomerase. MEP is finally converted to IPP and its isomer dimethylallyl di-phosphate (DMPP), through several pathway intermediates.

35. IPP is the building block for terpenoids, quinones, chlorophyll, plant hormones like gibberellins, cytokinins and various pigments including carotenoids (Roberts 2007; Tanaka et al. 2008; Giuliano et al. 2008). In the carotenoid biosynthesis pathway (Cunnigham 2002; Giuliano et al. 2008) four molecules of IPP are condensed to the 20 carbon molecule geranylgeranyl di-phosphate (GGPP) in a reaction catalysed by GGPP synthase. Two GGPP molecules are further coupled by the enzyme phytoene synthase to form phytoene, the first dedicated compound of the carotenoid pathway. Phytoene is then converted into lycopene (tetra-cis isoform) through sequential action of phytoene desaturase and ζ-carotene desaturase. Carotenoid cis-trans isomerase enzymes then convert the tetra-cis-lycopene to all-trans-lycopene. Bacterial phytoene desaturase can perform all these desaturase and isomerase functions and hence alone can convert phytoene to all-trans-lycopene (Ye et al. 2000). All-trans lycopene is the substrate for two competing enzymes, lycopene ε-cyclase and β-cyclase. These two enzymes can act in concert to convert all-trans lycopene to α-carotene by addition of one ε-ring at one end and one β-ring at the other. Alternatively lycopene β-cyclase alone can add two β-rings, one at each end of all-trans lycopene, to form β-carotene (DellaPenna & Pogson 2006). The relative abundance of α- or β-carotene thus depends on the relative activities of these two enzymes

This figure outlines the β(beta)-carotene biosynthesis pathway along with the enzymes and intermediate metabolites as described in Section 5.2.
      Figure 2. Biosynthesis of β-carotene.
Enzymes and encoding genes used in GM banana lines proposed for release are in orange. Refer to text (Section 5.2) for details.

5.3 Introduction to iron and its assimilation in plants

36. Iron (Fe) is an important micronutrient in all living organisms for its role in numerous cellular functions. In animals it serves as a carrier of oxygen from the lungs to the tissues by red blood cell haemoglobin; as a transport medium for electrons within cells; and as an integrated part of important enzyme systems in various tissues (FAO/WHO, 2001). At the same time excess iron has toxic effects in cells through generation of free radicals via the Fenton reaction (Hell & Stephan 2003; Palmer & Guerinot 2009).

37. Iron deficiency is a major concern to human nutrition all over the world. The RDI of iron is 18 mg/day in women aged 19 – 50 years and is 8 mg/day in adult men (19+ years) and women over 50. The upper level of intake (UL) for adult men and women is set at 45 mg/day (NH&MRC 2006).

38. Haem iron is found only in meat, fish and poultry and is absorbed much more easily than non-haem iron, which is found primarily in fruits, vegetables, dried beans, nuts and grain products (NH&MRC 2006). Some plant dietary sources of iron are given in Table 4. By comparison with these plant sources, the level of iron in lean red meat such as lamb or beef, which is regarded as the best dietary source of iron, varies from 2.2 – 3.3 mg/100 g (FSANZ 2006).

39. Due to high toxicity associated with excess iron, its uptake, transport and storage are highly regulated. One of the major ways of dealing with excess iron in cells is sequestration. Ferritin is the principal iron storage protein in all organisms which forms a nanocage capable of binding 4500 atoms of iron in its interior. When orally administered, ferritin can provide a source for iron. Over-expressing ferritin has been attempted in a bid to increase iron content in rice with modest success (Goto et al. 1999; Vasconcelos et al. 2003; Drakakaki et al. 2005). However, these studies also suggest that enhancing iron uptake as well as translocation is needed for further bio-fortification of rice with this essential mineral.

40. The main difficulty associated with iron uptake in plants is low solubility of the Fe3+ form prevalent in soil (Hell & Stephan 2003; Kim & Guerinot 2007). To circumvent this, non-graminiceous plants employ a reduction based strategy (Hell & Stephan 2003; Kim & Guerinot 2007). These plants release H+ ions into surrounding soils which reduce soil pH and increase solubility of Fe3+ ions. Soluble Fe3+ is then reduced to Fe2+ by Fe3+-chelate reductase and taken up into the cells by iron transporters. Graminiceous plants e.g. rice, wheat etc. employ an alternate chelation-based strategy (Hell & Stephan 2003; Kim & Guerinot 2007) where they secrete small molecular weight compounds known as mugineic acid (MA) to chelate Fe3+ in the rhizosphere. These Fe3+-MA complexes are then transported by specific transporter proteins.

41. In graminiceous plants mugineic acids e.g. deoxy-mugineic acid (DMA) are derived from nicotianamine (Hell & Stephan 2003; Kim & Guerinot 2007). Nicotianamine is synthesised from S-adenosyl methionine by the enzyme nicotianamine synthase. Apart from that, nicotianamine is also a universal chelator of metal ions including iron, zinc etc. and hence plays an important role in metal translocation and homeostasis. Thus, an increase in nicotianamine through genetic modification is envisaged as another strategy to increase iron in rice (Bashir et al. 2010).

5.4 The introduced genes and their encoded proteins

42. The genes and their encoded proteins are described in brief to illustrate their potential function within the GM banana lines. They have been grouped according to the trait associated with the introduced genes.

5.4.1 The introduced genes for pro-vitamin A biosynthesis and their encoded proteins

43. These genes encode various enzymes that are involved in biosynthesis of pro-vitamin A and are expected to increase pro-vitamin A content in GM banana. Some of these genes are currently being trialled under DIR 076/2007 as indicated below.

The DXS gene

44. The DXS gene encodes the deoxy-D-xylulose 5-phosphate synthase enzyme which catalyses the first step of the MEP pathway of IPP biosynthesis and is derived from A. thaliana. Constitutive over-expression of the endogenous gene in Arabidopsis enhanced levels of many plastidic isoprenoids including carotenoids (Estevez et al. 2001). Fruit-specific expression of E. coli DXS gene in tomato also resulted in up to 1.6 fold increase in phytoene and β-carotene (Enfissi et al. 2005). Over-expression of E. coli DXS gene also enhanced carotenoid levels in potato tubers (Morris et al. 2006). In contrast, expression of the daffodil DXS gene in rice did not achieve the expected increase in β-carotene level (Al-Babili & Beyer 2005).

The PsyB73, Psy1Q60 and APsy2a genes

45. Synthesis of phytoene by phytoene synthase is the first committed step of carotenoid biosynthesis and is generally accepted as the most important regulatory enzyme in the pathway (Cazzonelli & Pogson 2010). This enzyme is encoded by the phytoene synthase gene (Psy) and over-expression of this gene has been a successful strategy for increasing carotenoid levels in a number of crop plants.

46. The PsyB73 gene derived from maize inbred line B73 and the Psy1Q60 gene from maize cultivar Q60 both encode phytoene synthase enzymes and have identical amino acid sequences (Buckner et al. 1996). The PsyB73 gene was used in development of the high β-carotene Golden Rice II (Buckner et al. 1996; Paine et al. 2005). The APsy2a is the phytoene synthase II gene from banana cultivar Asupina and has not been reported in the literature. PsyB73 and APsy2a genes are currently being evaluated under DIR 076/2007 and preliminary data suggests that GM banana expressing either of these two genes have higher pro-vitamin A content (information supplied by the applicant). This also indicates that proteins encoded by PsyB73 and APsy2a genes perform the same function.

47. Over-expression of phytoene desaturase from a number of sources, e.g. daffodil (Narcissus pseudonarcissus), maize, tomato, rice, has been studied for modifying β-carotene levels in rice (Ye et al. 2000; Paine et al. 2005). The maize Psy1B73 gene was also used for increased carotenoid expression in GM wheat (Cong et al. 2009). The bacterial homolog of Psy gene, crtB from E. uredovora has been used for increasing total carotenoid as well as β-carotene content in tomato (Fray et al. 1995; Fraser et al. 2002), canola (Shewmaker et al. 1999), potato (Ducreux et al. 2005) and maize (Aluru et al. 2008).

The CrtI gene

48. CrtI gene from E. uredovora codes for phytoene desaturase and generates all-trans-lycopene directly from phytoene by combining the functions of phytoene desaturase, ζ-carotene desaturase and lycopene-cis-trans-isomerase (Giuliano et al. 2008). This gene will be expressed in GM banana as a fusion protein with the chloroplast targeting signal peptide of ribulose-1,5-bisphosphate carboxylase small subunit from chrysanthemum (CMSSP; Outchkourov et al. 2003) or pea (Rbcs-CTP; Coruzzi et al. 1984; Paine et al. 2005). The CrtI gene fused to Rbcs-CTP was employed along with Psy genes to enhance pro-vitamin A level in Golden Rice and Golden Rice II (Ye et al. 2000; Paine et al. 2005). A similar strategy has been successfully employed to increase total carotenoids and β-carotene levels in potato and maize and total carotenoids in wheat (Diretto et al. 2007; Aluru et al. 2008; Cong et al. 2009). This gene fused with either of the chloroplast targeting signals is currently being assessed under DIR 076/2007.

The LYC gene

49. The LYBC (lycopene beta cyclase) gene is obtained from rice and the encoded enzyme catalyses preferential conversion of lycopene to β-carotene over α-carotene (DellaPenna & Pogson 2006). However, over-expression of LYBC gene from daffodil was found not to be essential in increasing pro-vitamin A levels in Golden Rice as endogenous expression of this gene was sufficient for the conversion (Ye et al. 2000; Al-Babili & Beyer 2005). In contrast, plastidic expression of this daffodil gene in GM tomato resulted in higher conversion of lycopene to β-carotene (Apel & Bock 2009). The homologous CrtY gene from E. uredovora was also used in conjunction with CrtB and CrtI to enhance β-carotene levels in GM potato, although individual contributions from these genes were not studied in detail (Diretto et al. 2007).

5.4.2 The introduced genes for increasing iron content and their encoded proteins

The Ferritin gene

50. The Ferritin gene from wild soybean codes for a ferritin protein identical to the ferritin from cultivated soybean (Glycine max). Ferritin is a major iron-storage protein in both plants and animals. In plants, the level of ferritin is regulated at the transcriptional level and hence over-expression of this gene is a straightforward strategy to increase ferritin content. The Ferritin gene from cultivated soybean has been used for endosperm specific over-expression in rice, resulting in up to 3–fold increase in iron content in rice seeds (Goto et al. 1999; Vasconcelos et al. 2003). There was also an accompanying increase in zinc concentration (Vasconcelos et al. 2003). However, when constitutively over-expressed in rice and wheat, iron levels were increased in leaves but not grains (Drakakaki et al. 2000).

The NAS gene

51. Some of the GM bananas in the proposed release will contain a NAS gene from rice. The NAS gene encodes the enzyme nicotianamine synthase, which catalyses formation of nicotianamine (NA), a chelator of metal ions including iron and a precursor of plant deoxy-mugineic acid (DMA), another metal ion-chelator. Over-expression of the NAS gene as a strategy to increase NA levels, and hence iron uptake, has been attempted in several cases. Ectopic over-expression of the Arabidopsis NAS1 gene in tobacco resulted in higher NA concentration, higher iron, zinc and manganese concentration in leaves and better tolerance to iron deficiency (Douchkov et al. 2005). Similarly, rice plants over-expressing a barley NAS1 gene demonstrated higher levels of endogenous NA and DMA and a 2–3 fold increase in iron and zinc levels in seeds (Masuda et al. 2009). Wirth and colleagues (2009) reported up to 6-fold increase in seed iron content in GM rice co-expressing Arabidopsis NAS1 and bean (Phaseolus vulgaris) Ferritin genes. Endosperm specific over-expression of the endogenous NAS1 gene in GM rice not only lead to increased iron and zinc levels in unpolished grains, but also leads to increased bioavailability of iron in in-vitro studies (Zheng et al. 2010).

The FEA1 gene

52. The FEA1 gene is cloned from C. reinhardtii, a photosynthetic alga in which it is highly expressed under iron deficient conditions (Rubinelli et al. 2002). The encoded protein FEA (Fe assimilation) is secreted into the medium and is hypothesised to bind iron to facilitate its uptake by iron transporters (Allen et al. 2007). Over-expression of FEA1 in Arabidopsis rescued lethal iron transporter mutations and also GM plants demonstrated better growth than wild type under iron deficient conditions (Chiu 2007). Similar better performance under iron-limiting growth conditions suggesting better iron uptake was observed in GM cassava expressing FEA1 (information supplied by the applicant). GM banana plants proposed for release will contain a FEA1 gene that is codon optimized for expression in plants. However, the amino acid sequence of the encoded protein is identical to the C. reinhardtii FEA1 protein.

5.4.3 The reporter gene (uidA) and its encoded protein (GUS)

53. The uidA gene encodes the enzyme β-glucuronidase (GUS), which is derived from the common gut bacterium E. coli. The active GUS enzyme is a homo-tetramer of 68 kDa protein sub-units and catalyses the hydrolysis of β-glucuronides and, less efficiently, some β-galacturonides. Hence, E. coli can use these compounds as its main carbon and energy source. The uidA gene is a widely used reporter gene in GM plants (Miki & McHugh 2004) as it allows GM tissues to be identified using a simple visual assay.

5.4.4 The antibiotic resistance marker gene (nptII) and its encoded protein

54. The nptII gene was initially isolated from the transposon Tn5 that was present in the bacterium E. coli strain K12 and codes neomycin phosphotransferase type II. This is an aminoglycoside phosphotransferase enzyme which catalyses transfer of the γ-phosphate group of ATP to specific hydroxyl group of aminoglycoside antibiotics e.g. kanamycin, neomycin etc. and hence detoxifies them. Plants expressing the enzyme can tolerate certain concentrations of these antibiotics which lead to bleaching and growth inhibition in non-transformed plants. The nptII gene is used extensively as a selectable marker in the production of GM plants (Miki & McHugh 2004).

5.5 Toxicity/allergenicity of the proteins associated with the introduced genes

55. Homologues of the introduced proteins, and proteins with similar sequences and function, occur naturally in a range of organisms, including animals or plants widely consumed by people and animals (see Section 6.5). On this basis, people and other organisms have a long history of exposure to the proteins expressed by the introduced genes.

56. The GUS enzyme is from E. coli which lives in the digestive tract of vertebrates, including humans (Jefferson et al. 1986). GUS activity has been described in all vertebrates, numerous bacteria and many invertebrate species (Gilissen et al 1998). A number of GM crops containing the uidA gene have been approved for limited and controlled release in Australia, including GM bananas (DIR 076/2007, DIR 107), GM maize (DIR 086/2008) and GM sugarcane (DIR 095). In addition, the uidA gene is present in commercially approved Bollgard II® cotton (DIR 066/2006). No adverse effects on humans, animals or the environment have been reported from these releases. The US EPA does not consider the GUS protein to be toxic and has approved its exemption from the requirements to establish tolerance levels (EPA 2001). FSANZ has approved the use of food derived from GM plants containing the uidA gene (for example see FSANZ 2003; FSANZ 2002).

57. All of the GM banana lines contain the nptII gene. As discussed in previous DIR RARMPs, and in more detail in the RARMPs for DIR 070/2006 and DIR 074/2007 (available by contacting the OGTR), regulatory agencies in Australia and in other countries have assessed the use of the nptII gene in GM plants as not posing a risk to human or animal health or to the environment. The most recent detailed international evaluation of nptII in terms of human safety was by the European Food Safety Authority, which concluded that the use of the nptII gene as a selectable marker in GM plants (and derived food or feed) does not pose a risk to human or animal health or to the environment (EFSA 2009).

58. No toxicity/allergenicity tests have been performed on the other introduced proteins or on the GM banana lines as the proposed trial is still at proof of concept stage. Such tests may need to be conducted if approval was sought for the GMOs or their products to be used for human consumption in Australia (see discussion in Section 7.1.2). Recently a small scale human nutritional trial was conducted in which five adult human subjects consumed Golden Rice-II, a GM rice engineered for increased β-carotene (Paine et al. 2005; Tang et al. 2009). Golden Rice-II expresses PsyB73 gene from maize and CrtI gene from E. uredovora (Paine et al. 2005). No adverse effects, including allergic reactions and gastrointestinal disturbances, were noted in this trial (Tang et al. 2009).

59. Bioinformatic analysis may assist in the assessment process by predicting, on a purely theoretical basis, the toxic or allergenic potential of a protein based on similarity to known toxins and allergens. The results of such analyses are not definitive and are used to identify those proteins requiring more rigorous testing (Goodman et al. 2008). The applicant compared the predicted amino acid sequences of the proteins encoded by each of the genes introduced for enhanced nutrition to a database of known allergens, the Food Allergy Research and Resource Program allergen protein database AllergenOnline. According to this website, the most predictive search is the overall FASTA alignment, with identity matches greater than 50% indicating possible cross-reactivity. An additional precautionary search uses a sliding window of 80 amino acids, looking for identities greater than 35%, which is a threshold often used to highlight an allergenicity concern (Fiers et al. 2004). None of the proteins to be expressed in GM banana has any sequence/structure homology to any allergen catalogued at the AllergenOnline database.

60. A comprehensive search of the scientific literature yielded no further information to suggest that the encoded proteins are toxic or allergenic to people, or toxic to other organisms.

5.6 Toxicity of end products associated with the introduced genes for nutrient enhancement

61. Pro-vitamin A carotenoids (such as β–carotene) and iron occur naturally in the environment by virtue of their widespread presence in plant and/or animal material. In addition, carotene forms of vitamin A and iron are permitted food additives in Australia (FSANZ 2007). On this basis, people and other organisms have a long history of exposure to these end products. While soluble iron salts present in soils can be toxic to plants, there is no indication that the forms and levels of iron found naturally in plant or animal material are toxic (Table 4).

62. The mammalian toxicity of β–carotene and iron has been independently assessed by regulatory agencies or health advisory bodies in Australia and overseas for the purposes of determining safe levels of intake via food or supplements. Assessments of these compounds has been conducted by the National Health & Medical Research Council (NH&MRC 2006), the UK’s Expert Group on Vitamins and Minerals (Expert Group on Vitamins and Minerals 2003), the European Commission (EC 2000) and the Joint Food and Agricultural Organisation of the United Nations/World Health Organisation (FAO/WHO) Expert Committee on Food Additives (JECFA 1974; JECFA 1975; JECFA 1983; JECFA 1993). These assessments have examined both animal and human data.

63. Excess intake of preformed vitamin A (retinol) can cause hypervitaminosis A leading to birth defects, liver abnormalities, reduced mineral density in bones and central nervous system disorders. The upper intake levels (UL) for retinol is 600 µg/day for infants up to 3000 µg/day for lactating adults (NH&MRC 2006). However, the conversion of β-carotene to retinol in the body is tightly regulated by vitamin A status and there is not enough data to establish an UL for β-carotene (NH&MRC 2006).

64. Although β-carotenes can function as anti-oxidants, and epidemiological studies associated high β-carotene uptake with reduced risk of many chronic diseases, excess intake of β-carotene (20 mg/day) can cause some harm to smokers or people exposed to asbestos (Rao & Rao 2007; Gallicchio et al. 2008; Goralczyk 2009; Druesne-Pecollo et al. 2010).

65. A comprehensive search of the scientific literature found no information or data to suggest that pro-vitamin A carotenoids (such as β–carotene) have any toxicity potential in other organisms. However, as carotenoids can be converted to toxic cleavage products, high level of carotenoids in diets may have other negative effects particularly under oxidative stress conditions. American gold finches supplemented with high amounts of carotenoid (3000 µg/day) were more colourful, but showed more skeletal muscle deterioration and reduced flight performance during the post-supplementation period than birds receiving low doses of carotenoid supplements (30 µg/day) (Huggins et al. 2010). Thus, in the long term, maintaining high level of carotenoids for bright colouration may pose a challenge to these birds (Vinkler & Albrecht 2010).

66. There is considerable scope of iron toxicity as it can accumulate in the body and its effect can range from gastrointestinal irritation to systemic toxicity (NH&MRC 2006). This is more pronounced for people with haemochromatosis. The prescribed UL for iron is 20 mg/day for healthy infants and children (up to 3 yr), 40 mg/day for children up to 13 yr and 45 mg/day for people up to 50 years of age (NH&MRC 2006).

67. The levels of these compounds generated in the GM banana plants are not expected to exceed the levels found naturally in edible plants (Table 4). An indication of likely increases may be gained from similar studies with other GM plants. For example, Golden Rice II which expresses PsyB73 and CrtI genes in rice endosperm increased the total carotenoid content to a maximum of 37 µg/g (23-fold) of dry seed weight (Paine et al. 2005). Similar attempts in maize resulted in up to 34-fold increase in total carotenoids content (46.5 µg/g seed dry weight ) as compared to parental lines and a preferential accumulation of β-carotene (13.8µ g/g seed dry weight maximum; Aluru et al. 2008). Expressing bacterial CrtB (phytoene synthase), CrtI and CrtY (lycopene β-cyclase) in GM potato lead to a 20-fold increase in total carotenoids (Diretto et al. 2007) and the level of β-carotene was significantly higher than parental lines (47 µg/g dry weight, 3600 fold increase).

68. Expression of soybean Ferritin gene in rice lead to a 2–3 fold increase in iron content in unpolished GM rice seeds (Goto et al. 1999; Vasconcelos et al. 2003). Constitutive expression of the barley NAS gene in rice resulted in a maximum of 4.5-fold increase in iron content in seeds (Masuda et al. 2009). When the Ferritin gene was expressed in conjunction with NAS and Phytase genes, it led to more than six-fold increase in iron concentration in rice seeds (approximately 7 µg/g dry weight; Wirth et al. 2009); but is still over 10-fold less than iron-fortified breakfast cereals (81 µg/g fresh weight; Table 4). Over-expression of these genes also lead to increased accumulation of zinc, manganese and nickel (Vasconcelos et al. 2003; Douchkov et al. 2005; Pianelli et al. 2005; Masuda et al. 2009).

5.7 Other effects associated with genetic modifications for enhanced nutrition

69. Many of these genes involved in β-carotene biosynthesis and described under Section 5.4.1 have been expressed in various GM plants either singly or in combination, and effects on endogenous genes and metabolic pathways have been investigated. Constitutive expression of an additional Psy gene in GM tomato resulted in changed expression of other genes involved in the pathway (Fraser et al. 2007). Over-expression of a DXS gene in potato tubers also resulted in over-expression of the endogenous Psy1 gene (Morris et al. 2006). Re-creation of the β-carotene biosynthetic pathway in potato using E. caratovora genes generally resulted in down-regulation of the endogenous Psy1 gene but induction of some other genes e.g. phytoene desaturase in carotenoid pathway (Diretto et al. 2010). In contrast, no significant changes in expression of endogenous genes involved in carotenoid biosynthesis were observed in GM wheat expressing maize Psy1 and E. caratovora CrtI genes.

70. These changes in gene expression were also reflected in changes in metabolites including composition of total carotenoids (Lindgren et al. 2003; Ducreux et al. 2005). For example, in tomato, constitutive over-expression of Psy1 gene resulted in changes in total carotenoid compositions and other metabolites generally associated with ripening processes (Fraser et al. 2007). In contrast expression of the Psy1 gene from daffodil in rice led to an increase in phytoene but not the downstream carotenoids (Burkhardt et al. 1997). In the case of potato (Diretto et al. 2007; Diretto et al. 2010) expression of bacterial β-carotene biosynthesis genes also resulted in perturbation of leaf carotenoid metabolites. There were changes in expression levels of related pathway genes and high transcript-metabolite correlations reflecting a co-ordinated control of carotenoid biosynthesis pathways involving both positive and negative feedback mechanisms.

71. However, no novel metabolite has been reported in any of these studies and a comprehensive search of scientific literature also did not reveal any such cases.

72. Since isoprenes are also building blocks for a number of plant hormones including absicic acid (ABA), gibberelic acids (GA) and some cytokinins as well as pigments like chlorophyll, manipulation of the carotenoid biosynthetic pathway has also resulted in changes in hormone balances and phenotypes. Constitutive over-expression of the Psy1 gene in tomato resulted in dwarfism, loss of chlorophyll and up to 30-fold reduction in GA1 (Fray et al. 1995). Similar phenotypes were also observed in GM tobacco over-expressing endogenous Psy1 and Psy2 genes (Busch et al. 2002). Over-expression of bacterial DXS genes in potato tubers also led to significantly elongated tubers and a reduction in tuber dormancy (Morris et al. 2006). A significant increase in trans-zeatin riboside, a cytokinin, was observed in these lines while GA1 and ABA levels were not significantly affected. In contrast, seed specific over-expression of Psy1 and Psy2 in Arabidopsis resulted in delayed germination and increase in total carotenoids, chlorophyll and ABA (Lindgren et al. 2003).

73. The applicant has stated that in the case of GM banana lines released under DIR 076/2007 some changes in plant phenotype e.g. leaf yellowing, appearance of necrotic spots and stunting were observed in some of the lines. Also, some GM banana somatic embryos expressing Psy1 and/or CrtI genes developed orange colourations, presumably due to accumulation of high β-carotene, and failed to regenerate (information supplied by applicant).

74. Endosperm specific expression of ferritin in rice also resulted in concomitant increase in zinc (Vasconcelos et al. 2003). Similar increases in metal ions like zinc and manganese were also observed in tobacco expressing a soybean Ferritin gene (Yoshihara et al. 2003).

75. The mugineic acid family of metal chelators including NA can also bind to other metal ions and hence can contribute to their uptake in plants (Palmer & Guerinot 2009). Such increase in zinc content, although less pronounced than iron, was observed when Arabidopsis NAS and soybean Ferritin genes were expressed in rice (Wirth et al. 2009). Similarly, both iron and zinc levels were increased in rice seeds expressing the NAS gene from barley (Masuda et al. 2009). Over-expression of NAS also led to increased heavy metal tolerance (Douchkov et al. 2005; Pianelli et al. 2005) and paradoxically, in one case, increased sensitivity to iron deficiency which was due to decreased bioavailability of iron in NA over-accumulating plants (Cassin et al. 2009).

76. Additionally iron status of plants can influence the soil microbial community (Jin et al. 2010). Growth of tobacco plants over-expressing ferritin in plastids can lead to reduced iron availability to soil microflora, particularly in iron deficient soils, and can significantly alter the microbial community structure under such conditions (Robin et al. 2006a; Robin et al. 2006b). This led to selection of fluorescent pseudomonads that can synthesize siderophores with high affinity to iron and hence are less susceptible to iron stress (Robin et al. 2007).

5.8 The regulatory sequences

77. Promoters are DNA sequences that are required in order to allow RNA polymerase to bind and initiate correct gene transcription. Also required for gene expression in plants is a transcription termination region, including a polyadenylation signal. Other sequences, such as introns and protein targeting sequences, may contribute to the expression pattern of a given gene. The regulatory sequences used in the GM banana lines are listed in Table 2 and are detailed below.

78. Although some of regulatory sequences are derived from plant pathogens (CaMV and A. tumefaciens), the regulatory sequences comprise only a small part of the total genome, and are not in themselves capable of causing disease. Similarly, those regulatory sequences derived from plants that are associated with allergenic or toxic responses in humans (maize and castor bean) are not in themselves toxic or allergenic.

5.8.1 Regulatory sequences for expression of the introduced genes for enhanced nutrition

79. Expression of all the introduced genes are either driven by a constitutive promoter (Ubi) or promoters expected to provide high expression specifically in banana fruit tissues (Exp1, Exp4, Ext, MT2a, ACO and ACS).

80. The Ubi promoter is from the maize poly-ubiquitin gene. This is a constitutive promoter and is expected to direct the genes to be expressed in most plant tissues and throughout the plant life cycle (Christensen et al. 1992). This promoter has been extensively used to drive gene expression in a large number of GM monocotyledonous plants (Christensen & Quail 1996).

81. Exp1 and Exp4 promoters are derived from banana expansin genes (Exp1 and Exp4; Trivedi & Nath 2004; Asha et al. 2007). Expansins are proteins that play a role in cell wall loosening and extension and their expression is differentially regulated during stages of plant growth and development including fruit development and ripening (Choi et al. 2006). The Exp1 gene from banana is expressed only in fruit at later stages of ripening and the transcript accumulates steadily over the climacteric ripening phase (Trivedi & Nath 2004). In contrast Exp4 is expressed steadily in fruit throughout development and ripening (Asha et al. 2007). The promoter of the banana Exp1 gene was used in GM banana plants release under DIR 076/2007 but no data on expression pattern or level is yet available.

82. The Ext promoter is derived from a banana extensin gene. The Ext genes code for extensin proteins which are hydroxyproline-rich glycoproteins thought to be involved in cell wall extension (Wilson & Fry 1986). In banana, one Ext gene was observed to be down-regulated in pulp of ripening fruit (Medina-Suarez et al. 1997), while another study reported a peel specific up-regulation of another Ext gene during ripening (Drury et al. 1999). The Ext promoter element is from the Ext gene from banana which is up-regulated during ripening (Drury et al. 1999).

83. The MT2a promoter is derived from the corresponding metallothionein (MT) gene in banana fruit. Metallothioneins are ubiquitous, cysteine-rich heavy metal binding proteins that play an important role in heavy metal detoxification and homeostasis. They are grouped into different classes and types based on number and arrangement of cysteine containing motifs and are encoded by a family of genes with different temporal as well as spatial expression patterns and which also respond to different environmental cues (Cobbett & Goldsbrough 2002; Hassinen et al. 2011). For instance, in banana, MT2a gene expression is higher in green fruits and then declines as the fruit ripens whereas MT2b and MT3 genes are expressed later in ripening (Liu et al. 2002). Ethylene has an inhibitory effect on MT2a, but not on MT2b or MT3 and only MT3 expression is enhanced in response to heavy metals like copper, zinc or cadmium (Liu et al. 2002). Unpublished data from transient expression experiments suggest that the MT2a promoter can drive transgene expression in fruit (information supplied by applicant).

84. ACO and ACS promoters are derived from the banana ACC oxidase (ACO) and ACC synthase genes (ACS). The enzymes encoded by these genes are involved in biosynthesis of the fruit ripening hormone ethylene (Do et al. 2005; Huang et al. 2006). There are at least nine members of the ACS gene family in banana of which MaACS1 is strongly over-expressed in ripening banana fruits (Liu et al. 1999; Huang et al. 2006). Two members of the ACO gene family have been described so far and both show high expression in banana fruits during ripening (Liu et al. 1999; Do et al. 2005). Only fruit specific expression was observed when promoter regions of ACS1 and ACO1 genes were used to drive transient GUS expression in various tissues of banana (Wang & Peng 2001a; Wang & Peng 2001b). These promoters have been used to regulate gene expression in GM banana lines released under DIR 076/2007 but data on their expression pattern is not yet available.

85. All the introduced genes have the terminator region from A. tumefaciens nos gene (Depicker et al. 1982). The nos terminator has been used in a wide variety of constructs used for plant genetic modifications (Reiting et al. 2007).

5.8.2 Other genetic elements for expression of the introduced genes for enhanced nutrition

86. Two different plastid targeting signal sequences will be used to direct the phytoene desaturase enzyme from E. uredovora to chloroplasts where it functions. These signal sequences are obtained from Rubisco small subunit gene (RBCS1) either from chrysanthemum or pea (Coruzzi et al. 1984; Wong et al. 1992). These gene sequences will be translated but removed during post-translational modification, and hence will be absent from the mature functional protein.

5.8.3 Regulatory sequences for expression of the uidA reporter gene

87. The uidA reporter gene in GM bananas is driven by the Exp1 promoter and has a downstream terminator from the A. tumefaciens nos gene. The uidA gene also contains an intron from the castor bean catalase gene (Cat; Ohta et al. 1990). The Cat intron prevents expression in A. tumefaciens, ensuring that any expression of the reporter gene in the GMOs is occurring in eukaryotic cells rather than in Agrobacterium. The Cat intron can also enhance expression of introduced genes in plants (Tanaka et al. 1990).

5.8.4 Regulatory sequences for expression of the nptII marker genes

88. All of the GM bananas contain the nptII gene under the control of the promoter and mRNA termination region of the CaMV 35S gene (Odell et al. 1985) or A. tumefaciens nos gene (Depicker et al. 1982). These promoters are widely used to drive constitutive expression of genes in GMOs. The 35S and nos terminators have also been used widely in GM plants (Mitsuhara et al. 1996; Reiting et al. 2007).

5.9 Method of genetic modification

89. Agrobacterium tumefaciens-mediated transformation is being used to generate the GM banana lines in the proposed release. A. tumefaciens is a soil bacterium that causes gall formation on a wide range of plant species. The gall is induced by transfer of hormone-producing genes from the bacterial cell into the plant genome. The genes are carried on an extrachromosomal, circular DNA molecule found within the bacterial cell called a Tumour-inducing (Ti) plasmid. During the infection process, only a section of the Ti plasmid known as the Transfer DNA (T-DNA) is transferred to the plant. Molecular biologists have studied the infection and T-DNA transfer process of A. tumefaciens for many years and have used this natural process to facilitate genetic modification of plants. Well-characterised A. tumefaciens Ti plasmids have been produced that lack the genes responsible for tumour formation (disarmed plasmids) and instead enable genes of interest to be inserted between the T-DNA border sequences. When used to infect plants, A. tumefaciens cells carrying such plasmids cannot produce a tumour but will transfer the T-DNA sequence carrying the genes of interest into the plant cell where they stably integrate into the plant genome (Bevan 1984; Klee & Rogers 1989).

90. In addition to transfer of the T-DNA sequence, recent publications have shown that small segments of flanking Ti plasmid sequence and A. tumefaciens chromosomal sequence may be transferred into the plant genome at a low frequency during the transformation process (Smith 1998; Ulker et al. 2008). However, A. tumefaciens-mediated plant transformation has been used extensively in Australia and overseas and is not known to adversely affect human health and safety or the environment.

91. To genetically modify the banana lines in this application, A. tumefaciens strain AGL1 (ATCC® Number: BAA-101) is being used (Lazo et al. 1991), which shows high rates of T-DNA transfer when used with banana suspension cultures (Khanna et al. 2004).

92. For each transformation, embryonic cell suspensions of banana are co-cultivated with A. tumefaciens AGL1 carrying one of the gene constructs listed in Table 3, using a centrifugation assisted protocol (Khanna et al. 2004). Following transformation, the banana cells are cultured on medium containing both the antibiotics timentin (to remove A. tumefaciens) and kanamycin (to select for banana cells expressing the introduced nptII gene). Transformed embryos are developed on a regeneration medium and then grown into plantlets, which are hardened off in a shade house before being transferred to soil.

5.10 Characterisation of the GMOs

5.10.1 Stability and molecular characterisation

93. Not all of the GM banana lines proposed for release have been generated prior to submission of this application. All gene constructs used for transformation have been fully sequenced (information supplied by applicant).

94. As the project is at an early stage, full molecular characterisation of the GM banana lines has not been carried out. Of the lines proposed for release, lines previously released under DIR 076/2007 were screened for the presence of the introduced genes using PCR. These plants were also tested for presence of Agrobacterium by PCR using primers specific to non-T-DNA regions. As the remaining GM banana lines are being generated, they will also be screened for the introduced genes as well as Agrobacterium by PCR and no plant testing positive for Agrobacterium will be released.

95. Southern hybridisation for further confirmation and copy number determination have been performed with only some of the lines previously released under DIR 076/2007 and intended for further evaluation under the proposed release. From one up to 11 copies of the introduced genes have been detected in the lines tested.

96. The exact location of the inserted genes within the banana genome is not known. A. tumefaciens inserts genetic material into plant genomic DNA via illegitimate recombination, which can potentially result in insertion of the introduced genes anywhere in the host genome. Furthermore, the banana genome is poorly characterised, so that it would be difficult to generate meaningful data on the locations of introduced genes.

5.10.2 Characterisation of the phenotype of the GM banana lines

97. The purpose of the trial is to assess pro-vitamin A and/or iron levels in the GM banana fruit and agronomic performance of GM banana plants grown under field conditions. This would help identifying gene(s) and promoter(s) combinations that will enhance nutritional quality of banana fruits without the associated negative effects on GM plant growth and development observed in a previous trial. The applicant states that it is not possible to grow numerous large banana plants until fruiting in a glasshouse to assess pro-vitamin A and iron contents in fruit or to study the phenotypes of such large plants. Such nutrient content and phenotypic data will be collected during the proposed trial.

98. Most of the GM banana lines proposed for release are being developed and have not been phenotypically characterised. Of the GM banana plants released under DIR 076/2007, some over-expressing APsy2a or PsyB73 genes showed higher pro-vitamin A levels. However, some of these lines also showed leaf yellowing, stunting and development of necrotic spots. The lines selected for further evaluation under the proposed release have higher pro-vitamin A content but no adverse effect on growth or development. Nutrient content and phenotypic data for GM banana lines containing the CrtI and Ferritin genes are not yet available. However, almost all of the GM banana lines proposed to be released are expected to over-express the genes for pro-vitamin A biosynthesis and/or iron homeostasis in a fruit specific manner. This is intended to mitigate some of the negative effects e.g. chlorosis, stunting observed in GM banana plants trialled under DIR 076/2007. The applicant states that the GM banana plants will be monitored for aberrant phenotypes during the proposed trial.


Top of Page