The selection and analysis of transformants

Using either Agrobacterium or direct gene transfer systems, it is now possible to introduce DNA into virtually any regenerable plant cell type. However, only a minor fraction of the treated cells become transgenic while the majority of the cells remain untransformed. It is therefore essential to detect or select transformed cells among a large excess of untransformed cells, and to establish regeneration conditions allowing recovery of intact plants derived from single transformed cells.

Selectable genes
Selectable marker genes are essential for the introduction of agronomically important genes into important crop plants. The agronomic gene(s) of interest are invariably cointroduced with selectable marker genes and only cells that contain and express the selectable marker gene will survive the selective pressure imposed in the laboratory. Plants regenerated from the surviving cells will contain the selectable marker joined to the agronomic gene of interest.

The selection of transgenic plant cells has traditionally been accomplished by the introduction of an antibiotic or herbicide-resistant gene, enabling the transgenic cells to be selected on media containing the corresponding toxic compound. The antibiotics and herbicides selective agents are used only in the laboratory in the initial stages of the genetic modification process to select individual cells containing genes coding for agronomic traits of interest. The selective agents are not applied after the regeneration of whole plants from those cells nor during the subsequent growth of the crop in the field. Therefore, these plants and all subsequent plants and plant products will neither have been exposed to, nor contain the selective agent.

By far, the most widely used selectable gene is the neomycin phosphotransferase II (NPTII) gene [42] which confers resistance to the aminoglycoside antibiotics kanamycin, neomycin, paromomycin and G-418. [43–4] A number of other selective systems has been developed based on resistance to bleomycin, [45] bromoxynil, [46] chloramphenicol, [47] 2, 4-dichlorophenoxy-acetic acid, [48] glyphosate, [49] hygromycin, [50] or phosphinothricin. [51]

The increasing knowledge of modes of action of herbicides, and rapid progress in molecular genetics have led to the identification, isolation and modification of numerous genes encoding the target proteins for herbicides. Engineering herbicide tolerance into crops has proved useful not only as a selection system, but also as a valuable trait for commercial agriculture. To be useful in agriculture, herbicides must distinguish between crop plant and weed. Although they are designed to affect significant processes in plants such as photosynthesis and amino-acid biosynthesis, these processes are common to both crops and weeds. Consequently, at present, selectivity is based on differential herbicide uptake between weed and crop, or controlled timing and site of application of the herbicide by the crop plant. As to the different strategies employed to introduce herbicide tolerance in crops, the overexpression or modification of the biochemical target of the herbicide [52–4] and detoxificationdegradation of the herbicide before it reaches the biochemical target [55–6] are the general routes by which this trait is engineered in plants.

Reporter genes
Reporter genes are ‘scoreable’ markers which are useful for screening and labeling of transformed cells as well as for the investigation of transcriptional regulation of gene expression. Furthermore, reporter genes provide valuable tools to identify genetic modifications. They do not facilitate survival of transformed cells under particular laboratory conditions but rather, they identify or tag transformed cells. They are particularly important where the genetically modified plants cannot be regenerated from single cells and direct selection is not feasible or effective. They can also be important in quantifying both transformation efficiency and gene expression in transformants. The reporter gene should show low background activity in plants, should not have any detrimental effects on plant metabolism and should come with an assay system that is quantitative, sensitive, versatile, simple to carry out and inexpensive.

The gene encoding for the enzyme β-glucuronidase, GUS, has been developed as a reporter system for the transformation of plants.57–8 The β-glucuronidase enzyme is a hydrolase that catalyzes the cleavage of a wide variety of β-glucuronides, many of which are available commercially as spectrophotometric, fluorometric and histochemical substrates. There are several useful features of GUS which make it a superior reporter gene for plant studies. Firstly, many plants assayed to date lack detectable GUS activity, providing a null background in which to assay chimaeric gene expression. Secondly, glucuronidase is easily, sensitively and cheaply assayed both in vitro and in situ in gels and is robust enough to withstand fixation, enabling histochemical localization in cells and tissue sections. Thirdly, the enzyme tolerates large amino-terminal additions, enabling the construction of translational fusions.

The gene encoding firefly luciferase has proven to be highly effective as a reporter because the assay of enzyme activity is extremely sensitive, rapid, easy to perform and relatively inexpensive. [59] Light production by luciferase has the highest quantum efficiency known of any chemiluminescent reaction. Additionally, luciferase is a monomeric protein that does not require posttranslational processing for enzymatic activity. [60]

The use of green fluorescent protein (GFP) from the jellyfish Aequorea victoria to label plant cells has become an important reporter molecule for monitoring gene expression in vivo, in situ and in real time. GFP emits green light when excited with UV light. Unlike other reporters, GFP does not require any other proteins, substrates or cofactors. GFP is stable, species-independent and can be monitored noninvasively in living cells. It allows direct imaging of the fluorescent gene product in living cells without the need for prolonged and lethal histochemical staining procedures. In addition, GFP expression can be scored easily using a long-wave UV lamp if high levels of fluorescence intensity can be maintained in transformed plants. Another advantage of GFP is that it is relatively small (26 kDa) and can tolerate both N- and C-terminal protein fusions, lending itself to studies of protein localization and intracellular protein trafficking. [61] It has been reported that high levels of GFP expresion could be toxic to plant growth and development. [62] Solution to this problem comes from the utilization of GFP mutant genes. Among the various GFP mutations, the S65T (replacement of the serine in position 65 with a threonine) is one of the brightest chromophores characterized by its faster formation and greater resistance to photobleaching than wild-type GFP photobleaching. Furthermore, this mutant is characterized by having a single excitation peak ideal for fluorescin isothiocyanate filter sets [63] and also by its harmless action to the plant cell. [64]