10.3.3 Gene regulation during development

Printer-friendly version

(a)  Gene promoters: key control modules


Figure 10.29 Construction of a nested deletion series for a promoter analysis. Important regulatory DNA sequences within gene promoters are often found using techniques such as nested deletion analysis. A plant promoter sequence is fused to a reporter gene. The promoter DNA is the progressively digested from it's 5' end, using nuclease enzymes (exonuclease III and SI nuclease) and varying the incubation period. The 'nested' deletion series of plasmids is transformed into plants, which are tested for reporter gene activity. A difference in activity between different promoter fragment lengths indicates the presence of a regulatory element in the section of DNA which was deleted.

Having covered the basic mechanisms of transcription and translation of plant genes, we already know that sequences upstream (5') of the transcription start site are referred to as the promoter region. DNA-binding (so-called ‘trans acting’ or transcription factor) proteins interact with DNA sequences (so-called ‘cis regulatory elements’) within the promoter to allow differential expression of specific genes either in different cell types or at different times of development. Promoters often consist of sections or ‘modules’ each of which can be tested to determine its effect on spatial (cell/tissue type) or temporal (stage of development) expression. This is often achieved simply by coupling a reporter gene, such as those for glucuronidase (GUS), luciferase or green fluorescent protein (GFP) (see Section 10.4.1(c)), to the promoter, allowing histochemical or fluorescent detection of the location of expression.

Promoter analysis showing the presence of modules

One early systematic analysis of the nature of promoters which function in plants involved the dissection of the 35S promoter from cauliflower mosaic virus (CaMV35S). In most organs of transgenic plants, the region from –343 to + 8 nucleotides (0 corresponds to the transcription start point) of the CaMV35S promoter acts as a strong regulatory sequence ensuring high levels of transcription of genes which are fused to it. As transcription occurs in most organs and cell types of transgenic plants, the CaMV35S promoter is often described as being ‘constitutive’. Dissection of this promoter has revealed two major sections, referred to as ‘domains’, that result in different expression patterns of reporter genes when the latter are placed under transcriptional control of these domains (Benfey et al. 1989). Domain A (a region from –90 to + 8 of the CaMV35S promoter) gives expression predominantly in roots, whereas Domain B (–343 to -90) results in strong expression mainly in shoots. Within Domain A, a specific regulatory DNA element exists which ensures expression in roots. This element is called activation sequence 1 (as-1) and contains a tandem repeat of the sequence TGACG, separated by seven intervening nucleotides. Further deletion of Domain A from –90 to –72 nucleotides disrupts the integrity of the as-1 regulatory element and abolishes expression completely in the roots. A trans acting DNA-binding protein called ASF-1 (‘activation sequence factor’) has been shown to bind to as-1 in vitro (Lam et al. 1989). ASF-1 is abundant in roots but scarce in shoots, which neatly explains the root-specific expression due to Domain A. Within Domain B, there are five subdomains or modules each of which regulates a slightly different spatial and temporal expression pattern. The various modules within the entire CaMV35S promoter function synergistically to drive high levels of gene expression (Benfey and Chua 1990).


Figure 10.30 Mobility shift or gel retardation assays are used to test for interactions between DNA fragments and putative DNA-binding proteins. Speed of migration through the gel is inversely proportional to the molecular weight of the fragment. Track 1 contains only a radiolabelled DNA fragment suspected of having a sequence enabling specific interaction with transcription factor proteins. The DNA migrates freely through the gel matrix close to the gel front. Track 2. contains the same DNA fragment, but which is now mixed with nuclear protein extract prior to loading onto the gel. Some of the DNA molecules have bound to specific protein molecules, giving them an increased molecular mass, and so causing slower movement ('retardation') through the gel. Track 3. contains the same ingredients as in Track 2 but now mixed with an excess of of unlabelled ('cold') DNA of the same sequence. The cold DNA competes with the radiolabelled DNA for binding to the protein, and because the cold DNA is in excess, the radiolabelled DNA is displaced from the protein and so migrates again at the gel front. Track 4. contains the same ingredients as Track 3 except that the cold competing DNA contains a mutation in the region to which the binding protein attaches. This prevents it from interacting effectively with the protein, so the radiolabel is able to bind normally, and show the same retardation as in Track 2. By changing the position of the mutation, the exact binding protein site can be deduced.

Analysis of numerous plant promoters has employed a similar approach as that used for the CaMV35S promoter. Methods include construction of nested deletions of the promoter (‘deletion’ or ‘loss-of-function’ analysis), ‘gain-of-function’ analysis and ‘mobility shift/gel retardation’ analysis. In deletion analysis, a plasmid containing the promoter of interest is inserted upstream from a reporter gene, and nuclease enzymes are used to digest the promoter from its 5' end. Varying the digestion time results in differing lengths of fragments of the promoter, termed ‘nested deletions’ (Figure 10.29). In Section 10.3.4, a further example of deletion analysis shows how cold- and drought-inducible elements were detected in an Arabidopsis gene called RD29A.

These artificial DNA ‘constructs’ are then used to generate transgenic plants (see Section 10.4). To locate regions of specific function within the promoter, the level of expression from the reporter gene fused to each progressive deletion is assayed in separate transgenic plants. If, as a result of deletion of a DNA segment, the expression level is significantly altered in particular tissues, or in response to environmental or chemical cues, the missing region may contain regulatory sequences. This is conceptually similar to the analysis of the CaMV35S promoter noted above where the as-1 element was located by progressively deleting the DNA sequence. This loss-of-function approach has enabled identification of promoter elements in many plant genes that respond to hormones such as abscisic acid, gibberellins and auxins (see Table 9.3), and environmental stresses such as wounding or drought. Con-firmation that a region suspected of containing a DNA element necessary for gene expression usually requires a gain-of-function analysis. This involves fusion of the section of DNA suspected of harbouring the regulatory element with a minimal promoter fused to a reporter gene. A minimal promoter such as the –72 to +8 section of the CaMV35S promoter cannot in itself promote gene expression. However, by fusing DNA-binding elements to a minimal promoter, expression of the reporter gene can then proceed provided the corresponding transcription factor proteins are also present in the transformed cell, leading to gain of function.

Gel retardation (also known as ‘mobility shift’) assays can then confirm whether putative DNA-binding elements do in fact bind to proteins from nuclear extracts of the plant tissues of interest. This method is best suited to DNA fragments of 300 bp or less, such as regulatory elements from promoters. A DNA fragment suspected of containing such a regulatory element is radioactively labelled and incubated with nuclear protein extracts from the plant tissue of interest, followed by electrophoresis of the mixture. The position of DNA fragments in the gel is determined by exposure to X-ray film. Unbound DNA travels rapidly through the gel matrix and migrates close to the gel front. Specific binding of the DNA fragment to nuclear transcription factor proteins is indicated by a shift or retardation in the mobility of the DNA fragment within the gel matrix. The process is depicted in Figure 10.30. Further coverage is given in Brown (1994) and Alberts et al. (1994).

The DNA analysis methods noted above can identify and confirm whether suspected DNA fragments from plant promoters do indeed bind to specific nuclear proteins present in plant tissues at various stages in development or in response to environmental signals.

(b)  Patterns of gene expression

Many transcription factor proteins bind to promoters to regulate expression of certain genes at specific stages of development. Some of these genes may then have a role in controlling differentiation and development of specific organs. Several classes of regulatory genes encoding transcription factor proteins have been identified (Alberts et al. 1994). We can see the spatial control and sequential nature of gene expression during development by examining mRNA types and abundance in different cell types and tissues during differentiation. Here, we select two elegant examples which use Northern blot analysis (RNA gels probed with radio-actively labelled fragments of DNA or RNA, coding for specific genes of interest) to see whether the tissue was expressing these genes, and also in situ hybridisation which histochemically stains sections of plant tissue for the same mRNA types.


Figure 10.31 Location of gene expression in tomato shoot apical meristems reveals six different patterns. This study made use of in situ staining for presence of specific types of mRNA, shown as a fluorescence image in the central panel of each row of pictures. On the left is a normal light microscope image of each tissue section, and on the right a diagram interpreting the pattern of gene expression. From top to bottom, the genes tested and the patterns of expression were: (A, B, C) rpl2, a ribosomal protein (all meristem cells); (D, E, F) Ltp1, a lipid transfer protein (tunica cells, i.e. pre-epidermis); (G, H, I) arginine decarboxylase (corpus cells, i.e. non-epidermal); (J, K, L) rpl38, a different ribosomal protein (mainly meristem flanks); (M, N, O) H2A, a histone protein (discrete cell cluster throughout meristem); (P, Q, R) rbcS, Rubisco small subunit (leaf primordia but not meristem dome).

(Based on Fleming et al. 1993).


Figure 10.32 Spatial and temporal regulation of mRNAs during tobacco anther development. Here, analysis of extracted mRNA, and in situ hybridisation to specific mRNAs present in tissue sections were used to deduce when and where five different genes were expressed. The shading of each bar corresponds to the anther region in which each mRNA was found. Three shading levels were used for TA56 mRNA because it appears in several different anther tissues at different developmental stages. Key to anther tissues: C, connective tissue; S, stomium; T, tapetum; V, vascular bundle; W, anther wall. Stage 1 corresponds to the time of completion of meiosis, and Stage 12 is the time of anther dehiscence following opening of the flower.

(Based on Koltunow et al. 1990).

The first example is the work of Fleming et al. (1993) on patterns of gene expression within the shoot apical meristem of tomato (Figure 10.31). In Chapter 7 we discussed meri-stem organisation into layers and into zones of different rates of cell division. Gene expression patterns reveal many very reassuring similarities: some genes are expressed only in the surface layer (L1) and indeed relate to genes which we expect to find expressed in epidermal cells — lipid transfer proteins required for cuticle synthesis, and peroxidases probably associated with defence against wounding and pathogens. Other genes appear to be expressed only in the underlying layers, others again only in the faster dividing meristem flanks. Some genes are expressed throughout the apex, mostly with functions in core metabolism or gene expression. A further class is expressed only outside the main meristem dome, that is, in the leaf primordia, and indeed these genes, for example the Rubisco small subunit, are required only in photosynthetic tissues. The sixth pattern is the most surprising, best described as patchy expression in the apex, and suggests that small groups of cells, which perhaps could be termed ‘oligocells’, express some genes in a coordinated manner. Work by Lucas et al. (1995) indicates that macromolecules, including some kinds of mRNA or regulatory proteins, can pass between cells via plasmodesmata (see also Section 10.1.2).

Koltunow et al. (1990) took a similar approach to developing tobacco flowers and showed very precise patterns of mRNA in each of the anther tissues and changes in mRNA abundance with time (Figure 10.32). Northern analysis indicates that many of these genes are expressed only in anther cells and not elsewhere in the plant.