Plant science in lac: A continuation of using tools from Escherichia coli in studying gene function in heterologous systems

The assumption has been made for many years that genetic engineering will have its greatest benefit by producing valuable human and animal proteins and correcting genetic defects that cause certain diseases among a subset of the population. However, it now is becoming clear that the greatest potential will be in a second green revolution where food, fiber, proteins, and secondary products will be produced by crop plants in sufficient quantities with fewer chemicals and greater efficiencies. The reason for this revolution is that changes in the genetic information of plants are not limited to the adolescent period of a single life span as it is for humans. Actually, the tremendous amplification of beneficial traits through seeds of new crops is likely to provide the highest payoff for gene manipulation techniques. As plant sciences are entering the genomic DNA sequencing stage, it is expected that for a few examples like Arabidopsis thaliana (genome size = 125 megabases) and Oryza sativa (genome size = 400 megabases) all genes in their order and spacing within plant chromosomes will be known by the beginning of the millenium. However, knowledge-based design of new crop varieties will depend on the functional analysis of those DNA sequences during plant development.

Traditional breeding efforts have always relied on natural variability and classical plant genetics. Much has been learned about complex traits of important crops and plant morphogenesis in general from these efforts, but a deeper understanding of sequence function will require a more targeted analysis of selected DNA sequences. Critical to this analysis is DNA transformation of plant genomes, which now allows one to regulate hand-picked genes, even heterologous genes, and to study their role and their products during plant morphogenesis. In this issue, Moore et al. (1) developed a gene expression system that can be used to generate transgenic plants where a transgene is not expressed in the primary transformant. By crossing such a transgenic plant with a second one that produces a specific transcriptional activator, hybrid progeny then can be used to investigate the properties of the transgene from the primary transformant. Simple Mendelian segregation of the two transgenes, one carrying the target gene and the other one the activator gene, then can be used to reverse gene activation.

Using heterozygote plants to study control of gene action already has been the strategy in Barbara McClintock’s work (2) on transposable elements, where one parent plant would carry the autonomous transposable element and the other the nonautonomous transposable element. The nonautonomous element would be active only in the generation where the autonomous one would be introduced through crossing. The gene activation by crossing-scheme of Moore et al. (1) has certain appeal compared with previous designs of inducible promoters. Both the dexamethasone and tetracyline inducible promoter systems rely on a tight repressible promoter and an efficient way to control the level of the chemical inducer in the target cells (3, 4). Both aspects are not trivial, although the tetracycline-dependent gene expression system appears to have a very low leakiness under experimental conditions in tobacco. However, it is nearly impossible to target tetracycline to specific tissues, or the entire plant, or for a prolonged period of time. Chemical inducers have their primary applications in unicellular organisms or cell cultures, but cell differentiation requires a regulated system that takes advantage of developmentally controlled promoters. Although Moore et al.’s activator plasmids are driven by the standard ³⁵S promoter of cauliflower mosaic virus, it is conceivable to replace the viral promoter by tissue-specific promoters (e.g., endosperm or pollen-specific promoters) to limit expression of the reporter gene to certain stages of plant development.

Instrumental to the development of such a two parent-system is the use of protein domains and DNA sequences absent in plant genomes for the construction of chimeric promoters and transcription factors. Keegan et al. (5) already have described the modular nature of transcription factors, and Labow et al. (6) have shown how to convert a bacterial DNA-binding protein into a transcriptional activator for mammalian cells. Such a strategy in turn can be used to combine specific DNA binding sites with minimal plant promoter sequences, which are unlikely to be present in the plant genome. Moore et al. (1) have chosen the operator sequence of the Escherichia coli lactose operon as the specific DNA binding site. In the bacterial cell, this sequence is recognized by the product of the lacI gene, the lac repressor. When the lac repressor binds to the lac operator, RNA polymerase is sterically prevented from initiating transcription of the lacZ gene. Therefore, its function is that of a repressor. In Moore et al. (1), however, the lac repressor is converted into the opposite function, an activator. This conversion is achieved in two steps. First, the lac operator sequence is placed in front of a plant minimal promoter containing only the TATA box, but no sites for plant transcriptional activators. Second, the lac repressor is fused with a protein domain that is capable of activating RNA polymerase II-based transcription at the TATA box. This chimeric transcription factor is expressed from a second plasmid.

Interestingly, the demonstration of a bifunctional or chimeric protein precedes the era of recombinant DNA techniques and actually was done first with the lac repressor. Müller-Hill and Kania (7) selected by conventional genetic techniques the deletion of the lac operator region that lies between the lacI and the lacZ gene, thereby creating a gene fusion between these two genes. As a result the lacI promoter drove the expression of a single polypeptide chain of lac repressor and β-galactosidase that either could bind to the lac operator or cleave lactose into galactose and glucose. Assigning multiple functions to a single polypeptide chain has been instrumental in our understanding of protein structure and folding.

Obviously, the strength of a promoter design containing a unique DNA binding site depends on the affinity of the DNA binding protein to its target DNA sequence. By taking advantage of a mutation in the lac repressor protein that increases DNA binding and by including a tandem array of the operator sequence, the transgenic hybrid plants in Moore et al. (1) raise the association constant of the chimeric transcription factor with the target promoter.

The use of the bacterial lac operon sequences to regulate plant gene expression by activation rather than by repression adds to the impressive list of applications that have resulted from the knowledge gained over many years by basic studies of bacterial genetics, a fact that is readily forgotten nowadays in the steady flood of patents and technology transfer agreements (Fig. 1). For instance, the replication origin of pUC plasmids present today in every vector (and used in many biotechnology patents), where DNA yield is critical, was selected on its ability to titrate overproduction of lac repressor in bacterial cells with the lac operator sequence on the pUC plasmid because of the increased copy number of the plasmid (8). It simply reiterates the serendipity of research discovery and applications, which, however, is not possible without some openness of what research to support. Long-term research developments are critical to publicly funded research policies. Furthermore, here is an important example of why there is not basic and applied research; basic research is applied research over time.

Figure 1.

Multiple applications of the DNA binding protein lac repressor. The specific interactions between lac repressor and lac operator have been applied as various molecular tools. By protecting the operator sequence from sequence-specific DNA methylases, restriction enzymes can be used to cut genomic DNA at unique sites (9). Because a DNA/protein complex can bind to a nitrocellulose filter while DNA passes through, lac repressor has been used to purified restriction fragments containing lac operator DNA (10). The inducibility of the lac operon by isopropyl β-d-galactoside, which cannot be cleaved by β-galactosidase, has been used in the construction of single-stranded and double-stranded cloning vectors (8). lac repressor has been used as a transcriptional repressor (11) and activator (6) in eukaryotic gene expression systems.