Dale_Molecular Genetics of Bacteria 4th ed

antibiotic) on the other hand is usually produced during stationary phase and does not seem to be central to the growth of the cell.

8.2.1 Simple pathways

A primary metabolite can be considered as the end-product of a series of reactions (Figure 8.1) which converts a precursor substrate (S) into the final product (P). The first of the major factors that are likely to limit the production of P is the supply of the initial substrate (S); increased intracellular levels of S may be expected to increase the throughput of the pathway. However, since S is also used for other metabolic pathways within the cell, it may not be easy to influence its availability. One example (glutamate production) is considered later in this chapter.

Secondly, we come to the rates of the individual enzymic reactions. In each case, this will be influenced by the number of enzyme molecules, the catalytic activity of the enzyme and its affinity for the substrate. Theoretically, it is possible to alter the enzyme structure to increase its maximum activity and/or its substrate affinity, but in practice such mutations will be extremely rare. A more likely target is an increase in the rate of production of the enzyme, most commonly by an alteration in the promoter site so as to increase the transcription of the gene.

Feedback regulation

The third (and often the most important) factor is the regulation of the pathway. The pathway may be capable of high throughput, but as the level of product increases, feedback effects reduce the rate of product formation. This may be due to repression where the production of the enzyme is diminished or to inhibition where the activity of the enzyme is reduced. Either or both of these effects will mean that attempts to increase the production of P will simply lead to shutting down the pathway.

Feedback inhibition/repression

Figure 8.1 Regulation of primary metabolite production: a simple unbranched pathway

Therefore, if colonies are screened for overproduction of the final product, it is most likely that mutants which are deficient in the feedback regulation of the pathway will be obtained. If the production of tryptophan is taken as an example, the whole operon of five genes (in E. coli) is repressed by the presence of tryptophan (Chapter 3). Mutations that abolish these controls will prevent the bacterium from responding to the presence of tryptophan, leading to increased synthesis of the product.

Feedback inhibition on the other hand is (usually) caused by the end-product binding to a site on the enzyme which alters the shape of the protein so that it can no longer carry out its enzymic function. While it is possible to obtain mutants that are resistant to feedback inhibition, the required alterations to the enzyme are much more subtle than those that inactivate a repressor protein and are therefore more difficult to isolate by random screening.

Antimetabolites

Mutants resistant to feedback regulation can sometimes be isolated with the aid of antimetabolites. These are analogues of the end-product of the pathway that are lethal because they substitute for the genuine product in the feedback effect but are unable to substitute for it in its metabolic role within the cell. For example, the tryptophan analogue 5-methyl tryptophan represses the tryptophan pathway (so no tryptophan is made) but it does not substitute for tryptophan in protein synthesis. Mutants that are deficient in feedback regulation are able to grow in the presence of 5-methyl tryptophan, since the trp operon will be expressed and tryptophan will thus be produced for protein synthesis. Regulatory mutants can therefore be selected by plating the mutagenized culture on a medium containing the antimetabolite.

8. 2. 2 Branched pathways

Many amino acids are products not of a simple pathway but of a branched pathway. In Figure 8.2 the production of R is diverting resources from the product P. A mutant defective in enzyme 4, and thus unable to produce R, will be expected to make higher levels of product. An additional advantage of the absence of R is that a branched pathway often exhibits concerted (or multivalent) repression. In this example, enzyme 1 is only repressed if both P and R are present in sufficient quantity. For example, the attenuation of the ilv operon requires the simultaneous presence of leucine and valine (which are the end-points of a branched pathway), as well as isoleucine which requires the same enzymes as valine. Another example is the production of lysine, which is considered in detail below.

R

Concerted feedback inhibition/repression by P + R

Figure 8.2 Regulation of primary metabolite production: a branched pathway

In the general outline in Figure 8.2, if R, the unwanted by-product, is essential for growth of the cell, then a mutant that does not make enzyme 4 will be auxotrophic for R and can be isolated by replica plating (Chapter 2). Growth of these mutants of course requires the addition of R to the growth medium. It might be expected that this would result in feedback inhibition/repression of enzyme 1. However, it is usually possible to add such substrates in amounts that are sufficient for growth but do not result in feedback effects.

Lysine production

A specific example is the commercial production of lysine which is widely used for supplementation of cereal-based animal feeds. A simplified representation of the lysine production pathway of Corynebacterium glutamicum (Figure 8.3) shows that, as with other bacteria, the initial steps of lysine synthesis are shared with those of the pathways for the synthesis of threonine, isoleucine and methionine.

Mutants of C. glutamicum that are defective in the enzyme homoserine dehydrogenase are auxotrophic but can grow if homoserine (or a mixture of threonine and methionine) is provided. Such mutants produce large amounts of lysine (over 50 g l 1) due partly to the diversion of metabolites away from the other amino acids, but also to relief from feedback control. The first enzyme in the common pathway (aspartokinase) is subject to concerted feedback regulation by lysine and threonine. However, in the auxotroph, fed with limiting amounts of homoserine, the level of threonine will be too low to cause inhibition of the aspartokinase.

Amino acid analogues can be used to obtain feedback-resistant mutants of this system. For example, the lysine analogue S-(2-aminoethyl)-l-cysteine mimics the feedback inhibition by lysine of aspartokinase. Mutants that are resistant to this

they are restricted to naturally-occurring genes or relatively minor modifications of these genes. It is not possible using these techniques to make a product totally foreign to that organism.

Secondly, with classical techniques, work can only be carried out on the basis of the phenotype, i.e. mutants are selected by their effect on the observable characteristics of the organism. This severely limits the changes that can be selected.

The advent of gene cloning (also referred to as in vitro genetic manipulation or genetic engineering) has dramatically changed the picture in both respects. The basis of these techniques is the use of enzymes to cut and rejoin fragments of DNA. In this way, foreign DNA fragments can be inserted into a vector (a plasmid or a bacteriophage) which enables the DNA to be replicated within a bacterial cell. This section provides an introduction to the concepts and applications of gene cloning.

8.4.1 Cutting and joining DNA

The ability of restriction endonucleases to cut DNA at specific sites as described in Chapter 4, is a key factor in gene cloning since it enables the attainment of specific fragments of DNA. Such fragments can be joined together by DNA ligase. Any two ends generated with the same restriction enzyme can be joined in this way, but an EcoRI fragment cannot be ligated to say, a BamHI fragment. (A BglII fragment can however, be joined to one generated by BamHI or Sau3A; although the enzymes recognize different sequences, the sticky ends generated are identical: see Box 4.1). DNA ligase can also join together blunt-ended fragments. Although less efficient, blunt-end ligation can be useful because it does not require the fragments to have been generated with the same enzyme.

Replication of the DNA fragment is achieved by transforming a suitable host strain after ligation with a vector that is itself able to replicate, i.e. a plasmid or bacteriophage. Using a plasmid vector, it would be possible to recover a single colony from an agar plate and to use this to produce a bacterial culture in which each cell carries a copy of the original DNA fragment. Strictly speaking cloning is this process of purifying a single colony from the mixture of transformed cells (or a phage preparation from a single plaque). However, the term ‘gene cloning’ is often used to describe the whole process or even just the step of recombining the DNA fragment with the vector. The power of the technique arises firstly from the fact that the source of the DNA fragment is immaterial; DNA is DNA, and can be cloned and replicated in E. coli, irrespective of its source or its sequence. Secondly, it is not necessary to purify the fragment to be cloned. Provided that there is a method available for identifying which bacterial colonies carry the gene of interest, isolating those colonies in pure form (cloning them) achieves that purification in a very simple manner.

8. 4. 2 Plasmid vectors

Many of the features of a plasmid vector for gene cloning are illustrated by the plasmid pUC18 (Figure 8.5) which is widely used for gene cloning.

Origin of replication

The first requirement is that the plasmid must be able to replicate in the chosen host (in this case, E. coli). The basic replicon from which pUC18 was constructed is related to the plasmid ColE1; this is the region containing the origin of replication (oriV). This plasmid therefore follows the mode of replication described for ColE1 (see Chapter 5).

Selectable marker

Secondly, it is necessary to be able to select those cells which have received the plasmid (transformants). Plasmid cloning vectors are therefore constructed so as to carry one or more antibiotic resistance genes; with pUC18 this is a b-lactamase gene which confers resistance to ampicillin.

There is a second marker carried by pUC18, namely a b-galactosidase gene. (To be strictly accurate, pUC18 codes for a fragment of b-galactosidase with the rest being provided by a host gene so that a functional enzyme can be made; i.e. pUC18 complements the otherwise inactive host gene). Transformation of a suitable host strain with pUC18 will therefore yield cells that produce functional b-galactosidase. These can be detected using the chromogenic substrate X-gal which gives a blue product when hydrolysed by b-galactosidase. Thus, colonies containing pUC18 will be blue, while those without the plasmid will be white. Note that this is not a selectable marker – both types of cell will grow – but it does provide an additional and useful form of discrimination.

Cloning site

The third essential requirement of a cloning vector is that it must contain suitable recognition sites for cleavage by one or more restriction endonucleases. This is so that the circular plasmid can be opened up at that point and the ends ligated with the ends of the DNA fragment to be cloned (see Figure 8.5).

For example, if the DNA fragment to be cloned has been generated by BamH1 digestion and mixed with pUC18 DNA that has been linearized with the same enzyme, DNA ligase will join the BamH1 ends of the fragments, with the result being the recombinant plasmid shown in Figure 8.5.

Figure 8.5 Structure and use of the plasmid cloning vector pUC18

In pUC18, a piece of synthetic DNA has been inserted which contains recognition sites for a number of restriction enzymes. This multiple cloning site allows a considerable degree of flexibility in the choice of restriction enzyme. The plasmid has been engineered so that each of these sites is unique, i.e. it contains no other sites recognized by any of these enzymes.

Insertional inactivation

The position of the cloning site is important. Insertion of a DNA fragment will (usually) result in inactivation of the gene in which that site is found. The cloning site must therefore not be within an essential region of the plasmid. It can be seen in Figure 8.5 that in pUC18 the multiple cloning site is within the initial sequence of the b-galactosidase gene (but does not inactivate the b-galactosidase gene). If a DNA fragment is inserted into the multiple cloning site, it will (usually) prevent the production of b-galactosidase, either by interrupting transcription or by altering the reading frame; this is known as insertional inactivation. Genuine recombinants will therefore be white on a medium containing X-gal and can be distinguished from blue colonies containing the original pUC18.

8.4.3 Transformation

Having produced a recombinant plasmid, it is necessary to put it into a host cell. Naturally-occurring transformation systems, as described in Chapter 6, are not suitable for this purpose, partly because they often do not work well with circular plasmid DNA and partly because in many bacteria (including E. coli) competent cells, able to take up DNA, do not occur naturally. For these bacteria, transformation requires the artificial induction of competence. There are numerous ways of doing this, the simplest of which (Figure 8.6) involves washing the cells repeatedly with cold calcium chloride solution. The competent cells are mixed with the DNA solution and subjected to a heat shock treatment, such as heating them at 428 for 1–2 min and then transferring them back onto ice. They are then diluted into broth and incubated at 378 to allow expression of the newly acquired DNA before plating onto an appropriate medium.

Much effort has been put into optimizing this system for transforming E. coli, including the development of special strains that show high transformation efficiency, up to 109 transformants per g of DNA. Even with such a system, only a proportion of the host cells will have taken up the DNA, so it is necessary to have a good selectable marker on the transforming plasmid, as described above.

Other forms of transformation, notably electroporation, are especially useful for host bacteria other than E. coli and are considered later in this chapter.

8. 4. 4 Bacteriop hage lambda vectors

Insertion vectors

Vectors derived from bacteriophage have been widely used for gene cloning in E. coli. In the simplest of these, insertion vectors, the DNA of the vector has a

cold CaCl2

Log phase culture

Centrifuge

Antibiotic-resistant

transformants

Figure 8.6 Basic procedure for plasmid transformation of E. coli

single recognition site for the chosen restriction enzyme and the fragment to be cloned is inserted at that position (Figure 8.7).

The most obvious difference from a plasmid vector is that after introducing the DNA into a bacterial cell, the success of the procedure is determined by the appearance of phage plaques in a bacterial lawn (see Chapter 4) rather than by the selection of antibiotic-resistant colonies. Another important difference concerns the way in which the cells take up this DNA. In the natural lytic cycle of , lengths of DNA cut from a multiple length DNA molecule are packaged into the empty phage heads, as described in Chapter 4. It is possible to do this in vitro, using cell extracts which contain phage heads and tails and the required enzymes. This process, known as in vitro packaging, produces virus particles that are able to inject the DNA into a host cell.

The natural packaging limits of impose constraints on the size of fragments that can be cloned using insertion vectors. Regions of the genome that are not