Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
8 Properties of the Genetic Modification Powerful new molecular methods for DNA manipulation pro- vide a meaIls for constructing microorganisms with novel genotypes that cannot be duplicated by classical methods and would be highly unlikely to occur naturally. In addition, these new methods enable genotypes to be characterized with a degree of precision not previ- ously available. Genes now can be manipulated by methods for the synthesis, sequencing, cutting, and splicing of DNA molecules, and changes often can be characterized to the base sequence of the DNA. Unmodified or modified, genes can be inserted into virtually any microorganism, even if the donor and recipient microorganisms do not exchange genetic "formation under natural conditions. Molec- ular methods of DNA modification also permit deletion of a precise region of a genome and its replacement with a similar or altered segment of DNA. Thus, the new molecular methods support the production of novel genetic combinations so that microorganisms can perform new functions. The question has been raised whether the microorganisms produced by these methods present any risks not associated with microorganisms produced by classical microbial genetic techniques. The fact that many molecular methods provide far finer precision than has previously been available means that scientists can produce modified organisms with predictable genotypes. A general example will illustrate the superiority of the precision of molecular genetic 86
87 modification over classical genetic modification. In classical proce- dures, to transfer the ability to produce a particular protein Dom one bacterium (the donor) to another (the recipient), total DNA from the donor would be used to transform the recipient. Selection for the Meshed trait would yield a strain that produced the target protein, but also could contain additional DNA that contributed to the recipient's phenotype. Through the use of new molecular pros cedures, the target gene can be cloned from the donor and used alone to transform the recipient to produce a strain differing from its parent in only one gene. Precise knowledge of genotypic modi- fications would simplify experiments to determine the nature of the phenotypic changes. HISTORY To date, only ~ relatively few field trials have been conducted with microorganisms modified by molecular methods. It is instruc- tive to review three of these. Ic - Nucleation-Deficient Pseudomonas Mut~ts Certain epiphytic pseudomonads damage plants by acting as nu- clei for ice-crystal formation. To control this problem, it has been suggested that pseudomonads unable to form ice nuclei might pro- tect plants by excluding colonization by ice-positive (ice+) strains; both iced and ice negative (ice~) pseudomonads occur naturally on plants. The ice- mutation In Pseudomonas syringes was produced by deleting from an ice+ strain the specific gene encoding a protein that acts as the nucleus for crystallization (window, 1988~. The en- tire nucleotide sequence of this chromosomal gene is known, and, the protein has been isolated and well characterized. The ice~ cleletion mutant has been characterized genetically and physiologically (Lin- dow, 1988~. The modified genome contained no detectable foreign DNA sequences; it no longer produced the ice-nucleation protein; it did not Educe ice nucleation in vitro; and it was phenotypicaby indistinguishable from its parent in all other traits exarn~ned. It should be noted that it was selected for increased colonization, and changes in this characteristic were found. In field tests, the organism performed as anticipated (I,indow and Panopoulos, 1988~; that is, it reduced ice-nucleation on plants to which it was applied, it showed no detectable spread beyond the test zone, and after 18 days it could not be detected In the soil.
88 Tr~n~fer-Defective Mutant of Agrobacterium radiobacter Agrobacterium radiobacter is closely related to Agrobacterium tumefaciens and is a common soil inhabitant. Strain KS4 is used commercially to control crown gall, a bacterial disease of plants. It does this in part by producing a highly specific antibiotic, called agrocin 84, that kills susceptible pathogens (New and Kerr, 1972; Kerr and Htay, 1974; Ellis et al., 1979~. Production of the antibiotic, as well as immunity to the agent, are encoded by a plasmid present in strain K84 (Ellis et al., 1982; SIota and Farrand, 1982; Farrand, et al., 1985~. This plasmas is self-transmissible (~a+), and its transfer to pathogens could result in their becoming immune to control by strain K84 (Panogopoulos et al., 1979~. Molecular methods were used to delete from the plasmas the genes required for conjugal transfer (Jones et al., 1988~. The genetic alteration occurred as anticipated, and no foreign DNA sequences were detected ~ the new strain. The deletion had no eject on production of the antibiotic by straw K84, and, aside from the Tra~ character, the altered stram was pheno- typically indistinguishable from its parent. In field tests conclucted in Australia, the altered stram controlled crown gas as weD as its unaltered parent (Jones and Kerr, 1989~. Moreover, because it can no longer transfer the plasmas, it is considered a safer strain for long-term use in the field. [actos - Catabolizing Pseudomonas fIttorescens The field use of fluorescent pseudomonads to control certain plant diseases has generated much interest (Davison, 1988~. As a too! for the study of how such organisms behave ~ the environment, one such pseudomonad was marked with a genetic trait that would allow its direct selection from environmental samples. The marker chosen was the ability to utilize lactose; the gene for lactose utilization is rarely found In species of Pseudomonas. The genes for lactose utilization from Escherichia cold were inserted into the chromosome of the P. fluorescens strain with a transposon clelivery system (Barry, 1986; Drahos et al., 1986~. Transposons are mobile genetic elements that cause mutations by interrupting the DNA sequence of genes into which they insert. They sometimes act by preventing expression of genes distal to the insertion site, a phenomenon called polarity. Physical analysis confirmed the insertion, but the exact site was not known. The marker no longer was transposable, however, because the tr~nsposon was defective. The altered strain was indistinguishable
89 from its parent except for its associated ability to catabolize lactose and thus to form blue colonies on an agar medium containing an indicator dye. The strain has been used in field tests to deterrn~ne the efficacy of the iac marker for tracking purposes. Reports to date indicate that the marker works wed for its intended purpose and that the P. puoTescens stram shows no unexpected behavioral patterns in the environment (Drahos et al., 1988~. From these three examples, two key conclusions can be drawn. First, microorganisms can be engineered with the new molecular methods such that the genetic alterations are known precisely. Sec- ond, these modified rn~croorganisms have behaved in the environment as anticipated. That is, the genetic alterations have produced no doc- umented phenotypic alterations, either desirable or detrimental, that had not been predicted. GENETIC CONSIDERATIONS A number of unport ant considerations arise with respect to the actual genetic alteration of microorganisms destined for release into the environment. These considerations apply to manipulations car- ried out by classical microbial genetic techniques as well as to those performed with new molecular procedures. Ape of Genetic Alteration Two major types of alterations should be considered. The first is removal of a trait, either by mutational inactivation of the encoding gene or genes or by deletion of the DNA Leghorn encoding these de- term~nants. The second is the addition of new traits to an organism. Additions can be accomplished by inserting new genes into the chro- mosome or indigenous plasm~ds of the organism or by introducing a new plasm~d encoding the traits of interest. These types of manipulations connote nothing with respect to the sources of the genes or the particular methods used. Manipulated genes can be from a substrain of the organism, from a near or distant relative, or from some unrelated organism. For any of these sources, the alteration potentially could be effected by classics rn~crobial ge- netic techniques, new molecular methodologies, or some combination of the two. For mutational inactivation, commonly used techniques involve insertions of transposons. These elements often encode resistance to
·1 ~ ~ go one or more antibiotics, and expression of these resistance functions acts as a marker for the presence of the elements. Use of transposons has two drawbacks with respect to environmental releases. First, they are mobile and can be excised from their original target site, thus reversing the mutation. Second, their mobility allows them to move to other sites or, under proper conditions, to other organisms. The potential spread of such antibiotic resistance determinants has led many scientists to suggest that alternative methodologies for in- sertional mutations be used. Two solutions are available: Certain transposons have been modified to render them incapable of excision or further transposition after insertion at the site of interest (Barry, 1986~. Alternatively, Gene cassettes can carry fragments of DNA for insertion In sites within genes of interest (Close and Rodriguez, 1982; Prentki and Kirsch, 1984~. Such insertions disrupt gene func- tion in a manner similar to that of transposons, but they have the advantage of producing stable genetic alterations. It has also become possible to introduce precise mutations by changing bases at predetermined positions within a gene (Taylor et al., 1985~. Reintroduction of this modified gene into the host of origin, and its exchange for the wild-type allele, constitutes an extremely accurate method for constructing mutants lacking a par- ticular function. Such alterations can be permanent and do not introduce antibiotic resistance traits. However, the methodology rev quires detailed knowledge of the target gene. Mutations that result from Insertions or base substitutions are theoretically revertible, even ~ the absence of genetic exchange. Deletion mutations, on the other hand, are considered nonreversible because all or part of the gene in question has been physically re- moved. Perhaps the most useful technology to effect such deletions involves gene replacements (Ruvken and Ausubel, 1981~. The gene to be altered is cloned, and a small piece is removed; both steps use molecular methods. Because the cloned DNA is small, the exact nature of the deletion can be designated to the level of the nucleotide sequence. The cloned DNA is then reinserted into the target host and the altered gene is made to exchange for the wild-type copy. This produces an organism containing the altered gene, lacking the w~a-type copy, ant] expressing the phenotype of a cleletion mutant (Scolaik and Haselkorn, 1984; Jones et al., 198S, Lindow, 1988~. The technique can be performed in such a way that no other traits, such as resistance to an antibiotic, are added during the process (Jones et al., 1988~.
91 Addition of new genes can be accomplished by introducing them onto new plasmids or by inserting them into the chromosome or indigenous plasmas of the target organism. Introduction of the genes onto a new plasmas will invariably result in the acquisition of any additional traits encoded by that plasmid. The desirability of such ancillary traits must be considered within the context of the intended use of the organism. Plasrn~ds introduced into a target microorganism may be unmodified elements isolated from related or unrelated bacteria, constructs made by classical genetic means, or the products of new molecular methods. Several strategies can be followed to introduce new genes into chromosomes or indigenous plasmas of target organisms. ~ one, the genes are cloned into transposons, which are allowed to insert randomly into the plasrn~d or chromosome of the target organism. This strategy has the same disadvantages as using transposons as insertional mutagens namely, the mobility of transposons and their ability to be excised. Another strategy is to use defective transposons (Drahos et al., 1986~. Third is recombinational marker exchange; conceptually this is similar to gene replacement for producing dele- tions, as described above. The new genes are cloned into a segment of DNA previously removed from the target organism, and this modified segment then is reintroduced into the target organism. Obviously, the piece of DNA into which the new genes are inserted must be chosen carefully, so that manipulations do not inactivate essential genes. Such deleterious effects usually can be minimized by cloning new genes into dispensable or repeated sequences. Regulation of Gene Expression Two broad classes of genetically modified microorganisms have been discussed: ones ~ which function is lost because a gene has been deleted or inactivated (for example, construction of ice~ pseu- domonads) and ones acquiring a new function because a gene has been added (for example, production of an insecticidal toxin). Tm the former case, there is no concern about expression of the gene if the deletion or the inactivating event ~ permanent. With an added func- tion, molecular methods may allow precise control of the expression of that function. Expression of most bacterial genes is regulated at the level of RNA transcription (see, for example, Reznikoff and Gold, 1986~. Genes are turned on by promoting or permitting their transcription,
92 and they are turned off by preventing their transcription. Based on extensive work with E. cold (and to a lesser extent with Pseudomonas, Rhizobium, and Bacillus), it is known that a number of systems are available for controlling the expression of cloned genes. Some shy tems can be turned on, In the laboratory, by the addition of inducer molecules. For example, a gene in E. cold controlled by the lac pro- moter c" be switched on with isopropyl p-D-thiogalactopyra~oside, a lactose analog. Others can be turned on by ultraviolet irradiation or by heating. In Rhizobium species, it Is possible to Educe gene expression by adding secondary plant metabolites, such as flavonoids present in root exudates (Firemen et al., 1986; Peters et al., 1986; Redmond et al., 1986; Djor~jevic et al., 1987~. Similarly, certain genes in Agrobacterium strains are specifically induced by phenolic compounds, such as acetosyringone, present in plant wound exudates (Stache! et al., 1985; Stache! et al., 1986~. With inserted genes, the mode of regulation may be especially important. Is the inserted gene under the control of its own pro- moter? Does the modified microorganism produce the proper regu- latory signal, or is the gene for the regulatory component also present in the inserted genetic material? In some cases, it may be useful to place the gene or genes of interest under the control of a promoter that responds to inducing stunuli in a manner most appropriate for the intended function in the environment. If the m~croorg~ism Is modified to degrade a chemical pollutant, for example, expression of the degradative functions could be controlled by the presence of the appropriate chemical. Microor- gan~sms designed to interact with plants (for example, to protect them against pests) might be modified so that specific gene functions could be induced by a plant product such as a component of a root exudate. Under other circumstances, constitutive expression of the modi- fied traits may be advantageous, for example, in organisms designed for generalized functions under a broad range of environmental con- ditions. Constitutive expression may also be advantageous in organ- isms designed for specific tasks or habitats. For example, constitutive expression of a biodegradative function might allow maximal perfor- mance of biodegradation, while conferring a metabolic burden to put the organism at a competitive disadvantage upon depletion of the target chemucals. Thus, the microorganism might disappear from the ecosystem when its degradative task was complete.
93 Intended and Unintended Changes Whether the genetic alteration involves the addition or deletion of functions, it always has been important to determine if traits other than those intended have been affected. One of the great advantages of the molecular genetic techniques is that they are usually more precise, and hence their use m~nirn~zes the likelihood of unintended phenotypic changes. Nevertheless, even precise genetic manipulation may produce unintended genetic changes owing to the pleiotropic nature of gene action. For example, changing a single gene may make a bacterium resistant to a bacteriophage while also changing its ability to compete for limited resources. In this example, the surface receptors that determine phage sensitivity often are also involved in transport of nutrients into the bacterial cell (Braun and Hantke, 1977~. Similarly, ~ isolated instances a change in host range may be caused by introducing genes at other than the target locus (Staskawicz et al., 1984; Gabriel, et al., 1986~. Precision of Characterization Many of the new molecular methods permit a degree of genetic characterization unobtainable by classical genetic exchange and re- comb~nation processes. ~ some instances, it may be ~rnportant to have a thorough knowledge of the genes being manipulated, even to their base sequences. This knowledge can provide information on regulation of gene expression at the levels of transcription and trance ration, ~ well as on the properties, activities, and fates of the gene products. New molecular methods exist to make precise alterations in the genes, to verify these alterations, and to test their effects on the expression of the genes. This should not be taken to mean that only constructs charac- terized to base sequence are suitable for environmental release. This degree of characterization certainly is not available for organisms constructed by classical genetic techniques. Furthermore, one can conceive of microorganisms modified by new molecular methods, but having complex genetic systems, in which genetic characterization of the sequence would be exceedingly difficult. Again, we emphasize that it is the phenotypic characterization of the modified m~croor- ganism that is of primary concern.
94 Source of New Genes Considerable attention has been given to the sources from which new genes are derived for insertion into modified organisms. For ex- ample, might the acquisition of one or a few genes from a pathogenic source convert the recipient into a pathogen? This question is ad- dressed in depth In Chapter 9. The overall conclusion is that an · . . unrelated microorganism does not become a pathogen merely bet cause it receives a small portion of the DNA of a pathogenic species. Instead, pathogenesLs is a complex phenomenon requiring the coor- dinated action of many penes fMiller et al 1~) Nevertheless ~_ _ , J , a nonvirulent microorganism could become vir '~1~' ~ ;` __:_. AL_ __ ~ "luau t~ 1U l~lV~U Due necessary genes from a related pathogen that complemented the recipient's existing (but incomplete) viru- lence factors. For example, when a single gene encoding an enzyme ,L,~^L ;___l _.~L~ Lit 1__ ~ 1 · · . . -~" ^~-u~u~o o"= Pica ~lly~c'=exln' plSaLln, was transferred from the fungal pea pathogen, Nectria hematococca, to the fungal corn pathogen Cochliobolu~ heterostrophus (Olen Yoder, personal com- munication, 1989), the transformed fungus produced lesions on pea leaves, whereas before it had produced lesions only on the sterns of peas. Note that the recipient organism was a pathogen ~ its own right. Although the new genes increased the pathogenicity of the organism, in that it produced lesions on pea leaves, these genes did not convert a nonpathogen to a pathogen. Genetic Markers Marker genes, introduced for tracking or identifying the modified microorganism, can be introduced to evaluate the spread of the organism, its persistence in the intended environment, its ability to colonize a particular habitat, or its cal)acitv to exchange ~n~.`ir information with indigenous rnicroflora. Antibiotic resistance traits often have been used for such pur- poses, but they have drawbacks; resistance functions often impose substantial ~netabolic burdens (Zund and Lebek, 1980; Lee and Edlin, 1985), and their use as markers may contribute to the proliferation us ~c'u~c-res~srant microorganisms. As alternatives, marker genes encoding enzymes for the catabolism of substrates not normally uti- I~zed by the particular microorganism have been used successfully. For example, the lacZY genes of E. cold were introduced into a soil pseudomonad that could not catabol~ze lactose. This allowed it to ~, ~.,~_ ~_u^~ ~$ __L ~, . .
95 utilize the disaccharide and perrn~tted its differentiation from lac- pseudomonads because of its acquired ability to form blue colonies on plates containing the chromogenic substrate, X-g~ (Hemming and Drahos, 1984; Drahos et al., 1986, Drahos et al., 1988: also see above). Other candidate marker genes encoding catabolic functions include xytE and opine utilization determinants (Tempe and Petit, 1983; Dessaux et al., 1987~. Synthetic polynucleotides or DNA fragments encoding no func- tions but inserted into the chromosome or into plasmas can serve as markers. Nucleic acid hybridization with probes homologous to the introduced sequences then serve to identify the marked organ~srns (Attwood et al., 1988~. The polymerase chain reaction (PCR) pro- vides a variant on this strategy (Steffan ant] Atlas, 1988), in which a specific sequence unique to the organism to be tracked is greatly am- plified by cycling through a DNA synthesis system. This amplified DNA is detected by hybridization with probes specifically homol- ogous to the sequence. The probed sequence may be a naturally occurring part of the microorganism's genome, or it may be a seg- ment (synthetic or recombinant) inserted into the genome. However, it must be unique to the organism being tracked. The only other requirement is that techniques must be available for the effective extraction of DNA from the samples to be assayed. PCR methods have the potential for increasing the sensitivity of detection while maintaining a high degree of specificity. However, markers based on hybridization are limited in that they cannot be used for direct selection of the microorganism from environmental samples. Rather, they are useful only for screening sa~nples that may contain a large excess of other organisms. Biological Confinement No discussion on genetic alterations would be complete without consideration of genetic manipulations intended to biologically con- fine the modified microorganism and its genes. Bacteria can and do exchange genetic information in the environment with low fre- quency (Levy and Novick, 1986; Stotzky and Babich, 1986; Tenors et al., 1987; Schofield et al., 1987~. The scientific literature gives limited information on gene transfer in nature under realistic con- ditions of population densities and structures. However, it ~ clear that rn~croorganisms can exchange genetic information in soils and
96 water ~c] on, or within, plants and animals. The critical variables, namely the frequencies of such exchanges under natural conditions, are poorly known and warrant further studies with marker genes. BE cause exchange occurs, it is desirable in some instances to use gene combinations that minimize the possibility that these genes will be transferred to other organisms in the environment. Whenever pow sible, new genes should be introduced onto the chromosome of the target organism. As an alternative, nonconjugal and nonmobiliz able plasrn~ds (Levin and Rice, 1980), either indigenous or introduced, may be appropriate if they are not disseminated to indigenous mi- croorgan~sms under field conditions. Markers can be used to identify indigenous organisms that have acquired genes from an introduced strain. Situations exist in which the spread of a modified trait is desir- able. For example, consider hypovirulence for controlling a fungal disease such as chestnut blight caused by Cryptonectria parasitica. Certain strains of C. parasitica have been identified that are con- siderably less virulent than the pathogens that have destroyer] the Anierican chestnut forests. This hypovirulence is associated with the presence of double-stranded RNA (Fulbright, 1989~. Inoculation of infected chestnut trees with the hypov~rulent strams can effectively halt progression of the blight and allow the tree to overcome the in- faction. This phenomenon appears to be associated with the transfer of the double-stranded RNA molecules to the virulent pathogens. This induces their conversion to hypovirulence (E`ulbright, 1989~. The spread of the genetic elements-in this case naturally occur- ring double-stranded RNA molecules is a necessary feature of the system. One can envision that genetic modification of the doubly strandecl RNAs might lead to better control agents. If a microorganism might persist beyond the intended period of usefulness it may be necessary to utilize confinement, for example, biological confinement by suicide genes (see also Chapter 9~. For exa~nple, a modified microorganism could carry a ~suicide" function that Is repressed only when the introduced microorganism is per- form~ng its intended function. When that function is completed, the organism Is kiDed by derepression of a suicide gene (Molly et al., 1987; OTA, 1988; Curtiss, 19883. Alternatively, microorganisms intended for environmental introductions might be confined by incorporat- ing into their genomes restrictive nutritional requirements, which could prevent the microorganisms' persistence or spread beyond the intended time or space.
97 In practice, however, the goal may be to maintain the organism in the environment. This may be hard to achieve at times because competition with indigenous microorganisms may make it difficult for a modified microorganism to persist at a population density sup ficient for it to carry out its intended function. Modified strains may lose their environmental competence as a consequence of burdens of carriage and expression of additional functions, or as a consequence of the strands prior adaptation to laboratory conditions. SUMMARY POINTS 1. Modifying microorganisms by classical genetic methods may often be possible. ~ other cases, the development of novel genotypes may require molecular methods. The key concern, however, Is not the method by which the microorganism is modified, but rather the phenotypic properties conferred by the microorganism's new and preexisting genomic complements. 2. Molecular methods, used either to modify or to characterize genotypes, provide a degree of precision unavailable through classical microbial genetic techniques. This precision increases our knowledge about the genetic alteration and may improve our ability to predict how the organism will perform In the environment. 3. When evaluating rn~croorganisms for environmental mtro- ductions, the following characteristics should be considered: (1) the influence of the genetic alteration on the relevant phenotypes of the organism and (2) the genetic mobility of the altered traits. The presence of foreign DNA in a modified microorganism is, by itself, of little concern, but how it influences the expression of phenotypic traits and the mobility of genetic material is important. 4. The pathogenicity of a rnucroorganism results from a com- plex interaction among a number of genes and gene products of the pathogen and host. It is highly unlikely that moving one or a few genes Tom a pathogen to an unrelated nonpathogen wall confer on the recipient the ability to cause disease. If the recipient is closely related to the pathogenic donor, increased virulence may result if genes directly affecting virulence are transferred. The phenotypic properties of a microorganism intended for introduction into the en- vironment are of primary concern and the source of the gene has relevance only for understanding the nature of the modification.
98 5. ~ initial Geld testing of Oar geuetlcaDy modGed ~- croorg~, it m~ be desirable to mark the ~croorg=~m ~ tab can be monitored Her Us introduction into the Geld. It also m~ be desirable to enact genetic modiRc~ions designed to Cat pers~- teDce Ed minimize trouser of genetic material to the indigenous Crop.