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.
9 Phenotypic Properties of Source Microorganisms and Their Genetically Modified Derivatives ~ this chapter, we shift our attention from characterization of the genetic modifications to characterization of the ecologically important phenotypic properties of the unfamiliar microorganism intended for field testing and eventual introduction into the environ- ment. The most relevant phenotypic properties are those that relate to the persistence of the introduced microorganism (or the genetic material incorporated during its modification) and those that affect the ecosystem. Of course, for a microbial application that is familiar, such phenotypic information is likely to be already available or ren- dered unnecessary by a safe history of cast use in the environment. PERSISTENCE Persistence can be viewed as survival of the introduced modified microorganism or retention of particular genetic traits in new genetic combinations resulting from gene transfer. Persistence of the Microorganism If an organism cannot persist in a particular environment, it poses little threat of causing prolonged environmental impact. How- ever, short-term responses may be seen, as in the use of Bacillus 99
100 thuringiensis to control lepidopterous insect pests. An organism that Is killed or does not persist might be viewed as similar to a chern~cal treatment that produces no chemical residues. For exam- ple, some organisms wiD be developed to control pests or degrade pollutants and wiD be used once or reapplied, if necessary. On the other hand, the utility of some applications of genetically modified microorganisms wiD depend on their ability to persist in the environ- ment. For example, genetically modified microorganisms introduced into the rh~zosphere for plant growth promotion or disease control will be most beneficial to farmers if the organisms remain active for years. Persistence and spread are particularly relevant, however, if a proposed application is both unfamiliar and has some potential for adverse environmental effects. ~ such cases, it may be difficult to mitigate an adverse environmental effect simply by halting fur- ther application. Therefore, the potential for persistence and adverse effects should be considered together when establishing levels of con- cern for proposed field tests involving unfamiliar microorganisms; this is reflected in the framework presented in Chapter 11. For unfa- m~liar applications, it Is prudent to evaluate potential adverse effects prior to field testing (Vogel and McCarty, 1985; Tickle et al., 19893. The persistence of an unmodified m~croorg~ism in its usual habitat is largely predictable. Genetic modification could influence persistence if it changed the fitness of the modified organism or sig- nificantly altered its ecological niche. Fitness (the factor that reflects the rate at which a particular type of microorganism Increases or de- creases in number) might increase if a phenotypic change Increased resistance to a noxious substance present in the environment or in- creased the ability of the organism to metabolize a substrate in the environment. It Is often possible to deterrrune the fitness of an or- ganism through laboratory tests (Lenski, 1989), although field tests may be needed. Considerable evidence exists that nonindigenous bacterial pop- ulations, including genetically altered strains, decline rapidly after they are introduced into soil or aquatic environments (Scanferiato, et al., 1989~. This supports the well-documented fact that long- established microbial communities resist invasions by foreign organ- isms (Liang et al., 1982~. If persistence appears undesirable it may be avoided by choosing strains with reduced survival, reduced reproductive capacity, low re- sistance to a predictable change in the environment (such as seasonal
101 heat or cold), or a tendency to lose the specific function of concern. Biological confinement also can be provided by suicide genes (Molin et al., 1987; Curtiss, 1988; OTA, 1988) or by incorporation of addi- tional nutritional requirements. On the other hand, when persistence is desirable, it may be fostered to achieve the benefits for the intended function. Persistence of the Genetic Modification 1 Gene transfer can affect persistence and may occur either to or from the introduced microorganism. Only a new genetic combination with higher fitness than the introduced or indigenous genotypes has any likelihood of persisting. Conjugation, transduction, and transformation are mechan~srns of exchange of both chromosomal and extrachromosomal DNA be- tween bacteria (Freifelder, l9g7; Lenski, 1987; Miller, 1988~. Sim- ilarly, viruses can recombine when two or more types co-infect the same cell (Hershey and Rotman, 1949~. Genetic material is ex- changed most readily among clones of the same species, but ex- change among groups that are more distantly related can occur as well (Roberts et al., 1977; Morese et al., 19g6; Guerry and Colwell, 1977; Hememann and Sprague, 19893. Perhaps the most dramatic recombination mechanism is trance formation, in which DNA from injured or dead cells (Kieft et al., 1987) or extruded by living cells (Borenstein and Ephrati-~lizur, 1969; Orrego et al., 1978) is taken up by living cells of the same or other species (Graham and Istock, 1979; Duncan et al., 1989~. We know very little about the exchange of genetic "formation among closely or distantly related rn~croorganisrns under natural conditions. The possibility that these exchanges occur, however in- frequently, must be taken into consideration in any planned mtroduc- tion. However, In the extensive trials carried out by the Monsanto Company and Clemson University in South Carolina, in which the lacZY genes were used as a marker for a genetically altered strain of Pseudomonas pnorescens, there was no evidence of exchange of the genetic marker with other soil bacteria (Drahos et al, 1988~. Since those tests were lirn~ted both in time and location, it is evident that additional information must be generated on the frequency of gene exchange, particularly in managed ecosystems (window et al., 1989~. Effective genetic transfer between distantly related m~croorgan- isms is far less frequent than between closely related microorganisms
102 because of the improbability of each of a series of events: the in- sertion of DNA from a distantly related organism into the recipient organism; replication of the foreign DNA in the recipient; and set lection favoring the new recombinant genotype. While we cannot assume that any given transfer Is impossible in nature, a transfer of DNA among closely related organisms is far more probable than a transfer among distantly related ones. In principle, organisms ma- nipulated ~ the laboratory to cross genetic barriers may be able in the field to transfer genes secondarily to other organisms; but in fact, even transfers between closely related microorganisms in nature are infrequent and difficult to document. Not aD genetic transfer Is undesirable; in most instances it proba- bly wiD not matter. Unless the recipient organism has a selective ad- vantage, the genetic transfer will have little or no consequence, there will also be no consequence if an organism has a selective advan- tage, but Is innocuous. Examples of genetic modifications that pose little or no risk due to genetic transfer mclude ice- (ice-nucleation deficient) Pseudomonas syrtngae, used to protect plants from frost damage, and plants in the rhizosphere containing the [acZY marker genes, used for monitoring their movement and persistence in soil. PHENOTYPIC PROPERTIES AFFECTING THE ECOSYSTEM If an organism persists in the environment and has the potential to spread, it is important to consider its phenotypic properties that relate to its role in the ecosystem. These properties include compet- itiveness, substrate utilization, environmental range, and host range (if the organism is a pathogen or a symbiont). These properties are also prime candiciates for genetic modification, since many microbial applications depend on them. Examples include enhanced nitrogen fixation, Stereo hos~range biocontro} agents, or beneficial soil bac- teria that must out-compete indigenous m~croflora to be successful. Competitiveness Environments into which microorganisms are to be introduced are diverse ~d often support a complex indigenous microflora. As a result, competitive traits that are advantageous, such as a rapid utilization of abundant substrates, high maximum specific growth
103 rate, and antibiotic production, may be necessary if an introduced microorganism is to colonize successfully in competition with native · - microorganisms. Will the Introduced microorganism persist or spread to other environments? It Is difficult to envision a specialized microorganism faring unusually well in, for example, the plant rhizosphere because the soil microbial populations there are dense and incredibly diverse. However, when an organism is Introduced into an environment in which the indigenous population is restricted in density and diversity, or in which the introduced rn~croorganism has a monopoly on some resource (for example, a substrate difficult to metabolize), then the chances of proliferation to high density may be greater. However, spread probably will be limited if the resource is less abundant outside the target area. It has been suggested that genetically modified microorganisms will be competitively disadvantaged, relative to their wild-type coun- terparts, because of burdens associated with carriage and expression of additional functions. Indeed there are many we0-documented pa- pers to support this suggestion (Lenski and Nguyen, 1988; Brill, 1985; Davis, 1987; Zund ~d I`ebek, i980; I.ee and Edlin, 1985; Duval-~ah et al., 1981~. However, exceptions have been reported in which increased fitness has resulted from carriage of foreign genes in microorganisms (Hart] et al., 1983; Edwin et al., 1984; Bouma and Lenski, 19883. Most of these experunents have been performed ~ the laboratory and may not represent natural conditions. To date, how expression of additional functions affects fitness has not been clearly resolveci. It also has been suggested that an introduced microorganism might competitively exclude another potentially more valuable m~- croorga~sm (Tie~je et al., 1989~. An example cited ~ that Bradyrhi- zobium japonicum serogroup 123 is more competitive in some soils than ~ others and may exclude other more elective nitrogen fixers (Johnson et al., 1965; Moawad et al., 1984~. As no basis other than field tests exists for judging field competition and no clear infor- mation is available on what governs competitive ability among the rhizobia, selection of inoculant strams has been made empirically. Stram 123 forms a symbiotic relationship with soybeans resulting in nitrogen fixation. In the infection process it outcompetes many strains in the field that fix nitrogen more vigorously than 123 under aseptic greenhouse conditions.
104 Substrate Utilization It may be desirable either to expand or to restrict the range of substrates utilized for growth by an introduced microorganism. Expansion will have most appeal in applications for removal of sta- ble compounds from polluted environments. As introduced toxin- degrading microorganisms proliferate, they and the indigenous m~- croorganisms may co-exist owing to the removal of the toxic sum stance (Lenski and Hatt~ngh, 1986~. Genetic modifications also can be used to restrict substrate range, an appealing prospect for biolog- ical confinement of certain introduced microorganisms. Involvement In Ecosystem Processes Microorgan~srns play roles in ecosystem processes important to the sustained habitability of the planet, including major biogeochem- ical cycles: carbon (lignin and cellulose decomposition), nitrogen (ni- trogen fixation, vitrification, denitrification), sulfur (sulfur oxidation and sulfate reduction), and the cycling of elements such as phos- phorous, silicon, manganese, iron, and trace metals. Microorganisms also may produce or use gases that include CH4, NO2, H2, and CO2, and these gases may influence the earth's atmosphere. Field exper- iments should have no detectable influence on these processes but reasonable questions concerning potential adverse ecosystem ejects should be evaluated prior to field testing. lDuviron~nental Range One of the classical criteria for choice of microorganisms for applied purposes ~ their ability to function under a variety of cli- matic and management conditions. The cI~rnatic range of a num- ber of plant pathogens studied has been very restricted (for ex- almple, Pseudomonas solanacearum; Kelman, 1953~. Another well- documented example covers the restricted range of root nodule bac- teria. Bradyrhizobi?`m japonicum, serogroup 123, dominates the root nodules of soybeans grown in the north-central region of the United States, despite a diversity of other strains of the same species found in the soils (Ham, 1980~; however, serogroup 123 is rarely recov- ered from nodules of soybeans growing in the southeast states of the United States (Keyser et al., 1984~. Sirn~larly, Rhixobium trifolii TA1
105 has been a successful inoculant of clover species in eastern Australia but has been unsuccessful ~ western Australia (Parker et al., 1977~. It should be pouted out that non-~digenous or exotic sapro- phytic microorganisms have been introduced continually into soils in the United States for decades without documented cases of harmful responses. Examples of such introductions include those associated with shipment of plant propagation materials such as seedlings and bulbs that carry m~croorganisn~. Host Range Host specificity is well documented in symbiotic and pathogenic associations between plants and rn~croorganisTns. Since relatively few microorganisms enter these relationships, they clearly are specialized. Tm recent years geneticists have identified not only the genes needed to establish symbioses or pathogenesis, but also genes that determine host range (DjorUjevic et al., 1987; Keen and Staskawicz, 19883. From an agricultural perspective, it may be desirable either to restrict or to expand the host range of symbiotic microorganisms. With certain plant pathogens, the incidence of disease relates in a complex manner to the population dynamics of the pathogen ~d its interaction with susceptible host cultivars (Rouse et al., 1985). Two strategies can be envisioned for biocontro! of plant disease: (1) a biocontro] agent ~ introduced to colonize and occupy the target habitat prior to the pathogen's appearance; or (2) the biocontro} agent competitively displaces the pathogen. In the first case, biocon- tro} agents with general features for rapid colonization of the leaf will have broad appeal. In the second, the biocontro! agent will usually need to be a specialized microorganism recognizing the same host specific-signals as the pathogen and out-competing the latter. CHARACTERISTICS OF MICROBL\[ PATHOGENS One of the concerns often raised about genetic modification of rn~croorganisms is the possibility that inadvertent acquisition or loss of one or a few functions Knight convert unrelated noupathogenic organisms into potentially dangerous pathogens of humans, plants, or animals. However, pathogenicity is controlled by a large portion of the genome of prokaryotes. ~ certain plant-pathogenic bacteria,
106 for example, many pathogenicity functions are carried on the chro- mosome, while others are carried on very large plasmas. Thus, the acquisition or loss of a single function cannot convert a nonpathogen into a pathogen. In some cases, changes ~ virulence (increased ag- gressiveness on specific hosts) may be increased by acquisition of the ability to produce a toxin or by loss of an incompatibility func- tion, but the recipient organism must already possess aD the wide variety of genes that win allow it to Infect and colonize a particular host. Endowing rhizosphere bacteria with the ability to produce Bt toxin, as proposed by some biotechnology firms, does not represent conversion of a nonpathogen to a pathogen in the sense described above. Rather, a precise modification has been made that makes the bacterium toxic to certain insect larvae. Background Prokaryotes preceded eukaryotes during evolution. Thus, Tom the time higher plants and animals made their appearance on land, they have been surrounded by ~rucroorganisrns. Parasitic or sym- biotic relationships probably existed very early in evolution, but in spite of such long coevolution, relatively few microbial species have developer} the capacity to parasitize plants or animals (Sequeira, 1984~. For example, very few species of bacteria, representing only five major genera, parasitize plants. Among the gram-positive bacte- ria, which constitute a large portion of the soil ~rucroflora surrounding plant roots, there are no plant pathogens of significance with the ex- ception of a few members of the Corynebacteriae. Even fewer species have developed a symbiotic relationship that allows them to mul- tiply within plant cells. It is evident, therefore, that pathogenesis and symbiosis are highly complex relationships that developed, in relatively few instances, through eons of convolution. It is equally evident that the evolution of a compatible relationship with a host cannot be recreated by simply inserting or deleting a few genes in a nonpathogen by recombinant DNA techniques. Pathogen~cit~r Versus Virulence A distinction should be made at this point between genes that are involved in pathogenicity and those involved in virulence.
107 Pathogenicity is an attribute (the ability to cause disease) of an entire group, irrespective of the fact that particular strains or races may not be pathogenic to a given host. Virulence is the relative ability of an individual strain or race to cause disease under de- fined conditions (Federation of British Plant Pathologists, 1973~. Thus, virulence is a quantitative variable. While it may be am propriate to refer to a particular strain as nonvirulent, the term nonpathogenic may not be appropriate unless a wide range of hosts has been tested. Genes That Confer Pathogen~city Pathogenic microorganisms must possess two general types of genes: (1) those that are important for basic compatibility with the host; and, (2) those that are specific for virulence on particular hosts. To the first category belong housekeeping genes that control general metabolism and that allow the pathogen to grow on the nutrients available in the host. To this same group belong those genes that allow the pathogen to degrade, detoxify, or bypass preformed or induced substances that provide resistance for the host. Thus, one can assume a priori that a pathogen of a particular host must be able to utilize the nutrients available at the site of infection ~d must be insensitive to substances that inactivate or destroy the vast majority of the microorganisms that invade the host. The marked specificity of certain strains of bacteria that colonize leaf surfaces, for example, indicates a narrow adaptation to the nutritional and toxic components of the host environment. Also, these bacteria must be able to withstand intense solar radiation and extended periods of desiccation. Once inside the leaf, those that are pathogens now produce enzymes that degrade plant cell wads or toxins that may be very specific in terms of the cellular targets affected in certain tissues of particular hosts. These virulence functions are controlled by genes that belong to the second category. Gen~for-Gene interactions interactions between pathogens and their hosts are often under the control of apparently gen~for-gene systems (Flor, 1956; Barrett, 1985~. Such systems are common In plant-pathogen relationships, where host resistance depends on single, dominant genes that are sup erunposed on other, more general mechanisms of defense. The
108 gene-for-gene concept accounts for the fact that for every major ret sistance gene in the host there IS a corresponding gene for avirulence in the pathogen. For many pathogens, there is a great deal of evi- dence that av~rulence, rather than virulence, Is the positive function. hnport ant consequences follow from this seemingly contradictory statement. First, incompatibility of a pathogen on a particular host may result Tom the interaction of products of an avirulence gene of the pathogen and the corresponding resistance gene in the host (EHingboe, 1982~. Second, mutations that elirn~nate the av~rulence gene product may increase the pathogenic potential of an organism (Staskawicz et al., 1988; MelIano and Cooksey, 1988~. Ultraviolet treatment of plant parasitic fungi, for example, has long been known to yield mutants that have a wider host range than that of the parent line (Flor, 1958~. There are, of course, many examples of positive-control func- tions that also specify virulence of pathogens. Of particular interest are the host-specific toxins produced by some pathogenic fungi that have highly specific targets in particular varieties of the plant host. A°particular receptor for the toxin appears to confer susceptibility. The toxin Tom Helminthosporium maydis, for example, affects only the m~tochondria of hybrid corn lines carrying Texas male sterile cy- toplasm (Yoder, lOSO). Other species of Helminthospori?`m produce toxins that affect varieties of oats and sugar cane. Artificial hybrids among species of Helminthosporium may exhibit combined virulence to each of the hosts affected by the parental strains, as specified by the toxins they produce. GENERAL P:EIENOTYPIC CHARACTERISTICS OF PATHOGENS -One of the primary characteristics of pathogens is their ability to colonize the host. For many rnicroorganisrns, the ability to attach to the host surface, or to specific host cells in particular tissues, is the first step ~ colonization (Costerton et al., 1981~. Microor- ganisms have developed highly specialized structures (extracellular polysaccharides, fimbriae, appressoria) that provide adhesion to spe- cific host surfaces. Once attached, colonization depends not only on the ability of the organism to utilize nutrients at the site, but also on the ability to compete with other organisms for those same nutrients. A high degree of competitiveness, therefore, characterizes
109 a successful pathogen. Saprophytes share many of these properties with pathogens, but they may be unable to colonize particular eco- logical niches withm a host because of their inability to resist a wide range of host defense systems. A series of nonspecific and specific mechanisms for resistance present a hostile environment that limits the growth of the vast ma- jority of microorganisms that gain access to an uncompromised host (Falcone et al., 1984; Mims, 1982~. With the exception of cornrnensal species, only those organs s that have an array of virulence fac- tors are capable of growing in animal or plant host environments. Extracellular pathogens of anunals, for example, owe their virulence to their ability to resist or inhibit phagocytosis. The production of capsular polysaccharides Is important in this regard. Intracellu- lar pathogens, on the other hand, are highly adapted to ingestion by phagocytes and even use phagocytes as a means for dispersal within the host. There are evident analogies between plant and ani- mal pathogens in this regard. Whether intracellular or extracellul=, the ability to resist, inhibit, or degrade host defense compounds when colonizing the host is therefore one of the key properties of pathogens. Some plant pathogens, for example, are capable of chem- ically degrading phytoalexins, antimicrobial compounds produced by plants upon challenge by potential pathogens. Others are insensitive to phytoalexins from particular hosts. Yet others grow in the host in such a way that they electively bypass the tissues where phytoalexins accumulate. Pathogens can establish themselves in the host because of their ability to produce a large series of compounds, including hydrolytic and proteolytic enzymes, toxins, polysaccharides, and growth regula- tors that affect the host in ways that ultunately favor multiplication or transmission of the pathogen. The enzymes that destroy cell mem- branes or cell walls, the toxins that affect the normal metabolism of the cell, and the growth regulators that influence the ways host cells grow all have an unp act in determining pathogenesis. As a result, the host is damaged much more than would be expected from the strictly energetic drain on its metabolism. Pathogens must also have an elective mechanism for spreading to new hosts. Reproductive structures must be formed that can persist in unfavorable environments or to reach new hosts before they perish. Most pathogenic organisms depend on their hosts or on Rectors for survival, but many are capable of surviving for long periods ~ soil or water (Brubaker, 1985~. The chemotactic property
110 of m~croorganisrns often assumes ~rnporta~ce in this respect since many an~rnal pathogens must move through intercellular spaces to reach target tissues. Acquisition of Violence It ~ evident from this discussion that pathogenicity depends on an impressive array of different characteristics that relatively few microorganisms have acquired through extended coevolution with particular hosts. It is unlikely, therefore, that minor genetic mod- ifications can convert a nonpathogen into a pathogen. It also is clear that factors for virulence are not exchanged ~nd~criminately among microbial populations. Although certain unique properties of pathogens are encoded by plasmas and may be transferable under certain conditions, only a limited portion of the whole array of genes required for virulence would be acquired by the recipient organism. These plasmas, therefore, can confer virulence only to related strains that are already highly adapted to particular ecological niches on or in the host. For example, most of the genes necessary for tumori- genicity of Agrobacterium tumefaciens, the crown gall pathogen, are located in a plasmid (Ti). Acquisition of the Ti plasrn~d by strains of Agro bacterium radio la cter, a nonpathogenic soil inhabitant , auto- matica~y converts them into the pathogen A. tumefaciens. There is a great deal of evidence, however, that the two species are essentially identical except for the presence of the plastnid (Nester and Kosuge, 1981~. Similarly, E. cold is a normal component of human add some animal intestinal microflora, but certain strains are pathogenic and can cause different types of diarrhea or urinary tract infections. Certain commensal strains of E. cold become capable of causing disease upon acquisition of plasmas conferring enterotoxicity and adhesiveness (McConnell et al., 1981~. These pathogenic strains, however, often have short persistence tunes, presumably because they are less fit than strains lacking these elements (Duval-~ah et al., 1981~. In these examples, virulence is increased only when factors Dom very closely related species are acquired. When changes are eRected between unrelated species, pathogenicity is rarely acquired. For example, transfer of pectin lyase genes from Erwinia chrysanthemi, a potato soft-rot pathogen, to E. coZi does not convert the latter into
111 a pathogen, even though it is now capable of rotting potato tuber slices in the laboratory (CoHmer and Keen, 1986~. The overall conclusion of this discussion is that, as a rule, an unrelated Iriicroorganism does not become a pathogen merely because it has received a portion of the DNA of a pathogenic species. SUMMARY POINTS 1. E ram the standpoint of assessing potential effects of unfamil- iar application of microorganisms ~ managed ecosystems, the most relevant phenotypic properties are those that relate to the persis- tence of the microorganism (and its genetic modification) and those properties that may have adverse effects. Such assessments are not necessary, however, if a proposed microbial application is familiar and has a safe history of usage in the environment. 2. Key phenotypic properties include the fitness of a genetically modified microorganism relative to its unmodified counterpart; the potential for gene transfer between the introduced rn~croorg~sm and the indigenous microflora; the physiological tolerances of the introduced microorganism; the competitiveness of the introduced microorganism; the range of substrates available to the introduced microorganism; and, if applicable, the pathogenicity, virulence, and host range of the introduced microorganism. 3. Persistence of genetic modifications may occur in either of two ways. The introduced microorganism itself may survive and propagate to form a sel£sustaining population. Alternatively, the modification may persist in new genetic combinations resulting from gene transfer. The potential effects of undesired persistence can be avoided by the use of appropriate strategies for biological confinement of the introduced microorganism and its genetic material. 4. Microorganisms intended for environmental introduction may be modified for either highly specialized or more generalized phenotypic properties, such as competitiveness, substrate utiliza- tion, physiological tolerance, and host range (either pathogen or symbiont). Persistence after introduction is most likely for rn~croor- ganisms having generalized properties and for rn~croorgan~sms hav- ing properties that allow exploitation of previously unfilled ecological niches. 5. Pathogenicity is controlled by a large portion of the genome of rn~croorganisms. This is so because a pathogen must be able to cross a large number of potential barriers to be successful. The
112 successful pathogen must be able to attach to appropriate sites on or in the host, to compete for nutrients within the host, to resist various host defenses, to survive and persist when the host is not present, and to be effectively transmitted to new hosts. Therefore, acquisition of a small number of genes from a pathogenic source cannot convert an unrelated, nonpathogenic microorganism into a pathogen. 6. However, if the recipient is closely related to the pathogenic source, or if the recipient is itself a pathogen, increased virulence for particular hosts may result. In these special cases, the recipient al- ready contains a large complement of genes related to pathogenicity.