Living biosensors for the management and manipulation of microbial consortia.

INTRODUCTION

The science of microbiology grew rapidly after the realization that some microbial species were responsible for human disease, and accordingly much work has focused on medical microbiology. The isolation of pure microbial cultures, so that pathogens could be identified, was considered extremely important. Subsequently, microorganisms were recognized as being inherently interesting and were used to study physiology and genetics. They were ideal for these studies because they could be grown quickly (at exponential growth rates) and in convenient amounts, allowing repetitive experimentation and statistical analysis of results.

The study of pure cultures is still important, and the literature is replete with examples of how pure cultures of microorganisms can be used to gain insight into fundamental biological processes. However, pure cultures of bacteria rarely exist in nature and probably do not grow at exponential rates as they do in laboratory flasks. This realization has created a greater appreciation for the study of microbial ecology, i.e. the relationships between microorganisms and their environment. This field has practical benefits: microorganisms are intimately involved with important processes throughout Nature. For example, pathogenesis is a suitable model for ecological study, as pathogens successfully invading a host must adjust to a new ecological setting, one that may be very hostile. An understanding of how a pathogen successfully invades the niche of established flora and promotes disease would offer great insight into the disease process. Commensal relationships are also important, since they are critical to human health. In addition, plant-microbe relationships exhibit symbiotic relationships, and microbial communities are responsible for nutrient cycling, including the bioremediation of hazardous wastes.

The problem with studying microbial ecology is that the systems are complex. A pure culture of bacteria is somewhat predictable: if you have a liquid culture at a certain optical density, then you can be fairly certain how many cells are present. On the other hand, if you had a binary culture--two species mixed together--at the same optical density, you could have 99% of one and 1% of the other, or vice versa, or a completely different ratio. The presence of additional species compounds the problem. The techniques useful for pure culture studies are often unsuitable for describing mixed cultures.

As the interest in microbial ecology grows, so too does the attention it receives from many other scientific disciplines. Other disciplines may bring new methods for looking at complex systems, and especially for determining the impact of one species versus another.

This review aims to describe some approaches for examining microbial consortia, particularly those techniques that permit the collection of data without disturbing the consortium. If the consortium is undisturbed, continuous examination over time is possible, which may lead to an increased understanding of the growth and development of these complex systems.

TECHNIQUES FOR MICROBIAL CONSORTIA

Many valuable techniques have been used to describe microbial communities. Molecular techniques, which can distinguish between species, have proved especially valuable. The review by Jain et al (17) presents an overview of these different techniques, while the recent review by Kloepper & Beauchamp (19) discusses how they can be applied for examining bacteria in the rhizosphere. Other techniques have been developed to examine the microenvironment of consortia, such as the use of microelectrodes, or for the description of whole communities, such as the analysis of extracted lipids from samples (51). Biofilms can now be studied using nondestructive techniques (32; D Nivens, in preparation), which have been used to follow bacterial adhesion and colonization.

Methods allowing one to describe individual species within microbial consortia with greater detail and resolution find immediate application. Such examinations have two main objectives: to identify the presence of one species relative to all others, and to measure the physiological activity of the various species in the consortium. For example, Pichard & Paul (37) investigated whether a gene of interest was actively expressed in seawater. They recovered mRNA and DNA by a destructive sampling technique and quantified transcription level versus gene dose, to give a normalized result. Their method depends on the power of nucleic acid hybridization to produce a specific assay of activity and yields unambiguous results. However, hybridization techniques are not always sensitive enough for work with consortia. In other experiments, biofilms were examined using in situ hybridization (1, 38). Although this is also a destructive technique, it leaves the biofilm community reasonably intact. The great advantage of this system is that it allows the structure of the biofilm to be seen in relation to the placement of several bacterial species. It depends on the production of specific DNA probes for the 16S rRNA molecule, which are labeled with a fluorescent compound. The cells must be made permeable to the probes for a hybridization reaction to occur. Specific species within the biofilm can then be visualized using an epifluorescence microscope.

Bioreporter genes have a long history of usefulness in the study of genetics and physiology and have found some applications in microbial ecology. Bioreporters (such as lacZ, cat, neo, and xylE) supply an assayable gene product when the gene product of interest is difficult to assay. In pure culture analysis, the most frequently used bioreporter is the lacZ gene, for which a sensitive and rapid biochemical assay has been described (27). The lacZ gene has also been used for some environmental analyses, such as the tracking of a [lacZ.sup.+] strain of Pseudomonas fluorescens (14). This approach is unquestionably valuable, but the use of lacZ as a bioreporter is limited. The lacZ phenotype is not unusual, and it may compose a significant background in ecosystem experiments. Background can often be limited by using a strain that contains a mutation in the host lacZ system, such as many Escherichia coli strains, or by using a strain that is naturally [lacZ.sup.-], such as Pseudomonas spp. However, if other species in the consortium are [lacZ.sup.+], it is impossible to tell which fraction of the assay results from the bioreporter and which results from background. In the case mentioned above, the…