Plasmid transfer for enhancing degradation capabilities.

The kinetics of plasmid conjugation for the TOL and RP4 plasmids depend strongly on the donor cells' specific growth rate and substrate concentration, both of which determine the cells' energy availability. Although transfer rates can be large when energy availability is high, normal biological processes have low energy availability. Therefore, we propose and evaluate preliminarily a simple scheme to create a small zone of high energy availability.

One of the pressing goals for environmental engineers is to increase the rates of biodegradation of hazardous organic chemicals in processes for treating wastewaters, drinking waters, groundwaters, and soils. Because discharges of hazardous chemicals in effluent waters, off-gases, and wasted sludges or soils are increasingly restricted by law, the only alternative for controlling hazardous organic chemicals is biodegradation, which can destroy the contaminant molecules. When the degree of biodegradation is not sufficient with the normal means of biological treatment, more advanced strategies are needed to enhance the biodegradation rates.
This report summarizes research aimed at enhancing biodegradation rates by directly controlling the microorganisms' genetic capability through horizontal gene transfer. The application of this research involves using plasmid conjugation to extend or augment the biodegradative capabilities of bacteria that are well suited to function in treatment systems. The particular scientific goal of the research is to This

Plasmid and Horizontal Transfer
Horizontal transfer of genetic information is a naturally occurring phenomenon involving plasmids, which are covalently closed circular strands of DNA that exist and replicate autonomously from the host chromosome (1). Being accessory and mobile, plasmids can be introduced into bacteria indigenous to a treatment process without requiring a chromosomal change. Thus, the genetic elements responsible for making the indigenous species ecologically fit for a treatment system can be maintained while new elements needed for specific degradation reactions can be added via the plasmid. Furthermore, since replication of plasmid DNA is not dependent on cell division, the genetic information can be amplified and can proliferate throughout the microbial population without the need for creating a growth rate advantage for any added or indigenous species. Because the selective pressures inherent to treatment processes (e.g., low substrate concentrations of diverse substrates, slow specific growth rates, and aggregation) are not going to be alleviated, proliferation of critical, new genetic information to ecologically fit, indigenous bacteria is an extremely promising strategy for enhancing process performance for degradation. This is especially true when the target contaminants are present in concentrations too low (e.g., much less than 1 mg/I) to exert any significant selective pressure themselves.
The mechanisms affecting plasmid transfer and stability are illustrated schematically in Figure 1. Three types of cells exist: * donor cells, signified by D, contain the plasmid; * recipient cells, signified by R, do not contain the plasmid; they are chromosomally distinct from donor cells; and * transconjugant cells, noted by T, are recipient cells that have gained the plasmid. Thus, they are chromosomally R, but also contain the plasmid. Environmental Health Perspectives Figure 1 also illustrates the three plasmid-transfer processes: * DR transfer is a conjugation event in which D and R cells come into contact, the plasmid DNA is replicated, and the replicated plasmid is partitioned into both cells, creating D and T cells. Note that conjugation increases the number of plasmids and plasmid-containing cells, even though the total number of cells is the same; * TR transfer is a similar conjugation event, except that the plasmid is donated by a transconjugant and the result is two T cells; and * Plasmid loss from the transconjugant cell can occur through segregation (the improper partitioning of plasmid DNA during cell division), restriction of the plasmid, or unfaithful replication during cell division (2,3). Each transfer process occurs at a rate that is dependent on the plasmid, the bacterial species, and conditions within the system. Figure 1 indicates the rate expressions we employed for each process (4). Although we investigated all three transfer processes (5-7), we report here only on DR transfer, for which the rate expression is rDR= k,IDR [1] in which rDR is the rate of formation of transconjugants by DR  Trends for TR transfer were parallel to those for DR transfer (5-7), while loss rates were small for the systems studied (5,6). In the remainder of this report, we present experimental results for the value of kt, and what affected its value. Transfer of the RP4 Plasmid Our initial work (6,7) investigated the transfer kinetics for the RP4 plasmid, a promiscuous plasmid encoding resistance to antibiotics, including kanamycin. The donor strain was Rhodobacter capsulatus, and the recipient was a Pseudomonas species isolated from one of our laboratory biofilm reactors. Transconjugants were assayed by their ability to grow on plates with kanamycin-amended Luria broth medium.
The kinetic experiments were performed with donor cells harvested from the exponential phase of batch growth. The harvested cells were washed, mixed together with known cell numbers, and followed for the increase in transconjugant numbers.
The value of kt, was computed from the initial results, wh'ich always gave a linear increase in T with time.
The kt, values, summarized in Table 1, show moderately large values (4,8): from 0.0002 to 0.03 gT I (gD gR day)-. More striking is that the value of kt, systematically declined as the time for harvesting the donor cells progressed from early exponential to late exponential. We interpreted these results as evidence that the depletion of internal energy storage materials caused a reduction in plasmid-transfer kinetics because the conjugation process is energy dependent.
To further examine the role of energy availability, we performed kinetic tests with different concentrations of the donor's energy substrate, glucose. The donor cells for all tests were harvested simultaneously and had the same internal energy stores. In parallel to the results with different antecedent growth rates, tests performed with higher glucose concentrations gave systematically larger kt, values (7).

Transfer of the TOL Plasmid
We performed much more extensive work with the TOL plasmid, which codes for the degradation pathway of toluene. The donor was Pseudomonas putida PAWI (TOL), and a restriction-deficient recipient, Pseudomonas aeruginosa PAO 1162, was employed. Transconjugants were assayed routinely by their ability to grow on toluic acid, and hybridization of plasmid preps with oligonucleotide probes was used periodically to confirm the presence of the TOL plasmid (5).
The kinetic experiments with TOL were conducted in the same manner as described for RP4. As with RP4, we found that the donor's antecedent growth rate and substrate concentration during the kinetic experiment affected kt, dramatically. BF Smets (5) quantified the two effects with Equation 2: kt 0.021 exp{0.55 }14 (2.5/S)0.45 [2] in which kt, has the units gT 1 (gD gR day)-'; p is the donor's antecedent-specific growth rate (day-1); and S is the donor's substrate concentration (g/l) during the kinetic test. Equation 2 shows that increases in p and S increase kt,, and the (gD g, day)-', a modest value. Application to Treatment Processes Although k,, values can be quite large, the conditions normally found in biological treatment processes are quite the opposite of the conditions giving rise to high transfer rates. In order to achieve treatment goals and process stability, specific growth rates are very low (usually 0.2/day or smaller) and substrate concentrations small (typically only a few milligrams per liter). Thus, the normal conditions are much closer to those giving kt, = 0.001 gr 1 (gD gR day)-' than those giving 2.5 gT 1 (gD gR day)-'. Table 2 summarizes p, S, and kt, values for a range of typical steady-state process conditions (5). The kt, values probably are too small to sustain the plasmid in a significant fraction of the population (4,5) without prohibitively high costs for adding the donor cells.
While it is not feasible to operate biological treatment processes with uniformly high ji and S, we propose that the benefits of high plasmid-transfer rates may be "Parameters for generating these results taken from Smets (5).
Environmental Health Perspectives obtained by creating zones of high S and p within a treatment process that has low p and S values overall. This concept is shown schematically in Figure 2, which contrasts the normal format of the activated-sludge process with a modified process that creates a high-S zone in reactor 2. The modified process adds a small preliminary tank (denoted reactor 2) that receives the influent substrate and maintains a high substrate concentration. Table 3 shows that the substrate concentration in reactor 2 can be significantly augmented when reactor 2 is small enough, but the effluent substrate concentration remains low because of the long contact time in reactor 1.  Figure 2. Schematic comparison of a typical activated-sludge process (A) with a modified process (B) that has a preliminary tank (reactor 2) for creating a high-Szone.
The trends shown in Table 3 suggest that plasmid-transfer rates can be increased by creating a high-S zone. However, the practical impacts have not been assessed theoretically or experimentally. Further work should address whether the faster plasmid-transfer rates in the small high-S zone are significant enough to offset the low S and low Ii conditions in reactor 1.

Summary
Our research has demonstrated that plasmid transfer occurs at relatively fast rates among bacterial strains relevant to biological treatment. Most importantly, the rate coefficients for conjugation are not constants, but depend strongly on the donor cells' specific growth rate and substrate concentration, both of which affect the cells' energy availability. The relationship for how kt, increases for increasing it and S was quantified for the transfer of the TOL plasmid between two Pseudomonas strains (Equation 2) and the phenomena were qualitatively the same for transfer of RP4 from Rhodabacter capsulatus to Pseudomonas sp.
Although transfer rates can be high when p or S is large, normal biological processes have low values of p and S. Therefore, we propose that zones of high S or p be created within processes whose overall p and S values remain low. One simple scheme-a modification to the activated-sludge process-was presented as an example.