Wholecell Bioconversions Of Organic Compounds In Twoliquidphase Media With Solventresistant Organisms

3.2.1 Microbial Growth in Solvents and Emulsions

It has been known for at least a century that some microorganisms are capable of growing in the presence of organic solvents, and in some cases they can metabolize solvents as sources of energy and carbon (Fairhall 1920; Hopkins and Chibnall 1932). In nature, such organisms play a role in the biodegradation of aliphatic and aromatic components in oil. They have been isolated from natural oil deposits or soil, from human-made oil wells, from oil spills, and in general from oil-rich locations, such as machine shops or factories (Johnson 1967).

There is considerable variation in solvent resistance among microorganisms. Earlier work in our laboratory (Rajagopal 1996) has shown that it is possible to determine a unique and reproducible logP50 for a given microbial strain. This logP50 defines the solvent polarity (logP) that allows 50% survival of the microbes.f Cells survive fully in the presence of solvents with logP only slightly higher than logP50, while they die in the presence of solvents with a logP that is slightly lower than logP50. Thus, wild-type Escherichia coli W3110 — a K12 strain (Bachmann 1972) — can grow perfectly well on glucose in the presence of large volumes of long-and medium-chain-length alkanes down to n-octane (logP = 4.5), cyclooctane (logP = 4.1), and even n-heptane (logP = 4.0), but it cannot grow in the presence of 1-octene (logP = 3.7), n-hexane (logP = 3.5), cyclohexane (logP = 3.2), and other solvents with logP < 3.7 (Favre-Bulle et al. 1991).

An interesting example of the variability of biological responses to solvents is that provided by toluene. Toluene has been used extensively since the 1950s as a tool in rendering cells permeable to small molecules, thereby permitting enzyme assays of whole cells (Herzenberg 1959), and toluene has been quite useful in the study of multienzyme complexes that could not be isolated and studied in vitro. Early studies in our laboratory have shown that toluene extracts lipids from the cellular membranes (de Smet et al. 1978). Toluene-treated cells lose their intracel-lular metabolite pool in seconds, so that these cells are no longer able to metabolize external carbon and energy sources. Thus, despite the fact that most, if not all, intracellular enzymes remain active after toluene treatment, cells are killed by toluene because the membrane bilayer is impaired.

If cells could withstand toluene-engendered damage at the ultrastructural level, they might well survive. In fact, today many examples of such cell types are known. In 1989, Horikoshi and his coworkers first described a Pseudomonas isolate that is capable of growing in the presence of 50% (v/v) toluene, despite the fact that only 1% (v/v) toluene is lethal to other Pseudomonads (Inoue and Horikoshi 1983). Work is now underway to understand solvent resistance, and although it is not yet clear t The partition coefficient (Poct) for a given compound is equal to the ratio of the solubility of the compound in octanol (or some other standard solvent, to be denoted) to its solubility in water, under a specific set of standard conditions. The resulting ratio is usually expressed as the logarithm of the partition coefficient, denoted as logPoct or simply logP.

which factors account for the differences in toluene resistance between Horikoshi's solvent-resistant strain and most other solvent-sensitive microorganisms, likely factors include alterations in membrane and cell-wall composition and architecture, alterations in lipid synthesis and degradation, as well as active export of toluene by solvent pumps (Godoy et al. 2004; Isken and de Bont 1996, 1998; Kieboom et al. 1998a, 1998b; Ramos et al. 2002; Segura et al. 2003; Weber and de Bont 1996).

3.2.2 Whole-Cell Biocatalysis in Two-Liquid-Phase Media

The practical application of biooxidation reactions requires that mono- and dioxy-genases operate in functional and metabolically active cells (de Smet et al. 1981, 1983b; Held et al. 1998, 1999; Panke et al. 1999; Schmid et al. 1998b; Staijen and Witholt 1998). This is so because of the requirement of these enzymes for cofactors and cofactor regeneration. The nature of the two-liquid-phase system in which these cells must function will be dictated by the substrates used and the products formed; the conditions necessary for effective downstream processing; and the overall economics of the process as it emerges from the research and pilot development stages (Mathys et al. 1999; Panke et al. 1999). These requirements may determine the nature and quality of the solvent to be used as the apolar phase (the substrate may or may not be the solvent as well); the culture conditions; and in some cases, the state of recirculated separate phases, one or both of which may return from the downstream processing system. These various parameters in turn impose limitations on the host strain and the specific oxidation enzyme system to be used.

One particularly interesting strain was first isolated in the late 1930s by Lee and Chandler, then at the Rice Institute in Houston (Lee 1941). They examined water-oil emulsions at the Hughes Tool Co., also in Houston. Large amounts of emulsion were pumped over tools and metal objects during machining (milling and lathing) operations in company machine shops. The emulsion flowed over pieces as they were machined and then into collecting basins under the metalworking equipment. Workers stood in these basins and sometimes developed pustules and skin infections on their hands, which led to the notion that the emulsion might contain pathogenic microorganisms. Lee and Chandler set out to find and identify these organisms and found one single predominating organism that had all of the characteristics of a pseudomonad. They named it Pseudomonas oleovorans, an organism that has been worked with intensively since.

Significant interest in the biocatalytic potential of this and similar organisms began to develop in the early 1960s, when laboratory studies by van Linden and coworkers at the Shell laboratories in Amsterdam showed that certain pseudomonads are capable of oxidizing and utilizing alkanes for growth (van der Linden 1965). Subsequently, Coon and his coworkers at Michigan State University isolated an enzyme system responsible for the terminal oxidation of n-alkanes and fatty acids from Pseudomonas oleovorans in the 1960s and 1970s (Peterson et al. 1966; Ruettinger et al. 1977). In this same period, a growing group of researchers began to study the P450 monooxygenases from eukaryotic and bacterial cells. These monooxygenases oxidize a wide range of organic compounds, among them alkanes and complex cyclic compounds, including aromatic ring systems and steroids. Several hundred P450 monooxygenases have been described to date (Lewis 2003).

3.2.3 Two-Liquid-Phase Whole-Cell Biocatalysis Reactor

In the biocatalysis reactor, shown schematically in Figure 3.2, the emulsion is stirred sufficiently well that the apolar phase forms microdroplets with diameters on the order of 5 to 15 pm, as shown in Figure 3.3. As a result there is a very large surface layer between the aqueous phase and the apolar droplet phase. This permits substrates dissolved in the apolar droplets to exchange rapidly with the continuous aqueous phase, which contains the microorganisms that will take up the substrate and convert it to product, which will again migrate to the apolar droplets. Although the exchange of substrates and products is rapid, their concentration in the apolar and aqueous phases is determined by the partitioning of these compounds between the two phases. Many of the substrates of interest in bioreactions have logPs in the range of 3 to 6, which means that the solubility of these compounds in water is 10-3 to 10-6 lower than that in octanol or many other apolar organic solvents. .

Consequently, although substrates and products exchange rapidly between the two phases in a two-liquid-phase bioreaction system, the concentration of these

Two liquid phase biooxidation of styrene

Two liquid phase biooxidation of styrene

Biooxidation

FIGURE 3.2 Bioconversion in a two-liquid-phase culture. A recombinant E. coli JM101 host, equipped with a styrene-oxidation system on plasmid pSPZ3, converts styrene to styrene oxide. Both the substrate (styrene) and the product (styrene oxide) dissolve in the organic phase (golden brown), which in this experiment is hexadecane. A very small amount of styrene partitions into the aqueous phase (light blue) and enters the recombinant cells, where it is oxidized to styrene oxide. This product then partitions back into the apolar phase because it is much more soluble in hexadecane than in the aqueous phase. As a result, the cells are exposed to a very low concentration of the toxic substrate and product, and the cells remain biocatalytically active for 10 to 20 h. (Experiments reported by Sven Panke and Marcel Wubbolts, Institute of Biotechnology, ETH, Zurich, 1996.)

FIGURE 3.2 Bioconversion in a two-liquid-phase culture. A recombinant E. coli JM101 host, equipped with a styrene-oxidation system on plasmid pSPZ3, converts styrene to styrene oxide. Both the substrate (styrene) and the product (styrene oxide) dissolve in the organic phase (golden brown), which in this experiment is hexadecane. A very small amount of styrene partitions into the aqueous phase (light blue) and enters the recombinant cells, where it is oxidized to styrene oxide. This product then partitions back into the apolar phase because it is much more soluble in hexadecane than in the aqueous phase. As a result, the cells are exposed to a very low concentration of the toxic substrate and product, and the cells remain biocatalytically active for 10 to 20 h. (Experiments reported by Sven Panke and Marcel Wubbolts, Institute of Biotechnology, ETH, Zurich, 1996.)

Bioconversions

FIGURE 3.3 Light microscopy of Pseudomonas oleovorans growing in a two-liquid-phase culture, such as shown in Figure 3.2. The aqueous phase is the continuous phase and contains the rod-shaped bacteria. The organic phase consists of droplets that have a diameter of 5 to 15 |i,m. (This and similar photos were made by Marie-Jose de Smet [University of Groningen, 1980] and Andrew Schmid [Institute of Biotechnology, ETH, Zurich, 1995]).

FIGURE 3.3 Light microscopy of Pseudomonas oleovorans growing in a two-liquid-phase culture, such as shown in Figure 3.2. The aqueous phase is the continuous phase and contains the rod-shaped bacteria. The organic phase consists of droplets that have a diameter of 5 to 15 |i,m. (This and similar photos were made by Marie-Jose de Smet [University of Groningen, 1980] and Andrew Schmid [Institute of Biotechnology, ETH, Zurich, 1995]).

compounds in the aqueous phase is quite low. This means that even when such compounds are toxic for microorganisms, this toxicity is generally not a problem, because the effective concentrations of either substrates or products in the aqueous phase remain quite low. Thus, two-liquid-phase bioreaction systems are likely to play a significant role in future developments of white biotechnology.

With the basic problem of adapting biocatalytic processes to classical organic chemistry largely solved, there are several other requirements that must be met for all biocatalytic and chemocatalytic processes.

First, it is necessary to find and develop catalysts that carry out the desired reaction with the needed chemo-, regio-, and stereospecificities. This is dealt with via a variety of screening and selection methods, and once a promising biocatalyst is identified, attempts are made to improve it further via site-directed mutagenesis, DNA shuffling techniques, directed evolution, or in silico protein design techniques.

Second, high catalytic activities per unit bioreactor volume must be attained and maintained for long periods of time. One approach to meeting these requirements for two-liquid-phase biocatalytic processes is to maximize cellular enzyme activities by high-level expression of active enzymes. In addition, we attempt to grow cells to high densities in the aqueous phase of two-liquid-phase media.

3.2.4 Organisms Suitable as Hosts for Byconversions in Two-Liquid-Phase Media

Although organic solvent with low logPoct (<3.5) are generally lethal to bacteria (Rajagopal 1996), there are many organisms, including E. coli, that function reasonably well in the presence of organic solvents with intermediate logPoct (3.5 to 5.0) (Favre-Bulle et al. 1991). As work in this area develops in the next decade, it is likely that a few especially suitable species will emerge, which will gradually be improved to become standard two-liquid-phase bioconversion hosts. Such hosts might be endowed not only with good survival characteristics in a range of two-liquid-phase environments, but they are likely also to contain adequate cofactor production and regeneration systems (Wubbolts et al. 1990) and properties that favor high oxygen-transfer rates (Khosla et al. 1990; Magnolo et al. 1991) and facilitate downstream processing.

We have also developed continuous cultures in two-liquid-phase bioreactors, again for both P. oleovorans (Durner et al. 2000; Preusting 1993a) and E. coli recombinants. These cultures exhibit stable alkane oxidation activity for periods of a week to a month. Cell densities are lower (1 to 10 g dry mass per l), and cellular activities are generally higher than those found for the fed-batch system. For instance, for the production of polyhydroxyalkanoates (PHAs) with continuous growth on n-octane, a cell density of 10 to 12 g dry mass/l could be maintained for at least one month, with cellular PHA contents of 20 to 25% (m/m) (Favre-Bulle et al. 1993; Preusting 1993a).

3.2.5 Space-Time Yields and Limitations

The maximum productivity attainable is limited by the rate of oxygen transfer to cells under practical conditions. We usually try to improve the performance of the biocatalytic system through modifications of the biocatalysis enzyme-host combination and through technical efforts to maximize cell densities and performance in continuous and fed-batch cultures, while at the same time minimizing growth rates, since growth requires continuous input of nutrients. See Mathys et al. (1999) for a detailed analysis of large-scale bioconversion and downstream processing systems. In addition, it is necessary to channel the largest possible fraction of the substrate into product rather than cell-mass production, which we do by providing an aqueous-medium-based carbon and energy source, such as glucose or glycerol.

With fed-batch cultures, we have been able to reach cell densities of 40 g dry mass per l, both for P. oleovorans (Preusting 1993b) and for E. coli recombinants (Wubbolts 1996). Unfortunately, it is not trivial to retain high enzyme activities when cell densities are increased. Enzyme activities (U/g cell dry mass) typically decrease when cells are grown to high densities. In practice, this means that an optimum combination of cell density and enzyme activity must be sought, resulting in an overall volumetric activity (units of enzyme activity per liter of bioreactor medium). Typical maximum activities seen in the biooxidation systems we have worked with a range from 200 to 500 units per liter of aqueous medium component for 5 to 15 h (Buhler et al. 2003b; Panke et al. 2002). If such systems can be scaled up to industrial production with equivalent yearly activities for 40% of the time, this will result in space-time yields between 5 and 15 tons/m3 (aqueous volume)/yr.

3.2.6 Downstream Processing

The above activities result in two-liquid-phase bioconversion systems of varying efficacy. The next step has been to develop downstream processes to separate products from substrates and possible other (contaminating) compounds present in the fermentation or enzyme medium. Such processes have been developed to function either in batch mode or in series with a continuous culture (Mathys et al. 1999). They consist of several steps, beginning with a phase separation of the complete culture medium.

Figure 3.4 shows an example of a continuous two-liquid-phase culture of P. oleovorans (left panel), which is gradually collected in a large flask (right panel), where there is a slow phase separation into an aqueous (lower) phase and an organic (upper) phase that still contains aqueous medium, forming a mousse that must be further separated into an organic and an aqueous phase. Following this second separation, the product is removed from the cell-free organic phase by fractional distillation (Mathys et al. 1998a, 1998b) or large-scale high-performance liquid chromatography (HPLC). The remaining organic phase might consist of the substrate (Favre-Bulle and Witholt 1992), or it might consist of a carrier solvent such as dodecane (Wubbolts et al. 1996), hexadecene (Schmid et al. 1998a), or bis(2-ethylhexyl)phthalate (Panke et al. 2002).

Preusting Continual Fermentation

FIGURE 3.4 Two-liquid-phase continuous culture of Pseudomonas oleovorans, consisting of 15% (v/v) octane and 85% (v/v) aqueous medium containing ca. 10 g (cdm )/l cells (left panel). The culture is stirred at 1500 rpm, which results in an emulsion of small octane droplets (see Figure 3.3). The culture is harvested continuously (right panel). The aqueous phase with bacteria sinks to the bottom. The octane phase with some trapped water floats to the top and forms a mousse. Most of the water can be separated from the apolar octane phase by centrifugation, and the apolar phase can then be subjected to distillation to separate biocatalysis products from octane. (Photos supplied by Wil Hazenberg and Andrew Schmid, Institute of Biotechnology, ETH, Zurich, 1995.)

FIGURE 3.4 Two-liquid-phase continuous culture of Pseudomonas oleovorans, consisting of 15% (v/v) octane and 85% (v/v) aqueous medium containing ca. 10 g (cdm )/l cells (left panel). The culture is stirred at 1500 rpm, which results in an emulsion of small octane droplets (see Figure 3.3). The culture is harvested continuously (right panel). The aqueous phase with bacteria sinks to the bottom. The octane phase with some trapped water floats to the top and forms a mousse. Most of the water can be separated from the apolar octane phase by centrifugation, and the apolar phase can then be subjected to distillation to separate biocatalysis products from octane. (Photos supplied by Wil Hazenberg and Andrew Schmid, Institute of Biotechnology, ETH, Zurich, 1995.)

For industrial cost optimization, the remaining organic phase (either substrate or carrier or both, and generally the major fraction of the initially used organic phase) will be returned to the continuous reactor (Mathys et al. 1999). It may be necessary to remove inhibitory compounds that affect the bioconversion efficiency or other undesirable contaminants that may cause problems in repeated cycles through the complete bioconversion-downstream processing system.

In some cases, the bioconversion of organic compounds results in water-soluble compounds such as acids (Favre-Bulle and Witholt 1992), which need to be processed via precipitation, crystallization, filtration, extraction, chromatographic techniques, or other classical methods.

3.2.7 Integrated Bioconversion-Bioprocessing System

Each of the above topics provides ingredients for an integrated bioconversion system. Optimization of the entire biosynthesis and processing chain requires that all of the components in this complex process be sufficiently understood so that the activity of each component can be modulated with respect to that of all other steps such that the overall process is optimized. As an example, maximizing product accumulation in the two-liquid-phase bioreactor might appear desirable, since it facilitates downstream processing. However, some products, e.g., alkanoic acids, which accumulate in the water phase (Favre-Bulle et al. 1993), inhibit cell growth and enzyme activity. It might be preferable therefore to aim for a high substrate conversion rate, with a rather low steady-state product accumulation level. This has advantages for the volumetric productivity in the reactor, but now it becomes more difficult to remove the product from the reactor efflux. Here, it is useful to consider the product removal process. If the efflux stream is recycled after product removal, the extent of the removal need not be complete. It should be very specific, in the sense that the product removed meets the required purity criteria, but whatever is not removed from the efflux stream is returned to the reactor, where a steady state is maintained via (a) the continuous removal of an efflux stream containing the product and (b) the continuous influx of the remaining product via the return stream from the downstream processing unit plus the newly synthesized product in the bioreactor (Mathys et al. 1999). In the steady state, there is material (liquid phases, product in one of these phases, cell mass) cycling in the bioreactor-downstream processing unit, resulting in a net steady-state concentration of product in the bioreactor, and in a differential across the downstream processing unit. Optimizing the complete system requires finding the steady-state conditions that yield maximum overall productivities at required product purity criteria.

3.3 SPECIFIC COMPOUNDS OF INTEREST 3.3.1 Synthesis of Poly-3-Hydroxyalkanoates (PHAs)

Was this article helpful?

0 0

Post a comment