Bleached Kraft pulp mill effluents (BKME) are complex wastes containing organic matter and significant amounts of chlorinated compounds generated during chlorine bleaching of cellulose to obtain white paper. Among BKME chlorinated compounds, chlorinated aromatics like chloroguaiacols, chlorobenzoates and chlorophenols are more important because they are resistant to degradation and toxic. BKME are usually treated in aerobic systems where microbial activity substantially decreases the organic matter content, but it is less effective against chlorinated material.

Chloroaromatic compounds are difficult to degrade by bacteria because, in most cases, a specific intermediate is the bottleneck in the metabolic pathway due to impaired enzyme activity and/or toxic effects.  However, there are several catabolic strains that harbor specialized gene clusters that enable them to efficiently degrade chlorinated compounds. One of these microorganisms is the soil bacterium Ralstonia eutropha JMP134 (pJP4), able to grow in 2,4-dichlorophenoxyacetic acid (2,4-D), and 3-chlorobenzoate (3-CB), among other compounds. Most of its key catabolic abilities are encoded in the gene cluster tfd_ICDEF that allow conversion of chlorocatechols, a typical toxic intermediate, into ß-ketoadipate. The latter compound is channelled into the intermediate metabolism, and therefore, allowing the bacterium to use chlorinated pollutants as carbon sources.

In this work, we hypothesized that introduction by genetic engineering of the specialized gene cluster tfd_ICDEF into bacteria that accumulate chlorocatechols, and therefore become intoxicated, would allow improvement of their catabolic performance, for example in BKME microbial treatment systems. We chose three bacterial strains to test this idea, two of them, R. eutropha JMP222 lacking tfd genes, and P. putida KT2442 are well-characterized soil microorganisms that accumulate chlorocatechols during metabolism of chlorobenzoate. The third microorganism was first isolated from BKME and then metabolically characterized. It turned to be a new Acinetobacter lwoffii strain RB2, able to grow on guaiacol, and more importantly, accumulating chlorocatechols from 4- or 5-chloroguaiacol.

The tfd_ICDEF gene cluster was first cloned into a specialized minTn5-based delivery system which allows integration of the tfd cluster into the chromosome of the recipient strains. These minTn5-derived systems also allow the expression of the tfd-encoded functions under the control of a heterologous regulatory system induced by isopropyl-ß-D-thiogalactopyranoside (IPTG) or and homologous system that responds to 3-CB.

Several derivatives of each strain were obtained, and tested positive for the presence in the chromosome of tfd genes. They gave also positive results for the expression of the Tfd-encoded enzymes. However, the levels of enzyme activity were always lower (15-80%) than those found in the wild type strain JMP134. The dechlorination rates, that represent chloroaromatic degradation after release of chloride from the molecule, were in agreement with the enzyme tests. The low level of expression caused that derivatives of R. eutropha JMP222, P. putida KT2442, or A. lwoffii RB2, harboring chromosomal insertions of the tfd_ICDEF gene cluster did not grow on 3-CB or chloroguaiacols. A position effect of the introduced tfd genes was discarded because about hundred derivatives of each type were screened, and showed similar behaviour.

These observations, along with other results obtained in our laboratory strongly suggest that the lower level of tfd_ICDEF expression is due to a copy number effect. It should be mentioned that tfd genes in strain JMP134 are in about 5 copies per genome. In the case of A. lwoffii, it is also possible that inadequate gene regulation plays also a role.

Despite of these results, some derivatives were tested for catabolic performance of the tfd-encoded functions in microcosm systems that imitate the microbial treatment of BKME. In these microcosms, the catabolic performance of the engineered strains was also lower than that observed for the wild-type strain R. eutropha JMP134.

These results strongly indicate that very efficient catabolic strains should be selected for biological removal of chloroaromatics in BKME treatment systems.