Optimization of Pseudomonas putida KT2440 as host for the production of cis, cis-muconate from benzoate

Optimization of Pseudomonas putida KT2440 as host for the production of cis, cis-muconate

from benzoate P. putida KT2440 was used as biocatalyst given its versatile and energetically robust metabolism.

Therefore, a mutant was generated and a process developed based on which a life cycle assessment

(LCA) was performed. Additionally, the growth related parameters were experimentally obtained to

constrain the metabolic model iJP815 further.

The mutant Pseudomonas putida KT2440-JD1 was derived from P. putida KT2440 after NTGmutagenesis

and exposure to 3-fluorobenzoate. The strain was no longer able to grow with benzoate

as a single source of carbon and energy. Instead, benzoate was co-metabolized to cis, cis-muconate

that accumulated in the culture medium while the strain grew on glucose. In batch cultures, a

maximal production rate per gram biomass of 2.0 g cis, cis-muconate gDCW

-1 h-1 was obtained. This is 8-fold higher than thus far reported. The cat operon was no longer expressed in P. putida KT2440-JD1 due to a point mutation in the regulator gene catR. This operon contains the genes for the conversion of catechol to metabolites of the central metabolism, including catA, which encodes a catechol 1,2-dioxygenase. Consequently, benzoate is converted in the mutant by enzymes that are encoded on the ben operon. This operon includes a gene (catA2; PP_3166) that encodes an additional catechol 1,2-dioxygenase, thus allowing the conversion of benzoate to cis, cis-muconate.

In batch cultures the maximal growth rate of P. putida KT2440-JD1 in mineral medium with

glucose decreased linearly in the presence of increasing concentrations of benzoate and/or cis, cismuconate

and finally stopped at 6 g L-1 benzoate or 85 g L-1 cis, cis-muconate. The inhibitory

effects of both compounds were cumulative and no synergistic effects were observed. The maximal

uptake rate of benzoate was higher than the production rate of cis, cis-muconate per gram biomass

during growth on glucose in the presence of benzoate, indicating that a benzoate derivative

accumulated in the cells, which is likely to be catechol, as accumulation of this intermediate was

observed. Catechol is known to cause oxidative stress, and the accumulation of catechol and

benzoate should be prevented during the production of cis, cis-muconate. This is feasible by

coupling the addition of benzoate to the decrease of the pH of the culture medium in a so-called pHstat

process, as the cultivation medium acidifies only when benzoate has been converted to cis, cismuconate.

Such a pH-stat fed-batch process resulted in the production of 18.5 g L-1 cis, cismuconate

from benzoate with a molar yield of 96%.

The phenotype of P. putida KT2440 can be assessed by the constraint-based metabolic model

iJP815. The solution space of the model based on flux variability analysis was further constrained

by growth associated maintenance, non-growth associated maintenance, and biomass composition

determined form the experimentally measured growth-related factors of the strain that were

generated during continuous cultivations at various dilution rates (D) (0.05-0.49) h-1.

Transcriptomic profiles obtained at a D of 0.2 h-1 were consistent with model predictions of

expressed genes based on flux balance analysis. The growth-related macro molecular composition

of the biomass was similar as measured with E. coli K-12. However, growth parameters like the

maximum biomass yield and maintenance coefficient were different. The energy required for

assembly of P. putida KT2440 is higher compared to E. coli W3110, which will result in higher

costs of glucose in biotechnological processes. On the other hand, the metabolism is robust as even

at a D of 0.49 h-1 no overflow metabolism was observed. The lack of overflow metabolism is of

great importance as it underscores the capacity of P. putida KT2440 as biocatalyst e.g. for

conversions involving cofactor dependent oxygenase reactions. Further restriction of enzymatic

conversions with already published 13C measurements only decreased the maximum solution space

excluding solutions that are far from the experimental phenotype. Possibly, the best agreement was

reached within limitations of constraint-based modeling.

A LCA was performed on a combined biological and chemical process for the production of adipic

acid. The outcome was compared with the traditional chemical process. The LCA focused on the

cumulative energy demand, cumulative exergy demand and the CO2 equivalent emissions, with CO2

and N2O separate. Acidified cis, cis-muconate can be easily hydrogenated to adipic acid a resource

for nylon-6,6 used in carpets and the auto industry, because of its long lasting and strong features.

Feedstocks have a large effect on the overall environmental impact. The soil bacterium P. putida

KT2440 has a versatile metabolism and is able to convert various feedstocks to cis, cis-muconate.

Consequently, the use of feedstocks with a lower energy demand were taken into account besides

benzoate, including: the petrochemical based feedstocks impure aromatics and toluene, and the

biomass based feedstock phenol (from lignin). The effect of an increase of the final concentration

cis, cis-muconate in the fermentor broth from 1.85% to 4.26% was modeled as P. putida KT2440-

JD1 was able to consume benzoate up to 4.26% cis, cis-muconate. At a final concentration of 1.85%

cis, cis-muconate, the use of impure aromatics and lignin instead of benzoate reduced the energy

demand compared to the chemical production of adipic acid. The applicability of these feedstocks

depends on the metabolic robustness of P. putida KT2440-JD1 with impure aromatics, and/or the

development of an efficient process for the production of phenol from lignin. At a final

concentration of 4.26% cis, cis-muconate the process energy and CO2eq emissions were reduced for

all feedstocks.

The pH-stat fed-batch process had a higher production rate per gram biomass compared to other

processes although the speed was significantly lower as measured during the batch cultures. By

generating a high cell density in combination of a pH-regulated inflow of benzoate, the volumetric

productivity will increase as has been described for a cell-recycle process. While this only has a

minor impact on the energy consumption, it does have a major impact on the economic parameters

of an industrial scale bio-reactor. Recycling biomass with a membrane and using a solvent, as e.g.

cold diethyl ether, after the acidification of the medium (pH 2.5) for the extraction of cis, cismuconate,

could further reduce the environmental impact. The use of solvents will result in a higher

product concentration and/or the evaporation will require less energy since the boiling temperature

of solvents is most often lower than water.