The identification of Corynebacterium glutamicum as a glutamate producer in the

1950’s was the start of its career as an amino acid producer. C. glutamicum has now been employed

as cell factory for industrial amino acid production for over five decades and has a market size to

reach $20 billion by 2020. As C. glutamicum was isolated for its natural ability to produce

glutamate it makes it an excellent chassis for engineering it to produce its derivatives ornithine,

proline, putrescine, citrulline, and arginine. These products are becoming increasingly important as

they have a wide range of applications and the demand for these products is rising. The construction

of a platform strain can reduce the time and resources required for strain development. Another

approach to reduce the time and resources required for strain development is by developing new

tools for metabolic engineering. These strategies were both explored to ease the process of strain

construction.

Proline, citrulline, and putrescine can be synthesized directly from ornithine, and three more

reactions are required to synthesize arginine. Hence a strain engineered to produce ornithine can

serve as a platform to produce the other four compounds. The first step was to establish proline and

citrulline production by C. glutamicum as the rational engineering of C. glutamicum for production

of these two amino acids had not previously been described.

Proline is synthesized from glutamate via the proline biosynthetic pathway in most microorganisms,

but with the help of the enzyme ornithine cyclodeaminase (Ocd) it can be synthesized from

ornithine. Overexpression of putative ocd from C. glutamicum in an ornithine overproducing strain

did not result in detectable proline production. Plasmid-based expression of ocd from Pseudomonas

putida (ocdPp) on the other hand allowed accumulation of proline with a yield of 0.06 g proline / g

glucose (g pro/g glc). Interestingly replacing the stop codon TGA of ocdPp with TAA resulted in a

remarkable decrease in glutamate accumulation while the yield of proline increased to 0.25 g pro/g

glc. The byproduct accumulation was further reduced by medium optimization, which also entailed

a 25% higher proline yield. Lastly, the yield was boosted by overexpression of argBfbr encoding

feedback-alleviated N-acetylglutamate kinase to 0.36 ± 0.01 g pro/g glc.

A base strain suitable for the production of ornithine, proline, putrescine, citrulline, and arginine

was constructed by deleting the repressor of the arginine pathway ArgR. Additionally the genes

encoding enzymes of the arginine biosynthetic pathway were deleted, the deletion of argF allows

ornithine to accumulate, and the deletion of argG allows citrulline to accumulate when argF is

overexpressed. To add citrulline to the product spectrum of the platform strain we found it

necessary to overexpress argBfbr along with argF. The strain accumulated citrulline with a yield of

0.38 ± 0.01 g citrulline / g glucose. The potential and versatility of the strain was demonstrated by

producing citrulline from the alternative carbon sources starch, xylose, and glucosamine.

After the product range of the base strain had been extended to include citrulline, metabolic

engineering was performed to increase the ornithine yield of the strain. The initial assumption was

that an increased ornithine yield could be translated into increased yields of the other four bioproducts. The ornithine yield was improved by 71% with a yield of 0.52 g ornithine / g glucose

(g orn/g glc) compared to 0.31 g orn/g glc of the parent strain. This was achieved by feedback

alleviation of N-acetylglutamate kinase, tuning of the promoter of gdh encoding glutamate

dehydrogenase, lowering expression of pgi encoding phosphoglucoisomerase, along with the

introduction of a second copy of the arginine biosynthetic operon argCJBfbrD into the chromosome.

Strains capable of efficiently producing citrulline, proline, arginine or putrescine were derived from

ornithine producing strains by plasmid-based overexpression of appropriate pathway modules with

one to three genes. It was found that optimizing the base strain for ornithine production did not

increase citrulline and arginine yields any further, indicating that the reaction converting ornithine

into to citrulline is a bottleneck in citrulline and arginine production.

Finally the popular CRISPR/dCas9 technology was adapted for the use of metabolic engineering in

C. glutamicum. The system could be used to reversibly perturb gene expression. As proof of

concept we targeted pgi in the lysine overproducing strain DM1729, where an almost complete

repression of the transcription of the gene resulted in a 2-fold increase in the lysine titer. We also

targeted the genes pck and pyk to increase glutamate production in wild-type C. glutamicum where

we also observed nearly complete repression and increased glutamate titers.

With this work it was shown how valuable the concept of using a platform strain for production of

several industrially relevant bioproducts is. Moreover an efficient way to screen for new targets for

metabolic engineering in C. glutamicum was demonstrated. From the initial cloning in Escherichia

coli the C. glutamicum clones could be obtained in as little as four days by adapting the

CRISPR/dCas9 system.