Biocatalytical synthesis of aliphatic nitriles from aliphatic aldehydes via aldoximes

The results of this chapter were published by the author of this thesis and her coauthors in Journal of Organic Chemistry (Article 3)^{[ 6 4 ]} and were summarized in a book chapter in Applied Biocatalysis: The Chemist's Enzyme Toolkit (Article 4).^{[ 5 7 ]}

The next important intermediates in the cascade to aliphatic nitriles are the aldoximes, which can be obtained very easily by condensation of the aldehydes obtained from alcohols by nitroxyl radical-catalyzed oxidation, and hydroxylamine. For safety reasons hydroxylamine is formed in situ from hydroxylamine hydrochloride in the presence of sodium carbonate.

Using this method in aqueous reaction medium and subsequent filtration of the products, aliphatic aldoximes were obtained in a very pure form in isolated yields of 63 -96%. After successful synthesis of the aldoximes, the next step is the biocatalytic dehydration using Oxds to the nitriles. Aldoxime dehydratases were already found to be active for aliphatic aldoximes,^[30] but high substrate loadings which are required in industrial processes for bulk chemical syntheses were not tested so far. The process optimization towards

Aliphatic nitrile and amine syntheses starting from biorenewables

a highly productive nitrile synthesis using Oxds were performed in this thesis. Initial studies showed that aldoximes of chain length >12 were not or only in small extent converted by aldoxime dehydratases to the corresponding nitriles. Furthermore, OxdB was found to be the most active enzyme for the conversion of aliphatic aldoximes, wherefore subsequent experiments were conducted exclusively with this enzyme. Since aldoxime dehydratases were found to be relatively unstable,^[30] biotransformations were performed with OxdB in whole cells as catalyst.

Figure 4. Screening of conversion of various aliphatic aldoximes into the corresponding nitriles using the enzyme OxdB in whole cells on a 0.5 mL scale. This figure was taken from Article 3.^[64]

After an initial screening using different substrates and substrate loadings in small scale (Figure 4), 10 mL-scale experiments with isolation of the nitrile products were performed with C6-, C8- and C10-aldoximes. Very high conversions and yields were obtained with substrate loadings of up to 1.4 kg substrate per liter of aqueous reaction medium (Table 1). Besides the very high productivity of these biotransformations using OxdB in whole cells, also the product separation is very simple. After the reaction, an easy phase separation can be performed of the aqueous phase including OxdB whole cells and the nitrile product. Based on these results, an

up-scaling experiment for n-octanenitrile was performed by Sylvia Glinski, resulting in >182 g of pure n-octanenitrile from 250 mL aqueous reaction medium.^[62]

Table 1.Summary of 10 mL-scale biotransformation experiments using C6-, C8-, and C10-aldoximes.

# n Final substrate loading Conversion /% Isolated yield /%

1 1 288 g/L >99 81

2 3 665 g/L >99 98

3 3 1.4 kg/L 93 --

4 5 342 g/L >99 84

5 5 428 g/L 93 --

These results clearly show the potential of OxdB as whole cells catalyst for the dehydration of aliphatic aldoximes within in our planed cascade. In the following, the dehydration step was further optimized towards an easier implementation into Cascade 1 without the isolation of intermediates.

3.1.2.1 Immobilization of OxdB and OxdRE for the use in aqueous reaction media (Article 5) and in pure organic medium (Article 6)

The results of this chapter were submitted by the author of this thesis and her coauthors to Cataly sts (Article 5).^{[ 5 8 ]} The results of the superabsorber immobilization was published by the author and her coauthors in European Journal of Organic Chemistry (Article 6).^{[ 5 9 ]}

In Chapter 3.1.2 described and reported in Articles 4 and 5 Oxds can be used for the efficient synthesis of aliphatic nitriles from aldoximes. For a potential recycling of the biocatalyst or even simpler separation of the biocatalyst from the product, heterogenization of the biocatalyst would be a favorable option. For this purpose, different immobilization techniques for the immobilization of OxdB and OxdRE were tested using purified enzymes and whole cell catalysts. First, different immobilization carriers were investigated, which immobilize free

Aliphatic nitrile and amine syntheses starting from biorenewables

enzymes by hydrophobic interactions (three different were tested) or covalently (two epoxy carrier and one amino carrier were tested). It was found that the residual activity after immobilization with high immobilization efficiencies (~80-90%) were <20% in all cases. Since the immobilization of purified enzymes showed low activity in all cases the focus was further on the immobilization of whole cells. For this purpose, the encapsulation of whole cells into calcium alginate beads^[65–68] and absorbance into superabsorber^[69,70] were tested. Using an immobilization strategy in calcium alginate beads compared to calcium alginate beads coated with tetraethyl orthosilicate (TEOS), OxdB and OxdRE in whole cells were successfully immobilized with residual activities of up to ~70%. TEOS-coating changes the polarity of the beads' surfaces from very polar into non-polar, facilitating diffusion of the non-polar substrate molecules into the beads.^[65] Compared to the immobilization of purified enzymes, the residual activity of the whole cells entrapped in the alginate beads is remarkable. For both enzymes higher activities for the TEOS-coated beads in comparison to the uncoated beads were observed, which is probably due to a better diffusion of the substrate in the beads. It was found that the alginate beads, especially the TEOS-coated beads, are significantly more stable in ethanol-containing buffer with ~80% residual activity after 24 h incubation time in 10% ethanol of OxdB in TEOS-coated beads compared to ~20% for the free cells. Recycling of the beads led to decreased activity, however, the immobilizates can be used three times for 24 h reactions until a decrease of the activity to <85% is observed. An alternative immobilization-technique for OxdB in whole cells was published by the author in European Journal of Organic Chemistry.^[59] Whole cells in buffer are absorbed into polyacrylate (superabsorber), resulting in an immobilized aqueous phase. These immobilizates can then be used in pure organic medium, preferably in very non-polar organic solvents like cyclohexane, for the conversion of aldoximes to nitriles. 500 mM of n-octanaloxime were quantitatively converted to the corresponding nitrile within 24 h in pure cyclohexane as reaction medium. Since Oxds in non-immobilized whole cells do not show any activity in pure organic medium and almost no conversion is found, when using whole cells in a classical biphasic medium, a stabilizing effect of the superabsorber is expected. These remarkable results show that very labile Oxds can also be used in pure organic medium by a simple immobilization in superabsorber, which facilitates the combination of all steps into a cascade starting from alcohols to nitriles. All immobilization methods presented and discussed in this chapter are the first examples for heterogenization of Oxds and the results, especially of whole cell immobilization (Scheme 9), show the applicability of Oxd immobilizates in aqueous and organic reaction media. The latter, made possible by

immobilizing whole cells in superabsorber, opens up completely new possibilities for this class of enzymes, which previously could only be used in aqueous media.

Scheme 9. Whole cell immobilization of Oxd-catalysts in calcium alginate beads (green) and in superabsorber (blue), being usable in aqueous reaction medium or pure organic solvent.

3.1.2.2 OxdRE-mutagenesis study (Article 6)

A manuscript about OxdRE mutagenesis study was submitted by the author of this thesis and her coauthors to ChemBioChem (Article 7).^{[ 7 1 ]}

The low stability of aldoxime dehydratases, for example in the presence of high temperatures and organic solvents, is a well-known problem and is possibly a reason why this enzyme class is not used in industrial processes so far.^[30,59] Besides immobilization (Chapter 3.1.2.1), which can stabilize biocatalysts and makes them easier recyclable, enzymes can also be engineered towards higher stability. Whether or not recycling is possible successfully without loss of catalyst activity is mainly influenced by the stability of the catalyst. A mutagenesis study was conducted to increase the stability. Here OxdRE was used, because the X-ray structure of this enzyme is solved and thus the changes by directed evolution are better understandable. After implementation of a color-assay for the screening of a mutant library, error-prone PCR was conducted. Indeed, an OxdRE-variant with 10 mutations was found, which showed increased stability by treatment with acetonitrile and at higher temperatures. This result shows, that the stability of Oxds can be engineered. Since for this mutagenesis study OxdRE was used, which is compared to OxdB even more instable in contact with solvents and at high temperatures, this

Aliphatic nitrile and amine syntheses starting from biorenewables

OxdRE-variant was not used for the further development of the cascade. Future enzyme engineering for stable Oxds might be performed with OxdB instead of OxdRE to obtain a very stable OxdB-variant.