Experiments in loose cap shaking flasks

Shaking flasks with loose caps were used for larger culture volumes (1 L). As with the screw cap flasks, objective was to produce larger cell quantities for resting cells and crude extract experi-ments but especially larger quantities of the floating solids for characterization purposes. Similar to previous shaking flask experiments a fed-batch approach for the co-substrate was applied.

In these experiments, the solids previously observed were also present but as the incubation progressed (generally after 48 h), a complete separation from the liquid phase was no longer visible minutes after stopping agitation. Furthermore, small floating beads appeared after 24-36 h incubation which were clearly separated from the liquid phase or the previously observed float-ing solids. Figure 3.23 shows a loose cap shakfloat-ing flask culture (a) and a bead separated from the culture broth (b).

a) b)

Figure 3.23 - (a) Culture of mutant strain RMS5 growing in mineral medium supplemented with sodium pyruvate and n-eicosane at 12 h of incubation. (b) A bead separated from this culture after 60 h incubation (observed in co-substrate fed-batch experiments only). Experiment was conducted in a 2 L loose cap shaking flask during 60 hours.

Samples taken during incubation were also solvent extracted and analyzed by GC-MS. A similar peak profile to those previously observed was found in the GC-MS chromatograms at early stages of the experiment (generally until 24 h). However, as the experiment progressed and the de-scribed beads appeared, chromatograms showed additional peaks. As in the case of sealed serum vials or screw caps shaking flasks, no correlation was found between peak intensities and incuba-tion times, except for the decreasing alkane concentraincuba-tion.

Figure 3.24 (b) shows a chromatogram observed for the conversion of n-octadecane in a loose cap shaking flask after 48 h of incubation and it is compared to the chromatogram observed in a screw cap shaking flask after about 60 h of incubation (Figure 3.24 a).

a)

b)

c)

Peak Retention Time [min]

Substance

A 7.09 n-Octadecane

B 9.38 Hexadecanoic acid, trimethylsilyl ester C 11.13 Octadecanoic acid, trimethylsilyl ester D1 13.70 2-Monopalmitin, trimethylsilyl ester

D2 14.04 Hexadecanoic acid, 2,3-bis[(trimethylsilyl)oxy]propyl ester F 14.80 Octadecanedioic acid, bis(trimethylsilyl) ester

E1 15.50 2-Monostearin trimethylsilyl ether

E2 15.90 Octadecanoic acid, 2,3-bis[(trimethylsilyl)oxy]propyl ester

Figure 3.24 - Peaks detected from the silylated solvent extracts from samples taken during culti-vation of mutant strain RMS5 growing in mineral medium supplemented with 10 mM sodium pyruvate and n-octadecane: (a) After 60 h incubation in 1 L screw cap shaking flask; (b) after 48 h incubation in 2 L loose cap shaking flask; (c) peak identities determined by library match and later confirmed with reference substances. Peaks D3, E3 and E4 could not be identified.

Peak F in figure 3.24 (b) corresponded to 1,18-octadecanedioic acid, a bioconversion product of n-octadecane. This showed the ability of this mutant to produce α,ω-oxidized molecules from long-chain n-alkane. Further studies were performed with the strain to evaluate the bioconver-sion influencing factors and potential for the production of long-chain di-carboxylic acids. This is discussed in section 3.7.

Higher cell densities were observed in loose cap shaking flaks. In figure 3.25(a) growth curves of the wild-type strain, mutant RMS5 and a comparison to growth of this mutant in a screw cap flask are presented.

50 µm 0.0E+00

5.0E+08 1.0E+09 1.5E+09 2.0E+09 2.5E+09

0 20 40 60 80

Cell Density [Cells/mL]

Cultivation Time [h]

Thermus sp. ATN1 Mutant RMS5

Mutant RMS5 (Limited Oxygen)

Figure 3.25 - (a) Growth of Thermus sp. ATN1 wild-type strain and mutant RMS5 in mineral medi-um supplemented with 10 mM sodimedi-um pyruvate (fed-batch, adjusted to 10 mM each 12 h until 36 h) and 1.27 g/L (5mM) n-octadecane. Experiments were conducted in 2 L loose cap shaking flasks during 72 hours. Growth curve of mutant RMS5 growing on sodium pyruvate and n-octadecane in screw cap shaking flask (from figure 3.21) is also shown for comparison. (b) Cell morphology observed during growth of the wild-type strain in the different growth phases: (I.) Rods - first exponential phase; (II.) Small coccobacillus – plateau; (III.) Coccobacillus – second ex-ponential phase.

The growth curve of the wild-type strain in figure 3.25 (a), as in the case of sealed serum vials experiments, presented 2 defined exponential phases. Generation times were calculated to be 2.3 h for the first 12 h growth phase and about 23 h for the second phase observed after 48 h of incubation. The mutant strain RMS5 reached cell densities comparable to that of the wild-type strain after 36 h of incubation but the growth curve showed a flatter and longer exponential phase with a generation time calculated to about 9 h, 33% shorter than the generation time cal-culated for this mutant (13.5 h) when cultured in screw cap shaking flasks (limited oxygen) under comparable conditions, where the RMS5 mutant reached only half of the cell density observed with loose cap shaking flasks. It was assumed that increased oxygen availability enabled by im-proved gas exchange (because of the loose caps) was responsible for these changes.

a)

b)

I. II. III.

0.0E+00 5.0E+08 1.0E+09 1.5E+09 2.0E+09 2.5E+09

0 20 40 60 80

Cell Density [Cells/mL]

Cultivation Time [h]

Thermus sp. ATN1 Mutant RMS5 Mutant DG11

Cell morphology development for the mutant RMS5 during bioconversion experiments was simi-lar to that shown by the wild-type strain, except that for the mutant strain the change in mor-phology did not occur in two defined exponential phases and small coccobacillus were observed as early as 12 h after incubation (Figure 3.25 b). The wild-type strain and its mutants did not show this morphology change when cultured in complex media. Interestingly mutant RMS5 showed some filamentation in complex media, which might be an indication of unregulated ex-pression of the alkane hydroxylase complex. A similar phenomenon was observed with the physi-ology of P. putida Gpo1 and P. putida GPo12 alk recombinants after induction of expression of the alkBFGHJKL genes on glucose as a carbon source (Chen et al., 1996). It has been suggested that overproduction of the alkB hydroxylase, which is a membrane protein, may be responsible for the observed phenotype (Wentzel et al., 2007).