Strains
Transformation of Wildtype CBS7435 and Erg6-KO resulted in strains of both with Module 1, Module 1 and 2 and Module 1, 2 and 3. Only the last ones were of further interest for us. Success of Transformation is somewhat plausible, since they grew on media containing NTC, ZEO and HYG simultaneously for a long time, which should not be possible with linearised plasmids Morphology is also consistent with Pichia. However, no colony PCR was conducted and therefore no molecular biological evidence exists to confirm presence. No Testosterone or Pregenolone acetate could be confirmed in our samples.
Sterol quantification using HPLC
HPLC analysis could not ascertain the presence of Testosterone or Pregnenolone acetate.
As the figure shows, neither Testosterone nor Pregnenolone acetate could be found. However, the Erg6-KO shows comparatively little amounts of Ergosterol, meaning the plan to redirect the entire flux into the Testosterone pathway by blocking the Ergosterol pathway is not without merit. In the modified Wildtype a Peak in the Area of cholesterol appears. However, the retention time is slightly different, and the Wildtype shouldn’t be able to produce cholesterol.
Sterol characterization using GCMS
In both wildtype (ERG6+) DHCR7+module2+module3+ (WT123 denoted in selected-ion monitoring [SIM] chromatogram), wt control strains (WTc), Erg6ΔDHCR7+DHCR24+module2+module3+ (delta6III) strains and Erg6ΔDHCR7-DHCR24-module2-module3- (delta6c) strains, we could not find traces of Testosterone (Fig2a and b), suggesting that our process, so far, is insufficient for testosterone de novo synthesis. We then asked if our strains successfully synthesized pregnenolone. Peaks were present in all of our strains (WTc, WT123, delta6III and delta6c), resembling the retention time of that of pregnenolone (Fig,2c and d). However, when sample SIM chromatograms are compared with that of the pregnenolone acetate standard (Fig2.e and f), only delta6c (unmodified Erg6 deleted) showed peaks with retention time (t=19.782; 20.534) similar to that of pregnenolone acetate (t=19.679). Because Erg6 is a methyltransferase, and many of the enzymes in steroid synthesis pathways are known to be tolerate to similar molecules, deletion of Erg6 may result in characteristic metabolic intermediates. However, this is not observed in delta6III samples (Erg6ΔDHCR7+DHCR24+module2+module3+). Thus, we conclude that our process is unable to synthesize pregnenolone.
Next, we looked at if our strains were able to synthesize cholesterol. Interestingly all strain samples show peaks at the retention time of the cholesterol standard (Fig.2g and h), regardless of the strain (presence of cholesterol synthesis required enzymes). Notably, delta6III might be able to produce a significant amount of cholesterol compared to other strains (peak height: on the scale of 2×10^4), judged by the peak form and relative quantity (about 10% of the standard, not that the y-axis scale is different for different strains in the SIM chromatogram). This suggested that some of our Erg6ΔDHCR7+DHCR24+module2+module3+ (delta6III) strains might be able to synthesize cholesterol successfully. Importantly, further validation processes and investigation should either confirm our primary results. Finally, we characterised the ergosterol production in each of our strains, we found a significant disturbance in delta6III strains (Fig.2j). This is consistent with the function of Erg6. However, the ergosterol peak intensity is heterogenous among delta6c (control) samples, and one replicate showed considerably high ergosterol(Fig.2j). Further analytical steps should investigate is this is due to biological heterogeneity or errors in analytical procedures. In WT samples (123+ and control cells), ergosterol peak intensities are both homogenous in form, and the variance of the peak height (unfortunately, the Area under the peak is not available) is in magnitude of 10^1-10^2(Fig.2i).
Unfortunately, we could not get the designed PCR primers on time. Thus, we could not verify if the gene modules are successfully integrated into the P. pastoris genome and furthermore, determine whether the dysfunction of module 2 and 3 enzymes are due to disturbances on genetic or biochemical levels. From our GS-MS data, we are able to conclude prematurely that some of our delta3III(Erg6ΔDHCR7+DHCR24+module2+module3+) cells are able to produce cholesterol. Though, further analytical procedures and evidences are required for confirmation. For example, a colony PCR test could inform us if the genetic materials are correctly integrated into the yeast’s genome; and biochemical tests, such as westernblot could further provide us valuable information of the end-point expression levels of our desired product, hence further identify the key rate limiting and/or non-functional point.
Growth
The following diagram compares the growth of the produced mutant with steroid hormone pathway with similar strain lacking said pathway.
Unsurprisingly, the unmodified wildtype with its normal membrane lipid pathways shows the best growth. However, the growth of the other strains is surprising. First, the unmodified ΔErg6 strain shows the worst growth, significantly lower than the modified ΔErg6 strain. This is counterintuitive since one would assume that the added expression of 6 transgenic proteins and stress from xenobiotic cholesterol and steroid hormones slow growth. A mix up of modified and unmodified ΔErg6 strains was ruled out by growth assays in YPD with and without antibiotics. Another surprise is the very similar growth of both modified strains. Again, one would assume that a strain with an unmolested membrane lipid metabolism grows better than one where a key enzyme was knocked out. These unusual findings point toward some unrecognized experimental error and demand repeat of experiment with fresh pre cultures.
What to do now?
Current data shows clearly that our strains are unable to produce Testosterone. The question now is, what can be learned and what are possible ways forward. Some possibilities.
- The strains do not contain our plasmids. This can be ascertained using simple colony PCR. This would elucidate whether the problem is faulty genes or an unfeasible pathway.
- The plasmids were damaged during simultaneous integrationThis could be possible through homologous recombination with the homologous promotor and terminator sites. This could also be determined through colony PCR and if true possibly remedied by sequential transformation.
- The strain produces testosterone but very small amountsCells used for extraction were cultivated for only 24-36h before harvest. Together with the rather weak constitutional promotors this could mean that just not a lot of testosterone is produced, less than is lost through chemical/biochemical degradation. One of the possibilities is therefore a redesign of the used regulatory elements, a swich to stronger promoters, or methanol induced promotors. These would allow a biomass accumulation before production is turned on.