Final Report Summary - MICROPUFA (Developing a versatile and efficient microbial PUFA producer)
Final publishable summary report
Executive summary
The genetic modification of M. alpina has so far proved difficult and progress has been slow, with the lack of a robust molecular toolkit. A key reason is that the fungus is within the phylum Zygomycota, rather than the more widely studied fungal Ascomycota and Basidiomycota for which numerous protocols have been developed. A limited number of transformation procedures have been described for M. alpina. However, these methodologies have poor efficiency and there is a pressing need to optimize these and develop improved methods. We have designed a transformation procedure for M. alpina including a counter selectable marker. This toolbox was used to modify genes involved in PUFA biosynthesis or degradation which were identified by advanced transcriptomics experiments.
Previous biochemical work has already identified certain genes involved in PUFA production and regulation of PUFA production (malic enzyme). New microarray-based technologies now offer a powerful tool to identify genome-wide pathways mediating PUFA synthesis, with the exiting prospect of discovering novel key genes. DSM has considerable experience in microarray synthesis and interpretation as shown in nature biotechnology publications on A. niger and P. chrysogenum. The microarrays have been used to assess global changes in the gene expression profile of M. alpina grown under various PUFA production conditions. The integration of these data with fermentation and production data and extensive metabolite analysis allowed the identification of genes that are important in regulation of PUFA biosynthesis. By modifying these genes the PUFA content/composition can be steered.
With this research a greater fundamental understanding of the biosynthetic pathways and genetics of PUFA production has emerged. In addition the new molecular tools developed allowed the manipulation of M. alpina and facilitate it’s use as a cell factory for the production of various types of PUFAs. In summary, these insights give a great leap forward in the understanding of production of PUFAs in M. alpina and will provide invaluable information for the industrial application of this organism as a versatile PUFA producer.
Description of project context and objectives
The term PUFAs refers to a group of long chain fatty acids that include among them the essential omega-3 (n-3) and omega-6 (n-6) fatty acids. Humans are required to obtain the omega-3 linolenic acid and the omega-6 linoleic acid directly from dietary sources as they lack the enzymes capable of synthesizing these compounds from precursors (Wertz., 2009). These PUFAs are then metabolised into a number of key compounds that play vital roles in inflammation, cardiovascular health and normal brain and visual function. However, dietary intake of PUFAs, especially in the western world is insufficient and imbalanced with a ratio of n-6 to n-3 PUFAs of 20:1 (Wertz., 2009). The current recommendation for the ratio of n-6 to n-3 PUFAs is 1:1 with an increase in n-3 consumption as opposed to a decrease in n-6 intake advised (Wertz., 2009). In fact this imbalance of n-3 and n-6 acids has been implicated in a number of disease states (Wertz., 2009). This is most starkly highlighted by the lack of mortality due to cardiovascular disease in populations, like Japan, where a high intake of n-3 PUFAs and therefore a reduced ratio of n-6 to n-3 PUFAs is evident (Kita., 2004). PUFAs, both n-3 and n-6, can be obtained from many sources including flaxseed’s, animal tissues and algal cells however, these sources must be metabolised to become active and this process is inefficient (Williams et al., 2006). Therefore, it is recommended that most of the dietary intake of PUFAs should come from oily fish or derived supplements. However, some people do not like to eat oily fish and there is on-going concern about the levels of toxic contaminants that are ingested by fish and the sustainability of fish stocks (Brunner et al., 2009). This has led to an effort to find other sustainable and reliable sources of PUFA production. Therefore, in the last 25 years there has been a large amount of interest in the production of PUFAs via microbial means.
Micro-organisms that accumulate lipids, 'oleaginous' species, have been known for over 100 years (Ratledge., 2001). However, it is only in the past 25 years that these organisms have become a focal point for research into single cell oils (SCO). Oleaginous organisms are defined as being able to accumulate triacylglycerols at greater than 20% of their biomass. Typically these triacylglycerols contain high quantities of nutritionally essential n-3 or n-6 PUFAs (Ratledge., 2001). The mechanisms that these organisms use to accumulate triacylglycerols at such high levels is still not fully understood with current research indicating that nitrogen depletion of the medium in the presence of excess carbon is the main trigger (Evans et al., 1985; Ratledge., 2002).
One such oleaginous organism, Mortierella alpina, was discovered as a potential producer of the n-6 C20 fatty acid arachidonic acid (ARA) in 1987. Since then this organism has proved to be the most promising single-cell oil (SCO) source that is rich in C20 PUFAs (Totani and Oba
1987; Yamada et al., 1987; Amano et al., 1992; Shimizu and Jareonkitmongkol 1995). Subsequently extensive investigation into the production of ARA and other PUFAs by M. alpina has revealed the biosynthetic pathways involved (Figure 1) and led to the use of this fungus as a commercial producer of ARA (Sakuradani et al., 2009).
Pathways for the biosynthesis of PUFAs in M. alpina. Shown are the key intermediates and enzyme classes involved in the PUFA pathway. The final PUFA composition can be influenced by modifying enzyme activity. It is planned to develop a genetic toolbox for M. alpina to allow the regulation of such enzyme activities.
Although levels of industrial production are relatively satisfactory, there are commercial and global competitive pressures to strive for increased output and range of PUFA production. To facilitate an increased output and range of PUFAs from M. alpina robust molecular tools would be highly advantageous. However, the genetic modification of M. alpina has, so far, proved difficult and progress has been slow. A limited number of transformation procedures have been described for M. alpina. However, these methodologies have poor efficiency and there is a need to develop and optimize the procedures. To this end a program of work was devised that would allow for the development of a robust molecular “toolkit” for M. alpina.
Aim 1: To optimise a reliable and efficient transformation technique for M. alpine to improve and optimize previously reported transformation procedures for M. alpina. Transformation by both protoplast mediated and Agrobacterium methods will be assessed and optimized under various culture conditions.
Identification of a selectable marker Antibiotics are routinely used as positive selection markers in fungal transformations. They are useful as they preclude the need to generate auxotrophic strains and can be used in the wild type genetic background of the fungi. However, there is significant variation across fungal strains in their sensitivity to antibiotics and spontaneous resistance is known to occur. Therefore, careful screening of your fungi for sensitivity to antibiotics is an important 1st step in determining a selection protocol for transformation.
Previous studies have shown that there is significant variation in the sensitivity of different M. alpina strains to antibiotic selection markers. Therefore, the 1st step in optimising a transformation protocol for M. alpina was to determine the antibiotic sensitivity of the strains. To test the sensitivity a series of agar plates containing glucose (2%) and yeast extract (1%) medium, pH7.5 (GY medium) plus varying concentrations of a spectrum of antibiotics where inoculated. Growth on these plates was then monitored over 5 days and compared to GY agar containing no antibiotics.
The results from these experiments showed that M. alpina exhibit sensitivity to high concentrations of hygromycin B. This concentration of hygromycin B is particularly high as standard fungal transformations and other strains of M. alpina are inhibited by 0.20mg/ml of this antibiotic.
However, it was decided that despite the high concentrations required hygromycin B was a good candidate as a selectable marker.
Electroporation
Previous investigations have reported that any attempt to perform electroporation transformations on M. alpina had yielded no colonies, although in all cases no attempt at optimisation was made. As the major aim of this study is to develop a robust transformation procedure it was decided to attempt this procedure. These experiments where guided in part by an optimisation of the electroporation transformation of the related fungi Mucor circinelloides (Guttierez et al 2011). In this investigation a number of experiments where performed using different concentrations of DNA and different electroporation conditions. The experiments where performed using different plasmids.
Electroporation using an integrative plasmid
The results of using electroporation for transformation of M. alpina gives different results depending on the conditions used. The number of colonies varied from 1 -25 depending on the strain and conditions used. The false positive rate was determined as between 0-50% with the average being approx. 35-40%. Stability of the positive colonies ranged from 0-80% with the average approx.. 50-60%. Furthermore, at least 50% of all positive, stable colonies had a single random gene integration event with the remaining colonies containing 2-4 copies of the plasmid.
Analysis of the conditions tested showed that only the maximum capacitance value used produced any positive colonies and the use of linear DNA was preferable to circular DNA with a higher number of colonies observed per μg of DNA. In addition, increasing the concentration of DNA in the transformation increased the number of transformants, although this appeared to have no effect on the number of false positives, stability or copy number of the resultant strains.
Electroporation using an episomal plasmid
The transformation of M. alpina strains with the episomal plasmid was not very successful. Indeed, only a maximum of 2 colonies could be achieved using this method. Therefore this line of research was discontinued.
Chemical transformation of M. alpina
The earliest reported transformation of M. alpina was achieved using a chemical transformation procedure (Mackenzie et al, 2001). In that study a plasmid containing the HygB genes was transformed into M. alpina under the control of a homologous promoter. We have used the transformation procedure, modified to our strains, to assess the efficacy of transformation with both an integrative and episomal plasmid.
The results of the optimized chemical transformation with different plasmids produced consistent results for each of the M. alpina strains. In conclusion, we have demonstrated the viability of transformation both with chemical and electroporation. Moreover, these data show that the chemical transformation method is a more robust method for M. alpina than electroporation as it generates more colonies per μg of DNA, is reproducible, has no strain bias and delivers higher numbers of transformants. It was therefore decided that this method would be used in all subsequent transformations.
Protoplast regeneration
One of the key steps in the transformation protocols described above is the regeneration potential of the protoplasts. If the regeneration of protoplasts is poor the chances of achieving a high number of transformants is decreased as the pool of viable cell fragments is limited. To analyse the effect of media composition on protoplast regeneration, protoplasts were generated and assessed for viability on different media compositions. In these experiments we were able to increase the number of protoplast by using an optimised media formulation. Using this new medium we were able to increase protoplast potential from ~7% - ~40%. Therefore, the new media formulation will be used in all future experiments.
Use of optimised protoplast regeneration media in transformations
The previous transformations described above used GY as the base for the selection media. As demonstrated in the previous section this media results in inefficient protoplast regeneration and therefore potentially sub-optimal transformation performance. To investigate this possibility transformations were performed using the following conditions the chemical transformation method and medium optimised in the previous sections. In addition, an overlay of agar was added to the cultures in an attempt to increase the number of positive colonies observed.
The overlay media was identical to the media the transformed culture was plated onto the culture after after 2 days of growth.
In conclusion, we have demonstrated the viability of chemical transformation for M. alpina and optimised the media conditions, amount of DNA, type of DNA and the use of either episomal or integrative plasmids. These optimisation experiments have allowed us to formulate a transformation method that is both robust and simple and will be used to progress the rest of the project.
Creation of M. alpina specific vector
The previously described experiments have used heterologous plasmids not specifically designed for M. alpina expression. To create novel strains for the next-stage of the project new vectors were required that allow overexpression of genes and creation of knockout cassettes. To this end 2 new vectors were constructed that contained homologous M. alpina promoters that drive both GFP and HygB expression. Both vectors were designed by for ease of use and contain cloning sites for modification. These vectors are proprietary property of DSM pMORTv1 and pMORTv2
The vector named pMORTv1 was designed by the fellow and constructed by DNA2.0 whereas pMORTv2 was made by cloning of a 1kb region of M. alpina DNA into pMORTv1. This extra piece of DNA was added to aid the integration of the plasmid into the genome.
Both of these vectors where then used to transform DSM M. alpina strains using the optimised transformation described in a previous section.
In these initial experiments the plasmids were expressing the GFP gene.
Transformation of M. alpina with pMORTv1 (GFP) The results of the transformation with pMORTv1(GFP) show that the colony number is decreased from the number obtained from transformation with heterologous plasmids. However, the transformation was highly efficient with a false positive rate of only 10%.
In addition, to the above tests the positive colonies were also assessed for copy number and were found to contain either 1 or 2 copies per strain. GFP expression was conformed via fluorescent microscopy. This is the first known expression of GFP in this organism.
These results indicate that the pMORTv1 vector can be used to transform M. alpina with high efficiency and stability and is therefore a useful tool for overexpression of genes to increase or alter oil production in this organism. In addition, to our knowledge this is the first demonstration of GFP expression in M. alpina.
Transformation of M. alpina with pMORTv2 (GFP)
As described previously a 1kb region of M. alpina DNA was cloned into the pMORTv1 vector to help with genomic integration.
As described in the previous section GFP expression from pMORTv1 was weak and one possible reason for this could be the low copy numbers obtained for transformants (max of 2). Therefore a vector that had increased copy number may help overcome this weak expression and would be useful in later studies when expressing genes to alter or improve oil production.
Once constructed, the newly formed pMORTv2(GFP) vector was transformed into M.alpina. The results of the transformation shows that the total amount of colonies obtained is increased from the number obtained from transformation with pMORTv1(GFP) up to 42. Again, the transformation was highly efficient with a false positive rate of only 10%. Analysis of the copy number of positive colonies revealed a major difference in the copy number of pMORTv2 over pMORTv1. In positive colonies there was a greater variation in copy number than was previously observed for strains containing pMORTv1. As described previously for the pMORTv1 vector all colonies containing pMORTv2(GFP) were assessed for GFP expression via fluorescent microscopy and were seen to be expressing this gene.
These results indicate that the vector pMORTv2 can be used to transform M. alpina with high efficiency. In addition, pMORTv2 can also be used to obtain higher copy numbers than pMORTv1 and this may lead to higher expression of genes in this organism. The disadvantage of using this vector create high copy number strains is that these strains become less stable and are therefore more likely to revert. In conclusion, pMORTv2 is another vector with slightly different characteristics to pMORTv1 that can aid in the process of increasing or altering oil production in this M. alpina.
Differences in RNA levels for GFP between pMORTv1 and pMORTv2
As described in the previous section both pMORTv1 and pMORTv2 could successfully transform M. alpina with high efficiency. However, the one major difference between the two vectors was that pMORTv2 could be made to integrate at higher copy numbers. Although, these higher copy number cells were inherently less stable some of them were able to maintain up to 5 copies/cell. Analysis of these transformed strains under a fluorescent microscope appeared to show that these differences in copy number did not increase the level of protein production in the cell. To assess differences in the expression profile of cells containing different copy numbers of plasmid RNA was isolated and the GFP RNA transcript levels were measured.
This analysis shows that despite higher copy numbers there is no statistical difference in GFP expression across either plasmid or copy numbers. This suggests that the initial interpretation of the fluorescent images, that there was no difference in protein expression across the copy numbers was correct and that the overriding factor in weak GFP expression might be the promoter that was used.
Conclusion of transformation studies
In this section it has been demonstrated that we have developed a chemical based transformation procedure that is reliable, robust and generates sufficient colonies to be of practical use. In addition, we have created 2 vectors that can be used to overexpress heterologous genes in M. alpina at a range of copy numbers. Therefore, this aim has been successfully completed and even exceeded expectations.
Aim 2: To create novel auxotrophic strains for the development of selectable and counter selectable markers.
2.1 Use of classical mutagenesis to generate an auxotrophic strain
A common method for the generation of auxotrophic strains in fungi is the use of an anti-metabolite to identify cells deficient in certain biosynthetic pathways. Commonly the technique involves exposing cells to a mutagenic factor, either UV light or a chemical mutagen and then screening for mutants which have the ability to grow on media containing the anti-metabolite. In this study we have used the mutagen N-methyl- N’-nitro-N-nitroguanidine (NTG) to generate mutants that where resistant to a specific anti-metabolite and screen these mutants for deficiency in biosynthetic pathways. We have optimised the mutagenesis protocol and screened a number of colonies for the desired physiological trait.
From these colonies we identified strains with the desired traits and assessed the mutations that gave rise to this trait. Thus we have identified an auxotrophic strain that could be used in counter selectable marker studies. This strain was assessed for its physiological traits and was fully characterised. The strain was also returned to a wild type characteristic by transformation of the wild type gene to regain prototrophy.
2.2 Investigation of acetamide (amdS) as a counter selectable marker
The use of the acetamidase encoding gene amdS as a counter selectable marker in fungal genetics is commonplace. It is based on the premise that fungi are poor at utilizing acetamide as a nitrogen and carbon source. Therefore, once the fungi are transformed with the A. nidulans amdS gene positive selection can take place by testing for growth on acetamide. In addition, strains containing the amdS gene can be negatively selected for as they do not grow on media containing fluoroacetamide. This trait has been used, in conjunction with pyr selection, to select against ectopic integration events and therefore increase the efficiency of gene replacement in A. awamori (Michielse et al, 2005). The use of amdS had not been tested in M. alpina and was assessed in this study. The growth of M. alpina using acetamide as the sole nitrogen and sole nitrogen and carbon source was measured by plating. However, in all media tested both strains showed similar growth characteristics to growth on GY media. Therefore, this method of selection cannot be used for M. alpina transformations.
2.3 Investigate the viability of the Cre-Lox system of gene excision
Cre-Lox is a site-specific recombinase technology that can be used to carry out deletions, insertions, translocations and inversions in chromosomal DNA. The system consists of two parts, a DNA-recombinase Cre which catalyses the recombination between specific DNA sites and LoxP which contain the binding sites for Cre-recombinase. Therefore, when cells that contain LoxP sites express Cre recombinase a recombination event occurs between the two sites. Once bound Cre-recombinase cuts the double stranded DNA at both sites and recombines them. Depending on the orientation of the LoxP sites a deletion, inversion or transpositon event can take place. Activation of the Cre recombinase can be controlled via external stimuli, i.e. chemical or heat shock or transient transfection of the Cre-recombinase activity under the control of a constitutive promoter (Florea et al, 2009). In this study the Cre-Lox system will be used to excise marker genes so that they can be used in further rounds of transformations, important as only two markers are available. This technology was primarily used in the creation of gene knockouts so for a full discussion of the results see Aim 3.
Aim 3: To develop a reliable procedure for the creation of knockout strains of M. alpina.
A major difficulty in genetic manipulation of M. alpina has been the inability to make targeted gene replacements/deletions. To date only the method of RNAi has been described for disrupting genes in M. alpina. In this part of the study we will create vectors that will allow for targeted knockout of genes using homologous recombination. We will then verify this technique by first creating an auxotrophic strain that can be simply re-constituted with the cloned gene. Once verified this technique can then be harnessed to knockout genes that have been identified in the transcriptomic analysis below that will be beneficial for specific PUFA production.
3.1 Testing of RNAi technology
RNAi is a post-transcriptional gene silencing process that can be used to dampen, or silence, the expression of a target gene without the need for deletion or attenuation. Transformed cells expressing double stranded RNAi down regulate the gene of interest by causing the degradation of target RNA, provided there is 100% homology to the target sequence. This method has been demonstrated to work for the majority of fungi and provides distinct advantages over gene deletion methods, such as the ability to attenuate the expression of essential genes. In addition, it has also been demonstrated as an effective method in hetereokaryotic fungi with multi-nuclear hyphae such as the zygomycetes. This is of particular importance to this project as M. alpina is a member of the zygomycetes and has multi-nuclear hyphae. In this project RNAi is applied to silence either essential genes that maybe beneficial in boosting fatty acid production or for creating more efficient homologous recombination strains.
Using the pMORTv1 vector a construct has been created that contains an inverted repeat of a fatty acid gene from M. alpina. This construct once transformed will produce an RNA transcript should downregulate expression of this gene and therefore change the oil content of M. alpine (Takeno et al, 2005). The RNAi construct was synthesised by DNA2.0 and was cloned onto pMORTv1. This construct was then transformed into M. alpina and the fatty acid profile of the successful transformants was analysed. The analysed transformants showed no difference in their fatty acid profiles suggesting that this technique does not work.
3.2 Identification of Ku70/80 homlogs
To use the recently available whole genomic sequence to identify genes such as Ku70 and Ku80 that promote homologous recombination by homology mapping, and target these for down regulation using RNAi technology. Down regulation of ku70 promotes a higher frequency of homologous DNA recombination events and therefore increases transformation efficiency. We will apply this principal to M. alpina to create a strain that has down regulated ku70. This is predicted to improve the efficiency of homologous recombination events and therefore the efficiency of transformation procedures. At present there is no known homolog of Ku70/80 in M. alpina. Homology searches have yielded no results and therefore a degenerate primer approach has been undertaken to elucidate the genomic position of these genes. This work has been unsuccessful and due to time constraints within the project was Ku70/80 was not identified.
Creation of an auxotrophic strain using homologous recombination
To evaluate the feasibility of making gene knockouts in M. alpina a cassette was made for the deletion of a gene essential for biosynthesis of an essential metabolite. Transformation of the linear cassette was performed using the chemical transformation method described previously. In total 6 colonies where obtained from this transformation and all 6 where assessed for deletion of the gene. PCR of the genomic DNA from these colonies revealed that all 6 had a deletion of the gene and qPCR showed that it was a single copy integration event. All strains showed the same characteristics as the previously described auxotrophic strain. Re-transfromation of these strains with a cassette containing the wild type gene restored prototrophy and growth to wild-type levels. This finding is of vital importance for this project as it demonstrates the feasibility of creating gene knockouts without leaving any exogenous DNA. In addition, it provides an alternative method for strain construction and marker regeneration if the investigation of the Cre-Lox system proves unsuccessful. However, the efficiency of integration appears to be very low and the investigations into the Cre-Lox system maybe more efficient.
Creation of a gene knockout strain and testing of Cre-Lox
In order to test the Cre-Lox technique, a gene cassette containing LoxP sites needs to be transformed into the M. alpina genome. This cassette contains a LoxP flanked HygB as a selectable marker which will then be inserted between two DNA sequences identical to the flanking regions of the target gene as shown. When transforming this gene cassette, homologous recombination will occur where nucleotide sequences are exchanged between two similar or identical DNA fragments, replacing the target gene with the LoxP flanked HygB gene creating knockout strains. Two different approaches will be tested in this transformation: transform the full length cassette and transformation of two bipartite fragments with HygB gene overlap. Each bipartite fragment represents half of the full cassette from one homologous region till the HygB gene. Therefore HygB resistance can only be acquired when both of the fragments are integrated at the same locus thereby increasing the transformation efficiency. The inserted HygB gene can then be removed by introducing Cre-recombinase. If this proves successful it would allow the use of the same marker in multiple rounds of transformations, which is particularly useful because of the limited markers currently available for M. alpina. A major advantage of this method is that the removal of the marker gene used in transformations can eliminate the amount of unessential DNA that the cell carries.
The gene selected is an important enzyme on the path to the biosynthesis of arachidonic acid (ARA). Therefore removal of this gene will result in a decrease in the amount of arachidonic acid produced by M. alpina which allows for an easy identification of true knockouts via analysis of the arachidonic acid content of the cells.
The bipartite transformation generated a total of 5 colonies which were then assessed for the replacement of the gene by the cassette. Isolated genomic DNA was screened for the presence of the cassette and the gene by PCR. The results of this screen showed that the gene had been replaced by the cassette in all of the 5 colonies indicating that this method was viable for generating knockout strains.
After determining that the colonies obtained from the transformation were true knockouts the removal of the cassettes via Cre-Lox was assessed. To test this two strategies were assessed, the first was using transient transfection of the Cre-Lox plasmid whereby the plasmid was used in a transformation but without any selective pressure. The reasoning behind this approach is that it is advantageous to have the Crerecombinase activated for as short a time as possible to prevent random splicing events. The disadvantage is that the low transformation efficiency would equate to low Cre activity. Secondly, the plasmid would be transformed using the selective pressure of growth using the previously described auxotroph.
The results show that using the transient transformation procedure a confluent plate was observed. This was as expected as no selection pressure was used for the transformation. When the colonies from this plate were analysed for the presence of the plasmid it was found that no vector could be found. In addition, only 2% of all colonies analysed had the cassette removed via Cre-recombinase activity. However, once the addition of selection pressure is added to the transformation procedure 50% of all colonies contain the plasmid but only 20% of these i.e. 1 colony had the cassette removed. Removal was also confirmed by lack of growth on HygB containing media. These strains lack the target gene but also have no knockout cassette and therefore represent a “clean removal” of the gene.
Overall, these results show that the technique of Cre-lox removal of the cassette and therefore the selectable marker is feasible but is highly inefficient with only 1 – 2 colonies obtained per transformation.
After the removal of the knockout cassette via Cre-recombinase the wild type gene was cloned into the vector pMORTv2. This vector was then transformed into the cassette removed knockout strain to assess if activity could be returned to this strain. Once constructed all of the strains described in this section were sent for fatty acid analysis following shake flask fermentations.
The results of the removal of the gene via gene replacement reduced the amount of arachidonic acid the cells produced. This was also observed in the cells after Cre-recombinase removal of the knockout cassette, whereas re-introduction of the gene returned the level of arachidonic acid in the cells to WT levels. These results show that the creation of knockouts and marker recycling via Cre-recombinase is a viable genetic tool in M. alpina and could be used to perform multiple rounds of transformations with limited markers. However, it should be noted that this process is very inefficient and needs further optimisation. In conclusion we have shown for the first time that making knockouts and Cre-mediated marker removal in M. alpina is feasible and could therefore be used in the future to identify targets for improved product yield.
Aim 4: To use the newly acquired M. alpina genome sequence and microarrays to assess global transcript changes in response to various growth conditions in order to identify genes beneficial in the production of PUFAs. Previous biochemical work has already identified certain genes involved in PUFA production and regulation of PUFA production (malic enzyme). New microarray-based technologies now offer a powerful tool to identify genome-wide pathways mediating PUFA synthesis, with the existing prospect of discovering novel key genes. The microarrays (already available) will be used to assess global changes in the gene expression profile of M. alpina grown under various PUFA production conditions. The integration of these data with fermentation and production data and extensive metabolite analysis will then allow the identification of genes that are important in regulation of PUFA biosynthesis.
Identification of target genes
Transcriptomic data from production fermentations of M. alpina have identified a number of genes that are potential targets for modification. These include not only genes in fatty acid synthesis, but genes involved in degradation and co-factor and substrate availability. See aim 5 for more detail.
Aim 5: To use the molecular toolkit and microarray data to enhance the ability of M. alpina to synthesize PUFAs. The data from the microarray studies have identified a series of possible genes involving aspects of PUFA production. The new molecular tool kit developed in aims 1 and 2 will then be used as follows.
5.1 Manipulate the biosynthetic pathways of M. alpina to produce a plethora of strains producing predetermined or modified PUFAs as required.
Targeted genes
As described in aim 4 we have identified a number of genes for modification that will affect fatty acid synthesis. Vectors have been constructed to allow the overexpression of the a number of genes. The identified genes were cloned into pMORTv1 or pMORTv2. Once cloned the vectors were then transformed into M. alpina using the previously described transformation technique. Both vectors transformed with low false positives, i.e. 90% positive colonies. Fatty acid content was analysed for selected strains transformed with pMORTv1 or pMORTv2. Each strain was grown for 7 days in a scaled down fermentation process.
Overexpression of genes involved in fatty acid synthesis produced in all cases an increase in the amount of arachidonic acid. The largest increase observed was 76%. This indicates that there is scope for improvement in the yield of arachidonic acid that could be produced by this organism and suggests possible targets for a classical strain improvement program.
5.2 Targeted knockouts of the selected genes (e.g. those with high transcript changes) involved in PUFA biosynthesis from microarray analyses will be made to elucidate their function in this process.
Among the possible targets identified in aim 4 were genes other than biosynthesis genes of PUFA’s. We identified a specific target for deletion. Constructs were created as previously described and transformed into M. alpina. Resultant colonies were analyzed for the removal the gene and presence of the marker gene. Once again the efficiency of transformation was low, 5 colonies, but 100% of the colonies contained the cassette and were true knockouts. In addition, these knockouts were also made marker free via Cre-lox. This strain was then transformed with a newly constructed vector pMORTv2 plus the gene removed to test reconstitution. All of these strains were then analysed for arachidonic acid content following a shake flask fermentation In addition, a knockout strain was also transformed with pMORTv2 plus a gene involved in fatty acid synthesis to assess the effect of a knockout plus overexpression.
The results show that removal of the degradation gene increases the amount of arachidonic acid that is produced by the cell. It also shows that reintroducing the gene reduces the amount of ARA to wild type levels.
Significant results
Optimisation of a robust transformation procedure for M. alpina
Creation of an auxotrophic strain using targeted gene deletion
Creation of an auxotrophic strain using classical mutagenesis
Complementation of auxtrophic strains using targeted gene replacement
Expression of HygB gene from a heterologous promoter
Targeted gene deletion to improve oil yields
Altered the amount of oil produced via overexpression of genes
Training activities
SHE training
Extensive safety health and environment training has been undertaken with a focus on sustainability, safety at work and best practice in safety.
Fungal genetics training
In addition to working with M. alpina, the genetics of the many organisms that DSM utilise has been studied and applied to the current project.
This training has increased the fellow’s knowledge base and experience and will be of great benefit to his career.
Mentoring of a student
The fellow has undertaken the mentorship of a student and has gained valuable experience in leading a small research team.
Attendance at Fungal Genetics Conference
The fellow attended the Genetics Society of America fungal genetics conference.
Technology transfer and collaborations
The fellow has fostered close links with other DSM sites notably DSM Columbia where he visited for a week to transfer the transformation procedure.
Impact of the research
Fungi are widely used to commercially produce a very wide range of products that include enzymes, organic acids, pharmaceuticals and other chemicals. They are attractive as expression systems as they are capable of producing high yields and are relatively easy to culture and maintain. The results of the research described above has enhanced the understanding of PUFA production by M. alpina and allowed us to develop this fungus into an efficient and sustainable producer of arachidonic acid. The results described above have also unlocked the potential of this fungus to produce many other PUFAs which could be used in a number of applications from food technology to waterproof coatings. We have developed molecular tools and combined it with the transcriptome analysis to elucidate the biosynthetic pathways of PUFAs production and develop new strategies for improving the yields of PUFAs. This has afforded a better understanding and optimisation of existing industrial uses of M. alpina and is a prerequisite for the development of the strain as a microbial cell factory for novel future multiple PUFA production. As a consequence this research has led to the creation of a number of strains each of which produce a specific and high yield of PUFA that have the potential to be used in the food industry. Furthermore, due to the limited amount of information that is available on PUFA production in M. alpina the results of this research have greatly enhanced the understanding of the biosynthetic pathways involved including key regulators, ratelimiting steps and gene expression levels during PUFA production.
Our development of a genetic tool kit for M. alpina in conjunction with our transcriptomic analysis is a major contribution to the understanding of the biochemistry and molecular biology of oleaginous fungi and a key step towards sustainable and efficient production of PUFAs. The specific results of the project can be summarised as follows:
- New molecular tools for manipulation of an industrially relevant organism
- The first transcriptomic analysis of an oleaginous fungi
- Development of M. alpina as a microbial cell factory for the production of various PUFAs
- An improved understanding of the basic metabolism underpinning PUFA production in oleaginous fungi.
PUFAs have many beneficial effects in cardiovascular health, infant nutrition and disease prevention when given as a dietary supplement. In addition, an imbalance of n-3 and n-6 PUFAs in the diet has been implicated in a number of disease states. These factors, and the questions surrounding the sustainability of current production methods for PUFAs, mean that the development of robust and efficient systems for essential fatty acid production is of paramount importance. This project has used the zygomycete fungus M. alpina, (GRAS accredited) already a reliable producer of the C20 PUFA arachidonic acid, to develop a molecular tool kit and unlock its potential as a cell factory for PUFA production. The new genetic tools developed will allow the modification of this organism to direct the synthesis of PUFAs and therefore generate a sustainable platform technology for the production of these essential 'health giving' compounds.
Such a platform will increase the availability of PUFAs on a global scale. In addition, this would place Europe to the front end of science in this field. Much work is ongoing in Japan to improve lipid production in microorganisms, but the development of a robust, and versatile microbial lipid factory has not been achieved so far. Such a microbial lipid factory could also be an invaluable tool for other European research scientists who are interested in contributing to the further development of such a microbial factory not only in the field of strain construction, but also in the field of fermentation and down stream processing. By setting up such a lipid production platform, synergies are expected with nutritional institutes who can study potential health effects of different PUFAs in a variety of applications. The possibility to tailor make PUFAs and combinations of PUFAs will trigger studies on synergetic effects, counteraction and competitive inhibition of PUFAs on neuronal function and inflammation.
Furthermore, there is the very real threat that an obesity epidemic is about to hit the European union with obesity rates among men at 10 – 27% and up to 38% in women (International Obesity Task Force., 2005). Consequently the rates of morbidity from cardiovascular disease will increase. Therefore the awareness level of people regarding food consumption particularly regarding the health benefits of PUFAs in cardiovascular disease prevention should be increased. The construction of a microbial cell factory that can be made to make various kinds of PUFAs would lead to the production of new products and formulations for existing food and healthcare products that will lead to greater public awareness of the benefits of these compounds for optimal well being.
Comparison of results with initial timelines
Aim 1: Testing transformation systems
Progress: Assessed electroporation and chemical transformation methods. Agrobacterium method was not assessed due to worries about residual bacterial DNA and the efficiency of the chemical transformation method. Optimised the chemical transformation method. This aim has been successfully achieved.
Aim 2: Development of a counter-selectable marker
Progress: Uracil auxotroph created via two methods. Aim achieved.
Aim 3: Improve homologous recombination efficiency
Progress: Achieved targeted gene deletion of the URA3 gene therefore demonstartiing the feasibility of creating knockouts in M. alpina. RNAi experiments and Ku70 identification proved to be problematic and were unable to be completed successfully. Aim achieved, however with mixed results.
Aim 4: Microarray analysis
Progress: Targets for overexpression and deletion were identified and constructs made. Aim achieved.
Aim 5: Strain improvement using recombinant techniques.
Progress: Targets identified from aim 4 were either overexpressed or knocked out. These newly constructed strains were then assessed for fatty acid content. In each of the strains, detailed in below the fatty acid content of the organism had been altered. In addition, the knockout strains were constructed in such a way that the Cre-lox method of gene excision could be tested. This would allow the recycling of selectable markers. This method was shown to be viable. Aim achieved.
Executive summary
The genetic modification of M. alpina has so far proved difficult and progress has been slow, with the lack of a robust molecular toolkit. A key reason is that the fungus is within the phylum Zygomycota, rather than the more widely studied fungal Ascomycota and Basidiomycota for which numerous protocols have been developed. A limited number of transformation procedures have been described for M. alpina. However, these methodologies have poor efficiency and there is a pressing need to optimize these and develop improved methods. We have designed a transformation procedure for M. alpina including a counter selectable marker. This toolbox was used to modify genes involved in PUFA biosynthesis or degradation which were identified by advanced transcriptomics experiments.
Previous biochemical work has already identified certain genes involved in PUFA production and regulation of PUFA production (malic enzyme). New microarray-based technologies now offer a powerful tool to identify genome-wide pathways mediating PUFA synthesis, with the exiting prospect of discovering novel key genes. DSM has considerable experience in microarray synthesis and interpretation as shown in nature biotechnology publications on A. niger and P. chrysogenum. The microarrays have been used to assess global changes in the gene expression profile of M. alpina grown under various PUFA production conditions. The integration of these data with fermentation and production data and extensive metabolite analysis allowed the identification of genes that are important in regulation of PUFA biosynthesis. By modifying these genes the PUFA content/composition can be steered.
With this research a greater fundamental understanding of the biosynthetic pathways and genetics of PUFA production has emerged. In addition the new molecular tools developed allowed the manipulation of M. alpina and facilitate it’s use as a cell factory for the production of various types of PUFAs. In summary, these insights give a great leap forward in the understanding of production of PUFAs in M. alpina and will provide invaluable information for the industrial application of this organism as a versatile PUFA producer.
Description of project context and objectives
The term PUFAs refers to a group of long chain fatty acids that include among them the essential omega-3 (n-3) and omega-6 (n-6) fatty acids. Humans are required to obtain the omega-3 linolenic acid and the omega-6 linoleic acid directly from dietary sources as they lack the enzymes capable of synthesizing these compounds from precursors (Wertz., 2009). These PUFAs are then metabolised into a number of key compounds that play vital roles in inflammation, cardiovascular health and normal brain and visual function. However, dietary intake of PUFAs, especially in the western world is insufficient and imbalanced with a ratio of n-6 to n-3 PUFAs of 20:1 (Wertz., 2009). The current recommendation for the ratio of n-6 to n-3 PUFAs is 1:1 with an increase in n-3 consumption as opposed to a decrease in n-6 intake advised (Wertz., 2009). In fact this imbalance of n-3 and n-6 acids has been implicated in a number of disease states (Wertz., 2009). This is most starkly highlighted by the lack of mortality due to cardiovascular disease in populations, like Japan, where a high intake of n-3 PUFAs and therefore a reduced ratio of n-6 to n-3 PUFAs is evident (Kita., 2004). PUFAs, both n-3 and n-6, can be obtained from many sources including flaxseed’s, animal tissues and algal cells however, these sources must be metabolised to become active and this process is inefficient (Williams et al., 2006). Therefore, it is recommended that most of the dietary intake of PUFAs should come from oily fish or derived supplements. However, some people do not like to eat oily fish and there is on-going concern about the levels of toxic contaminants that are ingested by fish and the sustainability of fish stocks (Brunner et al., 2009). This has led to an effort to find other sustainable and reliable sources of PUFA production. Therefore, in the last 25 years there has been a large amount of interest in the production of PUFAs via microbial means.
Micro-organisms that accumulate lipids, 'oleaginous' species, have been known for over 100 years (Ratledge., 2001). However, it is only in the past 25 years that these organisms have become a focal point for research into single cell oils (SCO). Oleaginous organisms are defined as being able to accumulate triacylglycerols at greater than 20% of their biomass. Typically these triacylglycerols contain high quantities of nutritionally essential n-3 or n-6 PUFAs (Ratledge., 2001). The mechanisms that these organisms use to accumulate triacylglycerols at such high levels is still not fully understood with current research indicating that nitrogen depletion of the medium in the presence of excess carbon is the main trigger (Evans et al., 1985; Ratledge., 2002).
One such oleaginous organism, Mortierella alpina, was discovered as a potential producer of the n-6 C20 fatty acid arachidonic acid (ARA) in 1987. Since then this organism has proved to be the most promising single-cell oil (SCO) source that is rich in C20 PUFAs (Totani and Oba
1987; Yamada et al., 1987; Amano et al., 1992; Shimizu and Jareonkitmongkol 1995). Subsequently extensive investigation into the production of ARA and other PUFAs by M. alpina has revealed the biosynthetic pathways involved (Figure 1) and led to the use of this fungus as a commercial producer of ARA (Sakuradani et al., 2009).
Pathways for the biosynthesis of PUFAs in M. alpina. Shown are the key intermediates and enzyme classes involved in the PUFA pathway. The final PUFA composition can be influenced by modifying enzyme activity. It is planned to develop a genetic toolbox for M. alpina to allow the regulation of such enzyme activities.
Although levels of industrial production are relatively satisfactory, there are commercial and global competitive pressures to strive for increased output and range of PUFA production. To facilitate an increased output and range of PUFAs from M. alpina robust molecular tools would be highly advantageous. However, the genetic modification of M. alpina has, so far, proved difficult and progress has been slow. A limited number of transformation procedures have been described for M. alpina. However, these methodologies have poor efficiency and there is a need to develop and optimize the procedures. To this end a program of work was devised that would allow for the development of a robust molecular “toolkit” for M. alpina.
Aim 1: To optimise a reliable and efficient transformation technique for M. alpine to improve and optimize previously reported transformation procedures for M. alpina. Transformation by both protoplast mediated and Agrobacterium methods will be assessed and optimized under various culture conditions.
Identification of a selectable marker Antibiotics are routinely used as positive selection markers in fungal transformations. They are useful as they preclude the need to generate auxotrophic strains and can be used in the wild type genetic background of the fungi. However, there is significant variation across fungal strains in their sensitivity to antibiotics and spontaneous resistance is known to occur. Therefore, careful screening of your fungi for sensitivity to antibiotics is an important 1st step in determining a selection protocol for transformation.
Previous studies have shown that there is significant variation in the sensitivity of different M. alpina strains to antibiotic selection markers. Therefore, the 1st step in optimising a transformation protocol for M. alpina was to determine the antibiotic sensitivity of the strains. To test the sensitivity a series of agar plates containing glucose (2%) and yeast extract (1%) medium, pH7.5 (GY medium) plus varying concentrations of a spectrum of antibiotics where inoculated. Growth on these plates was then monitored over 5 days and compared to GY agar containing no antibiotics.
The results from these experiments showed that M. alpina exhibit sensitivity to high concentrations of hygromycin B. This concentration of hygromycin B is particularly high as standard fungal transformations and other strains of M. alpina are inhibited by 0.20mg/ml of this antibiotic.
However, it was decided that despite the high concentrations required hygromycin B was a good candidate as a selectable marker.
Electroporation
Previous investigations have reported that any attempt to perform electroporation transformations on M. alpina had yielded no colonies, although in all cases no attempt at optimisation was made. As the major aim of this study is to develop a robust transformation procedure it was decided to attempt this procedure. These experiments where guided in part by an optimisation of the electroporation transformation of the related fungi Mucor circinelloides (Guttierez et al 2011). In this investigation a number of experiments where performed using different concentrations of DNA and different electroporation conditions. The experiments where performed using different plasmids.
Electroporation using an integrative plasmid
The results of using electroporation for transformation of M. alpina gives different results depending on the conditions used. The number of colonies varied from 1 -25 depending on the strain and conditions used. The false positive rate was determined as between 0-50% with the average being approx. 35-40%. Stability of the positive colonies ranged from 0-80% with the average approx.. 50-60%. Furthermore, at least 50% of all positive, stable colonies had a single random gene integration event with the remaining colonies containing 2-4 copies of the plasmid.
Analysis of the conditions tested showed that only the maximum capacitance value used produced any positive colonies and the use of linear DNA was preferable to circular DNA with a higher number of colonies observed per μg of DNA. In addition, increasing the concentration of DNA in the transformation increased the number of transformants, although this appeared to have no effect on the number of false positives, stability or copy number of the resultant strains.
Electroporation using an episomal plasmid
The transformation of M. alpina strains with the episomal plasmid was not very successful. Indeed, only a maximum of 2 colonies could be achieved using this method. Therefore this line of research was discontinued.
Chemical transformation of M. alpina
The earliest reported transformation of M. alpina was achieved using a chemical transformation procedure (Mackenzie et al, 2001). In that study a plasmid containing the HygB genes was transformed into M. alpina under the control of a homologous promoter. We have used the transformation procedure, modified to our strains, to assess the efficacy of transformation with both an integrative and episomal plasmid.
The results of the optimized chemical transformation with different plasmids produced consistent results for each of the M. alpina strains. In conclusion, we have demonstrated the viability of transformation both with chemical and electroporation. Moreover, these data show that the chemical transformation method is a more robust method for M. alpina than electroporation as it generates more colonies per μg of DNA, is reproducible, has no strain bias and delivers higher numbers of transformants. It was therefore decided that this method would be used in all subsequent transformations.
Protoplast regeneration
One of the key steps in the transformation protocols described above is the regeneration potential of the protoplasts. If the regeneration of protoplasts is poor the chances of achieving a high number of transformants is decreased as the pool of viable cell fragments is limited. To analyse the effect of media composition on protoplast regeneration, protoplasts were generated and assessed for viability on different media compositions. In these experiments we were able to increase the number of protoplast by using an optimised media formulation. Using this new medium we were able to increase protoplast potential from ~7% - ~40%. Therefore, the new media formulation will be used in all future experiments.
Use of optimised protoplast regeneration media in transformations
The previous transformations described above used GY as the base for the selection media. As demonstrated in the previous section this media results in inefficient protoplast regeneration and therefore potentially sub-optimal transformation performance. To investigate this possibility transformations were performed using the following conditions the chemical transformation method and medium optimised in the previous sections. In addition, an overlay of agar was added to the cultures in an attempt to increase the number of positive colonies observed.
The overlay media was identical to the media the transformed culture was plated onto the culture after after 2 days of growth.
In conclusion, we have demonstrated the viability of chemical transformation for M. alpina and optimised the media conditions, amount of DNA, type of DNA and the use of either episomal or integrative plasmids. These optimisation experiments have allowed us to formulate a transformation method that is both robust and simple and will be used to progress the rest of the project.
Creation of M. alpina specific vector
The previously described experiments have used heterologous plasmids not specifically designed for M. alpina expression. To create novel strains for the next-stage of the project new vectors were required that allow overexpression of genes and creation of knockout cassettes. To this end 2 new vectors were constructed that contained homologous M. alpina promoters that drive both GFP and HygB expression. Both vectors were designed by for ease of use and contain cloning sites for modification. These vectors are proprietary property of DSM pMORTv1 and pMORTv2
The vector named pMORTv1 was designed by the fellow and constructed by DNA2.0 whereas pMORTv2 was made by cloning of a 1kb region of M. alpina DNA into pMORTv1. This extra piece of DNA was added to aid the integration of the plasmid into the genome.
Both of these vectors where then used to transform DSM M. alpina strains using the optimised transformation described in a previous section.
In these initial experiments the plasmids were expressing the GFP gene.
Transformation of M. alpina with pMORTv1 (GFP) The results of the transformation with pMORTv1(GFP) show that the colony number is decreased from the number obtained from transformation with heterologous plasmids. However, the transformation was highly efficient with a false positive rate of only 10%.
In addition, to the above tests the positive colonies were also assessed for copy number and were found to contain either 1 or 2 copies per strain. GFP expression was conformed via fluorescent microscopy. This is the first known expression of GFP in this organism.
These results indicate that the pMORTv1 vector can be used to transform M. alpina with high efficiency and stability and is therefore a useful tool for overexpression of genes to increase or alter oil production in this organism. In addition, to our knowledge this is the first demonstration of GFP expression in M. alpina.
Transformation of M. alpina with pMORTv2 (GFP)
As described previously a 1kb region of M. alpina DNA was cloned into the pMORTv1 vector to help with genomic integration.
As described in the previous section GFP expression from pMORTv1 was weak and one possible reason for this could be the low copy numbers obtained for transformants (max of 2). Therefore a vector that had increased copy number may help overcome this weak expression and would be useful in later studies when expressing genes to alter or improve oil production.
Once constructed, the newly formed pMORTv2(GFP) vector was transformed into M.alpina. The results of the transformation shows that the total amount of colonies obtained is increased from the number obtained from transformation with pMORTv1(GFP) up to 42. Again, the transformation was highly efficient with a false positive rate of only 10%. Analysis of the copy number of positive colonies revealed a major difference in the copy number of pMORTv2 over pMORTv1. In positive colonies there was a greater variation in copy number than was previously observed for strains containing pMORTv1. As described previously for the pMORTv1 vector all colonies containing pMORTv2(GFP) were assessed for GFP expression via fluorescent microscopy and were seen to be expressing this gene.
These results indicate that the vector pMORTv2 can be used to transform M. alpina with high efficiency. In addition, pMORTv2 can also be used to obtain higher copy numbers than pMORTv1 and this may lead to higher expression of genes in this organism. The disadvantage of using this vector create high copy number strains is that these strains become less stable and are therefore more likely to revert. In conclusion, pMORTv2 is another vector with slightly different characteristics to pMORTv1 that can aid in the process of increasing or altering oil production in this M. alpina.
Differences in RNA levels for GFP between pMORTv1 and pMORTv2
As described in the previous section both pMORTv1 and pMORTv2 could successfully transform M. alpina with high efficiency. However, the one major difference between the two vectors was that pMORTv2 could be made to integrate at higher copy numbers. Although, these higher copy number cells were inherently less stable some of them were able to maintain up to 5 copies/cell. Analysis of these transformed strains under a fluorescent microscope appeared to show that these differences in copy number did not increase the level of protein production in the cell. To assess differences in the expression profile of cells containing different copy numbers of plasmid RNA was isolated and the GFP RNA transcript levels were measured.
This analysis shows that despite higher copy numbers there is no statistical difference in GFP expression across either plasmid or copy numbers. This suggests that the initial interpretation of the fluorescent images, that there was no difference in protein expression across the copy numbers was correct and that the overriding factor in weak GFP expression might be the promoter that was used.
Conclusion of transformation studies
In this section it has been demonstrated that we have developed a chemical based transformation procedure that is reliable, robust and generates sufficient colonies to be of practical use. In addition, we have created 2 vectors that can be used to overexpress heterologous genes in M. alpina at a range of copy numbers. Therefore, this aim has been successfully completed and even exceeded expectations.
Aim 2: To create novel auxotrophic strains for the development of selectable and counter selectable markers.
2.1 Use of classical mutagenesis to generate an auxotrophic strain
A common method for the generation of auxotrophic strains in fungi is the use of an anti-metabolite to identify cells deficient in certain biosynthetic pathways. Commonly the technique involves exposing cells to a mutagenic factor, either UV light or a chemical mutagen and then screening for mutants which have the ability to grow on media containing the anti-metabolite. In this study we have used the mutagen N-methyl- N’-nitro-N-nitroguanidine (NTG) to generate mutants that where resistant to a specific anti-metabolite and screen these mutants for deficiency in biosynthetic pathways. We have optimised the mutagenesis protocol and screened a number of colonies for the desired physiological trait.
From these colonies we identified strains with the desired traits and assessed the mutations that gave rise to this trait. Thus we have identified an auxotrophic strain that could be used in counter selectable marker studies. This strain was assessed for its physiological traits and was fully characterised. The strain was also returned to a wild type characteristic by transformation of the wild type gene to regain prototrophy.
2.2 Investigation of acetamide (amdS) as a counter selectable marker
The use of the acetamidase encoding gene amdS as a counter selectable marker in fungal genetics is commonplace. It is based on the premise that fungi are poor at utilizing acetamide as a nitrogen and carbon source. Therefore, once the fungi are transformed with the A. nidulans amdS gene positive selection can take place by testing for growth on acetamide. In addition, strains containing the amdS gene can be negatively selected for as they do not grow on media containing fluoroacetamide. This trait has been used, in conjunction with pyr selection, to select against ectopic integration events and therefore increase the efficiency of gene replacement in A. awamori (Michielse et al, 2005). The use of amdS had not been tested in M. alpina and was assessed in this study. The growth of M. alpina using acetamide as the sole nitrogen and sole nitrogen and carbon source was measured by plating. However, in all media tested both strains showed similar growth characteristics to growth on GY media. Therefore, this method of selection cannot be used for M. alpina transformations.
2.3 Investigate the viability of the Cre-Lox system of gene excision
Cre-Lox is a site-specific recombinase technology that can be used to carry out deletions, insertions, translocations and inversions in chromosomal DNA. The system consists of two parts, a DNA-recombinase Cre which catalyses the recombination between specific DNA sites and LoxP which contain the binding sites for Cre-recombinase. Therefore, when cells that contain LoxP sites express Cre recombinase a recombination event occurs between the two sites. Once bound Cre-recombinase cuts the double stranded DNA at both sites and recombines them. Depending on the orientation of the LoxP sites a deletion, inversion or transpositon event can take place. Activation of the Cre recombinase can be controlled via external stimuli, i.e. chemical or heat shock or transient transfection of the Cre-recombinase activity under the control of a constitutive promoter (Florea et al, 2009). In this study the Cre-Lox system will be used to excise marker genes so that they can be used in further rounds of transformations, important as only two markers are available. This technology was primarily used in the creation of gene knockouts so for a full discussion of the results see Aim 3.
Aim 3: To develop a reliable procedure for the creation of knockout strains of M. alpina.
A major difficulty in genetic manipulation of M. alpina has been the inability to make targeted gene replacements/deletions. To date only the method of RNAi has been described for disrupting genes in M. alpina. In this part of the study we will create vectors that will allow for targeted knockout of genes using homologous recombination. We will then verify this technique by first creating an auxotrophic strain that can be simply re-constituted with the cloned gene. Once verified this technique can then be harnessed to knockout genes that have been identified in the transcriptomic analysis below that will be beneficial for specific PUFA production.
3.1 Testing of RNAi technology
RNAi is a post-transcriptional gene silencing process that can be used to dampen, or silence, the expression of a target gene without the need for deletion or attenuation. Transformed cells expressing double stranded RNAi down regulate the gene of interest by causing the degradation of target RNA, provided there is 100% homology to the target sequence. This method has been demonstrated to work for the majority of fungi and provides distinct advantages over gene deletion methods, such as the ability to attenuate the expression of essential genes. In addition, it has also been demonstrated as an effective method in hetereokaryotic fungi with multi-nuclear hyphae such as the zygomycetes. This is of particular importance to this project as M. alpina is a member of the zygomycetes and has multi-nuclear hyphae. In this project RNAi is applied to silence either essential genes that maybe beneficial in boosting fatty acid production or for creating more efficient homologous recombination strains.
Using the pMORTv1 vector a construct has been created that contains an inverted repeat of a fatty acid gene from M. alpina. This construct once transformed will produce an RNA transcript should downregulate expression of this gene and therefore change the oil content of M. alpine (Takeno et al, 2005). The RNAi construct was synthesised by DNA2.0 and was cloned onto pMORTv1. This construct was then transformed into M. alpina and the fatty acid profile of the successful transformants was analysed. The analysed transformants showed no difference in their fatty acid profiles suggesting that this technique does not work.
3.2 Identification of Ku70/80 homlogs
To use the recently available whole genomic sequence to identify genes such as Ku70 and Ku80 that promote homologous recombination by homology mapping, and target these for down regulation using RNAi technology. Down regulation of ku70 promotes a higher frequency of homologous DNA recombination events and therefore increases transformation efficiency. We will apply this principal to M. alpina to create a strain that has down regulated ku70. This is predicted to improve the efficiency of homologous recombination events and therefore the efficiency of transformation procedures. At present there is no known homolog of Ku70/80 in M. alpina. Homology searches have yielded no results and therefore a degenerate primer approach has been undertaken to elucidate the genomic position of these genes. This work has been unsuccessful and due to time constraints within the project was Ku70/80 was not identified.
Creation of an auxotrophic strain using homologous recombination
To evaluate the feasibility of making gene knockouts in M. alpina a cassette was made for the deletion of a gene essential for biosynthesis of an essential metabolite. Transformation of the linear cassette was performed using the chemical transformation method described previously. In total 6 colonies where obtained from this transformation and all 6 where assessed for deletion of the gene. PCR of the genomic DNA from these colonies revealed that all 6 had a deletion of the gene and qPCR showed that it was a single copy integration event. All strains showed the same characteristics as the previously described auxotrophic strain. Re-transfromation of these strains with a cassette containing the wild type gene restored prototrophy and growth to wild-type levels. This finding is of vital importance for this project as it demonstrates the feasibility of creating gene knockouts without leaving any exogenous DNA. In addition, it provides an alternative method for strain construction and marker regeneration if the investigation of the Cre-Lox system proves unsuccessful. However, the efficiency of integration appears to be very low and the investigations into the Cre-Lox system maybe more efficient.
Creation of a gene knockout strain and testing of Cre-Lox
In order to test the Cre-Lox technique, a gene cassette containing LoxP sites needs to be transformed into the M. alpina genome. This cassette contains a LoxP flanked HygB as a selectable marker which will then be inserted between two DNA sequences identical to the flanking regions of the target gene as shown. When transforming this gene cassette, homologous recombination will occur where nucleotide sequences are exchanged between two similar or identical DNA fragments, replacing the target gene with the LoxP flanked HygB gene creating knockout strains. Two different approaches will be tested in this transformation: transform the full length cassette and transformation of two bipartite fragments with HygB gene overlap. Each bipartite fragment represents half of the full cassette from one homologous region till the HygB gene. Therefore HygB resistance can only be acquired when both of the fragments are integrated at the same locus thereby increasing the transformation efficiency. The inserted HygB gene can then be removed by introducing Cre-recombinase. If this proves successful it would allow the use of the same marker in multiple rounds of transformations, which is particularly useful because of the limited markers currently available for M. alpina. A major advantage of this method is that the removal of the marker gene used in transformations can eliminate the amount of unessential DNA that the cell carries.
The gene selected is an important enzyme on the path to the biosynthesis of arachidonic acid (ARA). Therefore removal of this gene will result in a decrease in the amount of arachidonic acid produced by M. alpina which allows for an easy identification of true knockouts via analysis of the arachidonic acid content of the cells.
The bipartite transformation generated a total of 5 colonies which were then assessed for the replacement of the gene by the cassette. Isolated genomic DNA was screened for the presence of the cassette and the gene by PCR. The results of this screen showed that the gene had been replaced by the cassette in all of the 5 colonies indicating that this method was viable for generating knockout strains.
After determining that the colonies obtained from the transformation were true knockouts the removal of the cassettes via Cre-Lox was assessed. To test this two strategies were assessed, the first was using transient transfection of the Cre-Lox plasmid whereby the plasmid was used in a transformation but without any selective pressure. The reasoning behind this approach is that it is advantageous to have the Crerecombinase activated for as short a time as possible to prevent random splicing events. The disadvantage is that the low transformation efficiency would equate to low Cre activity. Secondly, the plasmid would be transformed using the selective pressure of growth using the previously described auxotroph.
The results show that using the transient transformation procedure a confluent plate was observed. This was as expected as no selection pressure was used for the transformation. When the colonies from this plate were analysed for the presence of the plasmid it was found that no vector could be found. In addition, only 2% of all colonies analysed had the cassette removed via Cre-recombinase activity. However, once the addition of selection pressure is added to the transformation procedure 50% of all colonies contain the plasmid but only 20% of these i.e. 1 colony had the cassette removed. Removal was also confirmed by lack of growth on HygB containing media. These strains lack the target gene but also have no knockout cassette and therefore represent a “clean removal” of the gene.
Overall, these results show that the technique of Cre-lox removal of the cassette and therefore the selectable marker is feasible but is highly inefficient with only 1 – 2 colonies obtained per transformation.
After the removal of the knockout cassette via Cre-recombinase the wild type gene was cloned into the vector pMORTv2. This vector was then transformed into the cassette removed knockout strain to assess if activity could be returned to this strain. Once constructed all of the strains described in this section were sent for fatty acid analysis following shake flask fermentations.
The results of the removal of the gene via gene replacement reduced the amount of arachidonic acid the cells produced. This was also observed in the cells after Cre-recombinase removal of the knockout cassette, whereas re-introduction of the gene returned the level of arachidonic acid in the cells to WT levels. These results show that the creation of knockouts and marker recycling via Cre-recombinase is a viable genetic tool in M. alpina and could be used to perform multiple rounds of transformations with limited markers. However, it should be noted that this process is very inefficient and needs further optimisation. In conclusion we have shown for the first time that making knockouts and Cre-mediated marker removal in M. alpina is feasible and could therefore be used in the future to identify targets for improved product yield.
Aim 4: To use the newly acquired M. alpina genome sequence and microarrays to assess global transcript changes in response to various growth conditions in order to identify genes beneficial in the production of PUFAs. Previous biochemical work has already identified certain genes involved in PUFA production and regulation of PUFA production (malic enzyme). New microarray-based technologies now offer a powerful tool to identify genome-wide pathways mediating PUFA synthesis, with the existing prospect of discovering novel key genes. The microarrays (already available) will be used to assess global changes in the gene expression profile of M. alpina grown under various PUFA production conditions. The integration of these data with fermentation and production data and extensive metabolite analysis will then allow the identification of genes that are important in regulation of PUFA biosynthesis.
Identification of target genes
Transcriptomic data from production fermentations of M. alpina have identified a number of genes that are potential targets for modification. These include not only genes in fatty acid synthesis, but genes involved in degradation and co-factor and substrate availability. See aim 5 for more detail.
Aim 5: To use the molecular toolkit and microarray data to enhance the ability of M. alpina to synthesize PUFAs. The data from the microarray studies have identified a series of possible genes involving aspects of PUFA production. The new molecular tool kit developed in aims 1 and 2 will then be used as follows.
5.1 Manipulate the biosynthetic pathways of M. alpina to produce a plethora of strains producing predetermined or modified PUFAs as required.
Targeted genes
As described in aim 4 we have identified a number of genes for modification that will affect fatty acid synthesis. Vectors have been constructed to allow the overexpression of the a number of genes. The identified genes were cloned into pMORTv1 or pMORTv2. Once cloned the vectors were then transformed into M. alpina using the previously described transformation technique. Both vectors transformed with low false positives, i.e. 90% positive colonies. Fatty acid content was analysed for selected strains transformed with pMORTv1 or pMORTv2. Each strain was grown for 7 days in a scaled down fermentation process.
Overexpression of genes involved in fatty acid synthesis produced in all cases an increase in the amount of arachidonic acid. The largest increase observed was 76%. This indicates that there is scope for improvement in the yield of arachidonic acid that could be produced by this organism and suggests possible targets for a classical strain improvement program.
5.2 Targeted knockouts of the selected genes (e.g. those with high transcript changes) involved in PUFA biosynthesis from microarray analyses will be made to elucidate their function in this process.
Among the possible targets identified in aim 4 were genes other than biosynthesis genes of PUFA’s. We identified a specific target for deletion. Constructs were created as previously described and transformed into M. alpina. Resultant colonies were analyzed for the removal the gene and presence of the marker gene. Once again the efficiency of transformation was low, 5 colonies, but 100% of the colonies contained the cassette and were true knockouts. In addition, these knockouts were also made marker free via Cre-lox. This strain was then transformed with a newly constructed vector pMORTv2 plus the gene removed to test reconstitution. All of these strains were then analysed for arachidonic acid content following a shake flask fermentation In addition, a knockout strain was also transformed with pMORTv2 plus a gene involved in fatty acid synthesis to assess the effect of a knockout plus overexpression.
The results show that removal of the degradation gene increases the amount of arachidonic acid that is produced by the cell. It also shows that reintroducing the gene reduces the amount of ARA to wild type levels.
Significant results
Optimisation of a robust transformation procedure for M. alpina
Creation of an auxotrophic strain using targeted gene deletion
Creation of an auxotrophic strain using classical mutagenesis
Complementation of auxtrophic strains using targeted gene replacement
Expression of HygB gene from a heterologous promoter
Targeted gene deletion to improve oil yields
Altered the amount of oil produced via overexpression of genes
Training activities
SHE training
Extensive safety health and environment training has been undertaken with a focus on sustainability, safety at work and best practice in safety.
Fungal genetics training
In addition to working with M. alpina, the genetics of the many organisms that DSM utilise has been studied and applied to the current project.
This training has increased the fellow’s knowledge base and experience and will be of great benefit to his career.
Mentoring of a student
The fellow has undertaken the mentorship of a student and has gained valuable experience in leading a small research team.
Attendance at Fungal Genetics Conference
The fellow attended the Genetics Society of America fungal genetics conference.
Technology transfer and collaborations
The fellow has fostered close links with other DSM sites notably DSM Columbia where he visited for a week to transfer the transformation procedure.
Impact of the research
Fungi are widely used to commercially produce a very wide range of products that include enzymes, organic acids, pharmaceuticals and other chemicals. They are attractive as expression systems as they are capable of producing high yields and are relatively easy to culture and maintain. The results of the research described above has enhanced the understanding of PUFA production by M. alpina and allowed us to develop this fungus into an efficient and sustainable producer of arachidonic acid. The results described above have also unlocked the potential of this fungus to produce many other PUFAs which could be used in a number of applications from food technology to waterproof coatings. We have developed molecular tools and combined it with the transcriptome analysis to elucidate the biosynthetic pathways of PUFAs production and develop new strategies for improving the yields of PUFAs. This has afforded a better understanding and optimisation of existing industrial uses of M. alpina and is a prerequisite for the development of the strain as a microbial cell factory for novel future multiple PUFA production. As a consequence this research has led to the creation of a number of strains each of which produce a specific and high yield of PUFA that have the potential to be used in the food industry. Furthermore, due to the limited amount of information that is available on PUFA production in M. alpina the results of this research have greatly enhanced the understanding of the biosynthetic pathways involved including key regulators, ratelimiting steps and gene expression levels during PUFA production.
Our development of a genetic tool kit for M. alpina in conjunction with our transcriptomic analysis is a major contribution to the understanding of the biochemistry and molecular biology of oleaginous fungi and a key step towards sustainable and efficient production of PUFAs. The specific results of the project can be summarised as follows:
- New molecular tools for manipulation of an industrially relevant organism
- The first transcriptomic analysis of an oleaginous fungi
- Development of M. alpina as a microbial cell factory for the production of various PUFAs
- An improved understanding of the basic metabolism underpinning PUFA production in oleaginous fungi.
PUFAs have many beneficial effects in cardiovascular health, infant nutrition and disease prevention when given as a dietary supplement. In addition, an imbalance of n-3 and n-6 PUFAs in the diet has been implicated in a number of disease states. These factors, and the questions surrounding the sustainability of current production methods for PUFAs, mean that the development of robust and efficient systems for essential fatty acid production is of paramount importance. This project has used the zygomycete fungus M. alpina, (GRAS accredited) already a reliable producer of the C20 PUFA arachidonic acid, to develop a molecular tool kit and unlock its potential as a cell factory for PUFA production. The new genetic tools developed will allow the modification of this organism to direct the synthesis of PUFAs and therefore generate a sustainable platform technology for the production of these essential 'health giving' compounds.
Such a platform will increase the availability of PUFAs on a global scale. In addition, this would place Europe to the front end of science in this field. Much work is ongoing in Japan to improve lipid production in microorganisms, but the development of a robust, and versatile microbial lipid factory has not been achieved so far. Such a microbial lipid factory could also be an invaluable tool for other European research scientists who are interested in contributing to the further development of such a microbial factory not only in the field of strain construction, but also in the field of fermentation and down stream processing. By setting up such a lipid production platform, synergies are expected with nutritional institutes who can study potential health effects of different PUFAs in a variety of applications. The possibility to tailor make PUFAs and combinations of PUFAs will trigger studies on synergetic effects, counteraction and competitive inhibition of PUFAs on neuronal function and inflammation.
Furthermore, there is the very real threat that an obesity epidemic is about to hit the European union with obesity rates among men at 10 – 27% and up to 38% in women (International Obesity Task Force., 2005). Consequently the rates of morbidity from cardiovascular disease will increase. Therefore the awareness level of people regarding food consumption particularly regarding the health benefits of PUFAs in cardiovascular disease prevention should be increased. The construction of a microbial cell factory that can be made to make various kinds of PUFAs would lead to the production of new products and formulations for existing food and healthcare products that will lead to greater public awareness of the benefits of these compounds for optimal well being.
Comparison of results with initial timelines
Aim 1: Testing transformation systems
Progress: Assessed electroporation and chemical transformation methods. Agrobacterium method was not assessed due to worries about residual bacterial DNA and the efficiency of the chemical transformation method. Optimised the chemical transformation method. This aim has been successfully achieved.
Aim 2: Development of a counter-selectable marker
Progress: Uracil auxotroph created via two methods. Aim achieved.
Aim 3: Improve homologous recombination efficiency
Progress: Achieved targeted gene deletion of the URA3 gene therefore demonstartiing the feasibility of creating knockouts in M. alpina. RNAi experiments and Ku70 identification proved to be problematic and were unable to be completed successfully. Aim achieved, however with mixed results.
Aim 4: Microarray analysis
Progress: Targets for overexpression and deletion were identified and constructs made. Aim achieved.
Aim 5: Strain improvement using recombinant techniques.
Progress: Targets identified from aim 4 were either overexpressed or knocked out. These newly constructed strains were then assessed for fatty acid content. In each of the strains, detailed in below the fatty acid content of the organism had been altered. In addition, the knockout strains were constructed in such a way that the Cre-lox method of gene excision could be tested. This would allow the recycling of selectable markers. This method was shown to be viable. Aim achieved.