Several oleaginous microalgal strains had been identified as potential candidates for biodiesel fuel production. These organisms became the target of genetic engineering efforts to manipulate the lipid biosynthetic pathways. Before the work on genetic transformation of algae at NREL, very little information was available on the molecular biology of these organisms. One of the first steps was to develop techniques to isolate and purify DNA from these organisms. A desirable protocol would disrupt the cell wall using methods gentle enough to prevent shearing of the genomic DNA. This was not trivial for some species, such as Monoraphidium, which has a very resistant wall that contains sporopollenin. A method that worked for most species tested (described in Jarvis et al. 1992) was developed based on a protocol used to isolate yeast DNA (Hoffman and Winston 1987). The cells were suspended in buffer that contained 2% Triton X — 100 and 1% SDS, then added to a tube that contained glass beads and an equal volume of phenol:cholorform:isoamyl alcohol (PCI). The cells were agitated for 1 minute using a vortex mixer. The DNA in the aqueous phase was purified by re-extraction with PCI, ethanol precipitation, and treated with RNase A. For some species, the DNA had to be purified further by using precipitation with hexadecyltrimethylammonium bromide (CTAB; Murray and Thompson 1980) to remove contaminating carbohydrates or by purifying the DNA on CsCl gradients. This procedure produced DNA that digested well with many common restriction endonucleases, but even highly purified DNA would not digest well with all restriction enzymes.

NREL researcher Eric Jarvis theorized that poor digestion of the DNA by some enzymes could be attributable to characteristics of the DNA. All DNA is composed of four nucleosides; deoxycytidine, deoxyguanosine, deoxythymidine, and deoxyadenosine, (abbreviated dC, dG, dT, dA); in double stranded DNA, dC is always paired with dG, and dT with dA. The percentage of each nucleoside (often represented as %GC) is variable between species. Restriction enzymes cut DNA at specific nucleotide sequences, generally recognizing 4-6 bp motifs. Therefore, the frequency of cutting by a particular enzyme will be affected by the total nucleotide composition of the DNA (i. e., an enzyme that recognizes CCGG would cut infrequently in an organism with a low %GC). The GC content is also reflected in the codon usage by each organism, as DNA with a high GC content would show a bias toward codons ending with G or C in the variable third position. DNA can also contain unusual modified nucleosides, including 5- hydroxymethyldeoxycytidine (hm5dC) and 5-hydroxymethyldeoxyuridine (hm5dU), although the biological significance is unclear. Another common modification is the presence of methylated nucleosides, in particular 5-methyldeoxycytidine (m5dC) and 6-methyldeoxyadenosine (m6dA). The degree of methylation has been associated with levels of gene expression. In addition, some microorganisms use DNA methylation as a defense mechanism, in that methylated DNA sequences are often not recognized by endonucleases from invading pathogens. Although the presence of methylated nucleosides is characteristic for some species, the degree of methylation can vary on a short time scale with changing environmental conditions. In contrast, the %GC and presence of modified nucleosides are characteristic for a particular organism. These characteristics only on an evolutionary time scale.

DNA was isolated from microalgae strains, including 10 species from 5 classes. The nucleoside composition was analyzed by reverse-phase HPLC and by digestion with restriction endonucleases. The results of the HPLC analysis are summarized in Table I. B.4-1. Although the diatoms showed a GC content typical for most eukaryotes (42%-48% GC), the GC content of the green algae (excepting Stichococcus) was significantly higher. In particular, Monoraphidium DNA contains 71% GC. The table also shows the presence of m5dC in the algal DNA. All species tested contained some level of this modified base, although once again Monoraphidium stands out with approximately 11% m5dC. The only other unusual feature was the presence of 12% hm5dU in the dinoflagellate C. cohnii (data not shown); dinoflagellates were not considered to be good candidates for biodiesel fuel production, so this observation was not explored further.

These data provided a good background for developing genetic transformation systems for these organisms. As mentioned above, the GC content of an organism can be reflected in the codon usage, suggesting that an organism with a high GC content such as Monoraphidium may not successfully express heterologous marker genes. This was found to be true for the green alga Chlamydomonas; successful transformation of this organism was achieved only by the use of homologous selectable markers (discussed in more detail later). Also, GC content should be considered when designing synthetic DNA probes based on protein sequences, i. e., for isolation of algal genes by PCR. In addition, DNA methylation can affect the ability to construct DNA libraries and to clone algal DNA, and may require the use of bacterial host strains that are insensitive to DNA methylation.