A note added to a chapter of the 1986 Annual Report (Lien and Roessler 1986) described preliminary data on the use of Si deficiency to trigger lipid accumulation in diatoms. Silicon is major component of diatom cell walls. Similar to the lipid trigger effect produced by N — deficiency, Si depletion also results in a decrease in cell growth and often is accompanied by an accumulation of lipid within the cells. However, Si (unlike N) is not a component of other cellular macromolecules (enzymes, membranes) or cell structures such as the photosynthetic

apparatus. Therefore, any changes in cellular biochemistry and lipid accumulation induced by Si deficiency might be more easily interpreted than changes induced by N starvation. This work initiated a series of experiments by Paul Roessler during the late 1980s and early 1990s on the biochemistry and molecular biology of lipid accumulation in Si-deficient diatoms.

The first set of experiments compared the effects of Si deficiency on lipid accumulation and cell physiology in several species of diatoms, including C. cryptica T13L, Thalassiosirapseudonana, and Cylindrothecafusiformis. Exponentially growing cultures were transferred to media that contained either excess Si or limited levels of Si so that the media became Si deficient while the cells were still growing exponentially. Cell growth, chlorophyll a content, AFDW, lipid, and photosynthetic capacity were monitored under both conditions. In all three species, cell division decreased as soon as the Si was depleted in the media. However the species responded differently with respect to other physiological parameters. In C. cryptica, chlorophyll a synthesis was almost completely inhibited after 12 hrs in Si-depleted media; C. fusiformis showed little change in chlorophyll a synthesis after 72 hrs. T. pseudonana exhibited an intermediate effect, with some decrease in chlorophyll a synthesis noted after 36 hours without Si. The effect on photosynthetic capacity, measured as O2 evolution, also varied between the three species. In C. fusiformis and C. cryptica, photosynthetic capacity decreased 33% and 58%, respectively, after 12 hours; T. pseudonana showed a steady decline in photosynthetic capability following Si — depletion. (However, photosynthetic capacity decreased in Si-replete cultures as well during the 72 hours time course of the experiment, presumably due to the increased ratio of antenna chlorophyll molecules versus reaction center molecules in the self-shaded, dense cultures).

The three species were also analyzed for accumulation of total biomass and lipid (Figure II. B.3.). In C. fusiformis, biomass accumulation (measured as AFDW) for the duration of the experiment was similar in cultures with or without sufficient Si, although lipids made up a higher percentage of the AFDW in the Si-deficient cultures (26% versus 21% in Si-replete cells). In T. pseudonana, synthesis of cell mass and lipid was not affected until 36 hours after Si depletion. At this point, biomass and lipid accumulation rates decreased; however, there was little difference in the percentage of total lipid in the cells with or without Si at the end of the 72 hours experimental period. The situation with C. cryptica was very different. Twelve hours following Si depletion, there was a 38% decrease in the growth rate of these cells compared to the Si — replete culture. However, lipid synthesis continued at the same rate in the Si-deficient cells as in the Si-replete cells, resulting in a significant increase in the lipid content of the Si-starved cells. Interestingly, after these initial changes, the Si deficient cultures of C. cryptica showed little gain in total AFDW or lipid during the remaining 72 hours of the experiment.

In order to determine if Si deprivation affected the composition of the lipids produced, the lipids were extracted and analyzed for the percentage of polar versus neutral lipids present. In all three species, the Si-deficient cultures showed a significant increase in the level of neutral lipids, primarily TAGs. For example, the percentage of neutral lipids in Si-deficient cultures of C. cryptica was 64%, compared to 32% in Si-replete cultures. In C. fusiformis, the percentage of neutral lipids increased from 17%-20% to 57% in Si-deprived cultures.

Based on these studies, C. cryptica was identified as the best candidate for further studies on the biochemistry of lipid accumulation. To determine the effects of Si deficiency on the synthesis of the cell components, the levels of protein, carbohydrate, and lipid were examined at various times after Si was depleted in the cultures. During the first 12 hours, protein and carbohydrate synthesis decreased. Lipid accumulation continued at a rate similar to that of the Si-replete cultures. This resulted in an increase in lipid content of the Si deficient cells from 19% to 27%. This observation was confirmed in subsequent studies that followed the incorporation of newly assimilated carbon (as H14CO3-) into the various cell components. Si depletion resulted in a net decrease in the rate of photosynthesis and carbon assimilation, but the individual cell fractions were affected differently. For example, the rate of 14C accumulation into lipids decreased by 48% in the first 4 hours of Si-deprivation; the uptake of 14C into chrysolaminarin, the major carbohydrate storage product in diatoms, decreased 84%. Therefore, the increase in lipid content of Si-deficient cells was not due to an increase in the rate of lipid synthesis, but to a relative decrease in the rate of synthesis of protein and carbohydrate.

Pulse-chase experiments were performed to test whether Si deficiency also caused the conversion of non-lipid cellular components into lipids. In these experiments, Si-replete cells were labeled with H14CO3- for 1 hour, then transferred into Si-deficient media without labeled bicarbonate. The amount of labeled carbon in the lipid fraction was determined at various times following transfer to Si-free media. This experiment showed that carbon was slowly redistributed from the nonlipid components of the cells into lipid under Si-deficient conditions, but not under Si-replete conditions. Therefore, the accumulation of lipids in diatoms in response to Si-deficiency is apparently due to two factors:

1. An increase in the proportion (but not the net amount) of newly assimilated carbon that is incorporated into lipids, resulting from a disproportionate decrease in the rate of lipid synthesis versus carbohydrate synthesis, and

2. A slow conversion of nonlipid cell material into lipids.

Fractionation of the lipids produced demonstrated that Si deprivation resulted in an increase in the proportion of total lipid as neutral lipids, primarily TAGs, from 43% to 63% after only 4 hours of Si deficiency. Analysis of the fatty acid composition of the accumulated lipids also showed changes induced by Si starvation. In Si-deficient cells, there was an increase in the proportions of mono — and unsaturated fatty acids (16:1, palmitoleic acid; 16:0, palmitic acid; and 14:0, myristic acid), and a reduction in the proportions of the three major polyunsaturated fatty acids, (16:3, 20:5, and 22:6). These results are consistent with the finding that the predominant fatty acids found in triacylglycerol storage lipids in C. cryptica are 16:1, 16:0, and 14:1. These shorter, more highly saturated fatty acids are also the most desirable substrates for conversion into fatty acid methyl esters (biodiesel), as they would be less likely to polymerize during combustion and “gum up” an engine.

Although Si depletion causes all diatoms tested to stop dividing, species responded differently with repsect to continued accumulation of biomass and lipid. C. cryptica showed a rapid

response to Si-depletion, with a decrease in growth accompanied by a significant increase in the proportion of the biomass as lipid within 12 hours (the response to N starvation was usually much slower, as the cells could utilize internal N stores). This result again emphasizes the need to understand the kinetics of lipid accumulation in individual species under specific conditions for cost-effective lipid production in the ponds.

Figure II. B.3. Changes in lipid mass, ash-free dry mass, and lipid content in Si-deficient cultures of three diatoms.

A. C. fusiformis B. C. cryptica C. T. pseudonana. Symbols: (■) Si-deficient cultures; (•) Si-replete cultures.

To better understand the processes involved in lipid accumulation in microalgae, and to identify potential molecular targets for genetic manipulation, studies were initiated to examine the effects of Si deficiency on the enzymatic pathways involved in lipid and carbohydrate synthesis in C. cryptica. One possibility is that the increased levels of storage lipid in cells exposed to Si starvation could result from shifts in the relative activities of one or more enzymes in the lipid biosynthesis pathway. Acetyl-coenzyme A (acetyl-CoA) is known to be the immediate precursor of fatty acid synthesis, but the source of this compound varies in different organisms. For example, in mammalian cells, acetyl-CoA used in cytosolic fatty acid synthesis is produced from citrate via the action of ATP citrate lyase. In plants, acetyl-CoA can be produced in the chloroplasts from pyruvate, catalyzed by pyruvate dehydrogenase. Alternatively, acetyl-CoA could be produced by the mitochondrial pyruvate dehydrogenase. In this case, the acetyl-CoA (which cannot diffuse across the organellar membranes) would be broken down to acetate and free CoA by acetyl-CoA hydrolase. Acetate would diffuse to the chloroplast and become incorporated into acetyl-CoA by the action of acetyl-CoA synthetase. Once acetyl-CoA is produced, it is then used as a substrate by acetyl-CoA carboxylase (ACCase) to produce malonyl CoA. Malonyl-CoA is a substrate for fatty acid synthase and this reaction is considered to be the first committed step in fatty acid synthesis.

These pathways had not previously been well-characterized in diatoms. To better understand the lipid synthesis pathways, Roessler first looked for the presence of these enzymes in extracts of C. cryptica, but found no citrate lyase activity. However, acetyl-CoA hydrolase, acetyl-CoA synthetase, and ACCase activity were all present. Enzyme activities were studied in Si-replete and Si-deficient cells (Figure II. B.4). The level of acetyl-CoA synthetase activity was similar under both conditions; however, the level of ACCase activity was two fold higher in Si-deficient cells after 4 hours, and four fold higher after 15 hours. Based on subsequent studies using protein synthesis inhibitors, the increased specific activity of the ACCase was believed to result from an increase in expression of the ACCase gene (Roessler 1988a; 1988c).

ACCase is a biotin-containing enzyme that catalyses the carboxylation of acetyl-CoA to form malonyl-CoA. This reaction entails two partial reactions: the carboxylation of biotin, followed by the transfer of the carboxyl group from biotin to acetyl-CoA. In bacteria, the enzyme is composed of four non-identical subunits. However, in eukaryotes, biotin binding, biotin carboxylation, and carboxyl-transfer all occur on a single large multifunctional protein; the functional ACCase occurs as a multimer of this polypeptide. ACCase had previously been shown to play a key regulatory role in the rates of fatty acid synthesis in both animal and plant systems. A project was initiated to isolate and characterize ACCase from C. cryptica to clarify the role of this enzyme in lipid accumulation induced by Si starvation, and to compare the microalgal enzyme with those isolated from plants, animals, yeast, and bacteria.

The enzyme was purified from C. cryptica by a combination of (NH4)2SO4 precipitation, gel filtration chromatography, and affinity chromatography based on the affinity of biotin to avidin. Consistent with ACCase enzymes isolated from other eukaryotes, C. cryptica ACCase was found
to consist of a homo-tetramer of 185 kDa subunits. The activity of the enzyme was assayed by the incorporation of 14C bicarbonate into malonyl-CoA, and other factors were identified that affect the stability and activity of the enzyme. As seen for other ACCases, the enzyme required a slightly alkaline pH for optimum activity (pH 8.2), although the enzyme was most stable when stored at pH 6.5. The enzyme was also stabilized by sulfhydryl reductants (i. e., dithiothreitol), citrate, NaCl, and KCl; divalent metal cations (Mg2+ or Mn2+) were required for activity. A number of cellular metabolites were also tested for their affects on ACCase activity. The enzyme was inhibited by products of the ACCase reaction, including malonyl-CoA, ADP, and NaH2PO4, and also by palmitoyl-CoA, but it was not affected by various glycolytic or photosynthetic intermediates or by free CoA. Two herbicides that inhibit ACCases from monocot plants were also had little or no effect on C. cryptica ACCase. Thus, the ACCase from this diatom was found to be similar to higher plant ACCase enzymes in that it is composed of multiple, identical, multifunctional subunits. In addition, the Kms for the ACCase substrates (acetyl-CoA, MgATP, and bicarbonate) in C. cryptica were similar to those found in plant ACCase enzymes (Roessler 1989; 1990).