Iron Solutions: Developing New Algal Growth Media for Increased Iron Uptake

Andrew Sweeney

U.S. Department of Energy Office of Science, Science Undergraduate Laboratory
Internship (SULI)

University of California San Diego

Lawrence Berkeley National Laboratory
Berkeley, California

August 6, 2015

Prepared in partial fulfillment of the requirements of the U.S. Department of Energy Office of Science, Science Undergraduate Laboratory Internship (SULI) under the direction of Dr. Nigel Quinn in the Earth Sciences Division at Lawrence Berkeley National Laboratory.
ABSTRACT

This study endeavored to improve sustained productivity of mass cultivated marine microalgae by using limitation of iron, a vital micronutrient, to create a growth medium that would prevent the growth of non-predatory invasive organisms. Iron’s aqueous chemistry is quite complex, and much of this study is focused on the chemical transformations of iron chelates and iron salts in the growth medium my group developed for Nannochloropsis oculata.. This algae has been identified ,because of its high proportion of unsaturated lipids, as a promising candidate for biofuels, specialty chemicals, and protein rich animal feed. Nannochloropsis oculata. also promises to be resource efficient as the cell’s small size ,and minimal agitation requirement, minimizes the loss of inorganic carbon through escaping CO2. The cells were grown in four different media (iron free, 30uM FeEDTA, 10um ferrous sulfate, and a combination of the two). The growth of the N oculata. was primarily measured using optical density testing during the logarithmic growth phase. After one week of growth the highest biomass density, 0.86, observed in the culture grown in 30uM ferric EDTA, was more than 2.5 times the lowest, 0.333, which was found in the culture containing 10uM ferrous sulfate, and substantially more than the no iron control solution which maxed out at 0.503. The 30uM ferric EDTA sample also had the lowest proportion of living cells at the end of the experiment, with a photo oximeter reading of only 41.5 initially, dropping to 18 after 10 minutes, and 8.5 after 20. The 10uM sulfate medium on the other hand started at 55, dropped to 34.4 after 8 minutes in the dark and climbed back to 50.9 after 12 minutes in the light. Both paled in comparison to the healthy cell culture which went from 90.2 to 71.7 in 5 minutes, and up to 144 in another 7. The larger scale test utilizing 30uM FeEDTA, the most commonly used iron compound, however, showed that the very little of this iron actually entered the cells, most simply clung to the inside of the flask. The results of this study also show the risk of excessive iron concentration leading to culture crash at the end of exponential growth, and iron toxicity from overconcentration. Further work is needed to develop improved ferrous iron compositions to selectively feed targeted microalgae in open ponds, both to increase the efficiency of iron uptake, and decrease the degree of contamination.

INTRODUCTION
It is no secret that the planet is running out of fossil fuels, and as the human population continues to grow and industrialize, the rate at which the supplies dwindle will only increase. Given these circumstances, biofuels are becoming an increasingly viable replacement, with the added benefit of carbon neutrality. Traditional biofuel crops, however such as oil palms and soybeans, have to compete with food crops for the limited supply of arable land. One possible solution to this dilemma is to produce biofuels from microalgae which, can have a lipid content of over 30%, yield up to 23,000 liters of oil annually per acre, and do not require fertile land to grow1. Microalgae, photosynthetic organisms that range in size from 2-4 um, also grow rapidly, are able to utilize agricultural runoff for nutrients, and, after being squeezed for lipids, the remaining biomass can be used to create many other products, such as protein rich animal feed2
For algal biofuels to be practical the algae must be able to both accumulate biomass rapidly, and produce a high percentage of unsaturated or polyunsaturated lipids. This proves problematic because, in their natural state, algae typically accumulate either biomass or lipids at the expense of the other1. If cultivated under stressful conditions, however, such as abnormal pH, high or low temperatures, or micronutrient limitation, the surviving algae will rapidly increase their lipid concentrations2. While more oil per acre can be generated, and it is much easier to maintain the necessary sterility for a good algae harvest in a closed bioreactor system, the high cost of this method means that open ponds are likely to be the method of choice for large scale biofuel operations. However open ponds have a high risk of contamination by bacteria, fungi, predatory protozoa and other algae. While microalgae do have defense mechanisms against these organisms, the cultivation process can also be tailored to enhance their resistance to contamination2. One of the most promising methods is the feast and famine method, growing a batch of algae in an iron rich solution, allowing them to absorb large quantities of luxury iron, before being transferred to an iron free solution2. The algae will be able to continue to grow through several doubling cycles using the iron they have stored, while any non-predatory organisms will not be able to survive in the depleted media3. As such it is theoretically possible to control culture contamination, by using nutrient-loaded algae as an inoculum in nutrient deprived growth media.

Iron is vital for algal growth as it facilitates nitrate/nitrite reduction, N2 fixation, and photosynthetic and respiratory electron transport4. It is however insoluble in most aerobic growth media, so the solutions currently in use rely on Fe(III)-EDTA or similar soluble chelates to provide the algae with an accessible source of this micronutrient5. Unfortunately, only a small proportion of the iron in these ferric solutions is released from the chelates, the rest remains locked up in the chelate and unusable, or forms insoluble ferric oxide after being released6. Not only does this iron wastage increase production costs, cell aggregation promoted by iron oxide/hydroxide colloids may occur with adverse effects on growth and product recovery ie. the algae retaining sufficiently high concentrations of unusable iron that their lipids cannot be refined into functional fuel. Some algal cells can, however directly uptake ferrous iron sources across the plasma membrane without the need for reduction of traces of “free” ferric species, or risk of iron depletion due to precipitation encountered with ferric sources7.
The species used in this experiment, Nannochloropsis oculata., is very similar to
Nannochloropsis oceanica., supplied by Cellana, the species used in Emma Tacardi’s experiment. Individual organisms are non-motile, completely lack chlorophyll a., have a simple ultrastructure compared to neighboring taxa, and are spherical with a diameter of between two and three micrometers8. N oculata. is an attractive species for biofuel cultivation because unlike many potential competitors it is capable of growing in sea water. Like most marine algae N oculata also continues to store iron throughout its exponential growth phase, which allows it to outcompete other, undesirable species, in the stationary (logarithmic) growth phase3. This species may be amenable to genetic modification, and even under normal conditions, accumulate large quantities, over 30% total biomass, of polyunsaturated fatty acids1. Its small size enables it to tolerate a near stagnant environment, and the spherical shape allows for reliable packed volume measurements.

MATERIALS AND METHODS

Packed Volume
The accumulation of algal biomass in the ten liter tank was tracked using the packed volume method which involved centrifuging 20ml of solution, pouring off the majority of the supernatant, transferring the remainder of the sample to 1ml Eppendorf tubes, and then centrifuging them at 10,000 rpm for two additional minutes to concentrate the biomass. The tubes were subsequently rotated 180 degrees and spun for another two minutes before the remaining biomass was combined into one Eppendorf tube, and centrifuged again. The solid biomass was re-suspended as a thick slurry, allowing it to be drawn up a hematocrit capillary tube. After two minutes in a clinical hematocrit centrifuge the length of the remaining biomass in the tube was measured, and a photograph taken to determine the health of the cells by studying the color and thickness of the layers of biomass. Dilution factors were also recorded to determine packed volumes of pre-centrifuged biomass with a conversion factor of 1ul per mm.
Choosing the sample
Three variables, the degree of contamination and aggregation, the pH (had to be below 9), and the clarity of the supernatant were used to identify the most suitable of the existing Nannochloropsis cultures to inoculate the experimental flasks. The pH was tested using a standard electrode, the degree of contamination and aggregation was analyzed by studying the samples under a microscope, and the clarity was measured using the lab’s spectrophotometer. After the best culture was identified two 40ml samples were repeatedly centrifuged, and washed with iron free seawater after each round to remove any remaining iron. The depletion of extracellular iron was confirmed with subsequently described our Ferene assay.

Iron (Ferene Assay)

The initial test for the amount of iron remaining in the algae involved measuring the packed volume and using the established carbon-iron and carbon-biomass ratios of N. Occulata to calculate the quantity of iron in the sample. For a more refined estimate we used a spectrophotometer to determine the absorbance of a one ml sample. The ferene assay consisted of 0.1ml of 2.5M sodium acetate, 0.1ml of 0.4M ascorbate, 0.05ml of 20mM ferene, and 0.75ml of the culture being measured. Before measuring the sample’s absorbance a baseline was set by making the same 1ml solution with iron free, artificial sea water. After the baseline was set the culture sample was transferred to a cuvette and tested with the spectrophotometer.

Oximeter Testing
The rate at which the ferrous iron is oxidized in the stock solution and the algal cultures was determined by measuring the rate of O2 consumption. A 2ml sample of either solution or culture was pipetted into the oximeter cuvet and stirred at a constant rate. After the O2 concentration stabilized the sample was exposed to light to measure the rate of increase brought about by photosynthesis. After the concentration reached a peak the solution was shrouded, and the per minute rate of O2 consumption via respiration was recorded (Table 3). After each experiment the oximeter was washed out, first with a dilute HCl solution then with DI water, to remove any iron remaining in the test chamber.
Optical Density Optical density was used as a rapid proxy measurement of biomass density, suitable for determining the daily growth of low volume samples. 0.5ml of culture and 0.5ml of supernatant were mixed, transferred a cuvette, and spectrophotometer readings were taken at 2, 5, 10, 30, 60, and 180 minutes (Figure 1).

RESULTS
Iron Precipitation

The rate of iron loss through precipitation in the biomass free growth medium was measured daily over the course of 17 days using the ferene assay (Figure 2). A substantial amount of iron precipitated out in the first 24 hours, with the spectrophotometer reading falling from 0.412 to 0.177 at the ten minute mark (Table 2). Over the next four days, however absorbance only dropped to 0.108, and for the following ten days the solution’s absorbance largely stabilized, with a reading of 1.16 on day 17.
Degree of Absorbance
The degree of absorbance was found to be the most suitable method of day to day biomass estimation. As the spectrophotometer is less accurate at measuring absorbance levels greater than one, a solution that was 0.5ml supernatant and 0.5ml culture was used. The readings were taken at 654nm, the point of minimum absorbance for the solutions in question (Figure 1). The preliminary results demonstrated that the cells supplied with iron using the tried and true chemical, FeEDTA, had a greater quantity of biomass after a week’s stable growth (absorbance 1.036) than the iron free solution (absorbance 0.656), and a far greater quantity than the ferrous sulfate (absorbance 0.46) that was being tested as a potential replacement.
Oxygen Consumption (Quantity of live cells)
The final oximeter test was conducted two weeks after the flasks were inoculated with N. Occulata, the stable growth phase for some cultures, and the death phase for others. Of the four cultures, only two, the 10uM ferrous sulfate and the iron free, showed any sign of oxygen production through photosynthesis (Figure 3). These two cultures rose from lows of 34.4 and 26.6 in darkness, to highs of 50.9 and 41.7, respectively, after being exposed to light. The 10uM sulfate medium started at 55, dropped at a rate of 2.65 per minute for 8 minutes to 34.4 in the dark and climbed at a rate of 1.375 per minute for 12 minutes in the light back to 50.9 (Figure 3). The 30uM ferric EDTA sample on the other hand had the lowest proportion of living cells at the end of the experiment, with photo oximeter readings that dropped by an average of 1.65 per minute over 20 minutes, and continued to decline when exposed to light. Both paled in comparison to the healthy cell culture which went from 90.2 to 71.7 in 5 minutes, 3.7 per min, and up to 144, 10.33 per min, in another 7.

DISCUSSION
Although practical algal biofuel production is still a few years down the road, the study detailed in this paper, and those to follow, will help continue to advance the search for a practical means of providing algae with luxury iron while creating usable fuel. In terms of absorbance and overall biomass growth during the steady growth phase, FeEDTA proved to be the most effective compound for delivering iron to the algae, its overall absorbance reading was nearly twice that of the no iron solution, and close to triple that of the ferrous sulfate after two weeks (Figure 1). The difference was even more pronounced a week earlier at the end of the exponential growth phase, where the FeEDTA culture was deep green while the sulfate culture was pale, and somewhat translucent (Figure 4). All that being said the results of the final oximeter test proved that the 30uM FeEDTA solution was an unsuitable candidate for scaling up to long term biofuel production, as after two weeks the culture had crashed, and there were virtually no living cells in the FeEDTA culture. Given that 30uM is an excessive concentration of iron, more than two orders of magnitude greater than normal seawater, it is entirely possible that oxidative stress, lead to a combination of chlorophyll a and protein biosynthesis inhibition, and ferritin synthesis, which caused the death of the culture6. Indeed a similar experiment by Xuxiong Huang et al. recorded the greatest growth of N Oculata. biomass in a solution of 1.2 x 10-1 FeEDTA with slightly lower growth in a 1.2 x 10-2 solution, and substantially lower growth in the more concentrated 1.2 and 12 mmolar solutions6. Although iron is not the most costly component of growth media, the need to keep it in solution as ferric chelates or simple ferric salts subject to gradual chemical transformations that can include toxic species, argues for developing efficient growth media that minimize the likelihood of iron overload or overt toxicity. Also, as only a small proportion of the iron actually entered the cells in our experimental cultures, the need for more effective targeting and transfer of iron nutrients to inoculated cells has become readily apparent. After the iron loaded algae had been transferred to their new medium, several tests were run to determine the quantity of intercellular iron. The concentration of biomass was too low to accurately measure the iron contained within the cells directly, and, while there was iron in the supernatant, it only registered an absorbance of 0.54 after 150 minutes despite the solution being concentrated nearly 50 fold from 40ml to 0.84ml. A stock solution of only 30uM would typically have a reading of around 1.0 after a similar interval. The majority of the iron in the solution was found to be clinging to the inside of the flask, as when tested with ferene there was a strong reaction, with the solution immediately turning a deep blue as it would in the presence of a 3mm iron solution. The ferrous sulfate did not provide the degree of rapid growth shown by the ferric EDTA, the absorbance was only 0.666 after one week, and 0.46 after two as opposed to 0.694 and 1.036, (Table 1) and visually the ferric solution was a deep green while the ferrous was yellowish and translucent (Figure 4). However the ferrous solution had a considerably higher quantity of living biomass at the end of the experiment with a O2 concentration of 50.9 after 20 minutes compared to the ferric solution’s 8.5 (Table 3). This indicates that while the ferrous solution had not suffered a complete culture crash from excess growth, some biomass had died off over the course of the second week. Both ferrous sulfate’s suitability and the idea of iron poisoning were substantiated in a brief follow on experiment, where after a week of growth the 1uM sulfate solution showed a substantially greater oxygen consumption, and therefore density of live cells, than the iron free solution.
When it came to measuring biomass the packed volume and optical density tests proved more useful than the dry weight method. While very accurate, dry weight testing would have taken considerably more time than the other methods, and would also have required accurate measurements of the evaporation from the flasks in order to determine the weight of the formerly dissolved ions in each sample. The packed volume test averted this issue as all of the ions remained in the supernatant. As both the carbon to iron and carbon to total biomass ratios of N oculata. are known, it was simple to determine the amount of biomass in each sample. The packed volume test had the added benefit of separating the healthy and dead biomass, which provided insight into the overall health of the culture rather than just its growth potential, and further marked FeEDTA as an unsuitable means of delivering iron. The packed volume assessment unfortunately required a minimum of 20ml of solution per test, meaning that at the 75ml scale it could only be run a couple of times per trial. This method is also somewhat unrefined as biomass often remained stuck to the bottom of the capillary tube, or the tube gets plugged with fully solid material though it is quick, simple, and requires few pieces of equipment. The process also does a good job of stratifying the sample, though the irregular spread in the capillary tube sometimes made it difficult to distinguish between dead and healthy cells. For day to day biomass measurements the optical density test proved a quick and accurate means of measuring the comparative densities of each culture.

Conclusions
The results from this study indicate that while FeEDTA is the most effective compound for delivering iron to the N oculata. culture in terms of initial biomass growth. However within two weeks the original inoculum had heavily aggregated and almost completely died off. The data from the other samples followed a similar, though less dramatic, pattern, exponential growth in the first week followed by biomass death in the second, though the ferrous sulfate solution had many surviving cells. This indicates that both solutions were unsustainable at the concentrations tested. The follow on experiment with 1uM ferrous sulfate, however showed sustainable growth, and lent credence to the feast and famine method as a whole. It will be helpful to conduct additional tests to determine the optimal, initial ferrous sulfate concentration, and whether or not it needs to be adjusted as the experiment progresses.
Future experiments to determine the optimum growth medium for iron transfer should focus on differing concentrations of ferrous iron sources, and on testing at lower pH levels to further reduce the risk of contamination. If issues with iron absorbance are not addressed then a combination of low biomass density, and iron laden oil will ensure that progress toward commercially viable algae production will be limited.
As a final note the development of the ferene assay proved to be a resounding success. The procedure allowed for accurate and rapid determination of the amount of iron remaining in a culture, by measuring the absorbance of a solution consisting of the culture sample, ascorbate, sodium acetate and ferene. The distribution of this process may well allow labs that cannot afford the equipment to perform iron radioisotope tests to build on this experiment.

Acknowledgements
I would like to thank my mentor Dr. Nigel Quinn for his guidance and support. I would also like to acknowledge Dr. Rolf Mehlhorn, for taking the time to share his knowledge, and for the opportunity to help one of his projects toward fruition. This project was made possible through the U.S. Department of Energy, Office of Science, Office of Workforce Development of Teachers and Scientists (WDTS) and the Workforce Development & Education under the Science Undergraduate Laboratory Internship (SULI) program at Lawrence Berkeley National Laboratory.

REFERENCES 1. Porphy, S.J. & Farid, M.M. 2012. Feasibility study for production of biofuel and chemicals from marine microalgae Nannochloropsis sp. based on basic mass and energy analysis. ISRN Renewable Energy Volume 2012: 1-11.

2. Anderson, M.A. & Morel F.M.M. 1982. The influence of aqueous iron chemistry on the uptake of iron by the costal diatom Thalassiosira weissfloggi. Limnology and Oceanography. 27(5): 789-813.

3. Sunda, W.G. & Huntsman, S.A. 1995. Iron uptake and growth limitation in oceanic and costal phytoplankton. Marine Chemistry. 50(1-4): 189-206.

4. Xuxiong, H., Likun, W., Zhengzheng, H., and Jiaqi, Y. 2013. Effect of high ferric ion concentrations on total lipids and lipid characteristics of Tetraselmis subcordiformis, Nannochloropsis oculata and Pavlova viridis. Journal of Applied Phycology. 26(1): 105-114.

5. Hugo, B., Et al. 2013. Different iron sources to study the physiology and biochemistry of iron metabolism in marine micro-algae. BioMetals. 27(1): 75-88.

6. Sunda WG, Price NM, Morel FMM. 2005. Trace metal ion buffers and their use in culture studies. In: Andersen RA, editor. Algal Culturing Techniques. 1st ed. San Diego: Elsiever Academic Press; p 35-65

7. Song, L., Qin, J.G., Su, S., Xu, J., Clarke, S., and Shan, Y., Evens T, editor. 2012. Micronutrient requirements for growth and Hydrocarbon production in the oil producing green alga Botryococcus braunii (Chlorophyta). PLoS One. 7(7): e41459

8. Taccardi, E. 2015. The effects of feed-and-starve nutrition on the growth of microalgae as a biofuel feedstock. HydroEcological Engineering Advanced Decision Support. 2015(Spring): 1-21

Figures and Tables

Degree of Absorbance (Two Fold Dilutions of Solution)

| Day 1 | Day 2 | Day 3 | Day 4 | Day 5 | Day 6 | Day 7 | Day 8 | 10L Tank | 0.42 | 0.413 | 0.403 | 0.466 | 0.404 | 0.511 | 0.71 | 0.658 | No Iron | 0.454 | 0.444 | 0.443 | 0.504 | 0.383 | 0.503 | 0.68 | 0.656 | FeEDTA | 0.694 | 0.712 | 0.727 | 0.841 | 0.738 | 0.86 | 0.54 | 1.036 | Ferrous Sulfate | 0.666 | 0.454 | 0.438 | 0.344 | 0.26 | 0.333 | 0.53 | 0.46 | Combination | 0.657 | 0.614 | 0.674 | 0.814 | 0.714 | 0.839 | 0.67 | 0.996 |
Table 1: The total amount of optical distortion measured for samples consisting of 50% culture and 50% supernatant from each of the four test cultures. The cultures were created on 7/10/15 and measurements began on 7/17/15. Plot shows high initial growth for both solutions containing 30uM FeEDTA, neutrality for the no iron solution, and steady decline for the 10uM ferrous sulfate solution

Figure 1: Graph of the optical density measurements of the four cultures and the original inoculum during the logarithmic growth period.
Iron Precipitation | Day 1 | Day 2 | Day 3 | Day 6 | Day7 | Day 8 | Day 9 | Day10 | Day 13 | Day 14 | Day 15 | Day 16 | Day 17 | 2 min | 0.343 | 0.099 | | 0.047 | 0.062 | 0.036 | | | 0.037 | 0.027 | | 0.053 | 0.045 | 5 min | 0.394 | 0.137 | 0.087 | 0.076 | 0.1 | 0.072 | 0.044 | 0.059 | 0.064 | 0.037 | 0.074 | 0.084 | 0.084 | 10 min | 0.412 | 0.177 | 0.124 | 0.108 | 0.143 | 0.109 | | 0.101 | 0.097 | 0.049 | 0.11 | 0.125 | 0.116 | 30 min | | 0.255 | 0.287 | 0.228 | 0.182 | 0.25 | 0.185 | 0.232 | 0.195 | | 0.235 | 0.273 | 0.244 | 60 min | | 0.347 | 0.334 | 0.271 | 0.231 | 0.329 | 0.232 | 0.336 | 0.288 | | 0.354 | 0.354 | 0.362 | 180 min | | 0.361 | 0.322 | 0.293 | 0.263 | 0.252 | | 0.348 | 0.297 | 0.065 | 0.403 | 0.403 | |

Table 2: The amount of iron, measured by degree of absorbance, remaining in a liter of biomass free seawater that was kept in the dark to measure the rate of iron precipitation sans external factors.

Figure 2: The degrees of absorbance for the biomass free seawater solution, tested using the ferene assay. The initial reading was lower than expected given the iron concentration. In the liter beaker little iron was lost over the two weeks, but as the plot shows this was not always the case in a smaller cuvette.

Time | No Iron | Ferric EDTA | Ferrous Sulfate | EDTA + Sulfate | Healthy Cells | 0 | 70.7 | 41.5 | 55 | 10.5 | 90.2 | 1 | 63.2 | 32.1 | 43.9 | 8.6 | 88.2 | 2 | 56.2 | 26.4 | 40.4 | 8.2 | 83.5 | 3 | 49.4 | 25.5 | 38.4 | 8.1 | 78.2 | 4 | 43.2 | 25.3 | 36.8 | 8 | 74.9 | 5 | 37.6 | 24.9 | 35.9 | 8.1 | 71.7 | 6 | 32 | 23.6 | 35.2 | 8 | 74.4 | 7 | 26.6 | 22.3 | 34.7 | 7.8 | 84.2 | 8 | 27 | 20.9 | 34.4 | 8.1 | 94.8 | 9 | 29.4 | 19.5 | 34.5 | 7.9 | 106.9 | 10 | 31.9 | 18 | 36.5 | 8.1 | 117.8 | 11 | 34.6 | 16.6 | 39.6 | 7.8 | 130.1 | 12 | 37.1 | 15.8 | 41.2 | | 144 | 13 | 39 | 14.4 | 43.5 | | 142.5 | 14 | 40.5 | 13.1 | 45.4 | | 137.4 | 15 | 41.4 | 11.6 | 47.2 | | 136.7 | 16 | 41.7 | 10.2 | 48.3 | | 136.2 | 17 | 41.3 | 9.1 | 49.3 | | 135.8 | 18 | | 8.7 | 50.1 | | 135.8 | 19 | | 8.5 | 50.6 | | 136.3 | 20 | | 8.5 | 50.9 | | 136.2 |

Table 3: The oxygen concentration of the four cultures after two weeks of growth. When descending the numbers indicate that the culture was kept in darkness, and consumed oxygen through respiration. When ascending they indicate that the culture was exposed to light and producing oxygen through photosynthesis. Note that both cultures that were fertilized with ferric EDTA had very low rates of photosynthesis, indicating cell death.

Figure 3: The rate of change in the O2 consumption in the four cultures and control, which corresponds to the rates of respiration and photosynthesis. The transition from decreasing to increasing signifies the light being turned on, and the sample switching from consuming to producing O2.

Figure 4: The four solutions after two weeks growth. Both solutions (2,4) that contained 30uM FeEDTA are deep brown indicating aggregation and cell death. The no iron solution (1) is still somewhat green and healthy, while the ferrous sulfate solution (3) has a comparatively small amount of biomass.