Author: Steven Back
Institution: Carnegie Mellon University
Date: April 2005
Abstract
Cryptophycin is a promising drug in many cancer therapies. Unfortunately, cryptophycin is expensive to produce synthetically because of its complex structure; however, certain strains of micro-algae naturally produce cryptophycin. This work studies the production of cryptophycin from two strains of Nostoc sp., ATCC 53789 and GSV 224. The production of cryptophycin was measured as a function of cellular life cycle and supplied media, normal BG-11, modified BG-11, and A3M7. Preliminary results show that GSV 224 yielded the highest quantity of intracellular cryptophycin on or before day 10 of its 30-day life cycle, producing 3.04 mg/L, as opposed to 0.4 mg/L on day 15 and 0.25 mg/L on day 20. Neither algal cell type culture released more than 1 mg/L of cryptophycin into the media during its life cycle. Normal BG-11 media elicited the highest relative level of intracellular cryptophycin production in both strains, with strain ATCC 53789 producing 1.30 mg/L and GSV 224 producing 2.55 mg/L.
Introduction
Advances in chemotherapy have led to the treatment of many different types of cancer; however, many still have no treatment options so these advances have no effect on countless others. Chemotherapy has proven to effectively cure acute lymphoblastic leukemia in 75% of cases for children and 40% for adults (Krakoff, 1987). Unfortunately, chemotherapy does not produce the same results in cancers that stem from "solid tumors", such as lung, breast, prostate, pancreas, ovary, brain, and colon cancer. These cancers account for 85% of all cancer deaths in the United States (Pratt et al. 1994; Kellen 1995).
Cryptophycin is a promising drug that may increase the success of treatment. Cryptophycin attacks tubulin microfilaments of eukaryotic cells and prevents cell division. It can destroy a wide variety of tumors including solid, multiple drug resistant (MDR) tumors like those mentioned above (Corbett et al. 1996). The following table illustrates the success of cryptophycin compared to other established cancer medications. However, the very complex structure of cryptophycin is difficult and expensive to produce synthetically. Its synthesis contains thirty-five steps with a yield of only 3.5% (Moore et al., 1996).
To make cryptophycin commercially viable, researchers must find a method to shorten the synthetic production or find a source of natural cryptophycin production. Researchers discovered that cyanobacteria, or blue-green algae, produce cryptophycin naturally (Moore et al., 1996). Based on this knowledge, two hypotheses were examined in this work: first, since cryptophycin has been proven to be a strong anti-fungal agent (Sesin et. al.), it is believed that the algae release cryptophycin into their environment to inhibit the growth of fungi that competes for food and sunlight. Therefore, cryptophycin will be present within the cells and media at different concentrations during different points of the lifecycle. Second, the algae will produce different levels of cryptophycin to adjust to different environmental conditions. Therefore, the different levels and kinds of nutrients in the media will elicit different levels of intracellular cryptophycin production.
The present study analyzes the production of cryptophycin in two strains of algae as a function of their life cycle and the supplied media. New information about cryptophycin production in blue green algae cultures, specifically, the production time of cryptophycin and the optimal growth medium, will broadly impact cancer treatment by facilitating the development of a cost efficient system of cryptophycin production and lowering the market price of an important and emerging anti-cancer drug.
Methods & Materials
Media Recipes
Modified BG-11 (Fisher): 1 g/L NaNO3; 0.2 g/L Na2CO3; 0.625 g/L MOPS; 0.175 mL/L 1 M K2HPO43H2O; 0.3 mL/L 1 M MgSO47H2O; 0.25 mL/L CaCl22H2O; 1 mL/L 1 M ASMT.
Normal BG-11 (Fisher): 1.5 g/L NaNO3; 0.2 g/L Na2CO3; 0.04 g/L K2HPO43H2O; 0.075 g/L MgSO47H2O; 0.036 g/L CaCl22H2O; 0.006 g/L citric acid; 0.006 g/L Ferric ammonium citrate; 0.001 g/L EDTA (disodium salt), and 1mL/L 1 M ASMT.
A3M7 (Fisher): 0.12 mL/L 1 M Arnon's solution; 0.19 mL/L 1 M NH4Cl; 0.175 mL/L 1 M K2HPO43H2O; 0.625 g/L MOPS; 0.2 mL/L 1 M MgSO47H2O; 0.1 mL/L CaCl22H2O; 1 mL/L ASMT; 2.3 ml/L 1 M NaNO3.
Prior to use, modified BG-11 was adjusted to a pH of 7.2, normal BG-11 to 7.1, and A3M7 to 7.0 using 1M HCl and 1M NaOH. The media was sterilized by autoclaving in situ for 75 minutes at 121º C and 20 psi.
Media Cultures: Inoculums
Inoculums for all experiments were prepared by culturing the specific strain of algae in twenty-liter carboys for 30 days in modified BG-11 media.
Cryptophycin Production as a Function of Lifecycle
Two strains of cells, ATCC 53789 and GSV 224, were grown separately in four 20 L (i.e., four replicates per strain) carboys containing modified BG-11 media. Each culture was inoculated with 1 L of inoculum described above. Premixed air with carbon dioxide (2%) was used as the aerating agent at an approximate flow rate of 0.4 mL/s.
Table 1. Comparison of cryptophycin effects in various tumors in mice to many of the cancer drugs on the market. N/A means "not available". See Materials and Methods for the conversion of Log Tumor Cell Kill in tumor-bearing mice to ++++, +++, ++, + and , activity ratings. * HCT15 - NCI-SRI data
Algae was harvested on day 10 with new samples taken every five days until day 30. Each carboy was shaken vigorously, and 1 L of cells and media was siphoned from each carboy to yield a total of 4 L of culture harvested per algal strain. The cells were separated from the media by centrifuging the mixture at 10,000 rpm at 10º C for 10 minutes. The supernatant was then filtered through adjacent layers of Calbiochem miracloth and Whatman 55 mm filter paper in order to remove cells from the media. The cell pellet was collected, frozen, and lyophilized at 25º C and 0.1 mbar for 48 hours to completely remove any water. Mass per unit volume was computed for lyophilized cells, and cryptophycin was isolated from the media by partitioning as described below. Cryptophycin was extracted from the algal cells, then isolated from the extract and purified.
Cryptophycin Production as a Function of Different Medias
Two strains of algae were cultivated in 2 L Erlenmeyer flasks using modified BG-11, normal BG-11, and A3M7 as the media. For each strain, 1 L of media was used, and two replicates were run per media type tested, yielding a total of six flask cultures for each strain of algae. Each flask was autoclaved for 40 minutes under the same conditions as the life cycle experiment. Identical amounts (0.2 L) of inoculum were added into each flask which was plugged with a foam stopper to permit aeration. The cultures were cultivated in batch mode for 21 days and harvested as described above.
Isolation of cryptophycin from media
150 mL of methylene chloride were added to one-liter samples of harvested media in a separatory funnel. The aqueous and organic layers were shaken thoroughly to ensure complete mixing, and were allowed to stand for 24 hours to ensure complete separation of both phases. The organic phase was removed and filtered through packed cotton to remove any precipitate that may have formed. 3 to 6 grams of anhydrous magnesium sulfate was added to the organic phase, which was then shaken and allowed to sit for several hours to remove any dissolved water. The organic phase was filtered through packed cotton to remove the magnesium sulfate precipitate and then evaporated under reduced pressure created by a water aspirator, at a temperature of approximately 38º C
Extraction of Cryptophycin
Table 2. Effect of media on cellular density and cryptophycin density. This chart compares the amount cellular growth, the isolated fraction and total purified amount of cryptophycin from ATCC 53789 and GSV 224 grown in Normal BG-11, Modified BG-11 and A3M7 media.
A 4:1 ratio of acetonitrile and methylene chloride was added to the dried cells (50 mL of organic solution per gram of biomass) and stirred for 48 hours. The solution was then filtered through adjacent layers of miracloth and Whatman paper, and the organic solvent evaporated under the reduced pressure created by a water aspirator at approximately 38ºC.
Isolation of Cryptophycin
A normal phase chromatography column was filled with an amount of silica gel that was approximately fifteen times greater than the mass of the extracted sample obtained from extraction. A known amount of methylene chloride was then circulated through the column to remove air bubbles while ensuring that the organic liquid level did not fall below the stationary phase. The organic liquid level was lowered to the top of the stationary phase, and the eluted methylene chloride was subtracted from the originally known amount of methylene chloride to determine the column volume. The extraction sample was dissolved in methylene chloride and pipetted to the top of the column. After loading the column, the sample was separated into three fractions by washing the column with three different organic solvents and collecting the eluted solvent in a different pre-weighed vial: three column volumes of methylene chloride, 1.5 column volumes of 4:1 ethyl acetate:isopropanol, and 0.5 column volume of 4:1 ethyl acetate:isopropanol. Each column volume was left in a hood to dry.
Purification of Cryptophycin
The second fraction from the isolation step was dissolved in approximately 0.5 mL of methylene chloride. This fraction was used because the polarity of solvent washes the cryptophycin off of the column; however each fraction was tested for cryptophycin using the NMR as described below. Approximately 0.5g of C-18 gel was added to this mixture, stirred, and allowed the dry in a safety hood. A reverse phase chromatography column was filled with an amount of C-18 gel that was approximately 15 to 20 times the mass of the sample. To remove air bubbles, 50% aqueous acetonitrile was circulated through the column. The column volume was determined as in the isolation step. The dried sample was added to the column and eluted with one column volume of 50% aqueous acetonitrile, two column volumes of 75% aqueous acetonitrile, two column volumes of 75% aqueous acetonitrile, and one column volume of isopropanol. Each fraction was dried and weighed.
Detection of Cryptophycin
Nuclear Magnetic Resonance (NMR) Spectroscopy was used to determine presence of cryptophycin is samples. 1 mg of sample was dissolved in 0.5 mL of CDCl3 and transferred to an NMR tube. The samples were then taken to a Varian, 300 MHz, proton NMR Spectrometer and analyzed.
Results
Note: To save time, replicate samples of culture at each time point were mixed together before extracting cryptophycin. Therefore, each value reported below represents an average level of cryptophycin found in replicate samples. Accordingly, only one plot was graphed at each time point, and no standard deviation was calculated.
Cryptophycin Production as a Function of Lifecycle
GSV 224 algal cells grew to 0.179 g/L by day 10, increased to 0.237 g/L by day 15, and leveled off at 0.246 g/L by day 20 (Figure 1). These results reflected typical exponential growth with an estimated growth rate of 0.1608 gL-1h-1. Cell growth appeared to plateau shortly after day 15. 5.32 mg/L by day 10, reached 9.65 mg/L by day 15 and leveled off at 9.75 mg/L by day 20 (Figure 2). The isolated fraction seemed to mimic the cellular growth, and plateaued shortly after day 15. The largest amount of pure cryptophycin was retrieved by day 10 at 3.04 mg/L, and sharply decayed to 0.4 mg/L by day 15 and to 0.25 mg/L by day 20 (Figure 3).
From NMR scans, levels of cryptophycin did exist in the media from both cultures of ATCC 53789 and GSV224; however, in all samples of extracellular media, less than 1 mg of cryptophycin was present.
Figure 1. Cellular growth rate of GSV 224. This graph illustrates the cellular growth of GSV 224 between day 10 and day 20 of the growth cycle.
Cryptophycin Production as a Function of Media
Figure 4 presents the effect of media on cellular levels of ATCC 53789. Biomass levels did not vary much between each type of media; although, the amount of isolated cryptophycin produced in the A3M7 media was almost twice that produced in both of the other types of media. However, ATCC 53789 produced the purest cryptophycin per liter of media in Normal BG-11 with 1.30 mg/L, followed by A3M7 with 1.25 mg/L and Modified BG-11 with 0.95 mg/L.
GSV 224 cellular levels did show some variation between each type of media. Normal BG-11 media elicited the highest cell count per liter of media. It also provoked the algal cells to produce the largest amount of isolated cryptophycin per liter of media by nearly 4 mg/L. This translated into the highest purified cryptophycin fraction with (mg/L) 2.55, 1.80, and 2.40 for Normal BG-11, A3M7, and Modified BG-11 respectively. These values are approximately twice of those achieved from ATCC 53789.
Discussion
Cryptophycin Production as a Function of Life Cycle
Cryptophycin levels could have reached a maximum on day 10 because cryptophycin production is most likely aligned with the reproductive cycle of the algae. Since cryptophycin is believed to be a defense chemical against algae, it may be produced to ensure that enough nutrients reach the algal cells during a critical period of growth.
Figure 2. Isolated cryptophycin density of GSV 224. This graph illustrates the amount of cryptophycin that was recovered from GSV 224 using normal phase chromatography during the isolation step of the life cycle experiment.
From the sharp decline in cryptophycin concentration after day 10 (Figure 3), the results suggest that only one-third of the life cycle must pass before the cryptophycin extraction, isolation, and purification processes occur, assuming a 30 day life cycle for GSV 224 algae. As a result, this production process will retrieve cryptophycin in one-third of the time, increasing speed of production and reducing production costs. In future trials, samples should be taken more frequently to find the precise time when cryptophycin concentrations are at a maximum.
Cryptophycin Production as a Function of Media
The results clearly indicate that the optimal media of those studied was Normal BG-11, followed by A3M7 and Modified BG-11 for both strains of algae. Although each nutrient varies between each type of media, the differences could be attributed to the nitrogen content in each media type. Normal BG-11 contained the most amount of nitrogen with 1.5 g/L from NaNO3, and elicited the most cryptophycin in both algal strains. However, A3M7 contained the least amount of nitrogen, with 0.205 g/L from NH4Cl and NaNO3, but elicited the second largest quantity of cryptophycin, which leads to the belief that differences in other nutrients certainly played an important role in production.
When cryptophycin is produced commercially from these strains of algae, the results of this experiment suggest that the Normal BG-11 media should be used to optimize production.
Figure 3. Purified cryptophycin density of GSV 224. This graph illustrates the purified amount of cryptophycin recovered from GSV 224 using reverse phase chromatography during the life cycle experiment.
Results from both the life cycle and media experiments further suggest that the most cryptophycin can be produced by growing GSV 224 in Normal BG-11 media and extracting approximately 10 days after inoculation. Interestingly, GSV 224 produced approximately twice the concentration of cryptophycin in all three types of media; therefore it is reasonable to suggest that future studies should focus only on maximizing production of cryptophycin using strain GSV 224.
Acknowledgements
The author would like to acknowledge funding support from NSF grant #0243600 and the University of Hawaii Sea Grant College Program, as well as USDA grant CSREES 2003-34135-14033. In addition, thanks are given to Dr. Jian Liang, Dr. Michael J. Cooney, and Louis Johnson for help and advice.
References
Corbett TH, et al. (1996). Preclinical anticancer activity of cryptophycin-8. J Experimental and Theoretical Oncology. 1:95-108.
Kellen JA. (1995). Alternative Mechanisms of Multi-drug Resistance, in Cancer, Kellen JA Ed. Birkhauser: Boston.
Krakoff I. (1987). Ca,A Cancer Journal for Clinicians. 37:93.
Moore RE, et al. (1996). The Search for new anti-tumor drugs from blue-green algae. Current Pharmaceutical Design. 2:317-330.
Pratt WB, et al. (1994). The Anticancer Drugs 2nd ed, Oxford University Press: New York.
Sesin A, et al. (1989). United States Patent 4,868,208.