Literature Review of the Microalga Prymnesium parvum and its Associated Toxicity

Sean Watson, Texas Parks and Wildlife Department, August 2001


Toxin Characteristics


The toxin of Prymnesium parvum has been found to be composed of a collection of substances and not a single component (Shilo and Sarig 1989). It was noted in one study that the P. parvum toxin was proteinaceous, acid-labile, thermostable, and non-dialyzable (Prescott 1968). Padilla (1970) noted the finding by Paster in 1968 that hemolysin, the hemolytic component of the P. parvum toxin, is a lipopolysaccharide. Padilla observed glycerol enhancement of hemolysin production, and suggested this shows that synthesis of hemolysin is dependent on carbohydrate and lipid metabolism. It was also implied by the author that hemolysin may be a structural component of P. parvum membranes; a notion supported by previous research that gave evidence that toxins of P. parvum are a heterogeneous mixture of phosphate-containing proteolipids. Dafni, Ulitzer and Shilo (1972) found a correlation between toxin formation (hemolysin) and presence of membrane vesicles. The authors noted that the observations of this study and past studies suggest that the P. parvum toxin appears in conditions where growth factors are limited and growth is disturbed. Because of this, they hypothesized that the toxin may be a product of imbalanced cell membrane metabolism. In one experiment, hemolysin was separated into six components with the major component, hemolysin I, determined to be a mixture of 1’-O-octadecatetraenoyl-3’-O-(6-O-B-D-galactopyranosyl-B-D-galactopyranosyl)-glycerol and 1’-O-octadecapentaenoyl-3’-O-(6-O-B-D-galactopyranosyl-B-D-galactopyranosyl)-glycerol (Kozakai et al. 1982).

Yariv and Hestrin (1961) noted that the P. parvum toxin, prymnesin, was soluble in methanol and n-propanol water solvent systems thereby distinguishing the toxin from simple protein and polysaccharides. They ascertained that prymnesin was a lipid with both non-polar and polar moieties. The authors recognized that the observed properties of prymnesin (ichthyotoxicity, hemolytic activity, non-dialysability against water, general solubility features, formation of insoluble inactive complexes with certain alcohols, and capacity to precipitate by Mg hydroxide and ammonium sulphate) were similar to the properties of saponins. However, they also noted that the action of prymnesin requires a cofactor whereas saponins do not require a cofactor. Paster (1973) described prymnesin as a high molecular weight glycolipid with a detergent-like structure. It has also been hypothesized that P. parvum toxins are plastid components or that toxin synthesis is partially plastid mediated (Guillard and Keller 1984). The hypothesis that the toxin, a proteophospholipid, is a membrane precursor is supported by the fact that there is a 10-to 20-fold increase in the toxins (ichthyotoxin, hemolysin and cytotoxin) when phosphate is limiting (Shilo and Sarig 1989).

Igarashi, Satake and Yasumoto (1999) have recently reported the structural elucidation (Figure 1) of the P. parvum toxin. They found that P. parvum produces two glycosidic toxins they named prymnesin-1 (C107H154Cl3NO44) and prymnesin-2 (C96H136Cl3NO35). The authors also concluded that prymnesin-1 and prymnesin-2 have biological activities that are almost the same. They noted that both prymnesin-1 and prymnesin-2 express potent hemolytic activity greater than that of a Merck plant saponin, and that both exhibit ichthyotoxicity.

Target and Action of Toxin

P. parvum produces soluble toxic principles: an ichthyotoxin, hemolysin and cytotoxin (Ulitzer and Shilo 1964). The P. parvum ichthyotoxin is toxic to gill-breathing species such as fish, mollusks, arthropods, and to the gill-breathing stage of amphibians (Paster 1973). The ichthyotoxin targets the permeability mechanism of the gill (Yariv and Hestrin 1961). Ulitzer and Shilo (1966) noted that toxicity occurs in two stages. The first stage is reversible damage to the gill tissues (i.e. permeability) that occurs only with a cation synergist and suitable pH. They described the second stage as mortality due to a response to toxicants already present in the water including the P. parvum toxin itself.

Dissolved potassium and calcium are necessary for developing extracellular micelles important for toxicity (Glass et al. 1991). The ichthyotoxin (now in micelles) requires activation by cofactors such as calcium, magnesium, streptomycin and sodium. (Shilo and Sarig 1989, Yariv and Hestrin 1961). Ulitzer and Shilo (1964) observed that neomycin, spermine and other polyamines can activate ichthyotoxicity with spermine being the most active. They concluded that, in the presence of more than one cofactor, the resulting toxicity was not always additive. Instead, the authors found that the toxicity depends on the specific activity of each cofactor present and their relative concentrations. They also observed that calcium has the ability to mask other cofactors. The authors found that the presence of calcium (low activity) in the presence of a cofactor that normally expresses high toxin activity will cause the activity of the toxin to decrease. Ulitzer and Shilo (1966) discovered that the P. parvum ichthyotoxin was also augmented by the cation DADPA (3,3-diaminodipropylamine) in lab with the ichthyotoxin increasing the sensitivity of Gambusia to toxicants already present in the media. Padilla and Martin (1973) noted that calcium and streptomycin have proven to be slightly synergistic, neomycin slightly more synergistic, spermine induces a four-fold increase in ichthyotoxicity, and DADPA induces a two-fold increase. They also speculated that the toxin/cation complex could interact with charged groups on toxin molecules reducing the degree of ionization and making them more reactive with the membrane.

 Prymnesin-1 (1) and Prymnesin-2 (2) structure.

Figure 1. Prymnesin-1 (1) and Prymnesin-2 (2) structure. Relative stereochemistry is shown for the rings A-N of prymnesin-1 and prymnesin-2 (From Igarashi et al. 1999).

Paster (1973) noted that the attachment of prymnesin to gill cell membranes most likely occurs where molecules such as lecithin and cholesterol are found, and attachment imposes a rearrangement on the membrane making it more permeable. The fact that prymnesin interacts with cholesterol in attack of erythrocyte membranes may support this idea (Padilla and Martin 1973). It has also been speculated that the cofactors may alter the permeability of the gills, thereby increasing the rate of absorption of the toxin into the circulation (Spiegelstein et al. 1969). Increased permeability of the gill membrane imposed by prymnesin causes fish to become more susceptible to compounds in water like CaCl and streptomycin sulphate (Yariv and Hestrin 1961). The increased permeability of the gills may even cause an increased susceptibility to the toxin’s cytotoxic and hemolytic activity (Ulitzer and Shilo 1966). Spiegelstein, Reich and Bergmann (1969) used two methods to observe the effects of the ichthyotoxin on Gambusia. They found that in the immersion method (fish in an ichthyotoxin solution), the toxicity occurs as follows: the toxin enters the gills (via capillaries), enters the dorsal aortas, and then travels to the brain. The authors noted that in the intraperitoneal injection method, the toxin first enters the circulation where it travels to the liver, then enters the hepatic vein, the heart, the aorta and finally the brain. They recalled that the toxin is acid-labile and suggested that it may be altered (inactivated) in the GI tract and liver. The authors suggested that this could be why the toxin is non-toxic to non-gill breathers, but toxic to gill breathers.


It has been reported that the ichthyotoxin accumulates during the stationary phase of growth, and the hemolytic toxin accumulates during log phase (Padilla 1970). Simonsen and Moestrup (1997) determined that the hemolytic compounds within the P. parvum cells are the highest in late exponential growth phase and decreased during stationary phase. The authors then discovered hemolytic activity in the medium during stationary phase.

Population Density

Shilo and Aschner (1953) discovered that fish peptone and egg yolk increases the density of P. parvum cultures. The authors ruled out the idea of toxigenic variants due to an observed decrease in toxicity accompanied by an increase in cell proliferation. They often observed an inverse relationship between cell count and toxicity. Shilo (1967) also
found a lack of correlation between toxicity and cell density. The author hypothesized that the ability to form toxins may be determined by genetic factors. The author based this hypothesis on the knowledge that different strains of Microcystis aeruginosa and Anabena flos-aquae were shown to differ markedly in their toxin productivity. The author also found non-toxigenic strains of this alga.


Reich and Rotberg claimed that the activity of the ichthyotoxin of P. parvum is inversely proportional to salt concentrations (Reich and Parnas 1962). Ulitzer and Shilo (1964) also found that a decrease in salinity equals an increase in ichthyotoxicity, and that ichthyotoxicity decreases as salinity increases. In a later study, increased salinity decreased the uptake of trypan blue (i.e. toxicity) in the gills of fish (Ulitzer and Shilo 1966). Paster (1973) observed that a NaCl range of 0.3%-5% was optimal for toxin production in P. parvum. However, Larsen and Bryant (1998) noted that variable salinity did not have significant effects on toxicity.


Shilo and Ashner (1953) found that at 80 C and 97 C toxicity declined rapidly while 62 C caused a slow decline in toxicity. They also observed that at room temperature and 4 C there was no decrease in toxicity. Yariv and Hestrin (1961) noted that prymnesin solutions in water showed a decrease in titer when kept at 35 C for 60 minutes, but returning the solution to a pH of 4 restored toxin activity. Paster also discerned the thermo-sensitivity of the toxin in 1968 (Stabell et al. 1993).

Ulitzer and Shilo (1964) noted that an increase in temperature in the range of 10 C to 30 C caused an increase in the rate of mortality of minnows with the titer of the toxin unaffected. However, Larsen and Bryant (1998) concluded that variable temperatures in the range of 5 C to 30 C do not have significant effects on toxicity thereby contradicting the research by Ulitzer and Shilo in 1964.


Shilo and Aschner (1953) concluded that light augments toxin production. These authors found water containing P. parvum to be more toxic in light than in dark. Parnas, Reich and Bergmann (1962) found that the P. parvum toxin is sensitive to light. The experiments conducted by these authors revealed that UV causes 100% inactivation of the toxin with the upper limit of inactivation by visible light at 520 nm (50% inactivation). Reich and Parnas (1962) noted that, in their first experiment, ichthyotoxicity decreased gradually with exposure to light. The authors also found that, in the dark, toxicity rises reaching a maximum in 7.5 hours. In the second experiment, the observed similar results: toxicity increased in the dark accompanied by a pH drop in dark to between 7.0 and 7.1. The pH rose in the light to 8.0-8.1 due to, the researchers speculated, photosynthetiactivity. They hypothesized that the desistance of ichthyotoxicity was either due to inactivation by light or to the delay of toxin formation due to increased photosynthesis in light.

Rahat and Jahn (1965) observed that dark cultures in their study were more toxic, even with less cells, than light cultures (this agrees with idea that the extracellular toxin is
inactivated by light). The authors also concluded that light is not needed to make the P. parvum toxin, and that previous assays of the toxin are only the net result of toxin production and inactivation. A study by Padan, Ginzburg and Shilo (1967) showed that the ichthyotoxin and hemolysin are both sensitive to inactivation by light. The authors also concluded that light is needed for the appearance of extracellular hemolysin. They noted that the equilibrium between the appearance of hemolytic activity and inactivation appears to be in favor of toxin accumulation in low light (60 foot candles). Spiegelstein, Reich and Bergmann (1969) determined that the ichthyotoxin is produced in the dark equally well as in the light (light not necessary). Paster (1973) observed inactivation of prymnesin by visible light (400nm-510nm) and UV light (225nm). The toxin of the closely related, flagellated algae, Phaeocystis pouchetii, is also believed to be photo-sensitive (Stabell et al. 1999). However, Larsen and Bryant (1998) have concluded that variable salinity, light and temperature do not have significant effects on toxicity. These authors believe that growth phase and nutrient status probably have a greater impact.


Shilo and Aschner (1953) deduced that P. parvum toxicity was independent of pH in the range of 7.5-9.0; toxicity decreased rapidly at pH less than 7.5 and was completely inactive at 6.0. The authors hypothesized that the inactivating effect was duto the hydrogen cation. McLaughlin (1958) concluded that toxicity decreased at pH 6.0-6.5 (or become non-toxic), and toxicity returns when the solution is brought back to a neutral pH. The author also found that acid grown cultures are less toxic than alkaline grown cultures. Ulitzer and Shilo(1964) noted that there is a correlation between elevated pH and toxicity. Ulitzer and Shilo (1966) observed that the gills of Gambusia became darkly stained at pH 9, but that no trypan blue staining (i.e. toxicity) was observed at a pH of 7. Padilla (1970) noted an increase in pH caused a decrease in hemolysis (pH range 5.5-8.0). Padilla and Martin (1973) noticed that maximum hemolytic activity occurred at pH 5.5. The authors also found that cytotoxicity is arrested by pH 6.4, and that maximum binding of the toxin was observed in the pH range of 4.6-5.5.

Shilo and Sarig (1989) found that a pH higher than 8 was necessary for cation activation, and a pH of 7 and lower equals little ichthyotoxic activity with the toxicity increasing to a pH of 9. They also determined that, when a cation is complexed with the toxin at high pH, the ichthyotoxicity is expressed even at low pH (6-7).

Phosphate and Nitrogen

Dafni, Ulitzer and Shilo (1972) found that a decrease in phosphate caused an increase in toxicity. The authors speculated that a phosphate-limiting environment could cause a disturbance in the formation of membrane phospholipids that may lead to leakiness (and the toxin escaping). They noted that the cell volume of P. parvum increased as the concentration of phosphate decreased, and it was hypothesized that swelling was due to osmotic imbalance (leakiness) or disturbance in regular cell division. Paster (1973) also found P. parvum to be more toxic in phosphate-poor media.

Holdway, Watson and Moss (1978) noted that, with substantial concentrations of nitrogen and phosphorous, P. parvum will not produce or release toxins. The fish kills in the Sandsfjord system in Norway were determined to be mostly due to phosphorous-limited growth of P. parvum; the phosphate-limited environment was considered to be the major factor influencing increased toxicity (Kaartvedt et al. 1991). Larsen, Eikrem and Paasche (1993) found that phosphate-limitation caused an increase in toxicity of a Denmark strain of P. parvum in lab. Simonsen and Moestrup (1997) observed an increase in the size of P. parvum, Chrysochromulina polylepsis, Chrysochromulina hirta, and Isochrysis spp. cells accompanied by increased toxicity when phosphate was limited. They also noted that the dinoflagellate Alexandrium tamarense has been noted to show increased toxicity with a decrease in phosphate concentrations. Johansson and Graneli (1999) discerned that nitrogen limitation causes an increase in toxicity, and also found that phosphate limitation causes increased toxicity as well. The authors hypothesized that the N:P ratio could be the governing factor of toxicity in P. parvum, and that a change in the N:P ratio by nutrient inputs could lead to toxicity (an unbalanced N:P ratio could result from eutrophication). The authors admitted that the reason for toxin production by P. parvum is uncertain, but speculated that the toxin could be produced because of the need to wipe out competition during nutrient limitation. Wynne and Rhee (1986) concluded that changes in the light regime can alter the optimum cellular N:P ratio in P. parvum thereby greatly influencing nutrient requirements and species interrelationships.


Glycerol was found to increase the growth rate and toxin synthesis in P. parvum (Padilla 1970). Cheng and Antia (1970) found that P. parvum is able to metabolize glycerol in high and low concentrations. The authors implied that glycerol pollution may stimulate P. parvum thereby causing blooms in light as well as in the absence of light. They found that P. parvum responded rapidly to high glycerol concentrations, and appeared to become ‘spent out’ with rapid cell lysis following an early growth peak.

Toxin Inhibitors

Padilla (1970) noticed that hemolysis is inhibited by high pHs with a maximum toxicity at pH 5.5, 50% at pH 7, and 10% at pH 8. Paster (1973) found that lecithin, cholesterol and cephalin inhibit the hemolytic affect in small quantities, and concluded that these lipid compounds must compete with the toxin for the target site. The author also noted that the bacteria Proteus vulgaris and Bacillus subtilis decrease the potency of P. parvum cultures. Padilla and Martin (1973) inferred that cholesterol, cephalin and the Gymnodinium breve toxin exert a protective influence. It has also been observed that NaCl inhibits P. parvum toxin activity (Shilo and Sarig 1989).

Successful and Possible Control Methods

Aditional Information:

Would you like to know more?
The Biology of Golden Alga summarizes what we know about the alga and its toxins.

Where does golden alga fit compared to other single-celled organisms?
The Golden Alga Family Tree gives examples of and information about golden alga and other protists.

What does golden alga look like?
TPWD Golden Alga Images has photos of fish kills, golden algal cells, and short videos of live golden alga. These images may be used for noncommercial/educational purposes as long as TPWD is given credit and other site policies are followed.

Golden Alga Information Card: TPWD has collaborated with TCEQ and other entities to produce a golden alga information card(pdf document). The purpose of this card is to educate the public on golden alga blooms and answer some common questions. Hard copies of this card are available for free by contacting 512 389-8750.

Back to Top
Back to Top