INHIBITION STUDIES OF PROTEIN PHOSPHATASE 2A BY KNOWN CATHARTIC PLANT DRUGS

Protein phosphatases play significant roles in signal transduction pathways that regulate cellular processes in response to external/internal stimuli. They are crucial for the growth, division, and differentiation of all organisms. Examples of cell functions involving reversible phosphorylation include ion transport, metabolism, cell cycle progression, developmental control, and stress responses. Serine/threoninespecific protein phosphatases are of particular interest in this study. There are two types of serine/threonine-specific protein phosphatases: type 1 & type 2. Both protein phosphatase 1 (PPl) and protein phosphatase 2A (PP2A) are inhibited by a structurally diverse group of natural toxins produced by marine organisms that cause diarrhetic shellfish poisoning. Recent research into diarrhetic shellfish poisoning has contributed to the understanding of some of the mechanism of actions of cathartics. It was learned that hydroxy acid moieties are essential for receptor binding of diarrhetic shellfish toxins, such as okadaic acid, to the protein phosphatase enzymes in the intestines which regulate ion channels. The loss of ion channel regulation in the intestines leads to an efflux of electrolytes and water, causing diarrhea. This study was undertaken in order to learn if four different crude cathartic plant drugs whose active components contain hydroxy-acid moieties, will act similarly to these diarrhetic shellfish toxins by inhibiting PP2A. First, two resin glycosides, jalapin and convolvulin, were isolated from Ipomoea purga and Pharbitis nil, of the Convolvulaceae family, respectively. After verifying their molecular structures by proton nuclear magnetic resonance (H NMR) spectroscopy, they were tested against PP2A. Pure ricinoleic acid from Ricinis communis and

podophyllotoxin from Podophyllum peltatum were purchased and also tested against PP2A. A fluorometric assay, recently developed in Dr. Shimizu's lab, was used to determine their activities toward protein phosphatase enzyme PP2A. The assay results indicated weak but significant inhibition of PP2A by these compounds. The doses of these crude drugs used to produce catharsis are six orders larger than the dose of okadaic acid which causes diarrhea in humans. Thus the weak activities of these compounds may be sufficient to account for the cathartic action of these drugs.
The bioavailability of these compounds in the human body is unknown. Since the traditional standard dosages of these cathartics are too large to prepare as samples for a micro-titer assay due to their solubilities, they were not able to be assayed at higher concentrations at which more or total inhibition might take place. The possibility that these crude cathartic drugs inhibit PP2A can not be discounted based on the results of this study. It is also possible that these compounds may affect other enzymatic activities which were not tested.     Figure 1). They may occur as 0-and C glycosides. Glycosides of anthranols and anthrones elicit a more drastic effect than do the anthraquinone glycosides causing a discomforting, griping action. (Stahl, E., 1973). These glycosides contribute significantly to the therapeutic activity of these crude plant drugs (Tyler, V. et al. , 1988). Taken orally, the free anthraquinone aglycones have little therapeutic activity, but in the glycosidic form the sugar moiety facilitates absorption and trans location of the aglycone to the site of action in the wall of the large intestine. Bacterial flora in the colon wall cleave the sugar residues and free the aglycones at the site of action. The anthraquinone glycosides and their derivatives are stimulant cathartics and exert their action by increasing the tone of the smooth muscle in the wall of the colon. The exact mechanisms of these actions are not yet known (Fairbairn, J.W., 1977).
Relatively recent studies involving croton oil, cholera toxin, and diarrhetic shellfish poisoning have shed some light on the mechanisms by which cathartics can act. These mechanisms of action involve interference with signal transduction pathways responsible for regulating electrolyte balance.
Signal transduction regulates cellular processes in response to external/internal stimuli and is crucial for the growth, division, and differentiation of all organisms. Reversible protein phosphorylation is an essential component of almost all signaling pathways in living cells. Changes in the phosphorylation state of a protein are conducted by two types of enzyme activities: protein kinases and protein phosphatases. Protein kinases catalyze the covalent attachment of a phosphate group to an amino acid side chain, whereas protein phosphatases reverse this process (Luan, 2000). The attachment to or removal of a phosphate group from a protein often has profound effects on its structure and thereby modifies the functional property of the protein.
Phosphorylation of key proteins with associated changes in their biological activity accounts for many physiological responses. The phosphate content of these proteins reflects a net balance of the protein kinases and protein phosphatases acting on them (Hanks et al., 1988).
Phosphorylation of hydroxyl-bearing amino acid side chains (serine, threonine, and tyrosine) is catalyzed by protein kinases (PKs) using ATP as a phosphoryl donor, whereas dephosphorylation is catalyzed by protein phosphatases, which induce changes in protein conformation, protein-protein or protein-ligand interactions, membrane permeability and solute gradients (Sheppeck, et al., 1997). This simple cycle acts as an 'on-off switch to selectively modulate the action of countless other proteins. Examples of enzyme regulation by phosphorylation include initiation of allosteric conformational changes which may directly block the access to an active site or regulation of the interaction among protein partners that must form complexes in order to function.
Examples of cell functions involving reversible phosphorylation include ion transport, metabolism, cell cycle progression, developmental control, and stress responses.
Research conducted on croton oil points to the mechanisms by which cathartics interfere with signal transduction. Croton oil is a naturally occurring plant oil obtained from Croton tiglium, a shrub-like tree of the Euphorbiaceae family (Evans, F .J ., 1986). Croton oil has been used traditionally in Western medicine as a powerful purgative (Trease, G.E. and Evans, W.C., 1972). Due to the highly potent tumor-promoting property of phorbol myristate acetate (PMA), the active ingredient of Croton tiglium oil, this drug was considered to be too toxic for human use and was eventually removed from modem pharmacopoeias (British Pharmaceutical Codex, 1934) (see Figure 2). This same compound, PMA, was later found to be a potent protein kinase C activator. This finding is consistent with the hypothesis that protein kinase C activation is part of the normal growth control process that becomes perturbed in turnorigenesis (Matthews, C. K., Van Holde, K. E., 1997). Activation of protein kinase C causes hyperphosphorylation of proteins that control sodium secretion by intestinal cells. The increased phosphorylation of cytoskeletal or junctional moieties that regulate solute permeability result in the passive loss of fluids as diarrhea (Dho et al., 1990).
There are many causes of diarrhea, but the overall alterations in intestinal function are similar in that the intestine ceases to be an organ of net absorption of water and electrolytes. The fluid produced exceeds the absorptive capacity of the remaining intestine and water passes into the stool. The aim of diarrhea treatment is to enhance intestinal absorption of water by reducing the content of luminal electrolytes (by increasing active absorption of Na+ or decreasing secretion of anions). Absorption of fluid by the colon is secondary to active transport of Na+ (Sellin, J. H., 1993).
The mechanism responsible for colonic absorption of Na+ is primarily electrogenic transport, which relies on a Na+/K+/ATPase activity in the basolateral membrane of the colonic epithelium. Neutral absorption of NaCl may also be involved. The colon absorbs er by an electrically neutral mechanism that involves the exchange of er for HC0 3 -and by neutral uptake of NaCl. Agents that elevate intracellular cAMP in colonic enterocytes stimulate electrogenic secretion of er and may inhibit NaCl uptake. This causes net fluid secretion. The colon also secretes K+, probably via an active mechanism that is stimulated by cAMP (Goodman & Gilman, 1996).
Most animal cells maintain large ionic gradients across their surface membranes such that intracellular fluid contains a higher concentration of K+ ions and low concentration of Na+ and Ca 2 + ions relative to the extracellular fluid.
These ionic gradients are maintained by the action of specific energy-dependent ion pumps. Ion channels mediating electrical signaling are intrinsic membrane proteins that form ion-selective pores through which ions can move down their electrochemical gradients into or out of cells. The responsiveness of voltagegated ion channels to membrane potential is regulated by G-protein-coupled receptors (Fine, K. D. et al., 1993). These regulatory processes are crucial in the control of hormone secretion, neurotransmitter release, muscle contraction, and gene transcription. Both direct binding of G proteins and phosphorylation of the ion-channel proteins are important effectors of this second order regulation of ionchannel function (Matthews, C. K., Van Holde, K. E., 1997).
Cholera toxin, a highly cathartic peptide produced by Vibrio cholerae, has been shown to stimulate er secretion in the small intestine and the colon by its ability to activate adenylate cyclase in the mucosa. The toxin consists of an A subunit surrounded by five B subunits. The B subunits attach the toxin to ganglioside GM 1 on the cell surface. The A subunit catalyzes ADP-ribosylation of the a-subunit of G proteins, reducing GTPase activity. Activating the asubunit of G proteins also catalyzes ADP-ribosylation of cell membrane adenylate cyclase (Chandana, S. et al., 2001). This activation interferes with the role of the Ga-subunit in regulating the maintenance of epithelial cell tight junctions. Tight junctions serve two functions: regulation of the permeability barrier to paracellular fluxes, and separation of the apical and basolateral membrane domains. Loss of regulation of epithelial cell tight junctions results in secretion of Cl-in the small intestine and colon, causing net fluid secretion (Wang, W. et al. , 2000).
Research into diarrhetic shellfish poisoning has also contributed to the understanding of the mechanism of action of cathartic drugs. Diarrhetic shellfish poisoning (DSP) is caused by consumption of shellfish which have been contaminated with natural toxins produced by dinoflagellates. The onset of the illness ranges from 30 minutes to several hours after consumption of the contaminated shellfish, but seldom exceeds 12 hours (Van Egmond et al. , 1993).
Victims suffer from diarrhea, nausea, and stomach pain, but recover within three days without serious after-effects. DSP toxins accumulated in the shellfish inhibit the activity of serine/threonine specific protein phosphatases 1 and 2A in colonic endothelial cells, resulting in rapid accumulation of phosphorylated proteins. This rapid accumulation of phosphorylated proteins results in disruption of the maintenance of electrolyte balance, preservation of membrane potential, and control of cellular volume in tissues. Oral ingestion of one microgram of okadaic acid is sufficient to produce the diarrhetic effect in a human being.
Both protein phosphatase 1 (PP 1) and protein phosphatase 2A (PP2A) are the intracellular targets for the toxins produced by marine organisms that cause diarrhetic shellfish poisoning. These DSP toxins include the polyketide inhibitors okadaic acid and dinophysistoxin-4 which are produced by dinoflagellates. Other natural protein phosphatase inhibitors include cyclic peptide inhibitors such as microcystins and nodularins, as well as calyculin and tautomycin (Gupta et al., 1997). The toxins, though structurally dissimilar from one another, all seem to bind at the active site of each phosphatase, where they contact multiple residues near the active site (Suganuma et al., 1989). A free carboxyl and hydroxyl group in the molecule is essential for receptor binding in the case of okadaic acid (see Figure 3). These moieties are important for such compounds to bind to a receptorial site on PP 1 and PP2A, thus inhibiting their activity and interfering with signal transduction processes, including ion channel regulation (Sasaki, K. et al., 1994).  (Catterall, W. A. , 1997).
As the diarrheic effect of okadaic acid and related toxins has been attributed to the accumulation of phosphorylated proteins that control sodium secretion in intestinal cells, a detection method for such cathartic compounds based on the inhibition of protein phosphatases is of particular interest. Using the specific inhibition of both PPl and PP2A catalytic subunits provides a sensitive method to detect diarrhetic shellfish poisoning. Recently, a highly sensitive fluorometric assay for DSP using PP 1 and PP2A was developed in Dr. Y.
After incubation of 4-MUP and purified PPl enzyme in a 96-well rnicrotiter plate, liberated 4-MU is measured with a fluorescent scanner.
Inhibition of PP 1 or PP2A by a specific inhibitor, e.g., okadaic acid, is quantified using the above protocol. The fluorescent inhibition assay involves introducing a test sample to the enzyme PPl or PP2A and then adding the substrate 4-MUP.
An inhibitory sample will result in a fluorescence reading that is less than a control known to be free of such inhibitors.
This inhibition assay has advantages over other assays, such as a radioactive phosphoprotein assay using 32 P ATP or a para-nitrophenyl phosphate colorimetric assay (pNPP assay), due to the elimination of clean up and the increase in specificity, sensitivity, precision, rapidness, reproducibility, and percentage ofrecovery of the 4-MUP assay. A drawback to this assay is that the commercially available PP2A enzymes which are obtained from human erythrocytes are very expensive and no recombinant enzymes are available for PP2A (Baden et al., 1995).
In this study, the 4-MUP assay was used to test the activity of jalapin and convolvulin, two hydroxy-acid containing resin glycosides from plants of the Convolvulaceae family, upon PP2A. Jalapin was isolated from roots of Ipomoea purga and convolvulin isolated from seeds of Pharbitis nil. At the same time other known cathartics whose active components contain hydroxy-acid moieties were also studied including ricinoleic acid (see Figure 5), isolated from the oil of Ricinus communis (castor oil), and podophyllotoxin (see Figure 6), isolated from podophyllum resin. The hydroxy acid moieties of these cathartic compounds, if released, may bind to a receptorial site on PP2A, as seen with the DSP toxins, inhibiting its activity and interfering with signal transduction processes such as ion channel regulation (Sasaki, K. et al., 1994).
Castor oil is cold-pressed from the seeds of Ricinis communis (Euphorbiaceae) and has traditionally been taken orally as a stimulant cathartic. It is composed of a mixture of triglycerides, of which about 7 5 % is triricinolein.
It has long been known that certain plant species of the Convolvulaceae family produce drastic cathartic and purgative effects when consumed. Two species ofthis family, Ipomoea purga (Mexican Jalap ), and Pharbitis nil (Morning Glory), were used in this study. Both of these plants produce resins, jalapin and convolvulin, respectively, which are both monomers of hydroxy-fatty acid oligoglycosides in which the sugar moiety is partially acylated by organic acids and can also combine with the carboxy group of the aglycone to form a macrocyclic ester in the case of jalapin (Noda, et al, 1987). Studies have identified these resins to be the active cathartic components of these crude drugs (Mannich, C. et al., 1938;Shellard, E. J., 1961). Jalapin is characterized by the presence of an oligoglycoside of 6-deoxyhexoses (rharnnose, fucose), whose aglycone is the hydroxylated fatty acid 1 lS-hydroxyhexadecanoic acid or jalapinolic acid. The aglycone found in convolvulin, also known as pharbitinis the hydroxy lated fatty acid, 11 S-hydroxytetradecanoic acid or convolvulinic acid (pharbitic acid) (Ono, 1990   Unless otherwise specified, all chemicals were reagent grade or better.
Solvents for extraction, partitioning, and chromatography were HPLC grade.
Distilled deionized water was used to make all solutions needed for the 4-MUP assay. Deionized water was used for all other procedures.
The following abbreviations have been used in this work:   et al. (1990). The outline of the extraction is shown in Figure 12. Fifty grams of root tubercles of lpomoea purga were pounded into a powder and macerated in methanol, sonicated, and extracted three times at room temperature, each time 200 ml methanol was used. Each methanol extraction was suctioned-filtered through a Buchner funnel lined with filter paper into a filtration flask, giving an orangecolored solution. The three filtered methanol extracts were combined and evaporated to dryness with a rotary evaporator, yielding a golden-orange resin (6.15 grams). Two grams of this methanol extractive was suspended in 25 ml of water in a separatory funnel and extracted three times with 1-butanol, each time 25 ml 1-butanol was used. The aqueous layer was a yellow color and was evaporated with a rotary evaporator to give a yellowish resin (330 milligrams).
The three 1-butanol extracts were orange in color and were combined together, Podophyllotoxin (4.14 milligrams) was pipetted into a small vial containing 1 ml of methanol to make a 10 mM solution. One hundred micro liters of 10 mM podophyllotoxin solution was pipetted into 900 µl of 50 mM Tris buffer, pH 8.50, to make a podophyllotoxin solution with a 1 mM concentration that contained ten percent methanol. Three more serial dilutions were made starting with the 1 mM podophyllotoxin solution. These three serial dilutions were diluted with 900 µl of 50 mM Tris buffer, pH 8.50. The concentrations of these podophyllotoxin solutions were 100 µM , 10 µM , and 1 µM, respectively.

4-Methylumbelliferone Phosphate Fluorometric Assay (4-MUP Assay)
In a 96-well microtiter plate, placed on ice, l µl of purified PP2A enzyme from Upstate Biotechnology was added to 10 µl of the compound being tested for potential inhibitory activity. The concentration of purified PP2A enzyme was 0.03 units per well. The mixture of the potential inhibitory compound with the enzyme was incubated at room temperature on a 3-D rotator (Lab-Line Instruments, Inc., Melrose Park, IL) for five minutes before adding the substrate.
Thirty-nine microliters of the substrate, 50 mM 4-MUP in 50 mM Tris buffer (20 mM MgCh, and 1 mM 2-mercaptoethanol, pH 8.5 adjusted with 1 N HCl), was added to each well to make a total volume of 50 µl per well. Each row of wells on the plate consisted of six replicates of the given reaction mixture for that row.
Fluorescence intensity measurements were performed using ICN Titertek model Fluoroskan II fluorescence reader using ex. 355 nm I em. 460 nm filters.
Readings were taken at time zero, thirty, and sixty minutes, after incubating the reaction mixtures at 37 °C. Inhibition of PP2A by a specific inhibitor, i.e. okadaic acid, was quantified using the above protocol. Four 96-well microtiter plates were used in this experiment.
Okadaic acid was used as a reference inhibitor on each of the four microtiter plates used. The okadaic acid sample was previously prepared in Dr.

A schematic representation of the extraction and isolation of jalapin from
Ipomoea purga roots is shown in Figure 7. Several purple spots were observed on silica-gel TLC of the 1-butanol-soluble portion of the methanol extract of the I. purga roots (1.65 grams). This 1-butanol extract was separated by silica-gel chromatography to afford a crude glycoside mixture. The fractions did not have color, had no UV absorbance, and displayed a purple color after being sprayed with the vanillin/sulfuric acid reagent. One of the TLC plates is shown in Figure   14. Based on their 1 H NMR spectra, these samples were determined to contain a mixture ofhomologues or very closely related compounds. Variations in the structures of these resin glycosides primarily come from the differing organic acid moieties such as tiglic acid, isobutyric acid, etc. which can occur. Variations also occur due to the different types and numbers of sugars within the glycosides.
However, these resin glycosides do share a basic structure which includes the hydroxy-acid moieties. The percentages of these resin glycosides present in the total crude extract are exceptionally high, reaching up to 26.9% of the crude methanol extract in some Convolvulaceae species (Noda, N., 1987 ). The predominance of these resin glycosides in the crude extract may account for the 1 H NMR data fitting the molecular structure so well, despite the fact that the separation procedure was not so thorough.
The 1 H NMR spectra of these two fractions were identical ( Figure 10) and gave similar 1 H NMR spectra as those compared to in the literature (H. Kogetsu et al., 1991 ), based on the peak assignments, integration, and ratios of hydrocarbon protons to protons of sugar and organic acids. This suggests that the resin glycosides are similar to what Mayer called the ether-soluble resin glycosides (Shellard, 1961 (8 5.51, 5.13, 4.82, 5.67 and 4.81) and four secondary methyls due to 6-deoxyhexose (8 1.56, 1.26, 1.45 and 1.65) as well as a 2-methylene (8 2.69) and a primary methyl (8 0.96) attributable to a jalapinolic acid moiety (Table 1 ). Based on this data, it is reasonable to presume that the Ipomoea resin glycoside isolated in this experiment is similar to the proposed chemical structure depicted in Figure 11.
A schematic representation of the extraction and isolation of convolvulin from Pharbitis nil seeds is shown in Figure 8. The crude methanol extract obtained from 100 grams of seeds weighed 4.17 grams. After partitioning 2.0 grams of this crude extract between 1-butanol and water, the amount ofwatersoluble extract was 730 milligrams (36.5 % of crude methanol extract). The amount of 1-butanol-soluble extract was 50 milligrams (2.5 % of crude methanol extract). These fractions had a slight yellowish-white color and had no UV absorbance. The water-soluble portion of the methanol extract from Pharbitis nil seeds exhibited several brownish-gold spots on silica-gel TLC. The LH-20 biogel column chromatography was employed to separate them. However these further purified fractions obtained from the chromatography were found to be pure sugars of di-and oligo-saccharides based on their 1 H NMR spectra, indicating that the convolvulinic acid aglycone had been hydrolyzed from the sugars and remained in the column. For this reason, the crude extracts obtained after the partitioning were identified by 1 H NMR and employed to run the PP2A 4-MUP inhibition assay. Two fractions, one being water-soluble and the other 1-butanol-soluble, each appeared as a group of three spots and had identical Rf values of 0.08, 0.14, and 0.32 (see Figure 12). These normal phase Rf values are reasonable values for high molecular weight amphoteric water-soluble resin glycosides containing lipophilic fatty acid and hydrophilic sugar moieties. Since the carboxy group of the aglycone (convolvulinic acid) is free and does not combine with a hydroxy group of the sugar moiety to form an intramolecular macrocyclic ester structure, as seen in Ipomoea purga resin glycosides, there are more free hydroxyl groups present, contributing to the water-soluble property of these resin glycosides (Ono, M. et al., 1990). Due to not being chromatographed, the purities of these samples were crude compared with the lpomoea purga samples. Based on the data from the 1 H NMR spectra, it was determined that they contained the resin glycosides as a mixture of homologues. It is possible that other contaminants may have been present, since many glycolipids were isolated in the extraction process.
Although the Pharbitis nil samples were rather crude preparations, their 1 H NMR spectra ( Figure 13) gave similar 1 H NMR spectra as those compared to in the literature, based on the peak assignments, integration, and ratios of hydrocarbon protons to protons of sugar and organic acids. Again this suggests that the resin glycoside was an overwhelmingly major component in these samples. They are similar to what Mayer called ether-insoluble resin glycosides with a free carboxy group of the convolulinic acid aglycone (Shellard, 1961 Table 2).
Based on this data, it is reasonable to presume that the Pharbitis resin glycoside isolated in this experiment is similar to the proposed chemical structure depicted in Figure 14.
In the assay experiments, the standard emission curve of 4-MU was established ( Figure 15). The fluorescence intensity of 4-MU was in a linear function with its concentration in the experimental range (0 to 10 µM). The enzyme activity could be analyzed by using the standard emission curve to compare the amount of 4-MU from the dephosphorylation of 4-MUP by PP2A.
The okadaic acid at 60 nM was used as a reference inhibitor to verify that the enzyme inhibition assay was working correctly. The results showed at least 99% significant inhibition of PP2A by 60 nM okadaic acid on all four plates, with the largest p value of the four sets of okadaic acid data being less than 0.01.
These results were expected since 60 nM of okadaic acid is a known concentration to completely inhibit PP2A activity. For each row of compounds tested on PP2A on each microtiter plate, a duplicate row of controls wells was prepared on the same plate that contained 1 µl of Tris buffer, pH 8.50 in place of the 1 µl of PP2A enzyme. These controls indicated that none of the compounds tested in this assay fluoresced on their own and therefore quenching could not occur that might produce false results. The row of control wells on each plate that contained 1 µl of PP2A enzyme, 39 µl of 4-MUP and 10 µl of Tris buffer, pH 8.50 was used to determine the normal activity that would be expected from the enzyme with no other compounds present. All of the wells on each plate containing potential inhibitory compounds were compared against these respective control wells to observe any change in activity of the enzyme.
The jalapin samples tested against PP2A were prepared in the following concentrations: 1.59 mg/ml, 0.159 mg/ml, 0.0159 mg/ml, and 0.00159 mg/ml.
These concentrations would correspond to 1 mM, 100 µM , 10 µM , and 1 µM if the samples were pure, according to the molecular weight of the proposed structure of the glycoside. Similarly, the convolvulin samples were prepared in the following concentrations: 1.48 mg/ml, 0.148 mg/ml, 0.0148 mg/ml, and 0.00148 mg/ml, which, likewise, would correspond to 1 mM, 100 µM , 10 µM , and 1 µM of pure glycoside. The molecular weight of the proposed jalapin structure is 1,590 and the molecular weight of the proposed convolvulin structure is 1, 480.
The results of the inhibition assays are shown in Figures 21 -24. A 4% increase in activation occurred at the 1.59 mg/ml concentration of jalapin, however the p value was equal to 0.3, so this data should be considered to be insignificant. Whereas, 12 % inhibition was observed at the concentrations of 0.159 mg/ml and 0.0159 mg/ml, with the p values both less than 0.01. The jalapin samples showed a maximum inhibition of 20% at the 0.00159 mg/ml concentration with a p value, 0.04. (see Figure 16). According to this data, the activity of PP2A did not change when the concentration of jalapin was decreased ten-fold from 0.159 mg/ml to 0.0159 mg/ml. The PP2A inhibition increased when the concentration ofjalapin was decreased ten-fold from 0.0159 mg/ml to 0.00159 mg/ml. However, the data from the 1.59 mg/ml concentration of jalapin was not significant, and there was no observable trend for the inhibition to increase as the concentration of jalapin decreased. The effective dose of jalapin (as a powdered resin) consumed orally which will produce catharsis is 2 grams (see Table 3). Because of the six order difference between the effective dose and the 0.00159 mg/ml concentration of jalapin which caused 20 % inhibition of PP2A in vitro, the possibility can not be excluded that jalapin inhibits PP2A .
The convolvulin showed a maximum inhibition of 50.9 % at the 1.48 mg/ml concentration with a p value less than 0.01, 5.7 % inhibition at the 0.148 mg/ml concentration with a p value of 0.045, 15.5 % inhibition at the 0.0148 mg/ml concentration with a p value of0.12, and 8% increase in enzyme activity occurred at the 0.00148 mg/ml convolvulin concentration with a p value less than 0.01 (see Figure 17). The inhibition of PP2A activity was significantly increased by 45.2% when the concentration of convolvulin was increased ten-fold from 0.148 to 1.48 mg/ml, increasing the inhibition from 5.7 % to 50.9 %. This was the largest significant change observed among two concentrations of the same compound with a ten-fold difference in concentration, out of the four cathartic plant drugs tested with this assay. The 15 .5 % inhibition of PP2A activity reported for the 0.0148 mg/ml concentration of convolvulin should not be considered significant based on the 0.12 p-value. An 8% increase in activation occurred at the 0.00148 mg/ml concentration of convolvulin. This does not seem to be consistent with the rest of the convolvulin data, despite the data's significance due to the low p-value below 0.01. The effective dose of convolvulin (as a powdered resin) consumed orally to produce catharsis is 2 grams (see Table   3). Because of the three order difference between the effective dose and the 1.48 mg/ml concentration of convolvulin which caused 50.9 % inhibition of PP2A in vitro, the possibility can not be excluded that convolvulin may inhibit PP2A .
The ricinoleic acid samples showed a maximum of 9. 7 % inhibition of enzyme activity at the 100 µM ricinoleic acid concentration with a p value of 0.02, a 6. 7 % inhibition at 10 µM with a p value of 0.01, and a 0.04 % inhibition of enzyme activity occurred at the 1 µM ricinoleic acid concentration with a p value less than 0.01 (see Figure 18). This data is significant, therefore ricinoleic acid may have a weak inhibitory effect on PP2A. According to this data, there was a slight trend for the inhibition of PP2A to increase as the concentration of ricinoleic acid increased. The effective dose of ricinoleic acid (as castor oil) consumed orally which will produce catharsis is 30 milliliters (see Table 3).
Because of the substantial difference between the effective dose and the concentration of ricinoleic acid which caused 9.7 % inhibition of PP2A in vitro, the possibility can not be excluded that ricinoleic acid may inhibit PP2A .
The podophyllotoxin samples showed a maximum inhibition of 7.2% at the 100 µM concentration with a p value of 0.03. There is 97% confidence that this data is significant, so podophyllotoxin may have a weak inhibitory effect on PP2A. A 3.4 % increase in activation occurred at the 10 µM concentration of podophyllotoxin, whereas the 1 µM concentrations showed an inhibition of 5.6 % PP2A activity (see Figure 19). However, these values should not be considered significant since there is only 15% and 25% confidence in this data based on the p values. Therefore it is unclear whether the inhibition is increasing or decreasing as the concentration of podophyllotoxin increases. The effective dose of podophyllotoxin consumed orally which will produce catharsis is 1.25 grams (see Table 3). Since there is a four-order difference between the effective dose and the concentration of podophyllotoxin which caused 7.2 % inhibition of PP2A in vitro, the possibility can not be excluded that podophyllotoxin may inhibit PP2A.
The cathartic compounds studied were chosen because of their partial structural resemblance to the DSP toxin okadaic acid and its derivatives. Their structural resemblance may cause them to act in the same manner as okadaic acid.
These plant-derived compounds all contain unique hydroxy acid moieties which can be easily cleaved in the intestines to produce catharsis. Although okadaic acid and its derivatives have complicated structures, the functional groups essential for binding to protein phosphatase 1 and 2A have been determined to be a carboxyl group and hydroxyl groups. They are located on the straight-chain carbon-carbon backbone. With the folding of the molecule, a free carboxyl group at carbon one and free hydroxyl groups at carbons 24 and 27 become available and are essential for receptor binding in the case of okadaic acid (see Figure 3).
This was determined by x-ray crystallography of okadaic acid bound to the protein phosphatase 1 and 2A enzymes (Sasaki, K. et al., 1994). These moieties are important for such compounds to bind to a receptorial site on protein phosphatases 1 (PPl) and 2A (PP2A), thus inhibiting their activity and interfering with signal transduction processes, such as ion channel regulation. The loss of ion channel regulation results in the efflux of electrolytes and water as diarrhea.
Resin glycosides are well known as the purgative ingredients of some crude drugs such as Pharbitidis Semen and Jalapae Tuber which originate from Convolvulaceae plants. Chemical investigations on these resin glycosides were conducted as early as 1840 by J. F. W. Johnston. When these resin glycosides are subjected to alkaline hydrolysis, a hydroxyfatty acid oligoglycoside (glycosidic acid) and some organic acids (isobutyric, 2-methylbutyric, tiglic acids, etc.) are provided. The glycosidic acid is cleaved by acid hydrolysis to yield a hydroxyfatty acid and several kinds of monosaccharides such as glucose rhamnose, quinovose, etc. (Noda, N. et al. , 1987).
Jalapae Tuber, the dried sliced root of Ipomoea purga, a Convolvulaceae species indigenous to Mexico, is well known as a purgative crude drug. The resin obtained from the root is called Ipomoea resin. Its resin glycoside is typically known as Mayer's "jalapin", an ether-soluble resin glycoside (Mayer, W. , 1852).
The hydroxyfatty acid (jalapinolic acid) obtained by alkaline and subsequent acid hydrolyses of Ipomoea resin was determined to be 11-hydroxyhexadecanoic acid (Asahina, Y. et al. , 1922). In 1961 , Shellard reexamined the components of this resin and identified seven organic acids (acetic, propionic, isobutyric, tiglic, 2methylbutyric, n-valeric and isovaleric acids), three sugars (glucose, fucose, and rhamnose) together with the jalapinolic acid from the ether-soluble portion. The same organic acids and sugars were isolated from the ether-insoluble portion along with ipurolic and convolvulinic acids. The parent glycosides were not isolated.
Pharbitidis Semen, the seeds of Pharbitis nil, a species of morning glory, is a cathartic crude drug. Its resin glycoside is typically known to be a Mayer' s "convolvulin," an ether-insoluble resin glycoside (Mayer, W., 1852). Early chemical investigations on the resin glycoside of this plant revealed the presence of a hydroxytetradecanoic acid ( convolvulinic acid) and tiglic acid (named for Croton tiglium, the source from which it was initially isolated), along with a glycosidic acid by alkaline hydrolysis and two crystalline fatty acids and Dglucose by acid hydrolysis of the glycosidic acid. (Kromer, N., 1896). More detailed investigations were later reported that alkaline hydrolysis of the crude glycoside named pharbitin (convolvulin) gave an organic acid named nilic acid (2-methyl-3-hydroxybutyric acid), together with tiglic acid and (+)-2methylbutyric acid (Asahina Y. et al. , 1922). Mannich and Schumann in 1938 presumed Mayer' s "convolvulin" to be a complex glycoside composed of a number of the repeating unit which is a glycosidic acid partially acylated by some organic acids at the sugar moiety. However, any pure resin glycoside had not yet been isolated and the chemical studies had been limited only to characterization of the component glycosidic acids and organic acids afforded by alkaline hydrolysis of a crude resin glycoside (Wagner, H. , 1974).
A reexamination of the chemical components of pharbitin was carried out again in 39 1990 by Ono, M. et al., who characterized the glycosidic acids as ipurolic acid 11-0-penta-and 11-0-hexaglycoside.
The effects of the hydroxy acid-containing cathartics on PP2A were determined using the PP2A 4-MUP inhibition assay. An assay that perhaps uses a crude PP2A preparation would be much more economical for further studies to be conducted. Testing these compounds on other enzymes may help to further explain their cathartic activities. Also, the development of an assay that is not performed on microscale might be more appropriate for future studies due to the solubilities of the test samples. Having a maximum volume of 50 µl per well is a limitation to this assay. This limitation prevents the drugs from being tested at a scale that is relevant to their actual therapeutic dosage and prevents them from being assayed at higher concentrations at which more or total inhibition might take place.
Another limitation of this experiment was the uncertainty of the purity of the jalapin and convolvulin samples. It may be possible that the slight activation observed in each data set from the jalapin and convolvulin samples could have been due to contamination of the enzyme assay by highly hydrophobic compounds which may have been present in the crude extracts. If this experiment were to be improved upon, separation of the crude extracts by HPLC would improve the ability to attain better purity of the isolated samples. 13 C NMR studies in addition to the 1 H NMR studies would enhance the ability to confirm the molecular structure of the isolated compounds. Performing replicates of each microtiter plate would provide more data, which would allow for better statistical analyses and more certainty in interpreting the results of the experiment.

CONCLUSIONS
The assay results indicated weak but significant inhibition of PP2A by these compounds. Okadaic acid almost completely inhibits PP2A in vitro at a 60 nM concentration and can produce diarrhea in humans by consumption of 1 µg of the toxin. The doses of these crude drugs used to produce catharsis are six orders larger than the dose of okadaic acid which causes diarrhea in humans. Thus the weak activities of these compounds may be sufficient to account for the cathartic action of these drugs. However, in the human body, the bioavailability of these compounds is unknown. Since the traditional standard dosages of these cathartics are too large to prepare as samples for a micro-titer assay due to their solubilities, they were not able to be assayed at higher concentrations at which more or total inhibition might take place. The possibility that these crude cathartic drugs inhibit PP2A can not be discounted based on the results ofthis study. It is also possible that these drugs may cause catharsis by affecting other enzymatic activities which were not tested. This matter is still open to question.      -. ..