Addition Of S-adenosyl-L-methionine Research Articles

Enzymatic preparation and identification of 5′-adenosyl-methylthiopropylamine for the impurity control in SAM fermentation

Abstract S-adenosyl-L-methionine (SAM) is an important biosynthetic drug used for various diseases such as depression, arthritis syndromes and cirrhosis in clinical therapeutics. In order to ensure the high quality of SAM as a clinical drug, the key impurity, 5′-adenosyl-methylthiopropylamine (decarboxylated SAM, dC-SAM) should be monitored during the whole production process and in end tablet product. In the present work, the SPE2 gene encoding SAM decarboxylase was inserted into pET28a to obtain the expression plasmid pET28a-SPE2, which was transferred into Escherichia coli BL21 to construct the recombinant strain. With low temperature (26 °C) cultivation and 0.1 mM IPTG induction, the recombinant cells were collected to prepare cell lysates, which can be utilized to effectively prepare dC-SAM in vitro with the addition of SAM as the substrate. Then dC-SAM sample with high-purity (99.7%) was isolated from this biotransformation mixture by semi-preparative high performance liquid chromatography. Further, the molecular structure of the purified dC-SAM sample was confirmed by liquid chromatography-mass spectrometry. Finally, prepared dC-SAM was employed to detect the content of the impurity during one-month industrial SAM fermentation processes. Our work will be very helpful to the control of impurities during the whole production process of SAM and its end products.

Identification and characterization of functional homologs of nitrogenase cofactor biosynthesis protein NifB from methanogens

Nitrogenase biosynthesis protein NifB catalyzes the radical S-adenosyl-L-methionine (SAM)-dependent insertion of carbide into the M cluster, the cofactor of the molybdenum nitrogenase from Azotobacter vinelandii. Here, we report the identification and characterization of two naturally "truncated" homologs of NifB from Methanosarcina acetivorans (NifB(Ma)) and Methanobacterium thermoautotrophicum (NifB(Mt)), which contain a SAM-binding domain at the N terminus but lack a domain toward the C terminus that shares homology with NifX, an accessory protein in M cluster biosynthesis. NifB(Ma) and NifB(Mt) are monomeric proteins containing a SAM-binding [Fe4S4] cluster (designated the SAM cluster) and a [Fe4S4]-like cluster pair (designated the K cluster) that can be processed into an [Fe8S9] precursor to the M cluster (designated the L cluster). Further, the K clusters in NifB(Ma) and NifB(Mt) can be converted to L clusters upon addition of SAM, which corresponds to their ability to heterologously donate L clusters to the biosynthetic machinery of A. vinelandii for further maturation into the M clusters. Perhaps even more excitingly, NifB(Ma) and NifB(Mt) can catalyze the removal of methyl group from SAM and the abstraction of hydrogen from this methyl group by 5'-deoxyadenosyl radical that initiates the radical-based incorporation of methyl-derived carbide into the M cluster. The successful identification of NifB(Ma) and NifB(Mt) as functional homologs of NifB not only enabled classification of a new subset of radical SAM methyltransferases that specialize in complex metallocluster assembly, but also provided a new tool for further characterization of the distinctive, NifB-catalyzed methyl transfer and conversion to an iron-bound carbide.

Effect of S-adenosyl-l-methionine (SAM), an allosteric activator of cystathionine-β-synthase (CBS) on colorectal cancer cell proliferation and bioenergetics in vitro

Recent data show that colon cancer cells selectively overexpress cystathionine-β-synthase (CBS), which produces hydrogen sulfide (H2S), to maintain cellular bioenergetics, support tumor growth and stimulate angiogenesis and vasorelaxation in the tumor microenvironment. The purpose of the current study was to investigate the effect of the allosteric CBS activator S-adenosyl-l-methionine (SAM) on the proliferation and bioenergetics of the CBS-expressing colon cancer cell line HCT116. The non-transformed, non-tumorigenic colon epithelial cell line NCM356 was used as control. For assessment of cell proliferation, the xCELLigence system was used. Bioenergetic function was measured by Extracellular Flux Analysis. Experiments using human recombinant CBS or HCT116 homogenates complemented the cell-based studies. SAM markedly enhanced CBS-mediated H2S production in vitro, especially when a combination of cysteine and homocysteine was used as substrates. Addition of SAM (0.1–3mM) to HCT116 cells induced a concentration-dependent increase H2S production. SAM exerted time- and concentration-dependent modulatory effects on cell proliferation. At 0.1–1mM SAM increased HCT116 proliferation between 0 and 12h, while the highest SAM concentration (3mM) inhibited proliferation. Over a longer time period (12–24h), only the lowest concentration of SAM used (0.1mM) stimulated cell proliferation; higher SAM concentrations produced a concentration-dependent inhibition. The short-term stimulatory effects of SAM were attenuated by the CBS inhibitor aminooxyacetic acid (AOAA) or by stable silencing of CBS. In contrast, the inhibitory effects of SAM on cell proliferation was unaffected by CBS inhibition or CBS silencing. In contrast to HCT116 cells, the lower rate of proliferation of the low-CBS expressor NCM356 cells was unaffected by SAM. Short-term (1h) exposure of HCT116 cells to SAM induced a concentration-dependent increase in oxygen consumption and bioenergetic function at 0.1–1mM, while 3mM was inhibitory. Longer-term (72h) exposure of HCT116 cells to all concentrations of SAM tested suppressed mitochondrial oxygen consumption rate, cellular ATP content and cell viability. The stimulatory effect of SAM on bioenergetics was attenuated in cells with stable CBS silencing, while the inhibitory effects were unaffected. In NCM356 cells SAM exerted smaller effects on cellular bioenergetics than in HCT116 cells. We have also observed a downregulation of CBS in response to prolonged exposure of SAM both in HCT116 and NCM356 cells. Taken together, the results demonstrate that H2S production in HCT116 cells is stimulated by the allosteric CBS activator, SAM. At low-to intermediate levels and early time periods the resulting H2S serves as an endogenous cancer cell growth and bioenergetic factor. In contrast, the inhibition of cell proliferation and bioenergetic function by SAM does not appear to relate to adverse autocrine effects of H2S resulting from CBS over-stimulation but, rather to CBS-independent pharmacological effects.

P03 Effect of S-adenosyl-l-methionine (SAM), an allosteric activator of cystathionine-β-synthase (CBS) on colorectal cancer cell proliferation and bioenergetics in vitro

Several series of in vitro and in vivo data presented at the current Meeting support the conclusion that colon cancer cells selectively overexpress CBS, which produces H2S, to maintain cellular bioenergetics, thereby supporting tumor growth and to promote angiogenesis and vasorelaxation. These studies, taken together, identify CBS-derived H2S as a tumor promoting factor and a potential future anticancer drug target. The purpose of the current study was to investigate the effect of S-adenosyl-l-methionine (SAM) on the proliferation and bioenergetics of the colon cancer cell line HCT116 in vitro. The non-transformed, the non-tumorigenic colon epithelial cell line NCM356, derived from the normal margin of a rectal cancer specimen, was used as a control. HCT116 cells were cultured in McCoy’s 5A medium, while NCM356 cells were cultured in DMEM supplemented with 10% fetal bovine serum. For assessment of cell proliferation, the xCELLigence system (Roche) was used. For the measurement of bioenergetic function, the XF24 Extracellular Flux Analyzer (Seahorse) was used. Oxygen consumption rate (OCR) after oligomycin (1.5 μg/ml) was used to assess ATP production rate and OCR after FCCP (0.5 μM) to assess maximal mitochondrial respiratory capacity. 2-deoxyglucose (100 mM) was used to estimate cellular glycolytic dependency and antimycin A (2 μg/m) and rotenone (2 μM) were used to inhibit the flux of electrons through complex III and I, to detect residual non-mitochondrial activity. Addition of SAM (0.1–1 mM) to HCT116 cells (which show a high expression of CBS) induced a concentration-dependent increase in cell proliferation between 0–50 h, while higher concentrations of SAM (3 mM) inhibited cell proliferation. However, at time periods between 50–100 h, the cells treated with intermediary concentrations of SAM (0.3–1 mM) showed a suppression of cell proliferation compared to the vehicle control; the proliferation-enhancing effect of SAM was only maintained throughout the entire experimental period with the 0.1 mM concentration of SAM. In contrast to HCT116 cells, NCM356 cells (which have only a low-level expression of CBS) were proliferating at a substantially slower rate, and SAM failed to stimulate their proliferation. Extracellular Flux Analysis of the effect of acute exposure of HCT116 cells to SAM (0.1–1 mM) induced a concentration-dependent increase in their oxygen consumption and bioenergetic function, while in NCM356 cells SAM failed to enhance oxygen consumption and cellular bioenergetics. Longer-term exposure of HCT116 cells to both low concentrations of SAM (0.1 mM) and high concentrations of SAM (1 mM) suppressed cellular bioenergetic responses. Studies presented at the current meeting demonstrate that the H2S-producing enzyme CBS is selectively overexpressed in human colorectal cancer cells, and significantly contributes to their proliferation, migration and invasion in vitro. The present data, demonstrating that short-term exposure of SAM to HCT116 cells exerts proliferative and positive bioenergetic effects that are in line with the effects of exogenously administered H2S in the same experimental system further support the view that CBS in colon cancer cells produces H2S in a regulated fashion, and that this H2S serves as an endogenous cancer cell proliferation and bioenergetic factor. Higher local concentrations of H2S are known to be antiproliferative, at least in part due to their inhibitory on mitochondrial function. Therefore, the longer-term inhibition of cell proliferation and bioenergetic function noted with SAM in the present experiments may either be attributed to the adverse autocrine effects of high concentrations of H2S, produced when CBS is ‘over-activated’ by SAM, or, alternatively, may be related to additional, CBS-independent pharmacological effects of SAM.

Kinetic Mechanism of Protein Arginine Methyltransferase 6 (PRMT6)

The protein arginine methyltransferases (PRMTs) are a family of enzymes that catalyze the mono- and dimethylation of arginine residues in a variety of proteins. Although these enzymes play important roles in a variety of cellular processes, aberrant PRMT activity is associated with several disease states, including heart disease and cancer. In an effort to guide the development of inhibitors targeting individual PRMTs, we initiated studies to characterize the molecular mechanisms of PRMT catalysis. Herein, we report studies on the kinetic mechanism of PRMT6. Initial velocity, product inhibition, and dead-end analog inhibition studies with the AcH4-21 and R1 peptides, as well as their monomethylated versions, indicate, in contrast to a previous report, that PRMT6 utilizes a rapid equilibrium random mechanism with dead-end EAP and EBQ complexes.

S-Adenosyl-Methionine and Betaine Improve Early Virological Response in Chronic Hepatitis C Patients with Previous Nonresponse

Background/AimsTreatment of chronic hepatitis C (CHC) with pegylated interferon α (pegIFNα) and ribavirin results in a sustained response in approximately half of patients. Viral interference with IFNα signal transduction through the Jak-STAT pathway might be an important factor underlying treatment failure. S-adenosyl-L-methionine (SAMe) and betaine potentiate IFNα signaling in cultured cells that express hepatitis C virus (HCV) proteins, and enhance the inhibitory effect of IFNα on HCV replicons. We have performed a clinical study with the aim to evaluate efficacy and safety of the addition of SAMe and betaine to treatment of CHC with pegIFNα/ribavirin.MethodsIn this open-label pilot study, 29 patients with CHC who failed previous therapy with (peg)IFNα/ribavirin were treated with SAMe, betaine, pegIFNα2b and ribavirin. Treatment duration was 6 or 12 months, depending on genotype, and the protocol comprised a stopping rule at week 12 if early virological response (EVR) was not achieved. Virological and biochemical response and safety were assessed throughout the treatment.Results29 patients were enrolled and treated according to the study protocol. 79% of the patients were infected with genotype 1, 72% had advanced fibrosis, 76% had previously received pegIFNα/ribavirin, and only 14% achieved EVR to the previous treatment. When treated with the study medications, 17 patients (59%) showed an EVR, only 3 (10%) however achieved a sustained virological response (SVR). SAMe and betaine were found to be safe when used with pegIFNα/ribavirin.ConclusionThe addition of SAMe and betaine to pegIFNα/ribavirin improves early virological response in CHC.Trial RegistrationClinicalTrials.gov NCT00310336

S-ADENOSYL-L-METHIONINE ATTENUATES OXIDATIVE STRESS AND HEPATIC STELLATE CELL ACTIVATION IN AN ETHANOL-LPS-INDUCED FIBROTIC RAT MODEL

Previous studies report S-adenosyl-L-methionine (SAMe) can exert hepatoprotective effects. At present, the role of SAMe in affecting the activation and/or proliferation of hepatic stellate cells (HSCs) during alcohol-induced fibrotic disease progression is poorly understood. In the human disease state, chronic ethanol intake increases hepatic exposure to LPS and magnifies the hepatic insult leading to fibrosis and cirrhosis. In this study, we developed a "2-hit" ethanol-LPS fibrotic liver rat model with which to investigate the effects of SAMe as a hepatic antifibrotic treatment. Male rats were maintained on liquid diets containing either ethanol or isocalorically matched controls for 8 weeks. Animals received ethanol alone (E), ethanol concomitant with twice weekly LPS injections (EL), or ethanol, LPS, and daily SAMe injections. When using this model, SAMe-treated animals demonstrated significantly decreased fibrosis, oxidative stress, steatosis, and improved liver function versus the EL group. In addition, the EL group showed increased HSC activation, an effect that was abrogated by the addition of SAMe. Analysis of the transforming growth factor-beta (TGF-beta) signaling pathways demonstrated increased hepatic TGF-beta and Smad3 messenger RNA expression in the E and EL groups, which was inhibited in the presence of SAMe. Conversely, SAMe led to increased Smad7 (an inhibitor of TGF-beta signaling) messenger RNA expression. These data demonstrate chronic ethanol feeding combined with LPS induces liver fibrosis, and the addition of SAMe significantly reduces hepatic injury and fibrosis through inhibition of oxidative stress and HSC activation.

S-ADENOSYL-L-METHIONINE ATTENUATES LIVER INJURY IN AN ALCOHOL-LPS MODEL

In vivo studies have shown that chronic alcoholism sensitizes the liver to endotoxemic shock leading to liver microcirculation disruption. Both LPS and chronic ethanol consumption increased hepatic sinusoidal contractile response to the constrictor effects of endothelin-1 (ET-1). The antioxidant S-adenosyl-L-methionine (SAMe) can attenuate and even prevent acute liver injury but very little is known about its role in chronic inflammatory disease progression. In this study we attempted to study the potential beneficial effects of SAMe on the liver microcirculation regulation in an alcohol-induced liver injury model. Male Sprague-Dawley rats were fed an alcohol liquid diet or a control isocaloric diet for 7 weeks with or without a daily injection of SAM (10 mg/kg b.w. i.p). Priming effects of ethanol were studied in a model with or without a semiweekly LPS treatment (1 mg/kg b.w. i.p.). At the end of the diet, animals receiving chronic ethanol and LPS (E-LPS) had an extensive steatosis and a mild infiltration as well as high levels of ALT and gained 66% more weight than animals under SAMe treatment (E-LPS-S). SAMe also significantly attenuated the steatohepatitis and liver injury as shown by the histology and the ALT levels. Alcohol feeding combined to a semiweekly LPS injection for 7 weeks resulted in HSC activation (upregulation of smooth muscle alpha actin, TGF beta and ET-1 RNA levels), and an upregulation of Caveolin-1. SAMe significantly decreased all these parameters when added to the alcohol and LPS treatment. In conclusion, our results demonstrate that chronic alcohol feeding combined with semiweekly LPS injection can induce a moderate liver fibrosis and injury combined with an imbalance in regulators of the liver microcirculation. Addition of SAMe to this chronic insult was able to significantly reduce hepatic injury and restore the balance of hepatic vasoregulators. These results suggest a close link between vascular dysregulation and progression of alcohol/LPS-induced liver injury. Supported by NIH grants AA14891 and DK38201.

DNA-(N4-cytosine)-methyltransferase from Bacillus amyloliquefaciens: kinetic and substrate binding properties

Interaction of DNA-(N4-cytosine)-methyltransferase from the Bacillus amyloliquefaciens (BamHI MTase, 49 kDa) with a 20-mer oligonucleotide duplex containing the palindrome recognition site GGATCC was studied by methods of steady-state and presteady-state kinetics of the methyl group transfer, gel retardation, and crosslinking of the enzyme subunits with glutaric aldehyde. In steady-state conditions, BamHI MTase displays a simple kinetic behavior toward a 20-mer oligonucleotide substrate. A linear dependence was observed for the reaction rate on the enzyme concentration and a Michaelis dependence of the reaction rate on the concentration of both substrates: S-adenosyl-L-methionine (SAM), the methyl group donor, and DNA, the methyl group acceptor. In independent experiments, the concentration of the 20-mer duplex or SAM was changed, the enzyme concentration being substantially lower then the concentrations of substrates. The kcat values determined in these conditions are in good agreement with one another and approximately equal to 0.05 s-1. The Km values for the duplex and SAM are 0.35 and 1.6 microM, respectively. An analysis of single turnover kinetics (at limiting concentration of the 20-mer oligonucleotide duplex) revealed the following characteristics of the BamHI MTase-dependent methylation of DNA. The value of rate constant of the DNA methylation step at the enzyme saturating concentration is on average 0.085 s-1, which is only 1.6 times higher than the value determined in steady-state conditions. Only one of two target cytidine residues was methylated in the course of the enzyme single turnover, which coincides with the earlier data on EcoRI MTase. Regardless of the order of the enzyme preincubation with SAM and DNA, both curves for the single turnover methylation are comparable. These results are consistent with the model of the random order of the productive ternary enzyme-substrate complex formation. In contrast to the relatively simple kinetic behavior of BamHI MTase in the steady-state reaction are the data on the enzyme binding of DNA. In gel retardation experiments, there was no stoichiometrically simple complexes with the oligonucleotide duplex even at low enzyme concentrations. The molecular mass of the complexes was so high that they did not enter 12% PAG. In experiments on crosslinking of the BamHI MTase subunits, it was shown that the enzyme in a free state exists as a dimer. Introduction of substoichiometric amounts of DNA into the reaction mixture results in pronounced multimerization of the enzyme. However, addition of SAM in saturating concentration at an excess of the oligonucleotide duplex over BamHI MTase converts most of the enzyme into a monomeric state.

Transmethylation and affective disorders

S-adenosyl-L-methionine (SAMe) is a naturally occurring metabolite and a major source of methyl groups in the brain. Methylation processes are important in the expression of mood and the failure of methylation may underlie some affective disorders. The administration of SAMe is associated with an increase in both 5-OH-indoleacetic acid and homovanillic acid, being the metabolites of serotonin and dopamine, respectively. SAMe indices mood elevation in depressed patients. Until 1995, 1,400 patients treated with SAMe were investigated in open or controlled trials. The efficacy of SAMe in treating depressive disorders is superior to placebo and comparable to standard tricyclic antidepressants. SAMe demonstrates a particularly low side effect profile. The increase of anxiety is the most common side effect. The addition of SAMe to standard antidepressants may shorten the time of treatment response compared to the use of antidepressants alone, especially in parenteral use (400–800 mg/day, iv.).

Methylation of Halogenated Phenols and Thiophenols by Cell Extracts of Gram-Positive and Gram-Negative Bacteria

O-methylation of 2,6-dibromophenol was studied in cell extracts prepared from Rhodococcus sp. strain 1395. O-methylation activity was enhanced by the addition of S-adenosyl-l-methionine but was not affected by the addition of 5-methyltetrahydrofolate nor by up to 10 mM MgCl(2) or EDTA. By using 2,6-dibromophenol, 4,5,6-trichloroguaiacol, and pentachlorothiophenol as the substrates, O-methylation activity was also demonstrated in extracts from two other Rhodococcus sp. strains, an Acinetobacter sp. strain, and a Pseudomonas sp. strain. A diverse range of chloro- and bromophenols, chlorothiophenols, chloro- and bromoguaiacols, and chloro- and bromocatechols were assayed as the substrates by using extracts prepared from strain 1395; all of the compounds were methylated to the corresponding anisoles, veratroles, or guaiacols, which have been identified previously from experiments using whole cells. The specific activity of the enzyme towards the thiophenols was significantly higher than it was towards all the other substrates-high activity was found with pentafluorothiophenol, although the activity with pentafluorophenol was undetectable with the incubation times used. For the chlorophenols, the position of the substituents was of cardinal importance. The enzyme had higher activity towards the halogenated catechols than towards the corresponding guaiacols, and selective O-methylation of the 3,4,5-trihalogenocatechols yielded predominantly the 3,4,5-trihalogenoguaiacols. As in experiments with whole cells, neither 2,4-dinitrophenol, hexachlorophene, nor 5-chloro- or 5-bromovanillin was O-methylated. The results showed conclusively that the methylation reactions were enzymatic and confirmed the conclusion from extensive studies using whole cells that methylation of halogenated phenols may be a significant alternative to biodegradation.

Kinetics of S-Adenosyl-L-methionine in isolated perfused rat liver and its effects on microsomal enzyme activity

S-Adenosyl-L-methionine (SAMe) levels in liver tissue were reduced by 35% after isolation and washing of the organ. The apparent half-life of SAMe (1,75-25 microM) during liver perfusion was between 120 and 460 min; however the uptake of the labelled methyl group by hepatic tissue of fed rats was low (6%). This value increased to 12% in organs isolated from 24-hour fasted animals. Addition of SAMe to the perfusion medium increased tissue levels of ATP. Except for N-demethylation of aminopyrine and cytochrome c reductase, liver microsomal activity was not affected by treatment with SAMe either in vivo or/and in perfused liver.

Affinity chromatography of Ruta graveolens L. O-methyltransferases. Studies demonstrating the potential of the technique in the mechanistic investigation of O-methyltransferases.

Two discrete furanocoumarin (5- and 8-)O-methyltransferases and a caffeic acid 3-O-methyl-transferase from cell cultures of Ruta graveoleus L. have been copurified by affinity chromatography on 1,6-diaminohexane agarose (AH-Sepharose 4B) linked with S-adenosyl-L-homocysteine (SAH). The furanocoumarin O-methyltransferases, which transfer a methyl group from S-adenosyl-L-methionine (SAM) to the 5- or 8-hydroxyls of linear furanocoumarins, were not retarded by 5-(3-carboxypropanamido)-xanthotoxin (CPAX) immobilized to AH-Sepharose 4B, but addition of SAM to the irrigant buffer led to complete retardation of both enzymes on this affinity system. An analogous phenomenon was observed for the caffeic acid O-methyltransferase, with a ferulic acid ligand coupled to the same insoluble support. SAH was as effective as SAM in promoting binding of the furanocoumarin O-methyltransferases to CPAX and caffeic acid 3-O-methyltransferase to immobilized ferulic acid, respectively. The strong and specific adsorption of these enzymes was abolished by exclusion of SAM or SAH from the irrigant buffer. It is concluded that the enzymes bind first to SAM or SAH, and that this binding process in turn induces the binding site for their specific phenolic substrates or their analogs. Based on these findings, a compulsory-ordered kinetic mechanism for the action of these O-methyltransferases is postulated.

Biosynthesis of puromycin in Streptomyces alboniger: regulation and properties of O-demethylpuromycin O-methyltransferase.

Mechanisms for regulation of puromycin biosynthesis in Streptomyces alboniger were studied by measuring the levels of S-adenosyl-l-methionine:O-demethylpuromycin O-methyltransferase. The enzyme was released in soluble form from mycelia by 3 to 5 min of sonication at 4 C. Maximal specific activities of 0.7 and 0.1 nmol/min per mg of protein were found in cells grown in corn steep liquor-corn starch and Hickey-Tresner media, respectively. In both media, the O-methyltransferase activity rose from low levels to a maximum during midlogarithmic growth and then declined or disappeared completely (in Hickey-Tresner medium) during stationary phase. Either glucose (1%) or ethidium bromide (5 muM) reduced O-methyltransferase formation to very low levels with no effect on overall growth. Complete glucose repression of antibiotic formation occurred on agar. Cells grown in the presence of ethidium bromide continued to produce low enzyme levels after regrowth in the absence of dye, but formed normal amounts of puromycin on Hickey-Tresner agar. The O-methyltransferase, either crude or purified, was rapidly inactivated at 37 C. Each substrate alone, or both together at lower concentrations, protected against this loss of activity. Puromycin inhibited the transferase. Regulation of O-methyltransferase synthesis in S. alboniger includes (i) induction early in growth that is susceptible to catabolite repression and differential inhibition by ethidium bromide, and (ii) protection of the enzyme from inactivation by increased intracellular levels of its substrates. The O-methyltransferase was purified 30- to 40-fold by a combination of protamine sulfate precipitation, ammonium sulfate fractionation, adsorption and gradient salt elution from diethylaminoethyl-cellulose and Sephadex G-200 gel filtration. The enzyme was very unstable, even at low temperatures, upon purification beyond the salt fractionation step, but was stabilized by the addition of S-adenosyl-l-methionine during later stages of purification.