An enzymatic derivative double isotope assay for L -methionine.

Abstract A highly sensitive and specific enzymatic derivative method for the measurement of l-methionine is described. The procedure depends upon the addition of tracer quantities of l-[methyl-3H]methionine and its conversion to [3H, 14C]-S-adenosyl-l-methionine in the presence of [8-14C]ATP and of purified Escherichia coli ATP:l-methionine S-adenosyltransferase (EC 2.4.2.13). The S-adenosyl-l-methionine is separated from the radioactive reactants and its 14C:3H ratio is a linear function of the quantity of l-methionine originally present. The method has been applied to the measurement of l-methionine in various tissues.

In the course of recent studies on the control of tissue levels and turnover of S-adenosyl-L-methionine (1) it became desirable to carry out mea,surements of the concentrations of free L-met.hionine in tissues. It has been previously reported by Schlenk (2) and Ijaldessarini (3) that the levels of L-methionine in animal tissues appear to exert an important coiltrolling influence on the concentrations of S-adenosyl-L-methionine. It has also been shown that excess dietary nL-methionine increases the activity of a number of enzymes, such as arginase, tryptophan pyrrolase, glutamate-oxalacetate aminotransferase, and ornithine &transaminase (reviewed in Reference 4). The increase in activity of ornithine b-transaminase has been specifically studied in rat liver and kidney (5) and in Chang's liver cells (4).
It is also of interest that the feeding of DL-mcthionine to animals at concentrations above 27; causes numerous, profound effects, most of which are pathological, including depression of growth, atrophy of liver, decreased nitrogen retention, and pancreatic damage (reviewed in Reference 4).
Among the 12 amino acids required for the growth of mouse L strain fibroblasts, only L-methionine and L-tryptophan inhibited the multiplication of the L cells when added in escess of concentrations required for maximal growth (6). Excess L-methionine is also toxic to Chang's liver cells in suspension culture (4). Virtually all of the above observations have been made without * These studies were supported by Grants GM 16492 and GM 1183 from the National Institutes of Health and by the Gustavus and Louise Pfeiffer Foundation of New York. direct measurement of L-methionine levels in the tissues in which the pathological or biochemical changes occur. Moreover, there appear to be no systematic studies on L-methionine levels in various animal tissues.
A comprehensive discussion of analytical procedures for the determination of methionine has been given by Greenstein and Winitz (7). A number of titrimetric and calorimetric methods are now largely of historical interest.
l/lore modern met,hotls for measuring L-methionine are the following: (a) the nitroprusside calorimetric technique of XIcCarthy and Paille (8), which has been modified by Vasantha, Aloorjani, and Sreenivasan (9) ; (b) microbiological assays which utilize bacterial production of lacatic acid which may be titrated (10) ; (c) paper and thin lager chromatographic separation, or high voltage electrophoretic selm-ation, followed by estimation wit,h ninhydrin (11, 12); (d) gas chromatographic separation (13) ; or (e) automatic amino a(+d analyzer techniques which utilize ion exchange chromatogr:tph) (14). Whereas the last four techniques possess sufficient sensitivity for tissue analyses, all are time-consuming, tedious, and unsuitable for large numbers of samples. A\loreover, some of the manipulations may result in the oxidation of L-methionine to its sulfoxide and sulfone which may not be measured by caertnin of the analytical techniques.
A rapid, sensitive, and highly specific method for measuring L methionine forms the subject of this paper. The principle of the method depends upon the addition of a tracer quantity of L-[methyZ-3H]methionine (of high specific radioactivity) to the reaction mixture containing the sample to be analyzed, followed by enzymatic conversion of the L-methionine to S-adenosyl-L-nlethionine in the presence of [Q4C]ATP and a purified preparation of hTP:L-methionine S-adenosyltransferase (EC 2.4.2.13) which may be conveniently obtained from Escherichia coli. Tkle reaction promoted by this enzyme, as first described by Cantoni and Durell (15,16), is the following: The cationic product (X-adenosyl-L-methionine) which carries a positive charge by virtue of its sulfonium function ma)-be easily separated from the radioactir-e reactants by retention and elution from ion exchange resins, and its radioactivity may be determined.
The ratio of the two isotopes remains constant unless 4466 Kstirnation of L-Methimine Vol. 246, No. 14 there has been dilution with unl:abeled L-methionine (or ATP). In order to mcasurc L-methionine, therefore, special precautions are required to eliminate interfcrcnce from ATP which is commolily present in tissue extracts at colisiderably higher levels than IA-mcthioninc.
IJnder specified caontlitions, the r&o of t,he two isotol)es in the S-:ldenos~l-I,-mc~thiollillc is a direct, and linear measure of the L-methionine conc~cntrntion and may be obtained front appropriate standard curves ill which known quantities of I,-mcthionine have been carried through the reaction system. The t heorctical justification for these derivations has been considered by Baldessarini and Kopin (I 7) in an analogous system for t,he measurement of S-atlellos~l-L-methiorline.
Apart from its sinlplicity and sen&ivity, the method has the advantages that it does not depend upon qu:rntitst,ive recoveries or conversions, and once the endogenous L-methionine has been diluted with tr:tc,er L-[methyl-31-I]methioaille, partial oxidations or losses of the L-methionine are without influence on the final result. The jljecificity of the method depends upon the high degree of substrate specificity of the adeao~\-ltrailsferase which has recently bee11 systematically defined (18).

EXPERIMENTAL PROCEDURE
.Ilaterials-All solutions were prepared from reagent grade cl~emicals in deionized, glass-distilled water.
Standard solutions of L-methionine and nL-norlcucine were prepared in 0.01 N I-ICI. Reduced glutathione was purchased from P-L Biochemicals.
The amino acid analyzer was calibrated with :t standard mixture of 20 amino acids obtained from Beckman-Spinco.
Spectroscopic quality pdiosane and glycerol were purchased from Matheson, Coleman, and Bell. Tetralithium The commercial [8-1%:]ATP was evaporated in a stream of N2, redissolved in a small volume of water, and passed over a small Dowex 5OW-X2 column in order to reduce the radioactivity of the cont,rols (see below).
f)etermination of Radioacl~z~il~-All measurements of radioactivit.v were carried out in glass scintillation vials containing 15 ml of Bray's solution and 2.0 ml of concentrated ammonium hydroxide.
Discriminator settings were selected in such a way that the counting efficiency of 14C was 42y0 and that of 3H was 14%. The spillover of 14C into the 3H channel under these circumstances was 27 to 2S0/& whereas that oC 311 into the 14C channel was 1 t>o 2%.
All measurements of radioactivity reported in this paper were obtained under these counting conditions. Sufficient counts were :rc,c,unlul:ltc,tl to reduce the error below +5c/,.
Calcztlnlions-The spillover of the 14C counts into the tritium channel was determined by adding S-adeaosyl-L-[1,zethyl-14C]met,hionine (20,000 cpm) in 10 ml of water to the Dowes 5OFV-X2 columns (following the same procedure as is used in t,he Lmethionine assay described below), washing with water, eluting with 2 ml of concentrated ammonium hydroxide, and counting with 15 ml of Bray's solution.
The percentage of spillover after passage through the column (27%) is slightly lower than that found by the direct addition of 2 ml of ammonium hydroxide to a scintillation vial rontaining X-adenosyl-L-[rnefhyl-'*C]rlletllionine (spill = 300;,) owing to t.he extra water contained in the columns.
Corrections for both the spillover of t.he 14C into the tritium channel and for the 3H spillover into the 14C channel were made by solving the following simultaneous equations on a Wang 300 series calculator, model 360K. In the sample calculations presented in Table I, it was not mandatory to correct for the tritium spillover into the carbon channel, since the carbon counts were quite high (> 5000 cpm) and the spillover (1.9% of tritium) was insignificant.
Elimination of the correction for the tritium spill simplifies the calculations as simultaneous equations are no longer necessary and thus the carbon spillover (27.2%) can be directly subtracted from the tritium counts. Ilowever, if the counts in the carbon channel are low ( < 1000 cpm), because of the presence of very low levels of L-methionine ( <O.l nmole) or the use of less active transferase preparations, both corrections are necessary.
Preparation OJ" I!. coli ATP:L-Methionine X-Adenosyltransferase-The enzynle was purified according to the procedure of Tabor and Tnbor (20) with certain modifications.
Operations were conducted near 4" unless otherwise indicated.
E. coli strain B cells (100 g) suspended in 400 ml of 50 mM Tris-HCl buffer, pH 7.0, containing 5 m&f 2-mercaptoethanol were homogenized in a Waring Blendor with 400 g of glass beads for five 2-min periods interspersed with 10.min cooling intervals.
The homogenate was centrifuged for 10 min at 10,000 X g, the precipitate was w-ashed three times with 200 ml of the same buffer and recentrifuged as above, and the supernatants were pooled. The combined supernatant fluids were fractionated with solid ammonium sulfate, and The supernatant fluid was again fractionated with ammonium sulfate (20 to 50yo saturation), dialyzed overnight against 50 KIM Tris-HCl buffer containing 5 InM 2-mercaptoethanol and 20% glycerol, and stored in small aliquots at -15". The purification was approximately 40-fold with a recovery of 30% of the initial activity.
The enzyme could be stored at -15" for at least 3 months without loss in activity.
The final specific activity obtained was 3.5 units per mg of protein when assayed according to the procedure of Lombardini, Coulter, and Talalay (18). One unit of enzyme causes the formation of 1 prnole of S-adenosyl-L-methionine in 30 min at 37". The enzyme solutions contained about 40 units per ml. Preparation oj Tissues for Assay of L-.l~ethiolzine-Tile animals were killed by cervical fracture and the tissues were immediately excised aud washed in cold 0.9% iYaC1 t,o remove extraneous blood. The tissues were then frozen in liquid nitrogen, and thereafter all opcratioiis were carried out in the cold. The tissues were weighed in the frozen state and homogenized with 2 volumes of cold 2% perchloric acid in motor driven Teflorl-glass homogenizers. The homogenates were centrifuged for 15 min at 10,000 X g, and the supernatant fluids vyerc neutralized with solid potassium bicarbonate and recentrifugcd as above. Small anion exchange columns (Dowcx ,4G l-X2,2.5 x 30 mm, 100 to 200 mesh; in the chloride form) were utilized to remove endogenous ATP from the tissue supernatants in order to avoid dilution of the specific activity of the exogenous [VCJATP added to the incubation mixture. After preparation of t,he columns with Dowex resin suspended in water, they were partially dried with a stream of nitrogen to remove trapped water. The above described neutralized supernatant (0.3 ml) was applied to the columns and eluted with water (0.3 ml), and again the trapped solu-tion in the columns was forced into the collecting tubes with a stream of nitrogen.
With t,he MC of this technique all of the sample and eluting water were reproducibly recovered (0.6 ml). The I,-methionine in the tissue:: was quantitatively retained in the eluate.
If the analyses were not to be performed on the same day that the animals were killed, the tissue cstracts were frozen after the first centrifugation.
Approximately 10 ~1 of thiodiglycol were added per ml of centrifuged extract to prevent oxidation of I,methionine.
The neutralization and column procedures were performed on the following day.
Assay for L-Methionine-The conversion of L-methionine to S-adenosyl-L-methionine was carried out in systems originally designed for measurement of adenosyltransferase activity by LIucld et al. Incubations were carried out at 37" with agitation for 2 hours. The reacf,ion was terminated by diluting with 10 ml of cold water and applying the total mixture to Dowes AG 5OWX2 columns (6 x 22 mm, 100 to 200 mesh) in tlrc nnmonium form. 12 minimunl of 200 ml of water was used to wash unreacted L-[methyl-31-I]mctIlioail~c and [8-14C]hTP through the columns.
The S-adenosyl-L-metlliolliile (containing 3H and N.?) was eluted from the columns with 2 ml of concentrated ammonium hydroxide, and counted for 31-I and 14C in a liquid scintillation spectrometer with 15 ml of Bray's scintillation fluid. Two controls in which eit.her the radioactive L-methionine or the ATP was omitted were also included.
A series of standard Lmet.hionine concentrations was also run with each set of detcrminations to provide a standard curve.
Amino Acid Analyzer-The lou, (r column of the Ueckman 120 C amino acid analyzer was us~~rl with the standard buffers to Typical measurements of radioactivity are recorded in Table I. The determinations for low levels of L-methionine are expanded in the inset. The graph does not pass through the origin because no correction for the quantity of radioactive tracer r,-mcthionine (about 0.1 nmole) has been made. determine the L-methionine content of the tissue homogenates. A c<alibrated amino acid mixture and m-norleucine were used to calibrate the instrument.

RESULTS
AXD DISCUSSIOI\I Standard Curve-A linear relationship between the 1% : 3H ratio of the S-adenosyl-L-methionine eluted from the columns and the L-methionine concentration is consistently obtained, as shown in Fig. 1, which covers the range of 0.1 to 10 nmoles of pure L-methionine.
Some of the actual measurements of radio-:&iv& are given in Table I in order to illustrate the accuracy of the method, to give closer insight into the degree of enzymatic conversion obtained, and to show the magnitudes of the experimental values. It should be noted that under these conditions about 75 to 807; of the total amount of L-methionine present in the system is converted to S-adenosyl-L-metllionine at low concentrations (0.1 to I .O nmole) and that this percentage decrc:~scs somewhat at higher concentrations of L-methionine.
Ill order to maintain the highest sensitivity it is desirable that the fraction of L-methionine converted to S-adenosyl-L-methionine be relatively high, although the degree of conversion does not influence the accuracy of the method, obher than by virtue of its influence on the accuracy of counting.
With each set of determinations, two controls are incubated containing all reactants except t'hat L-methionine (labeled and unlabeled) is omitted from one and [8J4C]ATP from the other.

Recovery
of L-methionine added to aliquot of whole rnt brain homogenate An aliquot, of rat brain homogenate (0.1 ml, equivalent to 16.7 mg of wet. tissue) was assayed for L-methionine and then reassayed after a known quantity of exogenous pure L-methionine (1.64 nmoles) was added. After the radioactivity was corrected as in Table I, the W:sH ratio of the S-adenosyl-L-methionine IT-hich ~-as ellzymatically formed was converted into nanomoles of Lmethionine by the use of the standard curve in Fig. 1. Details of t.he assay procedure and tissue preparation are discussed under "Experimental Procedure." Calculation of the content of I,methionine in this brain sample gives a value of 68.4 nmoles per g. A parallel det.ermination on the amino acid analyzer gave the value of 64.8 nmoles per g. Significant radioactivity is found in the column eluates from both of these controls (Table I). These incubated controls have higher radioactivity than those from which the enzyme was omitted.
It is presumed that the control values are partly attributable to the presence of contaminating enzyme activities that may degrade L-methioniue and ATP to cationic products which are retained on the Dowex 50 column.
The remainder of the radioactivity of the controls might be due to very slight chemical transformations of ATP and L-methionine that occur in the solutions or in the reaction system. These control values remain relatively constant over a short time span, but tend to increase as the solutions of L-methionine and ATP age. be uqed longer than about 2 months since it undergoes extensive radiochemical decomposition. The range of quantities of L-methionine that can be measured in the specified system can be extended up to at least 100 nmoles without significant deviation from linearity (results not shown). The sensitivity of the method permits measurement of 0.1 nmole of L-methionine or even less.
Since the standard curve shown in Fig. 1 is based only on the amount of exogenous L-[methyl-Wlmethionine which is added to the reaction system and not the L-[methyl-3H]methionine (about 0.1 nmole) which is also present, the graphical representation of the data as expected does not pass precisely through the origin.
Recovery of L-Methionine Added to Tissue-In order to evaluate the reliability of recovery of L-methionine added to a tissue extract,, a series of experiments was conducted, of which Table II gives a representative example.
Recoveries were studied when known quantities of L-methionine were added to a centrifuged, neutralized brain homogenate and assayed for L-methionine. When 1.64 nmoles of L-methionine were added to an aliquot of a brain tissue homogenate containing 1.14 nmoles of L-methionine, the total recovery of L-methionine was 102% of theory (Table  II).
Recoveries of added L-methionine have been carried out with each set of tissue analyses, and have ranged from 90 to 1057;.

Removal
of Endogenous dTP--Although the measurement of L-methionine in many solutions presents no problem, the enzymatic, isotopic proccdurc requires that the endogenous ATP present in tissues be quantitatively removed so as not to lower the specific activity of the added radioactive ATP.
This removal is accomplished by passing the neutralized, centrifuged tissue samples bhrough a l>owcs 1 column before subjecting them to the enzymatic conversion of L-methionine to S-adenosyl-Lmethionine. Table III shows the results of the passage of an aliquot of liver homogenate (prepared as described under "Experimental Procedure") to which has been added a known amount of radioactivity, either as [W4C]dTP or as L-[methyZ-14C]methionine, through Dowex AG l-X2 (100 to 200 mesh; chloride form). After addition to the column of a water wash equivalent in volume to that of the homogenate, ATP is removed and L-methionine is recovered quantitatively. Tissue Levels of L-Met&o&e-The L-methionine levels of various rat tissues are listed in Table IV. The results obtained   by the enzymatic,  double  isotope  procedure  are compared  to those obtained by direct amino acid analyzer measurement. Close agreement between the values obtained by the two procedures is evident.

Enzyme
Speci$city--The specificity of ATP: L-methionine adenosyltransferase for L-mcthionine is quite rigid as has been shown previously (18). The only I,-methionine analogues that will act as substrates for the E. coli transferase are the methyl and ethyl esters and the N-formy and Kacetyl derivatives, of which only the N-formyl~L-rllcthionine is known to occur naturally, although it may not exist in the free state but as the NH2- First, the method is extremely sensitive, estimating with ease 0.1 nmole; and second, the simplicity permits analysis of a large number (approximately 40) of tissue samples by one person in a single day. If necessary, the sensitivity citn be easily extended t,o 10 pmoles of L-methionine, or perhaps even lower.
All the reagents are commercially available except for the E. coli transferase which can be purified in about 2 working days.
It should be considered that, if analyses are carried out on tissue extracts both before and after passage through Dowex 1 columns for removal of ATP, it should be possible to measure both ATP and L-methionine levels in the same samples.