Quantitative amino acid analysis by mass spectrometry

Vol. 2, No.2

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Feb. 1960

QUANTITATIVEAMINCACIDANAIXSIS BYMASS SBECTRCMETBY K. Biemann and W. Vetter Department of Chemistry, Massachusetts Institute Cambridge, Mass.

of Technoloa

Received February 1, 1960

The methods commonlyused for amino acid analysis employ an unspecific calorimetric

assay of the individual

ration by chromatography.

componentsafter a very careful sepa-

This latter

step makes such a procedure very

time-consuming and even with the elegant, almost fully

automatic method of

Spachan, Stein and Moore (1958) only one sample per day can be analyzed. Since a much faster but still

accurate method would be advantageous for

studies of peptide structures, vestigated the possibility

amino acid metabolism, etc.,

we have in-

of massspectrometric analysis of mixtures of

amino acids, which was expected to be speedy and accurate because prior separation into the componentswould not be necessary. The spectra of ethyJ. esters of amino acicls exhibit very intense peaks which are strictly

relatively

related to the structure

few, but

of the individual

amino acid (Biemann, Seibl and Gapp, 1959); and even the spectra of isomers, like the esters of leucine and isoleucine,

are sufficiently

the branching of the side chains to permit quantitative the massspectrometric

quantitative

different

discrimination.

because of Thus,

analysis of such mixtures should be pos-

sible in much the sameway as the hydrocarbon analysis which is now widely used in the petroleum industry

(for a review see Dibeler,

However, the extension of this principle

to the analysis of free amino

acids is complicated by the necessity to esterify reproducibly)

1956).

quantitatively

(or at least

the very small amount of the mixture commonlyavailable

a determination,

for such

to convert it to the free esters and to introduce these into 93

Vol. 2, No. 2

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

the mass spectrometer without fractionation

in any one of these steps.

We have now achieved this in the following way: milligram)

is refluxed

Feb. 1960

with 2 ml. anhydrous ethanolic

The mixture (about one hydrochloric

acid (5-l@)

for three hours, the solvent evaporated in a vacuum desiccator and the residue dissolved (or susperidedin 2 ml. dichloromethane.

The hydrochlorides

are con-

verted to the free esters by passing anhydrous ammoniathrough the solution for a few seconds; the solution

is then filtered

into a small, long-necked bulb type distilling is distilled to -60°,

off.

tube and most of the solvent

The tube is then connected to the massspectrometer, cooled

and the remaining dichloromethane is pumpedoff through the oil dif-

fusion pump of the inlet into the inlet to 160°,

from the ammoniumchloride

system. The mixture of esters is then introduced

system of the mass spectrometeti by rapidly

heating the tube

and the spectrum is scanned from mass25 to 200.

Calibratirm

spectra of each individual

the samemanner, using an authentic

amino acid have been obtained in

sample. The relative

peaks used for the measurementof the individual

intensities

amino acids (its massnumber

is given in column 2, Fig. 1) were determined from synthetic

mixtures of known

caaposition containing aspartic acid or phenylalanine as internal The calculation

of the results is facilitated

amino esters exhibit

standard.

by the fact that marry of the

a strong peak with little

and thus can be subtracted one after

of the

or no interference

from others

the other from the spectrum.

The few

remaining ones are obtained by one or two successive approximations or by the use of simultaneous equations. Table I gives the results of somepreliminary mtitures

of amino acids, presented as the relative

rather than the absolute amount, are of interest the necessity of quantitative

transfer

determinations on synthetic molar ratios

in most cases, eliminating

of the sample during the preparation.

However, this could be done without major changes, if essential. found deviate l-5$ *

since these,

from the calculated

The values

ones; as can be seen from Table I, the

A c3c 21-103~ mass spectrometer, equipped with a heated inlet operated at 140°, was.used in these experiments. 94

system,

vol.

2, NO. 2

BIOCHEMICAL

Feb. 1960

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

TABIJ-3I Amino Acid

Mass Number' L

zl-Y ala abu val leu ileu

Mixture A Calcd. Foundh 2.64Q 2.63 1.73 1.85 1.87

1.30

v-0 ser thr

1.75 1.84 1.83

Mixture B Calcd. Foundi 2.16 2.09 1.28 1.31

Mixture C Calcd. Foundj

1.32

1.41

opr0 met

1.40

Me tyr asp glu orn

1.00

1.00

1.03 1.40

1.05 1.41

3.10

3.27

1.44

1.42

0.98 1.01

0.98 1.00

1.00

1.00

1.41 1.66 1.18

1.38 1.68 1.17

1.43

1.44

1.00

Mixture D Calcd. Foundj

1.02

1.60

1.65

1.00 0.80

1.03 0.76 1.21

1.19

(a) m/e of peak used for calculation. (b) (R-CR-NHa)+fragment of (c) originates through further fragmentation of (b). R-CR(NHH2)C0aCaH . 5 . (e) (H~-c?H~)+. (f) fragment of orn-lactam. (d) (CHs-S-C&.) (g) values represent molar ratios. (j) average of 3 determinations.

(h) average of 2, (i) average of 4,

accuracy varies somewhatfor the different

amino acids and also depends on the

particular

At present, we are trying

composition of the mixture.

accuracy obtained with ornithine

and tyrosine

the more commonamino acids not listed

to improve the

and to extend the method to someof

in the table,

like tryptophan,

histidine,

cysteine and, if possible, arginine. Webelieve the main advantage of this new approach to be the speed of the analysis;

each determination requires only 30-40 minutes instrument time, thus

allowing the analysis of a larger number of samples than hitherto following

possible.

the total

This

may be an important factor

in many studies, e.g.,

hydrolysis

of a peptide, multiplicate

analysis of mixtures to increase the confidence in the

results obtained, determination of the amino acid composition of the large number of small peptides obtained in the partial 95

hydrolysis

of proteins,

etc.

Further-

Vol. 2, No. 2

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

more, new amino acids can be analyzed accordingly, calibration

Feb. 1960

once a small, pure sample for

is available.

This investigation National Institutes

was supported by a research grant (RG-5472) from the

of Health, Public Health Service.

References K. Biemann, J. Seibl and F. Gapp, Biochem. Biophys. Res. Comm.,1, 307 (1959). V. H. Dibeler in "Organic Analysis,"

Vol. III,

Interscience

Pub., NewYork, 1956,

P. 387 ff. D. H. Spackman, W. H. stein and S. Moore, Anal. Chem., z(l, 1190 (1.958).

96