Identification, Source, and Metabolism of N-Ethylglutamate in Escherichia coli

Abstract

N-Ethylglutamate (NEG) was detected in Escherichia coli BL21 cells grown on LB broth, and it was found to occur at a concentration of ∼4 mM in these cells under these conditions. The same cells grown on M9 glucose medium contained no detectable amount of NEG. Analysis of the LB broth showed the presence of NEG, a compound never before reported as a natural product. Isotope dilution analysis showed that it occurred at a concentration of 160 μM in LB broth. Analyses of yeast extract and tryptone, the organic components of LB broth, both showed the presence NEG. It was demonstrated that NEG can be generated during the autolysis of the yeast used in the preparation of the yeast extract. Growth of these E. coli cells in LB broth prepared in deuterated water showed no incorporation of deuterium into NEG, demonstrating that E. coli cells did not generate the NEG. Cell growth rates were not affected by the addition of 5 mM NEG to either LB or M9 glucose medium. l-[ethyl-²H₄]NEG was found to be readily incorporated into the cells and metabolized by the cells. From these results, it was concluded that all of the NEG present in the cells was taken up from the medium. NEG could serve as the sole nitrogen source for E. coli when grown on M9 glucose medium in the presence of glucose but could not serve as the sole carbon source on M9 medium in the absence of glucose.

During work on developing methods for the analysis of the amino acids generated by recombinant archaeal mutases, I developed procedures for the recovery and analysis of the free amino acids present in cell extracts of Escherichia coli. When these methods were applied to analysis of E. coli grown on LB broth, I always found a large amount of an unknown amino acid. Here I report on the identification of this amino acid as N-ethylglutamate (NEG). NEG has never been reported as a natural product. I demonstrate that NEG is readily taken up by E. coli and can serve as the sole source of nitrogen when the cells are grown on M9 glucose medium.

MATERIALS AND METHODS

Extraction and analysis of NEG in E. coli cells.

The procedure used consisted of sonication of the cells [∼0.5 g (wet weight) BL21-CodonPlus(DE3)] to produce a soluble cell extract (3 ml), trichloroacetic acid (TCA) precipitation of the proteins by the addition of 0.1 ml of 2 M TCA to 1 ml of the extract, isolation of the soluble free amino acids by adsorption on a Dowex 50W-X8 H⁺ column, elution of the amino acids with aqueous ammonia, and formation of the trifluoroacetyl methyl (TM) derivatives followed by gas chromatography-mass spectrometry (GC-MS) analysis of the derivatives (13).

Preparation and characterization of NEG.

A sample of NEG was prepared by the ethylation of l-glutamate in base with ethyl iodide and by the reductive (NaBH₃CN) N-monoalkylation of l-glutamate with acetaldehyde (7). The ¹H-nuclear magnetic resonance (NMR) spectrum (D₂O, 400 MHz) of NEG prepared by the second method had δ 3.615 (1H, t, J = 5.9 Hz, H-2), 3.087 (2H, q, J = 7.3 Hz, CH₂CH₃), 2.395 (2H, m, H-4), 2.073 (2H, m, H-3), 1.30 (3H, t, J = 7.3 Hz, CH₂CH₃) consistent with the proposed structure (underlining indicates the hydrogens being considered). l-[ethyl-²H₄]NEG was prepared by the second method, using [U-²H₄]acetaldehyde. Measurement of the concentration of a prepared solution of the labeled NEG was accomplished by ¹H-NMR analysis of the solution by using a known amount of [2,2,3,3-²H₄]trimethylsilylpropionate (TSP) as the internal standard. The material from the ethyl iodide alkylation was purified by chromatography on the Dowex 50W-X8 H⁺ column with an HCl gradient and by preparative thin-layer chromatography (TLC) using the solvent system consisting of acetonitrile-water-formic acid (88%) (80:20:10, vol/vol/vol). NEG had an R_f of 0.52, compared to an R_f of 0.47 for glutamate in this solvent. The amino acids were observed on the TLC plate with ninhydrin spray reagent followed by heating.

Identification of NEG in LB broth.

The amino acids were recovered from LB broth by adsorption of the medium on a Dowex 50W-X8 H⁺ column, followed by washing of the column with water, and elution of the bound amino acids with aqueous ammonia. The recovered amino acid mixture containing NEG was then purified by preparative TLC and the isolated NEG converted into the TM derivative for GC-MS as described above. Alanine was selected as the internal standard, since it copurified with NEG with the methods used. To further support the presence of NEG in the samples, the Dowex 50-isolated amino acids were further fractionated by elution from a Dowex 50W-X8 H⁺ column with a gradient of HCl. The fractions expected to contain NEG, based on the elution position of the known NEG, were pooled and the HCl evaporated. The NEG was further purified by preparative TLC. NEG was observed by GC-MS, indicating that free NEG existed in the LB broth. Since it was possible that some of the NEG could be present in the LB broth as a peptide or some other covalent adduct, samples of the LB broth or the Dowex 50-isolated amino acids were subjected to acid hydrolysis in 6 M HCl at 100°C for 20 h. (A known sample of NEG was found to be stable under these hydrolysis conditions.) For each sample, the NEG was purified and assayed as described above. The same levels of NEG were detected.

RESULTS AND DISCUSSION

Identification of NEG.

Analysis of the amino acids from E. coli cells grown on LB broth consistently showed an unknown peak in the gas chromatography-mass spectrometry (GC-MS) data. The unknown compound, like the TM derivative, had a predicted mass 28 units higher than the glutamate derivative (Fig. 1). Possible candidates for the compound were either 2-aminopimelate, a known natural product found in plants (11) but never identified in bacteria, and NEG, an amino acid that has never been identified before in nature. The TM derivative of the synthetic NEG has the same mass spectrum and GC retention time as the TM derivative of the unknown compound from E. coli. The spectrum and retention time were different from those of the TM derivative of 2-aminopimelate (Fig. 1). Quantitation of the amount of the compound recovered from E. coli BL21 cells grown on LB broth (2.5 g per 100 ml; Fisher) for 4 to 5 h showed that the cytoplasm of the cells contained a calculated ∼4 mM concentration of NEG, assuming that 70% of the wet weight of the cells was derived from the cytoplasm. A concentration of 4 mM is high compared to the typical values, of around 0.1 to 2 mM, for free amino acids found in bacteria (8, 14). This is approximately the amount of glutamate occurring in these cell extracts (my unpublished results). Of the different amino acids, glutamate generally has the highest concentration. Growth of the cells for 14 h increased the level of NEG by 2- to 3-fold. This level is comparable to the levels of glutamate in LB broth (around 11.2 mM), where it is mostly bound in peptides (10) but some is free (5).

FIG. 1.

Electron impact mass spectra of the TM derivatives of an unknown compound (A), known N-ethylglutamate (NEG) (B), 2-aminopimelate (C), and glutamate (D). The NEG and 2-aminopimelate derivatives both have a molecular weight of 299, whereas the glutamate derivative has a molecular weight of 271. These ions are not indicated because of their low intensities.

Source and chemical nature of NEG.

As NEG was identified in E. coli, the question of its origin arose. Did it arise from the E. coli cells or from the medium? Growth of E. coli cells on M9 minimal medium produced cells with no detectable NEG. This indicated that the compound arose only when the cells were grown on LB broth, that the compound was present in the LB broth and taken up by the cells, or that the NEG was produced from constituents in the LB broth. Isolation and identification of NEG from LB broth (Fisher) demonstrated the presence of NEG (Table 1) . Based on the method used in the isolation, NEG was present in the LB broth in a free form, instead of bound as a peptide or other covalent adduct. Isotopic dilution analysis of the NEG in LB broth by using l-[ethyl-²H₄]NEG as the internal standard showed that it was present at a concentration of 190 μM.

TABLE 1.

Relative amounts of N-ethylglutamate in different sources

Source material	Ratio for amt of N-ethylglutamate/alaninea
E. coli grown on:
LB broth	0.89
LB broth with 5 mM l-[ethyl-2H4]NEG	22.1b
M9 glucose medium	<0.007
M9 glucose medium with 5 mMl-[ethyl-2H4]NEG	<0.007
LB brothc	0.08
Yeast extractc	0.014
Tryptonec	0.020
Yeast cellsd	<0.007
Autolyzed yeaste	0.015

^aThis ratio is based on the 180 m/z intensity of NEG or 184 m/z intensity of l-[ethyl-²H₄]NEG relative to the intensity of the 140 m/z ion for alanine. The choice of alanine as the internal standard was based on the fact that it copurified with NEG under the conditions used.

^bThe recovered NEG contained 96% of the molecules containing four deuteriums.

^cFrom Fisher Scientific Co.

^dThe yeast used was Fleishmann's RapidRise highly active yeast.

^eAutolysis was done by incubating Fleishmann's RapidRise highly active yeast in water for 48 h at 50°C.

Analysis of NEG in yeast extract and tryptone.

Since LB broth is a mixture of yeast extract and tryptone, each of these materials was assayed, and small amounts of NEG were detected in both. Analysis of Fleischmann's dried yeasts did not show the presence of NEG, but NEG was detectable after autolysis of the cells (Table 1). The amount, however, was much less that that found in the LB broth. Since bacterial contamination commonly occurs in the production of yeast extracts, this is also a possible source of the NEG in the yeast extract (1).

Possible routes to NEG.

One must thus conclude that NEG is present in the yeast extract, absent in dried yeasts, and concentrated into E. coli during growth on medium with yeast extract. N-Ethyl-containing natural products are very rare, so a route to formation of the chemical must be considered. The generation of NEG could occur either biochemically or chemically by one of the routes shown in Fig. 2. The most plausible route would be the reaction of l-glutamate with acetaldehyde to form the imine adduct that is subsequently reduced to NEG (Fig. 2, reaction 1). l-Glutamate is present in yeasts, and acetaldehyde could be produced from the decarboxylation of pyruvate or oxidation of ethanol. Alternately α-ketoglutarate could react with ethylamine and after reduction produce NEG (Fig. 2, reaction 2). Ethylamine could arise by the action of N-alkylglycine oxidase on N-ethylglycine (4) or from alanine catalyzed by an l-alanine decarboxylase (12). In either case, the reduction is likely to occur with a reduced pyridine nucleotide, although other ways are possible. A third route to NEG (Fig. 2, reaction 3) could be via the ethylation of glutamate. Exactly what the ethylating agent in this reaction would be is unknown, since S-adenosylethionine, the ethyl analog of S-adenosylmethionine (SAM), is not known as a natural product and is extremely toxic to cells.

FIG. 2.

Possible biosynthetic routes to N-ethylglutamate from glutamate or α-ketoglutarate. e⁻, electron.

Confirming that E. coli did not produce NEG.

Although the autolysis of yeasts was found to produce NEG, there was still the possibility that some of the NEG could be produced by E. coli, by one of the pathways discussed above. Thus, efforts were undertaken to test for its production by E. coli. This was done by incubating an E. coli cell extract (240 μl) containing 4 mM (each) NADH and NADPH, 8.3 mM ethylamine, and 8.3 mM α-ketoglutarate for 30 min at 37°C and looking for the formation of NEG. No NEG was detected.

A test for the production of labeled NEG was also conducted by growing the cells on LB broth supplemented with d/l-[2,4,4-²H₃]glutamate (40 mg/100 ml, 2.7 mM). In the case of the d/l-[2,4,4-²H₃]glutamate incubation, the glutamate recovered from the LB broth after growth was found to contain 93.2% of the total molecules with three deuterium atoms. The remaining unlabeled glutamate (6.8%) represents the glutamate from the medium. Bound cellular amino acids were recovered from the soluble proteins by TCA precipitation from cell extract followed by extensive washing with 0.2 M TCA to remove unbound amino acids. Acid hydrolysis (with 6 M HCl for 24 h at 110°C) of the washed material and GC-MS analysis of the TM derivatives of the amino acids were done to establish the amounts of deuterium in the amino acids. This measurement allowed the determination of whether any of the labeled glutamate had entered the cells and if it had been metabolized. The glutamate recovered contained 12% of the total molecules with two deuteriums and 6% of the total molecules with three deuteriums, indicating that only about 18% of the total glutamate in the cells was derived from the fed labeled glutamate, and of this, about two-thirds had undergone a reversible transamination with the loss of the C-2 deuterium. Label was also found in proline, alanine, lysine and, valine. These results showed that the fed glutamate was taken up and metabolized by the cells. However, the NEG contained <1% of the molecules with deuterium, indicating that neither glutamate nor α-ketoglutarate was a precursor to NEG in E. coli. If NEG arose in E. coli, then it must have come from precursors that were not deuterated by the glutamate. To prove for sure that NEG was not produced by E. coli, cells were grown on LB broth that contained 10% ²H₂O. Again, the recovered NEG had no detectable deuterium (<1%). The glutamate contained 13.6% of the molecules with a single deuterium. Alanine (9.5%), valine (8.7%), and Ile/Leu (2%) contained a single deuterium (amounts indicated in parentheses), showing that deuterium was being incorporated into many of the amino acids. This again indicates that all of the NEG was derived from the medium.

Uptake and utilization of NEG by E. coli.

Growth of E. coli on LB broth in the presence of 5 mM l-[ethyl-²H₄]NEG resulted in a large incorporation of labeled NEG into the cells (Table 1). Due to the large increase in the concentration of the NEG in the cells over the medium, it is clear that the NEG must be actively transported into the cells. E. coli has at least five separate transport systems for the incorporation of glutamate and aspartate (9). There are no data to indicate that any of these transport systems can or would transport NEG. N-Methyl-dl-aspartate, a compound with a structure similar to that of NEG was found to be a competitive inhibitor of aspartate transport in E. coli (6). Since NEG, like proline, contains a secondary amine group, it is also possible that it is transported by one of the three known proline transport systems (2). It is also possible that a transport system specific for NEG is present.

E. coli grew at the same rate and to the same extent when grown in LB broth alone and in LB broth in the presence of 5 mM NEG. The same was true for growth on M9 glucose medium. This indicated that the compound was not toxic to the cells and, at least at a concentration of 5 mM, did not inhibit protein synthesis. Cells grown on M9 glucose medium with 10 mM l-glutamate grew to a 10% higher absorbance level when the medium contained 5 mM NEG as well. To test for the uptake of NEG, E. coli cells were grown on M9 glucose medium (100 ml for 12 h at 37°C) in the presence of 5 mM l-[ethyl-²H₄]NEG either with or without 10 mM l-glutamate. To ensure that all external labeled NEG was removed from the cell pellet recovered from the medium, the cell pellets were first washed with M9 glucose medium containing 5 mM unlabeled NEG and then medium alone (three times) to ensure that all of the labeled NEG present in the medium was removed prior to the sonication of the cells to recover the intracellular NEG. No labeled NEG was found in either group of the cells or their medium after growth, showing that fed labeled NEG was completely metabolized by the cells. The cells grown with glutamate, however, had a high concentration of unlabeled NEG that had to be incorporated into the cells during the 1st washing. I propose that this incorporation resulted from the active transport of the NEG by the glutamate transport system induced when the cells were grown in the presence of glutamate. No incorporation was observed in cells grown in the absence of glutamate, since no glutamate transport system was in place. Repeating the experiment in LB broth with 5 mM l-[ethyl-²H₄]NEG (200 ml for 12 h at 37°C) produced a large increase in the amount of NEG in the cells (Table 1). The concentration of NEG in the medium dropped more than 3-fold over the time course of the growth. NEG could serve as the sole nitrogen source for E. coli when grown on M9 glucose medium without ammonium chloride but could not serve as the sole carbon source in the absence of glucose.

Conclusions.

The data presented here indicate that NEG is a naturally occurring compound that is likely taken up by the glutamate transport system of E. coli and metabolized with no apparent detrimental effects on the cells. The occurrence of NEG in LB broth is very unexpected, since examples of compounds containing N-ethylamino groups are very rare. The ethylamine moiety is also found in very few natural products, but one is theanine (γ-ethylamino-l-glutamic acid), the most abundant free amino acid in tea seedlings (3). The occurrence of NEG is yet another example of the presence of many unknown, as yet unidentified compounds that are present in living organisms.