Primordial reductive amination revisited

Article abstract of DOI:10.1016/S0040-4039(02)02863-0

Amino acids are formed efficiently by reductive amination of α-keto acids under aqueous, conditions with freshly precipitated FeS or Fe(OH)₂ and with NH₃, CH₃NH₂ or (CH₃)₂NH at pH values near their pK_a.

Full text of DOI:10.1016/S0040-4039(02)02863-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 1695–1697
Primordial reductive amination revisited
Claudia Hubera and Gu¨nter Wa¨chtersha¨userb,*
aDepartment for Organic Chemistry and Biochemistry, Technische Universita¨t Mu¨nchen, Lichtenbergstraße 4,
D-85747 Garching, Germany
bTal 29, D-80331 Mu¨nchen, Germany
Received 19 November 2002; accepted 16 December 2002
Abstract—Amino acids are formed efficiently by reductive amination of a-keto acids under aqueous, conditions with freshly
precipitated FeS or Fe(OH)2 and with NH3, CH3NH2 or (CH3)2NH at pH values near their pKa. © 2003 Elsevier Science Ltd.
All rights reserved.
Amino acids are essential constituents of the
metabolism. It is therefore not surprising that the prob-
lem of the primordial sources of amino acids plays a
major role in all theories on the origin of life. The
theory of a chemo-autotrophic origin of life1,2 assumes
a primordial biosynthesis of amino acids by reductive
amination of a-keto acids. It has been reported earlier
that such a primordial reaction channel does indeed
exist.3 We here report experiments to study the mecha-
nism of this reaction.
ducible enough to lead to conclusive comparisons and
for carrying out optimization experiments. It is note-
worthy that this procedure avoids a dehydration of FeS
and therefore it is geo-chemically more relevant than
the previously use of dried FeS. Furthermore, we dis-
covered that the reaction is sensitively dependent on the
pH of the aqueous reaction medium. Therefore, conclu-
sive comparisons could be made only by paying close
attention to the pH value.
Table 1 shows the results of representative experiments
on the formation of alanine, glutamic acid, phenylala-
nine, tyrosine, N-methyl-phenylalanine and N,N-
dimethyl-phenylalanine from 100 mmol of the
corresponding a-keto acid in the presence of 2 mmol
FeS and 15 mmol NH4Cl, NH4HCO3, CH3NH2 and
(CH3)2NH, respectively. Column 3 shows the maximum
yield of the amino acid at the pH given in parentheses.5
Column 4 shows yields at selected pH-values above and
We found that experiments on reductive amination of
a-keto acids in the presence of FeS in water are inflicted
with a reproducibility problem, which was found to be
due to the state of FeS. It was determined that repro-
ducibility can be improved if the reactions are carried
out with FeS that has been freshly precipitated from
FeSO4 for each reaction.4 With such freshly and identi-
cally prepared FeS, the results were found to be repro-
Table 1. Yield of various amino acids by reductive amination with various amino sources. The pH for each yield measure-
ment is given in parentheses. The pH-dependence of the yield is indicated for each pair of amino acid and amino source by
one lower and one higher yield (pH) measurement (Me-Phe N-methyl-phenylalanine; Me2-Phe N,N-dimethyl-phenylalanine)
Amino acid
Amino source
Yield max mol% (pH)
Yield range mol% (pH)
Temp. (°C)
Time (h)
Ala
Glu
Phe
Phe
Tyr
Me-Phe
Me2-Phe
NH4Cl
NH4Cl
NH4Cl
NH4HCO3
(NH4)2SO4
CH3NH2
(CH3)2NH
52 (8.8)
35 (8.9)
53 (9.2)
49 (9.3)
58 (9.1)
83 (10.1)
18 (11.1)
10 (8.6)–20 (9.3)
25 (8.4)–20 (9.6)
32 (8.5)–14 (10.5)
29 (8.6)–38 (10.1)
15 (8.3)–34 (9.8)
51 (7.3)–53 (11)
4 (9.5)–6 (12.7)
50
50
75
75
75
75
75
20
72
17
17
17
17
17
Keywords: reductive amination; ferrous sulfide; ferrous hydroxide.
* Corresponding author. Tel.: 0049-89-2199760; fax: 0049-89-223759; e-mail: info@patent.de
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02863-0

1696
C. Huber, G. Wa¨chtersha¨user / Tetrahedron Letters 44 (2003) 1695–1697
Table 2. Yield of amino acids in the presence of excess
sulfide
below the optimum pH. It appears that the optimum
pH is roughly identical with the pKa of the amino
source. For NH3, optimum pH values from 8.8 to 9.3
correspond to the pKa of ammonia of 9.25. For
CH3NH2 an optimum pH value of 10.1, corresponds to
the pKa of methylamine of 10.4. For (CH3)2NH an
optimum pH value of 11.1 corresponds to the pKa of
dimethylamine of 10.9.
Amino acid
Amino source
Temp./time
Phe
NH4Cl
75°C/17 h
Phe
NH4HCO3
75°C/17 h
Tyr
NH4HCO3
100°C/40 h
mmol Na2S
mol% (pH)
53 (9.2)
46 (9.0)
mol% (pH)
38 (8.6)
31 (8.6)
mol% (pH)
68 (8.8)
58 (8.8)
2
3
4
R%-CO-COOH+R2NH+2H++2e−“
R%-CH(NR2)-COOH+H2O.
31 (9.3)
21 (8.7)
26 (8.7)
No significant difference was detected between the
yields when NH4Cl or NH4HCO3 is used as amino
source. The previous3 failure of reductive amination in
the presence of NH4Cl without carbonate is not
explainable. It may have been due to a lack of pH-con-
trol or due to the state of FeS.
Table 3. Dependence of amino acid yield on the ionic
strength
N-Methyl-Phe
Ionic strength 10−4 mol/cm3
Yield mol%
All attempts have failed to convert oxaloacetic acid to
aspartic acid with FeS at room temperature (3 days),
50°C (3 days) and 100°C (17 h) in the pH range from
6.4 to 9.8, while it has been shown that aspartic acid is
quite stable under these conditions (e.g. 94% remaining
after 3 days at 50°C and pH 9.0). This may be due to
the predominance of a non-reactive intermediate enam-
ine structure in this case, while in the other cases a
reactive imino structure may be assumed to be the
dominant species. Instead of aspartic acid, small
amounts of alanine (up to 10 mol%) were formed,
probably due to decomposition of oxaloacetate to
pyruvate.
21
24
36
51
81
75
71
64
44
42
Table 4. Inhibition of reductive amination by air
Air vol%
Phe mmol
0
0.9
9
30
23
10
2
100
As shown in Table 2, excess amounts of Na2S over the
stoichiometric amount for 2 mmol FeSO4 lead to a
decrease of the yield of amino acids. This may be due
to a blockage of catalytic iron sites.
Table 5. Dependence of reductive amination on tempera-
ture
We have also found that the yield of amino acid
strongly decreases with increasing ionic strength of the
reaction medium. In Table 3, this is shown for N-
methyl-phenylalanine and pH 10.3 to 10.7. A similar
salt effect is found at other pH values and for other
amino acids. Such a negative salt effect is explained by
an ionic mechanism that involves an interaction of
oppositely charged species. It has the practical conse-
quence that the reaction rate can be increased by
decreasing the concentration of ions in the aqueous
medium. Further, it is an important hitherto hidden
parameter for carrying out comparisons.
Mol% Tyr
Temp. (°C)
1 h
17 h
40 h
20
50
75
9
17
39
87
15
66
66
81
26
70
69
86
100
50 mol% of the FeS is replaced by NiS the yield of
phenylalanine is about 50 mol% of that with FeS alone.
If FeS is replaced by freshly precipitated FeCO3 no
reductive amination is observed. Surprisingly, however,
if FeS is replaced by freshly precipitated Fe(OH)2,
efficient reductive amination is observed. In a typical
reaction series with NH4Cl we obtained a maximum
yield of 23 mol% Phe at pH 9.3. This maximum yield is
about half the maximum yield with FeS. The increased
rate in the presence of FeS may be due to a promotion
by sulfido ligands or to the mass effect of a removal of
ferric ions by the formation of pyrite from FeS and
ferric ions, which is well known to be a rapid reaction.6
Importantly, however, the result with Fe(OH)2 shows
that neither the presence of sulfido ligands nor the
formation of pyrite are necessary for reductive amina-
tion. It may be speculated that the primary oxidation
Table 4 shows the effect of an addition of increasing
amounts of air. The reaction is not inhibited by low
amounts of air, which are probably consumed by FeS.
Only large amounts of air inhibit the reaction.
The temperature dependence of the reductive amination
over time is shown in Table 5 for the case of tyrosine.
At 100°C the reaction is complete after 1 h. Even at
4°C, the reaction still proceeds, albeit at a very slow
rate.
If FeS is replaced by sulfides freshly precipitated from
CoSO4, NiSO4, MnSO4, ZnSO4, Ag2SO4, CuSO4,
MgSO4, CrCl3, no reductive amination is observed. If

C. Huber, G. Wa¨chtersha¨user / Tetrahedron Letters 44 (2003) 1695–1697
1697
products are ferric centers and that the polynuclear
colloidal state of freshly precipitated FeS and/or
References
Fe(OH)2 is a necessary requirement for the reaction.
1. Wa¨chtersha¨user, G. Microbiol. Rev. 1988, 52, 452–484.
2. Wa¨chtersha¨user, G. Prog. Biophys. Mol. Biol. 1992, 58,
85–201.
2FeS“FeS2+Fe2++2e−
2Fe(OH)2“2FeO(OH)+2H++2e−.
3. Hafenbradl, D.; Keller, M.; Wa¨chtersha¨user, G.; Stetter,
K. O. Tetrahedron Lett. 1995, 36, 5179–5182.
In spite of intense efforts, it was not possible to demon-
strate the previously suggested2 coupling of the reduc-
tive amination with an activation of the carboxylic acid
group of the amino acids. The carboxylic acid group is
apparently not involved in the reaction center. This was
confirmed by a set of experiments (data not given)
which show that methyl phenyl-glyoxylate undergoes
facile reductive amination to the methyl-ester of
phenyl-glycine. It has, however, been reported earlier
that under very similar conditions amino acids are
activated with CO to form peptides.7
4. All solutions were prepared from doubly distilled water,
which was boiled and cooled under a stream of argon.
Serum bottles (120 ml) were typically charged with 2 mmol
solid FeSO4·7H2O, 15 mmol NH4 source (as chloride,
hydrogen carbonate or sulfate) or alternatively the hydro-
chloride of CH3NH2 and (CH3)2NH, 100 mmol a-keto acid
(as sodium salt), closed with Viton stoppers (Ochs), deaer-
ated and subsequently charged with water and an appro-
priate amount of aqueous Na2S·9H2O for the in situ
precipitation of FeS. Subsequently the serum bottles were
charged with appropriate amounts of HCl or NaOH for
adjusting the pH. The total liquid volume was brought to
10 ml. Thereafter the reaction was carried out for 17 h at
75°C. All chemicals were purchased from Aldrich.
5. For analysis samples of the liquid reaction mixture were
centrifuged and the pH of the supernatant liquid was
measured. The yield of the resulting amino acid was
determined by HPLC 10C18 column with an H2O/MeOH
gradient (0–100% MeOH) and 0.1% phosphoric acid, using
a Merck–Hitachi Pump L-7100 and a Merck–Hitachi UV
detector L-7400 at the appropriate wavelength (254 nm for
aromatic amino acids and 190 nm for aliphatic amino
acids). In addition the yields of the amino acids were
confirmed by reaction with ortho-phthaldialdehyde and
HPLC analysis according to Bober, H. Beckman Report
1986, 12, 3–5.
Our results show that the mechanism of this reaction is
more complex than previously appreciated. They open
a new avenue of investigation regarding the mechanistic
relationship of FeS and Fe(OH)2 in the reductive ami-
nation to amino acids and in other reactions8 of a
presumptive primordial metabolism. Thus our results
seem to point in the direction of a revision of the theory
of a chemo-autotrophic origin of life and thus may
bring us closer to the ultimate achievement of the kind
of synthetic autocatalytic domino reactions, with which
life is perceived as having initiated itself.
Acknowledgements
6. Roberts, W. M. B.; Walker, A. L.; Buchman, A. S.
Mineral Deposits (Berl.) 1969, 4, 18–29.
7. Huber, C.; Wa¨chtersha¨user, G. Science 1998, 281, 670–
672.
We thank Adelbert Bacher and Helmut Simon for their
extraordinary generosity of providing us with the labo-
ratory and with support for carrying out this work. We
appreciate financial support by the Deutsche
Forschungsgemeinschaft.
8. Wa¨chtersha¨user, G. Science 2000, 289, 1307–1308.