J Mol Evol (2010) 70:572–582
DOI 10.1007/s00239-010-9353-z
Amino Acid Homochirality may be Linked to the Origin
of Phosphate-Based Life
Da Xiong Han • Hai Yan Wang • Zhi Liang Ji •
An Fu Hu • Yu Fen Zhao
Received: 16 June 2009 / Accepted: 6 May 2010 / Published online: 27 May 2010
Ó Springer Science+Business Media, LLC 2010
Abstract Phosphorylation has to have been one of the
key events in prebiotic evolution on earth. In this article,
the emergence of phosphoryl amino acid 50 -nucleosides
having a P–N bond is described as a model of the origin of
amino acid homochirality and Genetic Code. It is proposed
that the intramolecular interaction between the nucleotide
base and the amino acid side-chain influences the stability
of particular amino acid 50 -nucleotides, and the interaction
also selects for the chirality of amino acids. The differences
between L- and D-conformation energies (DEconf) are
evaluated by DFT methods at the B3LYP/6-31G(d) level.
Although, as expected, these DEconf values are not large,
they do give differences in energy that can distinguish the
chirality of amino acids. Based on our calculations, the
chiral selection of the earliest amino acids for L-enantio-
mers seems to be determined by a clear stereochemical/
physicochemical relationship. As later amino acids devel-
oped from the earliest amino acids, we deduce that the
chirality of these late amino acids was inherited from that
of the early amino acids. This idea reaches far back into
D. X. Han (&)
Department of Pharmacy, Medical College of Xiamen
University, Xiamen 361005, China
e-mail: daxiong@xmu.edu.cn
H. Y. Wang
The Third Institute of Oceanography, State Oceanic
Administration of China, Xiamen 361005, China
e-mail: why@xmu.edu.cn
Z. L. Ji  A. F. Hu  Y. F. Zhao
The Key Laboratory for Chemical Biology of Fujian Province,
College of Chemistry and Chemical Engineering, Xiamen
University, Xiamen 361005, China
123
evolution, and we hope that it will guide further experi-
ments in this area.
Keywords Homochirality  Genetic codes  Phosphate 
Origin of life  Chiral selection
Introduction
Homochirality means that the monomer units of proteins
are made up exclusively of L-enantiomers, whereas the
monomer units of the nucleic acid polymers DNA and
RNA as well as those of the biologically important poly-
saccharides are associated with D-sugars. It is now recog-
nized that all of the crucial biopolymers associated with
life are homochiral (Bonner 2000). Therefore, the source of
homochirality is one of the most important topics in the
research on the origin of life. In the case of these 20 protein
amino acids, the origin of selection for their homochirality
is still a puzzle. Several controversial hypotheses have been
invoked to explain the phenomenon. The homochirality of
amino acids might have stemmed from an asymmetric
adsorption on chiral mineral surfaces, such as quartz or
clays (Hazen 2001), from consequences at the molecular
level of the violation of parity in the weak interaction (the
so-called Vester-Ulbricht hypothesis; Bonner 2000), from
spontaneous resolution (Saghatelian et al. 2001), or from
asymmetric photoreactions (Bailey 2000, 2007; Jorissen
and Cerf 2002). These proposed physical and biochemical
mechanisms demonstrate how enantiomeric segregation
might have occurred in nature, however, the reason for the
preference for one enantiomer over the other has never
been explained satisfactorily.
In the present study, we demonstrate that amino acid
homochirality, as a unique feature of life, might have
J Mol Evol (2010) 70:572–582 573

originated synchronously with the Genetic Code. When O4

Uracil
Serine H5
amino acids entered into living matter, the formation of the
H3
Genetic Code might have provided the force for preferring
N3
one enantiomer of amino acids to the other. At the stage of Methyl O H6
chemical evolution, the primeval atmosphere of the Earth
must have been the source of a rich variety of pre-bio- O
O2
logical small molecules, so that nucleotides, amino acids, H
O H1'
oligonucleotides, and peptides present in life existed 5' position H2'
N
simultaneously. A primitive relationship between amino O
O
Linking center P
acids and nucleotides could also evolve. This relationship phosphate
is called the Genetic Code. Each amino acid relates to O
one or more trinucleotide sequences. A stereochemical Phosphodiester bond 2',3'-Isopropylidene
hypothesis suggests that the origin of the Genetic Code
Fig. 1 The structure model of the (N) amino acid-50 -nucleotide,
could probably be attributed to stereochemical and/or
similar to the structure of the Ser:Urd compound (dashed line stands
physicochemical interactions between anticodons (three for H-bond, double arrowhead stands for the NOE action)
successive nucleotide residues in RNA) or codons (three
successive nucleotide residues in DNA) and amino acids
(Di Giulio 1997, 1999; Inous et al. 1993; Speight et al.
a -
O O
2001). When the most archaic forms of the Genetic Code P
O O OH -
had established the physicochemical/stereochemical inter- O O- :NH2CHCOO- -O
O O O
P O P O P NHCHCHOO
actions between nucleotides and amino acids, the steric P P
R
O- O- O-
R
O
interaction must have been involved in the chiral selection O- O amino-acid P3-(N)amino acid
of amino acids. Based on the above hypothesis, it was P3m
suggested that the Genetic Code could initially related to
Base
mixed oligomers consisting either L-amino acids and b :OH R
O
D-ribonucleotides, or D-amino acids and D-ribonucleotides O
NHCHCHOO-
OH OH P
compounds, respectively (since D-ribose is generally con- -
O O O OH - -
O O
Base
P2
O P O P O P
sidered to have been the prior selection). Since the mixed O- O- O- NHCHCHOO- O

oligomers preceded RNA and L-proteins, the origin of R OH OH

amino acid homochirality evolved in an environment in (N)amino acid-5'-nucleotide

which the mixed oligomers were the dominant molecules.
Fig. 2 a Mechanism of phosphorylation of amino acids by P3m,
Then, in the process of oligomer polymerization to RNA, b formation of (N) amino acid-50 -nucleotides
D-ribose RNA was stereoselective for L-amino acids, and so
caused the chiral selection with the origin of the Genetic
Code.

A Model for the Origin of Homochirality

Uracil
Phosphorylation has to have been one of the key events in
prebiotic evolution on earth. Treatment of amino acids and
nucleosides with polyphosphate or cyclic trimetaphosphate
(P3m) gives good yields of phosphoryl amino acids and nucleophile attacking
nucleotides under reasonably plausible prebiological con-
ditions (Chung et al. 1971; Tsuhako et al. 1984; Westhei-
OH 2.58 Å OH
mer 1987; Yamagata et al. 1991; Inous et al. 1993; Cheng
OH
et al. 2004; Pasek et al. 2008; Ni et al. 2009). Once such a H-bond
2.82Å
single phosphoryl N-amino acid had formed, it makes O
O
possible a reaction with nucleosides by 50 -OH attack to P P
P
produce amino acid 50 -nucleotides (aaNs) (Figs. 1, 2, 3).
These phosphoryl monomer species then could react fur- penta-coordinated
phosphorus P3-L-Ala
ther and condense to form peptides and oligonucleotides.
In this study, we propose that the formation of the aaNs Fig. 3 Mechanism of the reaction of P3–(N) amino acid (such as
having a P–N bond could have been important as a model P3–L–Ala) with 50 -nucleoside (Uracil) by the orientation of H-bonds

123
574
Fig. 4 Schematic a (N)amino acid-5'-nucleoside
representation of the stack of
(N) amino acid-50 -nucleotides. side-chain
a The lateral view of aaN COOH
polymer, b the planform view of
L- and D-monomer
D-aas NH
P+
L-aas
nucleoside
O O
OH
OH
L-enantiomer
for chiral selection. The early bimolecular evolution would
have been carried out on organic surface layers rather than
in bulk solution. For example, positively charged mineral
layers concentrated simple phosphorylated nucleosides
from dilute solution and condensed them to form an oli-
gomer that contains a Genetic Code. Our current view on
this is that single nucleotides and amino acids were prob-
ably linked at random prior to the stereoselection of chiral
amino acids. As shown (Figs. 4, 5), the aaNs were arranged
orderly on flat surfaces under constraints that maximized
the degree of contact of polar groups with the surface, as
similarly described by Bailey (1998). When the arrange-
ment of nucleoside parts adopted a phospho-ribose back-
bone mode similar to the present RNA, the chiral
environment of RNA and steric interference would prevent
access of some isomers of amino acid parts in aaNs. Here,
only aaNs with a favorable symmetrical structure could
aggregate and polymerize. And in the aaN polymer, the
close proximity of the amino acid side-chain and nucleic
acid bases was also possibly in favor of Genetic Code
relationship formation. The aaNs as a model for chiral
selection presented here are speculative, but they have
served as guide for experiments. For example, we synthe-
sized some di-nucleoside phosphoramidates, and their
properties and bioavailability were investigated by Mass
Spectrometry (Qin et al. 2002) and NMR (Lin et al. 2003).
123
J Mol Evol (2010) 70:572–582
base
P+
P+
P+
surface-bound aaN polymer
OH
OH
D-enantiomer
It would be described later how the emergence of the
aaNs was featured as important reagents in the origin of
phosphate-based life.
Formation of aaN Compound
The genetic materials DNA and RNA, as well as many
intermediary metabolites are phosphodiesters, most of the
coenzymes are phosphate or pyrophosphate esters, the
principal reservoirs of biochemical energy such as adeno-
sine triphosphate (ATP) and creatine phosphate are also
phosphates. Thus, phosphates and pyrophosphates are
essential intermediates in biochemical syntheses and deg-
radation (Westheimer 1987). Based on these facts, we infer
that phosphates would have been selected as operators at an
early stage in chemical evolution. The spontaneous poly-
merization from free amino acids and nucleotides to pep-
tides and oligonucleotides occurs at a very low rate in
water. There is no doubt that clays, sand, and many min-
erals could concentrate biomolecules and catalyze their
polymerization (Ferris et al. 1996; Zaia 2004). In these
complexes, since minerals with metals have a stronger
catalyzing capability, the surface adsorptions of minerals
are very important. These positively charged minerals will
selectively concentrate negatively charged molecules for
ordering and surface-induced polymerization reactions. If
J Mol Evol (2010) 70:572–582
Fig. 5 The modeling graphic of a Code triplet
the (N) amino acid-50 -nucleotide
polymerization. 1, 2, 3 represents
a code triplet. a the planform
view of the modeling graphic,
b the lateral view of the modeling
graphic. The dashed rectangular
represents a monomer group 5'-
COOH
5'-
O
Surface affinity
biomolecules are to be negatively charged, prime consid-
eration should be given to phosphate as having a central
function for life. Trivalent orthophosphate ions not only
fulfill the functions of a connector for the phosphate-ribose
backbone, but also retain the negative charge necessary for
concentration and retention on mineral surface by elec-
trostatic interaction of positive and negative charges.
Consequently, as an important reagent for the origin of the
Genetic Code and amino acid homochirality, the aaNs
should similarly involve a phosphorylation procedure.
Both from experimental simulation and analysis of
volatile condensates in volcanic gas and volcanic activity
could produce water-soluble polyphosphates consisting of
PO43-, P3O93-, linear oligophosphate, and P3m through
partial hydrolysis of P4O10 (Yamagata et al. 1991). Under
sterile conditions at Mg/Ca ratios found in natural waters,
most phosphate from hydrolysis of polyphosphates would
precipitate as insoluble salts (Kolb et al. 1997). P3m, a six-
membered cyclic tri-phosphate, readily undergoes a ring-
opening reactions in the presence of amino acids to give
P3–(N)amino acids with a P–N bond (Tsuhako et al. 1984).
Thereafter, P3m reacts with nucleosides to form aaNs, fol-
lowed by the release of the product, P2 (Fig. 2). As a
reaction mechanism, we propose the following. P3–
(N)Amino acids are protected by their anionic charge from
rapid attack by water and other nucleophiles under an
575
1 2 3
Uracil
3'-
(N)amino acid-5'-nucleoside
OH
Serine
3'-
O Phosphate species
P- P- P-
M+ M+ M+ M+
Cationic minerals
aqueous environment. And the 20 ,30 -OH group of the
ribonucleoside could form two pairs of strong H-bonds to
phosphorus oxygen atoms of P3–(N)amino acids (Fig. 3).
The distance of 20 -O to 30 -O (about 2.58 Å) is close to that
of the two phosphorus oxygen atoms (about 2.82 Å), so the
two H-bonds (P)O–H(O) can be well formed simulta-
neously. The consequence of H-bond formation should
assist the nucleoside 50 -OH nucleophile attacking the
phosphorus atom of the phosphoryl group (this is the most
electrophilic center of the P3–(N)amino acid), forming
penta-coordinated phosphorus intermediates and opening
the triphosphate (Figs. 2b, 3). Through a study on the
mechanism of the biomimetic reactions of phosphorus
participation in our laboratory (Lu et al. 2002; Li et al.
2004; Tan et al. 2004; Fu et al. 1997, 1999), we have
proposed that the formation of aaNs with P3m could pro-
ceed through involvement of penta-coordinated phosphorus
intermediates, which are considered to be metastable
intermediates or transition states in the action of ribonu-
cleases and similar phosphodiesterases.
Since the production of aaNs has not been observed in a
water system in the presence of P3m, amino acids, and
nucleosides, we previously carried out a preliminary study
(Zhou et al. 1996). For example, the reaction products of
N-(O,O-diisopropyl) phosphoro-threonine (DIPPThr) and
the four ribonucleosides (adenosine (A), uridine (U),
123
576
cytidine (C), and guanosine (G)) at room temperature in
anhydrous pyridine gave the following products: the oligo-
peptide (Thr)2-4, DIPP(Thr)2-4, the denucleotides XpX,
(Xp)2, and the conjugates of amino-acids and denucleotides
(Xp)2-Thr, structures were determined by FAB-MS, HPLC,
UV, and CE-MS techniques (X represents A, U, C, and G,
respectively). The occurrence of the above oligomers might
suggest that peptides and oligonucleotides could have been
formed simultaneously on the primitive earth. In particular,
the occurrence of the (Xp)2-Thr product and the strong
hydrolytic stability of phosphodiester bonds (Arrhenius
et al. 1997) has led us to speculate that free amino acids
and nucleosides could have been connected by charged
phosphodiester bonds in the course of chemical evolution.
Consequently, the existence of the P–N amide linkage is
extremely favorable for interactions between the side-
chains of amino acids and nucleoside bases in comparison
with free states (the separated individual aa with free
nucleotides). The strength of the interaction would depend
on the orientation of the side-chains relative to the bases on
the nucleosides. Moreover, the asymmetry of D-ribose
would also display an abiotic selectivity for one enantiomer
of the amino acids. Thus, specific aaNs could be identified
as progenitors of homochirality.
To test the possibility of oligonucleotide-directed ste-
reoselectivity, we synthesized a series of aaN derivatives in
acetonitrile (CH3CN). The conformation of the synthesized
compounds was characterized by NMR nuclear Overhauser
effect (NOE) techniques (shown in following theoretical
studies). NOE peaks between amino acid side-chains and
nucleoside bases were not directly detected. This showed
that a fixed configuration of aaNs could not form in free
solution without rotational constraints. And it is difficult to
display stereoselectivity for either D- or L-enantiomers of
amino acids. Therefore, configurational constraints (an
analogous situation shown in Fig. 5) must have been
involved to enable RNA uniquely to show stereoselectivity
for L-enantiomers.
Materials and Methods
Modeling Build
It is supposed that the two structures of the aaN polymer in
Fig. 4a could occur. Their monomer structures (see Fig. 5)
were extracted for theoretical calculations. And their rela-
tive stability with D- and L-amino acids was then explored
to test the polymerizing structure directing the chiral
selection of amino acids. It is important firstly to analyze
the possible structure of the surface-bound aaN polymer
before theoretical calculations. We have proposed that the
oxygens of phosphate and the 50 -oxygen are coplanar and
123
J Mol Evol (2010) 70:572–582
in contact with a polar or mineral surface, providing a
relatively fixed aaN polymer backbone. The D-ribose ring
inclines to the polar face, whereas amino acids and the
bases of nucleosides project out of the polar face with
undefined orientations (Fig. 5). The nucleoside bases
mainly have the syn- and trans-conformation of RNAs.
Based on earlier study (McGuigan et al. 1993; Adelfins-
kaya and Herdewijn 2007), we synthesized a series of
nucleoside derivatives to ascertain the base conformation,
such as 20 ,30 -isopropylideneuridine 50 -methoxyserinyl
phosphamide in Fig. 6a (Ser:Urd compound). As can be
seen from NOE spectra (Fig. 6b), the uracil H6 gives
notable NOE signals with respect to H10 and H20 of the
ribose ring, showing that the Ser:Urd compound favors the
syn-conformation. This can be attributed to the two
H-bonds formed between carbonyl and hydroxyl groups of
serine and the uridine base, the first involving serine car-
bonyl oxygen and the base N–H3 and the second involving
the base O2 and the serine b-OH hydroxyl (Fig. 1). Like-
wise, steric interactions between bases and side-chains also
exist in the other synthetic nucleotide derivatives. For
nonpolar side-chains such Ala and Val, H-bonds can be
formed between carboxyl groups and base hydrogens,
whereas polar aromatic amino acids facilitate significant
aromatic–aromatic stacking with the bases. When polar
side-chains are replaced by nonpolar groups, the NOE
signals between Uracil H6 and H10 /H20 of the ribose
disappeared (Fig. 7), as seen for the NOE spectrum of
diisopropyl 20 ,30 -isopropylideneuridine 50 -O-phosphamide
(diisopropyl:Urd compound). These results provide an
extremely important experiment basis for the base syn
conformation of the aaN polymer model.
In the free state, such as in solution the amino acid side-
chains might have 360° rotational freedom around the P–N
bond. However, when the aaN molecules are constrained
on a surface monolayer, in an arranged order, the polar
groups of amino acids would automatically orientate to
nucleoside bases on the basis of our topological observa-
tions and the syn-conformation study of bases. For the aaN
polymer model, the positions of amino acid side-chains in
Figs. 4 and 5 are the most favorable, while other positions
would generate steric interference with neighboring
nucleosides.
Considering that the calculations for full length aaN
polymer molecules need a lot of computer time and
resources, a simpler and more practical model should be
defined from the structures of the aaN polymer. According
to the triplet code theory (Yarus 1998; Seligmann and
Amzallag 2002), the first two nucleotides generate a ste-
reospecific interaction with the amino acid side-chain, with
the second nucleotide being vital. For this purpose, amino
acids are phosphorylated at the 50 -position of the second
nucleotide to dissect the determined factors that regulates
J Mol Evol (2010) 70:572–582 577

Fig. 6 a The structure of a b

the Ser:Urd compound,
O
b The 400 MHz NOE spectra O H6 H1' H2'
of the Ser:Urd compound in O
OH
DMSO HN

O N H6
HN

HO P O
O

O H H 2'

H H1 '
O O

Fig. 7 a The structure of the a b

diisopropyl:Urd, b The O

400 MHz NOE spectra of H6 H1' H2'

the diisopropyl:Urd compound NH

in DMSO
H6
N O
N

HO P O
O

O H H 2'

H H1 '
O O

the selection on chirality. A starting structure of the
monomer group (the dashed rectangular in Fig. 5a) might
be abstracted to carry out the theory calculations. By
comparing the conformation energy, the relative stability
of the L- and D-monomers could be estimated; then, the aaN
polymer selection for amino acid chirality could be
deduced.
Calculations
All calculations in this article were performed on the
GAUSSIAN03 packages, utilizing 4 CPU parallel com-
puting in a cluster group of 20 Dell PC computers. The
starting geometric structure of the monomer group was
optimized using DFT methods at the B3LYP/6-31G(d)
level. A full convergence optimization was used with the
default cutoffs under GAUSSIAN03. The DFT theory
provides a partial account of correlation correction, so it is
a suitable level of theory for systems containing P and N
atoms. Although the convergence of system energies is
subject to the interaction between nucleoside bases and
amino acid side-chains, the conformations are strikingly
close to global minima. Calculations that use only HF
energies without adding the zero-point energy and thermal
energies (vibration, rotation, and translation) are explicitly
indicated with the conformation energy, E.
Differences of L- and D-conformer’s energies are eval-
uated for all aaN compounds: DEconf = EL - ED. EL
represents the conformation energy of L-monomer while
ED stands for D-monomer. The DEconf values are shown

123
578 J Mol Evol (2010) 70:572–582

Table 1 The differences

A C G U
between L- and D-conformation
energies (DEconf) of the 152 (N) Early aa
amino acid 50 -nucleotides
(kcal/mol) 1 Ala -7.10 -3.40 -5.50 -1.49
2 Ser -6.37 -5.59 -4.12 -4.11
3 Asp -4.82 -22.70 -3.36 -0.54
4 Glu -2.15 -22.50 -34.4 -11.09
5 Val -3.28 -3.12 -0.26 -1.40
Probably early aa
6 Pro -10.79 -0.27 -1.47 -5.59
7 Thr -6.54 -1.44 -3.70 -3.90
8 Leu -4.21 -4.15 -0.57 -0.99
9 Ile -3.86 -3.20 -0.42 -1.20
Probably late aa
10 Phe -8.61 -5.97 -2.05 -10.45
11 Tyr -1.76 -0.28 7.73 -0.53
12 Cys -2.72 -7.09 0.05 -4.67
Late aa
13 Arg -9.20 -2.89 -0.04 -2.63
DEconf = EL - ED. EL 14 Asn -8.81 -8.08 -2.48 -9.54
represents the conformation
energy of L-monomer while ED 15 Gln -12.23 -7.39 -14.90 -1.32
stands for D-monomer 16 Lys 5.07 -8.81 -22.60 1.06
Bold text represents that the 17 His 1.64 -12.88 5.31 -5.16
DEconf values are positive, the 18 Met -6.65 -9.16 1.03 -1.34
positive values are only 10.5%
19 Trp -2.66 0.38 -2.09 -0.81
of all data

(Table 1). For L-compounds, their formation potentials
(DEform) are simply equal to the differences between the
total potential of isolated fragments (including amino acid
Eaa, base Ebase, D-ribose Eribose, and phosphate Ephosphate)
and the potential of the aaN compounds (EL). The DEform
values (DEform = EL - (Eaa ? Ebase ? Eribose ? Ephosphate))
are shown (Table 2).
Results and Discussion
Selection Mechanism of Amino Acid Chirality
Although these DEconf values are not large, they are of the
order expected and give differences in energy to distin-
guish the chirality of amino acids. For negative DEconf
values, it was found that the aaN polymer consisting of
L-amino acids would be more stable than that of D-amino
acids. In other words, the asymmetric aaN polymer would
have a preference for L-amino acids selected from a race-
mic mixture. Statistical data indicate that the relative rate
of selection of L-amino acids for forming the aaN polymer
is 89.5% and that of the D-amino acids is 10.5% (Corre-
sponding to the Table 1, positive values are only 10.5%,
but negative values are 89.5%). Here, the DEconf notably
depends upon steric constraints and orientation between
side-chains and bases, and is an evaluating parameter for
the conventional thermodynamic intensity of H-bonds, Van
der Waals, electrostatic, and so on.
The cytidine aaNs with typical type of side-chains were
chosen as the examples. Their optimum conformations
clearly demonstrate that the DEconf arises mainly from the
difference of H-bonds as shown in Figs. 8 and 9, in which
the distance between the H atom of donors and the O or N
atom of acceptors is smaller than 2.0 Å. For valine (Val),
with a strictly nonpolar side-chain, the L-monomer forms
three H-bonds between the O atom of the Val carboxyl
group and the base H–N4, the H atom of carboxyl group
and the base N3, and the base O2 and the H of C20 -
hydroxyl group, respectively, whereas D-monomer only
forms one H-bond between the H atom of the Val amino
group and the base O2 (Fig. 8a). For the Serine (Ser) with
an uncharged polar side-chain, the L-monomer forms three
H-bonds: between the H atom of Ser carboxyl group and
the Ser hydroxyl group oxygen; the H atom of Ser amino
group and the base O2; and the H atom of Ser hydroxyl
group and the base O2, respectively. In contrast, the
D-monomer only forms two H-bonds between the H atom
of Ser amino group and the base O2 and the H atom of Ser
hydroxyl group and the base O2, respectively (Fig. 8b). For

123
J Mol Evol (2010) 70:572–582 579

Table 2 The relative formation

A C G U
potentials (DEform) of 78
L-enantiomers of (N) amino
Early aa
acid-50 -nucleotides (kcal/mol)
1 Ala -6.10 -6.36 -9.85 -1.93
2 Ser -9.69 -10.36 -9.21 -8.47
3 Asp -35.55 -53.64 -43.02 -31.57
4 Glu -35.55 -41.13 -57.21 -47.17
5 Val -6.87 -4.32 -5.831 -2.38
Probably early aa
6 Pro -6.74 -2.12 -7.35 -7.56
7 Thr -9.72 -11.89 -9.06 -8.04
8 Leu -5.65 -4.16 -4.63 -1.08
9 Ile -5.15 -4.34 -4.11 -0.31
Probably late aa
10 Phe -8.95 -1.96 -5.90 -6.20
11 Tyr -2.49 -0.94 -8.85 -3.73
12 Cys -5.00 -6.05 -4.15 -3.17
Late aa
13 Arg -28.28 -33.88 -25.98 -21.30
14 Asn -7.60 -3.76 -7.10 -9.41
15 Gln -19.01 -14.19 -17.07 -6.63
The formation potential of the
16 Lys -21.40 -38.12 -36.89 -27.17
Trp:Urd compound is regarded
as the reference value; bold text 17 His -6.15 -19.67 -12.32 -7.22
represents the minimum 18 Met -5.99 -12.36 -3.90 -3.11
potential for each amino acid 19 Trp -0.74 -4.44 -14.42 0
compound, respectively

aspartic acid (Asp) with a negatively charged side-chain,

a COOH H-N4 H-N4 the L-monomer forms three pairs of H-bonds: between the
O atom of Asp carboxyl group (side-chain) and the base
N3
N3 H–N4; the second O atom of Asp carboxyl group and the H
O2 atom of Asp amino group; and the base O2 and the H atom
NH

O2
COOH
NH
of C20 -hydroxyl group, respectively. Once again, the
P P C2'-OH D-monomer only forms two pairs of H-bonds between the
O atom of Asp carboxyl group (side-chain) and the base
C2'-OH H–N4 and the same O atom and the H atom of Asp amino
group (Fig. 9a). For Lysine (Lys) with a positive-charged
L-Val Cytidine D-Val Cytidine
side-chain, the L-monomer forms two pairs of H-bonds
between the H atom of Lys NH3 group (side-chain), H
b
atom of C20 -hydroxyl group and the base O2, whereas
COOH D-monomer only forms one H-bond between the H atom of
OH
OH O2
Lys NH3 group (side-chain) and the base O2 (Fig. 9b).
From the H-bond estimation of above representative sam-
O2 NH
ples, it can be seen from Fig. 4b that bordering upon
0
NH L-amino acid side-chains, the C2 -hydroxyl group of
COOH D-ribose with potential for H-bond formation would be an
P P
important factor to stabilize the L-monomer. However, the
furanose ring oxygen of D-ribose associating with D-amino
L-Ser Cytidine D-Ser Cytidine acid side-chains seems to be incapable of forming H-bonds.
Since the asymmetric structure of the aaN polymer envi-
Fig. 8 The differences of H-bonds in the optimum conformations of
L- and D-monomers (dashed line stands for H-bonds): a the Val and ronment (mainly D-ribose) serves as templates for the
Cytidine conjugate, b the Ser and Cytidine conjugate selection of amino acid enantiomers, the chiral selection

123
580
a COO-(side-chain)
COO-(side-chain)
H-N4
NH O2
NH
P P
C2'-OH
L-Asp Cytidine D-Asp
b NH3(side-chain)
NH 3(side-chain)
COOH
NH
O2
COOH
P NH
C2'-OH P five (excluding Gly), were selected as L-enantiomers. As
L-Lys Cytidine D-Lys
Fig. 9 The differences of H-bonds in the optimum conformations of
L- and D-monomers (dashed line stands for H-bonds): a the Asp and
Cytidine conjugate, b the Lys and Cytidine conjugate
seems to be determined by a strict stereochemical/physi-
cochemical preference, which is apparently related to the
origin of the Genetic Code. The stereochemical hypothesis
suggests that the origin of the Genetic Code can probably
be attributed to stereochemical and/or physicochemical
interactions between anticodons or codons and amino acids
(Di Giulio 1997, 1999; Szathmary 1993; Di Giulio and
Medugno 1998). Therefore, we have speculated that
L-amino acid homochirality was predetermined by the prior
evolution of D-ribose RNA. Similar to the opinion, bailey
has suggested that L-amino acid homochirality was prede-
termined by the prior evolution of D-ribonucleic acids and
probably was chirally directed by the orientation of early
RNA molecules on surface monolayers (Bailey 1998).
The emergent sequence of the amino acids from top
(sequence number = 1) to bottom (sequence num-
ber = 19) follows their original order in Table 1. Among
them, Ala, Ser, Asp, Glu, and Val are the earliest amino
acids (Trifonov 2000; Di Giulio 2004). Clearly, as far as
the earliest amino acids are concerned, all of the DEconf
values are negative. For some other amino acids, especially
the later amino acids, there appear a few positive DEconf
values (bold text in Table 1), indicating that L-enantiomers
did not have a selection priority to form aaN polymer over
D-enantiomers. For these cases, there are two possible
explanations. First, the positive DEconf values involving
Lys(5.07(A), 1.06(U)) and His(1.64(A), 5.31(G)) with
basic side-chains, Tyr (7.73(G)) and Trp (0.38(C)) with
hydrophobic aromatic side-chains, and Cys (0.05(G)) and
123
J Mol Evol (2010) 70:572–582
Met(1.03(G)) with sulfur containing side-chains, have an
H-N 4 unfavorable physicochemical character to form stronger
interactions such as H-bonds with the bases and C20 -
O2 hydroxyl group. In addition, the relatively bulkiness of
side-chains of the latter amino acids leads to difficulties in
attaining the optimal conformation of the aaN polymer
monomer, and that makes the DEconf values uncertain.
Second, the positive values for DEconf indicate that the
C2'-OH
stereochemical/physicochemical hypothesis is not suffi-
Cytidine cient on its own to account for the chiral selection of the
later amino acids. Similarly to the co-evolution theory of
the origin of the Genetic Code, the biosynthetic pathways
of amino acids played a fundamental role in defining the
O2 organization of the Genetic Code (Wong 1975; Wachter-
shauser 1997; Di Giulio 1998). We infer that the amino
C2'-OH acids which had first appeared (the precursors), possibly
other amino acids developed biosynthetically from these
Cytidine precursors, the chirality of the precursor amino acid in life
was inherited by the product amino acids. Finally, with the
evolution of amino acids, the homochirality in life gradu-
ally came into completion.
Selection of Nucleosides
Our modeling study showed that in addition to chiral pri-
orities, L-amino acids might have some selectivity for their
corresponding nucleosides based on the mode of aaN
polymer. As the products of all aaN reactions have the
same bond cleavage or formation, the formation potential
(DEform) is simply equal to the difference between the total
potential of isolated fragments (including amino acid, base,
D-ribose, and phosphate) and the potential of the compound
system (i.e. the potential difference before and after reac-
tion) (Yang and Han 2000). If the DEform of the Trp:Urd
compound (the tryptophane and uridine nucleotide conju-
gate) is set as to the reference value, the relative DEform
aptly exhibits the stability difference of the conformations
with the specific steric interactions between amino acids
and nucleotides. It seems unlikely that the first RNA con-
tained all four bases found in contemporary RNAs. The
codon amino acid assignment was hypothesized to evolve
from simplicity to complexity: the amino acids were only
codified by G and C in the early Genetic Code, and then
expanded from the G and C toward the complete G, C, A
(adenosine), and U (uridine) (Di Giulio 2004; Woese 1965;
Szathmary 1993). It is worth to point out that the study of
DEform values also supports this proposal. For the majority
L-enantiomers, the most negative DEform values are con-
centrated in Guanine (G) and Cytosine(C) regions (bold
text in Table 2). Although the code relationships are weak
in the case of the origin model of aaNs (amino acids vs.
nucleotides (1:1)), the lower DEform values for G and C
J Mol Evol (2010) 70:572–582
compounds imply that the selection of the G and C
nucleosides as the earlier precursor codons might be related
to their own chemical properties due to the number and
relative situation of the H-bond acceptors and donors in
nucleotide bases.
Conclusions
This study has investigated new aspects of the original
basis of a selection mechanism for amino acid homochi-
rality and conjectured on the results. Through the initial
formation of aaNs, followed by their condensation to
generate a aaN polymer template on a cationic mineral
surface, the D-ribose asymmetric entity leads to the chiral
amino acid distinguished. On the basis of the above model
for the origin of homochirality, meaningful results are
derived. First, it was proposed that in the earliest stage of
the origin of homochirality for amino acids, both steric
interactions between amino acid side-chains and D-nucle-
otide bases and D-ribose asymmetric entity could promote a
constant and strong selective drive in favor of L-enantio-
mers. Second, based on the theoretical model of nucleo-
tide:amino acid relationships (1:1), the binding system can
not only be significantly selective for L-amino acids over
their D-enantiomers, but can also differentiate the bases of
nucleotides, especially favoring G and C. From the pos-
tulates we present here, it is clear that there is a plausible
thermodynamic basis for specific interactions that could
have influenced the evolution of amino acid homochirality
and the Genetic Code. It is hoped that further study will
provide additional data to support these results.
Acknowledgments We thank Professor G. Michael Blackburn for
useful discussions. This study was supported by the National Science
Foundation of China (Grant No. 40976050 and Grant No. 40706043)
and the 908 Project Foundation of State Oceanic Administration of
China (FJ 908-02-03-05).
References
Adelfinskaya O, Herdewijn P (2007) Amino acid phosphoramidate
nucleotides as alternative substrates for HIV-1 reverse transcrip-
tase. Angew Chem Int Ed 46:4356–4358
Arrhenius G, Sales B, Mojzsis S, Lee T (1997) Entropy and charge in
molecular evolution-the case of phosphate. J Theor Biol 187:
503–522
Bailey JM (1998) RNA-directed amino acid homochirality. FASEB J
12:503–507
Bailey J (2000) Chirality and the origin of life. Acta Astronautica
46:627–631
Bailey J (2007) The inner solar system cataclysm, the origin of life,
and the return to the moon. In: Proceedings of 6th Australian
space science conference, Canberra, July 2006, pp 17–22
Bonner WA (2000) Parity violation and the evolution of biomolecular
homochirality. Chirality 12:114–126
581
Cheng CM, Liu XH, Li YM, Ma Y, Tan B, Wan R, Zhao YF (2004)
N-phosphoryl amino acids and biomolecular origins. Orig Life
Evol Biosph 34:455–464
Chung NM, Lohrmann R, Orgel LE, Rabinowitz J (1971) The
mechanism of the trimetaphosphate-induced peptide synthesis.
Tetrahedron 27:1205–1210
Di Giulio M (1997) On the origin of the genetic code. J Theor Biol
187:573–581
Di Giulio M (1998) Reflections on the origin of the genetic code: a
hypothesis. J Theor Biol 191:191–196
Di Giulio M (1999) Physicochemical optimization in the genetic code
origin as the number of codified amino acids increases. J Mol
Evol 49:1–10
Di Giulio M (2004) The coevolution theory of the origin of the
genetic code. Phys Life Rev 2:128–137
Di Giulio M, Medugno M (1998) The historical factor: the
biosynthetic relationships between amino acids and their phys-
icochemical properties in the origin of the genetic code. J Mol
Evol 46:615–621
Ferris JP, Hill AR, Liu RH, Orgel LE (1996) Synthesis of long
prebiotic oligomers on mineral surfaces. Nature 381:59–61
Fu H, Tu GZ, Li ZL, Zhao YF, Zhang RQ (1997) New and efficient
approach to the synthesis of pentacoordinate spirobicyclic
phosphoranes. J Chem Soc Perkin Trans 1:2021–2022
Fu H, Li ZL, Zhao YF, Tu GZ (1999) Oligomerization of N,
O-Bis(trimethylsilyl)-a-amino acids into peptides mediated by
o-phenylene phosphorochloridate. J Am Chem Soc 121:291–295
Hazen RM (2001) Life’s rocky start. Sci Am 284:77–85
Inous H, Baba Y, Furukawa T, Maeda Y, Tsuhako M (1993)
Formation of dipeptide in the reaction of amino acids with cyclo-
triphosphate. Chem Parm Bull 41:1895–1899
Jorissen A, Cerf C (2002) Asymmetric photoreactions as the origin of
biomolecular homochirality: a critical review. Orig Life Evol
Biosph 32:129–142
Kolb V, Zhang SB, Xu Y, Arrhenius G (1997) Mineral induced
phosphorylation of glycolate ion—a metaphor in chemical
evolution. Orig Life Evol Biosph 27:485–503
Li ZL, Fu H, Gong HG, Zhao YF (2004) Convenient solid-phase
synthesis of oligopeptides using pentacoordianated phosphoranes
with amino acid residue as building blocks. Bioorg Chem 32:
170–177
Lin CX, Fu H, Tu GZ, Zhao YF (2003) Synthesis and chiral
separation of dinucleotides(TpAZT) phosphoramidates. Chin
Chem Lett 14:779–782
Lu K, Tu GZ, Guo XF, Sun XB, Liu Y, Feng YP, Zhao YF (2002)
Structure and isomerization of O, O-phenylene penta-coordiant-
ed phosphoryl serine. J Mol Struct 610:65–72
McGuigan C, Pathirana RN, Balzarini J, De Clercq E (1993)
Intracellular delivery of bioactive AZT nucleotides by aryl
phosphate derivatives of AZT. J Med Chem 36:1048–1052
Ni F, Sun ST, Huang C, Zhao YF (2009) N-phosphorylation of amino
acids by trimetaphosphate in aqueous solution-learning from
prebiotic synthesis. Green Chem 11:569–573
Pasek MA, Kee TP, Bryant DE, Pavlov AA, Lunine JI (2008)
Production of potentially prebiotic condensed phosphates by
phosphorus redox chemistry. Angew Chem Int Ed 47:7918–7920
Qin ZH, Lin CX, Chen Y, Ju Y, Zhao YF (2002) Electrospray
ionization mass spectrometry of serine/alanine conjugated 50 -
UMP and 30 , 50 -dithymidine phosphoramidates. Rapid Commun
Mass Spectrom 16:1997–2002
Saghatelian A, Yokobayashi Y, Soltani K, Ghadiri MR (2001) A
chiroselective peptide replicator. Nature 409:797–801
Seligmann H, Amzallag GN (2002) Chemical interactions between
amino acid and RNA: multiplicity of the levels of specificity
explains origin of the genetic code. Naturwissenschaften 89:
542–551
123
582
Speight RE, Hart DJ, Sutherland JD, Blackburn JM (2001) A new
plasmid display technology for the in vitro selection of
functional phenotype-genotype linked proteins. Chem Biol
8:951–965
Szathmary E (1993) Coding coenzyme handles: a hypothesis for
the origin of the genetic code. Pro Natl Acad Sci 90:9916–
9920
Tan B, Lee MC, Cui M, Liu T, Chen ZZ, Li YM, Ju Y, Zhao YF,
Chen KX, Jiang HL (2004) A common intermediate for prebiotic
synthesis of proteins and nucleosides: a density functional theory
(DFT) study on the formation from N-phosphoryl amino acids. J
Mol Struct (Themchem) 672:51–60
Trifonov EN (2000) Consensus temporal order of amino acids and
evolution of the triplet code. Gene 261:139–151
Tsuhako M, Fujimoto M, Ohashi S, Nariai H, Motooka I (1984)
Phosphorylation of nucleosides with cyclo-triphosphate. Bull
Chem Soc Jpn 57:3274–3280
Wachtershauser G (1997) The origin of life and its methodological
challenge. J Theor Biol 187:483–494
123
J Mol Evol (2010) 70:572–582
Westheimer FH (1987) Why nature chose phosphates. Science
235:1173–1178
Woese CR (1965) On the origin of the genetic code. Proc Natl Acad
Sci 54:1546–1552
Wong JT (1975) A co-evolution theory of the genetic code. Proc Natl
Acad Sci 72:1909–1912
Yamagata Y, Watanabe H, Saitoh M, Namba T (1991) Volcanic
production of polyphosphates and its relevance to prebiotic
evolution. Nature 352:516–519
Yang P, Han DX (2000) Molecular modeling of the binding mode of
chiral metal complexes [Co(phen)2dppz]3 ? with B-DNA. Sci
China B 43:516–523
Yarus M (1998) Amino acids as RNA ligands: a direct-RNA-template
theory for the code’s origin. J Mol Evol 47:109–117
Zaia DAM (2004) A review of adsorption of amino acids an minerals:
as it important for origin of life? Amino Acids 27:113–118
Zhou WH, Ju Y, Zhao YF, Wang QG, Luo GA (1996) Simultaneous
formation of peptides and nucleotides from N-phosphothreonine.
Orig Life Evol Biosph 26:547–560
Copyright of Journal of Molecular Evolution is the property of Springer Science & Business Media B.V. and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.