Professional Documents
Culture Documents
Perspective
Evidence for the Likely Origin of Homochirality in Amino
Acids, Sugars, and Nucleosides on Prebiotic Earth
Ronald Breslow
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja3012897 • Publication Date (Web): 25 Mar 2012
Downloaded from http://pubs.acs.org on April 12, 2012
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
1
2
3
4
Evidence
for
the
Likely
Origin
of
Homochirality
in
Amino
Acids,
Sugars,
and
5 Nucleosides
on
Prebiotic
Earth
6
7 Ronald
Breslow
8 Department
of
Chemistry,
Columbia
University
9
10
New
York
NY
10024
11 rb33@columbia.edu
12
13 Abstract.
14 Over
the
past
century
the
origin
of
terrestrial
prebiotic
homochirality
has
been
the
15
16
subject
of
many
speculations.
For
life
to
start
on
earth
and
elsewhere,
it
is
critical
17 that
the
building
blocks
of
amino
acids,
sugars,
and
nucleosides
be
created
in
18 predominant
homochiral
form.
Recent
findings
of
a
modest
excess
L
chirality
of
α‐
19 methylamino
acids
in
some
meteorites
that
landed
on
earth
have
furnished
an
20
important
piece
of
evidence.
We
have
shown
how
these
meteoritic
components
can
21
22 furnish
normal
L
amino
acids,
and
therefrom
D
sugars
and
nucleosides,
in
high
23 chiral
excess
under
sensible
prebiotic
conditions.
Some
important
remaining
goals
24 are
also
described.
25
26
27
Introduction.
28 Currently
we
are
not
surprised
that
L
amino
acids
and
D
sugars
are
produced
in
29 biological
systems,
since
the
enzymes
that
produce
them
are
themselves
homochiral
30 (not
a
mixture
with
their
mirror
images.)
On
prebiotic
earth
no
such
chirally
31 selective
catalysts
were
there
to
make
the
first
amino
acids
or
sugars
or
nucleosides,
32
33
so
many
scientists
have
speculated
on
how
such
selectivity
could
have
arisen
in
a
34 previously
achiral
world.
Some
have
invoked
chiral
faces
of
quartz,
some
have
35 suggested
an
accident
that
was
then
promulgated,
but
the
question
has
had
no
36 important
new
impetus
until
recently.
In
1969
a
carbonaceous
chondritic
meteorite
37 landed
in
Murchison
Australia
carrying
many
organic
compounds
(see
an
extensive
38
39
recent
review
by
Pizzarello
and
Groy1
including
the
earliest
work
by
Kvenholden
et
40 al.2
and
by
Cronin
and
Pizzarello3
cited
in
the
review.)
These
compounds
were
41 apparently
able
to
survive
the
frictional
heating
as
the
meteorite
passed
through
our
42 atmosphere
since
they
were
initially
at
ca.
10K,
and
chondritic
meteorites
are
pieces
43 of
rock,
with
low
thermal
conductivity,
from
the
asteroid
belts
that
surround
the
44
45 sun.
When
the
meteorite
was
split
open
the
interior
was
still
cold
enough
to
freeze
46 water.
47
48 Among
the
compounds
identified
were
the
amino
acids
alanine,
valine,
aspartic
acid,
49 glutamic
acid,
proline,
and
leucine,
which
were
racemic,
with
equal
mixtures
of
the
L
50
51 and
D
forms,
along
with
achiral
glycine.
However,
five
amino
acids
were
found
that
52 had
methyl
groups
instead
of
hydrogens
on
their
alpha
positions
(Figure
1),
and
53 these
had
a
range
of
small
excesses
of
the
enantiomers
originally
described
as
the
L
54 amino
acids
(in
modern
terminology
they
are
the
S
enantiomers).
Since
that
time,
55
these
and
other
α‐methyl
amino
acids
with
small
excesses
of
the
S
enantiomer
have
56
57 been
found
in
the
Murchison,
Murray,
and
Orgueil
meteorites.1
58
59
60
1
ACS Paragon Plus Environment
Journal of the American Chemical Society Page 2 of 15
1
2
3
4
5 + + +
H 2N H3N CO2- H3N CO2-
6 CO2-
7 Me Pr Me Me Et
8
9
10 e.e. 2.8% S e.e. 2.8% S e.e. 15.2% S
11
12 + +
H 3N H 3N CO2-
13 CO2-
14 Me Me Me
Me (3R)
15 (3S)
Et Et
16
17 e.e. 9.1% S
18
e.e. 7% S
19
20 Figure
1.
Some
α‐methyl
amino
acids
discovered
in
the
Murchison
and
other
21 meteorites,
all
of
which
have
a
small
excess
of
the
S
configuration
that
was
22 described
as
L.
The
enantioexcesses
showed
a
range
of
values
in
this
and
later
23
24
work.
25
26 How
do
we
know
that
they
were
not
simply
contaminants
from
earth,
added
after
27 the
meteorites
landed?
There
are
three
arguments
against
this.
First
of
all,
the
28 actual
α‐methyl
amino
acids
isolated
from
the
Murchison
meteorite
are
not
found
in
29
30 our
biology
today,
but
the
meteorite
landed
here
only
41
years
ago.
Against
this,
31 there
are
some
α‐methyl
amino
acids
produced
by
microorganisms4‐6
but
not
all
the
32 ones
in
the
meteorites.
Secondly,
enzymes
do
not
produce
compounds
with
only
33 partial
chiral
selectivities;
such
partial
chiralities
are
really
symptomatic
of
species
34
35
that
were
originally
racemic
but
have
then
been
partially
deracemized.
If
36 microorganisms
had
invaded
the
meteorites
after
they
landed,
they
could
perhaps
37 selectively
cause
destruction
of
the
D
enantiomers
of
the
amino
acids
if
they
had
38 originally
been
racemic,
but
there
is
no
evidence
for
such
an
unlikely
process.
I
will
39 describe
an
alternative
scenario
for
such
partial
deracemization
below.
Finally,
the
40
41
α‐methyl
amino
acids
have
levels
of
13C
and
non‐exchangeable
deuterium
much
42 higher
than
are
seen
in
molecules
formed
on
Earth.
These
high
levels
of
heavy
43 isotopes
are
generally
seen
in
atoms
delivered
from
space,
where
isotopic
44 fractionation
is
performed
at
very
low
temperatures.
They
are
generally
accepted
to
45
be
definitive
proof
that
the
species
examined
are
not
of
terrestrial
origin.
46
47
48 Microwave
spectroscopy
has
detected
hundreds
of
molecules
in
the
interstellar
gas
49 clouds
in
space,
and
they
include
ammonia,
hydrogen
cyanide,
hydronium
ion,
and
50 various
ketones
and
aldehydes
and
the
acetylenes
from
which
they
can
be
derived.7
51
Thus
the
meteoritic
unmethylated
and
methylated
amino
acids
were
probably
52
53 formed
by
Strecker
reactions8
from
such
species.
The
reactants
are
aldehydes
for
54 normal
unmethylated
amino
acids
and
methyl
ketones
for
the
α‐methyl
amino
acids,
55 along
with
HCN,
NH3,
and
H3O+.
Methyl
ketones
are
the
products
from
reaction
of
56 terminal
acetylenes
with
hydronium
ion.
The
multimolecular
reactions
would
occur
57
58 on
solid
particles
whose
larger
versions
are
asteroids,
and
they
would
be
initially
59
60
2
ACS Paragon Plus Environment
Page 3 of 15 Journal of the American Chemical Society
1
2
3
4
formed
as
racemates.
They
could
develop
some
excess
of
the
L
forms
if
the
5 racemates
absorbed
circularly
polarized
light
that
selectively
destroyed
one
6 enantiomer,
perhaps
also
using
another
process
to
amplify
them.9
7
8 William
Bonner
had
shown
that
right
circularly
polarized
light
at
uv
wavelengths
9
10
would
selectively
destroy
the
D
component
of
racemic
leucine,
leaving
a
small
11 excess
of
the
L
amino
acid.10
Astronomers
have
detected
a
small
excess
of
circularly
12 polarized
infrared
light
in
space,
whose
energy
is
too
low
to
deracemize
amino
13 acids.11‐13
However,
one
author
indicated
that
the
same
processes
that
produce
the
14 identified
circularly
polarized
light
in
the
infrared
might
also
produce
it
in
the
15
16
ultraviolet.12
Very
high
energy
light
cannot
penetrate
our
atmosphere
for
17 observation
as
infrared
can,
and
could
not
penetrate
it
when
our
early
atmosphere
18 consisted
of
nitrogen
and
CO2.
As
one
possibility,
favored
by
many
astronomers,
19 circularly
polarized
light
could
be
formed
in
the
universe
by
cyclotron
processes
in
20 neutron
stars,
just
as
experimental
cyclotrons
produce
circularly
polarized
light
21
22 with
opposite
chirality
above
and
below
the
circulation
plane.
Some
other
23 astronomers
prefer
that
circularly
polarized
light
could
be
produced
by
magnetic
24 dwarf
stars.
One
of
the
important
challenges
for
astronomy
is
to
observe
outside
our
25 atmosphere
and
detect
circularly
polarized
short‐wavelength
light,
if
it
is
there.
For
26 instance,
observations
could
be
made
from
orbiting
satellites,
from
the
Hubble
27
28 telescope,
or
from
the
new
Web
Space
Telescope.
I
am
trying
to
get
such
29 observations
made.
30
31 If
there
was
also
(yet
undetected)
right
circularly
polarized
light
with
energy
in
the
32
uv
or
higher
irradiating
the
asteroid
belt
when
the
amino
acids
were
present
on
a
33
34 particle
that
later
came
to
Earth,
this
could
account
for
the
small
excesses
of
the
L
35 enantiomers
seen
in
the
α‐methyl
amino
acids.
If
this
also
happened
with
the
36 unmethylated
amino
acids
that
are
found
in
our
proteins
they
could
racemize
by
37 reversible
loss
of
their
alpha
protons
over
time,
explaining
why
they
are
found
in
38
39
racemic
form,
but
no
such
racemization
is
possible
with
the
α‐methyl
amino
acids.
40 (Isoleucine
has
two
chiral
centers,
and
it
was
not
completely
racemic.
The
α‐center
41 would
be
equilibrated
by
reversible
loss
of
its
somewhat
acidic
proton
but
the
β‐
42 center
could
not
be
equilibrated,
so
isoleucine
could
not
be
racemized
by
half
43
44 conversion
to
its
mirror
image.)
45
46 Thus
an
attractive
idea
is
that
the
L
amino
acids
and
the
D
sugars
that
are
now
the
47 basis
of
life
were
first
“seeded”
by
the
arrival
on
Earth
of
α‐methyl
amino
acids
in
48
49
meteorites,
formed
on
asteroid
fragments
as
racemates
and
then
partially
50 deracemized
by
circularly
polarized
light
of
short
wavelength.14
Light
of
the
needed
51 wavelength
could
not
penetrate
Earth’s
atmosphere—now
or
in
the
past—so
the
52 deracemization
occurred
in
space
and
the
result
was
then
delivered
to
Earth.
Such
53 homochirality
was
probably
important
since
mixture
of
enantiomers
would
not
54
55
form
well‐defined
structures
on
random
incorporation
into
polypeptides
or
56 polysaccharides
or
polynucleotides.
In
fact,
it
seems
likely
that
homochirality
is
57 generally
necessary
for
the
origin
of
life
anywhere,
so
without
the
kind
of
processes
58
59
60
3
ACS Paragon Plus Environment
Journal of the American Chemical Society Page 4 of 15
1
2
3
4
that
started
it
on
earth
other
planets
would
not
produce
life.
On
Earth
the
amino
5 acid
chirality
was
what
we
call
L,
but
on
other
planets
the
mirror
image
systems
6 could
occur
by
the
processes
we
describe.
7
8 We
have
recently
shown
how
such
α‐methyl
amino
acids
can
generate
normal
9
10 unmethylated
biological
amino
acids
with
some
chirality
transfer,
and
how
that
11 excess
chirality
can
be
amplified
under
credible
prebiotic
conditions,
producing
12 aqueous
solutions
with
very
high
enantioexcesses.
We
have
also
shown
how
the
D
13 sugars
and
D
ribonucleosides
could
have
arisen
on
prebiotic
Earth.
14
15
16
Results
and
Discussion
17
18 We
addressed
two
questions.
1)
What
credible
prebiotic
chemistry
could
use
the
19 small
excess
S
(L)
chirality
of
the
α‐methyl
amino
acids
in
a
meteorite
to
seed
the
20
21
formation
of
L
amino
acids
without
such
methyl
groups?
2)
How
could
credible
22 prebiotic
processes
amplify
small
enantioexcesses
to
available
high
enantiopurity
in
23 solution?
The
ideal
is
of
course
100%,
but
enantioexcesses
of
90%
or
better
could
24 probably
be
enough
for
primitive
biology
to
select
the
dominant
enantiomer.
25
26
27
Formation
of
nonmethylated
amino
acids
with
some
L
excess.
28 We
devised
a
process
of
decarboxylative
transamination:
α‐methyl
amino
acids
29 reacted
with
α‐keto
acids
to
form
unmethylated
amino
acids
on
simple
heating
in
30 the
presence
of
Cu(II)
(Figure
2).15
This
was
based
on
earlier
work
we
had
done
31
32
showing
that
α‐methyl
amino
acids
could
perform
such
decarboxylative
33 transaminations
with
pyridoxal.16
34
35 We
observed
some
chiral
transfer,
in
which
the
reaction
of
100%
e.e.
S‐α‐
36
37
methylvaline
with
sodium
phenylpyruvate
led
to
a
37%
enantioexcess
of
L‐
38 phenylalanine,
and
the
same
process
with
sodium
pyruvate
led
to
a
20%
39 enantioexcess
of
L‐alanine.
The
reaction
of
sodium
dimethylpyruvate
with
S‐α‐
40 methylisoleucine
afforded
L‐valine
with
17%
transfer
of
chirality.
All
these
41
processes
were
successful
only
in
the
presence
of
catalytic
Cu(II),
a
component
of
42
43 some
meteorites
and
surely
present
on
prebiotic
earth.
Without
it,
or
with
Zn(II)
44 instead,
the
reactions
were
much
less
successful,
and
the
correct
chiral
transfer
was
45 not
observed.
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
ACS Paragon Plus Environment
Page 5 of 15 Journal of the American Chemical Society
1
2
3 R
4 R CO2 -
+ - H2O
5 H 3N N - CO2
6 O- CO2- CO2-
O +
7 Me Me Cu(II)
O + H2O
8
9
10 R R H
CO2 - + H+ CO2 -
11 H2O R CO2 -
N N +
12
H NH3 +
13 Me Me O Me
14
15
16
17
18 Figure
2.
Transaminative
decarboxylation
of
the
imine
from
S‐α‐methylvaline
and
19 an
α
ketoacid.
In
the
presence
of
Cu(II),
which
catalyzes
the
reaction,
the
product
20 amino
acid
has
an
excess
of
the
L
enantiomer,
but
not
without
the
Cu(II).
This
is
21
22
expected
if
a
square
planar
complex
is
formed
between
two
imines
and
the
Cu(II),
so
23 after
decarboxylation
of
one
ligand
it
is
protonated
selectively,
guided
by
the
24 stereochemistry
of
the
other
ligand.
25
26 The
process
was
credible
under
prebiotic
conditions,
involving
only
heating
the
27
28
components
in
the
presence
of
some
catalytic
Cu(II)
ion.
As
we
described,15
DFT
29 calculations
indicated
that
a
square
planar
Cu(II)
complex
of
two
imines
that
were
30 formed
from
the
amino
acid
and
the
keto
acid
would
decarboxylate
one
of
them,
31 which
would
then
protonate
to
form
the
L
amino
acid
steered
by
the
chirality
of
the
32 second
ligand,
as
we
observed.
However,
the
few
percent
of
S
excess
in
the
33
34 meteoritic
amino
acids
would
produce
only
a
1%
or
so
enantioexcess
in
the
product
35 amino
acids.
Amplification
of
the
resulting
chiral
excess
would
be
critical
to
achieve
36 a
dominance
of
the
L
amino
acids
large
enough—of
the
order
of
a
90/10
ratio
or
37 better
of
L
to
D—that
new
organisms
or
pre‐organisms
would
select
them
for
38 evolutionary
success.
39
40
Amplification
of
the
Lamino
acid
excesses
in
solution.
41 In
1969
Morowitz
had
proposed
a
process
by
which
a
small
enantioexcess
of
an
42 amino
acid
could
be
amplified
to
form
solutions
with
large
enantioexcesses
under
43 simple
prebiotic
conditions.17
This
was
an
amplification
of
the
concentration
of
the
44 excess
component
in
solution
by
concentrating
it
and
removing
the
racemate,
not
by
45
46 making
more
of
it.
We
conceived
the
same
idea
in
2006,18
not
aware
of
his
work,
47 and
Blackmond
also
published
a
version
of
the
concept
in
2006
slightly
before
our
48 publication.19‐21
All
our
ideas
were
based
on
the
general
fact
that
most
amino
acids
49 form
racemic
compound
crystals
that
are
less
soluble,
and
higher
melting,
than
the
50
crystals
of
the
L
or
the
D
enantiomers.
In
such
racemic
crystals
neighboring
L
and
D
51
52 molecules
interact
to
lower
the
free
energy—either
by
better
crystal
packing
or
by
53 direct
stabilizing
interactions
in
the
crystals—compared
with
crystals
where
an
L
54 molecule
has
only
other
L
neighbors.
55
56
The
idea
is
that
a
mixture
of
D
and
L
amino
acids
with
some
excess
of
the
L
57
58 component
(or
elsewhere
in
the
universe
perhaps
the
D
component)
would
dissolve
59
60
5
ACS Paragon Plus Environment
Journal of the American Chemical Society Page 6 of 15
1
2
3
4
in
water,
and
as
the
water
evaporates
the
less
soluble
racemate
would
precipitate
5 leaving
a
solution
with
increased
richness
in
the
L
component.
Because
the
6 precipitation
of
the
racemate
depends
on
the
solubility
product
SP(DL)
=
[D][L],
7 while
that
of
the
homochiral
L
compound
depends
only
on
the
solubility
[L],
there
8 would
be
feedback
to
increase
the
precipitation
of
the
racemate
as
the
9
10
concentration
of
L
increases.
Thus
starting
with
a
small
excess
of
the
L
amino
acid,
11 the
final
L/D
ratio
would
become
very
large
with
even
a
modest
difference
in
12 solubilities.
The
Morowitz
treatment17
is
shown
below.
13
14 S(L)
=
[L]
15
16
SP(DL)
=
[D][L]
17 [D]
=
SP(DL)/[L]
18 [L]/[D]
=
S(L)2/SP(DL)
19
20 The
solubility
product
SP(DL)
is
equal
to
the
square
of
half
the
solubility
of
the
21
22 racemate.
By
this
equation,
if
the
homochiral
crystals
were
only
twice
as
soluble
as
23 were
the
racemates
the
final
ratio
[L]/[D]
would
be
16/1,
a
solution
with
94%
of
the
24 L
and
6%
of
the
D.
If
the
homochiral
crystals
were
three‐fold
as
soluble
as
the
25 racemate
the
final
L/D
ratio
would
be
36/1.
As
we
will
describe
later,
this
treatment
26 is
not
quite
correct
and
can
overestimate
the
selectivity
for
reasons
we
have
27
28 demonstrated.
29
30 The
previous
theories
and
experiments
involved
solutions
at
equilibria,
but
kinetics
31 can
often
afford
different
results.
With
equilibrium
solubilities
we
saw
L
tryptophan
32
amplified
to
94.5%
L
and
5.5%
D,
starting
with
any
arbitrarily
small
excess
(the
33
34 treatment
is
true
for
any
initial
ratio,
limited
only
by
the
practical
need
to
achieve
35 saturation
in
both
the
racemate
and
homochiral
compound,
so
smaller
amounts
of
36 water
would
be
needed
with
smaller
initial
L
excess).15
However,
we
found
that
this
37 ratio
was
raised
to
99/1
in
solution
when
we
poured
a
small
amount
of
water
38
through
the
final
dry
mixture,
imitating
the
effect
of
rainwater.15
Homochiral
and
39
40 racemic
crystals
differ
in
their
activation
energies
for
subtracting
(or
adding)
units,
41 a
form
of
dissociation
(or
binding),
and
in
our
case
the
homochiral
crystals
dissolve
42 faster
beyond
the
requirements
of
the
equilibrium
constants.
43
44
45
Forming
D
sugars
under
prebiotic
conditions
46 We
have
investigated
the
formation
of
the
simplest
sugar
belonging
to
the
D
series,
47 D‐glyceraldehyde.
All
other
sugars
in
the
D
family—including
D‐ribose,
D‐glucose,
48 D‐fructose,
D‐erythrose,
etc.—are
named
for
the
geometry
of
the
chiral
center
49 furthest
from
the
carbonyl
group
in
the
sugar,
and
this
is
the
unique
chiral
center
in
50
51
D‐glyceraldehyde.
Sugars
were
probably
synthesized
on
prebiotic
Earth
by
a
52 process
called
the
formose
reaction.22
D‐ribose
(C5H1005)
is
a
formal
pentamer
of
53 formaldehyde
(CH2O),
whose
dimer
is
glycolaldehyde
and
whose
trimer
is
54 glyceraldehyde.
In
the
formose
reaction
formaldehyde
is
treated
with
a
mild
base
55 such
as
calcium
hydroxide.
There
is
a
period
when
nothing
happens
until
there
is
56
57
the
sudden
conversion
of
the
formaldehyde
to
glycolaldehyde
and
glyceraldehyde
58
59
60
6
ACS Paragon Plus Environment
Page 7 of 15 Journal of the American Chemical Society
1
2
3
4
and
even
larger
sugars.
We
proposed
and
demonstrated
the
mechanism
of
the
5 formose
reaction
(Figure
3)
by
an
autocatalytic
cycle
many
years
ago.23
6
7 slow HO
8 2 H 2C O H2C CH=O
9 H2C O
10
11
12 HO H
13 2 H2C CH=O HO CH=O
14
15 CH2OH
16
17
18 HOH
19
HOHC CH=O
20
CH2OH O
21
22 HOH2C CH2OH
23
24
25 O
H2C O
26 HOHC CH2OH
27
CH2OH
28
29
30 Figure
3.
The
formose
reaction,
producing
sugars
from
formaldehyde.
31
32
33
A
first
molecule
of
glycolaldehyde
is
formed
from
formaldehyde
by
a
slow
unknown
34 process,
possibly
involving
ionization
by
cosmic
rays,
and
this
then
reacts
with
an
35 additional
formaldehyde
to
form
glyceraldehyde.
This
is
then
in
equilibrium
with
36 dihydroxyacetone
by
enolization
and
ketonization,
and
this
then
reacts
with
another
37 formaldehyde
to
form
a
four‐carbon
ketosugar.
By
enolization
and
aldehyde
38
39
formation
a
four‐carbon
aldehyde
is
formed
that
can
undergo
a
reverse
aldol
40 reaction
to
form
two
molecules
of
glycolaldehyde.
After
this,
four
glycolaldehydes
41 result
from
another
turn
of
the
cycle,
etc.
42
43 We
examined
the
formation
of
glyceraldehyde,
a
three‐carbon
sugar
that
is
the
44
45
simplest
one
with
a
chiral
center.24
Glyceraldehyde
is
formed
in
the
formose
cycle
46 by
the
reaction
of
formaldehyde
with
glycolaldehyde.
Without
a
chiral
catalyst
this
47 reaction
would
form
both
D‐glyceraldehyde
and
its
enantiomer
L‐glyceraldehyde
in
48 equal
amounts,
but
we
studied
what
would
happen
if
the
reaction
were
catalyzed
by
49 a
chiral
amino
acid.
We
can
trace
the
formation
of
L
amino
acids
back
to
the
small
50
51 excesses
of
S
α‐methyl
amino
acids
in
the
Murchison
meteorite,
so
the
question
was
52 simple.
Would
L
amino
acids
catalyze
the
formation
of
D‐glyceraldehyde
53 preferentially,
at
least
to
a
small
extent,
and
if
so
could
the
preference
be
amplified
54 to
a
sufficient
excess
that
the
dominant
D‐glyceraldehyde
would
be
selected
by
55
56
incipient
life?
If
so,
all
of
the
D
sugars—built
on
the
D‐glyceraldehyde
by
adding
57 pieces
to
its
aldehyde
group
that
do
not
change
the
geometry
of
the
D
center
in
the
58
59
60
7
ACS Paragon Plus Environment
Journal of the American Chemical Society Page 8 of 15
1
2
3
4
glyceraldehyde
original
unit—are
part
of
the
general
origin
of
homochirality
tracing
5 back
to
the
meteorites.
If
not,
a
new
source
of
homochirality
will
need
to
be
found
6 for
the
sugars.
As
Table
1
shows,
we
examined
various
L‐amino
acids
from
different
7 classes.
8
9
10
11 Table
1.
Ratio
of
D
to
L
glyceraldehyde
12
from
glycolaldehyde
and
formaldehyde
13 catalyzed
by
various
L‐amino
acids.
14 The
reaction
conditions
and
analytical
15
16
method
are
described
in
ref.
24.
17 _________________________________
18 Serine
(50.3/49.7)
Alanine
(50.8/49.2)
19 Phenylalanine
(52.2/47.8)
Valine
(52.2/47.8)
20 Glutamic
acid
(60.7/39.3)
Proline
(28.9/71.1)
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8
ACS Paragon Plus Environment
Page 9 of 15 Journal of the American Chemical Society
1
2
3
4
5
6 The
reaction
of
formaldehyde
with
glycolaldehyde
is
an
aldol
reaction,
and
many
7 such
reactions
are
known
to
be
catalyzed
by
amines.
They
would
react
with
the
8 glycolaldehyde
to
form
an
enamine,
and
this
adds
to
the
other
carbonyl
component.
9
10
Thus
we
carried
out
the
reaction
of
glycolaldehyde
with
formaldehyde
in
water
in
11 the
presence
of
a
variety
of
L
amino
acids.
We
found
(Table
1)
that
all
the
L
amino
12 acids
caused
the
formation
of
glycolaldehyde
with
a
small
excess
of
the
D
13 enantiomer,
with
one
exception:
L‐proline
catalyzed
the
formation
of
an
excess
of
L‐
14 glyceraldehyde.
(Some
recent
work
by
Blackmond25
indicates
that
in
basic
solution
15
16
this
proline
preference
is
reversed
when
the
carboxyl
group
is
a
carboxylate
ion,
so
17 proline
may
not
be
an
exception
after
all.)
18
19 The
preferences
were
modest.
There
was
a
60/40
ratio
of
D/L
glyceraldehyde
20 catalyzed
by
L‐glutamic
acid,
for
instance,
and
smaller
ratios
with
the
other
L‐amino
21
22 acids.
Proline
was
probably
not
present
in
large
amounts
among
the
early
amino
23 acids,
since
it
is
formed
in
two
secondary
reactions
from
glutamic
acid,
so
it
would
24 not
dominate
the
chiral
selectivity.
The
results
with
the
other
amino
acids
did
25 support
the
idea
that
the
preference
for
D‐glyceraldehyde—and
then
the
other
D
26 sugars
derived
from
it—was
simply
the
result
of
the
formation
of
the
other
L
amino
27
28 acids
on
prebiotic
Earth.
29
30 We
then
asked
whether
there
was
a
likely
process
for
amplifying
the
small
excess
of
31 D‐glyceraldehyde
formed
in
this
way,
preferably
by
using
the
selective
solubilities
32
that
had
worked
with
the
amino
acids.
This
turned
out
to
be
possible..
D‐
33
34 Glyceraldehyde
is
a
syrup
with
essentially
complete
water
solubility,
but
racemic
35 DL‐glyceraldehde
is
a
solid
with
a
melting
point
of
145°C
and
a
limited
water
36 solubility.
The
striking
difference
has
an
explanation.
A
previously
published
X‐ray
37 structure
determination
showed
that
DL‐glyceraldehyde
exists
as
a
chair‐form
six‐
38
membered
ring
dioxane
dimer
of
one
D
and
one
L
molecule,
with
all
the
substituents
39
40 equatorial
as
in
chair
cyclohexane,
so
this
flat
dimer
molecule
packs
well
into
a
41 crystal.26
With
the
same
structure
based
on
a
dimer
with
D‐glyceraldehyde
alone,
42 one
large
group
in
the
six‐membered
DD
ring
would
be
axial,
making
the
dimer
less
43 stable
and
also
making
crystal
packing
less
favorable.
The
result
was
that
we
could
44
45
take
the
60/40
ratio
of
D/L
glyceraldehyde
formed
by
catalysis
with
glutamic
acid
46 and
turn
it
into
a
92/8
D/L
ratio
in
water
solution
by
slow
evaporation
of
water,
47 causing
the
racemic
crystals
to
precipitate.
48
49 The
formation
and
amplification
of
Dribonucleosides.
50
51
An
aldol
condensation
of
D‐glyceraldehyde
wth
glycolaldehyde
would
lead
to
D‐
52 ribose.
We
are
studying
the
selective
catalysis
of
this
reaction,
in
which
two
new
53 chiral
centers
are
produced.
In
the
mean
time,
we
examined
the
possibility
that
any
54 excess
of
D
over
L
in
ribose
could
be
amplified
by
selective
solubilities,
as
in
the
55 amino
acid
and
glyceraldehyde
cases.
In
contrast
to
the
situation
with
amino
acids,
56
57
we
saw
that
D
ribose
and
DL
ribose
had
the
same
melting
point.27
Ribose
58 apparently
forms
a
racemate
that
is
a
solid
solution,
with
essentially
identical
59
60
9
ACS Paragon Plus Environment
Journal of the American Chemical Society Page 10 of 15
1
2
3
4
properties,
solubilities,
and
melting
points
at
all
D/L
ratios.
In
the
solids
a
D‐ribose
5 molecule
can
equally
well
have
either
a
D
or
an
L
as
its
neighbor.
We
did
not
6 examine
their
solubility,
since
in
all
cases
that
we
know
of
a
lower
solubility
7 correlated
with
a
higher
melting
point—both
melting
and
dissolving
break
up
the
8 crystals.
For
this
reason,
D‐ribose
cannot
be
amplified
by
the
selective
solubility
9
10
method.
Thus
we
examined
the
ribonucleosides
(Figure
5).27
11
12 It
had
been
shown
previously
that
ribose
reacts
with
purines
under
prebiotic
13 conditions
to
form
ribonucleosides.28
In
the
first
work
pyrophosphate
was
present
14 to
activate
the
ribose,
but
in
later
work
it
was
shown
that
the
pyrophosphate
could
15
16
be
replaced
by
the
residue
from
dried
seawater.
We
are
working
on
extending
these
17 results
to
pyrimidine
nucleosides
(see
another
proposal29
on
how
pyrimidine
18 nucleosides
could
have
been
formed),
and
in
the
meantime
we
examined
whether
19 ribonucleosides
with
small
excesses
of
the
D
form
could
be
amplified
by
the
20 solubility
method.
21
22
23 NH2 O
24 N NH
N
25
26 N N N O
27 HO HO
28 O O
29 H H H H
30 H H H H
OH OH OH OH
31
32
Adenosine Uracil
33
34
35
36 NH2 O
37
N N NH
38
39
N O N N NH2
40 HO
HO
41 O
O
42 H H H H
43 H H H H
OH OH OH OH
44
45
46 Cytidine Guanosine
47
48
49
50
51 Figure
4.
The
nucleosides
that
were
examined.
52
53 We
synthesized
L‐uridine
by
a
literature
method
and
made
a
1:1
mixture
with
D‐
54
55
uridine.
We
then
crystallized
the
racemate
to
purity
and
examined
its
properties.
56 The
crystals
of
uridine
racemate
had
a
melting
point
of
176°C,
higher
by
21°C
than
57 the
155°C
for
D‐uridine.
D‐Uridine
had
a
water
solubility
at
22°C
of
454
mg/mL,
58
59
60
10
ACS Paragon Plus Environment
Page 11 of 15 Journal of the American Chemical Society
1
2
3
4
while
the
racemic
uridine
had
a
solubility
of
only
87
mg/mL,
5.2
times
smaller.
By
5 the
Morowitz
calculation
this
should
afford
a
D/L
ratio
of
108/1
at
saturation
with
6 both
crystals.
We
then
dissolved
both
the
D‐uridine
and
the
DL‐uridine
crystals
to
7 saturation
together
in
water
at
22°C,
filtering
away
the
excess,
and
saw
96%
D
and
8 4%
L
in
solution,
a
ratio
of
24/1.
The
large
ratio
makes
the
D‐uridine
dominant,
but
9
10
it
is
less
than
predicted
by
the
Morowitz
theoretical
treatment.
11
12 We
then
examined
the
situation
with
adenosine,
synthesizing
L‐adenosine
and
13 mixing
it
with
D‐adenosine,
then
recrystallizing
the
racemic
crystals
to
purity.
In
14 this
case
the
melting
point
of
the
racemate
was
243
±
1°C,
while
for
D‐adenosine
the
15
16
m.p.
was
230
±
1°C.
The
solubility
of
D‐adenosine
in
water
at
22°C
was
5.2
mg/mL,
17 6.3
times
as
large
as
the
0.8
mg/mL
for
the
DL
crystal.
There
should
have
been
a
18 160/1
D/L
ratio
at
saturation
equilibrium
from
the
Morowitz
treatment,
but
we
19 observed
a
99/1
ratio
of
D/L
adenosine.
Again
very
large,
but
again
less
than
20 predicted
by
the
Morowitz
treatment.
With
cytidine
we
saw
decomposition
on
21
22 melting
of
both
the
D
and
the
DL
crystals,
so
no
melting
points
could
be
determined,
23 but
the
solubilities
predicted
an
even
larger
amplification
than
for
the
other
two
24 nucleosides.
D‐cytidine
had
a
solubility
of
192
mg/mL
in
22°C
water,
while
DL‐
25 cytidine
had
a
solubility
of
24.3
mg/mL.
The
Morowitz
equation
predicts
D/L
of
26 250/1
at
saturation,
while
we
observed
199/1.
27
28
29 All
three
of
these
nucleosides
can
be
amplified
to
very
high
D/L
ratios
by
selective
30 solubilities,
but
the
Morowitz
treatment
overestimated
the
ratios.
We
concluded
31 that
this
could
reflect
the
fact
that
the
solubilities
being
measured
were
in
pure
32
water,
but
the
solubilities
relevant
to
the
amplification
experiments
are
in
water
33
34 along
with
a
second
component.
In
the
experiment,
the
solubility
of
the
racemate
35 that
is
relevant
is
that
in
the
presence
of
the
homochiral
component
that
is
also
in
36 solution.
We
concluded
that
the
homochiral
dissolved
material
was
acting
as
a
37 cosolvent,
increasing
the
solubility
of
the
racemate
over
that
in
pure
water.14
The
38
dissolved
homochiral
component
has
two
roles.
In
one
it
helps
drive
the
racemate
39
40 out
of
solution
by
its
role
in
the
solubility
product
of
the
racemate,
but
in
the
other
41 role
it
increases
the
solubility
of
the
racemate
by
acting
as
an
antihydrophobic
42 cosolvent
like
ethanol,
with
which
relatively
hydrophobic
substrates
such
as
43 nucleosides
have
greater
solubility
than
in
pure
water.
If
this
is
true
the
solubility
of
44
45
the
homochiral
crystals
would
also
be
increased
by
the
presence
of
the
racemate
in
46 solution,
but
at
final
saturation—with
little
dissolved
racemate—this
effect
would
47 be
smaller.
48
49 To
test
this
idea,
we
examined
the
solubility
of
racemic
uridine
in
water
with
D‐
50
51
cytidine
added
to
saturation.14
This
can
mimic
the
cosolvent
effect
of
D‐uridine,
but
52 does
not
play
a
role
in
the
solubility
product
of
DL‐uridine.
We
found
that
the
D‐
53 cytidine
increased
the
solubility
of
DL‐uridine
by
40%
over
that
in
pure
water.
Thus
54 the
Morowitz
treatment
overestimates
the
amplifications
a
bit,
because
it
does
not
55 include
the
cosolvent
effect.
The
experimental
values
are
still
so
high
that
they
56
57
could
lead
to
the
dominance
of
the
D‐nucleosides
on
prebiotic
Earth,
even
starting
58 from
a
very
small
initial
excess.
59
60
11
ACS Paragon Plus Environment
Journal of the American Chemical Society Page 12 of 15
1
2
3
4
5 Guanosine
behaved
differently.27
D‐guanosine
and
the
DL‐guanosine
crystals
6 melted
with
decomposition,
but
we
could
still
examine
their
solubilities
in
22°C
7 water.
DL‐Guanosine
had
a
solubility
of
0.84
mg/mL
while
D‐guanosine
had
a
8 solubility
of
0.46
mg/mL,
essentially
half
that
of
the
racemate.
This
indicated
that
9
10
the
racemate
belongs
to
the
third
class
of
racemic
crystals.
It
is
not
a
solid
solution
11 as
we
saw
with
ribose,
nor
a
racemic
compound
crystal
like
those
of
uridine,
12 adenosine
and
cytidine.
It
is
a
racemic
conglomerate
of
D
and
L
crystals,
each
with
13 its
own
solubility.
Such
a
conglomerate
is
like
that
with
racemic
sodium
ammonium
14 tartrate,
that
allowed
Pasteur
to
separate
the
D
and
L
crystals
by
hand.
15
16
17 This
inversion
in
solubilities,
with
the
racemate
more
soluble
than
the
homochiral
18 compound,
makes
it
impossible
to
amplify
D‐guanosine
by
selective
solubilities,
but
19 of
course
the
D‐ribose
from
hydrolysis
of
the
other
three
nucleosides
could
have
20 been
used
to
synthesize
D‐guanosine
on
prebiotic
Earth.
The
D‐ribose
resulting
21
22 from
hydrolysis
of
the
other
three
nucleosides
could
also
play
additional
important
23 biochemical
roles,
including
conversion
to
D‐ribulose
whose
diphosphate
is
the
key
24 sugar
in
photosynthesis.
25
26
27
Conclusion.
This
work30,
31
answers
some
of
the
questions
in
the
general
idea
that
28 the
unusual
amino
acids
delivered
to
Earth
by
the
Murchison
meteorite
and
related
29 ones
could
have
led
to
the
dominance
of
L
amino
acids
and
D
sugars
on
early
Earth
30 that
would
permit
life
to
start.
Of
course
showing
that
it
could
have
happened
this
31
32 way
is
not
the
same
as
showing
that
it
did.
Proper
theories
need
the
possibility
of
33 falsification.
This
account
would
be
in
trouble
if
significant
examples
were
found
in
34 which
meteorites
that
landed
on
Earth
contained
R
α‐methyl
amino
acids
instead
of
35 the
S
examples
in
the
Murchison
meteorite
and
others
examined
so
far.
An
36
alternative
scenario
to
ours
would
use
the
Murchison
α‐methyl
amino
acids
to
37
38 catalyze
the
formation
of
some
D
sugars,
as
Pizzarello
and
Weber
have
shown32
(cf.
39 refs.
(33)
and
(34)
for
related
work),
and
then
use
those
sugars
to
catalyze
the
40 formation
of
the
normal
proteinogenic
L
amino
acids.
Good
credibly‐prebiotic
41 examples
of
the
latter
process
have
not
yet
been
produced.
42
43
44 Further
work
is
needed
to
show
prebiotic
versions
of
the
conversion
of
D‐
45 glyceraldehyde
to
D‐ribose,
D‐glucose,
D‐fructose,
and
D‐2‐deoxyribose,
efforts
46 underway
in
our
lab.
We
also
need
a
credible
way
in
which
nucleosides
and
47 nucleotides
could
be
formed
from
these
sugars
and
pyrimidines.,
not
just
purines
48 Finally,
of
course,
we
all
need
ways
in
which
these
and
other
sensible
building
49
50 blocks
could
assemble
into
structures
with
the
exciting
properties
of
life.
51
52
An
implication
from
this
work
is
that
elsewhere
in
the
universe
there
could
be
life
53 forms
based
on
D
amino
acids
and
L
sugars,
depending
on
the
chirality
of
circular
54 polarized
light
in
that
sector
of
the
universe
or
whatever
other
process
operated
to
55
56 favor
the
L
α‐methyl
amino
acids
in
the
meteorites
that
have
landed
on
Earth.
Such
57 life
forms
could
well
be
advanced
versions
of
dinosaurs,
if
mammals
did
not
have
58
59
60
12
ACS Paragon Plus Environment
Page 13 of 15 Journal of the American Chemical Society
1
2
3
4
the
good
fortune
to
have
the
dinosaurs
wiped
out
by
an
asteroidal
collision,
as
on
5 Earth.
We
would
be
better
off
not
meeting
them.
6
7
Acknowledgements
8
9
10 I
thank
my
coworkers
whose
names
appear
in
the
references,
and
the
U.
S.
National
11 Science
Foundation
for
support
of
this
work.
12
13
14
15
16
17 References
18 1.
Pizzarello,
S.;
Groy,
T.
Geochim.
Cosmochim.
2011,
75,
645‐656.
19 2.
Kvenvolden,
K.;
Lawless,
J.;
Pering,
K.;
Peterson,
E.;
Flores,
J.;
Ponnamperuma,
C;
20 Kaplan,
I.;
Moore,
C.
Nature
1970,
228,
923‐926.
21
22
3.
Cronin,
J.;
Pizzarello,
S.
Science
1997,
275,
951‐955.
23 4.
Degenkolb,
T.;
Gams,
W.;
Bruckner,
H.
Chem.
&
Biodiversity
2008,
5,
693‐706.
24 5.
Degenkolb,
T.;
Bruckner,
H.
Chem.
&
Biodiversity
2008,
5,
1817‐43.
25 6.
Bruckner,
H.;
Becker,
D.;
Gams,
W.;
Degenkolb,
T.
Chem.
&
Biodiversity
2009,
6,
26 38‐56.
27
28
7.
Herbst,
E.;
van
Dishoek,
E.
Ann
Rev.
Phys.
Astron.
&
Chem.
2009,
47,
427‐480.
29 8.
Strecker,
A.
Annal.
Chem.
Pharm.
1850,
75,
27‐45.
30 9.
Glavin,
D.;
Dworkin,
J.
Proc.
Natl.
Acad.
Sci.
USA
2009,
106,
5487–5492.
31 10.
Flores,
J.;
Bonner,
W.;
Massey,
G.
J.
Am.
Chem.
Soc.
1977,
99,
3622‐3625.
32 11.
Bailey,
J.;
Chrysostomou,
A.;
Hough,
J.;
Gledhill,
T.;
McCall,
A.;
Clark,
S.;
Menard,
F.;
33
34 Tamura,
M.
Science
1998,
281,
672‐674.
35 12.
Bailey,
J.
Orig.
Life
Evol.
Biosph.
2001,
31,
167‐183.
36 13.
Buschermohle,
M.;
Whittet,
D.;
Chrysostomou,
A.;
Hough,
J.;
Adamson,
A.;
37 Whitney,
B.;
Wolff,
M.
Astrophys.
J.
2005,
624,
821‐826.
38 14.
Breslow,
R.;
Levine,
M.;
Cheng,
Z.
Orig.
Life
Evol.
Biosph.
2010,
40,
11‐26.
39
40 15.
Levine,
M.;
Kenesky,
C.;
Mazori,
D.;
Breslow,
R.
Org.
Lett.
2008,
10,
2433‐2436.
41 16.
Chruma,
J.;
Liu,
L.;
Zhou,
W.;
Breslow,
R.
Bioorg.
Med.
Chem.
2005,
13,
5873‐
42 5883.
43 17.
Morowitz,
H.
J.
Theor.
Biol.
1969,
25,
491.
44
18.
Breslow,
R.;
Levine,
M.
Proc.
Natl.
Acad.
Sci.
USA
2006,
103,
12979‐12980.
45
46 19.
Klussmann,
M.;
Iwamura,
H.;
Mathew,
S.;
Wells,
D.;
Pandya,
U.;
Armstrong,
A.;
47 Blackmond,
D.
Nature
2006,
441,
621.
48 20.
Klussmann,
M.;
Izumi,
T.;
White,
A.;
Armstrong,
A.;
Blackmond,
D.
J.
Am.
Chem.
49 Soc.
2007,
129,
7657‐7660.
50
21.
Noorduin,
W.;
Izumi,
T.;
Millemaggi,
A.;
Leeman,
M.;
Meekes,
H.;
Van
Enckevort,
51
52 W.;
Kellogg,
R.;
Kaptein,
B.;
Vlieg,
E.;
Blackmond,
D.
J.
Am.
Chem.
Soc.
2008,
130,
53 1158‐1159.
54 22.
Langenbeck,
W.
Tetrahedron
1958,
3,
185‐196.
55 23.
Breslow,
R.
Tetrahedron
Lett.
1959,
1
(21),
22‐26.
56
57
24.
Breslow,
R.;
Cheng,
Z.
Proc.
Natl.
Acad.
Sci.
USA
2010,
107,
5723‐5725.
58 25.
Hein,
J.;
Blackmond,
D.
G.
Accts.
Chem.
Res.
2012,
30,
xxxx.
59
60
13
ACS Paragon Plus Environment
Journal of the American Chemical Society Page 14 of 15
1
2
3
4
26.
Senma,
M.;
Taira,
Z.;
Osaki,
K.;
Taga,
T.
J.
Chem.
Soc.
Chem.
Commun.
1973,
880‐
5 881.
6 27.
Breslow,
R.;
Cheng,
Z.
Proc.
Natl.
Acad.
Sci.
USA
2009,
106,
9144‐9146.
7 28.
Fuller,
W.
D.;
Sanchez,
R.
A.;
Orgel,
L.
E.
J.
Mol.
Evoln.
1972,
1,
249‐257.
W. D.
8 29.
Powner,
M.
W.;
Gerland,
B.;
Sutherland,
J.
D.
Nature
2009,
459,
239–242.
9
10
30.
Breslow,
R.
Tetrahedron
Lett.
2011,
52,
2028‐2032.
11 31.
Breslow,
R.
Israel
J.
Chem.,
2011,
51,
1‐7.
12 32.
Pizzarello,
S.;
Weber,
A.
Orig.
Life
Evol.
Biosph.
2010,
40,
3‐10.
13 33.
Burroughs,
L.;
Vale,
M.;
Gilks,
J.;
Hayes,
H.;
Christopher,
J.;
Clarke,
P.
Chem.Comm.
14 2010,
46,
4776‐4778.
15
16
34.
Fernandez‐Lopez,
R.;
Kofoed,
J.;
Machuqueiro,
M.;
Darbre,
T.
Eur.
J.
Org.
Chem.
17 2005,
24,
5268‐5276.
18
19 Figure
captions
20
21
22 Figure
1.
Some
α‐methyl
amino
acids
discovered
in
the
Murchison
and
other
23 meteorites,
all
of
which
have
a
small
excess
of
the
S
configuration
that
was
24 described
as
L.
The
enantioexcesses
showed
a
range
of
values
in
this
and
later
25 work.
26
27
28 Figure
2.
Transaminative
decarboxylation
of
the
imine
from
S‐α‐methylvaline
and
29 an
α
ketoacid.
In
the
presence
of
Cu(II),
which
catalyzes
the
reaction,
the
product
30 amino
acid
has
an
excess
of
the
L
enantiomer,
but
not
without
the
Cu(II).
This
is
31
32
expected
if
a
square
planar
complex
is
formed
between
two
imines
and
the
Cu(II),
so
33 after
decarboxylation
of
one
ligand
it
is
protonated
selectively,
guided
by
the
34 stereochemistry
of
the
other
ligand.
35
36 Figure
3.
The
formose
reaction,
producing
sugars
from
formaldehyde.
37
38
39
40 Figure
4.
The
nucleosides
that
were
examined
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
14
ACS Paragon Plus Environment
Page 15 of 15 Journal of the American Chemical Society
1
2
3
4
Graphic
Abstract
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
ACS Paragon Plus Environment