Amino Acid Analysis by Dansylation:
A Revised Method
An Honors Thesis (ID 499)
by
Kay Stephens
Thesis Director
Ball State University
Muncie, Indiana
August, 1986
Second Summer Session 1986
:--iF("!!
111£" :,{<
TABLE OF CONTENTS
Introduction
page 1
Literature Review
page 4
Theory
Chromatography
page 5
Dansylation
page 7
Materials and Equipment
page 10
Separation of Five Dansylated Amino Acids
Method
page 12
Chromatography
page 14
Results and Discussion
page 14
Separation of Seventeen Dansylated Amino Acids
-
Method
page 24
Results and Discussion
page 26
Conclusions
page 39
References
page 40
INTRODUCTION
Derivatization of amino acids by dansylation is
commonly used to identify N-terminal amino acids of peptides
and constituent amino acids of protein hydrolysates.
The
dansylation occurs when
1-(N,N-dimethylamino)naphthalene-5-sulfonyl chloride (dansyl
chloride or Dns-Cl) and an amino acid (aa) react to form the
dansylated amino acid (Dns-aa)
H~~a
Ht=1
II
,/
S-CI
1\
0
Dns-Cl
+
101
101,
I
~O
N-C-C
HI
'OH
h
shown.
3S
Hel
)
~
"J"'--Sc
H l
3
Amino Acid
-
,/
0
II HI R
I -c~O
S-N-C ,
~
~
(1)
0101
Dns-aa
A few problems have existed with this procedure.
A
large excess of dansyl chloride (necessary for amino acid
dansylation) results in a competing reaction that converts
Dns-aa's to dansylamide (Dns-NH 2 ) and various other
products, depending on the amino acid involved. To slow the
decomposition reaction, quenchers are added to use up the
excess Dns-Cl after the dansylation of amino acids has
occurred.
Amino acids derivatized collectively are
separated and each amino acid is identified by high
performance liquid chromatography (HPLC).
The principle of
2
-
the dansylation reaction and of HPLC are discussed in detail
later in this paper.
Information for the identification of
dansylated amino acids is in the form of an HPLC
chromatogram such as the one shown in Fig. 1.
d£
~8
~
~
A
B
"'.MeOH
:5
7'0
"'. THF
3
3
"'. AcOH
0.:57
"'. TEA
0.088 0088
100
I
100
~
.
...§
j
5
~~
____
~
to
______
ISS
~ .I'
....
....i ....
~/
,-
~~
.I'
/
.I'
~8
.I'
_.
~
':..
...
~
i
i
,-
0.:57
..
~l
.I'
____
20
TIME. minutes
~~
30
____
~o
40
Figure 1. Reverse phase separation of Dns-aa's on
a 5 micron C-18 Hypersil column, 15 cm x1~.6 mm.
Gradient conditions given in the figure.
In the reverse phase HPLC mode, polar molecules elute
first, followed by molecules of in order of increasing
nonpolarity.
In order to stop the dansylation reaction 3nd prevent
unwanted side reactions from occurring, a reaction quencher
is added.
As a res·ult, a dansylated quencher peak appears
in the chromatogram along with the amino acids.
The two
3
quenchers used by
.
Tapuh~,
et ale 11 ' 12 to stop the
dansylation reaction were N-methylamine hydrochloride
(NMe-HCl) and N-ethylamine hydrochloride (NEt-HCl).
To
prove useful to chromatographers, the reaction quencher must
result in a peak that elutes far removed from the dansylated
amino acid peaks.
It should not interfere with the
identification of the other peaks in any way.
The peaks
resulting from dansylation of the quenching material are
dansyl methylamide (Dns-NHMe) and dansyl ethylamide
(Dns-NHEt.)
In this study, separation of the same Dns-aa's
following an NEt-HCl quench did not occur as well as
reported by Tapuhi, et al. 11
Ammonium hydroxide can also be used to quench the
dansylation reaction.
large quantities.
This procedure produces Dns-NH
in
2
Chromatograms of this procedure display
enlarged Dns-NH 2 peaks. The Dns-NH peak is surrounded by
2
several dansylated amino acids in the chromatogram. The
increase in size of the Dns-NH 2 peak by the ammonia quench
repeatedly lead to a masking of the Dns-glycine peak with
the chromatography equipment used in this study.
be others who have had problems obtaining
There may
good separations
due to the inherent variability in chromatography equipment
and columns.
This project examines the use of pyridine as
the reaction quencher.
-
The aim was to simplify the
4
chromatogram by keeping the dansylation by-products to a
minimum whose chromatographic peaks might interfere with
those of dansylated amino acids.
In using pyridine, no
dansylated quencher is produced, and the amount of Dns-NH 2
produced is curtailed by the fast-working quencher.
LITERATURE REVIEW
Dansylation has been done for years to identify free
amino acids as well as N-terminal amino acids in proteins or
peptides. 16 Reversed phase high performance liquid
chromatography (HPLC) using precolumn dansyl chloride
derivatization was designed as a sensitive system for
identification of amino acids in the low femtomole
range. 1 ,6,10,11 ,12,14 A widely used precolumn derivatization
method is that of Tapuhi, et al. 12
This procedure is
reported to give high reaction yields, though it was found
by DeJong that changing the chromatography protocol (solvent
composition, gradient parameters, column temperature, etc.)
to suit the chromatographic system was necessary for
reproducible and relatively complete separation of the
dansylated common amino acids. 2 Naturally, other methods
have been devised in which both the dansylation reaction and
the chromatography are handled quite differently. 15
This
5
author, however, found the dansylation method of Tapuhi, et
al. 12 to be simpler and faster than most of the others. The
following study was undertaken to further simplify the
chromatograms obtained after following this procedure.
During this study, an advantage was found in using pyridine
as the dansylation reaction quencher rather than a primary
amine, as suggested by Tapuhi, et al. 12 This paper compares
the three methods of quenching (using N-ethylamine
hydrochloride, ammonium hydroxide, and pyridine) by analysis
of chromatograms obtained from comparable experiments.
The
clarified chromatograms obtained using pyridine indicate
that pyridine may be a better choice for quenching the
dansylation reaction than N-ethylamine hydrochloride or
ammonium hydroxide.
THEORY
Chromatography
Reverse phase high-performance liquid
chromatography separates molecules on the basis of
nonpolarity.
The chromatographic column is packed with
silica particles.
Some of the particles are bound to carbon
chains of various lengths, and many of the others are
end-capped with methyl groups.
The C-18 column has chains
eighteen carbons in length covalently attached to the silica
6
particles.
These chains create a nonpolar environment with
which nonpolar molecules can interact.
The nonpolar
alkyl-silica particles form the stationary phase of the
column.
When a sample of diverse molecules is injected on
the column, the relatively nonpolar molecules are attracted
to the nonpolar C-18 chains, while the more polar types are
easily washed from the column by a polar solvent.
Generally, in effecting a sample separation, a gradient
of two or more solvents is used in which the most polar
solvent begins in high concentration and is gradually
reduced until the more nonpolar solvent is prevalent.
In
this way individual molecules are eluted from the column
when the flowing solvent mixture approaches them in polarity
(nonpolarity.)
Upon leaving the column, these molecules can
be detected by various methods.
Ultraviolet/visible
absorption, fluorescence, and electrochemical detection are
three examples.
A chart recorder is utilized to graphically
display the elution profile of the sample, showing the
retention time and relative quantity of each of the
particular compounds.
7
-
The dansylation of amino acids is used in
Dansylation
conjunction with chromatography to identify amino acid
residues of peptides and proteins. 16 Dansyl chloride,
1-(N,N-dimethylamino)naphthalene-5-sulfonyl chloride,
covalently binds to free amino, phenol, imidazole and
sulfhydryl groups.13 The reaction mechanism is as follows:
0
H~>-S\
II
Ht=
\ I
S-CI
II
HCl
)
Hf'"S'
--'---:>'"
Hf'
- 0
II
+
\ I
o
Amino Acid
Dns-Cl
H
R 0
I _ I -<:.~ (1)
S-N C
~
~
\
OH
Dns-aa
These derivatives are easily produced for all amino acids,
and are intensely fluorescent and highly resistant to acid
and alkaline hydrolysis. 16 For these reasons, others have
used them for N-terminal amino acid identification in
proteins 4 ,5 or for detection of nanomolar quantities of
amino acids and peptides separated by thin-layer
chromatography.3,9
Using the dansylation procedure outlined by Tapuhi,12
reaction usually goes to more than 90% completion.
Dansylation of a mixture of amino acids results in
fluorescently tagged amino acids, still
chromatographically distinguishable by their varying
polarities.
-
In other words, this dansylation technique
8
allows the eluting amino acids to be "seen" by fluorescence
detection.
In the dansylation scheme, side reactions are present
which produce dansyl sulfonic acid (dansic acid, Dns-OH) and
dansylamide (Dns-NH 2 ):
Dns-Cl + H2 0
Dns-aa
+
Dns-Cl
Dns-NH 2
Dns-OH
+
+
HCl
various products.
Both reaction (1) side reaction (2) are high pH favored.
Side reaction (3) decomposes dansylated amino acids. If the
dansylation reaction is not terminated by using up the
excess dansyl chloride, this side reaction will essentially
undo the amino acid dansylation.
Termination is performed
by the addition of a quencher. Quenching the reaction also
allows the samples to be stored with no change in
fluorescence (about twelve hours) and prevents Dns-Cl from
reaching the column for on-column reaction Hith mobile phase
components. 12
Primary amine solutions such as ethylamine or
methylamine hydrochloride were used by Tapuhi, et al. 11,12
for dansylation reaction termination.
Reasons given for
using these quenchers state that the particular dansylated
quencher peak formed in either case a) serves as a indicator
to show that the dansylation reaction is occuring, b) elutes
at a position well separated from other dansylated amino
acids, and c) can be used as an internal chromatographic
9
-
standard. 12
Another logical choice for a quencher is
ammonium hydroxide since it would merely produce more
dansylamide and cause no additional peaks in the
chromatogram.
Pyridine, which produces no dansylated
quencher peak, is even more convenient with regards to the
chromatogram.
The proposed mechanism of quenching as
performed by pyridine (pyr) is as follows.
~~\
o· IIOH-~\
H:M13-\011
11.(,-\ I
- ?,
\/S-CI+Q~ H
\
g,-/.
o
HC
3
-
Dns-Cl
II
\ .
'"
J
-
\ I
+
pyr
~
fi\.~~
I
~
intermediate
~
~ (4)
+~
S-OH
~
Dns-OH
+
pyr
Pyridine reacts with the excess dansyl chloride in a
nucleophilic type addition to form an unstable intermediate
that is susceptible to attack by water and quickly converted
to dansic acid, thus regenerating the pyridine.
Pyridine
works quickly on the excess dansyl chloride, converting it
to dansic acid before more dansylamide can be produced in
side reactions.
The dansic acid peak is the first large
peak to appear on the chromatogram.
It elutes early, and
its peak is quite far removed from the peak of the first
dansylated amino acid, so that increasing its size presents
no chromatographic problems.
-
10
MATERIALS AND EQUIPMENT
HPLC grade H 0 was made by purifying deionized
2
feedwater to 16 megaohms resistance with a Barnstead Water-I
Lithium carbonate (Li 2 C0 ) was
3
purchased from J.T. Baker Chemical Company. Hydrochloric
purification system.
acid (HCl), ammonium hydroxide (NH 40H ), glacial acetic acid
(AcOH), and tetrahydrofuran (THF), all reagent grade, were
purchased from Fisher Scientific.
distillation over metallic sodium.
The THF was dried by
HPLC grade solvents
triethylamine (TEA), acetonitrile (CH CN), methanol (MeOH),
3
and pyridine were also purchased from Fisher Scientific. The
standard mixture of common L-amino acids: alanine, arginine,
aspartic acid, cystine, glutamic acid, glycine, histidine,
isoleucine, leucine, lysine, methionine, phenylalanine,
proline, serine, threonine, tyrosine, and valine containing
(NH4)2S04 was obtained from Pierce Chemical Company.
1-(N,N-Dimethylamino)naphthalene-5-sulfonyl chloride (dansyl
chloride) and the L-amino acids serine, glutamic acid,
lysine, histidine, and tyrosine were obtained from Sigma
Chemical Company.
L-arginine and glycine were purchased
from Matheson, Coleman & Bell.
L-alanine was from Aldrich,
and DL-threonine, L-proline, and N-ethylamine hydrochloride
(NEt-HCl) were purchased from Eastman Kodak Company.
11
The Beckman Model 334 Gradient Liquid Chromatograph and
Altex Ultrasphere C-18 column - 4.6 mm x 25 cm (Beckman
Instruments) performed the sample separations. For sample
detection, the Shimadzu RF530 dual monochromator
fluorescence detector was used.
A Linear Instruments chart
recorder produced the graphical data. A column water jacket
was used and thermostatted at 50 o C. Chromatography solvents
were filtered by Gelman filtration membranes (0.45 micron,
alpha and GA-8 series.)
Samples of a few ml were filtered
through Gelman Acrodisc LC13 (0.45 micron) syringe filters
prior to chromatography.
Microliters of sample were
filtered by Bioanalytical System centrifuge filtration
membranes (RC-58, 0.2 micron).
12
SEPARATION OF FIVE DANSYLATED AMINO ACIDS
METHOD
Stock solutions of arginine, glycine, threonine,
alanine, and proline were individually made at a
concentration of 1.1119 mM in 40 mM Li 2 C0 buffer, which had
3
been brought to pH 9.5 with HCl prior to the addition of the
amino acid. The alkaline pH of the Li 2 C0 buffer is
3
necessary to keep the amino groups in the unprotonated form,
which is required for reaction with dansyl chloride.
From
the amino acid stock solutions, a sixth solution was made
containing all five amino acids, each 5.5597 x 10- 2 mM, for
a total amino acid concentration of 0.2780 mM.
Dansyl
chloride reagent (5.56 mM in HPLC grade acetonitrile) was
made fresh daily as it collects water and undergoes
hydrolysis to dansic acid upon standing.
The dansyl
chloride reagent and all dansylation reaction vials were
foil-wrapped to protect the dansylated amino acids from
light.
The fluorescence yield decays with exposure to
ultraviolet radiation.
Dansyl chloride and all dansylated
amino acids were stored on ice or in the freezer at _20 o C.
For the first set of chromatograms, 2 mls of the five
amino acid mixture and 1 ml dansyl chloride reagent were
combined and allowed to react for 21 minutes at 4o C. These
13
reactions were run at 4 0 C instead of at room temperature
because, with such small quantities of amino acid present
(compared to dansyl chloride), side reaction 2, the
decomposition reaction, seemed to occur too rapidly at room
temperature.
These reactions were quenched by the addition
of 100 microliters of either 15M NH 4 0H, 4% pyridine
solution, or 4% NEt-HCl solution. After quenching, the
vials were returned to the ice bath for 2 minutes, and then
prepared for HPLC analysis by 0.45 micrometer membrane
filtratrion into clean, dry, foil-wrapped vials.
Three
blank trials with no amino acid present (one terminated with
each quencher) were made from the dansylation reaction on
Li 2 C0
buffer containing no amino acids. For the reaction
3
on individual amino acids, 0.1 ml of the stock solutions
were diluted to 2 mls with buffer so that the concentrations
of acids would equal that of each acid in the five amino
acid mixture reaction.
In these and the blank runs, the
procedure was exactly the same as described for the 5 amino
acid mixture.
Chromatograms of the individual dansylated
amino acids were also obtained.
Comparison of the blank
with the individual amino acids and the five amino acid
mixture served to identify the locations of the dansylated
amino acid peaks.
14
Chromatography
For HPLC analysis, 20 microliter
samples containing 0.7174 nmoles of each amino acid were
injected.
The separation was done at room temperature on an
Altex (Beckman) Ultrasphere C-18 column equilibrated with a
solvent mixture of 14% Band 86% A.
A 40 minute gradient of
14% B to 100% B starting at injection (time zero) eluted the
dansylated products from the column.
The solvent system
consisted of solvent A (5% MeOH, 3% THF, 0.57% AcOH, 0.088%
TEA, by volume) and solvent B (70% MeOH, 3% THF, 0.57% AcOH,
0.088% TEA, by volume), the balance of each made up to 100%
with H2 0.
RESULTS AND DISCUSSION
Figures 2 through 7 display some of the chromatograms
obtained in the above procedure.
Figure 2 shows how
individual dansylated amino acid peaks were labeled.
The
chromatogram of a sample containing no dansylated amino
acids (blank sample) was compared to chromatograms, such as
the ones in figures 2b and 3a, of samples which contained a
single dansylated amino acid.
A peak other than those
present in the blank represents the elution of the
additional substance (dansylated amino acid) in the sample.
Figures 2b and 3a show the retention times of Dns-arginine
and Dns-threonine, respectively.
After other Dns-aa peaks
)
)
100~ B
100$ B
-roo-
- -t----:::::==;::..--==;:- _--:+-=;--(-=--
J):~__
.-
=-7:
--+-_._-
.::;t==
E'
=-:7:.
-r---.
-i"-'~
tJ:
tJ·
to
I
to
I
Z
~
I
::s
::s
o
~z
--..j
--,--UjJ:::
-'--
-----+
I
!
I .--
-- o·
~ - ~'-- ~1---'
+
-,
..
I
tJ.---+--
-. - ::s
=r:::=.:J.
to
~
:
r---
....
--
-,
--,
~.
~
'"1
----.
-=+="
~
g
Ii
r-~~~~:
.
i--
-- -
$
-----.
--.
-
-"-
--.
--r
'-r-
.~-
~
=+===!
-
=-~
-.-~
~
~
;.--..,
~
;-----1
=+==::
.....-
I
~-.
14$
B
. ".,.____
i
'--:=:=::-=-::L
~
o
I
6'
- --
12
C
- _ L-
18 -
_
.
--+-
;-~~I
j:::: -
~
:±
~
.--1.
24
Minutes
30
36
±::::=
,-
;-
+----
i
t==+==
L--
l~~i
--.---.-
~'~~:L=:::-=== .==:==-:~-
o
6
-12'
18
24
!
30
,J::.::::::l
36
Minutes
Figure 2. Separation of 20 microliters, NHuOH-quenched dansylation
reaction samples containing a) no amino acids and b) 0.1174 nmoles
Dns-arginine. All chromatograms in figures 2 through 14 obtained by
reverse phase separation on an Altex Ultrasphere C-18 column. Figures 2
through 7 employed a 40 minute gradient at room temperature.
~
U1
)
)
100~
d·h--·
a~-
~~-.
~
]~-:r
.:-~.
--.-+0-,
~~/~
1-'.
t-
~
~
---_._-
I---.-jf.
=+===I
t:
~:--
::t:_
..--
.:.-: -: ~ $===*=.
:--;-c'
::t:.J-I.I--_._~·_N
··-~tI
;=::s : .
to
I
---- -::::l:.==:!===.
---'-''1
-'-. ~=::4()Q : t-I.t==:I=1=
m
r-.
.o==:!.
~.===r
-r-
("t'-~
::J":=:~,
.
'. '1
==: ,_
.~
.-~'.
·-"':'-':'f~
A.
:tt=::::t:
III .::J:'~~:::J=j
'
F-:
~ ~EV'
.i:---.
==':-~:z:
---'
I
.f---.
f-I-~
:s:
-,
::~
r---.
r=:::::-
---,...-
14" B
F=F~
to·
I
("t'
M:
"---"~~
F=I..l
;...--
J-
12
ltl
30
Minutes
36
~
.::s H-------l.
§-
--1-1-:-
.L
r------: .-J.1;:;::::::::::t:
I
0
~
f--~
t=='
;:==j
.::s
to
~
1== •
1001 B
--
~--t­
-
+--~
.............
r-'
_J-- __ I
~:i---4~,:::r::===t==
~ ~--
-;:=:::t~
---+--r'~
c---=.:
~
-
-f~.
~
~
~~.
b
····
r---
-r--.
;.-
.--'----
B
o
... :::t=::....,
6
12 '18
24
30
.=t==
36
Minutes
-...
0'1
Figure 3. Separation of 20 microliters, NHuOH-quenched dansylation
reaction samples containing a) 0.7174 nmoles Dns-threonine and
b) 0.7174 nmoles each of a mixture (see text) of five dansylated amino
acids, Dns-glycine peak not present.
17
were individually identified, a chromatogram of the
dansylated five amino acid mixture (figure 3b) was obtained.
Dns-glycine was not identified in this series of
chromatograms as no peak appeared here or in the individual
Dns-glycine chromatogram.
Figures 2 and 3 represent the
dansylation reaction as quenched by NH 4 0H. Note, in all
four chromatograms, the presence of the large Dns-NH peak
2
resulting from the addition of ammonia. The Dns-threonine
peak is so small in comparison to the Dns-NH 2 peak that it
was not noticed immediately.
Figure 4 shows the results of the same chromatogram
analysis procedure applied to samples of the dansylation
reaction terminated by NEt-HCl.
The large Dns-NHEt peaks
were caused by dansylation of the NEt-HCl quencher.
Also,
the Dns-NH 2 peak decreased in size compared to those of
figures 2 and 3 owing to the fact that NH 4 0H was not added.
In fact, it decreased enough to allow complete separation
from the Dns-threonine peak.
There were no significant
changes in the appearances of the Dns-arginine, Dns-alanine,
or Dns-proline peaks, but Dns-glycine became visible as a
leading shoulder of the Dns-NH 2 peak. Poor separation in
this area of the chromatogram occurred because there was
still too much Dns-NH 2 in the samples.
18
-
III 0
Q)
I"""f_
0..0
e
CU'O
III C
CU
::::
-
o III
......
.j.lCU
o CU
CU I
Q) III
s..
C
o
of"'4
C
Q
0
c-
.j.l
0
ro-.j.l
I"""fro::<
>.
Q)
aJC.j.l
Cof"'4
ro ro Q)
Q.j.lQ)
Cr:I)
0-
o
=r
0
0
I"""fQ.J
Q) U s..
s.. :c
::J
I .j.l
tlO.j.l ::<
::J
of"'4
tzl of"'4
ez..:z:e
19
Figures 5b, 6 and 7 show the results of quenching with
pyridine.
Figure 5 compares the chromatograms of blanks
quenched with NEt-HCI (fig. 5a) and pyridine (fig. 5b).
The
pyridine-quenched sample contained no Dns-NHEt and less
Dns-NH
than the sample that had been quenched with NEt-HCI.
2
The decrease in Dns-NH is thought to occur because pyridine
2
converts excess Dns-CI to Dns-OH (note the Dns-OH increase
in the chromatogram) so quickly that it prevents the
decomposition reaction from occurring to the extent NEt-HCI
allows.
The enlarged Dns-OH peak does not interfere with
Dns-aa identification, because it appears far removed from
the first Dns-aa peak.
Figure 6 was included to show that
the identification of the tiny peak labeled Dns-threonine in
figures 3 and 4b was confirmed by increasing the
concentration of threonine twenty-fold prior to the
dansylation reaction.
Figure 7a displays the best
separation of Dns-glycine from Dns-NH 2 that was achieved
with a 40 minute gradient. Here, Dns-glycine eluted as a
peak rather than a peak shoulder and this was possible
because the Dns-NH 2 content in the sample was actually
minimized by the fast quenching action of pyridine.
The height of the Dns-NH 2 is misleading in this
chromatogram. It is affected by the presence of the nearby
eluting substance, Dns-glycine.
-
The area under the peak
20
CIl
.!::s:
c::
ro
...-I
..a
c::
o •
::c
'r-i
~O
o
=:t
ro::c
s..
Q):Z:
-
c::..a
o
.r-i
"0
~c::
roro
...-I
>....-1
CIlU
c::::c
ro I
"O~
tzJ
.. :z:
CIl
s..Q) ro
~
..c
'r-i
...-I ~
O'r-i
~
s..
o
.r-i
S
"0
Q)
..c
00
NC::
Q)
c....::s
00'
c::o CIl
'r-i ~ro
ro ro
s.. I
ro CIl
o.c::
Q)O
tI)
o
. c::
l!"\C::
'r-i
Q)
ro
s..~
::s c::
tlOO
'r-i
-
0
tx,.-
)
)
8-
~a:-
100~
~
_'___' ___ -=:::J __-::---
f'--=~::~~~~-'r::
---r--'
r
t---
8-::'
',7"--
100~
B
B
.,--~---t-
I--
r
b
-------.
----=l
0 ..
::s-
to
;..-...-+. I
_.~
,~,.~
~~- i.!:::=::- _~
S, ~:-,=}t*­
---:= =7:: __ .__ .
__~I~
-__ b
-
r----
0-,
CD
.'
__
__
F-~ b -
~::: '- - -~--.----=~--~ '-,f.---~
--- '~=rJto ~_._ _
'::'
'
- ,I
i
•
_
_
_
I
~
:s=ft:f:::"
--,-,~~~
:::r:::::=
-----;-H·
' t:;z:,
~N::X:~=r::::::
f§g'
--t===-.
~-.-
.
..--,
,
--'--
f---
c--
'==1~
~
c-
~:=~~_~=_~
~ ~------
r-:_-'
~
,
---,--'
-'-
.---,
:0
-
_..=::s
to
~
::s
38' .~
o
:=~=f=3
._
'--jl-----
I
'---
::x:
i
:j
:=:::: ::r
Cf-=--.
.---,!.--:: '1
----+
I
~:z:
r--~
I
________ J
N
::Ill,
_
,.
Ir'--.
J-
r• 1..- -
14%B
.+F-r--
14 ~ -.!! -t---•.
':;iL
~,
~-.~-
gj=-~.6 -- .-12. -- 'fir
0'"
.
-.6-_
J------~ -.-~+_
I--_' __ ~
:.:..-.• ~--- ~:::.
'
24
Minutes
30
----.-----
36
~L' 1-
o
6
~-~~
~:
12
.
f.--;
18
24
30
36
Minutes
....
I\)
Figure 6. Identification of Dns-threonine peak. Injections of
20 microliters, pyridine-quenched dansylation reaction samples -containing
a) no Dns-aa's and b) 14.35 nmoles Dns-threonine.
22
-
CIl
s.. cv
cv.c
~~
.....
...-4 r.-t
o
s..
0
o.c
..... 0
E!
o
co
cv
C'\ICIl
-<1>
...-4
CIl 0
CV E!
...-4C
Co
E!.::t
•
cor-~
CIl ....
c
><
r-cv
.~
00
.....
CV
~-cv
o.oCl)
.~ IlU lilll'llllilillilii ltll JII
I 11 ttltl.
11
Jill ll' Ual
II
co
cv"O s.. C •
CO CV
C
s..
[I I llll'o
o Q) ::3
..... c~
~ ..... ><
coo .....
...-4 >- E!
~ jJ. I'! 111t11 IIl/t . .I~ .1.. p.l i· 1.IU
II',!
II I l
r , Ir II
II 'r 1.\
ij ll] II
J.
' I "1' I ,I
l 'II ; ~. ~i '
1 till I 11t I I11I . IIrl ..II ·f I··I I ./'1'1frt j.I +~. .~1 i~
I:nlIi II +,i1 .~iil'fj
g' :,1' Ij '.
I..
./1
I:
.... I!
I
I
90
I ' III
!1ft
I , L 7r,j11:II
!:il Iilii",~:,:~:~~.
; II
il:
I
I ,; ,
-
::
'I!
III
I 1'60
I
I
SO
.11 11:11
I
,
Iii
Iii
...
JO
i
I I,
I . J . . Iii
I
I. :111
-
i,
1IIIi'IIII I'li I'
I:co: !IHUJ J11 lill !llil fi
iI~::.
fli I III l
..
I
.c
J", ~
I,'
tl ';C'\I
. ...
I
I
Ii::) llJ
:J IIIJ III '11 It ii al ,:;1 III 0
~
.::t
..-
~
~
If ~
cx:l
:':!II lilli,' 1,1,',1'11111 1
II
iiil III III I' II! Till III f
I I II: I!j~ II+I~
iiii II,ll! I b~~~OH ,lltI.!l L' Ii !~~ WI",
1
..
t::\
"0
0
CV CIl C
c
Ie
1
1
"Occo
i
2~1
I
I .....
C'tl CIl 0
M
1
40
1
Cf.I l:lO"O
C
\0
..
I II
100111
'1.1
>-...-4
<1> •.-4
o ...-4 E! .
coco
CV e
:s C CV
0"
>
I .::t
'.-4
cvr-r.-t
c'.-4
"0
'.-4
s..
r- CV
•
0
.c
~
e
>-0..
0
co
s..
bOr.-t
• C
r- ..... CIl
s::Q) •.-4 C'tl
s.. co co
:s
~
I
bOS::r/l
'.-4 0 s::
ez..0t::\
23
indicates quantity.
The matter of dividing the area under
this type of peak for quantitation of each substance is not
trivial and was not addressed in this study.
The separation
alone of the Dns-glycine peak served as evidence that there
was less Dns-NH
in the pyridine samples than those of the
2
other two quenchers.
Finally, figure 7b is a display of the dansylated five
amino acid mixture, quenched with pyridine showing an
increase in the amounts of Dns-aa's except for Dns-proline.
This could be due to pyridine's curtailing of the Dns-aa
decomposition side reaction as proposed in reaction (4).
The significance of these chromatograms is that they
point out the possibility of important peaks being hidden by
the presence of large side reaction peaks.
To better
appreciate the advantage of using pyridine as the reaction'
quencher, spectra were next obtained on a mixture of 17
amino acids.
24
SEPARATION OF SEVENTEEN DANSYLATED AMINO ACIDS
METHOD
For the separation of the seventeen amino acids
commonly found in a protein hydrolysate, a standard mixture
of L-amino acids was used that contained serine, methionine,
aspartic acid, arginine, glutamic acid, glycine, lysine,
cysteine, threonine, alanine, valine, proline,
phenylalanine, isoleucine, leucine, histidine, and tyrosine
dissolved in 0.1 M HC1.
The mixture contained 1.25
micromoles cystine per ml and 2.5 micromoles/ml of each of
the other amino acids and (NH4)2S04.
For the reaction
stock, 100 microliters of the standard amino acid mixture
were diluted to 1 ml with 40 mM Li 2 C0 buffer, pH 9.5. The
3
resultant stock solution contained 0.0125 micromoles cystine
and 0.025 micromoles of the other amino acids per 100
microliter solution.
A Dns-Cl reagent solution was made at
a concentration of 43.75 mM in HPLC grade CH CN. Adding 50
3
microliters Dns-Cl reagent to 100 microliters amino acid
mixture was in keeping with the 5 : 1 (dansyl chloride :
total amino acid) mole ratio as suggested by Tapuhi, et
al. 12 As before, the reaction was run at 4 0 C on an ice bath
and was quenched after 21 minutes in one of three different
ways.
The addition of 10 microliters 2% NEt-HCl solution,
25
5 microliters 15 M NH 4 0H, or 10 microliters 2% pyridine
solution terminated the dansylation reaction.
After 2
additional minutes in the ice bath, the terminated reaction
mixtures were prepared for HPLC analysis by 0.45 micrometer
pore size membrane filtration.
Injections of 10 microliter samples into the HPLC
system contained 0.7813 nanomoles cystine and 1.5625
nanomoles each of the other amino acids (mostly dansylated,
assuming greater than 90% reaction to have occurred 12 ) for
NEt-HCI and pyridine-quenched trials.
The NH 0H-quenched
4
reaction contained 0.8065 nanomoles cystine and 1.6129
nanomoles of the other amino acids per 10 microliter
injection.
No more than 0.109 micromoles Dns-NEt-HCI or
0.113 micromoles Dns-NH 2 were produced in reactions
terminated by NEt-HCI or NH 4 0H, respectively, per 10
microliter injection. Pyridine-terminated reactions
contained 0.158 micromoles pyridine per 10 microliter sample
that does not appear in the chromatograms but could have
increased the amount of dansic acid by as much as 0.109
micromoles.
-
26
RESULTS AND DISCUSSION
Chromatograms of blanks (figures 8 through 10)
contained the same basic information as before, but were
obtained for a sixty minute gradient.
The dansylation
reaction samples that had been quenched with NEt-HCl (fig.
8) and NH 0H (fig. 9) contained dansylated quencher
4
molecules which appeared as Dns-NHEt and Dns-NH 2 in the
chromatograms.
The sample that had been quenched with
pyridine (fig. 10) contained no dansylated quencher
molecules, but contained an increase in the amount of Dns-OH
evident in the first 6 minutes of the chromatogram.
This
sample and the one quenched with NEt-HCl (fig. 8) contained
a very small amount of Dns-NH 2 , evident in the chromatogram
at about 24 minutes, from ammonia impurity.
Probable
sources of ammonia include the HCl used to buffer the Li C0
2 3
solution to pH 9.5 as well as the Li 2 C0 itself. Still
3
another possible source of ammonia is the air. When working
at the nanomole level, the ammonia in the air, especially
from a nearby smoker, is enough to contaminate a sample. 7
Other peaks occurred in these blanks that were not
identified, but they did not interfere in the identification
of the Dns-aa peaks of figures 11 through 14.
27
-
100% B
"
:'~-=
'--r-
.==
'---'
'"" c.J,
-
1---'
1---,
--_.:
f--- -
,
;-, ~ :::~ ~'. -~
-:I---~
,,."
o
I"-
,~+-
--
114"B10_ _II-
6
-
6
Figure 8. Reverse phase separation of 10 microliter
injection, NEt-HCl-quenched dansylation reaction blank.
Chromatograms in figures 8 through 14 were obtained from a
60 minute gradient with the solvent sgstem as shown in
figure 1, and the column heated to 50 C.
28
-
100% B
i5
o
-~-
I
.
--'3 ---·-__
-H\:::1
:s
. C/l
to
I
I
o
:::c
N
-
:z:
:::c
-== 1:-==-=======
===
====-~--'-..J
__
._.•
_.
.'::l
-
.- -i-""
_.
-.
..:.,
--4-jf--------. c -/-.
1- .
.-. -
--
~.
==.
-
- -- ..
...
- ..
----;:----~ . . .~
o
6
12
18
24
30
36
42
48
54
60
Minutes
Figure 9. Separation of 10 microliters, NH 0H-quenched
4
dansylation reaction blank .
-
.
29
-
100% B
30
36
42
48
54
60
Minutes
Figure 10. Separation of 10 microliters, pyridine-quenched
dansylation reaction blank.
-
30
-
Figure 11 displays the chromatographic problem
experienced in this study with dansylated samples that had
been quenched by NEt-Hel.
The dansylated quencher peak
eluted at 38 minutes, only slightly altering the retention
times of Dns-methionine and Dns-valine without hindering
their identification.
More troublesome was the fact that
two peaks reported in the literature 11 to appear before
Dns-NH
(epsilon-Dns-lysine and Dns-glycine) coeluted from
2
the column resulting in only one peak.
Figure 12 shows epsilon-Dns-lysine and Dns-glycine as
they eluted before the Dns-NH 2 peak in an NH 4 0H-quenched
sample. The reason for their separation in this
chromatogram is not clear, especially because it was not a
reproducible occurrence.
One explanation given 7 was that a
very large peak that appears in a chromatogram in the area
of several others may cause certain surrounding peaks to
behave differently in the chromatogram than they normally
would.
Whatever the reason, this was not a reliable method
due to poor reproducibility.
Figure 13 shows the chromatogram of a Dns-aa sample
that had been quenched with pyridine.
reproducible.
This chromatogram was
As predicted by reaction (4) there appeared
no Dns-quencher peak here.
Therefore, no interference with
Dns-aa identification from the presence of a Dns-quencher
-,
31
100% B
-
t1
===
(.Q
r== -
==
--
B
~
::I
oI
::c
I
===-1===
0----
I
t1
::I
CJl
I
f-'
'<-
===-
-Ut
--
~-
+
-f---
-
t1
~t1
===--===-
::Ii
--::I
f-CJl
(.Q ~o­
I ,0
- f--:I
f-'
- <-ru
----
:(1)
~f-'~
--
I
C
-- -
+- -
-
a-
-:
-
'<
-~
___
J
~
---
--
\::::I
::s
I-f-',.,J
o
I-
\::::I
1--
--
1--
t-
14%B
1-
1--
~-I---
~
--~\--
o
6
-~
12
18
24
30
--
- - -i--
l-
f--:--
l-
I-- -
1-1--
36
1--
::I
c-t-
I
-S
CJl~
f-
--
CJl
I
\--
f-:..,
f-: ~
t-.
::r1-"
~
r--
.x- 0
-0
42
48
54
00
Minutes
--
Figure 11. Separation of 10 microliter injection of
seventeen dansylated amino acids. The dansylation reaction
was quenched with NEt-HCl (details in text.) Each Dns-aa
peak represents 1.563 nmoles, except for Dns-cystine (0.7813
nmoles.)
32
-
I
"1
,
100% B
00
'0
0
\:1
::s
to
,
1== 8
HI
,
,
~
,
It
'='
::s
to
\:1
c7"- C»
to
-
::s
,
I
I
0
z
::c
N
--- 0
r'='
::s
::c
to
I
'0
0-
I
I
'='
::s
0
1 - '"
to
I
~~
It
to
'='
::s
to
I
8
<D
eT
I
\:1
I
c:
::s
~
::s
'<
.~
::s f----'to·
::s
to
:;
to
I
.-
<
3:=1= to 1-'
<D.=F= I rueT =t= ru "1_
I
. toOQ.
,
'0
N at::'
.
(J)
.,-"';":: ::s ."
to, .
14 % B
I
i=
It
U1
0
,
t::'
::s
'='
::So
to
I eT
0"1
,
c
~
t::'~
'=' ::s::
..=-I-::S to-
I
~
c-
ru
I
C
,
0
~
'~
til-
0-
to
w
~
\:1-
::s
,
to
I
ru
I
to
I
I
-t--'< ,
..-.:-- \:1
~>oJ
::s
0
~
,
ru
-
..
~l
~
::s
0
0
{r-
'<
"1
to
I
....to
0-
:
"1
,=,: ::st::'
,
::s:to
. I
o
6
12
18
0
30
36
42
48
54
60
Minutes
Figure 12. Separation of 10 microliter injection of
seventeen Dns-aa's from NH 0H-quenched dansylation reaction.
Each Dns-aa peak represents4 1.613 nmoles, except for
Dns-cystine (0.8065 nmoles.)
!
33
t::I
::3
to
l
-
o
-
t::I
::3
to
1
t::I
::3
to
I
---'
'-c=
===:Jc
o
--
.....
~-
Cl)
to
-
----- ~ --------~
----~+-------------- :g
f-- >-,1
-
f---
.
I--
t::I-
.....
-~'
, ----..-, ~~:
-~
,.,
---
-~-
===4k===~======~~~
:-- ~._._:- ---.
'~'- rk·
~--' -~.~
h'
Q.
_"
~~~~~~~~~~~~~~~~~.~~~~
---
t::I
::3
to l
. . . ====
-
-Cj~
::3
-
.....
::3 cT- t::I ~
to _.-::3~
'..J
t::I::3'
to
1
'<
o::;~~-
--__
~I---
-~
'
_S::
-b'"1=:::::::<
='= rJl c::=:=
I
,-
I
I Cl)
"d .,-::T . ':'
Cl) : t::I
::3
to
I ,
t::I~
I
::I:
100% B
t::I
::3 ~
to ......
<5
1---
~-''<.d
- to
-. - --£- --, -- b~'
, --~~ _
.-
, ::3
--
to
'"1 t::I .: ~. --t::I
~~. _~tl-?-~:~:-,:7-~_~'===-:'" ~()Q ~ ~;-= ~
-- - '/=-- -
rJl-_-
~-.---HV_~':
-
--,
~/--.
~:..=
j:.
-
,
t-
---,
--
1
•
1- .....
to ~= '<
~~_ _
.1-
:.; '"1.- 1-
f--
-.-'
.
1-:"-..... :
'rJl
"'~Cl)
.
r-I'
f-III
-
I~
. -
Ill'
1-
-nf__---I
---
' _ ------:,
IJ- ~
--
14%B~~==
.~~'ltJ~-I..:~-·-I::.II.II-~·II'I~~'-I==-I--·III-\:~.-~=I'~~-I·'·~-I-~'
~~~:~==~~~-,,~~l~
t-~ \;2....;/' ~i
- - ---
---
'===
o
- -
--
6
12
',::3
1
to
'"'-rJl~'1
18 1 ~ z
III .... ::t:
rJl
"d
30
36
42
48
54
60
N
Minutes
-
Figure 13. Separation of 10 microliter injection of
seventeen dansylated amino acids from pyridine-quenched
reaction. Each Dns-aa peak represents 1.563 nmoles, except
for Dns-cystine (0.7813 nmoles.)
34
-
100% B
,5
,:::-
===t:1
===
:1
===
=== I
-0
CIJ
-
--:;z.
/-
~~-;L
:;i:=
....
'
,C
t:1
,.!,0
:1.
CIJ
::::.- I
c1"
;,< .
---r--., ~ .
.---- .-
----l-~~
14% B-
~~.
o
6
12
18
24
30
36
42
48
54
60
Minutes
-
-
Figure 14. Identification of peaks. Separation of 10
microliter injection of four Dns-aa's from pyridine-quenched
reaction. Each Dns-aa peak represents 1.563 nmoles.
35
peak was possible.
Furthermore, the Dns-NH 2 peak was
greatly reduced and consistently allowed the separation of
Dns-threonine and Dns-glycine.
Epsilon-Dns-Iysine appeared
as a shoulder on the Dns-glycine peak in these chromatograms
but, with additional work on the chromatography protocol,
this problem should be overcome.
Figure 14 is included as an example of how the peaks
were identified.
It contains four dansylated amino acids
(Dns-serine, Dns-glutamic acid, Dns-Iysine, and
Dns-tyrosine) on which the same dansylation and
chromatography procedures were performed.
Insight as to the
location of several of the Dns-aa peaks was attained through
the literature 11 . This allowed peak identification
information to come from chromatograms of small groups of
amino acids as well as from chromatograms of individual
dansylated amino acids.
As found in the previous discussion section, several of
the dansylated amino acid peaks in the samples quenched with
pyridine are larger, possibly due to the effect pyridine has
on the decomposition side reaction (3).
The attentive reader will have noticed that a slight
shifting of Dns-NH 2 occurred in comparing blank
chromatograms to those of samples containing Dns-aa's.
is a common occurrence in chromatography at locations
This
36
several minutes after injection and well into the gradient
as in this case.
This shifting is especially evident when
comparing the location of an internal standard peak
(Dns-NH ) in a blank to its location in a sample where other
2
materials have eluted previous to it. This also occurred in
figures 2 through 7.
Certainly, because of potential problems with shifting
and reproducibility, more work should be done to enable
absolute identification of peaks, especially those eluting
near the Dns-NH 2 peak.
The labels of the peaks in the chromatograms of figures
11 through 14 were assigned by comparison with blank
chromatograms and chromatograms of small groups of amino
acids.
Aside from absolute peak identification, the
important issue should not be overlooked.
As predicted when
this study was initiated, terminating the dansylation
reaction by the addition of pyridine 1) eliminates the
presence of a Dns-quencher peak and 2) reduces the
inevitable Dns-NH 2 peak considerably. This second point is
not trivial and occurs to the extent that previously hidden
peaks can be resolved from Dns-NH .
2
37
-
Ideas for further work include improving
reproducibility.
This may involve reworking the solvent
system and gradient pattern, or experimenting with a variety
of HPLC columns. 2 Absolute identification of all amino acid
peaks is not very likely without good reproducibility in
retention times.
Another worthwhile project would be to get
better control over the amounts of materials in the
chromatogram.
When an unusually large peak is recorded on a
chromatogram, it typically distorts those near it.
This may
result in shifting peaks, coelution with the major peak, or
misrepresentation due to coelution with another peak of
similar size.
One approach attempted, though not long enough to
perfect the method, was the spiking of the standard amino
acid mixture prior to dansylation with individual amino
acids.
This process carefully done could help to identify
questionable peaks.
One needs to have an idea beforehand of
where to expect a particular amino acid to elute and then
perform several trials on each spiked mixture to obtain
convincing data with this procedure.
This is because
certain problems are inherent to pre-column derivatization,
according to DeJong, et al., one of which is widely
differing detector responses for equal amounts of dansylated
derivative. 2
.-
38
Identifying peaks that have been reported in the
literature as unknown, or those that arise in blank runs
would certainly be helpful.
As a first step to doing this,
one could reflux dansyl chloride in acetonitrile and get
valuable information about possible contaminants from a
chromatographed sample of this mixture. 8
Finally, a two-part question that should be addressed
concerns the derivatization reaction yield and the product
stability over time of the reaction quenched by pyridine.
39
CONCLUSIONS
A widely accepted method for accomplishing dansylated
amino acid separation has been studied, and a more
convenient way to terminate the derivatization reaction is
suggested.
This method employs pyridine as the quencher and
results in simpler chromatograms with less interference from
dansylated by-products and quenchers.
As chromatography
systems vary, improved or "personalized" elution schemes may
be needed to accurately interpret the results, especially
regarding the particular peak(s) uncovered with a given
system.
Although the elution scheme has not yet been
perfected for the system used here, this researcher feels
the significance of pyridine quenching of the dansylation
reaction has been demonstrated.
This study has shown that
the lack of a potentially interferring dansylated quencher
peak and the substantial decrease in the Dns-NH
peak are
2
the major advantages of quenching the dansylation reaction
with pyridine.
-
40
REFERENCES
1. Bayer, E., Grom, B., Kaltenegger, B., and Uhman, R.,
Analytical Chemistry, 48 (1976) pp. 1106-1109.
2. DeJong, C., Hughes, G.J., Van Wieringen, E., and Wilson,
K.J., Journal of Chromatography, 241 (1982) pp.
345-359.
3. Deyl, Z., and Rosmus, J., Journal of Chromatography, 20
(1965) p. 514.
4. Gray, W.R., and Hartley, B.S., Biochemical Journal, 89
(1963) p. 380.
5. Gray, W.R., Methods in Enzymology, 25 (1972)
p. 121.
6. Hsu, K.-T., and Currie, B.L., Journal of Chromatography,
166 (1978) pp. 555-561.
7. Johnson, E.R., Ball state University, July, 1986,
personal communication.
8. Kruger, T.L., Ball State University, July, 1986,
personal communication.
9. Morse, D., and Horecker, B., Analytical Biochemistry, 14
(1966) p.429.
10. Olson, D.C., Schmidt, G.J., and Slavin, W.,
Chromatography Newsletter, 7 (1979) pp. 22-25.
11. Tapuhi, Y., Miller, N., and Karger, B.L., Journal of
Chromatography, 205 (1981) pp. 325-337.
12. Tapuhi, Y., Schmidt, D.E., Lindner, W., and Karger,
B.L., Analytical Biochemistry, 115 (1981) pp.
123-129.
13. Weber, G., Biochemical Journal, 51 (1952)
p. 155.
14. \'leiner, S., and Tishbee, A., Journal of Chromatography,
213 (1981) pp. 501-506.
-
41
15. Wiedmeier, V.T., Porterfield, S.P., and Hendrich, C.E.,
Journal of Chromatography, 231 (1982) pp. 410-417.
16. Zanetta, J.P., Vincendon, G., Mandel, P., and Gombos,
G., Journal of Chromatography, 51 (1970) pp. 441-458.