CLIN. CHEM.22/9. 1497-1502(1976)
Reverse-Phase Chromatography of Polar Biological
Substances: Separation of Catechol Compounds
by High-Performance Liquid Chromatography
lmre Moln#{225}r
and Csaba Horv#{225}th
Catecholamines and their metabolites have been separated isocratically
by reverse-phase
chromatography
with
aqueous (no organic solvent admixed) eluents. Unlike
ion-exchange or ion-pair chromatography, mixtu’es of both
acidic and basic substances can be separated in a single
chromatographic run, because the retention is governed
by hydrophobic interactions between the nonpolar moiety
of the solute molecules and the octadecyl-silica stationary
phase. The relative retention values strongly depend on
the pH of the eluent, which governs the degree of dissociation of ionogenic solutes. The reproducibility of the
results and the stability and efficiency of the chromatographic systems make this approach’ particularly attractive
for use in clinical analysis.
Keyphrases: separation of acids and bases in a
AdditIonal
single run
pounds
uents
#{149}
#{149}
catecho!amines and metabolically
reverse-phase
chromatography
related comwith aqueous el-
Nonpolar bonded phases with hydro-organic eluents
have been increasingly used in high-performance liquid
chromatography for the separation of substances that
are sparingly soluble or insoluble in water. Because the
eluent is more polar than the stationary phase, the term
“reverse-phase chromatography,”
originally suggested
by Howard and Martin (1), is used to describe
this
technique.
In HPLC’
the
stationary
phase
is most
commonly a finely divided organo-silica with octadecyl
moieties covalently
bound to the solid surface,
and
aqueous methanol or acetonitrile are used as the mobile
phase. The popularity
of the method rests with its relative simplicity and reproducibility
(2), which are also
responsible for the recent application of reverse-phase
chromatography
in the analysis
of mixtures
of various
biological substances.
Biochemical
Engineering
Group,
Applied
Science, Mason Laboratory,
Department
of Engineering
and
Yale University,
New Haven,
Conn. 06520.
1 Nonstandard
chromatography;
phenylacetic
abbreviations
used: HPLC,
HVA, homovanillic
acid
acid);
VMA,
vanilmandelic
methoxymandelic
Received
April
high-performance
liquid
(4-hydroxy.3.methoxy.
acid
(4-hydroxy-3.
acid).
7, 1976; accepted
May 11, 1976.
In thisstudy we demonstrate by the results obtained
with the separation of catecholamines and their metabolitesthat octadecyl-silica
columns can also be used
successfullywith neat (i.e.,
not admixed with organic
solvent) aqueous eluents in HPLC
and that reversephase chromatography issuitableforthe separationand
analysis
of polar substances as well.This approach
represents an alternativenot only to the usual chromatographic techniques, in which a polar sorbent and
a lesspolar eluent are used, but also to ion-exchange
chromatography, which is most widely used to separate
polar biologicalsubstances.
The need for rapid analysis of catecholamines and
theirderivativesarisesfrom their important biological
role (3, 4). Moreover, the clinical significance of these
substances is increasingly recognized and a suitable
chromatographic technique could serve as a powerful
tool for clinical analysis to provide valuable diagnostic
information. At present, fluorescent analysis (5) is used
most commonly to estimate certain urinary catecholamines, and paper chromatography is used to determine
HVA and VMA (6). Ion-exchange chromatography has
been frequently used for such separations (7-9), but this
technique and other column-chromatographic
methods
have not been used routinely in clinical laboratories.
The main impediment in using column chromatography
for this is the very low physiological concentration of
these substances, which precludes their analysiswith
the conventional ultraviolet or fluorometric detectors.
However, the electrochemical detector recently introduced by Kissinger et al. (10, 11) offers a highly sensitive
way to detect catechol compounds and appears to be
eminently suitable for their analysis. Thus, further
progress requires the development of fast, reliable, and
simple separation methods.
Potentially, HPLC offers the necessary speed and
efficiency. It does not involve the formation of volatile
derivatives from polar substances such as the catecholamines and their metabolites, as is required in gas
chromatography
(12, 13). Nevertheless, the chromatographic systems used so far have not been fully satCLINICAL CHEMISTRY,
Vol. 22, No. 9, 1976
1497
Cation- or anion-exchange
only to the separation
isfactory.
chromatography
is applicable
of basic or acidic
components,
respectively.
The recently introduced ion-pair chromatography
(14) shows the same
discriminatory effect, because both techniques strongly
involve coulombic interactions in the chromatographic
process. Whereas the speed and efficiency of the latter
technique is superior as compared with conventional
ion-exchange columns, the inherent instability
of liquid-liquid
chromatographic
systems makes this approach less desirable in routine analytical work.
Here, we demonstrate that reverse-phase chromatography, which involves hydrophobic
rather than
coulombic or hydrogen bonding interactions between
the solutes and the stationary phase, allows one rapidly
to separate mixtures of biogenic amines and their acidic
metabolites in a single chromatographic run. In view of
the previously mentioned advantages of this technique
our approach may be useful for routine analysis of other
low-molecular-weight
biological substances.
sample
Type 316 stainless steel used in the instruments
and
for the column is not sufficiently
resistant to halides.
the instrument
Therefore
7.0.
Samples
All catecholamines
tained from the clinical
Materials and Methods
Chromatograph
sample
We used a Model 601 liquid chromatograph
with a
Model LC 55 detector (Perkin-Elmer Corp., Norwalk,
Conn. 06856) and a Rheodyne Model 7105 sampling
valve (Rheodyne, Berkeley, Calif. 94710). The instru-
components,
in the isocratic
was operated
mode and the ad-
sorbance of the eluent was monitored
use of a Perkin-Elmer
Chromatographic
Model
at 254 nm, with
56 strip-chart
recorder.
conditions are in the figure legends.
Columns
Most experiments
were done with Partisil-1025
ODS
(Whatman, Clifton, N. J. 07014), 0.46 cm (i.d.)
packed with octadecyl-silica particles of 10-tm
av diameter. We observed slight variations in the efficiency and retention properties of columns from different lots. Octadecyl-silica stationary
phases have been
prepared
(15) from 10- and 5-tm Partisil
(Whatman)
columns
X 25 cm,
by reacting
(Aldrich
the silica gel with trichloro-octadecyl
Chemical
Co., Milwaukee,
silane
Wis. 53233) without
subsequent
treatment
with
trimethyl
chlorosilane.
Columns
of the above dimensions
have been packed by
the isopycnic slurry technique
at 152 MN/m2
(1500
atm). The chromatographic properties of the columns
with the lO-Mm particles were similar to those of the
commercial
columns,
5-izm octadecyl
tion conditions.
silica
whereas
retention
were greater
values
under
with
the
the same elu-
Eluents
We used phosphate
solutions in all experiments.
Reagent-grade
phosphoric acid, KH2PO4,
and
Na2HPO4
were purchased
from Fisher
Scientific
Co.,
Pittsburgh, Pa. 15219. In certain experiments the ionic
strength of the eluent was adjusted with KC1 (Fisher,
reagent grade). In studies where the pH of the eluent
was varied, phosphate concentration was maintained
at 0.1 mol/liter over the pH range investigated.
1498
CLINICAL CHEMISTRY, Vol. 22, No. 9, 1976
and their derivativeswere pur-
chased from Aldrich Chemical Co. and the samples were
prepared by mixing stock solutions of the individual
components in distilled water. Urine samples were obHospital.
In most
ment
was exposed to KC1 solutions
only for the duration of the experiments and the system
was purged with distilled water thereafter.
The siliceous
column material cannot withstand
eluents that have a
pH exceeding 8 without degradation.
It appears that a
high concentration
of halides in the eluent also shortens
the useful life of the column material if it is long exposed
to such concentrations.
On the other hand, both the
stainless steel and the column material are remarkably
resistant to phosphate solutions in the pH range 1.9 to
chromatographic
solution,
Evaluation
laboratory
containing
experiments,
of
of the Chromatograms
k’, which
factor,
chromatogram
by using
-
20 il
2 to 10 izg of the individual
The retention of the individual
(tR
Haven
was injected.
the retardation
=
of Yale-New
peaks is expressed
was evaluated
from
by
the
the following
relationship:
k’
t0 and tR are the retention
time
solute and the solute in question,
t0)/t0, where
for an unretarded
respectively
(16).
Results
The chromatograms
trate
the separation
depicted in Figures 1 to 4 illusof substances
that
are involved
in
certain metabolic
pathways of neurotransmitters
(4)
and consequently
include both acidic and basic compounds. Figure 1 shows the species that occur in the
degradation
of phenylalanine
to norepinephrine.
Because catechol-O-methyltransferase
(EC 2.1.1.6) does
not act in this pathway, the phenolic hydroxyls of the
intermediates
are not methylated.
Thus, the order of
elution reflects the increasing polarity of the substances
produced successively in the metabolic process. Figure
2 shows the intermediates
involved in the metabolic
degradation
of dopamine
to HVA and homovanillic
alcohol. In this case the mixture is composed of both
catechol and 3-O-methyl-catechol
compounds that have
either acidic or basic functions,
or both. The chromatogram in Figure 3 shows a similar
picture for the
metabolic pathway from dopamine to VMA.
Because catechols can readily be removed from their
3-0-methyl
or 4-0-methyl
derivatives by adsorption on
alumina
(17), there is a considerable
interest in the
separation of the acidic and basic methylated
substances.
Figure
4 illustrates
that reverse-phase
chromatography is eminently suitable for such a separation,
which cannot be easily effected by ion-exchange chromatography.
Figure 5 demonstrates
that the tradi-
0.050
1oH
E
E
C
5)
C”
LU
0
‘I)
LU
0
z
2
0
0
C,)
U)
0
5
10
5
5
MINUTES
Fig. 1. Chromatogram of intermediates of L-phenylalanine metabolism to norepinephrine
Column, Partisil 1025 ODS; eluent, 0.2 mol/liter phosphoric acid-potassium
phosphate,pH, 1.9; flow rate, 0.5 mI/mm; inlet pressure, 5.05 MN/rn2 (50 atm);
temperature, 25 #{176}C
0
MINUTES
5
20
Fig. 3. Chromatogram of the intermediates of dopamine metabolism to vanilmandelic
acid
Column, Partisil 1025 ODS; eluent, 50 mmol/llter phosphoric acId-potassIum
phosphate; pH, 2.0; flow rate, 0.5 mI/mm; inlet pressure, 5.05 MN/rn2 (50 atm);
temperature, 23 #{176}C
a
5,
C”
E
LU
C
U,
m
w
0
z
0
U)
‘2
0
C’,
0
5
10
5
20
MINUTES
Fig. 4. Chromatogram
amine
of 3-0-methyl
metabolites
of dop-
Column, Partisil 1025 COS; eluent, 50 mmol/llter potassium phosphate; pH, 4.6;
flow rate, 0.66 mI/mm; inlet pressure 3.54 MN/rn2 (35 atm); temperature, 25
L
0
5
0
MI
5
OC
NU1#{128}
S
Fig. 2. Chromatogram
of the intermediates
of dopamine
metabolism to homovanillic acid
Column, PartIsil 1025 OOS; eluent, 50 mrnol/Ilter potassium phosphate;pH, 4.6;
flow rate, 2.0 mI/mm; inlet pressure 20.2 MN/rn2 (200 atm); temperature, 22
OC
tionally difficult separation of the physiologically
most
important
catecholamines
also is readily done by reverse-phase chromatography,
and that this technique
offers an efficient alternative
to cation-exchange
chromatography. This chromatogram
was obtained with use
of the 5-gm octadecyl-silica
column, which relatively
strongly retarded the sample components
even at pH
4.4, where they are dissociated. Figures 6 and 7 illustrate
the effect of pH on the chromatographic
retention of the
substances of interest. The pH of the eluent was varied,
but the ionic strength
did not change significantly.
Retention of the acidic compounds decreases and that
of the basic substances increases with increasing pH of
the eluent. Obviously the retention decreases with increasing dissociation
of the ionogenic functions of the
molecules. It is seen that the O-methylated
derivatives
are always more strongly retarded than are the correCLINICALCHEMISTRY,Vol. 22, No. 9, 1976
1499
a
C
5,
C”
a:
LU
0
0
I0
z
‘2
‘2
m
a:
4:
0
IL2
0
I-.
NH
C,)
‘2
‘2
0
a:
‘2
ILU
a:
0
5
MINUTES
10
Fig. 5. Chromatogram of biogenic amines
Column. 4.6mm id., 25cm long, packed with 5i octadecyl-silica; eluent, 0.1
rnol/liter potassium phosphate; pH, 4.4; flow rate. 0.5 mI/mm; Inlet pressure.
12.12 MN/rn2 (120 atm) temperature, 25 #{176}C
pH
Fig. 7. Effect of pH on the retention of biogenic amines
Conditions as in Figure 6. E, epmnephrmne;MDA, 3..O-methyl-dopamine;
metanephrmne; NE, noreplnephrmne; PN, paranephrmne; TA, tyramine
MN,
IIA
a:
0
I0
‘2
MDA
2
0
0 __-OPAC
I.-
‘2
0
a:
‘2
ILU
.0 5
a:
iii:ii;iii:iiii:iiiiiiiiii:iiii:iiii:iii:
TYC
Dopa
.11
0.5
Fig. 6. Effect of pH on the retention of acidic metabolites
catechol compounds. The retardation
factor
is evidently a measure of the relative hydrophobicity
of
the solutesas revealed by reverse-phase chromatography. The results have been used to select the optimum
pH of the eluent for the separation of a mixture containing known components.
sponding
on retention
also has been
It is seen in Figure 8 that retardation
in-
creases with increasing salt concentration, in contrast
to the effect of salt concentration in ion-exchange
1500
CLINICALCHEMISTRY,Vol. 22, No. 9, 1976
IN THE
2.0
ELUENT
The concenfration of KCI in a 50 rnmol/llter KH2P04 solution was varied. Column,
Partisil 1025 ODS; flow rate, 1 mI/mm; inlet pressure, 10.1 MN/rn2 (100 atm);
temperature, 25 #{176}C.
TYR, tyrosine; see Figures 6 and 7 for other symbols
Column, Partisil 1025 ODS; eluent, 0.1 mol/Ilter phosphate buffers; flow rate.
1 mI/mm; inlet pressure, 10.1 MN/rn2 (100 atm); temperature. 25#{176}C.
DOPAC,
3,4-dihydroxy-phenylacetic
acid; DOMA. 3,4-dihydroxy-mandelic acid; HVA.
homovanillic acid; VMA, vanilmandelic acid
The effect of ionic strength
[mol/Il
Fig.8.Effectof ionicstrengthon theretardation
factor
5
pH
investigated.
.5
1.0
KCI
chromatography;
this proves that ionic interactions
between the solutes and stationary phase are negligible.
On the other hand, the increase in values for the retardation factors with increasing salt concentration-an
increase of about the same magnitude for all solutes
investigated-implies
that the retention is due to hydrophobic interaction.
Van’t Hoff plots for different solutes are depicted in
Figure 9. The slopes of the straight lines are very similar
and it can be seen that the k’ values are approximately
halved when the temperature is increased by 30 to 40
OC.
MOPET
0.5
HVA
0
S
0
0’
0
-J
MOA
w
MET
m
U
z
4
0
Co
EETA
.5
-0.5
VMA
-
3.0
3.1
3.2
103/T
3.3
3.4
Column, Partisll 1025ODS;eluent, 50 rnrnol/liter KH2PO4; flow rate, 1 mI/mm.
DA, doparnine; MET, metanephrine; MOPET, homovanillic alcohol; for other
symbols see preceding Figures
the potential
of this method for urine
analysis,
we chromatographed
several acidified
urine
samples. As expected, the chromatograms
obtained with
either the native urine specimen or an ethyl acetate
extract of it show that many components
that absorb
light at 254 nm are present.One such chromatogram is
shown in Figure 10. The results indicate that reverse-
phase chromatography
is of potential
separation of the low-molecular-weight
cules in urine, but further
refinement
]
0
20
MINUTES
30
Fig. 10. Chromatogram of an acidified urine sample
(1<’)
Fig. 9. Effect of temperature on the retardation factor
To explore
0
use for rapid
organic moleof the technique
is necessary. Our interest is focused on the analysis of
urinary VMA and HVA because preliminary
studies
indicate
that reverse-phase
chromatography
could
represent an appropriate
alternative
to the presently
used paper chromatography.
Discussion
Relatively polar ionogenic substances can be separated by reverse-phase chromatography with neat
aqueous eluents, and the peaks do not show excessive
tailing even when the solutes are partially dissociated
in the mobile phase. Because the separation is the result
of hydrophobic interactions between the hydrocarbonaceous stationary phase and the hydrophobic
moiety
of the solute molecules, mixtures of acidic and basic
compounds can be easily separated in a single chromatographic run when the pH of the eluent is appropriate.
The pH dependence of the retention values shown in
Figures 6 and 7 can serve as a guide to select the optimum eluant pH for the isocratic separation of a particular mixture. As expected, retention values decrease
with increasing
dissociation
of the solutes, which attenuates
the magnitude
of their hydrophobic
interactions with the stationary
phase. If the sample consists
only of acidic components
the eluant pH should be low,
to suppress dissociation
and obtain stronger retardation
and better separation.
The opposite is true for the
separation
of amino compounds.
Retention
also can be
Column, Partisil 1025ODS;eluent.0.1 mol/llter phosphate buffer, pH 2.05, flow
rate, 1 mI/mm; inlet pressure. 10.1 MN/rn2 (100 atm); temperature, 25 #{176}C;
sample size, 10
detector sensItivity, 0.1 AUFS
prolonged by increasing the ionic strength of the eluent,
as shown in Figure 8, but this effect is not nearly as
dramatic as that of pH change and its magnitude is
similar for such closely related substances, because the
slopes of the curves almost parallel
one another.
Therefore,
the values for relative
retention
do not
change significantly
with ionic strength. The effect of
temperature
is such that the retention
values rapidly
decrease with increasing
temperature,
and again a
comparison of the slopes obtained for the various substances indicates that relative retention values are not
affected significantly
by temperature.
Most results presented here have been obtained with
commercially
available reverse-phase
columns. It is
known, however, that the chromatographic
behavior of
such columns depends on the structure and particle size
of the silica gel and the amount of covalently
bound
octadecyl moiety as well as on the method of its attachment.
Consequently,
different
results can be expected when other column materials are used. Nevertheless, columns packed with ocadecyl-silicas
prepared
in our laboratory showed retention properties not much
different
from those of the commercial
columns.
We expected that the efficiency of the column would
increase when the particle size of the stationary
phase
was reduced to 5 tm. As shown in Figure 5, however,
columns containing such a material are not significantly
more efficient; this result is attributed to the inadequate
packing structure. Further work is required, therefore,
to obtain a two- to threefold
increase in column efficiency by using 5-m instead of 10-am particles and
improved
column-packing
technology,
which is currently under study in our laboratory.
The retardation
factors we obtained with our 5-izm particles were significantly
higher than those attained with the 10-am
particles under identical conditions. Consequently,
the
early peaks were better separated.
It has been shown recently (18) that reverse-phase
chromatography
can also be used as “reverse ion-exchange chromatography,”
and this technique is termed
CLINICAL CHEMISTRY, Vol. 22, No. 9, 1976
1501
“paired-ion
nol/water
chromatography.”
mixtures
In this method metha-
containing
strong long-chain
aliphatic acids or bases are used as the eluent. This technique also represents a significant
broadening
of the
scope of reverse-phase
chromatography,
but like ionexchange
chromatography
it still can only be used efficiently
to separate acidic or basic components
in a
single chromatographic run. In addition,
the species
that interact with the stationary
phase are ion pairs
formed from the sample components
and the acid or
base in the mobile phase. Because the heptane sulfonate
or tributylammonium
ions, which are frequently
used
in this method, are of relativelyhigh molecular weight,
the selectivity
of the system for the separation,
of ion
pairs with small molecules is expected
to be inferior to
the direct method presented here, even when only acidic
or basic sample components
are analyzed.
In the conventional
mode, reverse-phase
chromatography
with aqueous methanol or acetonitrile
as eluent has been extensively
used to separate
nonpolar
biological
substances.
Our results show that chromatographic systems that involve a hydrophobic
stationary
phase have a much greater versatility
than envisoned
previously.
In addition to the conventional
technique
and the recently
introduced
“reverse ion-exchange”
chromatography,
the present
method can be used to
separate
biological
substances
directly
with neat
aqueous eluents. There are great practical advantages
to using a nonswelling stationary phase, which yields
stable columns, and aqueous eluents, which are readily
available with high purity and are not poisonous
or
flammable.
A major drawback of the siliceous column
materials now in use is that they are not resistantto
eluents having pH values higher than 7. Therefore
development
of hydrophobic
phases on a support such as
zirconia, which resists higher pH, is required to extend
the operating
range of the chromatographic
system.
We think that with improved column technology the
efficiency
can be significantly
increased
when column
materials
having a particle size range of 4 to 5 tm can
be used. Then short columns could yield rapid separations, and the sensitivity
of analysis could be enhanced
to meet the demanding requirements encountered in
clinical
analysis.
1502
CLINICAL CHEMISTRY, Vol. 22, No. 9, 1976
This work was supported
USPHS. We thank Dr. P.
by a grant GM 20993 from the NIH,
for valuable
discussions.
I. Jatlow
References
1. Howard, G. A., and Martin,
fatty acids by reversed-phase
J. 46, 532 (1950).
A. J. P., The separation
of the C12-C18
partition
chromatography.
Biochem.
2. Snyder, L. R., and Kirkland,
J. J., Introduction
to Modern
Chromatography.
Wiley, New York, N. Y., 1974, p 391.
3. Iversen,
of Biogenic
Liquid
L. L., Iversen, S. D., and Snyder, S. H., Eds. Biochemistry
Amines.
Plenum Press, New York, N. Y., 1975.
4. Seiler, N., Demisch, L., and Schneider,
H., Biochemie
und Funktion von biogenen Aminen im Zentralnervensystem.
Angew. Chem.
83,52 (1971).
5. Victoria, J. K., Baukal, A., and Wolff, F. W., New automated
orometric
methods for estimation
of small amounts of adrenaline
noradrenaline.
Anal. Biochem.
23, 513 (1968).
6. Hingerty,
D., and O’Boyle,
Medulla.
Charles C Thomas,
A., Clinical
Springfield,
fluand
Chemistry
of the Adrenal
Ill., 1972, pp 56-59.
7. Shibuya,
H., Watanabe,
S., and Toru, M., Separation
of homovanillic acid and vanillylmandelic
acid with ion-exchange
resin column.
Experientia,
31,623 (1975).
8. Haggendal,
J., Newer
developments
Pharmacol.
Rev. 18, 325 (1966).
of catecholamine
assay.
9.
Banda, P. W., Sherry, A. E., and Blois, M. S., An automatic
analyzer for the detection
of dihydroxyphenylalanine
metabolites
and
other reducing compounds
in urine. Anal. Chem. 46, 1772 (1974).
10. Kissinger, P. T., Felice, L. J., Riggin, R. M., et al., Electrochemical
detection
of selected
organic components
in the eluate from highperformance
liquid-chromatography.
Clin. Chem. 20, 992 (1974).
11. Kissinger, P. T., Riggin, R. M., Alcorn, R. L., and Rau, L. D., Estimation
of catecholamines
in urine by high performance
liquid
chromatography
with electrochemical
detection.
Biochem.
Med. 13,
299 (1975).
12. Bigdeli, M. G., and Collins, M. A., Tissue catecholamines
and
potential
tetrahydroisoquinoline
alkaloid
metabolites:
A gas chromatographic
assay method with electron capture detection.
Biochem.
Med. 12,55 (1975).
13. Cancalon,
P., and Klingman,
J. D., Simultaneous
determination
of biogenic and non-biogenic
amines by gas liquid chromatography.
J. Chromatogr. Sci. 10, 253 (1972).
14. Persson,
B. A., and Karger,
B. L., High performance
ion pair
partition
chromatography:
The separation
of biogenic amines and
their metabolites.
J. Chromatogr. Sci. 12, 521 (1974).
15. Majors, R. B., Techniques
for liquid chromatographic
columns
packed with small porous particles.
Anal. Chem. 45, 755 (1973).
16. Horvath,
C., High-performance
ion-exchange
chromatography
with narrow-bore
columns: Rapid analysis of nucleic acid constituents
at the subnanomole
level. Methods
Biochem.
Anal. 21, 82 (1973).
17. Hingerty, D., Thirty-minute
screening test for phaeochromocytoma. Lancet
i, 766 (1957).
18. Wittmer,
D. P., Nuessle, N. 0., and Haney, W. G., Jr., Simultaneous analysis of tartrazine
and its intermediates
by reversed phase
liquid chromatography.
Anal. Chem. 47, 1422 (1975).