Reexaminacion de melanina

Biochimica et Biophysica Acta 883 (1986) 155-161 155
Elsevier
BBA 22416
Reexamination of the structure of eumelanin
Shosuke Ito
School of Hygiene, Fujita-Gakuen Health University, Toyoake, Aichi 470-11 (Japan)
(Received March 7th, 1986)
Key words: Melanin; 3,4-Dihydroxyphenylalanine; Tyrosinase
The generally accepted concept that the black melanin eumelanin is made mostly from 5,6-dihydroxyindole
but not from 5,6-dihydroxyindole-2-carboxylic acid (DHIC) was reexamined by comparison of synthetic and
natural eumelanins. The analytical methods used were elemental analysis and determination of the carboxyl
group by acid treatment to yield CO 2 and by permanganate oxidation to yield pyrrole-2,3,5-tricarboxylic
acid. It was found that DHIC-derived monomer units comprise only approx. 10% of enzymically prepared
dopa-melanins but as much as a half of intact, natural eumelanins. The results also show that dopa-melanins
prepared at higher pH retain higher percentages of the carboxyl group of dopa and contain higher
percentages of pyrrole units, and that melanins are decomposed to a significant extent on acid treatment, the
method commonly used to isolate melanins from natural sources.
Introduction
The black melanin eumelanin is synthesized in
vivo from tyrosine. The classical Raper-Mason's
concept of melanin formation (Fig. 1) consists of:
(1) hydroxylation of tyrosine to dopa; (2) oxidation
of dopa to dopaquinone; (3) cyclization to
leucodopachrome; (4) oxidation to dopachrome
(by dopaquinone); (5) decarboxylation or rearrangement
to 5,6-dihydroxyindole (DHI) or
5,6-dihydroxyindole-2-carboxylic acid (DHIC);
and (6) oxidation to eumelanin [1-3]. It is generally
accepted that the first two steps are catalyzed
by the enzyme tyrosinase [4], and the following
steps proceed spontaneously, although factors regulating
these steps have recently been postulated
[5].
When a solution of dopachrome was kept
anaerobically at pH 5.6, the solution became col-
Abbreviations: DHIC, 5,6-dihydroxyindole-2-carboxylic acid;
DHI, 5,6-dihydroxyindole; PTCA, pyrrole-2,3,5-tricarboxylic
acid.
orless with the ultraviolet spectrum corresponding
to that of DHI [2]. Furthermore, the reaction was
reported to be accelerated by Zn 2+ [6]. On the
other hand, DHIC was formed as a product of
rearrangement of dopachrome at pH 1.3 [2]. These
observations were the basis for postulating DHI,
not DHIC, as the major, ultimate precursor of
eumelanin. However, the following findings which
accumulated in recent years suggest the significance
of DHIC as an alternative precursor of
eumelanin: (1) elevated levels of the methoxy derivatives
of DHIC were detected in urine of genetically
dark people [7] and of melanoma patients
[8]; (2) Zn 2+ catalyzes the rearrangement of
dopachrome to DHIC (not to DHI as previously
reported [6]) [9], Zn 2 + is rich in melanosomes [10],
and the dopachrome conversion factor also catalyzes
the same rearrangement [11].
Nicolaus and his associates studied by various
chemical methods the structure of melanin isolated
from sepia ink by acid treatment [12], and
Swan's group assessed different types of units
present in dopa-melanin [13]. The results of these
0304-4165/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

156
HO
Tyrosinose
/f'f \ \
/ \
/
i ~ COOH , H O ~ COOH \
• ~ ~,.,.C/ NH2 02 HO -r"-.,.~/ NH2 02
Tyrosi ne
Dopa
O ~ COOH
0"~'~
Dopaquinone
NH~
[ Eumelanin] ~
(o) H O ~
HO
(COOH)
H
5,6-Dihydroxyindole
(-2-carboxylic acid)
H
keucodopachrome
~ (o)
~~COOH
H
Dopachrome
Fig. 1. Raper-Mason's pathway of melanin formation from tyrosine by tyrosinase.
studies indicate that melanins are highly heterogeneous
polymers consisting of various monomer
units. These units include DH| unit (1), DHIC
unit (II), pyrrole unit (III), and pyrrole-carboxylic
acid unit (IV) (Fig. 2); the latter two units are
derived from the former two units by peroxidative
cleavage of o-quinone form.
HO
HO
NO%
The purpose of the present study was to estimate
the degree of incorporation of the DHIC-derived
units II and IV in intact natural eumelanins.
The analytical methods used were elemental analysis,
acid treatment to liberate CO 2, and permanganate
oxidation to form pyrrole-2,3,5-tricarboxylic
acid (PTCA) [14] (Fig. 3). The structural
differences in eumelanins synthesized from
various precursors under various conditions were
also compared.
COOH
Materials and Methods
t
I
oc
II
Materials
L-Dopa, dopamine, tyrosinase (2000 units/mg)
and catalase (40 000 units/rag) were purchsed from
Sigma Chemical Co. (St. Louis, MO). DHIC was
; H
III
OC
; H
IV
COOH
Fig. 2. Various monomer units present in eumelanins. I: DHI
unit, C/N ratio = 8; II: DHIC unit, C/N ratio = 9; Ill; pyrrole
unit, C/N ratio = 6; IV: pyrrolecarboxylic acid unit, C/N
ratio = 7. Units I and I! may be present in both reduced,
o-diphenolic form and oxidized, o-quinone form. One of the
two carbonyl groups in Ill and IV becomes a carboxyl group
when these units are located at the terminal of a polymer.
HO~H
H+7 CO 2 + Melanin
Jheat
NO COON KMn~04 NOOC~
DHIC-melanin
HOOC H
COOH
PTCA
Fig. 3. Methods for the analysis of DHIC-derived units in
eumelanins.

157
prepared by ferricyanide oxidation of dopa [14].
DHI was obtained by double sublimation of DHIC
under high vacuum [15]. The first sublimation was
done at 240°C and the second at 200°C. The
yield was approx. 10%. HPLC analysis revealed
that the crystals thus obtained were contaminated
by 9% of DHIC. PTCA was prepared by permanganate
oxidation of 5-hydroxyindole-2-carboxylic
acid (Aldrich Chemical Co., Milwaukee,
WI) [14]. Other chemicals were of analyticalgrade
from Wako Pure Chemicals (Osaka, Japan).
The melanosomes were prepared from the ink
sacs of sepia (cuttlefish) and from B16 mouse
melanomas and were kind gifts from Dr. G. Prota
(University of Naples) and Dr. K. Jimbow (Sapporo
Medical College), respectively. Elemental
analyses were performed in the Microanalytical
Laboratory, Faculty of Science, Osaka University.
Preparation of melanins
1 mmol of a precursor was dissolved in 80 ml
(12.5 mM) or 400 ml (2.5 mM) of a buffer. The
precursors used were L-dopa, L-tyrosine, dopamine,
DHI and DHIC. The buffers were 0.05 M
sodium phosphate (pH 6.8 and 8.0) and 0.05 M
sodium carbonate (pH 10.0). The mixture was
incubated at 37°C for 4 h in the presence of
tyrosinase (8 mg) under oxygen current or for 24 h
under air. Some reactions were carried out in the
presence of catalase (5 mg). After the incubation,
the mixture was acidified to pH 1 with 2 M HCI,
and the black precipitate was collected by centrifugation
and washed with 0.1 M HC1 (40 ml x 3).
The melanin was suspended in 0.1 M HC1 (100
ml) and kept in a refrigerator. To determine the
yield, a 10 ml aliquot was centrifuged, and the
melanin was dried over P205 and NaOH in a
desiccator. For the preparation of DHI-melanin,
the reaction scale was reduced to a half, but the
yield in Table I was doubled for the sake of
comparison.
Acid treatment of melanin and CO 2 determination
20 mg of a melanin or a related compound or
30-200 mg of a tissue sample was suspended in 20
ml 6 M HC1. The mixture was heated under reflux
with a constant bubbling of argon gas. The argon
gas that came out from the reaction mixture was
introduced into a test tube containing 5 ml of a
saturated Ba(OH)2 solution to trap the CO 2 gas
liberated from the sample. After heating for 24 h,
the remaining melanin was collected by centrifugation,
washed with 0.1 M HC1 (10 ml × 2) and
acetone (10 ml x 2; only for tissue samples), and
dried over P205 and naOH. The BaCO 3 formed
was collected by centrifugation, washed with water
(10 ml X 2), and dried over P205.
Results and Discussion
Preparation and elemental compositions of melanins
Melanins were prepared either by tyrosinase
oxidation or by air oxidation. Commercially available
mushroom tyrosinase was used as a tyrosinase,
although the preparation is not only impure, but
may have quite different substrate and enzymic
specificities than does animal tyrosinase. It is expected,
however, that eumelanins prepared by the
oxidation with tyrosinases from various sources
may not differ greatly, because the steps beyond
dopaquinone (Fig. 1) are spontaneous. Some preparations
were performed under two different precursor
concentrations and also in the presence of
catalase to suppress peroxidative cleavage of o-
quinone form. Table I summarizes the yields and
elemental compositions of the synthetic melanins.
For comparison, the elemental compositions of
natural melanins are also included.
The yields of melanins were higher when the
oxidation was performed at a higher substrate
concentration and in the presence of catalase.
Oxidation of the ultimate precursors of eumelanin,
DHI and DHIC afforded higher yields of
melanins.
The elemental composition of a melanin greatly
varies depending on the degree of hydration.
However, the C/N ratio was highly reproducible;
three separate analyses of acid-treated DHICmelanin
gave the C/N ratios of 8.31, 8.32, and
8.27. The C/N ratios of the synthetic dopamelanins
indicate that from one to nearly two
carbon atoms per monomer unit were lost in the
course of oxidation, and the carbon loss was more
pronounced at higher pH. Thus, extensive decomposition
of the indole ring occurred at pH 10.0,
most likely by peroxidative cleavage. The concentration
of dopa and the addition of catalase
had little effect on the C/N ratio. With respect to

158
TABLE I
PREPARATION AND ELEMENTAL COMPOSITIONS OF SYNTHETIC AND NATURAl.. MELANINS
A precursor (1 mmol) was oxidized either by tyrosinase at pH 6.8 or by' air in the presence or absence of calalase.
No. Melanin Yield Elemental composition " C/N
(mg) C H N S ratio
Monomer
Mr ~'
7
8
9
10
11
12
13
Synthetic melanin
dopa-melanin, 12.5 mM, tyrosinase
dopa-melanin, 2.5 mM, tyrosinase
dopa-melanin, 2.5 mM, tyrosinase + catalase
dopa-melanin, 12.5 raM, pH 8.0
dopa-melanin, 2.5 raM, pH 8.0
dopa-melanin, 12.5 raM, pH 10.0
dopa-melanin, 2.5 mM, pH 10.0
dopa-melanin, 2.5 mM, pH 10.0+catalase
tyrosine-melanin, 2.5 mM, tyrosinase
dopamine-melanin, 12.5 mM, tyrosinase
dopamine-melanin, 2.5 mM, tyrosinase
DHl-melanin, 12.5 mM, tyrosinase ,1
DHIC-melanin, 12.5 mM, tyrosinase ~
149 51.55 3.79 7.51 0.09 8.00 186
- (50.38 3.81 7.03 8.36 199) ~
114 50.03 3.78 7.37 - 7.91 190
141 51.92 3.53 7.67 7.89 183
120 55.25 3.32 8.13 - 7.92 172
120 52.28 3.25 7.92 - 7.70 177
110 47.26 3.28 7.62 - 7.24 184
- (49.67 4.13 7.37 - 7.86 190) ~
68 46.16 3.49 7.46 - 7.21 188
146 50.33 3.05 8.25 0.03 7.12 170
89 49.50 4.01 7.45 - 7.75 188
75 51.25 4.50 7.93 0.11 7.54 177
(52.14 4.49 6.98 - 8.71 201) "
61 51.62 4.22 7.84 - 7.67 179
172 51.12 3.84 7.68 - 7.76 182
- (51.90 3.70 7.47 - 8.10 188) "
212 44.24 4.16 6.16 0.00 8.37 227
- (45.00 4.61 6.31 - 8.31 222) "
Natural melanin
14 from sepia melanosomes
15 from B16 melanosomes
16 from C57BL black mouse hair
(52.25 3.40 7.88 0.20 7.73 178) ~
(52.28 4.52 7.27 0.59 8.39 193) "
(56.49 4.57 7.60 0.83 8.66 184) ~
" Analyzed after equilibration with moisture present in the air and corrected for ash
b Calculated by assuming that every monomer unit contains one nitrogen atom.
" After acid treatment at ll0°C for 24 h.
d Prepared from DHI containing 9% of DHIC.
Prepared in the presence of 0.05 mmol t-dopa.
content.
the C/N ratio, enzymically prepared dopamelanins,
tyrosine-melanin, and dopaminemelanins
were rather similar to each other and
also to DHI-melanin. The C/N ratios of DHImelanin
and DHIC-melanin suggest significant
degrees of decomposition of the indole ring. The
fact that DHIC-melanin was decomposed to a
higher extent may be related to its solubility;
DHIC-melanin remained mostly in solution during
oxidation, while DHI-melanin rapidly precipitated
out.
The molecular weights calculated from the
nitrogen contents indicate that synthetic melanins
contain approx. 2 molecules of water per monomer
unit.
Table I also shows that acid treatment of
dopa-melanins, dopamine-melanin, and DHImelanin
resulted in the increase in the C/N ratios.
When the loss of carbon atom in the form of CO2
(Table II) was taken into account, one might
expect the opposite results. This discrepancy may
partly be ascribed to decomposition of the indole
or pyrrole ring to form an NH 3 molecule; acid
treatment of dopa-melanin (Table II, No. 1),
dopamine-melanin (No. 10), DHI-melanin, and
DHIC-melanin liberated NH 3 in yields of 3,1, 3.5,
2.0, and 3.3 mol% per monomer unit, respectively,
as determined by amino acid analysis.
Natural melanins prepared by acid treatment
(hydrolysis) of melanosomes and black mouse hair
had rather diverse elemental compositions (Table
I). This fact made it difficult to compare similarity

TABLE II
ACID TREATMENT AND KMnO 4 OXIDATION OF SYNTHETIC MELANINS, EUMELANIC TISSUES. AND RELATED
COMPOUNDS.
Materials No. 1-16 correspond to those in Table I. KMnO 4 oxidation was performed as described by Ito and Fujita [14].
No. Material Acid treatment PTCA (/x g/mg)
melanin (%) a CO 2 (mol%) intact treated c
Synthetic melanin
1 dopa-melanin, 12.5 mM, tyrosinase 82 21 1.7 0.8
2 dopa-melanin, 2.5 mM, tyrosinase 79 28 1.3 0.7
3 dopa-melanin, 2.5 mM, tyrosinase + catalase 82 27 1.3 0.4
4 dopa-melanin, 12.5 mM, pH 8.0 73 30 1.3 0.7
5 dopa-melanin, 2.5 mM, pH 8.0 82 40 2.1 1.0
6 dopa-melanin, 12.5 mM, pH 10.0 75 46 3.1 0.8
7 dopa-melanin, 2.5 mM, pH 10.0 83 54 7.7 1.0
8 dopa-melanin, 2.5 mM, pH 10.0 + catalase 78 36 4.4 1.8
9 tyrosine-melanin, 2.5 mM, tyrosinase 80 16 1.3 0.7
10 dopamine-melanin, 12.5 mM, tyrosinase 85 11 0.7 0.6
11 dopamine-melanin, 2.5 raM, tyrosinase 67 16 0.9 0.8
12 DHI-melanin, 12.5 mM, tyrosinase a 82 24 1.3 0.2
13 DHIC-melanin, 12.5 mM, tyrosinase a 77 90 30 2.1
Eumelanic tissue
14 sepia melanosomes a 55 62 e 10 e 4.4
15 B16 melanosomes d 23 56 ¢ 8.5 e 2.4
16 C57BL black mouse hair a 6.9 65 ~.r 12 e 3.8
17
Related compounds
DHIC a
106 10
18 5-hydroxyindole-2-carboxylic acid a
98 32
19 pyrrole-2-carboxylic acid a
116 -
a Percent recovery of melanin after acid treatment (110°C, 24 h).
b Mol percent of CO 2 liberated per tool of monomer unit (M r in Table I), corrected for the formation of 2.1 mg of BaCO 3 in blank
experiments.
c Melanin obtained by acid treatment.
a Average of two determinations.
e Calculated by assuming that the recovery of melanin is 80% and the M r of monomer unit is 205 (averages for DHl-melanin and
DHIC-melanin).
f The yield of CO 2 was 2.8% (w/w) and was corrected for 1.6% (w/w) yield of CO 2 from A/J albino mouse hair.
159
(or diversity) between natural and synthetic
melanins based on the elemental composition. It is
unknown whether the natural melanins are in fact
quite different from each other in structure or
whether they are decomposed
during acid treatment.
in different ways
Acid treatment and permanganate oxidation of
melanins
The content of carboxyl group attached to an
indole or pyrrole ring can be determined by ther-
mal decarboxylation [12,13]. However, the method
appeared not to be directly applicable to tissue
samples. Thus, we examined acid treatment as an
alternative method for decarboxylation. When
melanins were heated in 6 M HC1 under reflux,
evolution of CO 2 occurred rapidly. Although most
of the CO 2 evolution was completed within a few
hours, the reaction time was fixed to 24 h so that
the method
could be applied to tissue samples.
The carboxyl group content may also be assessed
by permanganate oxidation to yield PTCA [14]
(Fig. 3). Table II summarizes the results of acid
tratment and permanganate oxidation of synthetic
melanins, eumelanic tissues, and related compounds.
The recoveries of melanins after acid treatment
of synthetic melanins were approx. 80% on the

160
average. The loss of weight should be ascribed to
degradative solubilization as well as to CO 2 liberation.
The 55% recovery of melanin from sepia
melanosomes corresponds to a 70% content of
melanin. The melanosomes contain 9% protein (by
amino acid analysis) and 4% ash (by elemental
analysis). The remaining 17% may be accounted
for by lipids, carbohydrates, and other tissue components
as well as water. Thus, the sepia melanosome
preparation appears to be fairly pure as a
eumelanin and should be considered as a good
model for the intact, natural eumelanin.
90% of the monomer units of DHIC-melanin
lost CO 2 on acid treatment. Furthermore, DHIC,
5-hydroxyindole-2-carboxylic acid and pyrrole-2-
carboxylic acid liberated amounts of CO 2 that
were equal to or more than the theoretical value.
Also, the method was fairly reproducible; differences
in two determinations were within 10%.
These results indicate that the amount of CO 2
liberated by acid treatment reflects well the content
of the carboxyl group attached to the indole
or pyrrole ring in melanin.
The CO 2 loss from dopa-melanins increased
with pH of the buffer, indicating that oxidation at
higher pH favors the retention of the C-2 carboxyl
group or the cleavage of the indole ring to form
the carboxyl group. Approx. 10% of the monomer
units of enzymically prepared dopa-melanins were
shown to liberate CO 2, when corrected for the
CO 2 liberation from dopamine-melanins whose
precursor dopamine lacks a carboxyl group. This
fact suggests that approx. 10% of the carboxyl
group of dopa was incorporated into dopamelanins
in the forms of the monomer units |i
and IV. Previous studies using carboxy-labeled
dopa revealed 5-20% incorporation of the carboxyl
group into melanins [4]. Tyrosine-melanin appeared
to be similar to dopamine-melanin with
respect to carboxyl content. This result was in
agreement with the previous finding that
tyrosine-melanin was more completely decarboxylated
than dopa-melanin [4].
Permanganate oxidation of DHIC-melanin gave
PTCA in a yield of 30 /~g/mg of melanin. Although
the yield was rather low, the monomers
DHIC and 5-hydroxyindole-2-carboxylic acid also
gave low yields (10 and 32 ~g/mg) of PTCA. The
monomer units that may give rise to PTCA are the
units !I and IV. The units ! and I11 with connection
at the C-2 position could also give rise to
PTCA. However, such a contribution should be
negligible, as acid-treated dopa-melanins as well
as dopamine-melanins gave only trace amounts of
PTCA. The yields of PTCA from enzymically
prepared dopa-melanins were less than 10% that
from DHIC-melanin, indicating that these
melanins consist mostly of the units I and III. The
results were in agreement with those obtained by
elemental analysis and CO2 determination. The
relatively high yield (7.7/~g/mg) of PTCA from
dopa-melanin prepared at pH 10.0 suggests some
retention of the carboxyl group of dopa. Furthermore,
this melanin showed the C/N ratio of 7.21
and liberated 54% of CO 2 per monomer unit.
From these results, it appears that the pyrrolecarboxylic
acid unit IV comprises a significant
fraction of this melanin.
Acid-treated melanins, either synthetic or natural,
gave much lower yields of PTCA than the
corresponding native melanins. This can be
ascribed to the loss of the C-2 carboxyl group in
the units II and IV. Natural melanins after acid
treatment gave relatively higher yields of PTCA.
The reason for this is not clear at present.
Acid treatment (hydrolysis) of melanosome
preparations and black mouse hair yielded approx.
60% of CO 2 per monomer unit of melanin.
The yields were intermediate values between those
from DHI-melanin and DHIC-melanin. Although
tissue components may also liberate CO 2 on acid
treatment, such a contribution should be minimal;
bovine serum albumin yielded only 0.4% (w/w) of
CO 2 on acid hydrolysis. Furthermore, the CO 2
value for the black mouse hair was corrected for
the tissue background (albino mouse hair). The
yields of PTCA from these eumelanic tissues were
30-40% of that from DHIC-melanin. Our previous
study showed that tissue components other
than eumelanin do not give rise to PTCA by
permanganate oxidation [14]. From these results,
it is concluded that DHIC-derived units !I and IV
may comprise one-third to a half of the monomer
units of intact, natural eumelanins.
Conclusions
The present study shows that natural
eumelanins are not homopolymers of DHI, but

161
rather copolymers of DHI and DHIC in various
ratios. It appears that DHIC could play a more
important role in the biosynthesis of eumelanins
than previously believed. In this connection, it
should be noted that oligomers of DHIC have
been found in the eye of catfish [16].
It is also shown that: (1) enzymically prepared
dopa-melanins are quite different from natural
eumelanins in terms of the content of carboxyl
group; (2) dopa-melanins prepared at higher pH
retain higher percentages of the carboxyl group of
dopa and contain higher percentages of pyrrole
units; and (3) melanins are decomposed to a significant
extent on acid treatment. Thus, the previous
results obtained for dopa-melanin by Swan
[13] and for sepia melanin by Nicolaus [12] cannot
be applied to the structure of intact, natural
eumelanins.
Eumelanins have the properties of polyanions
and may, in vivo and in vitro, bind with various
cations such as di- and trivalent metal ions [17],
polyamines [18], and anionic drugs [191. The binding
in vivo may be favored by the high carboxyl
group content in natural eumelanins.
Acknowledgements
This study was supported in part by Grant-in-
Aid No. 58390016 from the Ministry of Education,
Science and Culture. The author is grateful
to Professor Giuseppe Prota of the University of
Naples for reading the manuscript.
References
1 Raper, H.S. (1928) Physiol. Rev. 8, 245-258
2 Mason, H.S. (1948) J. Biol. Chem. 172, 83-99
3 Thompson, A., Land, E.J., Chedekel, M.R., Subbarao, K.V.
and Truscott, T.G. (1985) Biochim. Biophys. Acta 843,
49-57
4 Hearing, V.j., Ekel, T.M., Montague, P.M. and Nicholson,
J.M. (1980) Biochim. Biophys. Acta 611,251-268
5 Pawelek, J., Kornor, A.M., Bergstrom, A. and Bologna, J.
(1980) Nature 286, 617-619
6 Harley-Mason, J. and Bu'Lock, J.D. (1950) Nature 166,
1036-1037
7 Wirestrand, L.-E., Hansson, C., Rosengren, E. and Rorsman,
H. (1985) Acta Dermatovenereol. (Stockh.) 65,
345-347
8 Pavel, S. and Muskiert, F.A.J. (1983) Cancer Detect. Prev.
6, 311-316
9 Napolitano, A., Chioccara, F. and Prota, G. (1985) Gazz.
Chim. Ital. 115, 357-359
10 Hor6i6ko, J., Borovansk~,, J., Duchoh, J. and Prochazkova,
B. (1973) Hoppe-Seyler's Z. Physiol. Chem. 354, 203-204
11 Korner, A.M. and Gettins, P. (1985) J. Invest. Dermatol.
85, 229-231
12 Nicolaus, R.A. (1968) in Chemistry of Natural Products
(Lendered, E., ed.), Melanins, pp. 68-91, Harman, Paris
13 Swan, G.A. and Waggott, A. (1970) J. Chem. Soc. (C) 1970,
1409-1418
14 Ito, S. and Fujita, K. (1985) Anal. Biochem. 144, 527-536
15 Beer, R.J.S., Robertoson, L.M. and Woodier, A.B. (1949) J.
Chem. Soc. 2061-2066
16 Ito, S. and Nicol, J.A.C. (1974) Biochem. J. 143, 207-217
17 Sarna, T., Hyde, J.S. and Swartz, H.M. (1977) Science 192,
1132-1134
18 Tj~ilve, H., Nilsson, M. and Larsson, B. (1981) Acta Physiol.
Scand. 112, 209-214
19 Larsson, B. and Tj~ilve, H. (1979) Biochem. Pharmacol. 28,
1181-1187