Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
1~78951
Technical Field
This invention relates generally to chelating
agents, and more particularly to EDTA, ED3A and DTPA analogs
formed from amino acid precursors which may be attached to
biological molecules and which form physiologically stable
chelates with a variety of metal ions.
The invention described herein was made in the
course of, or under, a grant from the National Institutes of
Health.
10 Bac~ground Art
It is known that metal ions may be attached to
biological molecules by means of bifunctional chelating
agents. Such chelating agents are compounds which
incorporate a covalent bond-forming moiety, which may be
attached to a biological molecule, and a metal-binding
moiety which forms a chelate with metal ions.
In 1974 M.W. Sundberg, et al, demonstrated a
seven-step synthesîs yielding a bifunctional chelating
agent, p-nitrophenyl EDTA. J.Med.Chem., 17, 1304 ~1974).
20 Several years later, using the seven-step synthesis of
Sundberg, et al, such a para-substituted, l-phenyl-EDTA
compound was prepared, and chelates thereof were ormed with
lllindium. The radiolabelled chelates were demonstrated to
be stable in vivo and in vitro. C. Meares, et al, Proc.
Natl. Aca'd'.''S'ci'. U.S.A., 73, 3803 (1976). Such
`` 1~78~51
radiolabelled chelates may be attached to the protein
human serum albumin so that the product retains biological
activity, and can thus serve as an in vivo radiotracer
useful in clinical or diagnostic medicine. These para-
substituted, 1-phenyl EDTA compounds known to the art,
although forming a variety of stable chelates, have had
the disadvantage in that their synthesis has been rela-
tively difficult.
Various other chelating agents have also been used
for in vivo radiotracer studies, but have been less stable
than the l-phenyl EDTA compounds; thus, a large portion
of the radioactive metal ions has tended to be lost to
the serum protein transferrin. This leads to deposition
of radioactive metal ions in the liver and bone marrow.
Additionally, the prior known chelating agents have
normally been optically inactive: either because they
do not contain an asymmetric carbon atom, or because the
syntheses yield a racemic mixture. It is believed that
use of non-optically active chelating agents (or racemic
mixtures) with biological molecules may tend to adversely
affect the in vivo properties thereof.
The present invention is directed to overcoming one or
more of the problems as set forth above.
Disclosure Of The Invention
According to the invention there is provided a stereo-
specific method of making an optically active chelating
agent comprising: providing an amide derivative of an
optically active alpha amino acid, said amide derivative
having a carbonyl group and an alpha carbon, said alpha
carbon defining an L or D stereochemical configuration for
said amide derivative; reducing said carbonyl group of
'B
1178951
said amide derivative with borane to produce a diamine
having a pair of amine moieties with said alpha carbon
therebetween, said alpha carbon continuing to define
either said L or D stereochemical configuration for
said diamine; and, reacting said diamine with a carboxy-
methylating agent to form an ethylenediamine tetraacetic
acid analog, an ethylenediamine triacetic acid analog,
or a diethylenetriamine pentaacetic acid analog with said
alpha carbon retaining either said L or D stereochemical
configuration and recovering therefrom said ethylene-
diamine tetraacetic acid analog, said ethylenediamine
triacetic acid analog or said diethylenetriamine penta-
acetic acid analog as a substantially optically pure
product.
In another aspect of the present invention, a chelat-
ing agent comprises various particular EDTA, ED3A and DTPA
analogs which readily form stable chelates with a variety
of metal ions.
Detailed Description Of The Preferred Embodiments
The chelating agents of the present invention are
derived from amides of optically pure and active alpha
amino acids. The reactions leading to making the chela-
ting agents of the present invention are stereospecific.
Accordingly, the alpha amino acid amide precursors of the
inventive chelating agents may belong to either the L or
D stereochemical series, as follows.
All of the amino acids that have been found to
naturally occur in proteins, except glycine, contain
at least one asymmetric carbon atom, and are optically
active. By convention, all naturally occurring amino
acids found in proteins belong to the L stereochemical
~ 1 7~51
series, and are stereochemically related to L-
glyceraldehyde. As is well known, but defined herein
for purposes of clarity, these naturally occurring
amino acids may be generally represented as is herein
illustrated by Figure l(a), below.
- 3a -
.~
~.~78951
F;gure l(a~
COOH
I
~C~
H2N H
The enantiomer of Figure l(a) is represented by ~igure l(b),
below.
Figure l(b)
CcWH
/ ~ ~R
H NH2
As illustrated by Figures l(a) and l(b), all of these alpha
amino acids have a carboxyl group and an amino group at the
carbon atom alpha to the carboxyl group. The R moiety, well
know to the art, represents the variations among these
amino acids. For example, the R moiety of tyrosine is
-CH2-~-OH; whereas, the R moiety for phenylalanine is
C~2 0.
.~
~.178~5:~
Substantially all of the optically pure and active alpha
amino acids, when of either the L or D configuration, are
useful as precursors for preparing the inventive chelat-
ing agents. Among preferred precursor amino acids are:
tyrosine, phenylalanine, alanine, tryptophan, cysteine,
and lysine. Additionally, certain modified alpha amino
acids are useful as precursors for preparing the inven-
tive chelating agents. Among preferred precursor modified
alpha amino acids are optically pure and active methyl
tyrosine, methyl tryptophan and p-nitro phenylalanine.
In the best mode contemplated for carrying out the
present invention, the most preferred precursor amino
acids are tyrosine and phenylalanine of the L stereo-
chemical configuration. This is because the aromatic
rings thereof assist in enhancing certain end uses for
the inventive chelating agents, and because of the ready
availability and purity of the naturally occurring alpha
amino acids.
It is also believed that cysteine may provide par-
ticularly useful advantages for coupling with biologicalmolecules. The end uses and advantages shall be further
discussed after the following detailed description of the
formation of the inventive chelating agents.
Method
A method of making an optically active chelating agent
comprises providing an amide derivative of an optically
active amino acid. Such an amide derivative is generally
represented by either of the structures of Figure 2 below:
D
D
78951 ~-:
Figure 2
C- ~ C--~ H
/C~. R / j R
~; m e R moiety o~ Figure 2 is identical to, that is stems
from, the R moiety o~ the particular amino acid precursor.
The Rl moiety includes a variety of species, depending upon
the particular amine utilized in forming the amide
derivative. For example, if the amine is ammonia, then
Rl = H; if the amine is ethanolamine, then Rl = CH2CH20H; if
the amine is ethylenediamine, then Rl = CH2CH2N~2.
The amide derlvatlve has a carbonyl group. A
carbon atom is alpha to the carbonyl group and has the
original amino group of the alpha-amino acid bound thereto.
These amide derivatives may be prepared by conventional
peptide synthesis techniques or may in many instances be
purchased from comrnercial sources. In any event, the alpha
carbon continues to define the sarne op~ical configuration
(the L or D stereochemical configuration) for the amide
derivative as it did for the alpha amino acid precursor.
I~e carbonyl group of the aMide deriva'Give is then
chemically reduced with borane to produce a diamine
compound. Such a diamine compound is generally illustrated
-by Figure 3, below:
li 7895~
Figure 3
~,-C~ C~
The diamine compound thus has a pair of amine moieties with-
the alpha carbon therebetween. It has been found that the
alpha carbon continues to define the same optical
configuration for the diamine compound as it did for the
amide. That is, the reduction of the amide derivative wikh
borane preserves the original optical configuration. The R
moiety of Figure 3 ls substantially identical to the R
moiety of Figure 2, with an exception being that the
nitrogen atoms thereof will normally be protonated, each
with a H~ ion.
Ihis diamine compound is then reacted with one of
various conventlonal carboxymethylating agents. A chelating
agent in accordance with the present invention is recovered
therefrom, which has a general structure as represented by
Figure 4, below:
Figure 4
- C ~ ~ (C~,co~ C--~J (c~lc~D~\)
or ~ ~ ~C~ C~
1~789S~
The R moiety of the inventive chelating agent, as illustrated
by Figure 4 a~ove, is determined by the alpha amino acid
precursor. As shall be later described, the R moiety may
be subsequently modified if desired, so as to provide
covalent bonding opportunities for use of the chelating
agent as a bifunctional chelating agent. However, the
reaction of the diamine compound with the carboxymethylating
compound yields chelating agents having, that is continuing,
the L or D stereochemical configuration.
The R2 moiety is the same as, or is the carboxy-
methylated species from, the Rl moiety of Figures 2 and 3.
Thus, for examp}e, where Rl = H, then R2 = CH2COOH and
the chelating agent is an EDTA analog lwherein EDTA stands
for ethylenediamine tetraacetic acid]. Where Rl = CH2CH2OH,
then ~2 ~ CH2CH2OH, and the chelating agent is a hydroxyethyl-
ED3A analog lwherein ED3A stands for ethylenediamine triacetic
acid~. Where Rl = CH2CH2NH2, then R2 = CH2CH2N~CH2COOH)2, and
the chelating agent i8 a DTPA analog [wherein DTPA stands for
diethylenetriamine pentaacetic acid].
The inventive chelating agents and method shall
now be more fully described. The experimental procedures
and reagents were as follows.
~MR spectra were recorded at 60 MHz, on a Varian
EM360 Spectrometer with t-butanol (~ - 1.1) as internal
standard; pH meter readings were not corrected for the
deuterlum isotope effect. Fluorescence spectra were
recorded on a Perkin-Elmer/Hitachi MPF2A fluorescence
spectrophotometer, uncorrected for instrumental response.
Optical rotation measurements were made at 23 using a 1 dm
-- 8
-1~78951
tube in a JEOL DIP-180 polarimeter. All TLC analyses were
run in either solvent 1: n-amyl alcohol/pyridine/water
(43/37/20), V/V), solvent 2: 95% ethanol/25% NH40H (4/1,
V/V), solvent 3: 10% aq NH40Ac/methanol (1/1, V/V), or
solvent 4: 15 ml of 0.033 M HCl/35 ml of acetone, on
silica gel F-254 plates (Merck) with fluorescent indicator.
High voltage paper electrophoresis was run on Whatman 3 MM
paper with a Savant HV 5000 UL 115-3 power supply, FP22A
cooling plate, and RWC-50 UL recirculating water cooler.
Radioactive samples were counted on a Beckman Gamma 310
instrument.
Carrier-free lllInC13 was obtained from Medi-
Physics, Emeryville, California and purified. 5gFeC13
was obtained from New England Nuclear. Human serum
albumin was purchased from Cutter Laboratories as a 25%
solution, 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide
was obtained from Calbiochem. Borane-tetrahydrofuran
complex, L-phenylalanine, L-tryptophan, L-tyrosine, and
bromoacetic acid were purchased from Aldrich. L-alanine,
L-tyrosinamide, L-phenylalaninamide, and L-methyl tyrosine
were purchased from Sigma. Dowex* 50 and AG*l ion exchange
resins and Biogel* P-100 resin (porous polyacrylamide
beads) were Bio-Rad Laboratories, Inc. All other solvents
and reagents used were the purest commercially available
products.
Providing Step
The amide derivative of the alpha amino acid or
alpha amino acid analog may be provided by converting
the alpha amino acid or analog to an ester form, and then
converting this ester form to an amide form. Procedures 1
* Trade Mark
_ g _
B
117~395~
and 2, as follows, illustrate these preferred ester forms
and then the amide conversions for the providing step.
Procedure 1
a~ino acid - ~ amino acid methyl ester
The amino acid can be converted to the ester in
one of two ways. The amino acid is treated with Dowex 50W
resin (H I dried, 5.2 meq/g, in equimolar amounts in
methanol. This two phase system is allowed to stir for 1-2
days. Alternatively, the amino acid can -be treated with
10 methanol saturated with HCl~g), in approximately 20 ml of
methanol solution to lg of solid amino acid, and allowed to
sit for 1-2 days. The course of the reaction can be
followed by IR scan of the reaction solution, a change in
carbonyl from about 1650 cm 1 to about 1750 cm 1 is
expected, and the reaction is complete when the acid
¢arbonyl is gone. ~For the resin reaction, a sample of the
resin should be treated with HCl(g) to liberate the amino
acid ester into solution). The resin solution will be used
as it is. The methanol/HCl solution is reduced to a solid
20 under reduced pressure, solids recrystalized from ethanol.
TLC were run in solvent 2.
Esters of the following compounds have been made:
-- 10 --
~.~L7~3~5~
`.
a) ~ ~ ~. ~,
~i O J~ ~ O L~
.
o o o o o o o
Cq
~ - . '- ,.
` ~ , V
,1 :C ~ ~ P:
~ s ~
~d ~ ~Q o o o o o
o
~E
~ ~ ~ oo , ~ , ~U
~ o o o o o o o
a
s
o ~ ~
h ~ cq
.~ ~,, O
c~ h :~,
C~ Q~
~: ~ a
~ ~: ~1 0 S ~ ~
O ~q ~ h J~ ~ ~ h
~: O
td ~ h
1~ h ~ S ~ I O a~
¢ E~ ¢ 1
~7sssi
Procedure 2
~amino acid methyl ester ~ amino acid amîde)
The amino acid methyl ester can be converted to
the primary amide by treating either the resin bound ester
with repeated NH3(g) saturated methanol and filtering, or
the amino acid methyl ester hydrochloride with N~3(g)
saturated methanol, and leaving the methanol solution to sit
for 3-7 days. The course of the reaction can be followed by
monitoring the carbonyl peak on IR, as it changes from about
10 1750 ¢m 1 to about 1620 cm 1. Reaction is complete when the
ester carbonyl is gone. The solvent is removed under
reduced pressure, and TLC in solvent 2 shows one spot.
The primary amide derivatives of the following
amino acid esters were made:
Amino Acid EsterAmide RF Yield
Tyrosine .63 82%
Alanine .62 85%
Tryptophan .66 80%
Methyl tryptophan .62 65%
NO2 phenyl alanine .66 87%
Methyl tyrosine .64 90%
The amino acid methyl ester can be converted to a
secondary amide by reacting the amino acid methyl ester
hydrochloride with 3 equivalents of distilled ethanol amine
or ethylene diamine in methanol and allowing the reaction to
proceed for 2-3 days. The course of the reaction can be
followed ~y the disappearance of the carbonyl at about 1750
cm 1. Upon completion of the reaction, the solvent is
removed under reduced pressure. The oily residue is
~1
.. . .
1~7895~
dissolved in water and 6 N HCl to pH <3, then purified on an
AG 50w column, NH4 . The column is washed with water until
eluant is Ag /H negative, and followed by a 0-2 M NH40H
gradient, monitored at 254 nm. The major peak removed in
0.5 - 1.5 M, N~40~ is lyophilized. The following amino acid
esters have been converted to ethanol amides ~TLC in solvent
2):
Amino Acid Ester RF of Amide Yield
Tyrosine .74 89%
Phenyl alanine .68 93%
Methyl tryptophan .64 67%
N02 phenyl alanine .65 94%
The only amino acid ester treated with ethylene
diamine was P-N02 phenylalanine. When the reaction was
complete, the solvent was removed under reduced pressure.
The oily residue is dissolved in water, and the product i5
extracted wîth tetrahydrofuran. The tetrahydrofuran solution,
after back-extraction with water, yields pure product with RF
0.30 in TLC solvent 2, 60%.
Reducing Step
The reducing of the amide derivative with borane
to produce a diamine is illustrated by Procedure 3 as follows:
Procedure 3
amino acid amide ~1 and 2)- ) R-diamine (1 and 2)
All of the amino acid amides prepared in procedure
2 were reduced with borane in tetrahydrofuran using one
- 13 -
1~78951
equivalent of BH3 for every -OH, or -NH proton, two
equivalents for the carbonyl, and a surplus of one equivalent.
In addition, phenylalanylglycine was reduced with nine
equivalents of borane, to yield N'-Hydroxyethyl-l-benzyl
ethylene diamine, as an alternative procedure.
The procedure used in reducing the amides required
dissolving about lgm of the amide in dry tetrahydrofuran, or
in combination with dry 1, 2-diamethoxyethane, in a clean,
oven-dried, three-neck round bottom flask fitted with a
10 reflux condenser (capped by a CaC12 drying tu~e), a serum
stopper, and a ground glass stopper. The assembled
apparatus was flushed with dry N21g) for 15 minutes, while
the flask was cooled in an ice bath. The borane-tetrahydro-
furan solution tl.OM solution), was added slowly by syringe
through the serum stopper at a rate of about 1.5 ml/minute.
The syringe and the sides of the flask were washed with
additional dry tetrahydrofuran. In some cases a white
gelatinous solid appeared during the addition of the borane
reagent. The serum stopper was replaced by a glass stopper,
20 the ice bath was exchanged for a heating mantle, and the
mixture was refluxed for 5 hours. The solution was cooled
on ice for 20 minutes, and 50 ml of anhydrous methanol was
added to the unstoppered reaction vessel. The solution that
resulted was then saturated with HCl(g) and refluxed for
1 hour. The solvent was removed under reduced pressure. The
primary amino acid amides reduce to vicinal diamines which
solidify after removing reduction solvent. The vicinal
diamines are recrystalized from either absolute
- 14 -
1:1 789Sl
ethanol or anhydrous isopropyl alcohol. The amino acid
(~-hydroxyethyl)amides when reduced produce oils with
removal of reduction solvent. .These oils are purified by an
AG 50w column (NH4 ) monitored at 254 nm. The oily residue
was dissolved in water and 6 N HCl to pH 2, then applied to
the column. The column was then eluted with water till
effluent was negative to Ag~/H+, and then switched to
0-6.0 M NH40H linear gradient. The product, removed 3-5 M
NH40H was then lyophilized.
Finally, when the solvent was removed from P-N02
phenylalanine (B-amino ethyl)amide reduction a solid
resulted which was recrystalized from ethanol/ether. TLC in
solvent 2.
The following R-diamines were prepared by the above-
described Procedure 3 (wherein N' will always herein refer
to the nitrogen atom which was originally in the amide
linkage):
,~ . /
1.~789S~
~IC~Pc1Pc~Pc~Pc~Pc~P~Pc~P ~P dP c~ c~P
a cO u~ O O ~ c u~ O 0~ c.~
~r oo o o c~ co o co c~ ~ o ~o
o o o o o o o o o o o o
c~ n~ c
~ ~ 8 ~ ~
Z Z Z Z
-- 16 --
1~78951
.
.
m e products of the prlmary amldes, when reduced,
.
form viclnal diamines which react with disodium rhodizonate.
For test solutlons the pH should be 5-7. All of the
ethylenediamine derivatives prepared were characterized with
this test. All of the products of borane reduction were
characterized by the absence of carbonyl in IR scans. In
addition, l-(p-hydroxybenzyl)-ethylene diamine
dihydrochloride analyzed for C,H,N,Cl gave: -
,
Experimental: C-45.15%, H-6.53%, N-11.48%, Cl-29.35
Theory: C-45.20%, H-6.74%, N-11.72%, C1-29.65
m e C,H,N,Cl analysis for l-(p-nitrobenzyl)-
ethylenediamine dihydrochlo-ride was:
Experimental: C-40.04%, H-5.60%, N-15.67%, Cl-26.49%
Theory: ~-39.78%, H-5.87%, N-15.47%, Cl-26.19%
As an example of the retention of optical activity, Lt+)1,2-
diaminopropane was [ ~ ]23 = +42 + 1 (c 0.27,benzene).
This value is 7 higher than that reported for
determinatlons of [ ~ ~D from L~+)-1,2-diaminopropane
samples prepared by classical resolution of the racemic
mixture.
Reacting, or Carboxymethylating Step
Reactlng of the diamine with a carboxymethylating
agent is illustrated by Procedure 4, as follows.
.
Procedure 4
Treatment of the products of procedure 3 with
carboxymethylation reagents, such as bromoacetic acid,
produced chelates~ that var~ in the number of acid groups
from 3 to 5. To avoid heavy-metal-ion contamination, all
,
17
~1~78951
H2O used was deionized and distilled into acid-washed con-
tainers, and all glassware and transfer apparatus was washed
with conc. H2SO4/HNO3 (50/50, v/v) and rinsed with H2O.
The ethylene diamine dihydrochloride derivative of interest
was dissolved in a minimum volume of H2O in a wa~er-jacketed
reaction vessel co~nected to a 45C circulating water bath.
The magnetically stirred solution was adjusted to and main-
tained at pH 10 with 7 M KOH while 1.1 equivalents, to
replace every -N-H and phenol proton, of bromoacetic acid
10 was added in portions. By addition of KOH, the pH was kept
between 10-11 for approximately 20 hours (the pH was monitored
frequently during the first 3 hours). After 20 hours, the
product was purified by either of the following ways.
(l) Treating of the reaction solution with conc.
HCl till pH was 1.8, and then evaporating solution under
reduced pressure. The re~ulting solid was then extracted
with four 5-ml portions of boiling 95% ethanol; the
combined filtrates were dried under reduced pressure to an
oil. The oil, dissolved in H2O, was adjusted to pH 6.5,
20 with 6 N NaOH, and applied to an AG 1 anion exchange column
in the formate form, 10 meq of resin per meq of product.
Or, (2) Reaction solution may be applied directly
to the AG 1, formate, anion exchange column having 10 meq
of resin per meq of product. The product was eluted by a
linear gradient of formic acid from 1.5 to 7.0 M and the
column effluent was monitored at 254 nm, or in the nitro
derivatives 280 nm. The major peak, which eluted between
3.5 to 5.7 M formic acid was collected and lyophilized.
The resultant solid was analyzed by TLC, in solvent 3 and NMR.
The ~ollowqng di~nes-~ere cark~thylate~ (whe~ein, as with
EDTA and ED3A, carbon 1 of the DqPA product is and oontinues to be the
asymetric carbon in t~e alpha position of the sta ~ ng amine acid):
- 18 -
u~ ooo ~ ~ ~ ~:r
o ~ ~ o u~ o o ~
D;
o o o o o o o o
-- ~.q _ ~ -- -- _
0 ~ 1~ 0 0
O ~ N ~ O Lt~ ` 117
o o u~u~ o I I a~
z z z ~l r~
~ l
Zo~ ~
. ~ ~; o o
a~ ~ ,~ ~ I ~ :c u7 u~
~ ~ ~ ~ ~ 2 c~
~ : 3
rl a~
u~ ~ $
¢/
19 -
~78951
~1
a~ ,~
~;i
o~
~1 o
I'
- o ~ u~
Z a~
dO d
a~ , ~ ~
~ z ~ ~
~,~ u ~ u
D ~ . ~
- 20-
117895~
Examples I-VIII further illustrate preparation of the
inventive chelating agents, and Examples IX-XIII illus-
trate further modifications, within the scope of this
invention, to increase the flexibility in uses thereof.
EXAMPLE I
L-tvrosine --~ L-tyrosinamide
lOg (55 mmol) L-tyrosine was treated with 300
mL methanol and 11.3 g Dowex* 50Wx8 (H~) resin. The
reaction mixture was stirred for three days at room
temperature. After completion of the esterification,
the mixture containing L-tyrosine methylester was sat-
urated with ammonia; the resin was removed by filtration;
and the filtrate was stirred for 7 days. The yield of
L-tyrosinamide from L-tyrosir.e was 60% after recrystal
lization from water. The proton NMR, TLC, and mp (154)
of the product were identical to those of a sample of
L-tyrosinamide from Sigma. TLC ln solvent 1: Rf 0.40.
EXAMPLE II
L-alanine--~L-alaninamide
Starting with lOg (112 mmol) of L-alanine,
L-alaninamide was prepared by the procedure as in Example
I, with a difference being that four treatments with 200
mL portions of NH3 - saturated methanol were required to
remove l,lg (13 mmol) of alanine methyl ester from 22g of
Dowex resin.
* Trade Mark
- 21 -
B
(` ~i7895~ ~`
EXAMPLE III
L-phenylalanine --~ L-p-nitro-phenylalanine amide
10.0 g (60.6 mmoles) of L-phenylalanine was
dissolved in 16 ml concentrated H2S04 (95-98%, D. 1.84 g/ml)
at 0C. 3.0 ml o~ HNO3 (90%, ~. 1.5 g/ml) was added
dropwise to the stirring solution keeping temperature at
about 0C. After all of the HN03 has been added, the
solution was allowed to stir 10-15 minutes. ~eaction
solution was ~hen poured over about 200 ml o~ ice and then
diluted to about 700 ml with additional H20 This solut~on
was then ~eated to a boil, and neutralized with PbC03, about
80 g. The resulting precipitate was filtered and the
supernatant was treated with H2S(g) to precipitate remaining
Pb , and then filtered. The resulting filtrate was reduced
to 1/3 its volume. The solid which formed was filtered and
washed with 95% ethanol. Yield was 50-55% of p-
nitrophenylalanine when recrystallized from boiling H2O.
5 g of p-nitrophenylalanine (23.8 mmoles) was
dissolved with methanol saturated with HCl(g). The
resulting solution was allowed to stand for 1-2 days.
Solvent was removed under reduced pressure, and the
resulting solid, p-nitrophenylalanine methyl ester, was
recrystallized from absolute EtOH. TLC of product ester in
solvent 2 showed one spot, Rf .78 (p-nitrophenylalanine Rf
.58). Yield is 90-95% when recrystallized from absolute
ethanol.
4.5 g of p-nitrophenylalanine methyl ester . HCl
(17.3 mmoles) was treated with methanol saturated with
- NH3(g). Resulting solution was allowed to stand 3-4 days,
while maintaining NH3 saturation. Reaction solution was
~ 117~951
then rotoevaporated to dryness. The resulting solid,
p-nitrophenylalanine am~de, was recrystallized from absolute
EtOH. TLC in sol~ent 2 shows one spot, R~ .70 to the
starting ester Rf .78. Yield is 68-75%, m.p. 235C after
recrystallizing from absolute ethan ~.
Examples IV and VI, below, illustrate the
reducing step of the provided amide derivative with borane.
Example ~ illustrates that this reducing step preserves the
original optical configuration.
EXAMPLE IV
L-tyrosinamide ~ L-l-(p-hydroxybenzyl)
- 1,2-ethylenediamine dihydrochloride
All organic solvents used in this step were dried
over 3 A molecular sieves; all glassware was oven-dried and
flushed wlth dry N2(g). 1.0 g (5.5 mmol) of L-tyrosinamide
was dissolved in 220 mL of 1,2-dimethoxyethane and 50 mL
tetrahydrofuran in a 3-neck round-bottom flask fltted with a
reflux condens-er (capped by a ~aCl2 drying tube), a serum
stopper, and a ground glass stopper. The assembled
apparatus was flushed with dry N2(g) for 15 minutes, while
the flask was cooled in an ice bath. Borane-tetrahydrofuran
complex, 49 mL of a 1.0 M solution (49 mmol), was added
slowly by syringe through the serum stopper in a 30 minute
period. The syringe and the sides of the flask were ~lashed
with an additional ~0 mL tetrahydrofuran. A small amount of
white solid appeared during addition of the borane reagent.
The serum stopper was replaced by a glass stopper, the ice
bath was exchan~ed for a heating mantle, and the mixture was
refluxed for 5 h. The solution was cooled on ice for 20
minutes, and 50 mL of anhydrous methanol was added to the
unstoppered reaction vessel. The solution that resulted was
then saturated with HCl(g)-and refluxed for 1 h. The
solvent was removed under reduced pressure, and the residue
was recrystallized from 95% EtOH; yield, 990 mg (68%). The
product, L-1-(p-hydroxybenzyl)-1,2-ethylenediamine
dihydrochloride, gave one spot by TLC in solvent 2, Rf 0.34,
(tyrosinamide has Rf 0.74), a positive rhodizonate test for
vicinal diamines, and a positive folin test for phenols.
NMR (pH 6.4, D20) ~ = 2.3 - 3.3 (aliphatic, 5, m)~ 6.8
(aromatic aa'bb', 4, m). Anal. tCgH16N20Cl2~ C, H, N, C1.
Theory: C-45.20, H-6.74, N-11.72, Cl-29.65
Experimental: C-45.15, H-6.53, N-11.48, C1-29.35
EXAMPLE V
L-alaninamide -~ L-1,2 diaminopropane
L-alaninamide was treated in a manner analogous to
Example IV. Reduction of l.lg. (13 mmol) of L-alaninamide
with 65 mmol of borane complex proceeded without formation
of a precipitate. Upon addition of HCl(g), the product
precipitated from solution. After refluxing the mixture l
h, 850 mg (44%) of pure 1,2 diaminopropane was collected,
having [ a ]23 = ~ 42 ~ 1 (c 0.27, benzene), which in
fact is 7 hlgher than that reported for determinations of
[ a ~D from L (~) - 1,2 - diaminopropane samples prepared
by classical resolution of the racemic mixture.
Accordingly, this illustrates that reduction with borane of
the amide derivatives, in accordance with the present
invention, does not affect the optically active center of
the ~-amino acid amide.
_ 24
1 ~7895~
EXAMPLE VI
L-p-nitrophenylalanine amide -~ L-p-nitrobenzyl
ethylenediamine dihydrochloride~
1.50 g of p-nitrophenylalanine amide . HCl (6.2
mmoles) was dissolved in 75 ml tetrahydrofuran and cooled to
O~C. The system was closed to atmosphere and flushed with
N2(g) for 20 minutes. 43.3 ml of 1.0 M borane-
tetrahydrofuran . solution was-added slowly to the stirred
solution, with gases allowed to escape through a ~ent, over
a 30-45 minute period. After the last of borane solution
was added, the solution was heated to reflux ~or about 5
hrs, capped and vented with drying tube. After reflux, the
solution was cooled in ice for about 1/2 hr. To the cooled
solution, 50 ml MeOH i5 added to the opened container with
stirring. The solution was then saturated with HCl(g) and
refluxed again ~or about 1 hr. Reaction solution was
rotoevaporated to dryness and the resulting residue,
dissolved in H20, was pur~fied by Ag-50W resin (H ) column,
~ith O - 7 N HCl gradient. Ma~or peak, removed in 6-7 N HCl
region, was lyopnilized. [IR scan showed lack of C=O]. TLC
in solvent 4 showed Rf of product diamine .50 to the
starting amide R~ .78 and TLC in sol~ent 3 showed Rf of
diamine .50 to the amlde Rf .70. Yield was 65-70%.
Elemental Analysis: CgH15C12N302
Theory: C - 40.04%, H - 5.60%, N - 15.67%, Cl - 26.49%
Experimental: C - 39.78%, H - 5.87%, N - 15.43~, C1 - 26.19%
- 25
- 1~789Si
Examples VII and VIII, below, lllustrate the
reacting of the diamine compound with a carboxymethylating
agent to y~eld an EDTA analog in accordance with the present
invention.
EXAMPLE VII
L-1-(p-hydroxybenzyl)-1,2-ethylenediamine
dihydrochloride ~ ? L-l(p-carboxymethoxy benzyl EDTA
To avoid heavy-metal-lon contamination, all H20
used was deionized and distilled into acid-washed
containers, and all glassware and transfer apparatus was
washed with conc. H2S04/HN03 (50/50, V/V~ and rinsed with
H20. 734 rng (3.1 mmol) of the product from Example IV,
above, ~ras dissolved in a minimum volume of H20
(approximately 5 mL) in a water-jacketed reaction vessel
connected to a 45C circulating water bath. The
magnetically stirred solution was adjusted to and maintained
at pH 10 with 7N KOH while 2.3 g (16.7 mmol~ of BrCH2COOH
was added in portions. By addition of KOH, the pH was kept
between 10-11 for approximately 20 h (the pH was monitored
frequently during the first 3 h~. After 20 h, the reaction
solution was adjusted to pH 1.8 with conc. HCl, and the
solution was evaporated to dryness under reduced pressure.
The resulting solid was extracted with four 5-rnL portions of
boiling 95% ethanol; the combined filtrates were dried under
reduced pressure to an oil. The oil was dissolved in 5 mL
H20. The solution was adjusted to pH 6.5 with 6N NaOH and
applied to an AGlx4 anion exchange column in the for~ate
form (1.5 x 25.5 cm resin bed). The product was eluted by a
linear gradient of formic acid from 2.5 to 7.0 M (total
~951
volume 1.4L), and the column effluent was monitored at 254 nm.
The major peak, w~ich eluted between 4.1 and 5.7 M formic
acid, was collected and lyophilized. The resultant solid,
L-l~p-carboxymethoxy benzyl)-ethylenediaminetetraacetic acid,
when analyzed ~y TLC in solvent 2, showed one spot at Rf 0.16.
Yield, 900 mg C64%~. NMR (pH 7.5, D20) = 3.0-3.7 (aliphatic,
13,m), 4.5 ~ether methylene, 2, s), 6.9 (aromatic aa'bb',4,m).
[a]D = + 25.2 + 0.4 (C 0.94, H2O). Anal. ~ClgH24N2Oll) C,H,N-
Theoretical: C-50.00%, H-5.26%, N-6.14%
Experimental: C-50.36%, H-5.56%, N-6.31%
Structure is as illustrated by Figure 5:
Figure 5
NCCH2COOH)2
CHCH2N(CH2COOH)2
ICH2
0
I
OCH2COOH
EXAMPLE VIII
L-p-nitrobenzyl ethylenediamine
dihydroc~loride ~ L-p-nitrobenzyl EDTA
0.400 g of L-p-nitrobenzyl ethylenediamine .
(HCl)2 (1.58 mmole~), prepared as described by Example VI,
above, was dissolved in H2O/KOH to pH 10, at about 45C;
- 27 -
.~
~1~789~ `
0.915 g Or bromoacetic acid (6.63 mmoles) was added and pH
was adjusted to 10-ll range wlth 7 N KOH. Reactlon was
allowed to continue while stirrlng for 17-20 hrs, overnight.
Reactlon solution was then purified by AG-l resin column
(formate), with 2.0 - 7.0 M HCOOH gradient. Product was
removed in 4.8 - 6.2 M HCOOH region. Fractions collected
were combined and lyophilized. TLC in solvent 3. Rf of
product was .78, to the diamine .50 Rf. NMR in D2O/NaoD to
pH 10.0, ratio aliphatic to aromatic was 3.29 to 1,
theoretical is 3.25 to 1.
Elemental analysis results: Formula: C17H21N3Olo ~H2O
Experimental: C - 43.34%, H - 4.94%, N - 8.93%
Theory: C - 43.31%, H - 4.95%, N - 8.91%
,
Structure is as illustrated by Figure 6:
Figure 6
N(CH2COOH)2
- C H CH2N(CH2COOH)2
0
I
N2
~ Lq895~L
,
EXAMPLE IX
L-p-nitrobenzyl E~TA ~-~ L-p-aminobenzyl EDTA
0.4 n~noles, 170 mg, of p-nitrobenzyl EDTA was
dissolved in ~20/NaOH to pH 11.5 and 30 mg of 10~
Pd/Charcoal was added, while on ice. Reaction was allowed
to stand under 1 atmosphere of H2 for 3 hrs, in ice. A
green color was observed to occur and disappear during
course of the run.
After about 3 hrs the solution was filtered to
remove the catalyst, then lyophilized. NMR spectrum showed
expected changes in aromatic proton resonance frequencies.
Also material was positive to ninhydrin. Yield:
quantitatlve. Example IX is representative of hydrogenation
of nitro derivatives of the chelating agents, to form amine
derivatives thereof. Particularly, N~-hydroxyethyl-l-(p-
nitrobenzylj-EDTA has been modified by an idéntical
hydrogenat,ion to form the amlne thereof. Such amine of N~-
hydroxyethyl-l-(p-nitrobenzyl)-EDTA WQS ~urther modlfied in
a manner identical to and with resu~ts as in the descrlption
of Examples XII and XIII.
EXAMPLE X ,-
L-p-aminobenzyl EDTA-~ L-p-bromoacetamidobe_ yl EDTA
303.5'~mole of p-aminobenzyl EDTA was disso],ved in
500'~1 of ~2 and the pH of the solution was ad~u~ted to 6.5
with 20 ~1 of concentrated HCl. Bromoacetyl bromide was
added (46'~1) until the mixture was negative to
fluorescamine. This was followed by 10 extractions wlth 500
~1 of diethyl ether to remove excess bromoacetyi bromide.
_ ~a
117~gSl
After the extractions, the pH of the solution was 1.0; it
was adjusted to 2.2 by the addition of 100 ~1 of 1 M NaOH.
A white precipitate began to drop out at pH 2. The solution
was placed on ice and left overnight. The mixture was then
centrifuged, and the white solid washed twice with 600 ~1 of
ice cold 0.1 M HCl. The solid was dried under vacuum.
Yield was 32.17 mg (61.8 ~mol; 20.4%) of
L-p-bromoacetamidobenzyl EDTA. TLC of the product on silica
plates with solvent 3 showed only one spot with an Rf of .96
10 which was both fluorescence quenching and
4-(p-nitrobenzyl)pyridine positive. Structure is as
illustrated by Figure 7:
Figure 7
N(CH2COOH)2
CHCH2N(CH2COOH)2
ICH2
I oll ,
NHCCH2Br
EXAMPLE XI
20 l-(p-methoxybenzyl)-EDTA ~ l-(p-hydroxybenzyl)-EDTA
The chelating agent, l-(p-methoxybenzyl) EDTA, was
treated with 48% (aq) HBr/glacial ace~ic acid (1:5) and
refluxed for 2 hours. The reaction solution was then
evaporated under reduced pressure to give a white solid
which was purified on AG 1, formate, column with a linear
gradient of 0 - 7M formic acid. The major peak, removed
- 30 -
,
1~78951
with 2.9 - 3.3 M formic acid, was lyophilized. The white
solid that resulted was analyzed by NMR and showed lack of
peak at ~ 3.75. Additionally, a U.V. scan at pH 3 and pH 10
showed changes in wavelength of maximum absorption from 275 nm
to 295 nm, ~the starting material methoxybenzyl EDT~ did not
exhibit any changes in U.V.). Fluorescent studies showed the
material to bind Tb 3, and Eu+3. The products yield was 25%.
The product may be further modified through phosphorylation
to yield another useful chelate.
EXAM2LE XII
Acylation of L-(p-aminobenzyl)EDTA
The amîno chelate derivative from Example IX, 240
~moles, was converted to the tetraethylammonium salt by
eluting through an AG50w column, tetraethylammonium form,
with water. The material was then lyophilized. The
tetraethylammonium salt, when dried, was taken up in 1 ml of
acetonitrile, dired over 32 sieves, and then treated with
480 ~moles of any fatty acid chloride, in these cases
stearyl chloride, arachidoyl chloride, and behenoyl chloride
20 were used, dissolved in S00 ~1 of dry acetonitrile or
chloroform. The reaction solution was then shaken
continuously for 3-5 hours, while monitoring the solution
with fluorescamine. Reaction is complete when solution was
fluorescamine negative. The product, which solidified
during the course of the reaction, was separated by
centrifuging, and drawing off the supernatant. The solid
was then washed repeatedly with dry acetonitrile. The solid
was then treated with H2O and adjusted to pH 3 with 6 N HCl
and the solid/liquid extracted with chloroform. The solid
- 30 phase was then dried. Yield was quantitative.
- 31 -
. .
~'`1 ("
~78951
.
EXAMPLE XIII
The amino chelate derivatlve from Example IX, 214
~moles, was dissolved in H2O and added to 22 ~1 o~ 85%
thiophosgene in carbon tetrachloride, with stirring.
Solution was monitored with fluorescamine and the reaction
was complete when test was negative. The reaction solution
was then evaporated under reduced pressure, leaving a white
solid. This solid was then scanned on IR to detect presence
of SCN- stretch at 2100 cm 1. A TLC in solvent 3 showed
solid to have Rf of .9O to the starting amino chelate Rf of
.76. A sample of the solid dissolved in methanol was
treated with NH3(g) and an IR run after 20 minutes showed a
decrease in the 2100 cm 1 peak. Yield was quantitative.
' USES AS BIFUNCTIONAL CHELATING AGENTS
The EDTA and DTPA analogs of the inventive
chelating agents form stable chelates with a variety of
metal ion~. Such chelating is by the metal-binding portion
of the molecules (that is, with-the,plurality of carboxylate
and amine moieties thereof). Among the metal ions which may
be chelated are the ionic species of the elements
represented by Figure 8, below.
///////
//////
- 32
~.~78951
Figure 8
13
Al
26 27 28 ~ 31
Cr Fe Cb Ni Cu ¦ Ga
39 40 . 43 44 45 46 49 50
Y Zr Tc Ru Rh Pd 1 In &
57 72 76 77 80 83
La Hf Os Ir Hg . Bi
89
Ac
. I __ _ .
58 59 60 61 62 63 64 65 66 67 68 69 70 71
Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
91 92 93 94 95 96 97 98 g9 100 101 102 103
Th Pa U Np Pu Am Cm Bk OE Es ~m M~ No Lw
When it is desired that the EDTA analogs be used
as bifunctional chelating agents - that i9, that the EDTA
analogs include a covalent bond-forming moiety for attachment
of the EDTA analogs to biological molecules and the like -
the EDTA analogs may be, if necessary, easily modified to
include such a covalent bond-forming moiety. As may be
understood, the term "covalent bond-forming moiety" includes
for example, amino and carboxyl groups which are easily
attached by carbodiimide coupling for form peptide bonds
10 with biological molecules.
'~
- 1~78g5~ ~
Thus, for exampie, the EDTA analog represented by
Figure 6 of Example VIII may be readily converted from
- having the para-substituted nitro group to a
para-substituted amino group by reactlon under 1 atmosphere
of hydrogen with a catalyst such as lO~ Palladium/charcoal.
As illustrated by Example X, the amino group Or this EDTA
analog may be easily further modified so as to provide that
the covalent-bond forming moie~y functions as an alkylating
moiety for the EDTA analog.
The EDTA analog represented by Figure 5 of Example
VII already has a para-substituted carboxyl group available
for coupling to a biological molecule, and thus would not
usually be further modified. Example XIV,below, illustrates
the coupling of this EDTA analog in the Fe(III) chelate form
to human serum albumin.
EXAMPLE XIV
A. CouPling Of An EDTA Analog To Ethylenediamine
Formation of an Fe(III) chelate with
L-l(p-carboxymethoxy benzyl)-EDTA effectively protects, or
blocks, the EDTA carboxyl groups of the metal blnding
portion from the carbodiimide coupling reaction at the
covalent blnding portion.
- With a trace amount of radioactive ~9FeCl3, the
Fe(III) chelate was made by adding FeS04 to the cornpound as
in Figure 5 at pH 2.0; the colorless Fe(II) chelate rapidly
air oxidized to the yellow Fe(III) chelate upon vigorous
mixing. The Fe(III) chelate was protected frorn strong light
to prevent photodecomposition. The chelate solution was
analyzed for complete binding of metal ions by TLC in
13l789Sl
solvent 3; any unchelated metal ions were detected as
radioactivity left at the origln. If necessary, addltional
chelating agent was added until TLC showed <0.5% total
radioactivity left at the origin. Typically, after the
chelate was formed, 350 uL of a 45 mM chelate solution
(0.016 mmol) was added to 0.400 mmol of 1,2-diaminoethane
dihydrochloride, and the solution was adjusted to pH 5.0
with HCl and NaHC03. Half of the solution was saved as a
control. Using a freshly prepared 0.30 M 1-ethyl-3-(3-
dimethylaminopropyl)-carbodiimide solution, aliquots
containlng 1~3 mole of carbodiimide per mole chelate were
then added to the reaction solution. One hour after each
addition, high voltage paper electrophoresis analysis was
per~ormed in a 1.11 m acetic acid solution. The reaction
and control solutions were applied to the center of the
paper. Each electrophoresis waæ run at 4000 v for 30 min.
The paper was then dried and the position of colored iron
chelates noted. The paper waæ treated with ninhydrin~ and 2
cm sections were cut and counted. The desired coupling
product was located at the orlgin as determined by
radioactivity, a positive ninhydrin test, and by the yellow
color of the Fe(III) chelate.
B. Coupling Of the EDTA Analog
Chelate To ~uman Serum Albumin
Normal human serum albumin, 25% in saline
solutionJ was dialyzed at 4C for four days against 3
changes of 0.9% NaCl. The Fe(III) chelate from subpart A,
above, had a tracer amount of 59FeC13 added. Typically, 640
~L of the 0.22 M Fe(III) cnelate solution (141J~mol) at pH
5.0 was added to 240 ~L of a 2.7 mM albumin solution
- 35
~- 1178951
(0.65 ~mol) at the same pH in 0.9% NaCl. A total of 28 YL
of a freshly prepared 0.18 M carbodiimide solution (5.0
- mol) was added to the reaction solution at room temperature
in order to obtain approximately 0.5 chelate covalently
bound per albumin molecule. Polyacrylamide gel
electrophoresls indlcated no cross-linklng of albumin during
the coupllng reaction. Initlal analysis of the product
lnvolved Bio-Gel P-100 gel filtration of an aliquot of the
reaction mixture in a 0.7 x 9 cm column eluted with 0.9%
NaCl, and 59Fe counting of the fractions (0.5 + 0.1 chelate
per albumin). A nonspecific binding control was also
performed. Of the radioactivity shown by gel filtration to
be bound to protein, 95% was removed by dialysis of the
reaction mixture overnight at 4C and pH 5.0 a~ainst 2
changes of 50 mM EDTA, 50 mM citrate, and 34 mM ascorbate.
The labeled protein was dialyzed at 4C against 0.1 M
citrate, pH 5.0, for 8 days. A fluorescence titration using
terbium(III) yielded a value of 0.5 + 0.05 chelate per
albumin. Addition of radiolabelled indium(III) in excess to
an aliquot of protein, and subsequent analysis by Bio-Gel P-
100 column chromatography yielded a value of 0.5 + 0.05
chelate per albumin, in excellent agreement with the other
measurements. A control of normal albumin (dialyzed in
exactly the same manner) did not show any significant indium
binding.
In the case of the Fe(III) chelate of the EDTA
analog coupled to human serum albumin, described by subpart
B of Example XIV,above, iron may be easily removed after
reduction to Fe(II) by dialysis against a buffer containing
ascorbate, EDTA~ and citrate at pH 5. The iron-free
1~78951 ~;
protein-chelate con~ugate may then be labelled rapidly with
a short-lived, radionuclide such as lllindium, and the
~ radiolabelled protein used as a radlopharmaceutical.
The use of terbium or europium chelates of EDTA
analogs formed from L-tyrosinamide and L-p-nitro-
phenylalaninamide (such as the analogs of Figures 5 and 6)
ln fluorescence energy transfer studies is particualrly
attractive because the aromatic rings thereof lead to
- enhanced luminescence intensity from the chelated
lanthanides. Also, the use of an EDTA analog formed ~rom an
L-cysteine amide is attractive for coupllng with certain
b~ological molecules via disulfide bonds, ratner than the
peptide bond formation previously described.
In summary, the present invention discloses that
the amide derlvatives of C~-amino acids can be easily
converted into EDTA, ED3A or DTPA analogs with retention of
configuration at the asymmetrlc carbon thereof. Such EDTA,
ED3A and DTPA analogs are partlcularly useful as
bifunctional chelating agents w~erei-n a covalent bond
forming moiety of the EDTA analog, ED3A analog~ or of the
DTPA analog may be attached to a biological molecule in such
a way that the product retains biological activity, and can
serve as an in vivo radiotracer when the metal binding
moiety of the EDTA, ED3A or DTPA analog has formed a chelate
with an appropriate radioactive metal ion.
- 37
