Chemiluminescent acridinium-9-carboxamide boronic acid probes: Application to a homogeneous glycated hemoglobin assay

Article abstract of DOI:10.1016/j.bmcl.2005.11.062

Chemiluminescent acridinium-9-carboxamide probes containing 1, 3, 9, and 27 phenylboronic acids were prepared and their chemiluminescent properties evaluated. The relative chemiluminescent signal from the probes varied from 4 to 0.83 × 10¹⁹ cou

Full text of DOI:10.1016/j.bmcl.2005.11.062

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1324–1328
Chemiluminescent acridinium-9-carboxamide boronic acid
probes: Application to a homogeneous glycated hemoglobin assay
Maciej Adamczyk,* Yon-Yih Chen, Donald D. Johnson, Phillip G. Mattingly,
Jeffrey A. Moore, You Pan and Rajarathnam E. Reddy
Abbott Laboratories, Diagnostics Division, 100 Abbott Park Road, AP20, Abbott Park, IL 60064-6016, USA
Received 3 October 2005; revised 10 November 2005; accepted 15 November 2005
Available online 15 December 2005
Abstract—Chemiluminescent acridinium-9-carboxamide probes containing 1, 3, 9, and 27 phenylboronic acids were prepared and
their chemiluminescent properties evaluated. The relative chemiluminescent signal from the probes varied from 4 to
0.83 · 1019 counts/mol across the series, while the apparent affinity of the probes for the diabetes marker glycated hemoglobin
increased from 211 to 0.43 lM. The dose-dependent modulation of the chemiluminescent intensity of the probes upon binding
was used to demonstrate a homogeneous assay for glycated hemoglobin.
ꢀ 2006 Published by Elsevier Ltd.
The signal from chemiluminescent acridinium-9-carbox-
amide-modified ligands can be specifically quenched/
modulated upon non-covalent association with binding
proteins specific for the ligand. The ligand–binding pro-
tein pairs investigated thus far have included biotin–
avidin,1 folic acid–folate binding protein,2 and vitamin
B12—intrinsic factor.3 In each case, the degree of signal
modulation was dose-dependent and sufficient to dem-
onstrate a competitive homogeneous assay for the li-
gand of interest.
In this communication, we report a homogeneous
chemiluminescent assay for glycated hemoglobin, a
diagnostic marker for which no innate binding pro-
tein–ligand pairing is available.
Glycated hemoglobin (GHb) is the marker of choice in
assessing patient glycemic status during the diagnosis
and treatment of diabetes.4,5 The non-enzymatic reac-
tion of glucose and hemoglobin (Hb) primarily forms
a Schiff base with the b-chain, amino terminal valine
that subsequently undergoes an Amadori rearrangement
Figure 1. Formation of glycated hemoglobin (GHb) via the Amadori
rearrangement.
to give a stable N-(1-deoxy-D-fructos-1-yl)amino deriva-
tive (designated HbA1c) (Fig. 1). Other minor glycation
products are also formed and contribute to the total
in vivo is relatively long, the degree of glycation reflects
the time-averaged concentration of glucose over the
%GHb. Since the half-life of the hemoglobin molecule
previous 6–8 weeks.
Keywords: Chemiluminescence; Homogeneous assay; Multivalent
ligand; Glycated hemoglobin; Diabetes marker; Boronic acid.
*
The current trends in assaying this marker have been
recently reviewed.4 Among the most useful assays are
Corresponding author. Tel.: +1 847 937 0225; fax: +1 847 938
8927; e-mail: maciej.adamczyk@abbott.com
0960-894X/$ - see front matter ꢀ 2006 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2005.11.062

M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1324–1328
1325
O
a
H
H2O
OH
OH
OH
OH
OH
CbzNH(CH2)5CO2H
H2N
O
B
B
3
8
R
R
B
9
2H2O
cis-diol
CbzNH(CH2)5CONHC[(CH2)2COR]3
10 R = O-t- Bu
O
b
c
d
O
R
HO
11 R = OH
12 R = N-oxysuccinimidyl (HO)2B
Figure 2. Specific interaction of arylboronic acids with cis-diols.
13 R =
NH(CH2)2O(CH2)2O(CH2)2N
those that report total %GHb based on the specific
reaction of aryl boronic acids with the cis-diols of the
glycated molecule (Fig. 2).
O
O
R
R
O
e
NH
The stability of the boronate ester is rather low in
aqueous media (logK ꢀ 3) and typically is augmented
by the multivalent interactions provided by polymeric6
or dendritic boronic acids7 (see also any of the several
excellent reviews on boronic acid carbohydrate
sensors8–10).
(CH2)9C(O)NH(CH2)5CONH
R
N
O
14
O
NSO2
(CH2)3SO3
To demonstrate chemiluminescent signal modulation in
the detection of glycated hemoglobin, a series of acridi-
Scheme 2. Preparation of trivalent boronic acid chemiluminescent
probe 14. Reagents and conditions: (a) i—oxalyl chloride, CH2Cl2,
DMF; ii—BeheraÕs amine tert-butyl ester 9, Et3N, CH2Cl2, 1 h, 70%;
(b) TFA, CH2Cl2, 2 h, 88%; (c) N-hydroxysuccinimide, EDAC, DMF,
14 h, 62%; (d) 5, Et3N, DMF, 14 h, 79%; (e) i—HBr/AcOH, 45 min,
83%; ii—6, Et3N, DMF, 14 h, 86%.
a
t-BocNH(CH2CH2O)3H
t-BocNH(CH2CH2O)2CH2CH2NHCH3
1
2
Br Br
O
B
B
O
O
nium-9-carboxamide boronic acid reagents were pre-
pared containing from 1 to 27 boronic acid groups,
with the aim of identifying a reagent with the minimum
diol ÔaffinityÕ sufficient for an assay. The increased stabil-
ity and operable pH range offered by o-aminomethyl
arylboronic acids in the formation of the boronate
ester11 prompted the choice of compound 5 as the
common element for incorporation into the target
reagents.
B
b
Br
3
(HO)2B
N
O
RNH
O
4
R = t-Boc
R = H
c
Thus, as shown in Scheme 1,12 alcohol 113 was converted
to the methylamine linker 2, and then alkylated by the
o-bromomethylphenyl boronic acid anhydride 314,15 to
afford the t-Boc-protected intermediate 4. Subsequent
trifluoroacetic acid deprotection to 5 and acylation with
acridinium-9-carboxamide pentafluorophenyl active
ester 62 gave the desired mono-boronic acid chemilumi-
nescent probe 7.
5
(CH2)9CO2C6F5
N
d
O
N SO2
O3S(CH2)3
6
(HO)2B
N
O
Chemiluminescent arborol probes 14, 25, and 35 bearing
3, 9, and 27 boronic acid surface groups, respectively,
were prepared as shown in Schemes 2–4. The tert-butyl
ester of BeheraÕs amine16 (9, Scheme 2) served as the ini-
tial AB3 branching unit. The focal point was extended
by the addition of a 6-(Cbz-amino)caproic acid linker
to give 10. Acid deprotection, activation, and coupling
to boronic acid 5 afforded 13. Removal of the Cbz-pro-
tecting group by hydrogenolysis and acylation with
acridinium active ester 6 completed the preparation of
the trivalent reagent 14. Preparation of the 9-mer 25
O
(CH2)9CNH
O
N
7
O
NSO2
(CH2)3SO3
Scheme 1. Preparation of mono-boronic acid chemiluminescent probe
7. Reagents and conditions: (a) i—MsCl, Et3N, toluene, 0 ꢁC, 2 h;
ii—methylamine, 7 d, 68%; (b) 3, K2CO3, AcCN, 69%; (c) i—TFA,
CH2Cl2, 15 min; (d) 6, DMF/buffer (pH 8), 2 h, 50%.

1326
M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1324–1328
BnO2C(CH2)5CO2H
a
15
10 (Scheme 2)
BnO2C(CH2)5CONHC[(CH2)2COR]3
d
16 R = O-t- Bu
b
NH2(CH2)5CONHC[(CH2)2CO2t-Bu]3
17 R = OH
c
19
18 R = N-oxysuccinimidyl
e
BnO2C(CH2)5CONHC[(CH2)2C(O)NH(CH2)5C(O)NHC[(CH2)2COR]3]3
20 R = O-t- Bu
b
21 R = OH
c
(HO)2B
22 R = N-oxysuccinimidyl
f
NH(CH2)2O(CH2)2O(CH2)2N
23 R =
O
(CH2)9CNH(CH2)2NH2
N
g
24
6
h
O
NSO2
(CH2)3SO3
O
R
O
R
R
O
NH
O
Figure 3. Chemiluminescence of boronic acid probes. (A) Chemilumi-
nescence emission profiles. (B) Dilution linearity. Conditions: aliquots
(25 lL) of each sample were triggered using 200 lL of triggering
solution (0.35 N NaOH, 1.32% H2O2, 0.09% DTPA, and 2% Triton X-
100). The chemiluminescent response from each sample was integrated
over a 10s window.
NH
O
R
R
O
R
HN
HN
O
O
NH
O
O
NH
O
O
R
O
O
O
O
(CH2)9CNH(CH2)2NH
N
NH
R
O
The chemiluminescent response of the boronic acid
reagents 7, 14, 25, and 35 was next evaluated in the
presence of stabilized human glycated hemoglobin
(azido-methemoglobin, 40% glycation by vendor assay)
and Ônon-glycatedÕ hemoglobin (azido-methemoglobin,
5% glycation by vendor assay) both obtained from
American Biological Technology, Inc. (Fig. 4). The
chemiluminescent signal was recorded from samples
containing a fixed concentration of the chemilumines-
cent reagent (0.5 nM) and either glycated or non-glycat-
ed hemoglobin upon addition of basic hydrogen
peroxide. Total hemoglobin concentration was deter-
mined using a commercial kit following the manufactur-
erÕs protocol (Sigma Total Hemoglobin). Representative
raw data using the trivalent boronic acid 14 are seen in
Figure 4A.
R
O
O
NSO2
25
(CH2)3SO3
Scheme 3. Preparation of nonavalent-boronic acid chemiluminescent
probe 25. Reagents and conditions: (a) i—oxalyl chloride, CH2Cl2,
DMF; ii—BeheraÕs amine tert-butyl ester 9, Et3N, CH2Cl2, 1 h, 86%;
(b) TFA, CH2Cl2, 2 h, 72–84%; (c) N-hydroxysuccinimide, EDAC,
DMF, 14 h, 63–74%; (d) H2, Pd/C, MeOH, 100%; (e) 19, Et3N, DMF,
14 h, 79%; (f) 5, Et3N, DMF, 14 h, 86%; (g) NH2CH2CH2NH2, DMF/
buffer (pH 8), 1 h, 45%; (h) i—HBr/AcOH, 45 min, 80%;
ii—pentafluorophenyl trifluoroacetate, pyridine, DMF, 14 h, 57%;
iii—24, Et3N, DMF, 14 h, 48%.
(Scheme 3) and the 27-mer 35 (Scheme 4) proceeded
similarly, using the advanced intermediates from the
preceding schemes where possible.
The degree of signal modulation was dependent on the
total amount of hemoglobin present. Similar quenching
effects have been noted with fluorescent indicators in the
presence of hemoglobin.17 At each concentration, the
chemiluminescent signal from the glycated hemoglobin
was lower than that of the non-glycated one indicating
that upon formation of the boronate ester the signal
is further quenched. The concentration-dependent
quenching specific to the formation of the boronic
The chemiluminescence properties of each probe are
shown in Figure 3. The chemiluminescent signal gradu-
ally dropped from 4 · 1019 to 8.3 · 1018 counts/mol as
the number of boronic acids increased across the series,
while the emission profile for each displayed the ÔflashÕ
kinetics typical for the unsubstituted acridinium-9-
carboxamide.

M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1324–1328
1327
Scheme 4. Preparation of 27-valent boronic acid chemiluminescent probe 35. Reagents and conditions: (a) Et3N, DMF, 14 h; (b) TFA, CH2Cl2, 2 h,
72–85%; (c) N-hydroxysuccinimide, EDAC, DMF, 14 h, 63–72%; (d) 5, Et3N, DMF, 14 h, 75%; (e) HBr/HOAc, 2 h, 89%; (f) 6-(Cbz-amino)caproic
acid 8, HOBt, EDAC, CH2Cl2, 14 h, 72–87%; (g) H2 Pd/C, MeOH, 99%; (h) Et3N, DMF, 14 h, 50%; (i) i—HBr/HOAc, 45 min, 72%; ii—6, DMF/
buffer (pH 8), 5 h, 58%.
acid-glycated hemoglobin complex is presented in
Figure 4B.
absent. Second, the nonavalent boronic acid 25 had suf-
ficient affinity for GHb to operate in that concentration
range.
The effect of multivalency is clearly evident. The appar-
ent binding constants derived from the curves indicate a
500-fold improvement on going from the mono-boronic
acid 7 to the 27-mer 35. It is also interesting to note that
at total Hb concentrations below 10 lM, the chemilumi-
nescent signal from the weakest binder, mono-boronic
acid 7 is not quenched, while 25 and 35 just reach their
maximal signal attenuation at 10 lM. The conclusions
from this observation are twofold. First, at concentra-
tions below 10 lM Hb non-specific quenching of the
acridinium chemiluminescent signal by hemoglobin is
Testing of these observations with human whole blood
samples with established values for glycated hemoglobin
(as %HbA1c) resulted in the dose–response curve depict-
ed in Figure 5.
The five samples were first treated with potassium ferri-
cyanide (130 mol %) to convert hemoglobin to methe-
moglobin (analogous to the stabilized hemoglobin
used above) and then diluted in buffer to give a total
hemoglobin concentration of 10 lM. The data shown

1328
M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1324–1328
300000
250000
200000
150000
100000
50000
0
B
A
5% GHb "non-glycated Hb"
1
40% GHb "glycated Hb"
Kapp
0.8
0.6
0.4
0.2
0
7
211 µM
14 25 µM
25 2.1 µM
36 0.43 µM
1.95 3.90 7.80 15.6 31.2 62.5 125 250 500
0.01
0.1
1
10
100
1000 10000
Total Hb (µM)
Total Hb (µM)
Figure 4. Chemiluminescent signal attenuation by Ônon-glycatedÕ and ÔglycatedÕ hemoglobin. (A) Illustrative raw data using probe 14. (B) Apparent
affinity of probes 7, 14, 25, and 36. Conditions: non-glycated hemoglobin (ABT, Inc.) contained the normal physiological level of glycation (5%),
while the glycated samples contained a high level of modification (40%). Solutions containing the chemiluminescent boronic acids (0.5 nM) and
hemoglobin were prepared in buffer (10 mM HEPES, 2 mM EDTA, and 0.1% Tween 20, pH 7.3). Aliquots (25 lL) of each sample were triggered
(200 lL, 0.35 N NaOH, 1.32% H2O2, 0.09% DTPA, and 2% Triton X-100) and the chemiluminescent response from each sample was integrated over
a 2s window. Data points represent the average of triplicate values.
References and notes
1. Adamczyk, M.; Mattingly, P. G.; Moore, J. A.; Pan, Y.
Org. Lett. 2003, 5, 3779.
2. Adamczyk, M.; Fino, J. R.; Mattingly, P. G.; Moore, J.
A.; Pan, Y. Bioorg. Med. Chem. Lett. 2004, 14, 2313.
3. Adamczyk, M.; Johnson, D. D.; Mattingly, P. G.; Moore,
J. A.; Pan, Y. Bioorg. Med. Chem. Lett. 2004, 14,
3917.
4. Blincko, S.; Edwards, R. Clin. Lab. 2001, 47, 377.
5. Krishnamurti, U.; Steffes, M. W. Clin. Chem. 2001,
47, 1157.
6. Wilson, D. H.; Bogacz, J. P.; Forsythe, C. M.; Turk, P. J.;
Lane, T. L.; Gates, R. C.; Brandt, D. R. Clin. Chem. 1993,
39, 2090.
7. James, T. D.; Shimori, H.; Takeuchi, M.; Shinkai, S.
Chem. Commun. 1996, 705.
Figure 5. Glycated hemoglobin dose–response curve with chemilumi-
8. Hartley, J. H.; James, T. D.; Ward, C. J. J. Chem. Soc.,
nescent boronic acid probe 25. Conditions: whole blood samples
Perkin Trans. 1 2000, 3155.
(Trimar) were treated with K3Fe(CN)6 (130 mol %) and diluted to a
9. James, T. D.; Shinkai, S. Top. Curr. Chem. 2002,
final concentration of 10 lM along with 25 (0.5 nM) in buffer (10 mM
218, 159.
HEPES, 2 mM EDTA, and 0.1% Tween 20, pH 7.4). Aliquots (25 lL)
10. Wang, W.; Gao, X.; Wang, B. Curr. Org. Chem. 2002, 6,
of each sample were triggered (200 lL, 0.35 N NaOH, 1.32% H2O2,
0.09% DTPA, and 2% Triton X-100.) The chemiluminescent response
from each sample was subsequently integrated over a 2s window.
1285.
11. Wulff, G. Pure Appl. Chem. 1982, 54, 2093.
12. All new compounds were purified by reverse-phase HPLC
to greater than 95% purity and yielded the expected m/z by
electrospray mass spectrometry.
13. Lebeau, L.; Oudet, P.; Mioskowski, C. Helv. Chim. Acta
1991, 74, 1697.
14. Hawkins, R. T.; Snyder, H. R. J. Am. Chem. Soc. 1960,
82, 3863.
15. Cooper, C. R.; James, T. D. J. Chem. Soc., Perkin Trans. 1
2000, 963.
16. Newkome, G. R.; Behera, R. K.; Moorefield, C. N.;
here represent the average of triplicate reads for tripli-
cate samples (9 data points/sample). The response over
the clinically relevant %HbA1c range of 5–12% showed
good linearity (R = 0.9902) and slope, thus demonstrat-
ing the foundation for a homogeneous assay for glycat-
ed hemoglobin.
Baker, G. R. J. Org. Chem. 1991, 56, 7162.
17. Blincko, S.; Edwards, R. Ann. Clin. Biochem. 1998,
35, 140.
Further studies to expand the utility of homogeneous
chemiluminescent assays are in progress.