Hydrophilic pyrazine dyes as exogenous fluorescent tracer agents for real-time point-of-care measurement of glomerular filtration rate

Article abstract of DOI:10.1021/jm200257k

Various hydrophilic pyrazine-bis(carboxamides) derived from 3,5-diamino-pyrazine-2,5-dicarboxylic acid bearing neutral and anionic groups were prepared and evaluated for use as fluorescent glomerular filtration rate (GFR) tracer agents. Among these, the d

Full text of DOI:10.1021/jm200257k

ARTICLE
pubs.acs.org/jmc
Hydrophilic Pyrazine Dyes as Exogenous Fluorescent Tracer Agents
for Real-Time Point-of-Care Measurement of Glomerular
Filtration Rate
Raghavan Rajagopalan,* William L. Neumann,† Amruta R. Poreddy, Richard M. Fitch, John N. Freskos,
Bethel Asmelash, Kimberly R. Gaston, Karen P. Galen, Jeng-Jong Shieh, and Richard B. Dorshow
Covidien Pharmaceuticals, 675 McDonnell Boulevard, Hazelwood, Missouri 63042, United States
ABSTRACT: Various hydrophilic pyrazine-bis(carboxamides) derived from
3,5-diamino-pyrazine-2,5-dicarboxylic acid bearing neutral and anionic groups
were prepared and evaluated for use as fluorescent glomerular filtration rate
(GFR) tracer agents. Among these, the dianionic D-serine pyrazine derivatives
2d and 2j, and the neutral dihydroxypropyl 2h, exhibited favorable physico-
chemical and clearance properties. In vitro studies show that 2d, 2h, and 2j
have low plasma protein binding, a necessary condition for renal excretion. In vivo animal model results show that these three
compounds exhibit a plasma clearance equivalent to iothalamate (a commonly considered gold standard GFR agent). In addition,
these compounds have a higher urine recovery compared to iothalamate. Finally, the plasma clearance of 2d, 2h, and 2j remained
unchanged upon blockage of the tubular secretion pathway with probenecid, a necessary condition for establishment of clearance via
glomerular filtration only. Hence, 2d, 2h, and 2j are promising candidates for translation to the clinic as exogenous fluorescent tracer
agents in real-time point-of-care monitoring of GFR.
’ INTRODUCTION
has been considerable effort directed at developing fluorescent
GFR tracer agents.12À15
The assessment of renal function is a critical part of patient
care, and the accurate and real-time measurement and monitor-
ing of renal function is essential for minimizing the risk of kidney
failure due to chronic or acute clinical, physiological, and patho-
logical conditions. Glomerular filtration rate (GFR) is now widely
accepted as the best indicator of renal function, and current clinical
guidelines advocate its use in the staging of kidney disease.1
Endogenous plasma creatinine assay,2 and theoretical equations3
based on the plasma creatinine concentration, continue to be the
most common clinical method of assessing GFR despite the
known limitations.4À6 However, the GFR value obtained by
these methods is frequently misleading because it is affected by
age, state of hydration, renal perfusion, muscle mass, dietary
intake, and many other anthropometric and clinical variables.
Moreover, creatinine is partially cleared by tubular secretion
along with glomerular filtration and, as Diskin5 recently re-
marked, “Creatinine clearance is not and has never been synon-
ymous with GFR, and all of the regression analysis will not make
it so because the serum creatinine depends upon many factors
other than filtration.”
On the basis of the considerable empirical knowledge that
hydrophilic and anionic substances are preferentially cleared
through the renal system,16,17 we had originally proposed a
two-pronged approach for the rational design of fluorescent
exogenous GFR tracer agents.13 The first method involves
enhancing the fluorescence of known GFR tracer agents that
are intrinsically poor fluorescent emitters such as lanthanide
metal complexes, and the second involves transforming highly
fluorescent dyes (which are intrinsically lipophilic) into hydro-
philic, anionic species to force them to clear via the kidneys. In
the first approach, we have prepared and evaluated euro-
piumÀDTPA complexes endowed with various molecular “an-
tenna” to induce ligand-to-metal fluorescence resonance energy
transfer (FRET).13 Some success was obtained in that one of
these metal complexes exhibited a 2700-fold increase in euro-
pium fluorescence (compared to EuÀDTPA) and underwent
clearance exclusively through the kidneys. In this paper, we
present, for the first time, the successful results of developing
highly fluorescent exogenous GFR tracer agents based on the
second approach.18
The key requirements for the rational design of exogenous
fluorescent tracer agents are: (a) excitation and emission occur-
ring in the visible region (λ g ∼425 nm), (b) having highly
hydrophilic neutral or anionic character, (c) very low or no
plasma protein binding, (d) no in vivo metabolism, and (e)
The optimum measure of GFR is by the use of exogenous
tracer agents.7 However, the infrequently used conventional
GFR agents such as inulin,8 iothalamate,9,10 and 99 mTc-DTPA11
suffer from various undesirable properties (such as radioactivity,
need for ionizing radiation, laborious ex-vivo handling of blood
and urine samples, and lack of consistent and reliable supply) that
render them unsuitable for real-time point-of-care renal function
monitoring. Thus, to overcome these various limitations, there
Received: March 4, 2011
Published: June 13, 2011
r
2011 American Chemical Society
5048
dx.doi.org/10.1021/jm200257k J. Med. Chem. 2011, 54, 5048–5058
|

Journal of Medicinal Chemistry
ARTICLE
Figure 1. Pyrazine-based GFR tracer agents.
clearance exclusively via glomerular filtration (often demon-
strated by equality of plasma clearance with and without a tubular
secretion inhibitor such as probenecid19). The selection of the
lead clinical candidate(s) may be based on secondary considera-
tions such as the ease of synthesis, lack of toxicity, and stability.
The secondary screening criteria should further take into account
the tissue optics properties and the degree of extracellular
distribution of the fluorescent tracers. Volume of distribution is
an important parameter in the assessment of hydration state of
the patient, whereas the absorption/emission properties provide
essential information for the design of the probe.
Pyrazines are a class of photostable small molecules having
highly desirable photophysical properties useful for in biome-
dical applications. Pyrazine derivatives containing electron
withdrawing groups at the 2,5 positions and electron donating
groups at the 3,6 positions such as 3,6-diamino-2,5-pyrazine-
dicarboxylic acid (1) and the corresponding amides strongly
absorb and emit in the blue to orange regions with a large
Stokes shift on the order of ∼100 nm and with fluorescence
quantum yields of about 0.4.20,21 Conversion of the carboxyl
group in 1 to the secondary amide derivatives produces a
bathochromic shift of about 40 nm, and alkylation of the amino
group results in further red-shift of about 40 nm. Hence, the
pyrazine scaffold presents an attractive opportunity to “tune”
the electronic and renal clearance properties at once by
introducing hydrophilic substituents.
Table 1. Absorption, Emission, and Plasma Protein Binding
compd
λAbs (nm)
λEm (nm)
PPB (%)a
2a
436
432
435
435
436
490
486
484
486
488
NAb
558
558
559
557
558
599
600
594
597
597
NA
12
2
2b
2c
0
2d
0
2e
0
2f
74
15
6
2g
2h
2i
4
2j
0
10c
iothalamate
a Measured to (1%. b Not applicable. c Reference 23.
synthesized, and evaluated for use as GFR tracer agent candidates.
These pyrazine derivatives may be divided into two general
categories. Compounds 2aÀe bear primary amino groups and
absorb radiation in the blue region of the electromagnetic
spectrum (Table 1). Compounds 2fÀj are N-alkylated pyrazines
and absorb radiation in the green region of the spectrum.
’ RESULTS AND DISCUSSION
Syntheses. The neutral polyhydroxy derivatives 2a and 2b
were prepared by condensing the diacid 1 with protected
aminoethanol 3a or aminopropanediol 3b respectively by the
standard carbodiimide coupling method, followed by deprotec-
tion with hydrochloric acid (Scheme 1). The dianionic serine
derivatives 2c and 2d were prepared by coupling the diacid 1 with
L- or D-serine benzyl ester hydrochloride (3c or 3d), followed by
hydrogenolysis in the presence of 10%PdÀC as catalyst at
atmospheric pressure. The tetraanionic aspartate derivative 2e
was prepared by coupling the diacid 1 with dibenzyl aspartate 3e,
followed by transfer hydrogenation with ammonium formate and
10% PdÀC.
The pyrazine-dicarboxylic acid (1) and its known simple
carboxamide analogues are lipophilic and are insoluble in water
and, therefore, unsuitable as possible GFR tracer agents in this
form. However, the diacid 1 is an excellent scaffold for further
modification, and several hydrophilic amides bearing various
neutral and anionic substituents (Figure 1) were designed,
Alkyl substitution on the amino groups of the pyrazine scaffold
1 was known to significantly increase the wavelength of both
absorption and emission maxima, but the existing base-induced
alkylation methods resulted in rather poor yields of the desired
5049
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
Scheme 1a
a Reagents: (a) Et3N, EDC HCl, HOBt H2O, DMF; (b) 4 N HCl in dioxane; (c) EDC HCl, HOBt H2O, DMF; (d) 1 N HCl, THF; (e) DIPEA,
3
3
3
3
EDC HCl, HOBt H2O, DMF; (f) H2, 10% PdÀC, EtOHÀH2O (3:1, v/v); (g) HCO2NH4, 10% PdÀC, MeOHÀH2O (2:1, v/v), 60 °C. Note: the
3
3
percentages shown under each arrow refer to reaction yields.
products. Consequently, a convenient and robust reductive
amination method was developed for the synthesis of N-alkylated
pyrazines,22 which was adopted for the preparation of the analogues
2fÀj (Scheme 2). The neutral N-alkyalted pyrazine derivative 2g was
prepared by the reductive amination of D-glyceraldehyde acetonide
(7a) with 3,6-diamino-2,5-dicarboxamide (6), followed by de-
protection of the isopropylidene group with aqueous HCl. The
bis(propyl) derivative 2f, was prepared by reductive amination of
propionaldehyde with bis-benzyl ester 4d, followed by subse-
quent transfer hydrogenation of the intermediate 8a. The neutral
2h and 2i were made available by the initial reductive alkylation
of the bis-acetonide 4b with 7a or L-glyceraldehyde acetonide
(7b), respectively, followed by acidic hydrolysis of the acetonide
groups. Finally, a similar reductive alkylation of 4d with 7a
afforded 8e, which was then treated with aqueous HCl to remove
the acetonide groups followed by base hydrolysis of benzyl esters
leading to the dianionic compound 2j.
compounds prior to in vivo evaluation (Table 1). Compounds
2aÀe exhibited absorption maxima in the range of 430À440 nm
and emission maxima in the range of 550À560 nm. Compounds
2fÀj exhibited absorption maxima in the range of 480À490 nm
and emission maxima in the range of 590À600 nm. The
fluorescence quantum yield of 2c in DMSO was measured to
be 0.42. With the exception of 2f and 2g, all compounds exhibited
low plasma protein binding (PPB), comparable to or lower (i.e.,
better) than iothalamate. As would be expected, the introduction
of hydrophobic propyl group in 2f resulted in high PPB.
Surprisingly, the two regioisomeric compounds 2b and 2g
exhibited measurably different PPB. It is possible that the
primary carboxamide, which contains an additional site for
hydrogen bonding at the amide site in 2g is contributing to the
higher interaction with the protein.
In Vivo Studies. All in vivo studies were performed in the
SpragueÀDawley rat model. Plasma concentrations of the tracer
agents were determined by HPLC at 0, 2, 6, and 24 h post
administration of the injected dose. Urine was collected through
In Vitro Studies. Absorption(λabs) andemission(λem) maxima
and plasma protein binding (PPB) were determined for all
5050
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
Scheme 2a
a Reagents: (a) Na(OAc)3BH, HOAC, DCE; (b) HCO2NH4, 10% PdÀC, MeOHÀH2O (2:1, v/v), 60 °C; (c) 1 N HCl, THF; (d) (i) c, (ii) 1 N NaOH.
Note: the percentages shown under each arrow refer to reaction yields.
6 h post administration of the injected dose. Plasma clearance
profiles and percent injected dose (%ID) recovered in the urine
are given in Table 2. Among the neutral compounds, polyhy-
droxy derivatives 2h and 2i exhibit the highest %ID recovery
from urine. Among the charged compounds, the dianionic serine
derivatives 2d and 2j exhibit the highest %ID recovery from
urine. The tetraanionic compound 2e demonstrated the least %
ID recovery from urine among the anionic compounds. This is
consistent with the previous observation that optimal renal
clearance is observed for those compounds having two negative
charges, and that any increase in the charge adversely impacts
renal clearance.24,25 Additionally, an HPLC analysis of the
compounds 2c, 2d, 2h, and 2j prior to in vivo administration
and after recovery from urine did not reveal any appreciable
amounts of degradation products. Because of the substantially
lower amounts of 2b, 2e, and 2g recovered in the urine with
respect to the other pyrazine derivatives, plasma clearance studies
of these derivatives were not pursued.
To ascertain whether any of the compounds had clearance
affected by renal tubular secretion (a negative result being a
necessary condition for pure GFR clearance), renal proximal
tubule organic anion transport inhibition was next investigated.
The in vivo clearance of four selected pyrazine derivatives was
measured with and without probenecid (Figure 2). Iothalamate
was included in this study as a comparator. In addition, 99 mTc-
MAG3, a radioscintigraphic imaging agent that is known to clear
5051
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
via the tubular secretion pathway,19,25 was employed as a positive
control.
Probenecid significantly decreased the clearance rate of
the 60 min time point verified that the highly fluorescent spot in
the abdomen was the bladder.
99
To follow the time dependence of the fluorescence quantita-
tively, an arbitrary region of interest (ROI) was drawn encom-
passing the area around the bladder and an approximately equal
area was drawn near the shoulder on the series of optical images
of the mouse injected with compound 2h from 0 to 70 min. As a
function of time, the fluorescence signal is seen to monotonically
increase over the bladder region as expected (Figure 4). Over the
shoulder region, this signal is seen to increase post administration
of tracer agent and then wash out as a function of time as it is
removed from the body by the renal system. This fluorescence
decease is consistent with the decrease in plasma concentration
of 2h as indicated in Table 2.
mTc-MAG3 (p = 0.001)26 as expected from a non-GFR agent.
Probenecid did not significantly affect the clearance rate of
iothalamate (p > 0.05), also as expected from a GFR agent.
Compounds 2h and 2j did not exhibit differences in clearance
rates (p > 0.05). However, a statistically significant difference
(p < 0.05) was detected in the clearance rate of the L-serine
enantiomer 2c with respect to the presence or absence of
probenecid, whereas the D-serine enantiomer 2d did not display
a significant difference (p > 0.05). Therefore, renal tubular
secretion is not a significant elimination pathway for compounds
2d, 2h, and 2j in the rat.
Noninvasive real-time fluorescence monitoring of the clear-
ance of compound 2h is shown in Figure 3.27 Each panel contains
images of two mice. The mouse on the left was administered
300 μL of a 2 mM solution (in PBS) of compound 2h. The
mouse on the right received 200 μL of PBS only. The two
administrations were given simultaneously. Compound 2h is
seen to distribute quickly throughout the body and then clears
from the body and concentrates in the abdomen. Surgery after
’ CONCLUSION
On the basis of the fluorescence properties, plasma protein
binding data, the injected dose recovered in urine, the plasma
clearance data, and the renal tubular secretion studies, the
pyrazine serine derivatives 2d, 2h, and 2j are viable candidates
for translation to the clinic as exogenous fluorescent tracer agents
for the measurement of GFR. In the rat animal model, these
compounds display superior properties compared to iothala-
mate, which is currently an accepted standard for the measure-
ment of GFR. Formal preclinical development studies are in
progress and will be reported elsewhere.
Table 2. %ID Recovery in Urine, and Plasma Clearance Half-
Livesa
%ID recovered
in urine at 6 h
plasma half-lifeb
(min)
’ EXPERIMENTAL SECTION
compd
2b
61 ( 3 (3)c
82 ( 7 (3)
90 ( 1 (3)
61 ( 16 (3)
25 ( 4 (3)
88 ( 2 (3)
87 ( 4 (3)
85 ( 2 (3)
80 ( 2 (6)
NDd
36 ( 6 (8)
29 ( 1 (6)
ND
Chemistry. Unless otherwise noted, all solvents and reagents were
used as supplied. Organic extracts were dried over either anhyd Na2SO4
or anhyd MgSO4 and filtered using a fluted filter paper (P8) or a fritted
glass funnel. Solvents were removed on a rotary evaporator under
reduced pressure. Analytical TLC was performed on Analtech silica
gel GF plates (250 μm), and the components were visualized by UV
light. Flash chromatography was carried out either using EMD silica gel
60 (40À63 μm) or C18 silica gel (YMC ODS-A 120 Å I 25/40) in glass
columns. Automated flash chromatography was carried out on a Teldyne
Isco CombiFlash Rf system using prepacked silica gel columns. RP-LC/
MS (ESI, positive ion mode) analyses were carried out on either a BDS
Hypersil C18 3 μm (50 mm  4.6 mm) or a ThermoElectron Hypersil
Gold C18 3 μm (4.6 mm  50 mm) column. Compounds were injected
2c
2d
2e
2g
ND
2h
19 ( 1 (6)
ND
2i
2j
20 ( 1 (5)
32 ( 2 (4)
iothalamate
a The values are given as mean ( SEM (standard error of mean).
b Terminal phase half-life from two compartment modeling. c Numbers
in parentheses indicate number of test animals. d Not determined.
Figure 2. Clearance of 99 mTc-MAG3 with and without probenecid, and clearance of several pyrazine compounds and iothalamate comparator with and
without probenecid.
5052
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
Figure 4. Fluorescence measured over the bladder (dashed line) and
shoulder region (solid line) as a function of time (data was normalized to
peak value of each ROI).
spectra. Coupling constants (J) are reported in Hz. HRMS (ESI) data
was obtained on a ThermoFisher LTQ-Orbitrap mass spectrometer
equipped with an IonMax electrospray ionization source in FTMS mode
with resolution g30K. Elemental analyses were carried out by Atlantic
Microlab, Inc., Norcross, GA.
3,6-Diamino-N2,N5-bis(2-tert-butoxyethyl)pyrazine-2,5-
dicarboxamide (4a). A 100 mL round-bottom flask equipped with
a magnetic stir bar was charged with diacid 1 (0.420 g, 2.10 mmol),
2-tert-butoxyethylamine HCl (3a, 0.717 g, 4.66 mmol), EDC HCl
3
3
(0.894 g, 4.66 mmol), HOBt H2O (0.630 g, 4.66 mmol), and Et3N
3
(0.630 g, 6.36 mmol) in anhyd DMF (40 mL). The reaction mixture
was stirred overnight at rt and concentrated in vacuo to ∼10 mL
solution that was partitioned between EtOAc (125 mL) and satd
NaHCO3 (100 mL). The organic solution was washed with 10%
aqueous citric acid (100 mL) and brine (50 mL). The solution was
concentrated and purified via flash chromatography over silica gel
using EtOAcÀhexanes (2:1, v/v) as eluent to give 4a (0.290 g, 35%).
1H NMR (DMSO-d6) δ 8.30 (t, J = 5.9 Hz, 2 H), 6.57 (s, 4H), 3.42 (dt,
J = 5.3 Hz, 4 H), 3.36 (dt, J = 5.9 Hz, 4 H), 1.14 (s, 18 H). 13C NMR
(DMSO-d6) δ 164.8, 146.1, 126.1, 72.5, 59.7, 39.5, 27.2. HRMS (ESI)
m/z calcd for C18H33N6O4 (M + H)+ 397.2558, found 397.2558; calcd
for C18H32N6O4Na (M + Na)+ 419.2377, found 419.2376.
Figure 3. In vivo time-dependent fluorescence detection (with pixel
intensity range bar graph). The mouse on the left was treated with
compound 2h, and the mouse on the right was treated with PBS; (a)
preadministration, (b) 10 min post administration, (c) 20 min post
administration, (d) 30 min post administration, (e) 60 min post
administration. The ratio between most intense and least intense
fluorescence in the 60 min post administration image was approximately
a factor of 100.
3,6-Diamino-N2,N5-bis(2-hydroxyethyl)pyrazine-2,5-di-
carboxamide (2a). A 100 mL round-bottom flask equipped with a
magnetic stir bar was charged with the diether 4a (0.233 g, 0.588
mmol) and 4 N HClÀdioxane (8 mL) and stirred at rt overnight. The
reaction was concentrated in vacuo, and the crude product (∼0.170 g)
was purified via RP-HPLC (5À50%B). The product containing
fractions were concentrated in vacuo and vacuum-dried to afford 2a
(0.082 g, 49%) as an orange solid. 1H NMR (DMSO-d6) δ 8.30 (t J =
5.9 Hz, 2H), 7.60 (s, 4 H), 4.83 (t, J = 5.9 Hz, 2 H) 3.45 (dt, J = 5.4 Hz,
4 H), 3.35 (dt, J = 5.9 Hz, 4 H). 13C NMR (DMSO-d6) δ 165.4, 146.1,
126.7, 60.0, 41.8. HRMS (ESI) m/z calcd for C10H17N6O4 (M + H)+
285.1306, found 285.1307; calcd for C10H16N6O4Na (M + Na)+
307.1125, found 307.1126.
using a gradient condition (5 to 50À95%B/6À10 min) with a flow rate
of 1 mL/min (mobile phase A, 0.05% TFA in H2O; mobile phase B,
0.05% TFA in CH3CN). Preparative RP-HPLC was carried out using a
Waters XBridge Prep C18 5 μm OBD 30 mm  150 mm or 19  250
mm column [λmax, PDA (200À800 nm); flow, 50 mL/min; gradient,
5À50% B/10À17 min; mobile phase A, 0.1% TFA in H2O; mobile
phase B, 0.1% TFA in CH3CN]. RP-HPLC analyses were carried out
using a Phenomenex Luna 5 μm C18(2) 100 Å 250 mm  4.6 mm
column [detection, UV; flow, 1 mL/min; gradient, 5/10% B to 50/90%
B/20 min; mobile phase A, 0.1% TFA in H2O; mobile phase B, 0.1%
TFA in CH3CN)], and the chromatographic purities of all the com-
pounds were >95%. UV/vis and fluorescence spectra were measured on
a Shimadzu UV-3101 PC and Jobin Yvon Fluorolog-3 spectrometers,
respectively. NMR spectra were recorded on either a Varian VNMRS-
500 spectrometer. 1H Chemical shifts are expressed in parts per million
(δ) relative to TMS (δ = 0) as an internal standard. 13C Chemical shifts
are referenced to either TMS (δ = 0) or the residual solvent peaks in the
3,6-Diamino-N2,N5-bis[(2,2-dimethyl-1,3-dioxolan-4-yl)-
methyl]pyrazine-2,5-dicarboxamide (4b). A mixture of diacid 1
(0.350 g, 1.77 mmol), 2,2-dimethyl-1,3-dioxolane-4-methanamine (3b,
0.933 mL, 7.20 mmol), HOBt H2O (0.812 g, 5.30 mmol), and EDC
3
3
HCl (1.02 g, 5.32 mmol) were stirred together in DMF (20 mL) for 16 h
5053
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
at rt. The mixture was concentrated to dryness, and the residue was
partitioned between EtOAc and water. The layers were separated, and
the EtOAc solution was washed with satd NaHCO3 and brine. The
EtOAc solution was concentrated to afford the diastereomeric bis-amide
4b (0.665 g, 88%) as a yellow solid. 1H NMR (CDCl3) δ 8.38 (t, J = 5.8
Hz, 2 H), 6.55 (s, 4 H), 4.21 (quintet, J = 5.8 Hz, 2 H), 3.98 (dd, J = 8.4
Hz, 6.3 Hz, 2 H), 3.65 (dd, J = 8.4 Hz, J = 5.8 Hz, 2 H), 3.39 (apparent
quartetÀdiastereotopic mixture, J = 5.9 Hz, 4 H), 1.35 (s, 6 H), 1.26 (s,
6 H). 13C NMR (CDCl3) δ 165.7, 146.8, 126.8, 109.2, 74.8, 67.2, 42.2,
41.1, 27.6, 26.1. RP-LC/MS (ESI) m/z 425.4 (M + H)+ (tR = 4.00 min,
5À95%B/6 min). HRMS (ESI) m/z calcd for C18H28N6O6Na (M +
Na)+ 447.1963, found 447.1960. Anal. Calcd for C18H28N6O6: C, 50.93;
H, 6.65; N, 19.80. Found: C, 51.35; H, 7.12; N, 19.59.
3,6-Diamino-N2,N5-bis[(S)-1,3-dihydroxy-1-oxopropan-2-
yl]pyrazine-2,5-dicarboxamide (2c). The bis-amide 4c (3.87 g,
7.00 mmol) was weighed into a 500 mL round-bottom flask equipped
with a Claisen adapter and anhyd EtOH (225 mL), and milli-Q H2O
(50 mL) were added. The resulting orange suspension (partially
soluble) was briefly sonicated and 10% PdÀC (0.39 g) was added as a
slurry in H2O (25 mL). The reaction mixture was thoroughly purged
first with argon and then with H2 and stirred under H2 atmosphere (slow
bubbling) at rt for 5.5 h. The reaction mixture was once again purged
with Ar, and the catalyst was removed by filtration over a Celite bed. The
bed was thoroughly washed with EtOHÀH2O (400 mL; 1:1, v/v), and
the combined filtrates were concentrated in vacuo and further dried
under high vacuum. The resulting orange residue (2.45 g) was triturated
with CH3CN (100 mL; sonicated briefly to get the material off the walls
of the flask) to give 2c (2.29 g, 88%) as a red powder. 1H NMR (DMSO-
d6) δ 8.46 (d, J = 8.3 Hz, 2 H, exchangeable with D2O), 6.78 (br s, 4 H,
exchangeable with D2O), 4.48À4.45 (dt, J = 8.3, 3.9 Hz, 2 H; collapses to
a triplet at δ = 4.47 ppm with J = 3.7 Hz upon D2O exchange), 3.88 (dd,
J = 10.9, 3.8 Hz, 2 H), 3.74 (dd, J = 10.9, 3.6 Hz, 2 H). 13C NMR
(DMSO-d6) δ 171.5, 164.6, 146.3, 125.9, 61.1, 54.2. RP-LC/MS (ESI)
m/z 373.3 (M + H)+ (tR = 2.62 min, 5À95%B/6 min). Anal. Calcd
for C12H16N6O8: C, 38.71; H, 4.33; N, 22.57. Found: C, 38.67; H, 4.50;
N, 22.27.
3,6-Diamino-N2,N5-bis(2,3-dihydroxypropyl)pyrazine-2,5-
dicarboxamide (2b). The above bis-amide 4b was dissolved in THF
(100 mL) and treated with 1.0 N HCl (2 mL). After hydrolysis was
complete, the mixture was treated with K2CO3 (1 g) and stirred for 1 h
and filtered through a plug of C18 silica gel using methanol. The filtrate
was concentrated to dryness and the residue was triturated with MeOH
(50 mL). The solids were filtered and discarded, and the residue was
treated with ether (50 mL). The precipitate was collected by filtration
and dried under high vacuum. This material was purified by radial flash
chromatography using CHCl3ÀMeOH (19:1 to 1:1, v/v) to afford 2b
(0.221 g, 36%) as an orange solid. 1H NMR (DMSO-d6) δ 8.00 (br m,
6 H), 5.39 (br s, 2 H), 4.88 (br s, 2 H), 3.63À3.71 (complex m, 2 H),
3.40 (dd, J = 11.1, 5.1 Hz, 2 H), 3.28 (dd, J = 11.1, 6.6 Hz, 2 H), 2.92 (dd,
J = 12.6, 3.3 Hz, 2 H), 2.65 (dd, J = 12.6, 8.4 Hz, 2 H). RP-LC/MS (ESI)
m/z 345 (M + H)+ (tR = 4.13 min, 5À95% gradient acetonitrile in 0.1%
TFA over 10 min on a 30 mm column). HRMS (ESI) m/z calcd for
C12H20N6O6Na (M + Na)+ 367.1337, found 367.1336.
3,6-Diamino-N2,N5-bis[(R)-1-(benzyloxy)-3-hydroxy-1-ox-
opropan-2-yl]pyrazine-2,5-dicarboxamide (4d). The reaction
of diacid 1 (9.91 g, 50.0 mmol) with D-serine benzyl ester hydrochloride
(3d, 24.33 g, 105.0 mmol) in the presence of DIPEA (19.16 mL,
110.0 mmol), HOBt H2O (17.61 g, 115.0 mmol), and EDC HCl
3
3
(22.05 g, 115.0 mmol) in DMF (300 mL) was carried out overnight
(ca. 14 h) as described for the preparation of 4c. The dark solution was
concentrated to a syrupy residue under high vacuum (bath temp 60 °C)
that was partitioned between EtOAc and milli-Q H2O (400 mL each).
The layers were separated and the aqueous layer was further extracted
with EtOAc (3 Â 200 mL), and the combined EtOAc extracts were
successively washed with 0.50 M KHSO4, saturated NaHCO3, H2O, and
brine (250 mL each). Removal of the solvent left an orange slush that
was dried under high vacuum overnight to give 23.7 g of an orange solid.
The crude product was subjected to flash chromatography over silica gel
using CHCl3 to CHCl3ÀMeOH (97:3, v/v) as eluant in a gradient
fashion (sample adsorbed on silica gel was loaded on the column; carried
out in 4 runs) to give the bis-amide 4d (19.6 g, 71%) as an orange solid:
Rf 0.45 [CHCl3ÀMeOH (9:1, v/v)]. 1H NMR (DMSO-d6) δ 8.56 (d,
J = 8.0 Hz, 2 H, exchangeable with D2O), 7.40À7.33 (m, 10 H), 6.76 (s,
4 H, exchangeable with D2O), 5.37 (t, J = 5.5 Hz, 2 H), 5.20 (distorted
AB pair, 4 H), 4.66À4.63 (dt, J = 8.0, 4.0 Hz, 2 H), 3.97À3.93 (m, 2 H),
3.81À3.77 (m, 2 H). 13C NMR (DMSO-d6) δ 170.1, 164.9, 146.4,
135.8, 128.4, 128.0, 127.6, 125.9, 66.2, 61.1, 54.4. RP-LC/MS (ESI) m/z
553.3 (M + H)+ (tR = 4.44 min, 5À95%B/6 min). Anal. Calcd for
C26H28N6O8: C, 56.52; H, 5.11; N, 15.21. Found: C, 56.39; H, 5.11;
N, 14.99.
3,6-Diamino-N2,N5-bis[(S)-1-(benzyloxy)-3-hydroxy-1-ox-
opropan-2-yl]pyrazine-2,5-dicarboxamide (4c). A 250 mL
round-bottom flask equipped with a Claisen adapter and an addition
funnel was charged with L-serine benzyl ester hydrochloride (3c, 4.87 g,
21.0 mmol) and anhydrous DMF (100 mL) was cannulated into it. The
resulting colorless solution was cooled in an ice-bath and stirred for 15
min under N2 atmosphere. Then DIPEA (3.83 mL, 22.0 mmol) was
added dropwise via addition funnel over a 30 min period, and after 30
min, the cooling bath was removed and diacid 1 (1.98 g, 10.0 mmol) was
added in one portion. The brick-red suspension was allowed to stir for 30
min before the addition of HOBt H2O (3.37 g, 22.0 mmol) in one
3
portion. After another 15 min, the reaction flask was once again cooled in
an ice-bath, and EDC HCl (4.22 g, 22.0 mmol) was added in portions
3
over a 15 min period. The resulting somewhat lighter and brown
suspension was slowly allowed to warm to rt and stirred overnight
(ca. 17 h) under N2. The reaction mixture was diluted with EtOAc
(500 mL) and transferred to a 1 L separatory funnel. The solution was
washed with milli-Q H2O (150 mL), and the aq phase was further
extracted with EtOAc (100 mL). The combined organic extracts were
successively washed with 150 mL portions of water, 0.50 M KHSO4,
water, satd NaHCO3 (Â2), water, and brine. Removal of the solvent
followed by drying under high vacuum gave 4.75 g of the crude product
as an orange fluffy solid, which upon flash chromatography over silica gel
using CHCl3MeOH (19:1, v/v) as eluant (sample adsorbed on silica gel
was loaded on the column), afforded the bis-amide 4c (4.34 g, 79%) as
an orange solid: Rf 0.40 [CHCl3ÀMeOH (9:1, v/v)]. 1H NMR
(DMSO-d6)) δ 8.56 (d, J = 8.3 Hz, 2 H, exchangeable with D2O),
7.40À7.31 (m, 10 H), 6.76 (s, 4 H, exchangeable with D2O), 5.37 (t, J =
5.6 Hz 2 H), 5.20 (distorted AB pair, 4 H), 4.67À4.64 (dt, J = 8.3, 3.8 Hz,
2 H), 3.97À3.93 (m, 2 H), 3.81À3.77 (m, 2 H). 13C NMR (DMSO-d6)
δ 170.1, 164.9, 146.4, 135.8, 128.4, 128.0, 127.6, 125.9, 66.2, 61.1, 54.4.
RP-LC/MS (ESI) m/z 553.3 (M + H)+ (tR = 4.24 min, 15À95%B/
6 min). Anal. Calcd for C26H28N6O8: C, 56.52; H, 5.11; N, 15.21.
Found: C, 56.53; H, 5.05; N, 15.16.
3,6-Diamino-N2,N5-bis[(R)-1,3-dihydroxy-1-oxopropan-2-
yl]pyrazine-2,5-dicarboxamide (2d). The bis-amide 4d (7.74 g,
14.0 mmol) was hydrogenated in the presence of 10% PdÀC (0.774 g)
in EtOHÀH2O (560 mL; 3:1, v/v) for 5 h as described for the
preparation of 2c. After a similar work up and isolation procedure, 2d
(4.89 g, 94%) was obtained as an orange powder. 1H NMR (DMSO-d6)
δ 8.46 (d, J = 8.3 Hz, 2 H, exchangeable with D2O), 6.78 (br s, 4 H,
exchangeable with D2O), 4.48À4.45 (dt, J = 8.1, 3.9 Hz, 2 H; collapses to
a triplet at δ = 4.47 ppm with J = 3.7 Hz upon D2O exchange), 3.88 (dd,
J = 11.1, 3.9 Hz, 2 H), 3.74 (dd, J = 11.1, 3.7 Hz, 2 H). 13C NMR
(DMSO-d6) δ 171.6, 164.7, 146.4, 125.9, 61.2, 54.3. RP-LC/MS (ESI)
m/z 373.2 (M + H)+ (tR = 2.86 min, 5À95% B/6 min). Anal. Calcd
for C12H16N6O8: C, 38.71; H, 4.33; N, 22.57. Found: C, 38.44; H,
4.51: N, 22.33.
5054
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
3,6-Diamino-N2,N5-bis[(R)-1,2-(dibenzyloxycarbonyl)ethyl]-
pyrazine-2,5-dicarbox-amide (4e). A mixture of sodium salt of
(Sunfire Prep C18 OBD 5 μm 50 mm  10 mm, 5À95%B/13 min) to
give 2f (0.356 g, 71%) as dark-red powder. 1H NMR (DMSO-d6) δ 8.54
(d, J = 7.8 Hz, 2 H), 4.94À4.87 (m, 4 H), 4.77 (dd, J = 10.8, 5.8 Hz, 2 H),
4.45À4.42 (dt, J = 8.1, 3.8 Hz, 2 H), 3.89 (dd, J = 11.0, 3.7 Hz, 2 H), 3.79
(dd, J = 11.0, 3.4 Hz, 2 H), 3.40À3.31 (m, 4 H), 1.61À1.56 (m, 4 H),
0.94 (t, J = 7.4 Hz, 6 H). 13C NMR (DMSO-d6) δ 172.0, 165.5, 145.9,
126.1, 61.4, 54.7, 42.7, 22.8, 11.9. RP-LC/MS (ESI) m/z 457.4 (M +
H)+ (tR = 3.77 min, 5À95% B/6 min). HRMS (ESI) m/z calcd for
C18H29N6O8 (M + H)+ 457.2041, found 457.2044.
3,6-Bis{[(S)-2,2-dimethyl-1,3-dioxolan-4-yl]methylamino}-
pyrazine-2,5-dicarboxamide (8b). The reaction of 3,6-diamino-
pyrazine-2,5-dicarboxamide (6, 0.196 g, 1.00 mmol) with (R)-(+)-2,2-
dimethyl-1,3-dioxolane-4-carbaldehyde (7a, 0.390 g, 3.00 mmol) in the
presence of HOAc (0.180 g, 3.00 mmol) and Na(OAc)3BH (0.633 g,
3.00 mmol) in anhyd DCE (10 mL) was carried out overnight as
described for the preparation of 8a. A similar work up afforded 8b (0.36
g, 85%) as a red solid, which was used without further purification. 1H
NMR (CDCl3) δ 7.95 (br t, 2 H), 7.62 (br, 2 H), 5.60 (br, 2 H), 4.37 (m,
21 H), 4.12 (m, 2 H), 3.81 (m, 2 H), 3.58 (m, 4 H), 1.55 (s, 6 H), 1.38 (s,
6 H). 13C NMR (CDCl3) δ 168.1, 146.1, 126.1, 109.4, 74.9, 67.3, 43.4,
27.1, 25.5. RP-LC/MS (ESI) m/z 425.4 (M + H)+.
3,6-Bis[(S)-2,3-dihydroxypropylamino]pyrazine-2,5-di-
carboxamide (2g). The bis-acetonide 8b (0.36 g, 0.85 mmol) was
dissolved in THF (5 mL) and treated with 1 N HCl (1 mL). The
reaction mixture was stirred overnight, neutralized with 1 N NaOH
(1 mL), and evaporated to dryness. The crude product was subjected to
flash chromatography over C18 silica gel using water as eluent to afford
2g (0.200 g, 68%). 1H NMR (DMSO-d6) δ 7.91 (br t, 2 H), 7.60 (br,
2 H), 4.75 (br, 2 H), 4.60 (m, 2 H), 3.60 (m, 4 H), 3.20 (m, 4 H). 13C
NMR (DMSO-d6) δ 167.9, 145.5, 125.8, 70.4, 63.8, 43.6. HRMS (ESI)
m/z calcd for C12H21N6O6 (M + H)+ 345.1517, found 345.1457.
N2,N5-Bis[(2,2-dimethyl-1,3-dioxolan-4-yl)methyl]-3,6-
bis{[(S)-2,2-dimethyl-1,3-dioxolan-4-yl]methylamino}pyra-
zine-2,5-dicarboxamide (8c). The reaction of bis-acetonide 4b
(1.72 g, 4.05 mmol) with aldehyde 7a (2.64 g, 20.3 mmol) in the
presence of HOAc (1.13 mL, 19.6 mmol) and Na(OAc)3BH (4.18 g,
19.7 mmol) in anhyd DCE (50 mL) was carried out overnight (ca. 22 h)
as described for the preparation of 8a. The reaction was found to be
incomplete (RP-LC/MS) at this stage and subsequently treated with
additional aldehyde 7a (0.70 g, 5.38 mmol), HOAc (0.300 mL, 5.20
mmol), and Na(OAc)3BH (1.10 g, 5.19 mmol) and continued over-
night. After a similar work up described in the prepaaration of 8a, the
crude product (4.30 g) obtained was subjected to automated flash
chromatography [column: RediSep Rf 330 g silica gel; λmax, 280 nm;
flow, 80 mL/min; gradient, 0À2% B/40 min (mobile phase A, CHCl3;
mobile phase B, MeOH)] to afford 8c (1.90 g, 72%) as an orange
powder. 1H NMR (CDCl3) δ 8.10 (t, J = 6.0 Hz, 2 H), 8.06 (t, J = 6.0 Hz,
2 H), 4.36À4.31 (m, 4 H), 4.09À4.06 (m, 4 H), 3.79À3.76 (m, 2 H),
3.74À3.67 (m, 2 H), 3.66À3.52 (m, 8 H), 1.49 (s, 3 H), 1.48 (s, 3 H),
1.45 (s, 6 H), 1.37 (s, 6 H), 1.36 (s, 6 H). RP-LC/MS (ESI) m/z 653.3
(M + H)+ (tR = 4.83 min, 5À95% B/6 min). Anal. Calcd for
C30H48N6O10: C, 55.20; H, 7.41; N, 12.88. Found: C, 55.14; H, 7.37;
N, 12.60.
diacid 1 (0.600 g, 2.48 mmol), D-Asp(OBn)-OBn p-TsOH salt (3e,
3
2.43 g, 5.00 mmol), HOBt H2O (0.919 g, 6.00 mmol) and EDC HCl
3
3
(1.14 g, 5.95 mmol) in DMF (50 mL) was treated with Et3N (4 mL).
The resulting mixture was stirred overnight at rt The reaction mixture
was concentrated, and the residue was partitioned between water and
EtOAc. The EtOAc layer was separated and washed successively with
satd sodium bicarbonate, water, and brine. The EtOAc solution was
concentrated, and the residue was purified by flash chromatography
[SiO2, CHCl3ÀMeOH (50/1 to 10/1, v/v] to afford the bis-amide 4e
(1.15 g, 58% yield) as a yellow foam. 1H NMR (CDCl3) δ 8.61 (d, J =
8.4 Hz, 2 H), 7.29À7.39 (m, 20 H), 5.85 (br s, 4 H), 5.22 (ABq, J = 10.0
Hz, ΔνAB = 17.3 Hz, 4 H), 5.10 (ABq, J = 12.2 Hz, ΔνAB = 34.3 Hz, 4 H),
5.06À5.09 (m, 2 H), 3.11 (ABq of d, J = 17.0, 5.1 Hz, ΔνAB = 77.9 Hz,
4 H). 13C NMR (CDCl3) δ 170.7, 170.7, 165.4, 147.0, 135.7, 135.6,
129.0, 128.9, 128.8, 128.75, 128.7, 126.9, 68.0, 67.3, 49.1, 37.0. RP-LC/
MS (ESI) m/z 789 (M + H)+ (tR = 5.97 min, 50À95% gradient
acetonitrile in 0.1% TFA over 10 min on a 250 mm column).
3,6-Diamino-N2,N5-bis[(R)-1,2-(dicarboxy)ethyl]pyrazine-
2,5-dicarboxamide (2e). To a well stirred solution of tetra-benzyl
ester 4e (0.510 g, 0.650 mmol) in THF (20 mL) and water (10 mL) were
added 10% Pd/C (0.50 g) and ammonium formate (1.00 g, 15.9 mmol).
The resulting mixture was heated to 60 °C for 2 h and allowed to cool
to rt. The mixture was filtered through Celite and concentrated. The
resulting material was purified by reverse phase medium pressure
chromatography (C18, 10À70% manual gradient acetonitrile in 0.1%
TFA) to afford 2e (0.138 g, 54%) as an orange solid. 1H NMR (DMSO-
d6) δ 8.62 (d, J = 8.4 Hz, 2 H), 6.67 (br s, 4 H), 4.73 (dt, J = 8.4, 5.4 Hz,
2 H), 2.74À2.88 (complex m, 4 H). 13C NMR (DMSO-d6) δ 172.6,
165.2, 147.0, 126.6, 60.8, 49.1. RP-LC/MS (ESI) m/z 429 (M + H)+
(tR = 4.01 min, 5À95% gradient acetonitrile in 0.1% TFA over 10 min on
a 250 mm column). HRMS (ESI) m/z calcd for C14H14N6O10 (M À
2 H)2À 213.0391, found 213.0388.
N2,N5-Bis[(R)-1-(benzyloxy)-3-hydroxy-1-oxopropan-2-
yl]-3,6-bis(propylamino)pyrazine-2,5-dicarboxamide (8a).
To a solution of the bis-benzyl ester 4d (0.890 g, 1.61 mmol) and
propionaldehyde (5, 0.460 mL, 6.44 mmol) in anhyd DCE (65 mL) was
added HOAc (0.368 mL, 6.44 mmol) with stirring at 0 °C under argon
atmosphere. Then Na(OAc)3BH (1.36 g, 6.44 mmol) was introduced in
small portions and the reddish suspension was slowly allowed to warm to
rt and stirred overnight (ca. 24 h) under argon. The reaction was
quenched by a slow addition of satd NaHCO3 at 0 °C and extracted with
CHCl3 (150 mL). The CHCl3 extracts were successively washed with
H2O and brine (50 mL each). Removal of the solvent gave 1.04 g of a red
solid, which upon flash chromatography over silica gel [CHCl31ÀMeOH
(39:1, v/v)] afforded 8a (0.800 g, 78%) as a dark-red solid. H NMR
(CDCl3) δ 8.68 (d, J = 7.4 Hz, 2 H), 7.60 (t, J = 5.6 Hz, 2 H), 7.37À7.33
(m, 10 H), 5.26 (s, 4 H), 4.78À4.75 (dt, J = 7.3, 3.7 Hz, 2 H), 4.08 (dd,
J = 6.1, 3.9 Hz, 4 H), 3.39À3.28 (m, 4 H), 2.64 (t, J = 6.2 Hz, 2 H),
1.65À1.61 (m, 4 H), 0.97 (t, J = 7.4 Hz, 6 H). 13C NMR (CDCl3) δ
169.9, 166.5, 146.0, 135.1, 128.7, 128.6, 128.2, 125.8, 67.6, 63.6, 55.1,
42.9, 22.7, 11.7. RP-LC/MS (ESI) m/z 637.4 (M + H)+ (tR = 5.13 min,
5À95% B/6 min). Anal. Calcd for C32H40N6O8: C, 60.37; H, 6.33; N,
13.20. Found: C, 60.49; H, 6.42; N, 13.13.
N2,N5-Bis(2,3-dihydroxypropyl)-3,6-bis[(S)-2,3-dihydrox-
ypropylamino]pyrazine-2,5-dicarboxamide (2h). To an or-
ange solution of tetra-acetonide 8c (3.68 g, 5.64 mmol) in THF (50 mL)
was added 1.0 N HCl solution (45 mL) and stirred for 2 h at rt in an
atmosphere of argon. Most of the THF was removed; the aq mixture was
neutralized with 1.0 N NaOH, and evaporated to dryness. The residue
thus obtained was further dried under high vacuum to give the crude
product (5.86 g), which upon purification by gravity chromatography
over C18 silica gel (gradient elution with water to 35% MeOH in water),
afforded 2h (2.52 g, 91%) as a red powder. 1H NMR (DMSO-d6) δ 8.50
(t, J = 5.9 Hz, 2 H, exchangeable with D2O), 7.99 (t, J = 5.7 Hz, 2 H,
N2,N5-Bis[(R)-1,3-dihydroxy-1-oxopropan-2-yl]-3,6-bis-
(propylamino)pyrazine-2,5-dicarboxamide (2f). A solution of
the bis-benzyl ester 8a (0.700 g, 1.10 mmol) in MeOH (150 mL) was
purged with argon and a slurry of 10% Pd/C (0.350 g) in water (8 mL)
was added. Then ammonium formate (0.480 g, 7.61 mmol) was added,
and the resulting mixture was heated to 60 °C and stirred for 1 h. The
reaction mixture was allowed to cool to room temperature, filtered
through Celite, and concentrated on a rotary evaporator. The crude
product (0.559 g) was subjected to purification by preparative RP-HPLC
5055
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
exchangeable with D2O), 4.91 (dd, J = 4.9, 1.3 Hz, 2 H, exchangeable
with D2O), 4.81 (d, J = 4.9 Hz, 2 H, exchangeable with D2O), 4.69À4.63
(m, 4 H, exchangeable with D2O), 3.66À3.56 (m, 6 H), 3.47À3.31 (m,
10 H), 3.28À3.19 (m, 4 H). 13C NMR (DMSO-d6) δ 165.9, 145.9,
126.4, 70.8, 70.41, 70.36, 64.6, 64.31, 64.29, 44.15, 44.13, 42.9. RP-
HPLC (254 nm) 97.3% (tR = 11.76 min, 5À50% B/20 min). RP-LC/
MS (ESI) m/z 493.4 (M + H)+ (tR = 3.00 min, 5À95%B/6 min). HRMS
(ESI) m/z calcd for C18H32N6O10Na (M + Na)+ 515.2072, found
(s, 4 H), 4.83À4.80 (dt, J = 7.6, 3.8 Hz, 2 H), 4.33À4.29 (m, 2 H), 4.11
(dd, J = 11.5, 4.1 Hz, 2 H), 4.04À3.99 (m, 4 H), 3.89 (dd, J = 8.6, 4.7 Hz,
2 H), 3.77À3.72 (dt, J = 14.0, 6.7 Hz, 2 H), 3.44À3.39 (dt, J = 13.7, 5.6
Hz, 2 H), 3.17 (br, 2 H), 1.48 (s, 6 H), 1.35 (s, 6 H). 13C NMR (CDCl3)
δ 169.9, 165.8, 146.1, 135.2, 128.7, 128.5, 128.2, 125.9, 109.6, 75.4, 67.5,
67.1, 63.4, 54.9, 42.9, 27.2, 25.4. RP-LC/MS (ESI) m/z 781.4 (M + H)+
(tR = 4.80 min, 5À95%B/6 min). HRMS (ESI) m/z calcd for
C38H48N6O12Na (M + Na)+ 803.3222, found 803.3223; calcd for
C38H49N6O12 (M + H)+ 781.3403, found 781.3412.
515.2059. Anal. Calcd for C18H32N6O10 H2O: C, 42.35; H, 6.71; N,
3
N2,N5-Bis[(R)-1,3-dihydroxy-1-oxopropan-2-yl]-3,6-bis-
[(S)-2,3-dihydroxypropylamino]pyrazine-2,5-dicarboxa-
mide (2j). To a solution of bis-acetonide 8e (0.075 g, 0.096 mmol) in
THF (3 mL) was added 1.0 N HCl solution (0.5 mL) and stirred for 4 h
at rt under argon atmosphere. Most of the THF was removed, and the aq
reaction mixture was neutralized with 1.0 N NaOH solution and
evaporated to dryness. RP-LC/MS analysis of at this stage indicated
partial hydrolysis of benzyl ester groups even though the aliquot prior to
neutralization with NaOH showed the presence of single peak with
molecular ion at m/z 701.3 (M+H)+ corresponding to the desired
product. To complete the ester hydrolysis, the above residue was
dissolved in THFÀH2O (2 mL; 1:1, v/v) and treated with 1.0 N NaOH
(1 mL) for 1 h at rt. The reaction mixture was neutralized with 1.0 N
HCl, evaporated to dryness, and the crude product (0.200 g) was
subjected to purification by gravity chromatography over C18 silica gel
(gradient elution with water to 20% MeOH in water) to afford 2j (0.037
g, 74%) as a red powder. 1H NMR (DMSO-d6) δ 8.89 (d, J = 5.7 Hz, 2 H,
exchangeable with D2O), 7.90 (t, J = 5.7 Hz, 2 H, exchangeable with
D2O), 5.47 (br s, 2 H, exchangeable with D2O), 4.89 (d, J = 5.1 Hz, 2 H,
exchangeable with D2O), 4.75 (t, J = 5.7 Hz, 2 H, exchangeable with
D2O), 3.84À3.80 (dt, J = 7.4, 5.5 Hz, 2 H), 3.77À3.72 (m, 4 H),
3.52À3.23 (m, 10 H). 13C NMR (DMSO-d6) δ 172.67, 165.2, 145.9,
126.5, 70.3, 64.5, 62.7, 55.4, 44.6. RP-HPLC (280 nm) 98.2% (tR = 12.31
min, 5À50%B/20 min). RP-LC/MS (ESI) m/z 521.3 (M + H)+ (tR =
3.03 min, 5À95%B/6 min). HRMS (ESI) m/z calcd for C18H27N6O12
(M À H)À 519.1692, found 519.1680; calcd for C18H26N6O12 (M À
2H)2À 259.0810, found 259.0802.
16.46. Found: C, 42.31; H, 6.58; N, 16.43.
N2,N5-Bis[(2,2-dimethyl-1,3-dioxolan-4-yl)methyl]-3,6-
bis{[(R)-2,2-dimethyl-1,3-dioxolan-4-yl]methylamino}pyr-
azine-2,5-dicarboxamide (8d). The reaction of bis-acetonide 4b
(1.00 g, 2.36 mmol) with (S)-2,2-dimethyl-1,3-dioxolane-4-carbalde-
hyde (7b, 0.920 g, 7.07 mmol) in the presence of HOAc (0.400 mL,
6.94 mmol) and Na(OAc)3BH (1.46 g, 6.89 mmol) in anhyd DCE (40 mL)
was carried out overnight (ca. 22 h) as described for the preparation of 8c
[including further treatment with additional reagents aldehyde 7b
(0.920 g, 7.07 mmol), HOAc (0.400 mL, 6.94 mmol), and Na-
(OAc)3BH (1.46 g, 6.89 mmol) overnight]. After a similar work up,
the crude product (1.45 g) was subjected to an automated flash chroma-
tography [column: RediSep Rf 120 g silica gel; λmax, 280 nm; flow,
40 mL/min; gradient, 0À2% B/40 min, 2%B/90 min (mobile phase A,
CHCl3; mobile phase B, MeOH)] to afford 8d (1.38 g, 90%) as an
orangeÀred powder. 1H NMR (CDCl3) δ 8.10 (t, J = 6.0 Hz, 2 H), 8.06
(t, J = 6.0 Hz, 2 H), 4.36À4.31 (m, 4 H), 4.09À4.06 (m, 4 H), 3.79À3.76
(m, 2 H), 3.74À3.70 (m, 2 H), 3.66À3.52 (m, 8 H), 1.49 (s,
3 H), 1.48 (s, 3 H), 1.45 (s, 6 H), 1.37 (s, 6 H), 1.36 (s, 6 H). RP-
LC/MS (ESI) m/z 653.5 (M + H)+ (tR = 4.83 min, 5À95%B/6 min).
Anal. Calcd for C30H48N6O10: C, 55.20; H, 7.41; N, 12.88. Found: C,
54.98; H, 7.50; N, 12.73.
N2,N5-Bis(2,3-dihydroxypropyl)-3,6-bis[(R)-2,3-dihydrox-
ypropylamino]pyrazine-2,5-dicarboxamide (2i). The depro-
tection of tetra-acetonide 8d (1.00 g, 1.53 mmol) with 1.0 N HCl
(15 mL) in THF (25 mL) was carried out for 5 h as described for the
preparation of 2h. After a similar work up, the crude product (2.01 g)
was subjected to purification by gravity chromatography over C18 silica
gel (gradient elution with water to 25% MeOH in water) to give 2i (0.58
g, 77%) as an orange powder. 1H NMR (DMSO-d6) δ 8.50 (t, J = 5.8 Hz,
2 H, exchangeable with D2O), 7.99 (t, J = 5.4 Hz, 2 H, exchangeable with
D2O), 4.91 (d, J = 3.4, Hz, 2 H), 4.80 (d, J = 4.7 Hz, 2 H, exchangeable
with D2O), 4.68À4.63 (m, 4 H, exchangeable with D2O), 3.63À3.55
(m, 6 H), 3.47À3.31 (m, 10 H), 3.27À3.19 (m, 4 H). 13C NMR
(DMSO-d6) δ 165.9, 145.94, 145.93 126.4, 70.8, 70.41, 70.36, 64.7,
64.31, 64.29, 44.14, 44.12, 42.9. RP-HPLC (254 nm) 98.2% (tR = 11.78
min, 5À50%B/20 min). RP-LC/MS (ESI) m/z 493.3 (M + H)+ (tR =
2.97 min, 5À95%B/6 min). HRMS (ESI) m/z calcd for C18H32N6O10 Na
Measurement of Photophysical Properties and Protein
Binding. In general, each test compound was dissolved in PBS buffer to
form a 2 mM stock solution. The UV absorbance properties were
determined on a 100 μM solution in PBS using a UV-3101PC
UVÀvisÀNIR scanning spectrophotometer system from Shimadzu.
The fluorescence properties (λex, λem, and intensity maximum at λem
)
were determined on a 10 μM solution in PBS using a Fluorolog-3
spectrofluorometer system from Jobin Yvon Horiba. The percent
plasma protein binding was determined on a 20 μM compound solution
in rat plasma incubated at 37 °C for 1 h. The separation of free from
bound was made using an Amicon Centrifree YM-30 device
(Regenerated Cellulose 30000 MWCO) and a Z400K refrigerated
universal centrifuge from Hermle. The concentration of protein-free
was determined via HPLC analysis using a set of external calibration
standards and fluorescence detection.
Injected Dose Recovery in Urine Studies. Recovery of the
injected dose in urine studies were conducted in either conscious or
anesthetized SpragueÀDawley rats. The test compound (1 mL, 2 mM in
PBS) was administered by tail vein injection into conscious, restrained
rats, with subsequent collection of urine at the time points of 2, 4, and 6 h
post injection. The metabolic cages were washed with water to maximize
the recovery of urine discharged at each time point. Alternatively, rats
were anesthetized with 100 mg/kg Inactin intraperitoneally, a trachea
tube was inserted to maintain adequate respiration, and 1 mL of test
compound was injected into the lateral tail vein. Rats were placed on
37 °C heating pad during the entire experiment. At 6 h post injection,
the abdomen was opened, and the urine was removed from the bladder
using a 21 gauge needle and a 3 cc syringe. Quantitation of each compound
(M + Na)+ 515.2072, found 515.2069. Anal. Calcd for C18H32N6O10
3
H2O: C, 42.35; H, 6.71; N, 16.46. Found: C, 42.75; H, 6.94; N, 16.53.
N2,N5-Bis[(R)-1-(benzyloxy)-3-hydroxy-1-oxopropan-2-yl]-
3,6-bis{[(S)-2,2-dimethyl-1,3-dioxolan-4-yl]methylamino}-
pyrazine-2,5-dicarboxamide (8e). The reaction of bis-benzyl ester
4d (0.50 g, 0.905 mmol) with aldehyde 7a (0.71 g, 5.43 mmol) in the
presence of HOAc (0.31 mL, 5.43 mmol) and Na(OAc)3BH (1.15 g,
5.43 mmol) in anhyd DCE (20 mL) was carried out overnight as
described for the preparation of 8c [including further treatment with
additional reagents aldehyde 7a (0.71 g, 5.43 mmol), HOAc (0.71 g,
5.43 mmol), and Na(OAc)3BH (1.15 g, 5.43 mmol) overnight]. After a
similar work up, the crude product (0.80 g) was subjected to an
automated flash chromatography [column: RediSep Rf 40 g silica gel;
λmax, 280 nm; flow, 15 mL/min; gradient, 0À3% B/30 min, 3%B/50
min (mobile phase A, CHCl3; mobile phase B, MeOH)] to afford 8e
1
(0.13 g, 18%) as an orangeÀred powder. H NMR (CDCl3) δ 8.62
(d, J = 7.8 Hz, 2 H), 7.91 (t, J = 6.1 Hz, 2 H), 7.37À732 (m, 10 H), 5.25
5056
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
in urine was performed via HPLC analysis using a set of external calibra-
tion standards and fluorescence detection. The percent recovery of com-
pound in urine at each time point was calculated based on the balance
of mass.
’ REFERENCES
(1) Ekanoyan, G.; Levin, N. W. In Clinical Practice Guidelines for
Chronic Kidney Disease: Evaluation, Classification, and Stratification (K/
DOQI); National Kidney Foundation: Washington, D.C., 2002; pp 1À22.
(2) Rule, A. D.; Larson, T. S.; Bergstralh, E. J.; Slezak, J. M.; Jacobsen,
S. J.; Cosio, F. G. Using serum creatinine to estimate glomerular
filtration rate: accuracy in good health and in chronic kidney disease.
Ann. Intern. Med. 2004, 141, 929–937.
(3) Stevens, L. A.; Levey, A. S. Clinical implications of estimating
equations for glomerular filtration rate. Ann. Intern. Med. 2004, 141, 959–961.
(4) Bellomo, R.; Ronco, C.; Kellum, J. A.; Mehta, R. L.; Palevsky, P.
Acute Dialysis Quality Initiative workgroup. Acute renal failure—
definition, outcome measures, animal models, fluid therapy, and in-
formation technology needs: the Second International Consensus
Conference of the Acute Dialysis Quality Initiative (ADQI) Group.
Crit. Care 2004, 8, R204–R212.
(5) Diskin, C. J. Creatinine and GFR: an imperfect marriage of
convenience. Nephrol., Dial., Transplant. 2006, 21, 3338–3339.
(6) Carrie, B. J.; Goldbetz, H. V.; Michaels, A. S.; Myers, B. D.
Creatinine: an inadequate filtration marker in glomerular disease. Am.
J. Med. 1980, 69, 177–182.
(7) Stevens, L. A.; Levey, A. S. Measured GFR as a confirmatory test
for estimated GFR. J. Am. Soc. Nephrol. 2009, 20, 2305–2313.
(8) Sturgeon, C.; Sam, A. D.; Law, W. R. Rapid determination of
glomerular filtration rate by single-bolus inulin: a comparison of
estimation analyses. J. Appl. Physiol. 1998, 84, 2154–2162.
Invasive Pharmacokinetic Studies. Male SpragueÀDawley rats
(330À380 g) were anesthetized by Inactin (IP). Rats were surgically
instrumented with a trachea tube (PE-190) to facilitate breathing and
femoral artery and vein catheters (PE-50 filled with 20 units/mL heparinized
saline) for blood sampling and drug administration, respectively. After
administration of 1 mL of a 2 mM solution of agent, approximately 200
μL of blood was sampled and placed into a heparinized tube (Microtainer
Brand Tube w/Lithium Heparin, BD 365971) at 0, 1, 6, 12, 18, 30, 45, 60, 90,
120 min. The concentration of compound in each centrifuged plasma sample
was determined via HPLC analysis using a set of external calibration
standards and fluorescence detection. The resulting pharmacokinetic para-
meters of the compound were analyzed using WinNonLin pharmacokinetic
modeling software (Pharsight, Mountainview, CA).
Probenecid Inhibition Studies. Two groups of 4À6 rats male
SpragueÀDawley rats were treated in the same manner as described
above in the invasive pharmacokinetic studies. The treatment group
received 70 mg/kg probenecid 10 min prior to injection of the test
compound; this administration was flushed with 0.2 mL NaCl. The
control group was treated with the vehicle (PBS).
Fluorescence Detection Studies. A Maestro in vivo imaging
system from CRi was employed. Two mice were anesthetized and had their
abdomens shaved and remaining hair removed with depilatory. A home-
made catheter (30 gauge needle and PE 10 tubing) was put into the lateral
tail vein for administration of the test materials. The green filter set of the
Maestro instrument was used. A DyCE collection was set up with a 600 s
duration at 1 frame every 3 s with a 2.5 ms exposure. A second time series
was set up to acquire 1 frame every 60 s for a 3600 s duration. The collection
was started approximately 10 s before the compounds were injected.
(9) Wilson, D. M.; Bergert, J. M.; Larson, T. H.; Leidtke, R. R. GFR
determined by nonradiolabeled iothalamate using capillary electrophor-
esis. Am. J. Kidney Dis. 1997, 39, 646–652.
(10) Guesry, P.; Kaufman, L.; Orloff, S.; Nelson, J. A.; Swann, S.;
Holliday, M. Measurement of glomerular filtration rate by fluorescent
excitation of non-radioactive meglumine iothalamate. Clin. Nephrol.
1975, 3, 134–138.
(11) Lewis, N.; Kerr, R.; Van Buren, C. Comparative evaluation of
urographic contrast media, inulin, and 99mTc-DTPA clearance methods
for determination of glomerular filtration rate in clinical transplantation.
Transplantation 1989, 48, 790–796.
(12) Rabito, C .A.; Chen, Y.; Schomacker, K. T.; Modell, M. D.
Optical, real-time monitoring of the glomerular filtration rate. Appl. Opt.
2005, 44, 5956–5965.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: 314-654-3800. Fax: 314-654-8900. E-mail: Raghavan.
Rajagopalan@covidien.com.
(13) Chinen, L.; Galen, K. P.; Kuan, K. T.; Dyszlewski, M. E.; Ozaki,
H.; Sawai, H.; Pandurangi, R. S.; Jacobs, F. G.; Dorshow, R. B.;
Rajagopalan, R. Fluorescence-enhanced europium complexes for the
assessment of renal function. J. Med. Chem. 2008, 51, 957–962.
(14) Yu, W.; Sandoval, R. M.; Molitoris, B. A. Rapid determination
of renal filtration using an optical ratiometric imaging approach. Am.
J. Physiol. Renal Physiol. 2007, 292, F1873–F1880.
(15) Schock-Kusch, D.; Sadick, M.; Henninger, N.; Kraenzlin, B.;
Claus, G.; Kloetzer, H.-M.; Weiss, C.; Pill, J.; Gretz, N. Transcutaneous
measurement of glomerular filtration rate using FITC-sinistrin in rats.
Nephrol., Dial., Transplant. 2009, 24, 2997–3001.
Present Addresses
†Department of Pharmaceutical Sciences, Southern Illinois Uni-
versity, 220 University Park Drive, Edwardsville, Illinois, 62026,
United States.
’ ACKNOWLEDGMENT
We thank Dr. Carlos Rabito of the Harvard University Medical
School and Massachusetts General Hospital, Professor Thomas
Dowling of the University of Maryland, and Drs. Sevag Demirjian
and Joseph Nally of the Cleveland Clinic Foundation for their
insightful discussions and advice. We thank our Covidien Phar-
maceutical colleagues Christopher R. DeFusco, Jolette Wojdyla,
Tim Marzan, Micah Wilcox, Tasha Shoenstein, Rana Kumar, and
Bich Vu for their experimental and analytical support. We thank
our Covidien Pharmaceutical colleague Brian Donley for his help
with the statistical analysis of the probenecid inhibition study. We
thank Dr. Bryan Q. Spring of Wellman Center of Photomedicine
for his help in imaging processing.
(16) Roch-Ramel, F.; Besseghir, K.; Murer, H. Renal excretion and
tubular transport of organic anions and cations. In Handbook of
Physiology, Section 8, Neurological Physiology, Vol. II; Windhager,
E. E., Ed.; Oxford University Press: New York, 1992; pp 2189À2262.
(17) Roch-Ramel, F.; De Broe, M. E. Renal handling of drugs and
xenobiotics. In Clinical Nephrotoxins, 2nd ed.; De Broe, M. E., Porter,
G. A., Bennet, W. M., Verpooten, G. A., Eds.; Kluwer Academic
Publishers: Dordrect, Netherlands, 2003; pp 21À46.
(18) Preliminary results of the work were presented in part at the
SPIE, Biophotonics West annual conference, San Jose, CA, 2008.
Dorshow, R. B.; Asmelash, B.; Chinen, L. K.; Debreczeny, M. P.; Fitch,
R. M.; Freskos, J. N.; Galen, K. P.; Gaston, K. R.; Marzan, T. A.; Poreddy,
A. R.; Rajagopalan, R.; Shieh, J.-J; Neumann, W. L. New molecular
probes for the continuous monitoring of renal function. In Molecular
Probes for Biomedical Applications II, Vol. 6867; Achilefu, S., Barnhop,
D. J., Raghavachari, R., Eds.; SPIE Press: Bellingham, WA, 2008.
’ ABBREVIATIONS USED
GFR, glomerular filtration rate; %ID, percent injected dose; PPB,
plasma protein binding; PBS, phosphate buffered saline; DTPA,
diethylenetriaminepentaacete
5057
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058

Journal of Medicinal Chemistry
ARTICLE
(19) Fritzberg, A. R.; Kasina, S.; Eshima, D.; Johnson, D. L. Synthesis
and biological evaluation of technetium-99m-MAG3 as a hippuran
replacement. J. Nucl. Med. 1986, 27, 111–116.
(20) Shirai, K.; Yanagisawa, A.; Takahashi, H.; Fukunishi, K.; Matsuoka,
M. Synthesis and fluorescent properties of 2,5-diamino-3,6-dicyanopyrazine
dyes. Dyes Pigm. 1998, 39, 49–68.
(21) Kim, J. H.; Shin, S. R.; Matsuoka, M.; Fukunishi, K. Self-
assembling of aminopyrazine fluorescent dyes and their solid state
spectra. Dyes Pigm. 1998, 39, 341–357.
(22) Poreddy, A. R.; Asmelash, B.; Neumann, W. L.; Dorshow, R. B. A
highly efficient method for the N-alkylation of aminopyrazines: Synthesis
of hydrophilic red fluorescent dyes. Synthesis 2010, 2383–2392.
(23) Prueksaritanont, T.; Lui, C. Y.; Lee, M. G.; Chiou, W. L. Renal
and non-renal clearances of iothalamate. Biopharm. Drug Dispos. 1986,
7, 347–355.
(24) Bormans, G.; Cleynhens, B.; Adriaens, P.; de Roo, M.; Verbrug-
gen, A. Synthesis and labeling characteristics of 99mTc-mercaptoacetyltri-
peptides. J. Labelled Compound. Radiopharm. 1993, 33, 1065–1078.
(25) Bubeck, B.; Brandau, W.; Weber, E.; Kaelbe, T.; Parekh, N.;
Georgi, P. Pharmacokinetics of technetium-99m-MAG3 in humans
J. Nucl. Med. 1990, 31, 1285–1293.
(26) Two-sample t-test. Weiss, N. A. Introductory Statistics, 8th ed.;
Addison-Wesley: Reading, MA, 2007.
(27) Dorshow, R. B.; Bugaj, J. E.; Burleigh, B. D.; Duncan, J. R.;
Johnson, M. A.; Jones, W. B. Noninvasive fluorescence detection of
hepatic and renal function. J. Biomed. Opt. 1998, 3, 340–345.
5058
dx.doi.org/10.1021/jm200257k |J. Med. Chem. 2011, 54, 5048–5058