The Journal of Immunology
A2A Adenosine Receptors on Bone Marrow-Derived Cells
Protect Liver from Ischemia-Reperfusion Injury1
Yuan-Ji Day,*† Yuesheng Li,‡ Jayson M. Rieger,§ Susan I. Ramos,* Mark D. Okusa,‡ and
Joel Linden2*‡
Activation of the A2A adenosine receptor (A2AR) during reperfusion of various tissues has been found to markedly reduce
ischemia-reperfusion injury. In this study, we used bone marrow transplantation (BMT) to create chimeric mice that either
selectively lack or selectively express the A2AR on bone marrow-derived cells. Bolus i.p. injection of the selective A2A agonist,
4-{3-[6-amino-9-(5-cyclopropylcarbamoyl-3,4-dihydroxy-tetrahydro-furan-2-yl)-9H-purin-2-yl]-prop-2-ynyl}-piperidine-1-carboxylic
acid methyl ester (ATL313; 3 ␮g/kg), at the time of reperfusion protects wild-type (wt) mice from liver ischemia-reperfusion injury.
ATL313 also protects wt/wt (donor/recipient BMT mouse chimera) and wt/knockout chimera but produces modest protection of
knockout/wt chimera as assessed by alanine aminotransferase activity, induction of cytokine transcripts (RANTES, IFN-␥-inducible
protein-10, IL-1␣, IL-1-␤, IL-1R␣, IL-18, IL-6, and IFN-␥), or histological criteria. ATL313, which is highly selective for the A2AR,
produces more liver protection of chimeric BMT mice than 4-{3-[6-amino-9-(5-ethylcarbamoyl-3,4-dihydroxy-tetrahydro-furan-2-yl)9H-purin-2-yl]-prop-2-ynyl}-cyclohexanecarboxylic acid methyl ester, which is rapidly metabolized in mice to produce 4-{3-[6-amino9-(5-ethylcarbamoyl-3,4-dihydroxy-tetrahydro-furan-2-yl)-9H-purin-2-yl]-prop-2-ynyl}-cyclohexanecarboxylic acid, which has similar
affinity for the A2AR and the proinflammatory A3 adenosine receptor. GFP chimera mice were created to show that vascular endothelial
cells in the injured liver do not account for liver protection because they are not derived by transdifferentiation of bone marrow
precursors. The data suggest that activation of the A2AR on bone marrow-derived cells is primarily responsible for protecting the liver
from reperfusion injury. The Journal of Immunology, 2005, 174: 5040 –5046.
A
denosine is produced in response to ischemia or inflammation and protects tissues from injury. Exposure of tissues to adenosine before ischemia reduces subsequent
injury in the CNS (1) and myocardium (2) due to preconditioning
mediated by A1 or A3 adenosine receptors. In contrast, activation
of the A2A adenosine receptor (A2AR)3 before ischemia does not
reduce subsequent ischemic injury, but A2AR activation during
reperfusion reduces tissue damage to kidney (3), heart (4), skin (5),
lung (6), and spinal cord (7). Ohta and Sitkovsky (8) concluded
that endogenous adenosine protects the liver from injury caused by
toxins (Con A and endotoxin) by demonstrating enhanced hepatic
injury in mice lacking acloraza, the A2AR gene. Recently, we
found that endogenous adenosine also protects the liver from isch-
*Cardiovascular Research Center, Departments of †Molecular Physiology and Biological Physics and ‡Medicine, University of Virginia, Charlottesville, VA 22908; and
§
Adenosine Therapeutics, LLC, Charlottesville, VA 22911
Received for publication September 9, 2004. Accepted for publication February
3, 2005.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work is supported by National Institutes of Health Grant RO1-HL37942 and by
the Dr. Ralph and Marian Falk Medical Research Trust.
2
Address correspondence and reprint requests to Dr. Joel Linden, Cardiovascular
Research Center, MR5 Box 801394, University of Virginia Health System, Charlottesville, VA 22908. E-mail address: [email protected]
3
Abbreviations used in this paper: A2AR, A2A adenosine receptor; IRI, ischemiareperfusion injury; ATL146e, 4-{3-[6-amino-9-(5-ethylcarbamoyl-3,4-dihydroxy-tetrahydro-furan-2-yl)-9H-purin-2-yl]-prop-2-ynyl}-cyclohexanecarboxylic acid methyl
ester; wt, wild type; ko, knockout; ATL146a, 4-{3-[6-amino-9-(5-ethylcarbamoyl3,4-dihydroxy-tetrahydro-furan-2-yl)-9H-purin-2-yl]-prop-2-ynyl}-cyclo-hexanecarboxylic acid; BMT, bone marrow transplantation; FSG, fish skin gelatin; ATL313,
4-{3-[6-amino-9-(5-cyclopropylcarbamoyl-3,4-dihydroxy-tetrahydro-furan-2-yl)-9Hpurin-2-yl]-prop-2-ynyl}-piperidine-1-carboxylic acid methyl ester; ALT, alanine
aminotransferase; IP, IFN-␥-inducible protein.
Copyright © 2005 by The American Association of Immunologists, Inc.
emia-reperfusion injury (IRI), particularly of short duration, but
much greater protection is imparted to the liver by infusing the
synthetic A2A agonist, 4-{3-[6-amino-9-(5-ethylcarbamoyl-3,4dihydroxy-tetrahydro-furan-2-yl)-9H-purin-2-yl]-prop-2-ynyl}cyclohexanecarboxylic acid methyl ester (ATL146e), during reperfusion
following ischemia (9). ATL146e is effective in wild-type (wt), but
not A2AR knockout (ko), mice or in mice also treated with the
A2AR antagonist, 4(2-[7-amino-2-[2-furyl][1,2,4]triazolo[2,3a][1,3,5] triazin-5-yl-amino]-ethyl)phenol.
In the current study, we sought to investigate the contribution of
the A2AR on bone marrow-derived cells to protection of liver from
IRI. Our approach was to prepare chimeric mice by irradiation and
bone marrow transplantation (BMT), resulting in animals that either selectively lack or selectively express the A2AR on bone marrow-derived cells. Chimeric animals designated ko/wt were created by transplanting A2AR ko marrow to wt-irradiated recipients.
Chimeric animals designated wt/ko were created by transplanting
wt marrow to A2AR ko recipients. In addition, we created GFP/wt
BMT chimeric mice that were used to determine which cells are
derived from donor marrow in chimeric mice. The results indicate
that activation of the A2AR on bone marrow-derived cells is primarily responsible for protecting the liver from reperfusion injury.
Materials and Methods
Creating A2AR ko mice congenic to C57BL/6
The ko locus of B6;129P-adora2atm1chen mice that express an ablated A2AR
gene on a mixed genetic background (10) was moved to a C57BL/6 background by monitoring 96 microsatellites for five generations of markerassisted breeding. In the resulting mouse line, DNA derived from the 129
strain can be detected only in an 8-cm region between D10Mit31 and
D10Mit42 surrounding the adora2a locus on chromosome 10.
0022-1767/05/$02.00
The Journal of Immunology
Bone marrow transplantation
C57BL/6 mice (Hilltop) or congenic A2AR ko mice were irradiated and
used as recipients for BMT experiments. Donor marrow was derived from
four different mouse strains: 1) typical C57BL/6 mice that are negative for
CD45.1; 2) a substrain of C57BL/6 mice that are positive for CD45.1; 3)
A2AR ko mice; or 4) green mice expressing GFP in all cells. Male donor
mice (12 wk old; 25–28 g) were anesthetized with Nembutal (20 ␮g/g) and
sacrificed by cervical dislocation. The marrow from the tibia and femur
were harvested under sterile conditions. The marrow cavity was flushed
with RPMI 1640 medium (Invitrogen Life Technologies) ⫹ 10% FCS and
drawn through a 22-gauge needle and then through a 25-gauge needle to
obtain a suspension of ⬃50 million nucleated bone marrow cells. These
were washed, resuspended, and counted. Recipient mice (male, 8 –10 wk
old, 22–25 g) were irradiated twice at 4-h intervals with 600 rad. Immediately following the second irradiation period, 3 ⫻ 106 bone marrow cells
were injected i.v. Irradiated/transplanted mice were housed in microisolators for at least 8 wk before experimentation and given autoclaved food and
water containing 5 mM sulfamethoxazole and 0.86 mM trimethoprim.
Assessment of reconstitution efficacy
The origin of leukocytes in chimeric mice was determined 6 wk after BMT.
Leukocytes were prepared from peripheral blood or spleen digested for 30
min with collagenase D, filtered through 40-␮m nylon mesh, and washed
with PBS. Harvested spleen leukocytes were placed in 1.5 mM NH4Cl to
lyse RBC and then resuspended in 600 ␮l of PBS containing 1% BSA.
After blocking nonspecific Fc binding with anti-mouse CD16/CD32
(553141; BD Pharmingen), spleen leukocytes were incubated on ice for 30
min with R-PE-conjugated mouse anti-mouse CD45.1 mAb (clone A20,
553776; BD Pharmingen) and either allophycocyanin-conjugated rat antimouse CD11b (Mac-1␣-chain) mAb (553312; BD Pharmingen), FITC-conjugated rat anti-mouse CD8a mAb (553030; BD Pharmingen), or allophycocyanin-conjugated rat anti-mouse CD4 mAb (553051; BD Pharmingen).
Subsequent flow cytometry data acquisition and analysis were performed using a FACScan and CellQuest software (BD Biosciences).
Origin of endothelial cells following BMT
Mice expressing GFP in all cells were used as bone marrow donors to
produce GFP/wt mouse chimera expressing GFP only in bone marrowderived cells. GFP/wt chimeric mice were used to identify the source (donor or recipient) of liver vascular endothelial cells. Mouse livers were
excised and fixed by immersion in 3.7% paraformaldehyde in PBS for
20 –24 h and washed three times in PBS. Mouse livers were excised and
fixed by immersion in 3.7% paraformaldehyde in PBS for 20 –24 h, washed
three times in PBS, and placed in cassettes in 70% ethanol for paraffin
embedding. Five-micrometer sections were rehydrated and blocked for 20
min with 0.45% hydrogen peroxide in filtered deionized water. Sections
were preincubated in TBS (pH 7.8) and blocked with 0.5% fish skin gelatin
(TBS/FSG) ⫹ 10% normal serum (secondary host) ⫹ 2 drops/ml avidin
block (Vector Laboratories) for 60 min. A rabbit polyclonal anti-GFP Ab
(1:3500; Novus Biologicals) in TBS/FSG ⫹ 7.5% normal serum ⫹ 2
drops/ml biotin block was applied for 2–3 h at room temperature and then
overnight at 4°C. After washing four times for 5 min in TBS/FSG, secondary Ab (6 ␮g/ml) was applied for 60 min at room temperature and
washed again four times for 5 min in TBS before applying Vectastain Elite
ABC reagent (Vector Laboratories) for 30 min. Chromogenic visualizations was achieved using 1 mg/ml 3,3⬘-diaminobenzidine tetrahydrochloride (DAKO) for 5 min. Sections were counterstained with hematoxylin I
(Richard-Allan Scientific).
Liver IRI
Mice were anesthetized by i.p. injection of 100 mg/kg ketamine and 10
mg/kg xylazine and then injected s.c. with 50 ␮g/kg glycopyrrolate to
prevent excess salivation and possible suffocation. Eye lubricant was applied to prevent ocular dehydration. Mice were placed under a heating lamp
and on a 37°C heating pad. In some animals, rectal temperature was monitored with a TCAT-1A temperature control unit (Physitemp) and was
found to be 35°C ⫾ 1°C and did not differ between drug treatment groups.
Careful monitoring of body temperature is particularly important in these
experiments because A2AR agonists are vasodilators that can produce hypothermia, although this requires higher doses than were used here. We
used a partial hepatic ischemia model that spares the right lobe. This model
avoids intestinal congestion, sepsis, and peritonitis and is not lethal during
liver ischemia times up to 60 min. Each mouse was placed in a supine
position. A midline laparotomy and incision of the Linea Alba exposed the
peritoneal cavity. The stomach and duodenum were displaced caudally to
expose the hepatic triad and caudate lobes. The caudate lobe was separated
5041
gently from the left lobe and displaced from the right upper and lower lobes
caudally to clearly view the hepatic triad above the bifurcation of right
lobes, median lobe, and left lobe. A microaneurysm clip was applied to the
hepatic triad above the bifurcation to clamp the flow of the hepatic artery,
portal vein, and bile duct. The peritoneum was closed after superfusion
with 200 ␮l of warm saline supplemented with 50 U/kg heparin. After 60
min of ischemia, the peritoneum was reopened, and the microaneurysm
clip was removed. Immediately before reperfusion was initiated, each
mouse received either a single bolus i.p. injection of 4-{3-[6-amino-9-(5cyclopropylcarbamoyl-3,4-dihydroxy-tetrahydro-furan-2-yl)-9H-purin-2yl]-prop-2-ynyl}-piperidine-1-carboxylic acid methyl ester (ATL313; 3
␮g/kg) or an i.p. loading dose of ATL146e (1 ␮g/kg) and a s.c. osmotic
minipump (model 1003D; ALZET) to release 10 ng/kg/min ATL146e or
vehicle during the entire 24-h reperfusion period. The surgical wound was
closed with metal staples, and mice were maintained on the heating pad
until the anesthetic wore off.
Serum alanine aminotransferase (ALT) activity
Whole blood was harvested from an incision through an axillary artery and
placed on ice before centrifugation at 14,000 ⫻ g for 20 min. Serum was
collected and stored at ⫺80°C. ALT was determined by using an ALT kit
(Pointe Scientific) using a Spectra MAX 190 plate reader with SOFTmax
PRO software, version 3.1.2 (Molecular Devices). A 200-␮l aliquot of a
prewarmed (37°C) mixture of L-alanine and ␣-ketoglutaric acid was added
to 20 ␮l of undiluted and/or saline-diluted serum in a 96-well plate. After
a 1-min incubation at 37°C, the plate was scanned at 340 nm at 9-s intervals
for 60 s, and the rate of change in absorbance converted into Sigma-Frankel
units (1 IU ⫽ 0.482 Sigma-Frankel U).
RNase protection assays
Total liver RNA was extracted from homogenized tissue with RNAzol B
(Leedo Medical Laboratories) and subjected to 1.5% agarose gel electrophoresis to assess for the integrity of RNA before solution hybridization.
Cytokine and chemokine mRNA expression were assessed with the RiboQuant Multiprobe RNAase protection system (BD Pharmingen), according
to the manufacturer’s protocol. In brief, mRNA-specific RNA probes were
labeled with [32P]UTP using multiprobe template sets, mCK2b for IL12p35, IL-12p40, IL-10, IL-1␣, IL-1␤, IL-1Ra, IL-18, IL-6, and IFN-␥,
and mCK5c for Ltn, RANTES, MIP-1␤, MIP-1␣, MIP-2, IFN-␥-inducible
protein (IP)-10, MCP-1, TCA-3, and eotaxin. RNA was subjected to solution hybridization at 56°C and then digested by adding RNase. Protected
fragments were separated by electrophoresis through 5% polyacrylamide
gels at 100 mV. After ⬃4 h of electrophoresis, gels were exposed to Kodak
MS film at ⫺80°C with intensifying screens.
Histology
Livers were harvested after 24 h reperfusion, fixed in 4% paraformaldehyde in PBS (pH 7.4), and embedded in paraffin. Four-micrometer sections
were subjected to standard H&E staining.
Neutrophil immunohistochemistry
Four-micrometer tissue sections were incubated with rat anti-mouse neutrophil Abs (1 ␮g/ml) followed by a biotinylated goat anti-rat secondary
Ab. The peroxidase reaction was performed according to the manufacturer’s protocol (Vectastain ABC Elite kit), and reaction times for sections
from sham control and experimental animals were identical.
Metabolism of ATL146e
Blood samples collected from animals at various time after the i.v. injection of ATL146e or ATL313 were mixed with 10 U/ml heparin. In some
cases, ATL146e or ATL313 was added to blood in vitro. Blood aliquots
(200 ␮l) were pipetted into 400-␮l acetonitrile containing 0.3 ␮g/ml
(2S,3R,4S,5R)-5-(6-amino-2-iodo-9H-purin-9-yl)-N-cyclopropyl-tetrahydro-3,4-dihydroxyfuran-2-carboxamide as an internal standard, centrifuged
for 15 min at 14,000 ⫻ g, and the supernatants were collected and dried
under vacuum at 4°C with centrifugation. The residue was resuspended in
200 ␮l of MeOH/dH2O (50:50), sonicated for 15 min, applied to a
Millipore Microcon YM-100 filter cartridge, and centrifuged for 30 min at
14,000 ⫻ g. The filtrates were applied to a Waters Symmetry C8 2.1 ⫻
150-mm column and eluted with methanol/aqueous 0.1% formic acid. The
methanol concentration in the chromatography buffer as follows: 0 –2 min,
40% isocratic; 2–10 min, 40 – 60% linear gradient; and 10 –20 min, 60%
isocratic. Compounds were detected by tandem mass spectrometry in
positive ion mode using electrospray ionization with a Thermoelectron
LCQ Advantage mass spectroscopy detector. Peak integration of internal
5042
A2A ADENOSINE RECEPTORS AND LIVER PROTECTION
ation and BMT, the repopulation of B lymphocytes, monocytes/
macrophages, neutrophils, and dendritic spleen cells was ⬎97% as
shown in Fig. 2. Somewhat lower repopulation of CD4⫹ and
CD8⫹ T cells (83– 85%) can be attributed to radiation resistance of
T lymphoblasts.
The effects of ATL146e and ATL313 on liver IRI in chimeric
BMT mice
FIGURE 1. Effects of ATL146e on liver IRI. Livers from WT or A2AR
ko mice were subjected to 1 h of ischemia and 24 h of reperfusion. Serum
ALT was collected at 24 h. ATL146e or vehicle was added at the time of
reperfusion as described under Materials and Methods. Each bar is the
mean ⫾ SD of 10 –12 mice.
standard peaks and ATL146e, ATL146a, and ATL313 was by the Excalibur program, allowing for the calculation of sample concentration.
Statistical analyses
Differences in plasma ALT or liver myeloperoxidase in multiple groups of
chimeric mice were determined by one-way ANOVA and posttesting using
Bonferroni’s multiple comparison method.
Results
Effect of ATL146e on IRI of liver in wt and chimeric BMT mice
In wt C57BL/6 mice subjected to 60 min of liver ischemia and 24 h
of reperfusion, treatment with the selective A2A agonist ATL146e
during the reperfusion period resulted in an 84% decrease in liver
damage as assessed by serum levels of ALT (Fig. 1). As we have
reported previously (9), this protection is completely lost in congenic C57BL/6 mice lacking the A2AR gene, indicating that it is
mediated by the A2AR and not by other adenosine receptor subtypes. To determine which cells mediate protection by ATL146e,
we created chimeric mice through the use of lethal irradiation and
BMT. The repopulation efficiency of various hematopoietic lineages was determined by flow cytometry (n ⫽ 4) using CD45 as a
mouse strain marker (11). CD45 found in C57BL/6 mice is replaced by the alloantigen CD45.1 in B6.SJL-Ptprca Pep3b/BoyJ
mice (12). Cells arising from different mouse strains can be distinguished using selective Abs and FACS. Following lethal radi-
FIGURE 2. Repopulation of leukocytes in recipient
mice after BMT. Irradiated mice lacking CD45.1 received bone marrow cells from CD45.1-positive animals. Six weeks after transplantation, spleen cells were
harvested and examined by FACS for CD45.1 and leukocyte markers. The percentage of specific cell lineages
(to the right of vertical lines) and positive for the donor
epitope CD45.1 (above horizontal lines) is indicated on
each panel.
Because of the effects of radiation, BMT chimeric mice may be
somewhat less immunocompetent that normal mice. Therefore, we
used wt/wt (donor/recipient) BMT chimeric mice as controls for
the possible immunosuppressive effects of BMT. Fig. 3 shows that
ATL146e strongly inhibits liver IRI in wt/wt BMT chimera but
fails to exert any protection in ko/wt, wt/ko, or ko/ko BMT chimera. By contrast, another A2AR-selective agonist, ATL313, attenuates liver IRI in all BMT chimeric mice, except ko/ko chimera.
These finding raise two questions: why is ATL313 more effective at protecting liver from IRI than ATL146e, and why does
ATL146e fail to protect both ko/wt and wt/ko BMT chimeric
mice? A possible explanation is suggested by our finding that
ATL146e is rapidly (⬍30 s) converted to ATL146a when added to
mouse or rat blood (Table I). Interestingly, this rapid metabolism
occurs much more slowly (t1/2 ⬎ 30 min) in canine or human
blood, suggesting that an esterase that converts ATL146e to
ATL146a is much more active in rodent than in human or canine
blood. No metabolism of ATL313 was observed in blood from any
species, suggesting that replacement of the cyclohexyl moiety in
ATL146e with a piperidine moiety in ATL313 confers resistance to
esterase activity by converting the ester to the more stable nitrogencontaining carbamate. When injected into mice or rats, ATL146e was
also rapidly and completely converted into ATL146a. When injected
into dogs, ATL146e was converted to ATL146a with a t1/2 of ⬃10
min, possibly due to liver esterase activity (Fig. 4). The potency and
selectivity of ATL146e, ATL146a, and ATL313 to bind to recombinant adenosine receptor subtypes are summarized in Table I. Unlike
ATL313 and ATL146e, which are potent and selective A2AR agonists, ATL146a has equal affinity for the A2AR and the A3 adenosine
receptor. Hence, failure by ATL146e to inhibit liver injury may be
due to latent A3 adenosine receptor agonist activity that exerts a proinflammatory effect, particularly in chimeric mice that have a reduced
complement of cells that express the anti-inflammatory A2AR.
It is notable that ATL313 produces greater protection of the
liver from IRI in wt/ko than in ko/wt BMT mouse chimera. These
data indicate that most, but not all, liver protection is mediated by
The Journal of Immunology
FIGURE 3. Assessment of liver injury in BMT chimeric mice. Irradiated wt or A2AR ko mice received bone marrow transplants from wt or ko
mice to make wt/w, wt/ko, ko/wt, or ko/ko (donor/recipient) BMT chimeric
mice. Mice were subjected to liver ischemia for 1 h followed by reperfusion for 24 h, at which time serum ALT levels were measured. Vehicle,
ATL146e, or ATL313 was delivered immediately at reperfusion as described under Materials and Methods. Each bar represents the mean ⫾ SD
of 10 –12 animals. ⴱ, p ⬍ 0.05 vs vehicle; #, p ⬍ 0.001 vs vehicle.
A2ARs on bone marrow-derived cells. In contrast to the strong
protective effect from liver IRI produced by ATL146e in wt/wt
BMT chimeric mice, ATL146e appears to produce a small statistically insignificant increase in liver IRI in ko/wt BMT chimeric
mice based on serum ALT. H&E staining reveals more uniform
liver necrosis in ATL146e-treated than in vehicle-treated ko/wt
chimeric animals (Fig. 5, C and D). Similarly, immunohistochemical staining for neutrophils suggests a small proinflammatory response in ko/wt animals in response ATL146e (Fig. 6, C and D).
The clearest evidence that ATL146e causes a proinflammatory response in ko/wt mouse chimera is shown in Fig. 7. ATL146e increases IRI-stimulated induction of proinflammatory cytokines,
particularly RANTES, IP-10, IL-1␤, and IL-18. These data suggest
that ATL146e produces a latent proinflammatory response that is
unmasked when the A2AR is deleted partially in chimeric mice.
Evidence against transdifferentiation of transplanted bone
marrow cells into liver vascular endothelial cells
If the A2AR on endothelial cells mediate liver protection and if
bone marrow-derived cells repopulate liver endothelium, then the
deletion of the A2AR on endothelial cell precursors could account
for the absence of protection by ATL146e and ATL313 in ko/wt
BMT chimera. Hence, it was important to evaluate the degree to
which transdifferentiation of donor mouse-derived endothelial pro-
5043
FIGURE 4. Recovery of ATL146e and its metabolite, ATL146a, after
i.v. injection into a dog. ATL146e (10 ␮g/kg) was injected 5 min before
mixing a blood sample with acetonitrile. During HPLC, compounds were
monitored in tandem mass spectrometry-positive ion mode using a Thermofinnigan LCQ mass spectrometer as described in Materials and Methods. Results are typical of three experiments.
genitor cells repopulate the vascular endothelium in recipient mice
(Fig. 8). GFP/wt BMT chimera were examined 8 wk after BMT.
Some animals were subject to IRI, i.e., 1 h of ischemia followed by
24 h of reperfusion. Initial attempts to detect GFP by immunofluorescence in liver sections were unsuccessful because of high nonspecific background fluorescence in liver, especially in necrotic
areas. For this reason, GFP was detected using anti-GFP immunohistochemistry. Livers from wt mice were devoid of GFP immunoreactivity. All liver cells in sections of transgenic GFP livers
contained intense brown immunoperoxidase, including endothelial
cells (Fig. 8, A and B). Scattered GFP-immunoreactive cells were
seen in livers from GFP/wt BMT chimera, but no immunoreactivity was detected on endothelial cells lining the lumen of blood
vessels, either in uninjured livers (Fig. 8C) or following IRI with
reperfusion for 24 h (Fig. 8D). We conclude that the livers of mice
collected 8 wk after BMT with or without IRI of short duration (24
h) have no detectable endothelial cells derived from bone marrow
precursors. Therefore, the poor ability of ATL146e and ATL313 to
reduce IRI in ko/wt BMT chimera cannot be attributed to repopulation of liver endothelial cells by bone marrow precursors lacking
the A2AR.
Table I. Binding affinities of ATL146e, ATL146a, and ATL313 for
recombinant human adenosine receptor subtypesa
a
Data are the mean ⫾ SEM for binding to the high-affinity conformational state
of recombinant human adenosine receptors expressed in HEK-239 cells as determined
by competition for radioligand binding (34).
FIGURE 5. Effects of liver IRI and ATL146e on histological evidence
of liver damage. BMT mouse chimera, wt/wt (A and B) or ko/wt (C and D)
were subjected to liver IRI and treated with vehicle (A and C) or ATL146e
(B and D). Liver sections were stained with H&E after 24 h of reperfusion
to reveal living (purple granular) or necrotic (pink agranular) tissue. The
sections shown are representative of five similar experiments.
5044
FIGURE 6. Effects of liver IRI and ATL146e on immunohistological
evidence of neutrophil accumulation in liver. BMT mouse chimera, wt/wt
(A and B), or ko/wt (C and D) were subjected to liver IRI and treated with
vehicle (A and C) or ATL146e (B and D). Liver sections were immunostained for neutrophils (brown) after 24 h of reperfusion. The sections
shown are representative of five similar experiments.
Discussion
The purpose of this study was to determine whether bone marrowderived cells are responsible for the substantial liver protection
from IRI that is seen in response to A2AR activation during reperfusion. On the basis of the use of BMT to create chimeric mice, we
conclude that most of the effect of the selective A2AR agonist,
ATL313, to protect the liver from IRI are mediated by bone marrow-derived cells because little protection is noted in ko/wt chimeric mice that do not express the A2AR on bone marrow-derived
cells, and substantial protection is noted in wt/ko chimera that
express the A2AR only on bone marrow-derived cells. Nonbone
marrow-derived cells or bone marrow-derived cells that are not
efficiently repopulated within 8 wk of BMT also contribute somewhat to the effects of A2AR activation because ko/wt BMT chimeric mice are protected to a small degree by ATL313. Such cells
might consist of hepatocytes, endothelial cells, vascular smooth
muscle, or other parenchymal cells. They also could consist in part
of cells of bone marrow origin that are not efficiently repopulated,
such as a subset of ⬃15% of T lymphocytes (Fig. 2) or Kupffer
FIGURE 7. Effects of liver IRI and ATL146e on the induction of cytokine and chemokine transcripts in wt/wt and ko/wt BMT chimeric mice.
Mice were subjected to liver IRI and treated with vehicle (V) or ATL146e
for 24 h. Transcripts for various cytokines and chemokines were assessed
by RNase protection assays for BD Pharmingen multiprobe template set
mCK-2b and template set mCK-5c. The results shown are representative
for two to three replicate experiments with similar results. Little or no
transcripts for cytokines was visible in mRNA prepared from sham animals
not subjected to liver IRI (data not shown).
A2A ADENOSINE RECEPTORS AND LIVER PROTECTION
FIGURE 8. Source of endothelial cells in BMT chimeric mice. Liver
sections were stained with hematoxylin and immunostained with anti-GFP
Abs. Livers were derived from the following: (A) wt mouse following 1 h
of ischemia and 24 h of reperfusion; (B) transgenic GFP mice; (C)
GFP3 WT bone marrow chimera; (D) GFP3 WT bone marrow chimera
following 24 h of reperfusion after 60 min of ischemia. Black arrows denote examples of endothelial cells, and white arrows denote GFP-immunoreactive cells derived from donor GFP animals. The results are representative of three experiments.
cells, which appeared not to be repopulated in GFP/wt BMT chimera (Fig. 8).
ATL146e failed to protect either ko/wt or wt/ko BMT chimeric
mice from liver IRI. A possible explanation for this apparent paradox may be that ATL146e activates the A3 adenosine receptor,
which has been shown to be proinflammatory in rodents by enhancing hypoxia-induced mast cell degranulation (13) and to enhance endotoxin-induced killing of mice (14). Tissue resident mast
cell deregulation may contribute to inflammatory responses during
reperfusion after ischemia. Increased mast cell-specific proteases,
such as rat mast cell protease II, has been demonstrated in rat
intestine and mesentery subjected to IRI, suggesting that mast cell
degranulation occurs upon IRI and contributes to neutrophil
recruitment.
Although ATL146e is a highly selective agonist of the A2AR
based on binding to recombinant adenosine receptors (Table I), we
found in this study that it is rapidly (t1/2 ⬍ 30 s) converted to the
nonselective A2A/A3 agonist, ATL146a, when injected into rodents or mixed with rodent blood. By contrast, ATL146e has a t1/2
⬎ 30 min when mixed with human or canine blood and a t1/2 of
⬃10 min when injected into dogs. The anti-inflammatory effect of
the A2AR predominates over the proinflammatory effect of A3
adenosine receptor activation in mice expressing a normal complement of A2A and A3, i.e., in wt mice and in wt/wt BMT chimeric mice. In BMT chimeric mice, a latent proinflammatory component of ATL146a action appears to largely neutralize residual
anti-inflammatory activity. However, it is notable that there are no
data to indicate that activation of the A3 adenosine receptor has a
proinflammatory effect in nonrodent species. Unlike in rodent species, the adenosine receptor subtype that stimulates mast cell degranulation in dogs and human cells appears to be the A2B adenosine receptor (15, 16).
Neutrophil accumulation in the ischemic liver is reduced significantly by ATL146e administered at the time of reperfusion. Previous studies of liver damage using IRI models in rats and mice
have implicated neutrophils that infiltrate the injured tissue as a
source of reperfusion injury (17, 18). Activation of A2ARs suppresses the oxidative burst of neutrophils and inhibits leukocyte
The Journal of Immunology
extravasation into inflamed tissues. However, a delay in the administration of ATL146e for ⬎4 h after reperfusion results in loss
of liver protection by ATL146e, although most neutrophils extravasate into the liver ⬎ 4 h after the beginning of reperfusion injury
(9). Hence, the reduction in neutrophil accumulation with A2AR
activation may be primarily due to acute effects on a small number
of neutrophils or on other cells that reside in the liver or that
accumulate within a few hours of the beginning of reperfusion.
Some candidate targets of A2A agonists are mast cells (19),
platelets (20), macrophages (21), dendritic cells (22), and T lymphocytes (23). Macrophages and dendritic cells are APCs and
serve to initiate inflammation in response to invading microorganisms. In addition, they have been postulated to play a role in IRIinduced inflammation. Activation of the A2AR has been shown to
inhibit TNF-␣ and IL-12 release from macrophages (24). Immature dendritic cells do not make cAMP in response to adenosine,
but following treatment with LPS, mature dendritic cells respond
to adenosine, and this results in reduced IL-12 production in response to inflammation (25). Large numbers of T lymphocytes
accumulate late (48 –72 h) after reperfusion injury in tissues subjected to IRI. IFN-␥ production follows a similar temporal pattern.
The time course of T cell accumulation suggests that most of these
cells do not participate in liver protection by A2A agonists. However, a role for a small number of T cells in provoking injury was
suggested in a model of ex vivo cold IRI in which isolated livers
from nude mice were less injured and released less IFN-␥ (26) than
did livers from wt mice. Kupffer cells can recruit CD4⫹ T lymphocytes and neutrophils by releasing MIP-2, KC, and IP-10 (27).
We have recently shown that IFN-␥ release from isolated CD4⫹ T
cells is inhibited strongly by ATL146e and other A2A agonists
(23). These considerations suggest that activation of the A2AR on
liver macrophages and small numbers of T cells that are resident in
the liver or that accumulate shortly after reperfusion following
ischemia may contribute to the protection of the liver from reperfusion injury.
Early studies suggested that bone marrow-derived progenitor
cells can only differentiate into hemopoietic lineages. Subsequent
studies have shown that after transplantation, bone marrow cells
can differentiate into various differentiated cell types, including
vascular endothelial cells (28). Nevertheless, although hemopoietic stem cells clearly have the potential to transdifferentiate, this
has been noted only during the repair of damaged tissues and has
not be shown to occur in normal healthy tissues. For example,
transdifferentiation was observed after chemical injury to muscle
(29), in Duchenne’s muscular dystrophy (30), myocardial infarction (28), and tumor angiogenesis (31). Following BMT, endothelial cells derived from donor mice were found in blood vessels of
chimeric mice that had been injured by wire, graft vasculopathy, or
atherosclerosis (32, 33), but no donor cells were found in uninjured
vessels, even after radiation treatment. These results are consistent
with the conclusions of the current study in which GFP cells in
GFP/wt BMT mouse chimera cannot be found in the endothelium
or uninjured or even acutely injured (24 h of IRI) liver. Hence,
weak activity of ATL313 to protect liver from IRI in A2AR ko/wt
BMT chimera cannot be attributed to repopulation of liver blood
vessels with A2A ko endothelial cells derived from donor bone
marrow progenitors.
In conclusion, we have used BMT and A2AR ko mice to create
chimeric mice that selectively lack or selectively express the A2AR
on bone marrow-derived cells. The results suggest that most protection from liver reperfusion injury is due to activation of A2ARs
on bone marrow-derived cells because protection by ATL313 is
minimal in ko/wt BMT chimera and more substantial in wt/ko
BMT chimeric animals.
5045
Disclosures
J. Linden and M. Okusa own equity in Adenosine Therapeutics, LLC.
References
1. Reshef, A., O. Sperling, and E. Zoref-Shani. 1998. Adenosine-induced preconditioning of rat neuronal cultures against ischemia-reperfusion injury. Adv. Exp.
Med. Biol. 431:365.
2. Kurz, M. A., D. A. Bullough, C. J. Bugge, K. M. Mullane, and M. A. Young.
1997. Cardioprotection with a novel adenosine regulating agent mediated by
intravascular adenosine. Eur. J. Pharmacol. 322:211.
3. Day, Y. J., L. Huang, M. J. McDuffie, D. L. Rosin, H. Ye, J. F. Chen,
M. A. Schwarzschild, J. S. Fink, J. Linden, and M. D. Okusa. 2003. Renal protection from ischemia mediated by A2A adenosine receptors on bone marrowderived cells. J. Clin. Invest. 112:883.
4. Lasley, R. D., M. S. Jahania, and R. M. Mentzer. 2001. Beneficial effects of
adenosine A2a agonist CGS-21680 in infarcted and stunned porcine myocardium.
Am. J. Physiol. 280:H1660.
5. Peirce, S. M., T. C. Skalak, J. M. Rieger, T. L. MacDonald, and J. Linden. 2001.
Selective A2A adenosine receptors activation reduces skin pressure ulcer formation and inflammation. Am. J. Physiol. 281:H67.
6. Ross, S. D., C. G. Tribble, J. Linden, J. J. Gangemi, B. C. Lanpher, A. Y. Wang,
and I. L. Kron. 1999. Selective adenosine-A2A activation reduces lung reperfusion injury following transplantation. J. Heart Lung Transplant. 18:994.
7. Cassada, D. C., J. J. Gangemi, J. M. Rieger, J. Linden, A. K. Kaza, S. M. Long,
I. L. Kron, C. G. Tribble, and J. A. Kern. 2001. Systemic adenosine A2A agonist
ameliorates ischemic reperfusion injury in the rabbit spinal cord. Ann. Thorac.
Surg. 72:1245.
8. Ohta, A., and M. Sitkovsky. 2001. Role of G-protein-coupled adenosine receptors
in down-regulation of inflammation and protection from tissue damage. Nature
414:916.
9. Day, Y. J., M. A. Marshall, L. Huang, M. J. McDuffie, M. D. Okusa, and
J. Linden. 2004. Protection from ischemic liver injury by activation of A2A
adenosine receptors during reperfusion: inhibition of chemokine induction.
Am. J. Physiol. 286:G285.
10. Chen, J. F., Z. Huang, J. Ma, J. Zhu, R. Moratalla, D. Standaert,
M. A. Moskowitz, J. S. Fink, and M. A. Schwarzschild. 1999. A2A adenosine
receptor deficiency attenuates brain injury induced by transient focal ischemia in
mice. J. Neurosci. 19:9192.
11. Charbonneau, H., N. K. Tonks, K. A. Walsh, and E. H. Fischer. 1988. The
leukocyte common antigen (CD45): a putative receptor-linked protein tyrosine
phosphatase. Proc. Natl. Acad. Sci. USA 85:7182.
12. Morse, H. C., III. 1992. Genetic nomenclature for loci controlling surface antigens of mouse hemopoietic cells. J. Immunol. 149:3129.
13. Jin, H., R. Zastawny, S. R. George, and B. F. O’Dowd. 1997. Elimination of
palmitoylation sites in the human dopamine D1 receptor does not affect receptor-G protein interaction. Eur. J. Pharmacol. 324:109.
14. Sullivan, G. W., G. Fang, J. Linden, and W. M. Scheld. 2004. A2A adenosine
receptor activation improves survival in mouse models of endotoxemia and sepsis. J. Infect. Dis. 189:1897.
15. Auchampach, J. A., J. Jin, T. C. Wan, G. H. Caughey, and J. Linden. 1997.
Canine mast cell adenosine receptors: cloning and expression of the A3 receptors
and evidence that degranulation is mediated by the A2B receptor. Mol. Pharmacol. 52:846.
16. Linden, J., T. Thai, H. Figler, X. Jin, and A. S. Robeva. 1999. Characterization
of human A2B adenosine receptors: radioligand binding, Western blotting, and
coupling to Gq in human embryonic kidney 293 cells and HMC-1 mast cells. Mol.
Pharmacol. 56:705.
17. Harada, N., K. Okajima, K. Murakami, S. Usune, C. Sato, K. Ohshima, and
T. Katsuragi. 2000. Adenosine and selective A2A receptor agonists reduce ischemia/reperfusion injury of rat liver mainly by inhibiting leukocyte activation.
J. Pharmacol. Exp. Ther. 294:1034.
18. Jordan, J. E., Z. Q. Zhao, H. Sato, S. Taft, and J. Vinten-Johansen. 1997. Adenosine A2 receptor activation attenuates reperfusion injury by inhibiting neutrophil
accumulation, superoxide generation and coronary endothelial adherence.
J. Pharmacol. Exp. Ther. 280:301.
19. Fenster, M. S., R. K. Shepherd, J. Linden, and B. R. Duling. 2000. Activation of
adenosine A2 ␣ receptors inhibits mast cell degranulation and mast cell-dependent vasoconstriction. Microcirculation 7:129.
20. Hourani, S. M. 1996. Purinoceptors and platelet aggregation. J. Auton. Pharmacol. 16:349.
21. Day, Y. J., L. Huang, H. Ye, J. Linden, and M. D. Okusa. Renal ischemiareperfusion injury and adenosine 2A receptor-mediated tissue protection: the role
of macrophages. Am. J. Physiol. In press.
22. Dickenson, J. M., S. Reeder, B. Rees, S. Alexander, and D. Kendall. 2003. Functional expression of adenosine A2A and A3 receptors in the mouse dendritic cell
line XS-106. Eur. J. Pharmacol. 474:43.
23. Lappas, C. M., J. M. Rieger, and J. Linden. 2005. A2A adenosine receptor in-
5046
24.
25.
26.
27.
28.
duction inhibits IFN-␥ production in murine CD4⫹ T cells. J. Immunol.
174:1073.
Hasko, G., D. G. Kuhel, J. F. Chen, M. A. Schwarzschild, E. A. Deitch,
J. G. Mabley, A. Marton, and C. Szabo. 2000. Adenosine inhibits IL-12 and
TNF-␣ production via adenosine A2a receptor-dependent and independent mechanisms. FASEB J. 14:2065.
Panther, E., M. Idzko, Y. Herouy, H. Rheinen, P. J. Gebicke-Haerter,
U. Mrowietz, S. Dichmann, and J. Norgauer. 2001. Expression and function of
adenosine receptors in human dendritic cells. FASEB J. 15:1963.
Le Moine, O., H. Louis, A. Demols, F. Desalle, F. Demoor, E. Quertinmont,
M. Goldman, and J. Deviere. 2000. Cold liver ischemia-reperfusion injury critically depends on liver T cells and is improved by donor pretreatment with interleukin 10 in mice. Hepatology 31:1266.
Lentsch, A. B., H. Yoshidome, W. G. Cheadle, F. N. Miller, and M. J. Edwards.
1998. Chemokine involvement in hepatic ischemia/reperfusion injury in mice: roles
for macrophage inflammatory protein-2 and Kupffer cells. Hepatology 27:507.
Orlic, D., J. Kajstura, S. Chimenti, I. Jakoniuk, S. M. Anderson, B. Li, J. Pickel,
R. McKay, B. Nadal-Ginard, D. M. Bodine, A. Leri, and P. Anversa. 2001. Bone
marrow cells regenerate infarcted myocardium. Nature 410:701.
A2A ADENOSINE RECEPTORS AND LIVER PROTECTION
29. Ferrari, G., A. G. Cusella-De, M. Coletta, E. Paolucci, A. Stornaiuolo, G. Cossu,
and F. Mavilio. 1998. Muscle regeneration by bone marrow-derived myogenic
progenitors. Science 279:1528.
30. Gussoni, E., Y. Soneoka, C. D. Strickland, E. A. Buzney, M. K. Khan, A. F. Flint,
L. M. Kunkel, and R. C. Mulligan. 1999. Dystrophin expression in the mdx
mouse restored by stem cell transplantation. Nature 401:390.
31. Asahara, T., H. Masuda, T. Takahashi, C. Kalka, C. Pastore, M. Silver,
M. Kearne, M. Magner, and J. M. Isner. 1999. Bone marrow origin of endothelial
progenitor cells responsible for postnatal vasculogenesis in physiological and
pathological neovascularization. Circ. Res. 85:221.
32. Rookmaaker, M. B., H. Tolboom, R. Goldschmeding, J. J. Zwaginga,
T. J. Rabelink, and M. C. Verhaar. 2002. Bone-marrow-derived cells contribute
to endothelial repair after thrombotic microangiopathy. Blood 99:1095.
33. Sata, M., A. Saiura, A. Kunisato, A. Tojo, S. Okada, T. Tokuhisa, H. Hirai,
M. Makuuchi, Y. Hirata, and R. Nagai. 2002. Hematopoietic stem cells differentiate into vascular cells that participate in the pathogenesis of atherosclerosis.
Nat. Med. 8:403.
34. Murphree, L. J., M. A. Marshall, J. M. Rieger, T. L. MacDonald, and J. Linden.
2002. Human A2A adenosine receptors: high-affinity agonist binding to receptor-G protein complexes containing G␤4. Mol. Pharmacol. 61:455.