cardioprotective effects of a novel proteasome inhibitor following ischemia and reperfusion in the...
TRANSCRIPT
J Mol Cell Cardiol 31, 467–476 (1999)
Article No. jmcc.1998.0880, available online at http://www.idealibrary.com on
Cardioprotective Effects of a NovelProteasome Inhibitor Following Ischemiaand Reperfusion in the Isolated PerfusedRat HeartBarry Campbell, Julian Adams1, Yong K. Shin and Allan M. LeferDepartment of Physiology, Jefferson Medical College, Thomas Jefferson University, Philadelphia,PA 19107, USA and 1ProScript, Inc., 38 Sidney Street, Cambridge, MA 02139, USA
(Received 14 October, 1998, accepted in revised form 17 November 1998)
B. C, J. A, Y. K. S A. M. L. Cardioprotective Effects of a Novel Proteasome InhibitorFollowing Ischemia and Reperfusion in the Isolated Perfused Rat Heart. Journal of Molecular and Cellular Cardiology(1999) 31, 467–476. Ischemia followed by reperfusion in the presence of polymorphonuclear leukocytes (PMNs)results in cardiac contractile dysfunction as well as myocardial injury. These effects are due in large part toendothelial dysfunction leading to an upregulation of cell adhesion molecules and subsequent neutrophil inducedcardiac injury. The proteasome inhibitor, PS-519, has been shown to attenuate leukocyte-endothelial cellinteractions. We tested the effects of PS-519 on neutrophil mediated cardiac dysfunction in ischemia/reperfusion.This study examines the effects of PS-519 in a neutrophil dependent isolated perfused rat heart model of ischemia(I) (20 min) and reperfusion (R) (45 min). Administration of PS-519 (0.01, 0.1, 0.3, 1.0 mg/kg) to I/R heartsperfused with PMNs improved coronary flow, and preserved left ventricular developed pressure (LVDP) and+dP/dt max as indices of cardiac contractile function. At 1.0 mg/kg, PS-519 treated hearts exhibited a final LVDP of98±3% of initial compared to 52±8% in I/R hearts receiving only vehicle (P<0.001). In addition, PS-519significantly reduced PMN accumulation in the ischemic myocardium from 25.1±2.1 PMNs/mm2 in untreatedhearts to 7.3 PMNs/mm2, and attenuated P-selectin surface expression on coronary vascular endothelium from7.1±0.3% to 1.4±0.2% (P<0.01). These results provide evidence that PS-519 is a potent and effectivecardioprotective agent that inhibits P-selectin leukocyte-endothelial cell interactions and preserves cardiaccontractile function and coronary perfusion following myocardial ischemia and reperfusion.
1999 Academic Press
K W: Neutrophils; Myocardium; Ventricular function; NF-jB; P-selectin.
leukocytes (PMNs) are believed to play an importantIntroductionrole (Tsao et al., 1990; Entman et al., 1991). Al-though neutrophils play a vital role in the de-Early reperfusion of the ischemic myocardium plays
an important role in minimizing myocardial tissue struction of foreign antigens and the remodeling ofinjured tissues, many activated PMNs accumulateinjury associated with acute myocardial infarction.
However, reperfusion of the myocardium itself in the reperfused microvasculature upon re-perfusion, resulting in microvascular plugging andresults in enhanced myocardial injury (Tsao et al.,
1990; Forman et al., 1993) The process of re- an impairment in coronary perfusion (Engler et al.,1983; Lefer et al., 1991). The activated PMNsperfusion injury is characterized by an in-
flammatory response in which polymorphonuclear induce tissue injury by the release of a variety of
Please address all correspondence to: Dr Allan M. Lefer, Department of Physiology, Jefferson Medical College, 1020 Locust Street,Philadelphia, PA 19107, USA.
0022–2828/99/020467+10 $30.00/0 1999 Academic Press
B. Campbell et al.468
cytotoxic substances including oxygen derived free of PS-519 were given. We allowed 5 min for thePS-519 to reach an equilibrium concentration inradicals, inflammatory cytokines, and proteolytic
enzymes (Weiss, 1989). Many of these substances the body before proceeding. Following a midlinethoracotomy, the hearts were rapidly excised andare thought to mediate vascular endothelial dys-
function as well as contribute to myocardial injury the ascending aorta was cannulated. Retrogradeperfusion of the non-working heart was initiated(Buerke et al., 1994). This concept is also consistent
with evidence that either decreasing the number of with a modified Krebs buffer maintained at 37°Cand at a constant pressure of 80 mmHg. The Krebscirculating PMNs or administration of monoclonal
antibodies directed against cell adhesion molecules buffer had the following composition (in mmol/l):glucose, 17; NaCl, 120: NaHCO3, 25; CaCl2, 2.5;can lead to a significant cardioprotection against
reperfusion injury (Romson et al., 1983; Ma et al., EDTA, 0.5; KCl, 5.9; and MgCl2, 1.2. The perfusatewas aerated with 95% O2+5% CO2 which was1992; Weyrich et al., 1993; Lefer et al., 1994).
Upregulation of cell adhesion molecules (i.e. P- equilibrated at a pH of 7.3–7.4. Two sidearms ofthe apparatus in the perfusion line just proximalselectin, E-selectin, ICAM-1) has been shown to
increase leukocyte–endothelial cell interactions and to the heart inflow cannula allowed the infusion ofPMNs and autologous plasma. To assess cardiacto ultimately cause endothelial dysfunction (Lefer
et al., 1994). Activation of NF-jB has been shown contractile function, a 2.5 Fr microtip cathetertransducer (Millar Instruments, Inc., Houston, TX,to upregulate E- and P-selectin (Read et al., 1994;
Read et al., 1995; Adams and Stein, 1996; Weyrich USA) was inserted directly into the left ventricularcavity as previously reported (Pabla et al., 1996;et al., 1995; Meyer et al., 1997). However, P-selectin
may also be upregulated more rapidly (i.e. in 10–20 Lefer et al., 1997). The left ventricular pressure,maximal rate of development of left ventricularmin) by translocation to the endothelial cell surface
(McEver et al., 1989; Weyrich et al., 1993). Ad- developed pressure (+dP/dt max) and coronaryflow were all recorded using a MacLab data ac-ditional studies using proteasome inhibitors, which
play an important role in the activation of NF- quisition system (ADI Diagnostics Inc., Castle Hill,NSW, Australia) in conjunction with a Power Mac-jB, have also inhibited the activation of the cell
adhesion molecules (CAMs) (Read et al., 1995). In intosh 7600 computer (Apple Computers, Cu-pertino, CA, USA). All of the data were stored andthis study, we have shown that PS-519 protects
against cardiac contractile dysfunction and PMN analysed at the end of each experiment.accumulation associated with ischemia reperfusioninjury in a carefully controlled model of rat myo-cardial ischemia/reperfusion, which is dependent Rat neutrophil isolationupon neutrophils to mediate the cardiac contractiledysfunction. The study also demonstrates that PS- PMNs were isolated from whole blood by the method
of Williams et al. (1987) using the hetastarch ex-519 attenuates PMN adherence to the endothelium.change transfusion method in pentobarbital an-esthetized (40 mg/kg, i.p.) Sprague–Dawley rats(300–350 g). This method yielded 110–130×106Materials and methodsPMNs per rat which were >95% pure and >95%viable. These PMNs were washed five to six timesAll experiments reported in this study conform
to Thomas Jefferson University IACUC guidelines to remove the hetastarch prior to perfusion into rathearts. Finally, the PMNs were resuspended in Krebsregarding the use of animals in the laboratory as
well as to the standards established in the Guide buffer and counted using a hemocytometer andmicroscope.for the Care and Use of Laboratory Animals (NIH,
Publication Number 85-23, revised 1985).
Rat plasma preparationIsolated rat heart experiments
An intracardiac puncture was performed with a20 ml plastic syringe and a 20 gauge needle (BectonMale Sprague–Dawley rats (250–300 g) were an-
esthetized with 40 mg/kg sodium pentobarbital and Dickinson and Co., Franklin Lakes, NJ, USA)containing 2.0 ml of sodium citrate–administered 1000 U sodium heparin i.p. (Elkins
Sinn Inc., Cherry Hill, NJ, USA) . The PS-519 phosphate–dextrose solution in anesthetized rats toobtain whole blood. The blood was spun im-(ProScript, Inc., Cambridge, MA, USA) was diluted
in 25 ll DMSO and 175 ll saline, and i.v. injections mediately in a refrigerated centrifuge (Beckman
Cardioprotective Actions of a Proteasome Inhibitor 469
25
0
I/R
Cor
onar
y fl
ow (
ml/m
in)
20
15
10
5
I/R
+10
0 00
0 00
0 P
MN
s
I/R
+D
MS
O t
reat
edh
eart
+10
0 00
0 00
0 P
MN
s
I/R
+P
S-5
19 t
reat
edh
eart
(0.
01 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
1 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
3 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(1.
0 m
g/kg
) +
100
000
000
PM
Ns
7 7 7 7 5 6 6
NSP < 0.01
NS NS
NSP < 0.05 P < 0.05
Figure 1 Initial (Φ) and final (Ε) coronary flow (CF) expressed in ml/min in the isolated perfused rat hearts subjectedto global total ischemia for 20 min and 45 min of reperfusion. Ischemic hearts were perfused in the presence or absenceof PMNs (100 000 000). All values are expressed as means ±... Numbers at the bottom of the bars represent thenumber of hearts.
100
0
I/R
LVD
P (
mm
Hg)
80
60
40
20
I/R
+10
0 00
0 00
0 P
MN
s
I/R
+D
MS
O t
reat
edh
eart
+10
0 00
0 00
0 P
MN
s
I/R
+P
S-5
19 t
reat
edh
eart
(0.
01 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
1 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
3 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(1.
0 m
g/kg
) +
100
000
000
PM
Ns
7 7 7 5 6 6
NS P < 0.001P < 0.05 NSP < 0.001 P < 0.001
P < 0.01
7
Figure 2 Initial (Φ) and final (Ε) left ventricular developed pressure (LVDP) expressed in mm Hg in the isolatedperfused rat hearts prior to ischemia and following reperfusion. Ischemic hearts were perfused in the presence orabsence of PMNs. PMNs induced a marked contractile dysfunction which was attenuated by the PS-519. All valuesare expressed as means ±... Numbers at the bottom of the bars represent the number of hearts.
B. Campbell et al.470
5000
0
I/R
+ dP
/dt
max
(m
mH
g/s)
4000
3000
2000
1000
I/R
+10
0 00
0 00
0 P
MN
s
I/R
+D
MS
O t
reat
edh
eart
+10
0 00
0 00
0 P
MN
s
I/R
+P
S-5
19 t
reat
edh
eart
(0.
01 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
1 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
3 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(1.
0 m
g/kg
) +
100
000
000
PM
Ns
7 7 7 5 6 6
NS P < 0.001P < 0.05
NSP < 0.001P < 0.001
P < 0.01
7
Figure 3 Initial (Φ) and final (Ε) first derivative of LVDP (+dP/dt max) expressed in mmHg/s in rat hearts subjectto ischemia and reperfusion. Ischemic hearts were perfused in the presence or absence of PMNs. PMNs produced asignificant impairment which was eliminated by the PS-519. All values are expressed as means ±... Numbers atthe bottom of the bars represent the number of hearts.
Instruments, Inc., Palo Alto, CA, USA) at 2060×g The plasma was also placed in a 5.0 ml syringelocated just proximal to the inflow port to thefor 10 min, and the plasma was decanted. The
plasma was infused along with the PMNs in order coronary circulation. The hearts were allowed toreperfuse for a total of 45 min during which timeto more closely simulate conditions present in vivo.the data was collected every 5 min for the first 30min and at the 45 min time point. The proteasomeinhibitor, PS-519, was given at 0.01, 0.1, 0.3 andPerfused heart experimental protocol1.0 mg/kg in a series of five to seven rats in eachgroup.After the first 15 min of perfusion, baseline left
ventricular developed pressure LVDP,+dP/dt max,and coronary flow were measured every 5 min for15 min to ensure complete equilibration of the Histology and immunohistochemistryhearts. LVDP was defined as left ventricular systolicminus left ventricular end-diastolic pressure. In all In the original rats, some hearts were used to assay
the amount of PMNs in the heart by histologicalhearts, both the initial and final LV end-diastolicpressures were 4–8 mmHg. The first derivative of staining. After the final reading of reperfusion the
heart was removed from the perfusion apparatusleft ventricular pressure (+dP/dt max) was re-corded from instantaneous left ventricular pressure. and was placed in 4% paraformaldehyde overnight
at 4°C. The next day the tissue was cut into sectionsFlow of the Krebs buffer was then reduced to 0,creating a state of total global ischemia for 20 min. and dehydrated using graded acetone washes at
4°C. Tissue sections were embedded in plastic (Im-The flow was then allowed to return to values nearcontrol levels by re-establishing coronary perfusion munobed; Polysciences Inc., Warrington, PA, USA),
and 4 lm-thick sections were cut and transferred toat 80 mmHg. At reperfusion, 100×106 PMNs and5 ml of plasma was infused directly into the cor- Vectabond-coated slides (Vector Laboratories, Inc.,
Burlingame, CA, USA). The slides were rinsed inonary circulation over a period of 5 min via a setof separate side ports situated just proximal to ethanol for 10 min to remove some of the plastic
embedding and allow the tissue to stain. The tissuethe heart in the perfusion line. The PMNs weresuspended in 5.0 ml Krebs buffer in a 5.0 ml syringe. sections were then stained with either Hematoxylin
Cardioprotective Actions of a Proteasome Inhibitor 471
35
0
I/R
PM
Ns/
mm
2
20
15
10
5
I/R
+10
0 00
0 00
0 P
MN
s
I/R
+D
MS
O t
reat
edh
eart
+10
0 00
0 00
0 P
MN
s
I/R
+P
S-5
19 t
reat
edh
eart
(0.
01 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
1 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(0.
3 m
g/kg
) +
100
000
000
PM
Ns
I/R
+P
S-5
19 t
reat
edh
eart
(1.
0 m
g/kg
) +
100
000
000
PM
Ns
P < 0.001
30 30 30 30 30 30 30
NSP < 0.01
P < 0.001
30
25
Infiltrated PMNs in the heart
Figure 4 Histological analysis of trapped PMNs following hematoxylin-eosin staining of rat hearts subject to ischemia(20 min) and reperfusion (45 min). Ischemic hearts were perfused in the presence or absence of PMNs (100 000 000).The PMNs which had adhered or transmigrated were then counted. All values are expressed as mean number of PMNs/mm2 heart area ±... Numbers at the bottom of the bars represents the number of fields counted.
solution Gill No.3 for 10 min (Sigma Chemical Co.) neutrophils were then labeled with a Zynaxis PKH-and Giemsa stain for 3 min (Sigma Chemical Co.). 2 cell linker (Zynaxis Cell Science Inc., preparedThe slides were then observed microscopically, and for Sigma Immunochemicals, Malvern, PA, USA)the PMNs were counted. Immunohistochemistry according the procedure described by Yuan andfor P-selectin was performed on tissue sections in Fleming (1990). Two milliliters of diluent and 10 llfive to six hearts from each group according to dye were added to a loose cell pellet containingpreviously described techniques (Weyrich et al., approximately 40 000 000 cells. Following a 7 min1993). The basic method used was avidin-biotin incubation period, 200 ll of 0.2% bovine serumimmunoperoxidase technique with monoclonal albumin was added to stop the reaction, 2 ml phos-antibody PB1.3 (Cytel, San Diego, CA, USA) as the phate buffered saline was added, and the cells weremonoclonal antibody directed against P-selectin. resuspended. The mixture was then centrifuged forThis antibody recognizes only surface expressed P- 10 min at 700×g. Cells were washed three timesselectin (Weyrich et al., 1993). Positive staining after which the cells were resuspended in phosphatewas defined as a coronary microvessel displaying buffered saline, counted, and utilized in the ad-brown immunoperoxidase reaction product on herence assay.>50% of the circumference of its endothelium. Fifty Donor rats which had been injected with PS-519vessels per tissue sample were examined in each of (1 mg/kg, 0.1 mg/kg, 0.3 mg/kg, and 0.001 mg/five or six hearts per group. kg) 5 min prior to the removal of the superior
mesenteric arteries. They were then placed inwarmed K-H buffer, and cleaned of all externaladipose and loose connective tissue. Arterial seg-Neutrophil adherence to SMA endotheliumments were sectioned into 2–3 mm rings, opened,and placed into wells containing 2 ml Krebs–Neutrophils were isolated from donor rats utilizingHenseleit buffer. SMA segments were incubated forthe method previously described by Williams et al.10 min at 37°C with thrombin (2 U/ml) in a shaker(1987). Cells were washed five times prior to la-
beling. Following the washing cycles, isolated bath. Segments were then transferred to wells con-
B. Campbell et al.472
16
0I/R
% P
-Sel
ecti
n e
xpre
ssio
n14
12
10
8
I/R +100 000 000
PMNs
I/R +DMSO +
100 000 000 PMNs
I/R +PS-519 (1 mg/kg) +100 000 000 PMNs
Thrombin(1 U/ml)
6
4
2
6 6 5 6
P < 0.01
NS
NS
P < 0.01
5
Figure 5 P-selectin expression in rat coronary microvascular endothelium determined by immunohistochemistryfollowing heart perfusion under conditions of global ischemia/reperfusion (I/R) in the absence or presence of PMNs, orunder control non-ischemic conditions in the presence of 1 U/ml of thrombin.
taining fresh K–H buffer and incubated with PKH- model of myocardial ischemia reperfusion, we per-2-labeled neutrophils (106 cells). Following a 20 fused rat hearts at control flow for 80 min, or formin incubation period, sections were washed in 15 min of control flow followed by 20 min of totalK–H solution and placed endothelial side up on global ischemia and 45 min of reperfusion at controlglass microscope slides. Neutrophils adhering to the flows in the presence or absence of PMNs andendothelium were counted using epifluorescence plasma. Perfusion of rat hearts with PS-519, at allmicroscopy (Nikon Diaphot, Nikon Inc., Garden concentrations, was performed in control heartsCity, NY, USA). Five different fields of each endo- at a perfusion pressure of 80 mmHg for 80 min.thelial surface were counted and the results ex- Perfusion with PS-519 at 80 mmHg during shampressed as adherent neutrophils/mm2 of endothelial ischemia or during ischemia/reperfusion withoutsurface. PMNs resulted in no change in coronary flow (CF),
left ventricular developed pressure (LVDP), or thefirst derivative of LVDP (+dP/dt max), indicating
Statistical analysis that PS-519 did not exert any direct effects oncardiodynamics (Figs 1–3). Also, perfusion of non-
All data in the text and figures are presented as ischemic hearts with PMNs did not alter any of themean±... The data on left ventricular function cardiac function variables measured, indicating thatand coronary flow were analysed by ANOVA, in- PMNs do not induce cardiac dysfunction in normalcorporating repeated measures. Probability values non-ischemic hearts. Only in ischemic reperfusedof 0.05 or less were considered to be statistically rat hearts perfused with PMNs was there a markedsignificant. reduction in cardiac contractile function and CF.
Final CF decreased to 67±4% of initial in untreatedischemic-reperfused hearts and hearts treated with0.1, 0.3 and 1.0 mg/kg PS-519 significantly im-Resultsproved CF initial values (Fig. 1).
Changes in left ventricular function (i.e. LVDP)Experimental protocolwere similar to that of coronary flow. In ischemicreperfused rat hearts perfused with PMNs, LVDPTo determine whether PS-519 can attenuate leu-recovered only partially following reperfusion sta-kocyte–endothelial interactions and improve car-
diac contractile function in a well characterized bilizing at a LVDP of 56±4% of initial (P<0.001)
Cardioprotective Actions of a Proteasome Inhibitor 473
140
0
Non
-sti
mu
late
d
Adh
ered
PM
Ns/
mm
2
100
80
60
20
No
addi
tion
s
DM
SO
tre
ated
PS
-519
tre
ated
(0.0
1 m
g/kg
)
PS
-519
tre
ated
(0.1
mg/
kg)
PS
-519
tre
ated
(0.
3 m
g/kg
)
PS
-519
tre
ated
(1.0
mg/
kg)
NS
P < 0.001
120
40
8 8 6 8 4 8
P < 0.01
P < 0.001
8
Rat SMA endothelium
Figure 6 Rat superior mesenteric artery (SMA) adherence is expressed as PMNs/mm2. Rat SMA were isolated fromPS-519 treated rats or their vehicle and were either non-stimulated (Φ) or were stimulated with thrombin (Ε) (2 U/ml). The PMN which had adhered to the endothelium were counted and analysed. All values are expressed as mean±... Numbers at the bottom of the bars indicates the number of segments counted.
(Fig. 2). However, in hearts perfused with the same perfused rat hearts subjected to I/R+100 000 000PMNs exhibited a large number of adhered andnumber of PMNs under the same I/R conditions,transmigrated PMNs (P<0.001) compared to that of0.1, 0.3 and 1.0 mg/kg PS-519 showed a significantI/R hearts that did not receive the PMNs (5.3±2.1).protective effect in LVDP (Fig. 2). The lowest doseHearts which had been treated with the vehicleof PS-519 did not significantly maintain LVDP.were not significantly different from the I/R+ PMNThese same relationships were obtained with thegroup. At the three highest doses of PS-519, therefirst derivative of LVDP (+dP/dt max). There waswas a significant reduction of PMN accumulationa 50±3% reduction in the final +dP/dt max inin the cardiac muscle. The lowest dose of PS-519untreated PMN perfused hearts subject to ischemia/(0.01 mg/kg) did not show any difference from I/reperfusion (P<0.001) (Fig. 3). However, ischemic/R+PMNs alone.reperfused hearts given PS-519 at the 0.1, 0.3 or
Furthermore, I/R+PMNs resulted in a significant1 mg/kg dose showed a progressive and markedincrease in P-selectin surface expression on therecovery of cardiac contractility, comparable to thatcoronary vascular endothelium (Fig. 5). This trans-of control values (i.e. 98±2% of control) at thelocation of P-selectin from Weibel–Palade bodies1.0 mg/kg dose. The lowest dose of PS-519 (0.01to the endothelial cell surface was unaffected bymg/kg) was ineffective in this regard (Fig. 3).perfusion in DMSO treated rats, but was almostcompletely attenuated by rats given 1 mg/kg of PS-519 (P<0.01). Thus, PS-519 clearly suppresses
Histology and immunohistochemistry the P-selectin upregulation occurring in ischemia/reperfusion in the presence of PMNs. These data are
In order to determine whether PMNs actually in- compared to control non-ischemic hearts perfusedfiltrated into the myocardium, histological sections with 1 U/ml of thrombin for the same period of
time.of the cardiac muscle were observed (Fig. 4). Isolated
B. Campbell et al.474
SMA adherence effects of PMNs. Furthermore, a 90% recovery ofcardiac contractile function was observed in PS-
We also observed the effects of PS-519 on au- 519 (1.0 mg/kg) treated ischemic-reperfused heartsas compared to only 55% in those I/R hearts giventologous neutrophil adherence to rat superior mes-
enteric artery endothelial cells (Fig. 6). Few only its vehicle. A 75% inhibition of PMN ac-cumulation in the hearts was also observed, and thisneutrophils adhered to unstimulated superior mes-
enteric artery endothelium, suggesting that the attenuation of leukocyte-endothelium interactionappears to be the key cardioprotective effect ofisolation of neutrophils as well as of superior mes-
enteric artery segments failed to significantly stimu- PS-519. Consistent with this concept, this studydemonstrated that PS-519 exhibited a 70% pro-late either cell type. We therefore stimulated rat
superior mesenteric artery endothelium with 2 U/ tection in PMN adherence to thrombin stimulatedcoronary endothelium. Without PS-519, neutro-ml thrombin. Stimulation of rat superior mesenteric
artery endothelium by thrombin resulted in a sig- phils adhered to the vascular endothelium wherethey are able to release cytotoxic substances suchnificant (P<0.001) increase in neutrophil ad-
herence to the endothelium, indicating that the as proteases, eicosanoids, cytokines, and oxygen-derived free radicals (Weiss, 1989), each of whichendothelium was not injured by the isolation pro-
cedures. Incubation of superior mesenteric artery can mediate tissue injury and exacerbate endo-thelial dysfunction. These humoral mediators havesegments isolated from rats which had been treated
with PS-519 at a dose of 1 mg/kg (P<0.001) re- been found to lead to coronary endothelial injury,disruption of the basement membranes, PMN ex-duced adherence to thrombin stimulated segments
in a dose-dependent manner. Thus, PS-519 at- travasation, and myocardial necrosis (Smedley etal., 1986; Weiss, 1989; Lefer et al., 1991; Entmantenuates PMN adherence to the endothelium, an
action which may be important in explaining its and Smith, 1994).Recruitment of PMNs to inflammatory tissue con-cardioprotective effects.
sists of a series of sequential processes which aremediated by different cell adhesion molecules loc-ated either on the PMNs or the endothelial cellsDiscussion(Butcher, 1991). Upon reperfusion, an increasednumber of PMNs start to roll along the endothelialThe present study clearly demonstrates that PS-
519 exerts significant cardioprotective actions in a surface, a process mainly mediated by P-selectinupregulated on the endothelium and to a lesserLangendorff perfused heart model of myocardial
ischemia/reperfusion. The cardioprotective effects extent by constitutively expressed L-selectin onPMNs (Kishimoto et al., 1990; Butcher, 1991).of this proteasome inhibitor were characterized by
a significant maintenance of post-reperfusion cor- Rolling PMNs are then able to engage in firmadhesion to the endothelium mainly by b2 integrinonary flow, left ventricular developed pressure, and
the first derivative of left ventricular developed (i.e. CD11/CD18) interaction with intercellular ad-hesion molecule-1 (ICAM-1) (Butcher, 1991). Afterpressure (i.e.+dP/dt max), indicating a significant
attenuation of cardiac dysfunction. It is unlikely firm adhesion, many of the PMNs then becomeactivated, flatten out, and undergo transendothelialthat PS-519 exerted its cardioprotective effects by
directly influencing hemodynamics (i.e. inducing migration to the injured or inflamed tissue (Vapor-ciyan et al., 1993). It is known that the respiratorycoronary vasodilation or increased cardiac con-
tractility) since there were no increases in coronary bursts of PMNs (i.e., the production of oxygen-derived free radicals) at the onset of reperfusionflow, left ventricular developed pressure, or the first
derivative of left ventricular developed pressure in coincides with a marked reduction in nitric oxiderelease from the endothelium (Rubanyi and Van-non-ischemic perfused rat hearts given PS-519.
Since myocardial reperfusion injury has been houtte, 1986), which may in turn lead to anenhanced expression of P-selectin on the endothelialshown to be related to PMNs infiltrating into isch-
emic cardiac tissue (Engler et al., 1986; Tsao et surface (Larsen et al., 1989; Geng et al., 1990;Davenpeck et al., 1994). In recent studies, in vivoal., 1990; Weyrich et al., 1993), one important
component of the protection afforded by PS-519 administration of monoclonal antibodies directedagainst various cell adhesion molecules were shownis related to its inhibitory effect on cell adhesion
molecules (CAMs). This inhibition in turn at- to significantly attenuate PMN adherence to theendothelium and reduce myocardial necrosis aftertenuates PMN accumulation in the ischemic myo-
cardium. In this model of ischemia-reperfusion, myocardial ischemia-reperfusion (Ma et al., 1992;Weyrich et al., 1993). In the present study, wemost of the contractile dysfunction is due to the
Cardioprotective Actions of a Proteasome Inhibitor 475
B M, W AS, L AM, 1994. Isolated cardiacreport that PS-519 prevents the upregulation ofmyocytes are sensitized by hypoxia-reoxygenation toP-selectin onto the endothelial cell surface. Thisneutrophil released mediators. Am J Physiol 266: H128-
upregulation of P-selectin is most likely due to H136.inhibition of translocation of P-selectin to the endo- B EC, 1991. Leukocyte-endothelial cell re-
cognition: three (or more) steps to specificity and di-thelial cell surface, a process which requires onlyversity. Cell 67: 1033-1036.10–20 min (McEver et al., 1989). Whether in-
D KL, G TW, L AM, 1994. In-hibition of NF-jB by PS-519 also occurs in thishibition of endothelial-derived nitric oxide promotes P-
setting remains to be determined later. selectin expression and actions in the rat micro-In the present study, PS-519 protected cardiac circulation. Gastroenterology 107: 1050-1058.
E RL, S-S̈ GW, P RS, 1983.function after ischemia and reperfusion with PMNs.Leukocyte capillary plugging in myocardial ischemiaIt is known that endothelial dysfunction occursand reperfusion in the dog. Am J Pathol 111: 98-111.within 2.5–5 min of the onset of reperfusion (Tsao
E RL, D MD, M DD, P MA,et al., 1990). This early dysfunction leads to the S-S̈ GW, 1986. Role of leukocytes inupregulation of P-selectin within 10–20 min of re- response to acute myocardial ischemia and reflow in
dogs. Am J Physiol 251: H314-H323.perfusion (Weyrich et al., 1993). Earlier studiesE ML, S CW, 1994. Postreperfusion in-have shown that the loss of endothelium-derived
flammation: a model for reaction to injury in cardio-nitric oxide further facilitates PMN adherence tovascular disease. Cardiovasc Res 28: 1301-1311.
the vascular endothelium (Kubes et al., 1991; Lefer E ML, M L, R RD, D WJ, A-and Lefer, 1994). These adherent PMNs are ac- DC, T AA, S CW, 1991. Inflammation
in the course of early myocardial ischemia. FASEB Jtivated and release mediators which aggravate5: 2529-2537.endothelial dysfunction, resulting in an increased
F MB, K FD, J M, V R, 1993.PMN adherence to the vascular endothelium andEndothelial and myocardial injury during ischemia
myocardial necrosis (Kubes et al., 1991). Because and reperfusion: pathogenesis and therapeutic im-nitric oxide inhibits PMN adherence to the endo- plications. J Am Coll Cardiol 21: 1245-1253.
G JG, B MP , M KL, MI TM,thelium, the diminished PMN adherence to theP SM, K JM, B GA, Z GA,coronary endothelium after reperfusion may beME RP, 1990. Rapid neutrophil adhesion to ac-explained at least in part by preservation of thetivated endothelium mediated by GMP-140. Nature
endothelium by PS-519, eventually preventing 343: 757-760.myocardial necrosis and cardiac dysfunction (Lefer K TK, J MA, B EC, 1990. Iden-
tification of a human peripheral lymph node homingand Lefer, 1994).receptor: a rapidly down-regulated adhesion molecule.In conclusion, we have demonstrated that atProc Natl Acad Sci USA 87: 2244-2248.doses of 0.1–1.0 mg/kg, the proteasome inhibitor
K P, S M, G DN, 1991. Nitric oxide:PS-519 can attenuate cardiac contractile dys- an endogenous modulator of leukocyte adhesion. Procfunction related to myocardial ischemia-reperfusion Natl Acad Sci USA 88: 4651-4655.
L E, C A , G GE, F BC, E JK,injury. These cardioprotective effects appear to beB R, W DD, F B, 1989. PADGEMrelated to inhibition of cell adhesion molecule ex-protein: A receptor that mediates the interaction ofpression. This study provides strong evidence thatactivated platelets with neutrophils and monocytes.
PS-519 has important beneficial effects in acute Cell 59: 305-312.inflammatory states such as that occurring in the L AM, L DJ, 1994. Pharmacology of the endo-
thelium in ischemia-reperfusion and circulatory shock.reperfusion of the ischemic myocardium.Ann Rev Pharmacol Toxicol 33: 71-90.
L AM, T PS, L DJ, M X-L, 1991. Role ofAcknowledgments endothelial dysfunction in the pathogenesis of re-
perfusion injury after myocardial ischemia. FASEB J5: 2029-2034.This study was supported in part by Research Grant
L AM, W AS, B M, 1994. Role ofGM-45434 from the National Institute of Generalselectins: a new family of adhesion molecules, in isch-Medical Sciences, National Institute of Health. aemia-reperfusion injury. Cardiovas Res 28: 289-294.
The authors would like to gratefully acknowledge L DJ, S R, C B, N T, H R,Robert Craig for his expert technical assistance S M, G J, L AM, 1997. Peroxynitrite
inhibits leukocyte-endothelial cell interactions and pro-during the course of these investigations.tects against ischemia-reperfusion injury in rats. J ClinInvest 99: 684-691.
References M X-, L DJ, L AM, R R, 1992. Coronaryendothelial and cardiac protective effects of a mono-
A J, S R, 1996. Novel inhibitors of the pro- clonal antibody to intercellular adhesion molecule-1in myocardial ischemia and reperfusion. Circulation 86:teasome and their therapeutic use in inflammation.
Ann Rep Med Chem 31: 279-288. 937-946.
B. Campbell et al.476
ME RP, B JH, M KL, M-C cells. Enhancement by endotoxin and essential role ofneutrophil elastase. J Clin Invest 77: 1233-1243.L, B DF, 1989. GMP-140, a platelet a-granule
membrane protein, is also synthesized by vascular T PS, A N, L DJ, J III G, L AM,1990. Time course of endothelial dysfunction andendothelial cells and is localized in Weibel–Palade bod-
ies. J Clin Invest 84: 92-99. myocardial injury during myocardial ischemia andreperfusion in the cat. Circulation 82: 1402-1412.M S, K NG, J A, 1997. Cyclosporine A is
an uncompetitive inhibitor of proteasome activity and V A, DL A, Y HM, M HC II,T SR, J ML, W PA, A SM, 1993.prevents NF-jB activation. FEBS Lett 413: 354-358.
P R, B AJ, F DM, B SA, S AM, Involvement of platelet-endothelial cell adhesionmolecule-1 in neutrophil recruitment in vivo. ScienceC MJ, L DJ, 1996. Nitric oxide attenuates
neutrophil-mediated myocardial contractile dys- 262: 1580-1582.W SJ, 1989. Tissue destruction by neutrophils. N Englfunction after ischemia and reperfusion. Circ Res 78:
65-72. J Med 320: 365-376.W AS, M X-, L DJ, A KH, LR MA, N AS, L FW, P VJ,
M T, C T, 1995. The proteasome pathway AM, 1993. in vivo neutralization of P-selectin protectsfeline heart and endothelium in myocardial ischemiais required for cytokine-induced endothelial-leukocyte
adhesion molecule expression. Immunity 2: 493-506. and reperfusion injury. J Clin Invest 91: 2620-2629.W AS, MI TM, ME RP, P SM,R MA, W MZ, W AJ, C T, 1994.
NF-jB and IjBa: An inducible regulatory system in Z GA, 1995. Monocyte tethering by P-selectin regulates monocyte chemotactic protein-1 andendothelial activation. J Exp Med 179: 503-512.
R JL, H BGF, K SL, A GD, S tumor necrosis factor a secretion: Signal integrationand NF-jB translocation. J Clin Invest 95: 2297-2303.AS, L BR, 1983. Reduction of the extent of
ischemic myocardial injury by neutrophil depletion in W JH J, M KM, U T, C MS, 1987.Harvesting the circulating pool of polymorphonuclearthe dog. Circulation 67: 1016-1023.
R GM, V PM, 1986. Oxygen derived free leukocytes by hetastarch exchange transfusion (HET):Yield and function assessment. J Leuko Biol 42: 455-radicals, endothelium and responsiveness of vascular
smooth muscle. Am J Physiol 250: H815-H821. 462.Y Y, F BP, 1990. A method for isolation andS LA, T MG, S RA, H C,
G LA, J RB J H PM, W fluorescent labeling of rat neutrophils for intravitalmicrovascular studies. Microvasc Res 40: 218-229.GS, 1986. Neutrophil-mediated injury to endothelial