|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1Department of Surgery, Clinical Medical Faculty Mannheim and 2Experimental Surgery, Faculty of Medicine, University of Heidelberg, Heidelberg, and 3Department of Pediatric Surgery, University of Leipzig, Leipzig, Germany; and 4Department of Anesthesiology, University of Illinois, Chicago, Illinois
Submitted 2 January 2007 ; accepted in final form 4 April 2007
| ABSTRACT |
|---|
|
|
|---|
pulmonary endothelium; monoclonal antibodies
Reactive oxygen species (ROS) augment I/R injury of the lung by initiating a series of pathophysiological events. ROS are highly active oxygen compounds, causing injury to all cellular molecules. ROS do play a central role in cellular signal transduction, being inactivated by antioxidative enzymes like superoxide dismutase and catalase. Ischemia and postischemic reperfusion lead to an increased generation of free radicals as well as increased consumption and loss of antioxidative capacity. Several attempts to limit I/R injury have been employed, e.g., elimination of neutrophil granulocytes (an important source of ROS generation) and/or maximization of antioxidative capacity under supply of antioxidative enzymes, vitamins, and others (23, 24, 46). The results, however, have been inconsistent (12, 25, 26, 47). This might be because of insufficient uptake into the endothelial cells (10, 25). Therefore, the objective of our current investigation was to deliver antioxidant enzymes to the pulmonary endothelium by immunotargeting of lung-specific endothelial antigens (34).
The preferential expression of angiotensin-converting enzyme (ACE) in pulmonary capillaries makes it an ideal target for therapy directed toward the pulmonary endothelium (7, 13, 15). ACE is a membrane-anchored glycoprotein (carboxy dipeptidase) that is expressed at the luminal surface of endothelial cells. Monoclonal antibody (MAb)-recognizing human and rat ACE (designated 9B9) accumulates within the pulmonary endothelium after internalization. We postulated that anti-ACE MAb 9B9 could serve as a specific carrier for the delivery of therapeutic substances to the lung vasculature (3, 13, 15, 35).
The current study provides in vivo evidence of the protective effect of preischemic vascular immunotargeting by anti-ACE MAb 9B9 conjugated with CAT (9B9-CAT) against the alteration of pulmonary endothelium by ROS.
In the present research, we studied the effects of endothelial immunotargeting by anti-ACE MAb 9B9 conjugated with CAT (9B9-CAT) in a rat model of lung I/R in vivo. In test animals, right lungs were clamped for 60 min and then reperfused for 120 min. We demonstrated that catalase conjugated with anti-ACE MAb 9B9 effectively protects lung endothelium from I/R injury in vivo and provides a promising approach for protection of grafts for lung transplantation.
| METHODS |
|---|
|
|
|---|
Male Wistar rats (300 g) were used in all experiments, receiving human care in compliance with the European Convention on Animal Care (GV-SOLAS guidelines). All protocols on animal procedures have been critically evaluated and approved by a national ethics committee in accordance with the Institutional Animal Care and Use Committee. Animals were randomized in four groups to investigate the extent of reperfusion injury after 60 min of ischemia at 37°C. After induction of anesthesia and tracheotomy, pressure-controlled ventilation was performed at <15 cmH2O, 0.5 FIO2, 2 cmH2O positive end-expiratory pressure (PEEP) at 6080 cycles/min to receive physiological PCO2 values.
The thorax was opened from the anterior through the sixth intercostal space by a horizontal thoracotomy. All animals received 100 IU of heparin intravenously 15 min before ischemia to avoid thromboembolic events. After careful preparation, the hilus of the right lung, including upper, middle, and lower lobe, was laced by a silicon tube (outer diameter 0.96 mm; NeoLab, Heidelberg, Germany) to be used as a tourniquet. Organ temperature was set at 37 ± 0.25°C and monitored by a temperature probe situated between the lower right and middle lobe. The thorax was covered with a heating blanket coupled with the heating system below the animal to avoid temperature loss and assure stable ischemic temperature. Loss of temperature at the opened thorax can be significant, and hypothermia decreases enzyme activities 1.5- to 2.0-fold every 10°C, slowing metabolic function by 10% or more (17). Therefore, stable and comparable temperatures in models of warm ischemia must be ensured. No statistically significant differences in right lung temperature during ischemia have been observed between the groups (Fig. 1).
|
Furthermore, six animals of the group CAT received 150 µg of biotinylated catalase before induction of ischemia (n = 6). Another six animals received 300 µg of anti-ACE MAb 9B9 conjugated with the same amount of catalase before induction of ischemia (9B9-CAT, n = 6). MAb 9B9 and catalase were biotinylated using NHS biotin (Calbiochem) at a biotin-protein ratio of 10:1. Then, conjugate of biotinylated catalase and biotinylated MAb 9B9 was prepared by mixing with streptavidin exactly as previously described (33).
Rats were killed after 2 h of reperfusion. Thoracic organs were removed en block, and parts of the right lower lung lobe were shock-frozen in liquid nitrogen.
Analytical Parameters
Hemodynamics and oxygenation. Systemic pressure and blood gases were monitored before ischemia, at the start and end of ischemia, at the start of reperfusion, and at 30, 60, and 120 min of reperfusion.
Lung dry-wet ratio (expressed as a percentage of dry weight to wet weight) was determined at the end of each experiment by drying (at 100°C for 48 h) and weighing the right middle lobe.
Oxygenation was determined in arterial blood probes with a blood gas analyzer (ABL 50; Radiometer, Copenhagen, Denmark) before and at the end of ischemia and at 0, 30, 60, and 120 min of reperfusion.
Immunohistochemical Detection of Pulmonary Targeting of 9B9-CAT Conjugates
Histological evaluation was performed blinded for all parts of the right lower lung lobe by standardized hematoxylin and eosin and periodic acid-Schiff staining. The tissues were snap-frozen and stored in liquid nitrogen. Tissues were sectioned at 5 µm by a cryostat microtome (CM 1850, Leica). The slides were air-dried at room temperature for 1224 h and stored at 30°C.
To detect administered conjugates, in situ immunohistochemistry was performed using antibodies directed against mouse immunoglobulins (IgG). The mouse alkaline-phosphatase anti-alkaline phosphatase (APAAP) technique was applied according to the manufacturer's protocol (Dako Cytomation). The secondary rabbit anti-mouse IgG antibody (1:40 dilution, Dako Cytomation) was supplemented with reconstituted lyophilized rat serum (1:600 dilution, Dianova) to abolish any nonspecific binding. Alkaline phosphatase substrate reaction with new fuchsin (100 µg/ml) and levamisole (400 µg/ml) was performed for 20 min at room temperature. Sections were counterstained with hematoxylin and mounted in gelatin.
Quantification of mRNAs for Inducible Nitric Oxide Synthase and Hypoxia-Inducible Factor-1
in the Lung
Total RNA was extracted using the RNeasy mini kit (Qiagen, Hilden, Germany) with brief homogenization in a Micra D8 homogenizer (Art-moderne Labortechnik, Müllheim-Hügelheim, Germany) and usage of the QIAshredder (Qiagen). cDNA synthesis was performed using oligo(dT) primers and SuperScript II (Invitrogen, Karlsruhe, Germany) with 4 µg of RNA according to the manufacturer's instructions.
PCR for GAPDH, inducible nitric oxide synthase (iNOS), and hypoxia-inducible factor-1
(HIF-1
) were performed as follows. The amounts of PCR products were determined by ChemiDoc and Quantity One (Bio-Rad, Munich, Germany). The results were arbitrarily normalized to the signals of GAPDH cDNA and presented a ratio of iNOS on GAPDH values in each probe designated as arbitrary units (a.u.).
Specific oligonucleotide primers were synthesized for rat genes (MWG Biotech). To perform PCR, 5 µl of cDNA were added to the master mix of 50 µl, containing 35 µl of H2O, 1.5 µl of 50 mM MgCl2, 5 µl of PCR buffer, 1 µl of primer 1, 1 µl of primer 2, 1 µl of dNTP, and 0.5 µl of Taq polymerase.
For iNOS, the sequences of the 3'-5' and 5'-3' primer were CAG CTG GGC TGT ACA AAC CTT and CAT TGG AAG TGA AGC GTT TCG. Reverse transcription was performed at 94°C for 4 min, 72°C for 12 min, 58°C for 45 s, and 53°C for 45 s.
For HIF-1
sequences of the 3'- and 5'-primers, TCA AGT CAG CAA CGT GGA AGG and TAT CGA GGC TGT GTC GAC TG were used. Reverse transcription was performed at 94°C for 3.5 min, 55°C for 45 s, and 72°C for 11 min (28).
Quantification of Endothelin-1 and Cytokine-Induced Neutrophil Chemoattractant Protein in Serum
At the end of each experiment, 1 ml of arterial blood was collected and centrifuged at 1,500 g for 5 min to sediment cellular material. Supernatants were stored deep-frozen at 80°C for determination. Serum levels of cytokine-induced neutrophil chemoattractant-1 (CINC-1; Ref. 44) and endothelin-1 were determined by commercially available enzyme immunoabsorbent assays (CINC-1, R&D Systems; endothelin-1, Biomedica). Quotients on protein levels were formed.
Statistical Analysis
For group comparisons, the measured results were reported as means ± SD. Statistical differences between data sets during ischemia and reperfusion were calculated after ANOVA multivariant analysis by the double-sided post hoc test of Dunnet using SPSS scientific software (version 13.2). The null hypothesis was rejected when P
0.05. Figures 4 and 5 are displayed as box-whisker plots, including median, 5th, 25th, and 95th percentiles.
|
|
| RESULTS |
|---|
|
|
|---|
Injection of biotinylated CAT and 9B9-CAT into animals of groups 3 and 4 before ischemia resulted in an acute drop of blood pressure to 5070 mmHg in five of six animals. In both groups, animals recovered within 35 min without any intervention. Besides this exception, systemic pressures were stable throughout the experiments and did not differ significantly between all groups (Table 1).
|
|
In the normal rat lung, the leading cell type to express ACE are endothelial cells (15, 31). However, to a much lower extent, ACE is also detected in fibroblasts and in activated alveolar macrophages. The endothelium in almost the entire vasculature of the rat lung, including pulmonary arteries, pulmonary veins, and microcirculation, express ACE. In contrast, the bronchial vasculature showed a systemic-like expression pattern. Here, a strong expression of ACE occurs in endothelial cells of large arteries and veins, whereas within the microcirculation, only arterioles and
10% of the capillaries contain ACE (Danilov SM, Franke VE, unpublished data).
In the 9B9-CAT perfused rat lung, we detected immunohistochemically a homogenous staining of 9B9-CAT conjugates at all known expression sites of endothelial ACE throughout the pulmonary and bronchial circulation including arteries, arterioles, capillaries, and veins in all animals of group 4 (Fig. 2). As expected, alveolar macrophages and fibroblasts did not show any staining (because their ACE is not accessible for MAbs from circulation). This indicates that anti-ACE antibodies and their conjugates are suitable for lung endothelial targeting in vivo.
|
Until the end of reperfusion of 9B9-CAT-treated I/R animals, oxygenation in this group (221 ± 36 mmHg) was the same as in control, sham animals (215 ± 16 mmHg). Moreover, the oxygenation in this group was significantly higher compared with untreated I/R animals (110 ± 10 mmHg) and CAT-only-treated I/R animals (114 ± 30 mmHg; Fig. 3; P < 0.01).
|
RT-PCR expression analysis demonstrated that the lung level of iNOS mRNA was significantly higher in animals undergoing I/R (0.33 ± 0.08 a.u.) and I/R animals treated with CAT only (CAT, 0.26 ± 0.05 a.u.) compared with poorly positive expression of iNOS mRNA in the sham-operated controls (0.15 ± 0.06 a.u.) and in I/R rats treated with 9B9-CAT (0.14 ± 0.04 a.u.; Fig. 4A; P < 0.01).
The analysis of the HIF-1
mRNA expression in the lung showed inverse results: a higher expression level in sham (0.23 ± 0.05 a.u.) and I/R animals treated with 9B9-CAT (0.24 ± 0.07 a.u.) compared with I/R animals (0.11 ± 0.04 a.u.) and I/R animals treated with CAT only (0.11 ± 0.4 a.u.; Fig. 4B; P < 0.01).
Serum CINC-1 levels in the I/R animals (286 ± 115 pg/mg) have been significantly elevated compared with sham (132 ± 72 pg/mg) or I/R animals treated with 9B9-CAT (150 ± 45 pg/mg; P < 0.01). I/R animals treated with CAT (232 ± 96 pg/mg) did not differ significantly in CINC-1 serum level from all other groups (Fig. 5A).
Endothelin-1 serum level was significantly elevated in I/R animals (42 ± 12 fmol/mg) and I/R animals treated with CAT (36 ± 11 fmol/mg) compared with sham-operated controls (18 ± 9 fmol/mg) or I/R animals treated with 9B9-CAT (26 ± 9 fmol/mg; P < 0.01; Fig. 5B).
| DISCUSSION |
|---|
|
|
|---|
Oxidants have an important role in the initiation, augmentation, and potentiation of endothelial injury in inflammation, gas exposure, hypoxia, I/R, and other pathologies (1, 29, 40, 45). There have been attempts to achieve protection of oxidative injury in various organs by the administration of antioxidant enzymes or their derivates, pending adequate delivery and timing of therapeutic intervention (2, 12, 23, 26, 47). Only Cremer et al. (12) found a limitation of pulmonary reperfusion injury by giving the oxygen free radical scavengers in combination of catalase and superoxide dismutase during perfusion and continuously during the first 20 min in a canine model of heterotopic heart-lung transplantation. The main inability of exogenous catalase administration to provide beneficial physiological effects lies in the specific effects of endogenous lung catalase activity (25). Catalase is localized in the intracellular compartment. This is the site where most of the free radicals occur. Muzykantov et al. (34) have successively shown that even biotinylated catalase has a poor uptake into the pulmonary endothelium.
Vascular immunotargeting of anti-oxidant enzymes to endothelial antigens [e.g., platelet endothelial cell adhesion molecule (PECAM), ICAM-1, ACE] has been proposed to improve the therapeutic effect of these enzymes (2, 10, 11, 34). Recently, attenuation of in vivo lung injury has been achieved by anti-PECAM conjugates with catalase in a murine model of vascular oxidative stress (10, 11). The same group effectively demonstrated a strong protective effect of vascular immunotargeting of catalase in an in vivo model of lung transplantation in rats. Intravenously injected conjugates of PECAM and catalase accumulated in the pulmonary vasculature and retained their activity during prolonged cold storage and transplantation. Immunotargeting of catalase to donor rats augmented the antioxidant capacity of the pulmonary endothelium, reduced oxidative stress, ameliorated I/R injury, and even prolonged the acceptable cold ischemia period of lung grafts (27).
Importantly, PECAM is significantly expressed in other organs besides the lung, whereas ACE is expressed in 100% of the pulmonary capillaries compared with 1020% in other organs (7, 13). Therefore, we reasoned that conjugates of catalase with anti-ACE MAbs should offer even greater specific pulmonary endothelial immunotargeting than that seen with anti-PECAM MAb. We (15) have previously compared five MAbs to the different endothelial antigens as possible vehicles for lung endothelial targeting under identical conditions and convincingly demonstrated the advantages of anti-ACE MAb 9B9 over other anti-endothelial MAbs including PECAM and ICAM. It is important to recognize that these findings do not exclude the use of anti-PECAM or anti-ICAM MAbs for lung endothelial targeting. In fact, under certain conditions, delivery of the therapeutic enzymes into the lung capillaries using anti-PECAM and anti-ICAM MAbs was quite specific and effective (10, 27, 32).
In this study, we used anti-ACE MAb 9B9 conjugated with catalase (9B9-CAT), which has been shown to accumulate selectively in the rat lung endothelium (2, 34). It has also been demonstrated that the pulmonary uptake of CAT conjugated with MAb 9B9 is more than 100-fold stronger than that of biotinylated catalase (34). Nevertheless, it is still unclear whether antibody conjugates 9B9-CAT work in vivo without major side effects. We (16) discovered previously that 9B9-based targeting induced some ACE shedding from the endothelial surface; this might theoretically compete with membrane-bound ACE for the injected ligand. However, the amount of shed ACE from the surface of ACE-expressing endothelial cells induced by MAb 9B9 is relatively low compared with the high amount of membrane-bound ACE. Also, only two MAbs (from 8 anti-human ACE MAbs tested) induced ACE shedding (6). Most of the MAbs to rat ACE (4) and mouse ACE (5) did not induce ACE shedding but effectively localized in the lung endothelium after systemic injection.
Our goal is not to protect rat lungs but rather human lung transplants. The first candidate for that goal is MAb i2H5, which recognizes human and primate ACE (7) and does not significantly induce ACE shedding (6). We used MAb 9B9 in a rat model because it is a convenient and rational system to prove this principle. Therefore, despite some effect that MAb 9B9 may have on the ACE shedding, we believe the results of this study convincingly demonstrate the effective protection of lung endothelium from oxidative injury using conjugate of catalase with anti-ACE MAbs.
In this study, the model of in vivo warm lung ischemia and reperfusion was chosen to evaluate therapeutic possibilities of vascular immunotargeting via ACE to the pulmonary endothelium in general. Warm ischemia and reperfusion as in our model is not an exact simulation of the situation in lung transplantation. However, generation of ROS has been shown by simulated ischemia in flow-adapted endothelial cells (45). The cessation of flow within the pulmonary endothelium alters mechanotransduction leading to ischemia-mediated cell signaling (19, 21). Our model offers the opportunity to study effects and side effects in vivo in one organism. Furthermore, we propose that this approach might be relevant for other possible applications of anti-ACE conjugates in pulmonary diseases. The limiting effect of PECAM conjugates with catalase on I/R injury after rat lung transplantation described by Kozower et al. (27) is therefore not comparable to the situation of warm ischemia and reperfusion in our model.
After injection into the jugular vein, substances accumulate in the lung immediately. MAb 9B9-CAT has been shown to provide 18-fold increase in pulmonary uptake of enzyme compared with pulmonary uptake of native catalase (34). After binding, internalization of antibodies takes place with trafficking of catalase into the pulmonary endothelium (32). Biotinylation of catalase has been shown to elevate and prolong serum levels but does not increase pulmonary uptake. Also, conjugation of catalase with mouse IgG did not increase pulmonary uptake of catalase (34).
In our model, 15 min after injection, ischemia was induced, and we note that antibodies were detected immunohistochemically after reperfusion with strong staining of lung capillaries in all experimental animals treated with 9B9-CAT. The protective effect of the antibody-conjugated catalase is specific because unconjugated components of the conjugates as well as IgG-CAT have been shown to provide no protection by our group (2).
The initial drop of blood pressure after instillation of biotinylated catalase and conjugated antibodies has not been described before. We speculate that this might be because of an initial effect in the pulmonary endothelium or, much less likely, because of an allergic reaction to the bovine catalase. As it proved to be reversible (within 24 min) and self-limited, we did not interpret this to be a significant and ongoing hemodynamic effect of the 9B9-CAT. Arterial blood gas oxygenation dropped significantly during reperfusion in untreated animals and animals treated with biotinylated catalase. Animals treated with anti-ACE antibodies conjugated with catalase, however, did not differ from sham-operated controls. The model enabled us to determine oxygenation capacity only of the right lung injured by ischemia as the left lung was clamped during reperfusion. A shunting to the contralateral uninjured lung therefore can be excluded. At an FIO2 of 0.5, the initial PaO2 values in animals of all groups would be expected to be nearer 300 mmHg. It seems reasonable that operative manipulation and open chest surgery might have lowered the initial PaO2 values to those noted in our study.
Oxygenation results correlate with lung edema generation determined by drying and weighing the right middle lobe at the end of each experiment. The determination of dry-to-wet weight ratio, however, does not distinguish between extra- and intracellular edema formation and is strongly dependent on fluid balance and hemoglobin content. To minimize this potential variation, all animals received the same amount of venous infusion, and hemoglobin content was comparable between the investigated groups at the end of the experiments (Table 2).
Endothelin-1 is a potent vasoconstrictor, which is expressed predominantly in the pulmonary endothelium, accumulating in the first hours of reperfusion (17). The basal endothelin-1 production has been reported to be upregulated by oxidative stress in pulmonary endothelial cells in vitro (30). Clinical and experimental studies have shown that endothelin-1 accumulates in bronchoalveolar lavage and serum during the first hours of reperfusion (42, 43). Endothelin-1 plasma concentration has been reported to be significantly increased after reperfusion in a rat model of warm lung ischemia and reperfusion (37). The administration of an endothelin-1 antagonist was associated with an improvement in lung function, a lower expression of iNOS, and a lower proportion of apoptotic cells after lung reperfusion (41). In lungs, iNOS mainly occurs in the pulmonary epithelium and macrophages (17). Ovechkin et al. (38) showed a 1.5-fold increase of iNOS mRNA expression in a rabbit model of warm lung I/R (2 h of the right pulmonary artery occlusion followed by 2 h of reperfusion compared with sham-operated controls), similar to our model. By treating one group of animals with an iNOS inhibitor before ischemia, they successfully prevented platelet adhesion and arteriolar wall interactions. It was also shown that iNOS activity was highly correlated with ROS and P-selectin expression (38). The upregulation of iNOS has also been shown to parallel the onset and progression of posttransplant obliterative airway disease (39). The significantly lower expression profile of iNOS combined with the significantly lower levels of endothelin-1 in I/R animals treated with 9B9-CAT compared with untreated I/R animals or animals treated with biotinylated CAT only therefore underlines the significantly decreased vascular damage in the lung.
CINC (IL-8) is increased during lung reperfusion, where it originates from macrophage epithelial cells and fibroblasts by neutrophil chemotaxis and activation. IL-8 has been described as significantly increased during reperfusion in clinical lung transplantation and experimental models of lung I/R. CINC has been shown to peak after 2 h of reperfusion in a rat model of warm lung ischemia (18). It further correlates negatively to PaO2-FIO2 ratio and mean airway pressures. The potential importance of IL-8 has also been demonstrated in acute respiratory distress syndrome (ARDS) and liver transplantation. Increased IL-8 (CINC) levels have also been described in association with a rising risk of death from primary graft failure after transplantation (17). The significantly lower levels of CINC in I/R animals treated with 9B9-CAT might demonstrate not only diminished endothelial damage, but also diminished damage of epithelial and fibroblast cells during reperfusion after endothelial targeting with anti-ACE MAbs.
It has been postulated that HIF-1
is regulated by a hypoxic pathway and a nonhypoxic pathway in a ROS-sensitive mechanism (22). During lung I/R, the pulmonary endothelium gets injured and ACE is shed, increasing angiotensin II. Angiotensin II has been shown to induce HIF-1 expression and the hypertrophy of pulmonary arterial smooth muscle cells. This might have multiple roles in the pathogenesis of pulmonary vascular remodeling in I/R (40). Nevertheless, HIF-1
, to our knowledge, has not yet been determined in models of lung I/R.
Haddad (22) describes HIF-1
being involved in the formation and expression of iNOS and VEGF. VEGF protein and mRNA is increasing during ischemia of the lung independent of oxygen tension. HIF-1
protein increased only during hypoxic ischemia, whereas HIF-1
mRNA increased during hyperoxic and hypoxic ischemia (8).
Therefore, we postulated that HIF-1
mRNA expression could be a sensitive marker for oxidative injury of the lung during reperfusion. The achieved results, however, showed a significantly higher amount of HIF-1
mRNA on GAPDH expression in sham-operated animals and animals treated with anti-ACE MAbs (Fig. 5). This is unlike all other published results.
One could speculate that certain molecular interactions might potentiate the decrease of HIF-1
mRNA expression in I/R and CAT animals. For example, Madjdpour et al. (28) showed that HIF-1
was downregulated in lung inflammation after alveolar macrophage elimination by liposome-encapsulated clodronate in a hypoxia model of the rat. Therefore, a diminished activation of alveolar macrophages might be the reason for the decreased HIF-1
mRNA expression in I/R and CAT animals. Perhaps the protection from oxidative stress by anti-ACE-9B9 conjugates with catalase (9B9-CAT) leads to a higher function and vitality of alveolar macrophages and therefore to a higher amount of HIF-1
mRNA expression.
During I/R injury of the lung, a series of pathophysiological events lead to pulmonary edema, cytokine activation, and possibly an impairment of long-term function of grafts in clinical lung transplantation. During preservation, cooling, and finally reperfusion, a vast flush of oxidants stress the pulmonary tissue. In our model of warm lung ischemia and reperfusion, no preservation and cold storage of the lung was performed. The beneficial effect of the conjugate of the catalase with anti-ACE MAb has been demonstrated over a period of more than 3 h after induction of lung injury in an in vivo model for the first time.
Further studies in the range of cold ischemia and warm reperfusion are indicated to determine the clinical significance of our findings as well as whether these antibodies could possibly be a therapeutic strategy for human pulmonary diseases.
| GRANTS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
. Crit Care 7: 4754, 2003.[CrossRef][Web of Science][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |