| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print August 26, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Liver Research Institute, University of Arizona, Tucson
Key Words: endotoxemia • neutrophils
Dear Editor:
We read with great interest the recent paper by Li et al. [1 ], who reported evidence for a critical role of the CXC chemokines keratinocyte (KC) and macrophage-inflammatory protein-2 (MIP-2) in the pathophysiology of endotoxin (ET)-induced liver injury. The authors showed increased KC and MIP-2 formation during endotoxemia and found that neutralizing antibodies against these chemokines strongly attenuated parenchymal cell apoptosis, prevented neutrophil extravasation, and reduced liver injury after treatment of mice with galactosamine (Gal)/ET [1 ]. As neutrophil extravasation is a prerequisite of injury in this [2 ] and other inflammatory liver disease models [3 ], the data of Li et al. [1 ] appear to identify the missing link and signal, which trigger neutrophil extravasation and the attack on parenchymal cells [4 ]. Moreover, the authors showed a 68% reduction of hepatocellular apoptosis in animals treated with antibodies against KC and MIP-2, suggesting an involvement of CXC chemokines in apoptosis signaling, which would be a novel, unrecognized function of CXC chemokines. However, a closer look at the data raises substantial doubts about the interpretations of these results and the validity of the conclusions drawn from these studies.
The first concern relates to the inhibition of apoptosis in this model. The assessment of apoptosis with Hoechst stain is clearly not accurate; e.g., 6% apoptosis in control livers [1
] is ten- to 100-fold above baseline values [5
]. Conversely, the reduced caspase-3 activity, a critical mediator of Gal/ET-induced apoptosis [6
], supports the authors’ conclusion that anti-CXC antibodies attenuated apoptosis. However, before postulating a novel discovery, the authors should have checked tumor necrosis factor 
(TNF-) formation. As apoptosis in this model is entirely TNF-dependent [7
], the most likely explanation for reduced apoptosis is reduced TNF formation. In fact, injection of high doses of immunoglobulin G (IgG) has been shown to bind ET, which can reduce TNF formation by Kupffer cells [8
]. The observation that all IgG treatment attenuated apoptosis and that the combination of two antibodies (twice the IgG dose) had an additive effect [1
] is consistent with this explanation.
The second concern is related to the conclusion that CXC chemokines are responsible for neutrophil transmigration. The authors showed no effect of the anti-CXC chemokine antibodies on leukocyte adherence in postsinusoidal venules, but they did not investigate sinusoidal adherence. As has been shown in this model, neutrophil transmigration occurs from sinusoids and not from venules [2
]. Therefore, assessment of sinusoidal neutrophil sequestration would have been necessary to justify the conclusion that chemokine neutralization did not affect hepatic neutrophil accumulation. In fact, it is unclear from the authors’ description if they just counted the total number of neutrophils in the tissue as extravasated. However, if assessed in more detail, only 35–40% of neutrophils are actually extravasated [2
, 6
, 9
] at the time of injury. As a result of the difficulties in accurately assessing neutrophil localization in damaged tissue, the authors should have counted neutrophils before extravasation to get a reliable number of total hepatic neutrophil sequestration. As it is, the data support the interpretation that there are fewer neutrophils in the liver of anti-CXC chemokine antibody-treated animals. As hepatic neutrophil recruitment in this model is entirely TNF-dependent [10
], these observations would be consistent with reduced TNF- formation. A further inconsistency with the data is the puzzling fact that anti-KC and anti-MIP-2 antibodies by themselves are maximally effective in preventing "extravasation" and injury. Both chemokines are generated in large quantities, are potent chemotactic factors, and can trigger extravasation [11
]. Therefore, eliminating only one mediator should have had only a partial effect, and the combination of both antibodies should have had an additive effect. As it is, the observations by the authors are puzzling.
In addition to these concerns, our own data do not support a role of CXC chemokines in the pathophysiology of Gal/ET-induced liver failure for a number of reasons. First, extravasation of neutrophils occurs at a time (6 h) when KC and MIP-2 formation is reduced by more than 90% compared with peak levels at 1.5 h after Gal/ET [12 ]. Second, treatment with ET alone causes formation of similar amounts of CXC chemokines as with Gal/ET, but there is no apoptosis, transmigration, or injury [12 ]. Third, antibodies against MIP-2 and KC did not reduce apoptosis, transmigration, or injury after Gal/ET in our hands [12 ]. Fourth, mice deficient in the receptor used by KC and MIP-2 (CXC chemokine receptor 2) show similar apoptosis, transmigration, and injury as wild-type animals [12 ]. Fifth, inhibiting apoptosis with a pancaspase inhibitor reduced transmigration and injury without affecting CXC chemokine formation [12 ].
Taken together, there are a number of serious concerns with the interpretation of the data by Li et al. [1 ], which leads us to question the conclusion that CXC chemokines are the signal for transmigration in the Gal/ET model. We feel that neither our own data nor the data presented by Li et al. [1 ] support this hypothesis.
Received May 27, 2004; accepted May 28, 2004.
REFERENCES
-induced hepatic parenchymal cell apoptosis and neutrophil-mediated necrosis in a murine endotoxin shock model J. Immunol. 160,3480-3486Related Articles
This article has been cited by other articles:
|
|
 |
|
  J. P. Luyendyk, L. D. Lehman-McKeeman, D. M. Nelson, V. M. Bhaskaran, T. P. Reilly, B. D. Car, G. H. Cantor, X. Deng, J. F. Maddox, P. E. Ganey, et al. Coagulation-Dependent Gene Expression and Liver Injury in Rats Given Lipopolysaccharide with Ranitidine but Not with Famotidine J. Pharmacol. Exp. Ther., May 1, 2006; 317(2): 635 - 643. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Shayakhmetov, Z.-Y. Li, S. Ni, and A. Lieber Interference with the IL-1-Signaling Pathway Improves the Toxicity Profile of Systemically Applied Adenovirus Vectors J. Immunol., June 1, 2005; 174(11): 7310 - 7319. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
