| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Correspondence |
Columbia, Missouri
To the Editor-in-Chief:
Being interested in the histopathology of cardiac rejection,1
I was pleased to read the constructive article by Miura and co-workers2
about the transplantation of hearts from A/J (H-2a) mice to major histocompatibility complex mismatched recipients divided further into two groups: wild-type (WT)C57BI (H-2b) and interferon (IFN)-
deficient C57BI/IFN--/-. The allografts were rejected in about 8.5 days in WT receipients by a severe cellular acute rejection mediated by CD8+ T cells and in about 6 days in IFN- deficient hosts by a rejection manifesting intense infiltration by neutrophils. The analysis of Figure 2 of Miura et al’s2
article, visualizing basic histopathology of the rejecting cardiac allografts about 1 day before the end of rejection, is the subject of this letter.
In substance, I agree with the authors as far as Figure 2, a and c, are concerned. There are still some well-preserved cardiomyocyte nuclei in Figure 2c. I mention this fact because absolute magnification numbers accompanying published microphotographs are often imprecise due to printing procedures. I will compare, therefore, the approximate diameter of cardiomyocyte nuclei (4 µm) in Figure 2c (x200) with one structure in Figure 2d (x200) to determine its dimensions.
I also agree with the authors regarding the neutrophil infiltration in Figure 2, b and d. However, I do not see "diffuse intense intragraft thrombosis" in Figure 2, b and d. It is quite possible that small microthromboses exist in the heart in question, but I have not noticed any identifiable thrombus in this figure. What factual information supports the "thrombosis" statement?
I have been struck most by a formation of myocardial defects which are well visible in Figure 2b. One of them is seen in detail in Figure 2d. These defects may be mistaken for vessels at first sight. However, they manifest the features which are incompatible with vascular origin: they do not possess vessel walls; they often fuse gradually with the surrounding myocardium (Figure 2d, upper right quarter); they contain fragments of cardiomyoctyte cytoplasm (Figure 2d, lower left center, at the "7 o’clock" position); and they comprise cells with nuclei similar to the nuclei of cardiomyocytes and surrounded by a narrow rim of cytoplasm with cardiomyocyte tinctorial properties (Figure 2d). Some of these nuclei are practically "naked."3
What pathological process has created such myocardial defects within five days after transplantation? Is it the necrosis mediated by neutrophils and mentioned by the authors?2 Myocardial necrosis is phagocytized by macrophages. This is a process lasting days, weeks, and months which is followed immediately by healing reaction and scarring. Consequently, necrosis would not have formed myocardial defects filled with interstitial fluid (Figure 2, b and d), and one must look for another explanation for their appearance.
To do so, let’s pay attention to the "severe disseminated hemorrhagic necrosis," which is the third important pathological process mentioned by the authors in Figure 2, b and d.2 In cardiac pathology, the term "hemorrhagic necrosis" is used to designate necrotic myocardium with blood extravasated into the interstitial space. It is often described in hemorrhagic infarcts, hyperacute rejection, and other pathological processes. Its concept suffers, however, from numerous shortcomings. For example, interstitial spaces between cardiomyocytes are extremely narrow (from 0.2 µm to a few µm) and blood pressure is not high enough to dislodge cardiomyocytes from their original position.4 It is difficult, therefore, to account for large accumulations in red cells in the narrow interstitial spaces. Most often, alleged extravasated blood contains only erythrocytes and lacks an adequate amount of fibrin. Furthermore, it is a process which is accompanied by an unaccounted loss of cardiomyocytes.5 To explain these contradictions, one current theory proposes that alleged red cells present in "hemorrhagic necrosis" are mostly cardiomyocyte apoptotic bodies similar to erythrocytes.3–6 Only later, when the interstitium is no longer supported by intact cardiomyocytes, may vessels become injured and give rise to genuine hemorrhage.6
If this new hypothesis is correct, where are the cardiomyocyte apoptotic bodies in Figure 2, b and d? Undoubtedly, some were already phagocytized by macrophages, many were transported away by lymphatic outflow,3 and most are still in the tissue being considered to be red cells by the authors.2 One may see them best in the myocardial defect in Figure 2d. There, suspended in interstitial fluid, they are constricted neither by capillaries nor by cardiomyocytes and their features may be studied without interference of these factors. Being of variable dimensions, they are practically always smaller than the approximate diameter of cardiomyocyte nuclei in Figure 2c (4 µm) while red cells have the diameter of approximately 7.2 µm. In the myocardium surrounding the defects, both individual apoptotic bodies and their conglomerates may be seen. It is difficult to reconcile large dimensions of the conglomerates with narrow interstitial spaces. In reality, the conglomerates enter into intimate contact with cardiomyocyte cytoplasm and are sometimes entirely surrounded by it. All these features indicate that the alleged red cells are cardiomyocyte apoptotic bodies. Consequently, the main mechanism of cardiomyocyte apoptotic bodies has been described in the humoral rejection of human cardiac allografts.1
Does the experimental system visualized in Figure 2, b and d, permit cardiomyocyte apoptosis to take place? The answer is yes. Cardiomyocytes possess death receptors (Fas, tumor necrosis factor receptor, etc)7
and neutrophils have Fas ligand, tumor necrosis factor-
, etc8
In certain conditions, cardiomyocyte receptors and neutrophil ligands enter into contact resulting in cardiomyocyte apoptosis. Why hasn’t this striking phenomenon been described sooner? Firstly, it may have been overlooked and secondly, it may take place only in special situations such as a deficiency of interferon- in transplantation recipients.
References
. Am J Pathol 2003, 162:509-519The Cleveland Clinic, Cleveland, Ohio
Authors’ Reply:
We appreciate the interest and many positive comments made by Dr. Beranek concerning our recent report investigating the rejection of MHC-mismatched cardiac allografts in the absence of IFN-.1
The goal of this study was to investigate mechanisms mediating the rapid rejection of organ allografts in the absence of IFN-. This rapid rejection has been observed in murine models of renal, heart, and more recently liver allografts where either the recipients are unable to produce IFN- or the grafts are from IFN-R1-/- donors and are therefore unable to respond to IFN- produced by graft-infiltrating T cells.2–6
A finding that is always observed during the rapid rejection of renal, cardiac, and liver allografts in these models is severe graft tissue necrosis and hemorrhage that accompanies the rejection. Rejection of the control allografts in the presence of IFN- is delayed in comparison and characterized by increasing mononuclear cell infiltration that eventually results in graft failure. Based on the necrosis observed in the absence of IFN-, we hypothesized that neutrophils might play a critical role in this histopathology.
Dr. Beranek has astutely pointed out fine details of the histopathology in the allografts from IFN--/- recipients depicted in Figure 2 of our report. Several comments are in order regarding his evaluation of the sample sections shown. First, we were of the opinion from the first viewing that the structure shown in the lower left-hand corner of Figure 2d is not a vessel for many of the reasons stated by Dr. Beranek. Second, we did observe many small thromboses in vessels throughout the graft although these were not shown in the figure. Third, Dr. Beranek makes a good point with regard to the small bodies in the figure panel that may or may not be erythrocytes or apoptotic bodies. We have not investigated the presence or temporal aspects of myocardial apoptosis in this model. However, graft tissue hemorrhage has been observed in other solid organ allografts retrieved from IFN- deficient recipients with similar histopathological features shown in Figure 2, b and d.2–6
It will be of some importance to distinguish these features in the rejection of these heart allografts but as Dr. Beranek points out this may represent a very specialized case of tissue pathology. A potential solution might be the use of tissue factor staining to distinguish erythrocytes from myocardial apoptotic bodies as pointed out by Dr. Beranek in a recent letter to The American Journal of Pathology.7
Fourth, Dr. Beranek has asked what pathological process generates the myocardial defects shown in these grafts after only 5 days. The data of the report are strongly supportive of a neutrophil mediated mechanism that occurs as rapidly as shown. In our view, this histopathology looks like neutrophil-mediated necrosis and may include neutrophil-mediated apoptosis of cardiomyocytes. It is important to state again that similar patterns of histopathology are observed in other organ allograft models in the absence of IFN-.
Finally, Dr. Beranek concludes with what we feel is the most important point of this report. That is that IFN- is an important regulator of early innate immune attack on the allograft. In the absence of a source of IFN- the allografts are intensely infiltrated with neutrophils and quickly exhibit the tissue necrosis depicted in Figure 2 of the report. With this in mind one should be asking what regulatory aspects of IFN- protect the allograft from this pathology. When we initiated these studies we had expected to observe unregulated expression of neutrophil chemoattractants (eg, KC/Groa and MIP-2) in the allografts retrieved from the IFN--deficient recipients. As shown in Figure 5 of the report this is not the case. These results raise an important and unanswered question regarding the IFN--dependent mechanism(s) that restrict the temporal infiltration of neutrophils into the allograft to mediate this extreme histopathology. This continues to be a focus of our studies to fully understand and minimize this attack.
References
. Am J Pathol 2003, 162:509-519
acts directly on rejecting renal allografts to prevent graft necrosis. Am J Pathol 2001, 158:215-226 alters the pathology of graft rejection: protection from early necrosis. J Immunol 2001, 166:7072-7081 interferon regulates acute vascular xenogeneic rejection. Nat Med 2000, 6:549-555[Medline]
is an absolute requirement for spontaneous acceptance of liver allografts. Am J Transplant 2003, 3:942-951[Medline]
: CD8+ effector cells develop independently of CD4+ cells and CD40-CD40 ligand interactions. J Immunol 2001, 166:3248-3255
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
