| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,,1
* Department of Histology and Embryology, Peking University Health Science Center, Beijing, China;
Department of Pathology, School of Basic Medical Science, Peking University Health Science Center, Beijing, China;
Keck School of Medicine of University of Southern California, Los Angeles, California, USA; and
Department of Pathology, State University of New York-Heath Science Center at Brooklyn, Brooklyn, New York, USA
1Correspondence: Department of Pathology Dean, School of Basic Medical Sciences Director, Peking University Infectious Disease Center, 38 Xueyan Rd., Beijing 100083, China. E-mail: jianggu{at}bjmu.edu.cn
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: immunodeficiency • T lymphocytes • B lymphocytes • SARS
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. With a fatality rate ranging from 7% to 17% (1)
, this emergent disease was named severe acute respiratory syndrome (SARS) by the World Health Organization (WHO) (2
3
4)
. In 2004, sporadic SARS cases reappeared in laboratory workers in Asia, and it is believed that a resurgence of SARS continues to be a threat (5)
.
It has been reported that infection by the SARS virus is not confined to the lungs, but involves other parts of the respiratory tract and other organ systems (6)
. Our group has studied the respiratory system, intestinal tissues, circulating blood cells, reproductive system, and thyroid of SARS patients. These results have been reported elsewhere (7
8
9
10
11)
. In this study we focus on changes in the spleen that have not been described before in a systematic manner. Although the hallmark symptoms of SARS result from infection of the respiratory system, an immune-based mechanism of pathogenesis has been proposed (6
, 12)
in which damage to the spleen and other lymphoid tissue is thought to be a major event in the pathogenesis of the disease. This hypothesis provides a means of understanding the deterioration of clinical status and collapse of the immune system, which may eventuate in an infection which initially is mild. In this study, we demonstrated that SARS virus had infected all of these SARS patients’ spleens by in situ hybridization (ISH), and we evaluated the numbers of different immunocytes using seven monoclonal antibodies with immunohistochemistry (IHC). Semiquantitative analyses were performed by light microscopy to determine the extent of damage to different types of immunocytes in SARS spleens.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. These patients died between May 9 and July 27, 2003. The clinical diagnostic criteria included high fever (temperature>38°C), cough or shortness of breath, new pulmonary infiltrates on chest radiograph, and a positive serological test for the SARS-CoV antibody (Ab) or a positive RT-polymerase chain reaction (RT-PCR) test in sera, urine, pharyngeal aspirates, or sputum. The diagnoses were further confirmed by detecting the SARS-CoV sequence in postmortem tissue samples of the lungs using ISH. All patients except one (S03) were treated with steroids. Clinical and laboratory information for the confirmed SARS patients is presented in Table 1
.
|
Control patients
Six age-matched control spleens were obtained from patients who died of trauma or accident with no known respiratory or immune disorders. We confirmed that all six patients were free of SARS infection with ISH.
Other controls
Controls were used to exclude the possibility that fixation conditions might have affected immunocytes in the tissue samples. We chose one misdiagnosed SARS patient (S10) as a fixation control. The misdiagnosed SARS patient’s tissue samples were fixed in the same manner as that of the confirmed SARS patients. Information for patient S10 is presented in Table 1
. Tissue sections of the heart and pancreas from the confirmed SARS cases were examined to evaluate the possibility of autolysis-induced reduction of immune cells due to delayed autopsy.
Among the six confirmed SARS patients, S03 was the only one who did not receive steroid treatment. A comparison of SARS patients with and without steroid therapy served as an indicator of the effect of steroid therapy on immunocytes in these tissue samples.
The Research Administration Committees of Peking Medical University Health Science Center and local hospitals approved ethical issues related to this study.
Investigation
In situ hybridization
The technique for in situ hybridization was based on that described by Zhang et al. (14)
. Briefly, a fragment of 145 bp "gcg caagtattaa 15361 gtgagatggt catgtgtggc ggctcactat atgttaaacc aggtggaaca tcatccggtg 15421 atgctacaac tgcttatgct aatagtgtct ttaacatttg tcaagctgtt acagccaatg 15481 taaatgcact tc" was amplified by polymerase chain reaction (PCR) from the SARS conronavirus genome sequence (GenBank, Accession No. AY274119). The RNA probe was prepared by in vitro transcription in the presence of digoxigenin-uridine triphosphate. The in situ hybridization reaction was performed on formalin-fixed paraffin embedded tissue sections with the SARS probe at a concentration of 50 µg/ml. The hybridization cocktail was incubated at 55°C for 16 h. Alkaline phosphatase-labeled antidigoxigenin Ab (1:500) was incubated for 60 min, colorized with NBT/BCIP, and counterstained with methyl-green. An unrelated probe of the same length was used as a control.
Histopathologic examination
Small pieces (1–2 cm3) of tissue were fixed in 10% buffered formalin and embedded in paraffin. Paraffin-embedded blocks of tissue were sectioned at 6 µm, deparaffinized, and stained with hematoxylin and eosin (H&E). The slides were examined under a light microscope (Nikon Eclipse E800, Japan) with various objectives (Nikon Pan Fluor, Japan).
Immunohistochemistry
IHC was performed as described by Lin et al. (15)
. Briefly, paraffin-embedded tissue sections were deparaffinized and immersed in 3% hydrogen peroxide at 20°C for 30 min to eliminate endogenous peroxidase activity. Antigen retrieval was performed by heating the tissue sections at 96°C in 0.01 M citrate buffer (pH 6.0) for 20 min and allowing them to cool to room temperature in 0.01 M PBS (PBS, pH 7.4) for 5 min. Primary antibodies to specific cell markers were applied for various durations at different temperatures. Information about the specific antibodies used in this study is presented in Table 2
. In addition, Ab to insulin (Zymed, South San Francisco, CA, USA) was applied to sections of pancreas to control for the possibility of autolysis, as the pancreas may undergo autolysis quickly following death. Optimal contrast between the specific labeling and the background for each antigen was achieved with the PV9000 immunohistochemistry kit (Zymed). To visualize specific signals, the 3, 3'-diaminobenzidine tetrahydrochloride (DAB) substrate-chromogen kit (Zymed) was used, with incubation for 1–6 min until the desired color intensity had developed. Some slides were lightly counterstained with hematoxylin and all slides were then dehydrated, coverslipped, and examined using a light microscope (Nikon AFX-IIA). Negative controls included omission of the primary Ab and the use of an unrelated Ab as the primary Ab.
|
Image analysis
Manual counts of five random fields with a 4 x objective lens were performed for each spleen. The number of splenic follicles and peri-arterial lymphatic sheaths were noted. We then computed the average number of these structures for each patient.
We used a computerized image analysis system (Motic Med CMIAS Pathological Image Analysis System, Xiamen, P.R.C.) to evaluate the number and size of the different cell types in each spleen. Three parameters were measured: 1) the number of positively stained cells in the areas analyzed (CD3+, CD4+, CD8+ lymphocytes, CD56+ NK cells, CD68+ macrophages, and S-100+ dendritic cells); 2) the percent area of CD20+ lymphocytes as measured by area of positive signal in the fields examined; this method was employed because positive cells were too numerous and crowded to be distinguished from one another; 3) the average cell size of CD68+ cells, which reflects the functional state of macrophages in the spleen. All analyses were performed on images captured at a resolution of 600 x 800 pixels, corresponding to a tissue area of 0.1452 mm2 viewed with a 40 x objective lens and a 2.5 ocular lens (Nikon). Fifteen such randomly selected fields were analyzed for each spleen.
The Student’s t tests (independent samples/2-tailed) were used to study the correlation among the various groups by comparing universal population means. The statistical analyses were performed with SAS Version 8.0 (SAS, Heidelberg, Germany). We considered P < 0.05 to be statistically significant. Data are presented as mean ± SD.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). In controls for ISH in spleens from confirmed SARS patients, in whom an unrelated probe of the same length was substituted for the specific RNA probe, no positive signal was detected (Fig. 1B
). Specific positive signal was not found in normal spleens (Fig. 1C
).
|
Histopathologic examination
When analyzing the sections with light microscopy, we found that hemorrhagic necrosis was present in the spleens of all confirmed SARS cases. The white pulp was severely atrophic and portions of the white pulp were obliterated. Cells in the red pulp had undergone massive necrosis, and the cell borders and nuclei had become more difficult to identify than those of the normal spleens (Fig. 1A, C
, Fig. 2
). No obvious autolysis was seen in the heart or pancreas of any of the six confirmed SARS patients. Intact islet cells were clearly visible in the pancreas of a confirmed SARS patient S11 whose spleen was among the most severely damaged of all confirmed SARS patients.
|
Changes in number and distribution of immunocytes in confirmed SARS spleens
In the immunohistochemical evaluation of the following cell types, tissue sections that were incubated with an unrelated Ab or without the primary Ab as a control gave negative results.
T lymphocytes
CD3+ and CD4+ T cells in the normal control spleens were concentrated in the periarterial lymphatic sheaths and in the marginal zones of the white pulp. These cells were scattered in the red pulp of normal spleens (Fig. 2, 2A, 3A
). CD8+ T cells were much less concentrated and were spread throughout the entire spleen (Fig. 2, 4A
), although some accumulated in the marginal zones. In confirmed SARS spleens, the CD3+, CD4+, and CD8+ cells were markedly reduced or completely absent. Lymphoid atrophy of the periarterial lymphatic sheaths and the marginal zones was widespread (Fig. 2, 2C, 3C, 4C
).
Image analysis of 15 visual fields yielded the average number of these three cell types in each spleen. Statistical analysis showed a reduction in CD3+, CD4+, and CD8+ T cell counts by 68.5 ± 25.7%, 77.9 ± 23.44%, and 82.9 ± 19.8%, respectively. Detailed data are shown in Table 3
.
|
The number of periarterial lymphatic sheaths was decreased on average by 90.4% in confirmed SARS spleens compared with controls (Table 4
).
|
B lymphocytes
CD20+ B cells in the normal spleens were concentrated in the splenic nodules and in the marginal zones. These cells were distributed only sparsely in the periarterial lymphatic sheaths and the red pulp (Fig. 2, 5A
). However, in confirmed SARS spleens, B cells were found mainly in degenerated splenic nodules and in the marginal zones, which were thinner than normal (Fig. 2, 5C
). The number of positive cells in the red pulp was affected to varying degrees. These cells were markedly decreased in two confirmed SARS cases (S01 and S03); they were absent in three confirmed SARS cases (S05, S11, and S15) and were maintained at a normal level in one confirmed SARS case (S08).
We calculated the area occupied by positively stained B cells instead of counting cells individually, because many of these cells were clustered together and could not be distinguished. The area occupied by B-lymphocytes in confirmed SARS specimens was decreased on average by 90.70% compared with that in the normal spleens (Table 3)
.
The average number of splenic follicles was reduced by 82.07% in SARS spleens compared with controls (Table 4)
.
Natural killer (NK) cells
In normal spleens, CD56+ NK cells were scattered mostly in the red pulp and marginal zones; only a small number were scattered in the white pulp (Fig. 2, 6A
). In confirmed SARS spleens, NK cells were distributed mainly in the red pulp because the white pulp was largely destroyed (Fig. 2, 6C
). The number of CD56+ cells was reduced by 47.4% compared with controls (Table 3)
.
Dendritic cells
In normal spleens, S-100+ dendritic cells were distributed in three distinct regions, the periarterial T cell zones, the B cell zones, and the marginal zones (Fig. 2, 7A
). In confirmed SARS spleens, S-100+ cells were decreased markedly (Fig. 2, 7C
). The number of S-100+ cells in the white pulp was reduced by 80.4% (Table 3)
.
Macrophages
In both normal and confirmed SARS spleens, CD68+ macrophages were distributed mainly in the red pulp, with a few scattered in the white pulp (Fig. 2, 8A, C
). The number of CD68+ cells in the red pulp was reduced by 39.48% in SARS spleens, but the average size of individual macrophages was increased by 2.21 times (Fig. 2, 9A, C
) over that of normal controls (Table 3)
.
Effects of delayed fixation and steroid treatment on immunocytes in the spleen
The various immune cells under investigation were counted in patient S10, who was autopsied 3 days after death. There was no significant difference in results between patient S10 and normal controls autopsied within 24 h of death (Fig. 2B
). Detailed data are presented in Table 5
.
|
There was no significant difference in cell counts between patient S03 and the other five confirmed SARS patients. Detailed data are presented in Table 6
.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Other studies investigating the spleen have provided only descriptive information and lacked numerical data (17)
. In this study, both the presence and localization of the virus in spleen target cells were confirmed by ISH. It is evident that infection by the SARS virus is not confined only to the lungs. We also performed quantitative analysis on the numbers of different immune cell types in the spleens of fatal SARS cases, and documented the extent of immune cell injury in the spleen. The significant reduction in multiple immune cell types in SARS patients’ spleens, in conjunction with the decrease in peripheral lymphocytes of the same group of patients previously described (6)
, reflects severe damage to immune function in these patients. Extensive destruction of splenic components of the immune system, together with destruction of the immune cells in other organs (6
7
8)
, supports the hypothesis that immune deficiency is a key mechanism in SARS pathogenesis (6
, 12)
and argues strongly that SARS should be categorized as a viral immune deficiency disease. Immune system injury may help to explain why half of SARS patients in this study (cases S05, S08, S11) developed an opportunistic infection in the late stages of their clinical courses.
CD3+, CD4+, CD8+, CD20+, and CD56+ cells in SARS spleens were all decreased sharply (by 47.4%–90.7%), indicating that the SARS coronavirus did not target one specific cell type but involved lymphocytes of several types. This finding suggests there may be a common mechanism by which these cells are damaged. Recently, Li et al. discovered in vitro that angiotensin (ANG) -converting enzyme 2 (angiotensin 1-converting enzyme-2) is a functional receptor for the SARS coronavirus (18)
. Angiotensin 1-converting enzyme (ACE) -2 is not found on lymphocytes (19)
, but this finding along with the suspicion that there may be a common mechanism of injury to lymphocytes warrants a search for another SARS receptor common to immunocytic cells.
Dendritic cells (DCs) are antigen-presenting cells, and their ability to present antigen is more powerful than that of B cells and macrophages (20)
. DCs are able to initiate a primary immune response through recruitment and activation of naive T cells (20
, 21)
, and they play a key role in antigen-specific immune responses and tolerance induction (22
, 23)
. The severe damage to dendritic cells as demonstrated by the sharp reduction of these cells in the white pulp suggests that the normal process of immune defense is disarmed in SARS patients. This also supports the notion that the mechanism of pathogenesis in SARS is more of an immune deficiency than an immune over-reaction.
Although the number of CD68+ macrophages in confirmed SARS spleens was decreased by 39.48% in this study, these cells were nevertheless among the most prominent cells in SARS spleens. This was due in part to a lesser reduction in numbers of macrophages compared with lymphocytes, but in addition the average cell size of macrophages was more than doubled. This increase in size indicated that these macrophages were in an activated state (24)
. Similar cells have been observed in the lungs and other organs of SARS patients (6)
. Apart from acting as scavengers, the exact role of those cells in the pathogenesis of SARS remains to be defined.
Over the course of this study the working hypothesis was that the decrease in numbers of splenic immunocytes was most likely due to direct attack on these cells by the SARS virus. However, there are other possible explanations for the decrease in immunocytes that were considered. These included artifact of delayed postmortem examination, side effect of therapeutic steroids, and loss of immunocytes secondary to atrophy or zonal splenic necrosis that has resulted from some unindentified indirect mechanism such as viral targeting of vascular structures.
Of these three possibilities, it is most difficult to exclude loss of splenic immunocytes by an unidentified indirect mechanism. However, two points suggest there may be a direct mechanism of viral injury in loss of lymphocytes. First, it is clear from previous ISH studies that virus can infect immunocytes in large numbers. For example, in a previous study by our research group 51.5% of lymphocytes and 29.7% of monocytes in the peripheral blood were found to be infected by the SARS virus (6)
. Second, a decrease in immunocytes in the blood, where tissue necrosis and atrophy are not potential confounding factors, has been recognized as characteristic of the disease SARS (4
, 12)
. A third less compelling point is that the histological appearance of the SARS spleen was not typical of zonal necrosis. Only one of the SARS spleens showed homogeneous necrosis wherein most cellular elements, including lymphocytes, were nonviable, as might be expected with necrosis resulting from a vascular insult. All other spleens showed both lymphocyte nodules in the white pulp and numbers of single lymphocytes in the red pulp, which appeared histologically viable over a variable background of splenic degeneration or necrosis. Thus, although an unidentified indirect mechanism cannot be ruled out, these three points taken together suggest that the likely explanation for decrease in splenic immunocytes is direct injury by virus.
Delayed postmortem examination, fixation, and tissue processing were also considered as a potential cause of artifactual decrease in immune cells. However, misdiagnosed patient S10, whose body was treated in the exact same manner as the confirmed SARS cases, displayed no significant difference in splenic immune cells compared with normal control cadavers. Second, it is recognized that the pancreas and the heart are prone to autolysis, and as such may serve as a marker for delay in postmortem examination. However, no histologically identifiable autolysis was observed in these organs of any of the patients in this study. Therefore, we conclude that the marked reduction of immunocytes in the spleens of SARS patients was most likely not caused by delayed postmortem examination.
Steroid treatment was a third suspected potential cause for the observed loss of lymphoid tissue, as it is known to induce T cell apoptosis and suppression of the immune system (25)
. However, our results suggested that the changes in the spleen were not caused by steroid treatment. In case S03, no steroids were administered, but evaluation disclosed no significant difference in immunocyte counts between this case and the other SARS patients (Table 6)
. Moreover, misdiagnosed patient S10 was also given large dosages of steroids, and there were no significant differences in splenic immune cells counts between this patient and the normal controls (Table 5)
.
Based on the latter findings, it is unlikely that steroid administration affected the number of immune cells in the spleen. In addition, we would like to cite findings from an earlier publication that also support this line of argument (6)
. In that study, lymphocyte counts from peripheral blood obtained at various time points from 30 patients not treated with steroids and 70 patients who were treated with steroids were analyzed retrospectively. In the untreated group of patients, CD3, CD4, and CD8 cell counts were decreased from the onset of the disease in confirmed SARS cases (15 patients) but were not decreased in any of the misdiagnosed cases (15 cases). In patients treated with glucocorticoids, lymphocytes were decreased in confirmed SARS cases (50 patients) and in misdiagnosed cases (20 patients). However, the difference in lymphocyte counts between the two treated groups remained the same as the difference in counts between the two groups not treated with steroids (6)
. Taken together, the above data argue that the marked destruction of immune cells in SARS spleens is not caused by steroid treatment but by the SARS viral infection itself.
Comparing the pathology of SARS with that of H5N1 influenza, we find many characteristics are common to both, including atrophy of the spleen (particularly the white pulp) (26)
, lymphopenia (27
, 28)
, animal to human species crossover (26
, 27)
, and hemophagocytosis in the lungs (26
, 29)
. H5N1 influenza virus, which was isolated in 1997, was found to be a hyper-inducer of proinflammatory cytokines (30)
. Similarly, an array of cytokines [interleukin-1ß, IL-6, and interleukin 12 (IL-12)] was found in SARS to have significantly increased for at least 2 wk after onset of the initial symptoms (31)
. In SARS the two main sources of proinflammatory cytokines, namely macrophages and T cells, were directly infected by the virus. The possibility exists that changes in proinflammatory cytokines play a role in the pathogenesis of this new disease.
Among the six confirmed SARS patients, case S08 was different from the others in that the numbers of B cells, T cells, and dendritic cells approached a normal level. This patient’s length of illness (21 days) was the shortest among all six patients. On the other hand, case S15 had the longest duration of disease (92 days) and displayed one of the most severe decreases in numbers of immunocytes. Therefore, the duration of illness appears to be a relevant factor in the severity of immune damage. In this study, fatal SARS cases with longer illnesses had more severe immune system injury. This is consistent with the study by Chan et al., who reported that CD3+, CD4+, CD8+, and CD56+ cell counts were good prognostic indicators for admission to the intensive care unit for SARS patients (32)
.
In conclusion, the splenic immunocytes of SARS patients were found to be severely damaged in this study. Multiple cell types were reduced in number, indicating that this coronavirus affects a wide range of immune cell types. The extent of damage to the immune cells appeared to be related to the length of the disease, with longer duration conferring more severe damage. The data suggest that the pathogenesis of SARS is more likely to be immune injury rather than an immune over-reaction. This study provides fundamental information about SARS splenic pathology that may promote an understanding of the relationship of pathogenesis and the clinical manifestation of lymphopenia commonly observed in SARS patients.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 20, 2006. Accepted for publication June 2, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Gu and C. Korteweg Pathology and Pathogenesis of Severe Acute Respiratory Syndrome Am. J. Pathol., April 1, 2007; 170(4): 1136 - 1147. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
