Published online before print May 3, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Dermatology, Kanazawa University Graduate School of Medical Science, Ishikawa, Japan; and
Department of Dermatology, The Jikei University School of Medicine, Tokyo, Japan
1Correspondence: Department of Dermatology, Kanazawa University Graduate School of Medical Science, 13-1 Takaramachi, Kanazawa, Ishikawa 920-8641, Japan. E-mail: s-sato{at}med.kanazawa-u.ac.jp
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and interleukin-6. The results of this study indicate that PSGL-1 contributes to the Arthus reaction mainly as a ligand of P-selectin and partly as a ligand of E- and/or L-selectin by regulating neutrophil and mast-cell recruitment and that PSGL-1 would be a therapeutic target for human IC-mediated diseases.
Key Words: adhesion molecules • inflammation • mast cells • skin
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The selectins share a highly conserved N-terminal lectin domain that can interact with sialylated and fucosylated oligosaccharides such as sialyl-Lewis X [7 , 8 ]. Although various candidates have been identified as potential ligands for selectins, P-selectin glycoprotein ligand-1 (PSGL-1) is the best-characterized ligand, which is recognized by all three selectins [9 , 10 ]. PSGL-1 is a mucin-like, disulfide-linked homodimer expressed by all subsets of leukocytes [11 ] and is a high-affinity ligand for P-selectin [12 ]. PSGL-1 has also been shown to bind to E- and L-selectins, but their affinities are lower than P-selectin [9 , 13 14 15 16 ]. PSGL-1-deficient mice (PSGL-1–/–) are viable and fertile but exhibit reduced leukocyte infiltration in thioglycollate-induced peritonitis [17 ]. Similarly, functional blocking monoclonal antibody (mAb) to PSGL-1 also inhibits neutrophil accumulation in thioglycollate-induced peritonitis [18 ]. PSGL-1–/– mice show reduced T helper 1 cell infiltration in an oxazolane-induced contact hypersensitivity model [19 ]. Furthermore, anti-PSGL-1 mAb reduces neointima formation in mice by inhibiting leukocyte migration [20 ]. These previous studies suggest that blocking leukocyte accumulation through inhibition of selectins and PSGL-1 interaction provides a mean to the therapy of the inflammatory diseases.
The formation and local deposition of immune complexes (IC) induce an acute inflammatory response with significant tissue injury. The classical, experimental model for IC-mediated tissue injury is the Arthus reaction [21 ]. Previous studies have revealed that accumulation of neutrophils and mast cells is necessary for the progression of the IC-mediated vascular tissue damage, which results in edema and hemorrhage [22 23 24 25 26 27 ]. We recently reported that mice lacking P-selectin, E-selectin, or both exhibit reduced Arthus reaction that is associated with decreased infiltration of neutrophils and mast cells [28 ]. The finding that PSGL-1 is a well-characterized ligand for selectins suggests that PSGL-1 is involved in the development of IC-induced tissue injury by regulating leukocyte recruitment [9 , 10 ]. However, direct evidence for the importance of PSGL-1 in leukocyte recruitment into inflammatory sites during this pathogenic process is still insufficient. In this study, to clarify a role of selectins and PSGL-1 in the Arthus reaction, we examined inflammation induced by IC in mice treated with mAb to PSGL-1 or P- and/or E-selectin. The results of this study indicate that PSGL-1 contributes to IC-mediated vasculitis mainly as a ligand of P-selectin and partly as a ligand of E- and/or L-selectin by regulating the infiltration of neutrophils and mast cells. These results also suggest that blocking function of PSGL-1 as well as selectins would be a potential therapy for human IC-mediated diseases.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cutaneous reverse-passive Arthus reactions
For cutaneous reverse-passive Arthus reactions, mice anesthetized by inhalation of diethyl ether were shaved on their dorsal skin and wiped with 70% alcohol. Rabbit immunoglobulin G (IgG) anti-chicken egg albumin antibodies (60 µg/30 µl; Cappel, Aurora, OH) were injected intradermally (i.d.) with a 29-gauge needle, followed immediately thereafter by intravenous (i.v.) injection of chicken egg albumin (20 µg/g; Sigma-Aldrich, St. Louis, MO) [23
]. The i.d. injection of purified polyclonal rabbit IgG (60 µg/30 µl; Sigma-Aldrich) followed by i.v. installation of chicken egg albumin served as a negative control. The solution of chicken egg albumin contained 1% Evans blue dye (Sigma-Aldrich). For a blocking study using mAb to PSGL-1 or E- and/or P-selectin, mAb were injected i.v. 30 min before IC challenge. Antibodies used in this blocking study included mAb to murine PSGL-1 (2PH1, rat IgG1, 30 µg per mouse; BD PharMingen, San Diego, CA) [29
], mAb to murine P-selectin (RB40.34, rat IgG1, 30 µg per mouse; BD PharMingen) [30
], and mAb to murine E-selectin (10E9.6, rat IgG2a, 30 µg per mouse; BD PharMingen) [30
]. Irrelevant, isotype-matched, purified rat IgG1 mAb (R3-34) and rat IgG2a mAb (R35-95) served as positive controls (30 µg per mouse; BD PharMingen).
Total leukocyte counts and differentials in all experimental groups of mice were determined 8 h after IC challenge. Blood was obtained through the retro-orbital venous plexus into EDTA-anticoagulated tubes. White blood cell counts were determined using an automatic cell counter following the lysis of red blood cells. Blood smears were prepared and stained with Wright’s stain for differential leukocyte counts.
Peritoneal reverse-passive Arthus reactions
The peritoneal reverse-passive Arthus reaction was initiated by the i.v. injection of chicken egg albumin at 20 µg/g, followed immediately by the intraperitoneal (i.p.) injection of 800 µg rabbit IgG anti-chicken egg albumin antibody or negative-control, purified rabbit polyclonal IgG in a volume of 400 µl [23
]. Four hours or 8 h later, the peritoneum was exposed by a middle abdominal incision, and 5 ml ice-cold phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA) was injected into the peritoneal cavity via a 27-gauge needle. Cells in the recovered lavage fluid were centrifuged onto glass slides and stained with Giemsa for light microscopic examination to quantify neutrophil and mast-cell numbers.
Quantification of edema and hemorrhage
Edema was evaluated by measuring the vascular leak 4 h after IC challenge [23
]. Mice were killed, and the skin containing the injection site was removed at the level of fascia above skeletal muscle. The diameter of extravascular Evans blue dye on the fascia side of the injection site was measured directly. Evans blue dye binds to serum proteins and thereby can be used to quantify alterations in vascular permeability. The diameter of the major and minor axes of the blue spot was averaged for analysis. The amount of hemorrhage was assessed 8 h after IC challenge by direct macroscopic measurement of the purpuric spot. The diameter of the major and minor axes of the purpuric spot was averaged for analysis.
Histological examination
Skin tissues were harvested 4 or 8 h after IC challenge using a disposable, sterile, 6-mm punch biopsy (Maruho, Osaka, Japan) and were assessed for tissue damage and numbers of infiltrating neutrophils and mast cells. Tissues were cut into halves, fixed in 3.5% paraformaldehyde, and then paraffin-embedded. Sections (6 µm) were stained using hematoxylin and eosin (H&E) for neutrophil evaluation and toluidine blue for mast-cell staining. Neutrophil and mast-cell infiltration was evaluated by counting extravascular neutrophils and mast cells in the entire section and averaging the numbers present in 10 serial skin sections from the injection site. Three investigators in a blinded manner independently examined each section.
Cytokine enzyme-linked immunosorbent assay (ELISA)
Levels of murine tumor necrosis factor (TNF-) and interleukin (IL)-6 in the peritoneal lavage were determined by ELISA using rat mAb pairs for murine cytokines (BD PharMingen), according to the manufacturer’s instructions. Briefly, the plates were coated with cytokine-specific antibodies and incubated with appropriately diluted peritoneal lavage samples. After incubation with biotinylated, cytokine-specific mAb and streptavidine-horseradish peroxidase (HRP), the reaction was developed.
Statistical analysis
The Mann-Whitney U test was used for determining the level of significance of differences in sample means, and Bonferroni’s test was used for multiple comparisons.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (19K): [in a new window] |
Figure 1. The effect of E- and/or P-selectin (E-sel and/or P-sel, respectively) blockade and PSGL-1 blockade on edema and hemorrhage during the cutaneous reverse-passive Arthus reaction. Mice were injected i.d. with rabbit IgG anti-chicken egg albumin antibody, followed by systemic chicken egg albumin and 1% Evans blue dye. Cutaneous inflammation induced by the Arthus reaction can be separated into two distinct responses: edema, which reaches a maximum at 3–4 h after IC challenge, and hemorrhage, which peaks in intensity at 8 h. To assess a role of PSGL-1 and E- and P-selectins in the cutaneous Arthus reaction, edema and hemorrhage were evaluated 4 and 8 h after IC challenge, respectively, in mice treated with mAb to PSGL-1 or anti-P- and/or anti-E-selectin mAb. After 4 or 8 h, dorsal skins were assessed from each mouse. Edema was evaluated as the diameter of extravasated Evans blue spot (A). Hemorrhage after 8 h was assessed as the diameter of the purpuric spot (B). Mice that received an i.v. injection of irrelevant, isotype-matched, purified rat IgG1 mAb and rat IgG2a mAb served as positive controls. Mice that received an i.v. injection of chicken egg albumin followed by an i.d. injection of polyclonal rabbit IgG served as negative controls. Edema and hemorrhage were significantly inhibited in mice treated with anti-P-selectin mAb, E- and P-selectin mAb, or PSGL-1 mAb compared with positive control mice for both panels (P<0.001) but not in mice treated with anti-E-selectin mAb. Horizontal bars indicate mean values for each group of mice. N.S., Not significant.
|
Leukocyte infiltration in the cutaneous Arthus reaction
Extravascular neutrophils and mast cells were assessed in skin tissue sections after 4 and 8 h of IC formation (Figs. 2
3
4
). Treatment with irrelevant, isotype-matched mAb instead of anti-P- or anti-E-selectin or anti-PSGL-1 mAb, served as controls, did not affect neutrophil and mast-cell recruitment compared with untreated mice. After 4 h of IC challenge, neutrophil numbers were significantly reduced in mice with P-selectin blockade (32% decrease, P<0.05), PSGL-1 blockade (38%, P<0.05), or blockade of E- and P-selectins (68%, P<0.001) compared with control mice, and they were not affected by E-selectin blockade alone. There were no significant differences between blockade of PSGL-1 and that of P-selectin. The combination of E- and P-selectin blockade exhibited a reduction in neutrophil accumulation that was significantly lower than that found by P-selectin blockade as well as PSGL-1 blockade (P<0.01). Similar results were obtained after 8 h of IC formation.
 View larger version (32K): [in a new window] |
Figure 2. Arthus reaction-induced recruitment of neutrophils and mast cells in the skin (A) and the peritoneum (B) from mice treated with mAb to E- and/or P-selectin or PSGL-1 and control mice at 4 and 8 h after IC challenge. In the cutaneous Arthus reaction, mice were injected i.d. with rabbit IgG anti-chicken egg albumin antibody, followed by systemic chicken egg albumin. Numbers of neutrophils and mast cells per skin section were determined by counting in H&E- and toluidine blue-stained skin sections, respectively. The peritoneal reverse-passive Arthus reaction was induced by the i.v. injection of chicken egg albumin, followed immediately by the i.p. injection of rabbit IgG anti-chicken egg albumin antibody. Four hours or 8 h later, the peritoneum was exposed by a middle abdominal incision, and 5 ml ice-cold PBS containing 0.1% BSA was injected into the peritoneal cavity via a 27-gauge needle. Cells in the recovered lavage fluid were then centrifuged onto glass slides and stained with Giemsa to quantify neutrophil and mast-cell numbers. Mice that received an i.v. injection of irrelevant, isotype-matched, purified rat IgG1 mAb and rat IgG2a mAb served as controls. All values represent the mean ± SEM of results obtained from five to 10 mice in each group. Statistical analysis is provided in Results.
|
 View larger version (111K): [in a new window] |
Figure 3. Histological tissue sections showing neutrophil infiltration in the skin of mice treated with mAb to P-selectin, E- and P-selectins, or PSGL-1 and control mice at 4 (A) and 8 (B) h after IC challenge. Neutrophils were revealed by H&E staining. Original magnification, x100.
|
 View larger version (119K): [in a new window] |
Figure 4. Histological tissue sections showing mast-cell accumulation in the skin of mice treated with mAb to P-selectin, E- and P-selectins, or PSGL-1 and control mice at 4 (A) and 8 (B) h after IC challenge. Mast cells (arrows) were detected as cells with metachromatic staining of granules in toluidine blue-stained sections. Original magnification, x50.
|
Leukocyte infiltration in the peritoneal Arthus reaction
The i.p. injection of antibody with the i.v. injection of antigen elicits a reverse-passive Arthus reaction characterized by leukocyte influx into the peritoneal cavity [31
]. Treatment with irrelevant, isotype-matched mAb instead of anti-P- or anti-E-selectin or anti-PSGL-1 mAb, served as controls, did not affect peritoneal neutrophil and mast-cell recruitment compared with untreated mice (Fig. 2B)
. After 4 h of IC challenge, neutrophil numbers in the peritoneal cavity were significantly reduced in mice with P-selectin blockade (13%, P<0.05), PSGL-1 blockade (39%, P<0.001), or blockade of P- and E-selectins (46%, P<0.001) compared with control mice, and they were not affected by E-selectin blockade alone. Blocking mAb to PSGL-1 exhibited significantly diminished neutrophil numbers relative to blocking mAb to P-selectin (P<0.05). There was no significant difference between blockade of PSGL-1 and that of E- and P-selectins. Similar results were obtained after 8 h of IC formation.
Mast-cell numbers were significantly reduced in mice with P-selectin blockade (32%, P<0.05), PSGL-1 blockade (33%, P<0.01), or combined P- and E-selectin blockade (54%, P<0.001) compared with control mice, and they were not affected by E-selectin blockade alone (Fig. 2B) . Blocking mAb to PSGL-1 did not significantly reduce mast-cell numbers relative to blocking mAb to P-selectin. In general, results obtained for mast-cell accumulation in the peritoneum after 4 h were comparable with those in the skin (Fig. 2A) . Unlike the cutaneous Arthus reaction, mast-cell number remained increased after 8 h relative to that after 4 h. There was no neutrophil and mast-cell influx in mice following i.p. injection of rabbit polyclonal IgG with systemic chicken egg albumin (data not shown). Thus, the effect of PSGL-1 blockade on neutrophil recruitment during the peritoneal Arthus reaction was stronger than that observed in the cutaneous Arthus reaction, whereas the effect of PSGL-1 blockade on mast-cell recruitment during the peritoneal Arthus reaction was similar to that observed in the cutaneous Arthus reaction.
Cytokine levels in the peritoneal Arthus reaction
IC-induced inflammation in the peritoneum is associated with the production and release of proinflammatory cytokines, including TNF- and IL-6, by infiltrating leukocytes [23
, 32
, 33
]. In our previous study, reduced infiltration of neutrophils and mast cells was associated with the decreased production of TNF- and IL-6 levels in the peritoneal Arthus reaction [28
]. To assess relative roles of adhesion molecules in the release of TNF- and IL-6 during the peritoneal Arthus reaction, TNF- and IL-6 levels were measured in peritoneal lavage samples after 4 and 8 h of IC formation. Treatment with irrelevant, isotype-matched mAb instead of anti-P- or anti-E-selectin or anti-PSGL-1 mAb, served as controls, did not affect peritoneal TNF- and IL-6 levels compared with untreated mice (Fig. 5A
and 5B
). TNF- levels were significantly increased by 2.8-fold after 4 h compared with the levels before IC challenge in control mice (P<0.01) but returned to the baseline levels by 8 h (Fig. 5A)
. A similar increase in TNF- levels was observed in mice with E-selectin blockade (P<0.01) as well as P-selectin blockade (P<0.01). TNF- levels were also significantly elevated in mice with E- and P-selectin blockade and in mice with PSGL-1 blockade after 4 h (P<0.05) relative to the baseline levels but were significantly (20–36%) lower than those found in control mice at the same time-point (P<0.05).
 View larger version (21K): [in a new window] |
Figure 5. Arthus reaction-induced production and release of TNF- (A) and IL-6 (B) in the peritoneal lavage fluid from mice treated with mAb to E- and/or P-selectin or PSGL-1 and control mice before and at 4 and 8 h after IC challenge. The peritoneal reverse-passive Arthus reaction was induced by the i.v. injection of chicken egg albumin, followed immediately by the i.p. injection of rabbit IgG anti-chicken egg albumin antibody. Four hours or 8 h later, the peritoneum was exposed by a middle abdominal incision, and 5 ml ice-cold PBS containing 0.1% BSA was injected into the peritoneal cavity via a 27-gauge needle. TNF- and IL-6 levels in the peritoneal lavage samples were determined by ELISA. Mice that received an i.v. injection of irrelevant, isotype-matched, purified rat IgG1 mAb and rat IgG2a mAb served as positive controls. Mice that received an i.v. injection of chicken egg albumin followed by an i.p. injection of polyclonal rabbit IgG served as negative controls. All values represent the mean ± SEM of results obtained from five to eight mice in each group. *, P < 0.05, and **, P < 0.01, versus levels before IC challenge in each group of mice.
|
and IL-6 levels in mice following i.p. injection of rabbit polyclonal IgG with systemic chicken egg albumin (Fig. 5)
. Thus, the reduced, peritoneal, inflammatory responses by blockade of adhesion molecules were generally associated with the reduced release of TNF-and IL-6. |
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and IL-6 (Figs. 2B
and 5)
. Taken together, these results demonstrate that PSGL-1 contributes to the development of the Arthus reaction by regulating the accumulation of neutrophils and mast cells at the inflammatory site. In the current study, PSGL-1 blockade by mAb 2PH1 inhibited the Arthus reaction more strongly than P-selectin blockade by mAb RB40.34 (Figs. 1 2 and 5) . We previously reported that L-selectin deficiency significantly reduces leukocyte accumulation in the Arthus reaction [26 ]. Although the main ligand for L-selectin has not been identified yet, L-selectin also recognizes the N terminus of PSGL-1, and PSGL-1 is suggested to function as one of the ligands of L-selectin [34 , 35 ]. Consistently, the interaction of L-selectin and PSGL-1 is required for the capture of free-flowing leukocytes to already adherent leukocytes [34 , 35 ]. Thus, PSGL-1 blockade may partly inhibit the function of L-selectin during the Arthus reaction. In addition, as PSGL-1 also appears to be capable of binding E-selectin, PSGL-1 blockade can affect the function of E-selectin in the Arthus reaction [36 , 37 ]. However, it has also been reported that functional blocking mAb to PSGL-1 by 2PH1 does not block interactions with E-selectin [18 ]. Therefore, L-selectin may be a better candidate to explain the additional effects of 2PH1. Taken together, the additional, inhibiting effect of PSGL-1 on the Arthus reaction relative to P-selectin blockade may be mediated by E- and/or L-selectin.
Our present study demonstrated that PSGL-1 blockade resulted in a further reduction than P-selectin blockade in peritoneal neutrophil accumulation, and PSGL-1 blockade had equivalent, inhibitory effects to P-selectin blockade on cutaneous neutrophil infiltration and cutaneous and peritoneal mast-cell infiltration (Figs. 2 3 4) . By contrast, loss or blockade of P-selectin has profound, inhibitory effects on neutrophil infiltration relative to that of PSGL-1 in thioglycollate-induced peritonitis model [17 , 29 ]. These conflicting results suggest that the interaction between PSGL-1 and P-selectin may vary according to organs, a type of reaction, and related inflammatory cells. However, it is also possible that this may be a result of the differential contribution of additional P-selectin ligands other than PSGL-1, including CD24 [38 , 39 ], components of the peripheral node addressin complex [40 ], and sulfated glycolipids [41 ], which could be involved in leukocyte recruitment in the Arthus reaction.
In this study, IC challenge induced rapid mast-cell accumulation, which was observed in skin and peritoneum and was significantly inhibited by PSGL-1 blockade (Figs. 2 and 4) . Although the exact routes of mast cells into inflammatory sites have not been determined yet, immature mast cells derived from bone marrow progenitors are thought to migrate through the circulation into tissues and subsequently differ into mature mast cells [42 ]. In a chronic contact hypersensitivity model with repeated hapten sensitization for 24 days at 2-day intervals, mast-cell number was increased 30 min after elicitation on day 24 [43 ]. Additionally, our previous study has shown that peritoneal murine mature mast cells express significant levels of PSGL-1 [28 ]. These results suggest that mature mast cells are rapidly recruited to inflamed foci using PSGL-1 possibly through circulation.
The current study showed that TNF- and IL-6 levels after 4 h in the peritoneal lavage were significantly increased compared with baseline levels, and blockade of PSGL-1, P-selectin, or P- and E-selectins significantly reduced these cytokine levels (Fig. 5)
. These inhibitory effects generally correlated with numbers of neutrophils and mast cells (Fig. 2)
. Not only neutrophils but also mast cells produce and release proinflammatory cytokines, including TNF- and IL-6 [44
, 45
]. Therefore, these results suggest that PSGL-1 regulates neutrophil and mast cell, and thereby, TNF- and IL-6 levels are influenced in the Arthus reaction. However, PSGL-1 blockade profoundly reduced edema compared with P-selectin blockade in the cutaneous Arthus reaction, although the numbers of neutrophils and mast cells were similar between these groups (Figs. 1
and 2)
. Deficiency or blocking of adhesion molecules influences not only leukocyte trafficking but also survival and the state of activation by their role in out-side-in signaling [46
]. Therefore, it is possible that infiltrating neutrophils and mast cells may be functionally impaired for cytokine release in the blockade of PSGL-1.
Several studies reported that loss or blockade of PSGL-1 reduces inflammatory response in the thioglycollate-induced peritonitis, oxazolane-induced contact hypersensitivity model and carotid artery wire injury model [17 18 19 20 ]. Moreover, it has been reported that treatment with recombinant soluble PSGL-1, which is developed as an antagonist to P-selectin, is beneficial in animal models of deep-vein thrombosis [47 ], myocardial-ischemia reperfusion [48 , 49 ], renal-ischemia reperfusion [50 ], hepatic-ischemia reperfusion [51 ], initial hyperplasia after angioplasty [52 , 53 ], arterial thrombosis [54 , 55 ], and ocular allergy [56 ]. These previous studies suggest that blocking inflammatory response through inhibition of selectins and PSGL-1 interaction provides a mean to the therapy of these diseases. In addition, the results of the present study indicate that PSGL-1 contributes to IC-mediated tissue injury by mediating the accumulation of neutrophils and mast cells, which release proinflammatory cytokines, and also indicate that PSGL-1 blockade profoundly inhibits the inflammation compared with single blockade of E- or P-selectin. This suggests that PSGL-1 is a more efficient, therapeutic target for human IC-mediated diseases than E- or P-selectin.
Received December 23, 2003; revised March 25, 2004; accepted April 4, 2004.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
RI/III and C5aR in the reverse Arthus reaction J. Immunol. 164,1065-1070-receptors in the Arthus reaction Mol. Immunol. 36,893-903[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  T. Hara, K. Shimizu, F. Ogawa, K. Yanaba, Y. Iwata, E. Muroi, M. Takenaka, K. Komura, M. Hasegawa, M. Fujimoto, et al. Platelets Control Leukocyte Recruitment in a Murine Model of Cutaneous Arthus Reaction Am. J. Pathol., January 1, 2010; 176(1): 259 - 269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Orito, M. Fujimoto, N. Ishiura, K. Yanaba, T. Matsushita, M. Hasegawa, F. Ogawa, K. Takehara, and S. Sato Intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 cooperatively contribute to the cutaneous Arthus reaction J. Leukoc. Biol., May 1, 2007; 81(5): 1197 - 1204. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
