Severe inflammatory defect and reduced viability in CD18 and E-selectin double-mutant mice

CD18-deficient mice (CD18^–/– mice) have a severe leukocyte recruitment defect in some organs, and no detectable defect in other models. Mice lacking E-selectin (CD62E^–/– mice) have either no defect or a mild defect of neutrophil infiltration, depending on the model. CD18^–/–CD62E^–/–, but not CD18^–/–CD62P^–/–, mice generated by crossbreeding failed to thrive, reaching a maximum body weight of 10–15 grams. To explore the mechanisms underlying reduced viability, we investigated lethally irradiated CD62E^–/– mice that were reconstituted with CD18^–/– bone marrow. These mice, but not single-mutant controls, showed tenfold-increased rolling velocities in a TNF-α–induced model of inflammation. Leukocyte adhesion efficiency in CD18^–/–CD62E^–/– mice was reduced by 95%, and hematopoiesis was drastically altered, including severe bone marrow and blood neutrophilia and elevated G-CSF and GM-CSF levels. The greatly reduced viability of CD18^–/–CD62E^–/– mice appears to result from an inability to mount an adequate inflammatory response. Our data show that cooperation between E-selectin and CD18 integrins is necessary for neutrophil recruitment and that alternative adhesion pathways cannot compensate for the loss of these molecules.

Leukocyte recruitment to sites of inflammation requires leukocyte capture, rolling, and adhesion on activated endothelial cells. Leukocyte recruitment has been viewed as a cascade of events occurring in separable, sequential steps that are dependent on different adhesion molecules. Leukocyte capture and rolling are largely mediated by selectins and their ligands (1, 2); firm adhesion and transmigration require β₂ integrins (3, 4). The β₂ integrins are heterodimeric molecules composed of an α chain (CD11a, b, c, or d) and a common β chain (CD18). The development of various adhesion molecule–deficient mice has provided valuable information on the mechanisms of leukocyte recruitment.

The phenotypes of the different mice lacking one, two, or three selectins, one or all β₂ integrins, ICAM-1 alone, and both ICAM-1 and P-selectin have been described previously (3, 5–20). E-selectin–deficient mice (CD62E^–/– mice) show elevated leukocyte rolling velocities (5), increased leukocyte rolling flux (5), a reduced ability to produce adhesion in response to chemoattractants (21), and impairment of inflammatory function in at least three different models (5, 21, 22). CD18-null (CD18^–/–) mice have severe defects in leukocyte recruitment, adhesion, and activation (3). In TNF-α–induced inflammation, leukocyte rolling velocities in CD18^–/– mice are significantly elevated, indicating that CD18 integrins are involved in mediating leukocyte rolling (23). Leukocyte rolling velocities were increased even further, and the efficiency of leukocyte adherence was significantly decreased in CD18^–/– mice that were given an antibody that blocks E-selectin function (23). These data suggest that the steps in leukocyte recruitment are not discrete, separable events, but that the functions of selectins and integrins overlap to mediate efficient leukocyte recruitment. Therefore, we hypothesized that mice deficient in both CD18 and E-selectin (CD18^–/–CD62E^–/– mice) would display a severe phenotype, due to an inability to recruit leukocytes to sites of injury and inflammation. To test this hypothesis, we generated CD18^–/–CD62E^–/– double-mutant mice by crossbreeding. CD18^–/–CD62E^–/– mice demonstrated severely reduced viability. To obtain adult mice to investigate the inflammatory defect in CD18^–/–CD62E^–/– mice, we generated mice lacking both CD18 and E-selectin by transplantation of CD18^–/– bone marrow into adult CD62E^–/– mice. This study describes the phenotype of these mice and provides a mechanistic explanation for the severe inflammatory defect and lethality.

Phenotype of CD18^–/–CD62E^–/– and CD18^–/–CD62P^–/– mice. Mice deficient in CD18 and E-selectin (CD18^–/–CD62E^–/– mice) were generated by interbreeding CD18 and E-selectin single mutants. CD18^–/–CD62E^–/– mutants from CD18^+/–CD62E^–/– breeders were born at the expected Mendelian ratios (8 mutants of 30 mice, P = not significant), providing no evidence for in utero lethality. CD18^–/–CD62E^–/– mice that survived to weaning were runted compared with littermates (Figure 1). They failed to thrive, reaching maximum body weights of 10–15 grams. CD18^–/–CD62E^–/– mice required constant care for survival, including cutting of teeth and being given crushed food and antibiotics. Accurate survival curves could not be obtained because sick mice were killed in accordance with IACUC procedures. Most of the CD18^–/–CD62E^–/– mice that survived to 6 weeks of age developed mucocutaneous skin lesions and overgrowth of the incisors. They failed to breed and had reduced viability.

Figure 1

Mice deficient in CD18 integrins and E-selectin fail to thrive. CD18^–/–CD62E^–/– double-mutant mice are runted at birth and require constant care for survival. A typical CD18^–/–CD62E^–/– mutant (left) is shown with a CD18^+/+CD62E^–/– littermate at 12 days of age. The double mutant was significantly smaller (2.0 g) than its littermate (5.7 g).

In contrast, CD18^–/–CD62P^–/– mice generated by interbreeding single mutants were indistinguishable from littermates. CD18^–/–CD62P^–/– mice showed no differences in body weights at birth or at weaning compared with littermates. CD18^–/–CD62P^–/– mice did not have weaning weights that differed from controls (CD18^+/–CD62P^+/–) (11.3 ± 1.3 g and 12.1 ± 1.6 g, respectively; mean ± SD). CD18^–/–CD62P^–/– mice survived to adulthood (> 90% over 3 months) and bred successfully.

Since CD18 is exclusively expressed on bone marrow–derived blood cells, and E-selectin is restricted to endothelial cells, mice deficient in CD18 and E-selectin were generated by transplanting lethally irradiated CD62E^–/– mice with CD18^–/– bone marrow. As controls, wild-type, CD62E^–/–, and CD18^–/– mice were also generated through the appropriate bone marrow transplantation. All nontransplanted mice died within 2 weeks of irradiation (1.2 Gy). Despite residing in a pathogen-free barrier facility, approximately 50% of the CD18^–/–CD62E^–/– mice generated by bone marrow transplantation died within a few weeks of transplantation, compared with less than 5% for wild-type, CD18^–/–, and CD62E^–/– mice. CD18^–/–CD62E^–/– double-mutant mice became progressively sick, resulting in chronic ulcerative dermatitis and significant weight loss (body weights less than 20 grams, reduced from 25 grams at the time of transplantation). Circulating neutrophil counts in CD18^–/–CD62E^–/– mice became increasingly elevated as a function of time after transplantation (reaching 50,000–150,000/μl).

Four weeks after reconstitution, peripheral blood and bone marrow of chimeric mice were analyzed for CD18 expression (Figure 2). CD18 expression was undetectable on bone marrow and blood cells in mice transplanted with CD18^–/– bone marrow, confirming that reconstitution was efficient and complete.

Figure 2

Expression of CD18 (mAb C71/16) on unfractionated bone marrow cells 4 weeks after reconstitution. Neutrophils were identified and gated for expression of Gr-1 antigen. (a) Isotype control. Irradiated wild-type mice transplanted with wild-type bone marrow expressed high levels of CD18 (b). Transplantation of CD18^–/– mouse bone marrow into irradiated wild-type mice (c) and irradiated CD62E^–/– mice (d) resulted in a complete absence of CD18 expression on all bone marrow and blood cells.

Histopathology. CD18^–/–CD62E^–/– mice showed moderate to severe ulcerative dermatitis similar to that of CD18^–/– mice (3). Histopathology of the skin revealed significant hyperplasia of the epithelium, hyperkeratosis, acute ulceration with bacterial colonies under the ulcer, mild lymphocytic and neutrophilic infiltration throughout the dermis, and a thickened dermis (Figure 3a). Gram-stained histologic sections of CD18^–/–CD62E^–/– mice showed Gram-positive cocci (Figure 3b). Gram staining of the wild-type skin sections showed no evidence of organisms (data not shown). Bacterial cultures of the lung, liver, spleen, and blood of six CD18^–/–CD62E^–/– mice showed one spleen positive for a Corynebacterium species. None of the four wild-type mice showed positive cultures in any organs. CD18^–/–CD62E^–/– mice also displayed splenomegaly, with an approximately threefold increase in spleen weight due to myeloid hyperplasia within the red pulp consisting of primitive, intermediate, and differentiated cells. Myeloid hyperplasia was also present in the bone marrow of CD18^–/–CD62E^–/– mice.

Figure 3

(a) Skin section from the facial area of a CD18^–/–CD62E^–/– mouse generated through bone marrow transplantation, stained with hematoxylin and eosin. Section shows hyperplasia of the epithelium, hyperkeratosis, acute ulceration with bacterial colonies in the ulcer, mild lymphocytic and neutrophilic infiltration throughout the dermis, and a thickened dermis. (b) Gram stain of histologic section from the CD18^–/–CD62E^–/– mouse in a showed Gram-positive cocci (dark staining pattern) and a coagulum of necrotic epithelium and exudate (dark red staining). (c) Skin section from the face of a C57BL/6 control mouse generated by bone marrow transplantation, stained with hematoxylin and eosin. Original magnification, 30×.

To quantify neutrophil content in the spleen, MPO levels were measured in wild-type, CD62E^–/–, CD18^–/–, and CD18^–/–CD62E^–/– mice. Spleen weights were increased threefold in CD18^–/– and CD18^–/–CD62E^–/– mice compared with CD62E^–/– and wild-type mice (Table 1). Consistent with extramedullary hematopoiesis, a drastic increase in MPO activity was observed in the spleen of CD18^–/– and CD18^–/–CD62E^–/– mice compared with wild-type and CD62E^–/– mice (Table 1). In addition, neutrophil content in the lungs was elevated in the CD18^–/– and CD18^–/–CD62E^–/– mice (Table 1), suggesting abnormal neutrophil trafficking and margination.

Table 1

Systemic neutrophil counts, spleen weights, and tissue neutrophil infiltration

Analysis of hematopoiesis in CD18^–/–CD62E^–/– mice. Consistent with the highly elevated level of neutrophils in the peripheral blood, a morphological analysis of stained bone marrow cytospins showed a drastic increase in the percentage of neutrophils and neutrophil precursors compared with wild-type mice (data not shown). Unfractionated bone marrow and peripheral blood were analyzed for expression of the myeloid marker Gr-1 (31) using flow cytometry (Figure 4, a and b, respectively). CD18^–/– mice showed an increase in bone marrow Gr-1 expression of approximately 75%, compared with approximately 50% in wild-type and CD62E^–/– mice (Figure 4a). Remarkably, the percentage of Gr-1–expressing cells in the bone marrow and blood of CD18^–/–CD62E^–/– mice exceeded 90% (Figure 4, a and b). In vitro hematopoietic progenitor colony assays showed no increase in GM-CFU colony formation at day 7 in CD18^–/–CD62E^–/– mice compared with wild-type mice (data not shown).

Figure 4

Expression of Gr-1 antigen on unfractionated bone marrow cells (a) and peripheral blood leukocytes (b). Gr-1 expression in CD18^–/– mice and CD18^–/–CD62E^–/– mice is dramatically elevated (percent of cells expressing Gr-1 indicated; gate set to exclude 95% of cells stained with isotype control, not shown). (c) Concentration of G-CSF and (d) GM-CSF in plasma as measured by ELISA. Data expressed as mean ± SEM.

Serum G-CSF and GM-CSF levels. G-CSF levels in wild-type, CD62E^–/–, CD18^–/–, and CD18^–/–CD62E^–/– mice were measured to investigate the possible role of G-CSF in mediating the significant overproduction of neutrophils. Peripheral blood neutrophil counts were significantly elevated in CD18^–/– and CD18^–/–CD62E^–/– mice. Serum G-CSF levels in CD18^–/–CD62E^–/– mice were significantly higher (304 ± 75 pg/ml) than in wild-type, CD62E^–/–, and CD18^–/– mice (none detected) (Figure 4c), indicating that G-CSF may mediate the neutrophilia observed in CD18^–/–CD62E^–/– mice. To determine the possible involvement of GM-CSF in the neutrophilia, GM-CSF serum levels were also measured. GM-CSF was elevated in CD18^–/–CD62E^–/– mice (2.1 ± 1.2 pg/ml) compared with wild-type, CD62E^–/–, and CD18^–/– mice (none detected) (Figure 4d).

Intravital microscopy. Leukocyte rolling and adhesion were investigated in wild-type, CD62E^–/–, CD18^–/–, and CD18^–/–CD62E^–/– mice in the cremaster muscle 2 hours after injection of TNF-α. During anesthesia and surgery in preparation for intravital microscopy experiments performed at 4 weeks after transplantation, all CD18^–/–CD62E^–/– mice weighing less than 20 grams died, and were not used for experiments. Therefore, the CD18^–/–CD62E^–/– mice used for intravital microscopy probably represent a subgroup of healthier mice. Leukocyte rolling and adhesion were studied in hemodynamically similar venules over a shear-rate range from 100 s^–1 to approximately 1500 s^–1 (Table 2).

Table 2

Hemodynamic data

TNF-α induces the expression of E-selectin and enhances the expression of P-selectin in mouse cremaster muscle venules (32). To determine if CD18 and E-selectin cooperatively mediate efficient slow leukocyte rolling, leukocyte rolling velocities were measured in wild-type, CD62E^–/–, CD18^–/–, and CD18^–/–CD62E^–/– mice. The majority of leukocytes rolled at velocities under 5 μm/s in wild-type mice, resulting in an average leukocyte rolling velocity of 6.9 ± 0.2 μm/s (Figure 5). A noticeable shift in the leukocyte rolling velocity distribution (peak at 10–15 μm/s) and an increase in the average rolling velocity (21.1 ± 0.5 μm/s) occurred in CD62E^–/– mice (Figure 5). Leukocyte rolling under 5.0 μm/s in CD62E^–/– mice was very rare (1.6% of rolling leukocytes, compared with 50.2% in wild-type mice), confirming that E-selectin is necessary for slow leukocyte rolling (5). The average rolling velocity in CD18^–/– mice was increased to 22.7 ± 0.8 μm/s, due to a broadened rolling velocity distribution (leukocyte rolling velocities up to ∼100 μm/s; Figure 5). A further dramatic broadening of the rolling velocity distribution occurred in the CD18^–/–CD62E^–/– mice (Figure 5). Very few leukocytes rolled slower than 10 μm/s, but a large population rolled faster than 50 μm/s (36%), shifting the rolling velocity distribution peak to 15–20 μm/s. The average rolling velocity was 50.1 ± 1.4 μm/s in CD18^–/–CD62E^–/– mice, which was significantly higher than in either of the single mutants. These findings show that CD18 and E-selectin work cooperatively to efficiently mediate slow leukocyte rolling in TNF-α–treated venules.

Figure 5

Leukocyte rolling velocity distributions in TNF-α–treated cremaster muscle venules of wild-type, CD62E^–/–, CD18^–/–, and CD18^–/–CD62E^–/– mice. Arrowheads indicate mean leukocyte rolling velocity. All leukocytes rolling at velocities over 140 μm/s are represented by last bar. Leukocyte rolling velocities in wild-type mice and CD18^–/–CD62E^–/– mice were significantly different from those in other groups (P < 0.05).

In many systems, leukocyte rolling has been reported to be influenced by wall shear rate (33, 34). Therefore, we stratified the venules into low, intermediate, and high shear-rate groups (100–400 s^–1, 400–700 s^–1, and 700–1500 s^–1). Rolling velocity distributions changed little in wild-type mice as the shear rate was increased (Figure 6). Increasing the shear rate in CD62E^–/– mice produced a noticeable shift in the rolling velocity distribution (Figure 6). A similar shift occurred in the low and medium shear-rate ranges in CD18^–/– mice, as did a broadening of the velocity distribution, as reflected by the decreased slope of the cumulative rolling velocity histograms. A more drastic shift and broadening in rolling velocity distribution was seen in the high shear-rate range in CD18^–/– mice. In CD18^–/–CD62E^–/– mice, rolling velocity was dramatically elevated over all shear rates (Figure 6). Leukocyte rolling became very unstable in CD18^–/–CD62E^–/– at high shear rates. The leukocytes no longer rolled uniformly on the venule surface, but generally skipped along the surface, indicating repetitive capture and detachment events (inset in Figure 6). By contrast, leukocytes in wild-type mice showed slow, steady rolling at high wall shear rates. Table 3 summarizes the average leukocyte rolling velocities measured in the three shear-rate ranges. In all cases, the increase in average leukocyte rolling velocity in CD18^–/–CD62E^–/– mice was approximately additive of the increases seen in CD18^–/– and CD62E^–/– mice.

Figure 6

Effect of shear rate on leukocyte rolling velocity distributions in TNF-α–treated cremaster muscle venules of wild-type, CD62E^–/–, CD18^–/–, and CD18^–/–CD62E^–/– mice. Cumulative frequency curves of leukocyte rolling velocities in shear-rate ranges of 100–400 s^–1 (thin dotted line), 400–700 s^–1 (thin solid line), and 700–1500 s^–1 (thick solid line). All leukocytes rolling at velocities over 140 μm/s are represented by last data point. Median rolling velocity for each shear-rate range is indicated by vertical lines. Inset shows rolling velocities of three leukocytes rolling over a 200-μm distance in a venule of a wild-type mouse (closed circles, wall shear rate 754 s^–1) and of three leukocytes rolling in a CD18^–/–CD62E^–/– mouse (open circles, wall shear rate 729 s^–1).

Table 3

Average leukocyte rolling velocities (μm/s)

The ability of rolling leukocytes to adhere to TNF-α–stimulated endothelium was studied to determine if leukocyte rolling velocity is an important parameter mediating efficient leukocyte recruitment. Consistent with highly elevated systemic counts, the leukocyte rolling flux was greatly enhanced in CD18^–/– and CD18^–/–CD62E^–/– mice (Figure 7a). To show how efficiently rolling leukocytes became firmly adhered, leukocyte adhesion was normalized for leukocyte rolling flux (Figure 7b). In CD18^–/–CD62E^–/– double-mutant mice, 95% fewer leukocytes became adherent per 100 rolling leukocytes than in wild-type mice (199 ± 42 compared with 3,190 ± 423 in wild-type mice; 1,269 ± 448 in CD62E^–/– mice; and 1,099 ± 218 in CD18^–/– mice).

Figure 7

Leukocyte rolling and adhesion in TNF-α–treated cremaster muscle venules of wild-type, CD62E^–/–, CD18^–/–, and CD18^–/–CD62E^–/– mice. (a) Leukocyte rolling flux (number of rolling cells/min). (b) Adhesion efficiency (adherent cells/mm² per 100 rolling cells/min). Data represented as mean ± SEM, derived from 47–68 venules. ^ASignificantly different from wild type. ^BSignificantly different from CD62E^–/–. ^CSignificantly different from CD18^–/–CD62E^–/–. ^DSignificantly different from all other groups (P < 0.05).

Next, we tested the correlation between leukocyte rolling velocity and the ability of rolling leukocytes to become adherent (Figure 8). Leukocyte adhesion efficiency was significantly reduced in each mouse type as the leukocyte rolling velocity increased due to increased shear rate. Within each shear-rate class, firm leukocyte adhesion was modestly reduced by the absence of E-selectin or CD18, and was drastically reduced in CD18^–/–CD62E^–/– mice. The lack of firmly adherent leukocytes in CD18^–/–CD62E^–/– mice correlated with the highly elevated leukocyte rolling velocities in all three shear-rate classes. Adhesion efficiency was most severely affected when leukocyte rolling velocity exceeded approximately 20 μm/s, and when transit times fell below about 5 seconds for a 100-μm segment of venule.

Figure 8

Inverse correlation between leukocyte rolling velocity and leukocyte adhesion efficiency in TNF-α–treated cremaster muscle venules of wild-type mice (squares), CD62E^–/– mice (triangles), CD18^–/– mice (diamonds), and CD18^–/–CD62E^–/– mice (circles). Adhesion efficiency, represented as the ability of rolling leukocytes to become adherent, was drastically reduced as the leukocyte rolling velocity was increased. Data represented as mean ± SEM (SEM not shown if smaller than symbol).

Giemsa-stained cremaster whole mounts were used to differentiate intravascular and transmigrated leukocytes (Figure 9, a–d). We found a remarkable impairment in the ability of neutrophils to interact successfully with the vessel wall in CD18^–/–CD62E^–/– mice. Intravascular (rolling and firmly adhered) and transmigrated neutrophils were reduced by 80% in CD18^–/–CD62E^–/– mice (Figure 9h) compared with wild-type, CD62E^–/–, and CD18^–/– mice (Figure 9, e–g). No other adhesion molecule–mutant mouse demonstrated a comparable reduction of neutrophil recruitment in this assay.

Figure 9

Differentiation of intravenular leukocytes. Left column shows a representative venule for each genotype. Whole mounts were obtained from wild-type (a), CD62E^–/– (b), CD18^–/– (c), and CD18^–/–CD62E^–/– mice (d) and stained with Giemsa. Bar = 20 μm. (e–h) Differential leukocyte counts in blood and cremaster venules, and transmigrated leukocytes in cremaster tissue. Filled bars, neutrophils; open bars, mononuclear cells. ^AP < 0.05 compared with all other groups.

Our data show that mice lacking both CD18 integrins and E-selectin have severely impaired viability. When adult animals are generated by bone marrow transplantation, they show the most severe inflammatory defect seen in any adhesion molecule–deficient mouse described to date. Their inability to recruit neutrophils to sites of inflammation results in high systemic neutrophil counts, which appear to be maintained through elevated levels of G-CSF (a key regulator of neutrophil production) and GM-CSF. Using intravital microscopy, we were able to show a strong inverse relationship between the velocity of rolling leukocytes and the efficiency of adhesion.

Based on the phenotypes of the CD18 and E-selectin single-mutant mice, it could not be expected that the double mutation would result in severely reduced viability. In the first publication to characterize E-selectin–null mice, the mice were described as lacking a clear inflammatory phenotype (9), although defects in leukocyte rolling (5) and adhesion (21), and increased susceptibility to Streptococcus pneumoniae (22) were described later. CD18^–/– mice develop skin ulcerations and have elevated neutrophil counts and immunoglobulin levels, increased susceptibility to S. pneumoniae, and a severe defect in leukocyte adhesion and T-cell activation (3). They also have a defect in leukocyte recruitment to peritonitis (35), and a lack of neutrophil recruitment to the skin (36). The unexpected severity of the CD18^–/–CD62E^–/– mouse phenotype suggests that CD18 integrins and E-selectin serve a common function at a critical juncture or bottleneck in the leukocyte adhesion cascade, at the transition from rolling to firm adhesion.

Although we found many phenotypic differences between CD18^–/–CD62E^–/– mice and single mutants, our search for a cause of lethality was not successful. Neither the skin lesions, the neutrophil accumulation in the lungs, nor the hematopoietic changes and splenomegaly are likely causes for early death and failure to thrive. However, the combination of these defects may well be the cause of the reduced viability.

The severe phenotype seen in CD18^–/–CD62E^–/– mice is somewhat reminiscent of E- and P-selectin double-mutant (CD62E^–/–CD62P^–/–) mice. CD62E^–/–CD62P^–/– mice develop spontaneous mucocutaneous infections with skin ulcerations, even in a specific pathogen–free environment. They have very high circulating neutrophil counts (elevated 10- to 20-fold), which are thought to result from either an inability to recruit neutrophils to sites of inflammation, elevated circulating inflammatory cytokines (11), or alterations in the hematopoietic system (12, 37). CD62E^–/–CD62P^–/– mice demonstrate no leukocyte rolling after tissue trauma or after short-term cytokine treatment (11), but after prolonged (6–8 hour) cytokine treatment (6), these mice show a small number of rolling leukocytes and significant leukocyte recruitment. Similar to the current findings, E- and P-selectin were identified as serving an overlapping function at a critical juncture in the inflammatory adhesion cascade, namely, stable leukocyte rolling (11).

Two other mice with mutations in genes encoding for inflammatory adhesion molecules, α₄ integrins (38) and VCAM-1 (39), have produced a lethal phenotype. However, in these mice, lethality was not caused by an inflammatory defect, but by developmental defects affecting the formation of the chorioallantoic fusion (38, 39), epicardium, and coronary arteries (38). Although the phenotype of E- and P-selectin double-mutant mice is severe, it did not lead to reduced viability of juvenile mice. Therefore, the CD18^–/–CD62E^–/– mouse represents the most severe inflammatory defect yet described.

One of the striking phenotypic traits of the CD18^–/–CD62E^–/– mice is their high circulating neutrophil count. In previous studies with several other adhesion molecule–mutant mice, high circulating neutrophil counts have been observed. Mild elevations were seen in ICAM-1^–/–, LFA-1^–/–, and P-selectin–null mice (7, 18). Severe elevations were found in CD62E^–/–CD62P^–/– mice (11, 12) and CD18^–/– (3) mice. Some strains were investigated for possible causes of these elevated counts. In CD62E^–/–CD62P^–/– mice, elevated levels of GM-CSF and altered hematopoiesis (12) have been described. We also found elevated GM-CSF in CD18^–/–CD62E^–/– mice. However, GM-CSF is unlikely to be a primary causative factor for elevated neutrophil counts (40). We hypothesized that G-CSF would be a more likely candidate, based on the phenotype of the G-CSF–overexpressing transgenic mice and G-CSF–null mice (41), which have severe neutrophilia and neutropenia, respectively. G-CSF is considered to be the major regulator of granulocyte production (42), and our finding that CD18^–/–CD62E^–/– mice exhibit increased plasma G-CSF levels associated with neutrophilia supports this conclusion. The findings of extramedullary hematopoiesis in the spleen and the elevated percentage of neutrophil precursors are both consistent with a severe hematopoietic dysregulation in CD18^–/–CD62E^–/– mice.

Our intravital microscopy findings in CD18^–/–CD62E^–/– mice confirm and extend previous findings made in CD18^–/– mice treated with an E-selectin antibody (23). We now show that the defect of slow leukocyte rolling is dependent on wall shear rate, and, by inference, wall-shear stress (which is directly related to wall shear rate). In wild-type mice, rolling velocity is almost invariant over a wide range of wall shear rates in TNF-α−induced inflammation; even when E-selectin is missing, the dependence on wall shear rate is moderate, although the rolling velocities are increased throughout. However, eliminating CD18 has a large impact on rolling velocities at high wall shear rates. The shear-rate influence is strongest in the CD18^–/–CD62E^–/– double-mutant mice. In CD18^–/–CD62E^–/– mice, leukocyte recruitment is essentially absent in venules with moderate (400–700 s^–1) or high (> 700 s^–1) shear rates. Only a very modest rate of recruitment is possible in the venules with the slowest flow (wall shear rate < 400 s^–1). This is significant for two reasons. First, venules with shear rates above 400 s^–1 are common in most microcirculatory beds. Although complete samples are rarely reported (43, 44), it is estimated that about two-thirds of venules fall in this shear-rate range. Second, leukocyte delivery is directly proportional to flow rate, which has a linear dependence on blood-flow velocity and a quadratic dependence on vessel radius. On average, higher wall shear rates are observed more frequently in larger venules, whereas the small postcapillary venules show the lowest wall shear rates. This means that in CD18^–/–CD62E^–/– mice, the most important venules, carrying the most leukocytes, are eliminated from contributing to leukocyte recruitment.

In conclusion, we have identified an overlapping function of CD18 integrins and E-selectin at a crucial step in the leukocyte adhesion cascade. Apparently, no other molecules can substitute for the essential function of these molecules in converting rolling to firm adhesion. This leads to severe inflammatory impairment and severely reduced viability in CD18^–/–CD62E^–/– mice.

We thank Russell Lindsey (University of Alabama at Birmingham) for the detailed histological analysis. Funding for the study was provided by NIH grants HL-54136 to K. Ley and AI-32177 to A.L. Beaudet. S.B. Forlow was supported by NRSA HL-10447.

1. Butcher, EC. Leukocyte-endothelial cell recognition: three (or more) steps to specificity and diversity. Cell 1991. 67:1033-1036.
   View this article via: PubMed CrossRef Google Scholar
2. Kansas, GS. Selectins and their ligands: current concepts and controversies. Blood 1996. 88:3259-3287.
   View this article via: PubMed Google Scholar
3. Scharffetter-Kochanek, K, et al. Spontaneous skin ulceration and defective T cell function in CD18 null mice. J Exp Med 1998. 188:119-131.
   View this article via: PubMed CrossRef Google Scholar
4. Arfors, KE, et al. A monoclonal antibody to the membrane glycoprotein complex CD18 inhibits polymorphonuclear leukocyte accumulation and plasma leakage in vivo. Blood 1987. 69:338-340.
   View this article via: PubMed Google Scholar
5. Kunkel, EJ, Ley, K. Distinct phenotype of E-selectin-deficient mice. E-selectin is required for slow leukocyte rolling in vivo. Circ Res 1996. 79:1196-1204.
   View this article via: PubMed Google Scholar
6. Jung, U, Ramos, CL, Bullard, DC, Ley, K. Gene-targeted mice reveal importance of L-selectin-dependent rolling for neutrophil adhesion. Am J Physiol 1998. 274:H1785-H1791.
   View this article via: PubMed Google Scholar
7. Mayadas, TN, Johnson, RC, Rayburn, H, Hynes, RO, Wagner, DD. Leukocyte rolling and extravasation are severely compromised in P selectin-deficient mice. Cell 1993. 74:541-554.
   View this article via: PubMed CrossRef Google Scholar
8. Tedder, TF, Steeber, DA, Pizcueta, P. L-selectin-deficient mice have impaired leukocyte recruitment into inflammatory sites. J Exp Med 1995. 181:2259-2264.
   View this article via: PubMed CrossRef Google Scholar
9. Labow, MA, et al. Characterization of E-selectin-deficient mice: demonstration of overlapping function of the endothelial selectins. Immunity 1994. 1:709-720.
   View this article via: PubMed CrossRef Google Scholar
10. Arbones, ML, et al. Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin-deficient mice. Immunity 1994. 1:247-260.
    View this article via: PubMed CrossRef Google Scholar
11. Bullard, DC, et al. Infectious susceptibility and severe deficiency of leukocyte rolling and recruitment in E-selectin and P-selectin double mutant mice. J Exp Med 1996. 183:2329-2336.
    View this article via: PubMed CrossRef Google Scholar
12. Frenette, PS, Mayadas, TN, Rayburn, H, Hynes, RO, Wagner, DD. Susceptibility to infection and altered hematopoiesis in mice deficient in both P- and E-selectins. Cell 1996. 84:563-574.
    View this article via: PubMed CrossRef Google Scholar
13. Jung, U, Ley, K. Mice lacking two or all three selectins demonstrate overlapping and distinct functions for each selectin. J Immunol 1999. 162:6755-6762.
    View this article via: PubMed Google Scholar
14. Coxon, A, et al. A novel role for the β2 integrin CD11b/CD18 in neutrophil apoptosis: a homeostatic mechanism in inflammation. Immunity 1996. 5:653-666.
    View this article via: PubMed CrossRef Google Scholar
15. Schmits, R, et al. LFA-1-deficient mice show normal CTL responses to virus but fail to reject immunogenic tumor. J Exp Med 1996. 183:1415-1426.
    View this article via: PubMed CrossRef Google Scholar
16. Lu, H, et al. LFA-1 is sufficient in mediating neutrophil emigration in Mac-1-deficient mice. J Clin Invest 1997. 99:1340-1350.
    View this article via: JCI PubMed CrossRef Google Scholar
17. Xu, H, et al. Leukocytosis and resistance to septic shock in intercellular adhesion molecule 1-deficient mice. J Exp Med 1994. 180:95-109.
    View this article via: PubMed CrossRef Google Scholar
18. Sligh, JEJ, et al. Inflammatory and immune responses are impaired in mice deficient in intercellular adhesion molecule 1. Proc Natl Acad Sci USA 1993. 90:8529-8533.
    View this article via: PubMed CrossRef Google Scholar
19. Kunkel, EJ, et al. Absence of trauma-induced leukocyte rolling in mice deficient in both P-selectin and intercellular adhesion molecule 1. J Exp Med 1996. 183:57-65.
    View this article via: PubMed CrossRef Google Scholar
20. Bullard, DC, et al. P-selectin/ICAM-1 double mutant mice: acute emigration of neutrophils into the peritoneum is completely absent but is normal into pulmonary alveoli. J Clin Invest 1995. 95:1782-1788.
    View this article via: JCI PubMed CrossRef Google Scholar
21. Ley, K, Allietta, M, Bullard, DC, Morgan, S. Importance of E-selectin for firm leukocyte adhesion in vivo. Circ Res 1998. 83:287-294.
    View this article via: PubMed Google Scholar
22. Munoz, FM, Hawkins, EP, Bullard, DC, Beaudet, AL, Kaplan, SL. Host defense against systemic infection with Streptococcus pneumoniae is impaired in E-, P-, and E-/P-selectin-deficient mice. J Clin Invest 1997. 100:2099-2106.
    View this article via: JCI PubMed CrossRef Google Scholar
23. Jung, U, Norman, KE, Scharffetter-Kochanek, K, Beaudet, AL, Ley, K. Transit time of leukocytes rolling through venules controls cytokine-induced inflammatory cell recruitment in vivo. J Clin Invest 1998. 102:1526-1533.
    View this article via: JCI PubMed CrossRef Google Scholar
24. Pries, AR. A versatile video image analysis system for microcirculatory research. Int J Microcirc Clin Exp 1988. 7:327-345.
    View this article via: PubMed Google Scholar
25. Lipowsky, HH, Zweifach, BW. Application of the “two-slit” photometric technique to the measurement of microvascular volumetric flow rates. Microvasc Res 1978. 15:93-101.
    View this article via: PubMed CrossRef Google Scholar
26. Reneman, R.S., Woldhuis, B., oude Egbrink, M.G.A, Slaaf, D.W., and Tangelder, G.J. 1992. Concentration and velocity profiles of blood cells in the microcirculation. In Advances in cardiovascular engineering. N.H.C. Hwang, V.T. Turitto, and M.R.T. Yen, editors. Plenum Publishing Corp. New York, New York, USA. 25–40.
    View this article via: PubMed Google Scholar
27. Ley, K, et al. Sequential contribution of L- and P-selectin to leukocyte rolling in vivo. J Exp Med 1995. 181:669-675.
    View this article via: PubMed CrossRef Google Scholar
28. Schwarzenberger, P, et al. IL-17 stimulates granulopoiesis in mice: use of an alternate, novel gene therapy-derived method for in vivo evaluation of cytokines. J Immunol 1998. 161:6383-6389.
    View this article via: PubMed Google Scholar
29. Grisham, MB, Hernandez, LA, Granger, DN. Xanthine oxidase and neutrophil infiltration in intestinal ischemia. Am J Physiol 1986. 251:G567-G574.
    View this article via: PubMed Google Scholar
30. Hillegass, LM, Griswold, DE, Brickson, B, Albrightson-Winslow, C. Assessment of myeloperoxidase activity in whole rat kidney. J Pharmacol Methods 1990. 24:285-295.
    View this article via: PubMed CrossRef Google Scholar
31. Fleming, TJ, Fleming, ML, Malek, TR. Selective expression of Ly-6G on myeloid lineage cells in mouse bone marrow. RB6-8C5 mAb to granulocyte-differentiation antigen (Gr-1) detects members of the Ly-6 family. J Immunol 1993. 151:2399-2408.
    View this article via: PubMed Google Scholar
32. Jung, U, Ley, K. Regulation of E-selectin, P-selectin, and intercellular adhesion molecule 1 expression in mouse cremaster muscle vasculature. Microcirculation 1997. 4:311-319.
    View this article via: PubMed CrossRef Google Scholar
33. Damiano, ER, Westheider, J, Tozeren, A, Ley, K. Variation in the velocity, deformation, and adhesion energy density of leukocytes rolling within venules. Circ Res 1996. 79:1122-1130.
    View this article via: PubMed Google Scholar
34. Lipowsky, HH, Scott, DA, Cartmell, JS. Leukocyte rolling velocity and its relation to leukocyte-endothelium adhesion and cell deformability. Am J Physiol 1996. 270:H1371-H1380.
    View this article via: PubMed Google Scholar
35. Walzog, B, Scharffetter-Kochanek, K, Gaehtgens, P. Impairment of neutrophil emigration in CD18-null mice. Am J Physiol 1999. 276:G1125-G1130.
    View this article via: PubMed Google Scholar
36. Mizgerd, JP, et al. Neutrophil emigration in the skin, lungs, and peritoneum: different requirements for CD11/CD18 revealed by CD18-deficient mice. J Exp Med 1997. 186:1357-1364.
    View this article via: PubMed CrossRef Google Scholar
37. Frenette, PS, Subbarao, S, Mazo, IB, von Andrian, UH, Wagner, DD. Endothelial selectins and vascular cell adhesion molecule-1 promote hematopoietic progenitor homing to bone marrow. Proc Natl Acad Sci USA 1998. 95:14423-14428.
    View this article via: PubMed CrossRef Google Scholar
38. Yang, JT, Rayburn, H, Hynes, RO. Cell adhesion events mediated by alpha 4 integrins are essential in placental and cardiac development. Development 1995. 121:549-560.
    View this article via: PubMed Google Scholar
39. Gurtner, GC, et al. Targeted disruption of the murine VCAM1 gene: essential role of VCAM-1 in chorioallantoic fusion and placentation. Genes Dev 1995. 9:1-14.
    View this article via: PubMed CrossRef Google Scholar
40. Stanley, E, et al. Granulocyte/macrophage colony-stimulating factor-deficient mice show no major perturbation of hematopoiesis but develop a characteristic pulmonary pathology. Proc Natl Acad Sci USA 1994. 91:5592-5596.
    View this article via: PubMed CrossRef Google Scholar
41. Lieschke, GJ, et al. Mice lacking granulocyte colony-stimulating factor have chronic neutropenia, granulocyte and macrophage progenitor cell deficiency, and impaired neutrophil mobilization. Blood 1994. 84:1737-1746.
    View this article via: PubMed Google Scholar
42. Metcalf, D. The granulocyte-macrophage regulators: reappraisal by gene inactivation. Exp Hematol 1995. 23:569-572.
    View this article via: PubMed Google Scholar
43. Pries, AR, Fritzsche, A, Ley, K, Gaehtgens, P. Redistribution of red blood cell flow in microcirculatory networks by hemodilution. Circ Res 1992. 70:1113-1121.
    View this article via: PubMed Google Scholar
44. Zweifach, B.W., and Lipowsky, H.H. 1984. Pressure-flow relations in blood and lymph microcirculation. In Handbook of physiology. The cardiovascular system. Microcirculation. E.M. Renkin and C.C. Michel, editors. American Physiological Society. Bethesda, Maryland, USA. 251–307.
    View this article via: PubMed Google Scholar