Reduced Leukocyte Migration, but Normal Rolling and Arrest, in Interleukin-8 Receptor Homologue Knockout Mice

Abstract

purpose. To determine the role of the murine interleukin-8 receptor homologue (mIL-8Rh, neutrophil chemokine CXC receptor 2) in leukocyte migration using intravital microscopy in a standardized model of eye inflammation, endotoxin-induced uveitis (EIU).

methods. Two hundred fifty nanograms of E. coli endotoxin was injected into the vitreous of knockout mIL-8Rh^−/− (n = 7) mice or heterozygous littermate mIL-8Rh^+/− controls (n = 7). Intravital microscopic examination of iris microvasculature was performed at baseline and 6 and 24 hours after endotoxin injection. The numbers of rolling (cells/mm² endothelial surface/min), sticking (cells/mm² endothelial surface), and infiltrating cells (cells/mm² iris tissue) were evaluated by digital off-line quantification.

results. The number of infiltrating cells was significantly reduced in mIL-8Rh^−/− mice: 406 ± 77 cells/mm² at 6 hours and 242 ± 50 cells/mm² at 24 hours in mIL-8Rh^+/− mice versus 14 ± 4 cells/mm² at 6 hours and 38 ± 11 cells/mm² at 24 hours in mIL-8Rh^−/− mice (P < 0.001). In contrast, the absence of the IL-8 receptor homologue did not reduce rolling or sticking.

conclusions. Iris rhodamine angiography allows precise quantification of leukocyte–endothelial dynamics in the absence of surgical trauma. IL-8 and its homologues are known to be potent signals for leukocyte migration. Although IL-8 has previously been implicated in cell adhesion, video imaging in vivo demonstrated that deletion of the IL-8 receptor homologue had minimal effect on rolling or arrest in this model of inflammation.

Research on chemokines has provided considerable insight into the mechanism of diapedesis. Generation of attractant chemokines by endothelial and inflammatory cells has been recognized as a fundamental process. Lipopolysaccharide (LPS)-stimulated endothelial cells can synthesize a number of factors capable of affecting neutrophil adhesion or motility. These include platelet-activating factor, interleukin (IL)-6, granulocyte-macrophage colony-stimulating factor, and melanoma-growth stimulating activity. ¹ Recent investigations in animal models using either blocking antibodies against IL-8 or disruption of the gene encoding an IL-8 receptor have revealed the involvement of IL-8 and related chemokines in the recruitment of neutrophils and in neutrophil-associated tissue injury in acute inflammation. ^{2 3}  

Recruitment is achieved through a multistep paradigm that includes margination, selectin-mediated rolling,β ₂-integrin–mediated firm adhesion, and migration of leukocytes into the site of inflammation. ⁴ Binding of neutrophils to inflamed endothelium initiates an orchestrated series of events in which neutrophils shed L-selectin and engage a second set of adhesion-promoting glycoproteins, the leukocyteβ ₂-integrins. ^{5 6}  

Human IL-8 receptors are members of a family of G protein–coupled receptors and are abundantly expressed on neutrophils, the primary target for CXC chemokines. There are at least two different IL-8 receptor types with 77% identical amino acid sequences. The type-1 receptor (IL-8R1, CXCR1) specifically binds IL-8 and granulocyte-chemoattractant protein (GCP)-2 (reviewed by Baggiolini and co-workers. ^{2 7} ) The type-2 receptor (IL-8R2, CXCR2) also binds additional neutrophil-attracting CXC chemokines: MGSA (melanoma growth stimulatory activity), growth related oncogene (GRO), MIP (macrophage inflammatory protein)-2, and NAP (neutrophil-activating protein)-2. IL-8 receptors are also found on monocytes, basophils, and eosinophils, but the responses of these cells to IL-8 are much weaker than those of neutrophils. ² CXCR1 and CXCR2 are present in similar numbers on all neutrophils and monocytes but only on a minority of lymphocytes. 

No true homologue for human IL-8 has been found in mice. Mouse KC has been accepted as the closest murine homologue of human GRO-α. In analogy, the KC receptor is an IL-8 type-2 receptor homologue capable of binding KC, MIP-2, and GCP-2/liposaccharide-induced CXC chemokine (LIX) with high affinity. Each of these ligands may activate mouse neutrophils. It has been shown that KC is biologically active on human neutrophils and competes with ¹²⁵I–IL-8 binding to IL-8R2 but not IL-8R1. ^{8 9}  

So far, only one receptor for this family of chemokines (mIL-8Rh, KC-receptor) has been described in the mouse. ⁸ The mIL-8Rh binds at least three chemokines that activate neutrophils (KC, MIP-2, and GCP-2/LIX) and is structurally and functionally homologous to CXCR2. 

Although IL-8 and its homologues are clearly involved in leukocyte migration, the role of the IL-8 family of chemokines is less clearly understood with regard to leukocyte rolling and arrest. Experiments in vitro suggest that IL-8 released from stimulated endothelial cells facilitates the transendothelial migration of neutrophils ¹ by both regulating L-selectin and β₂-integrin expression and forming a transendothelial cell chemotactic gradient. In contrast, IL-8 does not affect the expression of intercellular adhesion molecule (ICAM)-1 or E-selectin on the endothelial cell surface. ¹ IL-8 activation of neutrophils in the microvascular compartment results in the concomitant shedding of L-selectin and the expression of theβ ₂-integrin CD11b/CD18 on the surface of the neutrophil, which leads to firm ICAM-1–dependent intravascular adhesion of neutrophils to the endothelium. Transgenic mice overexpressing the human gene for IL-8 have downregulated the expression of L-selectin. ¹⁰ Evidence for theβ ₂-integrin pathway has been given by a complete inhibition of IL-8–induced neutrophil transmigration by monoclonal antibodies to Mac-1. ¹¹  

Because IL-8 regulates expression of both P-selectin andβ ₂-integrins, molecules implicated in rolling and arrest, respectively, ⁴ it seems logical to hypothesize that both rolling and arrest would be impaired in mIL-8Rh–deficient mice. Testing this hypothesis requires an experimental system that can measure dynamic interactions between cells. We are aware of only two such studies. One of them implicated IL-8 in arrest of isolated monocytes rolling on cultured human umbilical cord endothelial cells expressing E-selectin that had been transduced with an adenovirus vector. ¹² A brief communication reported that the absence of the IL-8 receptor homologue resulted in reduced arrest in response to MIP-2 in cremasteric vessels, whereas the adhesion response to fMLP (formylmethionylleucylphenylalanine) did not differ between knockouts and controls. ¹³ Neither of these studies correlated the numbers of free-flowing, rolling, adherent, and infiltrating cells. To clarify the importance of mIL-8Rh in leukocyte rolling, sticking, and diapedesis, we studied leukocyte migration using intravital microscopy of iris microvasculature in receptor-knockout mice. 

Methods

Animals

Breeding pairs of BALB/c mice with a null mutation in the mIL-8Rh gene were a generous gift from Genentech (South San Francisco, CA). It has been reported that animals that lack the mIL-8R have a 12-fold increase of circulating neutrophils. ¹⁴ The mIL-8Rh^−/− mice we used for the present study, however, had only 2.6 times more circulating leukocytes than mIL-8Rh^+/− mice, primarily due to a 6-fold increase in the number of neutrophils. ¹⁵ Genotyping was performed by polymerase chain reaction analysis of tail DNA with the use of primer pairs specific for the natural receptor or chemokine genes and for the neomycin-receptor recombinant genes. Heterozygous mIL-8Rh^+/− and homozygous mIL-8Rh^−/− mice were mated to obtain both knockout and littermate control animals. They were fed standard laboratory chow, and sterile water was supplied ad libitum. All experiments were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Microscopy

The technique of digital live-tissue video microscopy of iris vessels in mice during endotoxin-induced uveitis has been described in detail elsewhere. ¹⁶ Therefore, we give here only a brief description. 

To visualize leukocytes, animals were intravenously injected with rhodamine 6G 10 minutes before experiments (35 mg/kg; Sigma Chemical, St. Louis, MO). Animals were anesthetized with isoflurane 1.7% in O₂ (1 l/min). The microcirculation of iris vessels was recorded with digital video microscopy at three different times (baseline and 6 and 24 hours after induction of inflammation) with an epifluorescence intravital microscope (modified Orthoplan; Leica, Wetzlar, Germany). Iris vessels were selected as described in detail elsewhere. ¹⁶ Only venules in the right eye of knockout (n = number of animals/total number of venules = 7/57, respectively) and heterozygous control animals (n = 7/53, respectively) were examined. Diameter and length of each vessel segment (in micrometers) and leukocyte behavior (free-flowing, rolling, sticking, and infiltrating) were analyzed off-line with image-analysis software. Free-flowing cells are expressed as number of cells passing a vessel segment per minute. Infiltrating cells are given as number of cells per square millimeter of iris tissue. For the calculation of rolling and sticking leukocytes, a cylindrical tube was assumed from the diameter and the length of each vessel segment. The number of rolling cells was given as number of cells per square millimeter of endothelial surface per minute. The number of sticking cells was given as number of cells per square millimeter of endothelial surface. 

Data Analysis

To describe leukocyte recruitment quantitatively and to consider the different sizes of the circulating leukocyte pool in homozygous and heterozygous animals, we normalized the above-mentioned parameters according to a recent proposal from Jung and coworkers. ¹⁷ Rolling efficiency is equivalent to rolling leukocyte flux (percent of rolling cells of all free-flowing and rolling cells). The adhesion efficiency is calculated as the number of sticking cells per square millimeter divided by the total number of rolling and sticking cells per square millimeter per minute. This parameter relates the number of sticking cells to the rolling leukocyte pool available in the same vessel. Probability values were calculated using a Mann–Whitney rank sum test. 

Induction of Inflammation

Uveitis was induced by intravitreal injection of 2 μl saline containing 250 ng LPS from Escherichia coli 055:B5 (List Biological Laboratories, Campbell, CA) and 0.25% human serum albumin (Baxter Healthcare, Glendale, CA). ¹⁵  

Wholemount Iris Immunohistology

Immunostaining of wholemount iris preparations has been described in detail elsewhere. ¹⁶ Irises were dissected and cut into 3 pieces. All pieces were stained with anti-platelet-endothelial cell adhesion molecule (PECAM) antibody to visualize vessels. In addition, 1 piece was labeled with anti-CD45 (leukocyte common antigen; PharMingen, San Diego, CA) to detect all leukocytes in the iris. 

Availability of Video Images

Because leukocyte rolling is a dynamic variable, this parameter is best appreciated with video rather than static photography. Video images related to this study can be viewed by visiting our web site at http://www.ohsu.edu/cei/Iris/Iris-Homepage.html. 

Results

Leukocyte Rolling

Absolute numbers of rolling leukocytes in mIL-8Rh^−/− mice increased from 94 ± 51 cells/mm² per minute at baseline to 296 ± 58 cells/mm² per minute at 6 hours and 441 ± 90 cells/mm² per minute at 24 hours after LPS injection (Fig. 1) . In contrast, rolling cells in control animals remained on a similar level (81 ± 37 cells/mm² per minute at baseline, 144 ± 40 cells/mm² per minute at 6 hours, and 164 ± 47 cells/mm² per minute at 24 hours). The increase in rolling cells in mIL-8Rh^−/− mice was significant at 24 hours (P = 0.01). 

Because mIL-8Rh^−/− mice have approximately 2.6-fold more circulating leukocytes than mIL-8Rh^+/− controls, ¹⁵ we normalized the number of rolling cells to the number of circulating cells and calculated a rolling efficiency. As shown in Figure 2 , rolling efficiencies at baseline were 8% ± 3% in mIL-8Rh^−/− mice and 4% ± 3% in mIL-8Rh^+/− mice, at 6 hours were 43% ± 4% in mIL-8Rh^−/− mice and 48% ± 7% in mIL-8Rh^+/− mice, and at 24 hours were 31% ± 4% in mIL-8Rh^−/− mice and 31% ± 6% in mIL-8Rh^+/− mice. There was no significant difference between the two groups. 

Leukocyte Sticking

Absolute numbers of sticking leukocytes showed no statistically significant differences between the groups (Fig. 3) . The numbers at baseline were 8 ± 4 cells/mm² in mIL-8Rh^−/− mice and 16 ± 7 cells/mm² in mIL-8Rh^+/− mice. At 6 hours they had increased to 237 ± 50 cells/mm² in knockouts and 235 ± 44 cells/mm² in controls. At 24 hours, the number of sticking leukocytes was down to 83 ± 21 cells/mm² in knockouts and 110 ± 32 cells/mm² in control animals. 

Adhesion efficiency normalized to the total number of rolling and sticking cells showed a tendency for a lower percentage of the rolling cells to stick in mIL-8Rh^−/− mice (47% ± 5% in knockouts versus 57% ± 6% in controls at 6 hours, 26% ± 4% in knockouts versus 36% ± 6% in controls), but this difference was not statistically significant (Fig. 4) . 

Infiltrating Leukocytes

The number of infiltrating cells was significantly reduced in mIL-8Rh^−/− mice at 6 and 24 hours as determined by in vivo evaluation (P < 0.001). The numbers of infiltrating cells were suppressed to14 ± 4 cells/mm² of iris tissue in mIL-8Rh^−/− mice, whereas heterozygous control animals showed a normal increase, to 406 ± 77 cells/mm² at 6 hours. Twenty-four hours after LPS injection the number of infiltrating cells was still reduced to 38 ± 11 cells/mm² in knockouts compared with 242 ± 50 cells/mm² in control animals (Fig. 5) . 

Infiltrating cell numbers determined by wholemount immunostaining (Figs. 6A at baseline, 6C at 6 hours after LPS injection) are qualitatively comparable to the in vivo results (Figs. 6B at baseline, 6D at 6 hours after LPS injection). 

Discussion

Our data indicate that the absence of the mouse IL-8 receptor homologue profoundly affects transmigration without a comparable impact on the number of rolling or arresting cells. The number of rolling cells is increased in mIL-8Rh–deficient mice, but this apparent increase disappears if one corrects for the number of circulating white cells. A limitation of our study is that we made no attempt to distinguish monocyte from neutrophil migration. However, in the LPS model, the neutrophil is the predominant migratory cell at the times studied. In our previous studies with mIL-8Rh knockout mice, we demonstrated with histology that this receptor affected neutrophil migration in uveitis induced by intravitreal LPS. ¹⁵ Our data differ from the conclusions of Gerszten and colleagues, ¹² who implicated IL-8 in monocyte adhesion. The difference may reflect the predominant leukocyte type that was studied. Morgan and colleagues ¹³ showed reduced arrest in IL-8Rh knockouts only if a ligand specific for the IL-8Rh was used as a stimulus. Endotoxin, like fMLP in the Morgan study, resulted in comparable arrest in knockouts and controls. Unlike prior studies, our technique allows rolling, arrest, and migration to be quantitated and correlated. The absence of the IL-8Rh inhibits migration independently of its effect on rolling or arrest. 

We believe that this is the first report describing the use of intravital microscopy of the eye in a knockout animal. The iris offers a rare opportunity for studying leukocyte adhesion and migration physiologically, without artifacts from surgical trauma. The transparency of the cornea and the absence of pigmentation in the iris of albino mice enable visualization of infiltrating leukocytes at the light-microscopic level. Additionally, the inherent thinness of the mouse iris facilitates a very precise quantification of cells that have migrated across the vascular endothelium into the tissue stroma. The infiltrating cells reside essentially in only two dimensions before they leave the iris and enter the aqueous humor. This allows infiltrating leukocytes to be counted and expressed as numbers of cells per square millimeter of iris tissue. Digitalization of the images facilitates quantification and enhances image quality. Traditional histologic methods cannot quantitate rolling or allow a study of the same vessel over multiple time points. 

Our observations clearly demonstrate that mice lacking the IL-8 receptor homologue have a severe impairment in leukocyte migration without a comparable defect in rolling or arrest. Our study shows the advantage of dynamic imaging of the immune response in vivo compared with histology as a means to study the stepwise process of inflammation. On the basis of our study, we predict that the clinical utility of an IL-8 receptor antagonist would be excellent for blocking neutrophil accumulation outside a vessel as in inflammation, but the same antagonist would be ineffective in blocking neutrophil arrest as might be the goal in limiting tissue damage due to ischemia. 

In a complex milieu induced by LPS, signals other than those that activate CXCR2 are sufficient to support leukocyte adhesion in vivo. However, these signals appear insufficient for subsequent extravasation into surrounding tissue. 

 Supported by a research grant from Deutsche Forschungsgemeinschaft, Germany Be 1795/3-1 (MDB); NIH Grants EY06484 and EY06477; and grants from Research to Prevent Blindness (JTR, SRP); and the Casey Eye Institute.

 Submitted for publication December 29, 1999; accepted February 10, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Rolling cells in iris venules expressed as the absolute number corrected for vessel diameter (mean ± SE). More rolling cells in mIL-8Rh^−/− (*P = 0.01 at 24 hours).“− /−” means mIL-8Rh^−/−, “+/−” means mIL-8Rh^+/−.

Figure 2.

 

Rolling efficiency (number of rolling cells related to all rolling and free-floating cells in a vessel). No difference was found between the groups. “−/−” means mIL-8Rh^−/−, “+/−” means mIL-8Rh^+/−.

Figure 3.

 

Sticking leukocytes in iris venules expressed as absolute vessel diameter-corrected numbers (mean ± SE). No difference exists between both groups. “−/−” means mIL-8Rh^−/−,“+ /−” means mIL-8Rh^+/−.

Figure 4.

 

Adhesion efficiency (number of sticking leukocytes related to all rolling and sticking cells in a vessel). The tendency of a lower percentage of rolling leukocytes to stick in mIL-8Rh^−/− mice was not statistically significant. “−/−” means mIL-8Rh^−/−, “+/−” means mIL-8Rh^+/−.

Figure 5.

 

In vivo visualization of infiltrating leukocytes revealed a significant difference between the groups (***P < 0.001 at 6 and 24 hours after endotoxin injection). “−/−” means mIL-8Rh^−/−, “+/−” means mIL-8Rh^+/−.

Figure 6.

 

Infiltrating cells as determined by immunohistology on wholemount iris tissue (A, C) and in vivo visualization (B, D). Leukocytes and vessels are stained with anti-CD45 and anti–PECAM-1 antibodies in (A) and (C). (A) and (B) represent baseline situation; (C) and (D) are from 6 hours after endotoxin injection. Scale bars, 200 μm.