Conditional vascular cell adhesion molecule 1 deletion in mice: impaired lymphocyte migration to bone marrow

Abstract

We generated vascular cell adhesion molecule (VCAM)-1 "knock-in" mice and Cre recombinase transgenic mice to delete the VCAM-1 gene (vcam-1) in whole mice, thereby overcoming the embryonic lethality seen with conventional vcam-1-deficient mice. vcam-1 knock-in mice expressed normal levels of VCAM-1 but showed loss of VCAM-1 on endothelial and hematopoietic cells when interbred with a "TIE2Cre" transgene. Analysis of peripheral blood from conditional vcam-1-deficient mice revealed mild leukocytosis, including elevated immature B cell numbers. Conversely, the bone marrow (BM) had reduced immature B cell numbers, but normal numbers of pro-B cells. vcam-1-deficient mice also had reduced mature IgD+ B and T cells in BM and a greatly reduced capacity to support short-term migration of transferred B cells, CD4+ T cells, CD8+ T cells, and preactivated CD4+ T cells to the BM. Thus, we report an until now unappreciated dominant role for VCAM-1 in lymphocyte homing to BM.

Figure 1

Targeting strategy and conditional deletion of the vcam-1^flox allele. (A) VCAM-1 knock-in mice (bottom locus map) contain Cre recombinase sites of recombination (loxP sites; black arrowheads). The vcam-1 coding sequence exons are depicted as striped bars, with the 5′ untranslated region as a white bar. Probes A and B are shown above the top map line as black bars. Endonuclease sites shown are BamHI (B), EcoRI (R), HindIII (H), SphI (S), and XhoI (X). (B) The vcam-1^neo allele (neo) is shown along with two of the possible outcomes of Cre recombinase–mediated deletion, the vcam-1^flox allele (flox) and the vcam-1 ^Δ allele (Δ). (C and D) BamHI Southern blot analysis of tail DNA using probes A and B, respectively. Deletion of the vcam-1^neo allele to generate the vcam-1^flox allele (flox) and the vcam-1 ^Δ allele (Δ) in whole mice was achieved by interbreeding with the splicer transgene. (E) Southern blot analysis of mouse tail DNA as in D. A vcam-1^flox/Δ mouse (track 1) is shown alongside two TIE2Cre⁺ mice born of vcam-1^flox/flox parents, but with the TIE2Cre⁺ parent being the father (track 2) or mother (tracks 1 and 3). (F and G) Southern blot analysis as in D of genomic DNA from a vcam-1^flox/flox/TIE2Cre⁺ mouse and a vcam-1^flox/Δ/TIE2Cre⁺ mouse, respectively. Tissues analyzed were: BM; Br, brain (cortex); H, heart; Ki, kidney; Li, liver; Lu, lung; LNs; M, muscle (thigh); Pa, pancreas; and Sp, spleen.

Figure 2

Cre recombinase–mediated deletion in ECs and hematopoietic cells. (A) Lung sections are shown for a vcam-1 ^+/+ mouse (WT) and a vcam-1^flox/flox/TIE2Cre⁺ mouse (KO) after OVA sensitization and aerosol challenge. A representative blood vessel is shown in each panel (arrow); original magnification: ×500. VCAM-1 staining is seen on ECs of the vcam-1 ^+/+ mouse lung (red) but not on ECs of the vcam-1^flox/flox/TIE2Cre⁺ mouse lung. (B) VCAM-1 (open histograms) was assessed on BM B220⁻ myeloid lineage cells (i.e., excluding high side-scatter granuloid cells and low forward-scatter erythroid cells, spleen CD11c⁺CD8a⁻ myeloid DCs, and spleen CD11c⁺CD8a⁺ lymphoid DCs by fluorocytometry. Anti-CD8b.2 (clone 53-5.8) was used as an isotype control (shaded histograms). The small peak of CD8b.2⁺ cells among CD11c⁺CD8a⁺ splenocytes is expected, since it is known that a small fraction of CD8⁺ T cells are CD11c⁺.

Figure 2

Cre recombinase–mediated deletion in ECs and hematopoietic cells. (A) Lung sections are shown for a vcam-1 ^+/+ mouse (WT) and a vcam-1^flox/flox/TIE2Cre⁺ mouse (KO) after OVA sensitization and aerosol challenge. A representative blood vessel is shown in each panel (arrow); original magnification: ×500. VCAM-1 staining is seen on ECs of the vcam-1 ^+/+ mouse lung (red) but not on ECs of the vcam-1^flox/flox/TIE2Cre⁺ mouse lung. (B) VCAM-1 (open histograms) was assessed on BM B220⁻ myeloid lineage cells (i.e., excluding high side-scatter granuloid cells and low forward-scatter erythroid cells, spleen CD11c⁺CD8a⁻ myeloid DCs, and spleen CD11c⁺CD8a⁺ lymphoid DCs by fluorocytometry. Anti-CD8b.2 (clone 53-5.8) was used as an isotype control (shaded histograms). The small peak of CD8b.2⁺ cells among CD11c⁺CD8a⁺ splenocytes is expected, since it is known that a small fraction of CD8⁺ T cells are CD11c⁺.

Figure 3

The TIE2Cre transgene activates the ROSA26R reporter allele and results in β-galactosidase activity in brain (Br), kidney (Ki), liver (Li), axillary LNs (LN), spleen (Sp), and thymus (Th), revealed by blue X-Gal staining (original magnification: ×100). Brain sections were 30 μm thick while all others were 10 μm. By way of example, thick and thin arrows in the spleen section indicate some larger vessels and arterioles, respectively.

Figure 4

Peripheral blood analysis reveals leukocytosis in vcam-1^flox/Δ/TIE2Cre⁺ mice. A representative experiment (n = 5 per group) with age- and sex-matched vcam-1^flox/flox mice (black bars), vcam-1^flox/Δ/TIE2Cre⁻ mice (striped bars), and their vcam-1^flox/Δ/TIE2Cre⁺ littermates (white bars). Total leukocyte numbers were determined from EDTA-treated blood in Turk's solution and are shown in millions of cells per milliliter of blood. Differential counts were established by fluorocytometry. PMNs were defined on the basis of high side-scatter, with the other cells defined as mononuclear (mono). The numbers of CD4⁺ T cells, IgD^loIgM^hi immature B cells, and IgD⁺ B cells shown on the right are not in addition to the mononuclear cell counts but are part of the latter. Significant differences between vcam-1^flox/flox mice and vcam-1^flox/Δ/TIE2Cre⁺ mice are indicated (Student's t test, *P < 0.05; **P < 0.01).

Figure 5

Reduced lymphocytes in BM of vcam-1^{flox /} Δ/TIE2Cre⁺ mice. (A) Representative IgD versus IgM profile of B220⁺ cells in the BM of a vcam-1^flox/flox mouse and a vcam-1^flox/Δ/TIE2Cre⁺ mouse showing reduced IgD^loIgM^hi immature B cells and IgD⁺ B cells in the latter. (B and C) Absolute lymphocyte numbers in the BM of age- and sex-matched vcam-1^flox/flox mice (black bars) and vcam-1^flox/Δ/TIE2Cre⁺ mice (white bars), showing normal B220⁺CD43⁺ pro-B cell numbers but reduced numbers of other lymphocyte types. Significant differences between the two groups (n = 4 each) are indicated (Student's t test, *P < 0.02; **P < 0.01; ***P < 0.002).

Figure 6

Relatively normal B cell responses in vcam-1^flox/Δ/TIE2Cre⁺ mice. Serial spleen sections from a vcam-1^flox/flox mouse (WT) and a vcam-1^flox/Δ/TIE2Cre⁺ mouse (KO) 10 d after intraperitoneal challenge with NP₁₃CG adsorbed to alum (original magnifications: ×100). All sections were stained with anti-IgD antibody (seen in brown) and either PNA, anti-CD8b, or anti–VCAM-1. PNA, CD8b, and VCAM-1 staining are in purple. Arrows in the KO section indicate typical VCAM-1 staining on FDCs within GCs.

Figure 7

Reduced short-term lymphocyte migration to BM in vcam-1^flox/Δ/TIE2Cre⁺ mice. A representative experiment with age- and sex-matched vcam-1^flox/flox recipients (black bars) and vcam-1^flox/Δ/TIE2Cre⁺ recipients (white bars). The frequency of CMFDA-labeled cells (IgD⁺, CD4⁺, and CD8a⁺) was used to determine the absolute numbers of each type of CMFDA-labeled cell in each lymphoid organ. These data were then represented as the percentage of homed cells, where 100% is the total number of recovered CMFDA-labeled cells in each individual mouse (n = 3 per group). *Significantly different from controls (Student's t test, P < 0.02). The slightly elevated migration to PLNs of vcam-1^flox/Δ/TIE2Cre⁺ recipients was not statistically significant.

Figure 8

Impaired B cell recirculation to BM results in elevated peripheral blood levels. Age- and sex-matched vcam-1^flox/flox recipients (black bars) and vcam-1^flox/Δ/TIE2Cre⁺ recipients (white bars) received CMFDA-labeled splenocytes from tcrα^{− /}− donors, and recipient lymphoid organs and blood were harvested 2 h later (n = 4 per group). The IgD versus IgM profile of CMFDA-labeled cells was determined as in the legend to Fig. 5 A, and the frequency of CMFDA-labeled cells was used to determine the absolute numbers of each type of CMFDA-labeled cell in each lymphoid organ. Significant differences are indicated (Student's t test, *P < 0.01; **P < 0.005; ***P < 0.001).

Figure 9

Reduced short-term migration by preactivated CD4⁺ T cells. Age- and sex-matched vcam-1^flox/flox recipients (black bars) and vcam-1^flox/Δ/TIE2Cre⁺ recipients (white bars) received CMFDA-labeled preactivated/experienced T cells as described in Materials and Methods, and recipient lymphoid organs and blood were harvested 2 h later (n = 4 per group). The CD4 versus CD62L profile of CMFDA-labeled cells was determined by fluorocytometry, and the frequency of CMFDA-labeled cells was used to determine the absolute numbers of each type of CMFDA-labeled cell in each lymphoid organ (note the differences in scale). Significant differences are indicated (Student's t test, *P < 0.002).