Monocytosis is a unique cellular abnormality associated with the Yaa (Y-linked autoimmune acceleration) mutation. The present study was designed to define the cellular mechanism responsible for the development of monocytosis and to characterize the effect of the Yaa mutation on the development of monocyte subsets.
We produced bone marrow chimeras reconstituted with a mixture of Yaa and non-Yaa bone marrow cells bearing distinct Ly-17 alloantigens, and determined whether monocytes of Yaa origin became dominant. Moreover, we defined the 2 major inflammatory (Gr-1+,CD62 ligand [CD62L]+) and resident (Gr-1−,CD62L−) subsets of blood monocytes in aged BXSB Yaa male mice, as compared with BXSB male mice lacking the Yaa mutation.
Analysis of the Ly17 allotype of blood monocytes in chimeric mice revealed that monocytes of both Yaa and non-Yaa origin were similarly involved in monocytosis. Significantly, the development of monocytosis paralleled a selective expansion of the resident monocyte subset compared with the inflammatory subset, and the former expressed CD11c, a marker of dendritic cells. Neither monocytosis nor the change in monocyte subpopulations, including CD11c expression, was observed in Yaa-bearing C57BL/6 mice, in which systemic lupus erythematosus (SLE) fails to develop.
Our results suggest that Yaa-associated monocytosis is not attributable to an intrinsic abnormality in the growth potential of monocyte lineage cells bearing the Yaa mutation and that the Yaa mutation could lead to the expansion of dendritic cells, thereby contributing to the accelerated development of SLE.
In the BXSB strain of mice, an autoimmune syndrome with features of systemic lupus erythematosus (SLE) develops spontaneously, and male mice are affected much earlier than female mice (1). The accelerated development of SLE in male BXSB mice is attributable to the presence of an as-yet-unidentified mutant gene located on the Y chromosome, designated Yaa (Y-linked autoimmune acceleration) (2). The Yaa gene by itself is unable to induce significant autoimmune responses in mice without an apparent SLE background, but in combination with autosomal susceptibility alleles that are present in lupus-prone mice, it can induce and accelerate the development of SLE (3, 4).
Monocytosis is a unique cellular abnormality associated with the Yaa mutation (5). In peripheral blood mononuclear cells (PBMCs) from 8-month-old male BXSB Yaa mice, the frequency of monocytes reaches >50%. The development of monocytosis is apparently dependent on the progression of SLE because monocytosis was not observed in BXSB.ll (ll for long-lived) and BXSB.H2d mice, both of which fail to develop SLE (6, 7). However, such a monocytosis is not a common feature of 2 lupus-prone strains of mice ([NZB × NZW]F1 and MRL-Faslpr), indicating that this abnormality is causally linked to the Yaa mutation.
Monocytes newly generated in the bone marrow circulate for a few days before entering tissues, where they differentiate into mature resident macrophages (8). Under conditions of inflammation, monocyte production in the bone marrow is increased, and the cells are rapidly recruited to sites of inflammation, where they differentiate into inflammatory macrophages (9). Recent studies in mice demonstrated that circulating monocytes could be divided into 2 phenotypically and functionally distinct subsets (10, 11). The first subset of monocytes, which is classified as “inflammatory” and is characterized by a Gr-1+,CX3CR1low,CCR2+,CD62 ligand (CD62L)+ phenotype, is preferentially recruited to inflamed tissue. The second subset of monocytes, Gr-1−,CX3CR1high, CCR2−,CD62L−, is considered to be a source of tissue resident macrophages and dendritic cells (DCs), and is classified as “resident.” These 2 subsets correspond, respectively, to the CCR2+,CD16− and CCR2−,CD16+ monocyte populations described in humans (12). Because blood monocytes are the major source of DCs, it is important to determine the effect of the Yaa mutation on development of the 2 different subsets of monocytes in BXSB Yaa male mice in relation to acceleration of the disease.
In the present study, we first analyzed whether the development of monocytosis is a defect intrinsic to monocyte lineage cells bearing the Yaa mutation or is associated with excessive production of a monocyte-specific growth factors(s) in Yaa-bearing mice. To address this question, we constructed radiation bone marrow chimeric mice reconstituted with a mixture of Yaa and non-Yaa bone marrow cells bearing distinct Ly17 alloantigens, and determined whether monocytes of Yaa origin became dominant. Second, we defined the 2 major subsets of blood monocytes in aged male BXSB Yaa mice as compared with male BXSB mice lacking the Yaa mutation. Our results demonstrated that monocytes of both Yaa and non-Yaa origin were similarly involved in the development of monocytosis. In addition, a selective expansion of resident monocytes expressing CD11c, a marker of DCs, suggests that the Yaa mutation may lead to excessive production of DCs, thereby contributing to acceleration of lupus autoimmune responses.
MATERIALS AND METHODS
BXSB and NZB (Ly-17.1) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Accelerated SLE develops in male BXSB mice, but not in female BXSB mice, partly because of the action of the Yaa gene (1). BXSB mice lacking the Yaa gene, C57BL/6 (B6; Ly17.2) mice bearing the Yaa mutation (B6.Yaa), and B6.NZB-Nba2 congenic mice homozygous for the Nba2 (New Zealand black autoimmunity 2) lupus-susceptibility locus and bearing the Ly17.1 allele (B6.Ly17.1) were generated as described previously (4, 13, 14). B6.Ly17.1 mice bearing the Yaa mutation were then generated by intercrossing B6.Ly17.1 females and B6.Yaa males. (NZB × B6.Ly17.1.Yaa)F1 (Ly17.1/Ly17.1) and (NZB × B6)F1 (Ly17.1/Ly17.2) mice were bred in the animal facility of the Centre Médical Universitaire, Geneva.
Flow cytometric analysis.
Flow cytometry was performed using 2- or 3-color staining of peripheral blood mononuclear cells (PBMCs) and analyzed with a FACSCalibur cytometer (BD Biosciences, San Jose, CA). The following antibodies were used: anti-F4/80, anti-CD11b (M1/70), anti–Ly-6C/G (Gr-1), anti-CD62L (MEL-14), anti–Ly-17.2 (K9.361), anti-CD11c (N418), anti–I-A (Y-3P), anti-CD80 (1G10), anti-CD86 (GL1), anti-CD4 (GK1.5), anti-CD8α (5H10-1), anti-B220 (RA3-6B2), anti-CD45RA (14.8), and anti-CD19 (1D3) monoclonal antibodies.
Serum levels of IgG anti-DNA autoantibodies were determined by enzyme-linked immunosorbent assay, with results expressed as titration units (units/ml) in reference to a standard curve obtained from lupus-prone MRL-Faslpr mice, as described previously (15).
In vivo monocyte proliferation assay.
Mice received 6 mg of bromodeoxyuridine (BrdU; Sigma-Aldrich, St. Louis, MO) intravenously. Two days later, peripheral blood monocytes were stained for CD11b, F4/80, and Gr-1 and then for BrdU incorporation (BrdU Flow kit; BD Biosciences).
Preparation of double bone marrow chimeras.
Three-to-four–month-old (NZB × B6)F1 recipient mice were irradiated (850 rads) and reconstituted with a mixture of bone marrow cells (5 × 106 from each donor) from 3-to-4–month-old male Ly17.1/Ly17.1 (NZB × B6.Ly17.1.Yaa)F1Yaa mice and male Ly17.1/Ly17.2 (NZB × B6)F1 non-Yaa mice or, as a control, a mixture of bone marrow cells from male Ly17.1/Ly17.1 (NZB × B6.Ly17.1)F1 and Ly17.1/Ly17.2 (NZB × B6)F1 non-Yaa mice, as described previously (16). Two months later, chimerism in recipients was assessed by staining peripheral blood B cells with anti–Ly-17.2 (K9.361) and anti-B220 (RA3-6B2) monoclonal antibodies.
Statistical analysis was performed with Wilcoxon's 2-sample test. P values less than or equal to 5% were considered significant.
Age-dependent development of monocytosis in lupus-prone male (NZB × B6. Yaa)F1 and BXSB Yaa mice, but not in male B6.Yaa mice.
As described previously (5), 8-month-old male BXSB Yaa mice had an increased percentage of monocytes in blood, while this age-dependent monocytosis was not observed in male BXSB non-Yaa mice, which are unable to develop a lupus-like autoimmune syndrome (13) (Table 1). Because the Yaa mutation induces severe SLE in male (NZB × B6.Yaa)F1 mice, with 50% mortality at 14 months of age (17), but fails to induce significant lupus-like autoimmune disease in nonautoimmune B6 mice (4), the development of monocytosis was assessed in these mice. Eight-month-old male (NZB × B6.Yaa)F1 mice displayed a significant monocytosis, with 26–52% of PBMCs being monocytes (P < 0.001) (Table 1). Monocytosis did not develop in male (NZB × B6)F1 mice lacking the Yaa mutation, and these mice also failed to develop disease. Notably, the absolute number of monocytes increased in male (NZB × B6.Yaa)F1 mice in which monocytosis developed (data not shown), as was observed in male BXSB Yaa mice (5). In contrast, CD11b+,F4/80+ monocytes in 8-month-old male B6.Yaa mice represented not more than 20% of PBMCs, and this percentage was comparable with that in control male B6 mice. These results thus confirmed a close association of monocytosis with the Yaa-mediated lupus-like autoimmune disease.
Table 1. Development of monocytosis, anti-DNA autoantibodies, and GN in lupus-prone male (NZB × B6.Yaa)F1 and BXSB mice bearing the Yaa mutation*
50% mortality due to GN
Values are the mean ± SD of 8–20 mice in each group. The percentage of CD11b+,F4/80+ monocytosis in peripheral blood mononuclear cells at 8 months of age was determined by flow cytometry. At 2 months of age, the percentage of blood monocytes in different groups of mice tested in the present study was ∼10% (data not shown). Serum levels of IgG anti-DNA, expressed as units/ml, at 8 months of age were determined by enzyme-linked immunosorbent assay. GN = glomerulonephritis.
46.9 ± 8.9
67 ± 36
14.4 ± 4.6
21 ± 9
(NZB × B6)F1
39.0 ± 8.1
81 ± 50
(NZB × B6)F1
14.7 ± 3.1
20 ± 8
14.0 ± 4.4
11 ± 8
10.3 ± 3.0
3 ± 2
Increased production of blood monocytes in aged male BXSB Yaa mice.
To investigate whether the development of monocytosis in aged male BXSB Yaa mice was indeed the result of increased generation of monocytes in the bone marrow rather than the result of abnormal persistence in the circulating blood, 8-month-old male BXSB mice bearing or lacking the Yaa mutation were treated intravenously with BrdU. Flow cytometric analysis of blood monocytes 2 days after BrdU injection revealed that the percentages of BrdU+,CD11b+ monocytes in PBMCs from male BXSB Yaa mice (mean ± SD 6.7 ± 1.7%) were ∼5-fold higher than those in PBMCs from male non-Yaa BXSB mice (mean ± SD 1.3 ± 0.4%; P < 0.001). Notably, the percentage of BrdU+ cells among CD11b+ monocytes was comparable in male BXSB Yaa mice (mean ± SD 10.6 ± 1.9%) and their non-Yaa counterparts (10.3 ± 3.8%). These results indicate that monocytosis observed in aged male BXSB Yaa mice is caused by an increased production of monocytes in bone marrow and their subsequent release into the circulation, and not by an aberrant accumulation of monocytes in the circulating blood due to differences in migration into tissues.
Comparable involvement of monocytes of both Yaa and non-Yaa origin in the development of monocytosis.
The increased production of monocytes in male BXSB Yaa mice could be attributable either to excessive production of monocyte-specific growth factor(s) or to hyperresponsiveness of monocyte lineage cells to normal levels of growth factors. To address this question, irradiated (NZB × B6)F1 mice were reconstituted with a mixture of bone marrow cells from Ly17.1 homozygous male (NZB × B6.Ly17.1.Yaa)F1Yaa mice and Ly17.1/Ly17.2 heterozygous male (NZB × B6)F1 non-Yaa mice, in which Ly17.1 and Ly17.2 are allelic markers of CD32. The enumeration of Ly-17.2+ circulating B cells (defined by B220 staining) 2 months after reconstitution showed that the percentage of Ly-17.2+ B cells (mean ± SD 22.7 ± 3.0% [n = 12 mice]) was comparable with the percentage of Ly-17.2− B cells (25.8 ± 4.1%), confirming an equal reconstitution of hematopoietic cells derived from both Yaa and non-Yaa origin (Figure 1A). Notably, similar results were obtained with control chimeras, in which irradiated F1 mice were reconstituted with a mixture of bone marrow cells from male (NZB × B6.Ly17.1)F1 and (NZB × B6)F1 mice, both of which lack the Yaa mutation (for Ly17.2+ B cells, mean ± SD 24.3 ± 5.4%; for Ly17.2− B cells, 26.2 ± 4.1% [n = 7 mice]).
In Yaa plus non-Yaa (B6 × NZB)F1 bone marrow chimeras, increased percentages of monocytes developed 8 months after reconstitution (mean ± SD 29.0 ± 7.2% [n = 12 mice]) as compared with control non-Yaa chimeras (mean ± SD 10.3 ± 2.9% [n = 7 mice]; P < 0.001) (Figure 1B). Analysis of the Ly-17 alloantigen on F4/80+ blood monocytes revealed that the percentage of Ly17.2− monocytes of Yaa origin (mean ± SD 15.7 ± 4.9%) was comparable with the percentage of Ly-17.2+ monocytes of non-Yaa origin (14.0 ± 26%) (Figure 1B). Notably, in control chimeras reconstituted with a mixture of non-YaaLy17.1 plus Ly17.1/Ly17.2 donor cells, the respective monocyte populations in circulating blood were of equal size (for Ly17.2+ monocytes, mean ± SD 4.4 ± 1.3%; for Ly17.2− monocytes, 5.4 ± 2.1%). These results indicate no selective production of monocytes from the Yaa-bearing bone marrow cells, thus providing evidence against selective hyperresponsiveness to growth factors of Yaa-bearing monocyte lineage cells.
Selective expansion of the resident monocyte subset in aged male BXSB Yaa mice but not in male B6 Yaa mice.
Recent studies have demonstrated the existence of 2 major subpopulations of blood monocytes, resident and inflammatory monocytes, which apparently represent 2 phenotypically and functionally distinct subsets (10, 11). To determine whether the Yaa mutation could exhibit an effect on the development of these 2 subsets, the expression of Gr-1 on blood monocytes from male BXSB Yaa and non-Yaa mice was analyzed by flow cytometry. In 2-month-old BXSB mice, independent of the Yaa genotype, the percentage of cells expressing the Gr-1+,CD11b+ monocyte phenotype in the circulation was almost the same as the percentage of cells expressing the Gr-1−,CD11b+ monocyte subset (Figure 2A and Table 2). Similar percentages of GR-1+ and Gr-1− monocytes were present in nonautoimmune B6 mice. However, in 8-month-old male BXSB Yaa mice developing monocytosis, the Gr-1− monocyte subset selectively increased (∼6-fold more than the Gr-1+ subset; P < 0.001) and became the dominant monocyte population in the circulating blood. Notably, this population also lacked expression of CD62L (Figure 2B), indicating that it belonged to the resident monocyte subset, while Gr-1+ monocytes were CD62L+, corresponding to the inflammatory monocyte subset, as described by Geissmann et al (10). In contrast, the balance of these 2 subsets did not change in aged male BXSB mice lacking the Yaa mutation or in aged male B6.Yaa mice (Figure 2A).
Table 2. Selective expansion of Gr-1−,CD11b+ monocytes in aged male BXSB Yaa, but not B6.Yaa, mice*
Percentage of Gr-1+ and Gr-1–,CD11b+ monocytes in peripheral blood mononuclear cells from male BXSB and B6 mice (mean ± SD of 5–10 mice from each group).
5.5 ± 1.0
5.5 ± 2.2
3.6 ± 0.9
4.1 ± 0.9
4.7 ± 1.7
29.5 ± 9.1
3.9 ± 0.7
6.0 ± 3.0
3.9 ± 1.1
3.6 ± 1.4
3.0 ± 1.0
3.8 ± 1.7
5.0 ± 0.7
4.9 ± 2.7
3.3 ± 0.7
4.6 ± 2.2
It was unclear whether these 2 monocyte subsets are derived from different monocyte lineages or whether they originate from the same precursor, with one differentiating into the other in the circulating blood. To address this question, 2 days after an intravenous injection of BrdU in 8-month-old male BXSB Yaa mice, surface expression of Gr-1 on newly generated BrdU+,F4/80+ monocytes was examined by flow cytometry. The majority of BrdU+,F4/80+ recent immigrant cells from the bone marrow stained positively with Gr-1 in both male BXSB Yaa and non-Yaa mice, independent of the proportion of Gr-1 phenotypes among preexisting unlabeled monocytes (Figure 3). This strongly suggests that Gr-1+ monocytes likely became the Gr-1− subset while still in the blood, thus providing evidence against the presence of 2 different monocyte lineages at the level of bone marrow.
Increased expression of CD11c on the resident monocyte subset in aged male BXSB Yaa, but not B6 Yaa, mice.
To further characterize the 2 different subpopulations of monocytes present in aged male BXSB Yaa mice, we determined the expression of different surface markers. As described previously (5, 10), both populations of monocytes expressed neither class II major histocompatibility complex (MHC) molecules nor costimulatory CD80 and CD86 (data not shown). However, the Gr-1− subset of monocytes from 8-month-old male BXSB Yaa mice displayed significant surface expression of CD11c, as compared with Gr-1+ monocytes, which remained negative for CD11c (Figure 4). This enhanced expression was not observed in Gr-1− monocytes from 2-month-old male BXSB Yaa mice. Gr-1− monocytes in 8-month-old male BXSB non-Yaa and B6.Yaa mice minimally expressed CD11c, at levels much lower than those seen in aged male BXSB Yaa mice (P < 0.001) (Figure 4). The age-dependent, selective increase of Gr-1−,CD11c+ monocytes was similarly observed in male (NZB × B6.Yaa)F1 mice in which SLE developed (data not shown). Notably, these Gr-1−,CD11c+ monocytes in aged male BXSB and (NZB × B6.Yaa) F1Yaa mice did not express CD4, CD8α, and B220 at a detectable level (data not shown).
Because a recent study identified a CD11clow, CD11b−,CD45RA+,B220+ population of cells, resembling precursors of plasmacytoid DCs, in the peripheral blood of mice (18), we explored the possible expansion of this particular subset in aged male BXSB Yaa mice. However, CD11b−,CD45RA+,B220+ cells were barely detectable among CD19− non–B cells in circulating blood from these mice (data not shown).
Monocytosis is a unique cellular abnormality associated with Yaa-mediated lupus-like autoimmune disease (5). The present study was designed to define the cellular mechanism responsible for the development of monocytosis and to characterize the effect of the Yaa mutation on the development of 2 different subsets of monocytes. We provide evidence that in mixed bone marrow chimeras, monocytes of both Yaa and non-Yaa origin were similarly involved in the development of monocytosis, suggesting that monocytosis is not attributable to an intrinsic abnormality in the growth potential of monocyte lineage cells in mice bearing the Yaa mutation. Furthermore, we observed that monocytosis resulted in selective expansion of a Gr-1−,CD62L− monocyte subset expressing the CD11c DC marker. Thus, the Yaa mutation could lead to the expansion of cells that develop into DCs in the tissues, thereby contributing to the acceleration of autoimmune responses in lupus-prone mice.
Analysis of Yaa plus non-Yaa mixed bone marrow chimeras clearly demonstrated that there was no selective production of monocytes of Yaa origin over those of non-Yaa origin. This result strongly suggests that monocytosis associated with the Yaa mutation is not attributable to hyperresponsiveness of monocyte lineage cells bearing the Yaa mutation to normal levels of monocyte-specific growth factor(s), but more likely, is a result of excessive production of growth factors during the course of lupus-like autoimmune syndrome. Because Yaa-linked monocytosis is associated with the development of SLE, one attractive hypothesis is that the activation of macrophages, for example by autoantigen-antibody immune complexes produced during the course of the disease, leads to an excessive production of monocyte-specific growth factor(s) by macrophages, resulting in the development of monocytosis in BXSB Yaa mice. It has been shown that the interaction of immune complexes with IgG Fcγ receptor (FcγR) on macrophages triggers the production of monocyte-specific growth factors, such as monocyte colony-stimulating factor and granulocyte–macrophage colony-stimulating factor (GM-CSF) (19, 20). Because lupus-prone mice that are deficient in activating FcγR spontaneously develop autoantibodies at levels comparable with those in wild-type animals (21, 22), it would be of interest to determine whether BXSB mice deficient in FcγR are still able to develop monocytosis, given the active production of lupus autoantibodies.
The second major observation in the present study is that the monocytosis that occurs in lupus-prone Yaa mice is associated with a selective expansion of only 1 of the 2 major monocyte subsets present in the circulation. This Gr-1−,CD62L− subset has been considered a source of resident macrophages and DCs in different tissues (10). Selective expansion of the Gr-1−,CD62L− monocyte subset in aged male BXSB Yaa mice is consistent with the earlier finding that these mice displayed hyperplasia of Kupffer cells (23), which are considered to be derived from the Gr-1− resident monocyte subset (10).
The analysis of BrdU labeling of monocytes in BXSB Yaa mice indicated that recent immigrants from bone marrow enter the circulation as Gr-1+ monocytes. This observation is consistent with the recent demonstration that monocytes repopulating the circulation after depletion of blood monocytes by liposome treatment were exclusively of the Gr-1+ subset (11). In addition, in vitro studies have demonstrated progressive down-regulation of Gr-1 during culture of blood monocytes (24). Thus, it is probable that Gr-1+ and Gr-1− monocytes represent 2 different stages of maturation, during which the Gr-1+ subset becomes the more mature Gr-1− subset while still in the bloodstream. Based on the differential expression of chemokine receptors and adhesion molecules by these 2 subsets, it has been suggested that Gr-1+ monocytes could more efficiently migrate into inflamed tissues (10). Because the production of monocytes in the bone marrow is stimulated during peripheral inflammation, it may be beneficial for the host to have high numbers of immature Gr-1+ monocytes, which have a high potential for migration into sites of inflammation (10, 25), released from the bone marrow.
Importantly, we observed that the Gr-1− subset expressed substantial levels of CD11c in aged lupus-prone male BXSB and (NZB × B6.Yaa)F1Yaa mice. In contrast, expression of CD11c was minimal or absent on the same subset in nonautoimmune mice as well as in young BXSB and (NZB × B6.Yaa)F1Yaa mice. The molecular mechanisms responsible for the expansion of Gr-1− monocytes and for the induction of CD11c expression remain to be determined. One hypothesis is that the presence of the Yaa mutation in lupus-prone mice may induce a unique cytokine environment, possibly through the activation of FcγR by immune complexes, as discussed above, not only promoting the rapid maturation toward the Gr-1− subset, but also inducing an increased level of CD11c expression. Although it has been established that CD11c is a marker of DCs, the expanding population of monocytes in aged male BXSB Yaa mice express neither class II MHC molecules nor costimulatory CD80 and CD86. It is possible that these cells could be the precursors of DCs.
A recent study identified 2 major populations of DC precursors in mouse blood, CD11cintermediate, CD11b+,CD45RA−,B220− and CD11clow,CD11b−, CD45RA+,B220+ (18). The surface phenotype of CD11c+ monocytes that is abundantly present in BXSB Yaa mice is comparable with that of the former cell type, which can give rise to myeloid DCs after incubation with GM-CSF and tumor necrosis factor α (18). Thus, one can speculate that the presence of an increased number of Gr-1−,CD11c+ monocytes may provide a source of resident DCs in the tissues, thereby accelerating autoimmune responses in male BXSB Yaa mice. It should be stressed that aged BXSB and (NZB × B6.Yaa)F1Yaa mice did not display an apparent expansion in blood of the second population of CD11b− DC precursors resembling precursors of plasmacytoid DCs, for which a role in the development of SLE has recently been proposed (26).
Based on the selective production of anti-DNA autoantibodies by B cells bearing the Yaa mutation in Yaa plus non-Yaa mixed bone marrow chimeras (16, 27), we previously proposed that the Yaa defect may decrease the threshold of B cell receptor–mediated signaling, thereby triggering and excessively stimulating autoreactive B cells (2). This is consistent with our recent finding that the Yaa mutation triggers the activation of autoreactive B cells in a T cell–independent manner early in life (28). In addition to this thesis, our present results suggest that Yaa could accelerate the progression of SLE through increased production of monocytes and DCs. Clearly, further understanding of the mechanisms responsible for Yaa-induced B cell activation and monocytosis should help identify the molecular nature of the Yaa mutation and the pathogenesis of SLE.
We thank Mr. G. Brighouse and Mr. G. Celetta for their excellent technical assistance.