Charlebois and colleagues recently reported that the antitumor activity of anti-ErbB2 mAb is enhanced by local polyI:C and CpG administration in murine breast tumor models (1) and concluded that this activity was dependent on IFNs, CD8+ T, and natural killer (NK) cells. The requirement for NK cells was based on in vivo depletion using anti-asialo-GM1 (ASGM1) Ab.
ASGM1 is a glycolipid highly expressed on the surface of mouse NK cells, but also expressed on T cells, NKT cells, eosinophils, basophils, and macrophages (2). Treatment with anti-ASGM1 Ab reduces NK-cell cytolytic ability in vitro, and in vivo, results in NK depletion by an unexplained mechanism (3). Anti-ASGM1 is available as polyimmunoglobulin (pIg) or IgG preparations. Most published studies, including Charlebois and colleagues' work, use the pIg. Our studies show that anti-ASGM1 pIg binds to macrophages at a higher level than anti-ASGM1 IgG (Fig. 1A and B) and that the IgG preparation is more efficient at depleting NK cells (Fig. 1C and D). When mice were treated with 50, 25, or 10 μL anti-ASGM1 on day 1, 10 μL of anti-ASGM1 IgG was as efficacious as 50 μL of pIg in depleting peripheral and splenic NK cells. Despite anti-ASGM1 binding to myeloid cells, we did not observe any significant changes in the number of these cells (Fig. 1E). Next, we examined the effects of anti-ASGM1 on macrophage function. Anti-ASGM1 pIG, but not IgG (at 10 μg/mL or a supersaturating level of 100 μg/mL), impaired macrophage-mediated phagocytosis of anti-CD20 (4)–opsonized human CD20–transgenic B cells in vitro (Fig. 2A and B). This effect was absent with deglycosylated anti-ASGM1 pIG, indicating that FcγR-dependent effector function was impaired. In vivo, anti-ASGM1 pIg reduced anti-CD20–mediated B-cell depletion to 60% in some mice (Fig. 2C). In contrast, the IgG preparation induced a less profound effect, especially at 10 μL, reducing depletion by only 10% at 48 hours.
Characterization of anti-ASGM1 pIg and IgG on NK and myeloid cells. A, BALB/c and C57BL/6J mice splenocytes were FcγR blocked for 10 minutes at room temperature, then untreated or treated with anti-ASGM1 pIg (pIg, 1/80 dilution, Wako Chemicals) or anti-ASGM1 IgG (IgG, 10 μg/mL, BioLegend) for 30 minutes, 4°C and washed twice. n.s., not significant. Cells were then stained with an Alexa 488–conjugated anti-rabbit IgG for 15 minutes at room temperature and washed twice before analysis on a flow cytometer. The histograms and median fluorescence intensity values are representative of 6 mice per strain. B, The graph shows the cumulative data described in A. Median fluorescence intensity values of individual “pIg” and “IgG” samples were divided by the average median fluorescence intensity value of “secondary only” samples in each experiment. P values were calculated using the t test, n = 6 per group. C, BALB/c mice were treated with rabbit serum (RS), anti-ASGM1 pIg, or anti-ASGM1 IgG intraperitoneally using the volumes shown. Tail blood was drawn immediately prior to injection (D1) and from D4–6 to stain for NK cells. n = 3, means and SD shown (t test). D, The spleens of the mice treated in B were harvested and NK cells analyzed by flow cytometry. n = 3, means and SD shown (t test). E, Cells from C were also stained for macrophages, monocytes, and neutrophils. n = 3, means and SD shown (t test). *, P < 0.05; **, P < 0.01; ***, P < 0.001. All animal experiments were approved by the local ethical committee and conducted according to the UK Home Office license guidelines.
Characterization of anti-ASGM1 pIg and IgG on NK and myeloid cells. A, BALB/c and C57BL/6J mice splenocytes were FcγR blocked for 10 minutes at room temperature, then untreated or treated with anti-ASGM1 pIg (pIg, 1/80 dilution, Wako Chemicals) or anti-ASGM1 IgG (IgG, 10 μg/mL, BioLegend) for 30 minutes, 4°C and washed twice. n.s., not significant. Cells were then stained with an Alexa 488–conjugated anti-rabbit IgG for 15 minutes at room temperature and washed twice before analysis on a flow cytometer. The histograms and median fluorescence intensity values are representative of 6 mice per strain. B, The graph shows the cumulative data described in A. Median fluorescence intensity values of individual “pIg” and “IgG” samples were divided by the average median fluorescence intensity value of “secondary only” samples in each experiment. P values were calculated using the t test, n = 6 per group. C, BALB/c mice were treated with rabbit serum (RS), anti-ASGM1 pIg, or anti-ASGM1 IgG intraperitoneally using the volumes shown. Tail blood was drawn immediately prior to injection (D1) and from D4–6 to stain for NK cells. n = 3, means and SD shown (t test). D, The spleens of the mice treated in B were harvested and NK cells analyzed by flow cytometry. n = 3, means and SD shown (t test). E, Cells from C were also stained for macrophages, monocytes, and neutrophils. n = 3, means and SD shown (t test). *, P < 0.05; **, P < 0.01; ***, P < 0.001. All animal experiments were approved by the local ethical committee and conducted according to the UK Home Office license guidelines.
Anti-ASGM1 reduces in vitro and in vivo macrophage-mediated antibody-dependent cellular phagocytosis. A, Murine bone marrow–derived macrophages were incubated in the presence of rabbit serum (RS), anti-ASGM1 pIg, or anti-ASGM1 IgG (native or deglycosylated) overnight and then cocultured with target human CD20 transgenic purified splenic B cells. Target cells were CFSE-labeled before opsonization with Ritux-m2a (10 μg/mL, produced inhouse as previously described) or an isotype control. Phagocytic cells were identified by double-positive CFSE+F4/80-APC+ staining, using flow cytometry. n = 4 to 13, means and SD shown (t test). n.s., not significant. B, The phagocytosis assay described in A was performed. To exclude differences in effect due to disparities in IgG concentrations in pIg and IgG preparations, murine bone marrow–derived macrophages were treated with rabbit serum, 10 μg/mL or a supersaturating concentration of 100 μg/mL anti-ASGM1 IgG. n = 4 to 6, means and SD shown. C, BALB/c mice were treated intraperitoneally with 50 μL RS, 50 μL anti-ASGM1 pIg, 50 μL or at an equally efficacious dose of 10 μL IgG to deplete NK cells (all given two doses and 4 days apart). Twenty-four hours after the second dose of anti-ASGM1, 100 μg anti-CD20 mAb (18B12 mIg2a, produced inhouse) was administered intraperitoneally to deplete B cells. Absolute peripheral blood B-cell counts were enumerated using a Coulter Counter and by flow cytometry at 24 and 48 hours after anti-CD20 administration. Percent B-cell depletion was derived by comparing absolute B-cell counts at individual conditions and time points over pre–anti-CD20 B-cell counts in respective mice. n = 9 to 12, means shown (t test).
Anti-ASGM1 reduces in vitro and in vivo macrophage-mediated antibody-dependent cellular phagocytosis. A, Murine bone marrow–derived macrophages were incubated in the presence of rabbit serum (RS), anti-ASGM1 pIg, or anti-ASGM1 IgG (native or deglycosylated) overnight and then cocultured with target human CD20 transgenic purified splenic B cells. Target cells were CFSE-labeled before opsonization with Ritux-m2a (10 μg/mL, produced inhouse as previously described) or an isotype control. Phagocytic cells were identified by double-positive CFSE+F4/80-APC+ staining, using flow cytometry. n = 4 to 13, means and SD shown (t test). n.s., not significant. B, The phagocytosis assay described in A was performed. To exclude differences in effect due to disparities in IgG concentrations in pIg and IgG preparations, murine bone marrow–derived macrophages were treated with rabbit serum, 10 μg/mL or a supersaturating concentration of 100 μg/mL anti-ASGM1 IgG. n = 4 to 6, means and SD shown. C, BALB/c mice were treated intraperitoneally with 50 μL RS, 50 μL anti-ASGM1 pIg, 50 μL or at an equally efficacious dose of 10 μL IgG to deplete NK cells (all given two doses and 4 days apart). Twenty-four hours after the second dose of anti-ASGM1, 100 μg anti-CD20 mAb (18B12 mIg2a, produced inhouse) was administered intraperitoneally to deplete B cells. Absolute peripheral blood B-cell counts were enumerated using a Coulter Counter and by flow cytometry at 24 and 48 hours after anti-CD20 administration. Percent B-cell depletion was derived by comparing absolute B-cell counts at individual conditions and time points over pre–anti-CD20 B-cell counts in respective mice. n = 9 to 12, means shown (t test).
In summary, we demonstrate that the anti-ASGM1 Ab used by Charlebois and colleagues can disrupt macrophage function. Although we do not dispute that NK cells might have a role in the described antitumor activity, our data suggest that the authors may have inadvertently excluded the contribution of macrophages, which can be affected by pIg anti-ASGM1 and which are known to be critical mediators of mAb function (5). Therefore, we highlight an important caveat of using the pIg preparation of anti-ASGM1 to deplete NK cells and recommend the use of a pure IgG preparation.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Grant Support
This work was supported by a Cancer Research U.K. Clinician Scientist Fellowship (C30010/A15269 to S.H. Lim) and a Bloodwise grant (12050 to M.S. Cragg).
RSS Feeds