Anti-Gr-1 antibody depletion fails to eliminate hepatic myeloid-derived suppressor cells in tumor-bearing mice

Abstract

Recent studies show that the liver is a preferred organ for the accumulation of MDSC. In this study, we examined the effect of systemic RB6-8C5 treatment on hepatic MDSC in tumor-bearing mice. EL4 tumor-bearing mice were injected i.p. with RB6-8C5, and hepatic, splenic, and blood MDSCs were analyzed by flow cytometry. Unexpectedly, hepatic MDSC remained in the liver, although RB6-8C5 completely eliminated them from the spleen and peripheral blood 24 h after treatment. Secondary antibody staining confirmed the presence of RB6-8C5-bound MDSC in the liver of mice with s.c. tumors. Similar observations were made in two other (colon and melanoma) tumor models. Whereas RB6-8C5 injection induced cell death of hepatic MDSC, as shown by Annexin V/7-AAD staining, these cells were replaced immediately, leading to a constant, increased frequency of hepatic MDSC. Adoptively transferred MDSC migrated preferentially to the liver after RB6-8C5 treatment, suggesting that hepatic MDSCs are reconstituted rapidly after depletion. Finally, hepatic MDSC remained immunosuppressive despite RB6-8C5 injection. Our study demonstrates that RB6-8C5 is not suitable for depletion of hepatic MDSCs and analysis of their function.

Figure 1.. Anti-Ly6C antibody stains RB6-8C5-bound MDSC.

Splenocytes from EL4 tumor bearing mice were isolated and preincubated with different concentration of RB6-8C5 antibody for 15 min and then stained with anti-CD11b-FITC plus anti-Ly6C-APC (HK1.4), anti-Ly6G-PE (1A8), or anti-Gr-1-APC (RB6-8C5), respectively. A rat IgG2b served as isotype control. N, Cells without antibody preincubation. Representative dot-plots from three independent experiments are shown in A, and competitive staining graph is shown in B. EL4 tumor-bearing mice were injected i.p. with 200 μg RB6-8C5 or isotype control. Two hours after treatment, mice were killed, and liver infiltrating cells were prepared and stained with Ly6C-APC, anti-Gr-1-APC (RB6-8C5), or a goat anti-rat IgG (2nd-Ab). Representative dot-plots from two independent experiments are shown in C. TB, Tumor-bearing mice without in vivo antibody treatment; Rb6, RB6-8C5-injected mice; Iso, rat IgG2b isotype antibody-injected mice.

Figure 2.. RB6-8C5 depletes MDSC in spleen and peripheral blood but not in liver of EL4 tumor-bearing mice.

EL4 tumor-bearing mice were injected i.p. with 200 μg RB6-8C5. Twenty-four hours later, mice were killed, and liver-infiltrating cells, splenocytes, and peripheral blood cells were prepared. MDSCs were detected by anti-Ly6C/anti-CD11b costaining. RB6-8C5-bound MDSCs were identified by staining with goat anti-rat IgG-biotin (2nd-Ab), followed by streptavidin/anti-CD11b costaining. A rat IgG2b antibody was injected as an isotype control. Representative dot-plots are shown in A and C. Cumulative results ± sem of four independent experiments are shown in B and D (naïve, n=4; TB, n=4; Rb6, n=7; Iso, n=7; *P<0.05).

Figure 3.. Hepatic MDSCs resist RB6-8C5 depletion in B16-GM-CSF and CT-26 tumor model.

Mice-bearing EL4, B16-GM-CSF, or CT-26 tumor cells were injected i.p. with 200 μg RB6-8C5. Twenty-four hours later, mice were killed, and liver-infiltrating cells and splenocytes were prepared. MDSCs were detected by anti-Ly6C/anti-CD11b costaining (A and B). RB6-8C5-bound MDSCs were identified by staining with goat anti-rat IgG-biotin (2nd-Ab), followed by streptavidin/anti-CD11b costaining (C). A rat IgG2b antibody was injected as an isotype control. Cumulative results ± sem are shown (for EL4 tumor: TB, n=4; Rb6, n=7; Iso, n=7; for B16-GM-CSF tumor: TB, n=4; Rb6, n=4; Iso, n=4; for CT-26 tumor: TB, n=4; Rb6, n=6; Iso, n=6; *P<0.05).

Figure 4.. RB6-8C5 stays on hepatic MDSC surface for at least 4 days in vivo.

EL4 tumor-bearing mice were injected i.p. with 200 μg RB6-8C5. At different time-points after treatment (1D, 1 day; 2D, 2 days; 4D, 4 days), mice were killed, and liver-infiltrating cells, splenocytes, and peripheral blood cells were prepared. MDSCs in different compartments were stained as described above. Representative dot-plots of hepatic MDSCs are shown in A and C. Cumulative results ± sem are shown in B and D (naïve, n=3; TB, n=3; 1D, n=6; 2D, n=4; 4D, n=3; *P<0.05).

Figure 5.. RB6-8C5-bound hepatic MDSCs show higher cell death rate but are replaced rapidly by new MDSCs.

CT-26 tumor-bearing mice were injected i.p. with 200 μg RB6-8C5 or isotype control. Annexin V/7-AAD staining was used to detect cell death 24 h after injection. Hepatic MDSCs were identified by gating on Ly6C⁺CD11b⁺ cells. Representative dot-plots are shown in A. Cumulative results ± sem from two independent experiments are shown (B; TB, n=4; Iso, n=4; Rb6, n=4; *P<0.05). Cell proliferation of hepatic MDSC before and after RB6-8C5 depletion was detected by Ki-67 staining. Splenic MDSCs were used as a control. Representative dot-plots from two independent experiments are shown in C and D. FSC, Forward-scatter. (E–G) Adoptive transfer was performed to study MDSC migration to liver upon RB6-8C5 treatment. Donor CD45.1 bone marrow MDSCs were isolated as described in Materials and Methods. RB6-8C5 or isotype control antibody (200 μg) was injected into CD45.2 EL4 tumor-bearing mice, and 24 h later, 5 × 10⁶ donor bone marrow-derived MDSCs were adoptively transferred. Three hours after cell transfer, mice were killed, and CD45.1 donor cells in liver were measured by flow cytometry. Representative dot-plots are shown in E. Cumulative results from two independent experiments (Iso, n=4; Rb6, n=4; *P<0.05) are indicated in F (relative numbers) and G (absolute numbers).

Figure 6.. RB6-8C5-bound hepatic MDSCs retain immunosuppressive function.

MDSCs from mice injected with RB6-8C5 or isotype antibody were isolated by autoMACS, as described in Materials and Methods. CFSE-labeled OT-I splenocytes were cocultued with splenic or hepatic MDSCs at indicated ratios. OVA peptide (0.1 μg/ml) was added to stimulate T cell proliferation. Two days later, proliferation of CD8 T cells was analyzed by flow cytometry. Representative flow cytometry histograms are shown in A. Cumulative results ± sem are from two independent experiments (B; TB, n=4; Iso, n=4; Rb6, n=4).