CD38 on tumor cells suppresses CD8+ T cell function.
(A) Growth of subcutaneous 344SQ tumors in immune competent
129/Sv mice treated with IgG control, anti-PD-L1 (clone 9G2; 200 μg per
mouse), anti-PD-L1 plus anti-CD8 (clone 2.43; 200 μg per mouse),
respectively. Mice (n = 5 or 8) were intraperitoneally treated with the antibody
once a week for 7 weeks beginning on day 7 after the tumor cell injection (1
× 106 cells per mouse). In the group of anti-PD-L1 plus
anti-CD8 treatment, mice were pretreated with anti-CD8 antibody (400 μg
per mouse) one week before tumor cell injection. Tumor sizes are presented as
mean ± SEM and statistical significance (ns, no significant difference;
*p < 0.05; **p < 0.01) determined on indicated weeks. The
statistically significant differences between the groups of control and
anti-PD-L1 are as follows: week 1, ns; week 2, *; week 3, **; week 4, *; week 5,
**; week 6, *; week 7, ns; week 8, ns. ANOVA test was used to calculate the
significant difference between two groups. For the analysis among multiple
groups at the end point (week 8), ANOVA was used to analyze, *p <
0.05.
(B) FACS analysis of % CD8+TIL, the
proliferation marker Ki67, surface CD44, CD62L, PD1, LAG3, and TIM3 marker
expression levels on CD8+ T cells from primary tumors in 129/Sv mice
(n = 5) treated with anti-PD-L1 antibody at week 5. Data are shown as mean
± sd. t-test was used to analyze.
(C) PD-L1KOCD38high531LN3 cells or
PD-L1KOCD38negative531LN3 cells (5 ×
106 cells per mouse) were subcutaneously injected into immune
competent 129/Sv mice (n = 5). The primary tumor sizes at week 4 are shown with
mean ± SEM. t-test was used to analyze.
(D) (Left panel) CD8 T cells in spleen were determined 2
weeks post initial anti-CD8 antibody injection to test CD8 T cell depletion
efficiency. (Middle panel) 5 × 106 of
PD-L1KOCD38high531LN3 cells were subcutaneously
injected into 129/Sv mice (n = 5) after CD8 T cells were depleted
(α-CD8). The T cell un-depleted group was included as the control (IgG).
The total tumors were measured 4 weeks post tumor cell transplantation and are
shown with mean ± SEM. t-test was used to analyze.
(Right panel) 5 × 106 of
PD-L1KOCD38negative 531LN3 cells were subcutaneously
injected into 129/Sv mice (n = 10) after CD8 T cells were depleted
(α-CD8). The T cell un-depleted group was included as the control (IgG).
The total tumors from primary site and peritoneal cavity were measured 5 weeks
post tumor cell transplantation and are shown with mean ± SEM.
t-test was used to analyze.
(E) To prepare CD8+ T cells, 129/Sv mice were
challenged with 0.5 × 106 344SQ for 2 weeks. CD8+ T
cells were isolated from these tumors, blood, and spleens. A separate cohort of
344SQ tumor-bearing mice were treated with anti-PD-L1 antibody or control (as
described in Figure 1A), then used as
recipients for the CD8 T cell adoptive transfer assay. At week 4, mice received
cyclophosphamide at 100 mg/kg intravenously 6 hours before CD8 T cell transfer
(6 × 106 per mouse, intravenously), followed by IL-2 (20,000
units, intraperitoneally) at 8 hours after T cell transfer then every 12 hours
for 3 days. The tumor growth curves are shown in the left panel. At the
endpoint, mice were necropsied to harvest primary tumors and lungs, which were
weighed, and to quantify distant metastases. The primary tumor weights and lung
metastatic nodules are shown in the middle and right panels. ANOVA test was used
to analyze.
(F) Cell lines with stable expression of a scramble control
(344SQ-scr) or shRNA against CD38 (344SQ-shCD38) were generated and subjected to
qPCR assays. Relative CD38 mRNA levels are normalized to L32 and shown in the
left panel, the Western blot assay for protein in the middle panel, versus
β-actin as a loading control. The surface expression of CD38 on the cell
lines was determined using FACS and are shown in the right panel.
(G) The stable cell lines 344SQ_vector (empty vector
control) and 344SQ_CD38 (CD38 overexpression) were generated and subjected to
qPCR assays. mRNA levels of CD38 are normalized to L32 and shown in the left
panel, Western blot assay for protein in the middle panel, versus β-actin
as a loading control. The surface expression of CD38 was determined using FACS
and is in the right panel.
(H-J) To prepare tumor specific
CD8+ T cells, 129/Sv mice were challenged with 0.5 ×
106 344SQ for 2 weeks. CD8+ T cells were isolated from
these tumors, blood, and spleens. CFSE-labeled CD8+ T cells were
co-cultured with indicated cancer cells in the presence of anti-CD3 (5
μg/ml) and anti-CD28 (5 μg/ml) for 4 days. CD8+ T cells
only was included as the control. T cell proliferation was quantified using FACS
analysis (H). The supernatants from each co-culture were subjected
to ELISA assay to measure IFN-γ and TNF-α (I). Tumor
cells and CD8+ T cells were co-cultured at the indicated ratios in
the presence of anti-CD3 (5 μg/ml) and anti-CD28 (5 μg/ml) for 4
days. Tumor cells only were used as the control for calculation. At day 4,
CD8+ T cells and dead tumor cells were washed away and viable
tumor cells were counted with 0.4% Trypan Blue staining. The CD8+ T
cell killing efficiency is shown in (J). The experiments were
repeated at least three times. p values were calculated with
ANOVA test.
(K-M) To prepare tumor specific
CD8+ T cells, C57BL/6 mice were challenged with 0.2 ×
106 LLC-JSP for 2 weeks. CD8+ T cells were isolated
from these tumors, blood, and spleens. CFSE-labeled CD8+ T cells were
co-cultured with indicated cancer cells in the presence of anti-CD3 (5
μg/ml) and anti-CD28 (5 μg/ml) for 4 days. CD8+ T cells
only was included as the control. T cell proliferation was quantified using FACS
analysis (K). The supernatants from each co-culture were subjected
to ELISA assay to measure IFN-γ and TNF-α (L). Tumor
cells and CD8+ T cells were co-cultured at the indicated ratios in
the presence of anti-CD3 (5 μg/ml) and anti-CD28 (5 μg/ml) for 4
days. Tumor cells only were used as the control for calculation. At day 4, after
taking photos, CD8+ T cells and some dead tumor cells were washed
away and viable tumor cells were counted with 0.4% Trypan Blue staining. The
CD8+ T cell killing efficiency is shown in (M). The
experiments were repeated at least three times. p values were calculated with
ANOVA test. Scale bars represent 100 μm.
(N and O) To prepare tumor specific
CD8+ T cells, 129/Sv mice were challenged with 0.5 ×
106 531LN3 for 2 weeks. CD8+ T cells were isolated
from these tumors, blood, and spleens. CFSE-labeled CD8+ T cells were
co-cultured with indicated cancer cells (sorted from PD-L1WT531LN3 or
PD-L1KO531LN3) in the presence of anti-CD3 (5 μg/ml) and
anti-CD28 (5 μg/ml) for 4 days. CD8+ T cells only was included
as the control. T cell proliferation was quantified using FACS analysis
(N). The supernatants from each co-culture were subjected to
ELISA assay to measure IFN-γ and TNF-α (N). Tumor
cells and CD8+ T cells were co-cultured at the indicated ratios in
the presence of anti-CD3 (5 μg/ml) and anti-CD28 (5 μg/ml) for 4
days. Tumor cells only were used as the control for calculation. At day 4,
CD8+ T cells and some dead tumor cells were washed away and
viable tumor cells were counted with 0.4% Trypan Blue staining. The
CD8+ T cell killing efficiency is shown in (O). The
experiments were repeated at least three times. p values were calculated with
ANOVA test.