PD-1 (PDCD1) and its ligand PD-L1 (CD274) are key immune checkpoint molecules that negatively regulate T-cell activation in the tumor microenvironment. VEGFA is a major pro-angiogenic factor that promotes tumor vascularization and immune suppression. Co-targeting PD-1/PD-L1 and VEGFA has become an important strategy in oncology, with several anti-PD-1/VEGF or anti-PD-L1/VEGF combination therapies and bispecific antibodies currently under clinical investigation.

In PD-1/hPD-L1/hVEGFA humanized mice (B-hPD-1/hPD-L1/hVEGFA), the endogenous mouse Pdcd1, Cd274 and Vegfa genes are replaced by their human counterparts on a C57BL/6 background. Homozygous animals express human PD-1, human PD-L1 and human VEGFA while maintaining normal immune cell composition in spleen, providing a physiologically relevant triple humanized PD-1/PD-L1/VEGFA mouse model. This allows direct in vivo evaluation of human-specific PD-1/PD-L1 checkpoint inhibitors, anti-VEGFA antibodies, and PD-1 x VEGF bispecific antibodies.

When implanted with MC38 cell (B-hVEGFA MC38) or MC38 syngeneic tumors (B-hVEGFA/hPD-L1 plus MC38), PD-1/hPD-L1/hVEGFA humanized mice (B-hPD-1/hPD-L1/hVEGFA) support robust antitumor responses to anti-PD-1/VEGFA bispecific antibodies, bevacizumab & Keytruda combination therapy, and ivonescimab (AK112), enabling translational assessment of tumor growth inhibition, body weight change, and dosing regimens.

Key Advantages

• Triple humanized checkpoint/angiogenesis axis:
  Simultaneous humanization of PD-1, PD-L1, and VEGFA on a C57BL/6 background provides a unique humanized PD-1/PD-L1/VEGFA mouse model to evaluate immuno-oncology and anti-angiogenic therapies within a single system.
• Validated expression of human PD-1, PD-L1 and VEGFA:
  Flow cytometry and ELISA confirm the loss of mouse PD-1, PD-L1, and VEGFA and exclusive expression of human PD-1, human PD-L1, and human VEGFA in the spleen and lung.
• Intact immune cell composition:
  Comparable frequencies of T cells, B cells, NK cells, dendritic cells, granulocytes, monocytes, macrophages, CD4⁺ and CD8⁺ T cells, and Tregs relative to wild-type mice demonstrate that triple humanization does not disrupt immune homeostasis.
• Ready-to-use syngeneic tumor models:
  Compatible with MC38 cell lines (B-hVEGFA and B-hVEGFA/hPD-L1 MC38), enabling fully human ligand–receptor interactions and robust antitumor responses for checkpoint inhibitor and anti-VEGFA drug evaluation.
• Demonstrated efficacy of clinical-relevant agents:
  The model supports in vivo efficacy testing of anti-PD-1/VEGF bispecific antibodies (ivonescimab analog, ivonescimab/AK112) and bevacizumab + Keytruda combination therapy, aligning preclinical findings with ongoing clinical strategies.
• Ideal for antibody validation and combination-strategy optimization:
  Suitable for dose finding, sequence and combination optimization, biomarker exploration and comparative studies versus other humanized PD-1 or PD-L1 mouse models.

PD-1

• The exon 2 of mouse Pdcd1 gene that encodes the IgV domain was replaced by human PDCD1 exon 2 in PD-1/hPD-L1/hVEGFA humanized mice (B-hPD-1/hPD-L1/hVEGFA).

PD-L1

• The exon 3 of mouse Cd274 gene that encodes the IgV domain was replaced by human CD274 exon 3 in PD-1/hPD-L1/hVEGFA humanized mice (B-hPD-1/hPD-L1/hVEGFA). Sequences of other regions still belonged to mice.

VEGFA

• The exons 1-8 of mouse Vegfa gene that encode the full-length protein were replaced by human VEGFA exons 1-8 in PD-1/PD-L1/VEGFA humanized mice (B-hPD-1/hPD-L1/hVEGFA).

The PD-1/PD-L1/VEGFA three knock-in model, was developed by breeding the PD-1 humanized mice (B-hPD-1), the PD-L1 humanized mice (B-hPD-L1) and the VEGFA humanized mice (B-hVEGFA).

• Mouse PD-1 was detected exclusively in wild-type C57BL/6 mice
• Human PD-1 was detected in homozygous PD-1/PD-L1/VEGFA humanized mice, but not in wild-type mice.

Strain-specific analysis of mouse and human PD-1 protein expression in wild-type C57BL/6 mice and PD-1/PD-L1/VEGFA humanized mice by flow cytometry. Splenocytes were collected from wild-type C57BL/6 mice (+/+) and homozygous PD-1/PD-L1/VEGFA humanized mice (H/H), with or without in vivo stimulation with an anti-mouse CD3ε antibody (7.5 μg, i.p.) for 24 hours (female, 6 weeks old, n = 1). Mouse and human PD-1 protein expression was analyzed by flow cytometry using species-specific anti-mouse PD-1 (BioLegend, 109104) and anti-human PD-1 (BioLegend, 329904) antibodies.

• Mouse PD-L1 was detected exclusively in wild-type C57BL/6 mice
• Human PD-L1 was detected in homozygous PD-1/PD-L1/VEGFA humanized mice, but not in wild-type mice.

Strain-specific analysis of mouse and human PD-L1 protein expression in wild-type C57BL/6 mice and PD-1/PD-L1/VEGFA humanized mice by flow cytometry. Splenocytes were collected from wild-type C57BL/6 mice (+/+) and homozygous PD-1/PD-L1/VEGFA humanized mice (H/H), with or without in vivo stimulation with an anti-mouse CD3ε antibody (7.5 μg, i.p.) for 24 hours (female, 6 weeks old, n = 1). Mouse and human PD-L1 protein expression was analyzed by flow cytometry using species-specific anti-mouse PD-L1 (BioLegend, 124312) and anti-human PD-L1 (BioLegend, 329706) antibodies.

• Mouse VEGFA was detected exclusively in wild-type C57BL/6 mice
• Human VEGFA was detected in homozygous PD-1/PD-L1/VEGFA humanized mice, but not in wild-type mice.

Strain-specific analysis of mouse and human VEGFA protein expression in wild-type C57BL/6 mice and PD-1/PD-L1/VEGFA humanized mice by ELISA. Lung tissue homogenates were collected from wild-type C57BL/6 mice (+/+) (male, 8 weeks old, n = 3) and homozygous PD-1/PD-L1/VEGFA humanized mice (H/H; H/H; H/H) (male, 8 weeks old, n = 3). Mouse and human VEGFA protein expression levels were measured by ELISA using species-specific mouse VEGFA (R&D Systems, MMV00) and human VEGFA (R&D Systems, DVE00) ELISA kits, respectively.

• The frequencies of T cells, B cells, NK cells, DCs, granulocyte, monocytes, macrophages, CD4+ T cells, CD8+ T cells and Tregs in homozygous PD-1/PD-L1/VEGFA humanized mice were similar to those in C57BL/6 mice.
• Humanization of PD-1, PD-L1 and VEGFA does not change the frequency or distribution of these cell types in spleen.

Flow cytometry analysis of leukocyte and T-cell subpopulations in the spleens of PD-1/PD-L1/VEGFA humanized mice. Splenocytes were isolated from wild-type C57BL/6 mice and homozygous PD-1/PD-L1/VEGFA humanized mice (10 weeks old, n = 3). (A) Flow cytometry analysis of leukocyte subpopulation frequencies in the spleen. (B) Analysis of T-cell subpopulation frequencies. Values are expressed as mean ± SEM. Statistical significance was determined using two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001.

Establishment of a human VEGFA-expressing MC38 tumor model and in vivo efficacy evaluation of an ivonescimab analog in PD-1/PD-L1/VEGFA humanized mice. Human VEGFA-expressing MC38 cells (B-hVEGFA MC38) were subcutaneously implanted into homozygous PD-1/PD-L1/VEGFA humanized mice (female, 8 weeks old, n = 6). When the average tumor volume reached approximately 70–90 mm³, mice were randomized into treatment groups and subsequently administered an ivonescimab analog via intraperitoneal injection.

In vivo antitumor efficacy of an ivonescimab analog in a human VEGFA-expressing MC38 tumor model using PD-1/PD-L1/VEGFA humanized mice. (A) Tumor growth curves. (B) Body weight changes during treatment. As shown in panel A, treatment with the in-house ivonescimab analog effectively inhibited tumor growth in PD-1/PD-L1/VEGFA humanized mice. Values are expressed as mean ± SEM. The tumor model establishment rate was 50%.

Establishment of a human VEGFA-expressing MC38 tumor model and in vivo efficacy evaluation of bevacizumab and pembrolizumab (Keytruda) combination therapy in PD-1/PD-L1/VEGFA humanized mice. Human VEGFA-expressing MC38 cells (B-hVEGFA MC38) were subcutaneously implanted into homozygous PD-1/PD-L1/VEGFA humanized mice (female, 8 weeks old, n = 6). When the average tumor volume reached approximately 70–90 mm³, mice were randomized into treatment groups and subsequently administered anti-PD-1 and anti-VEGFA antibodies via intraperitoneal injection.

In vivo antitumor efficacy of bevacizumab and pembrolizumab (Keytruda) combination therapy in a human VEGFA-expressing MC38 tumor model using PD-1/PD-L1/VEGFA humanized mice. (A) Tumor growth curves. (B) Body weight changes during treatment. As shown in panel A, bevacizumab effectively inhibited tumor growth in PD-1/PD-L1/VEGFA humanized mice. However, combination treatment with bevacizumab and pembrolizumab (Keytruda) did not demonstrate greater antitumor efficacy than monotherapy. Values are expressed as mean ± SEM. The tumor model establishment rate was 40%.

Establishment of a human VEGFA/PD-L1-expressing MC38 tumor model and in vivo efficacy evaluation of an ivonescimab analog in PD-1/PD-L1/VEGFA humanized mice. Human VEGFA/PD-L1-expressing MC38 cells (B-hVEGFA/hPD-L1 plus MC38) were subcutaneously implanted into homozygous PD-1/PD-L1/VEGFA humanized mice (male, 8 weeks old, n = 6). When the average tumor volume reached approximately 300 mm³, mice were randomized into treatment groups and subsequently administered anti-PD-1 and anti-VEGFA antibodies via intraperitoneal injection.

In vivo antitumor efficacy of ivonescimab (AK112) in a human VEGFA/PD-L1-expressing MC38 tumor model using PD-1/PD-L1/VEGFA humanized mice. (A) Tumor growth curves. (B) Body weight changes during treatment. As shown in panel A, ivonescimab (AK112) effectively inhibited tumor growth in PD-1/PD-L1/VEGFA humanized mice. Values are expressed as mean ± SEM. The tumor model establishment rate was 60%.

Establishment of a human VEGFA/PD-L1-expressing B16-F10 melanoma model and in vivo efficacy evaluation of ivonescimab in PD-1/PD-L1/VEGFA humanized mice. Human VEGFA/PD-L1-expressing B16-F10 mouse melanoma cells (B-hVEGFA/hPD-L1 B16-F10) were subcutaneously implanted into homozygous PD-1/PD-L1/VEGFA humanized mice (female, 8 weeks old, n = 6). When the average tumor volume reached approximately 150 mm³, mice were randomized into treatment groups and subsequently administered anti-PD-1 and anti-VEGFA antibodies via intraperitoneal injection.

In vivo antitumor efficacy of ivonescimab (AK112) in a human VEGFA/PD-L1-expressing B16-F10 melanoma model using PD-1/PD-L1/VEGFA humanized mice. (A) Tumor growth curves. (B) Body weight changes during treatment. As shown in panel A, treatment with ivonescimab (AK112) effectively inhibited tumor growth in PD-1/PD-L1/VEGFA humanized mice. Values are expressed as mean ± SEM. The tumor model establishment rate was 60%.
Note: One mouse in group G3 died on day 12.