Guide: Mouse models for the assessment of therapeutics targeting the PD-1/PD-L1 axis: practical considerations for translational research

Key messages (summary)

• The PD-1/PD-L1 pathway is one of the most clinically validated immune checkpoints in oncology.
• Most clinical anti-PD-1 and anti-PD-L1 antibodies bind human targets and do not fully cross-react with mouse PD-1 or PD-L1.
• Wild-type mice often require surrogate anti-mouse antibodies and cannot directly evaluate human-specific therapeutics.
• Humanized PD-1 and PD-L1 knock-in mice enable direct testing of human therapeutics in an immunocompetent setting.
• Physiological expression of humanized targets is critical for generating translatable efficacy and pharmacodynamic data.
• Model selection depends on whether the objective is target engagement, efficacy testing, mechanism-of-action studies, candidate ranking, combination therapy assessment, or evaluation in a human immune system.
• Humanized immune system (HIS) models provide access to human immune cells but introduce additional variability, cost, and complexity.

‍

‍

Introduction/background

What is the PD-1/PD-L1 pathway?

Programmed cell death protein 1 (PD-1, CD279) is an inhibitory receptor expressed primarily on activated T cells.

Its main ligands are:

• PD-L1 (CD274)
• PD-L2 (CD273)

Binding of PD-L1 or PD-L2 to PD-1:

• Suppresses T-cell activation
• Limits cytokine production
• Reduces cytotoxic activity
• Promotes T-cell exhaustion

Tumors frequently exploit this pathway to evade immune destruction.

Why is the PD-1/PD-L1 axis important in oncology?

PD-1/PD-L1 blockade has transformed cancer treatment.

Approved PD-1 therapeutics include:

• Pembrolizumab
• Nivolumab
• Cemiplimab
• Dostarlimab
• Atezolizumab
• Durvalumab
• Avelumab

These therapies can:

• Restore anti-tumor immunity
• Increase CD8+ T-cell activity
• Promote tumor rejection
• Generate durable clinical responses

What is a humanized PD-1 mouse model?

A humanized PD-1 mouse is a genetically engineered knock-in model in which murine PD-1 is replaced by its human counterpart while preserving physiological regulation of expression.

This allows:

• Direct binding of human anti-PD-1 therapeutics
• Normal endogenous immune-cell development
• Evaluation of human-specific checkpoint inhibitors in vivo

What is a humanized PD-1/PD-L1 mouse model?

These models carry humanized versions of both checkpoint partners (ligand and receptor).

Advantages include:

• Simultaneous human PD-1 and PD-L1 interactions
• Direct testing of anti-PD-1 and anti-PD-L1 antibodies
• Head-to-head comparison of checkpoint inhibitors
• More comprehensive modeling of target biology

The value of double-humanized PD-1/PD-L1 mice has been demonstrated for direct comparison of FDA-approved checkpoint inhibitors (Barham et al., 2023).

‍

‍

How to choose the correct model for a study

Ask yourself these questions first:

• Is my molecule targeting human PD-1 or human PD-L1?
• Am I evaluating efficacy, pharmacodynamics, safety, or all three?
• Do I require an intact immune system?
• Do I need human immune cells?
• Am I studying combination immunotherapy?
• Am I comparing multiple checkpoint inhibitors?
• Do I need human tumors?

‍

Comparison of mouse models for PD-1/PD-L1 therapy assessment

‍

Practical examples

Can I test my human anti-PD-1 antibody in a wild-type mouse?

Generally no.

Why?

• Human PD-1 antibodies are typically engineered against human PD-1 epitopes.
• Cross-reactivity with mouse PD-1 is often limited or absent.
• Surrogate anti-mouse antibodies may be required.

Preferred model:

• Humanized PD-1 knock-in mice (genO-hPD-1).

Do I need humanized PD-L1 as well?

Use PD-1-only models when:

• Developing anti-PD-1 therapeutics
• Confirming target engagement
• Exploring mechanisms of action

Use PD-1/PD-L1 double-humanized models when:

• Comparing anti-PD-1 versus anti-PD-L1 agents
• Developing anti-PD-L1 antibodies
• Studying checkpoint biology more comprehensively

Can the humanized PD-1 or PD-1/PD-L1 models be used to compare different FDA-approved antibodies head-to-head or as a positive control for new therapeutics?

Yes. The single-humanized genO-hPD-1 mouse model has been used to develop the FDA-approved Nivolumab (Opdivo). Double-humanized PD-1/PD-L1 models can also be used to directly compare approved checkpoint inhibitors and reveal differences in anti-tumor activity not apparent from clinical literature alone, or as a point of comparison (positive control) for the development of new therapeutics.

Can hPD-1 or hPD-1/hPD-L1 models evaluate combination immunotherapy?

Yes. Common combinations include:

• Anti-PD-1 + chemotherapy
• Anti-PD-1 + radiotherapy
• Anti-PD-1 + TIGIT blockade (with a TIGIT surrogate antibody)
• Anti-PD-1 + CTLA-4 blockade (with a CTLA-4 surrogate antibody)
• Anti-PD-1 + T-cell engagers (with TCEs targeting murine CD3)
• Anti-PD-1 + cytokine therapies

Can the humanized PD-1 mice model T-cell exhaustion?

Yes. PD-1 expression is a hallmark marker of exhausted T cells.

These models enable the study of:

• Checkpoint blockade
• T-cell reinvigoration
• Exhaustion reversal
• Memory responses
• Cytotoxic CD8+ T-cell activation

Can humanized PD-1 mice be used for inflammation and autoimmune disease research?

Yes. PD-1 is a key regulator of immune tolerance.

Beyond cancer, the PD-1/PD-L1 pathway plays a critical role in:

• Maintaining peripheral tolerance
• Limiting excessive T-cell activation
• Preventing tissue damage during chronic inflammation
• Controlling auto-reactive T cells

Disruption of this pathway is associated with multiple autoimmune and inflammatory diseases, including:

• Rheumatoid arthritis (RA)
• Systemic lupus erythematosus (SLE)
• Multiple sclerosis (MS)
• Inflammatory bowel disease (IBD)
• Type 1 diabetes (T1D)

‍

‍

FAQ

Which PD-1 model has the most physiological expression?

Models ensuring that the expression of humanized PD-1 is similar to the murine PD-1 expression generally provide the most physiological expression profile and preserve normal immune-cell development.

Are these mice immunocompetent?

The knock-in humanized PD-1 (genO-hPD-1) and PD-1/PD-L1 mice (genO-hPD-1/hPD-L1) are immunocompetent.

They retain:

• Functional T cells
• Functional B cells
• Normal immune architecture
• Endogenous immune development

Unlike HIS models, they do not require human cell engraftment.

Does replacing mouse PD-1 disrupt immune development?

Both genO-hPD-1 and genO-hPD-1/hPD-L1 mouse models show preserved immune development and functional immune responses, supporting their use for efficacy and MoA studies.

Does human PD-1 preserve downstream signaling?

The objective of knock-in humanization is to maintain physiological signaling while enabling binding of human therapeutics.

Key validation includes:

• Immune-cell frequencies
• PD-1 expression kinetics
• Functional T-cell responses
• Tumor growth inhibition after checkpoint blockade

What is the advantage of humanizing both PD-1 and PD-L1?

Humanizing both PD-1 and PD-L1 creates a model in which the entire therapeutic target axis is humanized, allowing direct evaluation of both anti-PD-1 and anti-PD-L1 antibodies under physiologically relevant conditions. This increases translational relevance compared with a PD-1-only model and avoids artifacts caused by human-mouse receptor mismatch.

Are there available cell lines that can be used in syngeneic tumor studies with these models?

Yes. Humanized PD-1 and PD-1/PD-L1 mice can be combined with many established murine syngeneic tumor models. Because these mice remain immunocompetent, they can be implanted with tumor cell lines originating from the same genetic background, allowing evaluation of anti-tumor immune responses in the presence of a functional immune system. genOway has developed a range of cell lines that can be used for tumor growth inhibition studies or MoA studies, such as: genO-MC38-hPD-L1-hCD47-LZ, genO-MC38-hPD-L1-hHER2-LZ, genO-MC38-hPD-L1-hFAP-LZ, genO-MC38-hFAP-LZ, genO-MC38-hPD-L1 and genO-MC38-hPD-L1-LZ for a C57BL/6 background, and genO-CT26-hPD-L1 and genO-CT26-Pd-l1KO for a BALB/c background.

Humanized PD-1 mouse versus HIS mouse: what is the difference?

Humanized PD-1 knock-in mouse (genO-hPD-1)

• Mouse immune system
• Human therapeutic target (PD-1)
• Lower variability
• Better reproducibility
• Lower cost
• Suitable for candidate ranking

HIS mouse (genO-BRGSF-HIS)

• Human immune cells
• Human tumor compatibility
• More clinically relevant immune interactions
• Greater variability
• Higher cost

How predictive are humanized PD-1 models of clinical efficacy?

They improve translational relevance by enabling:

• Human target engagement
• Physiological immune responses
• Direct evaluation of clinical antibodies

However, they do not completely recapitulate:

• Human tumor biology
• Human immune-cell diversity
• Human immune-related adverse events

Therefore, they are most powerful when integrated with complementary HIS or PDX studies.

Are there also variants of the genO-hPD-1 model that target other immune checkpoints?

Yes. genOway has developed a range of validated double humanized mouse models combining hPD-1 with other immunecheckpoints. For dual checkpoint blockade, models such as genO-hPD-1/hCTLA-4 have been exploited by AstraZeneca (Dovedi et al., 2021), while models such as genO-hPD-1/hLAG-3 (Weber et al., 2026), genO-hPD-1/hTIM-3 (Rohrberg et al., 2026), genO-hPD-1/hPD-L1, and genO-hPD-1/hVISTA also enable the evaluation of therapies targeting complementary inhibitory pathways involved in T-cell exhaustion and resistance to PD-1 blockade. For immune activation approaches, the genO-hPD-1/hICOS and genO-hPD-1/hGITR/hGITRL models support the assessment of agonistic therapies designed to enhance effector T-cell responses and modulate regulatory T-cell activity.

How can the Fc effector function of anti-PD-1 antibodies be assessed?

The genO-hPD-1 model can be intercrossed with the extensively validated genO-hFcγR model (Van Damme et al., 2026). The resulting model (genO-hFcγR/hPD-1) expresses the humanized target of the Fab region in the target cells, as well as all hFcγ receptors in effector cells. This provides a more comprehensive evaluation of therapeutic antibodies.

‍

‍

References

Barham W, et al. A Novel Humanized PD-1/PD-L1 Mouse Model Permits Direct Comparison of Antitumor Immunity Generated by FDA-Approved PD-1 and PD-L1 Inhibitors. ImmunoHorizons, 2023;7(1):125-139. doi:10.4049/immunohorizons.2200054

Dovedi S, et al. Design and Efficacy of a Monovalent Bispecific PD-1/CTLA4 Antibody That Enhances CTLA4 Blockade on PD-1+ Activated T Cells. Cancer Discov 1 May 2021; 11 (5): 1100–1117. https://doi.org/10.1158/2159-8290.CD-20-1445

Weber PAA, et al. Tobemstomig, a Novel Bispecific Antibody, Preferentially Blocks PD-1 and LAG-3 on CD8 TILs to Expand Stem-like T Cells for Sustained Tumor Control. Cancer Res Commun. 2026;6(7):1619-1639. doi:10.1158/2767-9764.CRC-26-0207

Rohrberg KS, et al. First-in-human study of lomvastomig, a PD-1-TIM-3 bispecific antibody, in patients with advanced and/or metastatic solid tumors. J Immunother Cancer. 2026;14(6):e012729. Published 2026 Jun 19. doi:10.1136/jitc-2025-012729

Van Damme KFA, et al. Cross-species cellular mapping and humanization of Fcγ receptors to advance antibody modeling. Sci Immunol. 2026;11(115):eady7328. doi:10.1126/sciimmunol.ady7328

‍

‍

‍

‍

‍