Anti-CD8 VHH Validation Series I: First 10 Candidates Complete tLNP Construction and Cell-Based Validation

In our previous update, we introduced 40 anti-CD8 VHH candidate sequences and raised a key question:

Among VHHs that can recognize CD8, which ones are truly suitable as targeting ligands for targeted lipid nanoparticles (tLNPs)?

Binding to CD8 is only the first requirement. For a VHH to function effectively on a tLNP, it must also tolerate conjugation, maintain target recognition after surface immobilization, preserve acceptable nanoparticle properties, and ultimately support functional payload delivery into CD8-positive cells.

To address this, LevoStar has initiated a systematic anti-CD8 VHH–tLNP compatibility evaluation program.

In the first round, 10 VHH candidates were selected for recombinant production, tLNP construction, nanoparticle characterization, and targeted mRNA delivery evaluation using engineered human and cynomolgus monkey CD8A-positive Jurkat cell models.

1. From CD8 Binding to tLNP Compatibility

For targeted LNP development, a VHH with good binding activity does not automatically become a good targeting ligand.

Once conjugated onto the LNP surface, several additional factors must be considered:

• Can the nanoparticle maintain suitable particle size and overall stability?
• Is the antibody successfully displayed on the particle surface?
• How many antibody molecules are present on each particle?
• Does mRNA encapsulation remain intact after conjugation?
• Can the conjugated VHH still recognize CD8?
• Most importantly, can it actually improve delivery into CD8-positive cells?

In this first batch, all 10 anti-CD8 VHH candidates were successfully conjugated to LNPs and advanced into particle characterization and cell-based evaluation.

2. Particle Characterization: Does VHH Conjugation Affect LNP Quality?

After antibody conjugation, the tLNPs were evaluated using nano-flow cytometry and encapsulation assays.

Key parameters included:

• Particle concentration
• Particle size
• Protein-positive particle rate
• Average antibody copy number per particle
• mRNA encapsulation efficiency
• Post-conjugation sample yield

Table 1. Nano-flow Cytometry and Encapsulation Efficiency Results for the First 10 Anti-CD8 tLNP Formulations

Across the first 10 anti-CD8A tLNP formulations, particle sizes were concentrated between 64.2 and 69.2 nm. Protein-positive particle rates were all above 75%, while mRNA encapsulation efficiency remained above 93%.

Although differences were observed among candidates in antibody copy number and sample recovery, the overall nanoparticle quality and mRNA encapsulation status remained suitable for subsequent functional testing.

This step is important because successful conjugation alone is not enough. A targeting ligand must be added without compromising the physical properties of the LNP or the integrity of the encapsulated payload.

3. Building Human and Cynomolgus CD8A Cell Models for Targeted Delivery Testing

Standard Jurkat cells do not express the CD8 molecule required for this study, so they cannot directly support anti-CD8 tLNP evaluation.

To establish a controlled validation system, LevoStar engineered two Jurkat-based cell models:

• Jurkat-hCD8A, expressing human CD8A
• Jurkat-cynoCD8A, expressing cynomolgus monkey CD8A

These models allow us to evaluate not only whether individual VHH candidates recognize CD8A from different species, but also whether they retain functional targeting capability after conjugation onto LNPs.

The 10 candidate VHHs were conjugated to LNPs encapsulating GFP mRNA, and the resulting tLNPs were tested in both cell models.

GFP expression was evaluated by fluorescence imaging and flow cytometry as a functional readout of targeted mRNA delivery.

Figure 1. Fluorescence Imaging of 10 Anti-CD8 tLNP Formulations in Jurkat-hCD8A Cells

Figure 2. Fluorescence Imaging of 10 Anti-CD8 tLNP Formulations in Jurkat-cynoCD8A Cells

Compared with unconjugated LNPs, all 10 anti-CD8 tLNP groups showed increased GFP fluorescence in Jurkat-hCD8A cells.

This indicates that, after conjugation to the LNP surface, all 10 candidate VHHs retained at least some ability to support recognition of CD8A-positive cells and facilitate mRNA delivery.

However, the results were not identical across candidates.

Some VHHs generated substantially stronger GFP expression than others, suggesting that even when two antibodies bind the same target, their performance as tLNP targeting ligands may differ significantly.

4. Greater Differences Emerged in Cynomolgus CD8A Cells

The differences among candidates became even more pronounced in Jurkat-cynoCD8A cells.

Some VHHs produced strong GFP signals, whereas others showed relatively weak delivery efficiency.

This suggests that cross-species binding alone does not necessarily predict equivalent functional delivery performance in human and cynomolgus systems.

Flow cytometry results were generally consistent with fluorescence imaging.

When results from both cell models were considered together, Candidates 02 and 03 showed particularly strong targeted delivery potential, while other candidates displayed different human-versus-cynomolgus delivery profiles.

This cross-species comparison may become particularly valuable when selecting ligands for programs expected to move from in vitro human-cell studies into non-human primate evaluation.

Table 2. Mean GFP Fluorescence Intensity of 10 Anti-CD8 tLNP Formulations in Two Jurkat-CD8A Cell Models

5. What Have We Learned from the First 10 Candidates?

The first round provides an important reminder for targeted LNP development :

A VHH that binds CD8 is not automatically a suitable tLNP targeting ligand.

All 10 candidates could be conjugated to the LNP surface while maintaining relatively stable particle size and high mRNA encapsulation efficiency. However, once functional delivery was evaluated in human and cynomolgus CD8A-positive cells, clear differences emerged among the candidates.

This means that ligand selection for tLNP development should go beyond conventional antibody-binding data.

A more complete evaluation should consider:

Binding → Conjugation compatibility → Particle quality → Surface antibody density → Target-cell recognition → Functional payload delivery

Only candidates that perform well across these stages are likely to become robust targeting ligands for downstream tLNP development.

What Comes Next?

This is only the first round.

Of the 40 anti-CD8 VHH candidates, the first 10 have now completed initial tLNP application testing. The next stage will integrate conjugation performance, targeted delivery efficiency, and additional functional assays to determine which candidates should advance, which require further evaluation, and which may not meet the requirements for tLNP applications.

Validation of the remaining anti-CD8 VHH candidates will also continue in batches.

40 candidates. First 10 tested. Which VHHs will advance to the next round?

Stay tuned for the next update in our Anti-CD8 VHH Validation Series.