Feeder-free expansion of natural killer cells using the Xuri™ cell expansion system

NK isolation

Fresh, leukapheresis (no plasma; ~ 1 × 10¹⁰ PBMC) (Charles River) from three healthy donors were used as the starting material. NK cells were isolated via CD3 depletion (Sepax™ C Pro system, BeadWash C-Pro application [Cytiva, 29264739]) followed by CliniMACS Plus depletion (Miltenyi Biotec, 170-076-652) and CD56 enrichment (Sepax™ C Pro BeadWash C-Pro application) followed by CliniMACS Plus enrichment (Miltenyi Biotec, 170-076-651). Flow cytometry was used to quantitate final NK cell (CD56+) purity using the Beckman Coulter CytoFLEX.

NK expansion

Isolated NK were seeded in a 2 L Xuri™ Cellbag™ basic at a concentration of 1 million cells/mL (400 million live NK in 400 mL volume) and gently rocked at 2 rpm and 2° within a Xuri™ cell expansion system W25. Complete NK MACS GMP Medium (Miltenyi Biotec) supplemented with 5% human serum albumin (GeminiBio), 500 IU/mL IL-2 (Akron), and 10 ng/mL IL-15 (PeproTech) was used throughout expansion. Donors with < 5% CD56^bright population (donor 2) were supplemented with IL-2 and IL-15 on day 3 to promote activation.

Cells were sampled daily starting on day 5. Analytics included cell count, viability, and cell size using the NucleoCounter NC200 (ChemoMetec), and metabolites using the Vi-CELL MetaFLEX Bioanalyte Analyzer (Beckman Coulter). On day 5, NK purity (CD56) and activation (NKG2D/KLRK1, NKp44/NCR2) were measured using flow cytometry. After sampling, cell concentration was reset to 6 × 10⁵ cells/mL via fed-batch media addition. If cells did not receive fresh media for > 3 d, the culture was supplemented with IL-2 and IL-15 to promote proliferation. NK purity was also measured at harvest.

NKs were transferred to 20 L and 50 L bags as required to maintain the target cell concentration, following bag capacity guidelines: maximum volume of 1 L for a 2 L bag; 10 L for a 20 L bag; and 25 L for a 50 L bag. Throughout culturing, Cellbag™ bioreactor bags were minimally rocked between 2 to 3 rpm and 2° to 10°.

Harvest

The FlexCell application on the Sefia™ S-2000 cell processing system was used to harvest (wash and reformulate) a maximum of 10 L per run (Table 1). If final volume was > 10 L, harvest was split into two runs (two 10 L input bags). NKs were washed and reformulated using default parameters with a g-force of 400 g for 5 min. Cells were washed using the CliniMACS PBS/EDTA buffer (Miltenyi Biotec) with 0.5% v/v human albumin solution (Gemini Bio) and resuspended in CryoStor CS10 cell freezing medium. Cell viability and recovery were quantitated using the NC200.

Table 1. FlexCell application parameters on the Sefia™ S-2000 cell processing system

Finally, NKs were cryopreserved using a VIA Freeze™ controlled rate freezer. The next day, cells were thawed using a VIA Thaw™ dry automated thawer and the resulting cell viability was measured.

Using a streamlined, feeder-free workflow for NK manufacturing, we report three key findings:

• Starting from 250-400 million cells from three individual donors, feeder-free expansion on with the Xuri™ bioreactor achieved > 15 billion live NK cells with over 99% CD56 purity.
• Cytokine-driven activation in NK MACS Medium supported early proliferation inside the Xuri™ bioreactor, enabling consistent expansion across donors. All donors reached multi‑billion cell yields within 16 to 19 days of feeder-free culture. For donor 2, who began with a lower initial NK input (246 million), supplemental cytokines were added during culture to achieve yields comparable to other donors.
• High post-harvest recovery (82 ± 2%) with > 92% average viability was achieved using the Sefia™ S-2000 cell processing system.
• High post-thaw viability (88 ± 2%) was achieved using automated procedures on the VIA Freeze™ and VIA Thaw™ systems.

Isolation via CD3 depletion and CD56 enrichment resulted in > 98% CD56+ input product. Based on this purity, 400 million live NK were inoculated in 400 mL of complete NK MACS Medium at a seeding density of 1 million live NK/mL in the 2 L Xuri™ Cellbag™ basic (2 rpm, 2°). One exception was donor 2, which only yielded 246 million live NK. In this case, all cells were seeded in 300 mL of NK media.

Fig 2. NK cell expansion results: (A) total live cells, (B) total fold expansion, (C) average viability, (D) average cell diameter, (E) NK purity at harvest, (F) NK activation.

Table 2. NK cell expansion results

Cells were sampled starting day 5 and reset to 6 × 10⁵ live cells/mL using fed-batch media addition. Donor 1 reached 27 billion live NK by day 16 (67-fold expansion); donor 2 reached 18 billion live NK by day 19 (38-fold expansion); donor 3 reached 17 billion live NK by day 16 (68-fold expansion). All samples remained > 91% viable throughout the culture and had > 99% NK purity (CD56+) at harvest (Fig 2, Table 2). Metabolic activity remained in the range of 7.1 to 7.4 pH and 2 to 12 mmol/L lactate throughout the culture period (note: fresh media addition throughout the culture impacted these values).

After expansion, cells were harvested and reformulated using the FlexCell application on the Sefia™ S-2000 system (Table 1). Average NK recovery was 82 ± 2% and average post-wash viability was 92 ± 0.3% (~ 2% viability drop compared to the initial product) (Table 3).

Cells were cryopreserved using the VIA Freeze™ system and thawed the next day using the VIA Thaw™ system, yielding an average post-thaw viability of 88 ± 2% (Table 3).

Table 3. NK cell harvest result summary

Discussion

Using a streamlined, feeder-free workflow with automated systems, starting material from three healthy donors (containing 250 to 400 million NK cells) were each expanded to > 15 billion NK cells, demonstrating clinically relevant quantities for allogeneic scale manufacturing.

Streamlined feeder-free expansion strategy

By maintaining a slow rocking rate on the Xuri™ bioreactor during activation, we could directly inoculate the 2 L Xuri™ Cellbag™ container—without the need for a smaller vessel or additional manipulation steps, simplifying the workflow and reducing touchpoints. Cytokines IL-2 and IL-15 alongside NK MACS medium accomplished sufficient cell activation (> 70% NKG2D, NKp44 co-expression) without the need for feeder cells or activation beads that complicate regulatory approval.

NK cell phenotype dependence

The data also sheds light on the relationship between initial CD56^bright populations and proliferative capabilities. The donors with the greatest total fold expansion (67- to 68-fold expansion for donors 1 and 3) had the greatest CD56^bright population on day 0 (approximately 15%). On the other hand, donor 2 reached 38-fold expansion and had just 1.32% CD56^bright subset on day 0.

Historically (1, 2), CD56^bright NK cell subsets show better in vitro proliferation in response to IL-2 and IL-15 compared to CD56^dim populations. This difference has been explained by the increased cytokine production and high-affinity receptors in CD56^bright subsets. Based on this relationship, we supplemented donor 2 with cytokines on day 3 to boost cytokine-related activation, allowing us to achieve > 15 billion cells despite initial phenotype.

In general, most healthy donors fall within the 5% to 15% CD56^bright subset population. We’ve demonstrated that the Xuri™ bioreactor can support clinically relevant allogeneic-scale expansion within 16 d from typical starting material and within 19 d from low CD56^bright inputs. Despite baseline phenotype differences, the closed process delivered consistent postharvest viability and recovery across donors.

Downstream NK processing

This work also confirms the suitability of the Sefia™ S-2000 cell processing system for NK cell harvest. Using automated procedures on the closed system to wash and formulate billions of NK cells, we achieved high cell recovery (> 82%) with minimal drop in viability.

Furthermore, controlled freezing and thawing procedures on the automated VIA Thaw™ and VIA Freeze™ systems enabled us to maintain high cell viability post-thaw, which is essential to support the logistics of allogeneic cell therapies.

1. Carson WE, Giri JG, Lindemann MJ, Linett ML, Ahdieh M, Paxton R, Anderson D, Eisenmann J, Grabstein K, Caligiuri MA. Interleukin (IL) 15 is a novel cytokine that activates human natural killer cells via components of the IL-2 receptor. J Exp Med. 1994 Oct 1;180(4):1395-403. doi: 10.1084/jem.180.4.1395. PMID: 7523571; PMCID: PMC2191697.
2. Baume DM, Robertson MJ, Levine H, Manley TJ, Schow PW, Ritz J. Differential responses to interleukin 2 define functionally distinct subsets of human natural killer cells. Eur J Immunol. 1992 Jan;22(1):1-6. doi: 10.1002/eji.1830220102. PMID: 1370410.

CY59080